From f14c4f141ad88290bb23d597a0a47833e7891fcd Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Tue, 27 Jul 2010 08:14:35 +0000 Subject: ARM: mach-shmobile: Runtime PM V3 This is the SH-Mobile ARM Runtime PM implementation V3. Only clocks are managed at this point. If Runtime PM is disabled then the clocks will be enabled before ->probe() and disabled after ->remove(). Signed-off-by: Magnus Damm Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/Makefile b/arch/arm/mach-shmobile/Makefile index 5e16b4c..ae416fe 100644 --- a/arch/arm/mach-shmobile/Makefile +++ b/arch/arm/mach-shmobile/Makefile @@ -3,7 +3,7 @@ # # Common objects -obj-y := timer.o console.o clock.o +obj-y := timer.o console.o clock.o pm_runtime.o # CPU objects obj-$(CONFIG_ARCH_SH7367) += setup-sh7367.o clock-sh7367.o intc-sh7367.o diff --git a/arch/arm/mach-shmobile/clock-sh7372.c b/arch/arm/mach-shmobile/clock-sh7372.c index fb4e9b1..4a1d019 100644 --- a/arch/arm/mach-shmobile/clock-sh7372.c +++ b/arch/arm/mach-shmobile/clock-sh7372.c @@ -428,7 +428,7 @@ static struct clk mstp_clks[MSTP_NR] = { [MSTP201] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 1, 0), /* SCIFA3 */ [MSTP200] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 0, 0), /* SCIFA4 */ [MSTP329] = MSTP(&r_clk, SMSTPCR3, 29, 0), /* CMT10 */ - [MSTP328] = MSTP(&div6_clks[DIV6_SPU], SMSTPCR3, 28, CLK_ENABLE_ON_INIT), /* FSIA */ + [MSTP328] = MSTP(&div6_clks[DIV6_SPU], SMSTPCR3, 28, 0), /* FSIA */ [MSTP323] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR3, 23, 0), /* IIC1 */ [MSTP322] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR3, 22, 0), /* USB0 */ [MSTP314] = MSTP(&div4_clks[DIV4_HP], SMSTPCR3, 14, 0), /* SDHI0 */ diff --git a/arch/arm/mach-shmobile/pm_runtime.c b/arch/arm/mach-shmobile/pm_runtime.c new file mode 100644 index 0000000..94912d3 --- /dev/null +++ b/arch/arm/mach-shmobile/pm_runtime.c @@ -0,0 +1,169 @@ +/* + * arch/arm/mach-shmobile/pm_runtime.c + * + * Runtime PM support code for SuperH Mobile ARM + * + * Copyright (C) 2009-2010 Magnus Damm + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef CONFIG_PM_RUNTIME +#define BIT_ONCE 0 +#define BIT_ACTIVE 1 +#define BIT_CLK_ENABLED 2 + +struct pm_runtime_data { + unsigned long flags; + struct clk *clk; +}; + +static void __devres_release(struct device *dev, void *res) +{ + struct pm_runtime_data *prd = res; + + dev_dbg(dev, "__devres_release()\n"); + + if (test_bit(BIT_CLK_ENABLED, &prd->flags)) + clk_disable(prd->clk); + + if (test_bit(BIT_ACTIVE, &prd->flags)) + clk_put(prd->clk); +} + +static struct pm_runtime_data *__to_prd(struct device *dev) +{ + return devres_find(dev, __devres_release, NULL, NULL); +} + +static void platform_pm_runtime_init(struct device *dev, + struct pm_runtime_data *prd) +{ + if (prd && !test_and_set_bit(BIT_ONCE, &prd->flags)) { + prd->clk = clk_get(dev, NULL); + if (!IS_ERR(prd->clk)) { + set_bit(BIT_ACTIVE, &prd->flags); + dev_info(dev, "clocks managed by runtime pm\n"); + } + } +} + +static void platform_pm_runtime_bug(struct device *dev, + struct pm_runtime_data *prd) +{ + if (prd && !test_and_set_bit(BIT_ONCE, &prd->flags)) + dev_err(dev, "runtime pm suspend before resume\n"); +} + +int platform_pm_runtime_suspend(struct device *dev) +{ + struct pm_runtime_data *prd = __to_prd(dev); + + dev_dbg(dev, "platform_pm_runtime_suspend()\n"); + + platform_pm_runtime_bug(dev, prd); + + if (prd && test_bit(BIT_ACTIVE, &prd->flags)) { + clk_disable(prd->clk); + clear_bit(BIT_CLK_ENABLED, &prd->flags); + } + + return 0; +} + +int platform_pm_runtime_resume(struct device *dev) +{ + struct pm_runtime_data *prd = __to_prd(dev); + + dev_dbg(dev, "platform_pm_runtime_resume()\n"); + + platform_pm_runtime_init(dev, prd); + + if (prd && test_bit(BIT_ACTIVE, &prd->flags)) { + clk_enable(prd->clk); + set_bit(BIT_CLK_ENABLED, &prd->flags); + } + + return 0; +} + +int platform_pm_runtime_idle(struct device *dev) +{ + /* suspend synchronously to disable clocks immediately */ + return pm_runtime_suspend(dev); +} + +static int platform_bus_notify(struct notifier_block *nb, + unsigned long action, void *data) +{ + struct device *dev = data; + struct pm_runtime_data *prd; + + dev_dbg(dev, "platform_bus_notify() %ld !\n", action); + + if (action == BUS_NOTIFY_BIND_DRIVER) { + prd = devres_alloc(__devres_release, sizeof(*prd), GFP_KERNEL); + if (prd) + devres_add(dev, prd); + else + dev_err(dev, "unable to alloc memory for runtime pm\n"); + } + + return 0; +} + +#else /* CONFIG_PM_RUNTIME */ + +static int platform_bus_notify(struct notifier_block *nb, + unsigned long action, void *data) +{ + struct device *dev = data; + struct clk *clk; + + dev_dbg(dev, "platform_bus_notify() %ld !\n", action); + + switch (action) { + case BUS_NOTIFY_BIND_DRIVER: + clk = clk_get(dev, NULL); + if (!IS_ERR(clk)) { + clk_enable(clk); + clk_put(clk); + dev_info(dev, "runtime pm disabled, clock forced on\n"); + } + break; + case BUS_NOTIFY_UNBOUND_DRIVER: + clk = clk_get(dev, NULL); + if (!IS_ERR(clk)) { + clk_disable(clk); + clk_put(clk); + dev_info(dev, "runtime pm disabled, clock forced off\n"); + } + break; + } + + return 0; +} + +#endif /* CONFIG_PM_RUNTIME */ + +static struct notifier_block platform_bus_notifier = { + .notifier_call = platform_bus_notify +}; + +static int __init sh_pm_runtime_init(void) +{ + bus_register_notifier(&platform_bus_type, &platform_bus_notifier); + return 0; +} +core_initcall(sh_pm_runtime_init); -- cgit v0.10.2 From a4909b52682c1e2ac858754c8bc46f2501e67dee Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Mon, 26 Jul 2010 16:20:53 +0000 Subject: ARM: mach-shmobile: add clock definitions for CEU and CSI2 Two more clocks to be managed by the runtime PM. Signed-off-by: Guennadi Liakhovetski Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/clock-sh7372.c b/arch/arm/mach-shmobile/clock-sh7372.c index 4a1d019..dcaac1b 100644 --- a/arch/arm/mach-shmobile/clock-sh7372.c +++ b/arch/arm/mach-shmobile/clock-sh7372.c @@ -395,7 +395,7 @@ static struct clk div6_reparent_clks[DIV6_REPARENT_NR] = { enum { MSTP001, MSTP131, MSTP130, - MSTP129, MSTP128, + MSTP129, MSTP128, MSTP127, MSTP126, MSTP118, MSTP117, MSTP116, MSTP106, MSTP101, MSTP100, MSTP223, @@ -413,6 +413,8 @@ static struct clk mstp_clks[MSTP_NR] = { [MSTP130] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 30, 0), /* VEU2 */ [MSTP129] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 29, 0), /* VEU1 */ [MSTP128] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 28, 0), /* VEU0 */ + [MSTP127] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 27, 0), /* CEU */ + [MSTP126] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 26, 0), /* CSI2 */ [MSTP118] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 18, 0), /* DSITX */ [MSTP117] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 17, 0), /* LCDC1 */ [MSTP116] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR1, 16, 0), /* IIC0 */ @@ -498,6 +500,8 @@ static struct clk_lookup lookups[] = { CLKDEV_DEV_ID("uio_pdrv_genirq.3", &mstp_clks[MSTP130]), /* VEU2 */ CLKDEV_DEV_ID("uio_pdrv_genirq.2", &mstp_clks[MSTP129]), /* VEU1 */ CLKDEV_DEV_ID("uio_pdrv_genirq.1", &mstp_clks[MSTP128]), /* VEU0 */ + CLKDEV_DEV_ID("sh_mobile_ceu.0", &mstp_clks[MSTP127]), /* CEU */ + CLKDEV_DEV_ID("sh-mobile-csi2.0", &mstp_clks[MSTP126]), /* CSI2 */ CLKDEV_DEV_ID("sh-mipi-dsi.0", &mstp_clks[MSTP118]), /* DSITX */ CLKDEV_DEV_ID("sh_mobile_lcdc_fb.1", &mstp_clks[MSTP117]), /* LCDC1 */ CLKDEV_DEV_ID("i2c-sh_mobile.0", &mstp_clks[MSTP116]), /* IIC0 */ -- cgit v0.10.2 From 2863e935a59b36d55522af8620112c64a56b4022 Mon Sep 17 00:00:00 2001 From: Arnd Hannemann Date: Sat, 14 Aug 2010 19:45:21 +0000 Subject: ARM: mach-shmobile: AP4EVB: use gpio-leds for leds. This allows control of the leds on the AP4EVB board by kernel events (triggers) like mmc0/mmc1. In userland, the leds won't be available anymore over /sys/class/gpio but via /sys/class/leds. Signed-off-by: Arnd Hannemann Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/board-ap4evb.c b/arch/arm/mach-shmobile/board-ap4evb.c index 23d472f..98c1cf7 100644 --- a/arch/arm/mach-shmobile/board-ap4evb.c +++ b/arch/arm/mach-shmobile/board-ap4evb.c @@ -39,6 +39,7 @@ #include #include #include +#include #include #include @@ -650,7 +651,44 @@ static struct platform_device hdmi_device = { }, }; +static struct gpio_led ap4evb_leds[] = { + { + .name = "led4", + .gpio = GPIO_PORT185, + .default_state = LEDS_GPIO_DEFSTATE_ON, + }, + { + .name = "led2", + .gpio = GPIO_PORT186, + .default_state = LEDS_GPIO_DEFSTATE_ON, + }, + { + .name = "led3", + .gpio = GPIO_PORT187, + .default_state = LEDS_GPIO_DEFSTATE_ON, + }, + { + .name = "led1", + .gpio = GPIO_PORT188, + .default_state = LEDS_GPIO_DEFSTATE_ON, + } +}; + +static struct gpio_led_platform_data ap4evb_leds_pdata = { + .num_leds = ARRAY_SIZE(ap4evb_leds), + .leds = &ap4evb_leds, +}; + +static struct platform_device leds_device = { + .name = "leds-gpio", + .id = 0, + .dev = { + .platform_data = &ap4evb_leds_pdata, + }, +}; + static struct platform_device *ap4evb_devices[] __initdata = { + &leds_device, &nor_flash_device, &smc911x_device, &sdhi0_device, @@ -840,20 +878,6 @@ static void __init ap4evb_init(void) gpio_request(GPIO_FN_CS5A, NULL); gpio_request(GPIO_FN_IRQ6_39, NULL); - /* enable LED 1 - 4 */ - gpio_request(GPIO_PORT185, NULL); - gpio_request(GPIO_PORT186, NULL); - gpio_request(GPIO_PORT187, NULL); - gpio_request(GPIO_PORT188, NULL); - gpio_direction_output(GPIO_PORT185, 1); - gpio_direction_output(GPIO_PORT186, 1); - gpio_direction_output(GPIO_PORT187, 1); - gpio_direction_output(GPIO_PORT188, 1); - gpio_export(GPIO_PORT185, 0); - gpio_export(GPIO_PORT186, 0); - gpio_export(GPIO_PORT187, 0); - gpio_export(GPIO_PORT188, 0); - /* enable Debug switch (S6) */ gpio_request(GPIO_PORT32, NULL); gpio_request(GPIO_PORT33, NULL); -- cgit v0.10.2 From 8050fbf268c3bc09752ceece40aff5d3f4e3e182 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 16 Aug 2010 14:54:35 +0900 Subject: ARM: mach-shmobile: Fix up gpio-leds pdata definition. Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/board-ap4evb.c b/arch/arm/mach-shmobile/board-ap4evb.c index 98c1cf7..8afdb1d 100644 --- a/arch/arm/mach-shmobile/board-ap4evb.c +++ b/arch/arm/mach-shmobile/board-ap4evb.c @@ -676,7 +676,7 @@ static struct gpio_led ap4evb_leds[] = { static struct gpio_led_platform_data ap4evb_leds_pdata = { .num_leds = ARRAY_SIZE(ap4evb_leds), - .leds = &ap4evb_leds, + .leds = ap4evb_leds, }; static struct platform_device leds_device = { -- cgit v0.10.2 From ec53fe3d655befce6420a1b8485af9b1145a0c90 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Eric=20B=C3=A9nard?= Date: Fri, 30 Jul 2010 22:58:40 +0200 Subject: eukrea-baseboards: fix the merge in one file MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Last change I did made a big mess in function names and CONFIG names, this patch fixes this so that the baseboard support is really built when selected. Signed-off-by: Eric Bénard Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx25/eukrea_mbimxsd-baseboard.c b/arch/arm/mach-mx25/eukrea_mbimxsd-baseboard.c index 91931dc..4aaadc7 100644 --- a/arch/arm/mach-mx25/eukrea_mbimxsd-baseboard.c +++ b/arch/arm/mach-mx25/eukrea_mbimxsd-baseboard.c @@ -215,7 +215,7 @@ struct imx_ssi_platform_data eukrea_mbimxsd_ssi_pdata = { * Add platform devices present on this baseboard and init * them from CPU side as far as required to use them later on */ -void __init eukrea_mbimxsd_baseboard_init(void) +void __init eukrea_mbimxsd25_baseboard_init(void) { if (mxc_iomux_v3_setup_multiple_pads(eukrea_mbimxsd_pads, ARRAY_SIZE(eukrea_mbimxsd_pads))) diff --git a/arch/arm/mach-mx25/mach-cpuimx25.c b/arch/arm/mach-mx25/mach-cpuimx25.c index 56b2e26..b5efee2 100644 --- a/arch/arm/mach-mx25/mach-cpuimx25.c +++ b/arch/arm/mach-mx25/mach-cpuimx25.c @@ -147,8 +147,8 @@ static void __init eukrea_cpuimx25_init(void) if (!otg_mode_host) mxc_register_device(&otg_udc_device, &otg_device_pdata); -#ifdef CONFIG_MACH_EUKREA_MBIMXSD_BASEBOARD - eukrea_mbimxsd_baseboard_init(); +#ifdef CONFIG_MACH_EUKREA_MBIMXSD25_BASEBOARD + eukrea_mbimxsd25_baseboard_init(); #endif } diff --git a/arch/arm/mach-mx3/eukrea_mbimxsd-baseboard.c b/arch/arm/mach-mx3/eukrea_mbimxsd-baseboard.c index 1dc5004..f8f15e3 100644 --- a/arch/arm/mach-mx3/eukrea_mbimxsd-baseboard.c +++ b/arch/arm/mach-mx3/eukrea_mbimxsd-baseboard.c @@ -216,7 +216,7 @@ struct imx_ssi_platform_data eukrea_mbimxsd_ssi_pdata = { * Add platform devices present on this baseboard and init * them from CPU side as far as required to use them later on */ -void __init eukrea_mbimxsd_baseboard_init(void) +void __init eukrea_mbimxsd35_baseboard_init(void) { if (mxc_iomux_v3_setup_multiple_pads(eukrea_mbimxsd_pads, ARRAY_SIZE(eukrea_mbimxsd_pads))) diff --git a/arch/arm/mach-mx3/mach-cpuimx35.c b/arch/arm/mach-mx3/mach-cpuimx35.c index 63f970f..e4268bb 100644 --- a/arch/arm/mach-mx3/mach-cpuimx35.c +++ b/arch/arm/mach-mx3/mach-cpuimx35.c @@ -201,8 +201,8 @@ static void __init mxc_board_init(void) if (!otg_mode_host) mxc_register_device(&mxc_otg_udc_device, &otg_device_pdata); -#ifdef CONFIG_MACH_EUKREA_MBIMXSD_BASEBOARD - eukrea_mbimxsd_baseboard_init(); +#ifdef CONFIG_MACH_EUKREA_MBIMXSD35_BASEBOARD + eukrea_mbimxsd35_baseboard_init(); #endif } diff --git a/arch/arm/plat-mxc/include/mach/eukrea-baseboards.h b/arch/arm/plat-mxc/include/mach/eukrea-baseboards.h index 634e3f4..656acb4 100644 --- a/arch/arm/plat-mxc/include/mach/eukrea-baseboards.h +++ b/arch/arm/plat-mxc/include/mach/eukrea-baseboards.h @@ -37,9 +37,9 @@ * mach-mx5/eukrea_mbimx51-baseboard.c for cpuimx51 */ -extern void eukrea_mbimx25_baseboard_init(void); +extern void eukrea_mbimxsd25_baseboard_init(void); extern void eukrea_mbimx27_baseboard_init(void); -extern void eukrea_mbimx35_baseboard_init(void); +extern void eukrea_mbimxsd35_baseboard_init(void); extern void eukrea_mbimx51_baseboard_init(void); #endif -- cgit v0.10.2 From 3d6e614952e3e4cf9e70e58893a740ffec220b24 Mon Sep 17 00:00:00 2001 From: "Hans J. Koch" Date: Thu, 12 Aug 2010 15:10:53 +0200 Subject: mx35: Fix boot ROM hang in internal boot mode MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If a watchdog reset occurs after booting in internal boot mode, the i.MX35 won't boot anymore. The boot ROM code seems to assume that some clocks are turned on (they are after a power-on reset). This patch turns on the necessary clocks. Signed-off-by: Hans J. Koch Acked-by: Uwe Kleine-König Reported-by: John Ogness Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c index d3af0fd..32adf8d 100644 --- a/arch/arm/mach-mx3/clock-imx35.c +++ b/arch/arm/mach-mx3/clock-imx35.c @@ -485,10 +485,10 @@ static struct clk_lookup lookups[] = { int __init mx35_clocks_init() { - unsigned int ll = 0; + unsigned int cgr2 = 3 << 26, cgr3 = 0; #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC) - ll = (3 << 16); + cgr2 |= 3 << 16; #endif clkdev_add_table(lookups, ARRAY_SIZE(lookups)); @@ -499,8 +499,20 @@ int __init mx35_clocks_init() __raw_writel((3 << 18), CCM_BASE + CCM_CGR0); __raw_writel((3 << 2) | (3 << 4) | (3 << 6) | (3 << 8) | (3 << 16), CCM_BASE + CCM_CGR1); - __raw_writel((3 << 26) | ll, CCM_BASE + CCM_CGR2); - __raw_writel(0, CCM_BASE + CCM_CGR3); + + /* + * Check if we came up in internal boot mode. If yes, we need some + * extra clocks turned on, otherwise the MX35 boot ROM code will + * hang after a watchdog reset. + */ + if (!(__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10))) { + /* Additionally turn on UART1, SCC, and IIM clocks */ + cgr2 |= 3 << 16 | 3 << 4; + cgr3 |= 3 << 2; + } + + __raw_writel(cgr2, CCM_BASE + CCM_CGR2); + __raw_writel(cgr3, CCM_BASE + CCM_CGR3); mxc_timer_init(&gpt_clk, MX35_IO_ADDRESS(MX35_GPT1_BASE_ADDR), MX35_INT_GPT); -- cgit v0.10.2 From 9a6f17fd1cd84811962735ff47111f84a8f05073 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 22 Apr 2010 12:37:13 +0200 Subject: i.MX35: remove get_3_3_div helper function In the v2 reference manual there are no dividers combined of two dividers. Instead, all dividers are simple 6bit dividers. I assume the combined dividers only exist in preliminary hardware. Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c index 32adf8d..d3a8d3f 100644 --- a/arch/arm/mach-mx3/clock-imx35.c +++ b/arch/arm/mach-mx3/clock-imx35.c @@ -176,16 +176,11 @@ static unsigned long get_rate_ipg(struct clk *clk) return get_rate_ahb(NULL) >> 1; } -static unsigned long get_3_3_div(unsigned long in) -{ - return (((in >> 3) & 0x7) + 1) * ((in & 0x7) + 1); -} - static unsigned long get_rate_uart(struct clk *clk) { unsigned long pdr3 = __raw_readl(CCM_BASE + CCM_PDR3); unsigned long pdr4 = __raw_readl(CCM_BASE + CCM_PDR4); - unsigned long div = get_3_3_div(pdr4 >> 10); + unsigned long div = ((pdr4 >> 10) & 0x3f) + 1; if (pdr3 & (1 << 14)) return get_rate_arm() / div; @@ -216,7 +211,7 @@ static unsigned long get_rate_sdhc(struct clk *clk) break; } - return rate / get_3_3_div(div); + return rate / (div + 1); } static unsigned long get_rate_mshc(struct clk *clk) @@ -270,7 +265,7 @@ static unsigned long get_rate_csi(struct clk *clk) else rate = get_rate_ppll(); - return rate / get_3_3_div((pdr2 >> 16) & 0x3f); + return rate / (((pdr2 >> 16) & 0x3f) + 1); } static unsigned long get_rate_otg(struct clk *clk) @@ -283,22 +278,21 @@ static unsigned long get_rate_otg(struct clk *clk) else rate = get_rate_ppll(); - return rate / get_3_3_div((pdr4 >> 22) & 0x3f); + return rate / (((pdr4 >> 22) & 0x3f) + 1); } static unsigned long get_rate_ipg_per(struct clk *clk) { unsigned long pdr0 = __raw_readl(CCM_BASE + CCM_PDR0); unsigned long pdr4 = __raw_readl(CCM_BASE + CCM_PDR4); - unsigned long div1, div2; + unsigned long div; if (pdr0 & (1 << 26)) { - div1 = (pdr4 >> 19) & 0x7; - div2 = (pdr4 >> 16) & 0x7; - return get_rate_arm() / ((div1 + 1) * (div2 + 1)); + div = (pdr4 >> 16) & 0x3f; + return get_rate_arm() / (div + 1); } else { - div1 = (pdr0 >> 12) & 0x7; - return get_rate_ahb(NULL) / div1; + div = (pdr0 >> 12) & 0x7; + return get_rate_ahb(NULL) / div; } } -- cgit v0.10.2 From c96c1e394fa72bd995154af7143873945a3fa89d Mon Sep 17 00:00:00 2001 From: Marc Kleine-Budde Date: Mon, 16 Aug 2010 15:54:05 +0200 Subject: clock-imx35: fix divider if ahb is source for ipg_per According to the Datasheet: "i.MX35 (MCIMX35) Multimedia Applications Processor Reference Manual, Rev. 2" "Table 14-6. PDR0 Field Descriptions" the divider is CCM_PER_AHB[3:0] + 1. This patch adds the missing + 1. Signed-off-by: Marc Kleine-Budde Signed-off-by: Michael Grzeschik Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c index d3a8d3f..e582ceb 100644 --- a/arch/arm/mach-mx3/clock-imx35.c +++ b/arch/arm/mach-mx3/clock-imx35.c @@ -292,7 +292,7 @@ static unsigned long get_rate_ipg_per(struct clk *clk) return get_rate_arm() / (div + 1); } else { div = (pdr0 >> 12) & 0x7; - return get_rate_ahb(NULL) / div; + return get_rate_ahb(NULL) / (div + 1); } } -- cgit v0.10.2 From c56702de106b4fde301282f7ea60412276481442 Mon Sep 17 00:00:00 2001 From: Marc Kleine-Budde Date: Mon, 16 Aug 2010 15:54:06 +0200 Subject: clock-imx35: correct arm and ahb clock calculation According to the Datasheet: "i.MX35 (MCIMX35) Multimedia Applications Processor Reference Manual, Rev. 2" "Figure 14-24. Clock Control And Gating" change the result of get_rate_ahb based on the frequency returned by get_rate_arm to calculate the proper rate. Signed-off-by: Marc Kleine-Budde Signed-off-by: Michael Grzeschik Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c index e582ceb..1d311b7 100644 --- a/arch/arm/mach-mx3/clock-imx35.c +++ b/arch/arm/mach-mx3/clock-imx35.c @@ -155,7 +155,7 @@ static unsigned long get_rate_arm(void) aad = &clk_consumer[(pdr0 >> 16) & 0xf]; if (aad->sel) - fref = fref * 2 / 3; + fref = fref * 3 / 4; return fref / aad->arm; } @@ -164,7 +164,7 @@ static unsigned long get_rate_ahb(struct clk *clk) { unsigned long pdr0 = __raw_readl(CCM_BASE + CCM_PDR0); struct arm_ahb_div *aad; - unsigned long fref = get_rate_mpll(); + unsigned long fref = get_rate_arm(); aad = &clk_consumer[(pdr0 >> 16) & 0xf]; -- cgit v0.10.2 From c735c40587cea45f9d2f83339f374595073c1923 Mon Sep 17 00:00:00 2001 From: Juergen Beisert Date: Mon, 16 Aug 2010 15:54:07 +0200 Subject: clock-imx35: Calculate the base clock rate for the IPU unit The mx3fb driver needs the clock the IPU runs in order to calculate the divider for the LCD clock. This patch adds the clock rate calculation routine for the i.MX35 CPU. Signed-off-by: Juergen Beisert Signed-off-by: Michael Grzeschik Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c index 1d311b7..7a62e74 100644 --- a/arch/arm/mach-mx3/clock-imx35.c +++ b/arch/arm/mach-mx3/clock-imx35.c @@ -296,6 +296,33 @@ static unsigned long get_rate_ipg_per(struct clk *clk) } } +static unsigned long get_rate_hsp(struct clk *clk) +{ + unsigned long hsp_podf = (__raw_readl(CCM_BASE + CCM_PDR0) >> 20) & 0x03; + unsigned long fref = get_rate_mpll(); + + if (fref > 400 * 1000 * 1000) { + switch (hsp_podf) { + case 0: + return fref >> 2; + case 1: + return fref >> 3; + case 2: + return fref / 3; + } + } else { + switch (hsp_podf) { + case 0: + case 2: + return fref / 3; + case 1: + return fref / 6; + } + } + + return 0; +} + static int clk_cgr_enable(struct clk *clk) { u32 reg; @@ -353,7 +380,7 @@ DEFINE_CLOCK(i2c1_clk, 0, CCM_CGR1, 10, get_rate_ipg_per, NULL); DEFINE_CLOCK(i2c2_clk, 1, CCM_CGR1, 12, get_rate_ipg_per, NULL); DEFINE_CLOCK(i2c3_clk, 2, CCM_CGR1, 14, get_rate_ipg_per, NULL); DEFINE_CLOCK(iomuxc_clk, 0, CCM_CGR1, 16, NULL, NULL); -DEFINE_CLOCK(ipu_clk, 0, CCM_CGR1, 18, get_rate_ahb, NULL); +DEFINE_CLOCK(ipu_clk, 0, CCM_CGR1, 18, get_rate_hsp, NULL); DEFINE_CLOCK(kpp_clk, 0, CCM_CGR1, 20, get_rate_ipg, NULL); DEFINE_CLOCK(mlb_clk, 0, CCM_CGR1, 22, get_rate_ahb, NULL); DEFINE_CLOCK(mshc_clk, 0, CCM_CGR1, 24, get_rate_mshc, NULL); -- cgit v0.10.2 From 3572b33efb044f76c224197408a812cda7d54fe4 Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Mon, 16 Aug 2010 10:40:49 +0000 Subject: ARM: mach-shmobile: Clock framework comment fix Update clock framework header comment for SH-Mobile ARM. Signed-off-by: Magnus Damm Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/clock.c b/arch/arm/mach-shmobile/clock.c index b7c705a..6b7c7c4 100644 --- a/arch/arm/mach-shmobile/clock.c +++ b/arch/arm/mach-shmobile/clock.c @@ -1,8 +1,10 @@ /* - * SH-Mobile Timer + * SH-Mobile Clock Framework * * Copyright (C) 2010 Magnus Damm * + * Used together with arch/arm/common/clkdev.c and drivers/sh/clk.c. + * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; version 2 of the License. -- cgit v0.10.2 From 4e1774f3a5de0885be2e59a8e8ef0e99c5f79e7a Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Tue, 17 Aug 2010 22:53:58 +0000 Subject: ARM: mach-shmobile: do not enable the PLLC2 clock on init AP4EVB has to reparent PLLC2 to provide a precise HDMI clock, this is only possible, if PLLC2 hasn't been enabled yet. Since no other driver currently uses PLLC2 we can safely remove the CLK_ENABLE_ON_INIT flag. This fixes the "Cannot set PLLC2 parent: -16, 1 users" error message, when trying to use HDMI on AP4EVB. Signed-off-by: Guennadi Liakhovetski Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/clock-sh7372.c b/arch/arm/mach-shmobile/clock-sh7372.c index dcaac1b..7594689 100644 --- a/arch/arm/mach-shmobile/clock-sh7372.c +++ b/arch/arm/mach-shmobile/clock-sh7372.c @@ -286,7 +286,6 @@ static struct clk_ops pllc2_clk_ops = { struct clk pllc2_clk = { .ops = &pllc2_clk_ops, - .flags = CLK_ENABLE_ON_INIT, .parent = &extal1_div2_clk, .freq_table = pllc2_freq_table, .parent_table = pllc2_parent, -- cgit v0.10.2 From 9f6f11b65ea2da8fab4d6902595d770e89e64d93 Mon Sep 17 00:00:00 2001 From: Kuninori Morimoto Date: Wed, 18 Aug 2010 03:37:19 +0000 Subject: ARM: mach-shmobile: ap4evb: modify FSI2 ID Signed-off-by: Kuninori Morimoto Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/board-ap4evb.c b/arch/arm/mach-shmobile/board-ap4evb.c index 8afdb1d..1b95f12 100644 --- a/arch/arm/mach-shmobile/board-ap4evb.c +++ b/arch/arm/mach-shmobile/board-ap4evb.c @@ -559,7 +559,7 @@ static struct resource fsi_resources[] = { static struct platform_device fsi_device = { .name = "sh_fsi2", - .id = 0, + .id = -1, .num_resources = ARRAY_SIZE(fsi_resources), .resource = fsi_resources, .dev = { -- cgit v0.10.2 From 7cacfa87d3018c68bef2defeda8948f753dfa9d0 Mon Sep 17 00:00:00 2001 From: David Gow Date: Thu, 19 Aug 2010 15:43:56 +0800 Subject: HID: Add support for chicony multitouch screens. Adds a hid quirk for the chicony multitouch screen found in the Acer Aspire 1820pt notebook. Signed-off-by: David Gow Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 85c6d13..2905142 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -149,6 +149,7 @@ #define USB_VENDOR_ID_CHICONY 0x04f2 #define USB_DEVICE_ID_CHICONY_TACTICAL_PAD 0x0418 +#define USB_DEVICE_ID_CHICONY_MULTI_TOUCH 0xb19d #define USB_VENDOR_ID_CIDC 0x1677 diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c index 2643d31..1b22239 100644 --- a/drivers/hid/usbhid/hid-quirks.c +++ b/drivers/hid/usbhid/hid-quirks.c @@ -77,6 +77,8 @@ static const struct hid_blacklist { { USB_VENDOR_ID_PI_ENGINEERING, USB_DEVICE_ID_PI_ENGINEERING_VEC_USB_FOOTPEDAL, HID_QUIRK_HIDINPUT_FORCE }, + { USB_VENDOR_ID_CHICONY, USB_DEVICE_ID_CHICONY_MULTI_TOUCH, HID_QUIRK_MULTI_INPUT }, + { 0, 0 } }; -- cgit v0.10.2 From 410d878bd05315a19894b8b4685107f2cd0712a5 Mon Sep 17 00:00:00 2001 From: Arnd Hannemann Date: Wed, 18 Aug 2010 16:24:13 +0000 Subject: ARM: mach-shmobile: ap4evb: fix write protect for SDHI1 SDHI1 on AP4EVB has no pin to detect write protection of a card. This had the result that a card inserted in this slot was always detected as read only. This patch adds the corresponding flag to disable write protection detection for SDHI1. Signed-off-by: Arnd Hannemann Signed-off-by: Paul Mundt diff --git a/arch/arm/mach-shmobile/board-ap4evb.c b/arch/arm/mach-shmobile/board-ap4evb.c index 1b95f12..95935c8 100644 --- a/arch/arm/mach-shmobile/board-ap4evb.c +++ b/arch/arm/mach-shmobile/board-ap4evb.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include @@ -308,6 +309,7 @@ static struct sh_mobile_sdhi_info sdhi1_info = { .dma_slave_tx = SHDMA_SLAVE_SDHI1_TX, .dma_slave_rx = SHDMA_SLAVE_SDHI1_RX, .tmio_ocr_mask = MMC_VDD_165_195, + .tmio_flags = TMIO_MMC_WRPROTECT_DISABLE, }; static struct resource sdhi1_resources[] = { -- cgit v0.10.2 From 51e3c1b558b31b11bf5fc66d3c6f5adacf3573f7 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Fri, 20 Aug 2010 10:36:34 +0300 Subject: x86, hwmon: Fix unsafe smp_processor_id() in thermal_throttle_add_dev Fix BUG: using smp_processor_id() in preemptible thermal_throttle_add_dev. We know the cpu number when calling thermal_throttle_add_dev, so we can remove smp_processor_id call in thermal_throttle_add_dev by supplying the cpu number as argument. This should resolve kernel bugzilla 16615/16629. Signed-off-by: Sergey Senozhatsky LKML-Reference: <20100820073634.GB5209@swordfish.minsk.epam.com> Cc: Fenghua Yu Cc: Joerg Roedel Cc: Maciej Rutecki Signed-off-by: H. Peter Anvin diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c index c2a8b26..d9368ee 100644 --- a/arch/x86/kernel/cpu/mcheck/therm_throt.c +++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c @@ -202,10 +202,11 @@ static int therm_throt_process(bool new_event, int event, int level) #ifdef CONFIG_SYSFS /* Add/Remove thermal_throttle interface for CPU device: */ -static __cpuinit int thermal_throttle_add_dev(struct sys_device *sys_dev) +static __cpuinit int thermal_throttle_add_dev(struct sys_device *sys_dev, + unsigned int cpu) { int err; - struct cpuinfo_x86 *c = &cpu_data(smp_processor_id()); + struct cpuinfo_x86 *c = &cpu_data(cpu); err = sysfs_create_group(&sys_dev->kobj, &thermal_attr_group); if (err) @@ -251,7 +252,7 @@ thermal_throttle_cpu_callback(struct notifier_block *nfb, case CPU_UP_PREPARE: case CPU_UP_PREPARE_FROZEN: mutex_lock(&therm_cpu_lock); - err = thermal_throttle_add_dev(sys_dev); + err = thermal_throttle_add_dev(sys_dev, cpu); mutex_unlock(&therm_cpu_lock); WARN_ON(err); break; @@ -287,7 +288,7 @@ static __init int thermal_throttle_init_device(void) #endif /* connect live CPUs to sysfs */ for_each_online_cpu(cpu) { - err = thermal_throttle_add_dev(get_cpu_sysdev(cpu)); + err = thermal_throttle_add_dev(get_cpu_sysdev(cpu), cpu); WARN_ON(err); } #ifdef CONFIG_HOTPLUG_CPU -- cgit v0.10.2 From a38b372fc2a3d8fe7ddbeab94d4c8bcb721f2403 Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Sat, 21 Aug 2010 16:24:03 +0800 Subject: mxc/tzic: add base address when accessing TZIC registers When we call tzic_enable_wake function, the kernel will crash because of access to an unmapped address. This is because two register access operations forgot to add base address. Signed-off-by: Jason Wang Signed-off-by: Sascha Hauer diff --git a/arch/arm/plat-mxc/tzic.c b/arch/arm/plat-mxc/tzic.c index b3da9aa..3703ab2 100644 --- a/arch/arm/plat-mxc/tzic.c +++ b/arch/arm/plat-mxc/tzic.c @@ -164,8 +164,9 @@ int tzic_enable_wake(int is_idle) return -EAGAIN; for (i = 0; i < 4; i++) { - v = is_idle ? __raw_readl(TZIC_ENSET0(i)) : wakeup_intr[i]; - __raw_writel(v, TZIC_WAKEUP0(i)); + v = is_idle ? __raw_readl(tzic_base + TZIC_ENSET0(i)) : + wakeup_intr[i]; + __raw_writel(v, tzic_base + TZIC_WAKEUP0(i)); } return 0; -- cgit v0.10.2 From 7f67bff58662604874778e6853e5c394f944e554 Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Sat, 21 Aug 2010 16:24:04 +0800 Subject: mx5/clock: fix clear bit fields issue in _clk_ccgr_disable function We can see MXC_CCM_CCGRx_MOD_OFF is defined as 0 while MXC_CCM_CCGRx_CG_MASK is defined as 0x3 in crm_regs.h, here in the _clk_ccgr_disable function, we want to clear the corresponding enable bit fields to disable this clock, so we should choose MASK instead of OFF otherwise clocks can't be disabled. Signed-off-by: Jason Wang Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx5/clock-mx51.c b/arch/arm/mach-mx5/clock-mx51.c index 6af69de..57c10a9 100644 --- a/arch/arm/mach-mx5/clock-mx51.c +++ b/arch/arm/mach-mx5/clock-mx51.c @@ -56,7 +56,7 @@ static void _clk_ccgr_disable(struct clk *clk) { u32 reg; reg = __raw_readl(clk->enable_reg); - reg &= ~(MXC_CCM_CCGRx_MOD_OFF << clk->enable_shift); + reg &= ~(MXC_CCM_CCGRx_CG_MASK << clk->enable_shift); __raw_writel(reg, clk->enable_reg); } -- cgit v0.10.2 From 14f0f512ba6f4eccc9b7cf67be47439ed2917afb Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Sat, 21 Aug 2010 16:24:05 +0800 Subject: ARM: imx: set cache line size to 64 bytes for i.MX5 The core of i.MX5 series is cortex-A8, its cache line size is 64 bytes instead of 32 bytes. Refer to the OMAP3's selection, we choose 64 bytes for i.MX5, this can increase a little bit performance when perform cache operations. Signed-off-by: Jason Wang Signed-off-by: Sascha Hauer diff --git a/arch/arm/plat-mxc/Kconfig b/arch/arm/plat-mxc/Kconfig index 0527e65..6785db4 100644 --- a/arch/arm/plat-mxc/Kconfig +++ b/arch/arm/plat-mxc/Kconfig @@ -43,6 +43,7 @@ config ARCH_MXC91231 config ARCH_MX5 bool "MX5-based" select CPU_V7 + select ARM_L1_CACHE_SHIFT_6 help This enables support for systems based on the Freescale i.MX51 family -- cgit v0.10.2 From 96aa1b419d47286db446f292cf898bb1a8b27f24 Mon Sep 17 00:00:00 2001 From: Ciju Rajan K Date: Mon, 23 Aug 2010 10:56:30 +0200 Subject: blkio: Fix return code for mkdir calls If the cgroup hierarchy for blkio control groups is deeper than two levels, kernel should not allow the creation of further levels. mkdir system call does not except EINVAL as a return value. This patch replaces EINVAL with more appropriate EPERM Signed-off-by: Ciju Rajan K Reviewed-by: KAMEZAWA Hiroyuki Signed-off-by: Jens Axboe diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c index a680964..2fef1ef 100644 --- a/block/blk-cgroup.c +++ b/block/blk-cgroup.c @@ -966,7 +966,7 @@ blkiocg_create(struct cgroup_subsys *subsys, struct cgroup *cgroup) /* Currently we do not support hierarchy deeper than two level (0,1) */ if (parent != cgroup->top_cgroup) - return ERR_PTR(-EINVAL); + return ERR_PTR(-EPERM); blkcg = kzalloc(sizeof(*blkcg), GFP_KERNEL); if (!blkcg) -- cgit v0.10.2 From 75230ff2751e88d594a13a70eae2c146f45e323b Mon Sep 17 00:00:00 2001 From: "Stephen M. Cameron" Date: Mon, 23 Aug 2010 11:02:17 +0200 Subject: cciss: disable doorbell reset on reset_devices The doorbell reset initially appears to work correctly, the controller resets, comes up, some i/o can even be done, but on at least some Smart Arrays in some servers, it eventually causes a subsequent controller lockup due to some kind of PCIe error, and kdump can end up leaving the root filesystem in an unbootable state. For this reason, until the problem is fixed, or at least isolated to certain hardware enough to be avoided, the doorbell reset should not be used at all. Signed-off-by: Stephen M. Cameron Signed-off-by: Jens Axboe diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 31064df..ce1a75d 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -4519,6 +4519,12 @@ static __devinit int cciss_kdump_hard_reset_controller(struct pci_dev *pdev) misc_fw_support = readl(&cfgtable->misc_fw_support); use_doorbell = misc_fw_support & MISC_FW_DOORBELL_RESET; + /* The doorbell reset seems to cause lockups on some Smart + * Arrays (e.g. P410, P410i, maybe others). Until this is + * fixed or at least isolated, avoid the doorbell reset. + */ + use_doorbell = 0; + rc = cciss_controller_hard_reset(pdev, vaddr, use_doorbell); if (rc) goto unmap_cfgtable; -- cgit v0.10.2 From b6508c1618e7aab085f191efb41b7b019a94ea38 Mon Sep 17 00:00:00 2001 From: Vivek Goyal Date: Mon, 23 Aug 2010 12:23:33 +0200 Subject: cfq-iosched: Do not idle if slice_idle=0 Do not idle either on cfq queue or service tree if slice_idle=0. User does not want any queue or service tree idling. Currently even if slice_idle=0, we were waiting for request to finish before expiring the queue and that can lead to lower queue depths. Acked-by: Jeff Moyer Signed-off-by: Vivek Goyal Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index eb4086f..8830569 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1839,6 +1839,9 @@ static bool cfq_should_idle(struct cfq_data *cfqd, struct cfq_queue *cfqq) BUG_ON(!service_tree); BUG_ON(!service_tree->count); + if (!cfqd->cfq_slice_idle) + return false; + /* We never do for idle class queues. */ if (prio == IDLE_WORKLOAD) return false; @@ -1879,7 +1882,7 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd) /* * idle is disabled, either manually or by past process history */ - if (!cfqd->cfq_slice_idle || !cfq_should_idle(cfqd, cfqq)) + if (!cfq_should_idle(cfqd, cfqq)) return; /* -- cgit v0.10.2 From 02b35081fc98f681411586d3acf9eaad8b8f6e07 Mon Sep 17 00:00:00 2001 From: Vivek Goyal Date: Mon, 23 Aug 2010 12:23:53 +0200 Subject: cfq-iosched: Do group share accounting in IOPS when slice_idle=0 o Implement another CFQ mode where we charge group in terms of number of requests dispatched instead of measuring the time. Measuring in terms of time is not possible when we are driving deeper queue depths and there are requests from multiple cfq queues in the request queue. o This mode currently gets activated if one sets slice_idle=0 and associated disk supports NCQ. Again the idea is that on an NCQ disk with idling disabled most of the queues will dispatch 1 or more requests and then cfq queue expiry happens and we don't have a way to measure time. So start providing fairness in terms of IOPS. o Currently IOPS mode works only with cfq group scheduling. CFQ is following different scheduling algorithms for queue and group scheduling. These IOPS stats are used only for group scheduling hence in non-croup mode nothing should change. o For CFQ group scheduling one can disable slice idling so that we don't idle on queue and drive deeper request queue depths (achieving better throughput), at the same time group idle is enabled so one should get service differentiation among groups. Signed-off-by: Vivek Goyal Acked-by: Jeff Moyer Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 8830569..3fc6be1 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -378,6 +378,21 @@ CFQ_CFQQ_FNS(wait_busy); &cfqg->service_trees[i][j]: NULL) \ +static inline bool iops_mode(struct cfq_data *cfqd) +{ + /* + * If we are not idling on queues and it is a NCQ drive, parallel + * execution of requests is on and measuring time is not possible + * in most of the cases until and unless we drive shallower queue + * depths and that becomes a performance bottleneck. In such cases + * switch to start providing fairness in terms of number of IOs. + */ + if (!cfqd->cfq_slice_idle && cfqd->hw_tag) + return true; + else + return false; +} + static inline enum wl_prio_t cfqq_prio(struct cfq_queue *cfqq) { if (cfq_class_idle(cfqq)) @@ -906,7 +921,6 @@ static inline unsigned int cfq_cfqq_slice_usage(struct cfq_queue *cfqq) slice_used = cfqq->allocated_slice; } - cfq_log_cfqq(cfqq->cfqd, cfqq, "sl_used=%u", slice_used); return slice_used; } @@ -914,19 +928,21 @@ static void cfq_group_served(struct cfq_data *cfqd, struct cfq_group *cfqg, struct cfq_queue *cfqq) { struct cfq_rb_root *st = &cfqd->grp_service_tree; - unsigned int used_sl, charge_sl; + unsigned int used_sl, charge; int nr_sync = cfqg->nr_cfqq - cfqg_busy_async_queues(cfqd, cfqg) - cfqg->service_tree_idle.count; BUG_ON(nr_sync < 0); - used_sl = charge_sl = cfq_cfqq_slice_usage(cfqq); + used_sl = charge = cfq_cfqq_slice_usage(cfqq); - if (!cfq_cfqq_sync(cfqq) && !nr_sync) - charge_sl = cfqq->allocated_slice; + if (iops_mode(cfqd)) + charge = cfqq->slice_dispatch; + else if (!cfq_cfqq_sync(cfqq) && !nr_sync) + charge = cfqq->allocated_slice; /* Can't update vdisktime while group is on service tree */ cfq_rb_erase(&cfqg->rb_node, st); - cfqg->vdisktime += cfq_scale_slice(charge_sl, cfqg); + cfqg->vdisktime += cfq_scale_slice(charge, cfqg); __cfq_group_service_tree_add(st, cfqg); /* This group is being expired. Save the context */ @@ -940,6 +956,8 @@ static void cfq_group_served(struct cfq_data *cfqd, struct cfq_group *cfqg, cfq_log_cfqg(cfqd, cfqg, "served: vt=%llu min_vt=%llu", cfqg->vdisktime, st->min_vdisktime); + cfq_log_cfqq(cfqq->cfqd, cfqq, "sl_used=%u disp=%u charge=%u iops=%u", + used_sl, cfqq->slice_dispatch, charge, iops_mode(cfqd)); cfq_blkiocg_update_timeslice_used(&cfqg->blkg, used_sl); cfq_blkiocg_set_start_empty_time(&cfqg->blkg); } -- cgit v0.10.2 From 80bdf0c78fff075dfa21576273c8b0c7db22bdfe Mon Sep 17 00:00:00 2001 From: Vivek Goyal Date: Mon, 23 Aug 2010 12:24:26 +0200 Subject: cfq-iosched: Implement tunable group_idle o Implement a new tunable group_idle, which allows idling on the group instead of a cfq queue. Hence one can set slice_idle = 0 and not idle on the individual queues but idle on the group. This way on fast storage we can get fairness between groups at the same time overall throughput improves. Signed-off-by: Vivek Goyal Acked-by: Jeff Moyer Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 3fc6be1..85e4819 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -30,6 +30,7 @@ static const int cfq_slice_sync = HZ / 10; static int cfq_slice_async = HZ / 25; static const int cfq_slice_async_rq = 2; static int cfq_slice_idle = HZ / 125; +static int cfq_group_idle = HZ / 125; static const int cfq_target_latency = HZ * 3/10; /* 300 ms */ static const int cfq_hist_divisor = 4; @@ -198,6 +199,8 @@ struct cfq_group { struct hlist_node cfqd_node; atomic_t ref; #endif + /* number of requests that are on the dispatch list or inside driver */ + int dispatched; }; /* @@ -271,6 +274,7 @@ struct cfq_data { unsigned int cfq_slice[2]; unsigned int cfq_slice_async_rq; unsigned int cfq_slice_idle; + unsigned int cfq_group_idle; unsigned int cfq_latency; unsigned int cfq_group_isolation; @@ -1884,7 +1888,7 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd) { struct cfq_queue *cfqq = cfqd->active_queue; struct cfq_io_context *cic; - unsigned long sl; + unsigned long sl, group_idle = 0; /* * SSD device without seek penalty, disable idling. But only do so @@ -1900,8 +1904,13 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd) /* * idle is disabled, either manually or by past process history */ - if (!cfq_should_idle(cfqd, cfqq)) - return; + if (!cfq_should_idle(cfqd, cfqq)) { + /* no queue idling. Check for group idling */ + if (cfqd->cfq_group_idle) + group_idle = cfqd->cfq_group_idle; + else + return; + } /* * still active requests from this queue, don't idle @@ -1928,13 +1937,21 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd) return; } + /* There are other queues in the group, don't do group idle */ + if (group_idle && cfqq->cfqg->nr_cfqq > 1) + return; + cfq_mark_cfqq_wait_request(cfqq); - sl = cfqd->cfq_slice_idle; + if (group_idle) + sl = cfqd->cfq_group_idle; + else + sl = cfqd->cfq_slice_idle; mod_timer(&cfqd->idle_slice_timer, jiffies + sl); cfq_blkiocg_update_set_idle_time_stats(&cfqq->cfqg->blkg); - cfq_log_cfqq(cfqd, cfqq, "arm_idle: %lu", sl); + cfq_log_cfqq(cfqd, cfqq, "arm_idle: %lu group_idle: %d", sl, + group_idle ? 1 : 0); } /* @@ -1950,6 +1967,7 @@ static void cfq_dispatch_insert(struct request_queue *q, struct request *rq) cfqq->next_rq = cfq_find_next_rq(cfqd, cfqq, rq); cfq_remove_request(rq); cfqq->dispatched++; + (RQ_CFQG(rq))->dispatched++; elv_dispatch_sort(q, rq); cfqd->rq_in_flight[cfq_cfqq_sync(cfqq)]++; @@ -2219,7 +2237,7 @@ static struct cfq_queue *cfq_select_queue(struct cfq_data *cfqd) cfqq = NULL; goto keep_queue; } else - goto expire; + goto check_group_idle; } /* @@ -2247,8 +2265,23 @@ static struct cfq_queue *cfq_select_queue(struct cfq_data *cfqd) * flight or is idling for a new request, allow either of these * conditions to happen (or time out) before selecting a new queue. */ - if (timer_pending(&cfqd->idle_slice_timer) || - (cfqq->dispatched && cfq_should_idle(cfqd, cfqq))) { + if (timer_pending(&cfqd->idle_slice_timer)) { + cfqq = NULL; + goto keep_queue; + } + + if (cfqq->dispatched && cfq_should_idle(cfqd, cfqq)) { + cfqq = NULL; + goto keep_queue; + } + + /* + * If group idle is enabled and there are requests dispatched from + * this group, wait for requests to complete. + */ +check_group_idle: + if (cfqd->cfq_group_idle && cfqq->cfqg->nr_cfqq == 1 + && cfqq->cfqg->dispatched) { cfqq = NULL; goto keep_queue; } @@ -3396,6 +3429,7 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq) WARN_ON(!cfqq->dispatched); cfqd->rq_in_driver--; cfqq->dispatched--; + (RQ_CFQG(rq))->dispatched--; cfq_blkiocg_update_completion_stats(&cfqq->cfqg->blkg, rq_start_time_ns(rq), rq_io_start_time_ns(rq), rq_data_dir(rq), rq_is_sync(rq)); @@ -3425,7 +3459,10 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq) * the queue. */ if (cfq_should_wait_busy(cfqd, cfqq)) { - cfqq->slice_end = jiffies + cfqd->cfq_slice_idle; + unsigned long extend_sl = cfqd->cfq_slice_idle; + if (!cfqd->cfq_slice_idle) + extend_sl = cfqd->cfq_group_idle; + cfqq->slice_end = jiffies + extend_sl; cfq_mark_cfqq_wait_busy(cfqq); cfq_log_cfqq(cfqd, cfqq, "will busy wait"); } @@ -3871,6 +3908,7 @@ static void *cfq_init_queue(struct request_queue *q) cfqd->cfq_slice[1] = cfq_slice_sync; cfqd->cfq_slice_async_rq = cfq_slice_async_rq; cfqd->cfq_slice_idle = cfq_slice_idle; + cfqd->cfq_group_idle = cfq_group_idle; cfqd->cfq_latency = 1; cfqd->cfq_group_isolation = 0; cfqd->hw_tag = -1; @@ -3943,6 +3981,7 @@ SHOW_FUNCTION(cfq_fifo_expire_async_show, cfqd->cfq_fifo_expire[0], 1); SHOW_FUNCTION(cfq_back_seek_max_show, cfqd->cfq_back_max, 0); SHOW_FUNCTION(cfq_back_seek_penalty_show, cfqd->cfq_back_penalty, 0); SHOW_FUNCTION(cfq_slice_idle_show, cfqd->cfq_slice_idle, 1); +SHOW_FUNCTION(cfq_group_idle_show, cfqd->cfq_group_idle, 1); SHOW_FUNCTION(cfq_slice_sync_show, cfqd->cfq_slice[1], 1); SHOW_FUNCTION(cfq_slice_async_show, cfqd->cfq_slice[0], 1); SHOW_FUNCTION(cfq_slice_async_rq_show, cfqd->cfq_slice_async_rq, 0); @@ -3975,6 +4014,7 @@ STORE_FUNCTION(cfq_back_seek_max_store, &cfqd->cfq_back_max, 0, UINT_MAX, 0); STORE_FUNCTION(cfq_back_seek_penalty_store, &cfqd->cfq_back_penalty, 1, UINT_MAX, 0); STORE_FUNCTION(cfq_slice_idle_store, &cfqd->cfq_slice_idle, 0, UINT_MAX, 1); +STORE_FUNCTION(cfq_group_idle_store, &cfqd->cfq_group_idle, 0, UINT_MAX, 1); STORE_FUNCTION(cfq_slice_sync_store, &cfqd->cfq_slice[1], 1, UINT_MAX, 1); STORE_FUNCTION(cfq_slice_async_store, &cfqd->cfq_slice[0], 1, UINT_MAX, 1); STORE_FUNCTION(cfq_slice_async_rq_store, &cfqd->cfq_slice_async_rq, 1, @@ -3996,6 +4036,7 @@ static struct elv_fs_entry cfq_attrs[] = { CFQ_ATTR(slice_async), CFQ_ATTR(slice_async_rq), CFQ_ATTR(slice_idle), + CFQ_ATTR(group_idle), CFQ_ATTR(low_latency), CFQ_ATTR(group_isolation), __ATTR_NULL @@ -4049,6 +4090,12 @@ static int __init cfq_init(void) if (!cfq_slice_idle) cfq_slice_idle = 1; +#ifdef CONFIG_CFQ_GROUP_IOSCHED + if (!cfq_group_idle) + cfq_group_idle = 1; +#else + cfq_group_idle = 0; +#endif if (cfq_slab_setup()) return -ENOMEM; -- cgit v0.10.2 From c4e7893ebc3a5c507b53f59b9de448db20849944 Mon Sep 17 00:00:00 2001 From: Vivek Goyal Date: Mon, 23 Aug 2010 12:25:03 +0200 Subject: cfq-iosched: blktrace print per slice sector stats o Divyesh had gotten rid of this code in the past. I want to re-introduce it back as it helps me a lot during debugging. Reviewed-by: Jeff Moyer Reviewed-by: Divyesh Shah Signed-off-by: Vivek Goyal Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 85e4819..f65c6f0 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -148,6 +148,8 @@ struct cfq_queue { struct cfq_queue *new_cfqq; struct cfq_group *cfqg; struct cfq_group *orig_cfqg; + /* Number of sectors dispatched from queue in single dispatch round */ + unsigned long nr_sectors; }; /* @@ -960,8 +962,9 @@ static void cfq_group_served(struct cfq_data *cfqd, struct cfq_group *cfqg, cfq_log_cfqg(cfqd, cfqg, "served: vt=%llu min_vt=%llu", cfqg->vdisktime, st->min_vdisktime); - cfq_log_cfqq(cfqq->cfqd, cfqq, "sl_used=%u disp=%u charge=%u iops=%u", - used_sl, cfqq->slice_dispatch, charge, iops_mode(cfqd)); + cfq_log_cfqq(cfqq->cfqd, cfqq, "sl_used=%u disp=%u charge=%u iops=%u" + " sect=%u", used_sl, cfqq->slice_dispatch, charge, + iops_mode(cfqd), cfqq->nr_sectors); cfq_blkiocg_update_timeslice_used(&cfqg->blkg, used_sl); cfq_blkiocg_set_start_empty_time(&cfqg->blkg); } @@ -1609,6 +1612,7 @@ static void __cfq_set_active_queue(struct cfq_data *cfqd, cfqq->allocated_slice = 0; cfqq->slice_end = 0; cfqq->slice_dispatch = 0; + cfqq->nr_sectors = 0; cfq_clear_cfqq_wait_request(cfqq); cfq_clear_cfqq_must_dispatch(cfqq); @@ -1971,6 +1975,7 @@ static void cfq_dispatch_insert(struct request_queue *q, struct request *rq) elv_dispatch_sort(q, rq); cfqd->rq_in_flight[cfq_cfqq_sync(cfqq)]++; + cfqq->nr_sectors += blk_rq_sectors(rq); cfq_blkiocg_update_dispatch_stats(&cfqq->cfqg->blkg, blk_rq_bytes(rq), rq_data_dir(rq), rq_is_sync(rq)); } -- cgit v0.10.2 From 6d6ac1c1a3d4f95953aa3b085e8f16692d3a7179 Mon Sep 17 00:00:00 2001 From: Vivek Goyal Date: Mon, 23 Aug 2010 12:25:29 +0200 Subject: cfq-iosched: Documentation help for new tunables Some documentation to provide help with tunables. Signed-off-by: Vivek Goyal Acked-by: Jeff Moyer Signed-off-by: Jens Axboe diff --git a/Documentation/block/cfq-iosched.txt b/Documentation/block/cfq-iosched.txt new file mode 100644 index 0000000..e578fee --- /dev/null +++ b/Documentation/block/cfq-iosched.txt @@ -0,0 +1,45 @@ +CFQ ioscheduler tunables +======================== + +slice_idle +---------- +This specifies how long CFQ should idle for next request on certain cfq queues +(for sequential workloads) and service trees (for random workloads) before +queue is expired and CFQ selects next queue to dispatch from. + +By default slice_idle is a non-zero value. That means by default we idle on +queues/service trees. This can be very helpful on highly seeky media like +single spindle SATA/SAS disks where we can cut down on overall number of +seeks and see improved throughput. + +Setting slice_idle to 0 will remove all the idling on queues/service tree +level and one should see an overall improved throughput on faster storage +devices like multiple SATA/SAS disks in hardware RAID configuration. The down +side is that isolation provided from WRITES also goes down and notion of +IO priority becomes weaker. + +So depending on storage and workload, it might be useful to set slice_idle=0. +In general I think for SATA/SAS disks and software RAID of SATA/SAS disks +keeping slice_idle enabled should be useful. For any configurations where +there are multiple spindles behind single LUN (Host based hardware RAID +controller or for storage arrays), setting slice_idle=0 might end up in better +throughput and acceptable latencies. + +CFQ IOPS Mode for group scheduling +=================================== +Basic CFQ design is to provide priority based time slices. Higher priority +process gets bigger time slice and lower priority process gets smaller time +slice. Measuring time becomes harder if storage is fast and supports NCQ and +it would be better to dispatch multiple requests from multiple cfq queues in +request queue at a time. In such scenario, it is not possible to measure time +consumed by single queue accurately. + +What is possible though is to measure number of requests dispatched from a +single queue and also allow dispatch from multiple cfq queue at the same time. +This effectively becomes the fairness in terms of IOPS (IO operations per +second). + +If one sets slice_idle=0 and if storage supports NCQ, CFQ internally switches +to IOPS mode and starts providing fairness in terms of number of requests +dispatched. Note that this mode switching takes effect only for group +scheduling. For non-cgroup users nothing should change. diff --git a/Documentation/cgroups/blkio-controller.txt b/Documentation/cgroups/blkio-controller.txt index 48e0b21..6919d62 100644 --- a/Documentation/cgroups/blkio-controller.txt +++ b/Documentation/cgroups/blkio-controller.txt @@ -217,6 +217,7 @@ Details of cgroup files CFQ sysfs tunable ================= /sys/block//queue/iosched/group_isolation +----------------------------------------------- If group_isolation=1, it provides stronger isolation between groups at the expense of throughput. By default group_isolation is 0. In general that @@ -243,6 +244,33 @@ By default one should run with group_isolation=0. If that is not sufficient and one wants stronger isolation between groups, then set group_isolation=1 but this will come at cost of reduced throughput. +/sys/block//queue/iosched/slice_idle +------------------------------------------ +On a faster hardware CFQ can be slow, especially with sequential workload. +This happens because CFQ idles on a single queue and single queue might not +drive deeper request queue depths to keep the storage busy. In such scenarios +one can try setting slice_idle=0 and that would switch CFQ to IOPS +(IO operations per second) mode on NCQ supporting hardware. + +That means CFQ will not idle between cfq queues of a cfq group and hence be +able to driver higher queue depth and achieve better throughput. That also +means that cfq provides fairness among groups in terms of IOPS and not in +terms of disk time. + +/sys/block//queue/iosched/group_idle +------------------------------------------ +If one disables idling on individual cfq queues and cfq service trees by +setting slice_idle=0, group_idle kicks in. That means CFQ will still idle +on the group in an attempt to provide fairness among groups. + +By default group_idle is same as slice_idle and does not do anything if +slice_idle is enabled. + +One can experience an overall throughput drop if you have created multiple +groups and put applications in that group which are not driving enough +IO to keep disk busy. In that case set group_idle=0, and CFQ will not idle +on individual groups and throughput should improve. + What works ========== - Currently only sync IO queues are support. All the buffered writes are -- cgit v0.10.2 From 4ee69851cd4880f574d22f5ce08bec35b01c94e3 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 23 Aug 2010 12:28:15 +0200 Subject: cciss: handle allocation failure If kmalloc() fails then cleanup and return failure (-1). Signed-off-by: Dan Carpenter Acked-by: Stephen M. Cameron Signed-off-by: Jens Axboe diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index ce1a75d..7191c16 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -4718,6 +4718,9 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, h->scatter_list = kmalloc(h->max_commands * sizeof(struct scatterlist *), GFP_KERNEL); + if (!h->scatter_list) + goto clean4; + for (k = 0; k < h->nr_cmds; k++) { h->scatter_list[k] = kmalloc(sizeof(struct scatterlist) * h->maxsgentries, -- cgit v0.10.2 From c87ffbb812cf6150097a5095b031f4013a8f78a5 Mon Sep 17 00:00:00 2001 From: Xiaotian Feng Date: Mon, 23 Aug 2010 12:30:29 +0200 Subject: block: put dev->kobj in blk_register_queue fail path kernel needs to kobject_put on dev->kobj if elv_register_queue fails. Signed-off-by: Xiaotian Feng Cc: "Martin K. Petersen" Cc: Stephen Hemminger Cc: Nikanth Karthikesan Cc: David Teigland Signed-off-by: Jens Axboe diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 001ab18..0749b89 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -511,6 +511,7 @@ int blk_register_queue(struct gendisk *disk) kobject_uevent(&q->kobj, KOBJ_REMOVE); kobject_del(&q->kobj); blk_trace_remove_sysfs(disk_to_dev(disk)); + kobject_put(&dev->kobj); return ret; } -- cgit v0.10.2 From 5e00d1b5b4c10fb839afd5ce61db8e24339454b0 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Thu, 12 Aug 2010 14:31:06 +0200 Subject: BLOCK: fix bio.bi_rw handling Return of the bi_rw tests is no longer bool after commit 74450be1. But results of such tests are stored in bools. This doesn't fit in there for some compilers (gcc 4.5 here), so either use !! magic to get real bools or use ulong where the result is assigned somewhere. Signed-off-by: Jiri Slaby Cc: Christoph Hellwig Reviewed-by: Jeff Moyer Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index ee1a1e7..32a1c12 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -1198,9 +1198,9 @@ static int __make_request(struct request_queue *q, struct bio *bio) int el_ret; unsigned int bytes = bio->bi_size; const unsigned short prio = bio_prio(bio); - const bool sync = (bio->bi_rw & REQ_SYNC); - const bool unplug = (bio->bi_rw & REQ_UNPLUG); - const unsigned int ff = bio->bi_rw & REQ_FAILFAST_MASK; + const bool sync = !!(bio->bi_rw & REQ_SYNC); + const bool unplug = !!(bio->bi_rw & REQ_UNPLUG); + const unsigned long ff = bio->bi_rw & REQ_FAILFAST_MASK; int rw_flags; if ((bio->bi_rw & REQ_HARDBARRIER) && diff --git a/drivers/block/loop.c b/drivers/block/loop.c index f3c636d..91797bb 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -477,7 +477,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio) pos = ((loff_t) bio->bi_sector << 9) + lo->lo_offset; if (bio_rw(bio) == WRITE) { - bool barrier = (bio->bi_rw & REQ_HARDBARRIER); + bool barrier = !!(bio->bi_rw & REQ_HARDBARRIER); struct file *file = lo->lo_backing_file; if (barrier) { -- cgit v0.10.2 From 72f465033702ebfe20db8f50edaad59f0f38b0f5 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Mon, 23 Aug 2010 13:35:09 +0200 Subject: bio-integrity.c: remove dependency on __GFP_NOFAIL The kmalloc() in bio_integrity_prep() is failable, so remove __GFP_NOFAIL from its mask. Signed-off-by: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Jens Axboe diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index 612a5c3..a8f4cc6 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c @@ -413,7 +413,7 @@ int bio_integrity_prep(struct bio *bio) /* Allocate kernel buffer for protection data */ len = sectors * blk_integrity_tuple_size(bi); - buf = kmalloc(len, GFP_NOIO | __GFP_NOFAIL | q->bounce_gfp); + buf = kmalloc(len, GFP_NOIO | q->bounce_gfp); if (unlikely(buf == NULL)) { printk(KERN_ERR "could not allocate integrity buffer\n"); return -EIO; -- cgit v0.10.2 From 220eb7fd984bfc7e6b4005fdf32efe9cd8af7cf2 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Mon, 23 Aug 2010 13:36:20 +0200 Subject: fs/bio-integrity.c: return -ENOMEM on kmalloc failure Cc: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Jens Axboe diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index a8f4cc6..4d0ff5e 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c @@ -416,7 +416,7 @@ int bio_integrity_prep(struct bio *bio) buf = kmalloc(len, GFP_NOIO | q->bounce_gfp); if (unlikely(buf == NULL)) { printk(KERN_ERR "could not allocate integrity buffer\n"); - return -EIO; + return -ENOMEM; } end = (((unsigned long) buf) + len + PAGE_SIZE - 1) >> PAGE_SHIFT; -- cgit v0.10.2 From 5dd531a03ad721b41911ddb32e6e0481404e7aaf Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 23 Aug 2010 13:52:19 +0200 Subject: block: add function call to switch the IO scheduler from a driver Currently drivers must do an elevator_exit() + elevator_init() to switch IO schedulers. There are a few problems with this: - Since commit 1abec4fdbb142e3ccb6ce99832fae42129134a96, elevator_init() requires a zeroed out q->elevator pointer. The two existing in-kernel users don't do that. - It will only work at initialization time, since using the above two-staged construct does not properly quisce the queue. So add elevator_change() which takes care of this, and convert the elv_iosched_store() sysfs interface to use this helper as well. Reported-by: Peter Oberparleiter Reported-by: Kevin Vigor Signed-off-by: Jens Axboe diff --git a/block/elevator.c b/block/elevator.c index ec585c9..205b09a 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -1009,18 +1009,19 @@ static int elevator_switch(struct request_queue *q, struct elevator_type *new_e) { struct elevator_queue *old_elevator, *e; void *data; + int err; /* * Allocate new elevator */ e = elevator_alloc(q, new_e); if (!e) - return 0; + return -ENOMEM; data = elevator_init_queue(q, e); if (!data) { kobject_put(&e->kobj); - return 0; + return -ENOMEM; } /* @@ -1043,7 +1044,8 @@ static int elevator_switch(struct request_queue *q, struct elevator_type *new_e) __elv_unregister_queue(old_elevator); - if (elv_register_queue(q)) + err = elv_register_queue(q); + if (err) goto fail_register; /* @@ -1056,7 +1058,7 @@ static int elevator_switch(struct request_queue *q, struct elevator_type *new_e) blk_add_trace_msg(q, "elv switch: %s", e->elevator_type->elevator_name); - return 1; + return 0; fail_register: /* @@ -1071,17 +1073,19 @@ fail_register: queue_flag_clear(QUEUE_FLAG_ELVSWITCH, q); spin_unlock_irq(q->queue_lock); - return 0; + return err; } -ssize_t elv_iosched_store(struct request_queue *q, const char *name, - size_t count) +/* + * Switch this queue to the given IO scheduler. + */ +int elevator_change(struct request_queue *q, const char *name) { char elevator_name[ELV_NAME_MAX]; struct elevator_type *e; if (!q->elevator) - return count; + return -ENXIO; strlcpy(elevator_name, name, sizeof(elevator_name)); e = elevator_get(strstrip(elevator_name)); @@ -1092,13 +1096,27 @@ ssize_t elv_iosched_store(struct request_queue *q, const char *name, if (!strcmp(elevator_name, q->elevator->elevator_type->elevator_name)) { elevator_put(e); - return count; + return 0; } - if (!elevator_switch(q, e)) - printk(KERN_ERR "elevator: switch to %s failed\n", - elevator_name); - return count; + return elevator_switch(q, e); +} +EXPORT_SYMBOL(elevator_change); + +ssize_t elv_iosched_store(struct request_queue *q, const char *name, + size_t count) +{ + int ret; + + if (!q->elevator) + return count; + + ret = elevator_change(q, name); + if (!ret) + return count; + + printk(KERN_ERR "elevator: switch to %s failed\n", name); + return ret; } ssize_t elv_iosched_show(struct request_queue *q, char *name) diff --git a/include/linux/elevator.h b/include/linux/elevator.h index 2c958f4..926b503 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -136,6 +136,7 @@ extern ssize_t elv_iosched_store(struct request_queue *, const char *, size_t); extern int elevator_init(struct request_queue *, char *); extern void elevator_exit(struct elevator_queue *); +extern int elevator_change(struct request_queue *, const char *); extern int elv_rq_merge_ok(struct request *, struct bio *); /* -- cgit v0.10.2 From 52cc2eef31587b22ce9fbe77b064a031a9613ab0 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 23 Aug 2010 14:02:44 +0200 Subject: block: switch s390 tape_block and mg_disk to elevator_change() Now that we have this API, switch the two in-kernel users to it. Resolves an oops introduced by commit 1abec4fdbb142e3ccb6ce99832fae42129134a96. Signed-off-by: Jens Axboe diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c index b82c5ce..76fa3de 100644 --- a/drivers/block/mg_disk.c +++ b/drivers/block/mg_disk.c @@ -974,8 +974,7 @@ static int mg_probe(struct platform_device *plat_dev) host->breq->queuedata = host; /* mflash is random device, thanx for the noop */ - elevator_exit(host->breq->elevator); - err = elevator_init(host->breq, "noop"); + err = elevator_change(host->breq, "noop"); if (err) { printk(KERN_ERR "%s:%d (elevator_init) fail\n", __func__, __LINE__); diff --git a/drivers/s390/char/tape_block.c b/drivers/s390/char/tape_block.c index b7de025..85cf607 100644 --- a/drivers/s390/char/tape_block.c +++ b/drivers/s390/char/tape_block.c @@ -217,8 +217,7 @@ tapeblock_setup_device(struct tape_device * device) if (!blkdat->request_queue) return -ENOMEM; - elevator_exit(blkdat->request_queue->elevator); - rc = elevator_init(blkdat->request_queue, "noop"); + rc = elevator_change(blkdat->request_queue, "noop"); if (rc) goto cleanup_queue; -- cgit v0.10.2 From 77f4b9fe050d59a30c3b11e267289630bb13f56a Mon Sep 17 00:00:00 2001 From: Shuduo Sang Date: Tue, 24 Aug 2010 14:35:17 +0100 Subject: intel_pmic_battery: Fix battery charging status on mrst The arguments got swapped on some functions which produces undefined results. The main one got fixed before submit but the other two were missed. Signed-off-by: Shuduo Sang Signed-off-by: Alan Cox Signed-off-by: Anton Vorontsov diff --git a/drivers/power/intel_mid_battery.c b/drivers/power/intel_mid_battery.c index c61ffec..2a10cd3 100644 --- a/drivers/power/intel_mid_battery.c +++ b/drivers/power/intel_mid_battery.c @@ -185,8 +185,8 @@ static int pmic_scu_ipc_battery_property_get(struct battery_property *prop) { u32 data[3]; u8 *p = (u8 *)&data[1]; - int err = intel_scu_ipc_command(IPC_CMD_BATTERY_PROPERTY, - IPCMSG_BATTERY, NULL, 0, data, 3); + int err = intel_scu_ipc_command(IPCMSG_BATTERY, + IPC_CMD_BATTERY_PROPERTY, NULL, 0, data, 3); prop->capacity = data[0]; prop->crnt = *p++; @@ -207,7 +207,7 @@ static int pmic_scu_ipc_battery_property_get(struct battery_property *prop) static int pmic_scu_ipc_set_charger(int charger) { - return intel_scu_ipc_simple_command(charger, IPCMSG_BATTERY); + return intel_scu_ipc_simple_command(IPCMSG_BATTERY, charger); } /** -- cgit v0.10.2 From 426409b1edcd7db922dd326911eba23d5a06d098 Mon Sep 17 00:00:00 2001 From: Decio Fonini Date: Tue, 24 Aug 2010 17:45:49 +0200 Subject: HID: Kanvus Note A5 tablet needs HID_QUIRK_MULTI_INPUT The Kanvus Note A5 tablet (with USB ID 5543:6001, USB vendor UC_Logic) needs the HID_QUIRK_MULTI_INPUT in order to work out of the box; otherwise, we get the usual "cursor stuck at the upper left corner of the screen". Signed-off-by: Decio Fonini Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 2905142..b309525 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -508,6 +508,7 @@ #define USB_VENDOR_ID_UCLOGIC 0x5543 #define USB_DEVICE_ID_UCLOGIC_TABLET_PF1209 0x0042 #define USB_DEVICE_ID_UCLOGIC_TABLET_WP4030U 0x0003 +#define USB_DEVICE_ID_UCLOGIC_TABLET_KNA5 0x6001 #define USB_VENDOR_ID_VERNIER 0x08f7 #define USB_DEVICE_ID_VERNIER_LABPRO 0x0001 diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c index 1b22239..89fa4ac 100644 --- a/drivers/hid/usbhid/hid-quirks.c +++ b/drivers/hid/usbhid/hid-quirks.c @@ -69,6 +69,7 @@ static const struct hid_blacklist { { USB_VENDOR_ID_TURBOX, USB_DEVICE_ID_TURBOX_KEYBOARD, HID_QUIRK_NOGET }, { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_PF1209, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_WP4030U, HID_QUIRK_MULTI_INPUT }, + { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_KNA5, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_WISEGROUP, USB_DEVICE_ID_DUAL_USB_JOYPAD, HID_QUIRK_NOGET | HID_QUIRK_MULTI_INPUT | HID_QUIRK_SKIP_OUTPUT_REPORTS }, { USB_VENDOR_ID_WISEGROUP, USB_DEVICE_ID_QUAD_USB_JOYPAD, HID_QUIRK_NOGET | HID_QUIRK_MULTI_INPUT }, -- cgit v0.10.2 From c29771c2d8ceb907ed45eb8c7fc0450308140aca Mon Sep 17 00:00:00 2001 From: Alan Ott Date: Tue, 17 Aug 2010 00:44:04 -0400 Subject: HID: Set Report ID properly for Output reports on the Control endpoint. When I made commit 29129a98e6fc89 ("HID: Send Report ID when numbered reports are sent over the control endpoint"), I didn't account for *buf not being the report ID anymore, as buf is incremented. Signed-off-by: Alan Ott Signed-off-by: Jiri Kosina diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c index b729c02..ffd6899 100644 --- a/drivers/hid/usbhid/hid-core.c +++ b/drivers/hid/usbhid/hid-core.c @@ -828,6 +828,7 @@ static int usbhid_output_raw_report(struct hid_device *hid, __u8 *buf, size_t co } } else { int skipped_report_id = 0; + int report_id = buf[0]; if (buf[0] == 0x0) { /* Don't send the Report ID */ buf++; @@ -837,7 +838,7 @@ static int usbhid_output_raw_report(struct hid_device *hid, __u8 *buf, size_t co ret = usb_control_msg(dev, usb_sndctrlpipe(dev, 0), HID_REQ_SET_REPORT, USB_DIR_OUT | USB_TYPE_CLASS | USB_RECIP_INTERFACE, - ((report_type + 1) << 8) | *buf, + ((report_type + 1) << 8) | report_id, interface->desc.bInterfaceNumber, buf, count, USB_CTRL_SET_TIMEOUT); /* count also the report id, if this was a numbered report. */ -- cgit v0.10.2 From b7d460897739e02f186425b7276e3fdb1595cea7 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Tue, 24 Aug 2010 22:44:12 -0700 Subject: x86, mm: Fix CONFIG_VMSPLIT_1G and 2G_OPT trampoline rc2 kernel crashes when booting second cpu on this CONFIG_VMSPLIT_2G_OPT laptop: whereas cloning from kernel to low mappings pgd range does need to limit by both KERNEL_PGD_PTRS and KERNEL_PGD_BOUNDARY, cloning kernel pgd range itself must not be limited by the smaller KERNEL_PGD_BOUNDARY. Signed-off-by: Hugh Dickins LKML-Reference: Signed-off-by: H. Peter Anvin diff --git a/arch/x86/kernel/trampoline.c b/arch/x86/kernel/trampoline.c index a874495..e2a5952 100644 --- a/arch/x86/kernel/trampoline.c +++ b/arch/x86/kernel/trampoline.c @@ -45,8 +45,7 @@ void __init setup_trampoline_page_table(void) /* Copy kernel address range */ clone_pgd_range(trampoline_pg_dir + KERNEL_PGD_BOUNDARY, swapper_pg_dir + KERNEL_PGD_BOUNDARY, - min_t(unsigned long, KERNEL_PGD_PTRS, - KERNEL_PGD_BOUNDARY)); + KERNEL_PGD_PTRS); /* Initialize low mappings */ clone_pgd_range(trampoline_pg_dir, -- cgit v0.10.2 From 750d857c682f4db60d14722d430c7ccc35070962 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Fri, 13 Aug 2010 16:29:04 +0200 Subject: oprofile: fix crash when accessing freed task structs This patch fixes a crash during shutdown reported below. The crash is caused by accessing already freed task structs. The fix changes the order for registering and unregistering notifier callbacks. All notifiers must be initialized before buffers start working. To stop buffer synchronization we cancel all workqueues, unregister the notifier callback and then flush all buffers. After all of this we finally can free all tasks listed. This should avoid accessing freed tasks. On 22.07.10 01:14:40, Benjamin Herrenschmidt wrote: > So the initial observation is a spinlock bad magic followed by a crash > in the spinlock debug code: > > [ 1541.586531] BUG: spinlock bad magic on CPU#5, events/5/136 > [ 1541.597564] Unable to handle kernel paging request for data at address 0x6b6b6b6b6b6b6d03 > > Backtrace looks like: > > spin_bug+0x74/0xd4 > ._raw_spin_lock+0x48/0x184 > ._spin_lock+0x10/0x24 > .get_task_mm+0x28/0x8c > .sync_buffer+0x1b4/0x598 > .wq_sync_buffer+0xa0/0xdc > .worker_thread+0x1d8/0x2a8 > .kthread+0xa8/0xb4 > .kernel_thread+0x54/0x70 > > So we are accessing a freed task struct in the work queue when > processing the samples. Reported-by: Benjamin Herrenschmidt Cc: stable@kernel.org Signed-off-by: Robert Richter diff --git a/drivers/oprofile/buffer_sync.c b/drivers/oprofile/buffer_sync.c index a9352b2..b7e755f 100644 --- a/drivers/oprofile/buffer_sync.c +++ b/drivers/oprofile/buffer_sync.c @@ -141,16 +141,6 @@ static struct notifier_block module_load_nb = { .notifier_call = module_load_notify, }; - -static void end_sync(void) -{ - end_cpu_work(); - /* make sure we don't leak task structs */ - process_task_mortuary(); - process_task_mortuary(); -} - - int sync_start(void) { int err; @@ -158,7 +148,7 @@ int sync_start(void) if (!zalloc_cpumask_var(&marked_cpus, GFP_KERNEL)) return -ENOMEM; - start_cpu_work(); + mutex_lock(&buffer_mutex); err = task_handoff_register(&task_free_nb); if (err) @@ -173,7 +163,10 @@ int sync_start(void) if (err) goto out4; + start_cpu_work(); + out: + mutex_unlock(&buffer_mutex); return err; out4: profile_event_unregister(PROFILE_MUNMAP, &munmap_nb); @@ -182,7 +175,6 @@ out3: out2: task_handoff_unregister(&task_free_nb); out1: - end_sync(); free_cpumask_var(marked_cpus); goto out; } @@ -190,11 +182,20 @@ out1: void sync_stop(void) { + /* flush buffers */ + mutex_lock(&buffer_mutex); + end_cpu_work(); unregister_module_notifier(&module_load_nb); profile_event_unregister(PROFILE_MUNMAP, &munmap_nb); profile_event_unregister(PROFILE_TASK_EXIT, &task_exit_nb); task_handoff_unregister(&task_free_nb); - end_sync(); + mutex_unlock(&buffer_mutex); + flush_scheduled_work(); + + /* make sure we don't leak task structs */ + process_task_mortuary(); + process_task_mortuary(); + free_cpumask_var(marked_cpus); } diff --git a/drivers/oprofile/cpu_buffer.c b/drivers/oprofile/cpu_buffer.c index 219f79e..f179ac2 100644 --- a/drivers/oprofile/cpu_buffer.c +++ b/drivers/oprofile/cpu_buffer.c @@ -120,8 +120,6 @@ void end_cpu_work(void) cancel_delayed_work(&b->work); } - - flush_scheduled_work(); } /* -- cgit v0.10.2 From 2a643ec67f9efc4b6921a3dd6e257f3b5360622b Mon Sep 17 00:00:00 2001 From: "Stephen M. Cameron" Date: Wed, 25 Aug 2010 19:58:53 +0200 Subject: cciss: fix reporting of max queue depth since init The ioctl path and the scsi tape path were not accounting for their additions to the queue depth. Signed-off-by: Stephen M. Cameron Signed-off-by: Jens Axboe diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 7191c16..6124c2f 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -297,6 +297,8 @@ static void enqueue_cmd_and_start_io(ctlr_info_t *h, spin_lock_irqsave(&h->lock, flags); addQ(&h->reqQ, c); h->Qdepth++; + if (h->Qdepth > h->maxQsinceinit) + h->maxQsinceinit = h->Qdepth; start_io(h); spin_unlock_irqrestore(&h->lock, flags); } -- cgit v0.10.2 From 6cb6cbeaa9d6573c674842b9471c0467967aae99 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 25 Aug 2010 08:58:50 -0700 Subject: MAINTAINERS: Add RCU's public git tree RCU now has a public git tree, so add it via the T: line in the MAINTAINERS file. Signed-off-by: Paul E. McKenney Cc: joe@perches.com LKML-Reference: <20100825155850.GA6601@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar diff --git a/MAINTAINERS b/MAINTAINERS index 5fa8451..614861a 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4799,6 +4799,7 @@ RCUTORTURE MODULE M: Josh Triplett M: "Paul E. McKenney" S: Supported +T: git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-2.6-rcu.git F: Documentation/RCU/torture.txt F: kernel/rcutorture.c @@ -4823,6 +4824,7 @@ M: Dipankar Sarma M: "Paul E. McKenney" W: http://www.rdrop.com/users/paulmck/rclock/ S: Supported +T: git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-2.6-rcu.git F: Documentation/RCU/ F: include/linux/rcu* F: include/linux/srcu* -- cgit v0.10.2 From f7b66e5e51d31c45c6039db9a6da7863fb75be1e Mon Sep 17 00:00:00 2001 From: John Ogness Date: Fri, 18 Jun 2010 18:59:47 +0200 Subject: mxc_nand: Do not do byte accesses to the NFC buffer. This patch avoids byte access to the NFC buffer. Byte access to the NFC is not allowed. The patch is against linux-next 20100618. Signed-off-by: John Ogness Signed-off-by: Sascha Hauer Tested-by: John Ogness Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c index fcf8ceb..26caa01 100644 --- a/drivers/mtd/nand/mxc_nand.c +++ b/drivers/mtd/nand/mxc_nand.c @@ -402,16 +402,16 @@ static void send_read_id_v1_v2(struct mxc_nand_host *host) /* Wait for operation to complete */ wait_op_done(host, true); + memcpy(host->data_buf, host->main_area0, 16); + if (this->options & NAND_BUSWIDTH_16) { - void __iomem *main_buf = host->main_area0; /* compress the ID info */ - writeb(readb(main_buf + 2), main_buf + 1); - writeb(readb(main_buf + 4), main_buf + 2); - writeb(readb(main_buf + 6), main_buf + 3); - writeb(readb(main_buf + 8), main_buf + 4); - writeb(readb(main_buf + 10), main_buf + 5); + host->data_buf[1] = host->data_buf[2]; + host->data_buf[2] = host->data_buf[4]; + host->data_buf[3] = host->data_buf[6]; + host->data_buf[4] = host->data_buf[8]; + host->data_buf[5] = host->data_buf[10]; } - memcpy(host->data_buf, host->main_area0, 16); } static uint16_t get_dev_status_v3(struct mxc_nand_host *host) -- cgit v0.10.2 From 5225c45899e872383ca39f5533d28ec63c54b39e Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 22 Aug 2010 04:29:17 +0200 Subject: perf: Initialize callchains roots's childen hits Each histogram entry has a callchain root that stores the callchain samples. However we forgot to initialize the tracking of children hits of these roots, which then got random values on their creation. The root children hits is multiplied by the minimum percentage of hits provided by the user, and the result becomes the minimum hits expected from children branches. If the random value due to the uninitialization is big enough, then this minimum number of hits can be huge and eventually filter every children branches. The end result was invisible callchains. All we need to fix this is to initialize the children hits of the root. Reported-by: Christoph Hellwig Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: 2.6.32.x-2.6.35.y diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h index 624a96c..6de4313 100644 --- a/tools/perf/util/callchain.h +++ b/tools/perf/util/callchain.h @@ -50,6 +50,7 @@ static inline void callchain_init(struct callchain_node *node) INIT_LIST_HEAD(&node->children); INIT_LIST_HEAD(&node->val); + node->children_hit = 0; node->parent = NULL; node->hit = 0; } -- cgit v0.10.2 From 6628bc74f1aa9c35dd386320bf7ec04f12edb1b3 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Fri, 27 Aug 2010 09:15:09 +0200 Subject: writeback: do not lose wakeup events when forking bdi threads This patch fixes the following issue: INFO: task mount.nfs4:1120 blocked for more than 120 seconds. "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. mount.nfs4 D 00000000fffc6a21 0 1120 1119 0x00000000 ffff880235643948 0000000000000046 ffffffff00000000 ffffffff00000000 ffff880235643fd8 ffff880235314760 00000000001d44c0 ffff880235643fd8 00000000001d44c0 00000000001d44c0 00000000001d44c0 00000000001d44c0 Call Trace: [] schedule_timeout+0x34/0xf1 [] ? wait_for_common+0x3f/0x130 [] ? trace_hardirqs_on+0xd/0xf [] wait_for_common+0xd2/0x130 [] ? default_wake_function+0x0/0xf [] ? _raw_spin_unlock+0x26/0x2a [] wait_for_completion+0x18/0x1a [] sync_inodes_sb+0xca/0x1bc [] __sync_filesystem+0x47/0x7e [] sync_filesystem+0x47/0x4b [] generic_shutdown_super+0x22/0xd2 [] kill_anon_super+0x11/0x4f [] nfs4_kill_super+0x3f/0x72 [nfs] [] deactivate_locked_super+0x21/0x41 [] deactivate_super+0x40/0x45 [] mntput_no_expire+0xb8/0xed [] release_mounts+0x9a/0xb0 [] put_mnt_ns+0x6a/0x7b [] nfs_follow_remote_path+0x19a/0x296 [nfs] [] nfs4_try_mount+0x75/0xaf [nfs] [] nfs4_get_sb+0x276/0x2ff [nfs] [] vfs_kern_mount+0xb8/0x196 [] do_kern_mount+0x48/0xe8 [] do_mount+0x771/0x7e8 [] sys_mount+0x83/0xbd [] system_call_fastpath+0x16/0x1b The reason of this hang was a race condition: when the flusher thread is forking a bdi thread, we use 'kthread_run()', so we run it _before_ we make it visible in 'bdi->wb.task'. The bdi thread runs, does all works, and goes sleep. 'bdi->wb.task' is still NULL. And this is a dangerous time window. If at this time someone queues a work for this bdi, he does not see the bdi thread and wakes up the forker thread instead! But the forker has already forked this bdi thread, but just did not make it visible yet! The result is that we lose the wake up event for this bdi thread and the NFS4 code waits forever. To fix the problem, we should use 'ktrhead_create()' for creating bdi threads, then make them visible in 'bdi->wb.task', and only after this wake them up. This is exactly what this patch does. Signed-off-by: Artem Bityutskiy Signed-off-by: Jens Axboe diff --git a/mm/backing-dev.c b/mm/backing-dev.c index eaa4a5b..c2bf86f 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -445,8 +445,8 @@ static int bdi_forker_thread(void *ptr) switch (action) { case FORK_THREAD: __set_current_state(TASK_RUNNING); - task = kthread_run(bdi_writeback_thread, &bdi->wb, "flush-%s", - dev_name(bdi->dev)); + task = kthread_create(bdi_writeback_thread, &bdi->wb, + "flush-%s", dev_name(bdi->dev)); if (IS_ERR(task)) { /* * If thread creation fails, force writeout of @@ -457,10 +457,13 @@ static int bdi_forker_thread(void *ptr) /* * The spinlock makes sure we do not lose * wake-ups when racing with 'bdi_queue_work()'. + * And as soon as the bdi thread is visible, we + * can start it. */ spin_lock_bh(&bdi->wb_lock); bdi->wb.task = task; spin_unlock_bh(&bdi->wb_lock); + wake_up_process(task); } break; -- cgit v0.10.2 From b76b4014f9d988d2412b873e4d4c13c7f9afc4e4 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 28 Aug 2010 08:52:10 +0200 Subject: writeback: Fix lost wake-up shutting down writeback thread Setting the task state here may cause us to miss the wake up from kthread_stop(), so we need to recheck kthread_should_stop() or risk sleeping forever in the following schedule(). Symptom was an indefinite hang on an NFSv4 mount. (NFSv4 may create multiple mounts in a temporary namespace while traversing the mount path, and since the temporary namespace is immediately destroyed, it may end up destroying a mount very soon after it was created, possibly making this race more likely.) INFO: task mount.nfs4:4314 blocked for more than 120 seconds. "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. mount.nfs4 D 0000000000000000 2880 4314 4313 0x00000000 ffff88001ed6da28 0000000000000046 ffff88001ed6dfd8 ffff88001ed6dfd8 ffff88001ed6c000 ffff88001ed6c000 ffff88001ed6c000 ffff88001e5003a0 ffff88001ed6dfd8 ffff88001e5003a8 ffff88001ed6c000 ffff88001ed6dfd8 Call Trace: [] schedule_timeout+0x1cd/0x2e0 [] ? mark_held_locks+0x6c/0xa0 [] ? _raw_spin_unlock_irq+0x30/0x60 [] ? trace_hardirqs_on_caller+0x14d/0x190 [] ? sub_preempt_count+0xe/0xd0 [] wait_for_common+0x120/0x190 [] ? default_wake_function+0x0/0x20 [] wait_for_completion+0x1d/0x20 [] kthread_stop+0x4a/0x150 [] ? thaw_process+0x70/0x80 [] bdi_unregister+0x10a/0x1a0 [] nfs_put_super+0x19/0x20 [] generic_shutdown_super+0x54/0xe0 [] kill_anon_super+0x16/0x60 [] nfs4_kill_super+0x39/0x90 [] deactivate_locked_super+0x45/0x60 [] deactivate_super+0x49/0x70 [] mntput_no_expire+0x84/0xe0 [] release_mounts+0x9f/0xc0 [] put_mnt_ns+0x65/0x80 [] nfs_follow_remote_path+0x1e6/0x420 [] nfs4_try_mount+0x6f/0xd0 [] nfs4_get_sb+0xa2/0x360 [] vfs_kern_mount+0x88/0x1f0 [] do_kern_mount+0x52/0x130 [] ? _lock_kernel+0x6a/0x170 [] do_mount+0x26e/0x7f0 [] ? copy_mount_options+0xea/0x190 [] sys_mount+0x98/0xf0 [] system_call_fastpath+0x16/0x1b 1 lock held by mount.nfs4/4314: #0: (&type->s_umount_key#24){+.+...}, at: [] deactivate_super+0x41/0x70 Signed-off-by: J. Bruce Fields Signed-off-by: Jens Axboe Acked-by: Artem Bityutskiy diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 7d9d06b..81e086d 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -808,7 +808,7 @@ int bdi_writeback_thread(void *data) wb->last_active = jiffies; set_current_state(TASK_INTERRUPTIBLE); - if (!list_empty(&bdi->work_list)) { + if (!list_empty(&bdi->work_list) || kthread_should_stop()) { __set_current_state(TASK_RUNNING); continue; } -- cgit v0.10.2 From d9f66c1a46163c7c83411058516a69da547262f8 Mon Sep 17 00:00:00 2001 From: Mike Auty Date: Sat, 28 Aug 2010 20:35:17 -0700 Subject: Input: wacom - fix mousewheel handling for old wacom tablets This fixes a regression introduced in 3b57ca0f80c5c8994b5b1e3d3f904cfe727951f2. The data[6] byte contains either 1 or -1 depending on the whether the mouse wheel on older wacom tablets is moved down (1) or up (-1). The patch introduced in the above commit changed the cast from (signed char) to (signed). When cast as a signed integer and negated, the value of -1 (stored in the byte as 0xff) became -255 rather than 1. This patch reverts the cast to a (signed char) and also removes an unnecessary (signed) cast, as all the values operated on are bitmasked. Signed-off-by: Mike Auty Reviewed-by: Ping Cheng Cc; stable@kernel.org Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/tablet/wacom_wac.c b/drivers/input/tablet/wacom_wac.c index 40d77ba..6e29bad 100644 --- a/drivers/input/tablet/wacom_wac.c +++ b/drivers/input/tablet/wacom_wac.c @@ -243,10 +243,10 @@ static int wacom_graphire_irq(struct wacom_wac *wacom) if (features->type == WACOM_G4 || features->type == WACOM_MO) { input_report_abs(input, ABS_DISTANCE, data[6] & 0x3f); - rw = (signed)(data[7] & 0x04) - (data[7] & 0x03); + rw = (data[7] & 0x04) - (data[7] & 0x03); } else { input_report_abs(input, ABS_DISTANCE, data[7] & 0x3f); - rw = -(signed)data[6]; + rw = -(signed char)data[6]; } input_report_rel(input, REL_WHEEL, rw); } -- cgit v0.10.2 From 2c4e9671edfef534e9726366707d64e63d44e7e6 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sat, 28 Aug 2010 20:35:17 -0700 Subject: Input: use PIT_TICK_RATE in vt beep ioctl The KIOCSOUND and KDMKTONE ioctls are based on the CLOCK_TICK_RATE, which is architecture and sometimes configuration specific. In practice, most user applications assume that it is actually defined as the i8253 PIT base clock of 1193182 Hz, which is true on some architectures but not on others. This patch makes the vt code use the PIT frequency on all architectures, which is much more well-defined. It will change the behavior of user applications sending the beep ioctl on all architectures that define CLOCK_TICK_RATE different from PIT_TICK_RATE. The original breakage was introduced in commit bcc8ca099 "Adapt drivers/char/vt_ioctl.c to non-x86". Hopefully, reverting this change will make the frequency correct in more cases than it will make it incorrect. Signed-off-by: Arnd Bergmann Acked-by: Alan Cox Signed-off-by: Dmitry Torokhov diff --git a/drivers/char/vt_ioctl.c b/drivers/char/vt_ioctl.c index 2bbeaae..38df8c1 100644 --- a/drivers/char/vt_ioctl.c +++ b/drivers/char/vt_ioctl.c @@ -533,11 +533,14 @@ int vt_ioctl(struct tty_struct *tty, struct file * file, case KIOCSOUND: if (!perm) goto eperm; - /* FIXME: This is an old broken API but we need to keep it - supported and somehow separate the historic advertised - tick rate from any real one */ + /* + * The use of PIT_TICK_RATE is historic, it used to be + * the platform-dependent CLOCK_TICK_RATE between 2.6.12 + * and 2.6.36, which was a minor but unfortunate ABI + * change. + */ if (arg) - arg = CLOCK_TICK_RATE / arg; + arg = PIT_TICK_RATE / arg; kd_mksound(arg, 0); break; @@ -553,11 +556,8 @@ int vt_ioctl(struct tty_struct *tty, struct file * file, */ ticks = HZ * ((arg >> 16) & 0xffff) / 1000; count = ticks ? (arg & 0xffff) : 0; - /* FIXME: This is an old broken API but we need to keep it - supported and somehow separate the historic advertised - tick rate from any real one */ if (count) - count = CLOCK_TICK_RATE / count; + count = PIT_TICK_RATE / count; kd_mksound(count, ticks); break; } -- cgit v0.10.2 From ba4d695a90c9176fca8e45d6c872bbf4e8bed315 Mon Sep 17 00:00:00 2001 From: Henrik Rydberg Date: Sat, 28 Aug 2010 21:33:50 -0700 Subject: Input: MT - initialize slots to unused For MT slots, the ABS_MT_TRACKING_ID determines whether a slot is in use, but currently leaves initialization up to the drivers. This patch sets the slot state to unused upon creation. Signed-off-by: Henrik Rydberg Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/input.c b/drivers/input/input.c index a9b025f..ab69820 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -1599,11 +1599,14 @@ EXPORT_SYMBOL(input_free_device); * @dev: input device supporting MT events and finger tracking * @num_slots: number of slots used by the device * - * This function allocates all necessary memory for MT slot handling - * in the input device, and adds ABS_MT_SLOT to the device capabilities. + * This function allocates all necessary memory for MT slot handling in the + * input device, and adds ABS_MT_SLOT to the device capabilities. All slots + * are initially marked as unused iby setting ABS_MT_TRACKING_ID to -1. */ int input_mt_create_slots(struct input_dev *dev, unsigned int num_slots) { + int i; + if (!num_slots) return 0; @@ -1614,6 +1617,10 @@ int input_mt_create_slots(struct input_dev *dev, unsigned int num_slots) dev->mtsize = num_slots; input_set_abs_params(dev, ABS_MT_SLOT, 0, num_slots - 1, 0, 0); + /* Mark slots as 'unused' */ + for (i = 0; i < num_slots; i++) + dev->mt[i].abs[ABS_MT_TRACKING_ID - ABS_MT_FIRST] = -1; + return 0; } EXPORT_SYMBOL(input_mt_create_slots); -- cgit v0.10.2 From 50e77fcd790e02a095f906df0f6fa72afef7e11c Mon Sep 17 00:00:00 2001 From: Eric Miao Date: Wed, 18 Aug 2010 11:51:13 +0800 Subject: ARM: pxa: remove __init from cpufreq_driver->init() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This is causing section mismatches when linking, as cpufreq_driver->init() is not supposed to be in init section. Reported-by: Tomáš 'Sleep_Walker' Čech Signed-off-by: Eric Miao diff --git a/arch/arm/mach-pxa/cpufreq-pxa2xx.c b/arch/arm/mach-pxa/cpufreq-pxa2xx.c index 268a9bc..50d5939 100644 --- a/arch/arm/mach-pxa/cpufreq-pxa2xx.c +++ b/arch/arm/mach-pxa/cpufreq-pxa2xx.c @@ -398,7 +398,7 @@ static int pxa_set_target(struct cpufreq_policy *policy, return 0; } -static __init int pxa_cpufreq_init(struct cpufreq_policy *policy) +static int pxa_cpufreq_init(struct cpufreq_policy *policy) { int i; unsigned int freq; diff --git a/arch/arm/mach-pxa/cpufreq-pxa3xx.c b/arch/arm/mach-pxa/cpufreq-pxa3xx.c index 27fa329..0a0d0fe 100644 --- a/arch/arm/mach-pxa/cpufreq-pxa3xx.c +++ b/arch/arm/mach-pxa/cpufreq-pxa3xx.c @@ -204,7 +204,7 @@ static int pxa3xx_cpufreq_set(struct cpufreq_policy *policy, return 0; } -static __init int pxa3xx_cpufreq_init(struct cpufreq_policy *policy) +static int pxa3xx_cpufreq_init(struct cpufreq_policy *policy) { int ret = -EINVAL; -- cgit v0.10.2 From be7a6d9dd67251675b83a29f4b15392eae62e70b Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 13 Aug 2010 11:39:07 +0800 Subject: ARM: pxa: fix CI_HSYNC and CI_VSYNC MFP defines for pxa300 According to PXA3xx Processor Family Developer Manuall Vol1. section "Pin Descriptions and Control", PXA30x and PXA31x Processor Alternate Function Table shows the Alt FN 0 for GPIO51 is CI_HSYNC and for GPIO52 is CI_VSYNC. This patch fixes the MFP defines and also corrects the order of MFD defines. Signed-off-by: Axel Lin Signed-off-by: Eric Miao diff --git a/arch/arm/mach-pxa/include/mach/mfp-pxa300.h b/arch/arm/mach-pxa/include/mach/mfp-pxa300.h index 7139e0d..4e12870 100644 --- a/arch/arm/mach-pxa/include/mach/mfp-pxa300.h +++ b/arch/arm/mach-pxa/include/mach/mfp-pxa300.h @@ -71,10 +71,10 @@ #define GPIO46_CI_DD_7 MFP_CFG_DRV(GPIO46, AF0, DS04X) #define GPIO47_CI_DD_8 MFP_CFG_DRV(GPIO47, AF1, DS04X) #define GPIO48_CI_DD_9 MFP_CFG_DRV(GPIO48, AF1, DS04X) -#define GPIO52_CI_HSYNC MFP_CFG_DRV(GPIO52, AF0, DS04X) -#define GPIO51_CI_VSYNC MFP_CFG_DRV(GPIO51, AF0, DS04X) #define GPIO49_CI_MCLK MFP_CFG_DRV(GPIO49, AF0, DS04X) #define GPIO50_CI_PCLK MFP_CFG_DRV(GPIO50, AF0, DS04X) +#define GPIO51_CI_HSYNC MFP_CFG_DRV(GPIO51, AF0, DS04X) +#define GPIO52_CI_VSYNC MFP_CFG_DRV(GPIO52, AF0, DS04X) /* KEYPAD */ #define GPIO3_KP_DKIN_6 MFP_CFG_LPM(GPIO3, AF2, FLOAT) -- cgit v0.10.2 From b3282ab116cd7ebd2829ac8088524b5b279330e5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Thu, 12 Aug 2010 08:18:28 +0200 Subject: ARM: pxa: Make id const in pwm_probe() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This fixes the compiler warning: arch/arm/plat-pxa/pwm.c: In function 'pwm_probe': arch/arm/plat-pxa/pwm.c:179: warning: initialization discards qualifiers from pointer target type Signed-off-by: Uwe Kleine-König Signed-off-by: Eric Miao diff --git a/arch/arm/plat-pxa/pwm.c b/arch/arm/plat-pxa/pwm.c index 0732c6c..ef32686 100644 --- a/arch/arm/plat-pxa/pwm.c +++ b/arch/arm/plat-pxa/pwm.c @@ -176,7 +176,7 @@ static inline void __add_pwm(struct pwm_device *pwm) static int __devinit pwm_probe(struct platform_device *pdev) { - struct platform_device_id *id = platform_get_device_id(pdev); + const struct platform_device_id *id = platform_get_device_id(pdev); struct pwm_device *pwm, *secondary = NULL; struct resource *r; int ret = 0; -- cgit v0.10.2 From ffb287c9daabc3734a92def0d59796f065ad29eb Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Thu, 26 Aug 2010 05:51:57 +0200 Subject: ARM: pxa168fb: fix section mismatch Signed-off-by: Marek Vasut Acked-by: Haojian Zhuang Signed-off-by: Eric Miao diff --git a/drivers/video/pxa168fb.c b/drivers/video/pxa168fb.c index c91a7f7..5d786bd 100644 --- a/drivers/video/pxa168fb.c +++ b/drivers/video/pxa168fb.c @@ -559,7 +559,7 @@ static struct fb_ops pxa168fb_ops = { .fb_imageblit = cfb_imageblit, }; -static int __init pxa168fb_init_mode(struct fb_info *info, +static int __devinit pxa168fb_init_mode(struct fb_info *info, struct pxa168fb_mach_info *mi) { struct pxa168fb_info *fbi = info->par; @@ -599,7 +599,7 @@ static int __init pxa168fb_init_mode(struct fb_info *info, return ret; } -static int __init pxa168fb_probe(struct platform_device *pdev) +static int __devinit pxa168fb_probe(struct platform_device *pdev) { struct pxa168fb_mach_info *mi; struct fb_info *info = 0; @@ -792,7 +792,7 @@ static struct platform_driver pxa168fb_driver = { .probe = pxa168fb_probe, }; -static int __devinit pxa168fb_init(void) +static int __init pxa168fb_init(void) { return platform_driver_register(&pxa168fb_driver); } -- cgit v0.10.2 From fa66f07aa1f0950e1dc78b7ab39728b3f8aa77a1 Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Thu, 26 Aug 2010 16:40:01 +0200 Subject: perf_events: Fix time tracking for events with pid != -1 and cpu != -1 Per-thread events with a cpu filter, i.e., cpu != -1, were not reporting correct timings when the thread never ran on the monitored cpu. The time enabled was reported as a negative value. This patch fixes the problem by updating tstamp_stopped, tstamp_running in event_sched_out() for events with filters and which are marked as INACTIVE. The function group_sched_out() is modified to systematically call into event_sched_out() to avoid duplicating the timing adjustment code twice. With the patch, I now get: $ task_cpu -i -e unhalted_core_cycles,unhalted_core_cycles noploop 2 noploop for 2 seconds CPU0 0 unhalted_core_cycles (ena=1,991,136,594, run=0) CPU0 0 unhalted_core_cycles (ena=1,991,136,594, run=0) CPU1 0 unhalted_core_cycles (ena=1,991,136,594, run=0) CPU1 0 unhalted_core_cycles (ena=1,991,136,594, run=0) CPU2 0 unhalted_core_cycles (ena=1,991,136,594, run=0) CPU2 0 unhalted_core_cycles (ena=1,991,136,594, run=0) CPU3 4,747,990,931 unhalted_core_cycles (ena=1,991,136,594, run=1,991,136,594) CPU3 4,747,990,931 unhalted_core_cycles (ena=1,991,136,594, run=1,991,136,594) Signed-off-by: Stephane Eranian Acked-by: Peter Zijlstra Cc: paulus@samba.org Cc: davem@davemloft.net Cc: fweisbec@gmail.com Cc: perfmon2-devel@lists.sf.net Cc: eranian@google.com LKML-Reference: <4c76802d.aae9d80a.115d.70fe@mx.google.com> Signed-off-by: Ingo Molnar diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 403d180..657555a 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -402,11 +402,31 @@ static void perf_group_detach(struct perf_event *event) } } +static inline int +event_filter_match(struct perf_event *event) +{ + return event->cpu == -1 || event->cpu == smp_processor_id(); +} + static void event_sched_out(struct perf_event *event, struct perf_cpu_context *cpuctx, struct perf_event_context *ctx) { + u64 delta; + /* + * An event which could not be activated because of + * filter mismatch still needs to have its timings + * maintained, otherwise bogus information is return + * via read() for time_enabled, time_running: + */ + if (event->state == PERF_EVENT_STATE_INACTIVE + && !event_filter_match(event)) { + delta = ctx->time - event->tstamp_stopped; + event->tstamp_running += delta; + event->tstamp_stopped = ctx->time; + } + if (event->state != PERF_EVENT_STATE_ACTIVE) return; @@ -432,9 +452,7 @@ group_sched_out(struct perf_event *group_event, struct perf_event_context *ctx) { struct perf_event *event; - - if (group_event->state != PERF_EVENT_STATE_ACTIVE) - return; + int state = group_event->state; event_sched_out(group_event, cpuctx, ctx); @@ -444,7 +462,7 @@ group_sched_out(struct perf_event *group_event, list_for_each_entry(event, &group_event->sibling_list, group_entry) event_sched_out(event, cpuctx, ctx); - if (group_event->attr.exclusive) + if (state == PERF_EVENT_STATE_ACTIVE && group_event->attr.exclusive) cpuctx->exclusive = 0; } -- cgit v0.10.2 From c3dc66de59531c921c4638b1285075ea1c831186 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Mon, 30 Aug 2010 15:43:25 +0200 Subject: HID: add support for another BTC Emprex remote control Add device ID for another variant of this remote control. Reported-by: Gregor Fuis Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 0c52899..b2b3bb9 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -1287,6 +1287,7 @@ static const struct hid_device_id hid_blacklist[] = { { HID_USB_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_GEYSER1_TP_ONLY) }, { HID_USB_DEVICE(USB_VENDOR_ID_BELKIN, USB_DEVICE_ID_FLIP_KVM) }, { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE) }, + { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE_2) }, { HID_USB_DEVICE(USB_VENDOR_ID_CANDO, USB_DEVICE_ID_CANDO_MULTI_TOUCH) }, { HID_USB_DEVICE(USB_VENDOR_ID_CANDO, USB_DEVICE_ID_CANDO_MULTI_TOUCH_11_6) }, { HID_USB_DEVICE(USB_VENDOR_ID_CHERRY, USB_DEVICE_ID_CHERRY_CYMOTION) }, diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index b309525..7c69f1e 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -128,6 +128,7 @@ #define USB_VENDOR_ID_BTC 0x046e #define USB_DEVICE_ID_BTC_EMPREX_REMOTE 0x5578 +#define USB_DEVICE_ID_BTC_EMPREX_REMOTE_2 0x5577 #define USB_VENDOR_ID_CANDO 0x2087 #define USB_DEVICE_ID_CANDO_MULTI_TOUCH 0x0a01 diff --git a/drivers/hid/hid-topseed.c b/drivers/hid/hid-topseed.c index 5771f85..956ed9a 100644 --- a/drivers/hid/hid-topseed.c +++ b/drivers/hid/hid-topseed.c @@ -64,6 +64,7 @@ static int ts_input_mapping(struct hid_device *hdev, struct hid_input *hi, static const struct hid_device_id ts_devices[] = { { HID_USB_DEVICE(USB_VENDOR_ID_TOPSEED, USB_DEVICE_ID_TOPSEED_CYBERLINK) }, { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE) }, + { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE_2) }, { HID_USB_DEVICE(USB_VENDOR_ID_TOPSEED2, USB_DEVICE_ID_TOPSEED2_RF_COMBO) }, { } }; -- cgit v0.10.2 From ebd11fecd3096b080c84fb35014916ae2b5ba64a Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Mon, 30 Aug 2010 13:11:42 +0200 Subject: HID: Add quirk for eGalax touch controler. This patch adds a quirk for the eGalax touch controller which reports two pairs of axes. Signed-off-by: Thierry Reding Signed-off-by: Jiri Kosina diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c index 89fa4ac..16808f1 100644 --- a/drivers/hid/usbhid/hid-quirks.c +++ b/drivers/hid/usbhid/hid-quirks.c @@ -33,6 +33,7 @@ static const struct hid_blacklist { { USB_VENDOR_ID_AASHIMA, USB_DEVICE_ID_AASHIMA_PREDATOR, HID_QUIRK_BADPAD }, { USB_VENDOR_ID_ALPS, USB_DEVICE_ID_IBM_GAMEPAD, HID_QUIRK_BADPAD }, { USB_VENDOR_ID_CHIC, USB_DEVICE_ID_CHIC_GAMEPAD, HID_QUIRK_BADPAD }, + { USB_VENDOR_ID_DWAV, USB_DEVICE_ID_EGALAX_TOUCHCONTROLLER, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_DWAV, USB_DEVICE_ID_DWAV_EGALAX_MULTITOUCH, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_MOJO, USB_DEVICE_ID_RETRO_ADAPTER, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_HAPP, USB_DEVICE_ID_UGCI_DRIVING, HID_QUIRK_BADPAD | HID_QUIRK_MULTI_INPUT }, -- cgit v0.10.2 From edce6820a9fdda85521211cb334a183e34cc455e Mon Sep 17 00:00:00 2001 From: Jeffrey Carlyle Date: Mon, 30 Aug 2010 19:55:09 +0200 Subject: scatterlist: prevent invalid free when alloc fails When alloc fails, free_table is being called. Depending on the number of bytes requested, we determine if we are going to call _get_free_page() or kmalloc(). When alloc fails, our math is wrong (due to sg_size - 1), and the last buffer is wrongfully assumed to have been allocated by kmalloc. Hence, kfree gets called and a panic occurs. Signed-off-by: Jeffrey Carlyle Signed-off-by: Olusanya Soyannwo Acked-by: Tejun Heo Signed-off-by: Jens Axboe diff --git a/lib/scatterlist.c b/lib/scatterlist.c index a5ec428..4ceb05d 100644 --- a/lib/scatterlist.c +++ b/lib/scatterlist.c @@ -248,8 +248,18 @@ int __sg_alloc_table(struct sg_table *table, unsigned int nents, left -= sg_size; sg = alloc_fn(alloc_size, gfp_mask); - if (unlikely(!sg)) - return -ENOMEM; + if (unlikely(!sg)) { + /* + * Adjust entry count to reflect that the last + * entry of the previous table won't be used for + * linkage. Without this, sg_kfree() may get + * confused. + */ + if (prv) + table->nents = ++table->orig_nents; + + return -ENOMEM; + } sg_init_table(sg, alloc_size); table->nents = table->orig_nents += sg_size; -- cgit v0.10.2 From b137b9942a07843c64a934cfdb7d43155e507e13 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Mon, 23 Aug 2010 20:25:32 +0000 Subject: ACPI: Don't report current_now if battery reports in mWh ACPI batteries can report in units of either current or energy. Right now we expose the current_now file even if the battery is reporting energy units, resulting in a file that should contain mA instead containing mW. Don't expose this value unless the battery is reporting current. Signed-off-by: Matthew Garrett Cc: Arjan van de Ven Acked-by: Rafael J. Wysocki Acked-by: Zhang Rui Signed-off-by: Len Brown diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c index dc58402..9841720 100644 --- a/drivers/acpi/battery.c +++ b/drivers/acpi/battery.c @@ -273,7 +273,6 @@ static enum power_supply_property energy_battery_props[] = { POWER_SUPPLY_PROP_CYCLE_COUNT, POWER_SUPPLY_PROP_VOLTAGE_MIN_DESIGN, POWER_SUPPLY_PROP_VOLTAGE_NOW, - POWER_SUPPLY_PROP_CURRENT_NOW, POWER_SUPPLY_PROP_POWER_NOW, POWER_SUPPLY_PROP_ENERGY_FULL_DESIGN, POWER_SUPPLY_PROP_ENERGY_FULL, -- cgit v0.10.2 From 10f0412f57f2a76a90eff4376f59cbb0a39e4e18 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Mon, 30 Aug 2010 10:56:18 +0200 Subject: oprofile, x86: fix init_sysfs error handling On failure init_sysfs() might not properly free resources. The error code of the function is not checked. And, when reinitializing the exit function might be called twice. This patch fixes all this. Cc: stable@kernel.org Signed-off-by: Robert Richter diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index f6b48f6..73a41d3 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -568,8 +568,13 @@ static int __init init_sysfs(void) int error; error = sysdev_class_register(&oprofile_sysclass); - if (!error) - error = sysdev_register(&device_oprofile); + if (error) + return error; + + error = sysdev_register(&device_oprofile); + if (error) + sysdev_class_unregister(&oprofile_sysclass); + return error; } @@ -695,6 +700,8 @@ int __init op_nmi_init(struct oprofile_operations *ops) char *cpu_type = NULL; int ret = 0; + using_nmi = 0; + if (!cpu_has_apic) return -ENODEV; @@ -774,7 +781,10 @@ int __init op_nmi_init(struct oprofile_operations *ops) mux_init(ops); - init_sysfs(); + ret = init_sysfs(); + if (ret) + return ret; + using_nmi = 1; printk(KERN_INFO "oprofile: using NMI interrupt.\n"); return 0; -- cgit v0.10.2 From 3aaba20f26f58843e8f20611e5c0b1c06954310f Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 23 Aug 2010 16:50:12 +0800 Subject: tracing: Fix a race in function profile While we are reading trace_stat/functionX and someone just disabled function_profile at that time, we can trigger this: divide error: 0000 [#1] PREEMPT SMP ... EIP is at function_stat_show+0x90/0x230 ... This fix just takes the ftrace_profile_lock and checks if rec->counter is 0. If it's 0, we know the profile buffer has been reset. Signed-off-by: Li Zefan Cc: stable@kernel.org LKML-Reference: <4C723644.4040708@cn.fujitsu.com> Signed-off-by: Steven Rostedt diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 0d88ce9..7cb1f45 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -381,12 +381,19 @@ static int function_stat_show(struct seq_file *m, void *v) { struct ftrace_profile *rec = v; char str[KSYM_SYMBOL_LEN]; + int ret = 0; #ifdef CONFIG_FUNCTION_GRAPH_TRACER - static DEFINE_MUTEX(mutex); static struct trace_seq s; unsigned long long avg; unsigned long long stddev; #endif + mutex_lock(&ftrace_profile_lock); + + /* we raced with function_profile_reset() */ + if (unlikely(rec->counter == 0)) { + ret = -EBUSY; + goto out; + } kallsyms_lookup(rec->ip, NULL, NULL, NULL, str); seq_printf(m, " %-30.30s %10lu", str, rec->counter); @@ -408,7 +415,6 @@ static int function_stat_show(struct seq_file *m, void *v) do_div(stddev, (rec->counter - 1) * 1000); } - mutex_lock(&mutex); trace_seq_init(&s); trace_print_graph_duration(rec->time, &s); trace_seq_puts(&s, " "); @@ -416,11 +422,12 @@ static int function_stat_show(struct seq_file *m, void *v) trace_seq_puts(&s, " "); trace_print_graph_duration(stddev, &s); trace_print_seq(m, &s); - mutex_unlock(&mutex); #endif seq_putc(m, '\n'); +out: + mutex_unlock(&ftrace_profile_lock); - return 0; + return ret; } static void ftrace_profile_reset(struct ftrace_profile_stat *stat) -- cgit v0.10.2 From 950eaaca681c44aab87a46225c9e44f902c080aa Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 31 Aug 2010 17:00:18 -0700 Subject: pid: make setpgid() system call use RCU read-side critical section [ 23.584719] [ 23.584720] =================================================== [ 23.585059] [ INFO: suspicious rcu_dereference_check() usage. ] [ 23.585176] --------------------------------------------------- [ 23.585176] kernel/pid.c:419 invoked rcu_dereference_check() without protection! [ 23.585176] [ 23.585176] other info that might help us debug this: [ 23.585176] [ 23.585176] [ 23.585176] rcu_scheduler_active = 1, debug_locks = 1 [ 23.585176] 1 lock held by rc.sysinit/728: [ 23.585176] #0: (tasklist_lock){.+.+..}, at: [] sys_setpgid+0x5f/0x193 [ 23.585176] [ 23.585176] stack backtrace: [ 23.585176] Pid: 728, comm: rc.sysinit Not tainted 2.6.36-rc2 #2 [ 23.585176] Call Trace: [ 23.585176] [] lockdep_rcu_dereference+0x99/0xa2 [ 23.585176] [] find_task_by_pid_ns+0x50/0x6a [ 23.585176] [] find_task_by_vpid+0x1d/0x1f [ 23.585176] [] sys_setpgid+0x67/0x193 [ 23.585176] [] system_call_fastpath+0x16/0x1b [ 24.959669] type=1400 audit(1282938522.956:4): avc: denied { module_request } for pid=766 comm="hwclock" kmod="char-major-10-135" scontext=system_u:system_r:hwclock_t:s0 tcontext=system_u:system_r:kernel_t:s0 tclas It turns out that the setpgid() system call fails to enter an RCU read-side critical section before doing a PID-to-task_struct translation. This commit therefore does rcu_read_lock() before the translation, and also does rcu_read_unlock() after the last use of the returned pointer. Reported-by: Andrew Morton Signed-off-by: Paul E. McKenney Acked-by: David Howells diff --git a/kernel/sys.c b/kernel/sys.c index e9ad444..7f5a0cd 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -931,6 +931,7 @@ SYSCALL_DEFINE2(setpgid, pid_t, pid, pid_t, pgid) pgid = pid; if (pgid < 0) return -EINVAL; + rcu_read_lock(); /* From this point forward we keep holding onto the tasklist lock * so that our parent does not change from under us. -DaveM @@ -984,6 +985,7 @@ SYSCALL_DEFINE2(setpgid, pid_t, pid, pid_t, pgid) out: /* All paths lead to here, thus we are safe. -DaveM */ write_unlock_irq(&tasklist_lock); + rcu_read_unlock(); return err; } -- cgit v0.10.2 From 57157becdd1d23e6c2b8661ffe6c78d7d605d121 Mon Sep 17 00:00:00 2001 From: Henrik Rydberg Date: Tue, 31 Aug 2010 17:27:02 -0700 Subject: Input: bcm5974 - adjust major/minor to scale By visual inspection, the reported touch_major and touch_minor axes are a factor of two too small. Presumably the device actually reports the semi-major and semi-minor axes. Corrected with this patch. Signed-off-by: Henrik Rydberg Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/mouse/bcm5974.c b/drivers/input/mouse/bcm5974.c index ea67c49..b952317 100644 --- a/drivers/input/mouse/bcm5974.c +++ b/drivers/input/mouse/bcm5974.c @@ -337,10 +337,14 @@ static void report_finger_data(struct input_dev *input, const struct bcm5974_config *cfg, const struct tp_finger *f) { - input_report_abs(input, ABS_MT_TOUCH_MAJOR, raw2int(f->force_major)); - input_report_abs(input, ABS_MT_TOUCH_MINOR, raw2int(f->force_minor)); - input_report_abs(input, ABS_MT_WIDTH_MAJOR, raw2int(f->size_major)); - input_report_abs(input, ABS_MT_WIDTH_MINOR, raw2int(f->size_minor)); + input_report_abs(input, ABS_MT_TOUCH_MAJOR, + raw2int(f->force_major) << 1); + input_report_abs(input, ABS_MT_TOUCH_MINOR, + raw2int(f->force_minor) << 1); + input_report_abs(input, ABS_MT_WIDTH_MAJOR, + raw2int(f->size_major) << 1); + input_report_abs(input, ABS_MT_WIDTH_MINOR, + raw2int(f->size_minor) << 1); input_report_abs(input, ABS_MT_ORIENTATION, MAX_FINGER_ORIENTATION - raw2int(f->orientation)); input_report_abs(input, ABS_MT_POSITION_X, raw2int(f->abs_x)); -- cgit v0.10.2 From af045b86662f17bf130239a65995c61a34f00a6b Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Tue, 31 Aug 2010 17:27:02 -0700 Subject: Input: i8042 - fix device removal on unload We need to call platform_device_unregister(i8042_platform_device) before calling platform_driver_unregister() because i8042_remove() resets i8042_platform_device to NULL. This leaves the platform device instance behind and prevents driver reload. Fixes https://bugzilla.kernel.org/show_bug.cgi?id=16613 Reported-by: Seryodkin Victor Cc: stable@kernel.org Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/serio/i8042.c b/drivers/input/serio/i8042.c index 46e4ba0..f585131 100644 --- a/drivers/input/serio/i8042.c +++ b/drivers/input/serio/i8042.c @@ -1485,8 +1485,8 @@ static int __init i8042_init(void) static void __exit i8042_exit(void) { - platform_driver_unregister(&i8042_driver); platform_device_unregister(i8042_platform_device); + platform_driver_unregister(&i8042_driver); i8042_platform_exit(); panic_blink = NULL; -- cgit v0.10.2 From 68d3f1d810500e8b975bdf0b20dd83d060076b4b Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 31 Aug 2010 23:00:07 -0400 Subject: lockup_detector: Sync touch_*_watchdog back to old semantics During my rewrite, the semantics of touch_nmi_watchdog and touch_softlockup_watchdog changed enough to break some drivers (mostly over preemptable regions). These are cases where long delays on one CPU (due to print_delay for example) can cause long delays on other CPUs - so we must 'touch' the nmi_watchdog flag of those other CPUs as well. This change brings those touch_*_watchdog() functions back in line with to how they used to work. Signed-off-by: Don Zickus Acked-by: Cyrill Gorcunov Cc: peterz@infradead.org Cc: fweisbec@gmail.com LKML-Reference: <1283310009-22168-2-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar diff --git a/kernel/watchdog.c b/kernel/watchdog.c index 0d53c8e..7f9c3c5 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -122,7 +122,7 @@ static void __touch_watchdog(void) void touch_softlockup_watchdog(void) { - __get_cpu_var(watchdog_touch_ts) = 0; + __raw_get_cpu_var(watchdog_touch_ts) = 0; } EXPORT_SYMBOL(touch_softlockup_watchdog); @@ -142,7 +142,14 @@ void touch_all_softlockup_watchdogs(void) #ifdef CONFIG_HARDLOCKUP_DETECTOR void touch_nmi_watchdog(void) { - __get_cpu_var(watchdog_nmi_touch) = true; + if (watchdog_enabled) { + unsigned cpu; + + for_each_present_cpu(cpu) { + if (per_cpu(watchdog_nmi_touch, cpu) != true) + per_cpu(watchdog_nmi_touch, cpu) = true; + } + } touch_softlockup_watchdog(); } EXPORT_SYMBOL(touch_nmi_watchdog); @@ -433,6 +440,9 @@ static int watchdog_enable(int cpu) wake_up_process(p); } + /* if any cpu succeeds, watchdog is considered enabled for the system */ + watchdog_enabled = 1; + return 0; } @@ -455,9 +465,6 @@ static void watchdog_disable(int cpu) per_cpu(softlockup_watchdog, cpu) = NULL; kthread_stop(p); } - - /* if any cpu succeeds, watchdog is considered enabled for the system */ - watchdog_enabled = 1; } static void watchdog_enable_all_cpus(void) -- cgit v0.10.2 From 25d3584797a39f57b69cd835722ac7c41113fb9a Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Mon, 16 Aug 2010 15:15:14 +0100 Subject: ARM: 6330/1: perf: reword comments relating to perf_event_do_pending This is purely a cosmetic change to the ARM perf backend because the current comments about the relationship between NMIs, interrupt context and perf_event_do_pending are misleading. This patch updates the comments so that they reflect what the code actually does (which is in line with other architectures). Acked-by: Jamie Iles Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/include/asm/perf_event.h b/arch/arm/include/asm/perf_event.h index 48837e6..b5799a3 100644 --- a/arch/arm/include/asm/perf_event.h +++ b/arch/arm/include/asm/perf_event.h @@ -17,7 +17,7 @@ * counter interrupts are regular interrupts and not an NMI. This * means that when we receive the interrupt we can call * perf_event_do_pending() that handles all of the work with - * interrupts enabled. + * interrupts disabled. */ static inline void set_perf_event_pending(void) diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index 417c392..e213925 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -1041,8 +1041,8 @@ armv6pmu_handle_irq(int irq_num, /* * Handle the pending perf events. * - * Note: this call *must* be run with interrupts enabled. For - * platforms that can have the PMU interrupts raised as a PMI, this + * Note: this call *must* be run with interrupts disabled. For + * platforms that can have the PMU interrupts raised as an NMI, this * will not work. */ perf_event_do_pending(); @@ -2017,8 +2017,8 @@ static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev) /* * Handle the pending perf events. * - * Note: this call *must* be run with interrupts enabled. For - * platforms that can have the PMU interrupts raised as a PMI, this + * Note: this call *must* be run with interrupts disabled. For + * platforms that can have the PMU interrupts raised as an NMI, this * will not work. */ perf_event_do_pending(); -- cgit v0.10.2 From 418cf646c9999f2f7d216b1a693c2c575ab8094c Mon Sep 17 00:00:00 2001 From: Mikael Pettersson Date: Wed, 25 Aug 2010 20:49:01 +0100 Subject: ARM: 6343/1: wire up fanotify and prlimit64 syscalls on ARM The 2.6.36-rc kernel added three new system calls: fanotify_init, fanotify_mark, and prlimit64. This patch wires them up on ARM. The only non-trivial issue here is the u64 argument to sys_fanotify_mark(), but it is the 3rd argument and thus passed in r2/r3 in both kernel and user space, so it causes no problems. Tested with a 2.6.36-rc2 EABI kernel on an ixp4xx machine. Tested-by: Anand Gadiyar Signed-off-by: Mikael Pettersson Signed-off-by: Russell King diff --git a/arch/arm/include/asm/unistd.h b/arch/arm/include/asm/unistd.h index d02cfb6..c891eb7 100644 --- a/arch/arm/include/asm/unistd.h +++ b/arch/arm/include/asm/unistd.h @@ -393,6 +393,9 @@ #define __NR_perf_event_open (__NR_SYSCALL_BASE+364) #define __NR_recvmmsg (__NR_SYSCALL_BASE+365) #define __NR_accept4 (__NR_SYSCALL_BASE+366) +#define __NR_fanotify_init (__NR_SYSCALL_BASE+367) +#define __NR_fanotify_mark (__NR_SYSCALL_BASE+368) +#define __NR_prlimit64 (__NR_SYSCALL_BASE+369) /* * The following SWIs are ARM private. diff --git a/arch/arm/kernel/calls.S b/arch/arm/kernel/calls.S index afeb71f..5c26ecc 100644 --- a/arch/arm/kernel/calls.S +++ b/arch/arm/kernel/calls.S @@ -376,6 +376,9 @@ CALL(sys_perf_event_open) /* 365 */ CALL(sys_recvmmsg) CALL(sys_accept4) + CALL(sys_fanotify_init) + CALL(sys_fanotify_mark) + CALL(sys_prlimit64) #ifndef syscalls_counted .equ syscalls_padding, ((NR_syscalls + 3) & ~3) - NR_syscalls #define syscalls_counted -- cgit v0.10.2 From af54decd6a2b8efa335020afc77254355c4c1bab Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 14 Aug 2010 11:03:16 +0200 Subject: regulator/ab8500: move dereference below the check for NULL I moved the dereference of "ab8500" below the check for NULL. Signed-off-by: Dan Carpenter Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c index dc3f1a4..cc7cbaf 100644 --- a/drivers/regulator/ab8500.c +++ b/drivers/regulator/ab8500.c @@ -344,13 +344,14 @@ static inline struct ab8500_regulator_info *find_regulator_info(int id) static __devinit int ab8500_regulator_probe(struct platform_device *pdev) { struct ab8500 *ab8500 = dev_get_drvdata(pdev->dev.parent); - struct ab8500_platform_data *pdata = dev_get_platdata(ab8500->dev); + struct ab8500_platform_data *pdata; int i, err; if (!ab8500) { dev_err(&pdev->dev, "null mfd parent\n"); return -EINVAL; } + pdata = dev_get_platdata(ab8500->dev); /* register all regulators */ for (i = 0; i < ARRAY_SIZE(ab8500_regulator_info); i++) { -- cgit v0.10.2 From b3fcf3e576749b911e984e752b6b390c326efb76 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 14 Aug 2010 21:31:01 +0800 Subject: regulator: ab3100 - fix the logic to remove already registered regulators in error path In current implementation, ab3100_regulators[0].rdev is not unregistered if the error happen at i > 0. This patch fixes the resource leak and also improves the readability. Signed-off-by: Axel Lin Acked-by: Linus Walleij Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/ab3100.c b/drivers/regulator/ab3100.c index 1179099..b349266 100644 --- a/drivers/regulator/ab3100.c +++ b/drivers/regulator/ab3100.c @@ -634,12 +634,9 @@ static int __devinit ab3100_regulators_probe(struct platform_device *pdev) "%s: failed to register regulator %s err %d\n", __func__, ab3100_regulator_desc[i].name, err); - i--; /* remove the already registered regulators */ - while (i > 0) { + while (--i >= 0) regulator_unregister(ab3100_regulators[i].rdev); - i--; - } return err; } -- cgit v0.10.2 From d4876a3bc041e8e40af20b8addbec6d0a42e3842 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 14 Aug 2010 21:44:04 +0800 Subject: regulator: ab8500 - fix the logic to remove already registered regulators in error path In current implementation, ab8500_regulator_info[0].regulator is not unregistered if the error happen at i > 0. This patch fixes the resource leak and also improves the readability. Signed-off-by: Axel Lin Acked-by: Linus Walleij Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c index cc7cbaf..3d09580 100644 --- a/drivers/regulator/ab8500.c +++ b/drivers/regulator/ab8500.c @@ -369,11 +369,9 @@ static __devinit int ab8500_regulator_probe(struct platform_device *pdev) dev_err(&pdev->dev, "failed to register regulator %s\n", info->desc.name); /* when we fail, un-register all earlier regulators */ - i--; - while (i > 0) { + while (--i >= 0) { info = &ab8500_regulator_info[i]; regulator_unregister(info->regulator); - i--; } return err; } -- cgit v0.10.2 From 3e352f9e02a37c11df695aabfe49faebf507971b Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 18 Aug 2010 11:37:21 +0800 Subject: regulator: max1586 - improve the logic of choosing selector A little bit improvement in the logic of choosing selector. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/max1586.c b/drivers/regulator/max1586.c index 8867c27..559cfa2 100644 --- a/drivers/regulator/max1586.c +++ b/drivers/regulator/max1586.c @@ -121,14 +121,14 @@ static int max1586_v6_set(struct regulator_dev *rdev, int min_uV, int max_uV) if (max_uV < MAX1586_V6_MIN_UV || max_uV > MAX1586_V6_MAX_UV) return -EINVAL; - if (min_uV >= 3000000) - selector = 3; - if (min_uV < 3000000) - selector = 2; - if (min_uV < 2500000) - selector = 1; if (min_uV < 1800000) selector = 0; + else if (min_uV < 2500000) + selector = 1; + else if (min_uV < 3000000) + selector = 2; + else if (min_uV >= 3000000) + selector = 3; if (max1586_v6_calc_voltage(selector) > max_uV) return -EINVAL; -- cgit v0.10.2 From 7112b2dfea4966c58d21b7197c3f099041248e59 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 14 Aug 2010 22:32:13 +0800 Subject: regulator: tps6507x - remove incorrect comments This driver is a platform driver, not a i2c driver. Thus remove incorrect tps6507x_remove comments. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/tps6507x-regulator.c b/drivers/regulator/tps6507x-regulator.c index c239f42..020f587 100644 --- a/drivers/regulator/tps6507x-regulator.c +++ b/drivers/regulator/tps6507x-regulator.c @@ -626,12 +626,6 @@ fail: return error; } -/** - * tps6507x_remove - TPS6507x driver i2c remove handler - * @client: i2c driver client device structure - * - * Unregister TPS driver as an i2c client device driver - */ static int __devexit tps6507x_pmic_remove(struct platform_device *pdev) { struct tps6507x_dev *tps6507x_dev = platform_get_drvdata(pdev); -- cgit v0.10.2 From 11fa0d1d20c7cc432c77369bc8bbfbc21030e457 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 19 Aug 2010 10:29:19 +0800 Subject: regulator: max8998 - fix memory allocation size for max8998->rdev We only use max8998->rdev[0] .. max8998->rdev[pdata->num_regulators-1], max8998->rdev[pdata->num_regulators] is not used. Thus fix the memory allocation size. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/max8998.c b/drivers/regulator/max8998.c index ab67298..fbcb385 100644 --- a/drivers/regulator/max8998.c +++ b/drivers/regulator/max8998.c @@ -549,7 +549,7 @@ static __devinit int max8998_pmic_probe(struct platform_device *pdev) if (!max8998) return -ENOMEM; - size = sizeof(struct regulator_dev *) * (pdata->num_regulators + 1); + size = sizeof(struct regulator_dev *) * pdata->num_regulators; max8998->rdev = kzalloc(size, GFP_KERNEL); if (!max8998->rdev) { kfree(max8998); @@ -583,7 +583,7 @@ static __devinit int max8998_pmic_probe(struct platform_device *pdev) return 0; err: - for (i = 0; i <= max8998->num_regulators; i++) + for (i = 0; i < max8998->num_regulators; i++) if (rdev[i]) regulator_unregister(rdev[i]); @@ -599,7 +599,7 @@ static int __devexit max8998_pmic_remove(struct platform_device *pdev) struct regulator_dev **rdev = max8998->rdev; int i; - for (i = 0; i <= max8998->num_regulators; i++) + for (i = 0; i < max8998->num_regulators; i++) if (rdev[i]) regulator_unregister(rdev[i]); -- cgit v0.10.2 From c356cbc2d4d99cf5a1429603fa1841e50987c4d3 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 22 Aug 2010 15:26:49 +0800 Subject: regulator: max8998 - set max8998->num_regulators Set max8998->num_regulators = pdata->num_regulators, otherwise it's default value is 0. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/max8998.c b/drivers/regulator/max8998.c index fbcb385..8b9bfdf 100644 --- a/drivers/regulator/max8998.c +++ b/drivers/regulator/max8998.c @@ -558,6 +558,7 @@ static __devinit int max8998_pmic_probe(struct platform_device *pdev) rdev = max8998->rdev; max8998->iodev = iodev; + max8998->num_regulators = pdata->num_regulators; platform_set_drvdata(pdev, max8998); for (i = 0; i < pdata->num_regulators; i++) { -- cgit v0.10.2 From 327531bada36e5786b13bb6918ad8afc545adfa2 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 22 Aug 2010 22:38:15 +0800 Subject: regulator: tps6586x-regulator - fix value range checking for val val is used as array index of ri->voltages. Thus the valid value range should be 0 .. ri->desc.n_voltages - 1. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/tps6586x-regulator.c b/drivers/regulator/tps6586x-regulator.c index 8cff141..facd439 100644 --- a/drivers/regulator/tps6586x-regulator.c +++ b/drivers/regulator/tps6586x-regulator.c @@ -133,7 +133,7 @@ static int tps6586x_ldo_get_voltage(struct regulator_dev *rdev) mask = ((1 << ri->volt_nbits) - 1) << ri->volt_shift; val = (val & mask) >> ri->volt_shift; - if (val > ri->desc.n_voltages) + if (val >= ri->desc.n_voltages) BUG(); return ri->voltages[val] * 1000; -- cgit v0.10.2 From 938b45927c240cf75a01ce29af3f173762e762f8 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 22 Aug 2010 22:42:42 +0800 Subject: regulator: tps6586x-regulator - fix bit_mask parameter for tps6586x_set_bits() The third parameter of tps6586x_set_bits() is the bit_mask, thus we should use (1 << ri->go_bit) instead of ri->go_bit. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/tps6586x-regulator.c b/drivers/regulator/tps6586x-regulator.c index facd439..51237fb 100644 --- a/drivers/regulator/tps6586x-regulator.c +++ b/drivers/regulator/tps6586x-regulator.c @@ -150,7 +150,7 @@ static int tps6586x_dvm_set_voltage(struct regulator_dev *rdev, if (ret) return ret; - return tps6586x_set_bits(parent, ri->go_reg, ri->go_bit); + return tps6586x_set_bits(parent, ri->go_reg, 1 << ri->go_bit); } static int tps6586x_regulator_enable(struct regulator_dev *rdev) -- cgit v0.10.2 From 747cc851dc42ffeac2872da066ca4293a6d90baf Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 27 Aug 2010 16:37:34 +0800 Subject: regulator: set max8998->dev to &pdev->dev. max8998->dev is NULL in current implementation, set it to &pdev->dev. regulator_register() still return success if max8998->dev is NULL, but rdev->dev.parent will be set to NULL which is incorrect. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/max8998.c b/drivers/regulator/max8998.c index 8b9bfdf..a1baf1f 100644 --- a/drivers/regulator/max8998.c +++ b/drivers/regulator/max8998.c @@ -557,6 +557,7 @@ static __devinit int max8998_pmic_probe(struct platform_device *pdev) } rdev = max8998->rdev; + max8998->dev = &pdev->dev; max8998->iodev = iodev; max8998->num_regulators = pdata->num_regulators; platform_set_drvdata(pdev, max8998); -- cgit v0.10.2 From 606b2f490fb80e55d05cf0e6cec0b6c0ff0fc18f Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Fri, 27 Aug 2010 22:33:38 +0100 Subject: regulator: Update e-mail address for Liam Girdwood Signed-off-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/Documentation/power/regulator/overview.txt b/Documentation/power/regulator/overview.txt index 9363e05..8ed1758 100644 --- a/Documentation/power/regulator/overview.txt +++ b/Documentation/power/regulator/overview.txt @@ -13,7 +13,7 @@ regulators (where voltage output is controllable) and current sinks (where current limit is controllable). (C) 2008 Wolfson Microelectronics PLC. -Author: Liam Girdwood +Author: Liam Girdwood Nomenclature -- cgit v0.10.2 From 58d463eec844f6381d63d04dc89d319ae3057ca9 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 1 Sep 2010 10:29:18 +0800 Subject: regulator: ad5398 - fix a memory leak In current implementation, the address return from regulator_register() is different from the address for regulator_unregister(). Signed-off-by: Axel Lin Acked-by: Sonic Zhang Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/ad5398.c b/drivers/regulator/ad5398.c index d59d2f2..df1fb53 100644 --- a/drivers/regulator/ad5398.c +++ b/drivers/regulator/ad5398.c @@ -25,7 +25,7 @@ struct ad5398_chip_info { unsigned int current_level; unsigned int current_mask; unsigned int current_offset; - struct regulator_dev rdev; + struct regulator_dev *rdev; }; static int ad5398_calc_current(struct ad5398_chip_info *chip, @@ -211,7 +211,6 @@ MODULE_DEVICE_TABLE(i2c, ad5398_id); static int __devinit ad5398_probe(struct i2c_client *client, const struct i2c_device_id *id) { - struct regulator_dev *rdev; struct regulator_init_data *init_data = client->dev.platform_data; struct ad5398_chip_info *chip; const struct ad5398_current_data_format *df = @@ -233,9 +232,10 @@ static int __devinit ad5398_probe(struct i2c_client *client, chip->current_offset = df->current_offset; chip->current_mask = (chip->current_level - 1) << chip->current_offset; - rdev = regulator_register(&ad5398_reg, &client->dev, init_data, chip); - if (IS_ERR(rdev)) { - ret = PTR_ERR(rdev); + chip->rdev = regulator_register(&ad5398_reg, &client->dev, + init_data, chip); + if (IS_ERR(chip->rdev)) { + ret = PTR_ERR(chip->rdev); dev_err(&client->dev, "failed to register %s %s\n", id->name, ad5398_reg.name); goto err; @@ -254,7 +254,7 @@ static int __devexit ad5398_remove(struct i2c_client *client) { struct ad5398_chip_info *chip = i2c_get_clientdata(client); - regulator_unregister(&chip->rdev); + regulator_unregister(chip->rdev); kfree(chip); i2c_set_clientdata(client, NULL); -- cgit v0.10.2 From b9e5d11a7e70000ace3ba92100bf1e81ff607604 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 1 Sep 2010 13:09:41 +0800 Subject: regulator: isl6271a-regulator - fix regulator_desc parameter for regulator_register() Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/isl6271a-regulator.c b/drivers/regulator/isl6271a-regulator.c index e49d2bd..d61ecb8 100644 --- a/drivers/regulator/isl6271a-regulator.c +++ b/drivers/regulator/isl6271a-regulator.c @@ -165,7 +165,7 @@ static int __devinit isl6271a_probe(struct i2c_client *i2c, mutex_init(&pmic->mtx); for (i = 0; i < 3; i++) { - pmic->rdev[i] = regulator_register(&isl_rd[0], &i2c->dev, + pmic->rdev[i] = regulator_register(&isl_rd[i], &i2c->dev, init_data, pmic); if (IS_ERR(pmic->rdev[i])) { dev_err(&i2c->dev, "failed to register %s\n", id->name); -- cgit v0.10.2 From cc0fc0bbeb17dd33ed7bfea58d0178e9c007ff67 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 1 Sep 2010 08:55:22 -0600 Subject: spi/spi_s3c64xx: Make probe more robust against missing board config The S3C64xx SPI driver requires the machine to call s3c64xx_spi_set_info() to select a few options, including the clock to use for the SPI controller. If this is not done then a NULL will be passed as the clock name for clk_get(), causing an obscure crash. Guard against this and other missing configuration by validating that the clock name has been filled in in the platform data that ets passed in. Signed-off-by: Mark Brown Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index 9736581..a0b63b7 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -919,6 +919,13 @@ static int __init s3c64xx_spi_probe(struct platform_device *pdev) return -ENODEV; } + sci = pdev->dev.platform_data; + if (!sci->src_clk_name) { + dev_err(&pdev->dev, + "Board init must call s3c64xx_spi_set_info()\n"); + return -EINVAL; + } + /* Check for availability of necessary resource */ dmatx_res = platform_get_resource(pdev, IORESOURCE_DMA, 0); @@ -946,8 +953,6 @@ static int __init s3c64xx_spi_probe(struct platform_device *pdev) return -ENOMEM; } - sci = pdev->dev.platform_data; - platform_set_drvdata(pdev, master); sdd = spi_master_get_devdata(master); -- cgit v0.10.2 From 8944f4f3d9695e04f7091ab61da9aab7a4be5846 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 1 Sep 2010 08:55:23 -0600 Subject: spi/spi_s3c64xx: Staticise non-exported functions Signed-off-by: Mark Brown Acked-by: Jassi Brar Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index a0b63b7..7e627f7 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -447,8 +447,8 @@ static void s3c64xx_spi_config(struct s3c64xx_spi_driver_data *sdd) writel(val, regs + S3C64XX_SPI_CLK_CFG); } -void s3c64xx_spi_dma_rxcb(struct s3c2410_dma_chan *chan, void *buf_id, - int size, enum s3c2410_dma_buffresult res) +static void s3c64xx_spi_dma_rxcb(struct s3c2410_dma_chan *chan, void *buf_id, + int size, enum s3c2410_dma_buffresult res) { struct s3c64xx_spi_driver_data *sdd = buf_id; unsigned long flags; @@ -467,8 +467,8 @@ void s3c64xx_spi_dma_rxcb(struct s3c2410_dma_chan *chan, void *buf_id, spin_unlock_irqrestore(&sdd->lock, flags); } -void s3c64xx_spi_dma_txcb(struct s3c2410_dma_chan *chan, void *buf_id, - int size, enum s3c2410_dma_buffresult res) +static void s3c64xx_spi_dma_txcb(struct s3c2410_dma_chan *chan, void *buf_id, + int size, enum s3c2410_dma_buffresult res) { struct s3c64xx_spi_driver_data *sdd = buf_id; unsigned long flags; -- cgit v0.10.2 From 9f1a1fca35066117353994bff80a5115cddad7a2 Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Wed, 1 Sep 2010 08:55:23 -0600 Subject: of: Fix missing includes - ll_temac It is the next patch which is fixing missing header which were removed from prom.h. Related patches: "of/address: Clean up function declarations" (sha1 id 22ae782f8) "of: Fix missing includes" (sha1 id f1ca09b2b) Signed-off-by: Michal Simek Signed-off-by: Grant Likely diff --git a/drivers/net/ll_temac_main.c b/drivers/net/ll_temac_main.c index bdf2149..87f0a93 100644 --- a/drivers/net/ll_temac_main.c +++ b/drivers/net/ll_temac_main.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include #include /* needed for sizeof(tcphdr) */ diff --git a/drivers/net/ll_temac_mdio.c b/drivers/net/ll_temac_mdio.c index 5ae28c9..8cf9d4f 100644 --- a/drivers/net/ll_temac_mdio.c +++ b/drivers/net/ll_temac_mdio.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include -- cgit v0.10.2 From 4e4438b86527e8bf1f49503a30d487e401e64f9c Mon Sep 17 00:00:00 2001 From: Anton Vorontsov Date: Wed, 1 Sep 2010 08:55:24 -0600 Subject: gpiolib: Add 'struct gpio_chip' forward declaration for !GPIOLIB case With CONFIG_GPIOLIB=n, the 'struct gpio_chip' is not declared, so the following pops up on PowerPC: cc1: warnings being treated as errors In file included from arch/powerpc/platforms/52xx/mpc52xx_common.c:19: include/linux/of_gpio.h:74: warning: 'struct gpio_chip' declared inside parameter list include/linux/of_gpio.h:74: warning: its scope is only this definition or declaration, which is probably not what you want include/linux/of_gpio.h:75: warning: 'struct gpio_chip' declared inside parameter list make[2]: *** [arch/powerpc/platforms/52xx/mpc52xx_common.o] Error 1 This patch fixes the issue by providing the proper forward declaration. Signed-off-by: Anton Vorontsov Signed-off-by: Grant Likely diff --git a/include/linux/gpio.h b/include/linux/gpio.h index 03f616b..e41f7dd 100644 --- a/include/linux/gpio.h +++ b/include/linux/gpio.h @@ -13,6 +13,7 @@ #include struct device; +struct gpio_chip; /* * Some platforms don't support the GPIO programming interface. -- cgit v0.10.2 From 34860089c9e8abcc77428d29743b37ff756197e7 Mon Sep 17 00:00:00 2001 From: David Lamparter Date: Mon, 30 Aug 2010 23:54:17 +0200 Subject: spi: free children in spi_unregister_master, not siblings introduced by 49dce689 ("spi doesn't need class_device") and bad-fixed by 350d0076 ("spi: fix double-free on spi_unregister_master"), spi_unregister_master would previously device_unregister all of the spi master's siblings (instead of its children). hilarity ensues. fix it to unregister children. Signed-off-by: David Lamparter Signed-off-by: Grant Likely diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c index a9e5c79..0bcf4c1 100644 --- a/drivers/spi/spi.c +++ b/drivers/spi/spi.c @@ -554,11 +554,9 @@ done: EXPORT_SYMBOL_GPL(spi_register_master); -static int __unregister(struct device *dev, void *master_dev) +static int __unregister(struct device *dev, void *null) { - /* note: before about 2.6.14-rc1 this would corrupt memory: */ - if (dev != master_dev) - spi_unregister_device(to_spi_device(dev)); + spi_unregister_device(to_spi_device(dev)); return 0; } @@ -576,8 +574,7 @@ void spi_unregister_master(struct spi_master *master) { int dummy; - dummy = device_for_each_child(master->dev.parent, &master->dev, - __unregister); + dummy = device_for_each_child(&master->dev, NULL, __unregister); device_unregister(&master->dev); } EXPORT_SYMBOL_GPL(spi_unregister_master); -- cgit v0.10.2 From 59482fe5959675f180469ae95e4fcd0a15920ced Mon Sep 17 00:00:00 2001 From: Akinobu Mita Date: Tue, 31 Aug 2010 17:43:51 +0900 Subject: powerpc/512x: fix clk_get() return value clk_get() should return an ERR_PTR value on error, not NULL. Signed-off-by: Akinobu Mita Signed-off-by: Grant Likely diff --git a/arch/powerpc/platforms/512x/clock.c b/arch/powerpc/platforms/512x/clock.c index 5b243bd..3dc2a8d 100644 --- a/arch/powerpc/platforms/512x/clock.c +++ b/arch/powerpc/platforms/512x/clock.c @@ -57,7 +57,7 @@ static struct clk *mpc5121_clk_get(struct device *dev, const char *id) int id_match = 0; if (dev == NULL || id == NULL) - return NULL; + return clk; mutex_lock(&clocks_mutex); list_for_each_entry(p, &clocks, node) { -- cgit v0.10.2 From 269f45c25028c75fe10e6d9be86e7202ab461fbc Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Wed, 1 Sep 2010 14:50:50 +0200 Subject: oprofile, x86: fix init_sysfs() function stub MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The use of the return value of init_sysfs() with commit 10f0412 oprofile, x86: fix init_sysfs error handling discovered the following build error for !CONFIG_PM: .../linux/arch/x86/oprofile/nmi_int.c: In function ‘op_nmi_init’: .../linux/arch/x86/oprofile/nmi_int.c:784: error: expected expression before ‘do’ make[2]: *** [arch/x86/oprofile/nmi_int.o] Error 1 make[1]: *** [arch/x86/oprofile] Error 2 This patch fixes this. Reported-by: Ingo Molnar Cc: stable@kernel.org Signed-off-by: Robert Richter diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index 73a41d3..cfe4faa 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -585,8 +585,10 @@ static void exit_sysfs(void) } #else -#define init_sysfs() do { } while (0) -#define exit_sysfs() do { } while (0) + +static inline int init_sysfs(void) { return 0; } +static inline void exit_sysfs(void) { } + #endif /* CONFIG_PM */ static int __init p4_init(char **cpu_type) -- cgit v0.10.2 From 048e78a5bc22c27410cb5ca9680c3c7ac400607f Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Thu, 2 Sep 2010 08:35:47 +0200 Subject: ALSA: hda - Add a new hp-laptop model for Conexant 5066, tested on HP G60 This new model adds the following functionality to HP G60: - Automute of internal speakers - Autoswitch of internal/external mics - Remove SPDIF not physically present BugLink: http://launchpad.net/bugs/587388 Cc: stable@kernel.org Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai diff --git a/Documentation/sound/alsa/HD-Audio-Models.txt b/Documentation/sound/alsa/HD-Audio-Models.txt index ce46fa1..37c6aad 100644 --- a/Documentation/sound/alsa/HD-Audio-Models.txt +++ b/Documentation/sound/alsa/HD-Audio-Models.txt @@ -296,6 +296,7 @@ Conexant 5051 Conexant 5066 ============= laptop Basic Laptop config (default) + hp-laptop HP laptops, e g G60 dell-laptop Dell laptops dell-vostro Dell Vostro olpc-xo-1_5 OLPC XO 1.5 diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c index 5cdb80e..4f06199 100644 --- a/sound/pci/hda/patch_conexant.c +++ b/sound/pci/hda/patch_conexant.c @@ -116,6 +116,7 @@ struct conexant_spec { unsigned int dell_vostro:1; unsigned int ideapad:1; unsigned int thinkpad:1; + unsigned int hp_laptop:1; unsigned int ext_mic_present; unsigned int recording; @@ -2299,6 +2300,18 @@ static void cxt5066_ideapad_automic(struct hda_codec *codec) } } +/* toggle input of built-in digital mic and mic jack appropriately */ +static void cxt5066_hp_laptop_automic(struct hda_codec *codec) +{ + unsigned int present; + + present = snd_hda_jack_detect(codec, 0x1b); + snd_printdd("CXT5066: external microphone present=%d\n", present); + snd_hda_codec_write(codec, 0x17, 0, AC_VERB_SET_CONNECT_SEL, + present ? 1 : 3); +} + + /* toggle input of built-in digital mic and mic jack appropriately order is: external mic -> dock mic -> interal mic */ static void cxt5066_thinkpad_automic(struct hda_codec *codec) @@ -2408,6 +2421,20 @@ static void cxt5066_ideapad_event(struct hda_codec *codec, unsigned int res) } /* unsolicited event for jack sensing */ +static void cxt5066_hp_laptop_event(struct hda_codec *codec, unsigned int res) +{ + snd_printdd("CXT5066_hp_laptop: unsol event %x (%x)\n", res, res >> 26); + switch (res >> 26) { + case CONEXANT_HP_EVENT: + cxt5066_hp_automute(codec); + break; + case CONEXANT_MIC_EVENT: + cxt5066_hp_laptop_automic(codec); + break; + } +} + +/* unsolicited event for jack sensing */ static void cxt5066_thinkpad_event(struct hda_codec *codec, unsigned int res) { snd_printdd("CXT5066_thinkpad: unsol event %x (%x)\n", res, res >> 26); @@ -2989,6 +3016,14 @@ static struct hda_verb cxt5066_init_verbs_portd_lo[] = { { } /* end */ }; + +static struct hda_verb cxt5066_init_verbs_hp_laptop[] = { + {0x14, AC_VERB_SET_CONNECT_SEL, 0x0}, + {0x19, AC_VERB_SET_UNSOLICITED_ENABLE, AC_USRSP_EN | CONEXANT_HP_EVENT}, + {0x1b, AC_VERB_SET_UNSOLICITED_ENABLE, AC_USRSP_EN | CONEXANT_MIC_EVENT}, + { } /* end */ +}; + /* initialize jack-sensing, too */ static int cxt5066_init(struct hda_codec *codec) { @@ -3004,6 +3039,8 @@ static int cxt5066_init(struct hda_codec *codec) cxt5066_ideapad_automic(codec); else if (spec->thinkpad) cxt5066_thinkpad_automic(codec); + else if (spec->hp_laptop) + cxt5066_hp_laptop_automic(codec); } cxt5066_set_mic_boost(codec); return 0; @@ -3031,6 +3068,7 @@ enum { CXT5066_DELL_VOSTO, /* Dell Vostro 1015i */ CXT5066_IDEAPAD, /* Lenovo IdeaPad U150 */ CXT5066_THINKPAD, /* Lenovo ThinkPad T410s, others? */ + CXT5066_HP_LAPTOP, /* HP Laptop */ CXT5066_MODELS }; @@ -3041,6 +3079,7 @@ static const char *cxt5066_models[CXT5066_MODELS] = { [CXT5066_DELL_VOSTO] = "dell-vostro", [CXT5066_IDEAPAD] = "ideapad", [CXT5066_THINKPAD] = "thinkpad", + [CXT5066_HP_LAPTOP] = "hp-laptop", }; static struct snd_pci_quirk cxt5066_cfg_tbl[] = { @@ -3052,6 +3091,7 @@ static struct snd_pci_quirk cxt5066_cfg_tbl[] = { SND_PCI_QUIRK(0x1028, 0x02d8, "Dell Vostro", CXT5066_DELL_VOSTO), SND_PCI_QUIRK(0x1028, 0x0402, "Dell Vostro", CXT5066_DELL_VOSTO), SND_PCI_QUIRK(0x1028, 0x0408, "Dell Inspiron One 19T", CXT5066_IDEAPAD), + SND_PCI_QUIRK(0x103c, 0x360b, "HP G60", CXT5066_HP_LAPTOP), SND_PCI_QUIRK(0x1179, 0xff50, "Toshiba Satellite P500-PSPGSC-01800T", CXT5066_OLPC_XO_1_5), SND_PCI_QUIRK(0x1179, 0xffe0, "Toshiba Satellite Pro T130-15F", CXT5066_OLPC_XO_1_5), SND_PCI_QUIRK(0x17aa, 0x21b2, "Thinkpad X100e", CXT5066_IDEAPAD), @@ -3116,6 +3156,23 @@ static int patch_cxt5066(struct hda_codec *codec) spec->num_init_verbs++; spec->dell_automute = 1; break; + case CXT5066_HP_LAPTOP: + codec->patch_ops.init = cxt5066_init; + codec->patch_ops.unsol_event = cxt5066_hp_laptop_event; + spec->init_verbs[spec->num_init_verbs] = + cxt5066_init_verbs_hp_laptop; + spec->num_init_verbs++; + spec->hp_laptop = 1; + spec->mixers[spec->num_mixers++] = cxt5066_mixer_master; + spec->mixers[spec->num_mixers++] = cxt5066_mixers; + /* no S/PDIF out */ + spec->multiout.dig_out_nid = 0; + /* input source automatically selected */ + spec->input_mux = NULL; + spec->port_d_mode = 0; + spec->mic_boost = 3; /* default 30dB gain */ + break; + case CXT5066_OLPC_XO_1_5: codec->patch_ops.init = cxt5066_olpc_init; codec->patch_ops.unsol_event = cxt5066_olpc_unsol_event; -- cgit v0.10.2 From 7b6717e144de6592e614fd7fc3b914b6bf686a9d Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Thu, 2 Sep 2010 17:13:15 +0800 Subject: ALSA: usb-audio: Assume first control interface is for audio For devices with more than one control interface, let's assume the first one contains the audio controls. Unfortunately, there is no field in any of the descriptors to tell us whether a control interface is for audio or MIDI controls, so a better check is not easy to implement. On a composite device with audio and MIDI functions, for example, the code currently overwrites chip->ctrl_intf, causing operations on the control interface to fail if they are issued after the device probe. Signed-off-by: Daniel Mack Signed-off-by: Takashi Iwai diff --git a/sound/usb/card.c b/sound/usb/card.c index 9feb00c..b443a33 100644 --- a/sound/usb/card.c +++ b/sound/usb/card.c @@ -465,7 +465,13 @@ static void *snd_usb_audio_probe(struct usb_device *dev, goto __error; } - chip->ctrl_intf = alts; + /* + * For devices with more than one control interface, we assume the + * first contains the audio controls. We might need a more specific + * check here in the future. + */ + if (!chip->ctrl_intf) + chip->ctrl_intf = alts; if (err > 0) { /* create normal USB audio interfaces */ -- cgit v0.10.2 From 026b5ca3b65f878019a8eb0c7a702cd5c20a4104 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 1 Sep 2010 14:33:29 +0100 Subject: ARM: 6344/1: Mark CPU_32v6K as depended on CPU_V7 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit CPU_32v6K is selected by CPU_V7 but it only depends on CPU_V6. Signed-off-by: Catalin Marinas Acked-by: Uwe Kleine-König Signed-off-by: Russell King diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index 33c3f57..a0a2928 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -398,7 +398,7 @@ config CPU_V6 # ARMv6k config CPU_32v6K bool "Support ARM V6K processor extensions" if !SMP - depends on CPU_V6 + depends on CPU_V6 || CPU_V7 default y if SMP && !(ARCH_MX3 || ARCH_OMAP2) help Say Y here if your ARMv6 processor supports the 'K' extension. -- cgit v0.10.2 From 65b4711ff513767341aa1915c822de6ec0de65cb Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Thu, 2 Sep 2010 09:32:08 +0100 Subject: ARM: 6352/1: perf: fix event validation The validate_event function in the ARM perf events backend has the following problems: 1.) Events that are disabled count towards the cost. 2.) Events associated with other PMUs [for example, software events or breakpoints] do not count towards the cost, but do fail validation, causing the group to fail. This patch changes validate_event so that it ignores events in the PERF_EVENT_STATE_OFF state or that are scheduled for other PMUs. Reported-by: Pawel Moll Acked-by: Jamie Iles Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index e213925..ecbb028 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -319,8 +319,8 @@ validate_event(struct cpu_hw_events *cpuc, { struct hw_perf_event fake_event = event->hw; - if (event->pmu && event->pmu != &pmu) - return 0; + if (event->pmu != &pmu || event->state <= PERF_EVENT_STATE_OFF) + return 1; return armpmu->get_event_idx(cpuc, &fake_event) >= 0; } -- cgit v0.10.2 From 8f34a430ac16d5fbd9d6b383184d35e152f5a963 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 2 Sep 2010 15:23:16 -0400 Subject: nfsd4: mask out non-access bits in nfs4_access_to_omode This fixes an unnecessary BUG(). Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 3dfef06..cf0d2ff 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -440,7 +440,7 @@ test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) { static int nfs4_access_to_omode(u32 access) { - switch (access) { + switch (access & NFS4_SHARE_ACCESS_BOTH) { case NFS4_SHARE_ACCESS_READ: return O_RDONLY; case NFS4_SHARE_ACCESS_WRITE: -- cgit v0.10.2 From d842a93c4b3a20d31f4fb97357e0964210d6066f Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Thu, 12 Aug 2010 14:31:05 +0200 Subject: [SCSI] fix bio.bi_rw handling Return of the bi_rw tests is no longer bool after commit 74450be1. So testing against constants doesn't make sense anymore. Fix this bug in osd_req_read by removing "== 1" in test. This is not a problem now, where REQ_WRITE is 1, but this can change in the future and we don't want to rely on that. Signed-off-by: Jiri Slaby Signed-off-by: James Bottomley diff --git a/drivers/scsi/osd/osd_initiator.c b/drivers/scsi/osd/osd_initiator.c index fda4de3..e88bbdd 100644 --- a/drivers/scsi/osd/osd_initiator.c +++ b/drivers/scsi/osd/osd_initiator.c @@ -865,7 +865,7 @@ void osd_req_read(struct osd_request *or, { _osd_req_encode_common(or, OSD_ACT_READ, obj, offset, len); WARN_ON(or->in.bio || or->in.total_bytes); - WARN_ON(1 == (bio->bi_rw & REQ_WRITE)); + WARN_ON(bio->bi_rw & REQ_WRITE); or->in.bio = bio; or->in.total_bytes = len; } -- cgit v0.10.2 From b15d05b0d358cedf9c4d420a60d2ee2d0f530788 Mon Sep 17 00:00:00 2001 From: Jayamohan Kallickal Date: Thu, 12 Aug 2010 23:36:06 +0530 Subject: [SCSI] be2iscsi: Fix for Login failure The current code in tree has problems with Login. This patch fixes the Login Failure . Signed-off-by: Jayamohan Kallickal [mnc: Can't believe I missed that.] Reviewed-by: Mike Christie Signed-off-by: James Bottomley diff --git a/drivers/scsi/be2iscsi/be_iscsi.c b/drivers/scsi/be2iscsi/be_iscsi.c index 7d4d227..7f11f3e 100644 --- a/drivers/scsi/be2iscsi/be_iscsi.c +++ b/drivers/scsi/be2iscsi/be_iscsi.c @@ -300,8 +300,7 @@ int beiscsi_get_host_param(struct Scsi_Host *shost, enum iscsi_host_param param, char *buf) { struct beiscsi_hba *phba = (struct beiscsi_hba *)iscsi_host_priv(shost); - int len = 0; - int status; + int status = 0; SE_DEBUG(DBG_LVL_8, "In beiscsi_get_host_param, param= %d\n", param); switch (param) { @@ -315,7 +314,7 @@ int beiscsi_get_host_param(struct Scsi_Host *shost, default: return iscsi_host_get_param(shost, param, buf); } - return len; + return status; } int beiscsi_get_macaddr(char *buf, struct beiscsi_hba *phba) diff --git a/drivers/scsi/be2iscsi/be_mgmt.c b/drivers/scsi/be2iscsi/be_mgmt.c index 26350e4..877324f 100644 --- a/drivers/scsi/be2iscsi/be_mgmt.c +++ b/drivers/scsi/be2iscsi/be_mgmt.c @@ -368,7 +368,7 @@ int mgmt_open_connection(struct beiscsi_hba *phba, memset(req, 0, sizeof(*req)); wrb->tag0 |= tag; - be_wrb_hdr_prepare(wrb, sizeof(*req), true, 1); + be_wrb_hdr_prepare(wrb, sizeof(*req), false, 1); be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_ISCSI, OPCODE_COMMON_ISCSI_TCP_CONNECT_AND_OFFLOAD, sizeof(*req)); -- cgit v0.10.2 From 36ed2176fedaa180b8ea3cdacf68c958e0090a3c Mon Sep 17 00:00:00 2001 From: "Stephen M. Cameron" Date: Wed, 25 Aug 2010 10:44:14 -0500 Subject: [SCSI] hpsa: disable doorbell reset on reset_devices The doorbell reset initially appears to work correctly, the controller resets, comes up, some i/o can even be done, but on at least some Smart Arrays in some servers, it eventually causes a subsequent controller lockup due to some kind of PCIe error, and kdump can end up leaving the root filesystem in an unbootable state. For this reason, until the problem is fixed, or at least isolated to certain hardware enough to be avoided, the doorbell reset should not be used at all. Signed-off-by: Stephen M. Cameron Signed-off-by: James Bottomley diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c index 4f5551b..c5d0606 100644 --- a/drivers/scsi/hpsa.c +++ b/drivers/scsi/hpsa.c @@ -3231,6 +3231,12 @@ static __devinit int hpsa_kdump_hard_reset_controller(struct pci_dev *pdev) misc_fw_support = readl(&cfgtable->misc_fw_support); use_doorbell = misc_fw_support & MISC_FW_DOORBELL_RESET; + /* The doorbell reset seems to cause lockups on some Smart + * Arrays (e.g. P410, P410i, maybe others). Until this is + * fixed or at least isolated, avoid the doorbell reset. + */ + use_doorbell = 0; + rc = hpsa_controller_hard_reset(pdev, vaddr, use_doorbell); if (rc) goto unmap_cfgtable; -- cgit v0.10.2 From 6f131ce1dfa9b283ddc212df42b015d152c670a5 Mon Sep 17 00:00:00 2001 From: Jean Sacren Date: Wed, 25 Aug 2010 17:58:09 -0600 Subject: [SCSI] Fix warning: zero-length gnu_printf format string warning: zero-length gnu_printf format string Fix the above warning by inserting a space into the literal string. Signed-off-by: Jean Sacren Signed-off-by: James Bottomley diff --git a/drivers/scsi/constants.c b/drivers/scsi/constants.c index cd05e04..d0c8234 100644 --- a/drivers/scsi/constants.c +++ b/drivers/scsi/constants.c @@ -1404,13 +1404,13 @@ void scsi_print_sense(char *name, struct scsi_cmnd *cmd) { struct scsi_sense_hdr sshdr; - scmd_printk(KERN_INFO, cmd, ""); + scmd_printk(KERN_INFO, cmd, " "); scsi_decode_sense_buffer(cmd->sense_buffer, SCSI_SENSE_BUFFERSIZE, &sshdr); scsi_show_sense_hdr(&sshdr); scsi_decode_sense_extras(cmd->sense_buffer, SCSI_SENSE_BUFFERSIZE, &sshdr); - scmd_printk(KERN_INFO, cmd, ""); + scmd_printk(KERN_INFO, cmd, " "); scsi_show_extd_sense(sshdr.asc, sshdr.ascq); } EXPORT_SYMBOL(scsi_print_sense); @@ -1453,7 +1453,7 @@ EXPORT_SYMBOL(scsi_show_result); void scsi_print_result(struct scsi_cmnd *cmd) { - scmd_printk(KERN_INFO, cmd, ""); + scmd_printk(KERN_INFO, cmd, " "); scsi_show_result(cmd->result); } EXPORT_SYMBOL(scsi_print_result); -- cgit v0.10.2 From 2e4c332913b5d39fef686b3964098f0d8fd97ead Mon Sep 17 00:00:00 2001 From: David Miller Date: Tue, 31 Aug 2010 13:35:31 -0700 Subject: [SCSI] sd, sym53c8xx: Remove warnings after vsprintf %pV introducation. GCC warns about empty printf format strings, and after the addition of %pV these existing such cases in the scsi driver layer were exposed enough for the compiler to start seeing them. Based almost entirely upon a patch by Joe Perches. [jejb: fix up sym53c8xx msg] Signed-off-by: David S. Miller Signed-off-by: James Bottomley diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index 2714bec..cd71f46 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -2625,15 +2625,15 @@ module_exit(exit_sd); static void sd_print_sense_hdr(struct scsi_disk *sdkp, struct scsi_sense_hdr *sshdr) { - sd_printk(KERN_INFO, sdkp, ""); + sd_printk(KERN_INFO, sdkp, " "); scsi_show_sense_hdr(sshdr); - sd_printk(KERN_INFO, sdkp, ""); + sd_printk(KERN_INFO, sdkp, " "); scsi_show_extd_sense(sshdr->asc, sshdr->ascq); } static void sd_print_result(struct scsi_disk *sdkp, int result) { - sd_printk(KERN_INFO, sdkp, ""); + sd_printk(KERN_INFO, sdkp, " "); scsi_show_result(result); } diff --git a/drivers/scsi/sym53c8xx_2/sym_hipd.c b/drivers/scsi/sym53c8xx_2/sym_hipd.c index a7bc8b7..2c3e89d 100644 --- a/drivers/scsi/sym53c8xx_2/sym_hipd.c +++ b/drivers/scsi/sym53c8xx_2/sym_hipd.c @@ -72,10 +72,7 @@ static void sym_printl_hex(u_char *p, int n) static void sym_print_msg(struct sym_ccb *cp, char *label, u_char *msg) { - if (label) - sym_print_addr(cp->cmd, "%s: ", label); - else - sym_print_addr(cp->cmd, ""); + sym_print_addr(cp->cmd, "%s: ", label); spi_print_msg(msg); printf("\n"); @@ -4558,7 +4555,8 @@ static void sym_int_sir(struct sym_hcb *np) switch (np->msgin [2]) { case M_X_MODIFY_DP: if (DEBUG_FLAGS & DEBUG_POINTER) - sym_print_msg(cp, NULL, np->msgin); + sym_print_msg(cp, "extended msg ", + np->msgin); tmp = (np->msgin[3]<<24) + (np->msgin[4]<<16) + (np->msgin[5]<<8) + (np->msgin[6]); sym_modify_dp(np, tp, cp, tmp); @@ -4585,7 +4583,7 @@ static void sym_int_sir(struct sym_hcb *np) */ case M_IGN_RESIDUE: if (DEBUG_FLAGS & DEBUG_POINTER) - sym_print_msg(cp, NULL, np->msgin); + sym_print_msg(cp, "1 or 2 byte ", np->msgin); if (cp->host_flags & HF_SENSE) OUTL_DSP(np, SCRIPTA_BA(np, clrack)); else -- cgit v0.10.2 From dc4e96ce2dceb649224ee84f83592aac8c54c9b7 Mon Sep 17 00:00:00 2001 From: Steve Wise Date: Sat, 28 Aug 2010 13:35:05 +0000 Subject: RDMA/cxgb3: Don't exceed the max HW CQ depth The max depth supported by T3 is 64K entries. This fixes a bug introduced in commit 9918b28d ("RDMA/cxgb3: Increase the max CQ depth") that causes stalls and possibly crashes in large MPI clusters. Signed-off-by: Steve Wise Cc: Signed-off-by: Roland Dreier diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.h b/drivers/infiniband/hw/cxgb3/cxio_hal.h index 8f0caf7..78fbe9f 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.h +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.h @@ -53,7 +53,7 @@ #define T3_MAX_PBL_SIZE 256 #define T3_MAX_RQ_SIZE 1024 #define T3_MAX_QP_DEPTH (T3_MAX_RQ_SIZE-1) -#define T3_MAX_CQ_DEPTH 262144 +#define T3_MAX_CQ_DEPTH 65536 #define T3_MAX_NUM_STAG (1<<15) #define T3_MAX_MR_SIZE 0x100000000ULL #define T3_PAGESIZE_MASK 0xffff000 /* 4KB-128MB */ -- cgit v0.10.2 From 2e556b5b320838fde98480a1f6cf220a5af200fc Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Thu, 2 Sep 2010 15:07:47 -0400 Subject: perf, x86: Fix accidentally ack'ing a second event on intel perf counter During testing of a patch to stop having the perf subsytem swallow nmis, it was uncovered that Nehalem boxes were randomly getting unknown nmis when using the perf tool. Moving the ack'ing of the PMI closer to when we get the status allows the hardware to properly re-set the PMU bit signaling another PMI was triggered during the processing of the first PMI. This allows the new logic for dealing with the shortcomings of multiple PMIs to handle the extra NMI by 'eat'ing it later. Now one can wonder why are we getting a second PMI when we disable all the PMUs in the begining of the NMI handler to prevent such a case, for that I do not know. But I know the fix below helps deal with this quirk. Tested on multiple Nehalems where the problem was occuring. With the patch, the code now loops a second time to handle the second PMI (whereas before it was not). Signed-off-by: Don Zickus Cc: peterz@infradead.org Cc: robert.richter@amd.com Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-2-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index d8d86d0..1297bf1 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -712,7 +712,7 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) struct perf_sample_data data; struct cpu_hw_events *cpuc; int bit, loops; - u64 ack, status; + u64 status; perf_sample_data_init(&data, 0); @@ -728,6 +728,7 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) loops = 0; again: + intel_pmu_ack_status(status); if (++loops > 100) { WARN_ONCE(1, "perfevents: irq loop stuck!\n"); perf_event_print_debug(); @@ -736,7 +737,6 @@ again: } inc_irq_stat(apic_perf_irqs); - ack = status; intel_pmu_lbr_read(); @@ -761,8 +761,6 @@ again: x86_pmu_stop(event); } - intel_pmu_ack_status(ack); - /* * Repeat if there is more work to be done: */ -- cgit v0.10.2 From de725dec9de7a7541996176d59cf8542365b8b0e Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 2 Sep 2010 15:07:49 -0400 Subject: perf, x86: Fix handle_irq return values Now that we rely on the number of handled overflows, ensure all handle_irq implementations actually return the right number. Signed-off-by: Peter Zijlstra Signed-off-by: Don Zickus Cc: peterz@infradead.org Cc: robert.richter@amd.com Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-4-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 1297bf1..ee05c90 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -713,6 +713,7 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) struct cpu_hw_events *cpuc; int bit, loops; u64 status; + int handled = 0; perf_sample_data_init(&data, 0); @@ -743,12 +744,16 @@ again: /* * PEBS overflow sets bit 62 in the global status register */ - if (__test_and_clear_bit(62, (unsigned long *)&status)) + if (__test_and_clear_bit(62, (unsigned long *)&status)) { + handled++; x86_pmu.drain_pebs(regs); + } for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) { struct perf_event *event = cpuc->events[bit]; + handled++; + if (!test_bit(bit, cpuc->active_mask)) continue; @@ -770,7 +775,7 @@ again: done: intel_pmu_enable_all(0); - return 1; + return handled; } static struct event_constraint * diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 7e578e9..b560db3 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -692,7 +692,7 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) inc_irq_stat(apic_perf_irqs); } - return handled > 0; + return handled; } /* -- cgit v0.10.2 From 4177c42a6301a34c20038ec2771a33dcc30bb338 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Thu, 2 Sep 2010 15:07:48 -0400 Subject: perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index f2da20f..3efdf28 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1154,7 +1154,7 @@ static int x86_pmu_handle_irq(struct pt_regs *regs) /* * event overflow */ - handled = 1; + handled++; data.period = event->hw.last_period; if (!x86_perf_event_set_period(event)) @@ -1200,12 +1200,20 @@ void perf_events_lapic_init(void) apic_write(APIC_LVTPC, APIC_DM_NMI); } +struct pmu_nmi_state { + unsigned int marked; + int handled; +}; + +static DEFINE_PER_CPU(struct pmu_nmi_state, pmu_nmi); + static int __kprobes perf_event_nmi_handler(struct notifier_block *self, unsigned long cmd, void *__args) { struct die_args *args = __args; - struct pt_regs *regs; + unsigned int this_nmi; + int handled; if (!atomic_read(&active_events)) return NOTIFY_DONE; @@ -1214,22 +1222,47 @@ perf_event_nmi_handler(struct notifier_block *self, case DIE_NMI: case DIE_NMI_IPI: break; - + case DIE_NMIUNKNOWN: + this_nmi = percpu_read(irq_stat.__nmi_count); + if (this_nmi != __get_cpu_var(pmu_nmi).marked) + /* let the kernel handle the unknown nmi */ + return NOTIFY_DONE; + /* + * This one is a PMU back-to-back nmi. Two events + * trigger 'simultaneously' raising two back-to-back + * NMIs. If the first NMI handles both, the latter + * will be empty and daze the CPU. So, we drop it to + * avoid false-positive 'unknown nmi' messages. + */ + return NOTIFY_STOP; default: return NOTIFY_DONE; } - regs = args->regs; - apic_write(APIC_LVTPC, APIC_DM_NMI); - /* - * Can't rely on the handled return value to say it was our NMI, two - * events could trigger 'simultaneously' raising two back-to-back NMIs. - * - * If the first NMI handles both, the latter will be empty and daze - * the CPU. - */ - x86_pmu.handle_irq(regs); + + handled = x86_pmu.handle_irq(args->regs); + if (!handled) + return NOTIFY_DONE; + + this_nmi = percpu_read(irq_stat.__nmi_count); + if ((handled > 1) || + /* the next nmi could be a back-to-back nmi */ + ((__get_cpu_var(pmu_nmi).marked == this_nmi) && + (__get_cpu_var(pmu_nmi).handled > 1))) { + /* + * We could have two subsequent back-to-back nmis: The + * first handles more than one counter, the 2nd + * handles only one counter and the 3rd handles no + * counter. + * + * This is the 2nd nmi because the previous was + * handling more than one counter. We will mark the + * next (3rd) and then drop it if unhandled. + */ + __get_cpu_var(pmu_nmi).marked = this_nmi + 1; + __get_cpu_var(pmu_nmi).handled = handled; + } return NOTIFY_STOP; } -- cgit v0.10.2 From ef5dc121d5a0bb1fa477c5395277259f07d318a3 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 2 Sep 2010 15:48:16 -0700 Subject: mutex: Fix annotations to include it in kernel-locking docbook Fix kernel-doc notation in linux/mutex.h and kernel/mutex.c, then add these 2 files to the kernel-locking docbook as the Mutex API reference chapter. Add one API function to mutex-design.txt and correct a typo in that file. Signed-off-by: Randy Dunlap Cc: Rusty Russell LKML-Reference: <20100902154816.6cc2f9ad.randy.dunlap@oracle.com> Signed-off-by: Ingo Molnar diff --git a/Documentation/DocBook/kernel-locking.tmpl b/Documentation/DocBook/kernel-locking.tmpl index 0b1a3f9..a0d479d 100644 --- a/Documentation/DocBook/kernel-locking.tmpl +++ b/Documentation/DocBook/kernel-locking.tmpl @@ -1961,6 +1961,12 @@ machines due to caching. + + Mutex API reference +!Iinclude/linux/mutex.h +!Ekernel/mutex.c + + Further reading diff --git a/Documentation/mutex-design.txt b/Documentation/mutex-design.txt index c91ccc0..38c10fd 100644 --- a/Documentation/mutex-design.txt +++ b/Documentation/mutex-design.txt @@ -9,7 +9,7 @@ firstly, there's nothing wrong with semaphores. But if the simpler mutex semantics are sufficient for your code, then there are a couple of advantages of mutexes: - - 'struct mutex' is smaller on most architectures: .e.g on x86, + - 'struct mutex' is smaller on most architectures: E.g. on x86, 'struct semaphore' is 20 bytes, 'struct mutex' is 16 bytes. A smaller structure size means less RAM footprint, and better CPU-cache utilization. @@ -136,3 +136,4 @@ the APIs of 'struct mutex' have been streamlined: void mutex_lock_nested(struct mutex *lock, unsigned int subclass); int mutex_lock_interruptible_nested(struct mutex *lock, unsigned int subclass); + int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); diff --git a/include/linux/mutex.h b/include/linux/mutex.h index 878cab4..f363bc8 100644 --- a/include/linux/mutex.h +++ b/include/linux/mutex.h @@ -78,6 +78,14 @@ struct mutex_waiter { # include #else # define __DEBUG_MUTEX_INITIALIZER(lockname) +/** + * mutex_init - initialize the mutex + * @mutex: the mutex to be initialized + * + * Initialize the mutex to unlocked state. + * + * It is not allowed to initialize an already locked mutex. + */ # define mutex_init(mutex) \ do { \ static struct lock_class_key __key; \ diff --git a/kernel/mutex.c b/kernel/mutex.c index 4c0b7b3..200407c 100644 --- a/kernel/mutex.c +++ b/kernel/mutex.c @@ -36,15 +36,6 @@ # include #endif -/*** - * mutex_init - initialize the mutex - * @lock: the mutex to be initialized - * @key: the lock_class_key for the class; used by mutex lock debugging - * - * Initialize the mutex to unlocked state. - * - * It is not allowed to initialize an already locked mutex. - */ void __mutex_init(struct mutex *lock, const char *name, struct lock_class_key *key) { @@ -68,7 +59,7 @@ EXPORT_SYMBOL(__mutex_init); static __used noinline void __sched __mutex_lock_slowpath(atomic_t *lock_count); -/*** +/** * mutex_lock - acquire the mutex * @lock: the mutex to be acquired * @@ -105,7 +96,7 @@ EXPORT_SYMBOL(mutex_lock); static __used noinline void __sched __mutex_unlock_slowpath(atomic_t *lock_count); -/*** +/** * mutex_unlock - release the mutex * @lock: the mutex to be released * @@ -364,8 +355,8 @@ __mutex_lock_killable_slowpath(atomic_t *lock_count); static noinline int __sched __mutex_lock_interruptible_slowpath(atomic_t *lock_count); -/*** - * mutex_lock_interruptible - acquire the mutex, interruptable +/** + * mutex_lock_interruptible - acquire the mutex, interruptible * @lock: the mutex to be acquired * * Lock the mutex like mutex_lock(), and return 0 if the mutex has @@ -456,15 +447,15 @@ static inline int __mutex_trylock_slowpath(atomic_t *lock_count) return prev == 1; } -/*** - * mutex_trylock - try acquire the mutex, without waiting +/** + * mutex_trylock - try to acquire the mutex, without waiting * @lock: the mutex to be acquired * * Try to acquire the mutex atomically. Returns 1 if the mutex * has been acquired successfully, and 0 on contention. * * NOTE: this function follows the spin_trylock() convention, so - * it is negated to the down_trylock() return values! Be careful + * it is negated from the down_trylock() return values! Be careful * about this when converting semaphore users to mutexes. * * This function must not be used in interrupt context. The -- cgit v0.10.2 From d34c4aa43df20bbf2730a67f14c1cf6d133d99e6 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Thu, 2 Sep 2010 23:05:09 +0200 Subject: HID: add no-get quirk for eGalax touch controller Add no-get quirk for eGalax touch controller to avoid timeout at probe. Signed-off-by: Johan Hovold Signed-off-by: Jiri Kosina diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c index 16808f1..70da318 100644 --- a/drivers/hid/usbhid/hid-quirks.c +++ b/drivers/hid/usbhid/hid-quirks.c @@ -33,7 +33,7 @@ static const struct hid_blacklist { { USB_VENDOR_ID_AASHIMA, USB_DEVICE_ID_AASHIMA_PREDATOR, HID_QUIRK_BADPAD }, { USB_VENDOR_ID_ALPS, USB_DEVICE_ID_IBM_GAMEPAD, HID_QUIRK_BADPAD }, { USB_VENDOR_ID_CHIC, USB_DEVICE_ID_CHIC_GAMEPAD, HID_QUIRK_BADPAD }, - { USB_VENDOR_ID_DWAV, USB_DEVICE_ID_EGALAX_TOUCHCONTROLLER, HID_QUIRK_MULTI_INPUT }, + { USB_VENDOR_ID_DWAV, USB_DEVICE_ID_EGALAX_TOUCHCONTROLLER, HID_QUIRK_MULTI_INPUT | HID_QUIRK_NOGET }, { USB_VENDOR_ID_DWAV, USB_DEVICE_ID_DWAV_EGALAX_MULTITOUCH, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_MOJO, USB_DEVICE_ID_RETRO_ADAPTER, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_HAPP, USB_DEVICE_ID_UGCI_DRIVING, HID_QUIRK_BADPAD | HID_QUIRK_MULTI_INPUT }, -- cgit v0.10.2 From a2acad8298a42b7be684a32fafaf83332bba9c2b Mon Sep 17 00:00:00 2001 From: Clemens Ladisch Date: Fri, 3 Sep 2010 10:53:11 +0200 Subject: ALSA: usb-audio: fix detection of vendor-specific device protocol settings The Audio Class v2 support code in 2.6.35 added checks for the bInterfaceProtocol field. However, there are devices (usually those detected by vendor-specific quirks) that do not have one of the predefined values in this field, which made the driver reject them. To fix this regression, restore the old behaviour, i.e., assume that a device with an unknown bInterfaceProtocol field (other than UAC_VERSION_2) has more or less UAC-v1-compatible descriptors. [compile warning fixes by tiwai] Signed-off-by: Clemens Ladisch Cc: Daniel Mack Cc: Signed-off-by: Takashi Iwai diff --git a/sound/usb/card.c b/sound/usb/card.c index b443a33..32e4be8 100644 --- a/sound/usb/card.c +++ b/sound/usb/card.c @@ -216,6 +216,11 @@ static int snd_usb_create_streams(struct snd_usb_audio *chip, int ctrlif) } switch (protocol) { + default: + snd_printdd(KERN_WARNING "unknown interface protocol %#02x, assuming v1\n", + protocol); + /* fall through */ + case UAC_VERSION_1: { struct uac1_ac_header_descriptor *h1 = control_header; @@ -253,10 +258,6 @@ static int snd_usb_create_streams(struct snd_usb_audio *chip, int ctrlif) break; } - - default: - snd_printk(KERN_ERR "unknown protocol version 0x%02x\n", protocol); - return -EINVAL; } return 0; diff --git a/sound/usb/clock.c b/sound/usb/clock.c index b853f8d..7754a10 100644 --- a/sound/usb/clock.c +++ b/sound/usb/clock.c @@ -295,12 +295,11 @@ int snd_usb_init_sample_rate(struct snd_usb_audio *chip, int iface, switch (altsd->bInterfaceProtocol) { case UAC_VERSION_1: + default: return set_sample_rate_v1(chip, iface, alts, fmt, rate); case UAC_VERSION_2: return set_sample_rate_v2(chip, iface, alts, fmt, rate); } - - return -EINVAL; } diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index 1a701f1..ef0a07e 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -275,6 +275,12 @@ int snd_usb_parse_audio_endpoints(struct snd_usb_audio *chip, int iface_no) /* get audio formats */ switch (protocol) { + default: + snd_printdd(KERN_WARNING "%d:%u:%d: unknown interface protocol %#02x, assuming v1\n", + dev->devnum, iface_no, altno, protocol); + protocol = UAC_VERSION_1; + /* fall through */ + case UAC_VERSION_1: { struct uac1_as_header_descriptor *as = snd_usb_find_csint_desc(alts->extra, alts->extralen, NULL, UAC_AS_GENERAL); @@ -336,11 +342,6 @@ int snd_usb_parse_audio_endpoints(struct snd_usb_audio *chip, int iface_no) dev->devnum, iface_no, altno, as->bTerminalLink); continue; } - - default: - snd_printk(KERN_ERR "%d:%u:%d : unknown interface protocol %04x\n", - dev->devnum, iface_no, altno, protocol); - continue; } /* get format type */ diff --git a/sound/usb/format.c b/sound/usb/format.c index 3a13754..6914821 100644 --- a/sound/usb/format.c +++ b/sound/usb/format.c @@ -49,7 +49,8 @@ static u64 parse_audio_format_i_type(struct snd_usb_audio *chip, u64 pcm_formats; switch (protocol) { - case UAC_VERSION_1: { + case UAC_VERSION_1: + default: { struct uac_format_type_i_discrete_descriptor *fmt = _fmt; sample_width = fmt->bBitResolution; sample_bytes = fmt->bSubframeSize; @@ -64,9 +65,6 @@ static u64 parse_audio_format_i_type(struct snd_usb_audio *chip, format <<= 1; break; } - - default: - return -EINVAL; } pcm_formats = 0; @@ -384,6 +382,10 @@ static int parse_audio_format_i(struct snd_usb_audio *chip, * audio class v2 uses class specific EP0 range requests for that. */ switch (protocol) { + default: + snd_printdd(KERN_WARNING "%d:%u:%d : invalid protocol version %d, assuming v1\n", + chip->dev->devnum, fp->iface, fp->altsetting, protocol); + /* fall through */ case UAC_VERSION_1: fp->channels = fmt->bNrChannels; ret = parse_audio_format_rates_v1(chip, fp, (unsigned char *) fmt, 7); @@ -392,10 +394,6 @@ static int parse_audio_format_i(struct snd_usb_audio *chip, /* fp->channels is already set in this case */ ret = parse_audio_format_rates_v2(chip, fp); break; - default: - snd_printk(KERN_ERR "%d:%u:%d : invalid protocol version %d\n", - chip->dev->devnum, fp->iface, fp->altsetting, protocol); - return -EINVAL; } if (fp->channels < 1) { @@ -438,6 +436,10 @@ static int parse_audio_format_ii(struct snd_usb_audio *chip, fp->channels = 1; switch (protocol) { + default: + snd_printdd(KERN_WARNING "%d:%u:%d : invalid protocol version %d, assuming v1\n", + chip->dev->devnum, fp->iface, fp->altsetting, protocol); + /* fall through */ case UAC_VERSION_1: { struct uac_format_type_ii_discrete_descriptor *fmt = _fmt; brate = le16_to_cpu(fmt->wMaxBitRate); @@ -456,10 +458,6 @@ static int parse_audio_format_ii(struct snd_usb_audio *chip, ret = parse_audio_format_rates_v2(chip, fp); break; } - default: - snd_printk(KERN_ERR "%d:%u:%d : invalid protocol version %d\n", - chip->dev->devnum, fp->iface, fp->altsetting, protocol); - return -EINVAL; } return ret; diff --git a/sound/usb/mixer.c b/sound/usb/mixer.c index c166db0..3ed3901 100644 --- a/sound/usb/mixer.c +++ b/sound/usb/mixer.c @@ -2175,7 +2175,15 @@ int snd_usb_create_mixer(struct snd_usb_audio *chip, int ctrlif, } host_iface = &usb_ifnum_to_if(chip->dev, ctrlif)->altsetting[0]; - mixer->protocol = get_iface_desc(host_iface)->bInterfaceProtocol; + switch (get_iface_desc(host_iface)->bInterfaceProtocol) { + case UAC_VERSION_1: + default: + mixer->protocol = UAC_VERSION_1; + break; + case UAC_VERSION_2: + mixer->protocol = UAC_VERSION_2; + break; + } if ((err = snd_usb_mixer_controls(mixer)) < 0 || (err = snd_usb_mixer_status_create(mixer)) < 0) diff --git a/sound/usb/pcm.c b/sound/usb/pcm.c index 3634ced..3b5135c 100644 --- a/sound/usb/pcm.c +++ b/sound/usb/pcm.c @@ -173,13 +173,12 @@ int snd_usb_init_pitch(struct snd_usb_audio *chip, int iface, switch (altsd->bInterfaceProtocol) { case UAC_VERSION_1: + default: return init_pitch_v1(chip, iface, alts, fmt); case UAC_VERSION_2: return init_pitch_v2(chip, iface, alts, fmt); } - - return -EINVAL; } /* -- cgit v0.10.2 From d06563cb860ab594889010889a7111c9e25d1051 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 4 Sep 2010 23:10:48 +0800 Subject: regulator: 88pm8607 - fix value range checking for accessing info->vol_table In choose_voltage(), we use i as array index of info->vol_table. The valid value range for i should be 0 .. ARRAY_SIZE(info->vol_table) - 1. Take LDO1 as example, ARRAY_SIZE(LDO1_table) is 4, vol_nbits of LDO1 is 2. for (i = 0; i < (2 << info->vol_nbits); i++) is equivalent to for (i = 0; i < 8; i++) which is wrong. The same value range checking also applies for index in pm8607_list_voltage(). Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/88pm8607.c b/drivers/regulator/88pm8607.c index 7d149a8..2ce2eb7 100644 --- a/drivers/regulator/88pm8607.c +++ b/drivers/regulator/88pm8607.c @@ -215,7 +215,7 @@ static int pm8607_list_voltage(struct regulator_dev *rdev, unsigned index) struct pm8607_regulator_info *info = rdev_get_drvdata(rdev); int ret = -EINVAL; - if (info->vol_table && (index < (2 << info->vol_nbits))) { + if (info->vol_table && (index < (1 << info->vol_nbits))) { ret = info->vol_table[index]; if (info->slope_double) ret <<= 1; @@ -233,7 +233,7 @@ static int choose_voltage(struct regulator_dev *rdev, int min_uV, int max_uV) max_uV = max_uV >> 1; } if (info->vol_table) { - for (i = 0; i < (2 << info->vol_nbits); i++) { + for (i = 0; i < (1 << info->vol_nbits); i++) { if (!info->vol_table[i]) break; if ((min_uV <= info->vol_table[i]) -- cgit v0.10.2 From 49990e6efe576b8707584398f93198b5aa182ab7 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 4 Sep 2010 23:06:41 +0800 Subject: regulator: ab8500 - fix off-by-one value range checking for selector selector is used as array index of info->supported_voltages Thus the valid value range should be 0 .. info->voltages_len -1 Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c index 3d09580..28c7ae6 100644 --- a/drivers/regulator/ab8500.c +++ b/drivers/regulator/ab8500.c @@ -157,7 +157,7 @@ static int ab8500_list_voltage(struct regulator_dev *rdev, unsigned selector) if (info->fixed_uV) return info->fixed_uV; - if (selector > info->voltages_len) + if (selector >= info->voltages_len) return -EINVAL; return info->supported_voltages[selector]; -- cgit v0.10.2 From cc1a8e523333bc3d990102140f4d12908fa83caf Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Sat, 4 Sep 2010 22:56:43 +0200 Subject: x86: Fix the address space annotations of iomap_atomic_prot_pfn() This patch fixes the sparse warnings when the return pointer of iomap_atomic_prot_pfn() is used as an argument of iowrite32() and friends. Signed-off-by: Francisco Jerez LKML-Reference: <1283633804-11749-1-git-send-email-currojerez@riseup.net> Cc: Andrew Morton Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/iomap.h b/arch/x86/include/asm/iomap.h index f35eb45..c4191b3 100644 --- a/arch/x86/include/asm/iomap.h +++ b/arch/x86/include/asm/iomap.h @@ -26,11 +26,11 @@ #include #include -void * +void __iomem * iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot); void -iounmap_atomic(void *kvaddr, enum km_type type); +iounmap_atomic(void __iomem *kvaddr, enum km_type type); int iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot); diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 84e236c..72fc70c 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -74,7 +74,7 @@ void *kmap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) /* * Map 'pfn' using fixed map 'type' and protections 'prot' */ -void * +void __iomem * iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) { /* @@ -86,12 +86,12 @@ iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) if (!pat_enabled && pgprot_val(prot) == pgprot_val(PAGE_KERNEL_WC)) prot = PAGE_KERNEL_UC_MINUS; - return kmap_atomic_prot_pfn(pfn, type, prot); + return (void __force __iomem *) kmap_atomic_prot_pfn(pfn, type, prot); } EXPORT_SYMBOL_GPL(iomap_atomic_prot_pfn); void -iounmap_atomic(void *kvaddr, enum km_type type) +iounmap_atomic(void __iomem *kvaddr, enum km_type type) { unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK; enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id(); -- cgit v0.10.2 From 29bc17ecb856ffb2b47c7009a71971c6f9334205 Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Sat, 4 Sep 2010 22:56:44 +0200 Subject: io-mapping: Fix the address space annotations Fixes a bunch of sparse warnings in io-mapping.h because of the inconsistent __iomem usage. Signed-off-by: Francisco Jerez LKML-Reference: <1283633804-11749-2-git-send-email-currojerez@riseup.net> Cc: Andrew Morton Signed-off-by: Ingo Molnar diff --git a/include/linux/io-mapping.h b/include/linux/io-mapping.h index 0a6b3d5..7fb5927 100644 --- a/include/linux/io-mapping.h +++ b/include/linux/io-mapping.h @@ -79,7 +79,7 @@ io_mapping_free(struct io_mapping *mapping) } /* Atomic map/unmap */ -static inline void * +static inline void __iomem * io_mapping_map_atomic_wc(struct io_mapping *mapping, unsigned long offset, int slot) @@ -94,12 +94,12 @@ io_mapping_map_atomic_wc(struct io_mapping *mapping, } static inline void -io_mapping_unmap_atomic(void *vaddr, int slot) +io_mapping_unmap_atomic(void __iomem *vaddr, int slot) { iounmap_atomic(vaddr, slot); } -static inline void * +static inline void __iomem * io_mapping_map_wc(struct io_mapping *mapping, unsigned long offset) { resource_size_t phys_addr; @@ -111,7 +111,7 @@ io_mapping_map_wc(struct io_mapping *mapping, unsigned long offset) } static inline void -io_mapping_unmap(void *vaddr) +io_mapping_unmap(void __iomem *vaddr) { iounmap(vaddr); } @@ -125,38 +125,38 @@ struct io_mapping; static inline struct io_mapping * io_mapping_create_wc(resource_size_t base, unsigned long size) { - return (struct io_mapping *) ioremap_wc(base, size); + return (struct io_mapping __force *) ioremap_wc(base, size); } static inline void io_mapping_free(struct io_mapping *mapping) { - iounmap(mapping); + iounmap((void __force __iomem *) mapping); } /* Atomic map/unmap */ -static inline void * +static inline void __iomem * io_mapping_map_atomic_wc(struct io_mapping *mapping, unsigned long offset, int slot) { - return ((char *) mapping) + offset; + return ((char __force __iomem *) mapping) + offset; } static inline void -io_mapping_unmap_atomic(void *vaddr, int slot) +io_mapping_unmap_atomic(void __iomem *vaddr, int slot) { } /* Non-atomic map/unmap */ -static inline void * +static inline void __iomem * io_mapping_map_wc(struct io_mapping *mapping, unsigned long offset) { - return ((char *) mapping) + offset; + return ((char __force __iomem *) mapping) + offset; } static inline void -io_mapping_unmap(void *vaddr) +io_mapping_unmap(void __iomem *vaddr) { } -- cgit v0.10.2 From 1389298f7d14f179c2fa4b4b343079bd56082c70 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Fri, 27 Aug 2010 11:20:07 +0200 Subject: x86, mcheck: Avoid duplicate sysfs links/files for thresholding banks kobject_add_internal failed for threshold_bank2 with -EEXIST, don't try to register things with the same name in the same directory: Pid: 1, comm: swapper Tainted: G W 2.6.31 #1 Call Trace: [] ? kobject_add_internal+0x156/0x180 [] ? kobject_add+0x66/0x6b [] ? kobject_init+0x42/0x82 [] ? kobject_create_and_add+0x34/0x63 [] ? threshold_create_bank+0x14f/0x259 [] ? mce_create_device+0x8d/0x1b8 [] ? threshold_init_device+0x3f/0x80 [] ? threshold_init_device+0x0/0x80 [] ? do_one_initcall+0x4f/0x143 [] ? kernel_init+0x14c/0x1a2 [] ? child_rip+0xa/0x20 [] ? kernel_init+0x0/0x1a2 [] ? child_rip+0x0/0x20 kobject_create_and_add: kobject_add error: -17 (Probably the for_each_cpu loop should be entirely removed.) Signed-off-by: Andreas Herrmann LKML-Reference: <20100827092006.GB5348@loge.amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index 224392d..5e97529 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c @@ -530,7 +530,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) err = -ENOMEM; goto out; } - if (!alloc_cpumask_var(&b->cpus, GFP_KERNEL)) { + if (!zalloc_cpumask_var(&b->cpus, GFP_KERNEL)) { kfree(b); err = -ENOMEM; goto out; @@ -543,7 +543,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) #ifndef CONFIG_SMP cpumask_setall(b->cpus); #else - cpumask_copy(b->cpus, c->llc_shared_map); + cpumask_set_cpu(cpu, b->cpus); #endif per_cpu(threshold_banks, cpu)[bank] = b; -- cgit v0.10.2 From b3bd3de66f60df4c9a2076e2886a622458929056 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 10 Aug 2010 14:17:51 -0700 Subject: gcc-4.6: kernel/*: Fix unused but set warnings No real bugs I believe, just some dead code. Signed-off-by: Andi Kleen Cc: Peter Zijlstra Cc: andi@firstfloor.org Signed-off-by: Andrew Morton Signed-off-by: Ingo Molnar diff --git a/kernel/debug/kdb/kdb_bp.c b/kernel/debug/kdb/kdb_bp.c index 75bd9b3..20059ef 100644 --- a/kernel/debug/kdb/kdb_bp.c +++ b/kernel/debug/kdb/kdb_bp.c @@ -274,7 +274,6 @@ static int kdb_bp(int argc, const char **argv) int i, bpno; kdb_bp_t *bp, *bp_check; int diag; - int free; char *symname = NULL; long offset = 0ul; int nextarg; @@ -305,7 +304,6 @@ static int kdb_bp(int argc, const char **argv) /* * Find an empty bp structure to allocate */ - free = KDB_MAXBPT; for (bpno = 0, bp = kdb_breakpoints; bpno < KDB_MAXBPT; bpno++, bp++) { if (bp->bp_free) break; diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index ce66917..1decafb 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -1091,11 +1091,10 @@ EXPORT_SYMBOL_GPL(hrtimer_cancel); */ ktime_t hrtimer_get_remaining(const struct hrtimer *timer) { - struct hrtimer_clock_base *base; unsigned long flags; ktime_t rem; - base = lock_hrtimer_base(timer, &flags); + lock_hrtimer_base(timer, &flags); rem = hrtimer_expires_remaining(timer); unlock_hrtimer_base(timer, &flags); diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index ab661eb..134f7ed 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1313,7 +1313,7 @@ static struct sched_group * find_idlest_group(struct sched_domain *sd, struct task_struct *p, int this_cpu, int load_idx) { - struct sched_group *idlest = NULL, *this = NULL, *group = sd->groups; + struct sched_group *idlest = NULL, *group = sd->groups; unsigned long min_load = ULONG_MAX, this_load = 0; int imbalance = 100 + (sd->imbalance_pct-100)/2; @@ -1348,7 +1348,6 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p, if (local_group) { this_load = avg_load; - this = group; } else if (avg_load < min_load) { min_load = avg_load; idlest = group; diff --git a/kernel/sysctl.c b/kernel/sysctl.c index ca38e8e..f88552c 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1713,10 +1713,7 @@ static __init int sysctl_init(void) { sysctl_set_parent(NULL, root_table); #ifdef CONFIG_SYSCTL_SYSCALL_CHECK - { - int err; - err = sysctl_check_table(current->nsproxy, root_table); - } + sysctl_check_table(current->nsproxy, root_table); #endif return 0; } diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 19cccc3..492197e 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -2985,13 +2985,11 @@ static void rb_advance_reader(struct ring_buffer_per_cpu *cpu_buffer) static void rb_advance_iter(struct ring_buffer_iter *iter) { - struct ring_buffer *buffer; struct ring_buffer_per_cpu *cpu_buffer; struct ring_buffer_event *event; unsigned length; cpu_buffer = iter->cpu_buffer; - buffer = cpu_buffer->buffer; /* * Check if we are at the end of the buffer. -- cgit v0.10.2 From feafb7b1714cf599a6d0fed45801ab3f66046cbd Mon Sep 17 00:00:00 2001 From: Arun Easi Date: Fri, 3 Sep 2010 14:57:00 -0700 Subject: [SCSI] qla2xxx: Fix vport delete issues Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_attr.c b/drivers/scsi/qla2xxx/qla_attr.c index 420238c..114bc5a 100644 --- a/drivers/scsi/qla2xxx/qla_attr.c +++ b/drivers/scsi/qla2xxx/qla_attr.c @@ -1838,26 +1838,33 @@ qla24xx_vport_delete(struct fc_vport *fc_vport) qla24xx_disable_vp(vha); + vha->flags.delete_progress = 1; + fc_remove_host(vha->host); scsi_remove_host(vha->host); - qla2x00_free_fcports(vha); + if (vha->timer_active) { + qla2x00_vp_stop_timer(vha); + DEBUG15(printk(KERN_INFO "scsi(%ld): timer for the vport[%d]" + " = %p has stopped\n", vha->host_no, vha->vp_idx, vha)); + } qla24xx_deallocate_vp_id(vha); + /* No pending activities shall be there on the vha now */ + DEBUG(msleep(random32()%10)); /* Just to see if something falls on + * the net we have placed below */ + + BUG_ON(atomic_read(&vha->vref_count)); + + qla2x00_free_fcports(vha); + mutex_lock(&ha->vport_lock); ha->cur_vport_count--; clear_bit(vha->vp_idx, ha->vp_idx_map); mutex_unlock(&ha->vport_lock); - if (vha->timer_active) { - qla2x00_vp_stop_timer(vha); - DEBUG15(printk ("scsi(%ld): timer for the vport[%d] = %p " - "has stopped\n", - vha->host_no, vha->vp_idx, vha)); - } - if (vha->req->id && !ha->flags.cpu_affinity_enabled) { if (qla25xx_delete_req_que(vha, vha->req) != QLA_SUCCESS) qla_printk(KERN_WARNING, ha, diff --git a/drivers/scsi/qla2xxx/qla_dbg.h b/drivers/scsi/qla2xxx/qla_dbg.h index 6cfc28a..b74e6b5 100644 --- a/drivers/scsi/qla2xxx/qla_dbg.h +++ b/drivers/scsi/qla2xxx/qla_dbg.h @@ -29,8 +29,6 @@ /* #define QL_DEBUG_LEVEL_17 */ /* Output EEH trace messages */ /* #define QL_DEBUG_LEVEL_18 */ /* Output T10 CRC trace messages */ -/* #define QL_PRINTK_BUF */ /* Captures printk to buffer */ - /* * Macros use for debugging the driver. */ diff --git a/drivers/scsi/qla2xxx/qla_def.h b/drivers/scsi/qla2xxx/qla_def.h index 3a432ea..d2a4e15 100644 --- a/drivers/scsi/qla2xxx/qla_def.h +++ b/drivers/scsi/qla2xxx/qla_def.h @@ -2641,6 +2641,7 @@ struct qla_hw_data { #define MBX_UPDATE_FLASH_ACTIVE 3 struct mutex vport_lock; /* Virtual port synchronization */ + spinlock_t vport_slock; /* order is hardware_lock, then vport_slock */ struct completion mbx_cmd_comp; /* Serialize mbx access */ struct completion mbx_intr_comp; /* Used for completion notification */ struct completion dcbx_comp; /* For set port config notification */ @@ -2828,6 +2829,7 @@ typedef struct scsi_qla_host { uint32_t management_server_logged_in :1; uint32_t process_response_queue :1; uint32_t difdix_supported:1; + uint32_t delete_progress:1; } flags; atomic_t loop_state; @@ -2922,6 +2924,8 @@ typedef struct scsi_qla_host { struct req_que *req; int fw_heartbeat_counter; int seconds_since_last_heartbeat; + + atomic_t vref_count; } scsi_qla_host_t; /* @@ -2932,6 +2936,22 @@ typedef struct scsi_qla_host { test_bit(LOOP_RESYNC_NEEDED, &ha->dpc_flags) || \ atomic_read(&ha->loop_state) == LOOP_DOWN) +#define QLA_VHA_MARK_BUSY(__vha, __bail) do { \ + atomic_inc(&__vha->vref_count); \ + mb(); \ + if (__vha->flags.delete_progress) { \ + atomic_dec(&__vha->vref_count); \ + __bail = 1; \ + } else { \ + __bail = 0; \ + } \ +} while (0) + +#define QLA_VHA_MARK_NOT_BUSY(__vha) do { \ + atomic_dec(&__vha->vref_count); \ +} while (0) + + #define qla_printk(level, ha, format, arg...) \ dev_printk(level , &((ha)->pdev->dev) , format , ## arg) diff --git a/drivers/scsi/qla2xxx/qla_init.c b/drivers/scsi/qla2xxx/qla_init.c index d863ed2..9c383ba 100644 --- a/drivers/scsi/qla2xxx/qla_init.c +++ b/drivers/scsi/qla2xxx/qla_init.c @@ -69,21 +69,29 @@ qla2x00_ctx_sp_free(srb_t *sp) { struct srb_ctx *ctx = sp->ctx; struct srb_iocb *iocb = ctx->u.iocb_cmd; + struct scsi_qla_host *vha = sp->fcport->vha; del_timer_sync(&iocb->timer); kfree(iocb); kfree(ctx); mempool_free(sp, sp->fcport->vha->hw->srb_mempool); + + QLA_VHA_MARK_NOT_BUSY(vha); } inline srb_t * qla2x00_get_ctx_sp(scsi_qla_host_t *vha, fc_port_t *fcport, size_t size, unsigned long tmo) { - srb_t *sp; + srb_t *sp = NULL; struct qla_hw_data *ha = vha->hw; struct srb_ctx *ctx; struct srb_iocb *iocb; + uint8_t bail; + + QLA_VHA_MARK_BUSY(vha, bail); + if (bail) + return NULL; sp = mempool_alloc(ha->srb_mempool, GFP_KERNEL); if (!sp) @@ -116,6 +124,8 @@ qla2x00_get_ctx_sp(scsi_qla_host_t *vha, fc_port_t *fcport, size_t size, iocb->timer.function = qla2x00_ctx_sp_timeout; add_timer(&iocb->timer); done: + if (!sp) + QLA_VHA_MARK_NOT_BUSY(vha); return sp; } @@ -1777,11 +1787,15 @@ qla2x00_init_rings(scsi_qla_host_t *vha) qla2x00_init_response_q_entries(rsp); } + spin_lock_irqsave(&ha->vport_slock, flags); /* Clear RSCN queue. */ list_for_each_entry(vp, &ha->vp_list, list) { vp->rscn_in_ptr = 0; vp->rscn_out_ptr = 0; } + + spin_unlock_irqrestore(&ha->vport_slock, flags); + ha->isp_ops->config_rings(vha); spin_unlock_irqrestore(&ha->hardware_lock, flags); @@ -3218,12 +3232,17 @@ qla2x00_find_all_fabric_devs(scsi_qla_host_t *vha, /* Bypass virtual ports of the same host. */ found = 0; if (ha->num_vhosts) { + unsigned long flags; + + spin_lock_irqsave(&ha->vport_slock, flags); list_for_each_entry_safe(vp, tvp, &ha->vp_list, list) { if (new_fcport->d_id.b24 == vp->d_id.b24) { found = 1; break; } } + spin_unlock_irqrestore(&ha->vport_slock, flags); + if (found) continue; } @@ -3343,6 +3362,7 @@ qla2x00_find_new_loop_id(scsi_qla_host_t *vha, fc_port_t *dev) struct qla_hw_data *ha = vha->hw; struct scsi_qla_host *vp; struct scsi_qla_host *tvp; + unsigned long flags = 0; rval = QLA_SUCCESS; @@ -3367,6 +3387,8 @@ qla2x00_find_new_loop_id(scsi_qla_host_t *vha, fc_port_t *dev) /* Check for loop ID being already in use. */ found = 0; fcport = NULL; + + spin_lock_irqsave(&ha->vport_slock, flags); list_for_each_entry_safe(vp, tvp, &ha->vp_list, list) { list_for_each_entry(fcport, &vp->vp_fcports, list) { if (fcport->loop_id == dev->loop_id && @@ -3379,6 +3401,7 @@ qla2x00_find_new_loop_id(scsi_qla_host_t *vha, fc_port_t *dev) if (found) break; } + spin_unlock_irqrestore(&ha->vport_slock, flags); /* If not in use then it is free to use. */ if (!found) { @@ -3791,14 +3814,27 @@ void qla2x00_update_fcports(scsi_qla_host_t *base_vha) { fc_port_t *fcport; - struct scsi_qla_host *tvp, *vha; + struct scsi_qla_host *vha; + struct qla_hw_data *ha = base_vha->hw; + unsigned long flags; + spin_lock_irqsave(&ha->vport_slock, flags); /* Go with deferred removal of rport references. */ - list_for_each_entry_safe(vha, tvp, &base_vha->hw->vp_list, list) - list_for_each_entry(fcport, &vha->vp_fcports, list) + list_for_each_entry(vha, &base_vha->hw->vp_list, list) { + atomic_inc(&vha->vref_count); + list_for_each_entry(fcport, &vha->vp_fcports, list) { if (fcport && fcport->drport && - atomic_read(&fcport->state) != FCS_UNCONFIGURED) + atomic_read(&fcport->state) != FCS_UNCONFIGURED) { + spin_unlock_irqrestore(&ha->vport_slock, flags); + qla2x00_rport_del(fcport); + + spin_lock_irqsave(&ha->vport_slock, flags); + } + } + atomic_dec(&vha->vref_count); + } + spin_unlock_irqrestore(&ha->vport_slock, flags); } void @@ -3806,7 +3842,7 @@ qla2x00_abort_isp_cleanup(scsi_qla_host_t *vha) { struct qla_hw_data *ha = vha->hw; struct scsi_qla_host *vp, *base_vha = pci_get_drvdata(ha->pdev); - struct scsi_qla_host *tvp; + unsigned long flags; vha->flags.online = 0; ha->flags.chip_reset_done = 0; @@ -3824,8 +3860,18 @@ qla2x00_abort_isp_cleanup(scsi_qla_host_t *vha) if (atomic_read(&vha->loop_state) != LOOP_DOWN) { atomic_set(&vha->loop_state, LOOP_DOWN); qla2x00_mark_all_devices_lost(vha, 0); - list_for_each_entry_safe(vp, tvp, &base_vha->hw->vp_list, list) + + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vp, &base_vha->hw->vp_list, list) { + atomic_inc(&vp->vref_count); + spin_unlock_irqrestore(&ha->vport_slock, flags); + qla2x00_mark_all_devices_lost(vp, 0); + + spin_lock_irqsave(&ha->vport_slock, flags); + atomic_dec(&vp->vref_count); + } + spin_unlock_irqrestore(&ha->vport_slock, flags); } else { if (!atomic_read(&vha->loop_down_timer)) atomic_set(&vha->loop_down_timer, @@ -3862,8 +3908,8 @@ qla2x00_abort_isp(scsi_qla_host_t *vha) uint8_t status = 0; struct qla_hw_data *ha = vha->hw; struct scsi_qla_host *vp; - struct scsi_qla_host *tvp; struct req_que *req = ha->req_q_map[0]; + unsigned long flags; if (vha->flags.online) { qla2x00_abort_isp_cleanup(vha); @@ -3970,10 +4016,21 @@ qla2x00_abort_isp(scsi_qla_host_t *vha) DEBUG(printk(KERN_INFO "qla2x00_abort_isp(%ld): succeeded.\n", vha->host_no)); - list_for_each_entry_safe(vp, tvp, &ha->vp_list, list) { - if (vp->vp_idx) + + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vp, &ha->vp_list, list) { + if (vp->vp_idx) { + atomic_inc(&vp->vref_count); + spin_unlock_irqrestore(&ha->vport_slock, flags); + qla2x00_vp_abort_isp(vp); + + spin_lock_irqsave(&ha->vport_slock, flags); + atomic_dec(&vp->vref_count); + } } + spin_unlock_irqrestore(&ha->vport_slock, flags); + } else { qla_printk(KERN_INFO, ha, "qla2x00_abort_isp: **** FAILED ****\n"); @@ -5185,7 +5242,7 @@ qla82xx_restart_isp(scsi_qla_host_t *vha) struct req_que *req = ha->req_q_map[0]; struct rsp_que *rsp = ha->rsp_q_map[0]; struct scsi_qla_host *vp; - struct scsi_qla_host *tvp; + unsigned long flags; status = qla2x00_init_rings(vha); if (!status) { @@ -5272,10 +5329,21 @@ qla82xx_restart_isp(scsi_qla_host_t *vha) DEBUG(printk(KERN_INFO "qla82xx_restart_isp(%ld): succeeded.\n", vha->host_no)); - list_for_each_entry_safe(vp, tvp, &ha->vp_list, list) { - if (vp->vp_idx) + + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vp, &ha->vp_list, list) { + if (vp->vp_idx) { + atomic_inc(&vp->vref_count); + spin_unlock_irqrestore(&ha->vport_slock, flags); + qla2x00_vp_abort_isp(vp); + + spin_lock_irqsave(&ha->vport_slock, flags); + atomic_dec(&vp->vref_count); + } } + spin_unlock_irqrestore(&ha->vport_slock, flags); + } else { qla_printk(KERN_INFO, ha, "qla82xx_restart_isp: **** FAILED ****\n"); diff --git a/drivers/scsi/qla2xxx/qla_mbx.c b/drivers/scsi/qla2xxx/qla_mbx.c index 6009b0c..a595ec8 100644 --- a/drivers/scsi/qla2xxx/qla_mbx.c +++ b/drivers/scsi/qla2xxx/qla_mbx.c @@ -2913,7 +2913,7 @@ qla24xx_report_id_acquisition(scsi_qla_host_t *vha, uint16_t stat = le16_to_cpu(rptid_entry->vp_idx); struct qla_hw_data *ha = vha->hw; scsi_qla_host_t *vp; - scsi_qla_host_t *tvp; + unsigned long flags; if (rptid_entry->entry_status != 0) return; @@ -2945,9 +2945,12 @@ qla24xx_report_id_acquisition(scsi_qla_host_t *vha, return; } - list_for_each_entry_safe(vp, tvp, &ha->vp_list, list) + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vp, &ha->vp_list, list) if (vp_idx == vp->vp_idx) break; + spin_unlock_irqrestore(&ha->vport_slock, flags); + if (!vp) return; diff --git a/drivers/scsi/qla2xxx/qla_mid.c b/drivers/scsi/qla2xxx/qla_mid.c index 987c5b0..bc1a745 100644 --- a/drivers/scsi/qla2xxx/qla_mid.c +++ b/drivers/scsi/qla2xxx/qla_mid.c @@ -30,6 +30,7 @@ qla24xx_allocate_vp_id(scsi_qla_host_t *vha) { uint32_t vp_id; struct qla_hw_data *ha = vha->hw; + unsigned long flags; /* Find an empty slot and assign an vp_id */ mutex_lock(&ha->vport_lock); @@ -44,7 +45,11 @@ qla24xx_allocate_vp_id(scsi_qla_host_t *vha) set_bit(vp_id, ha->vp_idx_map); ha->num_vhosts++; vha->vp_idx = vp_id; + + spin_lock_irqsave(&ha->vport_slock, flags); list_add_tail(&vha->list, &ha->vp_list); + spin_unlock_irqrestore(&ha->vport_slock, flags); + mutex_unlock(&ha->vport_lock); return vp_id; } @@ -54,12 +59,31 @@ qla24xx_deallocate_vp_id(scsi_qla_host_t *vha) { uint16_t vp_id; struct qla_hw_data *ha = vha->hw; + unsigned long flags = 0; mutex_lock(&ha->vport_lock); + /* + * Wait for all pending activities to finish before removing vport from + * the list. + * Lock needs to be held for safe removal from the list (it + * ensures no active vp_list traversal while the vport is removed + * from the queue) + */ + spin_lock_irqsave(&ha->vport_slock, flags); + while (atomic_read(&vha->vref_count)) { + spin_unlock_irqrestore(&ha->vport_slock, flags); + + msleep(500); + + spin_lock_irqsave(&ha->vport_slock, flags); + } + list_del(&vha->list); + spin_unlock_irqrestore(&ha->vport_slock, flags); + vp_id = vha->vp_idx; ha->num_vhosts--; clear_bit(vp_id, ha->vp_idx_map); - list_del(&vha->list); + mutex_unlock(&ha->vport_lock); } @@ -68,12 +92,17 @@ qla24xx_find_vhost_by_name(struct qla_hw_data *ha, uint8_t *port_name) { scsi_qla_host_t *vha; struct scsi_qla_host *tvha; + unsigned long flags; + spin_lock_irqsave(&ha->vport_slock, flags); /* Locate matching device in database. */ list_for_each_entry_safe(vha, tvha, &ha->vp_list, list) { - if (!memcmp(port_name, vha->port_name, WWN_SIZE)) + if (!memcmp(port_name, vha->port_name, WWN_SIZE)) { + spin_unlock_irqrestore(&ha->vport_slock, flags); return vha; + } } + spin_unlock_irqrestore(&ha->vport_slock, flags); return NULL; } @@ -93,6 +122,12 @@ qla24xx_find_vhost_by_name(struct qla_hw_data *ha, uint8_t *port_name) static void qla2x00_mark_vp_devices_dead(scsi_qla_host_t *vha) { + /* + * !!! NOTE !!! + * This function, if called in contexts other than vp create, disable + * or delete, please make sure this is synchronized with the + * delete thread. + */ fc_port_t *fcport; list_for_each_entry(fcport, &vha->vp_fcports, list) { @@ -194,12 +229,17 @@ qla24xx_configure_vp(scsi_qla_host_t *vha) void qla2x00_alert_all_vps(struct rsp_que *rsp, uint16_t *mb) { - scsi_qla_host_t *vha, *tvha; + scsi_qla_host_t *vha; struct qla_hw_data *ha = rsp->hw; int i = 0; + unsigned long flags; - list_for_each_entry_safe(vha, tvha, &ha->vp_list, list) { + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vha, &ha->vp_list, list) { if (vha->vp_idx) { + atomic_inc(&vha->vref_count); + spin_unlock_irqrestore(&ha->vport_slock, flags); + switch (mb[0]) { case MBA_LIP_OCCURRED: case MBA_LOOP_UP: @@ -215,9 +255,13 @@ qla2x00_alert_all_vps(struct rsp_que *rsp, uint16_t *mb) qla2x00_async_event(vha, rsp, mb); break; } + + spin_lock_irqsave(&ha->vport_slock, flags); + atomic_dec(&vha->vref_count); } i++; } + spin_unlock_irqrestore(&ha->vport_slock, flags); } int @@ -297,7 +341,7 @@ qla2x00_do_dpc_all_vps(scsi_qla_host_t *vha) int ret; struct qla_hw_data *ha = vha->hw; scsi_qla_host_t *vp; - struct scsi_qla_host *tvp; + unsigned long flags = 0; if (vha->vp_idx) return; @@ -309,10 +353,19 @@ qla2x00_do_dpc_all_vps(scsi_qla_host_t *vha) if (!(ha->current_topology & ISP_CFG_F)) return; - list_for_each_entry_safe(vp, tvp, &ha->vp_list, list) { - if (vp->vp_idx) + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vp, &ha->vp_list, list) { + if (vp->vp_idx) { + atomic_inc(&vp->vref_count); + spin_unlock_irqrestore(&ha->vport_slock, flags); + ret = qla2x00_do_dpc_vp(vp); + + spin_lock_irqsave(&ha->vport_slock, flags); + atomic_dec(&vp->vref_count); + } } + spin_unlock_irqrestore(&ha->vport_slock, flags); } int diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c index 8c80b49..1e4bff6 100644 --- a/drivers/scsi/qla2xxx/qla_os.c +++ b/drivers/scsi/qla2xxx/qla_os.c @@ -2341,16 +2341,28 @@ probe_out: static void qla2x00_remove_one(struct pci_dev *pdev) { - scsi_qla_host_t *base_vha, *vha, *temp; + scsi_qla_host_t *base_vha, *vha; struct qla_hw_data *ha; + unsigned long flags; base_vha = pci_get_drvdata(pdev); ha = base_vha->hw; - list_for_each_entry_safe(vha, temp, &ha->vp_list, list) { - if (vha && vha->fc_vport) + spin_lock_irqsave(&ha->vport_slock, flags); + list_for_each_entry(vha, &ha->vp_list, list) { + atomic_inc(&vha->vref_count); + + if (vha && vha->fc_vport) { + spin_unlock_irqrestore(&ha->vport_slock, flags); + fc_vport_terminate(vha->fc_vport); + + spin_lock_irqsave(&ha->vport_slock, flags); + } + + atomic_dec(&vha->vref_count); } + spin_unlock_irqrestore(&ha->vport_slock, flags); set_bit(UNLOADING, &base_vha->dpc_flags); @@ -2975,10 +2987,17 @@ static struct qla_work_evt * qla2x00_alloc_work(struct scsi_qla_host *vha, enum qla_work_type type) { struct qla_work_evt *e; + uint8_t bail; + + QLA_VHA_MARK_BUSY(vha, bail); + if (bail) + return NULL; e = kzalloc(sizeof(struct qla_work_evt), GFP_ATOMIC); - if (!e) + if (!e) { + QLA_VHA_MARK_NOT_BUSY(vha); return NULL; + } INIT_LIST_HEAD(&e->list); e->type = type; @@ -3135,6 +3154,9 @@ qla2x00_do_work(struct scsi_qla_host *vha) } if (e->flags & QLA_EVT_FLAG_FREE) kfree(e); + + /* For each work completed decrement vha ref count */ + QLA_VHA_MARK_NOT_BUSY(vha); } } -- cgit v0.10.2 From 970ee0c52a41cf27c1b5c346dd9475e9c236f3c5 Mon Sep 17 00:00:00 2001 From: Arun Easi Date: Fri, 3 Sep 2010 14:57:01 -0700 Subject: [SCSI] qla2xxx: make rport deletions explicit during vport removal Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_mid.c b/drivers/scsi/qla2xxx/qla_mid.c index bc1a745..2b69392 100644 --- a/drivers/scsi/qla2xxx/qla_mid.c +++ b/drivers/scsi/qla2xxx/qla_mid.c @@ -135,7 +135,6 @@ qla2x00_mark_vp_devices_dead(scsi_qla_host_t *vha) "loop_id=0x%04x :%x\n", vha->host_no, fcport->loop_id, fcport->vp_idx)); - atomic_set(&fcport->state, FCS_DEVICE_DEAD); qla2x00_mark_device_lost(vha, fcport, 0, 0); atomic_set(&fcport->state, FCS_UNCONFIGURED); } -- cgit v0.10.2 From efa786cc43a114d0bf2e4b95e856ea6911404d58 Mon Sep 17 00:00:00 2001 From: Lalit Chandivade Date: Fri, 3 Sep 2010 14:57:02 -0700 Subject: [SCSI] qla2xxx: Reset seconds_since_last_heartbeat correctly. The seconds_since_last_heartbeat should be checked for consecutive heartbeat checks. Currently it could happen that seconds_since_last_heartbeat gets set to max (2 seconds) for non-consecutive heartbeat checks. Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_nx.c b/drivers/scsi/qla2xxx/qla_nx.c index 915b77a..e54d2fa 100644 --- a/drivers/scsi/qla2xxx/qla_nx.c +++ b/drivers/scsi/qla2xxx/qla_nx.c @@ -3316,7 +3316,8 @@ qla82xx_check_fw_alive(scsi_qla_host_t *vha) complete(&ha->mbx_intr_comp); } } - } + } else + vha->seconds_since_last_heartbeat = 0; vha->fw_heartbeat_counter = fw_heartbeat_counter; } -- cgit v0.10.2 From 4142b1987f1f8ba90589642cb74566eaff3dc2e9 Mon Sep 17 00:00:00 2001 From: Giridhar Malavali Date: Fri, 3 Sep 2010 14:57:03 -0700 Subject: [SCSI] qla2xxx: Correctly set fw hung and complete only waiting mbx. The fw_hung flag should be set ir-respective of if there is a mbx command pending or not. Also the complete should be called if there is a mbx waiting. Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_nx.c b/drivers/scsi/qla2xxx/qla_nx.c index e54d2fa..ad290dc 100644 --- a/drivers/scsi/qla2xxx/qla_nx.c +++ b/drivers/scsi/qla2xxx/qla_nx.c @@ -3307,13 +3307,15 @@ qla82xx_check_fw_alive(scsi_qla_host_t *vha) set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags); } qla2xxx_wake_dpc(vha); + ha->flags.fw_hung = 1; if (ha->flags.mbox_busy) { - ha->flags.fw_hung = 1; ha->flags.mbox_int = 1; DEBUG2(qla_printk(KERN_ERR, ha, - "Due to fw hung, doing premature " - "completion of mbx command\n")); - complete(&ha->mbx_intr_comp); + "Due to fw hung, doing premature " + "completion of mbx command\n")); + if (test_bit(MBX_INTR_WAIT, + &ha->mbx_cmd_flags)) + complete(&ha->mbx_intr_comp); } } } else @@ -3419,13 +3421,15 @@ void qla82xx_watchdog(scsi_qla_host_t *vha) "%s(): Adapter reset needed!\n", __func__); set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags); qla2xxx_wake_dpc(vha); + ha->flags.fw_hung = 1; if (ha->flags.mbox_busy) { - ha->flags.fw_hung = 1; ha->flags.mbox_int = 1; DEBUG2(qla_printk(KERN_ERR, ha, - "Need reset, doing premature " - "completion of mbx command\n")); - complete(&ha->mbx_intr_comp); + "Need reset, doing premature " + "completion of mbx command\n")); + if (test_bit(MBX_INTR_WAIT, + &ha->mbx_cmd_flags)) + complete(&ha->mbx_intr_comp); } } else { qla82xx_check_fw_alive(vha); -- cgit v0.10.2 From 0374f55ed882a46cd4825dde16ca2392d4c367f6 Mon Sep 17 00:00:00 2001 From: Lalit Chandivade Date: Fri, 3 Sep 2010 14:57:04 -0700 Subject: [SCSI] qla2xxx: Cover UNDERRUN case where SCSI status is set. Currently, if target sets the SCSI Status (with Check condition) and there is no FCP residual bit set then driver does not check for dropped frame. This could lead to data corruption. Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c index 6982ba7..28f65be 100644 --- a/drivers/scsi/qla2xxx/qla_isr.c +++ b/drivers/scsi/qla2xxx/qla_isr.c @@ -1706,19 +1706,20 @@ qla2x00_status_entry(scsi_qla_host_t *vha, struct rsp_que *rsp, void *pkt) cp->result = DID_ERROR << 16; break; } - } else if (!lscsi_status) { + } else { DEBUG2(qla_printk(KERN_INFO, ha, "scsi(%ld:%d:%d) Dropped frame(s) detected (0x%x " "of 0x%x bytes).\n", vha->host_no, cp->device->id, cp->device->lun, resid, scsi_bufflen(cp))); - cp->result = DID_ERROR << 16; - break; + cp->result = DID_ERROR << 16 | lscsi_status; + goto check_scsi_status; } cp->result = DID_OK << 16 | lscsi_status; logit = 0; +check_scsi_status: /* * Check to see if SCSI Status is non zero. If so report SCSI * Status. -- cgit v0.10.2 From 1bd58b89e84b15283aaa3148fee4969abe19af8d Mon Sep 17 00:00:00 2001 From: Giridhar Malavali Date: Fri, 3 Sep 2010 14:57:05 -0700 Subject: [SCSI] qla2xxx: Check for empty slot in request queue before posting Command type 6 request. For ISP82xx, the check for empty slot in request queue before posting command type 6 request was missing. This could lead to request queue entry corruptions causing IO timeouts. Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_nx.c b/drivers/scsi/qla2xxx/qla_nx.c index ad290dc..0a71cc7 100644 --- a/drivers/scsi/qla2xxx/qla_nx.c +++ b/drivers/scsi/qla2xxx/qla_nx.c @@ -2672,6 +2672,19 @@ qla82xx_start_scsi(srb_t *sp) sufficient_dsds: req_cnt = 1; + if (req->cnt < (req_cnt + 2)) { + cnt = (uint16_t)RD_REG_DWORD_RELAXED( + ®->req_q_out[0]); + if (req->ring_index < cnt) + req->cnt = cnt - req->ring_index; + else + req->cnt = req->length - + (req->ring_index - cnt); + } + + if (req->cnt < (req_cnt + 2)) + goto queuing_error; + ctx = sp->ctx = mempool_alloc(ha->ctx_mempool, GFP_ATOMIC); if (!sp->ctx) { DEBUG(printk(KERN_INFO -- cgit v0.10.2 From 0fb576d8251c10f498ed4c6938aeeed8d0c93cfe Mon Sep 17 00:00:00 2001 From: Madhuranath Iyengar Date: Fri, 3 Sep 2010 14:57:06 -0700 Subject: [SCSI] qla2xxx: Update version number to 8.03.04-k0. Signed-off-by: Giridhar Malavali Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_version.h b/drivers/scsi/qla2xxx/qla_version.h index e75ccb9..8edbccb 100644 --- a/drivers/scsi/qla2xxx/qla_version.h +++ b/drivers/scsi/qla2xxx/qla_version.h @@ -7,9 +7,9 @@ /* * Driver version */ -#define QLA2XXX_VERSION "8.03.03-k0" +#define QLA2XXX_VERSION "8.03.04-k0" #define QLA_DRIVER_MAJOR_VER 8 #define QLA_DRIVER_MINOR_VER 3 -#define QLA_DRIVER_PATCH_VER 3 +#define QLA_DRIVER_PATCH_VER 4 #define QLA_DRIVER_BETA_VER 0 -- cgit v0.10.2 From e260999c66768c2fccd9da8c3918b4e0e5121b3a Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Mon, 6 Sep 2010 16:48:13 +0800 Subject: regulator: wm831x-ldo - fix the logic to set REGULATOR_MODE_IDLE and REGULATOR_MODE_STANDBY modes Problem description in current implementation: When setting REGULATOR_MODE_IDLE mode, current implementation set WM831X_LDO1_LP_MODE bit of ctrl_reg (which is wrong, it should clear the bit). But due to a missing break statement for case REGULATOR_MODE_IDLE, the code fall through to case REGULATOR_MODE_STANDBY and then clear WM831X_LDO1_LP_MODE bit. So it still looks OK when checking the status by wm831x_gp_ldo_get_mode(). When setting REGULATOR_MODE_STANDBY mode, it just does not work. wm831x_gp_ldo_get_mode() will still return REGULATOR_MODE_IDLE because the accordingly WM831X_LDO1_LP_MODE bit is clear. Correct behavior should be: Clear WM831X_LDO1_LP_MODE bit of ctrl_reg for REGULATOR_MODE_IDLE mode. Set WM831X_LDO1_LP_MODE bit of ctrl_reg for REGULATOR_MODE_STANDBY mode. Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/wm831x-ldo.c b/drivers/regulator/wm831x-ldo.c index e686cdb..9edf8f6 100644 --- a/drivers/regulator/wm831x-ldo.c +++ b/drivers/regulator/wm831x-ldo.c @@ -215,8 +215,7 @@ static int wm831x_gp_ldo_set_mode(struct regulator_dev *rdev, case REGULATOR_MODE_IDLE: ret = wm831x_set_bits(wm831x, ctrl_reg, - WM831X_LDO1_LP_MODE, - WM831X_LDO1_LP_MODE); + WM831X_LDO1_LP_MODE, 0); if (ret < 0) return ret; @@ -225,10 +224,12 @@ static int wm831x_gp_ldo_set_mode(struct regulator_dev *rdev, WM831X_LDO1_ON_MODE); if (ret < 0) return ret; + break; case REGULATOR_MODE_STANDBY: ret = wm831x_set_bits(wm831x, ctrl_reg, - WM831X_LDO1_LP_MODE, 0); + WM831X_LDO1_LP_MODE, + WM831X_LDO1_LP_MODE); if (ret < 0) return ret; -- cgit v0.10.2 From 8ecee36adc9d2cf19471c395af6ef70264dec251 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Mon, 6 Sep 2010 14:06:07 +0800 Subject: regulator: wm8350-regulator - fix the logic of checking REGULATOR_MODE_STANDBY mode In wm8350_dcdc_set_mode(), we set DCx_SLEEP bit of WM8350_DCDC_SLEEP_OPTIONS register for REGULATOR_MODE_STANDBY mode. ( DCx_SLEEP bits: 0: Normal DC-DC operation 1: Select LDO mode ) In wm8350_dcdc_get_mode(), current logic to determinate REGULATOR_MODE_STANDBY mode is just reverse. ( sleep is set should mean REGULATOR_MODE_STANDBY mode. ) Signed-off-by: Axel Lin Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/drivers/regulator/wm8350-regulator.c b/drivers/regulator/wm8350-regulator.c index 0e6ed7d..fe4b8a8 100644 --- a/drivers/regulator/wm8350-regulator.c +++ b/drivers/regulator/wm8350-regulator.c @@ -1129,7 +1129,7 @@ static unsigned int wm8350_dcdc_get_mode(struct regulator_dev *rdev) mode = REGULATOR_MODE_NORMAL; } else if (!active && !sleep) mode = REGULATOR_MODE_IDLE; - else if (!sleep) + else if (sleep) mode = REGULATOR_MODE_STANDBY; return mode; -- cgit v0.10.2 From 4d155641c81203440da64c4633b4efaab75f63b3 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 7 Sep 2010 11:58:30 +0200 Subject: ALSA: hda - Add quirk for Lenovo T400s Lenovo T400s requires the quirk to make automatic HP/mic switching working. Reported-by: Frank Becker Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c index 4f06199..71f9d64 100644 --- a/sound/pci/hda/patch_conexant.c +++ b/sound/pci/hda/patch_conexant.c @@ -3094,6 +3094,7 @@ static struct snd_pci_quirk cxt5066_cfg_tbl[] = { SND_PCI_QUIRK(0x103c, 0x360b, "HP G60", CXT5066_HP_LAPTOP), SND_PCI_QUIRK(0x1179, 0xff50, "Toshiba Satellite P500-PSPGSC-01800T", CXT5066_OLPC_XO_1_5), SND_PCI_QUIRK(0x1179, 0xffe0, "Toshiba Satellite Pro T130-15F", CXT5066_OLPC_XO_1_5), + SND_PCI_QUIRK(0x17aa, 0x20f2, "Lenovo T400s", CXT5066_THINKPAD), SND_PCI_QUIRK(0x17aa, 0x21b2, "Thinkpad X100e", CXT5066_IDEAPAD), SND_PCI_QUIRK(0x17aa, 0x21b3, "Thinkpad Edge 13 (197)", CXT5066_IDEAPAD), SND_PCI_QUIRK(0x17aa, 0x21b4, "Thinkpad Edge", CXT5066_IDEAPAD), -- cgit v0.10.2 From 595afaf9e6ee1b48e13ec4b8bcc8c7dee888161a Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 7 Sep 2010 13:42:41 +0200 Subject: fuse: flush background queue on connection close David Bartly reported that fuse can hang in fuse_get_req_nofail() when the connection to the filesystem server is no longer active. If bg_queue is not empty then flush_bg_queue() called from request_end() can put more requests on to the pending queue. If this happens while ending requests on the processing queue then those background requests will be queued to the pending list and never ended. Another problem is that fuse_dev_release() didn't wake up processes sleeping on blocked_waitq. Solve this by: a) flushing the background queue before calling end_requests() on the pending and processing queues b) setting blocked = 0 and waking up processes waiting on blocked_waitq() Thanks to David for an excellent bug report. Reported-by: David Bartley Signed-off-by: Miklos Szeredi CC: stable@kernel.org diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 69ad053..b4fc47f 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1769,6 +1769,14 @@ __acquires(&fc->lock) } } +static void end_queued_requests(struct fuse_conn *fc) +{ + fc->max_background = UINT_MAX; + flush_bg_queue(fc); + end_requests(fc, &fc->pending); + end_requests(fc, &fc->processing); +} + /* * Abort all requests. * @@ -1795,8 +1803,7 @@ void fuse_abort_conn(struct fuse_conn *fc) fc->connected = 0; fc->blocked = 0; end_io_requests(fc); - end_requests(fc, &fc->pending); - end_requests(fc, &fc->processing); + end_queued_requests(fc); wake_up_all(&fc->waitq); wake_up_all(&fc->blocked_waitq); kill_fasync(&fc->fasync, SIGIO, POLL_IN); @@ -1811,8 +1818,9 @@ int fuse_dev_release(struct inode *inode, struct file *file) if (fc) { spin_lock(&fc->lock); fc->connected = 0; - end_requests(fc, &fc->pending); - end_requests(fc, &fc->processing); + fc->blocked = 0; + end_queued_requests(fc); + wake_up_all(&fc->blocked_waitq); spin_unlock(&fc->lock); fuse_conn_put(fc); } -- cgit v0.10.2 From b9ca67b2ddf021491a3741d9555e8ff59b2175ba Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 7 Sep 2010 13:42:41 +0200 Subject: fuse: fix lock annotations Sparse doesn't understand lock annotations of the form __releases(&foo->lock). Change them to __releases(foo->lock). Same for __acquires(). Signed-off-by: Miklos Szeredi diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index b4fc47f..d367af1 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -276,7 +276,7 @@ static void flush_bg_queue(struct fuse_conn *fc) * Called with fc->lock, unlocks it */ static void request_end(struct fuse_conn *fc, struct fuse_req *req) -__releases(&fc->lock) +__releases(fc->lock) { void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; req->end = NULL; @@ -306,8 +306,8 @@ __releases(&fc->lock) static void wait_answer_interruptible(struct fuse_conn *fc, struct fuse_req *req) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { if (signal_pending(current)) return; @@ -325,8 +325,8 @@ static void queue_interrupt(struct fuse_conn *fc, struct fuse_req *req) } static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { if (!fc->no_interrupt) { /* Any signal may interrupt this */ @@ -905,8 +905,8 @@ static int request_pending(struct fuse_conn *fc) /* Wait until a request is available on the pending list */ static void request_wait(struct fuse_conn *fc) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { DECLARE_WAITQUEUE(wait, current); @@ -934,7 +934,7 @@ __acquires(&fc->lock) */ static int fuse_read_interrupt(struct fuse_conn *fc, struct fuse_copy_state *cs, size_t nbytes, struct fuse_req *req) -__releases(&fc->lock) +__releases(fc->lock) { struct fuse_in_header ih; struct fuse_interrupt_in arg; @@ -1720,8 +1720,8 @@ static unsigned fuse_dev_poll(struct file *file, poll_table *wait) * This function releases and reacquires fc->lock */ static void end_requests(struct fuse_conn *fc, struct list_head *head) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { while (!list_empty(head)) { struct fuse_req *req; @@ -1744,8 +1744,8 @@ __acquires(&fc->lock) * locked). */ static void end_io_requests(struct fuse_conn *fc) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { while (!list_empty(&fc->io)) { struct fuse_req *req = @@ -1770,6 +1770,8 @@ __acquires(&fc->lock) } static void end_queued_requests(struct fuse_conn *fc) +__releases(fc->lock) +__acquires(fc->lock) { fc->max_background = UINT_MAX; flush_bg_queue(fc); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 147c1f7..c822458 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1144,8 +1144,8 @@ static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req) /* Called under fc->lock, may release and reacquire it */ static void fuse_send_writepage(struct fuse_conn *fc, struct fuse_req *req) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { struct fuse_inode *fi = get_fuse_inode(req->inode); loff_t size = i_size_read(req->inode); @@ -1183,8 +1183,8 @@ __acquires(&fc->lock) * Called with fc->lock */ void fuse_flush_writepages(struct inode *inode) -__releases(&fc->lock) -__acquires(&fc->lock) +__releases(fc->lock) +__acquires(fc->lock) { struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_inode *fi = get_fuse_inode(inode); -- cgit v0.10.2 From 7e443312403ad1ff40ef3177590e96d1fe747c79 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 7 Sep 2010 11:27:52 -0400 Subject: [SCSI] sd: fix medium-removal bug Commit 409f3499a2cfcd1e9c2857c53af7fcce069f027f (scsi/sd: remove big kernel lock) introduced a bug in the sd_release routine. Medium removal should be allowed when the number of open file references drops to 0, not when it becomes non-zero. This patch (as1414) adjusts the test to fix the bug. Signed-off-by: Alan Stern Signed-off-by: James Bottomley diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index cd71f46..ffa0689 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -870,7 +870,7 @@ static int sd_release(struct gendisk *disk, fmode_t mode) SCSI_LOG_HLQUEUE(3, sd_printk(KERN_INFO, sdkp, "sd_release\n")); - if (atomic_dec_return(&sdkp->openers) && sdev->removable) { + if (atomic_dec_return(&sdkp->openers) == 0 && sdev->removable) { if (scsi_block_when_processing_errors(sdev)) scsi_set_medium_removal(sdev, SCSI_REMOVAL_ALLOW); } -- cgit v0.10.2 From 8d2602e0778299e2d6084f03086b716d6e7a1e1e Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Fri, 20 Aug 2010 16:44:33 +0200 Subject: AT91: change dma resource index Reported-by: Dan Liang Signed-off-by: Nicolas Ferre diff --git a/arch/arm/mach-at91/at91sam9g45_devices.c b/arch/arm/mach-at91/at91sam9g45_devices.c index 809114d..501ac6f 100644 --- a/arch/arm/mach-at91/at91sam9g45_devices.c +++ b/arch/arm/mach-at91/at91sam9g45_devices.c @@ -46,7 +46,7 @@ static struct resource hdmac_resources[] = { .end = AT91_BASE_SYS + AT91_DMA + SZ_512 - 1, .flags = IORESOURCE_MEM, }, - [2] = { + [1] = { .start = AT91SAM9G45_ID_DMA, .end = AT91SAM9G45_ID_DMA, .flags = IORESOURCE_IRQ, -- cgit v0.10.2 From bc41606aefa8b17000619f510d5809e6c4003d65 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 7 Sep 2010 21:51:02 +0100 Subject: Revert "drm/i915: Enable RC6 on Ironlake." This reverts commit ce17178094f368d9e3f39b2cb4303da5ed633dd4. This commit has been independently bisected a few times as being the cause of a s2ram failure. Reported-and-tested-by: Kyle McMartin Reported-and-tested-by: Andy Isaacson Cc: Zou Nan hai Signed-off-by: Chris Wilson diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 40cc5da..e0d1952 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -5696,8 +5696,7 @@ void intel_init_clock_gating(struct drm_device *dev) ILK_DPFC_DIS2 | ILK_CLK_FBC); } - if (IS_GEN6(dev)) - return; + return; } else if (IS_G4X(dev)) { uint32_t dspclk_gate; I915_WRITE(RENCLK_GATE_D1, 0); @@ -5758,11 +5757,9 @@ void intel_init_clock_gating(struct drm_device *dev) OUT_RING(MI_FLUSH); ADVANCE_LP_RING(); } - } else { + } else DRM_DEBUG_KMS("Failed to allocate render context." - "Disable RC6\n"); - return; - } + "Disable RC6\n"); } if (I915_HAS_RC6(dev) && drm_core_check_feature(dev, DRIVER_MODESET)) { -- cgit v0.10.2 From e819ff519b2d74373eca4a9a2b417ebf4c1e1b29 Mon Sep 17 00:00:00 2001 From: John Johansen Date: Fri, 27 Aug 2010 18:33:26 -0700 Subject: AppArmor: Drop hack to remove appended " (deleted)" string The 2.6.36 kernel has refactored __d_path() so that it no longer appends " (deleted)" to unlinked paths. So drop the hack that was used to detect and remove the appended string. Signed-off-by: John Johansen Signed-off-by: James Morris diff --git a/security/apparmor/path.c b/security/apparmor/path.c index 19358dc..8239605 100644 --- a/security/apparmor/path.c +++ b/security/apparmor/path.c @@ -59,8 +59,7 @@ static int d_namespace_path(struct path *path, char *buf, int buflen, { struct path root, tmp; char *res; - int deleted, connected; - int error = 0; + int connected, error = 0; /* Get the root we want to resolve too, released below */ if (flags & PATH_CHROOT_REL) { @@ -74,19 +73,8 @@ static int d_namespace_path(struct path *path, char *buf, int buflen, } spin_lock(&dcache_lock); - /* There is a race window between path lookup here and the - * need to strip the " (deleted) string that __d_path applies - * Detect the race and relookup the path - * - * The stripping of (deleted) is a hack that could be removed - * with an updated __d_path - */ - do { - tmp = root; - deleted = d_unlinked(path->dentry); - res = __d_path(path, &tmp, buf, buflen); - - } while (deleted != d_unlinked(path->dentry)); + tmp = root; + res = __d_path(path, &tmp, buf, buflen); spin_unlock(&dcache_lock); *name = res; @@ -98,21 +86,17 @@ static int d_namespace_path(struct path *path, char *buf, int buflen, *name = buf; goto out; } - if (deleted) { - /* On some filesystems, newly allocated dentries appear to the - * security_path hooks as a deleted dentry except without an - * inode allocated. - * - * Remove the appended deleted text and return as string for - * normal mediation, or auditing. The (deleted) string is - * guaranteed to be added in this case, so just strip it. - */ - buf[buflen - 11] = 0; /* - (len(" (deleted)") +\0) */ - if (path->dentry->d_inode && !(flags & PATH_MEDIATE_DELETED)) { + /* Handle two cases: + * 1. A deleted dentry && profile is not allowing mediation of deleted + * 2. On some filesystems, newly allocated dentries appear to the + * security_path hooks as a deleted dentry except without an inode + * allocated. + */ + if (d_unlinked(path->dentry) && path->dentry->d_inode && + !(flags & PATH_MEDIATE_DELETED)) { error = -ENOENT; goto out; - } } /* Determine if the path is connected to the expected root */ -- cgit v0.10.2 From 3a2dc8382a3e85a51ed9c6f57ea80665ea7a0c95 Mon Sep 17 00:00:00 2001 From: John Johansen Date: Mon, 6 Sep 2010 10:10:20 -0700 Subject: AppArmor: Fix security_task_setrlimit logic for 2.6.36 changes 2.6.36 introduced the abilitiy to specify the task that is having its rlimits set. Update mediation to ensure that confined tasks can only set their own group_leader as expected by current policy. Add TODO note about extending policy to support setting other tasks rlimits. Signed-off-by: John Johansen Signed-off-by: James Morris diff --git a/security/apparmor/include/resource.h b/security/apparmor/include/resource.h index 3c88be9..02baec7 100644 --- a/security/apparmor/include/resource.h +++ b/security/apparmor/include/resource.h @@ -33,8 +33,8 @@ struct aa_rlimit { }; int aa_map_resource(int resource); -int aa_task_setrlimit(struct aa_profile *profile, unsigned int resource, - struct rlimit *new_rlim); +int aa_task_setrlimit(struct aa_profile *profile, struct task_struct *, + unsigned int resource, struct rlimit *new_rlim); void __aa_transition_rlimits(struct aa_profile *old, struct aa_profile *new); diff --git a/security/apparmor/lsm.c b/security/apparmor/lsm.c index f73e2c2..cf1de44 100644 --- a/security/apparmor/lsm.c +++ b/security/apparmor/lsm.c @@ -614,7 +614,7 @@ static int apparmor_task_setrlimit(struct task_struct *task, int error = 0; if (!unconfined(profile)) - error = aa_task_setrlimit(profile, resource, new_rlim); + error = aa_task_setrlimit(profile, task, resource, new_rlim); return error; } diff --git a/security/apparmor/resource.c b/security/apparmor/resource.c index 4a368f1..a4136c1 100644 --- a/security/apparmor/resource.c +++ b/security/apparmor/resource.c @@ -72,6 +72,7 @@ int aa_map_resource(int resource) /** * aa_task_setrlimit - test permission to set an rlimit * @profile - profile confining the task (NOT NULL) + * @task - task the resource is being set on * @resource - the resource being set * @new_rlim - the new resource limit (NOT NULL) * @@ -79,18 +80,21 @@ int aa_map_resource(int resource) * * Returns: 0 or error code if setting resource failed */ -int aa_task_setrlimit(struct aa_profile *profile, unsigned int resource, - struct rlimit *new_rlim) +int aa_task_setrlimit(struct aa_profile *profile, struct task_struct *task, + unsigned int resource, struct rlimit *new_rlim) { int error = 0; - if (profile->rlimits.mask & (1 << resource) && - new_rlim->rlim_max > profile->rlimits.limits[resource].rlim_max) - - error = audit_resource(profile, resource, new_rlim->rlim_max, - -EACCES); + /* TODO: extend resource control to handle other (non current) + * processes. AppArmor rules currently have the implicit assumption + * that the task is setting the resource of the current process + */ + if ((task != current->group_leader) || + (profile->rlimits.mask & (1 << resource) && + new_rlim->rlim_max > profile->rlimits.limits[resource].rlim_max)) + error = -EACCES; - return error; + return audit_resource(profile, resource, new_rlim->rlim_max, error); } /** -- cgit v0.10.2 From 04ccd53f09741c4bc54ab36db000bc1383e4812e Mon Sep 17 00:00:00 2001 From: John Johansen Date: Fri, 27 Aug 2010 18:33:28 -0700 Subject: AppArmor: Fix splitting an fqname into separate namespace and profile names As per Dan Carpenter If we have a ns name without a following profile then in the original code it did "*ns_name = &name[1];". "name" is NULL so "*ns_name" is 0x1. That isn't useful and could cause an oops when this function is called from aa_remove_profiles(). Beyond this the assignment of the namespace name was wrong in the case where the profile name was provided as it was being set to &name[1] after name = skip_spaces(split + 1); Move the ns_name assignment before updating name for the split and also add skip_spaces, making the interface more robust. Signed-off-by: John Johansen Signed-off-by: James Morris diff --git a/security/apparmor/lib.c b/security/apparmor/lib.c index 6e85cdb..506d2ba 100644 --- a/security/apparmor/lib.c +++ b/security/apparmor/lib.c @@ -40,6 +40,7 @@ char *aa_split_fqname(char *fqname, char **ns_name) *ns_name = NULL; if (name[0] == ':') { char *split = strchr(&name[1], ':'); + *ns_name = skip_spaces(&name[1]); if (split) { /* overwrite ':' with \0 */ *split = 0; @@ -47,7 +48,6 @@ char *aa_split_fqname(char *fqname, char **ns_name) } else /* a ns name without a following profile is allowed */ name = NULL; - *ns_name = &name[1]; } if (name && *name == 0) name = NULL; -- cgit v0.10.2 From 999b4f0aa2314b76857775334cb94bafa053db64 Mon Sep 17 00:00:00 2001 From: John Johansen Date: Fri, 27 Aug 2010 18:33:29 -0700 Subject: AppArmor: Fix locking from removal of profile namespace The locking for profile namespace removal is wrong, when removing a profile namespace, it needs to be removed from its parent's list. Lock the parent of namespace list instead of the namespace being removed. Signed-off-by: John Johansen Signed-off-by: James Morris diff --git a/security/apparmor/policy.c b/security/apparmor/policy.c index 3cdc1ad..52cc865 100644 --- a/security/apparmor/policy.c +++ b/security/apparmor/policy.c @@ -1151,12 +1151,14 @@ ssize_t aa_remove_profiles(char *fqname, size_t size) /* released below */ ns = aa_get_namespace(root); - write_lock(&ns->lock); if (!name) { /* remove namespace - can only happen if fqname[0] == ':' */ + write_lock(&ns->parent->lock); __remove_namespace(ns); + write_unlock(&ns->parent->lock); } else { /* remove profile */ + write_lock(&ns->lock); profile = aa_get_profile(__lookup_profile(&ns->base, name)); if (!profile) { error = -ENOENT; @@ -1165,8 +1167,8 @@ ssize_t aa_remove_profiles(char *fqname, size_t size) } name = profile->base.hname; __remove_profile(profile); + write_unlock(&ns->lock); } - write_unlock(&ns->lock); /* don't fail removal if audit fails */ (void) audit_policy(OP_PROF_RM, GFP_KERNEL, name, info, error); -- cgit v0.10.2 From e950598d43dce8d97e7d5270808393425d1e5cbd Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Tue, 31 Aug 2010 09:38:51 -0400 Subject: ima: always maintain counters commit 8262bb85da allocated the inode integrity struct (iint) before any inodes were created. Only after IMA was initialized in late_initcall were the counters updated. This patch updates the counters, whether or not IMA has been initialized, to resolve 'imbalance' messages. This patch fixes the bug as reported in bugzilla: 15673. When the i915 is builtin, the ring_buffer is initialized before IMA, causing the imbalance message on suspend. Reported-by: Thomas Meyer Signed-off-by: Mimi Zohar Tested-by: Thomas Meyer Tested-by: David Safford Cc: Stable Kernel Signed-off-by: James Morris diff --git a/security/integrity/ima/ima.h b/security/integrity/ima/ima.h index 16d100d..3fbcd1d 100644 --- a/security/integrity/ima/ima.h +++ b/security/integrity/ima/ima.h @@ -35,6 +35,7 @@ enum tpm_pcrs { TPM_PCR0 = 0, TPM_PCR8 = 8 }; #define IMA_MEASURE_HTABLE_SIZE (1 << IMA_HASH_BITS) /* set during initialization */ +extern int iint_initialized; extern int ima_initialized; extern int ima_used_chip; extern char *ima_hash; diff --git a/security/integrity/ima/ima_iint.c b/security/integrity/ima/ima_iint.c index 7625b85..afba4ae 100644 --- a/security/integrity/ima/ima_iint.c +++ b/security/integrity/ima/ima_iint.c @@ -22,9 +22,10 @@ RADIX_TREE(ima_iint_store, GFP_ATOMIC); DEFINE_SPINLOCK(ima_iint_lock); - static struct kmem_cache *iint_cache __read_mostly; +int iint_initialized = 0; + /* ima_iint_find_get - return the iint associated with an inode * * ima_iint_find_get gets a reference to the iint. Caller must @@ -141,6 +142,7 @@ static int __init ima_iintcache_init(void) iint_cache = kmem_cache_create("iint_cache", sizeof(struct ima_iint_cache), 0, SLAB_PANIC, init_once); + iint_initialized = 1; return 0; } security_initcall(ima_iintcache_init); diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c index f936413..e662b89 100644 --- a/security/integrity/ima/ima_main.c +++ b/security/integrity/ima/ima_main.c @@ -148,12 +148,14 @@ void ima_counts_get(struct file *file) struct ima_iint_cache *iint; int rc; - if (!ima_initialized || !S_ISREG(inode->i_mode)) + if (!iint_initialized || !S_ISREG(inode->i_mode)) return; iint = ima_iint_find_get(inode); if (!iint) return; mutex_lock(&iint->mutex); + if (!ima_initialized) + goto out; rc = ima_must_measure(iint, inode, MAY_READ, FILE_CHECK); if (rc < 0) goto out; @@ -213,7 +215,7 @@ void ima_file_free(struct file *file) struct inode *inode = file->f_dentry->d_inode; struct ima_iint_cache *iint; - if (!ima_initialized || !S_ISREG(inode->i_mode)) + if (!iint_initialized || !S_ISREG(inode->i_mode)) return; iint = ima_iint_find_get(inode); if (!iint) @@ -230,7 +232,7 @@ static int process_measurement(struct file *file, const unsigned char *filename, { struct inode *inode = file->f_dentry->d_inode; struct ima_iint_cache *iint; - int rc; + int rc = 0; if (!ima_initialized || !S_ISREG(inode->i_mode)) return 0; -- cgit v0.10.2 From dc696aced9f09f05b1f927b93f5a7918017a3e49 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Thu, 12 Aug 2010 16:24:25 -0700 Subject: ocfs2: Fix metaecc error messages Like tools, the checksum validate function now prints the values in hex. Signed-off-by: Sunil Mushran Singed-off-by: Tao Ma diff --git a/fs/ocfs2/blockcheck.c b/fs/ocfs2/blockcheck.c index ec6d123..c7ee03c 100644 --- a/fs/ocfs2/blockcheck.c +++ b/fs/ocfs2/blockcheck.c @@ -439,7 +439,7 @@ int ocfs2_block_check_validate(void *data, size_t blocksize, ocfs2_blockcheck_inc_failure(stats); mlog(ML_ERROR, - "CRC32 failed: stored: %u, computed %u. Applying ECC.\n", + "CRC32 failed: stored: 0x%x, computed 0x%x. Applying ECC.\n", (unsigned int)check.bc_crc32e, (unsigned int)crc); /* Ok, try ECC fixups */ @@ -453,7 +453,7 @@ int ocfs2_block_check_validate(void *data, size_t blocksize, goto out; } - mlog(ML_ERROR, "Fixed CRC32 failed: stored: %u, computed %u\n", + mlog(ML_ERROR, "Fixed CRC32 failed: stored: 0x%x, computed 0x%x\n", (unsigned int)check.bc_crc32e, (unsigned int)crc); rc = -EIO; -- cgit v0.10.2 From f5ce5a08a40f2086435858ddc80cb40394b082eb Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Thu, 12 Aug 2010 16:24:26 -0700 Subject: ocfs2: Fix incorrect checksum validation error For local mounts, ocfs2_read_locked_inode() calls ocfs2_read_blocks_sync() to read the inode off the disk. The latter first checks to see if that block is cached in the journal, and, if so, returns that block. That is ok. But ocfs2_read_locked_inode() goes wrong when it tries to validate the checksum of such blocks. Blocks that are cached in the journal may not have had their checksum computed as yet. We should not validate the checksums of such blocks. Fixes ossbz#1282 http://oss.oracle.com/bugzilla/show_bug.cgi?id=1282 Signed-off-by: Sunil Mushran Cc: stable@kernel.org Singed-off-by: Tao Ma diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index 0492464..eece3e0 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -488,7 +488,11 @@ static int ocfs2_read_locked_inode(struct inode *inode, OCFS2_BH_IGNORE_CACHE); } else { status = ocfs2_read_blocks_sync(osb, args->fi_blkno, 1, &bh); - if (!status) + /* + * If buffer is in jbd, then its checksum may not have been + * computed as yet. + */ + if (!status && !buffer_jbd(bh)) status = ocfs2_validate_inode_block(osb->sb, bh); } if (status < 0) { -- cgit v0.10.2 From f63afdb2c32db850fa1bfccf84643a8885cbeb61 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Sat, 17 Jul 2010 21:45:49 +0800 Subject: ocfs2: make __ocfs2_page_mkwrite handle file end properly. __ocfs2_page_mkwrite now is broken in handling file end. 1. the last page should be the page contains i_size - 1. 2. the len in the last page is also calculated wrong. So change them accordingly. Acked-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c index af2b8fe..4c18f4a 100644 --- a/fs/ocfs2/mmap.c +++ b/fs/ocfs2/mmap.c @@ -74,9 +74,11 @@ static int __ocfs2_page_mkwrite(struct inode *inode, struct buffer_head *di_bh, /* * Another node might have truncated while we were waiting on * cluster locks. + * We don't check size == 0 before the shift. This is borrowed + * from do_generic_file_read. */ - last_index = size >> PAGE_CACHE_SHIFT; - if (page->index > last_index) { + last_index = (size - 1) >> PAGE_CACHE_SHIFT; + if (unlikely(!size || page->index > last_index)) { ret = -EINVAL; goto out; } @@ -107,7 +109,7 @@ static int __ocfs2_page_mkwrite(struct inode *inode, struct buffer_head *di_bh, * because the "write" would invalidate their data. */ if (page->index == last_index) - len = size & ~PAGE_CACHE_MASK; + len = ((size - 1) & ~PAGE_CACHE_MASK) + 1; ret = ocfs2_write_begin_nolock(mapping, pos, len, 0, &locked_page, &fsdata, di_bh, page); -- cgit v0.10.2 From 04eda1a18019bb387dc7e97ee99979dd88dc608a Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 5 Aug 2010 20:32:45 +0200 Subject: ocfs2: Flush drive's caches on fdatasync When 'barrier' mount option is specified, we have to issue a cache flush during fdatasync(2). We have to do this even if inode doesn't have I_DIRTY_DATASYNC set because we still have to get written *data* to disk so that they are not lost in case of crash. Acked-by: Tao Ma Signed-off-by: Jan Kara Singed-off-by: Tao Ma diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 81296b4..6b2be0f 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -36,6 +36,7 @@ #include #include #include +#include #define MLOG_MASK_PREFIX ML_INODE #include @@ -190,8 +191,16 @@ static int ocfs2_sync_file(struct file *file, int datasync) if (err) goto bail; - if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) + if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) { + /* + * We still have to flush drive's caches to get data to the + * platter + */ + if (osb->s_mount_opt & OCFS2_MOUNT_BARRIER) + blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, + NULL, BLKDEV_IFL_WAIT); goto bail; + } journal = osb->journal->j_journal; err = jbd2_journal_force_commit(journal); -- cgit v0.10.2 From 889f004a8c83d515f275078687f859bc0d5ede9d Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Thu, 2 Sep 2010 13:10:10 +0800 Subject: ocfs2: Use the right group in nfs sync check. We have added discontig block group now, and now an inode can be allocated in an discontig block group. So get it in ocfs2_get_suballoc_slot_bit. The old ocfs2_test_suballoc_bit gets group block no from the allocation inode which is wrong. Fix it by passing the right group. Acked-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index a8e6a95..8a009ee 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -2567,7 +2567,8 @@ out: * suballoc_bit. */ static int ocfs2_get_suballoc_slot_bit(struct ocfs2_super *osb, u64 blkno, - u16 *suballoc_slot, u16 *suballoc_bit) + u16 *suballoc_slot, u64 *group_blkno, + u16 *suballoc_bit) { int status; struct buffer_head *inode_bh = NULL; @@ -2604,6 +2605,8 @@ static int ocfs2_get_suballoc_slot_bit(struct ocfs2_super *osb, u64 blkno, *suballoc_slot = le16_to_cpu(inode_fe->i_suballoc_slot); if (suballoc_bit) *suballoc_bit = le16_to_cpu(inode_fe->i_suballoc_bit); + if (group_blkno) + *group_blkno = le64_to_cpu(inode_fe->i_suballoc_loc); bail: brelse(inode_bh); @@ -2621,7 +2624,8 @@ bail: */ static int ocfs2_test_suballoc_bit(struct ocfs2_super *osb, struct inode *suballoc, - struct buffer_head *alloc_bh, u64 blkno, + struct buffer_head *alloc_bh, + u64 group_blkno, u64 blkno, u16 bit, int *res) { struct ocfs2_dinode *alloc_di; @@ -2642,10 +2646,8 @@ static int ocfs2_test_suballoc_bit(struct ocfs2_super *osb, goto bail; } - if (alloc_di->i_suballoc_loc) - bg_blkno = le64_to_cpu(alloc_di->i_suballoc_loc); - else - bg_blkno = ocfs2_which_suballoc_group(blkno, bit); + bg_blkno = group_blkno ? group_blkno : + ocfs2_which_suballoc_group(blkno, bit); status = ocfs2_read_group_descriptor(suballoc, alloc_di, bg_blkno, &group_bh); if (status < 0) { @@ -2680,6 +2682,7 @@ bail: int ocfs2_test_inode_bit(struct ocfs2_super *osb, u64 blkno, int *res) { int status; + u64 group_blkno = 0; u16 suballoc_bit = 0, suballoc_slot = 0; struct inode *inode_alloc_inode; struct buffer_head *alloc_bh = NULL; @@ -2687,7 +2690,7 @@ int ocfs2_test_inode_bit(struct ocfs2_super *osb, u64 blkno, int *res) mlog_entry("blkno: %llu", (unsigned long long)blkno); status = ocfs2_get_suballoc_slot_bit(osb, blkno, &suballoc_slot, - &suballoc_bit); + &group_blkno, &suballoc_bit); if (status < 0) { mlog(ML_ERROR, "get alloc slot and bit failed %d\n", status); goto bail; @@ -2715,7 +2718,7 @@ int ocfs2_test_inode_bit(struct ocfs2_super *osb, u64 blkno, int *res) } status = ocfs2_test_suballoc_bit(osb, inode_alloc_inode, alloc_bh, - blkno, suballoc_bit, res); + group_blkno, blkno, suballoc_bit, res); if (status < 0) mlog(ML_ERROR, "test suballoc bit failed %d\n", status); -- cgit v0.10.2 From b2b6ebf5f740e015b2155343958f067e594323ea Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Thu, 26 Aug 2010 13:06:50 -0700 Subject: ocfs2: properly set and use inode group alloc hint We were setting ac->ac_last_group in ocfs2_claim_suballoc_bits from res->sr_bg_blkno. Unfortunately, res->sr_bg_blkno is going to be zero under normal (non-fragmented) circumstances. The discontig block group patches effectively turned off that feature. Fix this by correctly calculating what the next group hint should be. Acked-by: Tao Ma Signed-off-by: Mark Fasheh Tested-by: Goldwyn Rodrigues Signed-off-by: Tao Ma diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index 8a009ee..b93d7e7 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -62,6 +62,17 @@ struct ocfs2_suballoc_result { unsigned int sr_bits; /* How many bits we claimed */ }; +static u64 ocfs2_group_from_res(struct ocfs2_suballoc_result *res) +{ + if (res->sr_blkno == 0) + return 0; + + if (res->sr_bg_blkno) + return res->sr_bg_blkno; + + return ocfs2_which_suballoc_group(res->sr_blkno, res->sr_bit_offset); +} + static inline void ocfs2_debug_bg(struct ocfs2_group_desc *bg); static inline void ocfs2_debug_suballoc_inode(struct ocfs2_dinode *fe); static inline u16 ocfs2_find_victim_chain(struct ocfs2_chain_list *cl); @@ -1845,6 +1856,7 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_alloc_context *ac, int status; u16 victim, i; u16 bits_left = 0; + u64 hint = ac->ac_last_group; struct ocfs2_chain_list *cl; struct ocfs2_dinode *fe; @@ -1872,7 +1884,7 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_alloc_context *ac, goto bail; } - res->sr_bg_blkno = ac->ac_last_group; + res->sr_bg_blkno = hint; if (res->sr_bg_blkno) { /* Attempt to short-circuit the usual search mechanism * by jumping straight to the most recently used @@ -1896,8 +1908,10 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_alloc_context *ac, status = ocfs2_search_chain(ac, handle, bits_wanted, min_bits, res, &bits_left); - if (!status) + if (!status) { + hint = ocfs2_group_from_res(res); goto set_hint; + } if (status < 0 && status != -ENOSPC) { mlog_errno(status); goto bail; @@ -1920,8 +1934,10 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_alloc_context *ac, ac->ac_chain = i; status = ocfs2_search_chain(ac, handle, bits_wanted, min_bits, res, &bits_left); - if (!status) + if (!status) { + hint = ocfs2_group_from_res(res); break; + } if (status < 0 && status != -ENOSPC) { mlog_errno(status); goto bail; @@ -1936,7 +1952,7 @@ set_hint: if (bits_left < min_bits) ac->ac_last_group = 0; else - ac->ac_last_group = res->sr_bg_blkno; + ac->ac_last_group = hint; } bail: -- cgit v0.10.2 From 9b4c0ff32ccd87ab52d4c5bd0a0536febce11370 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 24 Aug 2010 14:28:03 +0200 Subject: ocfs2: Fix deadlock when allocating page We cannot call grab_cache_page() when holding filesystem locks or with a transaction started as grab_cache_page() calls page allocation with GFP_KERNEL flag and thus page reclaim can recurse back into the filesystem causing deadlocks or various assertion failures. We have to use find_or_create_page() instead and pass it GFP_NOFS as we do with other allocations. Acked-by: Mark Fasheh Signed-off-by: Jan Kara Signed-off-by: Tao Ma diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 215e12c..592fae5 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -6672,7 +6672,7 @@ int ocfs2_grab_pages(struct inode *inode, loff_t start, loff_t end, last_page_bytes = PAGE_ALIGN(end); index = start >> PAGE_CACHE_SHIFT; do { - pages[numpages] = grab_cache_page(mapping, index); + pages[numpages] = find_or_create_page(mapping, index, GFP_NOFS); if (!pages[numpages]) { ret = -ENOMEM; mlog_errno(ret); diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 6b2be0f..2caa3a7 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -783,7 +783,7 @@ static int ocfs2_write_zero_page(struct inode *inode, u64 abs_from, BUG_ON(abs_to > (((u64)index + 1) << PAGE_CACHE_SHIFT)); BUG_ON(abs_from & (inode->i_blkbits - 1)); - page = grab_cache_page(mapping, index); + page = find_or_create_page(mapping, index, GFP_NOFS); if (!page) { ret = -ENOMEM; mlog_errno(ret); diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index 73a11cc..0afeda831 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -2960,7 +2960,7 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle, if (map_end & (PAGE_CACHE_SIZE - 1)) to = map_end & (PAGE_CACHE_SIZE - 1); - page = grab_cache_page(mapping, page_index); + page = find_or_create_page(mapping, page_index, GFP_NOFS); /* * In case PAGE_CACHE_SIZE <= CLUSTER_SIZE, This page @@ -3179,7 +3179,8 @@ static int ocfs2_cow_sync_writeback(struct super_block *sb, if (map_end > end) map_end = end; - page = grab_cache_page(context->inode->i_mapping, page_index); + page = find_or_create_page(context->inode->i_mapping, + page_index, GFP_NOFS); BUG_ON(!page); wait_on_page_writeback(page); -- cgit v0.10.2 From 81c8c82b5a39f9127e8b239e9b406a6c3a41b228 Mon Sep 17 00:00:00 2001 From: Tristan Ye Date: Thu, 19 Aug 2010 15:15:00 +0800 Subject: Ocfs2: Fix a regression bug from mainline commit(6b933c8e6f1a2f3118082c455eef25f9b1ac7b45). The patch is to fix the regression bug brought from commit 6b933c8...( 'ocfs2: Avoid direct write if we fall back to buffered I/O'): http://oss.oracle.com/bugzilla/show_bug.cgi?id=1285 The commit 6b933c8e6f1a2f3118082c455eef25f9b1ac7b45 changed __generic_file_aio_write to generic_file_buffered_write, which didn't call filemap_{write,wait}_range to flush the pagecaches when we were falling O_DIRECT writes back to buffered ones. it did hurt the O_DIRECT semantics somehow in extented odirect writes. This patch tries to guarantee O_DIRECT writes of 'fall back to buffered' to be correctly flushed. Signed-off-by: Tristan Ye Signed-off-by: Tao Ma diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 2caa3a7..9a03c15 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2338,7 +2338,7 @@ out_dio: BUG_ON(ret == -EIOCBQUEUED && !(file->f_flags & O_DIRECT)); if (((file->f_flags & O_DSYNC) && !direct_io) || IS_SYNC(inode) || - ((file->f_flags & O_DIRECT) && has_refcount)) { + ((file->f_flags & O_DIRECT) && !direct_io)) { ret = filemap_fdatawrite_range(file->f_mapping, pos, pos + count - 1); if (ret < 0) -- cgit v0.10.2 From 021960cab320ae3cc4e9aba9cca42f9f5ce785f3 Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Fri, 13 Aug 2010 15:15:15 -0700 Subject: ocfs2: split out inode alloc code from ocfs2_mknod_locked Do this by splitting the bulk of the function away from the inode allocation code at the very tom of ocfs2_mknod_locked(). Existing callers don't need to change and won't see any difference. The new function created, __ocfs2_mknod_locked() will be used shortly. Signed-off-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index f171b51..2aa66b6 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -472,32 +472,23 @@ leave: return status; } -static int ocfs2_mknod_locked(struct ocfs2_super *osb, - struct inode *dir, - struct inode *inode, - dev_t dev, - struct buffer_head **new_fe_bh, - struct buffer_head *parent_fe_bh, - handle_t *handle, - struct ocfs2_alloc_context *inode_ac) +static int __ocfs2_mknod_locked(struct inode *dir, + struct inode *inode, + dev_t dev, + struct buffer_head **new_fe_bh, + struct buffer_head *parent_fe_bh, + handle_t *handle, + struct ocfs2_alloc_context *inode_ac, + u64 fe_blkno, u64 suballoc_loc, u16 suballoc_bit) { int status = 0; + struct ocfs2_super *osb = OCFS2_SB(dir->i_sb); struct ocfs2_dinode *fe = NULL; struct ocfs2_extent_list *fel; - u64 suballoc_loc, fe_blkno = 0; - u16 suballoc_bit; u16 feat; *new_fe_bh = NULL; - status = ocfs2_claim_new_inode(handle, dir, parent_fe_bh, - inode_ac, &suballoc_loc, - &suballoc_bit, &fe_blkno); - if (status < 0) { - mlog_errno(status); - goto leave; - } - /* populate as many fields early on as possible - many of * these are used by the support functions here and in * callers. */ @@ -591,6 +582,34 @@ leave: return status; } +static int ocfs2_mknod_locked(struct ocfs2_super *osb, + struct inode *dir, + struct inode *inode, + dev_t dev, + struct buffer_head **new_fe_bh, + struct buffer_head *parent_fe_bh, + handle_t *handle, + struct ocfs2_alloc_context *inode_ac) +{ + int status = 0; + u64 suballoc_loc, fe_blkno = 0; + u16 suballoc_bit; + + *new_fe_bh = NULL; + + status = ocfs2_claim_new_inode(handle, dir, parent_fe_bh, + inode_ac, &suballoc_loc, + &suballoc_bit, &fe_blkno); + if (status < 0) { + mlog_errno(status); + return status; + } + + return __ocfs2_mknod_locked(dir, inode, dev, new_fe_bh, + parent_fe_bh, handle, inode_ac, + fe_blkno, suballoc_loc, suballoc_bit); +} + static int ocfs2_mkdir(struct inode *dir, struct dentry *dentry, int mode) -- cgit v0.10.2 From d51349829c378c06ba4aa7d4b16ca23739858608 Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Fri, 13 Aug 2010 15:15:16 -0700 Subject: ocfs2: use ocfs2_alloc_dinode_update_counts() instead of open coding ocfs2_search_chain() makes the same updates as ocfs2_alloc_dinode_update_counts to the alloc inode. Instead of open coding the bitmap update, use our helper function. Signed-off-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index b93d7e7..e7edda8 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -1719,7 +1719,6 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, { int status; u16 chain; - u32 tmp_used; u64 next_group; struct inode *alloc_inode = ac->ac_inode; struct buffer_head *group_bh = NULL; @@ -1807,22 +1806,14 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, } } - /* Ok, claim our bits now: set the info on dinode, chainlist - * and then the group */ - status = ocfs2_journal_access_di(handle, - INODE_CACHE(alloc_inode), - ac->ac_bh, - OCFS2_JOURNAL_ACCESS_WRITE); - if (status < 0) { + status = ocfs2_alloc_dinode_update_counts(alloc_inode, handle, + ac->ac_bh, res->sr_bits, + chain); + if (status) { mlog_errno(status); goto bail; } - tmp_used = le32_to_cpu(fe->id1.bitmap1.i_used); - fe->id1.bitmap1.i_used = cpu_to_le32(res->sr_bits + tmp_used); - le32_add_cpu(&cl->cl_recs[chain].c_free, -res->sr_bits); - ocfs2_journal_dirty(handle, ac->ac_bh); - status = ocfs2_block_group_set_bits(handle, alloc_inode, bg, -- cgit v0.10.2 From e49e27674d1dd2717ad90b21ece8f83102153315 Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Fri, 13 Aug 2010 15:15:17 -0700 Subject: ocfs2: allow return of new inode block location before allocation of the inode This allows code which needs to know the eventual block number of an inode but can't allocate it yet due to transaction or lock ordering. For example, ocfs2_create_inode_in_orphan() currently gives a junk blkno for preparation of the orphan dir because it can't yet know where the actual inode is placed - that code is actually in ocfs2_mknod_locked. This is a problem when the orphan dirs are indexed as the junk inode number will create an index entry which goes unused (and fails the later removal from the orphan dir). Now with these interfaces, ocfs2_create_inode_in_orphan() can run the block group search (and get back the inode block number) *before* any actual allocation occurs. Signed-off-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index e7edda8..8a286f5 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -57,6 +57,12 @@ struct ocfs2_suballoc_result { u64 sr_bg_blkno; /* The bg we allocated from. Set to 0 when a block group is contiguous. */ + u64 sr_bg_stable_blkno; /* + * Doesn't change, always + * set to target block + * group descriptor + * block. + */ u64 sr_blkno; /* The first allocated block */ unsigned int sr_bit_offset; /* The bit in the bg */ unsigned int sr_bits; /* How many bits we claimed */ @@ -149,6 +155,10 @@ void ocfs2_free_ac_resource(struct ocfs2_alloc_context *ac) brelse(ac->ac_bh); ac->ac_bh = NULL; ac->ac_resv = NULL; + if (ac->ac_find_loc_priv) { + kfree(ac->ac_find_loc_priv); + ac->ac_find_loc_priv = NULL; + } } void ocfs2_free_alloc_context(struct ocfs2_alloc_context *ac) @@ -1689,6 +1699,15 @@ static int ocfs2_search_one_group(struct ocfs2_alloc_context *ac, if (!ret) ocfs2_bg_discontig_fix_result(ac, gd, res); + /* + * sr_bg_blkno might have been changed by + * ocfs2_bg_discontig_fix_result + */ + res->sr_bg_stable_blkno = group_bh->b_blocknr; + + if (ac->ac_find_loc_only) + goto out_loc_only; + ret = ocfs2_alloc_dinode_update_counts(alloc_inode, handle, ac->ac_bh, res->sr_bits, le16_to_cpu(gd->bg_chain)); @@ -1702,6 +1721,7 @@ static int ocfs2_search_one_group(struct ocfs2_alloc_context *ac, if (ret < 0) mlog_errno(ret); +out_loc_only: *bits_left = le16_to_cpu(gd->bg_free_bits_count); out: @@ -1780,6 +1800,11 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, if (!status) ocfs2_bg_discontig_fix_result(ac, bg, res); + /* + * sr_bg_blkno might have been changed by + * ocfs2_bg_discontig_fix_result + */ + res->sr_bg_stable_blkno = group_bh->b_blocknr; /* * Keep track of previous block descriptor read. When @@ -1806,6 +1831,9 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, } } + if (ac->ac_find_loc_only) + goto out_loc_only; + status = ocfs2_alloc_dinode_update_counts(alloc_inode, handle, ac->ac_bh, res->sr_bits, chain); @@ -1828,6 +1856,7 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, mlog(0, "Allocated %u bits from suballocator %llu\n", res->sr_bits, (unsigned long long)le64_to_cpu(fe->i_blkno)); +out_loc_only: *bits_left = le16_to_cpu(bg->bg_free_bits_count); bail: brelse(group_bh); @@ -2023,6 +2052,136 @@ static inline void ocfs2_save_inode_ac_group(struct inode *dir, OCFS2_I(dir)->ip_last_used_slot = ac->ac_alloc_slot; } +int ocfs2_find_new_inode_loc(struct inode *dir, + struct buffer_head *parent_fe_bh, + struct ocfs2_alloc_context *ac, + u64 *fe_blkno) +{ + int ret; + handle_t *handle = NULL; + struct ocfs2_suballoc_result *res; + + BUG_ON(!ac); + BUG_ON(ac->ac_bits_given != 0); + BUG_ON(ac->ac_bits_wanted != 1); + BUG_ON(ac->ac_which != OCFS2_AC_USE_INODE); + + res = kzalloc(sizeof(*res), GFP_NOFS); + if (res == NULL) { + ret = -ENOMEM; + mlog_errno(ret); + goto out; + } + + ocfs2_init_inode_ac_group(dir, parent_fe_bh, ac); + + /* + * The handle started here is for chain relink. Alternatively, + * we could just disable relink for these calls. + */ + handle = ocfs2_start_trans(OCFS2_SB(dir->i_sb), OCFS2_SUBALLOC_ALLOC); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + handle = NULL; + mlog_errno(ret); + goto out; + } + + /* + * This will instruct ocfs2_claim_suballoc_bits and + * ocfs2_search_one_group to search but save actual allocation + * for later. + */ + ac->ac_find_loc_only = 1; + + ret = ocfs2_claim_suballoc_bits(ac, handle, 1, 1, res); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + ac->ac_find_loc_priv = res; + *fe_blkno = res->sr_blkno; + +out: + if (handle) + ocfs2_commit_trans(OCFS2_SB(dir->i_sb), handle); + + if (ret) + kfree(res); + + return ret; +} + +int ocfs2_claim_new_inode_at_loc(handle_t *handle, + struct inode *dir, + struct ocfs2_alloc_context *ac, + u64 *suballoc_loc, + u16 *suballoc_bit, + u64 di_blkno) +{ + int ret; + u16 chain; + struct ocfs2_suballoc_result *res = ac->ac_find_loc_priv; + struct buffer_head *bg_bh = NULL; + struct ocfs2_group_desc *bg; + struct ocfs2_dinode *di = (struct ocfs2_dinode *) ac->ac_bh->b_data; + + /* + * Since di_blkno is being passed back in, we check for any + * inconsistencies which may have happened between + * calls. These are code bugs as di_blkno is not expected to + * change once returned from ocfs2_find_new_inode_loc() + */ + BUG_ON(res->sr_blkno != di_blkno); + + ret = ocfs2_read_group_descriptor(ac->ac_inode, di, + res->sr_bg_stable_blkno, &bg_bh); + if (ret) { + mlog_errno(ret); + goto out; + } + + bg = (struct ocfs2_group_desc *) bg_bh->b_data; + chain = le16_to_cpu(bg->bg_chain); + + ret = ocfs2_alloc_dinode_update_counts(ac->ac_inode, handle, + ac->ac_bh, res->sr_bits, + chain); + if (ret) { + mlog_errno(ret); + goto out; + } + + ret = ocfs2_block_group_set_bits(handle, + ac->ac_inode, + bg, + bg_bh, + res->sr_bit_offset, + res->sr_bits); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + mlog(0, "Allocated %u bits from suballocator %llu\n", res->sr_bits, + (unsigned long long)di_blkno); + + atomic_inc(&OCFS2_SB(ac->ac_inode->i_sb)->alloc_stats.bg_allocs); + + BUG_ON(res->sr_bits != 1); + + *suballoc_loc = res->sr_bg_blkno; + *suballoc_bit = res->sr_bit_offset; + ac->ac_bits_given++; + ocfs2_save_inode_ac_group(dir, ac); + +out: + brelse(bg_bh); + + return ret; +} + int ocfs2_claim_new_inode(handle_t *handle, struct inode *dir, struct buffer_head *parent_fe_bh, diff --git a/fs/ocfs2/suballoc.h b/fs/ocfs2/suballoc.h index a017dd3..b8afabf 100644 --- a/fs/ocfs2/suballoc.h +++ b/fs/ocfs2/suballoc.h @@ -56,6 +56,9 @@ struct ocfs2_alloc_context { u64 ac_max_block; /* Highest block number to allocate. 0 is is the same as ~0 - unlimited */ + int ac_find_loc_only; /* hack for reflink operation ordering */ + struct ocfs2_suballoc_result *ac_find_loc_priv; /* */ + struct ocfs2_alloc_reservation *ac_resv; }; @@ -197,4 +200,22 @@ int ocfs2_lock_allocators(struct inode *inode, struct ocfs2_extent_tree *et, struct ocfs2_alloc_context **meta_ac); int ocfs2_test_inode_bit(struct ocfs2_super *osb, u64 blkno, int *res); + + + +/* + * The following two interfaces are for ocfs2_create_inode_in_orphan(). + */ +int ocfs2_find_new_inode_loc(struct inode *dir, + struct buffer_head *parent_fe_bh, + struct ocfs2_alloc_context *ac, + u64 *fe_blkno); + +int ocfs2_claim_new_inode_at_loc(handle_t *handle, + struct inode *dir, + struct ocfs2_alloc_context *ac, + u64 *suballoc_loc, + u16 *suballoc_bit, + u64 di_blkno); + #endif /* _CHAINALLOC_H_ */ -- cgit v0.10.2 From dd43bcde23c527f64897eef41aa1fed2c9905ea9 Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Fri, 13 Aug 2010 15:15:18 -0700 Subject: ocfs2: split out ocfs2_prepare_orphan_dir() into locking and prep functions We do this because ocfs2_create_inode_in_orphan() wants to order locking of the orphan dir with respect to locking of the inode allocator *before* making any changes to the directory. Signed-off-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 2aa66b6..54c6298 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -1871,61 +1871,117 @@ bail: return status; } -static int ocfs2_prepare_orphan_dir(struct ocfs2_super *osb, - struct inode **ret_orphan_dir, - u64 blkno, - char *name, - struct ocfs2_dir_lookup_result *lookup) +static int ocfs2_lookup_lock_orphan_dir(struct ocfs2_super *osb, + struct inode **ret_orphan_dir, + struct buffer_head **ret_orphan_dir_bh) { struct inode *orphan_dir_inode; struct buffer_head *orphan_dir_bh = NULL; - int status = 0; - - status = ocfs2_blkno_stringify(blkno, name); - if (status < 0) { - mlog_errno(status); - return status; - } + int ret = 0; orphan_dir_inode = ocfs2_get_system_file_inode(osb, ORPHAN_DIR_SYSTEM_INODE, osb->slot_num); if (!orphan_dir_inode) { - status = -ENOENT; - mlog_errno(status); - return status; + ret = -ENOENT; + mlog_errno(ret); + return ret; } mutex_lock(&orphan_dir_inode->i_mutex); - status = ocfs2_inode_lock(orphan_dir_inode, &orphan_dir_bh, 1); - if (status < 0) { - mlog_errno(status); - goto leave; + ret = ocfs2_inode_lock(orphan_dir_inode, &orphan_dir_bh, 1); + if (ret < 0) { + mutex_unlock(&orphan_dir_inode->i_mutex); + iput(orphan_dir_inode); + + mlog_errno(ret); + return ret; } - status = ocfs2_prepare_dir_for_insert(osb, orphan_dir_inode, - orphan_dir_bh, name, - OCFS2_ORPHAN_NAMELEN, lookup); - if (status < 0) { - ocfs2_inode_unlock(orphan_dir_inode, 1); + *ret_orphan_dir = orphan_dir_inode; + *ret_orphan_dir_bh = orphan_dir_bh; - mlog_errno(status); - goto leave; + return 0; +} + +static int __ocfs2_prepare_orphan_dir(struct inode *orphan_dir_inode, + struct buffer_head *orphan_dir_bh, + u64 blkno, + char *name, + struct ocfs2_dir_lookup_result *lookup) +{ + int ret; + struct ocfs2_super *osb = OCFS2_SB(orphan_dir_inode->i_sb); + + ret = ocfs2_blkno_stringify(blkno, name); + if (ret < 0) { + mlog_errno(ret); + return ret; + } + + ret = ocfs2_prepare_dir_for_insert(osb, orphan_dir_inode, + orphan_dir_bh, name, + OCFS2_ORPHAN_NAMELEN, lookup); + if (ret < 0) { + mlog_errno(ret); + return ret; + } + + return 0; +} + +/** + * ocfs2_prepare_orphan_dir() - Prepare an orphan directory for + * insertion of an orphan. + * @osb: ocfs2 file system + * @ret_orphan_dir: Orphan dir inode - returned locked! + * @blkno: Actual block number of the inode to be inserted into orphan dir. + * @lookup: dir lookup result, to be passed back into functions like + * ocfs2_orphan_add + * + * Returns zero on success and the ret_orphan_dir, name and lookup + * fields will be populated. + * + * Returns non-zero on failure. + */ +static int ocfs2_prepare_orphan_dir(struct ocfs2_super *osb, + struct inode **ret_orphan_dir, + u64 blkno, + char *name, + struct ocfs2_dir_lookup_result *lookup) +{ + struct inode *orphan_dir_inode = NULL; + struct buffer_head *orphan_dir_bh = NULL; + int ret = 0; + + ret = ocfs2_lookup_lock_orphan_dir(osb, &orphan_dir_inode, + &orphan_dir_bh); + if (ret < 0) { + mlog_errno(ret); + return ret; + } + + ret = __ocfs2_prepare_orphan_dir(orphan_dir_inode, orphan_dir_bh, + blkno, name, lookup); + if (ret < 0) { + mlog_errno(ret); + goto out; } *ret_orphan_dir = orphan_dir_inode; -leave: - if (status) { +out: + brelse(orphan_dir_bh); + + if (ret) { + ocfs2_inode_unlock(orphan_dir_inode, 1); mutex_unlock(&orphan_dir_inode->i_mutex); iput(orphan_dir_inode); } - brelse(orphan_dir_bh); - - mlog_exit(status); - return status; + mlog_exit(ret); + return ret; } static int ocfs2_orphan_add(struct ocfs2_super *osb, -- cgit v0.10.2 From 97b8f4a9dfd932997677136e11980eb2fafea91d Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Fri, 13 Aug 2010 15:15:19 -0700 Subject: ocfs2: Fix orphan add in ocfs2_create_inode_in_orphan ocfs2_create_inode_in_orphan() is used by reflink to create the newly reflinked inode simultaneously in the orphan dir. This allows us to easily handle partially-reflinked files during recovery cleanup. We have a problem though - the orphan dir stringifies inode # to determine a unique name under which the orphan entry dirent can be created. Since ocfs2_create_inode_in_orphan() needs the space allocated in the orphan dir before it can allocate the inode, we currently call into the orphan code: /* * We give the orphan dir the root blkno to fake an orphan name, * and allocate enough space for our insertion. */ status = ocfs2_prepare_orphan_dir(osb, &orphan_dir, osb->root_blkno, orphan_name, &orphan_insert); Using osb->root_blkno might work fine on unindexed directories, but the orphan dir can have an index. When it has that index, the above code fails to allocate the proper index entry. Later, when we try to remove the file from the orphan dir (using the actual inode #), the reflink operation will fail. To fix this, I created a function ocfs2_alloc_orphaned_file() which uses the newly split out orphan and inode alloc code to figure out what the inode block number will be (once allocated) and then prepare the orphan dir from that data. Signed-off-by: Mark Fasheh Signed-off-by: Tao Ma diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 54c6298..a00dda2 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -2128,6 +2128,99 @@ leave: return status; } +/** + * ocfs2_prep_new_orphaned_file() - Prepare the orphan dir to recieve a newly + * allocated file. This is different from the typical 'add to orphan dir' + * operation in that the inode does not yet exist. This is a problem because + * the orphan dir stringifies the inode block number to come up with it's + * dirent. Obviously if the inode does not yet exist we have a chicken and egg + * problem. This function works around it by calling deeper into the orphan + * and suballoc code than other callers. Use this only by necessity. + * @dir: The directory which this inode will ultimately wind up under - not the + * orphan dir! + * @dir_bh: buffer_head the @dir inode block + * @orphan_name: string of length (CFS2_ORPHAN_NAMELEN + 1). Will be filled + * with the string to be used for orphan dirent. Pass back to the orphan dir + * code. + * @ret_orphan_dir: orphan dir inode returned to be passed back into orphan + * dir code. + * @ret_di_blkno: block number where the new inode will be allocated. + * @orphan_insert: Dir insert context to be passed back into orphan dir code. + * @ret_inode_ac: Inode alloc context to be passed back to the allocator. + * + * Returns zero on success and the ret_orphan_dir, name and lookup + * fields will be populated. + * + * Returns non-zero on failure. + */ +static int ocfs2_prep_new_orphaned_file(struct inode *dir, + struct buffer_head *dir_bh, + char *orphan_name, + struct inode **ret_orphan_dir, + u64 *ret_di_blkno, + struct ocfs2_dir_lookup_result *orphan_insert, + struct ocfs2_alloc_context **ret_inode_ac) +{ + int ret; + u64 di_blkno; + struct ocfs2_super *osb = OCFS2_SB(dir->i_sb); + struct inode *orphan_dir = NULL; + struct buffer_head *orphan_dir_bh = NULL; + struct ocfs2_alloc_context *inode_ac = NULL; + + ret = ocfs2_lookup_lock_orphan_dir(osb, &orphan_dir, &orphan_dir_bh); + if (ret < 0) { + mlog_errno(ret); + return ret; + } + + /* reserve an inode spot */ + ret = ocfs2_reserve_new_inode(osb, &inode_ac); + if (ret < 0) { + if (ret != -ENOSPC) + mlog_errno(ret); + goto out; + } + + ret = ocfs2_find_new_inode_loc(dir, dir_bh, inode_ac, + &di_blkno); + if (ret) { + mlog_errno(ret); + goto out; + } + + ret = __ocfs2_prepare_orphan_dir(orphan_dir, orphan_dir_bh, + di_blkno, orphan_name, orphan_insert); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + +out: + if (ret == 0) { + *ret_orphan_dir = orphan_dir; + *ret_di_blkno = di_blkno; + *ret_inode_ac = inode_ac; + /* + * orphan_name and orphan_insert are already up to + * date via prepare_orphan_dir + */ + } else { + /* Unroll reserve_new_inode* */ + if (inode_ac) + ocfs2_free_alloc_context(inode_ac); + + /* Unroll orphan dir locking */ + mutex_unlock(&orphan_dir->i_mutex); + ocfs2_inode_unlock(orphan_dir, 1); + iput(orphan_dir); + } + + brelse(orphan_dir_bh); + + return 0; +} + int ocfs2_create_inode_in_orphan(struct inode *dir, int mode, struct inode **new_inode) @@ -2143,6 +2236,8 @@ int ocfs2_create_inode_in_orphan(struct inode *dir, struct buffer_head *new_di_bh = NULL; struct ocfs2_alloc_context *inode_ac = NULL; struct ocfs2_dir_lookup_result orphan_insert = { NULL, }; + u64 uninitialized_var(di_blkno), suballoc_loc; + u16 suballoc_bit; status = ocfs2_inode_lock(dir, &parent_di_bh, 1); if (status < 0) { @@ -2151,20 +2246,9 @@ int ocfs2_create_inode_in_orphan(struct inode *dir, return status; } - /* - * We give the orphan dir the root blkno to fake an orphan name, - * and allocate enough space for our insertion. - */ - status = ocfs2_prepare_orphan_dir(osb, &orphan_dir, - osb->root_blkno, - orphan_name, &orphan_insert); - if (status < 0) { - mlog_errno(status); - goto leave; - } - - /* reserve an inode spot */ - status = ocfs2_reserve_new_inode(osb, &inode_ac); + status = ocfs2_prep_new_orphaned_file(dir, parent_di_bh, + orphan_name, &orphan_dir, + &di_blkno, &orphan_insert, &inode_ac); if (status < 0) { if (status != -ENOSPC) mlog_errno(status); @@ -2191,17 +2275,20 @@ int ocfs2_create_inode_in_orphan(struct inode *dir, goto leave; did_quota_inode = 1; - inode->i_nlink = 0; - /* do the real work now. */ - status = ocfs2_mknod_locked(osb, dir, inode, - 0, &new_di_bh, parent_di_bh, handle, - inode_ac); + status = ocfs2_claim_new_inode_at_loc(handle, dir, inode_ac, + &suballoc_loc, + &suballoc_bit, di_blkno); if (status < 0) { mlog_errno(status); goto leave; } - status = ocfs2_blkno_stringify(OCFS2_I(inode)->ip_blkno, orphan_name); + inode->i_nlink = 0; + /* do the real work now. */ + status = __ocfs2_mknod_locked(dir, inode, + 0, &new_di_bh, parent_di_bh, handle, + inode_ac, di_blkno, suballoc_loc, + suballoc_bit); if (status < 0) { mlog_errno(status); goto leave; -- cgit v0.10.2 From 4c25b93223340deff73381cc47f9244fb379a74d Mon Sep 17 00:00:00 2001 From: Clemens Ladisch Date: Tue, 7 Sep 2010 13:37:10 +0200 Subject: ALSA: virtuoso: work around missing reset in the Xonar DS Windows driver For the WM8776 chip, this driver uses a different sample format and more features than the Windows driver. When rebooting from Linux into Windows, the latter driver does not reset the chip but assumes all its registers have their default settings, so we get garbled sound or, if the output happened to be muted before rebooting, no sound. To make that driver happy, hook our driver's cleanup function into the shutdown notifier and ensure that the chip gets reset. Signed-off-by: Clemens Ladisch Reported-and-tested-by: Nathan Schagen Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/oxygen/oxygen.h b/sound/pci/oxygen/oxygen.h index 6147216..a3409ed 100644 --- a/sound/pci/oxygen/oxygen.h +++ b/sound/pci/oxygen/oxygen.h @@ -155,6 +155,7 @@ void oxygen_pci_remove(struct pci_dev *pci); int oxygen_pci_suspend(struct pci_dev *pci, pm_message_t state); int oxygen_pci_resume(struct pci_dev *pci); #endif +void oxygen_pci_shutdown(struct pci_dev *pci); /* oxygen_mixer.c */ diff --git a/sound/pci/oxygen/oxygen_lib.c b/sound/pci/oxygen/oxygen_lib.c index fad03d6..7e93cf8 100644 --- a/sound/pci/oxygen/oxygen_lib.c +++ b/sound/pci/oxygen/oxygen_lib.c @@ -519,16 +519,21 @@ static void oxygen_init(struct oxygen *chip) } } -static void oxygen_card_free(struct snd_card *card) +static void oxygen_shutdown(struct oxygen *chip) { - struct oxygen *chip = card->private_data; - spin_lock_irq(&chip->reg_lock); chip->interrupt_mask = 0; chip->pcm_running = 0; oxygen_write16(chip, OXYGEN_DMA_STATUS, 0); oxygen_write16(chip, OXYGEN_INTERRUPT_MASK, 0); spin_unlock_irq(&chip->reg_lock); +} + +static void oxygen_card_free(struct snd_card *card) +{ + struct oxygen *chip = card->private_data; + + oxygen_shutdown(chip); if (chip->irq >= 0) free_irq(chip->irq, chip); flush_scheduled_work(); @@ -778,3 +783,13 @@ int oxygen_pci_resume(struct pci_dev *pci) } EXPORT_SYMBOL(oxygen_pci_resume); #endif /* CONFIG_PM */ + +void oxygen_pci_shutdown(struct pci_dev *pci) +{ + struct snd_card *card = pci_get_drvdata(pci); + struct oxygen *chip = card->private_data; + + oxygen_shutdown(chip); + chip->model.cleanup(chip); +} +EXPORT_SYMBOL(oxygen_pci_shutdown); diff --git a/sound/pci/oxygen/virtuoso.c b/sound/pci/oxygen/virtuoso.c index f03a2f2..06c863e 100644 --- a/sound/pci/oxygen/virtuoso.c +++ b/sound/pci/oxygen/virtuoso.c @@ -95,6 +95,7 @@ static struct pci_driver xonar_driver = { .suspend = oxygen_pci_suspend, .resume = oxygen_pci_resume, #endif + .shutdown = oxygen_pci_shutdown, }; static int __init alsa_card_xonar_init(void) diff --git a/sound/pci/oxygen/xonar_wm87x6.c b/sound/pci/oxygen/xonar_wm87x6.c index dbc4b89..0b89932 100644 --- a/sound/pci/oxygen/xonar_wm87x6.c +++ b/sound/pci/oxygen/xonar_wm87x6.c @@ -193,6 +193,7 @@ static void xonar_ds_init(struct oxygen *chip) static void xonar_ds_cleanup(struct oxygen *chip) { xonar_disable_output(chip); + wm8776_write(chip, WM8776_RESET, 0); } static void xonar_ds_suspend(struct oxygen *chip) -- cgit v0.10.2 From fe6ce80ae25953d95ebaf9bce27b585218cda25c Mon Sep 17 00:00:00 2001 From: Clemens Ladisch Date: Tue, 7 Sep 2010 13:38:49 +0200 Subject: ALSA: virtuoso: fix setting of Xonar DS line-in/mic-in controls The Line and Mic inputs cannot be used at the same time, so the driver has to automatically disable one of them if both are set. However, it forgot to notify userspace about this change, so the mixer state would be inconsistent. To fix this, check if the other control gets muted, and send a notification event in this case. Signed-off-by: Clemens Ladisch Reported-and-tested-by: Nathan Schagen Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/oxygen/xonar_wm87x6.c b/sound/pci/oxygen/xonar_wm87x6.c index 0b89932..b82c1cf 100644 --- a/sound/pci/oxygen/xonar_wm87x6.c +++ b/sound/pci/oxygen/xonar_wm87x6.c @@ -53,6 +53,8 @@ struct xonar_wm87x6 { struct xonar_generic generic; u16 wm8776_regs[0x17]; u16 wm8766_regs[0x10]; + struct snd_kcontrol *line_adcmux_control; + struct snd_kcontrol *mic_adcmux_control; struct snd_kcontrol *lc_controls[13]; }; @@ -604,6 +606,7 @@ static int wm8776_input_mux_put(struct snd_kcontrol *ctl, { struct oxygen *chip = ctl->private_data; struct xonar_wm87x6 *data = chip->model_data; + struct snd_kcontrol *other_ctl; unsigned int mux_bit = ctl->private_value; u16 reg; int changed; @@ -611,8 +614,18 @@ static int wm8776_input_mux_put(struct snd_kcontrol *ctl, mutex_lock(&chip->mutex); reg = data->wm8776_regs[WM8776_ADCMUX]; if (value->value.integer.value[0]) { - reg &= ~0x003; reg |= mux_bit; + /* line-in and mic-in are exclusive */ + mux_bit ^= 3; + if (reg & mux_bit) { + reg &= ~mux_bit; + if (mux_bit == 1) + other_ctl = data->line_adcmux_control; + else + other_ctl = data->mic_adcmux_control; + snd_ctl_notify(chip->card, SNDRV_CTL_EVENT_MASK_VALUE, + &other_ctl->id); + } } else reg &= ~mux_bit; changed = reg != data->wm8776_regs[WM8776_ADCMUX]; @@ -964,7 +977,13 @@ static int xonar_ds_mixer_init(struct oxygen *chip) err = snd_ctl_add(chip->card, ctl); if (err < 0) return err; + if (!strcmp(ctl->id.name, "Line Capture Switch")) + data->line_adcmux_control = ctl; + else if (!strcmp(ctl->id.name, "Mic Capture Switch")) + data->mic_adcmux_control = ctl; } + if (!data->line_adcmux_control || !data->mic_adcmux_control) + return -ENXIO; BUILD_BUG_ON(ARRAY_SIZE(lc_controls) != ARRAY_SIZE(data->lc_controls)); for (i = 0; i < ARRAY_SIZE(lc_controls); ++i) { ctl = snd_ctl_new1(&lc_controls[i], chip); -- cgit v0.10.2 From 76195fb096ca6db2f8bbaffb96e3025aaf1649a0 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 8 Sep 2010 08:27:02 +0200 Subject: ALSA: usb - Release capture substream URBs properly Due to the wrong "return" in the loop, a capture substream won't be released at disconnection properly if the device is capture only and has no playback substream. This caused Oops occasionally at the device reconnection. Reported-by: Kim Minhyoung Cc: Signed-off-by: Takashi Iwai diff --git a/sound/usb/card.c b/sound/usb/card.c index 32e4be8..4eabafa 100644 --- a/sound/usb/card.c +++ b/sound/usb/card.c @@ -126,7 +126,7 @@ static void snd_usb_stream_disconnect(struct list_head *head) for (idx = 0; idx < 2; idx++) { subs = &as->substream[idx]; if (!subs->num_formats) - return; + continue; snd_usb_release_substream_urbs(subs, 1); subs->interface = -1; } -- cgit v0.10.2 From a769cbcf60cee51f4431c0938acd39e7e5b76b8d Mon Sep 17 00:00:00 2001 From: Brian Austin Date: Tue, 7 Sep 2010 14:36:22 -0500 Subject: ALSA: hda - Add errata initverb sequence for CS42xx codecs Add init verb sequence for errata ER880C3 http://www.cirrus.com/en/pubs/errata/ER880C3.pdf Signed-off-by: Brian Austin Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_cirrus.c b/sound/pci/hda/patch_cirrus.c index 4ef5efa..488fd9a 100644 --- a/sound/pci/hda/patch_cirrus.c +++ b/sound/pci/hda/patch_cirrus.c @@ -972,6 +972,53 @@ static struct hda_verb cs_coef_init_verbs[] = { {} /* terminator */ }; +/* Errata: CS4207 rev C0/C1/C2 Silicon + * + * http://www.cirrus.com/en/pubs/errata/ER880C3.pdf + * + * 6. At high temperature (TA > +85°C), the digital supply current (IVD) + * may be excessive (up to an additional 200 μA), which is most easily + * observed while the part is being held in reset (RESET# active low). + * + * Root Cause: At initial powerup of the device, the logic that drives + * the clock and write enable to the S/PDIF SRC RAMs is not properly + * initialized. + * Certain random patterns will cause a steady leakage current in those + * RAM cells. The issue will resolve once the SRCs are used (turned on). + * + * Workaround: The following verb sequence briefly turns on the S/PDIF SRC + * blocks, which will alleviate the issue. + */ + +static struct hda_verb cs_errata_init_verbs[] = { + {0x01, AC_VERB_SET_POWER_STATE, 0x00}, /* AFG: D0 */ + {0x11, AC_VERB_SET_PROC_STATE, 0x01}, /* VPW: processing on */ + + {0x11, AC_VERB_SET_COEF_INDEX, 0x0008}, + {0x11, AC_VERB_SET_PROC_COEF, 0x9999}, + {0x11, AC_VERB_SET_COEF_INDEX, 0x0017}, + {0x11, AC_VERB_SET_PROC_COEF, 0xa412}, + {0x11, AC_VERB_SET_COEF_INDEX, 0x0001}, + {0x11, AC_VERB_SET_PROC_COEF, 0x0009}, + + {0x07, AC_VERB_SET_POWER_STATE, 0x00}, /* S/PDIF Rx: D0 */ + {0x08, AC_VERB_SET_POWER_STATE, 0x00}, /* S/PDIF Tx: D0 */ + + {0x11, AC_VERB_SET_COEF_INDEX, 0x0017}, + {0x11, AC_VERB_SET_PROC_COEF, 0x2412}, + {0x11, AC_VERB_SET_COEF_INDEX, 0x0008}, + {0x11, AC_VERB_SET_PROC_COEF, 0x0000}, + {0x11, AC_VERB_SET_COEF_INDEX, 0x0001}, + {0x11, AC_VERB_SET_PROC_COEF, 0x0008}, + {0x11, AC_VERB_SET_PROC_STATE, 0x00}, + + {0x07, AC_VERB_SET_POWER_STATE, 0x03}, /* S/PDIF Rx: D3 */ + {0x08, AC_VERB_SET_POWER_STATE, 0x03}, /* S/PDIF Tx: D3 */ + /*{0x01, AC_VERB_SET_POWER_STATE, 0x03},*/ /* AFG: D3 This is already handled */ + + {} /* terminator */ +}; + /* SPDIF setup */ static void init_digital(struct hda_codec *codec) { @@ -991,6 +1038,9 @@ static int cs_init(struct hda_codec *codec) { struct cs_spec *spec = codec->spec; + /* init_verb sequence for C0/C1/C2 errata*/ + snd_hda_sequence_write(codec, cs_errata_init_verbs); + snd_hda_sequence_write(codec, cs_coef_init_verbs); if (spec->gpio_mask) { -- cgit v0.10.2 From 080dc7bc2562615a5be0a705a9d1a8c24eb198d4 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 8 Sep 2010 08:38:41 +0200 Subject: ALSA: hda - Enable PC-beep for EeePC with ALC269 codec EeePC 1001HAG has a similar problem like other ASUS machine, which doesn't set the codec SSID properly for indicating the beep capability. To enable PC-beep again, put this to the whitelist. Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 627bf99..bcbf916 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -5334,6 +5334,7 @@ static void fillup_priv_adc_nids(struct hda_codec *codec, hda_nid_t *nids, static struct snd_pci_quirk beep_white_list[] = { SND_PCI_QUIRK(0x1043, 0x829f, "ASUS", 1), + SND_PCI_QUIRK(0x1043, 0x83ce, "EeePC", 1), SND_PCI_QUIRK(0x8086, 0xd613, "Intel", 1), {} }; -- cgit v0.10.2 From 12e8ba25ef52f19e7a42e61aecb3c1fef83b2a82 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 7 Sep 2010 23:39:28 +0100 Subject: Revert "drm/i915: Allow LVDS on pipe A on gen4+" This reverts commit 0f3ee801b332d6ff22285386675fe5aaedf035c3. Enabling LVDS on pipe A was causing excessive wakeups on otherwise idle systems due to i915 interrupts. So restrict the LVDS to pipe B once more, whilst the issue is properly diagnosed. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=16307 Reported-and-tested-by: Enrico Bandiello Poked-by: Florian Mickler Signed-off-by: Chris Wilson Cc: Adam Jackson Cc: stable@kernel.org diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index b819c10..4fbb016 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -875,8 +875,6 @@ void intel_lvds_init(struct drm_device *dev) intel_encoder->clone_mask = (1 << INTEL_LVDS_CLONE_BIT); intel_encoder->crtc_mask = (1 << 1); - if (IS_I965G(dev)) - intel_encoder->crtc_mask |= (1 << 0); drm_encoder_helper_add(encoder, &intel_lvds_helper_funcs); drm_connector_helper_add(connector, &intel_lvds_connector_helper_funcs); connector->display_info.subpixel_order = SubPixelHorizontalRGB; -- cgit v0.10.2 From e4ee8dd8afcbcbe502fa8a3d3af6eb09c96dd806 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 8 Sep 2010 09:58:12 +0200 Subject: ALSA: msnd-classic: Fix invalid cfg parameter The driver doesn't probe the device properly because of left-over cfg[] that isn't used at all for msnd-classic device. This is only for msnd- pinnacle. Signed-off-by: Takashi Iwai diff --git a/sound/isa/msnd/msnd_pinnacle.c b/sound/isa/msnd/msnd_pinnacle.c index 5f3e684..91d6023 100644 --- a/sound/isa/msnd/msnd_pinnacle.c +++ b/sound/isa/msnd/msnd_pinnacle.c @@ -764,9 +764,9 @@ static long io[SNDRV_CARDS] = SNDRV_DEFAULT_PORT; static int irq[SNDRV_CARDS] = SNDRV_DEFAULT_IRQ; static long mem[SNDRV_CARDS] = SNDRV_DEFAULT_PORT; +#ifndef MSND_CLASSIC static long cfg[SNDRV_CARDS] = SNDRV_DEFAULT_PORT; -#ifndef MSND_CLASSIC /* Extra Peripheral Configuration (Default: Disable) */ static long ide_io0[SNDRV_CARDS] = SNDRV_DEFAULT_PORT; static long ide_io1[SNDRV_CARDS] = SNDRV_DEFAULT_PORT; @@ -894,7 +894,11 @@ static int __devinit snd_msnd_isa_probe(struct device *pdev, unsigned int idx) struct snd_card *card; struct snd_msnd *chip; - if (has_isapnp(idx) || cfg[idx] == SNDRV_AUTO_PORT) { + if (has_isapnp(idx) +#ifndef MSND_CLASSIC + || cfg[idx] == SNDRV_AUTO_PORT +#endif + ) { printk(KERN_INFO LOGNAME ": Assuming PnP mode\n"); return -ENODEV; } -- cgit v0.10.2 From c3add4b63438555d5e88c5893d238ab80d1f5959 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 8 Sep 2010 09:14:08 +0100 Subject: Revert "drm/i915: Warn if we run out of FIFO space for a mode" This reverts commit b9421ae8f30958deea98d71477b4a77a066856b4. This warning was so prelevant, even for apparently working machines, that it was just causing fear, anxiety and panic. The root cause still remains, so we will add some better debugging when we focus on fixing it. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=17021 Reported-by: Maciej Rutecki Signed-off-by: Chris Wilson diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index e0d1952..7c91030 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2767,14 +2767,8 @@ static unsigned long intel_calculate_wm(unsigned long clock_in_khz, /* Don't promote wm_size to unsigned... */ if (wm_size > (long)wm->max_wm) wm_size = wm->max_wm; - if (wm_size <= 0) { + if (wm_size <= 0) wm_size = wm->default_wm; - DRM_ERROR("Insufficient FIFO for plane, expect flickering:" - " entries required = %ld, available = %lu.\n", - entries_required + wm->guard_size, - wm->fifo_size); - } - return wm_size; } -- cgit v0.10.2 From 27f7ad53829f79e799a253285318bff79ece15bd Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 6 Sep 2010 09:13:45 +0200 Subject: ALSA: seq/oss - Fix double-free at error path of snd_seq_oss_open() The error handling in snd_seq_oss_open() has several bad codes that do dereferecing released pointers and double-free of kmalloc'ed data. The object dp is release in free_devinfo() that is called via private_free callback. The rest shouldn't touch this object any more. The patch changes delete_port() to call kfree() in any case, and gets rid of unnecessary calls of destructors in snd_seq_oss_open(). Fixes CVE-2010-3080. Reported-and-tested-by: Tavis Ormandy Cc: Signed-off-by: Takashi Iwai diff --git a/sound/core/seq/oss/seq_oss_init.c b/sound/core/seq/oss/seq_oss_init.c index 6857122..69cd7b3 100644 --- a/sound/core/seq/oss/seq_oss_init.c +++ b/sound/core/seq/oss/seq_oss_init.c @@ -281,13 +281,10 @@ snd_seq_oss_open(struct file *file, int level) return 0; _error: - snd_seq_oss_writeq_delete(dp->writeq); - snd_seq_oss_readq_delete(dp->readq); snd_seq_oss_synth_cleanup(dp); snd_seq_oss_midi_cleanup(dp); - delete_port(dp); delete_seq_queue(dp->queue); - kfree(dp); + delete_port(dp); return rc; } @@ -350,8 +347,10 @@ create_port(struct seq_oss_devinfo *dp) static int delete_port(struct seq_oss_devinfo *dp) { - if (dp->port < 0) + if (dp->port < 0) { + kfree(dp); return 0; + } debug_printk(("delete_port %i\n", dp->port)); return snd_seq_event_port_detach(dp->cseq, dp->port); -- cgit v0.10.2 From 1d220334d6a8a711149234dc5f98d34ae02226b8 Mon Sep 17 00:00:00 2001 From: Anton Vorontsov Date: Wed, 8 Sep 2010 00:10:26 +0400 Subject: apm_power: Add missing break statement The missing break statement causes wrong capacity calculation for batteries that report energy. Reported-by: d binderman Cc: Signed-off-by: Anton Vorontsov diff --git a/drivers/power/apm_power.c b/drivers/power/apm_power.c index 936bae5..dc628cb 100644 --- a/drivers/power/apm_power.c +++ b/drivers/power/apm_power.c @@ -233,6 +233,7 @@ static int calculate_capacity(enum apm_source source) empty_design_prop = POWER_SUPPLY_PROP_ENERGY_EMPTY_DESIGN; now_prop = POWER_SUPPLY_PROP_ENERGY_NOW; avg_prop = POWER_SUPPLY_PROP_ENERGY_AVG; + break; case SOURCE_VOLTAGE: full_prop = POWER_SUPPLY_PROP_VOLTAGE_MAX; empty_prop = POWER_SUPPLY_PROP_VOLTAGE_MIN; -- cgit v0.10.2 From 0485e18bc4112a3b548baa314c24bfbece4d156b Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 5 Sep 2010 12:19:50 +0100 Subject: Revert "[ARM] pxa: remove now unnecessary dma_needs_bounce()" This reverts commit 4fa5518, which causes a compilation regression for IXP4xx platforms. Reported-by: Richard Cochran Acked-by: Eric Miao Signed-off-by: Russell King diff --git a/arch/arm/common/it8152.c b/arch/arm/common/it8152.c index 6c09135..7974baa 100644 --- a/arch/arm/common/it8152.c +++ b/arch/arm/common/it8152.c @@ -263,6 +263,14 @@ static int it8152_pci_platform_notify_remove(struct device *dev) return 0; } +int dma_needs_bounce(struct device *dev, dma_addr_t dma_addr, size_t size) +{ + dev_dbg(dev, "%s: dma_addr %08x, size %08x\n", + __func__, dma_addr, size); + return (dev->bus == &pci_bus_type) && + ((dma_addr + size - PHYS_OFFSET) >= SZ_64M); +} + int __init it8152_pci_setup(int nr, struct pci_sys_data *sys) { it8152_io.start = IT8152_IO_BASE + 0x12000; diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h index c226fe1..c568da7 100644 --- a/arch/arm/include/asm/dma-mapping.h +++ b/arch/arm/include/asm/dma-mapping.h @@ -288,15 +288,7 @@ extern void dmabounce_unregister_dev(struct device *); * DMA access and 1 if the buffer needs to be bounced. * */ -#ifdef CONFIG_SA1111 extern int dma_needs_bounce(struct device*, dma_addr_t, size_t); -#else -static inline int dma_needs_bounce(struct device *dev, dma_addr_t addr, - size_t size) -{ - return 0; -} -#endif /* * The DMA API, implemented by dmabounce.c. See below for descriptions. -- cgit v0.10.2 From a387f0f5409276c5cf75eecb61ef6e6896f851b4 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Fri, 3 Sep 2010 17:14:54 +0100 Subject: ARM: 6359/1: ep93xx: move clock initialization earlier Commit 7cfe24947 ("ARM: AMBA: Add pclk support to AMBA bus infrastructure") changed AMBA bus to handle the PCLK automatically. However, in EP93xx clock initialization is arch_initcall which is done later than AMBA device identification. This causes amba_get_enable_pclk() to fail resulting device where UARTs are not functional. So change ep93xx_clock_init() to be postcore_initcall. Signed-off-by: Mika Westerberg Acked-by: H Hartley Sweeten Signed-off-by: Russell King diff --git a/arch/arm/mach-ep93xx/clock.c b/arch/arm/mach-ep93xx/clock.c index 8bf3cec..4566bd1 100644 --- a/arch/arm/mach-ep93xx/clock.c +++ b/arch/arm/mach-ep93xx/clock.c @@ -560,4 +560,4 @@ static int __init ep93xx_clock_init(void) clkdev_add_table(clocks, ARRAY_SIZE(clocks)); return 0; } -arch_initcall(ep93xx_clock_init); +postcore_initcall(ep93xx_clock_init); -- cgit v0.10.2 From cec15a0ece19116b6c2c53fedf9696c20124d491 Mon Sep 17 00:00:00 2001 From: Roland Baum Date: Wed, 8 Sep 2010 14:27:55 +0200 Subject: HID: add device ID for new Asus Multitouch Controller MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The following patch instructs usbhid/hid-mosart to handle a new multitouch controller, built-in by some Asus EeePC T101MT models. Signed-off-by: Roland Baum Tested-by: Roland Baum Acked-by: Stéphane Chatty CC: Stéphane Chatty Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index b2b3bb9..2baf959 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -1285,6 +1285,7 @@ static const struct hid_device_id hid_blacklist[] = { { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_ALU_WIRELESS_2009_JIS) }, { HID_USB_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_FOUNTAIN_TP_ONLY) }, { HID_USB_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_GEYSER1_TP_ONLY) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ASUS, USB_DEVICE_ID_ASUSTEK_MULTITOUCH_YFO) }, { HID_USB_DEVICE(USB_VENDOR_ID_BELKIN, USB_DEVICE_ID_FLIP_KVM) }, { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE) }, { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE_2) }, diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 7c69f1e..765a4f5 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -105,6 +105,7 @@ #define USB_VENDOR_ID_ASUS 0x0486 #define USB_DEVICE_ID_ASUS_T91MT 0x0185 +#define USB_DEVICE_ID_ASUSTEK_MULTITOUCH_YFO 0x0186 #define USB_VENDOR_ID_ASUSTEK 0x0b05 #define USB_DEVICE_ID_ASUSTEK_LCM 0x1726 diff --git a/drivers/hid/hid-mosart.c b/drivers/hid/hid-mosart.c index e91437c..ac5421d 100644 --- a/drivers/hid/hid-mosart.c +++ b/drivers/hid/hid-mosart.c @@ -239,6 +239,7 @@ static void mosart_remove(struct hid_device *hdev) static const struct hid_device_id mosart_devices[] = { { HID_USB_DEVICE(USB_VENDOR_ID_ASUS, USB_DEVICE_ID_ASUS_T91MT) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ASUS, USB_DEVICE_ID_ASUSTEK_MULTITOUCH_YFO) }, { } }; MODULE_DEVICE_TABLE(hid, mosart_devices); -- cgit v0.10.2 From eaca1386207a9e0314647d3a88967acb17cc30e3 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Wed, 8 Sep 2010 14:31:47 +0200 Subject: HID: fixup blacklist entry for Asus T91MT The device is handled by hid-mosart driver, and therefore should be present in hid_blacklist[], not hid_ignore_list[]. Cc: Stephane Chatty Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 2baf959..3f72924 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -1285,6 +1285,7 @@ static const struct hid_device_id hid_blacklist[] = { { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_ALU_WIRELESS_2009_JIS) }, { HID_USB_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_FOUNTAIN_TP_ONLY) }, { HID_USB_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_GEYSER1_TP_ONLY) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ASUS, USB_DEVICE_ID_ASUS_T91MT) }, { HID_USB_DEVICE(USB_VENDOR_ID_ASUS, USB_DEVICE_ID_ASUSTEK_MULTITOUCH_YFO) }, { HID_USB_DEVICE(USB_VENDOR_ID_BELKIN, USB_DEVICE_ID_FLIP_KVM) }, { HID_USB_DEVICE(USB_VENDOR_ID_BTC, USB_DEVICE_ID_BTC_EMPREX_REMOTE) }, @@ -1580,7 +1581,6 @@ static const struct hid_device_id hid_ignore_list[] = { { HID_USB_DEVICE(USB_VENDOR_ID_AIPTEK, USB_DEVICE_ID_AIPTEK_24) }, { HID_USB_DEVICE(USB_VENDOR_ID_AIRCABLE, USB_DEVICE_ID_AIRCABLE1) }, { HID_USB_DEVICE(USB_VENDOR_ID_ALCOR, USB_DEVICE_ID_ALCOR_USBRS232) }, - { HID_USB_DEVICE(USB_VENDOR_ID_ASUS, USB_DEVICE_ID_ASUS_T91MT)}, { HID_USB_DEVICE(USB_VENDOR_ID_ASUSTEK, USB_DEVICE_ID_ASUSTEK_LCM)}, { HID_USB_DEVICE(USB_VENDOR_ID_ASUSTEK, USB_DEVICE_ID_ASUSTEK_LCM2)}, { HID_USB_DEVICE(USB_VENDOR_ID_AVERMEDIA, USB_DEVICE_ID_AVER_FM_MR800) }, -- cgit v0.10.2 From 122661b67899980f1372812d907e73ebcfb3d037 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 8 Sep 2010 14:57:04 +0200 Subject: ALSA: hda - Fix wrong HP pin detection in snd_hda_parse_pin_def_config() snd_hda_parse_pin_def_config() has some workaround for re-assigning some pins declared as headphones to line-outs. This didn't work properly for some cases because it used memmove() stupidly wrongly. Reference: Novell bnc#637263 https://bugzilla.novell.com/show_bug.cgi?id=637263 Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/hda_codec.c b/sound/pci/hda/hda_codec.c index 3827092..1482921 100644 --- a/sound/pci/hda/hda_codec.c +++ b/sound/pci/hda/hda_codec.c @@ -4536,7 +4536,7 @@ int snd_hda_parse_pin_def_config(struct hda_codec *codec, cfg->hp_outs--; memmove(cfg->hp_pins + i, cfg->hp_pins + i + 1, sizeof(cfg->hp_pins[0]) * (cfg->hp_outs - i)); - memmove(sequences_hp + i - 1, sequences_hp + i, + memmove(sequences_hp + i, sequences_hp + i + 1, sizeof(sequences_hp[0]) * (cfg->hp_outs - i)); } } -- cgit v0.10.2 From febc88c5948f81114f64c3412011d695aecae233 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 7 Sep 2010 14:46:37 +0200 Subject: semaphore: Add DEFINE_SEMAPHORE The full cleanup of init_MUTEX[_LOCKED] and DECLARE_MUTEX has not been done. Some of the users are real semaphores and we should name them as such instead of confusing everyone with "MUTEX". Provide the infrastructure to get finally rid of init_MUTEX[_LOCKED] and DECLARE_MUTEX. Signed-off-by: Thomas Gleixner Cc: Peter Zijlstra Cc: Christoph Hellwig LKML-Reference: <20100907125054.795929962@linutronix.de> diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h index 7415839..5310d27 100644 --- a/include/linux/semaphore.h +++ b/include/linux/semaphore.h @@ -26,6 +26,9 @@ struct semaphore { .wait_list = LIST_HEAD_INIT((name).wait_list), \ } +#define DEFINE_SEMAPHORE(name) \ + struct semaphore name = __SEMAPHORE_INITIALIZER(name, 1) + #define DECLARE_MUTEX(name) \ struct semaphore name = __SEMAPHORE_INITIALIZER(name, 1) -- cgit v0.10.2 From 9c55cb12c1c172e2d51e85fbb5a4796ca86b77e7 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 8 Sep 2010 11:20:37 -0400 Subject: tracing: Do not allow llseek to set_ftrace_filter Reading the file set_ftrace_filter does three things. 1) shows whether or not filters are set for the function tracer 2) shows what functions are set for the function tracer 3) shows what triggers are set on any functions 3 is independent from 1 and 2. The way this file currently works is that it is a state machine, and as you read it, it may change state. But this assumption breaks when you use lseek() on the file. The state machine gets out of sync and the t_show() may use the wrong pointer and cause a kernel oops. Luckily, this will only kill the app that does the lseek, but the app dies while holding a mutex. This prevents anyone else from using the set_ftrace_filter file (or any other function tracing file for that matter). A real fix for this is to rewrite the code, but that is too much for a -rc release or stable. This patch simply disables llseek on the set_ftrace_filter() file for now, and we can do the proper fix for the next major release. Reported-by: Robert Swiecki Cc: Chris Wright Cc: Tavis Ormandy Cc: Eugene Teo Cc: vendor-sec@lst.de Cc: Signed-off-by: Steven Rostedt diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 7cb1f45..83a16e9 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -2416,7 +2416,7 @@ static const struct file_operations ftrace_filter_fops = { .open = ftrace_filter_open, .read = seq_read, .write = ftrace_filter_write, - .llseek = ftrace_regex_lseek, + .llseek = no_llseek, .release = ftrace_filter_release, }; -- cgit v0.10.2 From 61a527362234ac3352a91ac67c50c6f7cd248eb1 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 27 Aug 2010 20:38:46 +0900 Subject: tracing/kprobe: Fix a memory leak in error case Fix a memory leak which happens when a field name conflicts with others. In error case, free_trace_probe() will free all arguments until nr_args, so this increments nr_args the begining of the loop instead of the end. Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Mathieu Desnoyers LKML-Reference: <20100827113846.22882.12670.stgit@ltc236.sdl.hitachi.co.jp> Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 8b27c98..0116c03 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -992,6 +992,9 @@ static int create_trace_probe(int argc, char **argv) /* parse arguments */ ret = 0; for (i = 0; i < argc && i < MAX_TRACE_ARGS; i++) { + /* Increment count for freeing args in error case */ + tp->nr_args++; + /* Parse argument name */ arg = strchr(argv[i], '='); if (arg) @@ -1021,11 +1024,8 @@ static int create_trace_probe(int argc, char **argv) ret = parse_probe_arg(arg, tp, &tp->args[i], is_return); if (ret) { pr_info("Parse error at argument%d. (%d)\n", i, ret); - kfree(tp->args[i].name); goto error; } - - tp->nr_args++; } ret = register_trace_probe(tp); -- cgit v0.10.2 From 04ddd04b044d8896a4f8a921b23ba09d365df196 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 27 Aug 2010 20:38:53 +0900 Subject: perf probe: Fix return probe support Fix a bug to support %return probe syntax again. Previous commit 4235b04 has a bug which disables the %return syntax on perf probe. Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Ingo Molnar Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker LKML-Reference: <20100827113852.22882.87447.stgit@ltc236.sdl.hitachi.co.jp> Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index e72f05c..fcc16e4 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -1539,6 +1539,7 @@ static int convert_to_probe_trace_events(struct perf_probe_event *pev, goto error; } tev->point.offset = pev->point.offset; + tev->point.retprobe = pev->point.retprobe; tev->nargs = pev->nargs; if (tev->nargs) { tev->args = zalloc(sizeof(struct probe_trace_arg) diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 5251366..9b0e1b1 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -783,6 +783,16 @@ static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) /* This function has no name. */ tev->point.offset = (unsigned long)pf->addr; + /* Return probe must be on the head of a subprogram */ + if (pf->pev->point.retprobe) { + if (tev->point.offset != 0) { + pr_warning("Return probe must be on the head of" + " a real function\n"); + return -EINVAL; + } + tev->point.retprobe = true; + } + pr_debug("Probe point found: %s+%lu\n", tev->point.symbol, tev->point.offset); -- cgit v0.10.2 From 367e94c10092469c896a226a77ef13cf6da757e4 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 27 Aug 2010 20:38:59 +0900 Subject: perf probe: Fix handling of arguments names Don't make argument names from raw parameters (means the parameters are written in kprobe-tracer syntax), because the argument syntax may include special characters. Just leave it, then kprobe-tracer gives a new name. Reported-by: Srikar Dronamraju Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Ingo Molnar Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker LKML-Reference: <20100827113859.22882.75598.stgit@ltc236.sdl.hitachi.co.jp> Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 9b0e1b1..32b81f7 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -686,6 +686,25 @@ static int find_variable(Dwarf_Die *sp_die, struct probe_finder *pf) char buf[32], *ptr; int ret, nscopes; + if (!is_c_varname(pf->pvar->var)) { + /* Copy raw parameters */ + pf->tvar->value = strdup(pf->pvar->var); + if (pf->tvar->value == NULL) + return -ENOMEM; + if (pf->pvar->type) { + pf->tvar->type = strdup(pf->pvar->type); + if (pf->tvar->type == NULL) + return -ENOMEM; + } + if (pf->pvar->name) { + pf->tvar->name = strdup(pf->pvar->name); + if (pf->tvar->name == NULL) + return -ENOMEM; + } else + pf->tvar->name = NULL; + return 0; + } + if (pf->pvar->name) pf->tvar->name = strdup(pf->pvar->name); else { @@ -700,19 +719,6 @@ static int find_variable(Dwarf_Die *sp_die, struct probe_finder *pf) if (pf->tvar->name == NULL) return -ENOMEM; - if (!is_c_varname(pf->pvar->var)) { - /* Copy raw parameters */ - pf->tvar->value = strdup(pf->pvar->var); - if (pf->tvar->value == NULL) - return -ENOMEM; - if (pf->pvar->type) { - pf->tvar->type = strdup(pf->pvar->type); - if (pf->tvar->type == NULL) - return -ENOMEM; - } - return 0; - } - pr_debug("Searching '%s' variable in context.\n", pf->pvar->var); /* Search child die for local variables and parameters. */ -- cgit v0.10.2 From aba91595cfcebd193425e20aabc407531526a1c5 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 27 Aug 2010 20:39:06 +0900 Subject: tracing/kprobes: Fix handling of argument names Set "argN" name for each argument automatically if it has no specified name. Since dynamic trace event(kprobe_events) accepts special characters for its argument, its format can show those special characters (e.g. '$', '%', '+'). However, perf can't parse those format because of the character (especially '%') mess up the format. This sets "argX" name for those arguments if user omitted the argument names. E.g. # echo 'p do_fork %ax IP=%ip $stack' > tracing/kprobe_events # cat tracing/kprobe_events p:kprobes/p_do_fork_0 do_fork arg1=%ax IP=%ip arg3=$stack Reported-by: Srikar Dronamraju Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Mathieu Desnoyers LKML-Reference: <20100827113906.22882.59312.stgit@ltc236.sdl.hitachi.co.jp> Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 0116c03..a39251e 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -997,15 +997,18 @@ static int create_trace_probe(int argc, char **argv) /* Parse argument name */ arg = strchr(argv[i], '='); - if (arg) + if (arg) { *arg++ = '\0'; - else + tp->args[i].name = kstrdup(argv[i], GFP_KERNEL); + } else { arg = argv[i]; + /* If argument name is omitted, set "argN" */ + snprintf(buf, MAX_EVENT_NAME_LEN, "arg%d", i + 1); + tp->args[i].name = kstrdup(buf, GFP_KERNEL); + } - tp->args[i].name = kstrdup(argv[i], GFP_KERNEL); if (!tp->args[i].name) { - pr_info("Failed to allocate argument%d name '%s'.\n", - i, argv[i]); + pr_info("Failed to allocate argument[%d] name.\n", i); ret = -ENOMEM; goto error; } @@ -1014,7 +1017,7 @@ static int create_trace_probe(int argc, char **argv) *tmp = '_'; /* convert : to _ */ if (conflict_field_name(tp->args[i].name, tp->args, i)) { - pr_info("Argument%d name '%s' conflicts with " + pr_info("Argument[%d] name '%s' conflicts with " "another field.\n", i, argv[i]); ret = -EINVAL; goto error; @@ -1023,7 +1026,7 @@ static int create_trace_probe(int argc, char **argv) /* Parse fetch argument */ ret = parse_probe_arg(arg, tp, &tp->args[i], is_return); if (ret) { - pr_info("Parse error at argument%d. (%d)\n", i, ret); + pr_info("Parse error at argument[%d]. (%d)\n", i, ret); goto error; } } -- cgit v0.10.2 From da34634fd39958725310d2c30c9b4543945f968b Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 27 Aug 2010 20:39:12 +0900 Subject: tracing/kprobe: Fix handling of C-unlike argument names Check the argument name whether it is invalid (not C-like symbol name). This makes event format simple. Reported-by: Srikar Dronamraju Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Mathieu Desnoyers LKML-Reference: <20100827113912.22882.62313.stgit@ltc236.sdl.hitachi.co.jp> Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index a39251e..544301d 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -514,8 +514,8 @@ static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs); static int kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs); -/* Check the name is good for event/group */ -static int check_event_name(const char *name) +/* Check the name is good for event/group/fields */ +static int is_good_name(const char *name) { if (!isalpha(*name) && *name != '_') return 0; @@ -557,7 +557,7 @@ static struct trace_probe *alloc_trace_probe(const char *group, else tp->rp.kp.pre_handler = kprobe_dispatcher; - if (!event || !check_event_name(event)) { + if (!event || !is_good_name(event)) { ret = -EINVAL; goto error; } @@ -567,7 +567,7 @@ static struct trace_probe *alloc_trace_probe(const char *group, if (!tp->call.name) goto error; - if (!group || !check_event_name(group)) { + if (!group || !is_good_name(group)) { ret = -EINVAL; goto error; } @@ -883,7 +883,7 @@ static int create_trace_probe(int argc, char **argv) int i, ret = 0; int is_return = 0, is_delete = 0; char *symbol = NULL, *event = NULL, *group = NULL; - char *arg, *tmp; + char *arg; unsigned long offset = 0; void *addr = NULL; char buf[MAX_EVENT_NAME_LEN]; @@ -1012,9 +1012,13 @@ static int create_trace_probe(int argc, char **argv) ret = -ENOMEM; goto error; } - tmp = strchr(tp->args[i].name, ':'); - if (tmp) - *tmp = '_'; /* convert : to _ */ + + if (!is_good_name(tp->args[i].name)) { + pr_info("Invalid argument[%d] name: %s\n", + i, tp->args[i].name); + ret = -EINVAL; + goto error; + } if (conflict_field_name(tp->args[i].name, tp->args, i)) { pr_info("Argument[%d] name '%s' conflicts with " -- cgit v0.10.2 From 2be23c475af8ae4e25f8bab08d815b17593bd547 Mon Sep 17 00:00:00 2001 From: Russell King Date: Wed, 8 Sep 2010 16:27:56 +0100 Subject: ARM: Ensure PTE modifications via dma_alloc_coherent are visible Dave Hylands reports: | We've observed a problem with dma_alloc_writecombine when the system | is under heavy load (heavy bus traffic). We've managed to reduce the | problem to the following snippet, which is run from a kthread in a | continuous loop: | | void *virtAddr; | dma_addr_t physAddr; | unsigned int numBytes = 256; | | for (;;) { | virtAddr = dma_alloc_writecombine(NULL, | numBytes, &physAddr, GFP_KERNEL); | if (virtAddr == NULL) { | printk(KERN_ERR "Running out of memory\n"); | break; | } | | /* access DMA memory allocated */ | tmp = virtAddr; | *tmp = 0x77; | | /* free DMA memory */ | dma_free_writecombine(NULL, | numBytes, virtAddr, physAddr); | | ...sleep here... | } | | By itself, the code will run forever with no issues. However, as we | increase our bus traffic (typically using DMA) then the *tmp = 0x77 | line will eventually cause a page fault. If we add a small delay (a | few microseconds) before the *tmp = 0x77, then we don't see a page | fault, even under heavy load. A dsb() is required after modifying the PTE entries to ensure that they will always be visible. Add this dsb(). Reported-by: Dave Hylands Tested-by: Dave Hylands Signed-off-by: Russell King diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index c704eed..4bc43e5 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -229,6 +229,8 @@ __dma_alloc_remap(struct page *page, size_t size, gfp_t gfp, pgprot_t prot) } } while (size -= PAGE_SIZE); + dsb(); + return (void *)c->vm_start; } return NULL; -- cgit v0.10.2 From d2a787fc57142ba8757142f1569603b4d0b714a4 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 7 Sep 2010 11:29:17 +0100 Subject: spi/spi_s3c64xx: Move to subsys_initcall() Allow the use of the S3C64xx SPI controller with things like PMICs by moving the init up to subsys_initcall(). Signed-off-by: Mark Brown Acked-by: Jassi Brar Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index 7e627f7..a4c4805 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -1175,7 +1175,7 @@ static int __init s3c64xx_spi_init(void) { return platform_driver_probe(&s3c64xx_spi_driver, s3c64xx_spi_probe); } -module_init(s3c64xx_spi_init); +subsys_initcall(s3c64xx_spi_init); static void __exit s3c64xx_spi_exit(void) { -- cgit v0.10.2 From 9d8f86b56093d7b06d81d4063d5b9a4cbf887e75 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 7 Sep 2010 16:37:52 +0100 Subject: spi/spi_s3c64xx: Increase dead reckoning time in wait_for_xfer() For small transfers at high speeds the expected transfer time can easily be well under 1ms, causing the delay in wait_for_xfer() to be only the dead reckoning fudge factor of 5ms currently included. Experiments on some of my systems shows that this is marginal for some transfers so double it to 10ms. Signed-off-by: Mark Brown Acked-by: Jassi Brar Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index a4c4805..8130f02 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -320,7 +320,7 @@ static int wait_for_xfer(struct s3c64xx_spi_driver_data *sdd, /* millisecs to xfer 'len' bytes @ 'cur_speed' */ ms = xfer->len * 8 * 1000 / sdd->cur_speed; - ms += 5; /* some tolerance */ + ms += 10; /* some tolerance */ if (dma_mode) { val = msecs_to_jiffies(ms) + 10; -- cgit v0.10.2 From cbcc062abb16d39839b3d8d4e3d20360fc21eb58 Mon Sep 17 00:00:00 2001 From: Yong Wang Date: Tue, 7 Sep 2010 15:27:27 +0800 Subject: spi/dw_spi: Allow interrupt sharing Allow interrupt sharing since exclusive interrupt line for DW SPI controller is not provided on every platform. Signed-off-by: Yong Wang Signed-off-by: Grant Likely diff --git a/drivers/spi/dw_spi.c b/drivers/spi/dw_spi.c index d256cb0..11fbbf6f 100644 --- a/drivers/spi/dw_spi.c +++ b/drivers/spi/dw_spi.c @@ -396,6 +396,11 @@ static irqreturn_t interrupt_transfer(struct dw_spi *dws) static irqreturn_t dw_spi_irq(int irq, void *dev_id) { struct dw_spi *dws = dev_id; + u16 irq_status, irq_mask = 0x3f; + + irq_status = dw_readw(dws, isr) & irq_mask; + if (!irq_status) + return IRQ_NONE; if (!dws->cur_msg) { spi_mask_intr(dws, SPI_INT_TXEI); @@ -883,7 +888,7 @@ int __devinit dw_spi_add_host(struct dw_spi *dws) dws->dma_inited = 0; dws->dma_addr = (dma_addr_t)(dws->paddr + 0x60); - ret = request_irq(dws->irq, dw_spi_irq, 0, + ret = request_irq(dws->irq, dw_spi_irq, IRQF_SHARED, "dw_spi", dws); if (ret < 0) { dev_err(&master->dev, "can not get IRQ\n"); -- cgit v0.10.2 From e3e55ff5854655d8723ad8b307f02515aecc3df5 Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Tue, 7 Sep 2010 15:52:06 +0800 Subject: spi/dw_spi: clean the cs_control code commit 052dc7c45i "spi/dw_spi: conditional transfer mode change" introduced cs_control code, which has a bug by using bit offset for spi mode to set transfer mode in control register. Also it forces devices who don't need cs_control to re-configure the control registers for each spi transfer. This patch will fix them Signed-off-by: Feng Tang Signed-off-by: Grant Likely diff --git a/drivers/spi/dw_spi.c b/drivers/spi/dw_spi.c index 11fbbf6f..5624785 100644 --- a/drivers/spi/dw_spi.c +++ b/drivers/spi/dw_spi.c @@ -181,10 +181,6 @@ static void flush(struct dw_spi *dws) wait_till_not_busy(dws); } -static void null_cs_control(u32 command) -{ -} - static int null_writer(struct dw_spi *dws) { u8 n_bytes = dws->n_bytes; @@ -322,7 +318,7 @@ static void giveback(struct dw_spi *dws) struct spi_transfer, transfer_list); - if (!last_transfer->cs_change) + if (!last_transfer->cs_change && dws->cs_control) dws->cs_control(MRST_SPI_DEASSERT); msg->state = NULL; @@ -549,13 +545,13 @@ static void pump_transfers(unsigned long data) */ if (dws->cs_control) { if (dws->rx && dws->tx) - chip->tmode = 0x00; + chip->tmode = SPI_TMOD_TR; else if (dws->rx) - chip->tmode = 0x02; + chip->tmode = SPI_TMOD_RO; else - chip->tmode = 0x01; + chip->tmode = SPI_TMOD_TO; - cr0 &= ~(0x3 << SPI_MODE_OFFSET); + cr0 &= ~SPI_TMOD_MASK; cr0 |= (chip->tmode << SPI_TMOD_OFFSET); } @@ -704,9 +700,6 @@ static int dw_spi_setup(struct spi_device *spi) chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL); if (!chip) return -ENOMEM; - - chip->cs_control = null_cs_control; - chip->enable_dma = 0; } /* diff --git a/include/linux/spi/dw_spi.h b/include/linux/spi/dw_spi.h index cc813f9..c91302f 100644 --- a/include/linux/spi/dw_spi.h +++ b/include/linux/spi/dw_spi.h @@ -14,7 +14,9 @@ #define SPI_MODE_OFFSET 6 #define SPI_SCPH_OFFSET 6 #define SPI_SCOL_OFFSET 7 + #define SPI_TMOD_OFFSET 8 +#define SPI_TMOD_MASK (0x3 << SPI_TMOD_OFFSET) #define SPI_TMOD_TR 0x0 /* xmit & recv */ #define SPI_TMOD_TO 0x1 /* xmit only */ #define SPI_TMOD_RO 0x2 /* recv only */ -- cgit v0.10.2 From 915b96191f01d53e30d74083dcf4aebfb5b7ce10 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sat, 4 Sep 2010 09:13:17 +0200 Subject: powerpc/5200: efika.c: Add of_node_put to avoid memory leak This function is implemented as though the function of_get_next_child does not increment the reference count of its result, but actually it does. Thus the patch adds of_node_put in error handling code and drops a call to of_node_get. The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @r exists@ local idexpression x; expression E1; position p1,p2; @@ x@p1 = of_get_next_child(...); ... when != x = E1 of_node_get@p2(x) @script:python@ p1 << r.p1; p2 << r.p2; @@ cocci.print_main("call",p1) cocci.print_secs("get",p2) // Signed-off-by: Julia Lawall Signed-off-by: Grant Likely diff --git a/arch/powerpc/platforms/52xx/efika.c b/arch/powerpc/platforms/52xx/efika.c index 45c0cb9..18c1048 100644 --- a/arch/powerpc/platforms/52xx/efika.c +++ b/arch/powerpc/platforms/52xx/efika.c @@ -99,7 +99,7 @@ static void __init efika_pcisetup(void) if (bus_range == NULL || len < 2 * sizeof(int)) { printk(KERN_WARNING EFIKA_PLATFORM_NAME ": Can't get bus-range for %s\n", pcictrl->full_name); - return; + goto out_put; } if (bus_range[1] == bus_range[0]) @@ -111,12 +111,12 @@ static void __init efika_pcisetup(void) printk(" controlled by %s\n", pcictrl->full_name); printk("\n"); - hose = pcibios_alloc_controller(of_node_get(pcictrl)); + hose = pcibios_alloc_controller(pcictrl); if (!hose) { printk(KERN_WARNING EFIKA_PLATFORM_NAME ": Can't allocate PCI controller structure for %s\n", pcictrl->full_name); - return; + goto out_put; } hose->first_busno = bus_range[0]; @@ -124,6 +124,9 @@ static void __init efika_pcisetup(void) hose->ops = &rtas_pci_ops; pci_process_bridge_OF_ranges(hose, pcictrl, 0); + return; +out_put: + of_node_put(pcictrl); } #else -- cgit v0.10.2 From 16518d5ada690643453eb0aef3cc7841d3623c2d Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Thu, 26 Aug 2010 14:31:30 +0300 Subject: KVM: x86 emulator: fix regression with cmpxchg8b on i386 hosts operand::val and operand::orig_val are 32-bit on i386, whereas cmpxchg8b operands are 64-bit. Fix by adding val64 and orig_val64 union members to struct operand, and using them where needed. Signed-off-by: Avi Kivity Signed-off-by: Marcelo Tosatti diff --git a/arch/x86/include/asm/kvm_emulate.h b/arch/x86/include/asm/kvm_emulate.h index 51cfd73..1f99ecf 100644 --- a/arch/x86/include/asm/kvm_emulate.h +++ b/arch/x86/include/asm/kvm_emulate.h @@ -152,9 +152,14 @@ struct x86_emulate_ops { struct operand { enum { OP_REG, OP_MEM, OP_IMM, OP_NONE } type; unsigned int bytes; - unsigned long orig_val, *ptr; + union { + unsigned long orig_val; + u64 orig_val64; + }; + unsigned long *ptr; union { unsigned long val; + u64 val64; char valptr[sizeof(unsigned long) + 2]; }; }; diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index b38bd8b..66ca98a 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -1870,17 +1870,16 @@ static inline int emulate_grp9(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops) { struct decode_cache *c = &ctxt->decode; - u64 old = c->dst.orig_val; + u64 old = c->dst.orig_val64; if (((u32) (old >> 0) != (u32) c->regs[VCPU_REGS_RAX]) || ((u32) (old >> 32) != (u32) c->regs[VCPU_REGS_RDX])) { - c->regs[VCPU_REGS_RAX] = (u32) (old >> 0); c->regs[VCPU_REGS_RDX] = (u32) (old >> 32); ctxt->eflags &= ~EFLG_ZF; } else { - c->dst.val = ((u64)c->regs[VCPU_REGS_RCX] << 32) | - (u32) c->regs[VCPU_REGS_RBX]; + c->dst.val64 = ((u64)c->regs[VCPU_REGS_RCX] << 32) | + (u32) c->regs[VCPU_REGS_RBX]; ctxt->eflags |= EFLG_ZF; } @@ -2616,7 +2615,7 @@ x86_emulate_insn(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops) c->src.valptr, c->src.bytes); if (rc != X86EMUL_CONTINUE) goto done; - c->src.orig_val = c->src.val; + c->src.orig_val64 = c->src.val64; } if (c->src2.type == OP_MEM) { -- cgit v0.10.2 From ae0635b3580844e320a253a16985f4ce8b5bd0b6 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Tue, 27 Jul 2010 14:51:44 +0300 Subject: KVM: fix i8259 oops when no vcpus are online If there are no vcpus, found will be NULL. Check before doing anything with it. Signed-off-by: Avi Kivity diff --git a/arch/x86/kvm/i8259.c b/arch/x86/kvm/i8259.c index 8d10c06..4b7b73c 100644 --- a/arch/x86/kvm/i8259.c +++ b/arch/x86/kvm/i8259.c @@ -64,6 +64,9 @@ static void pic_unlock(struct kvm_pic *s) if (!found) found = s->kvm->bsp_vcpu; + if (!found) + return; + kvm_vcpu_kick(found); } } -- cgit v0.10.2 From eebb5f31b8d9a2620dcf32297096f8ce1240b179 Mon Sep 17 00:00:00 2001 From: Gleb Natapov Date: Mon, 30 Aug 2010 12:49:13 +0300 Subject: KVM: i8259: fix migration Top of kvm_kpic_state structure should have the same memory layout as kvm_pic_state since it is copied by memcpy. Signed-off-by: Gleb Natapov Signed-off-by: Avi Kivity diff --git a/arch/x86/kvm/irq.h b/arch/x86/kvm/irq.h index ffed068..63c3145 100644 --- a/arch/x86/kvm/irq.h +++ b/arch/x86/kvm/irq.h @@ -43,7 +43,6 @@ struct kvm_kpic_state { u8 irr; /* interrupt request register */ u8 imr; /* interrupt mask register */ u8 isr; /* interrupt service register */ - u8 isr_ack; /* interrupt ack detection */ u8 priority_add; /* highest irq priority */ u8 irq_base; u8 read_reg_select; @@ -56,6 +55,7 @@ struct kvm_kpic_state { u8 init4; /* true if 4 byte init */ u8 elcr; /* PIIX edge/trigger selection */ u8 elcr_mask; + u8 isr_ack; /* interrupt ack detection */ struct kvm_pic *pics_state; }; -- cgit v0.10.2 From fa32154e47a203688453e53c1369fcbc63b06a21 Mon Sep 17 00:00:00 2001 From: Eric Millbrandt Date: Fri, 3 Sep 2010 13:27:38 -0400 Subject: powerpc/5200: tighten up ac97 reset timing Tighten up time timing around the gpio reset functionality. Add a 200ns delay before remuxing the pins back to ac97 to comply with the ac97 spec. Signed-off-by: Eric Millbrandt Signed-off-by: Grant Likely diff --git a/arch/powerpc/platforms/52xx/mpc52xx_common.c b/arch/powerpc/platforms/52xx/mpc52xx_common.c index 6e90531..41f3a7e 100644 --- a/arch/powerpc/platforms/52xx/mpc52xx_common.c +++ b/arch/powerpc/platforms/52xx/mpc52xx_common.c @@ -325,12 +325,16 @@ int mpc5200_psc_ac97_gpio_reset(int psc_number) clrbits32(&simple_gpio->simple_dvo, sync | out); clrbits8(&wkup_gpio->wkup_dvo, reset); - /* wait at lease 1 us */ - udelay(2); + /* wait for 1 us */ + udelay(1); /* Deassert reset */ setbits8(&wkup_gpio->wkup_dvo, reset); + /* wait at least 200ns */ + /* 7 ~= (200ns * timebase) / ns2sec */ + __delay(7); + /* Restore pin-muxing */ out_be32(&simple_gpio->port_config, mux); -- cgit v0.10.2 From 251ee478f2c877a9a80362e094c542fbac7f5651 Mon Sep 17 00:00:00 2001 From: Jassi Brar Date: Fri, 3 Sep 2010 10:36:26 +0900 Subject: spi/s3c64xx: Fix compilation warning Fix compilation warning by typecasting the tx_buf pointer. [I'm not thrilled with resorting to a cast; but I cannot see a better way to go about this. I don't want to drop the const from struct spi_transfer ~~glikely] Signed-off-by: Jassi Brar Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index 8130f02..240b690 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -508,8 +508,9 @@ static int s3c64xx_spi_map_mssg(struct s3c64xx_spi_driver_data *sdd, list_for_each_entry(xfer, &msg->transfers, transfer_list) { if (xfer->tx_buf != NULL) { - xfer->tx_dma = dma_map_single(dev, xfer->tx_buf, - xfer->len, DMA_TO_DEVICE); + xfer->tx_dma = dma_map_single(dev, + (void *)xfer->tx_buf, xfer->len, + DMA_TO_DEVICE); if (dma_mapping_error(dev, xfer->tx_dma)) { dev_err(dev, "dma_map_single Tx failed\n"); xfer->tx_dma = XFER_DMAADDR_INVALID; -- cgit v0.10.2 From c3f139b65585a5f29df47b2302ff8dbd9bdad0b0 Mon Sep 17 00:00:00 2001 From: Jassi Brar Date: Fri, 3 Sep 2010 10:36:46 +0900 Subject: spi/s3c64xx: Fix incorrect reuse of 'val' local variable. Instead of, wrongly, reusing the 'val' variable, use a dedicated one for reading the status register. Signed-off-by: Jassi Brar Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index 240b690..ef9daca 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -326,10 +326,11 @@ static int wait_for_xfer(struct s3c64xx_spi_driver_data *sdd, val = msecs_to_jiffies(ms) + 10; val = wait_for_completion_timeout(&sdd->xfer_completion, val); } else { + u32 status; val = msecs_to_loops(ms); do { - val = readl(regs + S3C64XX_SPI_STATUS); - } while (RX_FIFO_LVL(val, sci) < xfer->len && --val); + status = readl(regs + S3C64XX_SPI_STATUS); + } while (RX_FIFO_LVL(status, sci) < xfer->len && --val); } if (!val) -- cgit v0.10.2 From be7852a839b6dcd86db1a2d25b9a1a99f38db2db Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 23 Aug 2010 17:40:56 +0100 Subject: spi/spi_s3c64xx: Warn if PIO transfers time out When using PIO we have a timeout for the TX and RX FIFOs to ensure that the data actually gets transferred. Warn if we hit that timeout - it should never happen, but this makes sure we'll find out if it does. Signed-off-by: Mark Brown Acked-by: Jassi Brar Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c index ef9daca..c3038da 100644 --- a/drivers/spi/spi_s3c64xx.c +++ b/drivers/spi/spi_s3c64xx.c @@ -200,6 +200,9 @@ static void flush_fifo(struct s3c64xx_spi_driver_data *sdd) val = readl(regs + S3C64XX_SPI_STATUS); } while (TX_FIFO_LVL(val, sci) && loops--); + if (loops == 0) + dev_warn(&sdd->pdev->dev, "Timed out flushing TX FIFO\n"); + /* Flush RxFIFO*/ loops = msecs_to_loops(1); do { @@ -210,6 +213,9 @@ static void flush_fifo(struct s3c64xx_spi_driver_data *sdd) break; } while (loops--); + if (loops == 0) + dev_warn(&sdd->pdev->dev, "Timed out flushing RX FIFO\n"); + val = readl(regs + S3C64XX_SPI_CH_CFG); val &= ~S3C64XX_SPI_CH_SW_RST; writel(val, regs + S3C64XX_SPI_CH_CFG); -- cgit v0.10.2 From 545074fb953e1753f6b8409db533ad7998789efb Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Sat, 21 Aug 2010 11:07:36 +0200 Subject: spi/pl022: fix APB pclk power regression on U300 With the introduction of an AMBA PrimeCell per-cell block clock, the pclk was left on after probe() unless explicitly disabled. This clock is wired to the same clock on PL022 causing it to stay always on since. Fix this up properly by clocking the pclk whenever we want to write into any PL022 registers and clocking the external clock whenever we want to transmit messages on the bus. Signed-off-by: Linus Walleij Tested-by : Kevin Wells Signed-off-by: Grant Likely diff --git a/drivers/spi/amba-pl022.c b/drivers/spi/amba-pl022.c index acd35d1..80613ab 100644 --- a/drivers/spi/amba-pl022.c +++ b/drivers/spi/amba-pl022.c @@ -503,8 +503,9 @@ static void giveback(struct pl022 *pl022) msg->state = NULL; if (msg->complete) msg->complete(msg->context); - /* This message is completed, so let's turn off the clock! */ + /* This message is completed, so let's turn off the clocks! */ clk_disable(pl022->clk); + amba_pclk_disable(pl022->adev); } /** @@ -1139,9 +1140,10 @@ static void pump_messages(struct work_struct *work) /* Setup the SPI using the per chip configuration */ pl022->cur_chip = spi_get_ctldata(pl022->cur_msg->spi); /* - * We enable the clock here, then the clock will be disabled when + * We enable the clocks here, then the clocks will be disabled when * giveback() is called in each method (poll/interrupt/DMA) */ + amba_pclk_enable(pl022->adev); clk_enable(pl022->clk); restore_state(pl022); flush(pl022); @@ -1786,11 +1788,9 @@ pl022_probe(struct amba_device *adev, struct amba_id *id) } /* Disable SSP */ - clk_enable(pl022->clk); writew((readw(SSP_CR1(pl022->virtbase)) & (~SSP_CR1_MASK_SSE)), SSP_CR1(pl022->virtbase)); load_ssp_default_config(pl022); - clk_disable(pl022->clk); status = request_irq(adev->irq[0], pl022_interrupt_handler, 0, "pl022", pl022); @@ -1818,6 +1818,8 @@ pl022_probe(struct amba_device *adev, struct amba_id *id) goto err_spi_register; } dev_dbg(dev, "probe succeded\n"); + /* Disable the silicon block pclk and clock it when needed */ + amba_pclk_disable(adev); return 0; err_spi_register: @@ -1879,9 +1881,9 @@ static int pl022_suspend(struct amba_device *adev, pm_message_t state) return status; } - clk_enable(pl022->clk); + amba_pclk_enable(adev); load_ssp_default_config(pl022); - clk_disable(pl022->clk); + amba_pclk_disable(adev); dev_dbg(&adev->dev, "suspended\n"); return 0; } -- cgit v0.10.2 From 970f4be85ae6ecf97b711a3a2a1d5cecd3ea0534 Mon Sep 17 00:00:00 2001 From: Heikki Lindholm Date: Mon, 6 Sep 2010 22:30:45 +0300 Subject: firewire: ohci: activate cycle timer register quirk on Ricoh chips The Ricoh FireWire controllers appear to have the non-atomic cycle timer register access bug, so, activate the driver workaround by default. The behaviour was observed on: Ricoh Co Ltd R5C552 IEEE 1394 Controller [1180:0552] and Ricoh Co Ltd R5C832 IEEE 1394 Controller [1180:0832] (rev 04). Signed-off-by: Heikki Lindholm Signed-off-by: Stefan Richter diff --git a/drivers/firewire/ohci.c b/drivers/firewire/ohci.c index be29b0b..1b05896 100644 --- a/drivers/firewire/ohci.c +++ b/drivers/firewire/ohci.c @@ -263,6 +263,7 @@ static const struct { {PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB38X_FW, QUIRK_NO_MSI}, {PCI_VENDOR_ID_NEC, PCI_ANY_ID, QUIRK_CYCLE_TIMER}, {PCI_VENDOR_ID_VIA, PCI_ANY_ID, QUIRK_CYCLE_TIMER}, + {PCI_VENDOR_ID_RICOH, PCI_ANY_ID, QUIRK_CYCLE_TIMER}, {PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_UNI_N_FW, QUIRK_BE_HEADERS}, }; -- cgit v0.10.2 From 05f25abcf6043952fb2a2d98735dec58ba1fcadb Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sun, 29 Aug 2010 11:52:41 +0200 Subject: powerpc/5200: mpc52xx_uart.c: Add of_node_put to avoid memory leak Add a call to of_node_put in the error handling code following a call to of_find_compatible_node. The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @r exists@ local idexpression x; expression E,E1; statement S; @@ *x = (of_find_node_by_path |of_find_node_by_name |of_find_node_by_phandle |of_get_parent |of_get_next_parent |of_get_next_child |of_find_compatible_node |of_match_node )(...); ... if (x == NULL) S <... when != x = E *if (...) { ... when != of_node_put(x) when != if (...) { ... of_node_put(x); ... } ( return <+...x...+>; | * return ...; ) } ...> of_node_put(x); // Signed-off-by: Julia Lawall Acked-by: Wolfram Sang Signed-off-by: Grant Likely diff --git a/drivers/serial/mpc52xx_uart.c b/drivers/serial/mpc52xx_uart.c index 8dedb26..c4399e2 100644 --- a/drivers/serial/mpc52xx_uart.c +++ b/drivers/serial/mpc52xx_uart.c @@ -500,6 +500,7 @@ static int __init mpc512x_psc_fifoc_init(void) psc_fifoc = of_iomap(np, 0); if (!psc_fifoc) { pr_err("%s: Can't map FIFOC\n", __func__); + of_node_put(np); return -ENODEV; } -- cgit v0.10.2 From 7100ae97266e387d25d0c8a5d9934931f0b07dbc Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 8 Sep 2010 20:54:49 +0000 Subject: Revert "[CIFS] Eliminate unused variable warning" The change to kernel crypto and fixes to ntlvm2 and ntlmssp series, introduced a regression. Deferring this patch series to 2.6.37 after Shirish fixes it. This reverts commit c89e5198b26a869ce2842bad8519264f3394dee9. Signed-off-by: Steve French Acked-by: Jeff Layton CC: Shirish Pargaonkar diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 795095f..4788e16 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -620,6 +620,7 @@ CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, struct key *spnego_key = NULL; __le32 phase = NtLmNegotiate; /* NTLMSSP, if needed, is multistage */ bool first_time; + char *ntlmsspblob; if (ses == NULL) return -EINVAL; @@ -867,8 +868,6 @@ ssetup_ntlmssp_authenticate: iov[1].iov_base = &pSMB->req.SecurityBlob[0]; } else if (phase == NtLmAuthenticate) { int blob_len; - char *ntlmsspblob; - ntlmsspblob = kmalloc(5 * sizeof(struct _AUTHENTICATE_MESSAGE), GFP_KERNEL); -- cgit v0.10.2 From 56234e2767496c125a858f880f1b3a62e04a3406 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 8 Sep 2010 20:57:05 +0000 Subject: Revert "Eliminate sparse warning - bad constant expression" This reverts commit 2d20ca835867d93ead6ce61780d883a4b128106d. The change to kernel crypto and fixes to ntlvm2 and ntlmssp series, introduced a regression. Deferring this patch series to 2.6.37 after Shirish fixes it. Signed-off-by: Steve French Acked-by: Jeff Layton CC: Shirish Pargaonkar diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index 709f229..eef78c2 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -45,38 +45,39 @@ extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, char *signature) { - int rc; + int rc = 0; + struct { + struct shash_desc shash; + char ctx[crypto_shash_descsize(server->ntlmssp.md5)]; + } sdesc; if (cifs_pdu == NULL || server == NULL || signature == NULL) return -EINVAL; - if (!server->ntlmssp.sdescmd5) { - cERROR(1, - "cifs_calculate_signature: can't generate signature\n"); - return -1; - } + sdesc.shash.tfm = server->ntlmssp.md5; + sdesc.shash.flags = 0x0; - rc = crypto_shash_init(&server->ntlmssp.sdescmd5->shash); + rc = crypto_shash_init(&sdesc.shash); if (rc) { - cERROR(1, "cifs_calculate_signature: oould not init md5\n"); + cERROR(1, "could not initialize master crypto API hmacmd5\n"); return rc; } if (server->secType == RawNTLMSSP) - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, server->session_key.data.ntlmv2.key, CIFS_NTLMV2_SESSKEY_SIZE); else - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, (char *)&server->session_key.data, server->session_key.len); - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, cifs_pdu->Protocol, cifs_pdu->smb_buf_length); - rc = crypto_shash_final(&server->ntlmssp.sdescmd5->shash, signature); + rc = crypto_shash_final(&sdesc.shash, signature); - return rc; + return 0; } @@ -114,28 +115,30 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, struct TCP_Server_Info *server, char *signature) { int i; - int rc; + int rc = 0; + struct { + struct shash_desc shash; + char ctx[crypto_shash_descsize(server->ntlmssp.md5)]; + } sdesc; if (iov == NULL || server == NULL || signature == NULL) return -EINVAL; - if (!server->ntlmssp.sdescmd5) { - cERROR(1, "cifs_calc_signature2: can't generate signature\n"); - return -1; - } + sdesc.shash.tfm = server->ntlmssp.md5; + sdesc.shash.flags = 0x0; - rc = crypto_shash_init(&server->ntlmssp.sdescmd5->shash); + rc = crypto_shash_init(&sdesc.shash); if (rc) { - cERROR(1, "cifs_calc_signature2: oould not init md5\n"); + cERROR(1, "could not initialize master crypto API hmacmd5\n"); return rc; } if (server->secType == RawNTLMSSP) - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, server->session_key.data.ntlmv2.key, CIFS_NTLMV2_SESSKEY_SIZE); else - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, (char *)&server->session_key.data, server->session_key.len); @@ -143,7 +146,7 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, if (iov[i].iov_len == 0) continue; if (iov[i].iov_base == NULL) { - cERROR(1, "cifs_calc_signature2: null iovec entry"); + cERROR(1, "null iovec entry"); return -EIO; } /* The first entry includes a length field (which does not get @@ -151,16 +154,16 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, if (i == 0) { if (iov[0].iov_len <= 8) /* cmd field at offset 9 */ break; /* nothing to sign or corrupt header */ - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, iov[i].iov_base + 4, iov[i].iov_len - 4); } else - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, + crypto_shash_update(&sdesc.shash, iov[i].iov_base, iov[i].iov_len); } - rc = crypto_shash_final(&server->ntlmssp.sdescmd5->shash, signature); + rc = crypto_shash_final(&sdesc.shash, signature); - return rc; + return 0; } int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, @@ -310,48 +313,43 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, wchar_t *user; wchar_t *domain; wchar_t *server; - - if (!ses->server->ntlmssp.sdeschmacmd5) { - cERROR(1, "calc_ntlmv2_hash: can't generate ntlmv2 hash\n"); - return -1; - } + struct { + struct shash_desc shash; + char ctx[crypto_shash_descsize(ses->server->ntlmssp.hmacmd5)]; + } sdesc; /* calculate md4 hash of password */ E_md4hash(ses->password, nt_hash); + sdesc.shash.tfm = ses->server->ntlmssp.hmacmd5; + sdesc.shash.flags = 0x0; + crypto_shash_setkey(ses->server->ntlmssp.hmacmd5, nt_hash, CIFS_NTHASH_SIZE); - rc = crypto_shash_init(&ses->server->ntlmssp.sdeschmacmd5->shash); + rc = crypto_shash_init(&sdesc.shash); if (rc) { - cERROR(1, "calc_ntlmv2_hash: could not init hmacmd5\n"); + cERROR(1, "could not initialize master crypto API hmacmd5\n"); return rc; } /* convert ses->userName to unicode and uppercase */ len = strlen(ses->userName); user = kmalloc(2 + (len * 2), GFP_KERNEL); - if (user == NULL) { - cERROR(1, "calc_ntlmv2_hash: user mem alloc failure\n"); - rc = -ENOMEM; + if (user == NULL) goto calc_exit_2; - } len = cifs_strtoUCS((__le16 *)user, ses->userName, len, nls_cp); UniStrupr(user); - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - (char *)user, 2 * len); + crypto_shash_update(&sdesc.shash, (char *)user, 2 * len); /* convert ses->domainName to unicode and uppercase */ if (ses->domainName) { len = strlen(ses->domainName); domain = kmalloc(2 + (len * 2), GFP_KERNEL); - if (domain == NULL) { - cERROR(1, "calc_ntlmv2_hash: domain mem alloc failure"); - rc = -ENOMEM; + if (domain == NULL) goto calc_exit_1; - } len = cifs_strtoUCS((__le16 *)domain, ses->domainName, len, nls_cp); /* the following line was removed since it didn't work well @@ -359,19 +357,15 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, Maybe converting the domain name earlier makes sense */ /* UniStrupr(domain); */ - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - (char *)domain, 2 * len); + crypto_shash_update(&sdesc.shash, (char *)domain, 2 * len); kfree(domain); } else if (ses->serverName) { len = strlen(ses->serverName); server = kmalloc(2 + (len * 2), GFP_KERNEL); - if (server == NULL) { - cERROR(1, "calc_ntlmv2_hash: server mem alloc failure"); - rc = -ENOMEM; + if (server == NULL) goto calc_exit_1; - } len = cifs_strtoUCS((__le16 *)server, ses->serverName, len, nls_cp); /* the following line was removed since it didn't work well @@ -379,20 +373,16 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, Maybe converting the domain name earlier makes sense */ /* UniStrupr(domain); */ - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - (char *)server, 2 * len); + crypto_shash_update(&sdesc.shash, (char *)server, 2 * len); kfree(server); } - - rc = crypto_shash_final(&ses->server->ntlmssp.sdeschmacmd5->shash, - ses->server->ntlmv2_hash); - calc_exit_1: kfree(user); calc_exit_2: /* BB FIXME what about bytes 24 through 40 of the signing key? compare with the NTLM example */ + rc = crypto_shash_final(&sdesc.shash, ses->server->ntlmv2_hash); return rc; } @@ -452,33 +442,34 @@ CalcNTLMv2_response(const struct TCP_Server_Info *server, char *v2_session_response) { int rc; + struct { + struct shash_desc shash; + char ctx[crypto_shash_descsize(server->ntlmssp.hmacmd5)]; + } sdesc; - if (!server->ntlmssp.sdeschmacmd5) { - cERROR(1, "calc_ntlmv2_hash: can't generate ntlmv2 hash\n"); - return -1; - } + sdesc.shash.tfm = server->ntlmssp.hmacmd5; + sdesc.shash.flags = 0x0; crypto_shash_setkey(server->ntlmssp.hmacmd5, server->ntlmv2_hash, CIFS_HMAC_MD5_HASH_SIZE); - rc = crypto_shash_init(&server->ntlmssp.sdeschmacmd5->shash); + rc = crypto_shash_init(&sdesc.shash); if (rc) { - cERROR(1, "CalcNTLMv2_response: could not init hmacmd5"); + cERROR(1, "could not initialize master crypto API hmacmd5\n"); return rc; } memcpy(v2_session_response + CIFS_SERVER_CHALLENGE_SIZE, server->cryptKey, CIFS_SERVER_CHALLENGE_SIZE); - crypto_shash_update(&server->ntlmssp.sdeschmacmd5->shash, + crypto_shash_update(&sdesc.shash, v2_session_response + CIFS_SERVER_CHALLENGE_SIZE, sizeof(struct ntlmv2_resp) - CIFS_SERVER_CHALLENGE_SIZE); if (server->tilen) - crypto_shash_update(&server->ntlmssp.sdeschmacmd5->shash, + crypto_shash_update(&sdesc.shash, server->tiblob, server->tilen); - rc = crypto_shash_final(&server->ntlmssp.sdeschmacmd5->shash, - v2_session_response); + rc = crypto_shash_final(&sdesc.shash, v2_session_response); return rc; } @@ -489,6 +480,10 @@ setup_ntlmv2_rsp(struct cifsSesInfo *ses, char *resp_buf, { int rc = 0; struct ntlmv2_resp *buf = (struct ntlmv2_resp *)resp_buf; + struct { + struct shash_desc shash; + char ctx[crypto_shash_descsize(ses->server->ntlmssp.hmacmd5)]; + } sdesc; buf->blob_signature = cpu_to_le32(0x00000101); buf->reserved = 0; @@ -516,24 +511,21 @@ setup_ntlmv2_rsp(struct cifsSesInfo *ses, char *resp_buf, return rc; } - if (!ses->server->ntlmssp.sdeschmacmd5) { - cERROR(1, "calc_ntlmv2_hash: can't generate ntlmv2 hash\n"); - return -1; - } - crypto_shash_setkey(ses->server->ntlmssp.hmacmd5, ses->server->ntlmv2_hash, CIFS_HMAC_MD5_HASH_SIZE); - rc = crypto_shash_init(&ses->server->ntlmssp.sdeschmacmd5->shash); + sdesc.shash.tfm = ses->server->ntlmssp.hmacmd5; + sdesc.shash.flags = 0x0; + + rc = crypto_shash_init(&sdesc.shash); if (rc) { - cERROR(1, "setup_ntlmv2_rsp: could not init hmacmd5\n"); + cERROR(1, "could not initialize master crypto API hmacmd5\n"); return rc; } - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - resp_buf, CIFS_HMAC_MD5_HASH_SIZE); + crypto_shash_update(&sdesc.shash, resp_buf, CIFS_HMAC_MD5_HASH_SIZE); - rc = crypto_shash_final(&ses->server->ntlmssp.sdeschmacmd5->shash, + rc = crypto_shash_final(&sdesc.shash, ses->server->session_key.data.ntlmv2.key); memcpy(&ses->server->session_key.data.ntlmv2.resp, resp_buf, @@ -586,65 +578,24 @@ cifs_crypto_shash_release(struct TCP_Server_Info *server) if (server->ntlmssp.hmacmd5) crypto_free_shash(server->ntlmssp.hmacmd5); - - kfree(server->ntlmssp.sdeschmacmd5); - - kfree(server->ntlmssp.sdescmd5); } int cifs_crypto_shash_allocate(struct TCP_Server_Info *server) { - int rc; - unsigned int size; - server->ntlmssp.hmacmd5 = crypto_alloc_shash("hmac(md5)", 0, 0); if (!server->ntlmssp.hmacmd5 || IS_ERR(server->ntlmssp.hmacmd5)) { - cERROR(1, "could not allocate crypto hmacmd5\n"); + cERROR(1, "could not allocate master crypto API hmacmd5\n"); return 1; } server->ntlmssp.md5 = crypto_alloc_shash("md5", 0, 0); if (!server->ntlmssp.md5 || IS_ERR(server->ntlmssp.md5)) { - cERROR(1, "could not allocate crypto md5\n"); - rc = 1; - goto cifs_crypto_shash_allocate_ret1; - } - - size = sizeof(struct shash_desc) + - crypto_shash_descsize(server->ntlmssp.hmacmd5); - server->ntlmssp.sdeschmacmd5 = kmalloc(size, GFP_KERNEL); - if (!server->ntlmssp.sdeschmacmd5) { - cERROR(1, "cifs_crypto_shash_allocate: can't alloc hmacmd5\n"); - rc = -ENOMEM; - goto cifs_crypto_shash_allocate_ret2; - } - server->ntlmssp.sdeschmacmd5->shash.tfm = server->ntlmssp.hmacmd5; - server->ntlmssp.sdeschmacmd5->shash.flags = 0x0; - - - size = sizeof(struct shash_desc) + - crypto_shash_descsize(server->ntlmssp.md5); - server->ntlmssp.sdescmd5 = kmalloc(size, GFP_KERNEL); - if (!server->ntlmssp.sdescmd5) { - cERROR(1, "cifs_crypto_shash_allocate: can't alloc md5\n"); - rc = -ENOMEM; - goto cifs_crypto_shash_allocate_ret3; + crypto_free_shash(server->ntlmssp.hmacmd5); + cERROR(1, "could not allocate master crypto API md5\n"); + return 1; } - server->ntlmssp.sdescmd5->shash.tfm = server->ntlmssp.md5; - server->ntlmssp.sdescmd5->shash.flags = 0x0; return 0; - -cifs_crypto_shash_allocate_ret3: - kfree(server->ntlmssp.sdeschmacmd5); - -cifs_crypto_shash_allocate_ret2: - crypto_free_shash(server->ntlmssp.md5); - -cifs_crypto_shash_allocate_ret1: - crypto_free_shash(server->ntlmssp.hmacmd5); - - return rc; } diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index c9d0cfc..49563e0 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -123,19 +123,12 @@ struct cifs_cred { struct cifs_ace *aces; }; -struct sdesc { - struct shash_desc shash; - char ctx[]; -}; - struct ntlmssp_auth { __u32 client_flags; __u32 server_flags; unsigned char ciphertext[CIFS_CPHTXT_SIZE]; struct crypto_shash *hmacmd5; struct crypto_shash *md5; - struct sdesc *sdeschmacmd5; - struct sdesc *sdescmd5; }; /* -- cgit v0.10.2 From 745e507a9c79c6e1385d3414d5e56f3d4621a375 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 8 Sep 2010 21:09:27 +0000 Subject: Revert "missing changes during ntlmv2/ntlmssp auth and sign" This reverts commit 3ec6bbcdb4e85403f2c5958876ca9492afdf4031. The change to kernel crypto and fixes to ntlvm2 and ntlmssp series, introduced a regression. Deferring this patch series to 2.6.37 after Shirish fixes it. Signed-off-by: Steve French Acked-by: Jeff Layton CC: Shirish Pargaonkar diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index eef78c2..051d000 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -553,8 +553,6 @@ calc_seckey(struct TCP_Server_Info *server) return 1; } - desc.tfm = tfm_arc4; - crypto_blkcipher_setkey(tfm_arc4, server->session_key.data.ntlmv2.key, CIFS_CPHTXT_SIZE); sg_init_one(&sgin, sec_key, CIFS_CPHTXT_SIZE); diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 4788e16..41fc532 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -408,8 +408,6 @@ static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, /* BB spec says that if AvId field of MsvAvTimestamp is populated then we must set the MIC field of the AUTHENTICATE_MESSAGE */ - ses->server->ntlmssp.server_flags = le32_to_cpu(pblob->NegotiateFlags); - tioffset = cpu_to_le16(pblob->TargetInfoArray.BufferOffset); tilen = cpu_to_le16(pblob->TargetInfoArray.Length); ses->server->tilen = tilen; @@ -442,13 +440,12 @@ static void build_ntlmssp_negotiate_blob(unsigned char *pbuffer, /* BB is NTLMV2 session security format easier to use here? */ flags = NTLMSSP_NEGOTIATE_56 | NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE | - NTLMSSP_NEGOTIATE_NTLM; + NTLMSSP_NEGOTIATE_NT_ONLY | NTLMSSP_NEGOTIATE_NTLM; if (ses->server->secMode & - (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { - flags |= NTLMSSP_NEGOTIATE_SIGN | - NTLMSSP_NEGOTIATE_KEY_XCH | - NTLMSSP_NEGOTIATE_EXTENDED_SEC; - } + (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) + flags |= NTLMSSP_NEGOTIATE_SIGN; + if (ses->server->secMode & SECMODE_SIGN_REQUIRED) + flags |= NTLMSSP_NEGOTIATE_ALWAYS_SIGN; sec_blob->NegotiateFlags |= cpu_to_le32(flags); -- cgit v0.10.2 From c8e56f1f4fb9f82f63e4ce6d73a14501d0432c76 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 8 Sep 2010 21:10:58 +0000 Subject: Revert "[CIFS] Fix ntlmv2 auth with ntlmssp" This reverts commit 9fbc590860e75785bdaf8b83e48fabfe4d4f7d58. The change to kernel crypto and fixes to ntlvm2 and ntlmssp series, introduced a regression. Deferring this patch series to 2.6.37 after Shirish fixes it. Signed-off-by: Steve French Acked-by: Jeff Layton CC: Shirish Pargaonkar diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index 0da1deb..917b7d4 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -2,8 +2,6 @@ config CIFS tristate "CIFS support (advanced network filesystem, SMBFS successor)" depends on INET select NLS - select CRYPTO_MD5 - select CRYPTO_ARC4 help This is the client VFS module for the Common Internet File System (CIFS) protocol which is the successor to the Server Message Block diff --git a/fs/cifs/asn1.c b/fs/cifs/asn1.c index 21f0fbd..cfd1ce3 100644 --- a/fs/cifs/asn1.c +++ b/fs/cifs/asn1.c @@ -597,13 +597,13 @@ decode_negTokenInit(unsigned char *security_blob, int length, if (compare_oid(oid, oidlen, MSKRB5_OID, MSKRB5_OID_LEN)) server->sec_mskerberos = true; - if (compare_oid(oid, oidlen, KRB5U2U_OID, + else if (compare_oid(oid, oidlen, KRB5U2U_OID, KRB5U2U_OID_LEN)) server->sec_kerberosu2u = true; - if (compare_oid(oid, oidlen, KRB5_OID, + else if (compare_oid(oid, oidlen, KRB5_OID, KRB5_OID_LEN)) server->sec_kerberos = true; - if (compare_oid(oid, oidlen, NTLMSSP_OID, + else if (compare_oid(oid, oidlen, NTLMSSP_OID, NTLMSSP_OID_LEN)) server->sec_ntlmssp = true; diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index 051d000..847628d 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -27,7 +27,6 @@ #include "md5.h" #include "cifs_unicode.h" #include "cifsproto.h" -#include "ntlmssp.h" #include #include @@ -43,44 +42,21 @@ extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, unsigned char *p24); static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, - struct TCP_Server_Info *server, char *signature) + const struct mac_key *key, char *signature) { - int rc = 0; - struct { - struct shash_desc shash; - char ctx[crypto_shash_descsize(server->ntlmssp.md5)]; - } sdesc; + struct MD5Context context; - if (cifs_pdu == NULL || server == NULL || signature == NULL) + if ((cifs_pdu == NULL) || (signature == NULL) || (key == NULL)) return -EINVAL; - sdesc.shash.tfm = server->ntlmssp.md5; - sdesc.shash.flags = 0x0; - - rc = crypto_shash_init(&sdesc.shash); - if (rc) { - cERROR(1, "could not initialize master crypto API hmacmd5\n"); - return rc; - } - - if (server->secType == RawNTLMSSP) - crypto_shash_update(&sdesc.shash, - server->session_key.data.ntlmv2.key, - CIFS_NTLMV2_SESSKEY_SIZE); - else - crypto_shash_update(&sdesc.shash, - (char *)&server->session_key.data, - server->session_key.len); - - crypto_shash_update(&sdesc.shash, - cifs_pdu->Protocol, cifs_pdu->smb_buf_length); - - rc = crypto_shash_final(&sdesc.shash, signature); + cifs_MD5_init(&context); + cifs_MD5_update(&context, (char *)&key->data, key->len); + cifs_MD5_update(&context, cifs_pdu->Protocol, cifs_pdu->smb_buf_length); + cifs_MD5_final(signature, &context); return 0; } - int cifs_sign_smb(struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, __u32 *pexpected_response_sequence_number) { @@ -102,7 +78,8 @@ int cifs_sign_smb(struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, server->sequence_number++; spin_unlock(&GlobalMid_Lock); - rc = cifs_calculate_signature(cifs_pdu, server, smb_signature); + rc = cifs_calculate_signature(cifs_pdu, &server->mac_signing_key, + smb_signature); if (rc) memset(cifs_pdu->Signature.SecuritySignature, 0, 8); else @@ -112,36 +89,16 @@ int cifs_sign_smb(struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, } static int cifs_calc_signature2(const struct kvec *iov, int n_vec, - struct TCP_Server_Info *server, char *signature) + const struct mac_key *key, char *signature) { + struct MD5Context context; int i; - int rc = 0; - struct { - struct shash_desc shash; - char ctx[crypto_shash_descsize(server->ntlmssp.md5)]; - } sdesc; - if (iov == NULL || server == NULL || signature == NULL) + if ((iov == NULL) || (signature == NULL) || (key == NULL)) return -EINVAL; - sdesc.shash.tfm = server->ntlmssp.md5; - sdesc.shash.flags = 0x0; - - rc = crypto_shash_init(&sdesc.shash); - if (rc) { - cERROR(1, "could not initialize master crypto API hmacmd5\n"); - return rc; - } - - if (server->secType == RawNTLMSSP) - crypto_shash_update(&sdesc.shash, - server->session_key.data.ntlmv2.key, - CIFS_NTLMV2_SESSKEY_SIZE); - else - crypto_shash_update(&sdesc.shash, - (char *)&server->session_key.data, - server->session_key.len); - + cifs_MD5_init(&context); + cifs_MD5_update(&context, (char *)&key->data, key->len); for (i = 0; i < n_vec; i++) { if (iov[i].iov_len == 0) continue; @@ -154,18 +111,18 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, if (i == 0) { if (iov[0].iov_len <= 8) /* cmd field at offset 9 */ break; /* nothing to sign or corrupt header */ - crypto_shash_update(&sdesc.shash, - iov[i].iov_base + 4, iov[i].iov_len - 4); + cifs_MD5_update(&context, iov[0].iov_base+4, + iov[0].iov_len-4); } else - crypto_shash_update(&sdesc.shash, - iov[i].iov_base, iov[i].iov_len); + cifs_MD5_update(&context, iov[i].iov_base, iov[i].iov_len); } - rc = crypto_shash_final(&sdesc.shash, signature); + cifs_MD5_final(signature, &context); return 0; } + int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, __u32 *pexpected_response_sequence_number) { @@ -188,7 +145,8 @@ int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, server->sequence_number++; spin_unlock(&GlobalMid_Lock); - rc = cifs_calc_signature2(iov, n_vec, server, smb_signature); + rc = cifs_calc_signature2(iov, n_vec, &server->mac_signing_key, + smb_signature); if (rc) memset(cifs_pdu->Signature.SecuritySignature, 0, 8); else @@ -198,14 +156,14 @@ int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, } int cifs_verify_signature(struct smb_hdr *cifs_pdu, - struct TCP_Server_Info *server, + const struct mac_key *mac_key, __u32 expected_sequence_number) { - int rc; + unsigned int rc; char server_response_sig[8]; char what_we_think_sig_should_be[20]; - if (cifs_pdu == NULL || server == NULL) + if ((cifs_pdu == NULL) || (mac_key == NULL)) return -EINVAL; if (cifs_pdu->Command == SMB_COM_NEGOTIATE) @@ -234,7 +192,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, cpu_to_le32(expected_sequence_number); cifs_pdu->Signature.Sequence.Reserved = 0; - rc = cifs_calculate_signature(cifs_pdu, server, + rc = cifs_calculate_signature(cifs_pdu, mac_key, what_we_think_sig_should_be); if (rc) @@ -251,7 +209,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, } /* We fill in key by putting in 40 byte array which was allocated by caller */ -int cifs_calculate_session_key(struct session_key *key, const char *rn, +int cifs_calculate_mac_key(struct mac_key *key, const char *rn, const char *password) { char temp_key[16]; @@ -265,6 +223,63 @@ int cifs_calculate_session_key(struct session_key *key, const char *rn, return 0; } +int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *ses, + const struct nls_table *nls_info) +{ + char temp_hash[16]; + struct HMACMD5Context ctx; + char *ucase_buf; + __le16 *unicode_buf; + unsigned int i, user_name_len, dom_name_len; + + if (ses == NULL) + return -EINVAL; + + E_md4hash(ses->password, temp_hash); + + hmac_md5_init_limK_to_64(temp_hash, 16, &ctx); + user_name_len = strlen(ses->userName); + if (user_name_len > MAX_USERNAME_SIZE) + return -EINVAL; + if (ses->domainName == NULL) + return -EINVAL; /* BB should we use CIFS_LINUX_DOM */ + dom_name_len = strlen(ses->domainName); + if (dom_name_len > MAX_USERNAME_SIZE) + return -EINVAL; + + ucase_buf = kmalloc((MAX_USERNAME_SIZE+1), GFP_KERNEL); + if (ucase_buf == NULL) + return -ENOMEM; + unicode_buf = kmalloc((MAX_USERNAME_SIZE+1)*4, GFP_KERNEL); + if (unicode_buf == NULL) { + kfree(ucase_buf); + return -ENOMEM; + } + + for (i = 0; i < user_name_len; i++) + ucase_buf[i] = nls_info->charset2upper[(int)ses->userName[i]]; + ucase_buf[i] = 0; + user_name_len = cifs_strtoUCS(unicode_buf, ucase_buf, + MAX_USERNAME_SIZE*2, nls_info); + unicode_buf[user_name_len] = 0; + user_name_len++; + + for (i = 0; i < dom_name_len; i++) + ucase_buf[i] = nls_info->charset2upper[(int)ses->domainName[i]]; + ucase_buf[i] = 0; + dom_name_len = cifs_strtoUCS(unicode_buf+user_name_len, ucase_buf, + MAX_USERNAME_SIZE*2, nls_info); + + unicode_buf[user_name_len + dom_name_len] = 0; + hmac_md5_update((const unsigned char *) unicode_buf, + (user_name_len+dom_name_len)*2, &ctx); + + hmac_md5_final(ses->server->ntlmv2_hash, &ctx); + kfree(ucase_buf); + kfree(unicode_buf); + return 0; +} + #ifdef CONFIG_CIFS_WEAK_PW_HASH void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, char *lnm_session_key) @@ -309,29 +324,21 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, { int rc = 0; int len; - char nt_hash[CIFS_NTHASH_SIZE]; + char nt_hash[16]; + struct HMACMD5Context *pctxt; wchar_t *user; wchar_t *domain; - wchar_t *server; - struct { - struct shash_desc shash; - char ctx[crypto_shash_descsize(ses->server->ntlmssp.hmacmd5)]; - } sdesc; - /* calculate md4 hash of password */ - E_md4hash(ses->password, nt_hash); + pctxt = kmalloc(sizeof(struct HMACMD5Context), GFP_KERNEL); - sdesc.shash.tfm = ses->server->ntlmssp.hmacmd5; - sdesc.shash.flags = 0x0; + if (pctxt == NULL) + return -ENOMEM; - crypto_shash_setkey(ses->server->ntlmssp.hmacmd5, nt_hash, - CIFS_NTHASH_SIZE); + /* calculate md4 hash of password */ + E_md4hash(ses->password, nt_hash); - rc = crypto_shash_init(&sdesc.shash); - if (rc) { - cERROR(1, "could not initialize master crypto API hmacmd5\n"); - return rc; - } + /* convert Domainname to unicode and uppercase */ + hmac_md5_init_limK_to_64(nt_hash, 16, pctxt); /* convert ses->userName to unicode and uppercase */ len = strlen(ses->userName); @@ -340,8 +347,7 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, goto calc_exit_2; len = cifs_strtoUCS((__le16 *)user, ses->userName, len, nls_cp); UniStrupr(user); - - crypto_shash_update(&sdesc.shash, (char *)user, 2 * len); + hmac_md5_update((char *)user, 2*len, pctxt); /* convert ses->domainName to unicode and uppercase */ if (ses->domainName) { @@ -357,243 +363,65 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, Maybe converting the domain name earlier makes sense */ /* UniStrupr(domain); */ - crypto_shash_update(&sdesc.shash, (char *)domain, 2 * len); + hmac_md5_update((char *)domain, 2*len, pctxt); kfree(domain); - } else if (ses->serverName) { - len = strlen(ses->serverName); - - server = kmalloc(2 + (len * 2), GFP_KERNEL); - if (server == NULL) - goto calc_exit_1; - len = cifs_strtoUCS((__le16 *)server, ses->serverName, len, - nls_cp); - /* the following line was removed since it didn't work well - with lower cased domain name that passed as an option. - Maybe converting the domain name earlier makes sense */ - /* UniStrupr(domain); */ - - crypto_shash_update(&sdesc.shash, (char *)server, 2 * len); - - kfree(server); } calc_exit_1: kfree(user); calc_exit_2: /* BB FIXME what about bytes 24 through 40 of the signing key? compare with the NTLM example */ - rc = crypto_shash_final(&sdesc.shash, ses->server->ntlmv2_hash); - - return rc; -} - -static int -find_domain_name(struct cifsSesInfo *ses) -{ - int rc = 0; - unsigned int attrsize; - unsigned int type; - unsigned char *blobptr; - struct ntlmssp2_name *attrptr; - - if (ses->server->tiblob) { - blobptr = ses->server->tiblob; - attrptr = (struct ntlmssp2_name *) blobptr; - - while ((type = attrptr->type) != 0) { - blobptr += 2; /* advance attr type */ - attrsize = attrptr->length; - blobptr += 2; /* advance attr size */ - if (type == NTLMSSP_AV_NB_DOMAIN_NAME) { - if (!ses->domainName) { - ses->domainName = - kmalloc(attrptr->length + 1, - GFP_KERNEL); - if (!ses->domainName) - return -ENOMEM; - cifs_from_ucs2(ses->domainName, - (__le16 *)blobptr, - attrptr->length, - attrptr->length, - load_nls_default(), false); - } - } - blobptr += attrsize; /* advance attr value */ - attrptr = (struct ntlmssp2_name *) blobptr; - } - } else { - ses->server->tilen = 2 * sizeof(struct ntlmssp2_name); - ses->server->tiblob = kmalloc(ses->server->tilen, GFP_KERNEL); - if (!ses->server->tiblob) { - ses->server->tilen = 0; - cERROR(1, "Challenge target info allocation failure"); - return -ENOMEM; - } - memset(ses->server->tiblob, 0x0, ses->server->tilen); - attrptr = (struct ntlmssp2_name *) ses->server->tiblob; - attrptr->type = cpu_to_le16(NTLMSSP_DOMAIN_TYPE); - } + hmac_md5_final(ses->server->ntlmv2_hash, pctxt); + kfree(pctxt); return rc; } -static int -CalcNTLMv2_response(const struct TCP_Server_Info *server, - char *v2_session_response) -{ - int rc; - struct { - struct shash_desc shash; - char ctx[crypto_shash_descsize(server->ntlmssp.hmacmd5)]; - } sdesc; - - sdesc.shash.tfm = server->ntlmssp.hmacmd5; - sdesc.shash.flags = 0x0; - - crypto_shash_setkey(server->ntlmssp.hmacmd5, server->ntlmv2_hash, - CIFS_HMAC_MD5_HASH_SIZE); - - rc = crypto_shash_init(&sdesc.shash); - if (rc) { - cERROR(1, "could not initialize master crypto API hmacmd5\n"); - return rc; - } - - memcpy(v2_session_response + CIFS_SERVER_CHALLENGE_SIZE, - server->cryptKey, CIFS_SERVER_CHALLENGE_SIZE); - crypto_shash_update(&sdesc.shash, - v2_session_response + CIFS_SERVER_CHALLENGE_SIZE, - sizeof(struct ntlmv2_resp) - CIFS_SERVER_CHALLENGE_SIZE); - - if (server->tilen) - crypto_shash_update(&sdesc.shash, - server->tiblob, server->tilen); - - rc = crypto_shash_final(&sdesc.shash, v2_session_response); - - return rc; -} - -int -setup_ntlmv2_rsp(struct cifsSesInfo *ses, char *resp_buf, +void setup_ntlmv2_rsp(struct cifsSesInfo *ses, char *resp_buf, const struct nls_table *nls_cp) { - int rc = 0; + int rc; struct ntlmv2_resp *buf = (struct ntlmv2_resp *)resp_buf; - struct { - struct shash_desc shash; - char ctx[crypto_shash_descsize(ses->server->ntlmssp.hmacmd5)]; - } sdesc; + struct HMACMD5Context context; buf->blob_signature = cpu_to_le32(0x00000101); buf->reserved = 0; buf->time = cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); get_random_bytes(&buf->client_chal, sizeof(buf->client_chal)); buf->reserved2 = 0; - - if (!ses->domainName) { - rc = find_domain_name(ses); - if (rc) { - cERROR(1, "could not get domain/server name rc %d", rc); - return rc; - } - } + buf->names[0].type = cpu_to_le16(NTLMSSP_DOMAIN_TYPE); + buf->names[0].length = 0; + buf->names[1].type = 0; + buf->names[1].length = 0; /* calculate buf->ntlmv2_hash */ rc = calc_ntlmv2_hash(ses, nls_cp); - if (rc) { - cERROR(1, "could not get v2 hash rc %d", rc); - return rc; - } - rc = CalcNTLMv2_response(ses->server, resp_buf); - if (rc) { + if (rc) cERROR(1, "could not get v2 hash rc %d", rc); - return rc; - } - - crypto_shash_setkey(ses->server->ntlmssp.hmacmd5, - ses->server->ntlmv2_hash, CIFS_HMAC_MD5_HASH_SIZE); - - sdesc.shash.tfm = ses->server->ntlmssp.hmacmd5; - sdesc.shash.flags = 0x0; - - rc = crypto_shash_init(&sdesc.shash); - if (rc) { - cERROR(1, "could not initialize master crypto API hmacmd5\n"); - return rc; - } - - crypto_shash_update(&sdesc.shash, resp_buf, CIFS_HMAC_MD5_HASH_SIZE); + CalcNTLMv2_response(ses, resp_buf); - rc = crypto_shash_final(&sdesc.shash, - ses->server->session_key.data.ntlmv2.key); + /* now calculate the MAC key for NTLMv2 */ + hmac_md5_init_limK_to_64(ses->server->ntlmv2_hash, 16, &context); + hmac_md5_update(resp_buf, 16, &context); + hmac_md5_final(ses->server->mac_signing_key.data.ntlmv2.key, &context); - memcpy(&ses->server->session_key.data.ntlmv2.resp, resp_buf, - sizeof(struct ntlmv2_resp)); - ses->server->session_key.len = 16 + sizeof(struct ntlmv2_resp); - - return rc; + memcpy(&ses->server->mac_signing_key.data.ntlmv2.resp, resp_buf, + sizeof(struct ntlmv2_resp)); + ses->server->mac_signing_key.len = 16 + sizeof(struct ntlmv2_resp); } -int -calc_seckey(struct TCP_Server_Info *server) -{ - int rc; - unsigned char sec_key[CIFS_NTLMV2_SESSKEY_SIZE]; - struct crypto_blkcipher *tfm_arc4; - struct scatterlist sgin, sgout; - struct blkcipher_desc desc; - - get_random_bytes(sec_key, CIFS_NTLMV2_SESSKEY_SIZE); - - tfm_arc4 = crypto_alloc_blkcipher("ecb(arc4)", - 0, CRYPTO_ALG_ASYNC); - if (!tfm_arc4 || IS_ERR(tfm_arc4)) { - cERROR(1, "could not allocate " "master crypto API arc4\n"); - return 1; - } - - crypto_blkcipher_setkey(tfm_arc4, - server->session_key.data.ntlmv2.key, CIFS_CPHTXT_SIZE); - sg_init_one(&sgin, sec_key, CIFS_CPHTXT_SIZE); - sg_init_one(&sgout, server->ntlmssp.ciphertext, CIFS_CPHTXT_SIZE); - rc = crypto_blkcipher_encrypt(&desc, &sgout, &sgin, CIFS_CPHTXT_SIZE); - - if (!rc) - memcpy(server->session_key.data.ntlmv2.key, - sec_key, CIFS_NTLMV2_SESSKEY_SIZE); - - crypto_free_blkcipher(tfm_arc4); - - return 0; -} - -void -cifs_crypto_shash_release(struct TCP_Server_Info *server) -{ - if (server->ntlmssp.md5) - crypto_free_shash(server->ntlmssp.md5); - - if (server->ntlmssp.hmacmd5) - crypto_free_shash(server->ntlmssp.hmacmd5); -} - -int -cifs_crypto_shash_allocate(struct TCP_Server_Info *server) +void CalcNTLMv2_response(const struct cifsSesInfo *ses, + char *v2_session_response) { - server->ntlmssp.hmacmd5 = crypto_alloc_shash("hmac(md5)", 0, 0); - if (!server->ntlmssp.hmacmd5 || - IS_ERR(server->ntlmssp.hmacmd5)) { - cERROR(1, "could not allocate master crypto API hmacmd5\n"); - return 1; - } + struct HMACMD5Context context; + /* rest of v2 struct already generated */ + memcpy(v2_session_response + 8, ses->server->cryptKey, 8); + hmac_md5_init_limK_to_64(ses->server->ntlmv2_hash, 16, &context); - server->ntlmssp.md5 = crypto_alloc_shash("md5", 0, 0); - if (!server->ntlmssp.md5 || IS_ERR(server->ntlmssp.md5)) { - crypto_free_shash(server->ntlmssp.hmacmd5); - cERROR(1, "could not allocate master crypto API md5\n"); - return 1; - } + hmac_md5_update(v2_session_response+8, + sizeof(struct ntlmv2_resp) - 8, &context); - return 0; + hmac_md5_final(v2_session_response, &context); +/* cifs_dump_mem("v2_sess_rsp: ", v2_session_response, 32); */ } diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 49563e0..0cdfb8c 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -25,9 +25,6 @@ #include #include "cifs_fs_sb.h" #include "cifsacl.h" -#include -#include - /* * The sizes of various internal tables and strings */ @@ -100,7 +97,7 @@ enum protocolEnum { /* Netbios frames protocol not supported at this time */ }; -struct session_key { +struct mac_key { unsigned int len; union { char ntlm[CIFS_SESS_KEY_SIZE + 16]; @@ -123,14 +120,6 @@ struct cifs_cred { struct cifs_ace *aces; }; -struct ntlmssp_auth { - __u32 client_flags; - __u32 server_flags; - unsigned char ciphertext[CIFS_CPHTXT_SIZE]; - struct crypto_shash *hmacmd5; - struct crypto_shash *md5; -}; - /* ***************************************************************** * Except the CIFS PDUs themselves all the @@ -193,14 +182,11 @@ struct TCP_Server_Info { /* 16th byte of RFC1001 workstation name is always null */ char workstation_RFC1001_name[RFC1001_NAME_LEN_WITH_NULL]; __u32 sequence_number; /* needed for CIFS PDU signature */ - struct session_key session_key; + struct mac_key mac_signing_key; char ntlmv2_hash[16]; unsigned long lstrp; /* when we got last response from this server */ u16 dialect; /* dialect index that server chose */ /* extended security flavors that server supports */ - unsigned int tilen; /* length of the target info blob */ - unsigned char *tiblob; /* target info blob in challenge response */ - struct ntlmssp_auth ntlmssp; /* various keys, ciphers, flags */ bool sec_kerberos; /* supports plain Kerberos */ bool sec_mskerberos; /* supports legacy MS Kerberos */ bool sec_kerberosu2u; /* supports U2U Kerberos */ diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index 320e0fd..14d036d 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -134,12 +134,6 @@ * Size of the session key (crypto key encrypted with the password */ #define CIFS_SESS_KEY_SIZE (24) -#define CIFS_CLIENT_CHALLENGE_SIZE (8) -#define CIFS_SERVER_CHALLENGE_SIZE (8) -#define CIFS_HMAC_MD5_HASH_SIZE (16) -#define CIFS_CPHTXT_SIZE (16) -#define CIFS_NTLMV2_SESSKEY_SIZE (16) -#define CIFS_NTHASH_SIZE (16) /* * Maximum user name length @@ -669,6 +663,7 @@ struct ntlmv2_resp { __le64 time; __u64 client_chal; /* random */ __u32 reserved2; + struct ntlmssp2_name names[2]; /* array of name entries could follow ending in minimum 4 byte struct */ } __attribute__((packed)); diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 1378d91..1f54508 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -361,15 +361,15 @@ extern int cifs_sign_smb(struct smb_hdr *, struct TCP_Server_Info *, __u32 *); extern int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *, __u32 *); extern int cifs_verify_signature(struct smb_hdr *, - struct TCP_Server_Info *server, + const struct mac_key *mac_key, __u32 expected_sequence_number); -extern int cifs_calculate_session_key(struct session_key *key, const char *rn, +extern int cifs_calculate_mac_key(struct mac_key *key, const char *rn, const char *pass); -extern int setup_ntlmv2_rsp(struct cifsSesInfo *, char *, +extern int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *, + const struct nls_table *); +extern void CalcNTLMv2_response(const struct cifsSesInfo *, char *); +extern void setup_ntlmv2_rsp(struct cifsSesInfo *, char *, const struct nls_table *); -extern int cifs_crypto_shash_allocate(struct TCP_Server_Info *); -extern void cifs_crypto_shash_release(struct TCP_Server_Info *); -extern int calc_seckey(struct TCP_Server_Info *); #ifdef CONFIG_CIFS_WEAK_PW_HASH extern void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, char *lnm_session_key); diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 4bda920..c65c341 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -604,14 +604,11 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) else rc = -EINVAL; - if (server->secType == Kerberos) { - if (!server->sec_kerberos && - !server->sec_mskerberos) - rc = -EOPNOTSUPP; - } else if (server->secType == RawNTLMSSP) { - if (!server->sec_ntlmssp) - rc = -EOPNOTSUPP; - } else + if (server->sec_kerberos || server->sec_mskerberos) + server->secType = Kerberos; + else if (server->sec_ntlmssp) + server->secType = RawNTLMSSP; + else rc = -EOPNOTSUPP; } } else diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index ec0ea4a..0ea52e9 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1708,7 +1708,6 @@ cifs_put_smb_ses(struct cifsSesInfo *ses) CIFSSMBLogoff(xid, ses); _FreeXid(xid); } - cifs_crypto_shash_release(server); sesInfoFree(ses); cifs_put_tcp_session(server); } @@ -1788,23 +1787,13 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) ses->linux_uid = volume_info->linux_uid; ses->overrideSecFlg = volume_info->secFlg; - rc = cifs_crypto_shash_allocate(server); - if (rc) { - cERROR(1, "could not setup hash structures rc %d", rc); - goto get_ses_fail; - } - server->tilen = 0; - server->tiblob = NULL; - mutex_lock(&ses->session_mutex); rc = cifs_negotiate_protocol(xid, ses); if (!rc) rc = cifs_setup_session(xid, ses, volume_info->local_nls); mutex_unlock(&ses->session_mutex); - if (rc) { - cifs_crypto_shash_release(ses->server); + if (rc) goto get_ses_fail; - } /* success, put it on the list */ write_lock(&cifs_tcp_ses_lock); diff --git a/fs/cifs/ntlmssp.h b/fs/cifs/ntlmssp.h index 1db0f07..49c9a4e 100644 --- a/fs/cifs/ntlmssp.h +++ b/fs/cifs/ntlmssp.h @@ -61,19 +61,6 @@ #define NTLMSSP_NEGOTIATE_KEY_XCH 0x40000000 #define NTLMSSP_NEGOTIATE_56 0x80000000 -/* Define AV Pair Field IDs */ -#define NTLMSSP_AV_EOL 0 -#define NTLMSSP_AV_NB_COMPUTER_NAME 1 -#define NTLMSSP_AV_NB_DOMAIN_NAME 2 -#define NTLMSSP_AV_DNS_COMPUTER_NAME 3 -#define NTLMSSP_AV_DNS_DOMAIN_NAME 4 -#define NTLMSSP_AV_DNS_TREE_NAME 5 -#define NTLMSSP_AV_FLAGS 6 -#define NTLMSSP_AV_TIMESTAMP 7 -#define NTLMSSP_AV_RESTRICTION 8 -#define NTLMSSP_AV_TARGET_NAME 9 -#define NTLMSSP_AV_CHANNEL_BINDINGS 10 - /* Although typedefs are not commonly used for structure definitions */ /* in the Linux kernel, in this particular case they are useful */ /* to more closely match the standards document for NTLMSSP from */ diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 41fc532..0a57cb7 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -383,9 +383,6 @@ static int decode_ascii_ssetup(char **pbcc_area, int bleft, static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, struct cifsSesInfo *ses) { - unsigned int tioffset; /* challeng message target info area */ - unsigned int tilen; /* challeng message target info area length */ - CHALLENGE_MESSAGE *pblob = (CHALLENGE_MESSAGE *)bcc_ptr; if (blob_len < sizeof(CHALLENGE_MESSAGE)) { @@ -408,18 +405,6 @@ static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, /* BB spec says that if AvId field of MsvAvTimestamp is populated then we must set the MIC field of the AUTHENTICATE_MESSAGE */ - tioffset = cpu_to_le16(pblob->TargetInfoArray.BufferOffset); - tilen = cpu_to_le16(pblob->TargetInfoArray.Length); - ses->server->tilen = tilen; - if (tilen) { - ses->server->tiblob = kmalloc(tilen, GFP_KERNEL); - if (!ses->server->tiblob) { - cERROR(1, "Challenge target info allocation failure"); - return -ENOMEM; - } - memcpy(ses->server->tiblob, bcc_ptr + tioffset, tilen); - } - return 0; } @@ -466,12 +451,10 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, struct cifsSesInfo *ses, const struct nls_table *nls_cp, bool first) { - int rc; - unsigned int size; AUTHENTICATE_MESSAGE *sec_blob = (AUTHENTICATE_MESSAGE *)pbuffer; __u32 flags; unsigned char *tmp; - struct ntlmv2_resp ntlmv2_response = {}; + char ntlm_session_key[CIFS_SESS_KEY_SIZE]; memcpy(sec_blob->Signature, NTLMSSP_SIGNATURE, 8); sec_blob->MessageType = NtLmAuthenticate; @@ -494,25 +477,19 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, sec_blob->LmChallengeResponse.Length = 0; sec_blob->LmChallengeResponse.MaximumLength = 0; - sec_blob->NtChallengeResponse.BufferOffset = cpu_to_le32(tmp - pbuffer); - rc = setup_ntlmv2_rsp(ses, (char *)&ntlmv2_response, nls_cp); - if (rc) { - cERROR(1, "error rc: %d during ntlmssp ntlmv2 setup", rc); - goto setup_ntlmv2_ret; - } - size = sizeof(struct ntlmv2_resp); - memcpy(tmp, (char *)&ntlmv2_response, size); - tmp += size; - if (ses->server->tilen > 0) { - memcpy(tmp, ses->server->tiblob, ses->server->tilen); - tmp += ses->server->tilen; - } else - ses->server->tilen = 0; + /* calculate session key, BB what about adding similar ntlmv2 path? */ + SMBNTencrypt(ses->password, ses->server->cryptKey, ntlm_session_key); + if (first) + cifs_calculate_mac_key(&ses->server->mac_signing_key, + ntlm_session_key, ses->password); - sec_blob->NtChallengeResponse.Length = cpu_to_le16(size + - ses->server->tilen); + memcpy(tmp, ntlm_session_key, CIFS_SESS_KEY_SIZE); + sec_blob->NtChallengeResponse.BufferOffset = cpu_to_le32(tmp - pbuffer); + sec_blob->NtChallengeResponse.Length = cpu_to_le16(CIFS_SESS_KEY_SIZE); sec_blob->NtChallengeResponse.MaximumLength = - cpu_to_le16(size + ses->server->tilen); + cpu_to_le16(CIFS_SESS_KEY_SIZE); + + tmp += CIFS_SESS_KEY_SIZE; if (ses->domainName == NULL) { sec_blob->DomainName.BufferOffset = cpu_to_le32(tmp - pbuffer); @@ -524,6 +501,7 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, len = cifs_strtoUCS((__le16 *)tmp, ses->domainName, MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ + len += 2; /* trailing null */ sec_blob->DomainName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->DomainName.Length = cpu_to_le16(len); sec_blob->DomainName.MaximumLength = cpu_to_le16(len); @@ -540,6 +518,7 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, len = cifs_strtoUCS((__le16 *)tmp, ses->userName, MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ + len += 2; /* trailing null */ sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->UserName.Length = cpu_to_le16(len); sec_blob->UserName.MaximumLength = cpu_to_le16(len); @@ -551,26 +530,9 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, sec_blob->WorkstationName.MaximumLength = 0; tmp += 2; - if ((ses->server->ntlmssp.server_flags & NTLMSSP_NEGOTIATE_KEY_XCH) && - !calc_seckey(ses->server)) { - memcpy(tmp, ses->server->ntlmssp.ciphertext, CIFS_CPHTXT_SIZE); - sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - pbuffer); - sec_blob->SessionKey.Length = cpu_to_le16(CIFS_CPHTXT_SIZE); - sec_blob->SessionKey.MaximumLength = - cpu_to_le16(CIFS_CPHTXT_SIZE); - tmp += CIFS_CPHTXT_SIZE; - } else { - sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - pbuffer); - sec_blob->SessionKey.Length = 0; - sec_blob->SessionKey.MaximumLength = 0; - } - - ses->server->sequence_number = 0; - -setup_ntlmv2_ret: - if (ses->server->tilen > 0) - kfree(ses->server->tiblob); - + sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - pbuffer); + sec_blob->SessionKey.Length = 0; + sec_blob->SessionKey.MaximumLength = 0; return tmp - pbuffer; } @@ -584,14 +546,15 @@ static void setup_ntlmssp_neg_req(SESSION_SETUP_ANDX *pSMB, return; } -static int setup_ntlmssp_auth_req(char *ntlmsspblob, +static int setup_ntlmssp_auth_req(SESSION_SETUP_ANDX *pSMB, struct cifsSesInfo *ses, const struct nls_table *nls, bool first_time) { int bloblen; - bloblen = build_ntlmssp_auth_blob(ntlmsspblob, ses, nls, + bloblen = build_ntlmssp_auth_blob(&pSMB->req.SecurityBlob[0], ses, nls, first_time); + pSMB->req.SecurityBlobLength = cpu_to_le16(bloblen); return bloblen; } @@ -617,7 +580,6 @@ CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, struct key *spnego_key = NULL; __le32 phase = NtLmNegotiate; /* NTLMSSP, if needed, is multistage */ bool first_time; - char *ntlmsspblob; if (ses == NULL) return -EINVAL; @@ -728,7 +690,7 @@ ssetup_ntlmssp_authenticate: if (first_time) /* should this be moved into common code with similar ntlmv2 path? */ - cifs_calculate_session_key(&ses->server->session_key, + cifs_calculate_mac_key(&ses->server->mac_signing_key, ntlm_session_key, ses->password); /* copy session key */ @@ -767,21 +729,12 @@ ssetup_ntlmssp_authenticate: cpu_to_le16(sizeof(struct ntlmv2_resp)); /* calculate session key */ - rc = setup_ntlmv2_rsp(ses, v2_sess_key, nls_cp); - if (rc) { - kfree(v2_sess_key); - goto ssetup_exit; - } + setup_ntlmv2_rsp(ses, v2_sess_key, nls_cp); /* FIXME: calculate MAC key */ memcpy(bcc_ptr, (char *)v2_sess_key, sizeof(struct ntlmv2_resp)); bcc_ptr += sizeof(struct ntlmv2_resp); kfree(v2_sess_key); - if (ses->server->tilen > 0) { - memcpy(bcc_ptr, ses->server->tiblob, - ses->server->tilen); - bcc_ptr += ses->server->tilen; - } if (ses->capabilities & CAP_UNICODE) { if (iov[0].iov_len % 2) { *bcc_ptr = 0; @@ -812,15 +765,15 @@ ssetup_ntlmssp_authenticate: } /* bail out if key is too long */ if (msg->sesskey_len > - sizeof(ses->server->session_key.data.krb5)) { + sizeof(ses->server->mac_signing_key.data.krb5)) { cERROR(1, "Kerberos signing key too long (%u bytes)", msg->sesskey_len); rc = -EOVERFLOW; goto ssetup_exit; } if (first_time) { - ses->server->session_key.len = msg->sesskey_len; - memcpy(ses->server->session_key.data.krb5, + ses->server->mac_signing_key.len = msg->sesskey_len; + memcpy(ses->server->mac_signing_key.data.krb5, msg->data, msg->sesskey_len); } pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; @@ -862,26 +815,12 @@ ssetup_ntlmssp_authenticate: if (phase == NtLmNegotiate) { setup_ntlmssp_neg_req(pSMB, ses); iov[1].iov_len = sizeof(NEGOTIATE_MESSAGE); - iov[1].iov_base = &pSMB->req.SecurityBlob[0]; } else if (phase == NtLmAuthenticate) { int blob_len; - ntlmsspblob = kmalloc(5 * - sizeof(struct _AUTHENTICATE_MESSAGE), - GFP_KERNEL); - if (!ntlmsspblob) { - cERROR(1, "Can't allocate NTLMSSP"); - rc = -ENOMEM; - goto ssetup_exit; - } - - blob_len = setup_ntlmssp_auth_req(ntlmsspblob, - ses, - nls_cp, - first_time); + blob_len = setup_ntlmssp_auth_req(pSMB, ses, + nls_cp, + first_time); iov[1].iov_len = blob_len; - iov[1].iov_base = ntlmsspblob; - pSMB->req.SecurityBlobLength = - cpu_to_le16(blob_len); /* Make sure that we tell the server that we are using the uid that it just gave us back on the response (challenge) */ @@ -891,6 +830,7 @@ ssetup_ntlmssp_authenticate: rc = -ENOSYS; goto ssetup_exit; } + iov[1].iov_base = &pSMB->req.SecurityBlob[0]; /* unicode strings must be word aligned */ if ((iov[0].iov_len + iov[1].iov_len) % 2) { *bcc_ptr = 0; diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index e0588cd..82f78c4 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -543,7 +543,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))) { rc = cifs_verify_signature(midQ->resp_buf, - ses->server, + &ses->server->mac_signing_key, midQ->sequence_number+1); if (rc) { cERROR(1, "Unexpected SMB signature"); @@ -731,7 +731,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))) { rc = cifs_verify_signature(out_buf, - ses->server, + &ses->server->mac_signing_key, midQ->sequence_number+1); if (rc) { cERROR(1, "Unexpected SMB signature"); @@ -981,7 +981,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))) { rc = cifs_verify_signature(out_buf, - ses->server, + &ses->server->mac_signing_key, midQ->sequence_number+1); if (rc) { cERROR(1, "Unexpected SMB signature"); -- cgit v0.10.2 From 639e7a913d81f918bfbf506e6ecd54664f787cbd Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 3 Sep 2010 11:50:09 -0400 Subject: cifs: eliminate redundant xdev check in cifs_rename The VFS always checks that the source and target of a rename are on the same vfsmount, and hence have the same superblock. So, this check is redundant. Remove it and simplify the error handling. Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 86a164f..93f77d4 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1462,29 +1462,18 @@ int cifs_rename(struct inode *source_dir, struct dentry *source_dentry, { char *fromName = NULL; char *toName = NULL; - struct cifs_sb_info *cifs_sb_source; - struct cifs_sb_info *cifs_sb_target; + struct cifs_sb_info *cifs_sb; struct cifsTconInfo *tcon; FILE_UNIX_BASIC_INFO *info_buf_source = NULL; FILE_UNIX_BASIC_INFO *info_buf_target; int xid, rc, tmprc; - cifs_sb_target = CIFS_SB(target_dir->i_sb); - cifs_sb_source = CIFS_SB(source_dir->i_sb); - tcon = cifs_sb_source->tcon; + cifs_sb = CIFS_SB(source_dir->i_sb); + tcon = cifs_sb->tcon; xid = GetXid(); /* - * BB: this might be allowed if same server, but different share. - * Consider adding support for this - */ - if (tcon != cifs_sb_target->tcon) { - rc = -EXDEV; - goto cifs_rename_exit; - } - - /* * we already have the rename sem so we do not need to * grab it again here to protect the path integrity */ @@ -1519,17 +1508,16 @@ int cifs_rename(struct inode *source_dir, struct dentry *source_dentry, info_buf_target = info_buf_source + 1; tmprc = CIFSSMBUnixQPathInfo(xid, tcon, fromName, info_buf_source, - cifs_sb_source->local_nls, - cifs_sb_source->mnt_cifs_flags & + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (tmprc != 0) goto unlink_target; - tmprc = CIFSSMBUnixQPathInfo(xid, tcon, - toName, info_buf_target, - cifs_sb_target->local_nls, - /* remap based on source sb */ - cifs_sb_source->mnt_cifs_flags & + tmprc = CIFSSMBUnixQPathInfo(xid, tcon, toName, + info_buf_target, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (tmprc == 0 && (info_buf_source->UniqueId == -- cgit v0.10.2 From 4266d9118f85b050a341992f0cfab40d392ef32c Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 8 Sep 2010 21:17:29 +0000 Subject: [CIFS] ntlmv2/ntlmssp remove-unused-function CalcNTLMv2_partial_mac_key This function is not used, so remove the definition and declaration. Reviewed-by: Jeff Layton Signed-off-by: Shirish Pargaonkar Signed-off-by: Steve French diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index 847628d..35042d8 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -223,63 +223,6 @@ int cifs_calculate_mac_key(struct mac_key *key, const char *rn, return 0; } -int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *ses, - const struct nls_table *nls_info) -{ - char temp_hash[16]; - struct HMACMD5Context ctx; - char *ucase_buf; - __le16 *unicode_buf; - unsigned int i, user_name_len, dom_name_len; - - if (ses == NULL) - return -EINVAL; - - E_md4hash(ses->password, temp_hash); - - hmac_md5_init_limK_to_64(temp_hash, 16, &ctx); - user_name_len = strlen(ses->userName); - if (user_name_len > MAX_USERNAME_SIZE) - return -EINVAL; - if (ses->domainName == NULL) - return -EINVAL; /* BB should we use CIFS_LINUX_DOM */ - dom_name_len = strlen(ses->domainName); - if (dom_name_len > MAX_USERNAME_SIZE) - return -EINVAL; - - ucase_buf = kmalloc((MAX_USERNAME_SIZE+1), GFP_KERNEL); - if (ucase_buf == NULL) - return -ENOMEM; - unicode_buf = kmalloc((MAX_USERNAME_SIZE+1)*4, GFP_KERNEL); - if (unicode_buf == NULL) { - kfree(ucase_buf); - return -ENOMEM; - } - - for (i = 0; i < user_name_len; i++) - ucase_buf[i] = nls_info->charset2upper[(int)ses->userName[i]]; - ucase_buf[i] = 0; - user_name_len = cifs_strtoUCS(unicode_buf, ucase_buf, - MAX_USERNAME_SIZE*2, nls_info); - unicode_buf[user_name_len] = 0; - user_name_len++; - - for (i = 0; i < dom_name_len; i++) - ucase_buf[i] = nls_info->charset2upper[(int)ses->domainName[i]]; - ucase_buf[i] = 0; - dom_name_len = cifs_strtoUCS(unicode_buf+user_name_len, ucase_buf, - MAX_USERNAME_SIZE*2, nls_info); - - unicode_buf[user_name_len + dom_name_len] = 0; - hmac_md5_update((const unsigned char *) unicode_buf, - (user_name_len+dom_name_len)*2, &ctx); - - hmac_md5_final(ses->server->ntlmv2_hash, &ctx); - kfree(ucase_buf); - kfree(unicode_buf); - return 0; -} - #ifdef CONFIG_CIFS_WEAK_PW_HASH void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, char *lnm_session_key) diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 1f54508..f399b16 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -365,8 +365,6 @@ extern int cifs_verify_signature(struct smb_hdr *, __u32 expected_sequence_number); extern int cifs_calculate_mac_key(struct mac_key *key, const char *rn, const char *pass); -extern int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *, - const struct nls_table *); extern void CalcNTLMv2_response(const struct cifsSesInfo *, char *); extern void setup_ntlmv2_rsp(struct cifsSesInfo *, char *, const struct nls_table *); -- cgit v0.10.2 From 522bbe65a2415fabce618186fc7777eb4c502989 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 3 Sep 2010 12:00:49 -0400 Subject: cifs: prevent cifsd from exiting prematurely When cifs_demultiplex_thread exits, it does a number of cleanup tasks including freeing the TCP_Server_Info struct. Much of the existing code in cifs assumes that when there is a cisfSesInfo struct, that it holds a reference to a valid TCP_Server_Info struct. We can never allow cifsd to exit when a cifsSesInfo struct is still holding a reference to the server. The server pointers will then point to freed memory. This patch eliminates a couple of questionable conditions where it does this. The idea here is to make an -EINTR return from kernel_recvmsg behave the same way as -ERESTARTSYS or -EAGAIN. If the task was signalled from cifs_put_tcp_session, then tcpStatus will be CifsExiting, and the kernel_recvmsg call will return quickly. There's also another condition where this can occur too -- if the tcpStatus is still in CifsNew, then it will also exit if the server closes the socket prematurely. I think we'll probably also need to fix that situation, but that requires a bit more consideration. Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 0ea52e9..5f68b96 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -400,7 +400,9 @@ incomplete_rcv: cFYI(1, "call to reconnect done"); csocket = server->ssocket; continue; - } else if ((length == -ERESTARTSYS) || (length == -EAGAIN)) { + } else if (length == -ERESTARTSYS || + length == -EAGAIN || + length == -EINTR) { msleep(1); /* minimum sleep to prevent looping allowing socket to clear and app threads to set tcpStatus CifsNeedReconnect if server hung */ @@ -422,10 +424,6 @@ incomplete_rcv: and so simply return error to mount */ break; } - if (!try_to_freeze() && (length == -EINTR)) { - cFYI(1, "cifsd thread killed"); - break; - } cFYI(1, "Reconnect after unexpected peek error %d", length); cifs_reconnect(server); @@ -522,8 +520,7 @@ incomplete_rcv: total_read += length) { length = kernel_recvmsg(csocket, &smb_msg, &iov, 1, pdu_length - total_read, 0); - if ((server->tcpStatus == CifsExiting) || - (length == -EINTR)) { + if (server->tcpStatus == CifsExiting) { /* then will exit */ reconnect = 2; break; @@ -534,8 +531,9 @@ incomplete_rcv: /* Now we will reread sock */ reconnect = 1; break; - } else if ((length == -ERESTARTSYS) || - (length == -EAGAIN)) { + } else if (length == -ERESTARTSYS || + length == -EAGAIN || + length == -EINTR) { msleep(1); /* minimum sleep to prevent looping, allowing socket to clear and app threads to set tcpStatus -- cgit v0.10.2 From 7332f2a6217ee6925f83ef0e725013067ed316ba Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 3 Sep 2010 12:00:49 -0400 Subject: cifs: eliminate some more premature cifsd exits If the tcpStatus is still CifsNew, the main cifs_demultiplex_loop can break out prematurely in some cases. This is wrong as we will almost always have other structures with pointers to the TCP_Server_Info. If the main loop breaks under any other condition other than tcpStatus == CifsExiting, then it'll face a use-after-free situation. I don't see any reason to treat a CifsNew tcpStatus differently than CifsGood. I believe we'll still want to attempt to reconnect in either case. What should happen in those situations is that the MIDs get marked as MID_RETRY_NEEDED. This will make CIFSSMBNegotiate return -EAGAIN, and then the caller can retry the whole thing on a newly reconnected socket. If that fails again in the same way, the caller of cifs_get_smb_ses should tear down the TCP_Server_Info struct. Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 5f68b96..5fde83f 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -416,14 +416,6 @@ incomplete_rcv: } else continue; } else if (length <= 0) { - if (server->tcpStatus == CifsNew) { - cFYI(1, "tcp session abend after SMBnegprot"); - /* some servers kill the TCP session rather than - returning an SMB negprot error, in which - case reconnecting here is not going to help, - and so simply return error to mount */ - break; - } cFYI(1, "Reconnect after unexpected peek error %d", length); cifs_reconnect(server); @@ -464,27 +456,18 @@ incomplete_rcv: an error on SMB negprot response */ cFYI(1, "Negative RFC1002 Session Response Error 0x%x)", pdu_length); - if (server->tcpStatus == CifsNew) { - /* if nack on negprot (rather than - ret of smb negprot error) reconnecting - not going to help, ret error to mount */ - break; - } else { - /* give server a second to - clean up before reconnect attempt */ - msleep(1000); - /* always try 445 first on reconnect - since we get NACK on some if we ever - connected to port 139 (the NACK is - since we do not begin with RFC1001 - session initialize frame) */ - server->addr.sockAddr.sin_port = - htons(CIFS_PORT); - cifs_reconnect(server); - csocket = server->ssocket; - wake_up(&server->response_q); - continue; - } + /* give server a second to clean up */ + msleep(1000); + /* always try 445 first on reconnect since we get NACK + * on some if we ever connected to port 139 (the NACK + * is since we do not begin with RFC1001 session + * initialize frame) + */ + server->addr.sockAddr.sin_port = htons(CIFS_PORT); + cifs_reconnect(server); + csocket = server->ssocket; + wake_up(&server->response_q); + continue; } else if (temp != (char) 0) { cERROR(1, "Unknown RFC 1002 frame"); cifs_dump_mem(" Received Data: ", (char *)smb_buffer, -- cgit v0.10.2 From 32670396e7fc6e4f37451a69339968985461a374 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 3 Sep 2010 12:00:50 -0400 Subject: cifs: prevent possible memory corruption in cifs_demultiplex_thread cifs_demultiplex_thread sets the addr.sockAddr.sin_port without any regard for the socket family. While it may be that the error in question here never occurs on an IPv6 socket, it's probably best to be safe and set the port properly if it ever does. Break the port setting code out of cifs_fill_sockaddr and into a new function, and call that from cifs_demultiplex_thread. Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index f399b16..1d60c65 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -87,8 +87,9 @@ extern unsigned int smbCalcSize_LE(struct smb_hdr *ptr); extern int decode_negTokenInit(unsigned char *security_blob, int length, struct TCP_Server_Info *server); extern int cifs_convert_address(struct sockaddr *dst, const char *src, int len); +extern int cifs_set_port(struct sockaddr *addr, const unsigned short int port); extern int cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, - unsigned short int port); + const unsigned short int port); extern int map_smb_to_linux_error(struct smb_hdr *smb, int logErr); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifsTconInfo *, int /* length of diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 5fde83f..67dad54 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -463,7 +463,8 @@ incomplete_rcv: * is since we do not begin with RFC1001 session * initialize frame) */ - server->addr.sockAddr.sin_port = htons(CIFS_PORT); + cifs_set_port((struct sockaddr *) + &server->addr.sockAddr, CIFS_PORT); cifs_reconnect(server); csocket = server->ssocket; wake_up(&server->response_q); diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index f978511..9aad47a 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -206,26 +206,30 @@ cifs_convert_address(struct sockaddr *dst, const char *src, int len) } int -cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, - const unsigned short int port) +cifs_set_port(struct sockaddr *addr, const unsigned short int port) { - if (!cifs_convert_address(dst, src, len)) - return 0; - - switch (dst->sa_family) { + switch (addr->sa_family) { case AF_INET: - ((struct sockaddr_in *)dst)->sin_port = htons(port); + ((struct sockaddr_in *)addr)->sin_port = htons(port); break; case AF_INET6: - ((struct sockaddr_in6 *)dst)->sin6_port = htons(port); + ((struct sockaddr_in6 *)addr)->sin6_port = htons(port); break; default: return 0; } - return 1; } +int +cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, + const unsigned short int port) +{ + if (!cifs_convert_address(dst, src, len)) + return 0; + return cifs_set_port(dst, port); +} + /***************************************************************************** convert a NT status code to a dos class/code *****************************************************************************/ -- cgit v0.10.2 From 70c9db0fdfd703781c3b8c2caf9287806f642e02 Mon Sep 17 00:00:00 2001 From: Chien Tung Date: Tue, 7 Sep 2010 16:31:20 +0000 Subject: RDMA/nes: Write correct register write to set TX pause param Setting TX pause param writes to the wrong register location causing the adapter to hang. Correct the define used to write the reigster. Addresses: https://bugs.openfabrics.org/show_bug.cgi?id=2116 Reported-by: Shiri Franchi Signed-off-by: Chien Tung Signed-off-by: Roland Dreier diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c index 6dfdd49..10560c7 100644 --- a/drivers/infiniband/hw/nes/nes_nic.c +++ b/drivers/infiniband/hw/nes/nes_nic.c @@ -1446,14 +1446,14 @@ static int nes_netdev_set_pauseparam(struct net_device *netdev, NES_IDX_MAC_TX_CONFIG + (nesdev->mac_index*0x200)); u32temp |= NES_IDX_MAC_TX_CONFIG_ENABLE_PAUSE; nes_write_indexed(nesdev, - NES_IDX_MAC_TX_CONFIG_ENABLE_PAUSE + (nesdev->mac_index*0x200), u32temp); + NES_IDX_MAC_TX_CONFIG + (nesdev->mac_index*0x200), u32temp); nesdev->disable_tx_flow_control = 0; } else if ((et_pauseparam->tx_pause == 0) && (nesdev->disable_tx_flow_control == 0)) { u32temp = nes_read_indexed(nesdev, NES_IDX_MAC_TX_CONFIG + (nesdev->mac_index*0x200)); u32temp &= ~NES_IDX_MAC_TX_CONFIG_ENABLE_PAUSE; nes_write_indexed(nesdev, - NES_IDX_MAC_TX_CONFIG_ENABLE_PAUSE + (nesdev->mac_index*0x200), u32temp); + NES_IDX_MAC_TX_CONFIG + (nesdev->mac_index*0x200), u32temp); nesdev->disable_tx_flow_control = 1; } if ((et_pauseparam->rx_pause == 1) && (nesdev->disable_rx_flow_control == 1)) { -- cgit v0.10.2 From dae58728dc64e9ad71c40ac90b463bff6ecce271 Mon Sep 17 00:00:00 2001 From: Faisal Latif Date: Sat, 14 Aug 2010 21:04:56 +0000 Subject: RDMA/nes: Fix double CLOSE event indication crash During a stress testing in a large cluster, multiple close event are detected and BUG() is hit in the iWARP core. The cause is that the active node gave up while waiting for an MPA response from the peer and tried to close the connection by sending RST. The passive node driver receives the RST but is waiting for MPA response from the user. When the MPA accept is received, the driver offloads the connection and sends a CLOSE event. The driver gets an AE indicating RESET received and also sends a CLOSE event, hitting a BUG(). Fix this by correcting RESET handling and sending CLOSE events. Signed-off-by: Faisal Latif Signed-off-by: Roland Dreier diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c index 443cea5..61e0efd 100644 --- a/drivers/infiniband/hw/nes/nes_cm.c +++ b/drivers/infiniband/hw/nes/nes_cm.c @@ -502,7 +502,9 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb, static void nes_retrans_expired(struct nes_cm_node *cm_node) { struct iw_cm_id *cm_id = cm_node->cm_id; - switch (cm_node->state) { + enum nes_cm_node_state state = cm_node->state; + cm_node->state = NES_CM_STATE_CLOSED; + switch (state) { case NES_CM_STATE_SYN_RCVD: case NES_CM_STATE_CLOSING: rem_ref_cm_node(cm_node->cm_core, cm_node); @@ -511,7 +513,6 @@ static void nes_retrans_expired(struct nes_cm_node *cm_node) case NES_CM_STATE_FIN_WAIT1: if (cm_node->cm_id) cm_id->rem_ref(cm_id); - cm_node->state = NES_CM_STATE_CLOSED; send_reset(cm_node, NULL); break; default: @@ -1439,9 +1440,6 @@ static void handle_rst_pkt(struct nes_cm_node *cm_node, struct sk_buff *skb, break; case NES_CM_STATE_MPAREQ_RCVD: passive_state = atomic_add_return(1, &cm_node->passive_state); - if (passive_state == NES_SEND_RESET_EVENT) - create_event(cm_node, NES_CM_EVENT_RESET); - cm_node->state = NES_CM_STATE_CLOSED; dev_kfree_skb_any(skb); break; case NES_CM_STATE_ESTABLISHED: @@ -1456,6 +1454,7 @@ static void handle_rst_pkt(struct nes_cm_node *cm_node, struct sk_buff *skb, case NES_CM_STATE_CLOSED: drop_packet(skb); break; + case NES_CM_STATE_FIN_WAIT2: case NES_CM_STATE_FIN_WAIT1: case NES_CM_STATE_LAST_ACK: cm_node->cm_id->rem_ref(cm_node->cm_id); @@ -2777,6 +2776,12 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) return -EINVAL; } + passive_state = atomic_add_return(1, &cm_node->passive_state); + if (passive_state == NES_SEND_RESET_EVENT) { + rem_ref_cm_node(cm_node->cm_core, cm_node); + return -ECONNRESET; + } + /* associate the node with the QP */ nesqp->cm_node = (void *)cm_node; cm_node->nesqp = nesqp; @@ -2979,9 +2984,6 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) printk(KERN_ERR "%s[%u] OFA CM event_handler returned, " "ret=%d\n", __func__, __LINE__, ret); - passive_state = atomic_add_return(1, &cm_node->passive_state); - if (passive_state == NES_SEND_RESET_EVENT) - create_event(cm_node, NES_CM_EVENT_RESET); return 0; } -- cgit v0.10.2 From 67d70721151726286763209ecadc3fce86abfdce Mon Sep 17 00:00:00 2001 From: Faisal Latif Date: Sat, 14 Aug 2010 21:05:04 +0000 Subject: RDMA/nes: Change state to closing after FIN When the driver receives an AE for FIN received, it closes the connection without changing the state of the connection in the hardware to closing. By changing the state to closing, hardware will do a normal close sequence. Signed-off-by: Faisal Latif Signed-off-by: Roland Dreier diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c index f8233c8..ba93a8b 100644 --- a/drivers/infiniband/hw/nes/nes_hw.c +++ b/drivers/infiniband/hw/nes/nes_hw.c @@ -3468,6 +3468,18 @@ static void nes_process_iwarp_aeqe(struct nes_device *nesdev, return; /* Ignore it, wait for close complete */ if (atomic_inc_return(&nesqp->close_timer_started) == 1) { + if ((tcp_state == NES_AEQE_TCP_STATE_CLOSE_WAIT) && + (nesqp->ibqp_state == IB_QPS_RTS)) { + spin_lock_irqsave(&nesqp->lock, flags); + nesqp->hw_iwarp_state = iwarp_state; + nesqp->hw_tcp_state = tcp_state; + nesqp->last_aeq = async_event_id; + next_iwarp_state = NES_CQP_QP_IWARP_STATE_CLOSING; + nesqp->hw_iwarp_state = NES_AEQE_IWARP_STATE_CLOSING; + spin_unlock_irqrestore(&nesqp->lock, flags); + nes_hw_modify_qp(nesdev, nesqp, next_iwarp_state, 0, 0); + nes_cm_disconn(nesqp); + } nesqp->cm_id->add_ref(nesqp->cm_id); schedule_nes_timer(nesqp->cm_node, (struct sk_buff *)nesqp, NES_TIMER_TYPE_CLOSE, 1, 0); @@ -3477,7 +3489,6 @@ static void nes_process_iwarp_aeqe(struct nes_device *nesdev, nesqp->hwqp.qp_id, atomic_read(&nesqp->refcount), async_event_id, nesqp->last_aeq, tcp_state); } - break; case NES_AEQE_AEID_LLP_CLOSE_COMPLETE: if (nesqp->term_flags) { -- cgit v0.10.2 From 29da03b9d1c6f24548d98cebda1e15a25d80ee1b Mon Sep 17 00:00:00 2001 From: Faisal Latif Date: Wed, 1 Sep 2010 15:43:11 +0000 Subject: RDMA/nes: Fix hang with modified FIN handling on A0 cards Changing state to CLOSING when FIN is received causes A0 cards to hang. Fix this by checking for A0 cards in FIN handling. Signed-off-by: Faisal Latif Signed-off-by: Roland Dreier diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c index ba93a8b..1980a46 100644 --- a/drivers/infiniband/hw/nes/nes_hw.c +++ b/drivers/infiniband/hw/nes/nes_hw.c @@ -3469,7 +3469,8 @@ static void nes_process_iwarp_aeqe(struct nes_device *nesdev, if (atomic_inc_return(&nesqp->close_timer_started) == 1) { if ((tcp_state == NES_AEQE_TCP_STATE_CLOSE_WAIT) && - (nesqp->ibqp_state == IB_QPS_RTS)) { + (nesqp->ibqp_state == IB_QPS_RTS) && + ((nesadapter->eeprom_version >> 16) != NES_A0)) { spin_lock_irqsave(&nesqp->lock, flags); nesqp->hw_iwarp_state = iwarp_state; nesqp->hw_tcp_state = tcp_state; diff --git a/drivers/infiniband/hw/nes/nes_hw.h b/drivers/infiniband/hw/nes/nes_hw.h index aa9183d..1204c34 100644 --- a/drivers/infiniband/hw/nes/nes_hw.h +++ b/drivers/infiniband/hw/nes/nes_hw.h @@ -45,6 +45,7 @@ #define NES_PHY_TYPE_KR 9 #define NES_MULTICAST_PF_MAX 8 +#define NES_A0 3 enum pci_regs { NES_INT_STAT = 0x0000, -- cgit v0.10.2 From 152e1d592071c8b312bb898bc1118b64e4aea535 Mon Sep 17 00:00:00 2001 From: Colin Cross Date: Fri, 3 Sep 2010 01:24:07 +0200 Subject: PM: Prevent waiting forever on asynchronous resume after failing suspend During suspend, the power.completion is expected to be set when a device has not yet started suspending. Set it on init to fix a corner case where a device is resumed when its parent has never suspended. Consider three drivers, A, B, and C. The parent of A is C, and C has async_suspend set. On boot, C->power.completion is initialized to 0. During the first suspend: suspend_devices_and_enter(...) dpm_resume(...) device_suspend(A) device_suspend(B) returns error, aborts suspend dpm_resume_end(...) dpm_resume(...) device_resume(A) dpm_wait(A->parent == C) wait_for_completion(C->power.completion) The wait_for_completion will never complete, because complete_all(C->power.completion) will only be called from device_suspend(C) or device_resume(C), neither of which is called if suspend is aborted before C. After a successful suspend->resume cycle, where B doesn't abort suspend, C->power.completion is left in the completed state by the call to device_resume(C), and the same call path will work if B aborts suspend. Signed-off-by: Colin Cross Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 5419a49..276d5a7 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -59,6 +59,7 @@ void device_pm_init(struct device *dev) { dev->power.status = DPM_ON; init_completion(&dev->power.completion); + complete_all(&dev->power.completion); dev->power.wakeup_count = 0; pm_runtime_init(dev); } -- cgit v0.10.2 From b19f8200392db88d529576a4f8178d08b70047be Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 8 Sep 2010 20:57:59 -0700 Subject: sparc: Kill all BKL usage. They were all bogus artifacts and completely unnecessary. Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/sys_sparc_32.c b/arch/sparc/kernel/sys_sparc_32.c index ee995b7..de45de4 100644 --- a/arch/sparc/kernel/sys_sparc_32.c +++ b/arch/sparc/kernel/sys_sparc_32.c @@ -166,7 +166,6 @@ sparc_breakpoint (struct pt_regs *regs) { siginfo_t info; - lock_kernel(); #ifdef DEBUG_SPARC_BREAKPOINT printk ("TRAP: Entering kernel PC=%x, nPC=%x\n", regs->pc, regs->npc); #endif @@ -180,7 +179,6 @@ sparc_breakpoint (struct pt_regs *regs) #ifdef DEBUG_SPARC_BREAKPOINT printk ("TRAP: Returning to space: PC=%x nPC=%x\n", regs->pc, regs->npc); #endif - unlock_kernel(); } asmlinkage int diff --git a/arch/sparc/kernel/unaligned_32.c b/arch/sparc/kernel/unaligned_32.c index f8514e2..12b9f35 100644 --- a/arch/sparc/kernel/unaligned_32.c +++ b/arch/sparc/kernel/unaligned_32.c @@ -323,7 +323,6 @@ asmlinkage void user_unaligned_trap(struct pt_regs *regs, unsigned int insn) { enum direction dir; - lock_kernel(); if(!(current->thread.flags & SPARC_FLAG_UNALIGNED) || (((insn >> 30) & 3) != 3)) goto kill_user; @@ -377,5 +376,5 @@ asmlinkage void user_unaligned_trap(struct pt_regs *regs, unsigned int insn) kill_user: user_mna_trap_fault(regs, insn); out: - unlock_kernel(); + ; } diff --git a/arch/sparc/kernel/windows.c b/arch/sparc/kernel/windows.c index f24d298..b351770 100644 --- a/arch/sparc/kernel/windows.c +++ b/arch/sparc/kernel/windows.c @@ -112,7 +112,6 @@ void try_to_clear_window_buffer(struct pt_regs *regs, int who) struct thread_info *tp = current_thread_info(); int window; - lock_kernel(); flush_user_windows(); for(window = 0; window < tp->w_saved; window++) { unsigned long sp = tp->rwbuf_stkptrs[window]; @@ -123,5 +122,4 @@ void try_to_clear_window_buffer(struct pt_regs *regs, int who) do_exit(SIGILL); } tp->w_saved = 0; - unlock_kernel(); } -- cgit v0.10.2 From 25c8e03bdb769dfe2381f8b7942f05b0eb4bdf31 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Mon, 6 Sep 2010 11:02:12 +0200 Subject: spi/pl022: move probe call to subsys_initcall() The PL022 SPI bus is sometimes used for early stuff like regulators that need to be present at module_init() time, so we move this to a subsys_initcall(). Signed-off-by: Linus Walleij Signed-off-by: Grant Likely diff --git a/drivers/spi/amba-pl022.c b/drivers/spi/amba-pl022.c index 80613ab..4c37c4e2 100644 --- a/drivers/spi/amba-pl022.c +++ b/drivers/spi/amba-pl022.c @@ -1983,7 +1983,7 @@ static int __init pl022_init(void) return amba_driver_register(&pl022_driver); } -module_init(pl022_init); +subsys_initcall(pl022_init); static void __exit pl022_exit(void) { -- cgit v0.10.2 From a7a13d0676335a7dc9dd72264cca02606e43aaba Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 9 Sep 2010 00:11:41 +0200 Subject: ALSA: rawmidi: fix the get next midi device ioctl If we pass in a device which is higher than SNDRV_RAWMIDI_DEVICES then the "next device" should be -1. This function just returns device + 1. But the main thing is that "device + 1" can lead to a (harmless) integer overflow and that annoys static analysis tools. [fix the case for device == SNDRV_RAWMIDI_DEVICE by tiwai] Signed-off-by: Dan Carpenter Signed-off-by: Takashi Iwai diff --git a/sound/core/rawmidi.c b/sound/core/rawmidi.c index eb68326..a7868ad 100644 --- a/sound/core/rawmidi.c +++ b/sound/core/rawmidi.c @@ -829,6 +829,8 @@ static int snd_rawmidi_control_ioctl(struct snd_card *card, if (get_user(device, (int __user *)argp)) return -EFAULT; + if (device >= SNDRV_RAWMIDI_DEVICES) /* next device is -1 */ + device = SNDRV_RAWMIDI_DEVICES - 1; mutex_lock(®ister_mutex); device = device < 0 ? 0 : device + 1; while (device < SNDRV_RAWMIDI_DEVICES) { -- cgit v0.10.2 From 7839d956fc6aecbb66d645b4050e8e88e2e821cd Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 9 Sep 2010 00:02:03 +0100 Subject: drm/i915: Double check that the wait_request is not pending before warning If we are busy, then we may have woken up the wait_request handler but not yet serviced it before the hang check fires. So in hang check, double check that the i915_gem_do_wait_request() is still pending the wake-up before declaring all hope lost. Fixes regression with e78d73b16bcde921c9cf458d2e4de8e4fc2518f3. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30073 Reported-and-tested-by: Sitsofe Wheeler Signed-off-by: Chris Wilson diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 59457e8..744225e 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -1350,17 +1350,25 @@ void i915_hangcheck_elapsed(unsigned long data) i915_seqno_passed(i915_get_gem_seqno(dev, &dev_priv->render_ring), i915_get_tail_request(dev)->seqno)) { + bool missed_wakeup = false; + dev_priv->hangcheck_count = 0; /* Issue a wake-up to catch stuck h/w. */ - if (dev_priv->render_ring.waiting_gem_seqno | - dev_priv->bsd_ring.waiting_gem_seqno) { - DRM_ERROR("Hangcheck timer elapsed... GPU idle, missed IRQ.\n"); - if (dev_priv->render_ring.waiting_gem_seqno) - DRM_WAKEUP(&dev_priv->render_ring.irq_queue); - if (dev_priv->bsd_ring.waiting_gem_seqno) - DRM_WAKEUP(&dev_priv->bsd_ring.irq_queue); + if (dev_priv->render_ring.waiting_gem_seqno && + waitqueue_active(&dev_priv->render_ring.irq_queue)) { + DRM_WAKEUP(&dev_priv->render_ring.irq_queue); + missed_wakeup = true; + } + + if (dev_priv->bsd_ring.waiting_gem_seqno && + waitqueue_active(&dev_priv->bsd_ring.irq_queue)) { + DRM_WAKEUP(&dev_priv->bsd_ring.irq_queue); + missed_wakeup = true; } + + if (missed_wakeup) + DRM_ERROR("Hangcheck timer elapsed... GPU idle, missed IRQ.\n"); return; } -- cgit v0.10.2 From d530148ae8bffe1b33f50d1776d185a6e85dc774 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Fri, 20 Aug 2010 16:49:43 +0800 Subject: dquot: do full inode dirty in allocating space Alex Shi found a regression when doing ffsb test. The test has several threads, and each thread creates a small file, write to it and then delete it. ffsb reports about 20% regression and Alex bisected it to 43d2932d88e4. The test will call __mark_inode_dirty 3 times. without this commit, we only take inode_lock one time, while with it, we take the lock 3 times with flags ( I_DIRTY_SYNC,I_DIRTY_PAGES,I_DIRTY). Perf shows the lock contention increased too much. Below proposed patch fixes it. fs is allocating blocks, which usually means file writes and the inode will be dirtied soon. We fully dirty the inode to reduce some inode_lock contention in several calls of __mark_inode_dirty. Jan Kara: Added comment. Signed-off-by: Shaohua Li Signed-off-by: Alex Shi Signed-off-by: Jan Kara diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index d50ba85..d1a9193 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -274,8 +274,14 @@ static inline int dquot_alloc_space(struct inode *inode, qsize_t nr) int ret; ret = dquot_alloc_space_nodirty(inode, nr); - if (!ret) - mark_inode_dirty_sync(inode); + if (!ret) { + /* + * Mark inode fully dirty. Since we are allocating blocks, inode + * would become fully dirty soon anyway and it reportedly + * reduces inode_lock contention. + */ + mark_inode_dirty(inode); + } return ret; } -- cgit v0.10.2 From 3a5c19c23db65a554f2e4f5df5f307c668277056 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Mon, 16 Aug 2010 10:06:26 -0500 Subject: [SCSI] fix use-after-free in scsi_init_io() we're using a pointer through a freed command to reset the request, which has shown up as an oops with slab poisoning: Reported-by: Tejun Heo Reported-by: Alexey Dobriyan Signed-off-by: James Bottomley diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 9ade720..ee02d38 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1011,8 +1011,8 @@ int scsi_init_io(struct scsi_cmnd *cmd, gfp_t gfp_mask) err_exit: scsi_release_buffers(cmd); - scsi_put_command(cmd); cmd->request->special = NULL; + scsi_put_command(cmd); return error; } EXPORT_SYMBOL(scsi_init_io); -- cgit v0.10.2 From 39aa3cb3e8250db9188a6f1e3fb62ffa1a717678 Mon Sep 17 00:00:00 2001 From: Stefan Bader Date: Tue, 31 Aug 2010 15:52:27 +0200 Subject: mm: Move vma_stack_continue into mm.h So it can be used by all that need to check for that. Signed-off-by: Stefan Bader Signed-off-by: Linus Torvalds diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 439fc1f..271afc4 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -224,7 +224,8 @@ static void show_map_vma(struct seq_file *m, struct vm_area_struct *vma) /* We don't show the stack guard page in /proc/maps */ start = vma->vm_start; if (vma->vm_flags & VM_GROWSDOWN) - start += PAGE_SIZE; + if (!vma_stack_continue(vma->vm_prev, vma->vm_start)) + start += PAGE_SIZE; seq_printf(m, "%08lx-%08lx %c%c%c%c %08llx %02x:%02x %lu %n", start, diff --git a/include/linux/mm.h b/include/linux/mm.h index e6b1210..74949fb 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -864,6 +864,12 @@ int set_page_dirty(struct page *page); int set_page_dirty_lock(struct page *page); int clear_page_dirty_for_io(struct page *page); +/* Is the vma a continuation of the stack vma above it? */ +static inline int vma_stack_continue(struct vm_area_struct *vma, unsigned long addr) +{ + return vma && (vma->vm_end == addr) && (vma->vm_flags & VM_GROWSDOWN); +} + extern unsigned long move_page_tables(struct vm_area_struct *vma, unsigned long old_addr, struct vm_area_struct *new_vma, unsigned long new_addr, unsigned long len); diff --git a/mm/mlock.c b/mm/mlock.c index cbae7c5..b70919c 100644 --- a/mm/mlock.c +++ b/mm/mlock.c @@ -135,12 +135,6 @@ void munlock_vma_page(struct page *page) } } -/* Is the vma a continuation of the stack vma above it? */ -static inline int vma_stack_continue(struct vm_area_struct *vma, unsigned long addr) -{ - return vma && (vma->vm_end == addr) && (vma->vm_flags & VM_GROWSDOWN); -} - static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr) { return (vma->vm_flags & VM_GROWSDOWN) && -- cgit v0.10.2 From a73f8844e1fc54c3762555c1cf1f71774142ca91 Mon Sep 17 00:00:00 2001 From: Jonathan Corbet Date: Wed, 8 Sep 2010 16:54:54 -0600 Subject: lglock: make lg_lock_global() actually lock globally lg_lock_global() currently only acquires spinlocks for online CPUs, but it's meant to lock all possible CPUs. Lglock-protected resources may be associated with removed CPUs - and, indeed, that could happen with the per-superblock open files lists. At Nick's suggestion, change for_each_online_cpu() to for_each_possible_cpu() to protect accesses to those resources. Cc: Al Viro Acked-by: Nick Piggin Signed-off-by: Jonathan Corbet Signed-off-by: Linus Torvalds diff --git a/include/linux/lglock.h b/include/linux/lglock.h index b288cb7..f549056 100644 --- a/include/linux/lglock.h +++ b/include/linux/lglock.h @@ -150,7 +150,7 @@ int i; \ preempt_disable(); \ rwlock_acquire(&name##_lock_dep_map, 0, 0, _RET_IP_); \ - for_each_online_cpu(i) { \ + for_each_possible_cpu(i) { \ arch_spinlock_t *lock; \ lock = &per_cpu(name##_lock, i); \ arch_spin_lock(lock); \ @@ -161,7 +161,7 @@ void name##_global_unlock(void) { \ int i; \ rwlock_release(&name##_lock_dep_map, 1, _RET_IP_); \ - for_each_online_cpu(i) { \ + for_each_possible_cpu(i) { \ arch_spinlock_t *lock; \ lock = &per_cpu(name##_lock, i); \ arch_spin_unlock(lock); \ -- cgit v0.10.2 From da908f2fb4e783c2a4de751fb90f11a0dd041161 Mon Sep 17 00:00:00 2001 From: Zachary Amsden Date: Thu, 19 Aug 2010 22:07:27 -1000 Subject: KVM: x86: Perform hardware_enable in CPU_STARTING callback The CPU_STARTING callback was added upstream with the intention of being used for KVM, specifically for the hardware enablement that must be done before we can run in hardware virt. It had bugs on the x86_64 architecture at the time, where it was called after CPU_ONLINE. The arches have since merged and the bug is gone. It might be noted other features should probably start making use of this callback; microcode updates in particular which might be fixing important erratums would be best applied before beginning to run user tasks. Signed-off-by: Zachary Amsden Signed-off-by: Marcelo Tosatti diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index b78b794..d4853a5 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -1958,10 +1958,10 @@ static int kvm_cpu_hotplug(struct notifier_block *notifier, unsigned long val, cpu); hardware_disable(NULL); break; - case CPU_ONLINE: + case CPU_STARTING: printk(KERN_INFO "kvm: enabling virtualization on CPU%d\n", cpu); - smp_call_function_single(cpu, hardware_enable, NULL, 1); + hardware_enable(NULL); break; } return NOTIFY_OK; @@ -2096,7 +2096,6 @@ int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx, static struct notifier_block kvm_cpu_notifier = { .notifier_call = kvm_cpu_hotplug, - .priority = 20, /* must be > scheduler priority */ }; static int vm_stat_get(void *_offset, u64 *val) -- cgit v0.10.2 From 5afddee415c829704d3a91031ae634008bc332e9 Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Thu, 9 Sep 2010 19:58:23 +0200 Subject: AT91: clock: peripheral clocks can have other parent than mck While registering clock allow to set parent clock other than mck. It is useful for clocks than can be seen as child clock of a peripheral. Signed-off-by: Nicolas Ferre Acked-by: Jean-Christophe PLAGNIOL-VILLARD Acked-by: Andrew Victor diff --git a/arch/arm/mach-at91/clock.c b/arch/arm/mach-at91/clock.c index 7f7da43..7525cee 100644 --- a/arch/arm/mach-at91/clock.c +++ b/arch/arm/mach-at91/clock.c @@ -501,7 +501,8 @@ postcore_initcall(at91_clk_debugfs_init); int __init clk_register(struct clk *clk) { if (clk_is_peripheral(clk)) { - clk->parent = &mck; + if (!clk->parent) + clk->parent = &mck; clk->mode = pmc_periph_mode; list_add_tail(&clk->node, &clocks); } -- cgit v0.10.2 From ab64511cbbd03196d84bcc32c6e7b9d46543df7b Mon Sep 17 00:00:00 2001 From: Fabian Godehardt Date: Fri, 3 Sep 2010 13:31:33 +0100 Subject: AT91: SAM9G45 - add a separate clock entry for every single TC block Without this patch you will not be able to register the first block because of the second association call on at91_add_device_tc(). Signed-off-by: Fabian Godehardt [nicolas.ferre@atmel.com: change tcb1_clk to fake child clock of tcb0_clk] Acked-by: Jean-Christophe PLAGNIOL-VILLARD Acked-by: Andrew Victor Signed-off-by: Nicolas Ferre diff --git a/arch/arm/mach-at91/at91sam9g45.c b/arch/arm/mach-at91/at91sam9g45.c index 753c0d3..c67b47f 100644 --- a/arch/arm/mach-at91/at91sam9g45.c +++ b/arch/arm/mach-at91/at91sam9g45.c @@ -121,8 +121,8 @@ static struct clk ssc1_clk = { .pmc_mask = 1 << AT91SAM9G45_ID_SSC1, .type = CLK_TYPE_PERIPHERAL, }; -static struct clk tcb_clk = { - .name = "tcb_clk", +static struct clk tcb0_clk = { + .name = "tcb0_clk", .pmc_mask = 1 << AT91SAM9G45_ID_TCB, .type = CLK_TYPE_PERIPHERAL, }; @@ -192,6 +192,14 @@ static struct clk ohci_clk = { .parent = &uhphs_clk, }; +/* One additional fake clock for second TC block */ +static struct clk tcb1_clk = { + .name = "tcb1_clk", + .pmc_mask = 0, + .type = CLK_TYPE_PERIPHERAL, + .parent = &tcb0_clk, +}; + static struct clk *periph_clocks[] __initdata = { &pioA_clk, &pioB_clk, @@ -208,7 +216,7 @@ static struct clk *periph_clocks[] __initdata = { &spi1_clk, &ssc0_clk, &ssc1_clk, - &tcb_clk, + &tcb0_clk, &pwm_clk, &tsc_clk, &dma_clk, @@ -221,6 +229,7 @@ static struct clk *periph_clocks[] __initdata = { &mmc1_clk, // irq0 &ohci_clk, + &tcb1_clk, }; /* diff --git a/arch/arm/mach-at91/at91sam9g45_devices.c b/arch/arm/mach-at91/at91sam9g45_devices.c index 501ac6f..5e71ccd 100644 --- a/arch/arm/mach-at91/at91sam9g45_devices.c +++ b/arch/arm/mach-at91/at91sam9g45_devices.c @@ -835,9 +835,9 @@ static struct platform_device at91sam9g45_tcb1_device = { static void __init at91_add_device_tc(void) { /* this chip has one clock and irq for all six TC channels */ - at91_clock_associate("tcb_clk", &at91sam9g45_tcb0_device.dev, "t0_clk"); + at91_clock_associate("tcb0_clk", &at91sam9g45_tcb0_device.dev, "t0_clk"); platform_device_register(&at91sam9g45_tcb0_device); - at91_clock_associate("tcb_clk", &at91sam9g45_tcb1_device.dev, "t0_clk"); + at91_clock_associate("tcb1_clk", &at91sam9g45_tcb1_device.dev, "t0_clk"); platform_device_register(&at91sam9g45_tcb1_device); } #else -- cgit v0.10.2 From 673424c0890a00e22398017c9adf999577526220 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Mon, 30 Aug 2010 17:37:05 +0200 Subject: pata_artop: Fix device ID parity check x % 1 always evaluates to 0, which clearly isn't the intent. The author probably had "% 2" or "& 1" in mind, and mispelled it. Signed-off-by: Jean Delvare Cc: Jeff Garzik Cc: Alan Cox Signed-off-by: Jeff Garzik diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c index ba43f0f..2215632 100644 --- a/drivers/ata/pata_artop.c +++ b/drivers/ata/pata_artop.c @@ -74,7 +74,8 @@ static int artop6260_pre_reset(struct ata_link *link, unsigned long deadline) struct pci_dev *pdev = to_pci_dev(ap->host->dev); /* Odd numbered device ids are the units with enable bits (the -R cards) */ - if (pdev->device % 1 && !pci_test_config_bits(pdev, &artop_enable_bits[ap->port_no])) + if ((pdev->device & 1) && + !pci_test_config_bits(pdev, &artop_enable_bits[ap->port_no])) return -ENOENT; return ata_sff_prereset(link, deadline); -- cgit v0.10.2 From f1f5a807b051eddd3f302e503d39214e5bde0ef2 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 27 Aug 2010 11:09:15 +0200 Subject: ahci: fix hang on failed softreset ahci_do_softreset() compared the current time and deadline in reverse when calculating timeout for SRST issue. The result is that if @deadline is in future, SRST is issued with 0 timeout, which hasn't caused any problem because it later waits for DRDY with the correct timeout. If deadline is already exceeded by the time SRST is about to be issued, the timeout calculation underflows and if the device doesn't respond, timeout doesn't trigger for a _very_ long time. Reverse the incorrect comparison order. Signed-off-by: Tejun Heo Reported-by: Anssi Hannula Tested-by: Gwendal Grignou Cc: stable@kernel.org Signed-off-by: Jeff Garzik diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c index 666850d..68dc678 100644 --- a/drivers/ata/libahci.c +++ b/drivers/ata/libahci.c @@ -1326,7 +1326,7 @@ int ahci_do_softreset(struct ata_link *link, unsigned int *class, /* issue the first D2H Register FIS */ msecs = 0; now = jiffies; - if (time_after(now, deadline)) + if (time_after(deadline, now)) msecs = jiffies_to_msecs(deadline - now); tf.ctl |= ATA_SRST; -- cgit v0.10.2 From 9cb627d5f38830ca19aa0dca52d1d3a633018bf7 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 1 Sep 2010 12:58:43 +0200 Subject: perf, trace: Fix module leak Commit 1c024eca (perf, trace: Optimize tracepoints by using per-tracepoint-per-cpu hlist to track events) caused a module refcount leak. Reported-And-Tested-by: Avi Kivity Signed-off-by: Peter Zijlstra LKML-Reference: <4C7E1F12.8030304@cn.fujitsu.com> Signed-off-by: Ingo Molnar diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 000e6e8..31cc4cb 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -91,6 +91,8 @@ int perf_trace_init(struct perf_event *p_event) tp_event->class && tp_event->class->reg && try_module_get(tp_event->mod)) { ret = perf_trace_event_init(tp_event, p_event); + if (ret) + module_put(tp_event->mod); break; } } @@ -146,6 +148,7 @@ void perf_trace_destroy(struct perf_event *p_event) } } out: + module_put(tp_event->mod); mutex_unlock(&event_mutex); } -- cgit v0.10.2 From 5e11637e2c929e34dcc0fbbfb48bdb638937701a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 11 Jun 2010 13:35:08 +0200 Subject: perf: Fix CPU hotplug Since we have UP_PREPARE, we should also have UP_CANCELED. Signed-off-by: Peter Zijlstra Cc: paulus LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 657555a..db5b560 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -5761,15 +5761,15 @@ perf_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) { unsigned int cpu = (long)hcpu; - switch (action) { + switch (action & ~CPU_TASKS_FROZEN) { case CPU_UP_PREPARE: - case CPU_UP_PREPARE_FROZEN: + case CPU_DOWN_FAILED: perf_event_init_cpu(cpu); break; + case CPU_UP_CANCELED: case CPU_DOWN_PREPARE: - case CPU_DOWN_PREPARE_FROZEN: perf_event_exit_cpu(cpu); break; -- cgit v0.10.2 From da2b71edd8a7db44fe1746261410a981f3e03632 Mon Sep 17 00:00:00 2001 From: Suresh Siddha Date: Mon, 23 Aug 2010 13:42:51 -0700 Subject: sched: Move sched_avg_update() to update_cpu_load() Currently sched_avg_update() (which updates rt_avg stats in the rq) is getting called from scale_rt_power() (in the load balance context) which doesn't take rq->lock. Fix it by moving the sched_avg_update() to more appropriate update_cpu_load() where the CFS load gets updated as well. Signed-off-by: Suresh Siddha Signed-off-by: Peter Zijlstra LKML-Reference: <1282596171.2694.3.camel@sbsiddha-MOBL3> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index 09b574e..ed09d4f 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1294,6 +1294,10 @@ static void resched_task(struct task_struct *p) static void sched_rt_avg_update(struct rq *rq, u64 rt_delta) { } + +static void sched_avg_update(struct rq *rq) +{ +} #endif /* CONFIG_SMP */ #if BITS_PER_LONG == 32 @@ -3182,6 +3186,8 @@ static void update_cpu_load(struct rq *this_rq) this_rq->cpu_load[i] = (old_load * (scale - 1) + new_load) >> i; } + + sched_avg_update(this_rq); } static void update_cpu_load_active(struct rq *this_rq) diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index ab661eb..f53ec75 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -2268,8 +2268,6 @@ unsigned long scale_rt_power(int cpu) struct rq *rq = cpu_rq(cpu); u64 total, available; - sched_avg_update(rq); - total = sched_avg_period() + (rq->clock - rq->age_stamp); available = total - rq->rt_avg; -- cgit v0.10.2 From 0e52247a2ed1f211f0c4f682dc999610a368903f Mon Sep 17 00:00:00 2001 From: Cam Macdonell Date: Tue, 7 Sep 2010 17:25:20 -0700 Subject: PCI: fix pci_resource_alignment prototype This fixes the prototype for both pci_resource_alignment() and pci_sriov_resource_alignment(). Patch started as debugging effort from Cam Macdonell. Cc: Cam Macdonell Cc: Avi Kivity [chrisw: add iov bits] Signed-off-by: Chris Wright Signed-off-by: Jesse Barnes diff --git a/drivers/pci/iov.c b/drivers/pci/iov.c index ce6a366..553d8ee 100644 --- a/drivers/pci/iov.c +++ b/drivers/pci/iov.c @@ -608,7 +608,7 @@ int pci_iov_resource_bar(struct pci_dev *dev, int resno, * the VF BAR size multiplied by the number of VFs. The alignment * is just the VF BAR size. */ -int pci_sriov_resource_alignment(struct pci_dev *dev, int resno) +resource_size_t pci_sriov_resource_alignment(struct pci_dev *dev, int resno) { struct resource tmp; enum pci_bar_type type; diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index 7754a67..6beb11b 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -264,7 +264,8 @@ extern int pci_iov_init(struct pci_dev *dev); extern void pci_iov_release(struct pci_dev *dev); extern int pci_iov_resource_bar(struct pci_dev *dev, int resno, enum pci_bar_type *type); -extern int pci_sriov_resource_alignment(struct pci_dev *dev, int resno); +extern resource_size_t pci_sriov_resource_alignment(struct pci_dev *dev, + int resno); extern void pci_restore_iov_state(struct pci_dev *dev); extern int pci_iov_bus_range(struct pci_bus *bus); @@ -320,7 +321,7 @@ static inline int pci_ats_enabled(struct pci_dev *dev) } #endif /* CONFIG_PCI_IOV */ -static inline int pci_resource_alignment(struct pci_dev *dev, +static inline resource_size_t pci_resource_alignment(struct pci_dev *dev, struct resource *res) { #ifdef CONFIG_PCI_IOV -- cgit v0.10.2 From 85e00b55517a5696282b80836655e6b6f5702df0 Mon Sep 17 00:00:00 2001 From: Jovi Zhang Date: Thu, 9 Sep 2010 13:30:59 -0300 Subject: perf symbols: Fix multiple initialization of symbol system By returning immediately if it was already initialized, do it as well at symbol__exit, refusing multiple deinitializations. This fixes problems in the kmem, sched and timechart commands. Reported-by: Davidlohr Bueso Cc: Davidlohr Bueso Signed-off-by: Jovi Zhang LKML-Reference: AANLkTi=9Cn=R8SPMCRp5z+gEjXbaBHeb-AaOtRbuwwcn@mail.gmail.com Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 1a36773..b2f5ae9 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -2268,6 +2268,9 @@ static int setup_list(struct strlist **list, const char *list_str, int symbol__init(void) { + if (symbol_conf.initialized) + return 0; + elf_version(EV_CURRENT); if (symbol_conf.sort_by_name) symbol_conf.priv_size += (sizeof(struct symbol_name_rb_node) - @@ -2293,6 +2296,7 @@ int symbol__init(void) symbol_conf.sym_list_str, "symbol") < 0) goto out_free_comm_list; + symbol_conf.initialized = true; return 0; out_free_dso_list: @@ -2304,11 +2308,14 @@ out_free_comm_list: void symbol__exit(void) { + if (!symbol_conf.initialized) + return; strlist__delete(symbol_conf.sym_list); strlist__delete(symbol_conf.dso_list); strlist__delete(symbol_conf.comm_list); vmlinux_path__exit(); symbol_conf.sym_list = symbol_conf.dso_list = symbol_conf.comm_list = NULL; + symbol_conf.initialized = false; } int machines__create_kernel_maps(struct rb_root *self, pid_t pid) diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index b7a8da4..ea95c27 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -69,7 +69,8 @@ struct symbol_conf { show_nr_samples, use_callchain, exclude_other, - show_cpu_utilization; + show_cpu_utilization, + initialized; const char *vmlinux_name, *source_prefix, *field_sep; -- cgit v0.10.2 From 9e84ed63dc71e13b62cea5ec6b0049260cca0b7a Mon Sep 17 00:00:00 2001 From: Russell King Date: Thu, 9 Sep 2010 22:39:41 +0100 Subject: ARM: Partially revert "Auto calculate ZRELADDR and provide option for exceptions" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Partially revert e69edc7, which introduced automatic zreladdr support. The change in the way the manual definition is defined seems to be error and conflict prone. Go back to the original way we were handling this for the time being, while keeping the automatic zreladdr facility. Acked-by: Uwe Kleine-König diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 16bc8eb..553b7cf 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1576,97 +1576,6 @@ config AUTO_ZRELADDR 0xf8000000. This assumes the zImage being placed in the first 128MB from start of memory. -config ZRELADDR - hex "Physical address of the decompressed kernel image" - depends on !AUTO_ZRELADDR - default 0x00008000 if ARCH_BCMRING ||\ - ARCH_CNS3XXX ||\ - ARCH_DOVE ||\ - ARCH_EBSA110 ||\ - ARCH_FOOTBRIDGE ||\ - ARCH_INTEGRATOR ||\ - ARCH_IOP13XX ||\ - ARCH_IOP33X ||\ - ARCH_IXP2000 ||\ - ARCH_IXP23XX ||\ - ARCH_IXP4XX ||\ - ARCH_KIRKWOOD ||\ - ARCH_KS8695 ||\ - ARCH_LOKI ||\ - ARCH_MMP ||\ - ARCH_MV78XX0 ||\ - ARCH_NOMADIK ||\ - ARCH_NUC93X ||\ - ARCH_NS9XXX ||\ - ARCH_ORION5X ||\ - ARCH_SPEAR3XX ||\ - ARCH_SPEAR6XX ||\ - ARCH_TEGRA ||\ - ARCH_U8500 ||\ - ARCH_VERSATILE ||\ - ARCH_W90X900 - default 0x08008000 if ARCH_MX1 ||\ - ARCH_SHARK - default 0x10008000 if ARCH_MSM ||\ - ARCH_OMAP1 ||\ - ARCH_RPC - default 0x20008000 if ARCH_S5P6440 ||\ - ARCH_S5P6442 ||\ - ARCH_S5PC100 ||\ - ARCH_S5PV210 - default 0x30008000 if ARCH_S3C2410 ||\ - ARCH_S3C2400 ||\ - ARCH_S3C2412 ||\ - ARCH_S3C2416 ||\ - ARCH_S3C2440 ||\ - ARCH_S3C2443 - default 0x40008000 if ARCH_STMP378X ||\ - ARCH_STMP37XX ||\ - ARCH_SH7372 ||\ - ARCH_SH7377 ||\ - ARCH_S5PV310 - default 0x50008000 if ARCH_S3C64XX ||\ - ARCH_SH7367 - default 0x60008000 if ARCH_VEXPRESS - default 0x80008000 if ARCH_MX25 ||\ - ARCH_MX3 ||\ - ARCH_NETX ||\ - ARCH_OMAP2PLUS ||\ - ARCH_PNX4008 - default 0x90008000 if ARCH_MX5 ||\ - ARCH_MX91231 - default 0xa0008000 if ARCH_IOP32X ||\ - ARCH_PXA ||\ - MACH_MX27 - default 0xc0008000 if ARCH_LH7A40X ||\ - MACH_MX21 - default 0xf0008000 if ARCH_AAEC2000 ||\ - ARCH_L7200 - default 0xc0028000 if ARCH_CLPS711X - default 0x70008000 if ARCH_AT91 && (ARCH_AT91CAP9 || ARCH_AT91SAM9G45) - default 0x20008000 if ARCH_AT91 && !(ARCH_AT91CAP9 || ARCH_AT91SAM9G45) - default 0xc0008000 if ARCH_DAVINCI && ARCH_DAVINCI_DA8XX - default 0x80008000 if ARCH_DAVINCI && !ARCH_DAVINCI_DA8XX - default 0x00008000 if ARCH_EP93XX && EP93XX_SDCE3_SYNC_PHYS_OFFSET - default 0xc0008000 if ARCH_EP93XX && EP93XX_SDCE0_PHYS_OFFSET - default 0xd0008000 if ARCH_EP93XX && EP93XX_SDCE1_PHYS_OFFSET - default 0xe0008000 if ARCH_EP93XX && EP93XX_SDCE2_PHYS_OFFSET - default 0xf0008000 if ARCH_EP93XX && EP93XX_SDCE3_ASYNC_PHYS_OFFSET - default 0x00008000 if ARCH_GEMINI && GEMINI_MEM_SWAP - default 0x10008000 if ARCH_GEMINI && !GEMINI_MEM_SWAP - default 0x70008000 if ARCH_REALVIEW && REALVIEW_HIGH_PHYS_OFFSET - default 0x00008000 if ARCH_REALVIEW && !REALVIEW_HIGH_PHYS_OFFSET - default 0xc0208000 if ARCH_SA1100 && SA1111 - default 0xc0008000 if ARCH_SA1100 && !SA1111 - default 0x30108000 if ARCH_S3C2410 && PM_H1940 - default 0x28E08000 if ARCH_U300 && MACH_U300_SINGLE_RAM - default 0x48008000 if ARCH_U300 && !MACH_U300_SINGLE_RAM - help - ZRELADDR is the physical address where the decompressed kernel - image will be placed. ZRELADDR has to be specified when the - assumption of AUTO_ZRELADDR is not valid, or when ZBOOT_ROM is - selected. - endmenu menu "CPU Power Management" diff --git a/arch/arm/boot/Makefile b/arch/arm/boot/Makefile index f705213..4a590f4 100644 --- a/arch/arm/boot/Makefile +++ b/arch/arm/boot/Makefile @@ -14,16 +14,18 @@ MKIMAGE := $(srctree)/scripts/mkuboot.sh ifneq ($(MACHINE),) --include $(srctree)/$(MACHINE)/Makefile.boot +include $(srctree)/$(MACHINE)/Makefile.boot endif # Note: the following conditions must always be true: +# ZRELADDR == virt_to_phys(PAGE_OFFSET + TEXT_OFFSET) # PARAMS_PHYS must be within 4MB of ZRELADDR # INITRD_PHYS must be in RAM +ZRELADDR := $(zreladdr-y) PARAMS_PHYS := $(params_phys-y) INITRD_PHYS := $(initrd_phys-y) -export INITRD_PHYS PARAMS_PHYS +export ZRELADDR INITRD_PHYS PARAMS_PHYS targets := Image zImage xipImage bootpImage uImage @@ -65,7 +67,7 @@ quiet_cmd_uimage = UIMAGE $@ ifeq ($(CONFIG_ZBOOT_ROM),y) $(obj)/uImage: LOADADDR=$(CONFIG_ZBOOT_ROM_TEXT) else -$(obj)/uImage: LOADADDR=$(CONFIG_ZRELADDR) +$(obj)/uImage: LOADADDR=$(ZRELADDR) endif ifeq ($(CONFIG_THUMB2_KERNEL),y) diff --git a/arch/arm/boot/compressed/Makefile b/arch/arm/boot/compressed/Makefile index 68775e3..b23f6bc 100644 --- a/arch/arm/boot/compressed/Makefile +++ b/arch/arm/boot/compressed/Makefile @@ -79,6 +79,10 @@ endif EXTRA_CFLAGS := -fpic -fno-builtin EXTRA_AFLAGS := -Wa,-march=all +# Supply ZRELADDR to the decompressor via a linker symbol. +ifneq ($(CONFIG_AUTO_ZRELADDR),y) +LDFLAGS_vmlinux := --defsym zreladdr=$(ZRELADDR) +endif ifeq ($(CONFIG_CPU_ENDIAN_BE8),y) LDFLAGS_vmlinux += --be8 endif diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S index 6af9907..6825c34 100644 --- a/arch/arm/boot/compressed/head.S +++ b/arch/arm/boot/compressed/head.S @@ -177,7 +177,7 @@ not_angel: and r4, pc, #0xf8000000 add r4, r4, #TEXT_OFFSET #else - ldr r4, =CONFIG_ZRELADDR + ldr r4, =zreladdr #endif subs r0, r0, r1 @ calculate the delta offset -- cgit v0.10.2 From a14d0404088a7b0d51bb5219baf33d77e5592007 Mon Sep 17 00:00:00 2001 From: Russell King Date: Thu, 9 Sep 2010 22:43:02 +0100 Subject: ARM: Update mach-types Signed-off-by: Russell King diff --git a/arch/arm/tools/mach-types b/arch/arm/tools/mach-types index 48cbdcb..55590a4 100644 --- a/arch/arm/tools/mach-types +++ b/arch/arm/tools/mach-types @@ -12,7 +12,7 @@ # # http://www.arm.linux.org.uk/developer/machines/?action=new # -# Last update: Mon Jul 12 21:10:14 2010 +# Last update: Thu Sep 9 22:43:01 2010 # # machine_is_xxx CONFIG_xxxx MACH_TYPE_xxx number # @@ -2622,7 +2622,7 @@ kraken MACH_KRAKEN KRAKEN 2634 gw2388 MACH_GW2388 GW2388 2635 jadecpu MACH_JADECPU JADECPU 2636 carlisle MACH_CARLISLE CARLISLE 2637 -lux_sf9 MACH_LUX_SFT9 LUX_SFT9 2638 +lux_sf9 MACH_LUX_SF9 LUX_SF9 2638 nemid_tb MACH_NEMID_TB NEMID_TB 2639 terrier MACH_TERRIER TERRIER 2640 turbot MACH_TURBOT TURBOT 2641 @@ -2950,3 +2950,97 @@ davinci_dm365_dvr MACH_DAVINCI_DM365_DVR DAVINCI_DM365_DVR 2963 netviz MACH_NETVIZ NETVIZ 2964 flexibity MACH_FLEXIBITY FLEXIBITY 2965 wlan_computer MACH_WLAN_COMPUTER WLAN_COMPUTER 2966 +lpc24xx MACH_LPC24XX LPC24XX 2967 +spica MACH_SPICA SPICA 2968 +gpsdisplay MACH_GPSDISPLAY GPSDISPLAY 2969 +bipnet MACH_BIPNET BIPNET 2970 +overo_ctu_inertial MACH_OVERO_CTU_INERTIAL OVERO_CTU_INERTIAL 2971 +davinci_dm355_mmm MACH_DAVINCI_DM355_MMM DAVINCI_DM355_MMM 2972 +pc9260_v2 MACH_PC9260_V2 PC9260_V2 2973 +ptx7545 MACH_PTX7545 PTX7545 2974 +tm_efdc MACH_TM_EFDC TM_EFDC 2975 +omap3_waldo1 MACH_OMAP3_WALDO1 OMAP3_WALDO1 2977 +flyer MACH_FLYER FLYER 2978 +tornado3240 MACH_TORNADO3240 TORNADO3240 2979 +soli_01 MACH_SOLI_01 SOLI_01 2980 +omapl138_europalc MACH_OMAPL138_EUROPALC OMAPL138_EUROPALC 2981 +helios_v1 MACH_HELIOS_V1 HELIOS_V1 2982 +netspace_lite_v2 MACH_NETSPACE_LITE_V2 NETSPACE_LITE_V2 2983 +ssc MACH_SSC SSC 2984 +premierwave_en MACH_PREMIERWAVE_EN PREMIERWAVE_EN 2985 +wasabi MACH_WASABI WASABI 2986 +vivow MACH_VIVOW VIVOW 2987 +mx50_rdp MACH_MX50_RDP MX50_RDP 2988 +universal MACH_UNIVERSAL UNIVERSAL 2989 +real6410 MACH_REAL6410 REAL6410 2990 +spx_sakura MACH_SPX_SAKURA SPX_SAKURA 2991 +ij3k_2440 MACH_IJ3K_2440 IJ3K_2440 2992 +omap3_bc10 MACH_OMAP3_BC10 OMAP3_BC10 2993 +thebe MACH_THEBE THEBE 2994 +rv082 MACH_RV082 RV082 2995 +armlguest MACH_ARMLGUEST ARMLGUEST 2996 +tjinc1000 MACH_TJINC1000 TJINC1000 2997 +dockstar MACH_DOCKSTAR DOCKSTAR 2998 +ax8008 MACH_AX8008 AX8008 2999 +gnet_sgce MACH_GNET_SGCE GNET_SGCE 3000 +pxwnas_500_1000 MACH_PXWNAS_500_1000 PXWNAS_500_1000 3001 +ea20 MACH_EA20 EA20 3002 +awm2 MACH_AWM2 AWM2 3003 +ti8148evm MACH_TI8148EVM TI8148EVM 3004 +tegra_seaboard MACH_TEGRA_SEABOARD TEGRA_SEABOARD 3005 +linkstation_chlv2 MACH_LINKSTATION_CHLV2 LINKSTATION_CHLV2 3006 +tera_pro2_rack MACH_TERA_PRO2_RACK TERA_PRO2_RACK 3007 +rubys MACH_RUBYS RUBYS 3008 +aquarius MACH_AQUARIUS AQUARIUS 3009 +mx53_ard MACH_MX53_ARD MX53_ARD 3010 +mx53_smd MACH_MX53_SMD MX53_SMD 3011 +lswxl MACH_LSWXL LSWXL 3012 +dove_avng_v3 MACH_DOVE_AVNG_V3 DOVE_AVNG_V3 3013 +sdi_ess_9263 MACH_SDI_ESS_9263 SDI_ESS_9263 3014 +jocpu550 MACH_JOCPU550 JOCPU550 3015 +msm8x60_rumi3 MACH_MSM8X60_RUMI3 MSM8X60_RUMI3 3016 +msm8x60_ffa MACH_MSM8X60_FFA MSM8X60_FFA 3017 +yanomami MACH_YANOMAMI YANOMAMI 3018 +gta04 MACH_GTA04 GTA04 3019 +cm_a510 MACH_CM_A510 CM_A510 3020 +omap3_rfs200 MACH_OMAP3_RFS200 OMAP3_RFS200 3021 +kx33xx MACH_KX33XX KX33XX 3022 +ptx7510 MACH_PTX7510 PTX7510 3023 +top9000 MACH_TOP9000 TOP9000 3024 +teenote MACH_TEENOTE TEENOTE 3025 +ts3 MACH_TS3 TS3 3026 +a0 MACH_A0 A0 3027 +fsm9xxx_surf MACH_FSM9XXX_SURF FSM9XXX_SURF 3028 +fsm9xxx_ffa MACH_FSM9XXX_FFA FSM9XXX_FFA 3029 +frrhwcdma60w MACH_FRRHWCDMA60W FRRHWCDMA60W 3030 +remus MACH_REMUS REMUS 3031 +at91cap7xdk MACH_AT91CAP7XDK AT91CAP7XDK 3032 +at91cap7stk MACH_AT91CAP7STK AT91CAP7STK 3033 +kt_sbc_sam9_1 MACH_KT_SBC_SAM9_1 KT_SBC_SAM9_1 3034 +oratisrouter MACH_ORATISROUTER ORATISROUTER 3035 +armada_xp_db MACH_ARMADA_XP_DB ARMADA_XP_DB 3036 +spdm MACH_SPDM SPDM 3037 +gtib MACH_GTIB GTIB 3038 +dgm3240 MACH_DGM3240 DGM3240 3039 +atlas_i_lpe MACH_ATLAS_I_LPE ATLAS_I_LPE 3040 +htcmega MACH_HTCMEGA HTCMEGA 3041 +tricorder MACH_TRICORDER TRICORDER 3042 +tx28 MACH_TX28 TX28 3043 +bstbrd MACH_BSTBRD BSTBRD 3044 +pwb3090 MACH_PWB3090 PWB3090 3045 +idea6410 MACH_IDEA6410 IDEA6410 3046 +qbc9263 MACH_QBC9263 QBC9263 3047 +borabora MACH_BORABORA BORABORA 3048 +valdez MACH_VALDEZ VALDEZ 3049 +ls9g20 MACH_LS9G20 LS9G20 3050 +mios_v1 MACH_MIOS_V1 MIOS_V1 3051 +s5pc110_crespo MACH_S5PC110_CRESPO S5PC110_CRESPO 3052 +controltek9g20 MACH_CONTROLTEK9G20 CONTROLTEK9G20 3053 +tin307 MACH_TIN307 TIN307 3054 +tin510 MACH_TIN510 TIN510 3055 +bluecheese MACH_BLUECHEESE BLUECHEESE 3057 +tem3x30 MACH_TEM3X30 TEM3X30 3058 +harvest_desoto MACH_HARVEST_DESOTO HARVEST_DESOTO 3059 +msm8x60_qrdc MACH_MSM8X60_QRDC MSM8X60_QRDC 3060 +spear900 MACH_SPEAR900 SPEAR900 3061 +pcontrol_g20 MACH_PCONTROL_G20 PCONTROL_G20 3062 -- cgit v0.10.2 From f574c843191728d9407b766a027f779dcd27b272 Mon Sep 17 00:00:00 2001 From: Tony Luck Date: Thu, 9 Sep 2010 15:16:56 -0700 Subject: [IA64] fix siglock When ia64 converted to using ticket locks, an inline implementation of trylock/unlock in fsys.S was missed. This was not noticed because in most circumstances it simply resulted in using the slow path because the siglock was apparently not available (under old spinlock rules). Problems occur when the ticket spinlock has value 0x0 (when first initialised, or when it wraps around). At this point the fsys.S code acquires the lock (changing the 0x0 to 0x1. If another process attempts to get the lock at this point, it will change the value from 0x1 to 0x2 (using new ticket lock rules). Then the fsys.S code will free the lock using old spinlock rules by writing 0x0 to it. From here a variety of bad things can happen. Signed-off-by: Tony Luck diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S index 3567d54..471a1e7 100644 --- a/arch/ia64/kernel/fsys.S +++ b/arch/ia64/kernel/fsys.S @@ -424,14 +424,26 @@ EX(.fail_efault, ld8 r14=[r33]) // r14 <- *set andcm r14=r14,r17 // filter out SIGKILL & SIGSTOP #ifdef CONFIG_SMP - mov r17=1 - ;; - cmpxchg4.acq r18=[r31],r17,ar.ccv // try to acquire the lock + // __ticket_spin_trylock(r31) + ld4 r17=[r31] mov r8=EINVAL // default to EINVAL ;; + extr r9=r17,17,15 + ;; + xor r18=r17,r9 + adds r19=1,r17 + ;; + extr.u r18=r18,0,15 + ;; + cmp.eq p0,p7=0,r18 +(p7) br.cond.spnt.many .lock_contention + mov.m ar.ccv=r17 + ;; + cmpxchg4.acq r9=[r31],r19,ar.ccv + ;; + cmp4.eq p0,p7=r9,r17 +(p7) br.cond.spnt.many .lock_contention ld8 r3=[r2] // re-read current->blocked now that we hold the lock - cmp4.ne p6,p0=r18,r0 -(p6) br.cond.spnt.many .lock_contention ;; #else ld8 r3=[r2] // re-read current->blocked now that we hold the lock @@ -490,7 +502,17 @@ EX(.fail_efault, ld8 r14=[r33]) // r14 <- *set (p6) br.cond.spnt.few 1b // yes -> retry #ifdef CONFIG_SMP - st4.rel [r31]=r0 // release the lock + // __ticket_spin_unlock(r31) + adds r31=2,r31 + ;; + ld2.bias r2=[r31] + mov r3=65534 + ;; + adds r2=2,r2 + ;; + and r3=r3,r2 + ;; + st2.rel [r31]=r3 #endif SSM_PSR_I(p0, p9, r31) ;; @@ -512,7 +534,17 @@ EX(.fail_efault, (p15) st8 [r34]=r3) .sig_pending: #ifdef CONFIG_SMP - st4.rel [r31]=r0 // release the lock + // __ticket_spin_unlock(r31) + adds r31=2,r31 + ;; + ld2.bias r2=[r31] + mov r3=65534 + ;; + adds r2=2,r2 + ;; + and r3=r3,r2 + ;; + st2.rel [r31]=r3 #endif SSM_PSR_I(p0, p9, r17) ;; -- cgit v0.10.2 From f3c65b2870f2481f3646bc410a58a12989ecc704 Mon Sep 17 00:00:00 2001 From: David Vrabel Date: Thu, 9 Sep 2010 16:37:24 -0700 Subject: mmc: avoid getting CID on SDIO-only cards The introduction of support for SD combo cards breaks the initialization of all CSR SDIO chips. The GO_IDLE (CMD0) in mmc_sd_get_cid() causes CSR chips to be reset (this is non-standard behavior). When initializing an SDIO card check for a combo card by using the memory present bit in the R4 response to IO_SEND_OP_COND (CMD5). This avoids the call to mmc_sd_get_cid() on an SDIO-only card. Signed-off-by: David Vrabel Acked-by: Michal Mirolaw Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c index bd2755e..f332c52 100644 --- a/drivers/mmc/core/sdio.c +++ b/drivers/mmc/core/sdio.c @@ -362,9 +362,8 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, goto err; } - err = mmc_sd_get_cid(host, host->ocr & ocr, card->raw_cid); - - if (!err) { + if (ocr & R4_MEMORY_PRESENT + && mmc_sd_get_cid(host, host->ocr & ocr, card->raw_cid) == 0) { card->type = MMC_TYPE_SD_COMBO; if (oldcard && (oldcard->type != MMC_TYPE_SD_COMBO || diff --git a/include/linux/mmc/sdio.h b/include/linux/mmc/sdio.h index 329a8fa..245cdac 100644 --- a/include/linux/mmc/sdio.h +++ b/include/linux/mmc/sdio.h @@ -38,6 +38,8 @@ * [8:0] Byte/block count */ +#define R4_MEMORY_PRESENT (1 << 27) + /* SDIO status in R5 Type -- cgit v0.10.2 From e0bf1024b36be90da241af3c2767311e055b612c Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 9 Sep 2010 16:37:26 -0700 Subject: kfifo: add parenthesis for macro parameter reference Some macro parameter references inside typeof() operator are not enclosed with parenthesis. It should be safer to add them. Signed-off-by: Huang Ying Acked-by: Stefani Seibold Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kfifo.h b/include/linux/kfifo.h index 4aa95f2..62dbee5 100644 --- a/include/linux/kfifo.h +++ b/include/linux/kfifo.h @@ -214,7 +214,7 @@ __kfifo_must_check_helper(unsigned int val) */ #define kfifo_reset(fifo) \ (void)({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ __tmp->kfifo.in = __tmp->kfifo.out = 0; \ }) @@ -228,7 +228,7 @@ __kfifo_must_check_helper(unsigned int val) */ #define kfifo_reset_out(fifo) \ (void)({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ __tmp->kfifo.out = __tmp->kfifo.in; \ }) @@ -238,7 +238,7 @@ __kfifo_must_check_helper(unsigned int val) */ #define kfifo_len(fifo) \ ({ \ - typeof(fifo + 1) __tmpl = (fifo); \ + typeof((fifo) + 1) __tmpl = (fifo); \ __tmpl->kfifo.in - __tmpl->kfifo.out; \ }) @@ -248,7 +248,7 @@ __kfifo_must_check_helper(unsigned int val) */ #define kfifo_is_empty(fifo) \ ({ \ - typeof(fifo + 1) __tmpq = (fifo); \ + typeof((fifo) + 1) __tmpq = (fifo); \ __tmpq->kfifo.in == __tmpq->kfifo.out; \ }) @@ -258,7 +258,7 @@ __kfifo_must_check_helper(unsigned int val) */ #define kfifo_is_full(fifo) \ ({ \ - typeof(fifo + 1) __tmpq = (fifo); \ + typeof((fifo) + 1) __tmpq = (fifo); \ kfifo_len(__tmpq) > __tmpq->kfifo.mask; \ }) @@ -269,7 +269,7 @@ __kfifo_must_check_helper(unsigned int val) #define kfifo_avail(fifo) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmpq = (fifo); \ + typeof((fifo) + 1) __tmpq = (fifo); \ const size_t __recsize = sizeof(*__tmpq->rectype); \ unsigned int __avail = kfifo_size(__tmpq) - kfifo_len(__tmpq); \ (__recsize) ? ((__avail <= __recsize) ? 0 : \ @@ -284,7 +284,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_skip(fifo) \ (void)({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ if (__recsize) \ @@ -302,7 +302,7 @@ __kfifo_must_check_helper( \ #define kfifo_peek_len(fifo) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ (!__recsize) ? kfifo_len(__tmp) * sizeof(*__tmp->type) : \ @@ -325,7 +325,7 @@ __kfifo_must_check_helper( \ #define kfifo_alloc(fifo, size, gfp_mask) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ __is_kfifo_ptr(__tmp) ? \ __kfifo_alloc(__kfifo, size, sizeof(*__tmp->type), gfp_mask) : \ @@ -339,7 +339,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_free(fifo) \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ if (__is_kfifo_ptr(__tmp)) \ __kfifo_free(__kfifo); \ @@ -358,7 +358,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_init(fifo, buffer, size) \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ __is_kfifo_ptr(__tmp) ? \ __kfifo_init(__kfifo, buffer, size, sizeof(*__tmp->type)) : \ @@ -379,8 +379,8 @@ __kfifo_must_check_helper( \ */ #define kfifo_put(fifo, val) \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ - typeof(val + 1) __val = (val); \ + typeof((fifo) + 1) __tmp = (fifo); \ + typeof((val) + 1) __val = (val); \ unsigned int __ret; \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -421,8 +421,8 @@ __kfifo_must_check_helper( \ #define kfifo_get(fifo, val) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ - typeof(val + 1) __val = (val); \ + typeof((fifo) + 1) __tmp = (fifo); \ + typeof((val) + 1) __val = (val); \ unsigned int __ret; \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -462,8 +462,8 @@ __kfifo_must_check_helper( \ #define kfifo_peek(fifo, val) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ - typeof(val + 1) __val = (val); \ + typeof((fifo) + 1) __tmp = (fifo); \ + typeof((val) + 1) __val = (val); \ unsigned int __ret; \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -501,8 +501,8 @@ __kfifo_must_check_helper( \ */ #define kfifo_in(fifo, buf, n) \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ - typeof(buf + 1) __buf = (buf); \ + typeof((fifo) + 1) __tmp = (fifo); \ + typeof((buf) + 1) __buf = (buf); \ unsigned long __n = (n); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -554,8 +554,8 @@ __kfifo_must_check_helper( \ #define kfifo_out(fifo, buf, n) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ - typeof(buf + 1) __buf = (buf); \ + typeof((fifo) + 1) __tmp = (fifo); \ + typeof((buf) + 1) __buf = (buf); \ unsigned long __n = (n); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -611,7 +611,7 @@ __kfifo_must_check_helper( \ #define kfifo_from_user(fifo, from, len, copied) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ const void __user *__from = (from); \ unsigned int __len = (len); \ unsigned int *__copied = (copied); \ @@ -639,7 +639,7 @@ __kfifo_must_check_helper( \ #define kfifo_to_user(fifo, to, len, copied) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ void __user *__to = (to); \ unsigned int __len = (len); \ unsigned int *__copied = (copied); \ @@ -666,7 +666,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_dma_in_prepare(fifo, sgl, nents, len) \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ struct scatterlist *__sgl = (sgl); \ int __nents = (nents); \ unsigned int __len = (len); \ @@ -690,7 +690,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_dma_in_finish(fifo, len) \ (void)({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ unsigned int __len = (len); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -717,7 +717,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_dma_out_prepare(fifo, sgl, nents, len) \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ struct scatterlist *__sgl = (sgl); \ int __nents = (nents); \ unsigned int __len = (len); \ @@ -741,7 +741,7 @@ __kfifo_must_check_helper( \ */ #define kfifo_dma_out_finish(fifo, len) \ (void)({ \ - typeof(fifo + 1) __tmp = (fifo); \ + typeof((fifo) + 1) __tmp = (fifo); \ unsigned int __len = (len); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ @@ -766,8 +766,8 @@ __kfifo_must_check_helper( \ #define kfifo_out_peek(fifo, buf, n) \ __kfifo_must_check_helper( \ ({ \ - typeof(fifo + 1) __tmp = (fifo); \ - typeof(buf + 1) __buf = (buf); \ + typeof((fifo) + 1) __tmp = (fifo); \ + typeof((buf) + 1) __buf = (buf); \ unsigned long __n = (n); \ const size_t __recsize = sizeof(*__tmp->rectype); \ struct __kfifo *__kfifo = &__tmp->kfifo; \ -- cgit v0.10.2 From 110b7e9698601b28f313c2c560d51a8b1c742002 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Thu, 9 Sep 2010 16:37:27 -0700 Subject: rtc-bfin: fix inverted logic in suspend path The int_clear helper takes a bitmask of interrupts to keep, not to disable. When suspending without wakeup enabled, we want to disable all interrupts, so use 0 (keep none) instead of -1 (keep all). Signed-off-by: Mike Frysinger Acked-by: Alessandro Zummo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-bfin.c b/drivers/rtc/rtc-bfin.c index 72b2bcc..65facfb 100644 --- a/drivers/rtc/rtc-bfin.c +++ b/drivers/rtc/rtc-bfin.c @@ -426,7 +426,7 @@ static int bfin_rtc_suspend(struct platform_device *pdev, pm_message_t state) enable_irq_wake(IRQ_RTC); bfin_rtc_sync_pending(&pdev->dev); } else - bfin_rtc_int_clear(-1); + bfin_rtc_int_clear(0); return 0; } -- cgit v0.10.2 From b6de860651d5a9e56ba4f4e3edc1aa52ac2ac849 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Thu, 9 Sep 2010 16:37:29 -0700 Subject: rtc-bfin: fix state restoration when resuming Much (but not all) of the RTC state is kept in the RTC peripheral which has its own power domain. Periodically (1 HZ), that state is synced from one power domain to the other (peripheral->core). When we are resuming, we need to wait for the sync to occur so that we don't get a mismatch of reading undefined state in the rest of the driver. Further, once the externally maintained bits have been synced back into the core, we then need to restore the bits maintained in the core. In our particular case, that is just the write completion interrupt bit. If we don't do any of this, working with the RTC causes ~5 second delays from time to time after waking up due to the write completion interrupt never firing. Reported-by: Michael Dean Reported-by: Michael Hennerich Signed-off-by: Mike Frysinger Acked-by: Alessandro Zummo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-bfin.c b/drivers/rtc/rtc-bfin.c index 65facfb..d4fb82d 100644 --- a/drivers/rtc/rtc-bfin.c +++ b/drivers/rtc/rtc-bfin.c @@ -435,8 +435,17 @@ static int bfin_rtc_resume(struct platform_device *pdev) { if (device_may_wakeup(&pdev->dev)) disable_irq_wake(IRQ_RTC); - else - bfin_write_RTC_ISTAT(-1); + + /* + * Since only some of the RTC bits are maintained externally in the + * Vbat domain, we need to wait for the RTC MMRs to be synced into + * the core after waking up. This happens every RTC 1HZ. Once that + * has happened, we can go ahead and re-enable the important write + * complete interrupt event. + */ + while (!(bfin_read_RTC_ISTAT() & RTC_ISTAT_SEC)) + continue; + bfin_rtc_int_set(RTC_ISTAT_WRITE_COMPLETE); return 0; } -- cgit v0.10.2 From 841a23ae1e7a8f8758ac915ceaf2e11a1cc235c5 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Thu, 9 Sep 2010 16:37:30 -0700 Subject: MAINTAINERS: remove drivers/char/hpet.c bob.picco@hp.com doesn't work any more and Bob says that he's unlikely to work on hpet.c in the future. Signed-off-by: Jiri Slaby Acked-by: Bob Picco Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 7e189b5..05e5537 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2787,11 +2787,6 @@ S: Maintained F: arch/x86/kernel/hpet.c F: arch/x86/include/asm/hpet.h -HPET: ACPI -M: Bob Picco -S: Maintained -F: drivers/char/hpet.c - HPFS FILESYSTEM M: Mikulas Patocka W: http://artax.karlin.mff.cuni.cz/~mikulas/vyplody/hpfs/index-e.cgi -- cgit v0.10.2 From 4e70598c3b56e6fec551454c495d4d4025834749 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 9 Sep 2010 16:37:31 -0700 Subject: hp_accel: add quirks for HP ProBook 532x and HP Mini 5102 Added missing axis-mapping for HP ProBook 532x and HP Mini 5102. Signed-off-by: Takashi Iwai Cc: Eric Piel Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/hwmon/hp_accel.c b/drivers/hwmon/hp_accel.c index 7580f55..36e9575 100644 --- a/drivers/hwmon/hp_accel.c +++ b/drivers/hwmon/hp_accel.c @@ -221,6 +221,8 @@ static struct dmi_system_id lis3lv02d_dmi_ids[] = { AXIS_DMI_MATCH("HPB442x", "HP ProBook 442", xy_rotated_left), AXIS_DMI_MATCH("HPB452x", "HP ProBook 452", y_inverted), AXIS_DMI_MATCH("HPB522x", "HP ProBook 522", xy_swap), + AXIS_DMI_MATCH("HPB532x", "HP ProBook 532", y_inverted), + AXIS_DMI_MATCH("Mini5102", "HP Mini 5102", xy_rotated_left_usd), { NULL, } /* Laptop models without axis info (yet): * "NC6910" "HP Compaq 6910" -- cgit v0.10.2 From 7a801ac6f5067539ceb5fad0fe90ec49fc156e47 Mon Sep 17 00:00:00 2001 From: Jeff Moyer Date: Thu, 9 Sep 2010 16:37:33 -0700 Subject: O_DIRECT: fix the splitting up of contiguous I/O commit c2c6ca4 (direct-io: do not merge logically non-contiguous requests) introduced a bug whereby all O_DIRECT I/Os were submitted a page at a time to the block layer. The problem is that the code expected dio->block_in_file to correspond to the current page in the dio. In fact, it corresponds to the previous page submitted via submit_page_section. This was purely an oversight, as the dio->cur_page_fs_offset field was introduced for just this purpose. This patch simply uses the correct variable when calculating whether there is a mismatch between contiguous logical blocks and contiguous physical blocks (as described in the comments). I also switched the if conditional following this check to an else if, to ensure that we never call dio_bio_submit twice for the same dio (in theory, this should not happen, anyway). I've tested this by running blktrace and verifying that a 64KB I/O was submitted as a single I/O. I also ran the patched kernel through xfstests' aio tests using xfs, ext4 (with 1k and 4k block sizes) and btrfs and verified that there were no regressions as compared to an unpatched kernel. Signed-off-by: Jeff Moyer Acked-by: Josef Bacik Cc: Christoph Hellwig Cc: Chris Mason Cc: [2.6.35.x] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/direct-io.c b/fs/direct-io.c index 51f270b..48d74c7 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -634,7 +634,7 @@ static int dio_send_cur_page(struct dio *dio) int ret = 0; if (dio->bio) { - loff_t cur_offset = dio->block_in_file << dio->blkbits; + loff_t cur_offset = dio->cur_page_fs_offset; loff_t bio_next_offset = dio->logical_offset_in_bio + dio->bio->bi_size; @@ -659,7 +659,7 @@ static int dio_send_cur_page(struct dio *dio) * Submit now if the underlying fs is about to perform a * metadata read */ - if (dio->boundary) + else if (dio->boundary) dio_bio_submit(dio); } -- cgit v0.10.2 From 2f327dad14aa8bc939a4f0a2d3fdcf64a2d8c09e Mon Sep 17 00:00:00 2001 From: Simon Horman Date: Thu, 9 Sep 2010 16:37:33 -0700 Subject: MAINTAINERS: kexec-tools has moved Kexec tools has been moved to http://kernel.org/pub/linux/utils/kernel/kexec/ as user-space code shouldn't be in /pub/linux/kernel Signed-off-by: Simon Horman Acked-by: "Eric W. Biederman" Cc: "H. Peter Anvin" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 05e5537..a3da87f 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3399,7 +3399,7 @@ F: drivers/s390/kvm/ KEXEC M: Eric Biederman -W: http://ftp.kernel.org/pub/linux/kernel/people/horms/kexec-tools/ +W: http://kernel.org/pub/linux/utils/kernel/kexec/ L: kexec@lists.infradead.org S: Maintained F: include/linux/kexec.h -- cgit v0.10.2 From 85a0fdfd0f967507f3903e8419bc7e408f5a59de Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Thu, 9 Sep 2010 16:37:35 -0700 Subject: gcov: fix null-pointer dereference for certain module types The gcov-kernel infrastructure expects that each object file is loaded only once. This may not be true, e.g. when loading multiple kernel modules which are linked to the same object file. As a result, loading such kernel modules will result in incorrect gcov results while unloading will cause a null-pointer dereference. This patch fixes these problems by changing the gcov-kernel infrastructure so that multiple profiling data sets can be associated with one debugfs entry. It applies to 2.6.36-rc1. Signed-off-by: Peter Oberparleiter Reported-by: Werner Spies Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/gcov/fs.c b/kernel/gcov/fs.c index ef3c3f8..f83972b 100644 --- a/kernel/gcov/fs.c +++ b/kernel/gcov/fs.c @@ -33,10 +33,11 @@ * @children: child nodes * @all: list head for list of all nodes * @parent: parent node - * @info: associated profiling data structure if not a directory - * @ghost: when an object file containing profiling data is unloaded we keep a - * copy of the profiling data here to allow collecting coverage data - * for cleanup code. Such a node is called a "ghost". + * @loaded_info: array of pointers to profiling data sets for loaded object + * files. + * @num_loaded: number of profiling data sets for loaded object files. + * @unloaded_info: accumulated copy of profiling data sets for unloaded + * object files. Used only when gcov_persist=1. * @dentry: main debugfs entry, either a directory or data file * @links: associated symbolic links * @name: data file basename @@ -51,10 +52,11 @@ struct gcov_node { struct list_head children; struct list_head all; struct gcov_node *parent; - struct gcov_info *info; - struct gcov_info *ghost; + struct gcov_info **loaded_info; + struct gcov_info *unloaded_info; struct dentry *dentry; struct dentry **links; + int num_loaded; char name[0]; }; @@ -136,16 +138,37 @@ static const struct seq_operations gcov_seq_ops = { }; /* - * Return the profiling data set for a given node. This can either be the - * original profiling data structure or a duplicate (also called "ghost") - * in case the associated object file has been unloaded. + * Return a profiling data set associated with the given node. This is + * either a data set for a loaded object file or a data set copy in case + * all associated object files have been unloaded. */ static struct gcov_info *get_node_info(struct gcov_node *node) { - if (node->info) - return node->info; + if (node->num_loaded > 0) + return node->loaded_info[0]; - return node->ghost; + return node->unloaded_info; +} + +/* + * Return a newly allocated profiling data set which contains the sum of + * all profiling data associated with the given node. + */ +static struct gcov_info *get_accumulated_info(struct gcov_node *node) +{ + struct gcov_info *info; + int i = 0; + + if (node->unloaded_info) + info = gcov_info_dup(node->unloaded_info); + else + info = gcov_info_dup(node->loaded_info[i++]); + if (!info) + return NULL; + for (; i < node->num_loaded; i++) + gcov_info_add(info, node->loaded_info[i]); + + return info; } /* @@ -163,9 +186,10 @@ static int gcov_seq_open(struct inode *inode, struct file *file) mutex_lock(&node_lock); /* * Read from a profiling data copy to minimize reference tracking - * complexity and concurrent access. + * complexity and concurrent access and to keep accumulating multiple + * profiling data sets associated with one node simple. */ - info = gcov_info_dup(get_node_info(node)); + info = get_accumulated_info(node); if (!info) goto out_unlock; iter = gcov_iter_new(info); @@ -225,12 +249,25 @@ static struct gcov_node *get_node_by_name(const char *name) return NULL; } +/* + * Reset all profiling data associated with the specified node. + */ +static void reset_node(struct gcov_node *node) +{ + int i; + + if (node->unloaded_info) + gcov_info_reset(node->unloaded_info); + for (i = 0; i < node->num_loaded; i++) + gcov_info_reset(node->loaded_info[i]); +} + static void remove_node(struct gcov_node *node); /* * write() implementation for gcov data files. Reset profiling data for the - * associated file. If the object file has been unloaded (i.e. this is - * a "ghost" node), remove the debug fs node as well. + * corresponding file. If all associated object files have been unloaded, + * remove the debug fs node as well. */ static ssize_t gcov_seq_write(struct file *file, const char __user *addr, size_t len, loff_t *pos) @@ -245,10 +282,10 @@ static ssize_t gcov_seq_write(struct file *file, const char __user *addr, node = get_node_by_name(info->filename); if (node) { /* Reset counts or remove node for unloaded modules. */ - if (node->ghost) + if (node->num_loaded == 0) remove_node(node); else - gcov_info_reset(node->info); + reset_node(node); } /* Reset counts for open file. */ gcov_info_reset(info); @@ -378,7 +415,10 @@ static void init_node(struct gcov_node *node, struct gcov_info *info, INIT_LIST_HEAD(&node->list); INIT_LIST_HEAD(&node->children); INIT_LIST_HEAD(&node->all); - node->info = info; + if (node->loaded_info) { + node->loaded_info[0] = info; + node->num_loaded = 1; + } node->parent = parent; if (name) strcpy(node->name, name); @@ -394,9 +434,13 @@ static struct gcov_node *new_node(struct gcov_node *parent, struct gcov_node *node; node = kzalloc(sizeof(struct gcov_node) + strlen(name) + 1, GFP_KERNEL); - if (!node) { - pr_warning("out of memory\n"); - return NULL; + if (!node) + goto err_nomem; + if (info) { + node->loaded_info = kcalloc(1, sizeof(struct gcov_info *), + GFP_KERNEL); + if (!node->loaded_info) + goto err_nomem; } init_node(node, info, name, parent); /* Differentiate between gcov data file nodes and directory nodes. */ @@ -416,6 +460,11 @@ static struct gcov_node *new_node(struct gcov_node *parent, list_add(&node->all, &all_head); return node; + +err_nomem: + kfree(node); + pr_warning("out of memory\n"); + return NULL; } /* Remove symbolic links associated with node. */ @@ -441,8 +490,9 @@ static void release_node(struct gcov_node *node) list_del(&node->all); debugfs_remove(node->dentry); remove_links(node); - if (node->ghost) - gcov_info_free(node->ghost); + kfree(node->loaded_info); + if (node->unloaded_info) + gcov_info_free(node->unloaded_info); kfree(node); } @@ -477,7 +527,7 @@ static struct gcov_node *get_child_by_name(struct gcov_node *parent, /* * write() implementation for reset file. Reset all profiling data to zero - * and remove ghost nodes. + * and remove nodes for which all associated object files are unloaded. */ static ssize_t reset_write(struct file *file, const char __user *addr, size_t len, loff_t *pos) @@ -487,8 +537,8 @@ static ssize_t reset_write(struct file *file, const char __user *addr, mutex_lock(&node_lock); restart: list_for_each_entry(node, &all_head, all) { - if (node->info) - gcov_info_reset(node->info); + if (node->num_loaded > 0) + reset_node(node); else if (list_empty(&node->children)) { remove_node(node); /* Several nodes may have gone - restart loop. */ @@ -564,37 +614,115 @@ err_remove: } /* - * The profiling data set associated with this node is being unloaded. Store a - * copy of the profiling data and turn this node into a "ghost". + * Associate a profiling data set with an existing node. Needs to be called + * with node_lock held. */ -static int ghost_node(struct gcov_node *node) +static void add_info(struct gcov_node *node, struct gcov_info *info) { - node->ghost = gcov_info_dup(node->info); - if (!node->ghost) { - pr_warning("could not save data for '%s' (out of memory)\n", - node->info->filename); - return -ENOMEM; + struct gcov_info **loaded_info; + int num = node->num_loaded; + + /* + * Prepare new array. This is done first to simplify cleanup in + * case the new data set is incompatible, the node only contains + * unloaded data sets and there's not enough memory for the array. + */ + loaded_info = kcalloc(num + 1, sizeof(struct gcov_info *), GFP_KERNEL); + if (!loaded_info) { + pr_warning("could not add '%s' (out of memory)\n", + info->filename); + return; + } + memcpy(loaded_info, node->loaded_info, + num * sizeof(struct gcov_info *)); + loaded_info[num] = info; + /* Check if the new data set is compatible. */ + if (num == 0) { + /* + * A module was unloaded, modified and reloaded. The new + * data set replaces the copy of the last one. + */ + if (!gcov_info_is_compatible(node->unloaded_info, info)) { + pr_warning("discarding saved data for %s " + "(incompatible version)\n", info->filename); + gcov_info_free(node->unloaded_info); + node->unloaded_info = NULL; + } + } else { + /* + * Two different versions of the same object file are loaded. + * The initial one takes precedence. + */ + if (!gcov_info_is_compatible(node->loaded_info[0], info)) { + pr_warning("could not add '%s' (incompatible " + "version)\n", info->filename); + kfree(loaded_info); + return; + } } - node->info = NULL; + /* Overwrite previous array. */ + kfree(node->loaded_info); + node->loaded_info = loaded_info; + node->num_loaded = num + 1; +} - return 0; +/* + * Return the index of a profiling data set associated with a node. + */ +static int get_info_index(struct gcov_node *node, struct gcov_info *info) +{ + int i; + + for (i = 0; i < node->num_loaded; i++) { + if (node->loaded_info[i] == info) + return i; + } + return -ENOENT; } /* - * Profiling data for this node has been loaded again. Add profiling data - * from previous instantiation and turn this node into a regular node. + * Save the data of a profiling data set which is being unloaded. */ -static void revive_node(struct gcov_node *node, struct gcov_info *info) +static void save_info(struct gcov_node *node, struct gcov_info *info) { - if (gcov_info_is_compatible(node->ghost, info)) - gcov_info_add(info, node->ghost); + if (node->unloaded_info) + gcov_info_add(node->unloaded_info, info); else { - pr_warning("discarding saved data for '%s' (version changed)\n", + node->unloaded_info = gcov_info_dup(info); + if (!node->unloaded_info) { + pr_warning("could not save data for '%s' " + "(out of memory)\n", info->filename); + } + } +} + +/* + * Disassociate a profiling data set from a node. Needs to be called with + * node_lock held. + */ +static void remove_info(struct gcov_node *node, struct gcov_info *info) +{ + int i; + + i = get_info_index(node, info); + if (i < 0) { + pr_warning("could not remove '%s' (not found)\n", info->filename); + return; } - gcov_info_free(node->ghost); - node->ghost = NULL; - node->info = info; + if (gcov_persist) + save_info(node, info); + /* Shrink array. */ + node->loaded_info[i] = node->loaded_info[node->num_loaded - 1]; + node->num_loaded--; + if (node->num_loaded > 0) + return; + /* Last loaded data set was removed. */ + kfree(node->loaded_info); + node->loaded_info = NULL; + node->num_loaded = 0; + if (!node->unloaded_info) + remove_node(node); } /* @@ -609,30 +737,18 @@ void gcov_event(enum gcov_action action, struct gcov_info *info) node = get_node_by_name(info->filename); switch (action) { case GCOV_ADD: - /* Add new node or revive ghost. */ - if (!node) { + if (node) + add_info(node, info); + else add_node(info); - break; - } - if (gcov_persist) - revive_node(node, info); - else { - pr_warning("could not add '%s' (already exists)\n", - info->filename); - } break; case GCOV_REMOVE: - /* Remove node or turn into ghost. */ - if (!node) { + if (node) + remove_info(node, info); + else { pr_warning("could not remove '%s' (not found)\n", info->filename); - break; } - if (gcov_persist) { - if (!ghost_node(node)) - break; - } - remove_node(node); break; } mutex_unlock(&node_lock); -- cgit v0.10.2 From ed430fec756ad65f7cfba24f8ad17c3d5a403290 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 9 Sep 2010 16:37:36 -0700 Subject: proc: export uncached bit properly in /proc/kpageflags Fix the left-over old ifdef for PG_uncached in /proc/kpageflags. Now it's used by x86, too. Signed-off-by: Takashi Iwai Cc: Wu Fengguang Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/proc/page.c b/fs/proc/page.c index 180cf5a..3b8b456 100644 --- a/fs/proc/page.c +++ b/fs/proc/page.c @@ -146,7 +146,7 @@ u64 stable_page_flags(struct page *page) u |= kpf_copy_bit(k, KPF_HWPOISON, PG_hwpoison); #endif -#ifdef CONFIG_IA64_UNCACHED_ALLOCATOR +#ifdef CONFIG_ARCH_USES_PG_UNCACHED u |= kpf_copy_bit(k, KPF_UNCACHED, PG_uncached); #endif -- cgit v0.10.2 From 31583bb0cf6cc40f2a468a4d2f3b9cbefd24f891 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 9 Sep 2010 16:37:37 -0700 Subject: cgroups: fix API thinko Add cgroup_attach_task_all() The existing cgroup_attach_task_current_cg() API is called by a thread to attach another thread to all of its cgroups; this is unsuitable for cases where a privileged task wants to attach itself to the cgroups of a less privileged one, since the call must be made from the context of the target task. This patch adds a more generic cgroup_attach_task_all() API that allows both the source task and to-be-moved task to be specified. cgroup_attach_task_current_cg() becomes a specialization of the more generic new function. [menage@google.com: rewrote changelog] [akpm@linux-foundation.org: address reviewer comments] Signed-off-by: Michael S. Tsirkin Tested-by: Alex Williamson Acked-by: Paul Menage Cc: Li Zefan Cc: Ben Blum Cc: Sridhar Samudrala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index ed3e92e..0c99102 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -578,7 +578,12 @@ struct task_struct *cgroup_iter_next(struct cgroup *cgrp, void cgroup_iter_end(struct cgroup *cgrp, struct cgroup_iter *it); int cgroup_scan_tasks(struct cgroup_scanner *scan); int cgroup_attach_task(struct cgroup *, struct task_struct *); -int cgroup_attach_task_current_cg(struct task_struct *); +int cgroup_attach_task_all(struct task_struct *from, struct task_struct *); + +static inline int cgroup_attach_task_current_cg(struct task_struct *tsk) +{ + return cgroup_attach_task_all(current, tsk); +} /* * CSS ID is ID for cgroup_subsys_state structs under subsys. This only works @@ -636,6 +641,11 @@ static inline int cgroupstats_build(struct cgroupstats *stats, } /* No cgroups - nothing to do */ +static inline int cgroup_attach_task_all(struct task_struct *from, + struct task_struct *t) +{ + return 0; +} static inline int cgroup_attach_task_current_cg(struct task_struct *t) { return 0; diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 192f88c..c9483d8 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -1791,19 +1791,20 @@ out: } /** - * cgroup_attach_task_current_cg - attach task 'tsk' to current task's cgroup + * cgroup_attach_task_all - attach task 'tsk' to all cgroups of task 'from' + * @from: attach to all cgroups of a given task * @tsk: the task to be attached */ -int cgroup_attach_task_current_cg(struct task_struct *tsk) +int cgroup_attach_task_all(struct task_struct *from, struct task_struct *tsk) { struct cgroupfs_root *root; - struct cgroup *cur_cg; int retval = 0; cgroup_lock(); for_each_active_root(root) { - cur_cg = task_cgroup_from_root(current, root); - retval = cgroup_attach_task(cur_cg, tsk); + struct cgroup *from_cg = task_cgroup_from_root(from, root); + + retval = cgroup_attach_task(from_cg, tsk); if (retval) break; } @@ -1811,7 +1812,7 @@ int cgroup_attach_task_current_cg(struct task_struct *tsk) return retval; } -EXPORT_SYMBOL_GPL(cgroup_attach_task_current_cg); +EXPORT_SYMBOL_GPL(cgroup_attach_task_all); /* * Attach task with pid 'pid' to cgroup 'cgrp'. Call with cgroup_mutex -- cgit v0.10.2 From b78d6c5f51935ba89df8db33a57bacb547aa7325 Mon Sep 17 00:00:00 2001 From: Yusuke Goda Date: Thu, 9 Sep 2010 16:37:39 -0700 Subject: tmio_mmc: don't clear unhandled pending interrupts Previously, it was possible for ack_mmc_irqs() to clear pending interrupt bits in the CTL_STATUS register, even though the interrupt handler had not been called. This was because of a race that existed when doing a read-modify-write sequence on CTL_STATUS. After the read step in this sequence, if an interrupt occurred (causing one of the bits in CTL_STATUS to be set) the write step would inadvertently clear it. Observed with the TMIO_STAT_RXRDY bit together with CMD53 on AR6002 and BCM4318 SDIO cards in polled mode. This patch eliminates this race by only writing to CTL_STATUS and clearing the interrupts that were passed as an argument to ack_mmc_irqs()." [matt@console-pimps.org: rewrote changelog] Signed-off-by: Yusuke Goda Acked-by: Magnus Damm " Tested-by: Arnd Hannemann " Acked-by: Ian Molton Cc: Matt Fleming Cc: Samuel Ortiz Cc: Paul Mundt Cc: Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h index 64f7d5d..7944604 100644 --- a/drivers/mmc/host/tmio_mmc.h +++ b/drivers/mmc/host/tmio_mmc.h @@ -82,10 +82,7 @@ #define ack_mmc_irqs(host, i) \ do { \ - u32 mask;\ - mask = sd_ctrl_read32((host), CTL_STATUS); \ - mask &= ~((i) & TMIO_MASK_IRQ); \ - sd_ctrl_write32((host), CTL_STATUS, mask); \ + sd_ctrl_write32((host), CTL_STATUS, ~(i)); \ } while (0) -- cgit v0.10.2 From 5600efb1bc2745d93ae0bc08130117a84f2b9d69 Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Thu, 9 Sep 2010 16:37:43 -0700 Subject: mmc: fix the use of kunmap_atomic() in tmio_mmc.h kunmap_atomic() takes the cookie, returned by the kmap_atomic() as its argument and not the page address, used as an argument to kmap_atomic(). This patch fixes the compile error: In file included from drivers/mmc/host/tmio_mmc.c:37: drivers/mmc/host/tmio_mmc.h: In function 'tmio_mmc_kunmap_atomic': drivers/mmc/host/tmio_mmc.h:192: error: negative width in bit-field '' Signed-off-by: Guennadi Liakhovetski Acked-by: Eric Miao Tested-by: Magnus Damm Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/tmio_mmc.c b/drivers/mmc/host/tmio_mmc.c index ee7d0a5..69d98e3 100644 --- a/drivers/mmc/host/tmio_mmc.c +++ b/drivers/mmc/host/tmio_mmc.c @@ -164,6 +164,7 @@ tmio_mmc_start_command(struct tmio_mmc_host *host, struct mmc_command *cmd) static void tmio_mmc_pio_irq(struct tmio_mmc_host *host) { struct mmc_data *data = host->data; + void *sg_virt; unsigned short *buf; unsigned int count; unsigned long flags; @@ -173,8 +174,8 @@ static void tmio_mmc_pio_irq(struct tmio_mmc_host *host) return; } - buf = (unsigned short *)(tmio_mmc_kmap_atomic(host, &flags) + - host->sg_off); + sg_virt = tmio_mmc_kmap_atomic(host->sg_ptr, &flags); + buf = (unsigned short *)(sg_virt + host->sg_off); count = host->sg_ptr->length - host->sg_off; if (count > data->blksz) @@ -191,7 +192,7 @@ static void tmio_mmc_pio_irq(struct tmio_mmc_host *host) host->sg_off += count; - tmio_mmc_kunmap_atomic(host, &flags); + tmio_mmc_kunmap_atomic(sg_virt, &flags); if (host->sg_off == host->sg_ptr->length) tmio_mmc_next_sg(host); diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h index 7944604..0fedc78 100644 --- a/drivers/mmc/host/tmio_mmc.h +++ b/drivers/mmc/host/tmio_mmc.h @@ -174,19 +174,17 @@ static inline int tmio_mmc_next_sg(struct tmio_mmc_host *host) return --host->sg_len; } -static inline char *tmio_mmc_kmap_atomic(struct tmio_mmc_host *host, +static inline char *tmio_mmc_kmap_atomic(struct scatterlist *sg, unsigned long *flags) { - struct scatterlist *sg = host->sg_ptr; - local_irq_save(*flags); return kmap_atomic(sg_page(sg), KM_BIO_SRC_IRQ) + sg->offset; } -static inline void tmio_mmc_kunmap_atomic(struct tmio_mmc_host *host, +static inline void tmio_mmc_kunmap_atomic(void *virt, unsigned long *flags) { - kunmap_atomic(sg_page(host->sg_ptr), KM_BIO_SRC_IRQ); + kunmap_atomic(virt, KM_BIO_SRC_IRQ); local_irq_restore(*flags); } -- cgit v0.10.2 From 60c2c0d5658082468b569d039f4d0dc24f92c66b Mon Sep 17 00:00:00 2001 From: Jiri Pinkava Date: Tue, 25 May 2010 09:48:58 +0200 Subject: ARM: SAMSUNG: MMC: fix build error when both DMA and PIO mode selected [cjb: fix line-wrapped patch] Signed-off-by: Jiri Pinkava Signed-off-by: Chris Ball Cc: Matt Fleming Cc: Russell King Cc: Ben Dooks Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/s3cmci.c b/drivers/mmc/host/s3cmci.c index 2e16e0a..976330d 100644 --- a/drivers/mmc/host/s3cmci.c +++ b/drivers/mmc/host/s3cmci.c @@ -1600,7 +1600,7 @@ static int __devinit s3cmci_probe(struct platform_device *pdev) host->pio_active = XFER_NONE; #ifdef CONFIG_MMC_S3C_PIODMA - host->dodma = host->pdata->dma; + host->dodma = host->pdata->use_dma; #endif host->mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); -- cgit v0.10.2 From 16d9b130783c54c30cab80e24810ab1ab9596e11 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Thu, 9 Sep 2010 16:37:46 -0700 Subject: omap_hsmmc: remove unused local `state' This fixes the following warning: drivers/mmc/host/omap_hsmmc.c: In function 'omap_hsmmc_suspend': drivers/mmc/host/omap_hsmmc.c:2275: warning: unused variable 'state' Introduced by commit ID: commit 1a13f8fa76c880be41d6b1e6a2b44404bcbfdf9e Author: Matt Fleming Date: Wed May 26 14:42:08 2010 -0700 mmc: remove the "state" argument to mmc_suspend_host() The unique usage of this var was removed there, and missed removing the respective declaration aswell. Signed-off-by: Sergio Aguirre Signed-off-by: Chris Ball Acked-by: Matt Fleming Cc: Madhusudhan Chikkature Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c index 4a8776f..fb6446a 100644 --- a/drivers/mmc/host/omap_hsmmc.c +++ b/drivers/mmc/host/omap_hsmmc.c @@ -2305,7 +2305,6 @@ static int omap_hsmmc_suspend(struct device *dev) int ret = 0; struct platform_device *pdev = to_platform_device(dev); struct omap_hsmmc_host *host = platform_get_drvdata(pdev); - pm_message_t state = PMSG_SUSPEND; /* unused by MMC core */ if (host && host->suspended) return 0; -- cgit v0.10.2 From 23ef309a6e070490da0a37b9b6383819f8170ea3 Mon Sep 17 00:00:00 2001 From: Marc Kleine-Budde Date: Thu, 9 Sep 2010 16:37:48 -0700 Subject: mmc: at91_mci: add missing linux/highmem.h include Fix the following error: at91_mci.c: In function 'at91_mci_sg_to_dma': at91_mci.c:236: error: implicit declaration of function 'kmap_atomic' at91_mci.c:236: error: 'KM_BIO_SRC_IRQ' undeclared (first use in this function) at91_mci.c:236: error: (Each undeclared identifier is reported only once at91_mci.c:236: error: for each function it appears in.) at91_mci.c:236: warning: assignment makes pointer from integer without a cast at91_mci.c:252: error: implicit declaration of function 'kunmap_atomic' at91_mci.c: In function 'at91_mci_post_dma_read': at91_mci.c:302: error: 'KM_BIO_SRC_IRQ' undeclared (first use in this function) at91_mci.c:302: warning: assignment makes pointer from integer without a cast at91_mci.c:317: error: implicit declaration of function 'flush_kernel_dcache_page' Signed-off-by: Marc Kleine-Budde Signed-off-by: Chris Ball Cc: Nicolas Ferre Cc: Andrew Victor Cc: Wolfgang Muees Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/at91_mci.c b/drivers/mmc/host/at91_mci.c index 5f3a599..87226cd 100644 --- a/drivers/mmc/host/at91_mci.c +++ b/drivers/mmc/host/at91_mci.c @@ -66,6 +66,7 @@ #include #include #include +#include #include -- cgit v0.10.2 From e7cb756fc3c7c32040283963572258381b342dff Mon Sep 17 00:00:00 2001 From: Ethan Du Date: Thu, 9 Sep 2010 16:37:49 -0700 Subject: omap hsmmc: fix a racing case between kmmcd and omap_hsmmc_suspend If suspend called when kmmcd is doing host->ops->disable, as kmmcd already increased host->en_dis_recurs to 1, the mmc_host_enable in suspend function will return directly without increase the nesting_cnt, which will cause the followed register access carried out to the disabled host. mmc_suspend_host will enable host itself. No need to enable host before it. Also works on kmmcd will get flushed in mmc_suspend_host, enable host after it will be safe. So make the mmc_host_enable after it. [cjb: rebase against current Linus] Signed-off-by: Ethan Signed-off-by: Chris Ball Acked-by: Adrian Hunter Acked-by: Madhusudhan Chikkature Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c index fb6446a..4526d27 100644 --- a/drivers/mmc/host/omap_hsmmc.c +++ b/drivers/mmc/host/omap_hsmmc.c @@ -2323,8 +2323,8 @@ static int omap_hsmmc_suspend(struct device *dev) } } cancel_work_sync(&host->mmc_carddetect_work); - mmc_host_enable(host->mmc); ret = mmc_suspend_host(host->mmc); + mmc_host_enable(host->mmc); if (ret == 0) { omap_hsmmc_disable_irq(host); OMAP_HSMMC_WRITE(host->base, HCTL, -- cgit v0.10.2 From 7c5367f205f7d53659fb19b9fdf65b7bc1a592c6 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Thu, 9 Sep 2010 16:37:50 -0700 Subject: drivers/mmc/host/imxmmc.c: adjust confusing if indentation Move the second if (reg & ...) test into the branch indicated by its indentation. The test was previously always executed after the if containing that branch, but it was always false unless the if branch was taken. The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @r disable braces4@ position p1,p2; statement S1,S2; @@ ( if (...) { ... } | if (...) S1@p1 S2@p2 ) @script:python@ p1 << r.p1; p2 << r.p2; @@ if (p1[0].column == p2[0].column): cocci.print_main("branch",p1) cocci.print_secs("after",p2) // Signed-off-by: Julia Lawall Signed-off-by: Chris Ball Cc: Pavel Pisa Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/imxmmc.c b/drivers/mmc/host/imxmmc.c index 9a68ff4..5a950b1 100644 --- a/drivers/mmc/host/imxmmc.c +++ b/drivers/mmc/host/imxmmc.c @@ -148,11 +148,12 @@ static int imxmci_start_clock(struct imxmci_host *host) while (delay--) { reg = readw(host->base + MMC_REG_STATUS); - if (reg & STATUS_CARD_BUS_CLK_RUN) + if (reg & STATUS_CARD_BUS_CLK_RUN) { /* Check twice before cut */ reg = readw(host->base + MMC_REG_STATUS); if (reg & STATUS_CARD_BUS_CLK_RUN) return 0; + } if (test_bit(IMXMCI_PEND_STARTED_b, &host->pending_events)) return 0; -- cgit v0.10.2 From 4969c1192d15afa3389e7ae3302096ff684ba655 Mon Sep 17 00:00:00 2001 From: Andrea Arcangeli Date: Thu, 9 Sep 2010 16:37:52 -0700 Subject: mm: fix swapin race condition The pte_same check is reliable only if the swap entry remains pinned (by the page lock on swapcache). We've also to ensure the swapcache isn't removed before we take the lock as try_to_free_swap won't care about the page pin. One of the possible impacts of this patch is that a KSM-shared page can point to the anon_vma of another process, which could exit before the page is freed. This can leave a page with a pointer to a recycled anon_vma object, or worse, a pointer to something that is no longer an anon_vma. [riel@redhat.com: changelog help] Signed-off-by: Andrea Arcangeli Acked-by: Hugh Dickins Reviewed-by: Rik van Riel Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/ksm.h b/include/linux/ksm.h index 74d691e..3319a69 100644 --- a/include/linux/ksm.h +++ b/include/linux/ksm.h @@ -16,6 +16,9 @@ struct stable_node; struct mem_cgroup; +struct page *ksm_does_need_to_copy(struct page *page, + struct vm_area_struct *vma, unsigned long address); + #ifdef CONFIG_KSM int ksm_madvise(struct vm_area_struct *vma, unsigned long start, unsigned long end, int advice, unsigned long *vm_flags); @@ -70,19 +73,14 @@ static inline void set_page_stable_node(struct page *page, * We'd like to make this conditional on vma->vm_flags & VM_MERGEABLE, * but what if the vma was unmerged while the page was swapped out? */ -struct page *ksm_does_need_to_copy(struct page *page, - struct vm_area_struct *vma, unsigned long address); -static inline struct page *ksm_might_need_to_copy(struct page *page, +static inline int ksm_might_need_to_copy(struct page *page, struct vm_area_struct *vma, unsigned long address) { struct anon_vma *anon_vma = page_anon_vma(page); - if (!anon_vma || - (anon_vma->root == vma->anon_vma->root && - page->index == linear_page_index(vma, address))) - return page; - - return ksm_does_need_to_copy(page, vma, address); + return anon_vma && + (anon_vma->root != vma->anon_vma->root || + page->index != linear_page_index(vma, address)); } int page_referenced_ksm(struct page *page, @@ -115,10 +113,10 @@ static inline int ksm_madvise(struct vm_area_struct *vma, unsigned long start, return 0; } -static inline struct page *ksm_might_need_to_copy(struct page *page, +static inline int ksm_might_need_to_copy(struct page *page, struct vm_area_struct *vma, unsigned long address) { - return page; + return 0; } static inline int page_referenced_ksm(struct page *page, diff --git a/mm/ksm.c b/mm/ksm.c index e2ae004..b1873cf 100644 --- a/mm/ksm.c +++ b/mm/ksm.c @@ -1504,8 +1504,6 @@ struct page *ksm_does_need_to_copy(struct page *page, { struct page *new_page; - unlock_page(page); /* any racers will COW it, not modify it */ - new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); if (new_page) { copy_user_highpage(new_page, page, address, vma); @@ -1521,7 +1519,6 @@ struct page *ksm_does_need_to_copy(struct page *page, add_page_to_unevictable_list(new_page); } - page_cache_release(page); return new_page; } diff --git a/mm/memory.c b/mm/memory.c index 6b2ab10..71b161b 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -2623,7 +2623,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned int flags, pte_t orig_pte) { spinlock_t *ptl; - struct page *page; + struct page *page, *swapcache = NULL; swp_entry_t entry; pte_t pte; struct mem_cgroup *ptr = NULL; @@ -2679,10 +2679,23 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, lock_page(page); delayacct_clear_flag(DELAYACCT_PF_SWAPIN); - page = ksm_might_need_to_copy(page, vma, address); - if (!page) { - ret = VM_FAULT_OOM; - goto out; + /* + * Make sure try_to_free_swap didn't release the swapcache + * from under us. The page pin isn't enough to prevent that. + */ + if (unlikely(!PageSwapCache(page))) + goto out_page; + + if (ksm_might_need_to_copy(page, vma, address)) { + swapcache = page; + page = ksm_does_need_to_copy(page, vma, address); + + if (unlikely(!page)) { + ret = VM_FAULT_OOM; + page = swapcache; + swapcache = NULL; + goto out_page; + } } if (mem_cgroup_try_charge_swapin(mm, page, GFP_KERNEL, &ptr)) { @@ -2735,6 +2748,18 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, if (vm_swap_full() || (vma->vm_flags & VM_LOCKED) || PageMlocked(page)) try_to_free_swap(page); unlock_page(page); + if (swapcache) { + /* + * Hold the lock to avoid the swap entry to be reused + * until we take the PT lock for the pte_same() check + * (to avoid false positives from pte_same). For + * further safety release the lock after the swap_free + * so that the swap count won't change under a + * parallel locked swapcache. + */ + unlock_page(swapcache); + page_cache_release(swapcache); + } if (flags & FAULT_FLAG_WRITE) { ret |= do_wp_page(mm, vma, address, page_table, pmd, ptl, pte); @@ -2756,6 +2781,10 @@ out_page: unlock_page(page); out_release: page_cache_release(page); + if (swapcache) { + unlock_page(swapcache); + page_cache_release(swapcache); + } return ret; } -- cgit v0.10.2 From 152e0659fc001029c70fa4373af1792b1ae0d01c Mon Sep 17 00:00:00 2001 From: Andrea Arcangeli Date: Thu, 9 Sep 2010 16:37:53 -0700 Subject: mm: avoid warning when COMPACTION is selected COMPACTION enables MIGRATION, but MIGRATION spawns a warning if numa or memhotplug aren't selected. However MIGRATION doesn't depend on them. I guess it's just trying to be strict doing a double check on who's enabling it, but it doesn't know that compaction also enables MIGRATION. Signed-off-by: Andrea Arcangeli Acked-by: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/Kconfig b/mm/Kconfig index f4e516e..f0fb912 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -189,7 +189,7 @@ config COMPACTION config MIGRATION bool "Page migration" def_bool y - depends on NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE + depends on NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION help Allows the migration of the physical location of pages of processes while the virtual addresses are not changed. This is useful in -- cgit v0.10.2 From ee3aebdd8f5f8eac41c25c80ceee3d728f920f3b Mon Sep 17 00:00:00 2001 From: Jan Sembera Date: Thu, 9 Sep 2010 16:37:54 -0700 Subject: binfmt_misc: fix binfmt_misc priority Commit 74641f584da ("alpha: binfmt_aout fix") (May 2009) introduced a regression - binfmt_misc is now consulted after binfmt_elf, which will unfortunately break ia32el. ia32 ELF binaries on ia64 used to be matched using binfmt_misc and executed using wrapper. As 32bit binaries are now matched by binfmt_elf before bindmt_misc kicks in, the wrapper is ignored. The fix increases precedence of binfmt_misc to the original state. Signed-off-by: Jan Sembera Cc: Ivan Kokshaysky Cc: Al Viro Cc: Richard Henderson [2.6.everything.x] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index a7528b9..fd0cc0b 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -724,7 +724,7 @@ static int __init init_misc_binfmt(void) { int err = register_filesystem(&bm_fs_type); if (!err) { - err = register_binfmt(&misc_format); + err = insert_binfmt(&misc_format); if (err) unregister_filesystem(&bm_fs_type); } -- cgit v0.10.2 From 408929bed7841686ce5fdd06366fb652cb653d6c Mon Sep 17 00:00:00 2001 From: Atsushi Nemoto Date: Thu, 9 Sep 2010 16:37:56 -0700 Subject: rtc: m41t80: do not use rtc_valid_tm in m41t80_rtc_read_alarm Commit b485fe5ea ("rtc/m41t80: use rtc_valid_tm() to check returned tm") added rtc_valid_tm to m41t80_rtc_read_alarm() but it was wrong while the t->time does not contain complete date/time. This patch also fixes a warning: warning: passing argument 1 of 'rtc_valid_tm' from incompatible pointer type Signed-off-by: Atsushi Nemoto Cc: Wan ZongShun Cc: Alessandro Zummo Cc: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-m41t80.c b/drivers/rtc/rtc-m41t80.c index 66377f3..d60557c 100644 --- a/drivers/rtc/rtc-m41t80.c +++ b/drivers/rtc/rtc-m41t80.c @@ -364,7 +364,7 @@ static int m41t80_rtc_read_alarm(struct device *dev, struct rtc_wkalrm *t) t->time.tm_isdst = -1; t->enabled = !!(reg[M41T80_REG_ALARM_MON] & M41T80_ALMON_AFE); t->pending = !!(reg[M41T80_REG_FLAGS] & M41T80_FLAGS_AF); - return rtc_valid_tm(t); + return 0; } static struct rtc_class_ops m41t80_rtc_ops = { -- cgit v0.10.2 From 94131e174fedd9f3f9bb148cee4be12f2d46d68e Mon Sep 17 00:00:00 2001 From: "Ira W. Snyder" Date: Thu, 9 Sep 2010 16:37:58 -0700 Subject: arch/powerpc/include/asm/fsldma.h needs slab.h The slab.h header is required to use the kmalloc() family of functions. Due to recent kernel changes, this header must be directly included by code that calls into the memory allocator. Without this patch, any code which includes this header fails to build. Signed-off-by: Ira W. Snyder Cc: Dan Williams Cc: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/powerpc/include/asm/fsldma.h b/arch/powerpc/include/asm/fsldma.h index a67aeed..debc5ed 100644 --- a/arch/powerpc/include/asm/fsldma.h +++ b/arch/powerpc/include/asm/fsldma.h @@ -11,6 +11,7 @@ #ifndef __ARCH_POWERPC_ASM_FSLDMA_H__ #define __ARCH_POWERPC_ASM_FSLDMA_H__ +#include #include /* -- cgit v0.10.2 From 1c24de60e50fb19b94d94225458da17c720f0729 Mon Sep 17 00:00:00 2001 From: Jerome Marchand Date: Thu, 9 Sep 2010 16:37:59 -0700 Subject: kernel/groups.c: fix integer overflow in groups_search gid_t is a unsigned int. If group_info contains a gid greater than MAX_INT, groups_search() function may look on the wrong side of the search tree. This solves some unfair "permission denied" problems. Signed-off-by: Jerome Marchand Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/groups.c b/kernel/groups.c index 53b1916..253dc0f 100644 --- a/kernel/groups.c +++ b/kernel/groups.c @@ -143,10 +143,9 @@ int groups_search(const struct group_info *group_info, gid_t grp) right = group_info->ngroups; while (left < right) { unsigned int mid = (left+right)/2; - int cmp = grp - GROUP_AT(group_info, mid); - if (cmp > 0) + if (grp > GROUP_AT(group_info, mid)) left = mid + 1; - else if (cmp < 0) + else if (grp < GROUP_AT(group_info, mid)) right = mid; else return 1; -- cgit v0.10.2 From bc6930457460788e14b2c0808ed4632a1592bd61 Mon Sep 17 00:00:00 2001 From: Minchan Kim Date: Thu, 9 Sep 2010 16:38:00 -0700 Subject: mm: compaction: handle active and inactive fairly in too_many_isolated Iram reported that compaction's too_many_isolated() loops forever. (http://www.spinics.net/lists/linux-mm/msg08123.html) The meminfo when the situation happened was inactive anon is zero. That's because the system has no memory pressure until then. While all anon pages were in the active lru, compaction could select active lru as well as inactive lru. That's a different thing from vmscan's isolated. So we has been two too_many_isolated. While compaction can isolate pages in both active and inactive, current implementation of too_many_isolated only considers inactive. It made Iram's problem. This patch handles active and inactive fairly. That's because we can't expect where from and how many compaction would isolated pages. This patch changes (nr_isolated > nr_inactive) with nr_isolated > (nr_active + nr_inactive) / 2. Signed-off-by: Minchan Kim Reported-by: Iram Shahzad Acked-by: Mel Gorman Acked-by: Wu Fengguang Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/compaction.c b/mm/compaction.c index 94cce51..4d709ee 100644 --- a/mm/compaction.c +++ b/mm/compaction.c @@ -214,15 +214,16 @@ static void acct_isolated(struct zone *zone, struct compact_control *cc) /* Similar to reclaim, but different enough that they don't share logic */ static bool too_many_isolated(struct zone *zone) { - - unsigned long inactive, isolated; + unsigned long active, inactive, isolated; inactive = zone_page_state(zone, NR_INACTIVE_FILE) + zone_page_state(zone, NR_INACTIVE_ANON); + active = zone_page_state(zone, NR_ACTIVE_FILE) + + zone_page_state(zone, NR_ACTIVE_ANON); isolated = zone_page_state(zone, NR_ISOLATED_FILE) + zone_page_state(zone, NR_ISOLATED_ANON); - return isolated > inactive; + return isolated > (inactive + active) / 2; } /* -- cgit v0.10.2 From 0dcc48c15f63ee86c2fcd33968b08d651f0360a5 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Thu, 9 Sep 2010 16:38:01 -0700 Subject: memory hotplug: fix next block calculation in is_removable next_active_pageblock() is for finding next _used_ freeblock. It skips several blocks when it finds there are a chunk of free pages lager than pageblock. But it has 2 bugs. 1. We have no lock. page_order(page) - pageblock_order can be minus. 2. pageblocks_stride += is wrong. it should skip page_order(p) of pages. Signed-off-by: KAMEZAWA Hiroyuki Cc: Michal Hocko Cc: Wu Fengguang Cc: Mel Gorman Cc: KAMEZAWA Hiroyuki Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index a4cfcdc..dd186c1 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -584,19 +584,19 @@ static inline int pageblock_free(struct page *page) /* Return the start of the next active pageblock after a given page */ static struct page *next_active_pageblock(struct page *page) { - int pageblocks_stride; - /* Ensure the starting page is pageblock-aligned */ BUG_ON(page_to_pfn(page) & (pageblock_nr_pages - 1)); - /* Move forward by at least 1 * pageblock_nr_pages */ - pageblocks_stride = 1; - /* If the entire pageblock is free, move to the end of free page */ - if (pageblock_free(page)) - pageblocks_stride += page_order(page) - pageblock_order; + if (pageblock_free(page)) { + int order; + /* be careful. we don't have locks, page_order can be changed.*/ + order = page_order(page); + if ((order < MAX_ORDER) && (order >= pageblock_order)) + return page + (1 << order); + } - return page + (pageblocks_stride * pageblock_nr_pages); + return page + pageblock_nr_pages; } /* Checks if this range of memory is likely to be hot-removable. */ -- cgit v0.10.2 From 5affb607720d734ca572b8a77c5c7d62d3042b6f Mon Sep 17 00:00:00 2001 From: Gregory Bean Date: Thu, 9 Sep 2010 16:38:02 -0700 Subject: gpio: sx150x: correct and refine reset-on-probe behavior Replace the arbitrary software-reset call from the device-probe method, because: - It is defective. To work correctly, it should be two byte writes, not a single word write. As it stands, it does nothing. - Some devices with sx150x expanders installed have their NRESET pins ganged on the same line, so resetting one causes the others to reset - not a nice thing to do arbitrarily! - The probe, usually taking place at boot, implies a recent hard-reset, so a software reset at this point is just a waste of energy anyway. Therefore, make it optional, defaulting to off, as this will match the common case of probing at powerup and also matches the current broken no-op behavior. Signed-off-by: Gregory Bean Reviewed-by: Jean Delvare Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/gpio/sx150x.c b/drivers/gpio/sx150x.c index b42f42c..823559a 100644 --- a/drivers/gpio/sx150x.c +++ b/drivers/gpio/sx150x.c @@ -459,17 +459,33 @@ static int sx150x_init_io(struct sx150x_chip *chip, u8 base, u16 cfg) return err; } -static int sx150x_init_hw(struct sx150x_chip *chip, - struct sx150x_platform_data *pdata) +static int sx150x_reset(struct sx150x_chip *chip) { - int err = 0; + int err; - err = i2c_smbus_write_word_data(chip->client, + err = i2c_smbus_write_byte_data(chip->client, chip->dev_cfg->reg_reset, - 0x3412); + 0x12); if (err < 0) return err; + err = i2c_smbus_write_byte_data(chip->client, + chip->dev_cfg->reg_reset, + 0x34); + return err; +} + +static int sx150x_init_hw(struct sx150x_chip *chip, + struct sx150x_platform_data *pdata) +{ + int err = 0; + + if (pdata->reset_during_probe) { + err = sx150x_reset(chip); + if (err < 0) + return err; + } + err = sx150x_i2c_write(chip->client, chip->dev_cfg->reg_misc, 0x01); diff --git a/include/linux/i2c/sx150x.h b/include/linux/i2c/sx150x.h index ee3049c..52baa79 100644 --- a/include/linux/i2c/sx150x.h +++ b/include/linux/i2c/sx150x.h @@ -63,6 +63,9 @@ * IRQ lines will appear. Similarly to gpio_base, the expander * will create a block of irqs beginning at this number. * This value is ignored if irq_summary is < 0. + * @reset_during_probe: If set to true, the driver will trigger a full + * reset of the chip at the beginning of the probe + * in order to place it in a known state. */ struct sx150x_platform_data { unsigned gpio_base; @@ -73,6 +76,7 @@ struct sx150x_platform_data { u16 io_polarity; int irq_summary; unsigned irq_base; + bool reset_during_probe; }; #endif /* __LINUX_I2C_SX150X_H */ -- cgit v0.10.2 From c956126c137d97acb6f4d56fa9572d0bcc84e4ed Mon Sep 17 00:00:00 2001 From: David Brownell Date: Thu, 9 Sep 2010 16:38:03 -0700 Subject: gpio: doc updates There's been some recent confusion about error checking GPIO numbers. briefly, it should be handled mostly during setup, when gpio_request() is called, and NEVER by expectig gpio_is_valid to report more than never-usable GPIO numbers. [akpm@linux-foundation.org: terminate unterminated comment] Signed-off-by: David Brownell Cc: Eric Miao" Cc: "Ryan Mallon" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/gpio.txt b/Documentation/gpio.txt index d96a6db..9633da0 100644 --- a/Documentation/gpio.txt +++ b/Documentation/gpio.txt @@ -109,17 +109,19 @@ use numbers 2000-2063 to identify GPIOs in a bank of I2C GPIO expanders. If you want to initialize a structure with an invalid GPIO number, use some negative number (perhaps "-EINVAL"); that will never be valid. To -test if a number could reference a GPIO, you may use this predicate: +test if such number from such a structure could reference a GPIO, you +may use this predicate: int gpio_is_valid(int number); A number that's not valid will be rejected by calls which may request or free GPIOs (see below). Other numbers may also be rejected; for -example, a number might be valid but unused on a given board. - -Whether a platform supports multiple GPIO controllers is currently a -platform-specific implementation issue. +example, a number might be valid but temporarily unused on a given board. +Whether a platform supports multiple GPIO controllers is a platform-specific +implementation issue, as are whether that support can leave "holes" in the space +of GPIO numbers, and whether new controllers can be added at runtime. Such issues +can affect things including whether adjacent GPIO numbers are both valid. Using GPIOs ----------- @@ -480,12 +482,16 @@ To support this framework, a platform's Kconfig will "select" either ARCH_REQUIRE_GPIOLIB or ARCH_WANT_OPTIONAL_GPIOLIB and arrange that its includes and defines three functions: gpio_get_value(), gpio_set_value(), and gpio_cansleep(). -They may also want to provide a custom value for ARCH_NR_GPIOS. -ARCH_REQUIRE_GPIOLIB means that the gpio-lib code will always get compiled +It may also provide a custom value for ARCH_NR_GPIOS, so that it better +reflects the number of GPIOs in actual use on that platform, without +wasting static table space. (It should count both built-in/SoC GPIOs and +also ones on GPIO expanders. + +ARCH_REQUIRE_GPIOLIB means that the gpiolib code will always get compiled into the kernel on that architecture. -ARCH_WANT_OPTIONAL_GPIOLIB means the gpio-lib code defaults to off and the user +ARCH_WANT_OPTIONAL_GPIOLIB means the gpiolib code defaults to off and the user can enable it and build it into the kernel optionally. If neither of these options are selected, the platform does not support diff --git a/include/asm-generic/gpio.h b/include/asm-generic/gpio.h index c7376bf..8ca18e2 100644 --- a/include/asm-generic/gpio.h +++ b/include/asm-generic/gpio.h @@ -16,15 +16,27 @@ * While the GPIO programming interface defines valid GPIO numbers * to be in the range 0..MAX_INT, this library restricts them to the * smaller range 0..ARCH_NR_GPIOS-1. + * + * ARCH_NR_GPIOS is somewhat arbitrary; it usually reflects the sum of + * builtin/SoC GPIOs plus a number of GPIOs on expanders; the latter is + * actually an estimate of a board-specific value. */ #ifndef ARCH_NR_GPIOS #define ARCH_NR_GPIOS 256 #endif +/* + * "valid" GPIO numbers are nonnegative and may be passed to + * setup routines like gpio_request(). only some valid numbers + * can successfully be requested and used. + * + * Invalid GPIO numbers are useful for indicating no-such-GPIO in + * platform data and other tables. + */ + static inline int gpio_is_valid(int number) { - /* only some non-negative numbers are valid */ return ((unsigned)number) < ARCH_NR_GPIOS; } -- cgit v0.10.2 From b7bbbf354ed0edbbffdc70c9c17f1a5d4d78204c Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Thu, 9 Sep 2010 16:38:04 -0700 Subject: MAINTAINERS: correct entry for legacy RTC-driver Because no one dared to remove it so far, let's keep the entry correct, at least. Signed-off-by: Wolfram Sang Acked-by: Mike Frysinger Cc: Paul Gortmaker Cc: Alessandro Zummo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index a3da87f..9800de5 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4838,12 +4838,10 @@ F: kernel/rcu* F: kernel/srcu* X: kernel/rcutorture.c -REAL TIME CLOCK DRIVER +REAL TIME CLOCK DRIVER (LEGACY) M: Paul Gortmaker S: Maintained -F: Documentation/rtc.txt -F: drivers/rtc/ -F: include/linux/rtc.h +F: drivers/char/rtc.c REAL TIME CLOCK (RTC) SUBSYSTEM M: Alessandro Zummo -- cgit v0.10.2 From 47016434257b90445113eed1c5b8b57eb2d35330 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Thu, 9 Sep 2010 16:38:04 -0700 Subject: drivers/rtc/rtc-pl031.c: do not mark PL031 IRQ as shared It was a mistake to mark the PL031 IRQ as shared (for the U8500), we misread the datasheet. Get rid of this. Signed-off-by: Linus Walleij Cc: Jonas Aberg Cc: Mian Yousaf Kaukab Acked-by: Alessandro Zummo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-pl031.c b/drivers/rtc/rtc-pl031.c index 6c418fe..b7a6690 100644 --- a/drivers/rtc/rtc-pl031.c +++ b/drivers/rtc/rtc-pl031.c @@ -403,7 +403,7 @@ static int pl031_probe(struct amba_device *adev, struct amba_id *id) } if (request_irq(adev->irq[0], pl031_interrupt, - IRQF_DISABLED | IRQF_SHARED, "rtc-pl031", ldata)) { + IRQF_DISABLED, "rtc-pl031", ldata)) { ret = -EIO; goto out_no_irq; } -- cgit v0.10.2 From ac8456d6f9a3011c824176bd6084d39e5f70a382 Mon Sep 17 00:00:00 2001 From: Gary King Date: Thu, 9 Sep 2010 16:38:05 -0700 Subject: bounce: call flush_dcache_page() after bounce_copy_vec() I have been seeing problems on Tegra 2 (ARMv7 SMP) systems with HIGHMEM enabled on 2.6.35 (plus some patches targetted at 2.6.36 to perform cache maintenance lazily), and the root cause appears to be that the mm bouncing code is calling flush_dcache_page before it copies the bounce buffer into the bio. The bounced page needs to be flushed after data is copied into it, to ensure that architecture implementations can synchronize instruction and data caches if necessary. Signed-off-by: Gary King Cc: Tejun Heo Cc: Russell King Acked-by: Jens Axboe Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/bounce.c b/mm/bounce.c index 13b6dad..1481de6 100644 --- a/mm/bounce.c +++ b/mm/bounce.c @@ -116,8 +116,8 @@ static void copy_to_high_bio_irq(struct bio *to, struct bio *from) */ vfrom = page_address(fromvec->bv_page) + tovec->bv_offset; - flush_dcache_page(tovec->bv_page); bounce_copy_vec(tovec, vfrom); + flush_dcache_page(tovec->bv_page); } } -- cgit v0.10.2 From 910321ea817a202ff70fac666e37e2c8e2f88823 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 9 Sep 2010 16:38:07 -0700 Subject: swap: revert special hibernation allocation Please revert 2.6.36-rc commit d2997b1042ec150616c1963b5e5e919ffd0b0ebf "hibernation: freeze swap at hibernation". It complicated matters by adding a second swap allocation path, just for hibernation; without in any way fixing the issue that it was intended to address - page reclaim after fixing the hibernation image might free swap from a page already imaged as swapcache, letting its swap be reallocated to store a different page of the image: resulting in data corruption if the imaged page were freed as clean then swapped back in. Pages freed to si->swap_map were still in danger of being reallocated by the alternative allocation path. I guess it inadvertently fixed slow SSD swap allocation for hibernation, as reported by Nigel Cunningham: by missing out the discards that occur on the usual swap allocation path; but that was unintentional, and needs a separate fix. Signed-off-by: Hugh Dickins Cc: KAMEZAWA Hiroyuki Cc: KOSAKI Motohiro Cc: "Rafael J. Wysocki" Cc: Ondrej Zary Cc: Andrea Gelmini Cc: Balbir Singh Cc: Andrea Arcangeli Cc: Nigel Cunningham Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/swap.h b/include/linux/swap.h index 2fee51a..bf4eb62 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -315,6 +315,7 @@ extern long nr_swap_pages; extern long total_swap_pages; extern void si_swapinfo(struct sysinfo *); extern swp_entry_t get_swap_page(void); +extern swp_entry_t get_swap_page_of_type(int); extern int valid_swaphandles(swp_entry_t, unsigned long *); extern int add_swap_count_continuation(swp_entry_t, gfp_t); extern void swap_shmem_alloc(swp_entry_t); @@ -331,13 +332,6 @@ extern int reuse_swap_page(struct page *); extern int try_to_free_swap(struct page *); struct backing_dev_info; -#ifdef CONFIG_HIBERNATION -void hibernation_freeze_swap(void); -void hibernation_thaw_swap(void); -swp_entry_t get_swap_for_hibernation(int type); -void swap_free_for_hibernation(swp_entry_t val); -#endif - /* linux/mm/thrash.c */ extern struct mm_struct *swap_token_mm; extern void grab_swap_token(struct mm_struct *); diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c index c779639..8dc31e0 100644 --- a/kernel/power/hibernate.c +++ b/kernel/power/hibernate.c @@ -338,7 +338,6 @@ int hibernation_snapshot(int platform_mode) goto Close; suspend_console(); - hibernation_freeze_swap(); saved_mask = clear_gfp_allowed_mask(GFP_IOFS); error = dpm_suspend_start(PMSG_FREEZE); if (error) diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index 5e7edfb..f6cd6fa 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -1086,7 +1086,6 @@ void swsusp_free(void) buffer = NULL; alloc_normal = 0; alloc_highmem = 0; - hibernation_thaw_swap(); } /* Helper functions used for the shrinking of memory. */ diff --git a/kernel/power/swap.c b/kernel/power/swap.c index 5d0059e..e6a5bdf 100644 --- a/kernel/power/swap.c +++ b/kernel/power/swap.c @@ -136,10 +136,10 @@ sector_t alloc_swapdev_block(int swap) { unsigned long offset; - offset = swp_offset(get_swap_for_hibernation(swap)); + offset = swp_offset(get_swap_page_of_type(swap)); if (offset) { if (swsusp_extents_insert(offset)) - swap_free_for_hibernation(swp_entry(swap, offset)); + swap_free(swp_entry(swap, offset)); else return swapdev_block(swap, offset); } @@ -163,7 +163,7 @@ void free_all_swap_pages(int swap) ext = container_of(node, struct swsusp_extent, node); rb_erase(node, &swsusp_extents); for (offset = ext->start; offset <= ext->end; offset++) - swap_free_for_hibernation(swp_entry(swap, offset)); + swap_free(swp_entry(swap, offset)); kfree(ext); } diff --git a/mm/swapfile.c b/mm/swapfile.c index 1f3f9c5..f08d165 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -47,8 +47,6 @@ long nr_swap_pages; long total_swap_pages; static int least_priority; -static bool swap_for_hibernation; - static const char Bad_file[] = "Bad swap file entry "; static const char Unused_file[] = "Unused swap file entry "; static const char Bad_offset[] = "Bad swap offset entry "; @@ -453,8 +451,6 @@ swp_entry_t get_swap_page(void) spin_lock(&swap_lock); if (nr_swap_pages <= 0) goto noswap; - if (swap_for_hibernation) - goto noswap; nr_swap_pages--; for (type = swap_list.next; type >= 0 && wrapped < 2; type = next) { @@ -487,6 +483,28 @@ noswap: return (swp_entry_t) {0}; } +/* The only caller of this function is now susupend routine */ +swp_entry_t get_swap_page_of_type(int type) +{ + struct swap_info_struct *si; + pgoff_t offset; + + spin_lock(&swap_lock); + si = swap_info[type]; + if (si && (si->flags & SWP_WRITEOK)) { + nr_swap_pages--; + /* This is called for allocating swap entry, not cache */ + offset = scan_swap_map(si, 1); + if (offset) { + spin_unlock(&swap_lock); + return swp_entry(type, offset); + } + nr_swap_pages++; + } + spin_unlock(&swap_lock); + return (swp_entry_t) {0}; +} + static struct swap_info_struct *swap_info_get(swp_entry_t entry) { struct swap_info_struct *p; @@ -746,74 +764,6 @@ int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep) #endif #ifdef CONFIG_HIBERNATION - -static pgoff_t hibernation_offset[MAX_SWAPFILES]; -/* - * Once hibernation starts to use swap, we freeze swap_map[]. Otherwise, - * saved swap_map[] image to the disk will be an incomplete because it's - * changing without synchronization with hibernation snap shot. - * At resume, we just make swap_for_hibernation=false. We can forget - * used maps easily. - */ -void hibernation_freeze_swap(void) -{ - int i; - - spin_lock(&swap_lock); - - printk(KERN_INFO "PM: Freeze Swap\n"); - swap_for_hibernation = true; - for (i = 0; i < MAX_SWAPFILES; i++) - hibernation_offset[i] = 1; - spin_unlock(&swap_lock); -} - -void hibernation_thaw_swap(void) -{ - spin_lock(&swap_lock); - if (swap_for_hibernation) { - printk(KERN_INFO "PM: Thaw Swap\n"); - swap_for_hibernation = false; - } - spin_unlock(&swap_lock); -} - -/* - * Because updateing swap_map[] can make not-saved-status-change, - * we use our own easy allocator. - * Please see kernel/power/swap.c, Used swaps are recorded into - * RB-tree. - */ -swp_entry_t get_swap_for_hibernation(int type) -{ - pgoff_t off; - swp_entry_t val = {0}; - struct swap_info_struct *si; - - spin_lock(&swap_lock); - - si = swap_info[type]; - if (!si || !(si->flags & SWP_WRITEOK)) - goto done; - - for (off = hibernation_offset[type]; off < si->max; ++off) { - if (!si->swap_map[off]) - break; - } - if (off < si->max) { - val = swp_entry(type, off); - hibernation_offset[type] = off + 1; - } -done: - spin_unlock(&swap_lock); - return val; -} - -void swap_free_for_hibernation(swp_entry_t ent) -{ - /* Nothing to do */ -} - /* * Find the swap type that corresponds to given device (if any). * -- cgit v0.10.2 From b73d7fcecd93dc15eaa3c45c8c587b613f6673c4 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 9 Sep 2010 16:38:09 -0700 Subject: swap: prevent reuse during hibernation Move the hibernation check from scan_swap_map() into try_to_free_swap(): to catch not only the common case when hibernation's allocation itself triggers swap reuse, but also the less likely case when concurrent page reclaim (shrink_page_list) might happen to try_to_free_swap from a page. Hibernation already clears __GFP_IO from the gfp_allowed_mask, to stop reclaim from going to swap: check that to prevent swap reuse too. Signed-off-by: Hugh Dickins Cc: KAMEZAWA Hiroyuki Cc: KOSAKI Motohiro Cc: "Rafael J. Wysocki" Cc: Ondrej Zary Cc: Andrea Gelmini Cc: Balbir Singh Cc: Andrea Arcangeli Cc: Nigel Cunningham Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/swapfile.c b/mm/swapfile.c index f08d165..ed51510 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -318,10 +318,8 @@ checks: if (offset > si->highest_bit) scan_base = offset = si->lowest_bit; - /* reuse swap entry of cache-only swap if not hibernation. */ - if (vm_swap_full() - && usage == SWAP_HAS_CACHE - && si->swap_map[offset] == SWAP_HAS_CACHE) { + /* reuse swap entry of cache-only swap if not busy. */ + if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { int swap_was_freed; spin_unlock(&swap_lock); swap_was_freed = __try_to_reclaim_swap(si, offset); @@ -688,6 +686,24 @@ int try_to_free_swap(struct page *page) if (page_swapcount(page)) return 0; + /* + * Once hibernation has begun to create its image of memory, + * there's a danger that one of the calls to try_to_free_swap() + * - most probably a call from __try_to_reclaim_swap() while + * hibernation is allocating its own swap pages for the image, + * but conceivably even a call from memory reclaim - will free + * the swap from a page which has already been recorded in the + * image as a clean swapcache page, and then reuse its swap for + * another page of the image. On waking from hibernation, the + * original page might be freed under memory pressure, then + * later read back in from swap, now with the wrong data. + * + * Hibernation clears bits from gfp_allowed_mask to prevent + * memory reclaim from writing to disk, so check that here. + */ + if (!(gfp_allowed_mask & __GFP_IO)) + return 0; + delete_from_swap_cache(page); SetPageDirty(page); return 1; -- cgit v0.10.2 From 8f2ae0faa3a119158c4dcfe89926d6fad5f5332c Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 9 Sep 2010 16:38:10 -0700 Subject: swap: do not send discards as barriers The swap code already uses synchronous discards, no need to add I/O barriers. This fixes the worst of the terrible slowdown in swap allocation for hibernation, reported on 2.6.35 by Nigel Cunningham; but does not entirely eliminate that regression. [tj@kernel.org: superflous newlines removed] Signed-off-by: Christoph Hellwig Tested-by: Nigel Cunningham Signed-off-by: Tejun Heo Signed-off-by: Hugh Dickins Cc: Jens Axboe Cc: James Bottomley Cc: "Martin K. Petersen" Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/swapfile.c b/mm/swapfile.c index ed51510..1894dea 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -139,8 +139,7 @@ static int discard_swap(struct swap_info_struct *si) nr_blocks = ((sector_t)se->nr_pages - 1) << (PAGE_SHIFT - 9); if (nr_blocks) { err = blkdev_issue_discard(si->bdev, start_block, - nr_blocks, GFP_KERNEL, - BLKDEV_IFL_WAIT | BLKDEV_IFL_BARRIER); + nr_blocks, GFP_KERNEL, BLKDEV_IFL_WAIT); if (err) return err; cond_resched(); @@ -151,8 +150,7 @@ static int discard_swap(struct swap_info_struct *si) nr_blocks = (sector_t)se->nr_pages << (PAGE_SHIFT - 9); err = blkdev_issue_discard(si->bdev, start_block, - nr_blocks, GFP_KERNEL, - BLKDEV_IFL_WAIT | BLKDEV_IFL_BARRIER); + nr_blocks, GFP_KERNEL, BLKDEV_IFL_WAIT); if (err) break; @@ -191,8 +189,7 @@ static void discard_swap_cluster(struct swap_info_struct *si, start_block <<= PAGE_SHIFT - 9; nr_blocks <<= PAGE_SHIFT - 9; if (blkdev_issue_discard(si->bdev, start_block, - nr_blocks, GFP_NOIO, BLKDEV_IFL_WAIT | - BLKDEV_IFL_BARRIER)) + nr_blocks, GFP_NOIO, BLKDEV_IFL_WAIT)) break; } -- cgit v0.10.2 From 3399446632739fcd05fd8b272b476a69c6e6d14a Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 9 Sep 2010 16:38:11 -0700 Subject: swap: discard while swapping only if SWAP_FLAG_DISCARD Tests with recent firmware on Intel X25-M 80GB and OCZ Vertex 60GB SSDs show a shift since I last tested in December: in part because of firmware updates, in part because of the necessary move from barriers to awaiting completion at the block layer. While discard at swapon still shows as slightly beneficial on both, discarding 1MB swap cluster when allocating is now disadvanteous: adds 25% overhead on Intel, adds 230% on OCZ (YMMV). Surrender: discard as presently implemented is more hindrance than help for swap; but might prove useful on other devices, or with improvements. So continue to do the discard at swapon, but make discard while swapping conditional on a SWAP_FLAG_DISCARD to sys_swapon() (which has been using only the lower 16 bits of int flags). We can add a --discard or -d to swapon(8), and a "discard" to swap in /etc/fstab: matching the mount option for btrfs, ext4, fat, gfs2, nilfs2. Signed-off-by: Hugh Dickins Cc: Christoph Hellwig Cc: Nigel Cunningham Cc: Tejun Heo Cc: Jens Axboe Cc: James Bottomley Cc: "Martin K. Petersen" Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/swap.h b/include/linux/swap.h index bf4eb62..7cdd633 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -19,6 +19,7 @@ struct bio; #define SWAP_FLAG_PREFER 0x8000 /* set if swap priority specified */ #define SWAP_FLAG_PRIO_MASK 0x7fff #define SWAP_FLAG_PRIO_SHIFT 0 +#define SWAP_FLAG_DISCARD 0x10000 /* discard swap cluster after use */ static inline int current_is_kswapd(void) { @@ -142,7 +143,7 @@ struct swap_extent { enum { SWP_USED = (1 << 0), /* is slot in swap_info[] used? */ SWP_WRITEOK = (1 << 1), /* ok to write to this swap? */ - SWP_DISCARDABLE = (1 << 2), /* blkdev supports discard */ + SWP_DISCARDABLE = (1 << 2), /* swapon+blkdev support discard */ SWP_DISCARDING = (1 << 3), /* now discarding a free cluster */ SWP_SOLIDSTATE = (1 << 4), /* blkdev seeks are cheap */ SWP_CONTINUED = (1 << 5), /* swap_map has count continuation */ diff --git a/mm/swapfile.c b/mm/swapfile.c index 1894dea..7c703ff 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -2047,7 +2047,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) p->flags |= SWP_SOLIDSTATE; p->cluster_next = 1 + (random32() % p->highest_bit); } - if (discard_swap(p) == 0) + if (discard_swap(p) == 0 && (swap_flags & SWAP_FLAG_DISCARD)) p->flags |= SWP_DISCARDABLE; } -- cgit v0.10.2 From 3ab04d5cf9736b7a4e9dfcf28285d8663b01aa0e Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 9 Sep 2010 16:38:12 -0700 Subject: vfs: take O_NONBLOCK out of the O_* uniqueness test O_NONBLOCK on parisc has a dual value: #define O_NONBLOCK 000200004 /* HPUX has separate NDELAY & NONBLOCK */ It is caught by the O_* bits uniqueness check and leads to a parisc compile error. The fix would be to take O_NONBLOCK out. Signed-off-by: Wu Fengguang Signed-off-by: James Bottomley Cc: Jamie Lokier Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/fcntl.c b/fs/fcntl.c index 6769fd0..f8cc34f 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -769,11 +769,15 @@ EXPORT_SYMBOL(kill_fasync); static int __init fcntl_init(void) { - /* please add new bits here to ensure allocation uniqueness */ - BUILD_BUG_ON(19 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( + /* + * Please add new bits here to ensure allocation uniqueness. + * Exceptions: O_NONBLOCK is a two bit define on parisc; O_NDELAY + * is defined as O_NONBLOCK on some platforms and not on others. + */ + BUILD_BUG_ON(18 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( O_RDONLY | O_WRONLY | O_RDWR | O_CREAT | O_EXCL | O_NOCTTY | - O_TRUNC | O_APPEND | O_NONBLOCK | + O_TRUNC | O_APPEND | /* O_NONBLOCK | */ __O_SYNC | O_DSYNC | FASYNC | O_DIRECT | O_LARGEFILE | O_DIRECTORY | O_NOFOLLOW | O_NOATIME | O_CLOEXEC | -- cgit v0.10.2 From 5ee28a447625b9fe64fbf7cff026561084fc5f16 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Thu, 9 Sep 2010 16:38:14 -0700 Subject: vmstat: update zone stat threshold when onlining a cpu refresh_zone_stat_thresholds() calculates parameter based on the number of online cpus. It's called at cpu offlining but needs to be called at onlining, too. Signed-off-by: KAMEZAWA Hiroyuki Cc: Christoph Lameter Acked-by: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/vmstat.c b/mm/vmstat.c index f389168..a8d6b59 100644 --- a/mm/vmstat.c +++ b/mm/vmstat.c @@ -998,6 +998,7 @@ static int __cpuinit vmstat_cpuup_callback(struct notifier_block *nfb, switch (action) { case CPU_ONLINE: case CPU_ONLINE_FROZEN: + refresh_zone_stat_thresholds(); start_cpu_timer(cpu); node_set_state(cpu_to_node(cpu), N_CPU); break; -- cgit v0.10.2 From 72853e2991a2702ae93aaf889ac7db743a415dd3 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Thu, 9 Sep 2010 16:38:16 -0700 Subject: mm: page allocator: update free page counters after pages are placed on the free list When allocating a page, the system uses NR_FREE_PAGES counters to determine if watermarks would remain intact after the allocation was made. This check is made without interrupts disabled or the zone lock held and so is race-prone by nature. Unfortunately, when pages are being freed in batch, the counters are updated before the pages are added on the list. During this window, the counters are misleading as the pages do not exist yet. When under significant pressure on systems with large numbers of CPUs, it's possible for processes to make progress even though they should have been stalled. This is particularly problematic if a number of the processes are using GFP_ATOMIC as the min watermark can be accidentally breached and in extreme cases, the system can livelock. This patch updates the counters after the pages have been added to the list. This makes the allocator more cautious with respect to preserving the watermarks and mitigates livelock possibilities. [akpm@linux-foundation.org: avoid modifying incoming args] Signed-off-by: Mel Gorman Reviewed-by: Rik van Riel Reviewed-by: Minchan Kim Reviewed-by: KAMEZAWA Hiroyuki Reviewed-by: Christoph Lameter Reviewed-by: KOSAKI Motohiro Acked-by: Johannes Weiner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/page_alloc.c b/mm/page_alloc.c index a9649f4..452e2ba0 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -588,13 +588,13 @@ static void free_pcppages_bulk(struct zone *zone, int count, { int migratetype = 0; int batch_free = 0; + int to_free = count; spin_lock(&zone->lock); zone->all_unreclaimable = 0; zone->pages_scanned = 0; - __mod_zone_page_state(zone, NR_FREE_PAGES, count); - while (count) { + while (to_free) { struct page *page; struct list_head *list; @@ -619,8 +619,9 @@ static void free_pcppages_bulk(struct zone *zone, int count, /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */ __free_one_page(page, zone, 0, page_private(page)); trace_mm_page_pcpu_drain(page, 0, page_private(page)); - } while (--count && --batch_free && !list_empty(list)); + } while (--to_free && --batch_free && !list_empty(list)); } + __mod_zone_page_state(zone, NR_FREE_PAGES, count); spin_unlock(&zone->lock); } @@ -631,8 +632,8 @@ static void free_one_page(struct zone *zone, struct page *page, int order, zone->all_unreclaimable = 0; zone->pages_scanned = 0; - __mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order); __free_one_page(page, zone, order, migratetype); + __mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order); spin_unlock(&zone->lock); } -- cgit v0.10.2 From aa45484031ddee09b06350ab8528bfe5b2c76d1c Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 9 Sep 2010 16:38:17 -0700 Subject: mm: page allocator: calculate a better estimate of NR_FREE_PAGES when memory is low and kswapd is awake Ordinarily watermark checks are based on the vmstat NR_FREE_PAGES as it is cheaper than scanning a number of lists. To avoid synchronization overhead, counter deltas are maintained on a per-cpu basis and drained both periodically and when the delta is above a threshold. On large CPU systems, the difference between the estimated and real value of NR_FREE_PAGES can be very high. If NR_FREE_PAGES is much higher than number of real free page in buddy, the VM can allocate pages below min watermark, at worst reducing the real number of pages to zero. Even if the OOM killer kills some victim for freeing memory, it may not free memory if the exit path requires a new page resulting in livelock. This patch introduces a zone_page_state_snapshot() function (courtesy of Christoph) that takes a slightly more accurate view of an arbitrary vmstat counter. It is used to read NR_FREE_PAGES while kswapd is awake to avoid the watermark being accidentally broken. The estimate is not perfect and may result in cache line bounces but is expected to be lighter than the IPI calls necessary to continually drain the per-cpu counters while kswapd is awake. Signed-off-by: Christoph Lameter Signed-off-by: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 6e6e626..3984c4e 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -284,6 +284,13 @@ struct zone { unsigned long watermark[NR_WMARK]; /* + * When free pages are below this point, additional steps are taken + * when reading the number of free pages to avoid per-cpu counter + * drift allowing watermarks to be breached + */ + unsigned long percpu_drift_mark; + + /* * We don't know if the memory that we're going to allocate will be freeable * or/and it will be released eventually, so to avoid totally wasting several * GB of ram we must reserve some of the lower zone memory (otherwise we risk @@ -441,6 +448,12 @@ static inline int zone_is_oom_locked(const struct zone *zone) return test_bit(ZONE_OOM_LOCKED, &zone->flags); } +#ifdef CONFIG_SMP +unsigned long zone_nr_free_pages(struct zone *zone); +#else +#define zone_nr_free_pages(zone) zone_page_state(zone, NR_FREE_PAGES) +#endif /* CONFIG_SMP */ + /* * The "priority" of VM scanning is how much of the queues we will scan in one * go. A value of 12 for DEF_PRIORITY implies that we will scan 1/4096th of the diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h index 7f43ccd..eaaea37 100644 --- a/include/linux/vmstat.h +++ b/include/linux/vmstat.h @@ -170,6 +170,28 @@ static inline unsigned long zone_page_state(struct zone *zone, return x; } +/* + * More accurate version that also considers the currently pending + * deltas. For that we need to loop over all cpus to find the current + * deltas. There is no synchronization so the result cannot be + * exactly accurate either. + */ +static inline unsigned long zone_page_state_snapshot(struct zone *zone, + enum zone_stat_item item) +{ + long x = atomic_long_read(&zone->vm_stat[item]); + +#ifdef CONFIG_SMP + int cpu; + for_each_online_cpu(cpu) + x += per_cpu_ptr(zone->pageset, cpu)->vm_stat_diff[item]; + + if (x < 0) + x = 0; +#endif + return x; +} + extern unsigned long global_reclaimable_pages(void); extern unsigned long zone_reclaimable_pages(struct zone *zone); diff --git a/mm/mmzone.c b/mm/mmzone.c index f5b7d17..e35bfb8 100644 --- a/mm/mmzone.c +++ b/mm/mmzone.c @@ -87,3 +87,24 @@ int memmap_valid_within(unsigned long pfn, return 1; } #endif /* CONFIG_ARCH_HAS_HOLES_MEMORYMODEL */ + +#ifdef CONFIG_SMP +/* Called when a more accurate view of NR_FREE_PAGES is needed */ +unsigned long zone_nr_free_pages(struct zone *zone) +{ + unsigned long nr_free_pages = zone_page_state(zone, NR_FREE_PAGES); + + /* + * While kswapd is awake, it is considered the zone is under some + * memory pressure. Under pressure, there is a risk that + * per-cpu-counter-drift will allow the min watermark to be breached + * potentially causing a live-lock. While kswapd is awake and + * free pages are low, get a better estimate for free pages + */ + if (nr_free_pages < zone->percpu_drift_mark && + !waitqueue_active(&zone->zone_pgdat->kswapd_wait)) + return zone_page_state_snapshot(zone, NR_FREE_PAGES); + + return nr_free_pages; +} +#endif /* CONFIG_SMP */ diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 452e2ba0..b2d21e0 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1462,7 +1462,7 @@ int zone_watermark_ok(struct zone *z, int order, unsigned long mark, { /* free_pages my go negative - that's OK */ long min = mark; - long free_pages = zone_page_state(z, NR_FREE_PAGES) - (1 << order) + 1; + long free_pages = zone_nr_free_pages(z) - (1 << order) + 1; int o; if (alloc_flags & ALLOC_HIGH) @@ -2424,7 +2424,7 @@ void show_free_areas(void) " all_unreclaimable? %s" "\n", zone->name, - K(zone_page_state(zone, NR_FREE_PAGES)), + K(zone_nr_free_pages(zone)), K(min_wmark_pages(zone)), K(low_wmark_pages(zone)), K(high_wmark_pages(zone)), diff --git a/mm/vmstat.c b/mm/vmstat.c index a8d6b59..355a9e6 100644 --- a/mm/vmstat.c +++ b/mm/vmstat.c @@ -138,11 +138,24 @@ static void refresh_zone_stat_thresholds(void) int threshold; for_each_populated_zone(zone) { + unsigned long max_drift, tolerate_drift; + threshold = calculate_threshold(zone); for_each_online_cpu(cpu) per_cpu_ptr(zone->pageset, cpu)->stat_threshold = threshold; + + /* + * Only set percpu_drift_mark if there is a danger that + * NR_FREE_PAGES reports the low watermark is ok when in fact + * the min watermark could be breached by an allocation + */ + tolerate_drift = low_wmark_pages(zone) - min_wmark_pages(zone); + max_drift = num_online_cpus() * threshold; + if (max_drift > tolerate_drift) + zone->percpu_drift_mark = high_wmark_pages(zone) + + max_drift; } } @@ -813,7 +826,7 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat, "\n scanned %lu" "\n spanned %lu" "\n present %lu", - zone_page_state(zone, NR_FREE_PAGES), + zone_nr_free_pages(zone), min_wmark_pages(zone), low_wmark_pages(zone), high_wmark_pages(zone), -- cgit v0.10.2 From 9ee493ce0a60bf42c0f8fd0b0fe91df5704a1cbf Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Thu, 9 Sep 2010 16:38:18 -0700 Subject: mm: page allocator: drain per-cpu lists after direct reclaim allocation fails When under significant memory pressure, a process enters direct reclaim and immediately afterwards tries to allocate a page. If it fails and no further progress is made, it's possible the system will go OOM. However, on systems with large amounts of memory, it's possible that a significant number of pages are on per-cpu lists and inaccessible to the calling process. This leads to a process entering direct reclaim more often than it should increasing the pressure on the system and compounding the problem. This patch notes that if direct reclaim is making progress but allocations are still failing that the system is already under heavy pressure. In this case, it drains the per-cpu lists and tries the allocation a second time before continuing. Signed-off-by: Mel Gorman Reviewed-by: Minchan Kim Reviewed-by: KAMEZAWA Hiroyuki Reviewed-by: KOSAKI Motohiro Reviewed-by: Christoph Lameter Cc: Dave Chinner Cc: Wu Fengguang Cc: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/page_alloc.c b/mm/page_alloc.c index b2d21e0..a8cfa9c 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1847,6 +1847,7 @@ __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order, struct page *page = NULL; struct reclaim_state reclaim_state; struct task_struct *p = current; + bool drained = false; cond_resched(); @@ -1865,14 +1866,25 @@ __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order, cond_resched(); - if (order != 0) - drain_all_pages(); + if (unlikely(!(*did_some_progress))) + return NULL; - if (likely(*did_some_progress)) - page = get_page_from_freelist(gfp_mask, nodemask, order, +retry: + page = get_page_from_freelist(gfp_mask, nodemask, order, zonelist, high_zoneidx, alloc_flags, preferred_zone, migratetype); + + /* + * If an allocation failed after direct reclaim, it could be because + * pages are pinned on the per-cpu lists. Drain them and try again + */ + if (!page && !drained) { + drain_all_pages(); + drained = true; + goto retry; + } + return page; } -- cgit v0.10.2 From eee743fd7eac9f2ea69ad06d093dfb5a12538fe5 Mon Sep 17 00:00:00 2001 From: "Jorge Boncompte [DTI2]" Date: Thu, 9 Sep 2010 16:38:19 -0700 Subject: minix: fix regression in minix_mkdir() Commit 9eed1fb721c ("minix: replace inode uid,gid,mode init with helper") broke directory creation on minix filesystems. Fix it by passing the needed mode flag to inode init helper. Signed-off-by: Jorge Boncompte [DTI2] Cc: Dmitry Monakhov Cc: Al Viro Cc: [2.6.35.x] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/minix/namei.c b/fs/minix/namei.c index e20ee85..f3f3578 100644 --- a/fs/minix/namei.c +++ b/fs/minix/namei.c @@ -115,7 +115,7 @@ static int minix_mkdir(struct inode * dir, struct dentry *dentry, int mode) inode_inc_link_count(dir); - inode = minix_new_inode(dir, mode, &err); + inode = minix_new_inode(dir, S_IFDIR | mode, &err); if (!inode) goto out_dir; -- cgit v0.10.2 From 40c6023031369ae5573e622ca54fa3ffe89fb865 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 9 Sep 2010 17:13:31 +0200 Subject: libata,pata_via: revert ata_wait_idle() removal from ata_sff/via_tf_load() Commit 978c0666 (libata: Remove excess delay in the tf_load path) removed ata_wait_idle() from ata_sff_tf_load() and via_tf_load(). This caused obscure detection problems in sata_sil. https://bugzilla.kernel.org/show_bug.cgi?id=16606 The commit was pure performance optimization. Revert it for now. Reported-by: Dieter Plaetinck Reported-by: Jan Beulich Bisected-by: gianluca Cc: stable@kernel.org Signed-off-by: Jeff Garzik diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c index 3b82d8e..dee3c2c 100644 --- a/drivers/ata/libata-sff.c +++ b/drivers/ata/libata-sff.c @@ -418,6 +418,7 @@ void ata_sff_tf_load(struct ata_port *ap, const struct ata_taskfile *tf) if (ioaddr->ctl_addr) iowrite8(tf->ctl, ioaddr->ctl_addr); ap->last_ctl = tf->ctl; + ata_wait_idle(ap); } if (is_addr && (tf->flags & ATA_TFLAG_LBA48)) { @@ -453,6 +454,8 @@ void ata_sff_tf_load(struct ata_port *ap, const struct ata_taskfile *tf) iowrite8(tf->device, ioaddr->device_addr); VPRINTK("device 0x%X\n", tf->device); } + + ata_wait_idle(ap); } EXPORT_SYMBOL_GPL(ata_sff_tf_load); diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c index 5e65988..ac8d7d9 100644 --- a/drivers/ata/pata_via.c +++ b/drivers/ata/pata_via.c @@ -417,6 +417,8 @@ static void via_tf_load(struct ata_port *ap, const struct ata_taskfile *tf) tf->lbam, tf->lbah); } + + ata_wait_idle(ap); } static int via_port_start(struct ata_port *ap) -- cgit v0.10.2 From 238e149c7a92eb79ab9f48c171e907a5bde18333 Mon Sep 17 00:00:00 2001 From: Seth Heasley Date: Thu, 9 Sep 2010 09:42:40 -0700 Subject: ata_piix: IDE Mode SATA patch for Intel Patsburg DeviceIDs This patch adds the Intel Patsburg (PCH) IDE mode SATA Controller DeviceIDs. Signed-off-by: Seth Heasley Signed-off-by: Jeff Garzik diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c index 3971bc0..d712675 100644 --- a/drivers/ata/ata_piix.c +++ b/drivers/ata/ata_piix.c @@ -302,6 +302,10 @@ static const struct pci_device_id piix_pci_tbl[] = { { 0x8086, 0x1c08, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich8_2port_sata }, /* SATA Controller IDE (CPT) */ { 0x8086, 0x1c09, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich8_2port_sata }, + /* SATA Controller IDE (PBG) */ + { 0x8086, 0x1d00, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich8_sata }, + /* SATA Controller IDE (PBG) */ + { 0x8086, 0x1d08, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich8_2port_sata }, { } /* terminate list */ }; -- cgit v0.10.2 From 992b3fb9b5391bc4de5b42bb810dc6dd583a6c4a Mon Sep 17 00:00:00 2001 From: Seth Heasley Date: Thu, 9 Sep 2010 09:44:56 -0700 Subject: ahci: AHCI and RAID mode SATA patch for Intel Patsburg DeviceIDs This patch adds the Intel Patsburg (PCH) SATA AHCI and RAID Controller DeviceIDs. Signed-off-by: Seth Heasley Signed-off-by: Jeff Garzik diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 013727b..ff1c945 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -253,6 +253,9 @@ static const struct pci_device_id ahci_pci_tbl[] = { { PCI_VDEVICE(INTEL, 0x1c05), board_ahci }, /* CPT RAID */ { PCI_VDEVICE(INTEL, 0x1c06), board_ahci }, /* CPT RAID */ { PCI_VDEVICE(INTEL, 0x1c07), board_ahci }, /* CPT RAID */ + { PCI_VDEVICE(INTEL, 0x1d02), board_ahci }, /* PBG AHCI */ + { PCI_VDEVICE(INTEL, 0x1d04), board_ahci }, /* PBG RAID */ + { PCI_VDEVICE(INTEL, 0x1d06), board_ahci }, /* PBG RAID */ /* JMicron 360/1/3/5/6, match class to avoid IDE function */ { PCI_VENDOR_ID_JMICRON, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, -- cgit v0.10.2 From e2f3d75fc0e4a0d03c61872bad39ffa2e74a04ff Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 7 Sep 2010 14:05:31 +0200 Subject: libata: skip EH autopsy and recovery during suspend For some mysterious reason, certain hardware reacts badly to usual EH actions while the system is going for suspend. As the devices won't be needed until the system is resumed, ask EH to skip usual autopsy and recovery and proceed directly to suspend. Signed-off-by: Tejun Heo Tested-by: Stephan Diestelhorst Cc: stable@kernel.org Signed-off-by: Jeff Garzik diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index c035b3d..932eaee 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -5418,6 +5418,7 @@ static int ata_host_request_pm(struct ata_host *host, pm_message_t mesg, */ int ata_host_suspend(struct ata_host *host, pm_message_t mesg) { + unsigned int ehi_flags = ATA_EHI_QUIET; int rc; /* @@ -5426,7 +5427,18 @@ int ata_host_suspend(struct ata_host *host, pm_message_t mesg) */ ata_lpm_enable(host); - rc = ata_host_request_pm(host, mesg, 0, ATA_EHI_QUIET, 1); + /* + * On some hardware, device fails to respond after spun down + * for suspend. As the device won't be used before being + * resumed, we don't need to touch the device. Ask EH to skip + * the usual stuff and proceed directly to suspend. + * + * http://thread.gmane.org/gmane.linux.ide/46764 + */ + if (mesg.event == PM_EVENT_SUSPEND) + ehi_flags |= ATA_EHI_NO_AUTOPSY | ATA_EHI_NO_RECOVERY; + + rc = ata_host_request_pm(host, mesg, 0, ehi_flags, 1); if (rc == 0) host->dev->power.power_state = mesg; return rc; diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index c9ae299..e48302e 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -3235,6 +3235,10 @@ static int ata_eh_skip_recovery(struct ata_link *link) if (link->flags & ATA_LFLAG_DISABLED) return 1; + /* skip if explicitly requested */ + if (ehc->i.flags & ATA_EHI_NO_RECOVERY) + return 1; + /* thaw frozen port and recover failed devices */ if ((ap->pflags & ATA_PFLAG_FROZEN) || ata_link_nr_enabled(link)) return 0; diff --git a/include/linux/libata.h b/include/linux/libata.h index f010f18..7de282d 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -335,6 +335,7 @@ enum { ATA_EHI_HOTPLUGGED = (1 << 0), /* could have been hotplugged */ ATA_EHI_NO_AUTOPSY = (1 << 2), /* no autopsy */ ATA_EHI_QUIET = (1 << 3), /* be quiet */ + ATA_EHI_NO_RECOVERY = (1 << 4), /* no recovery */ ATA_EHI_DID_SOFTRESET = (1 << 16), /* already soft-reset this port */ ATA_EHI_DID_HARDRESET = (1 << 17), /* already soft-reset this port */ -- cgit v0.10.2 From ea3c64506ea7965f86f030155e6fdef381de10e2 Mon Sep 17 00:00:00 2001 From: Gwendal Grignou Date: Tue, 31 Aug 2010 16:20:36 -0700 Subject: libata-sff: Reenable Port Multiplier after libata-sff remodeling. Keep track of the link on the which the current request is in progress. It allows support of links behind port multiplier. Not all libata-sff is PMP compliant. Code for native BMDMA controller does not take in accound PMP. Tested on Marvell 7042 and Sil7526. Signed-off-by: Gwendal Grignou Signed-off-by: Jeff Garzik diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c index dee3c2c..e30c537 100644 --- a/drivers/ata/libata-sff.c +++ b/drivers/ata/libata-sff.c @@ -1045,7 +1045,8 @@ static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq) int ata_sff_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc, u8 status, int in_wq) { - struct ata_eh_info *ehi = &ap->link.eh_info; + struct ata_link *link = qc->dev->link; + struct ata_eh_info *ehi = &link->eh_info; unsigned long flags = 0; int poll_next; @@ -1301,8 +1302,14 @@ fsm_start: } EXPORT_SYMBOL_GPL(ata_sff_hsm_move); -void ata_sff_queue_pio_task(struct ata_port *ap, unsigned long delay) +void ata_sff_queue_pio_task(struct ata_link *link, unsigned long delay) { + struct ata_port *ap = link->ap; + + WARN_ON((ap->sff_pio_task_link != NULL) && + (ap->sff_pio_task_link != link)); + ap->sff_pio_task_link = link; + /* may fail if ata_sff_flush_pio_task() in progress */ queue_delayed_work(ata_sff_wq, &ap->sff_pio_task, msecs_to_jiffies(delay)); @@ -1324,14 +1331,18 @@ static void ata_sff_pio_task(struct work_struct *work) { struct ata_port *ap = container_of(work, struct ata_port, sff_pio_task.work); + struct ata_link *link = ap->sff_pio_task_link; struct ata_queued_cmd *qc; u8 status; int poll_next; + BUG_ON(ap->sff_pio_task_link == NULL); /* qc can be NULL if timeout occurred */ - qc = ata_qc_from_tag(ap, ap->link.active_tag); - if (!qc) + qc = ata_qc_from_tag(ap, link->active_tag); + if (!qc) { + ap->sff_pio_task_link = NULL; return; + } fsm_start: WARN_ON_ONCE(ap->hsm_task_state == HSM_ST_IDLE); @@ -1348,11 +1359,16 @@ fsm_start: msleep(2); status = ata_sff_busy_wait(ap, ATA_BUSY, 10); if (status & ATA_BUSY) { - ata_sff_queue_pio_task(ap, ATA_SHORT_PAUSE); + ata_sff_queue_pio_task(link, ATA_SHORT_PAUSE); return; } } + /* + * hsm_move() may trigger another command to be processed. + * clean the link beforehand. + */ + ap->sff_pio_task_link = NULL; /* move the HSM */ poll_next = ata_sff_hsm_move(ap, qc, status, 1); @@ -1379,6 +1395,7 @@ fsm_start: unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc) { struct ata_port *ap = qc->ap; + struct ata_link *link = qc->dev->link; /* Use polling pio if the LLD doesn't handle * interrupt driven pio and atapi CDB interrupt. @@ -1399,7 +1416,7 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc) ap->hsm_task_state = HSM_ST_LAST; if (qc->tf.flags & ATA_TFLAG_POLLING) - ata_sff_queue_pio_task(ap, 0); + ata_sff_queue_pio_task(link, 0); break; @@ -1412,7 +1429,7 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc) if (qc->tf.flags & ATA_TFLAG_WRITE) { /* PIO data out protocol */ ap->hsm_task_state = HSM_ST_FIRST; - ata_sff_queue_pio_task(ap, 0); + ata_sff_queue_pio_task(link, 0); /* always send first data block using the * ata_sff_pio_task() codepath. @@ -1422,7 +1439,7 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc) ap->hsm_task_state = HSM_ST; if (qc->tf.flags & ATA_TFLAG_POLLING) - ata_sff_queue_pio_task(ap, 0); + ata_sff_queue_pio_task(link, 0); /* if polling, ata_sff_pio_task() handles the * rest. otherwise, interrupt handler takes @@ -1444,7 +1461,7 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc) /* send cdb by polling if no cdb interrupt */ if ((!(qc->dev->flags & ATA_DFLAG_CDB_INTR)) || (qc->tf.flags & ATA_TFLAG_POLLING)) - ata_sff_queue_pio_task(ap, 0); + ata_sff_queue_pio_task(link, 0); break; default: @@ -2737,6 +2754,7 @@ EXPORT_SYMBOL_GPL(ata_bmdma_dumb_qc_prep); unsigned int ata_bmdma_qc_issue(struct ata_queued_cmd *qc) { struct ata_port *ap = qc->ap; + struct ata_link *link = qc->dev->link; /* defer PIO handling to sff_qc_issue */ if (!ata_is_dma(qc->tf.protocol)) @@ -2765,7 +2783,7 @@ unsigned int ata_bmdma_qc_issue(struct ata_queued_cmd *qc) /* send cdb by polling if no cdb interrupt */ if (!(qc->dev->flags & ATA_DFLAG_CDB_INTR)) - ata_sff_queue_pio_task(ap, 0); + ata_sff_queue_pio_task(link, 0); break; default: diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c index 8198259..a9fd970 100644 --- a/drivers/ata/sata_mv.c +++ b/drivers/ata/sata_mv.c @@ -2284,7 +2284,7 @@ static unsigned int mv_qc_issue_fis(struct ata_queued_cmd *qc) } if (qc->tf.flags & ATA_TFLAG_POLLING) - ata_sff_queue_pio_task(ap, 0); + ata_sff_queue_pio_task(link, 0); return 0; } diff --git a/include/linux/libata.h b/include/linux/libata.h index 7de282d..45fb296 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -724,6 +724,7 @@ struct ata_port { struct ata_ioports ioaddr; /* ATA cmd/ctl/dma register blocks */ u8 ctl; /* cache of ATA control register */ u8 last_ctl; /* Cache last written value */ + struct ata_link* sff_pio_task_link; /* link currently used */ struct delayed_work sff_pio_task; #ifdef CONFIG_ATA_BMDMA struct ata_bmdma_prd *bmdma_prd; /* BMDMA SG list */ @@ -1595,7 +1596,7 @@ extern void ata_sff_irq_on(struct ata_port *ap); extern void ata_sff_irq_clear(struct ata_port *ap); extern int ata_sff_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc, u8 status, int in_wq); -extern void ata_sff_queue_pio_task(struct ata_port *ap, unsigned long delay); +extern void ata_sff_queue_pio_task(struct ata_link *link, unsigned long delay); extern unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc); extern bool ata_sff_qc_fill_rtf(struct ata_queued_cmd *qc); extern unsigned int ata_sff_port_intr(struct ata_port *ap, -- cgit v0.10.2 From df09162550fbb53354f0c88e85b5d0e6129ee9cc Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Thu, 9 Sep 2010 16:34:59 -0700 Subject: tracing: t_start: reset FTRACE_ITER_HASH in case of seek/pread Be sure to avoid entering t_show() with FTRACE_ITER_HASH set without having properly started the iterator to iterate the hash. This case is degenerate and, as discovered by Robert Swiecki, can cause t_hash_show() to misuse a pointer. This causes a NULL ptr deref with possible security implications. Tracked as CVE-2010-3079. Cc: Robert Swiecki Cc: Eugene Teo Cc: Signed-off-by: Chris Wright Signed-off-by: Steven Rostedt diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 83a16e9..fa7ece6 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1510,6 +1510,8 @@ static void *t_start(struct seq_file *m, loff_t *pos) if (*pos > 0) return t_hash_start(m, pos); iter->flags |= FTRACE_ITER_PRINTALL; + /* reset in case of seek/pread */ + iter->flags &= ~FTRACE_ITER_HASH; return iter; } -- cgit v0.10.2 From be14eb619108fa8b7120eb2c42d66d5f623ae10e Mon Sep 17 00:00:00 2001 From: Brian King Date: Fri, 10 Sep 2010 09:03:21 +0200 Subject: block: Range check cpu in blk_cpu_to_group While testing CPU DLPAR, the following problem was discovered. We were DLPAR removing the first CPU, which in this case was logical CPUs 0-3. CPUs 0-2 were already marked offline and we were in the process of offlining CPU 3. After marking the CPU inactive and offline in cpu_disable, but before the cpu was completely idle (cpu_die), we ended up in __make_request on CPU 3. There we looked at the topology map to see which CPU to complete the I/O on and found no CPUs in the cpu_sibling_map. This resulted in the block layer setting the completion cpu to be NR_CPUS, which then caused an oops when we tried to complete the I/O. Fix this by sanity checking the value we return from blk_cpu_to_group to be a valid cpu value. Signed-off-by: Brian King Signed-off-by: Jens Axboe diff --git a/block/blk.h b/block/blk.h index 6e7dc87..d6b911a 100644 --- a/block/blk.h +++ b/block/blk.h @@ -142,14 +142,18 @@ static inline int queue_congestion_off_threshold(struct request_queue *q) static inline int blk_cpu_to_group(int cpu) { + int group = NR_CPUS; #ifdef CONFIG_SCHED_MC const struct cpumask *mask = cpu_coregroup_mask(cpu); - return cpumask_first(mask); + group = cpumask_first(mask); #elif defined(CONFIG_SCHED_SMT) - return cpumask_first(topology_thread_cpumask(cpu)); + group = cpumask_first(topology_thread_cpumask(cpu)); #else return cpu; #endif + if (likely(group < NR_CPUS)) + return group; + return cpu; } /* -- cgit v0.10.2 From 1879c45cce6411fe97b0b65ccad38f5206700dda Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Fri, 10 Sep 2010 11:38:43 +0200 Subject: AT91: dm9000 initialization update Add information in dm9000 mac/phy chip initialization: - irq resource details - platform data details Signed-off-by: Nicolas Ferre diff --git a/arch/arm/mach-at91/board-sam9261ek.c b/arch/arm/mach-at91/board-sam9261ek.c index c4c8865..b4307ab 100644 --- a/arch/arm/mach-at91/board-sam9261ek.c +++ b/arch/arm/mach-at91/board-sam9261ek.c @@ -93,11 +93,12 @@ static struct resource dm9000_resource[] = { .start = AT91_PIN_PC11, .end = AT91_PIN_PC11, .flags = IORESOURCE_IRQ + | IORESOURCE_IRQ_LOWEDGE | IORESOURCE_IRQ_HIGHEDGE, } }; static struct dm9000_plat_data dm9000_platdata = { - .flags = DM9000_PLATF_16BITONLY, + .flags = DM9000_PLATF_16BITONLY | DM9000_PLATF_NO_EEPROM, }; static struct platform_device dm9000_device = { -- cgit v0.10.2 From 64d72bbeeb68b536a27319471f212d413f58bf6f Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Fri, 10 Sep 2010 11:26:42 +0200 Subject: AT91: at91sam9261ek board: remove warnings related to use of SPI or SD/MMC The sd/mmc data structure is not used if SPI is selected. The configuration of PIO on the board prevent from using both interfaces at the same time (board dependent). Remove the warnings at compilation time adding a preprocessor condition. Signed-off-by: Nicolas Ferre diff --git a/arch/arm/mach-at91/board-sam9261ek.c b/arch/arm/mach-at91/board-sam9261ek.c index b4307ab..19d4d2c 100644 --- a/arch/arm/mach-at91/board-sam9261ek.c +++ b/arch/arm/mach-at91/board-sam9261ek.c @@ -169,17 +169,6 @@ static struct at91_udc_data __initdata ek_udc_data = { /* - * MCI (SD/MMC) - */ -static struct at91_mmc_data __initdata ek_mmc_data = { - .wire4 = 1, -// .det_pin = ... not connected -// .wp_pin = ... not connected -// .vcc_pin = ... not connected -}; - - -/* * NAND flash */ static struct mtd_partition __initdata ek_nand_partition[] = { @@ -247,6 +236,10 @@ static void __init ek_add_device_nand(void) at91_add_device_nand(&ek_nand_data); } +/* + * SPI related devices + */ +#if defined(CONFIG_SPI_ATMEL) || defined(CONFIG_SPI_ATMEL_MODULE) /* * ADS7846 Touchscreen @@ -357,6 +350,21 @@ static struct spi_board_info ek_spi_devices[] = { #endif }; +#else /* CONFIG_SPI_ATMEL_* */ + +/* + * MCI (SD/MMC) + * spi0 and mmc/sd share the same PIO pins: cannot be used at the same time + */ +static struct at91_mmc_data __initdata ek_mmc_data = { + .wire4 = 1, +// .det_pin = ... not connected +// .wp_pin = ... not connected +// .vcc_pin = ... not connected +}; + +#endif /* CONFIG_SPI_ATMEL_* */ + /* * LCD Controller -- cgit v0.10.2 From 4deb22a6001eb0d4058295c3645c36190a0ec603 Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Fri, 10 Sep 2010 14:36:06 +0200 Subject: AT91: at91sam9261ek: remove C99 comments but keep information Signed-off-by: Nicolas Ferre diff --git a/arch/arm/mach-at91/board-sam9261ek.c b/arch/arm/mach-at91/board-sam9261ek.c index 19d4d2c..65eb094 100644 --- a/arch/arm/mach-at91/board-sam9261ek.c +++ b/arch/arm/mach-at91/board-sam9261ek.c @@ -351,16 +351,14 @@ static struct spi_board_info ek_spi_devices[] = { }; #else /* CONFIG_SPI_ATMEL_* */ +/* spi0 and mmc/sd share the same PIO pins: cannot be used at the same time */ /* * MCI (SD/MMC) - * spi0 and mmc/sd share the same PIO pins: cannot be used at the same time + * det_pin, wp_pin and vcc_pin are not connected */ static struct at91_mmc_data __initdata ek_mmc_data = { .wire4 = 1, -// .det_pin = ... not connected -// .wp_pin = ... not connected -// .vcc_pin = ... not connected }; #endif /* CONFIG_SPI_ATMEL_* */ -- cgit v0.10.2 From a122eb2fdfd78b58c6dd992d6f4b1aaef667eef9 Mon Sep 17 00:00:00 2001 From: Dan Rosenberg Date: Mon, 6 Sep 2010 18:24:57 -0400 Subject: xfs: prevent reading uninitialized stack memory The XFS_IOC_FSGETXATTR ioctl allows unprivileged users to read 12 bytes of uninitialized stack memory, because the fsxattr struct declared on the stack in xfs_ioc_fsgetxattr() does not alter (or zero) the 12-byte fsx_pad member before copying it back to the user. This patch takes care of it. Signed-off-by: Dan Rosenberg Reviewed-by: Eric Sandeen Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index 4fec427..3b9e626 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -785,6 +785,8 @@ xfs_ioc_fsgetxattr( { struct fsxattr fa; + memset(&fa, 0, sizeof(struct fsxattr)); + xfs_ilock(ip, XFS_ILOCK_SHARED); fa.fsx_xflags = xfs_ip2xflags(ip); fa.fsx_extsize = ip->i_d.di_extsize << ip->i_mount->m_sb.sb_blocklog; -- cgit v0.10.2 From dd8849c8f59ec1cee4809a0c5e603e045abe860e Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Thu, 9 Sep 2010 11:58:02 -0700 Subject: drm/i915: don't enable self-refresh on Ironlake We don't know how to enable it safely, especially as outputs turn on and off. When disabling LP1 we also need to make sure LP2 and 3 are already disabled. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29173 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29082 Reported-by: Chris Lord Signed-off-by: Jesse Barnes Tested-by: Daniel Vetter Cc: stable@kernel.org Signed-off-by: Chris Wilson diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index d094e91..4f5e155 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -2206,9 +2206,17 @@ #define WM1_LP_SR_EN (1<<31) #define WM1_LP_LATENCY_SHIFT 24 #define WM1_LP_LATENCY_MASK (0x7f<<24) +#define WM1_LP_FBC_LP1_MASK (0xf<<20) +#define WM1_LP_FBC_LP1_SHIFT 20 #define WM1_LP_SR_MASK (0x1ff<<8) #define WM1_LP_SR_SHIFT 8 #define WM1_LP_CURSOR_MASK (0x3f) +#define WM2_LP_ILK 0x4510c +#define WM2_LP_EN (1<<31) +#define WM3_LP_ILK 0x45110 +#define WM3_LP_EN (1<<31) +#define WM1S_LP_ILK 0x45120 +#define WM1S_LP_EN (1<<31) /* Memory latency timer register */ #define MLTR_ILK 0x11222 diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 7c91030..19daead 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -3382,8 +3382,7 @@ static void ironlake_update_wm(struct drm_device *dev, int planea_clock, reg_value = I915_READ(WM1_LP_ILK); reg_value &= ~(WM1_LP_LATENCY_MASK | WM1_LP_SR_MASK | WM1_LP_CURSOR_MASK); - reg_value |= WM1_LP_SR_EN | - (ilk_sr_latency << WM1_LP_LATENCY_SHIFT) | + reg_value |= (ilk_sr_latency << WM1_LP_LATENCY_SHIFT) | (sr_wm << WM1_LP_SR_SHIFT) | cursor_wm; I915_WRITE(WM1_LP_ILK, reg_value); @@ -5669,6 +5668,9 @@ void intel_init_clock_gating(struct drm_device *dev) I915_WRITE(DISP_ARB_CTL, (I915_READ(DISP_ARB_CTL) | DISP_FBC_WM_DIS)); + I915_WRITE(WM3_LP_ILK, 0); + I915_WRITE(WM2_LP_ILK, 0); + I915_WRITE(WM1_LP_ILK, 0); } /* * Based on the document from hardware guys the following bits -- cgit v0.10.2 From 9d1ac65a9698513d00e5608d93fca0c53f536c14 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 10 Sep 2010 09:59:46 +0100 Subject: KEYS: Fix RCU no-lock warning in keyctl_session_to_parent() There's an protected access to the parent process's credentials in the middle of keyctl_session_to_parent(). This results in the following RCU warning: =================================================== [ INFO: suspicious rcu_dereference_check() usage. ] --------------------------------------------------- security/keys/keyctl.c:1291 invoked rcu_dereference_check() without protection! other info that might help us debug this: rcu_scheduler_active = 1, debug_locks = 0 1 lock held by keyctl-session-/2137: #0: (tasklist_lock){.+.+..}, at: [] keyctl_session_to_parent+0x60/0x236 stack backtrace: Pid: 2137, comm: keyctl-session- Not tainted 2.6.36-rc2-cachefs+ #1 Call Trace: [] lockdep_rcu_dereference+0xaa/0xb3 [] keyctl_session_to_parent+0xed/0x236 [] sys_keyctl+0xb4/0xb6 [] system_call_fastpath+0x16/0x1b The code should take the RCU read lock to make sure the parents credentials don't go away, even though it's holding a spinlock and has IRQ disabled. Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c index b2b0998..3868c67 100644 --- a/security/keys/keyctl.c +++ b/security/keys/keyctl.c @@ -1272,6 +1272,7 @@ long keyctl_session_to_parent(void) keyring_r = NULL; me = current; + rcu_read_lock(); write_lock_irq(&tasklist_lock); parent = me->real_parent; @@ -1319,6 +1320,7 @@ long keyctl_session_to_parent(void) set_ti_thread_flag(task_thread_info(parent), TIF_NOTIFY_RESUME); write_unlock_irq(&tasklist_lock); + rcu_read_unlock(); if (oldcred) put_cred(oldcred); return 0; @@ -1327,6 +1329,7 @@ already_same: ret = 0; not_permitted: write_unlock_irq(&tasklist_lock); + rcu_read_unlock(); put_cred(cred); return ret; -- cgit v0.10.2 From 3d96406c7da1ed5811ea52a3b0905f4f0e295376 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 10 Sep 2010 09:59:51 +0100 Subject: KEYS: Fix bug in keyctl_session_to_parent() if parent has no session keyring Fix a bug in keyctl_session_to_parent() whereby it tries to check the ownership of the parent process's session keyring whether or not the parent has a session keyring [CVE-2010-2960]. This results in the following oops: BUG: unable to handle kernel NULL pointer dereference at 00000000000000a0 IP: [] keyctl_session_to_parent+0x251/0x443 ... Call Trace: [] ? keyctl_session_to_parent+0x67/0x443 [] ? __do_fault+0x24b/0x3d0 [] sys_keyctl+0xb4/0xb8 [] system_call_fastpath+0x16/0x1b if the parent process has no session keyring. If the system is using pam_keyinit then it mostly protected against this as all processes derived from a login will have inherited the session keyring created by pam_keyinit during the log in procedure. To test this, pam_keyinit calls need to be commented out in /etc/pam.d/. Reported-by: Tavis Ormandy Signed-off-by: David Howells Acked-by: Tavis Ormandy Signed-off-by: Linus Torvalds diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c index 3868c67..60924f6 100644 --- a/security/keys/keyctl.c +++ b/security/keys/keyctl.c @@ -1305,7 +1305,8 @@ long keyctl_session_to_parent(void) goto not_permitted; /* the keyrings must have the same UID */ - if (pcred->tgcred->session_keyring->uid != mycred->euid || + if ((pcred->tgcred->session_keyring && + pcred->tgcred->session_keyring->uid != mycred->euid) || mycred->tgcred->session_keyring->uid != mycred->euid) goto not_permitted; -- cgit v0.10.2 From 27c379f7f89a4d558c685b5d89b5ba2fe79ae701 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 10 Sep 2010 13:47:29 +0200 Subject: generic-ipi: Fix deadlock in __smp_call_function_single Just got my 6 way machine to a state where cpu 0 is in an endless loop within __smp_call_function_single. All other cpus are idle. The call trace on cpu 0 looks like this: __smp_call_function_single scheduler_tick update_process_times tick_sched_timer __run_hrtimer hrtimer_interrupt clock_comparator_work do_extint ext_int_handler ----> timer irq cpu_idle __smp_call_function_single() got called from nohz_balancer_kick() (inlined) with the remote cpu being 1, wait being 0 and the per cpu variable remote_sched_softirq_cb (call_single_data) of the current cpu (0). Then it loops forever when it tries to grab the lock of the call_single_data, since it is already locked and enqueued on cpu 0. My theory how this could have happened: for some reason the scheduler decided to call __smp_call_function_single() on it's own cpu, and sends an IPI to itself. The interrupt stays pending since IRQs are disabled. If then the hypervisor schedules the cpu away it might happen that upon rescheduling both the IPI and the timer IRQ are pending. If then interrupts are enabled again it depends which one gets scheduled first. If the timer interrupt gets delivered first we end up with the local deadlock as seen in the calltrace above. Let's make __smp_call_function_single() check if the target cpu is the current cpu and execute the function immediately just like smp_call_function_single does. That should prevent at least the scenario described here. It might also be that the scheduler is not supposed to call __smp_call_function_single with the remote cpu being the current cpu, but that is a different issue. Signed-off-by: Heiko Carstens Acked-by: Peter Zijlstra Acked-by: Jens Axboe Cc: Venkatesh Pallipadi Cc: Suresh Siddha LKML-Reference: <20100910114729.GB2827@osiris.boeblingen.de.ibm.com> Signed-off-by: Ingo Molnar diff --git a/kernel/smp.c b/kernel/smp.c index 75c970c..ed6aacf 100644 --- a/kernel/smp.c +++ b/kernel/smp.c @@ -365,9 +365,10 @@ call: EXPORT_SYMBOL_GPL(smp_call_function_any); /** - * __smp_call_function_single(): Run a function on another CPU + * __smp_call_function_single(): Run a function on a specific CPU * @cpu: The CPU to run on. * @data: Pre-allocated and setup data structure + * @wait: If true, wait until function has completed on specified CPU. * * Like smp_call_function_single(), but allow caller to pass in a * pre-allocated data structure. Useful for embedding @data inside @@ -376,8 +377,10 @@ EXPORT_SYMBOL_GPL(smp_call_function_any); void __smp_call_function_single(int cpu, struct call_single_data *data, int wait) { - csd_lock(data); + unsigned int this_cpu; + unsigned long flags; + this_cpu = get_cpu(); /* * Can deadlock when called with interrupts disabled. * We allow cpu's that are not yet online though, as no one else can @@ -387,7 +390,15 @@ void __smp_call_function_single(int cpu, struct call_single_data *data, WARN_ON_ONCE(cpu_online(smp_processor_id()) && wait && irqs_disabled() && !oops_in_progress); - generic_exec_single(cpu, data, wait); + if (cpu == this_cpu) { + local_irq_save(flags); + data->func(data->info); + local_irq_restore(flags); + } else { + csd_lock(data); + generic_exec_single(cpu, data, wait); + } + put_cpu(); } /** -- cgit v0.10.2 From 1b528181b2ffa14721fb28ad1bd539fe1732c583 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Tue, 7 Sep 2010 19:35:49 -0700 Subject: setup_arg_pages: diagnose excessive argument size The CONFIG_STACK_GROWSDOWN variant of setup_arg_pages() does not check the size of the argument/environment area on the stack. When it is unworkably large, shift_arg_pages() hits its BUG_ON. This is exploitable with a very large RLIMIT_STACK limit, to create a crash pretty easily. Check that the initial stack is not too large to make it possible to map in any executable. We're not checking that the actual executable (or intepreter, for binfmt_elf) will fit. So those mappings might clobber part of the initial stack mapping. But that is just userland lossage that userland made happen, not a kernel problem. Signed-off-by: Roland McGrath Reviewed-by: KOSAKI Motohiro Signed-off-by: Linus Torvalds diff --git a/fs/exec.c b/fs/exec.c index 2d94552..1b63237 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -594,6 +594,11 @@ int setup_arg_pages(struct linux_binprm *bprm, #else stack_top = arch_align_stack(stack_top); stack_top = PAGE_ALIGN(stack_top); + + if (unlikely(stack_top < mmap_min_addr) || + unlikely(vma->vm_end - vma->vm_start >= stack_top - mmap_min_addr)) + return -ENOMEM; + stack_shift = vma->vm_end - stack_top; bprm->p -= stack_shift; -- cgit v0.10.2 From 7993bc1f4663c0db67bb8f0d98e6678145b387cd Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Tue, 7 Sep 2010 19:36:28 -0700 Subject: execve: improve interactivity with large arguments This adds a preemption point during the copying of the argument and environment strings for execve, in copy_strings(). There is already a preemption point in the count() loop, so this doesn't add any new points in the abstract sense. When the total argument+environment strings are very large, the time spent copying them can be much more than a normal user time slice. So this change improves the interactivity of the rest of the system when one process is doing an execve with very large arguments. Signed-off-by: Roland McGrath Reviewed-by: KOSAKI Motohiro Signed-off-by: Linus Torvalds diff --git a/fs/exec.c b/fs/exec.c index 1b63237..6f2d777 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -419,6 +419,8 @@ static int copy_strings(int argc, const char __user *const __user *argv, while (len > 0) { int offset, bytes_to_copy; + cond_resched(); + offset = pos % PAGE_SIZE; if (offset == 0) offset = PAGE_SIZE; -- cgit v0.10.2 From 9aea5a65aa7a1af9a4236dfaeb0088f1624f9919 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Tue, 7 Sep 2010 19:37:06 -0700 Subject: execve: make responsive to SIGKILL with large arguments An execve with a very large total of argument/environment strings can take a really long time in the execve system call. It runs uninterruptibly to count and copy all the strings. This change makes it abort the exec quickly if sent a SIGKILL. Note that this is the conservative change, to interrupt only for SIGKILL, by using fatal_signal_pending(). It would be perfectly correct semantics to let any signal interrupt the string-copying in execve, i.e. use signal_pending() instead of fatal_signal_pending(). We'll save that change for later, since it could have user-visible consequences, such as having a timer set too quickly make it so that an execve can never complete, though it always happened to work before. Signed-off-by: Roland McGrath Reviewed-by: KOSAKI Motohiro Signed-off-by: Linus Torvalds diff --git a/fs/exec.c b/fs/exec.c index 6f2d777..828dd24 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -376,6 +376,9 @@ static int count(const char __user * const __user * argv, int max) argv++; if (i++ >= max) return -E2BIG; + + if (fatal_signal_pending(current)) + return -ERESTARTNOHAND; cond_resched(); } } @@ -419,6 +422,10 @@ static int copy_strings(int argc, const char __user *const __user *argv, while (len > 0) { int offset, bytes_to_copy; + if (fatal_signal_pending(current)) { + ret = -ERESTARTNOHAND; + goto out; + } cond_resched(); offset = pos % PAGE_SIZE; -- cgit v0.10.2 From 36ac4b987bea9a95217e1af552252f275ca7fc44 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Fri, 10 Sep 2010 10:08:08 -0500 Subject: x86, UV: Fix initialization of max_pnode Fix calculation of "max_pnode" for systems where the the highest blade has neither cpus or memory. (And, yes, although rare this does occur). Signed-off-by: Jack Steiner LKML-Reference: <20100910150808.GA19802@sgi.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 7b598b8..f744f54 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -698,9 +698,11 @@ void __init uv_system_init(void) for (j = 0; j < 64; j++) { if (!test_bit(j, &present)) continue; - uv_blade_info[blade].pnode = (i * 64 + j); + pnode = (i * 64 + j); + uv_blade_info[blade].pnode = pnode; uv_blade_info[blade].nr_possible_cpus = 0; uv_blade_info[blade].nr_online_cpus = 0; + max_pnode = max(pnode, max_pnode); blade++; } } @@ -738,7 +740,6 @@ void __init uv_system_init(void) uv_cpu_hub_info(cpu)->scir.offset = uv_scir_offset(apicid); uv_node_to_blade[nid] = blade; uv_cpu_to_blade[cpu] = blade; - max_pnode = max(pnode, max_pnode); } /* Add blade/pnode info for nodes without cpus */ @@ -750,7 +751,6 @@ void __init uv_system_init(void) pnode = (paddr >> m_val) & pnode_mask; blade = boot_pnode_to_blade(pnode); uv_node_to_blade[nid] = blade; - max_pnode = max(pnode, max_pnode); } map_gru_high(max_pnode); -- cgit v0.10.2 From 51749e47e191db8e588ad5cebea731caf7b705d7 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Wed, 8 Sep 2010 09:00:22 +0000 Subject: xfs: log IO completion workqueue is a high priority queue The workqueue implementation in 2.6.36-rcX has changed, resulting in the workqueues no longer having dedicated threads for work processing. This has caused severe livelocks under heavy parallel create workloads because the log IO completions have been getting held up behind metadata IO completions. Hence log commits would stall, memory allocation would stall because pages could not be cleaned, and lock contention on the AIL during inode IO completion processing was being seen to slow everything down even further. By making the log Io completion workqueue a high priority workqueue, they are queued ahead of all data/metadata IO completions and processed before the data/metadata completions. Hence the log never gets stalled, and operations needed to clean memory can continue as quickly as possible. This avoids the livelock conditions and allos the system to keep running under heavy load as per normal. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index d72cf2b..286e36e 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -1932,7 +1932,8 @@ xfs_buf_init(void) if (!xfs_buf_zone) goto out; - xfslogd_workqueue = create_workqueue("xfslogd"); + xfslogd_workqueue = alloc_workqueue("xfslogd", + WQ_RESCUER | WQ_HIGHPRI, 1); if (!xfslogd_workqueue) goto out_free_buf_zone; -- cgit v0.10.2 From 5e64b0d9e86ffff8b299556341d85319117539e9 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 7 Sep 2010 13:30:05 +0800 Subject: ocfs2/lockdep: Move ip_xattr_sem out of ocfs2_xattr_get_nolock. As the name shows, we shouldn't have any lock in ocfs2_xattr_get_nolock. so lift ip_xattr_sem to the caller. This should be safe for us since the only 2 callers are: 1. ocfs2_xattr_get which will lock the resources. 2. ocfs2_mknod which don't need this locking. And this also resolves the following lockdep warning. ======================================================= [ INFO: possible circular locking dependency detected ] 2.6.35+ #5 ------------------------------------------------------- reflink/30027 is trying to acquire lock: (&oi->ip_alloc_sem){+.+.+.}, at: [] ocfs2_reflink_ioctl+0x69a/0x1226 [ocfs2] but task is already holding lock: (&oi->ip_xattr_sem){++++..}, at: [] ocfs2_reflink_ioctl+0x68b/0x1226 [ocfs2] which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&oi->ip_xattr_sem){++++..}: [] __lock_acquire+0x79a/0x7f1 [] lock_acquire+0xc6/0xed [] down_read+0x34/0x47 [] ocfs2_xattr_get_nolock+0xa0/0x4e6 [ocfs2] [] ocfs2_get_acl_nolock+0x5c/0x132 [ocfs2] [] ocfs2_init_acl+0x60/0x243 [ocfs2] [] ocfs2_mknod+0xae8/0xfea [ocfs2] [] ocfs2_create+0x9d/0x105 [ocfs2] [] vfs_create+0x9b/0xf4 [] do_last+0x2fd/0x5be [] do_filp_open+0x1fb/0x572 [] do_sys_open+0x5a/0xe7 [] sys_open+0x1b/0x1d [] system_call_fastpath+0x16/0x1b -> #2 (jbd2_handle){+.+...}: [] __lock_acquire+0x79a/0x7f1 [] lock_acquire+0xc6/0xed [] start_this_handle+0x4a3/0x4bc [jbd2] [] jbd2__journal_start+0xba/0xee [jbd2] [] jbd2_journal_start+0xe/0x10 [jbd2] [] ocfs2_start_trans+0xb7/0x19b [ocfs2] [] ocfs2_mknod+0x73e/0xfea [ocfs2] [] ocfs2_create+0x9d/0x105 [ocfs2] [] vfs_create+0x9b/0xf4 [] do_last+0x2fd/0x5be [] do_filp_open+0x1fb/0x572 [] do_sys_open+0x5a/0xe7 [] sys_open+0x1b/0x1d [] system_call_fastpath+0x16/0x1b -> #1 (&journal->j_trans_barrier){.+.+..}: [] __lock_acquire+0x79a/0x7f1 [] lock_release_non_nested+0x1e5/0x24b [] lock_release+0x158/0x17a [] __mutex_unlock_slowpath+0xbf/0x11b [] mutex_unlock+0x9/0xb [] ocfs2_free_ac_resource+0x31/0x67 [ocfs2] [] ocfs2_free_alloc_context+0x11/0x1d [ocfs2] [] ocfs2_write_begin_nolock+0x141e/0x159b [ocfs2] [] ocfs2_write_begin+0x11e/0x1e7 [ocfs2] [] generic_file_buffered_write+0x10c/0x210 [] ocfs2_file_aio_write+0x4cc/0x6d3 [ocfs2] [] do_sync_write+0xc2/0x106 [] vfs_write+0xae/0x131 [] sys_write+0x47/0x6f [] system_call_fastpath+0x16/0x1b -> #0 (&oi->ip_alloc_sem){+.+.+.}: [] validate_chain+0x727/0xd68 [] __lock_acquire+0x79a/0x7f1 [] lock_acquire+0xc6/0xed [] down_write+0x31/0x52 [] ocfs2_reflink_ioctl+0x69a/0x1226 [ocfs2] [] ocfs2_ioctl+0x61a/0x656 [ocfs2] [] vfs_ioctl+0x2a/0x9d [] do_vfs_ioctl+0x45d/0x4ae [] sys_ioctl+0x57/0x7a [] system_call_fastpath+0x16/0x1b Signed-off-by: Tao Ma Signed-off-by: Joel Becker diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index d03469f..06fa5e7 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -1286,13 +1286,11 @@ int ocfs2_xattr_get_nolock(struct inode *inode, xis.inode_bh = xbs.inode_bh = di_bh; di = (struct ocfs2_dinode *)di_bh->b_data; - down_read(&oi->ip_xattr_sem); ret = ocfs2_xattr_ibody_get(inode, name_index, name, buffer, buffer_size, &xis); if (ret == -ENODATA && di->i_xattr_loc) ret = ocfs2_xattr_block_get(inode, name_index, name, buffer, buffer_size, &xbs); - up_read(&oi->ip_xattr_sem); return ret; } @@ -1316,8 +1314,10 @@ static int ocfs2_xattr_get(struct inode *inode, mlog_errno(ret); return ret; } + down_read(&OCFS2_I(inode)->ip_xattr_sem); ret = ocfs2_xattr_get_nolock(inode, di_bh, name_index, name, buffer, buffer_size); + up_read(&OCFS2_I(inode)->ip_xattr_sem); ocfs2_inode_unlock(inode, 0); -- cgit v0.10.2 From 07eaac9438b13ec0b863111698b91ccec8f3b8d4 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 7 Sep 2010 13:30:06 +0800 Subject: ocfs2: Fix lockdep warning in reflink. This patch change mutex_lock to a new subclass and add a new inode lock subclass for the target inode which caused this lockdep warning. ============================================= [ INFO: possible recursive locking detected ] 2.6.35+ #5 --------------------------------------------- reflink/11086 is trying to acquire lock: (Meta){+++++.}, at: [] ocfs2_reflink_ioctl+0x898/0x1229 [ocfs2] but task is already holding lock: (Meta){+++++.}, at: [] ocfs2_reflink_ioctl+0x5d3/0x1229 [ocfs2] other info that might help us debug this: 6 locks held by reflink/11086: #0: (&sb->s_type->i_mutex_key#15/1){+.+.+.}, at: [] lookup_create+0x26/0x97 #1: (&sb->s_type->i_mutex_key#15){+.+.+.}, at: [] ocfs2_reflink_ioctl+0x4d3/0x1229 [ocfs2] #2: (Meta){+++++.}, at: [] ocfs2_reflink_ioctl+0x5d3/0x1229 [ocfs2] #3: (&oi->ip_xattr_sem){+.+.+.}, at: [] ocfs2_reflink_ioctl+0x68b/0x1229 [ocfs2] #4: (&oi->ip_alloc_sem){+.+.+.}, at: [] ocfs2_reflink_ioctl+0x69a/0x1229 [ocfs2] #5: (&sb->s_type->i_mutex_key#15/2){+.+...}, at: [] ocfs2_reflink_ioctl+0x882/0x1229 [ocfs2] stack backtrace: Pid: 11086, comm: reflink Not tainted 2.6.35+ #5 Call Trace: [] validate_chain+0x56e/0xd68 [] ? mark_held_locks+0x49/0x69 [] __lock_acquire+0x79a/0x7f1 [] lock_acquire+0xc6/0xed [] ? ocfs2_reflink_ioctl+0x898/0x1229 [ocfs2] [] __ocfs2_cluster_lock+0x975/0xa0d [ocfs2] [] ? ocfs2_reflink_ioctl+0x898/0x1229 [ocfs2] [] ? ocfs2_wait_for_recovery+0x15/0x8a [ocfs2] [] ocfs2_inode_lock_full_nested+0x1ac/0xdc5 [ocfs2] [] ? ocfs2_reflink_ioctl+0x898/0x1229 [ocfs2] [] ? trace_hardirqs_on_caller+0x10b/0x12f [] ? debug_mutex_free_waiter+0x4f/0x53 [] ocfs2_reflink_ioctl+0x898/0x1229 [ocfs2] [] ? ocfs2_file_lock_res_init+0x66/0x78 [ocfs2] [] ? might_fault+0x40/0x8d [] ocfs2_ioctl+0x61a/0x656 [ocfs2] [] ? mntput_no_expire+0x1d/0xb0 [] ? path_put+0x2c/0x31 [] vfs_ioctl+0x2a/0x9d [] do_vfs_ioctl+0x45d/0x4ae [] ? _raw_spin_unlock+0x26/0x2a [] ? sysret_check+0x27/0x62 [] sys_ioctl+0x57/0x7a [] system_call_fastpath+0x16/0x1b Signed-off-by: Tao Ma Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h index d1ce48e..1d596d8 100644 --- a/fs/ocfs2/dlmglue.h +++ b/fs/ocfs2/dlmglue.h @@ -84,6 +84,7 @@ enum { OI_LS_PARENT, OI_LS_RENAME1, OI_LS_RENAME2, + OI_LS_REFLINK_TARGET, }; int ocfs2_dlm_init(struct ocfs2_super *osb); diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index 0afeda831..efdd756 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -4201,8 +4201,9 @@ static int __ocfs2_reflink(struct dentry *old_dentry, goto out; } - mutex_lock(&new_inode->i_mutex); - ret = ocfs2_inode_lock(new_inode, &new_bh, 1); + mutex_lock_nested(&new_inode->i_mutex, I_MUTEX_CHILD); + ret = ocfs2_inode_lock_nested(new_inode, &new_bh, 1, + OI_LS_REFLINK_TARGET); if (ret) { mlog_errno(ret); goto out_unlock; -- cgit v0.10.2 From 0f4da216b8c3c35c90ecd18e1899c6f125957c2b Mon Sep 17 00:00:00 2001 From: Tristan Ye Date: Wed, 8 Sep 2010 17:12:38 +0800 Subject: Ocfs2: Re-access the journal after ocfs2_insert_extent() in dxdir codes. In ocfs2_dx_dir_rebalance(), we need to rejournal_acess the blocks after calling ocfs2_insert_extent() since growing an extent tree may trigger ocfs2_extend_trans(), which makes previous journal_access meaningless. Signed-off-by: Tristan Ye Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index f04ebcf..c49f6de 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -3931,6 +3931,15 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir, goto out_commit; } + cpos = split_hash; + ret = ocfs2_dx_dir_new_cluster(dir, &et, cpos, handle, + data_ac, meta_ac, new_dx_leaves, + num_dx_leaves); + if (ret) { + mlog_errno(ret); + goto out_commit; + } + for (i = 0; i < num_dx_leaves; i++) { ret = ocfs2_journal_access_dl(handle, INODE_CACHE(dir), orig_dx_leaves[i], @@ -3939,15 +3948,14 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir, mlog_errno(ret); goto out_commit; } - } - cpos = split_hash; - ret = ocfs2_dx_dir_new_cluster(dir, &et, cpos, handle, - data_ac, meta_ac, new_dx_leaves, - num_dx_leaves); - if (ret) { - mlog_errno(ret); - goto out_commit; + ret = ocfs2_journal_access_dl(handle, INODE_CACHE(dir), + new_dx_leaves[i], + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret) { + mlog_errno(ret); + goto out_commit; + } } ocfs2_dx_dir_transfer_leaf(dir, split_hash, handle, tmp_dx_leaf, -- cgit v0.10.2 From 228ac6357718df2d5c8d70210fa51b2225aab5ee Mon Sep 17 00:00:00 2001 From: Tristan Ye Date: Fri, 10 Sep 2010 10:16:33 +0800 Subject: Ocfs2: Handle empty list in lockres_seq_start() for dlmdebug.c This patch tries to handle the case in which list 'dlm->tracking_list' is empty, to avoid accessing an invalid pointer. It fixes the following oops: http://oss.oracle.com/bugzilla/show_bug.cgi?id=1287 Signed-off-by: Tristan Ye Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index 5efdd37..901ca52 100644 --- a/fs/ocfs2/dlm/dlmdebug.c +++ b/fs/ocfs2/dlm/dlmdebug.c @@ -636,8 +636,14 @@ static void *lockres_seq_start(struct seq_file *m, loff_t *pos) spin_lock(&dlm->track_lock); if (oldres) track_list = &oldres->tracking; - else + else { track_list = &dlm->tracking_list; + if (list_empty(track_list)) { + dl = NULL; + spin_unlock(&dlm->track_lock); + goto bail; + } + } list_for_each_entry(res, track_list, tracking) { if (&res->tracking == &dlm->tracking_list) @@ -660,6 +666,7 @@ static void *lockres_seq_start(struct seq_file *m, loff_t *pos) } else dl = NULL; +bail: /* passed to seq_show */ return dl; } -- cgit v0.10.2 From 0109c2c48d062a04685638926a35ed20153fedc8 Mon Sep 17 00:00:00 2001 From: mark gross Date: Thu, 9 Sep 2010 23:20:09 +0200 Subject: PM QoS: Correct pr_debug() misuse and improve parameter checks Correct some pr_debug() misuse and add a stronger parameter check to pm_qos_write() for the ASCII hex value case. Thanks to Dan Carpenter for pointing out the problem! Signed-off-by: mark gross Signed-off-by: Rafael J. Wysocki diff --git a/kernel/pm_qos_params.c b/kernel/pm_qos_params.c index b7e4c36..645e541 100644 --- a/kernel/pm_qos_params.c +++ b/kernel/pm_qos_params.c @@ -389,10 +389,12 @@ static ssize_t pm_qos_power_write(struct file *filp, const char __user *buf, } else if (count == 11) { /* len('0x12345678/0') */ if (copy_from_user(ascii_value, buf, 11)) return -EFAULT; + if (strlen(ascii_value) != 10) + return -EINVAL; x = sscanf(ascii_value, "%x", &value); if (x != 1) return -EINVAL; - pr_debug(KERN_ERR "%s, %d, 0x%x\n", ascii_value, x, value); + pr_debug("%s, %d, 0x%x\n", ascii_value, x, value); } else return -EINVAL; -- cgit v0.10.2 From 5ee5e97ee9bca919af11c562beeaf61741ad33f1 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 10 Sep 2010 22:32:53 +0200 Subject: x86, tsc: Fix a preemption leak in restore_sched_clock_state() A real life genuine preemption leak.. Reported-and-tested-by: Jeff Chua Signed-off-by: Peter Zijlstra Acked-by: Suresh Siddha Signed-off-by: Linus Torvalds diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index d632934..26a863a 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -655,7 +655,7 @@ void restore_sched_clock_state(void) local_irq_save(flags); - get_cpu_var(cyc2ns_offset) = 0; + __get_cpu_var(cyc2ns_offset) = 0; offset = cyc2ns_suspend - sched_clock(); for_each_possible_cpu(cpu) -- cgit v0.10.2 From 55496c896b8a695140045099d4e0175cf09d4eae Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 10 Sep 2010 22:32:53 +0200 Subject: x86, tsc: Fix a preemption leak in restore_sched_clock_state() Doh, a real life genuine preemption leak.. This caused a suspend failure. Reported-bisected-and-tested-by-the-invaluable: Jeff Chua Acked-by: Suresh Siddha Signed-off-by: Peter Zijlstra Cc: Rafael J. Wysocki Cc: Nico Schottelius Cc: Jesse Barnes Cc: Linus Torvalds Cc: Florian Pritz Cc: Suresh Siddha Cc: Len Brown Cc: # Greg, please apply after: cd7240c ("x86, tsc, sched: Recompute cyc2ns_offset's during resume from") sleep states LKML-Reference: <1284150773.402.122.camel@laptop> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index d632934..26a863a 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -655,7 +655,7 @@ void restore_sched_clock_state(void) local_irq_save(flags); - get_cpu_var(cyc2ns_offset) = 0; + __get_cpu_var(cyc2ns_offset) = 0; offset = cyc2ns_suspend - sched_clock(); for_each_possible_cpu(cpu) -- cgit v0.10.2 From ca04d9c3ec721e474f00992efc1b1afb625507f5 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 26 Aug 2010 16:12:01 -0700 Subject: ceph: fix null pointer deref on anon root dentry release When we release a root dentry, particularly after a splice, the parent (actually our) inode was evaluating to NULL and was getting dereferenced by ceph_snap(). This is reproduced by something as simple as mount -t ceph monhost:/a/b mnt mount -t ceph monhost:/a mnt2 ls mnt2 A splice_dentry() would kill the old 'b' inode's root dentry, and we'd crash while releasing it. Fix by checking for both the ROOT and NULL cases explicitly. We only need to invalidate the parent dir when we have a correct parent to invalidate. Signed-off-by: Sage Weil diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 6e4f43f..a1986eb 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -1021,11 +1021,15 @@ out_touch: static void ceph_dentry_release(struct dentry *dentry) { struct ceph_dentry_info *di = ceph_dentry(dentry); - struct inode *parent_inode = dentry->d_parent->d_inode; - u64 snapid = ceph_snap(parent_inode); + struct inode *parent_inode = NULL; + u64 snapid = CEPH_NOSNAP; + if (!IS_ROOT(dentry)) { + parent_inode = dentry->d_parent->d_inode; + if (parent_inode) + snapid = ceph_snap(parent_inode); + } dout("dentry_release %p parent %p\n", dentry, parent_inode); - if (parent_inode && snapid != CEPH_SNAPDIR) { struct ceph_inode_info *ci = ceph_inode(parent_inode); -- cgit v0.10.2 From 3d4401d9d0aef5c40706350685ddea3df6708496 Mon Sep 17 00:00:00 2001 From: Yehuda Sadeh Date: Fri, 3 Sep 2010 12:57:11 -0700 Subject: ceph: fix pagelist kunmap tail A wrong parameter was passed to the kunmap. Signed-off-by: Yehuda Sadeh Signed-off-by: Sage Weil diff --git a/fs/ceph/pagelist.c b/fs/ceph/pagelist.c index b6859f4..46a368b 100644 --- a/fs/ceph/pagelist.c +++ b/fs/ceph/pagelist.c @@ -5,10 +5,18 @@ #include "pagelist.h" +static void ceph_pagelist_unmap_tail(struct ceph_pagelist *pl) +{ + struct page *page = list_entry(pl->head.prev, struct page, + lru); + kunmap(page); +} + int ceph_pagelist_release(struct ceph_pagelist *pl) { if (pl->mapped_tail) - kunmap(pl->mapped_tail); + ceph_pagelist_unmap_tail(pl); + while (!list_empty(&pl->head)) { struct page *page = list_first_entry(&pl->head, struct page, lru); @@ -26,7 +34,7 @@ static int ceph_pagelist_addpage(struct ceph_pagelist *pl) pl->room += PAGE_SIZE; list_add_tail(&page->lru, &pl->head); if (pl->mapped_tail) - kunmap(pl->mapped_tail); + ceph_pagelist_unmap_tail(pl); pl->mapped_tail = kmap(page); return 0; } -- cgit v0.10.2 From 3612abbd5df6baa9ca3e0777c6c8646e202d3f66 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 7 Sep 2010 15:59:27 -0700 Subject: ceph: fix reconnect encoding for old servers Fix the reconnect encoding to encode the cap record when the MDS does not have the FLOCK capability (i.e., pre v0.22). Signed-off-by: Sage Weil diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index f091b13..fad95f8 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -2374,6 +2374,8 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap, num_fcntl_locks, num_flock_locks); unlock_kernel(); + } else { + err = ceph_pagelist_append(pagelist, &rec, reclen); } out_free: -- cgit v0.10.2 From a77d9f7dce7600058d56f0670ed29d77abffcde2 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sat, 11 Sep 2010 10:55:25 -0700 Subject: ceph: fix file offset wrapping at 4GB on 32-bit archs Cast the value before shifting so that we don't run out of bits with a 32-bit unsigned long. This fixes wrapping of high file offsets into the low 4GB of a file on disk, and the subsequent data corruption for large files. Signed-off-by: Sage Weil diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 4cfce1e..50461b8 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -766,7 +766,8 @@ get_more_pages: /* ok */ if (locked_pages == 0) { /* prepare async write request */ - offset = page->index << PAGE_CACHE_SHIFT; + offset = (unsigned long long)page->index + << PAGE_CACHE_SHIFT; len = wsize; req = ceph_osdc_new_request(&client->osdc, &ci->i_layout, -- cgit v0.10.2 From 6715045ddc7472a22be5e49d4047d2d89b391f45 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sat, 11 Sep 2010 20:58:27 +0200 Subject: PM / Hibernate: Avoid hitting OOM during preallocation of memory There is a problem in hibernate_preallocate_memory() that it calls preallocate_image_memory() with an argument that may be greater than the total number of available non-highmem memory pages. If that's the case, the OOM condition is guaranteed to trigger, which in turn can cause significant slowdown to occur during hibernation. To avoid that, make preallocate_image_memory() adjust its argument before calling preallocate_image_pages(), so that the total number of saveable non-highem pages left is not less than the minimum size of a hibernation image. Change hibernate_preallocate_memory() to try to allocate from highmem if the number of pages allocated by preallocate_image_memory() is too low. Modify free_unnecessary_pages() to take all possible memory allocation patterns into account. Reported-by: KOSAKI Motohiro Signed-off-by: Rafael J. Wysocki Tested-by: M. Vefa Bicakci diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index 5e7edfb..5209b39 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -1122,9 +1122,19 @@ static unsigned long preallocate_image_pages(unsigned long nr_pages, gfp_t mask) return nr_alloc; } -static unsigned long preallocate_image_memory(unsigned long nr_pages) +static unsigned long preallocate_image_memory(unsigned long nr_pages, + unsigned long avail_normal) { - return preallocate_image_pages(nr_pages, GFP_IMAGE); + unsigned long alloc; + + if (avail_normal <= alloc_normal) + return 0; + + alloc = avail_normal - alloc_normal; + if (nr_pages < alloc) + alloc = nr_pages; + + return preallocate_image_pages(alloc, GFP_IMAGE); } #ifdef CONFIG_HIGHMEM @@ -1170,15 +1180,22 @@ static inline unsigned long preallocate_highmem_fraction(unsigned long nr_pages, */ static void free_unnecessary_pages(void) { - unsigned long save_highmem, to_free_normal, to_free_highmem; + unsigned long save, to_free_normal, to_free_highmem; - to_free_normal = alloc_normal - count_data_pages(); - save_highmem = count_highmem_pages(); - if (alloc_highmem > save_highmem) { - to_free_highmem = alloc_highmem - save_highmem; + save = count_data_pages(); + if (alloc_normal >= save) { + to_free_normal = alloc_normal - save; + save = 0; + } else { + to_free_normal = 0; + save -= alloc_normal; + } + save += count_highmem_pages(); + if (alloc_highmem >= save) { + to_free_highmem = alloc_highmem - save; } else { to_free_highmem = 0; - to_free_normal -= save_highmem - alloc_highmem; + to_free_normal -= save - alloc_highmem; } memory_bm_position_reset(©_bm); @@ -1259,7 +1276,7 @@ int hibernate_preallocate_memory(void) { struct zone *zone; unsigned long saveable, size, max_size, count, highmem, pages = 0; - unsigned long alloc, save_highmem, pages_highmem; + unsigned long alloc, save_highmem, pages_highmem, avail_normal; struct timeval start, stop; int error; @@ -1296,6 +1313,7 @@ int hibernate_preallocate_memory(void) else count += zone_page_state(zone, NR_FREE_PAGES); } + avail_normal = count; count += highmem; count -= totalreserve_pages; @@ -1310,12 +1328,21 @@ int hibernate_preallocate_memory(void) */ if (size >= saveable) { pages = preallocate_image_highmem(save_highmem); - pages += preallocate_image_memory(saveable - pages); + pages += preallocate_image_memory(saveable - pages, avail_normal); goto out; } /* Estimate the minimum size of the image. */ pages = minimum_image_size(saveable); + /* + * To avoid excessive pressure on the normal zone, leave room in it to + * accommodate an image of the minimum size (unless it's already too + * small, in which case don't preallocate pages from it at all). + */ + if (avail_normal > pages) + avail_normal -= pages; + else + avail_normal = 0; if (size < pages) size = min_t(unsigned long, pages, max_size); @@ -1336,16 +1363,34 @@ int hibernate_preallocate_memory(void) */ pages_highmem = preallocate_image_highmem(highmem / 2); alloc = (count - max_size) - pages_highmem; - pages = preallocate_image_memory(alloc); - if (pages < alloc) - goto err_out; - size = max_size - size; - alloc = size; - size = preallocate_highmem_fraction(size, highmem, count); - pages_highmem += size; - alloc -= size; - pages += preallocate_image_memory(alloc); - pages += pages_highmem; + pages = preallocate_image_memory(alloc, avail_normal); + if (pages < alloc) { + /* We have exhausted non-highmem pages, try highmem. */ + alloc -= pages; + pages += pages_highmem; + pages_highmem = preallocate_image_highmem(alloc); + if (pages_highmem < alloc) + goto err_out; + pages += pages_highmem; + /* + * size is the desired number of saveable pages to leave in + * memory, so try to preallocate (all memory - size) pages. + */ + alloc = (count - pages) - size; + pages += preallocate_image_highmem(alloc); + } else { + /* + * There are approximately max_size saveable pages at this point + * and we want to reduce this number down to size. + */ + alloc = max_size - size; + size = preallocate_highmem_fraction(alloc, highmem, count); + pages_highmem += size; + alloc -= size; + size = preallocate_image_memory(alloc, avail_normal); + pages_highmem += preallocate_image_highmem(alloc - size); + pages += pages_highmem + size; + } /* * We only need as many page frames for the image as there are saveable -- cgit v0.10.2 From 1f3a66889c4c80c821f3eadf899c140e91452f8e Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sat, 11 Sep 2010 15:55:12 -0700 Subject: kernel-doc: ignore case when stripping attributes There are valid attributes that could have upper case letters, but we still want to remove, like for example __attribute__((aligned(NETDEV_ALIGN))) as encountered in the wireless code. Signed-off-by: Johannes Berg Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/scripts/kernel-doc b/scripts/kernel-doc index 102e123..a68240c 100755 --- a/scripts/kernel-doc +++ b/scripts/kernel-doc @@ -1679,7 +1679,7 @@ sub check_sections($$$$$$) { foreach $px (0 .. $#prms) { $prm_clean = $prms[$px]; $prm_clean =~ s/\[.*\]//; - $prm_clean =~ s/__attribute__\s*\(\([a-z,_\*\s\(\)]*\)\)//; + $prm_clean =~ s/__attribute__\s*\(\([a-z,_\*\s\(\)]*\)\)//i; # ignore array size in a parameter string; # however, the original param string may contain # spaces, e.g.: addr[6 + 2] -- cgit v0.10.2 From eda603f6cdba4b14dbf80531fab2fe545232e7a0 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sat, 11 Sep 2010 15:55:22 -0700 Subject: docbook: warn on unused doc entries When you don't use !E or !I but only !F, then it's very easy to miss including some functions, structs etc. in documentation. To help finding which ones were missed, allow printing out the unused ones as warnings. For example, using this on mac80211 yields a lot of warnings like this: Warning: didn't use docs for DOC: mac80211 workqueue Warning: didn't use docs for ieee80211_max_queues Warning: didn't use docs for ieee80211_bss_change Warning: didn't use docs for ieee80211_bss_conf when generating the documentation for it. Signed-off-by: Johannes Berg Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/Documentation/kernel-doc-nano-HOWTO.txt b/Documentation/kernel-doc-nano-HOWTO.txt index 27a52b3..3d8a977 100644 --- a/Documentation/kernel-doc-nano-HOWTO.txt +++ b/Documentation/kernel-doc-nano-HOWTO.txt @@ -345,5 +345,10 @@ documentation, in , for the functions listed. section titled
from . Spaces are allowed in
; do not quote the
. +!C is replaced by nothing, but makes the tools check that +all DOC: sections and documented functions, symbols, etc. are used. +This makes sense to use when you use !F/!P only and want to verify +that all documentation is included. + Tim. */ diff --git a/scripts/basic/docproc.c b/scripts/basic/docproc.c index 79ab973..fc3b18d 100644 --- a/scripts/basic/docproc.c +++ b/scripts/basic/docproc.c @@ -34,12 +34,14 @@ * */ +#define _GNU_SOURCE #include #include #include #include #include #include +#include #include #include @@ -54,6 +56,7 @@ typedef void FILEONLY(char * file); FILEONLY *internalfunctions; FILEONLY *externalfunctions; FILEONLY *symbolsonly; +FILEONLY *findall; typedef void FILELINE(char * file, char * line); FILELINE * singlefunctions; @@ -65,12 +68,30 @@ FILELINE * docsection; #define KERNELDOCPATH "scripts/" #define KERNELDOC "kernel-doc" #define DOCBOOK "-docbook" +#define LIST "-list" #define FUNCTION "-function" #define NOFUNCTION "-nofunction" #define NODOCSECTIONS "-no-doc-sections" static char *srctree, *kernsrctree; +static char **all_list = NULL; +static int all_list_len = 0; + +static void consume_symbol(const char *sym) +{ + int i; + + for (i = 0; i < all_list_len; i++) { + if (!all_list[i]) + continue; + if (strcmp(sym, all_list[i])) + continue; + all_list[i] = NULL; + break; + } +} + static void usage (void) { fprintf(stderr, "Usage: docproc {doc|depend} file\n"); @@ -248,6 +269,7 @@ static void docfunctions(char * filename, char * type) struct symfile * sym = &symfilelist[i]; for (j=0; j < sym->symbolcnt; j++) { vec[idx++] = type; + consume_symbol(sym->symbollist[j].name); vec[idx++] = sym->symbollist[j].name; } } @@ -287,6 +309,11 @@ static void singfunc(char * filename, char * line) vec[idx++] = &line[i]; } } + for (i = 0; i < idx; i++) { + if (strcmp(vec[i], FUNCTION)) + continue; + consume_symbol(vec[i + 1]); + } vec[idx++] = filename; vec[idx] = NULL; exec_kernel_doc(vec); @@ -306,6 +333,10 @@ static void docsect(char *filename, char *line) if (*s == '\n') *s = '\0'; + asprintf(&s, "DOC: %s", line); + consume_symbol(s); + free(s); + vec[0] = KERNELDOC; vec[1] = DOCBOOK; vec[2] = FUNCTION; @@ -315,6 +346,84 @@ static void docsect(char *filename, char *line) exec_kernel_doc(vec); } +static void find_all_symbols(char *filename) +{ + char *vec[4]; /* kerneldoc -list file NULL */ + pid_t pid; + int ret, i, count, start; + char real_filename[PATH_MAX + 1]; + int pipefd[2]; + char *data, *str; + size_t data_len = 0; + + vec[0] = KERNELDOC; + vec[1] = LIST; + vec[2] = filename; + vec[3] = NULL; + + if (pipe(pipefd)) { + perror("pipe"); + exit(1); + } + + switch (pid=fork()) { + case -1: + perror("fork"); + exit(1); + case 0: + close(pipefd[0]); + dup2(pipefd[1], 1); + memset(real_filename, 0, sizeof(real_filename)); + strncat(real_filename, kernsrctree, PATH_MAX); + strncat(real_filename, "/" KERNELDOCPATH KERNELDOC, + PATH_MAX - strlen(real_filename)); + execvp(real_filename, vec); + fprintf(stderr, "exec "); + perror(real_filename); + exit(1); + default: + close(pipefd[1]); + data = malloc(4096); + do { + while ((ret = read(pipefd[0], + data + data_len, + 4096)) > 0) { + data_len += ret; + data = realloc(data, data_len + 4096); + } + } while (ret == -EAGAIN); + if (ret != 0) { + perror("read"); + exit(1); + } + waitpid(pid, &ret ,0); + } + if (WIFEXITED(ret)) + exitstatus |= WEXITSTATUS(ret); + else + exitstatus = 0xff; + + count = 0; + /* poor man's strtok, but with counting */ + for (i = 0; i < data_len; i++) { + if (data[i] == '\n') { + count++; + data[i] = '\0'; + } + } + start = all_list_len; + all_list_len += count; + all_list = realloc(all_list, sizeof(char *) * all_list_len); + str = data; + for (i = 0; i < data_len && start != all_list_len; i++) { + if (data[i] == '\0') { + all_list[start] = str; + str = data + i + 1; + start++; + } + } +} + /* * Parse file, calling action specific functions for: * 1) Lines containing !E @@ -322,7 +431,8 @@ static void docsect(char *filename, char *line) * 3) Lines containing !D * 4) Lines containing !F * 5) Lines containing !P - * 6) Default lines - lines not matching the above + * 6) Lines containing !C + * 7) Default lines - lines not matching the above */ static void parse_file(FILE *infile) { @@ -365,6 +475,12 @@ static void parse_file(FILE *infile) s++; docsection(line + 2, s); break; + case 'C': + while (*s && !isspace(*s)) s++; + *s = '\0'; + if (findall) + findall(line+2); + break; default: defaultline(line); } @@ -380,6 +496,7 @@ static void parse_file(FILE *infile) int main(int argc, char *argv[]) { FILE * infile; + int i; srctree = getenv("SRCTREE"); if (!srctree) @@ -415,6 +532,7 @@ int main(int argc, char *argv[]) symbolsonly = find_export_symbols; singlefunctions = noaction2; docsection = noaction2; + findall = find_all_symbols; parse_file(infile); /* Rewind to start from beginning of file again */ @@ -425,8 +543,16 @@ int main(int argc, char *argv[]) symbolsonly = printline; singlefunctions = singfunc; docsection = docsect; + findall = NULL; parse_file(infile); + + for (i = 0; i < all_list_len; i++) { + if (!all_list[i]) + continue; + fprintf(stderr, "Warning: didn't use docs for %s\n", + all_list[i]); + } } else if (strcmp("depend", argv[1]) == 0) { @@ -439,6 +565,7 @@ int main(int argc, char *argv[]) symbolsonly = adddep; singlefunctions = adddep2; docsection = adddep2; + findall = adddep; parse_file(infile); printf("\n"); } diff --git a/scripts/kernel-doc b/scripts/kernel-doc index a68240c..cdb6dc1 100755 --- a/scripts/kernel-doc +++ b/scripts/kernel-doc @@ -44,12 +44,13 @@ use strict; # Note: This only supports 'c'. # usage: -# kernel-doc [ -docbook | -html | -text | -man ] [ -no-doc-sections ] +# kernel-doc [ -docbook | -html | -text | -man | -list ] [ -no-doc-sections ] # [ -function funcname [ -function funcname ...] ] c file(s)s > outputfile # or # [ -nofunction funcname [ -function funcname ...] ] c file(s)s > outputfile # # Set output format using one of -docbook -html -text or -man. Default is man. +# The -list format is for internal use by docproc. # # -no-doc-sections # Do not output DOC: sections @@ -210,9 +211,16 @@ my %highlights_text = ( $type_constant, "\$1", $type_param, "\$1" ); my $blankline_text = ""; +# list mode +my %highlights_list = ( $type_constant, "\$1", + $type_func, "\$1", + $type_struct, "\$1", + $type_param, "\$1" ); +my $blankline_list = ""; sub usage { - print "Usage: $0 [ -v ] [ -docbook | -html | -text | -man ] [ -no-doc-sections ]\n"; + print "Usage: $0 [ -v ] [ -docbook | -html | -text | -man | -list ]\n"; + print " [ -no-doc-sections ]\n"; print " [ -function funcname [ -function funcname ...] ]\n"; print " [ -nofunction funcname [ -nofunction funcname ...] ]\n"; print " c source file(s) > outputfile\n"; @@ -318,6 +326,10 @@ while ($ARGV[0] =~ m/^-(.*)/) { $output_mode = "xml"; %highlights = %highlights_xml; $blankline = $blankline_xml; + } elsif ($cmd eq "-list") { + $output_mode = "list"; + %highlights = %highlights_list; + $blankline = $blankline_list; } elsif ($cmd eq "-gnome") { $output_mode = "gnome"; %highlights = %highlights_gnome; @@ -1361,6 +1373,42 @@ sub output_blockhead_text(%) { } } +## list mode output functions + +sub output_function_list(%) { + my %args = %{$_[0]}; + + print $args{'function'} . "\n"; +} + +# output enum in list +sub output_enum_list(%) { + my %args = %{$_[0]}; + print $args{'enum'} . "\n"; +} + +# output typedef in list +sub output_typedef_list(%) { + my %args = %{$_[0]}; + print $args{'typedef'} . "\n"; +} + +# output struct as list +sub output_struct_list(%) { + my %args = %{$_[0]}; + + print $args{'struct'} . "\n"; +} + +sub output_blockhead_list(%) { + my %args = %{$_[0]}; + my ($parameter, $section); + + foreach $section (@{$args{'sectionlist'}}) { + print "DOC: $section\n"; + } +} + ## # generic output function for all types (function, struct/union, typedef, enum); # calls the generated, variable output_ function name based on -- cgit v0.10.2 From 39d709392ffc1d05f825191b07b7d865d434f282 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 11 Sep 2010 15:55:26 -0700 Subject: docbook: skip files with no docs since they generate scary warnings Fix docbook templates that reference files that do not contain the expected kernel-doc notation. Fixes these warnings: Warning(arch/x86/include/asm/unaligned.h): no structured comments found Warning(lib/vsprintf.c): no structured comments found These cause errors in the generated html output, like below, so drop these lines. Name arch/x86/include/asm/unaligned.h - Document generation inconsistency Oops Warning The template for this document tried to insert the structured comment from the file arch/x86/include/asm/unaligned.h at this point, but none was found. This dummy section is inserted to allow generation to continue. Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/Documentation/DocBook/device-drivers.tmpl b/Documentation/DocBook/device-drivers.tmpl index ecd35e9..feca075 100644 --- a/Documentation/DocBook/device-drivers.tmpl +++ b/Documentation/DocBook/device-drivers.tmpl @@ -46,7 +46,6 @@ Atomic and pointer manipulation !Iarch/x86/include/asm/atomic.h -!Iarch/x86/include/asm/unaligned.h Delaying, scheduling, and timer routines diff --git a/Documentation/DocBook/kernel-api.tmpl b/Documentation/DocBook/kernel-api.tmpl index a20c6f6..6899f47 100644 --- a/Documentation/DocBook/kernel-api.tmpl +++ b/Documentation/DocBook/kernel-api.tmpl @@ -57,7 +57,6 @@ String Conversions -!Ilib/vsprintf.c !Elib/vsprintf.c String Manipulation -- cgit v0.10.2 From 897493504addc5609f04a2c4f73c37ab972c29b2 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Sun, 12 Sep 2010 18:25:19 +0100 Subject: drm/i915: Ensure that the crtcinfo is populated during mode_fixup() This should fix the mysterious mode setting failures reported during boot up and after resume, generally for i8xx class machines. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=16478 Reported-and-tested-by: Xavier Chantry Buzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29413 Tested-by: Daniel Vetter Signed-off-by: Chris Wilson Cc: stable@kernel.org diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 19daead..b5bf51a 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2463,11 +2463,19 @@ static bool intel_crtc_mode_fixup(struct drm_crtc *crtc, struct drm_display_mode *adjusted_mode) { struct drm_device *dev = crtc->dev; + if (HAS_PCH_SPLIT(dev)) { /* FDI link clock is fixed at 2.7G */ if (mode->clock * 3 > IRONLAKE_FDI_FREQ * 4) return false; } + + /* XXX some encoders set the crtcinfo, others don't. + * Obviously we need some form of conflict resolution here... + */ + if (adjusted_mode->crtc_htotal == 0) + drm_mode_set_crtcinfo(adjusted_mode, 0); + return true; } -- cgit v0.10.2 From 49553c2ef88749dd502687f4eb9c258bb10a4f44 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 12 Sep 2010 16:07:37 -0700 Subject: Linux 2.6.36-rc4 diff --git a/Makefile b/Makefile index 4df9873..92ab33f 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 2 PATCHLEVEL = 6 SUBLEVEL = 36 -EXTRAVERSION = -rc3 +EXTRAVERSION = -rc4 NAME = Sheep on Meth # *DOCUMENTATION* -- cgit v0.10.2 From f2d47d02fd84343a3c5452daca6ed12c75618aff Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 12 Sep 2010 19:55:25 -0400 Subject: Fix null dereference in call_allocate In call_allocate we need to reach the auth in order to factor au_cslack into the allocation. As of a17c2153d2e271b0cbacae9bed83b0eaa41db7e1 "SUNRPC: Move the bound cred to struct rpc_rqst", call_allocate attempts to do this by dereferencing tk_client->cl_auth, however this is not guaranteed to be defined--cl_auth can be zero in the case of gss context destruction (see rpc_free_auth). Reorder the client state machine to bind credentials before allocating, so that we can instead reach the auth through the cred. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust Cc: stable@kernel.org diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 2388d83..657aac6 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -931,7 +931,7 @@ call_reserveresult(struct rpc_task *task) task->tk_status = 0; if (status >= 0) { if (task->tk_rqstp) { - task->tk_action = call_allocate; + task->tk_action = call_refresh; return; } @@ -972,7 +972,7 @@ call_reserveresult(struct rpc_task *task) static void call_allocate(struct rpc_task *task) { - unsigned int slack = task->tk_client->cl_auth->au_cslack; + unsigned int slack = task->tk_rqstp->rq_cred->cr_auth->au_cslack; struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = task->tk_xprt; struct rpc_procinfo *proc = task->tk_msg.rpc_proc; @@ -980,7 +980,7 @@ call_allocate(struct rpc_task *task) dprint_status(task); task->tk_status = 0; - task->tk_action = call_refresh; + task->tk_action = call_bind; if (req->rq_buffer) return; @@ -1042,7 +1042,7 @@ call_refreshresult(struct rpc_task *task) dprint_status(task); task->tk_status = 0; - task->tk_action = call_bind; + task->tk_action = call_allocate; if (status >= 0 && rpcauth_uptodatecred(task)) return; switch (status) { -- cgit v0.10.2 From 5a67657a2e90c9e4a48518f95d4ba7777aa20fbb Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 12 Sep 2010 19:55:25 -0400 Subject: SUNRPC: Fix race corrupting rpc upcall If rpc_queue_upcall() adds a new upcall to the rpci->pipe list just after rpc_pipe_release calls rpc_purge_list(), but before it calls gss_pipe_release (as rpci->ops->release_pipe(inode)), then the latter will free a message without deleting it from the rpci->pipe list. We will be left with a freed object on the rpc->pipe list. Most frequent symptoms are kernel crashes in rpc.gssd system calls on the pipe in question. Reported-by: J. Bruce Fields Signed-off-by: Trond Myklebust Cc: stable@kernel.org diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index dcfc66b..12c4859 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -745,17 +745,18 @@ gss_pipe_release(struct inode *inode) struct rpc_inode *rpci = RPC_I(inode); struct gss_upcall_msg *gss_msg; +restart: spin_lock(&inode->i_lock); - while (!list_empty(&rpci->in_downcall)) { + list_for_each_entry(gss_msg, &rpci->in_downcall, list) { - gss_msg = list_entry(rpci->in_downcall.next, - struct gss_upcall_msg, list); + if (!list_empty(&gss_msg->msg.list)) + continue; gss_msg->msg.errno = -EPIPE; atomic_inc(&gss_msg->count); __gss_unhash_msg(gss_msg); spin_unlock(&inode->i_lock); gss_release_msg(gss_msg); - spin_lock(&inode->i_lock); + goto restart; } spin_unlock(&inode->i_lock); diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 95ccbcf..41a762f 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -48,7 +48,7 @@ static void rpc_purge_list(struct rpc_inode *rpci, struct list_head *head, return; do { msg = list_entry(head->next, struct rpc_pipe_msg, list); - list_del(&msg->list); + list_del_init(&msg->list); msg->errno = err; destroy_msg(msg); } while (!list_empty(head)); @@ -208,7 +208,7 @@ rpc_pipe_release(struct inode *inode, struct file *filp) if (msg != NULL) { spin_lock(&inode->i_lock); msg->errno = -EAGAIN; - list_del(&msg->list); + list_del_init(&msg->list); spin_unlock(&inode->i_lock); rpci->ops->destroy_msg(msg); } @@ -268,7 +268,7 @@ rpc_pipe_read(struct file *filp, char __user *buf, size_t len, loff_t *offset) if (res < 0 || msg->len == msg->copied) { filp->private_data = NULL; spin_lock(&inode->i_lock); - list_del(&msg->list); + list_del_init(&msg->list); spin_unlock(&inode->i_lock); rpci->ops->destroy_msg(msg); } -- cgit v0.10.2 From 006abe887c5e637d059c44310de6c92f36aded3b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 12 Sep 2010 19:55:25 -0400 Subject: SUNRPC: Fix a race in rpc_info_open There is a race between rpc_info_open and rpc_release_client() in that nothing stops a process from opening the file after the clnt->cl_kref goes to zero. Fix this by using atomic_inc_unless_zero()... Reported-by: J. Bruce Fields Signed-off-by: Trond Myklebust Cc: stable@kernel.org diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 569dc72..85f38a63 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -30,7 +30,7 @@ struct rpc_inode; * The high-level client handle */ struct rpc_clnt { - struct kref cl_kref; /* Number of references */ + atomic_t cl_count; /* Number of references */ struct list_head cl_clients; /* Global list of clients */ struct list_head cl_tasks; /* List of tasks */ spinlock_t cl_lock; /* spinlock */ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 657aac6..3a8f53e 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -226,7 +226,7 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru goto out_no_principal; } - kref_init(&clnt->cl_kref); + atomic_set(&clnt->cl_count, 1); err = rpc_setup_pipedir(clnt, program->pipe_dir_name); if (err < 0) @@ -390,14 +390,14 @@ rpc_clone_client(struct rpc_clnt *clnt) if (new->cl_principal == NULL) goto out_no_principal; } - kref_init(&new->cl_kref); + atomic_set(&new->cl_count, 1); err = rpc_setup_pipedir(new, clnt->cl_program->pipe_dir_name); if (err != 0) goto out_no_path; if (new->cl_auth) atomic_inc(&new->cl_auth->au_count); xprt_get(clnt->cl_xprt); - kref_get(&clnt->cl_kref); + atomic_inc(&clnt->cl_count); rpc_register_client(new); rpciod_up(); return new; @@ -465,10 +465,8 @@ EXPORT_SYMBOL_GPL(rpc_shutdown_client); * Free an RPC client */ static void -rpc_free_client(struct kref *kref) +rpc_free_client(struct rpc_clnt *clnt) { - struct rpc_clnt *clnt = container_of(kref, struct rpc_clnt, cl_kref); - dprintk("RPC: destroying %s client for %s\n", clnt->cl_protname, clnt->cl_server); if (!IS_ERR(clnt->cl_path.dentry)) { @@ -495,12 +493,10 @@ out_free: * Free an RPC client */ static void -rpc_free_auth(struct kref *kref) +rpc_free_auth(struct rpc_clnt *clnt) { - struct rpc_clnt *clnt = container_of(kref, struct rpc_clnt, cl_kref); - if (clnt->cl_auth == NULL) { - rpc_free_client(kref); + rpc_free_client(clnt); return; } @@ -509,10 +505,11 @@ rpc_free_auth(struct kref *kref) * release remaining GSS contexts. This mechanism ensures * that it can do so safely. */ - kref_init(kref); + atomic_inc(&clnt->cl_count); rpcauth_release(clnt->cl_auth); clnt->cl_auth = NULL; - kref_put(kref, rpc_free_client); + if (atomic_dec_and_test(&clnt->cl_count)) + rpc_free_client(clnt); } /* @@ -525,7 +522,8 @@ rpc_release_client(struct rpc_clnt *clnt) if (list_empty(&clnt->cl_tasks)) wake_up(&destroy_wait); - kref_put(&clnt->cl_kref, rpc_free_auth); + if (atomic_dec_and_test(&clnt->cl_count)) + rpc_free_auth(clnt); } /** @@ -588,7 +586,7 @@ void rpc_task_set_client(struct rpc_task *task, struct rpc_clnt *clnt) if (clnt != NULL) { rpc_task_release_client(task); task->tk_client = clnt; - kref_get(&clnt->cl_kref); + atomic_inc(&clnt->cl_count); if (clnt->cl_softrtry) task->tk_flags |= RPC_TASK_SOFT; /* Add to the client's list of all tasks */ diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 41a762f..8c8eef2 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -371,21 +371,23 @@ rpc_show_info(struct seq_file *m, void *v) static int rpc_info_open(struct inode *inode, struct file *file) { - struct rpc_clnt *clnt; + struct rpc_clnt *clnt = NULL; int ret = single_open(file, rpc_show_info, NULL); if (!ret) { struct seq_file *m = file->private_data; - mutex_lock(&inode->i_mutex); - clnt = RPC_I(inode)->private; - if (clnt) { - kref_get(&clnt->cl_kref); + + spin_lock(&file->f_path.dentry->d_lock); + if (!d_unhashed(file->f_path.dentry)) + clnt = RPC_I(inode)->private; + if (clnt != NULL && atomic_inc_not_zero(&clnt->cl_count)) { + spin_unlock(&file->f_path.dentry->d_lock); m->private = clnt; } else { + spin_unlock(&file->f_path.dentry->d_lock); single_release(inode, file); ret = -EINVAL; } - mutex_unlock(&inode->i_mutex); } return ret; } -- cgit v0.10.2 From 55576244eba805307a2b2b6a145b8f85f8c7c124 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 12 Sep 2010 19:55:25 -0400 Subject: SUNRPC: cleanup state-machine ordering This is just a minor cleanup: net/sunrpc/clnt.c clarifies the rpc client state machine by commenting each state and by laying out the functions implementing each state in the order that each state is normally executed (in the absence of errors). The previous patch "Fix null dereference in call_allocate" changed the order of the states. Move the functions and update the comments to reflect the change. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 3a8f53e..fa55490 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -964,7 +964,48 @@ call_reserveresult(struct rpc_task *task) } /* - * 2. Allocate the buffer. For details, see sched.c:rpc_malloc. + * 2. Bind and/or refresh the credentials + */ +static void +call_refresh(struct rpc_task *task) +{ + dprint_status(task); + + task->tk_action = call_refreshresult; + task->tk_status = 0; + task->tk_client->cl_stats->rpcauthrefresh++; + rpcauth_refreshcred(task); +} + +/* + * 2a. Process the results of a credential refresh + */ +static void +call_refreshresult(struct rpc_task *task) +{ + int status = task->tk_status; + + dprint_status(task); + + task->tk_status = 0; + task->tk_action = call_allocate; + if (status >= 0 && rpcauth_uptodatecred(task)) + return; + switch (status) { + case -EACCES: + rpc_exit(task, -EACCES); + return; + case -ENOMEM: + rpc_exit(task, -ENOMEM); + return; + case -ETIMEDOUT: + rpc_delay(task, 3*HZ); + } + task->tk_action = call_refresh; +} + +/* + * 2b. Allocate the buffer. For details, see sched.c:rpc_malloc. * (Note: buffer memory is freed in xprt_release). */ static void @@ -1015,47 +1056,6 @@ call_allocate(struct rpc_task *task) rpc_exit(task, -ERESTARTSYS); } -/* - * 2a. Bind and/or refresh the credentials - */ -static void -call_refresh(struct rpc_task *task) -{ - dprint_status(task); - - task->tk_action = call_refreshresult; - task->tk_status = 0; - task->tk_client->cl_stats->rpcauthrefresh++; - rpcauth_refreshcred(task); -} - -/* - * 2b. Process the results of a credential refresh - */ -static void -call_refreshresult(struct rpc_task *task) -{ - int status = task->tk_status; - - dprint_status(task); - - task->tk_status = 0; - task->tk_action = call_allocate; - if (status >= 0 && rpcauth_uptodatecred(task)) - return; - switch (status) { - case -EACCES: - rpc_exit(task, -EACCES); - return; - case -ENOMEM: - rpc_exit(task, -ENOMEM); - return; - case -ETIMEDOUT: - rpc_delay(task, 3*HZ); - } - task->tk_action = call_refresh; -} - static inline int rpc_task_need_encode(struct rpc_task *task) { -- cgit v0.10.2 From b1bde04c6d9a120dec602cc8a70b8a7f21600883 Mon Sep 17 00:00:00 2001 From: Fabio Olive Leite Date: Sun, 12 Sep 2010 19:55:25 -0400 Subject: Remove incorrect do_vfs_lock message The do_vfs_lock function on fs/nfs/file.c is only called if NLM is not being used, via the -onolock mount option. Therefore it cannot really be "out of sync with lock manager" when the local locking function called returns an error, as there will be no corresponding call to the NLM. For details, simply check the if/else on do_setlk and do_unlk on fs/nfs/file.c. Signed-Off-By: Fabio Olive Leite Reviewed-by: Jeff Layton Signed-off-by: Trond Myklebust diff --git a/fs/nfs/file.c b/fs/nfs/file.c index eb51bd6..05bf3c0 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -723,10 +723,6 @@ static int do_vfs_lock(struct file *file, struct file_lock *fl) default: BUG(); } - if (res < 0) - dprintk(KERN_WARNING "%s: VFS is out of sync with lock manager" - " - error %d!\n", - __func__, res); return res; } -- cgit v0.10.2 From ce8477e1176389ed920550f4c925ad4a815b22d5 Mon Sep 17 00:00:00 2001 From: Bian Naimeng Date: Sun, 12 Sep 2010 19:55:25 -0400 Subject: gss:krb5 miss returning error to caller when import security context krb5 miss returning error to up layer when import security context, it may be return ok though it has failed to import security context. Signed-off-by: Bian Naimeng Signed-off-by: Trond Myklebust diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 0326446..778e5df 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -237,6 +237,7 @@ get_key(const void *p, const void *end, if (!supported_gss_krb5_enctype(alg)) { printk(KERN_WARNING "gss_kerberos_mech: unsupported " "encryption key algorithm %d\n", alg); + p = ERR_PTR(-EINVAL); goto out_err; } p = simple_get_netobj(p, end, &key); @@ -282,15 +283,19 @@ gss_import_v1_context(const void *p, const void *end, struct krb5_ctx *ctx) ctx->enctype = ENCTYPE_DES_CBC_RAW; ctx->gk5e = get_gss_krb5_enctype(ctx->enctype); - if (ctx->gk5e == NULL) + if (ctx->gk5e == NULL) { + p = ERR_PTR(-EINVAL); goto out_err; + } /* The downcall format was designed before we completely understood * the uses of the context fields; so it includes some stuff we * just give some minimal sanity-checking, and some we ignore * completely (like the next twenty bytes): */ - if (unlikely(p + 20 > end || p + 20 < p)) + if (unlikely(p + 20 > end || p + 20 < p)) { + p = ERR_PTR(-EFAULT); goto out_err; + } p += 20; p = simple_get_bytes(p, end, &tmp, sizeof(tmp)); if (IS_ERR(p)) @@ -619,6 +624,7 @@ gss_import_v2_context(const void *p, const void *end, struct krb5_ctx *ctx, if (ctx->seq_send64 != ctx->seq_send) { dprintk("%s: seq_send64 %lx, seq_send %x overflow?\n", __func__, (long unsigned)ctx->seq_send64, ctx->seq_send); + p = ERR_PTR(-EINVAL); goto out_err; } p = simple_get_bytes(p, end, &ctx->enctype, sizeof(ctx->enctype)); -- cgit v0.10.2 From 651b2933b22a0c060e6bb940c4104eb447a61f9a Mon Sep 17 00:00:00 2001 From: Bian Naimeng Date: Sun, 12 Sep 2010 19:55:26 -0400 Subject: gss:spkm3 miss returning error to caller when import security context spkm3 miss returning error to up layer when import security context, it may be return ok though it has failed to import security context. Signed-off-by: Bian Naimeng Signed-off-by: Trond Myklebust diff --git a/net/sunrpc/auth_gss/gss_spkm3_mech.c b/net/sunrpc/auth_gss/gss_spkm3_mech.c index dc3f1f5..adade3d 100644 --- a/net/sunrpc/auth_gss/gss_spkm3_mech.c +++ b/net/sunrpc/auth_gss/gss_spkm3_mech.c @@ -100,6 +100,7 @@ gss_import_sec_context_spkm3(const void *p, size_t len, if (version != 1) { dprintk("RPC: unknown spkm3 token format: " "obsolete nfs-utils?\n"); + p = ERR_PTR(-EINVAL); goto out_err_free_ctx; } @@ -135,8 +136,10 @@ gss_import_sec_context_spkm3(const void *p, size_t len, if (IS_ERR(p)) goto out_err_free_intg_alg; - if (p != end) + if (p != end) { + p = ERR_PTR(-EFAULT); goto out_err_free_intg_key; + } ctx_id->internal_ctx_id = ctx; -- cgit v0.10.2 From db5fe26541b6b48460104a0d949a27cdc7786957 Mon Sep 17 00:00:00 2001 From: Miquel van Smoorenburg Date: Sun, 12 Sep 2010 19:55:26 -0400 Subject: sunrpc: increase MAX_HASHTABLE_BITS to 14 The maximum size of the authcache is now set to 1024 (10 bits), but on our server we need at least 4096 (12 bits). Increase MAX_HASHTABLE_BITS to 14. This is a maximum of 16384 entries, each containing a pointer (8 bytes on x86_64). This is exactly the limit of kmalloc() (128K). Signed-off-by: Miquel van Smoorenburg Signed-off-by: Trond Myklebust diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index 36cb660..e9eaaf7 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -38,7 +38,7 @@ static const struct rpc_authops *auth_flavors[RPC_AUTH_MAXFLAVOR] = { static LIST_HEAD(cred_unused); static unsigned long number_cred_unused; -#define MAX_HASHTABLE_BITS (10) +#define MAX_HASHTABLE_BITS (14) static int param_set_hashtbl_sz(const char *val, const struct kernel_param *kp) { unsigned long num; -- cgit v0.10.2 From b20d37ca9561711c6a3c4b859c2855f49565e061 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 12 Sep 2010 19:55:26 -0400 Subject: NFS: Fix a typo in nfs_sockaddr_match_ipaddr6 Reported-by: Ben Greear Signed-off-by: Trond Myklebust Cc: stable@kernel.org diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 4e7df2a..e734072 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -275,7 +275,7 @@ static int nfs_sockaddr_match_ipaddr6(const struct sockaddr *sa1, sin1->sin6_scope_id != sin2->sin6_scope_id) return 0; - return ipv6_addr_equal(&sin1->sin6_addr, &sin1->sin6_addr); + return ipv6_addr_equal(&sin1->sin6_addr, &sin2->sin6_addr); } #else /* !defined(CONFIG_IPV6) && !defined(CONFIG_IPV6_MODULE) */ static int nfs_sockaddr_match_ipaddr6(const struct sockaddr *sa1, -- cgit v0.10.2 From fbf3fdd2443965d9ba6fb4b5fecd1f6e0847218f Mon Sep 17 00:00:00 2001 From: Menyhart Zoltan Date: Sun, 12 Sep 2010 19:55:26 -0400 Subject: statfs() gives ESTALE error Hi, An NFS client executes a statfs("file", &buff) call. "file" exists / existed, the client has read / written it, but it has already closed it. user_path(pathname, &path) looks up "file" successfully in the directory-cache and restarts the aging timer of the directory-entry. Even if "file" has already been removed from the server, because the lookupcache=positive option I use, keeps the entries valid for a while. nfs_statfs() returns ESTALE if "file" has already been removed from the server. If the user application repeats the statfs("file", &buff) call, we are stuck: "file" remains young forever in the directory-cache. Signed-off-by: Zoltan Menyhart Signed-off-by: Trond Myklebust Cc: stable@kernel.org diff --git a/fs/nfs/super.c b/fs/nfs/super.c index ec3966e..f4cbf0c 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -431,7 +431,15 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf) goto out_err; error = server->nfs_client->rpc_ops->statfs(server, fh, &res); + if (unlikely(error == -ESTALE)) { + struct dentry *pd_dentry; + pd_dentry = dget_parent(dentry); + if (pd_dentry != NULL) { + nfs_zap_caches(pd_dentry->d_inode); + dput(pd_dentry); + } + } nfs_free_fattr(res.fattr); if (error < 0) goto out_err; -- cgit v0.10.2 From 827e3457022d0bb0b1bb8a0eb88501876fe7dcf0 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 12 Sep 2010 19:57:50 -0400 Subject: SUNRPC: Fix the NFSv4 and RPCSEC_GSS Kconfig dependencies MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The NFSv4 client's callback server calls svc_gss_principal(), which is defined in the auth_rpcgss.ko The NFSv4 server has the same dependency, and in addition calls svcauth_gss_flavor(), gss_mech_get_by_pseudoflavor(), gss_pseudoflavor_to_service() and gss_mech_put() from the same module. The module auth_rpcgss itself has no dependencies aside from sunrpc, so we only need to select RPCSEC_GSS. Reported-by: Uwe Kleine-König Signed-off-by: Trond Myklebust diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index 6c2aad4..f7e13db 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -63,6 +63,7 @@ config NFS_V3_ACL config NFS_V4 bool "NFS client support for NFS version 4" depends on NFS_FS + select SUNRPC_GSS help This option enables support for version 4 of the NFS protocol (RFC 3530) in the kernel's NFS client. diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig index 95932f5..4264377 100644 --- a/fs/nfsd/Kconfig +++ b/fs/nfsd/Kconfig @@ -69,6 +69,7 @@ config NFSD_V4 depends on NFSD && PROC_FS && EXPERIMENTAL select NFSD_V3 select FS_POSIX_ACL + select SUNRPC_GSS help This option enables support in your system's NFS server for version 4 of the NFS protocol (RFC 3530). -- cgit v0.10.2 From b343ae51c116dffaef07a8596661774c12212b66 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Sun, 12 Sep 2010 17:20:24 -0700 Subject: sparc64: Support RAW perf events. Encoding is "(encoding << 16) | pic_mask" Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/perf_event.c b/arch/sparc/kernel/perf_event.c index 357ced3..6318e62 100644 --- a/arch/sparc/kernel/perf_event.c +++ b/arch/sparc/kernel/perf_event.c @@ -1038,6 +1038,7 @@ static int __hw_perf_event_init(struct perf_event *event) if (atomic_read(&nmi_active) < 0) return -ENODEV; + pmap = NULL; if (attr->type == PERF_TYPE_HARDWARE) { if (attr->config >= sparc_pmu->max_events) return -EINVAL; @@ -1046,9 +1047,18 @@ static int __hw_perf_event_init(struct perf_event *event) pmap = sparc_map_cache_event(attr->config); if (IS_ERR(pmap)) return PTR_ERR(pmap); - } else + } else if (attr->type != PERF_TYPE_RAW) return -EOPNOTSUPP; + if (pmap) { + hwc->event_base = perf_event_encode(pmap); + } else { + /* User gives us "(encoding << 16) | pic_mask" for + * PERF_TYPE_RAW events. + */ + hwc->event_base = attr->config; + } + /* We save the enable bits in the config_base. */ hwc->config_base = sparc_pmu->irq_bit; if (!attr->exclude_user) @@ -1058,8 +1068,6 @@ static int __hw_perf_event_init(struct perf_event *event) if (!attr->exclude_hv) hwc->config_base |= sparc_pmu->hv_bit; - hwc->event_base = perf_event_encode(pmap); - n = 0; if (event->group_leader != event) { n = collect_events(event->group_leader, -- cgit v0.10.2 From cea310e8f8702226f982f09386cfd3c5793c5e2f Mon Sep 17 00:00:00 2001 From: Seth Heasley Date: Fri, 10 Sep 2010 16:29:56 -0700 Subject: ALSA: hda_intel: ALSA HD Audio patch for Intel Patsburg DeviceIDs This patch adds the Intel Patsburg (PCH) HD Audio Controller DeviceIDs. Signed-off-by: Seth Heasley Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index 1053fff..34940a0 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -126,6 +126,7 @@ MODULE_SUPPORTED_DEVICE("{{Intel, ICH6}," "{Intel, ICH10}," "{Intel, PCH}," "{Intel, CPT}," + "{Intel, PBG}," "{Intel, SCH}," "{ATI, SB450}," "{ATI, SB600}," @@ -2749,6 +2750,8 @@ static DEFINE_PCI_DEVICE_TABLE(azx_ids) = { { PCI_DEVICE(0x8086, 0x3b57), .driver_data = AZX_DRIVER_ICH }, /* CPT */ { PCI_DEVICE(0x8086, 0x1c20), .driver_data = AZX_DRIVER_PCH }, + /* PBG */ + { PCI_DEVICE(0x8086, 0x1d20), .driver_data = AZX_DRIVER_PCH }, /* SCH */ { PCI_DEVICE(0x8086, 0x811b), .driver_data = AZX_DRIVER_SCH }, /* ATI SB 450/600 */ -- cgit v0.10.2 From eac15a429a27cb74115daaf4c1127c5e854d50e4 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Sat, 28 Aug 2010 01:45:00 -0400 Subject: mtd: Blackfin NFC: fix build error after nand_scan_ident() change Seems some patches got out sync when being merged. The Blackfin NFC driver was updated to use nand_scan_ident(), but it missed the change where nand_scan_ident() now takes 3 arguments. So update this driver to fix build failures. Signed-off-by: Mike Frysinger Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/bf5xx_nand.c b/drivers/mtd/nand/bf5xx_nand.c index a382e3d..162c5ea 100644 --- a/drivers/mtd/nand/bf5xx_nand.c +++ b/drivers/mtd/nand/bf5xx_nand.c @@ -710,7 +710,7 @@ static int bf5xx_nand_scan(struct mtd_info *mtd) struct nand_chip *chip = mtd->priv; int ret; - ret = nand_scan_ident(mtd, 1); + ret = nand_scan_ident(mtd, 1, NULL); if (ret) return ret; -- cgit v0.10.2 From 8b865d5efd9205b131dd9a43a6f450c05d38aaa1 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Sat, 28 Aug 2010 16:42:04 -0400 Subject: mtd: Blackfin NFC: fix invalid free in remove() Since info->mtd isn't dynamically allocated, we shouldn't attempt to kfree() it. Otherwise we get random fun corruption when unloading the driver built as a module. Signed-off-by: Mike Frysinger Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/bf5xx_nand.c b/drivers/mtd/nand/bf5xx_nand.c index 162c5ea..6fbeefa 100644 --- a/drivers/mtd/nand/bf5xx_nand.c +++ b/drivers/mtd/nand/bf5xx_nand.c @@ -682,7 +682,6 @@ static int __devinit bf5xx_nand_add_partition(struct bf5xx_nand_info *info) static int __devexit bf5xx_nand_remove(struct platform_device *pdev) { struct bf5xx_nand_info *info = to_nand_info(pdev); - struct mtd_info *mtd = NULL; platform_set_drvdata(pdev, NULL); @@ -690,11 +689,7 @@ static int __devexit bf5xx_nand_remove(struct platform_device *pdev) * and their partitions, then go through freeing the * resources used */ - mtd = &info->mtd; - if (mtd) { - nand_release(mtd); - kfree(mtd); - } + nand_release(&info->mtd); peripheral_free_list(bfin_nfc_pin_req); bf5xx_nand_dma_remove(info); -- cgit v0.10.2 From 9aba97ad004ed0cde9747a9daf5b1484edb746cd Mon Sep 17 00:00:00 2001 From: Kyungmin Park Date: Fri, 27 Aug 2010 11:55:44 +0900 Subject: mtd: OneNAND: Fix 2KiB pagesize handling at Samsung SoCs Wrong assumption bufferram can be switched between BufferRAM0 and BufferRAM1 Signed-off-by: Kyungmin Park Signed-off-by: David Woodhouse diff --git a/drivers/mtd/onenand/samsung.c b/drivers/mtd/onenand/samsung.c index cb443af..094dfcb 100644 --- a/drivers/mtd/onenand/samsung.c +++ b/drivers/mtd/onenand/samsung.c @@ -571,13 +571,12 @@ static int s5pc110_read_bufferram(struct mtd_info *mtd, int area, unsigned char *buffer, int offset, size_t count) { struct onenand_chip *this = mtd->priv; - void __iomem *bufferram; void __iomem *p; void *buf = (void *) buffer; dma_addr_t dma_src, dma_dst; int err; - p = bufferram = this->base + area; + p = this->base + area; if (ONENAND_CURRENT_BUFFERRAM(this)) { if (area == ONENAND_DATARAM) p += this->writesize; @@ -621,7 +620,7 @@ static int s5pc110_read_bufferram(struct mtd_info *mtd, int area, normal: if (count != mtd->writesize) { /* Copy the bufferram to memory to prevent unaligned access */ - memcpy(this->page_buf, bufferram, mtd->writesize); + memcpy(this->page_buf, p, mtd->writesize); p = this->page_buf + offset; } -- cgit v0.10.2 From 53d1e137d5ddd8a1c5ec54c5375cb2832e1cbcd1 Mon Sep 17 00:00:00 2001 From: Kyungmin Park Date: Fri, 27 Aug 2010 11:55:37 +0900 Subject: mtd: OneNAND: Fix loop hang when DMA error at Samsung SoCs When DMA error occurs. it's loop hang since it can't exit the loop. and it's the right DMA handling code as Spec. Signed-off-by: Kyungmin Park Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/onenand/samsung.c b/drivers/mtd/onenand/samsung.c index 094dfcb..a460f1b 100644 --- a/drivers/mtd/onenand/samsung.c +++ b/drivers/mtd/onenand/samsung.c @@ -554,14 +554,13 @@ static int s5pc110_dma_ops(void *dst, void *src, size_t count, int direction) do { status = readl(base + S5PC110_DMA_TRANS_STATUS); + if (status & S5PC110_DMA_TRANS_STATUS_TE) { + writel(S5PC110_DMA_TRANS_CMD_TEC, + base + S5PC110_DMA_TRANS_CMD); + return -EIO; + } } while (!(status & S5PC110_DMA_TRANS_STATUS_TD)); - if (status & S5PC110_DMA_TRANS_STATUS_TE) { - writel(S5PC110_DMA_TRANS_CMD_TEC, base + S5PC110_DMA_TRANS_CMD); - writel(S5PC110_DMA_TRANS_CMD_TDC, base + S5PC110_DMA_TRANS_CMD); - return -EIO; - } - writel(S5PC110_DMA_TRANS_CMD_TDC, base + S5PC110_DMA_TRANS_CMD); return 0; -- cgit v0.10.2 From b8db2f51dcccbaf4d23ed44cd0c64856c58a16c0 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Mon, 9 Aug 2010 15:04:19 +0200 Subject: mtd: mxc_nand: configure pages per block for v2 controller This patch initializes the pages per block field in CONFIG1 for v2 controllers. It also sets the FP_INT field. This is the last field not correctly initialized, so we can switch from read/modify/write the CONFIG1 reg to just write the correct value. Signed-off-by: Sascha Hauer Acked-by: John Ogness Tested-by: John Ogness Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c index 26caa01..b2828e8 100644 --- a/drivers/mtd/nand/mxc_nand.c +++ b/drivers/mtd/nand/mxc_nand.c @@ -67,7 +67,9 @@ #define NFC_V1_V2_CONFIG1_BIG (1 << 5) #define NFC_V1_V2_CONFIG1_RST (1 << 6) #define NFC_V1_V2_CONFIG1_CE (1 << 7) -#define NFC_V1_V2_CONFIG1_ONE_CYCLE (1 << 8) +#define NFC_V2_CONFIG1_ONE_CYCLE (1 << 8) +#define NFC_V2_CONFIG1_PPB(x) (((x) & 0x3) << 9) +#define NFC_V2_CONFIG1_FP_INT (1 << 11) #define NFC_V1_V2_CONFIG2_INT (1 << 15) @@ -729,27 +731,30 @@ static void preset_v1_v2(struct mtd_info *mtd) { struct nand_chip *nand_chip = mtd->priv; struct mxc_nand_host *host = nand_chip->priv; - uint16_t tmp; - - /* enable interrupt, disable spare enable */ - tmp = readw(NFC_V1_V2_CONFIG1); - tmp &= ~NFC_V1_V2_CONFIG1_INT_MSK; - tmp &= ~NFC_V1_V2_CONFIG1_SP_EN; - if (nand_chip->ecc.mode == NAND_ECC_HW) { - tmp |= NFC_V1_V2_CONFIG1_ECC_EN; - } else { - tmp &= ~NFC_V1_V2_CONFIG1_ECC_EN; - } + uint16_t config1 = 0; + + if (nand_chip->ecc.mode == NAND_ECC_HW) + config1 |= NFC_V1_V2_CONFIG1_ECC_EN; + + if (nfc_is_v21()) + config1 |= NFC_V2_CONFIG1_FP_INT; + + if (!cpu_is_mx21()) + config1 |= NFC_V1_V2_CONFIG1_INT_MSK; if (nfc_is_v21() && mtd->writesize) { + uint16_t pages_per_block = mtd->erasesize / mtd->writesize; + host->eccsize = get_eccsize(mtd); if (host->eccsize == 4) - tmp |= NFC_V2_CONFIG1_ECC_MODE_4; + config1 |= NFC_V2_CONFIG1_ECC_MODE_4; + + config1 |= NFC_V2_CONFIG1_PPB(ffs(pages_per_block) - 6); } else { host->eccsize = 1; } - writew(tmp, NFC_V1_V2_CONFIG1); + writew(config1, NFC_V1_V2_CONFIG1); /* preset operation */ /* Unlock the internal RAM Buffer */ -- cgit v0.10.2 From 99d389640b58052884fb231bce9dbffb4f595aa4 Mon Sep 17 00:00:00 2001 From: "Mark F. Brown" Date: Thu, 26 Aug 2010 04:56:51 -0400 Subject: mtd: pxa3xx: fix build error when CONFIG_MTD_PARTITIONS is not defined Signed-off-by: Mark F. Brown Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c index 4d89f37..4d01cda6 100644 --- a/drivers/mtd/nand/pxa3xx_nand.c +++ b/drivers/mtd/nand/pxa3xx_nand.c @@ -1320,6 +1320,7 @@ static int pxa3xx_nand_probe(struct platform_device *pdev) goto fail_free_irq; } +#ifdef CONFIG_MTD_PARTITIONS if (mtd_has_cmdlinepart()) { static const char *probes[] = { "cmdlinepart", NULL }; struct mtd_partition *parts; @@ -1332,6 +1333,9 @@ static int pxa3xx_nand_probe(struct platform_device *pdev) } return add_mtd_partitions(mtd, pdata->parts, pdata->nr_parts); +#else + return 0; +#endif fail_free_irq: free_irq(irq, info); @@ -1364,7 +1368,9 @@ static int pxa3xx_nand_remove(struct platform_device *pdev) platform_set_drvdata(pdev, NULL); del_mtd_device(mtd); +#ifdef CONFIG_MTD_PARTITIONS del_mtd_partitions(mtd); +#endif irq = platform_get_irq(pdev, 0); if (irq >= 0) free_irq(irq, info); -- cgit v0.10.2 From c54fce6eff197d9c57c97afbf6c9722ce434fc8f Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 10 Sep 2010 16:51:36 +0200 Subject: workqueue: add documentation Update copyright notice and add Documentation/workqueue.txt. Randy Dunlap, Dave Chinner: misc fixes. Signed-off-by: Tejun Heo Reviewed-By: Florian Mickler Cc: Ingo Molnar Cc: Christoph Lameter Cc: Randy Dunlap Cc: Dave Chinner diff --git a/Documentation/workqueue.txt b/Documentation/workqueue.txt new file mode 100644 index 0000000..e4498a2 --- /dev/null +++ b/Documentation/workqueue.txt @@ -0,0 +1,380 @@ + +Concurrency Managed Workqueue (cmwq) + +September, 2010 Tejun Heo + Florian Mickler + +CONTENTS + +1. Introduction +2. Why cmwq? +3. The Design +4. Application Programming Interface (API) +5. Example Execution Scenarios +6. Guidelines + + +1. Introduction + +There are many cases where an asynchronous process execution context +is needed and the workqueue (wq) API is the most commonly used +mechanism for such cases. + +When such an asynchronous execution context is needed, a work item +describing which function to execute is put on a queue. An +independent thread serves as the asynchronous execution context. The +queue is called workqueue and the thread is called worker. + +While there are work items on the workqueue the worker executes the +functions associated with the work items one after the other. When +there is no work item left on the workqueue the worker becomes idle. +When a new work item gets queued, the worker begins executing again. + + +2. Why cmwq? + +In the original wq implementation, a multi threaded (MT) wq had one +worker thread per CPU and a single threaded (ST) wq had one worker +thread system-wide. A single MT wq needed to keep around the same +number of workers as the number of CPUs. The kernel grew a lot of MT +wq users over the years and with the number of CPU cores continuously +rising, some systems saturated the default 32k PID space just booting +up. + +Although MT wq wasted a lot of resource, the level of concurrency +provided was unsatisfactory. The limitation was common to both ST and +MT wq albeit less severe on MT. Each wq maintained its own separate +worker pool. A MT wq could provide only one execution context per CPU +while a ST wq one for the whole system. Work items had to compete for +those very limited execution contexts leading to various problems +including proneness to deadlocks around the single execution context. + +The tension between the provided level of concurrency and resource +usage also forced its users to make unnecessary tradeoffs like libata +choosing to use ST wq for polling PIOs and accepting an unnecessary +limitation that no two polling PIOs can progress at the same time. As +MT wq don't provide much better concurrency, users which require +higher level of concurrency, like async or fscache, had to implement +their own thread pool. + +Concurrency Managed Workqueue (cmwq) is a reimplementation of wq with +focus on the following goals. + +* Maintain compatibility with the original workqueue API. + +* Use per-CPU unified worker pools shared by all wq to provide + flexible level of concurrency on demand without wasting a lot of + resource. + +* Automatically regulate worker pool and level of concurrency so that + the API users don't need to worry about such details. + + +3. The Design + +In order to ease the asynchronous execution of functions a new +abstraction, the work item, is introduced. + +A work item is a simple struct that holds a pointer to the function +that is to be executed asynchronously. Whenever a driver or subsystem +wants a function to be executed asynchronously it has to set up a work +item pointing to that function and queue that work item on a +workqueue. + +Special purpose threads, called worker threads, execute the functions +off of the queue, one after the other. If no work is queued, the +worker threads become idle. These worker threads are managed in so +called thread-pools. + +The cmwq design differentiates between the user-facing workqueues that +subsystems and drivers queue work items on and the backend mechanism +which manages thread-pool and processes the queued work items. + +The backend is called gcwq. There is one gcwq for each possible CPU +and one gcwq to serve work items queued on unbound workqueues. + +Subsystems and drivers can create and queue work items through special +workqueue API functions as they see fit. They can influence some +aspects of the way the work items are executed by setting flags on the +workqueue they are putting the work item on. These flags include +things like CPU locality, reentrancy, concurrency limits and more. To +get a detailed overview refer to the API description of +alloc_workqueue() below. + +When a work item is queued to a workqueue, the target gcwq is +determined according to the queue parameters and workqueue attributes +and appended on the shared worklist of the gcwq. For example, unless +specifically overridden, a work item of a bound workqueue will be +queued on the worklist of exactly that gcwq that is associated to the +CPU the issuer is running on. + +For any worker pool implementation, managing the concurrency level +(how many execution contexts are active) is an important issue. cmwq +tries to keep the concurrency at a minimal but sufficient level. +Minimal to save resources and sufficient in that the system is used at +its full capacity. + +Each gcwq bound to an actual CPU implements concurrency management by +hooking into the scheduler. The gcwq is notified whenever an active +worker wakes up or sleeps and keeps track of the number of the +currently runnable workers. Generally, work items are not expected to +hog a CPU and consume many cycles. That means maintaining just enough +concurrency to prevent work processing from stalling should be +optimal. As long as there are one or more runnable workers on the +CPU, the gcwq doesn't start execution of a new work, but, when the +last running worker goes to sleep, it immediately schedules a new +worker so that the CPU doesn't sit idle while there are pending work +items. This allows using a minimal number of workers without losing +execution bandwidth. + +Keeping idle workers around doesn't cost other than the memory space +for kthreads, so cmwq holds onto idle ones for a while before killing +them. + +For an unbound wq, the above concurrency management doesn't apply and +the gcwq for the pseudo unbound CPU tries to start executing all work +items as soon as possible. The responsibility of regulating +concurrency level is on the users. There is also a flag to mark a +bound wq to ignore the concurrency management. Please refer to the +API section for details. + +Forward progress guarantee relies on that workers can be created when +more execution contexts are necessary, which in turn is guaranteed +through the use of rescue workers. All work items which might be used +on code paths that handle memory reclaim are required to be queued on +wq's that have a rescue-worker reserved for execution under memory +pressure. Else it is possible that the thread-pool deadlocks waiting +for execution contexts to free up. + + +4. Application Programming Interface (API) + +alloc_workqueue() allocates a wq. The original create_*workqueue() +functions are deprecated and scheduled for removal. alloc_workqueue() +takes three arguments - @name, @flags and @max_active. @name is the +name of the wq and also used as the name of the rescuer thread if +there is one. + +A wq no longer manages execution resources but serves as a domain for +forward progress guarantee, flush and work item attributes. @flags +and @max_active control how work items are assigned execution +resources, scheduled and executed. + +@flags: + + WQ_NON_REENTRANT + + By default, a wq guarantees non-reentrance only on the same + CPU. A work item may not be executed concurrently on the same + CPU by multiple workers but is allowed to be executed + concurrently on multiple CPUs. This flag makes sure + non-reentrance is enforced across all CPUs. Work items queued + to a non-reentrant wq are guaranteed to be executed by at most + one worker system-wide at any given time. + + WQ_UNBOUND + + Work items queued to an unbound wq are served by a special + gcwq which hosts workers which are not bound to any specific + CPU. This makes the wq behave as a simple execution context + provider without concurrency management. The unbound gcwq + tries to start execution of work items as soon as possible. + Unbound wq sacrifices locality but is useful for the following + cases. + + * Wide fluctuation in the concurrency level requirement is + expected and using bound wq may end up creating large number + of mostly unused workers across different CPUs as the issuer + hops through different CPUs. + + * Long running CPU intensive workloads which can be better + managed by the system scheduler. + + WQ_FREEZEABLE + + A freezeable wq participates in the freeze phase of the system + suspend operations. Work items on the wq are drained and no + new work item starts execution until thawed. + + WQ_RESCUER + + All wq which might be used in the memory reclaim paths _MUST_ + have this flag set. This reserves one worker exclusively for + the execution of this wq under memory pressure. + + WQ_HIGHPRI + + Work items of a highpri wq are queued at the head of the + worklist of the target gcwq and start execution regardless of + the current concurrency level. In other words, highpri work + items will always start execution as soon as execution + resource is available. + + Ordering among highpri work items is preserved - a highpri + work item queued after another highpri work item will start + execution after the earlier highpri work item starts. + + Although highpri work items are not held back by other + runnable work items, they still contribute to the concurrency + level. Highpri work items in runnable state will prevent + non-highpri work items from starting execution. + + This flag is meaningless for unbound wq. + + WQ_CPU_INTENSIVE + + Work items of a CPU intensive wq do not contribute to the + concurrency level. In other words, runnable CPU intensive + work items will not prevent other work items from starting + execution. This is useful for bound work items which are + expected to hog CPU cycles so that their execution is + regulated by the system scheduler. + + Although CPU intensive work items don't contribute to the + concurrency level, start of their executions is still + regulated by the concurrency management and runnable + non-CPU-intensive work items can delay execution of CPU + intensive work items. + + This flag is meaningless for unbound wq. + + WQ_HIGHPRI | WQ_CPU_INTENSIVE + + This combination makes the wq avoid interaction with + concurrency management completely and behave as a simple + per-CPU execution context provider. Work items queued on a + highpri CPU-intensive wq start execution as soon as resources + are available and don't affect execution of other work items. + +@max_active: + +@max_active determines the maximum number of execution contexts per +CPU which can be assigned to the work items of a wq. For example, +with @max_active of 16, at most 16 work items of the wq can be +executing at the same time per CPU. + +Currently, for a bound wq, the maximum limit for @max_active is 512 +and the default value used when 0 is specified is 256. For an unbound +wq, the limit is higher of 512 and 4 * num_possible_cpus(). These +values are chosen sufficiently high such that they are not the +limiting factor while providing protection in runaway cases. + +The number of active work items of a wq is usually regulated by the +users of the wq, more specifically, by how many work items the users +may queue at the same time. Unless there is a specific need for +throttling the number of active work items, specifying '0' is +recommended. + +Some users depend on the strict execution ordering of ST wq. The +combination of @max_active of 1 and WQ_UNBOUND is used to achieve this +behavior. Work items on such wq are always queued to the unbound gcwq +and only one work item can be active at any given time thus achieving +the same ordering property as ST wq. + + +5. Example Execution Scenarios + +The following example execution scenarios try to illustrate how cmwq +behave under different configurations. + + Work items w0, w1, w2 are queued to a bound wq q0 on the same CPU. + w0 burns CPU for 5ms then sleeps for 10ms then burns CPU for 5ms + again before finishing. w1 and w2 burn CPU for 5ms then sleep for + 10ms. + +Ignoring all other tasks, works and processing overhead, and assuming +simple FIFO scheduling, the following is one highly simplified version +of possible sequences of events with the original wq. + + TIME IN MSECS EVENT + 0 w0 starts and burns CPU + 5 w0 sleeps + 15 w0 wakes up and burns CPU + 20 w0 finishes + 20 w1 starts and burns CPU + 25 w1 sleeps + 35 w1 wakes up and finishes + 35 w2 starts and burns CPU + 40 w2 sleeps + 50 w2 wakes up and finishes + +And with cmwq with @max_active >= 3, + + TIME IN MSECS EVENT + 0 w0 starts and burns CPU + 5 w0 sleeps + 5 w1 starts and burns CPU + 10 w1 sleeps + 10 w2 starts and burns CPU + 15 w2 sleeps + 15 w0 wakes up and burns CPU + 20 w0 finishes + 20 w1 wakes up and finishes + 25 w2 wakes up and finishes + +If @max_active == 2, + + TIME IN MSECS EVENT + 0 w0 starts and burns CPU + 5 w0 sleeps + 5 w1 starts and burns CPU + 10 w1 sleeps + 15 w0 wakes up and burns CPU + 20 w0 finishes + 20 w1 wakes up and finishes + 20 w2 starts and burns CPU + 25 w2 sleeps + 35 w2 wakes up and finishes + +Now, let's assume w1 and w2 are queued to a different wq q1 which has +WQ_HIGHPRI set, + + TIME IN MSECS EVENT + 0 w1 and w2 start and burn CPU + 5 w1 sleeps + 10 w2 sleeps + 10 w0 starts and burns CPU + 15 w0 sleeps + 15 w1 wakes up and finishes + 20 w2 wakes up and finishes + 25 w0 wakes up and burns CPU + 30 w0 finishes + +If q1 has WQ_CPU_INTENSIVE set, + + TIME IN MSECS EVENT + 0 w0 starts and burns CPU + 5 w0 sleeps + 5 w1 and w2 start and burn CPU + 10 w1 sleeps + 15 w2 sleeps + 15 w0 wakes up and burns CPU + 20 w0 finishes + 20 w1 wakes up and finishes + 25 w2 wakes up and finishes + + +6. Guidelines + +* Do not forget to use WQ_RESCUER if a wq may process work items which + are used during memory reclaim. Each wq with WQ_RESCUER set has one + rescuer thread reserved for it. If there is dependency among + multiple work items used during memory reclaim, they should be + queued to separate wq each with WQ_RESCUER. + +* Unless strict ordering is required, there is no need to use ST wq. + +* Unless there is a specific need, using 0 for @max_active is + recommended. In most use cases, concurrency level usually stays + well under the default limit. + +* A wq serves as a domain for forward progress guarantee (WQ_RESCUER), + flush and work item attributes. Work items which are not involved + in memory reclaim and don't need to be flushed as a part of a group + of work items, and don't require any special attribute, can use one + of the system wq. There is no difference in execution + characteristics between using a dedicated wq and a system wq. + +* Unless work items are expected to consume a huge amount of CPU + cycles, using a bound wq is usually beneficial due to the increased + level of locality in wq operations and work item execution. diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h index f11100f..25e02c9 100644 --- a/include/linux/workqueue.h +++ b/include/linux/workqueue.h @@ -235,6 +235,10 @@ static inline unsigned int work_static(struct work_struct *work) { return 0; } #define work_clear_pending(work) \ clear_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work)) +/* + * Workqueue flags and constants. For details, please refer to + * Documentation/workqueue.txt. + */ enum { WQ_NON_REENTRANT = 1 << 0, /* guarantee non-reentrance */ WQ_UNBOUND = 1 << 1, /* not bound to any cpu */ diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 727f24e..f77afd9 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -1,19 +1,26 @@ /* - * linux/kernel/workqueue.c + * kernel/workqueue.c - generic async execution with shared worker pool * - * Generic mechanism for defining kernel helper threads for running - * arbitrary tasks in process context. + * Copyright (C) 2002 Ingo Molnar * - * Started by Ingo Molnar, Copyright (C) 2002 + * Derived from the taskqueue/keventd code by: + * David Woodhouse + * Andrew Morton + * Kai Petzke + * Theodore Ts'o * - * Derived from the taskqueue/keventd code by: + * Made to use alloc_percpu by Christoph Lameter. * - * David Woodhouse - * Andrew Morton - * Kai Petzke - * Theodore Ts'o + * Copyright (C) 2010 SUSE Linux Products GmbH + * Copyright (C) 2010 Tejun Heo * - * Made to use alloc_percpu by Christoph Lameter. + * This is the generic async execution mechanism. Work items as are + * executed in process context. The worker pool is shared and + * automatically managed. There is one worker pool for each CPU and + * one extra for works which are better served by workers which are + * not bound to any specific CPU. + * + * Please read Documentation/workqueue.txt for details. */ #include -- cgit v0.10.2 From 551402a30efa45560e23c22a7aa04453861602c3 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 6 Sep 2010 23:53:47 +0100 Subject: drm: Fix regression in disable polling e58f637 I broke out my trusty i845 and found a new boot failure, which upon inspection turned out to be a recursion within: drm_helper_probe_single_connector_modes() -> drm_helper_hpd_irq_event() -> intel_crt_detect() -> drm_helper_probe_single_connector_modes() Calling drm_kms_helper_poll_enable() instead performs the desired re-initialisation of the polling should the user have toggled the parameter, without the recursive side-effect. Signed-off-by: Chris Wilson Cc: Dave Airlie Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index d2ab01e..4238a19 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -104,7 +104,7 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector, connector->funcs->force(connector); } else { connector->status = connector->funcs->detect(connector); - drm_helper_hpd_irq_event(dev); + drm_kms_helper_poll_enable(dev); } if (connector->status == connector_status_disconnected) { -- cgit v0.10.2 From ff32a59daea4f3eae980ae8d0932b135c633ec5d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 7 Sep 2010 13:26:39 -0400 Subject: drm/radeon/kms: fix regression in RMX code (v2) caused by d65d65b175a29bd7ea2bb69c046419329c4a5db7 need to update the radeon crtc priv native mode before using it. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30049 v2: integrate v/h copy paste typo Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 6dd434a..8c987c9 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1140,14 +1140,14 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc, radeon_crtc->rmx_type = radeon_encoder->rmx_type; else radeon_crtc->rmx_type = RMX_OFF; - src_v = crtc->mode.vdisplay; - dst_v = radeon_crtc->native_mode.vdisplay; - src_h = crtc->mode.hdisplay; - dst_h = radeon_crtc->native_mode.vdisplay; /* copy native mode */ memcpy(&radeon_crtc->native_mode, &radeon_encoder->native_mode, sizeof(struct drm_display_mode)); + src_v = crtc->mode.vdisplay; + dst_v = radeon_crtc->native_mode.vdisplay; + src_h = crtc->mode.hdisplay; + dst_h = radeon_crtc->native_mode.hdisplay; /* fix up for overscan on hdmi */ if (ASIC_IS_AVIVO(rdev) && -- cgit v0.10.2 From aa74fbb4c905c6c746b79d4d7d8c95d8bbd4360c Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 7 Sep 2010 14:41:30 -0400 Subject: drm/radeon/kms: add connector table for Mac x800 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=28671 Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c index bd74e42..a04b7a6 100644 --- a/drivers/gpu/drm/radeon/radeon_combios.c +++ b/drivers/gpu/drm/radeon/radeon_combios.c @@ -1485,6 +1485,11 @@ bool radeon_get_legacy_connector_info_from_table(struct drm_device *dev) /* PowerMac8,1 ? */ /* imac g5 isight */ rdev->mode_info.connector_table = CT_IMAC_G5_ISIGHT; + } else if ((rdev->pdev->device == 0x4a48) && + (rdev->pdev->subsystem_vendor == 0x1002) && + (rdev->pdev->subsystem_device == 0x4a48)) { + /* Mac X800 */ + rdev->mode_info.connector_table = CT_MAC_X800; } else #endif /* CONFIG_PPC_PMAC */ #ifdef CONFIG_PPC64 @@ -1961,6 +1966,48 @@ bool radeon_get_legacy_connector_info_from_table(struct drm_device *dev) CONNECTOR_OBJECT_ID_VGA, &hpd); break; + case CT_MAC_X800: + DRM_INFO("Connector Table: %d (mac x800)\n", + rdev->mode_info.connector_table); + /* DVI - primary dac, internal tmds */ + ddc_i2c = combios_setup_i2c_bus(rdev, DDC_DVI, 0, 0); + hpd.hpd = RADEON_HPD_1; /* ??? */ + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_DFP1_SUPPORT, + 0), + ATOM_DEVICE_DFP1_SUPPORT); + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_CRT1_SUPPORT, + 1), + ATOM_DEVICE_CRT1_SUPPORT); + radeon_add_legacy_connector(dev, 0, + ATOM_DEVICE_DFP1_SUPPORT | + ATOM_DEVICE_CRT1_SUPPORT, + DRM_MODE_CONNECTOR_DVII, &ddc_i2c, + CONNECTOR_OBJECT_ID_SINGLE_LINK_DVI_I, + &hpd); + /* DVI - tv dac, dvo */ + ddc_i2c = combios_setup_i2c_bus(rdev, DDC_MONID, 0, 0); + hpd.hpd = RADEON_HPD_2; /* ??? */ + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_DFP2_SUPPORT, + 0), + ATOM_DEVICE_DFP2_SUPPORT); + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_CRT2_SUPPORT, + 2), + ATOM_DEVICE_CRT2_SUPPORT); + radeon_add_legacy_connector(dev, 1, + ATOM_DEVICE_DFP2_SUPPORT | + ATOM_DEVICE_CRT2_SUPPORT, + DRM_MODE_CONNECTOR_DVII, &ddc_i2c, + CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_I, + &hpd); + break; default: DRM_INFO("Connector table: %d (invalid)\n", rdev->mode_info.connector_table); diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index efbe975..17a6602 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -204,7 +204,7 @@ struct radeon_i2c_chan { /* mostly for macs, but really any system without connector tables */ enum radeon_connector_table { - CT_NONE, + CT_NONE = 0, CT_GENERIC, CT_IBOOK, CT_POWERBOOK_EXTERNAL, @@ -215,6 +215,7 @@ enum radeon_connector_table { CT_IMAC_G5_ISIGHT, CT_EMAC, CT_RN50_POWER, + CT_MAC_X800, }; enum radeon_dvo_chip { -- cgit v0.10.2 From e6db0da02ea753968d15ae3e835059c207647e78 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 10 Sep 2010 03:19:05 -0400 Subject: drm/radeon/kms: don't enable underscan with interlaced modes They aren't compatible. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 8c987c9..127a395f 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1151,6 +1151,7 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc, /* fix up for overscan on hdmi */ if (ASIC_IS_AVIVO(rdev) && + (!(mode->flags & DRM_MODE_FLAG_INTERLACE)) && ((radeon_encoder->underscan_type == UNDERSCAN_ON) || ((radeon_encoder->underscan_type == UNDERSCAN_AUTO) && drm_detect_hdmi_monitor(radeon_connector->edid) && -- cgit v0.10.2 From 356ad3cd616185631235ffb48b3efbf39f9923b3 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 9 Sep 2010 09:41:32 +0100 Subject: drm: Only decouple the old_fb from the crtc is we call mode_set* Otherwise when disabling the output we switch to the new fb (which is likely NULL) and skip the call to mode_set -- leaking driver private state on the old_fb. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29857 Reported-by: Sitsofe Wheeler Signed-off-by: Chris Wilson Cc: Dave Airlie Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index 4238a19..de152a5 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -637,13 +637,13 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) mode_changed = true; if (mode_changed) { - old_fb = set->crtc->fb; - set->crtc->fb = set->fb; set->crtc->enabled = (set->mode != NULL); if (set->mode != NULL) { DRM_DEBUG_KMS("attempting to set mode from" " userspace\n"); drm_mode_debug_printmodeline(set->mode); + old_fb = set->crtc->fb; + set->crtc->fb = set->fb; if (!drm_crtc_helper_set_mode(set->crtc, set->mode, set->x, set->y, old_fb)) { -- cgit v0.10.2 From 27849044ca6ff9c52f63271b511282acf6d1c251 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 9 Sep 2010 11:31:13 -0400 Subject: drm/radeon: add some missing copyright headers Noticed while adding evergreen blit support. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c index d13622a..9ceb2a1 100644 --- a/drivers/gpu/drm/radeon/r600_blit_kms.c +++ b/drivers/gpu/drm/radeon/r600_blit_kms.c @@ -1,3 +1,28 @@ +/* + * Copyright 2009 Advanced Micro Devices, Inc. + * Copyright 2009 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + */ + #include "drmP.h" #include "drm.h" #include "radeon_drm.h" diff --git a/drivers/gpu/drm/radeon/r600_blit_shaders.h b/drivers/gpu/drm/radeon/r600_blit_shaders.h index fdc3b37..f437d36 100644 --- a/drivers/gpu/drm/radeon/r600_blit_shaders.h +++ b/drivers/gpu/drm/radeon/r600_blit_shaders.h @@ -1,3 +1,27 @@ +/* + * Copyright 2009 Advanced Micro Devices, Inc. + * Copyright 2009 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + */ #ifndef R600_BLIT_SHADERS_H #define R600_BLIT_SHADERS_H -- cgit v0.10.2 From 7b334fcb45b757ffb093696ca3de1b0c8b4a33f1 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 9 Sep 2010 23:51:02 +0100 Subject: drm: Use a nondestructive mode for output detect when polling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Destructive load-detection is very expensive and due to failings elsewhere can trigger system wide stalls of up to 600ms. A simple first step to correcting this is not to invoke such an expensive and destructive load-detection operation automatically. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29536 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=16265 Reported-by: Bruno Prémont Tested-by: Sitsofe Wheeler Signed-off-by: Chris Wilson Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index de152a5..fb6b70f 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -103,7 +103,7 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector, if (connector->funcs->force) connector->funcs->force(connector); } else { - connector->status = connector->funcs->detect(connector); + connector->status = connector->funcs->detect(connector, false); drm_kms_helper_poll_enable(dev); } @@ -866,7 +866,7 @@ static void output_poll_execute(struct work_struct *work) !(connector->polled & DRM_CONNECTOR_POLL_HPD)) continue; - status = connector->funcs->detect(connector); + status = connector->funcs->detect(connector, true); if (old_status != status) changed = true; } diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c index 86118a7..85da4c4 100644 --- a/drivers/gpu/drm/drm_sysfs.c +++ b/drivers/gpu/drm/drm_sysfs.c @@ -159,7 +159,7 @@ static ssize_t status_show(struct device *device, struct drm_connector *connector = to_drm_connector(device); enum drm_connector_status status; - status = connector->funcs->detect(connector); + status = connector->funcs->detect(connector, true); return snprintf(buf, PAGE_SIZE, "%s\n", drm_get_connector_status_name(status)); } diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 4b77351..0350e5d 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -400,7 +400,9 @@ intel_crt_load_detect(struct drm_crtc *crtc, struct intel_encoder *intel_encoder return status; } -static enum drm_connector_status intel_crt_detect(struct drm_connector *connector) +static enum drm_connector_status +intel_crt_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_device *dev = connector->dev; struct drm_encoder *encoder = intel_attached_encoder(connector); @@ -419,6 +421,9 @@ static enum drm_connector_status intel_crt_detect(struct drm_connector *connecto if (intel_crt_detect_ddc(encoder)) return connector_status_connected; + if (nondestructive) + return connector->status; + /* for pre-945g platforms use load detect */ if (encoder->crtc && encoder->crtc->enabled) { status = intel_crt_load_detect(encoder->crtc, intel_encoder); diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 51d1429..e1a2a05 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1386,7 +1386,8 @@ ironlake_dp_detect(struct drm_connector *connector) * \return false if DP port is disconnected. */ static enum drm_connector_status -intel_dp_detect(struct drm_connector *connector) +intel_dp_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_encoder *encoder = intel_attached_encoder(connector); struct intel_dp *intel_dp = enc_to_intel_dp(encoder); diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c index a399f4b..f0de1ad 100644 --- a/drivers/gpu/drm/i915/intel_dvo.c +++ b/drivers/gpu/drm/i915/intel_dvo.c @@ -221,7 +221,9 @@ static void intel_dvo_mode_set(struct drm_encoder *encoder, * * Unimplemented. */ -static enum drm_connector_status intel_dvo_detect(struct drm_connector *connector) +static enum drm_connector_status +intel_dvo_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_encoder *encoder = intel_attached_encoder(connector); struct intel_dvo *intel_dvo = enc_to_intel_dvo(encoder); diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index ccd4c97..2ea123d 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -139,7 +139,8 @@ static bool intel_hdmi_mode_fixup(struct drm_encoder *encoder, } static enum drm_connector_status -intel_hdmi_detect(struct drm_connector *connector) +intel_hdmi_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_encoder *encoder = intel_attached_encoder(connector); struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder); diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 4fbb016..fb1bed8 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -445,7 +445,9 @@ static void intel_lvds_mode_set(struct drm_encoder *encoder, * connected and closed means disconnected. We also send hotplug events as * needed, using lid status notification from the input layer. */ -static enum drm_connector_status intel_lvds_detect(struct drm_connector *connector) +static enum drm_connector_status +intel_lvds_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_device *dev = connector->dev; enum drm_connector_status status = connector_status_connected; @@ -540,7 +542,9 @@ static int intel_lid_notify(struct notifier_block *nb, unsigned long val, * the LID nofication event. */ if (connector) - connector->status = connector->funcs->detect(connector); + connector->status = connector->funcs->detect(connector, + true); + /* Don't force modeset on machines where it causes a GPU lockup */ if (dmi_check_system(intel_no_modeset_on_lid)) return NOTIFY_OK; diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index e3b7a7e..db6b6d4 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -1417,7 +1417,7 @@ intel_analog_is_connected(struct drm_device *dev) if (!analog_connector) return false; - if (analog_connector->funcs->detect(analog_connector) == + if (analog_connector->funcs->detect(analog_connector, true) == connector_status_disconnected) return false; @@ -1486,7 +1486,9 @@ intel_sdvo_hdmi_sink_detect(struct drm_connector *connector) return status; } -static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connector) +static enum drm_connector_status +intel_sdvo_detect(struct drm_connector *connector, + bool nondestructive) { uint16_t response; struct drm_encoder *encoder = intel_attached_encoder(connector); diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index c671f60..d20b550 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1341,7 +1341,8 @@ static void intel_tv_find_better_format(struct drm_connector *connector) * we have a pipe programmed in order to probe the TV. */ static enum drm_connector_status -intel_tv_detect(struct drm_connector *connector) +intel_tv_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_display_mode mode; struct drm_encoder *encoder = intel_attached_encoder(connector); @@ -1353,7 +1354,7 @@ intel_tv_detect(struct drm_connector *connector) if (encoder->crtc && encoder->crtc->enabled) { type = intel_tv_detect_type(intel_tv); - } else { + } else if (nondestructive) { struct drm_crtc *crtc; int dpms_mode; @@ -1364,10 +1365,9 @@ intel_tv_detect(struct drm_connector *connector) intel_release_load_detect_pipe(&intel_tv->base, connector, dpms_mode); } else - type = -1; - } - - intel_tv->type = type; + return connector_status_unknown; + } else + return connector->status; if (type < 0) return connector_status_disconnected; diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c index a1473ff..67d515c 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.c +++ b/drivers/gpu/drm/nouveau/nouveau_connector.c @@ -168,7 +168,8 @@ nouveau_connector_set_encoder(struct drm_connector *connector, } static enum drm_connector_status -nouveau_connector_detect(struct drm_connector *connector) +nouveau_connector_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_device *dev = connector->dev; struct nouveau_connector *nv_connector = nouveau_connector(connector); @@ -246,7 +247,8 @@ detect_analog: } static enum drm_connector_status -nouveau_connector_detect_lvds(struct drm_connector *connector) +nouveau_connector_detect_lvds(struct drm_connector *connector, + bool nondestructive) { struct drm_device *dev = connector->dev; struct drm_nouveau_private *dev_priv = dev->dev_private; @@ -267,7 +269,7 @@ nouveau_connector_detect_lvds(struct drm_connector *connector) /* Try retrieving EDID via DDC */ if (!dev_priv->vbios.fp_no_ddc) { - status = nouveau_connector_detect(connector); + status = nouveau_connector_detect(connector, nondestructive); if (status == connector_status_connected) goto out; } diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index a9dd784..31d309a 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -481,7 +481,9 @@ static int radeon_lvds_mode_valid(struct drm_connector *connector, return MODE_OK; } -static enum drm_connector_status radeon_lvds_detect(struct drm_connector *connector) +static enum drm_connector_status +radeon_lvds_detect(struct drm_connector *connector, + bool nondestructive) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct drm_encoder *encoder = radeon_best_single_encoder(connector); @@ -594,7 +596,9 @@ static int radeon_vga_mode_valid(struct drm_connector *connector, return MODE_OK; } -static enum drm_connector_status radeon_vga_detect(struct drm_connector *connector) +static enum drm_connector_status +radeon_vga_detect(struct drm_connector *connector, + bool nondestructive) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct drm_encoder *encoder; @@ -691,7 +695,9 @@ static int radeon_tv_mode_valid(struct drm_connector *connector, return MODE_OK; } -static enum drm_connector_status radeon_tv_detect(struct drm_connector *connector) +static enum drm_connector_status +radeon_tv_detect(struct drm_connector *connector, + bool nondestructive) { struct drm_encoder *encoder; struct drm_encoder_helper_funcs *encoder_funcs; @@ -748,7 +754,9 @@ static int radeon_dvi_get_modes(struct drm_connector *connector) * we have to check if this analog encoder is shared with anyone else (TV) * if its shared we have to set the other connector to disconnected. */ -static enum drm_connector_status radeon_dvi_detect(struct drm_connector *connector) +static enum drm_connector_status +radeon_dvi_detect(struct drm_connector *connector, + bool nondestructive) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct drm_encoder *encoder = NULL; @@ -972,7 +980,9 @@ static int radeon_dp_get_modes(struct drm_connector *connector) return ret; } -static enum drm_connector_status radeon_dp_detect(struct drm_connector *connector) +static enum drm_connector_status +radeon_dp_detect(struct drm_connector *connector, + bool nondestructive) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); enum drm_connector_status ret = connector_status_disconnected; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c index 2ff5cf7..a527c91 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c @@ -335,7 +335,8 @@ static void vmw_ldu_connector_restore(struct drm_connector *connector) } static enum drm_connector_status - vmw_ldu_connector_detect(struct drm_connector *connector) + vmw_ldu_connector_detect(struct drm_connector *connector, + bool nondestructive) { if (vmw_connector_to_ldu(connector)->pref_active) return connector_status_connected; diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index c9f3cc5..5536223 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -386,7 +386,8 @@ struct drm_connector_funcs { void (*dpms)(struct drm_connector *connector, int mode); void (*save)(struct drm_connector *connector); void (*restore)(struct drm_connector *connector); - enum drm_connector_status (*detect)(struct drm_connector *connector); + enum drm_connector_status (*detect)(struct drm_connector *connector, + bool nondestructive); int (*fill_modes)(struct drm_connector *connector, uint32_t max_width, uint32_t max_height); int (*set_property)(struct drm_connector *connector, struct drm_property *property, uint64_t val); -- cgit v0.10.2 From b741be82cf2079f71553af595610f17a3a3a752a Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 9 Sep 2010 19:15:23 -0400 Subject: drm/radeon/kms/evergreen: fix backend setup This patch fixes rendering errors on some evergreen boards. Hardcoding the backend map is not an optimal solution, but a better fix is being worked on. Similar to the fix for rv740 (6271901d828b34b27607314026deaf417f9f9b75). Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=29986 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index b8b7f01..79082d4 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1160,14 +1160,25 @@ static void evergreen_gpu_init(struct radeon_device *rdev) EVERGREEN_MAX_BACKENDS_MASK)); break; } - } else - gb_backend_map = - evergreen_get_tile_pipe_to_backend_map(rdev, - rdev->config.evergreen.max_tile_pipes, - rdev->config.evergreen.max_backends, - ((EVERGREEN_MAX_BACKENDS_MASK << - rdev->config.evergreen.max_backends) & - EVERGREEN_MAX_BACKENDS_MASK)); + } else { + switch (rdev->family) { + case CHIP_CYPRESS: + case CHIP_HEMLOCK: + gb_backend_map = 0x66442200; + break; + case CHIP_JUNIPER: + gb_backend_map = 0x00006420; + break; + default: + gb_backend_map = + evergreen_get_tile_pipe_to_backend_map(rdev, + rdev->config.evergreen.max_tile_pipes, + rdev->config.evergreen.max_backends, + ((EVERGREEN_MAX_BACKENDS_MASK << + rdev->config.evergreen.max_backends) & + EVERGREEN_MAX_BACKENDS_MASK)); + } + } rdev->config.evergreen.tile_config = gb_addr_config; WREG32(GB_BACKEND_MAP, gb_backend_map); -- cgit v0.10.2 From ec00efb72f4b88078427d01f38f664c67c7ca0c0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Marek=20Ol=C5=A1=C3=A1k?= Date: Sun, 12 Sep 2010 05:09:12 +0200 Subject: drm/radeon/kms: increase lockup detection interval to 10 sec for r100-r500 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit One subtest of mesa/demos/gltestperf takes 9 seconds to complete, so to prevent an unnecessary gpu reset followed by a hardlock, I am increasing the interval to 10 seconds after which a GPU is considered in a locked-up state. This is on RV530. However, with a little slower GPU, we would surpass the interval easily, so this is not a good fix for gltestperf. Nevertheless, this commit also fixes hardlocks in the applications which render at speed of less than 1 frame per second, where the whole frame consists of only one command stream. The game Tiny & Big is an example. This bar is now lowered to 0.1 fps. Now the question comes down to whether we should (often unsuccessfully) reset the GPU at all? Once we have stable enough drivers, we won't have to. Has the time come already? If possible, this commit should go to stable as well. Signed-off-by: Marek Olšák Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index e817a0b..ec64b36 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -2020,18 +2020,7 @@ bool r100_gpu_cp_is_lockup(struct radeon_device *rdev, struct r100_gpu_lockup *l return false; } elapsed = jiffies_to_msecs(cjiffies - lockup->last_jiffies); - if (elapsed >= 3000) { - /* very likely the improbable case where current - * rptr is equal to last recorded, a while ago, rptr - * this is more likely a false positive update tracking - * information which should force us to be recall at - * latter point - */ - lockup->last_cp_rptr = cp->rptr; - lockup->last_jiffies = jiffies; - return false; - } - if (elapsed >= 1000) { + if (elapsed >= 10000) { dev_err(rdev->dev, "GPU lockup CP stall for more than %lumsec\n", elapsed); return true; } -- cgit v0.10.2 From a41ceb1c17af06a17c0d88e987215ef20b93c471 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Marek=20Ol=C5=A1=C3=A1k?= Date: Sun, 12 Sep 2010 05:09:13 +0200 Subject: drm/radeon/kms: fix the colorbuffer CS checker for r300-r500 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This commit fixes bogus CS rejection if it contains a sequence of the following operations: - Set the color buffer 0. track->cb[i].robj becomes non-NULL. - Render. - Set a larger zbuffer than the previously-set color buffer. - Set a larger scissor area as well. - Set the color channel mask to 0 to do depth-only rendering. - Render. --> rejected, because track->cb[i].robj remained non-NULL, therefore the conditional checking for the color channel mask and friends is not performed, and the larger scissor area causes the rejection. This fixes bugs: - https://bugs.freedesktop.org/show_bug.cgi?id=29762 - https://bugs.freedesktop.org/show_bug.cgi?id=28869 And maybe some others which seem to look the same. If possible, this commit should go to stable as well. Signed-off-by: Marek Olšák Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index ec64b36..e151f16 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -3297,13 +3297,14 @@ int r100_cs_track_check(struct radeon_device *rdev, struct r100_cs_track *track) unsigned long size; unsigned prim_walk; unsigned nverts; + unsigned num_cb = track->num_cb; - for (i = 0; i < track->num_cb; i++) { + if (!track->zb_cb_clear && !track->color_channel_mask && + !track->blend_read_enable) + num_cb = 0; + + for (i = 0; i < num_cb; i++) { if (track->cb[i].robj == NULL) { - if (!(track->zb_cb_clear || track->color_channel_mask || - track->blend_read_enable)) { - continue; - } DRM_ERROR("[drm] No buffer for color buffer %d !\n", i); return -EINVAL; } -- cgit v0.10.2 From 2ca9cac965e81da4b74f2dcec4b87ebfd106b357 Mon Sep 17 00:00:00 2001 From: Anisse Astier Date: Fri, 10 Sep 2010 15:47:55 +0200 Subject: ALSA: hda - Add quirk for Toshiba C650D using a Conexant CX20585 Add a quirk for laptop Toshiba Satellite C650D to have proper external HP and external Mic support. Signed-off-by: Anisse Astier Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c index 71f9d64..972e7c4 100644 --- a/sound/pci/hda/patch_conexant.c +++ b/sound/pci/hda/patch_conexant.c @@ -3092,6 +3092,7 @@ static struct snd_pci_quirk cxt5066_cfg_tbl[] = { SND_PCI_QUIRK(0x1028, 0x0402, "Dell Vostro", CXT5066_DELL_VOSTO), SND_PCI_QUIRK(0x1028, 0x0408, "Dell Inspiron One 19T", CXT5066_IDEAPAD), SND_PCI_QUIRK(0x103c, 0x360b, "HP G60", CXT5066_HP_LAPTOP), + SND_PCI_QUIRK(0x1179, 0xff1e, "Toshiba Satellite C650D", CXT5066_IDEAPAD), SND_PCI_QUIRK(0x1179, 0xff50, "Toshiba Satellite P500-PSPGSC-01800T", CXT5066_OLPC_XO_1_5), SND_PCI_QUIRK(0x1179, 0xffe0, "Toshiba Satellite Pro T130-15F", CXT5066_OLPC_XO_1_5), SND_PCI_QUIRK(0x17aa, 0x20f2, "Lenovo T400s", CXT5066_THINKPAD), -- cgit v0.10.2 From 62b2be591a9b12c550308ef7718a31abfc815b50 Mon Sep 17 00:00:00 2001 From: Latchesar Ionkov Date: Tue, 24 Aug 2010 18:13:59 +0000 Subject: fs/9p, net/9p: memory leak fixes Four memory leak fixes in the 9P code. Signed-off-by: Latchesar Ionkov Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index c7c23ea..84159cf 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -1128,6 +1128,7 @@ v9fs_vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, v9fs_stat2inode(st, dentry->d_inode, dentry->d_inode->i_sb); generic_fillattr(dentry->d_inode, stat); + p9stat_free(st); kfree(st); return 0; } @@ -1489,6 +1490,7 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen) retval = strnlen(buffer, buflen); done: + p9stat_free(st); kfree(st); return retval; } diff --git a/net/9p/client.c b/net/9p/client.c index dc6f2f2..9eb7250 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -331,8 +331,10 @@ static void p9_tag_cleanup(struct p9_client *c) } } - if (c->tagpool) + if (c->tagpool) { + p9_idpool_put(0, c->tagpool); /* free reserved tag 0 */ p9_idpool_destroy(c->tagpool); + } /* free requests associated with tags */ for (row = 0; row < (c->max_tag/P9_ROW_MAXTAG); row++) { @@ -944,6 +946,7 @@ struct p9_fid *p9_client_walk(struct p9_fid *oldfid, int nwname, char **wnames, int16_t nwqids, count; err = 0; + wqids = NULL; clnt = oldfid->clnt; if (clone) { fid = p9_fid_create(clnt); @@ -994,9 +997,11 @@ struct p9_fid *p9_client_walk(struct p9_fid *oldfid, int nwname, char **wnames, else fid->qid = oldfid->qid; + kfree(wqids); return fid; clunk_fid: + kfree(wqids); p9_client_clunk(fid); fid = NULL; -- cgit v0.10.2 From 5c25f347a7b00b2ebe0a55c4a3cfe4c3e1e8725e Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Tue, 24 Aug 2010 10:30:49 +0000 Subject: fs/9p: Fix error handling in v9fs_get_sb This was introduced by 7cadb63d58a932041afa3f957d5cbb6ce69dcee5 Signed-off-by: Aneesh Kumar K.V Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index f931107..1d12ba0 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -122,6 +122,10 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, fid = v9fs_session_init(v9ses, dev_name, data); if (IS_ERR(fid)) { retval = PTR_ERR(fid); + /* + * we need to call session_close to tear down some + * of the data structure setup by session_init + */ goto close_session; } @@ -144,7 +148,6 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, retval = -ENOMEM; goto release_sb; } - sb->s_root = root; if (v9fs_proto_dotl(v9ses)) { @@ -152,7 +155,7 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); if (IS_ERR(st)) { retval = PTR_ERR(st); - goto clunk_fid; + goto release_sb; } v9fs_stat2inode_dotl(st, root->d_inode); @@ -162,7 +165,7 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, st = p9_client_stat(fid); if (IS_ERR(st)) { retval = PTR_ERR(st); - goto clunk_fid; + goto release_sb; } root->d_inode->i_ino = v9fs_qid2ino(&st->qid); @@ -174,19 +177,24 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, v9fs_fid_add(root, fid); -P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); + P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); simple_set_mnt(mnt, sb); return 0; clunk_fid: p9_client_clunk(fid); - close_session: v9fs_session_close(v9ses); kfree(v9ses); return retval; - release_sb: + /* + * we will do the session_close and root dentry release + * in the below call. But we need to clunk fid, because we haven't + * attached the fid to dentry so it won't get clunked + * automatically. + */ + p9_client_clunk(fid); deactivate_locked_super(sb); return retval; } -- cgit v0.10.2 From 62726a7ab3a6a3624256172af055ff0a38c6ffa2 Mon Sep 17 00:00:00 2001 From: jvrao Date: Wed, 25 Aug 2010 16:26:21 +0000 Subject: 9p: Check for NULL fid in v9fs_dir_release() NULL fid should be handled in cases where we endup calling v9fs_dir_release() before even we instantiate the fid in filp. Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index 16c8a2a..899f168 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -292,9 +292,11 @@ int v9fs_dir_release(struct inode *inode, struct file *filp) fid = filp->private_data; P9_DPRINTK(P9_DEBUG_VFS, - "inode: %p filp: %p fid: %d\n", inode, filp, fid->fid); + "v9fs_dir_release: inode: %p filp: %p fid: %d\n", + inode, filp, fid ? fid->fid : -1); filemap_write_and_wait(inode->i_mapping); - p9_client_clunk(fid); + if (fid) + p9_client_clunk(fid); return 0; } -- cgit v0.10.2 From 3c30750ffafbc32af040b09f777b67aa2486b063 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Mon, 30 Aug 2010 16:04:35 +0000 Subject: fs/9p: Use the correct dentry operations We should use the cached dentry operation only if caching mode is enabled Signed-off-by: Aneesh Kumar K.V Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 84159cf..a6990bb 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -730,7 +730,10 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int mode, P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err); goto error; } - dentry->d_op = &v9fs_cached_dentry_operations; + if (v9ses->cache) + dentry->d_op = &v9fs_cached_dentry_operations; + else + dentry->d_op = &v9fs_dentry_operations; d_instantiate(dentry, inode); err = v9fs_fid_add(dentry, fid); if (err < 0) -- cgit v0.10.2 From 1d76e3135733a06aa12bb35891c05f306b27b2d6 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Mon, 30 Aug 2010 17:43:07 +0000 Subject: fs/9p: Don't use dotl version of mknod for dotu inode operations We should not use dotlversion for the dotu inode operations Signed-off-by: Aneesh Kumar K.V Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index a6990bb..9e670d5 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -1947,7 +1947,7 @@ static const struct inode_operations v9fs_dir_inode_operations_dotu = { .unlink = v9fs_vfs_unlink, .mkdir = v9fs_vfs_mkdir, .rmdir = v9fs_vfs_rmdir, - .mknod = v9fs_vfs_mknod_dotl, + .mknod = v9fs_vfs_mknod, .rename = v9fs_vfs_rename, .getattr = v9fs_vfs_getattr, .setattr = v9fs_vfs_setattr, -- cgit v0.10.2 From 0bf377bbb0bea6130f35613491887cc622e42a8b Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sun, 12 Sep 2010 08:14:52 +0200 Subject: sched: Improve latencies under load by decreasing minimum scheduling granularity MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Mathieu reported bad latencies with make -j10 kind of kbuild workloads - which is mostly caused by us scheduling with a too coarse granularity. Reduce the minimum granularity some more, to make sure we can meet the latency target. I got the following results (make -j10 kbuild load, average of 3 runs): vanilla: maximum latency: 38278.9 µs average latency: 7730.1 µs patched: maximum latency: 22702.1 µs average latency: 6684.8 µs Mathieu also measured it: | | * wakeup-latency.c (SIGEV_THREAD) with make -j10 | | - Mainline 2.6.35.2 kernel | | maximum latency: 45762.1 µs | average latency: 7348.6 µs | | - With only Peter's smaller min_gran (shown below): | | maximum latency: 29100.6 µs | average latency: 6684.1 µs | Reported-by: Mathieu Desnoyers Reported-by: Linus Torvalds Acked-by: Mathieu Desnoyers Suggested-by: Peter Zijlstra Acked-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 9b5b4f8..a171138 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -54,13 +54,13 @@ enum sched_tunable_scaling sysctl_sched_tunable_scaling * Minimal preemption granularity for CPU-bound tasks: * (default: 2 msec * (1 + ilog(ncpus)), units: nanoseconds) */ -unsigned int sysctl_sched_min_granularity = 2000000ULL; -unsigned int normalized_sysctl_sched_min_granularity = 2000000ULL; +unsigned int sysctl_sched_min_granularity = 750000ULL; +unsigned int normalized_sysctl_sched_min_granularity = 750000ULL; /* * is kept at sysctl_sched_latency / sysctl_sched_min_granularity */ -static unsigned int sched_nr_latency = 3; +static unsigned int sched_nr_latency = 8; /* * After fork, child runs first. If set to 0 (default) then -- cgit v0.10.2 From edc805b7c5e45c3d7202cea5239985d204178e24 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Mon, 16 Aug 2010 14:05:20 +0200 Subject: m68k,m68knommu: Wire up fanotify_init, fanotify_mark, and prlimit64 Signed-off-by: Geert Uytterhoeven Acked-by: Greg Ungerer diff --git a/arch/m68k/include/asm/unistd.h b/arch/m68k/include/asm/unistd.h index 60b15d0..b43b36b 100644 --- a/arch/m68k/include/asm/unistd.h +++ b/arch/m68k/include/asm/unistd.h @@ -340,10 +340,13 @@ #define __NR_set_thread_area 334 #define __NR_atomic_cmpxchg_32 335 #define __NR_atomic_barrier 336 +#define __NR_fanotify_init 337 +#define __NR_fanotify_mark 338 +#define __NR_prlimit64 339 #ifdef __KERNEL__ -#define NR_syscalls 337 +#define NR_syscalls 340 #define __ARCH_WANT_IPC_PARSE_VERSION #define __ARCH_WANT_OLD_READDIR diff --git a/arch/m68k/kernel/entry.S b/arch/m68k/kernel/entry.S index 2391bdf..6360c43 100644 --- a/arch/m68k/kernel/entry.S +++ b/arch/m68k/kernel/entry.S @@ -765,4 +765,7 @@ sys_call_table: .long sys_set_thread_area .long sys_atomic_cmpxchg_32 /* 335 */ .long sys_atomic_barrier + .long sys_fanotify_init + .long sys_fanotify_mark + .long sys_prlimit64 diff --git a/arch/m68knommu/kernel/syscalltable.S b/arch/m68knommu/kernel/syscalltable.S index b30b3eb..79b1ed1 100644 --- a/arch/m68knommu/kernel/syscalltable.S +++ b/arch/m68knommu/kernel/syscalltable.S @@ -355,6 +355,9 @@ ENTRY(sys_call_table) .long sys_set_thread_area .long sys_atomic_cmpxchg_32 /* 335 */ .long sys_atomic_barrier + .long sys_fanotify_init + .long sys_fanotify_mark + .long sys_prlimit64 .rept NR_syscalls-(.-sys_call_table)/4 .long sys_ni_syscall -- cgit v0.10.2 From 467c525109d5d542d7d416b0c11bdd54610fe2f4 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Mon, 13 Sep 2010 11:39:20 -0700 Subject: ceph: fix dn offset during readdir_prepopulate When adding the readdir results to the cache, ceph_set_dentry_offset was clobbered our just-set offset. This can cause the readdir result offsets to get out of sync with the server. Add an argument to the helper so that it does not. This bug was introduced by 1cd3935bedccf592d44343890251452a6dd74fc4. Signed-off-by: Sage Weil diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index e7cca414..62377ec 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -845,7 +845,7 @@ static void ceph_set_dentry_offset(struct dentry *dn) * the caller) if we fail. */ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in, - bool *prehash) + bool *prehash, bool set_offset) { struct dentry *realdn; @@ -877,7 +877,8 @@ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in, } if ((!prehash || *prehash) && d_unhashed(dn)) d_rehash(dn); - ceph_set_dentry_offset(dn); + if (set_offset) + ceph_set_dentry_offset(dn); out: return dn; } @@ -1062,7 +1063,7 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, d_delete(dn); goto done; } - dn = splice_dentry(dn, in, &have_lease); + dn = splice_dentry(dn, in, &have_lease, true); if (IS_ERR(dn)) { err = PTR_ERR(dn); goto done; @@ -1105,7 +1106,7 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, goto done; } dout(" linking snapped dir %p to dn %p\n", in, dn); - dn = splice_dentry(dn, in, NULL); + dn = splice_dentry(dn, in, NULL, true); if (IS_ERR(dn)) { err = PTR_ERR(dn); goto done; @@ -1237,7 +1238,7 @@ retry_lookup: err = PTR_ERR(in); goto out; } - dn = splice_dentry(dn, in, NULL); + dn = splice_dentry(dn, in, NULL, false); if (IS_ERR(dn)) dn = NULL; } -- cgit v0.10.2 From 05ed160e89baf7f5fb3432d527fec467d2493626 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 13 Sep 2010 19:25:41 +0900 Subject: kprobes: Fix Kconfig dependency Fix Kconfig dependency among Kprobes, optprobe and kallsyms. Kprobes uses kallsyms_lookup for finding target function and checking instruction boundary, thus CONFIG_KPROBES should select CONFIG_KALLSYMS. Optprobe is an optional feature which is supported on x86 arch, and it also uses kallsyms_lookup for checking instructions in the target function. Since KALLSYMS_ALL just adds symbols of kernel variables, it doesn't need to select KALLSYMS_ALL. Signed-off-by: Masami Hiramatsu Acked-by: Randy Dunlap , Cc: Ananth N Mavinakayanahalli Cc: Felipe Contreras Cc: Randy Dunlap Cc: Ananth N Mavinakayanahalli Cc: Anil S Keshavamurthy Cc: David S. Miller Cc: akpm LKML-Reference: <20100913102541.20260.85700.stgit@ltc236.sdl.hitachi.co.jp> Signed-off-by: Ingo Molnar diff --git a/arch/Kconfig b/arch/Kconfig index 4877a8c..fe48fc7 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -32,8 +32,9 @@ config HAVE_OPROFILE config KPROBES bool "Kprobes" - depends on KALLSYMS && MODULES + depends on MODULES depends on HAVE_KPROBES + select KALLSYMS help Kprobes allows you to trap at almost any kernel address and execute a callback function. register_kprobe() establishes @@ -45,7 +46,6 @@ config OPTPROBES def_bool y depends on KPROBES && HAVE_OPTPROBES depends on !PREEMPT - select KALLSYMS_ALL config HAVE_EFFICIENT_UNALIGNED_ACCESS bool -- cgit v0.10.2 From 147fcf1c211f1a87bf4d0711b7e9637f3d6ce080 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Sat, 11 Sep 2010 22:10:59 -0700 Subject: sound: Remove pr_ uses of KERN_ Signed-off-by: Joe Perches Acked-by: Mark Brown Acked-by: Geoff Levand Signed-off-by: Takashi Iwai diff --git a/sound/ppc/snd_ps3.c b/sound/ppc/snd_ps3.c index 2f12da4..581a670 100644 --- a/sound/ppc/snd_ps3.c +++ b/sound/ppc/snd_ps3.c @@ -579,7 +579,7 @@ static int snd_ps3_delay_to_bytes(struct snd_pcm_substream *substream, rate * delay_ms / 1000) * substream->runtime->channels; - pr_debug(KERN_ERR "%s: time=%d rate=%d bytes=%ld, frames=%d, ret=%d\n", + pr_debug("%s: time=%d rate=%d bytes=%ld, frames=%d, ret=%d\n", __func__, delay_ms, rate, diff --git a/sound/soc/s3c24xx/s3c-dma.c b/sound/soc/s3c24xx/s3c-dma.c index 1b61c23..f1b1bc4 100644 --- a/sound/soc/s3c24xx/s3c-dma.c +++ b/sound/soc/s3c24xx/s3c-dma.c @@ -94,8 +94,7 @@ static void s3c_dma_enqueue(struct snd_pcm_substream *substream) if ((pos + len) > prtd->dma_end) { len = prtd->dma_end - pos; - pr_debug(KERN_DEBUG "%s: corrected dma len %ld\n", - __func__, len); + pr_debug("%s: corrected dma len %ld\n", __func__, len); } ret = s3c2410_dma_enqueue(prtd->params->channel, -- cgit v0.10.2 From 2fd818642a3c3a257f3b1262145a52bc1bfebd26 Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Mon, 30 Aug 2010 09:45:40 +0900 Subject: x86, cpufeature: Suppress compiler warning with gcc 3.x Gcc 3.x generates a warning arch/x86/include/asm/cpufeature.h: In function `__static_cpu_has': arch/x86/include/asm/cpufeature.h:326: warning: asm operand 1 probably doesn't match constraints on each file. But static_cpu_has() for gcc 3.x does not need __static_cpu_has(). Signed-off-by: Tetsuo Handa LKML-Reference: <201008300127.o7U1RC6Z044051@www262.sakura.ne.jp> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index 781a50b..c6fbb7b 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -296,6 +296,7 @@ extern const char * const x86_power_flags[32]; #endif /* CONFIG_X86_64 */ +#if __GNUC__ >= 4 /* * Static testing of CPU features. Used the same as boot_cpu_has(). * These are only valid after alternatives have run, but will statically @@ -304,7 +305,7 @@ extern const char * const x86_power_flags[32]; */ static __always_inline __pure bool __static_cpu_has(u16 bit) { -#if __GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 5) +#if __GNUC__ > 4 || __GNUC_MINOR__ >= 5 asm goto("1: jmp %l[t_no]\n" "2:\n" ".section .altinstructions,\"a\"\n" @@ -345,7 +346,6 @@ static __always_inline __pure bool __static_cpu_has(u16 bit) #endif } -#if __GNUC__ >= 4 #define static_cpu_has(bit) \ ( \ __builtin_constant_p(boot_cpu_has(bit)) ? \ -- cgit v0.10.2 From 08c2b394b98786ebb067e2a54d08f1f6f0d247da Mon Sep 17 00:00:00 2001 From: "basile@opensource.dyc.edu" Date: Mon, 13 Sep 2010 06:13:19 -0400 Subject: x86, build: Disable -fPIE when compiling with CONFIG_CC_STACKPROTECTOR=y The arch/x86/Makefile uses scripts/gcc-x86_$(BITS)-has-stack-protector.sh to check if cc1 supports -fstack-protector. When -fPIE is passed to cc1, these scripts fail causing stack protection to be disabled even when it is available. This fix is similar to commit c47efe5548abbf53c2f66e06dcb46183b11d6b22 Reported-by: Kai Dietrich Signed-off-by: Magnus Granberg LKML-Reference: <20100913101319.748A1148E216@opensource.dyc.edu> Signed-off-by: Anthony G. Basile Cc: Andrew Morton Signed-off-by: H. Peter Anvin diff --git a/arch/x86/Makefile b/arch/x86/Makefile index 8aa1b59..e8c8881 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -74,7 +74,7 @@ endif ifdef CONFIG_CC_STACKPROTECTOR cc_has_sp := $(srctree)/scripts/gcc-x86_$(BITS)-has-stack-protector.sh - ifeq ($(shell $(CONFIG_SHELL) $(cc_has_sp) $(CC) $(biarch)),y) + ifeq ($(shell $(CONFIG_SHELL) $(cc_has_sp) $(CC) $(KBUILD_CPPFLAGS) $(biarch)),y) stackp-y := -fstack-protector KBUILD_CFLAGS += $(stackp-y) else -- cgit v0.10.2 From dbee032295dac88742734ee9988e08a0e4f2f732 Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Thu, 2 Sep 2010 11:28:39 +0000 Subject: ide: Fix ordering of procfs registry. We must ensure that ide_proc_port_register_devices() occurs on an interface before ide_proc_register_driver() executes for that interfaces drives. Therefore defer the registry of the driver device objects backed by ide_bus_type until after ide_proc_port_register_devices() has run and thus all of the drive->proc procfs directory pointers have been setup. Signed-off-by: Wolfram Sang Signed-off-by: David S. Miller diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c index 4c3d1bf..068cef0 100644 --- a/drivers/ide/ide-probe.c +++ b/drivers/ide/ide-probe.c @@ -1448,19 +1448,13 @@ int ide_host_register(struct ide_host *host, const struct ide_port_info *d, if (hwif == NULL) continue; - if (hwif->present) - hwif_register_devices(hwif); - } - - ide_host_for_each_port(i, hwif, host) { - if (hwif == NULL) - continue; - ide_sysfs_register_port(hwif); ide_proc_register_port(hwif); - if (hwif->present) + if (hwif->present) { ide_proc_port_register_devices(hwif); + hwif_register_devices(hwif); + } } return j ? 0 : -1; -- cgit v0.10.2 From 0fe7f885043ea0e3e026cc98cfd7f8715018c023 Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Thu, 9 Sep 2010 21:29:52 +0900 Subject: ARM: S5P: Fix DMA coherent mask for FIMC FIMC driver uses DMA_coherent allocator, which requires proper dma mask to be set. Signed-off-by: Marek Szyprowski Signed-off-by: Kyungmin Park [kgene.kim@samsung.com: minor title fix] Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-s5p/dev-fimc0.c b/arch/arm/plat-s5p/dev-fimc0.c index d3f1a9b..f9203fa 100644 --- a/arch/arm/plat-s5p/dev-fimc0.c +++ b/arch/arm/plat-s5p/dev-fimc0.c @@ -10,6 +10,7 @@ */ #include +#include #include #include #include @@ -28,9 +29,15 @@ static struct resource s5p_fimc0_resource[] = { }, }; +static u64 s5p_fimc0_dma_mask = DMA_BIT_MASK(32); + struct platform_device s5p_device_fimc0 = { .name = "s5p-fimc", .id = 0, .num_resources = ARRAY_SIZE(s5p_fimc0_resource), .resource = s5p_fimc0_resource, + .dev = { + .dma_mask = &s5p_fimc0_dma_mask, + .coherent_dma_mask = DMA_BIT_MASK(32), + }, }; diff --git a/arch/arm/plat-s5p/dev-fimc1.c b/arch/arm/plat-s5p/dev-fimc1.c index 41bd698..7ed91a3 100644 --- a/arch/arm/plat-s5p/dev-fimc1.c +++ b/arch/arm/plat-s5p/dev-fimc1.c @@ -10,6 +10,7 @@ */ #include +#include #include #include #include @@ -28,9 +29,15 @@ static struct resource s5p_fimc1_resource[] = { }, }; +static u64 s5p_fimc1_dma_mask = DMA_BIT_MASK(32); + struct platform_device s5p_device_fimc1 = { .name = "s5p-fimc", .id = 1, .num_resources = ARRAY_SIZE(s5p_fimc1_resource), .resource = s5p_fimc1_resource, + .dev = { + .dma_mask = &s5p_fimc1_dma_mask, + .coherent_dma_mask = DMA_BIT_MASK(32), + }, }; diff --git a/arch/arm/plat-s5p/dev-fimc2.c b/arch/arm/plat-s5p/dev-fimc2.c index dfddeda..5543d63 100644 --- a/arch/arm/plat-s5p/dev-fimc2.c +++ b/arch/arm/plat-s5p/dev-fimc2.c @@ -10,6 +10,7 @@ */ #include +#include #include #include #include @@ -28,9 +29,15 @@ static struct resource s5p_fimc2_resource[] = { }, }; +static u64 s5p_fimc2_dma_mask = DMA_BIT_MASK(32); + struct platform_device s5p_device_fimc2 = { .name = "s5p-fimc", .id = 2, .num_resources = ARRAY_SIZE(s5p_fimc2_resource), .resource = s5p_fimc2_resource, + .dev = { + .dma_mask = &s5p_fimc2_dma_mask, + .coherent_dma_mask = DMA_BIT_MASK(32), + }, }; -- cgit v0.10.2 From 80e2f36aab7e69a193926814b73b3609c2f69f20 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Thu, 9 Sep 2010 21:31:46 +0900 Subject: ARM: S5P: Decrease IO Registers memory region size on FIMC IO registers region size of all FIMC versions is less than 1kB so there is no need to reserve 1M. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Marek Szyprowski [kgene.kim@samsung.com: minor title fix] Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-s5p/dev-fimc0.c b/arch/arm/plat-s5p/dev-fimc0.c index f9203fa..608770f 100644 --- a/arch/arm/plat-s5p/dev-fimc0.c +++ b/arch/arm/plat-s5p/dev-fimc0.c @@ -19,7 +19,7 @@ static struct resource s5p_fimc0_resource[] = { [0] = { .start = S5P_PA_FIMC0, - .end = S5P_PA_FIMC0 + SZ_1M - 1, + .end = S5P_PA_FIMC0 + SZ_4K - 1, .flags = IORESOURCE_MEM, }, [1] = { diff --git a/arch/arm/plat-s5p/dev-fimc1.c b/arch/arm/plat-s5p/dev-fimc1.c index 7ed91a3..76e3a97 100644 --- a/arch/arm/plat-s5p/dev-fimc1.c +++ b/arch/arm/plat-s5p/dev-fimc1.c @@ -19,7 +19,7 @@ static struct resource s5p_fimc1_resource[] = { [0] = { .start = S5P_PA_FIMC1, - .end = S5P_PA_FIMC1 + SZ_1M - 1, + .end = S5P_PA_FIMC1 + SZ_4K - 1, .flags = IORESOURCE_MEM, }, [1] = { diff --git a/arch/arm/plat-s5p/dev-fimc2.c b/arch/arm/plat-s5p/dev-fimc2.c index 5543d63..24d2981 100644 --- a/arch/arm/plat-s5p/dev-fimc2.c +++ b/arch/arm/plat-s5p/dev-fimc2.c @@ -19,7 +19,7 @@ static struct resource s5p_fimc2_resource[] = { [0] = { .start = S5P_PA_FIMC2, - .end = S5P_PA_FIMC2 + SZ_1M - 1, + .end = S5P_PA_FIMC2 + SZ_4K - 1, .flags = IORESOURCE_MEM, }, [1] = { -- cgit v0.10.2 From f1c894de475c2455e6cedcc91d0ee4eb04388f3b Mon Sep 17 00:00:00 2001 From: MyungJoo Ham Date: Sat, 21 Aug 2010 09:18:19 +0900 Subject: ARM: S5PV210: Update I2C-1 Clock Register Property. CLK_GATE_IP3[8] is RESERVED. The port "I2C_HDMI_DDC" of CLK_GATE_IP3[10] is used as another I2C port. Therefore, defined the unused I2C-1 as another I2C there was left undefined but used. Signed-off-by: MyungJoo Ham Signed-off-by: Kyungmin Park Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s5pv210/clock.c b/arch/arm/mach-s5pv210/clock.c index af91fef..825b868 100644 --- a/arch/arm/mach-s5pv210/clock.c +++ b/arch/arm/mach-s5pv210/clock.c @@ -357,7 +357,7 @@ static struct clk init_clocks_disable[] = { .id = 1, .parent = &clk_pclk_psys.clk, .enable = s5pv210_clk_ip3_ctrl, - .ctrlbit = (1<<8), + .ctrlbit = (1 << 10), }, { .name = "i2c", .id = 2, -- cgit v0.10.2 From a203a13a8843c3ec2d4dbc6285d000079a12afa5 Mon Sep 17 00:00:00 2001 From: Kyungmin Park Date: Sat, 21 Aug 2010 11:48:45 +0900 Subject: ARM: S5PV210: Reduce the iodesc length of systimer It's enough to use 4KiB. Signed-off-by: Kyungmin Park Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s5pv210/cpu.c b/arch/arm/mach-s5pv210/cpu.c index b9f4d67..77f456c 100644 --- a/arch/arm/mach-s5pv210/cpu.c +++ b/arch/arm/mach-s5pv210/cpu.c @@ -47,7 +47,7 @@ static struct map_desc s5pv210_iodesc[] __initdata = { { .virtual = (unsigned long)S5P_VA_SYSTIMER, .pfn = __phys_to_pfn(S5PV210_PA_SYSTIMER), - .length = SZ_1M, + .length = SZ_4K, .type = MT_DEVICE, }, { .virtual = (unsigned long)VA_VIC2, -- cgit v0.10.2 From 8fe294caf8c868edd9046251824a0af91991bf43 Mon Sep 17 00:00:00 2001 From: Guillaume Chazarain Date: Sun, 12 Sep 2010 21:32:35 +0200 Subject: HID: fix hiddev's use of usb_find_interface My macbook infrared remote control was broken by commit bd25f4dd6972755579d0ea50d1a5ace2e9b00d1a ("HID: hiddev: use usb_find_interface, get rid of BKL"). This device appears in dmesg as: apple 0003:05AC:8242.0001: hiddev0,hidraw0: USB HID v1.11 Device [Apple Computer, Inc. IR Receiver] on usb-0000:00:1d.2-1/input0 It stopped working as lircd was getting ENODEV when opening /dev/usb/hiddev0. AFAICS hiddev_driver is a dummy driver so usb_find_interface(&hiddev_driver) does not find anything. The device is associated with the usbhid driver, so let's do usb_find_interface(&hid_driver) instead. $ ls -l /sys/devices/pci0000:00/0000:00:1d.2/usb7/7-1/7-1:1.0/usb/hiddev0/device/driver lrwxrwxrwx 1 root root 0 2010-09-12 16:28 /sys/devices/pci0000:00/0000:00:1d.2/usb7/7-1/7-1:1.0/usb/hiddev0/device/driver -> ../../../../../../bus/usb/drivers/usbhid Signed-off-by: Guillaume Chazarain Signed-off-by: Jiri Kosina diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c index ffd6899..599041a 100644 --- a/drivers/hid/usbhid/hid-core.c +++ b/drivers/hid/usbhid/hid-core.c @@ -1446,6 +1446,11 @@ static const struct hid_device_id hid_usb_table[] = { { } }; +struct usb_interface *usbhid_find_interface(int minor) +{ + return usb_find_interface(&hid_driver, minor); +} + static struct hid_driver hid_usb_driver = { .name = "generic-usb", .id_table = hid_usb_table, diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c index 0a29c51..681e620 100644 --- a/drivers/hid/usbhid/hiddev.c +++ b/drivers/hid/usbhid/hiddev.c @@ -270,7 +270,7 @@ static int hiddev_open(struct inode *inode, struct file *file) struct hiddev *hiddev; int res; - intf = usb_find_interface(&hiddev_driver, iminor(inode)); + intf = usbhid_find_interface(iminor(inode)); if (!intf) return -ENODEV; hid = usb_get_intfdata(intf); diff --git a/drivers/hid/usbhid/usbhid.h b/drivers/hid/usbhid/usbhid.h index 693fd3e..89d2e84 100644 --- a/drivers/hid/usbhid/usbhid.h +++ b/drivers/hid/usbhid/usbhid.h @@ -42,6 +42,7 @@ void usbhid_submit_report (struct hid_device *hid, struct hid_report *report, unsigned char dir); int usbhid_get_power(struct hid_device *hid); void usbhid_put_power(struct hid_device *hid); +struct usb_interface *usbhid_find_interface(int minor); /* iofl flags */ #define HID_CTRL_RUNNING 1 -- cgit v0.10.2 From da01c2f733f5eaa1b0670e7b584f431934c60306 Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Fri, 10 Sep 2010 19:43:12 +0900 Subject: ARM: S5PV210: Add FIMC clocks These clocks enables FIMC driver to operate on machines, which bootloader power gated FIMC devices to save power on boot. Signed-off-by: Marek Szyprowski Signed-off-by: Kyungmin Park [kgene.kim@samsung.com: minor title fix] Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s5pv210/clock.c b/arch/arm/mach-s5pv210/clock.c index 825b868..cfecd70 100644 --- a/arch/arm/mach-s5pv210/clock.c +++ b/arch/arm/mach-s5pv210/clock.c @@ -281,6 +281,24 @@ static struct clk init_clocks_disable[] = { .enable = s5pv210_clk_ip0_ctrl, .ctrlbit = (1<<29), }, { + .name = "fimc", + .id = 0, + .parent = &clk_hclk_dsys.clk, + .enable = s5pv210_clk_ip0_ctrl, + .ctrlbit = (1 << 24), + }, { + .name = "fimc", + .id = 1, + .parent = &clk_hclk_dsys.clk, + .enable = s5pv210_clk_ip0_ctrl, + .ctrlbit = (1 << 25), + }, { + .name = "fimc", + .id = 2, + .parent = &clk_hclk_dsys.clk, + .enable = s5pv210_clk_ip0_ctrl, + .ctrlbit = (1 << 26), + }, { .name = "otg", .id = -1, .parent = &clk_hclk_psys.clk, -- cgit v0.10.2 From 0770e5280efd6f0001972a2b5127dffad0e22591 Mon Sep 17 00:00:00 2001 From: Kukjin Kim Date: Mon, 30 Aug 2010 10:45:54 +0900 Subject: ARM: SAMSUNG: Fix on drive strength value This patch fixes on defined drive strength value for GPIO. According to data sheet, if we want drive strength 1x, the value should be 00(b), if 2x should be 10(b), if 3x should be 01(b), and if 4x should be 11(b). Also fixes comment(from S5C to S5P). Reported-by: Janghyuck Kim Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-samsung/include/plat/gpio-cfg.h b/arch/arm/plat-samsung/include/plat/gpio-cfg.h index db4112c..1c6b929 100644 --- a/arch/arm/plat-samsung/include/plat/gpio-cfg.h +++ b/arch/arm/plat-samsung/include/plat/gpio-cfg.h @@ -143,12 +143,12 @@ extern s3c_gpio_pull_t s3c_gpio_getpull(unsigned int pin); /* Define values for the drvstr available for each gpio pin. * * These values control the value of the output signal driver strength, - * configurable on most pins on the S5C series. + * configurable on most pins on the S5P series. */ -#define S5P_GPIO_DRVSTR_LV1 ((__force s5p_gpio_drvstr_t)0x00) -#define S5P_GPIO_DRVSTR_LV2 ((__force s5p_gpio_drvstr_t)0x01) -#define S5P_GPIO_DRVSTR_LV3 ((__force s5p_gpio_drvstr_t)0x10) -#define S5P_GPIO_DRVSTR_LV4 ((__force s5p_gpio_drvstr_t)0x11) +#define S5P_GPIO_DRVSTR_LV1 ((__force s5p_gpio_drvstr_t)0x0) +#define S5P_GPIO_DRVSTR_LV2 ((__force s5p_gpio_drvstr_t)0x2) +#define S5P_GPIO_DRVSTR_LV3 ((__force s5p_gpio_drvstr_t)0x1) +#define S5P_GPIO_DRVSTR_LV4 ((__force s5p_gpio_drvstr_t)0x3) /** * s5c_gpio_get_drvstr() - get the driver streght value of a gpio pin -- cgit v0.10.2 From cbd2780fcefc370732c1f1526fb784573308177d Mon Sep 17 00:00:00 2001 From: Kukjin Kim Date: Mon, 30 Aug 2010 12:07:58 +0900 Subject: ARM: SAMSUNG: Fix on s5p_gpio_[get,set]_drvstr This patch fixes bug on gpio drive strength helper function. The offset should be like follwoing. - off = chip->chip.base - pin; + off = pin - chip->chip.base; In the s5p_gpio_get_drvstr(), the second line is unnecessary, because overwrite drvstr. drvstr = __raw_readl(reg); - drvstr = 0xffff & (0x3 << shift); And need 2bit masking before return the drvstr value. drvstr = drvstr >> shift; + drvstr &= 0x3; In the s5p_gpio_set_drvstr(), need relevant bit clear. tmp = __raw_readl(reg); + tmp &= ~(0x3 << shift); tmp |= drvstr << shift; Reported-by: Jaecheol Lee Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-samsung/gpio-config.c b/arch/arm/plat-samsung/gpio-config.c index 57b68a5..e3d41ea 100644 --- a/arch/arm/plat-samsung/gpio-config.c +++ b/arch/arm/plat-samsung/gpio-config.c @@ -273,13 +273,13 @@ s5p_gpio_drvstr_t s5p_gpio_get_drvstr(unsigned int pin) if (!chip) return -EINVAL; - off = chip->chip.base - pin; + off = pin - chip->chip.base; shift = off * 2; reg = chip->base + 0x0C; drvstr = __raw_readl(reg); - drvstr = 0xffff & (0x3 << shift); drvstr = drvstr >> shift; + drvstr &= 0x3; return (__force s5p_gpio_drvstr_t)drvstr; } @@ -296,11 +296,12 @@ int s5p_gpio_set_drvstr(unsigned int pin, s5p_gpio_drvstr_t drvstr) if (!chip) return -EINVAL; - off = chip->chip.base - pin; + off = pin - chip->chip.base; shift = off * 2; reg = chip->base + 0x0C; tmp = __raw_readl(reg); + tmp &= ~(0x3 << shift); tmp |= drvstr << shift; __raw_writel(tmp, reg); -- cgit v0.10.2 From 23a07eb0e8ea7d5b6ac41f2cc647e7ae53859292 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 9 Sep 2010 11:12:37 +0900 Subject: ARM: S3C64XX: Fix dev-spi build The irqs.h usage here got missed in the Samsung platform reorganisation. Signed-off-by: Mark Brown Acked-by: Jassi Brar Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s3c64xx/dev-spi.c b/arch/arm/mach-s3c64xx/dev-spi.c index a492b98..bfa3ef7 100644 --- a/arch/arm/mach-s3c64xx/dev-spi.c +++ b/arch/arm/mach-s3c64xx/dev-spi.c @@ -18,10 +18,10 @@ #include #include #include +#include #include #include -#include static char *spi_src_clks[] = { [S3C64XX_SPI_SRCCLK_PCLK] = "pclk", -- cgit v0.10.2 From 930a9e283516a3a3595c0c515113f1b78d07f695 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 14 Sep 2010 11:07:23 +0100 Subject: drm: Use a nondestructive mode for output detect when polling (v2) v2: Julien Cristau pointed out that @nondestructive results in double-negatives and confusion when trying to interpret the parameter, so use @force instead. Much easier to type as well. ;-) And fix the miscompilation of vmgfx reported by Sedat Dilek. Signed-off-by: Chris Wilson Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index fb6b70f..dcbeb98 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -103,7 +103,7 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector, if (connector->funcs->force) connector->funcs->force(connector); } else { - connector->status = connector->funcs->detect(connector, false); + connector->status = connector->funcs->detect(connector, true); drm_kms_helper_poll_enable(dev); } @@ -866,7 +866,7 @@ static void output_poll_execute(struct work_struct *work) !(connector->polled & DRM_CONNECTOR_POLL_HPD)) continue; - status = connector->funcs->detect(connector, true); + status = connector->funcs->detect(connector, false); if (old_status != status) changed = true; } diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 0350e5d..a02a8df 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -401,8 +401,7 @@ intel_crt_load_detect(struct drm_crtc *crtc, struct intel_encoder *intel_encoder } static enum drm_connector_status -intel_crt_detect(struct drm_connector *connector, - bool nondestructive) +intel_crt_detect(struct drm_connector *connector, bool force) { struct drm_device *dev = connector->dev; struct drm_encoder *encoder = intel_attached_encoder(connector); @@ -421,7 +420,7 @@ intel_crt_detect(struct drm_connector *connector, if (intel_crt_detect_ddc(encoder)) return connector_status_connected; - if (nondestructive) + if (!force) return connector->status; /* for pre-945g platforms use load detect */ diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index e1a2a05..1a51ee0 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1386,8 +1386,7 @@ ironlake_dp_detect(struct drm_connector *connector) * \return false if DP port is disconnected. */ static enum drm_connector_status -intel_dp_detect(struct drm_connector *connector, - bool nondestructive) +intel_dp_detect(struct drm_connector *connector, bool force) { struct drm_encoder *encoder = intel_attached_encoder(connector); struct intel_dp *intel_dp = enc_to_intel_dp(encoder); diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c index f0de1ad..7c9ec14 100644 --- a/drivers/gpu/drm/i915/intel_dvo.c +++ b/drivers/gpu/drm/i915/intel_dvo.c @@ -222,8 +222,7 @@ static void intel_dvo_mode_set(struct drm_encoder *encoder, * Unimplemented. */ static enum drm_connector_status -intel_dvo_detect(struct drm_connector *connector, - bool nondestructive) +intel_dvo_detect(struct drm_connector *connector, bool force) { struct drm_encoder *encoder = intel_attached_encoder(connector); struct intel_dvo *intel_dvo = enc_to_intel_dvo(encoder); diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 2ea123d..926934a 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -139,8 +139,7 @@ static bool intel_hdmi_mode_fixup(struct drm_encoder *encoder, } static enum drm_connector_status -intel_hdmi_detect(struct drm_connector *connector, - bool nondestructive) +intel_hdmi_detect(struct drm_connector *connector, bool force) { struct drm_encoder *encoder = intel_attached_encoder(connector); struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder); diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index fb1bed8..6ec39a8 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -446,8 +446,7 @@ static void intel_lvds_mode_set(struct drm_encoder *encoder, * needed, using lid status notification from the input layer. */ static enum drm_connector_status -intel_lvds_detect(struct drm_connector *connector, - bool nondestructive) +intel_lvds_detect(struct drm_connector *connector, bool force) { struct drm_device *dev = connector->dev; enum drm_connector_status status = connector_status_connected; @@ -543,7 +542,7 @@ static int intel_lid_notify(struct notifier_block *nb, unsigned long val, */ if (connector) connector->status = connector->funcs->detect(connector, - true); + false); /* Don't force modeset on machines where it causes a GPU lockup */ if (dmi_check_system(intel_no_modeset_on_lid)) diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index db6b6d4..e8e902d 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -1417,7 +1417,7 @@ intel_analog_is_connected(struct drm_device *dev) if (!analog_connector) return false; - if (analog_connector->funcs->detect(analog_connector, true) == + if (analog_connector->funcs->detect(analog_connector, false) == connector_status_disconnected) return false; @@ -1487,8 +1487,7 @@ intel_sdvo_hdmi_sink_detect(struct drm_connector *connector) } static enum drm_connector_status -intel_sdvo_detect(struct drm_connector *connector, - bool nondestructive) +intel_sdvo_detect(struct drm_connector *connector, bool force) { uint16_t response; struct drm_encoder *encoder = intel_attached_encoder(connector); diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index d20b550..4a117e3 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1341,8 +1341,7 @@ static void intel_tv_find_better_format(struct drm_connector *connector) * we have a pipe programmed in order to probe the TV. */ static enum drm_connector_status -intel_tv_detect(struct drm_connector *connector, - bool nondestructive) +intel_tv_detect(struct drm_connector *connector, bool force) { struct drm_display_mode mode; struct drm_encoder *encoder = intel_attached_encoder(connector); @@ -1354,7 +1353,7 @@ intel_tv_detect(struct drm_connector *connector, if (encoder->crtc && encoder->crtc->enabled) { type = intel_tv_detect_type(intel_tv); - } else if (nondestructive) { + } else if (force) { struct drm_crtc *crtc; int dpms_mode; diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c index 67d515c..87186a4 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.c +++ b/drivers/gpu/drm/nouveau/nouveau_connector.c @@ -168,8 +168,7 @@ nouveau_connector_set_encoder(struct drm_connector *connector, } static enum drm_connector_status -nouveau_connector_detect(struct drm_connector *connector, - bool nondestructive) +nouveau_connector_detect(struct drm_connector *connector, bool force) { struct drm_device *dev = connector->dev; struct nouveau_connector *nv_connector = nouveau_connector(connector); @@ -247,8 +246,7 @@ detect_analog: } static enum drm_connector_status -nouveau_connector_detect_lvds(struct drm_connector *connector, - bool nondestructive) +nouveau_connector_detect_lvds(struct drm_connector *connector, bool force) { struct drm_device *dev = connector->dev; struct drm_nouveau_private *dev_priv = dev->dev_private; @@ -269,7 +267,7 @@ nouveau_connector_detect_lvds(struct drm_connector *connector, /* Try retrieving EDID via DDC */ if (!dev_priv->vbios.fp_no_ddc) { - status = nouveau_connector_detect(connector, nondestructive); + status = nouveau_connector_detect(connector, force); if (status == connector_status_connected) goto out; } diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 31d309a..ecc1a8f 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -482,8 +482,7 @@ static int radeon_lvds_mode_valid(struct drm_connector *connector, } static enum drm_connector_status -radeon_lvds_detect(struct drm_connector *connector, - bool nondestructive) +radeon_lvds_detect(struct drm_connector *connector, bool force) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct drm_encoder *encoder = radeon_best_single_encoder(connector); @@ -597,8 +596,7 @@ static int radeon_vga_mode_valid(struct drm_connector *connector, } static enum drm_connector_status -radeon_vga_detect(struct drm_connector *connector, - bool nondestructive) +radeon_vga_detect(struct drm_connector *connector, bool force) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct drm_encoder *encoder; @@ -696,8 +694,7 @@ static int radeon_tv_mode_valid(struct drm_connector *connector, } static enum drm_connector_status -radeon_tv_detect(struct drm_connector *connector, - bool nondestructive) +radeon_tv_detect(struct drm_connector *connector, bool force) { struct drm_encoder *encoder; struct drm_encoder_helper_funcs *encoder_funcs; @@ -755,8 +752,7 @@ static int radeon_dvi_get_modes(struct drm_connector *connector) * if its shared we have to set the other connector to disconnected. */ static enum drm_connector_status -radeon_dvi_detect(struct drm_connector *connector, - bool nondestructive) +radeon_dvi_detect(struct drm_connector *connector, bool force) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct drm_encoder *encoder = NULL; @@ -981,8 +977,7 @@ static int radeon_dp_get_modes(struct drm_connector *connector) } static enum drm_connector_status -radeon_dp_detect(struct drm_connector *connector, - bool nondestructive) +radeon_dp_detect(struct drm_connector *connector, bool force) { struct radeon_connector *radeon_connector = to_radeon_connector(connector); enum drm_connector_status ret = connector_status_disconnected; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c index a527c91..7083b1a 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c @@ -336,7 +336,7 @@ static void vmw_ldu_connector_restore(struct drm_connector *connector) static enum drm_connector_status vmw_ldu_connector_detect(struct drm_connector *connector, - bool nondestructive) + bool force) { if (vmw_connector_to_ldu(connector)->pref_active) return connector_status_connected; @@ -517,7 +517,7 @@ static int vmw_ldu_init(struct vmw_private *dev_priv, unsigned unit) drm_connector_init(dev, connector, &vmw_legacy_connector_funcs, DRM_MODE_CONNECTOR_LVDS); - connector->status = vmw_ldu_connector_detect(connector); + connector->status = vmw_ldu_connector_detect(connector, true); drm_encoder_init(dev, encoder, &vmw_legacy_encoder_funcs, DRM_MODE_ENCODER_LVDS); @@ -611,7 +611,7 @@ int vmw_kms_ldu_update_layout(struct vmw_private *dev_priv, unsigned num, ldu->pref_height = 600; ldu->pref_active = false; } - con->status = vmw_ldu_connector_detect(con); + con->status = vmw_ldu_connector_detect(con, true); } mutex_unlock(&dev->mode_config.mutex); diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h index 5536223..3e5a51a 100644 --- a/include/drm/drm_crtc.h +++ b/include/drm/drm_crtc.h @@ -386,8 +386,15 @@ struct drm_connector_funcs { void (*dpms)(struct drm_connector *connector, int mode); void (*save)(struct drm_connector *connector); void (*restore)(struct drm_connector *connector); + + /* Check to see if anything is attached to the connector. + * @force is set to false whilst polling, true when checking the + * connector due to user request. @force can be used by the driver + * to avoid expensive, destructive operations during automated + * probing. + */ enum drm_connector_status (*detect)(struct drm_connector *connector, - bool nondestructive); + bool force); int (*fill_modes)(struct drm_connector *connector, uint32_t max_width, uint32_t max_height); int (*set_property)(struct drm_connector *connector, struct drm_property *property, uint64_t val); -- cgit v0.10.2 From b64c115eb22516ecd187c74ad6de3f1693f1dc7b Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Tue, 14 Sep 2010 20:14:38 +1000 Subject: drm: fix race between driver loading and userspace open. Not 100% sure this is due to BKL removal, its most likely a combination of that + userspace timing changes in udev/plymouth. The drm adds the sysfs device before the driver has completed internal loading, this causes udev to make the node and plymouth to open it before we've completed loading. The proper solution is to delay the sysfs manipulation until later in loading however this causes knock on issues with sysfs connector nodes, so we can use the global mutex to serialise loading and userspace opens. Reported-by: Toni Spets (hifi on #radeon) Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_pci.c b/drivers/gpu/drm/drm_pci.c index e20f78b..f5bd9e5 100644 --- a/drivers/gpu/drm/drm_pci.c +++ b/drivers/gpu/drm/drm_pci.c @@ -164,6 +164,8 @@ int drm_get_pci_dev(struct pci_dev *pdev, const struct pci_device_id *ent, dev->hose = pdev->sysdata; #endif + mutex_lock(&drm_global_mutex); + if ((ret = drm_fill_in_dev(dev, ent, driver))) { printk(KERN_ERR "DRM: Fill_in_dev failed.\n"); goto err_g2; @@ -199,6 +201,7 @@ int drm_get_pci_dev(struct pci_dev *pdev, const struct pci_device_id *ent, driver->name, driver->major, driver->minor, driver->patchlevel, driver->date, pci_name(pdev), dev->primary->index); + mutex_unlock(&drm_global_mutex); return 0; err_g4: @@ -210,6 +213,7 @@ err_g2: pci_disable_device(pdev); err_g1: kfree(dev); + mutex_unlock(&drm_global_mutex); return ret; } EXPORT_SYMBOL(drm_get_pci_dev); diff --git a/drivers/gpu/drm/drm_platform.c b/drivers/gpu/drm/drm_platform.c index 460e9a3..92d1d0f 100644 --- a/drivers/gpu/drm/drm_platform.c +++ b/drivers/gpu/drm/drm_platform.c @@ -53,6 +53,8 @@ int drm_get_platform_dev(struct platform_device *platdev, dev->platformdev = platdev; dev->dev = &platdev->dev; + mutex_lock(&drm_global_mutex); + ret = drm_fill_in_dev(dev, NULL, driver); if (ret) { @@ -87,6 +89,8 @@ int drm_get_platform_dev(struct platform_device *platdev, list_add_tail(&dev->driver_item, &driver->device_list); + mutex_unlock(&drm_global_mutex); + DRM_INFO("Initialized %s %d.%d.%d %s on minor %d\n", driver->name, driver->major, driver->minor, driver->patchlevel, driver->date, dev->primary->index); @@ -100,6 +104,7 @@ err_g2: drm_put_minor(&dev->control); err_g1: kfree(dev); + mutex_unlock(&drm_global_mutex); return ret; } EXPORT_SYMBOL(drm_get_platform_dev); -- cgit v0.10.2 From f90087eea44ce5fad139f086bc9d89ca37b0edc2 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 7 Sep 2010 11:42:45 -0400 Subject: drm/radeon/kms: force legacy pll algo for RV620 LVDS There has been periodic evidence that LVDS, on at least some panels, prefers the dividers selected by the legacy pll algo. This patch forces the use of the legacy pll algo on RV620 LVDS panels. The old behavior (new pll algo) can be selected by setting the new_pll module parameter to 1. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30029 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index 464a81a..cd0290f 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -539,14 +539,15 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, pll->algo = PLL_ALGO_LEGACY; pll->flags |= RADEON_PLL_PREFER_CLOSEST_LOWER; } - /* There is some evidence (often anecdotal) that RV515 LVDS + /* There is some evidence (often anecdotal) that RV515/RV620 LVDS * (on some boards at least) prefers the legacy algo. I'm not * sure whether this should handled generically or on a * case-by-case quirk basis. Both algos should work fine in the * majority of cases. */ if ((radeon_encoder->active_device & (ATOM_DEVICE_LCD_SUPPORT)) && - (rdev->family == CHIP_RV515)) { + ((rdev->family == CHIP_RV515) || + (rdev->family == CHIP_RV620))) { /* allow the user to overrride just in case */ if (radeon_new_pll == 1) pll->algo = PLL_ALGO_NEW; -- cgit v0.10.2 From 3894335876a6257ac46e14845bd37ae6fb0f7c87 Mon Sep 17 00:00:00 2001 From: Stephen Warren Date: Tue, 14 Sep 2010 10:48:59 -0600 Subject: ALSA: patch_nvhdmi.c: Fix supported sample rate list. 22050 isn't a valid HDMI sample rate. 32000 is. Signed-off-by: Stephen Warren Acked-By: Wei Ni Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_nvhdmi.c b/sound/pci/hda/patch_nvhdmi.c index 69b950d..baa108b 100644 --- a/sound/pci/hda/patch_nvhdmi.c +++ b/sound/pci/hda/patch_nvhdmi.c @@ -84,7 +84,7 @@ static struct hda_verb nvhdmi_basic_init_7x[] = { #else /* support all rates and formats */ #define SUPPORTED_RATES \ - (SNDRV_PCM_RATE_22050 | SNDRV_PCM_RATE_44100 | SNDRV_PCM_RATE_48000 |\ + (SNDRV_PCM_RATE_32000 | SNDRV_PCM_RATE_44100 | SNDRV_PCM_RATE_48000 |\ SNDRV_PCM_RATE_88200 | SNDRV_PCM_RATE_96000 | SNDRV_PCM_RATE_176400 |\ SNDRV_PCM_RATE_192000) #define SUPPORTED_MAXBPS 24 -- cgit v0.10.2 From 8bef9239ee1a42eb37d3f83bacf6a75f019c028d Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 14 Sep 2010 15:45:44 -0700 Subject: ceph: correctly set 'follows' in flushsnap messages The 'follows' should match the seq for the snap context for the given snap cap, which is the context under which we have been dirtying and writing data and metadata. The snapshot that _contains_ those updates thus _follows_ that context's seq #. Signed-off-by: Sage Weil diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 4868b9d..9e836af 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -467,7 +467,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci) INIT_LIST_HEAD(&capsnap->ci_item); INIT_LIST_HEAD(&capsnap->flushing_item); - capsnap->follows = snapc->seq - 1; + capsnap->follows = snapc->seq; capsnap->issued = __ceph_caps_issued(ci, NULL); capsnap->dirty = dirty; -- cgit v0.10.2 From cfc0bf6640dfd0f43bf8bfec5a475284809baa4d Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 14 Sep 2010 15:50:59 -0700 Subject: ceph: stop sending FLUSHSNAPs when we hit a dirty capsnap Stop sending FLUSHSNAP messages when we hit a capsnap that has dirty_pages or is still writing. We'll send the newer capsnaps only after the older ones complete. Signed-off-by: Sage Weil diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index a2069b6..9fbe901 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1227,7 +1227,7 @@ retry: * pages to be written out. */ if (capsnap->dirty_pages || capsnap->writing) - continue; + break; /* * if cap writeback already occurred, we should have dropped @@ -1276,8 +1276,8 @@ retry: &session->s_cap_snaps_flushing); spin_unlock(&inode->i_lock); - dout("flush_snaps %p cap_snap %p follows %lld size %llu\n", - inode, capsnap, next_follows, capsnap->size); + dout("flush_snaps %p cap_snap %p follows %lld tid %llu\n", + inode, capsnap, capsnap->follows, capsnap->flush_tid); send_cap_msg(session, ceph_vino(inode).ino, 0, CEPH_CAP_OP_FLUSHSNAP, capsnap->issued, 0, capsnap->dirty, 0, capsnap->flush_tid, 0, mseq, -- cgit v0.10.2 From 54ff7e595d763d894104d421b103a89f7becf47c Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 14 Sep 2010 22:10:21 +0200 Subject: x86: hpet: Work around hardware stupidity This more or less reverts commits 08be979 (x86: Force HPET readback_cmp for all ATI chipsets) and 30a564be (x86, hpet: Restrict read back to affected ATI chipsets) to the status of commit 8da854c (x86, hpet: Erratum workaround for read after write of HPET comparator). The delta to commit 8da854c is mostly comments and the change from WARN_ONCE to printk_once as we know the call path of this function already. This needs really in depth explanation: First of all the HPET design is a complete failure. Having a counter compare register which generates an interrupt on matching values forces the software to do at least one superfluous readback of the counter register. While it is nice in theory to program "absolute" time events it is practically useless because the timer runs at some absurd frequency which can never be matched to real world units. So we are forced to calculate a relative delta and this forces a readout of the actual counter value, adding the delta and programming the compare register. When the delta is small enough we run into the danger that we program a compare value which is already in the past. Due to the compare for equal nature of HPET we need to read back the counter value after writing the compare rehgister (btw. this is necessary for absolute timeouts as well) to make sure that we did not miss the timer event. We try to work around that by setting the minimum delta to a value which is larger than the theoretical time which elapses between the counter readout and the compare register write, but that's only true in theory. A NMI or SMI which hits between the readout and the write can easily push us beyond that limit. This would result in waiting for the next HPET timer interrupt until the 32bit wraparound of the counter happens which takes about 306 seconds. So we designed the next event function to look like: match = read_cnt() + delta; write_compare_ref(match); return read_cnt() < match ? 0 : -ETIME; At some point we got into trouble with certain ATI chipsets. Even the above "safe" procedure failed. The reason was that the write to the compare register was delayed probably for performance reasons. The theory was that they wanted to avoid the synchronization of the write with the HPET clock, which is understandable. So the write does not hit the compare register directly instead it goes to some intermediate register which is copied to the real compare register in sync with the HPET clock. That opens another window for hitting the dreaded "wait for a wraparound" problem. To work around that "optimization" we added a read back of the compare register which either enforced the update of the just written value or just delayed the readout of the counter enough to avoid the issue. We unfortunately never got any affirmative info from ATI/AMD about this. One thing is sure, that we nuked the performance "optimization" that way completely and I'm pretty sure that the result is worse than before some HW folks came up with those. Just for paranoia reasons I added a check whether the read back compare register value was the same as the value we wrote right before. That paranoia check triggered a couple of years after it was added on an Intel ICH9 chipset. Venki added a workaround (commit 8da854c) which was reading the compare register twice when the first check failed. We considered this to be a penalty in general and restricted the readback (thus the wasted CPU cycles) to the known to be affected ATI chipsets. This turned out to be a utterly wrong decision. 2.6.35 testers experienced massive problems and finally one of them bisected it down to commit 30a564be which spured some further investigation. Finally we got confirmation that the write to the compare register can be delayed by up to two HPET clock cycles which explains the problems nicely. All we can do about this is to go back to Venki's initial workaround in a slightly modified version. Just for the record I need to say, that all of this could have been avoided if hardware designers and of course the HPET committee would have thought about the consequences for a split second. It's out of my comprehension why designing a working timer is so hard. There are two ways to achieve it: 1) Use a counter wrap around aware compare_reg <= counter_reg implementation instead of the easy compare_reg == counter_reg Downsides: - It needs more silicon. - It needs a readout of the counter to apply a relative timeout. This is necessary as the counter does not run in any useful (and adjustable) frequency and there is no guarantee that the counter which is used for timer events is the same which is used for reading the actual time (and therefor for calculating the delta) Upsides: - None 2) Use a simple down counter for relative timer events Downsides: - Absolute timeouts are not possible, which is not a problem at all in the context of an OS and the expected max. latencies/jitter (also see Downsides of #1) Upsides: - It needs less or equal silicon. - It works ALWAYS - It is way faster than a compare register based solution (One write versus one write plus at least one and up to four reads) I would not be so grumpy about all of this, if I would not have been ignored for many years when pointing out these flaws to various hardware folks. I really hate timers (at least those which seem to be designed by janitors). Though finally we got a reasonable explanation plus a solution and I want to thank all the folks involved in chasing it down and providing valuable input to this. Bisected-by: Nix Reported-by: Artur Skawina Reported-by: Damien Wyart Reported-by: John Drescher Cc: Venkatesh Pallipadi Cc: Ingo Molnar Cc: H. Peter Anvin Cc: Arjan van de Ven Cc: Andreas Herrmann Cc: Borislav Petkov Cc: stable@kernel.org Acked-by: Suresh Siddha Signed-off-by: Thomas Gleixner diff --git a/arch/x86/include/asm/hpet.h b/arch/x86/include/asm/hpet.h index 004e6e2..1d5c08a 100644 --- a/arch/x86/include/asm/hpet.h +++ b/arch/x86/include/asm/hpet.h @@ -68,7 +68,6 @@ extern unsigned long force_hpet_address; extern u8 hpet_blockid; extern int hpet_force_user; extern u8 hpet_msi_disable; -extern u8 hpet_readback_cmp; extern int is_hpet_enabled(void); extern int hpet_enable(void); extern void hpet_disable(void); diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c index e5cc7e8..ebdb85c 100644 --- a/arch/x86/kernel/early-quirks.c +++ b/arch/x86/kernel/early-quirks.c @@ -18,7 +18,6 @@ #include #include #include -#include static void __init fix_hypertransport_config(int num, int slot, int func) { @@ -192,21 +191,6 @@ static void __init ati_bugs_contd(int num, int slot, int func) } #endif -/* - * Force the read back of the CMP register in hpet_next_event() - * to work around the problem that the CMP register write seems to be - * delayed. See hpet_next_event() for details. - * - * We do this on all SMBUS incarnations for now until we have more - * information about the affected chipsets. - */ -static void __init ati_hpet_bugs(int num, int slot, int func) -{ -#ifdef CONFIG_HPET_TIMER - hpet_readback_cmp = 1; -#endif -} - #define QFLAG_APPLY_ONCE 0x1 #define QFLAG_APPLIED 0x2 #define QFLAG_DONE (QFLAG_APPLY_ONCE|QFLAG_APPLIED) @@ -236,8 +220,6 @@ static struct chipset early_qrk[] __initdata = { PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_bugs }, { PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_SBX00_SMBUS, PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_bugs_contd }, - { PCI_VENDOR_ID_ATI, PCI_ANY_ID, - PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_hpet_bugs }, {} }; diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 351f9c0..410fdb3 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -35,7 +35,6 @@ unsigned long hpet_address; u8 hpet_blockid; /* OS timer block num */ u8 hpet_msi_disable; -u8 hpet_readback_cmp; #ifdef CONFIG_PCI_MSI static unsigned long hpet_num_timers; @@ -395,23 +394,27 @@ static int hpet_next_event(unsigned long delta, * at that point and we would wait for the next hpet interrupt * forever. We found out that reading the CMP register back * forces the transfer so we can rely on the comparison with - * the counter register below. + * the counter register below. If the read back from the + * compare register does not match the value we programmed + * then we might have a real hardware problem. We can not do + * much about it here, but at least alert the user/admin with + * a prominent warning. * - * That works fine on those ATI chipsets, but on newer Intel - * chipsets (ICH9...) this triggers due to an erratum: Reading - * the comparator immediately following a write is returning - * the old value. + * An erratum on some chipsets (ICH9,..), results in + * comparator read immediately following a write returning old + * value. Workaround for this is to read this value second + * time, when first read returns old value. * - * We restrict the read back to the affected ATI chipsets (set - * by quirks) and also run it with hpet=verbose for debugging - * purposes. + * In fact the write to the comparator register is delayed up + * to two HPET cycles so the workaround we tried to restrict + * the readback to those known to be borked ATI chipsets + * failed miserably. So we give up on optimizations forever + * and penalize all HPET incarnations unconditionally. */ - if (hpet_readback_cmp || hpet_verbose) { - u32 cmp = hpet_readl(HPET_Tn_CMP(timer)); - - if (cmp != cnt) + if (unlikely((u32)hpet_readl(HPET_Tn_CMP(timer)) != cnt)) { + if (hpet_readl(HPET_Tn_CMP(timer)) != cnt) printk_once(KERN_WARNING - "hpet: compare register read back failed.\n"); + "hpet: compare register read back failed.\n"); } return (s32)(hpet_readl(HPET_COUNTER) - cnt) >= 0 ? -ETIME : 0; -- cgit v0.10.2 From c41d68a513c71e35a14f66d71782d27a79a81ea6 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Tue, 7 Sep 2010 16:16:18 -0700 Subject: compat: Make compat_alloc_user_space() incorporate the access_ok() compat_alloc_user_space() expects the caller to independently call access_ok() to verify the returned area. A missing call could introduce problems on some architectures. This patch incorporates the access_ok() check into compat_alloc_user_space() and also adds a sanity check on the length. The existing compat_alloc_user_space() implementations are renamed arch_compat_alloc_user_space() and are used as part of the implementation of the new global function. This patch assumes NULL will cause __get_user()/__put_user() to either fail or access userspace on all architectures. This should be followed by checking the return value of compat_access_user_space() for NULL in the callers, at which time the access_ok() in the callers can also be removed. Reported-by: Ben Hawkes Signed-off-by: H. Peter Anvin Acked-by: Benjamin Herrenschmidt Acked-by: Chris Metcalf Acked-by: David S. Miller Acked-by: Ingo Molnar Acked-by: Thomas Gleixner Acked-by: Tony Luck Cc: Andrew Morton Cc: Arnd Bergmann Cc: Fenghua Yu Cc: H. Peter Anvin Cc: Heiko Carstens Cc: Helge Deller Cc: James Bottomley Cc: Kyle McMartin Cc: Martin Schwidefsky Cc: Paul Mackerras Cc: Ralf Baechle Cc: diff --git a/arch/ia64/include/asm/compat.h b/arch/ia64/include/asm/compat.h index f90edc8..9301a28 100644 --- a/arch/ia64/include/asm/compat.h +++ b/arch/ia64/include/asm/compat.h @@ -199,7 +199,7 @@ ptr_to_compat(void __user *uptr) } static __inline__ void __user * -compat_alloc_user_space (long len) +arch_compat_alloc_user_space (long len) { struct pt_regs *regs = task_pt_regs(current); return (void __user *) (((regs->r12 & 0xffffffff) & -16) - len); diff --git a/arch/mips/include/asm/compat.h b/arch/mips/include/asm/compat.h index 613f691..dbc5106 100644 --- a/arch/mips/include/asm/compat.h +++ b/arch/mips/include/asm/compat.h @@ -145,7 +145,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -static inline void __user *compat_alloc_user_space(long len) +static inline void __user *arch_compat_alloc_user_space(long len) { struct pt_regs *regs = (struct pt_regs *) ((unsigned long) current_thread_info() + THREAD_SIZE - 32) - 1; diff --git a/arch/parisc/include/asm/compat.h b/arch/parisc/include/asm/compat.h index 02b77ba..efa0b60 100644 --- a/arch/parisc/include/asm/compat.h +++ b/arch/parisc/include/asm/compat.h @@ -147,7 +147,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -static __inline__ void __user *compat_alloc_user_space(long len) +static __inline__ void __user *arch_compat_alloc_user_space(long len) { struct pt_regs *regs = ¤t->thread.regs; return (void __user *)regs->gr[30]; diff --git a/arch/powerpc/include/asm/compat.h b/arch/powerpc/include/asm/compat.h index 396d21a..a11d4ea 100644 --- a/arch/powerpc/include/asm/compat.h +++ b/arch/powerpc/include/asm/compat.h @@ -134,7 +134,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -static inline void __user *compat_alloc_user_space(long len) +static inline void __user *arch_compat_alloc_user_space(long len) { struct pt_regs *regs = current->thread.regs; unsigned long usp = regs->gpr[1]; diff --git a/arch/s390/include/asm/compat.h b/arch/s390/include/asm/compat.h index 104f200..a875c2f 100644 --- a/arch/s390/include/asm/compat.h +++ b/arch/s390/include/asm/compat.h @@ -181,7 +181,7 @@ static inline int is_compat_task(void) #endif -static inline void __user *compat_alloc_user_space(long len) +static inline void __user *arch_compat_alloc_user_space(long len) { unsigned long stack; diff --git a/arch/sparc/include/asm/compat.h b/arch/sparc/include/asm/compat.h index 5016f76..6f57325 100644 --- a/arch/sparc/include/asm/compat.h +++ b/arch/sparc/include/asm/compat.h @@ -167,7 +167,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -static inline void __user *compat_alloc_user_space(long len) +static inline void __user *arch_compat_alloc_user_space(long len) { struct pt_regs *regs = current_thread_info()->kregs; unsigned long usp = regs->u_regs[UREG_I6]; diff --git a/arch/tile/include/asm/compat.h b/arch/tile/include/asm/compat.h index 5a34da6..345d81c 100644 --- a/arch/tile/include/asm/compat.h +++ b/arch/tile/include/asm/compat.h @@ -195,7 +195,7 @@ static inline unsigned long ptr_to_compat_reg(void __user *uptr) return (long)(int)(long __force)uptr; } -static inline void __user *compat_alloc_user_space(long len) +static inline void __user *arch_compat_alloc_user_space(long len) { struct pt_regs *regs = task_pt_regs(current); return (void __user *)regs->sp - len; diff --git a/arch/x86/include/asm/compat.h b/arch/x86/include/asm/compat.h index 306160e..1d9cd27 100644 --- a/arch/x86/include/asm/compat.h +++ b/arch/x86/include/asm/compat.h @@ -205,7 +205,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -static inline void __user *compat_alloc_user_space(long len) +static inline void __user *arch_compat_alloc_user_space(long len) { struct pt_regs *regs = task_pt_regs(current); return (void __user *)regs->sp - len; diff --git a/include/linux/compat.h b/include/linux/compat.h index 9ddc878..5778b55 100644 --- a/include/linux/compat.h +++ b/include/linux/compat.h @@ -360,5 +360,8 @@ extern ssize_t compat_rw_copy_check_uvector(int type, const struct compat_iovec __user *uvector, unsigned long nr_segs, unsigned long fast_segs, struct iovec *fast_pointer, struct iovec **ret_pointer); + +extern void __user *compat_alloc_user_space(unsigned long len); + #endif /* CONFIG_COMPAT */ #endif /* _LINUX_COMPAT_H */ diff --git a/kernel/compat.c b/kernel/compat.c index e167efc..c9e2ec0 100644 --- a/kernel/compat.c +++ b/kernel/compat.c @@ -1126,3 +1126,24 @@ compat_sys_sysinfo(struct compat_sysinfo __user *info) return 0; } + +/* + * Allocate user-space memory for the duration of a single system call, + * in order to marshall parameters inside a compat thunk. + */ +void __user *compat_alloc_user_space(unsigned long len) +{ + void __user *ptr; + + /* If len would occupy more than half of the entire compat space... */ + if (unlikely(len > (((compat_uptr_t)~0) >> 1))) + return NULL; + + ptr = arch_compat_alloc_user_space(len); + + if (unlikely(!access_ok(VERIFY_WRITE, ptr, len))) + return NULL; + + return ptr; +} +EXPORT_SYMBOL_GPL(compat_alloc_user_space); -- cgit v0.10.2 From 36d001c70d8a0144ac1d038f6876c484849a74de Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Tue, 14 Sep 2010 12:42:41 -0700 Subject: x86-64, compat: Test %rax for the syscall number, not %eax On 64 bits, we always, by necessity, jump through the system call table via %rax. For 32-bit system calls, in theory the system call number is stored in %eax, and the code was testing %eax for a valid system call number. At one point we loaded the stored value back from the stack to enforce zero-extension, but that was removed in checkin d4d67150165df8bf1cc05e532f6efca96f907cab. An actual 32-bit process will not be able to introduce a non-zero-extended number, but it can happen via ptrace. Instead of re-introducing the zero-extension, test what we are actually going to use, i.e. %rax. This only adds a handful of REX prefixes to the code. Reported-by: Ben Hawkes Signed-off-by: H. Peter Anvin Cc: Cc: Roland McGrath Cc: Andrew Morton diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index b86feab..84e3a4e 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -153,7 +153,7 @@ ENTRY(ia32_sysenter_target) testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r10) CFI_REMEMBER_STATE jnz sysenter_tracesys - cmpl $(IA32_NR_syscalls-1),%eax + cmpq $(IA32_NR_syscalls-1),%rax ja ia32_badsys sysenter_do_call: IA32_ARG_FIXUP @@ -195,7 +195,7 @@ sysexit_from_sys_call: movl $AUDIT_ARCH_I386,%edi /* 1st arg: audit arch */ call audit_syscall_entry movl RAX-ARGOFFSET(%rsp),%eax /* reload syscall number */ - cmpl $(IA32_NR_syscalls-1),%eax + cmpq $(IA32_NR_syscalls-1),%rax ja ia32_badsys movl %ebx,%edi /* reload 1st syscall arg */ movl RCX-ARGOFFSET(%rsp),%esi /* reload 2nd syscall arg */ @@ -248,7 +248,7 @@ sysenter_tracesys: call syscall_trace_enter LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST - cmpl $(IA32_NR_syscalls-1),%eax + cmpq $(IA32_NR_syscalls-1),%rax ja int_ret_from_sys_call /* sysenter_tracesys has set RAX(%rsp) */ jmp sysenter_do_call CFI_ENDPROC @@ -314,7 +314,7 @@ ENTRY(ia32_cstar_target) testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r10) CFI_REMEMBER_STATE jnz cstar_tracesys - cmpl $IA32_NR_syscalls-1,%eax + cmpq $IA32_NR_syscalls-1,%rax ja ia32_badsys cstar_do_call: IA32_ARG_FIXUP 1 @@ -367,7 +367,7 @@ cstar_tracesys: LOAD_ARGS32 ARGOFFSET, 1 /* reload args from stack in case ptrace changed it */ RESTORE_REST xchgl %ebp,%r9d - cmpl $(IA32_NR_syscalls-1),%eax + cmpq $(IA32_NR_syscalls-1),%rax ja int_ret_from_sys_call /* cstar_tracesys has set RAX(%rsp) */ jmp cstar_do_call END(ia32_cstar_target) @@ -425,7 +425,7 @@ ENTRY(ia32_syscall) orl $TS_COMPAT,TI_status(%r10) testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r10) jnz ia32_tracesys - cmpl $(IA32_NR_syscalls-1),%eax + cmpq $(IA32_NR_syscalls-1),%rax ja ia32_badsys ia32_do_call: IA32_ARG_FIXUP @@ -444,7 +444,7 @@ ia32_tracesys: call syscall_trace_enter LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST - cmpl $(IA32_NR_syscalls-1),%eax + cmpq $(IA32_NR_syscalls-1),%rax ja int_ret_from_sys_call /* ia32_tracesys has set RAX(%rsp) */ jmp ia32_do_call END(ia32_syscall) -- cgit v0.10.2 From eefdca043e8391dcd719711716492063030b55ac Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Tue, 14 Sep 2010 12:22:58 -0700 Subject: x86-64, compat: Retruncate rax after ia32 syscall entry tracing In commit d4d6715, we reopened an old hole for a 64-bit ptracer touching a 32-bit tracee in system call entry. A %rax value set via ptrace at the entry tracing stop gets used whole as a 32-bit syscall number, while we only check the low 32 bits for validity. Fix it by truncating %rax back to 32 bits after syscall_trace_enter, in addition to testing the full 64 bits as has already been added. Reported-by: Ben Hawkes Signed-off-by: Roland McGrath Signed-off-by: H. Peter Anvin diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index 84e3a4e..518bb99 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -50,7 +50,12 @@ /* * Reload arg registers from stack in case ptrace changed them. * We don't reload %eax because syscall_trace_enter() returned - * the value it wants us to use in the table lookup. + * the %rax value we should see. Instead, we just truncate that + * value to 32 bits again as we did on entry from user mode. + * If it's a new value set by user_regset during entry tracing, + * this matches the normal truncation of the user-mode value. + * If it's -1 to make us punt the syscall, then (u32)-1 is still + * an appropriately invalid value. */ .macro LOAD_ARGS32 offset, _r9=0 .if \_r9 @@ -60,6 +65,7 @@ movl \offset+48(%rsp),%edx movl \offset+56(%rsp),%esi movl \offset+64(%rsp),%edi + movl %eax,%eax /* zero extension */ .endm .macro CFI_STARTPROC32 simple -- cgit v0.10.2 From 460cf3411b858ad509d5255e0dfaf862a83c0299 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 14 Sep 2010 11:38:24 -0400 Subject: cifs: fix potential double put of TCP session reference cifs_get_smb_ses must be called on a server pointer on which it holds an active reference. It first does a search for an existing SMB session. If it finds one, it'll put the server reference and then try to ensure that the negprot is done, etc. If it encounters an error at that point then it'll return an error. There's a potential problem here though. When cifs_get_smb_ses returns an error, the caller will also put the TCP server reference leading to a double-put. Fix this by having cifs_get_smb_ses only put the server reference if it found an existing session that it could use and isn't returning an error. Cc: stable@kernel.org Reviewed-by: Suresh Jayaraman Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 67dad54..88c84a3 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1706,9 +1706,6 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) if (ses) { cFYI(1, "Existing smb sess found (status=%d)", ses->status); - /* existing SMB ses has a server reference already */ - cifs_put_tcp_session(server); - mutex_lock(&ses->session_mutex); rc = cifs_negotiate_protocol(xid, ses); if (rc) { @@ -1731,6 +1728,9 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) } } mutex_unlock(&ses->session_mutex); + + /* existing SMB ses has a server reference already */ + cifs_put_tcp_session(server); FreeXid(xid); return ses; } -- cgit v0.10.2 From 75e1c70fc31490ef8a373ea2a4bea2524099b478 Mon Sep 17 00:00:00 2001 From: Jeff Moyer Date: Fri, 10 Sep 2010 14:16:00 -0700 Subject: aio: check for multiplication overflow in do_io_submit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Tavis Ormandy pointed out that do_io_submit does not do proper bounds checking on the passed-in iocb array:        if (unlikely(nr < 0))                return -EINVAL;        if (unlikely(!access_ok(VERIFY_READ, iocbpp, (nr*sizeof(iocbpp)))))                return -EFAULT;                      ^^^^^^^^^^^^^^^^^^ The attached patch checks for overflow, and if it is detected, the number of iocbs submitted is scaled down to a number that will fit in the long.  This is an ok thing to do, as sys_io_submit is documented as returning the number of iocbs submitted, so callers should handle a return value of less than the 'nr' argument passed in. Reported-by: Tavis Ormandy Signed-off-by: Jeff Moyer Signed-off-by: Linus Torvalds diff --git a/fs/aio.c b/fs/aio.c index 3006b5b..1320b2a 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -1659,6 +1659,9 @@ long do_io_submit(aio_context_t ctx_id, long nr, if (unlikely(nr < 0)) return -EINVAL; + if (unlikely(nr > LONG_MAX/sizeof(*iocbpp))) + nr = LONG_MAX/sizeof(*iocbpp); + if (unlikely(!access_ok(VERIFY_READ, iocbpp, (nr*sizeof(*iocbpp))))) return -EFAULT; -- cgit v0.10.2 From a4128b03ff1ff5bc6ea922518a3f36a09b914379 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 14 Sep 2010 17:59:15 +0100 Subject: MN10300: Fix up the IRQ names for the on-chip serial ports Fix up the IRQ names for the MN10300 on-chip serial ports in the driver as request_interrupt() no longer allows names containing slashes, giving a warning like the following if one is encountered: ------------[ cut here ]------------ WARNING: at fs/proc/generic.c:323 __xlate_proc_name+0x62/0x7c() name 'ttySM0/Rx' Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/arch/mn10300/kernel/mn10300-serial.c b/arch/mn10300/kernel/mn10300-serial.c index 9d49073..db509dd 100644 --- a/arch/mn10300/kernel/mn10300-serial.c +++ b/arch/mn10300/kernel/mn10300-serial.c @@ -156,17 +156,17 @@ struct mn10300_serial_port mn10300_serial_port_sif0 = { ._intr = &SC0ICR, ._rxb = &SC0RXB, ._txb = &SC0TXB, - .rx_name = "ttySM0/Rx", - .tx_name = "ttySM0/Tx", + .rx_name = "ttySM0:Rx", + .tx_name = "ttySM0:Tx", #ifdef CONFIG_MN10300_TTYSM0_TIMER8 - .tm_name = "ttySM0/Timer8", + .tm_name = "ttySM0:Timer8", ._tmxmd = &TM8MD, ._tmxbr = &TM8BR, ._tmicr = &TM8ICR, .tm_irq = TM8IRQ, .div_timer = MNSCx_DIV_TIMER_16BIT, #else /* CONFIG_MN10300_TTYSM0_TIMER2 */ - .tm_name = "ttySM0/Timer2", + .tm_name = "ttySM0:Timer2", ._tmxmd = &TM2MD, ._tmxbr = (volatile u16 *) &TM2BR, ._tmicr = &TM2ICR, @@ -209,17 +209,17 @@ struct mn10300_serial_port mn10300_serial_port_sif1 = { ._intr = &SC1ICR, ._rxb = &SC1RXB, ._txb = &SC1TXB, - .rx_name = "ttySM1/Rx", - .tx_name = "ttySM1/Tx", + .rx_name = "ttySM1:Rx", + .tx_name = "ttySM1:Tx", #ifdef CONFIG_MN10300_TTYSM1_TIMER9 - .tm_name = "ttySM1/Timer9", + .tm_name = "ttySM1:Timer9", ._tmxmd = &TM9MD, ._tmxbr = &TM9BR, ._tmicr = &TM9ICR, .tm_irq = TM9IRQ, .div_timer = MNSCx_DIV_TIMER_16BIT, #else /* CONFIG_MN10300_TTYSM1_TIMER3 */ - .tm_name = "ttySM1/Timer3", + .tm_name = "ttySM1:Timer3", ._tmxmd = &TM3MD, ._tmxbr = (volatile u16 *) &TM3BR, ._tmicr = &TM3ICR, @@ -260,9 +260,9 @@ struct mn10300_serial_port mn10300_serial_port_sif2 = { .uart.lock = __SPIN_LOCK_UNLOCKED(mn10300_serial_port_sif2.uart.lock), .name = "ttySM2", - .rx_name = "ttySM2/Rx", - .tx_name = "ttySM2/Tx", - .tm_name = "ttySM2/Timer10", + .rx_name = "ttySM2:Rx", + .tx_name = "ttySM2:Tx", + .tm_name = "ttySM2:Timer10", ._iobase = &SC2CTR, ._control = &SC2CTR, ._status = &SC2STR, -- cgit v0.10.2 From fe725d4f22f6bd1e7a5e7074bdf53a8fe0a954ee Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 14 Sep 2010 10:10:47 -0400 Subject: drm/radeon/kms: only warn on mipmap size checks in r600 cs checker (v2) The texture base address registers are in units of 256 bytes. The original CS checker treated these offsets as bytes, so the original check was wrong. I fixed the units in a patch during the 2.6.36 cycle, but this ended up breaking some existing userspace (probably due to a bug in either userspace texture allocation or the drm texture mipmap checker). So for now, until we come up with a better fix, just warn if the mipmap size it too large. This will keep existing userspace working and it should be just as safe as before when we were checking the wrong units. These are GPU MC addresses, so if they fall outside of the VRAM or GART apertures, they end up at the GPU default page, so this should be safe from a security perspective. v2: Just disable the warning. It just spams the log and there's nothing the user can do about it. Signed-off-by: Alex Deucher Cc: Jerome Glisse Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c index d886494..250a3a9 100644 --- a/drivers/gpu/drm/radeon/r600_cs.c +++ b/drivers/gpu/drm/radeon/r600_cs.c @@ -1170,9 +1170,8 @@ static inline int r600_check_texture_resource(struct radeon_cs_parser *p, u32 i /* using get ib will give us the offset into the mipmap bo */ word0 = radeon_get_ib_value(p, idx + 3) << 8; if ((mipmap_size + word0) > radeon_bo_size(mipmap)) { - dev_warn(p->dev, "mipmap bo too small (%d %d %d %d %d %d -> %d have %ld)\n", - w0, h0, bpe, blevel, nlevels, word0, mipmap_size, radeon_bo_size(texture)); - return -EINVAL; + /*dev_warn(p->dev, "mipmap bo too small (%d %d %d %d %d %d -> %d have %ld)\n", + w0, h0, bpe, blevel, nlevels, word0, mipmap_size, radeon_bo_size(texture));*/ } return 0; } -- cgit v0.10.2 From e75e863dd5c7d96b91ebbd241da5328fc38a78cc Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Tue, 14 Sep 2010 16:35:14 +0200 Subject: sched: Fix user time incorrectly accounted as system time on 32-bit We have 32-bit variable overflow possibility when multiply in task_times() and thread_group_times() functions. When the overflow happens then the scaled utime value becomes erroneously small and the scaled stime becomes i erroneously big. Reported here: https://bugzilla.redhat.com/show_bug.cgi?id=633037 https://bugzilla.kernel.org/show_bug.cgi?id=16559 Reported-by: Michael Chapman Reported-by: Ciriaco Garcia de Celis Signed-off-by: Stanislaw Gruszka Signed-off-by: Peter Zijlstra Cc: Hidetoshi Seto Cc: # 2.6.32.19+ (partially) and 2.6.33+ LKML-Reference: <20100914143513.GB8415@redhat.com> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index ed09d4f..dc85ceb 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -3513,9 +3513,9 @@ void task_times(struct task_struct *p, cputime_t *ut, cputime_t *st) rtime = nsecs_to_cputime(p->se.sum_exec_runtime); if (total) { - u64 temp; + u64 temp = rtime; - temp = (u64)(rtime * utime); + temp *= utime; do_div(temp, total); utime = (cputime_t)temp; } else @@ -3546,9 +3546,9 @@ void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *st) rtime = nsecs_to_cputime(cputime.sum_exec_runtime); if (total) { - u64 temp; + u64 temp = rtime; - temp = (u64)(rtime * cputime.utime); + temp *= cputime.utime; do_div(temp, total); utime = (cputime_t)temp; } else -- cgit v0.10.2 From e6e6c46d759cd013cb57eba112a4129a3a353c4b Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Wed, 15 Sep 2010 11:16:05 -0400 Subject: arch/tile: finish const-ifying sys_execve() The sys_execve() implementation was properly const-ified but not the declaration, the syscall wrappers, or the compat version. This change completes the constification process. Signed-off-by: Chris Metcalf diff --git a/arch/tile/include/asm/compat.h b/arch/tile/include/asm/compat.h index 5a34da6..070ad0a 100644 --- a/arch/tile/include/asm/compat.h +++ b/arch/tile/include/asm/compat.h @@ -214,8 +214,9 @@ extern int compat_setup_rt_frame(int sig, struct k_sigaction *ka, struct compat_sigaction; struct compat_siginfo; struct compat_sigaltstack; -long compat_sys_execve(char __user *path, compat_uptr_t __user *argv, - compat_uptr_t __user *envp); +long compat_sys_execve(const char __user *path, + const compat_uptr_t __user *argv, + const compat_uptr_t __user *envp); long compat_sys_rt_sigaction(int sig, struct compat_sigaction __user *act, struct compat_sigaction __user *oact, size_t sigsetsize); diff --git a/arch/tile/include/asm/syscalls.h b/arch/tile/include/asm/syscalls.h index af165a7..ce99ffe 100644 --- a/arch/tile/include/asm/syscalls.h +++ b/arch/tile/include/asm/syscalls.h @@ -62,10 +62,12 @@ long sys_fork(void); long _sys_fork(struct pt_regs *regs); long sys_vfork(void); long _sys_vfork(struct pt_regs *regs); -long sys_execve(char __user *filename, char __user * __user *argv, - char __user * __user *envp); -long _sys_execve(char __user *filename, char __user * __user *argv, - char __user * __user *envp, struct pt_regs *regs); +long sys_execve(const char __user *filename, + const char __user *const __user *argv, + const char __user *const __user *envp); +long _sys_execve(const char __user *filename, + const char __user *const __user *argv, + const char __user *const __user *envp, struct pt_regs *regs); /* kernel/signal.c */ long sys_sigaltstack(const stack_t __user *, stack_t __user *); @@ -86,10 +88,13 @@ int _sys_cmpxchg_badaddr(unsigned long address, struct pt_regs *); #endif #ifdef CONFIG_COMPAT -long compat_sys_execve(char __user *path, compat_uptr_t __user *argv, - compat_uptr_t __user *envp); -long _compat_sys_execve(char __user *path, compat_uptr_t __user *argv, - compat_uptr_t __user *envp, struct pt_regs *regs); +long compat_sys_execve(const char __user *path, + const compat_uptr_t __user *argv, + const compat_uptr_t __user *envp); +long _compat_sys_execve(const char __user *path, + const compat_uptr_t __user *argv, + const compat_uptr_t __user *envp, + struct pt_regs *regs); long compat_sys_sigaltstack(const struct compat_sigaltstack __user *uss_ptr, struct compat_sigaltstack __user *uoss_ptr); long _compat_sys_sigaltstack(const struct compat_sigaltstack __user *uss_ptr, diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c index 985cc28..88be49e 100644 --- a/arch/tile/kernel/process.c +++ b/arch/tile/kernel/process.c @@ -561,8 +561,9 @@ out: } #ifdef CONFIG_COMPAT -long _compat_sys_execve(char __user *path, compat_uptr_t __user *argv, - compat_uptr_t __user *envp, struct pt_regs *regs) +long _compat_sys_execve(const char __user *path, + const compat_uptr_t __user *argv, + const compat_uptr_t __user *envp, struct pt_regs *regs) { long error; char *filename; -- cgit v0.10.2 From 74fca9da097b74117ae2cef9e5f0d9b0e28ccbb7 Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Wed, 15 Sep 2010 11:16:08 -0400 Subject: arch/tile: Change struct sigcontext to be more useful Rather than just using pt_regs, it now contains the actual saved state explicitly, similar to pt_regs. By doing it this way, we provide a cleaner API for userspace (or equivalently, we avoid the need for libc to provide its own definition of sigcontext). While we're at it, move PT_FLAGS_xxx to where they are not visible from userspace. And always pass siginfo and mcontext to signal handlers, even if they claim they don't need it, since sometimes they actually try to use it anyway in practice. Signed-off-by: Chris Metcalf diff --git a/arch/tile/include/asm/ptrace.h b/arch/tile/include/asm/ptrace.h index acdae81..4a02bb0 100644 --- a/arch/tile/include/asm/ptrace.h +++ b/arch/tile/include/asm/ptrace.h @@ -51,10 +51,7 @@ typedef uint_reg_t pt_reg_t; /* * This struct defines the way the registers are stored on the stack during a - * system call/exception. It should be a multiple of 8 bytes to preserve - * normal stack alignment rules. - * - * Must track and + * system call or exception. "struct sigcontext" has the same shape. */ struct pt_regs { /* Saved main processor registers; 56..63 are special. */ @@ -80,11 +77,6 @@ struct pt_regs { #endif /* __ASSEMBLY__ */ -/* Flag bits in pt_regs.flags */ -#define PT_FLAGS_DISABLE_IRQ 1 /* on return to kernel, disable irqs */ -#define PT_FLAGS_CALLER_SAVES 2 /* caller-save registers are valid */ -#define PT_FLAGS_RESTORE_REGS 4 /* restore callee-save regs on return */ - #define PTRACE_GETREGS 12 #define PTRACE_SETREGS 13 #define PTRACE_GETFPREGS 14 @@ -101,6 +93,11 @@ struct pt_regs { #ifdef __KERNEL__ +/* Flag bits in pt_regs.flags */ +#define PT_FLAGS_DISABLE_IRQ 1 /* on return to kernel, disable irqs */ +#define PT_FLAGS_CALLER_SAVES 2 /* caller-save registers are valid */ +#define PT_FLAGS_RESTORE_REGS 4 /* restore callee-save regs on return */ + #ifndef __ASSEMBLY__ #define instruction_pointer(regs) ((regs)->pc) diff --git a/arch/tile/include/asm/sigcontext.h b/arch/tile/include/asm/sigcontext.h index 7cd7672..5e2d033 100644 --- a/arch/tile/include/asm/sigcontext.h +++ b/arch/tile/include/asm/sigcontext.h @@ -15,13 +15,21 @@ #ifndef _ASM_TILE_SIGCONTEXT_H #define _ASM_TILE_SIGCONTEXT_H -/* NOTE: we can't include due to #include dependencies. */ -#include - -/* Must track */ +#include +/* + * struct sigcontext has the same shape as struct pt_regs, + * but is simplified since we know the fault is from userspace. + */ struct sigcontext { - struct pt_regs regs; + uint_reg_t gregs[53]; /* General-purpose registers. */ + uint_reg_t tp; /* Aliases gregs[TREG_TP]. */ + uint_reg_t sp; /* Aliases gregs[TREG_SP]. */ + uint_reg_t lr; /* Aliases gregs[TREG_LR]. */ + uint_reg_t pc; /* Program counter. */ + uint_reg_t ics; /* In Interrupt Critical Section? */ + uint_reg_t faultnum; /* Fault number. */ + uint_reg_t pad[5]; }; #endif /* _ASM_TILE_SIGCONTEXT_H */ diff --git a/arch/tile/include/asm/signal.h b/arch/tile/include/asm/signal.h index eb0253f..c1ee1d6 100644 --- a/arch/tile/include/asm/signal.h +++ b/arch/tile/include/asm/signal.h @@ -24,6 +24,7 @@ #include #if defined(__KERNEL__) && !defined(__ASSEMBLY__) +struct pt_regs; int restore_sigcontext(struct pt_regs *, struct sigcontext __user *, long *); int setup_sigcontext(struct sigcontext __user *, struct pt_regs *); void do_signal(struct pt_regs *regs); diff --git a/arch/tile/kernel/signal.c b/arch/tile/kernel/signal.c index 45b66a3..ce183aa 100644 --- a/arch/tile/kernel/signal.c +++ b/arch/tile/kernel/signal.c @@ -61,13 +61,19 @@ int restore_sigcontext(struct pt_regs *regs, /* Always make any pending restarted system calls return -EINTR */ current_thread_info()->restart_block.fn = do_no_restart_syscall; + /* + * Enforce that sigcontext is like pt_regs, and doesn't mess + * up our stack alignment rules. + */ + BUILD_BUG_ON(sizeof(struct sigcontext) != sizeof(struct pt_regs)); + BUILD_BUG_ON(sizeof(struct sigcontext) % 8 != 0); + for (i = 0; i < sizeof(struct pt_regs)/sizeof(long); ++i) - err |= __get_user(((long *)regs)[i], - &((long __user *)(&sc->regs))[i]); + err |= __get_user(regs->regs[i], &sc->gregs[i]); regs->faultnum = INT_SWINT_1_SIGRETURN; - err |= __get_user(*pr0, &sc->regs.regs[0]); + err |= __get_user(*pr0, &sc->gregs[0]); return err; } @@ -112,8 +118,7 @@ int setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs) int i, err = 0; for (i = 0; i < sizeof(struct pt_regs)/sizeof(long); ++i) - err |= __put_user(((long *)regs)[i], - &((long __user *)(&sc->regs))[i]); + err |= __put_user(regs->regs[i], &sc->gregs[i]); return err; } @@ -203,19 +208,17 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, * Set up registers for signal handler. * Registers that we don't modify keep the value they had from * user-space at the time we took the signal. + * We always pass siginfo and mcontext, regardless of SA_SIGINFO, + * since some things rely on this (e.g. glibc's debug/segfault.c). */ regs->pc = (unsigned long) ka->sa.sa_handler; regs->ex1 = PL_ICS_EX1(USER_PL, 1); /* set crit sec in handler */ regs->sp = (unsigned long) frame; regs->lr = restorer; regs->regs[0] = (unsigned long) usig; - - if (ka->sa.sa_flags & SA_SIGINFO) { - /* Need extra arguments, so mark to restore caller-saves. */ - regs->regs[1] = (unsigned long) &frame->info; - regs->regs[2] = (unsigned long) &frame->uc; - regs->flags |= PT_FLAGS_CALLER_SAVES; - } + regs->regs[1] = (unsigned long) &frame->info; + regs->regs[2] = (unsigned long) &frame->uc; + regs->flags |= PT_FLAGS_CALLER_SAVES; /* * Notify any tracer that was single-stepping it. diff --git a/arch/tile/kernel/stack.c b/arch/tile/kernel/stack.c index 38a68b0b4..ea2e0ce 100644 --- a/arch/tile/kernel/stack.c +++ b/arch/tile/kernel/stack.c @@ -175,7 +175,7 @@ static struct pt_regs *valid_sigframe(struct KBacktraceIterator* kbt) pr_err(" \n", frame->info.si_signo); } - return &frame->uc.uc_mcontext.regs; + return (struct pt_regs *)&frame->uc.uc_mcontext; } return NULL; } -- cgit v0.10.2 From a802fc685426303ab627b7ad3fd5c97b5dea7e00 Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Wed, 15 Sep 2010 11:16:10 -0400 Subject: arch/tile: Save and restore extra user state for tilegx During context switch, save and restore a couple of additional bits of tilegx user state that can be persistently modified by userspace. Signed-off-by: Chris Metcalf diff --git a/arch/tile/include/arch/chip_tile64.h b/arch/tile/include/arch/chip_tile64.h index 1246573..261aaba 100644 --- a/arch/tile/include/arch/chip_tile64.h +++ b/arch/tile/include/arch/chip_tile64.h @@ -150,6 +150,9 @@ /** Is the PROC_STATUS SPR supported? */ #define CHIP_HAS_PROC_STATUS_SPR() 0 +/** Is the DSTREAM_PF SPR supported? */ +#define CHIP_HAS_DSTREAM_PF() 0 + /** Log of the number of mshims we have. */ #define CHIP_LOG_NUM_MSHIMS() 2 diff --git a/arch/tile/include/arch/chip_tilepro.h b/arch/tile/include/arch/chip_tilepro.h index e864c47..7001769 100644 --- a/arch/tile/include/arch/chip_tilepro.h +++ b/arch/tile/include/arch/chip_tilepro.h @@ -150,6 +150,9 @@ /** Is the PROC_STATUS SPR supported? */ #define CHIP_HAS_PROC_STATUS_SPR() 1 +/** Is the DSTREAM_PF SPR supported? */ +#define CHIP_HAS_DSTREAM_PF() 0 + /** Log of the number of mshims we have. */ #define CHIP_LOG_NUM_MSHIMS() 2 diff --git a/arch/tile/include/asm/processor.h b/arch/tile/include/asm/processor.h index d942d09..ccd5f84 100644 --- a/arch/tile/include/asm/processor.h +++ b/arch/tile/include/asm/processor.h @@ -103,6 +103,18 @@ struct thread_struct { /* Any other miscellaneous processor state bits */ unsigned long proc_status; #endif +#if !CHIP_HAS_FIXED_INTVEC_BASE() + /* Interrupt base for PL0 interrupts */ + unsigned long interrupt_vector_base; +#endif +#if CHIP_HAS_TILE_RTF_HWM() + /* Tile cache retry fifo high-water mark */ + unsigned long tile_rtf_hwm; +#endif +#if CHIP_HAS_DSTREAM_PF() + /* Data stream prefetch control */ + unsigned long dstream_pf; +#endif #ifdef CONFIG_HARDWALL /* Is this task tied to an activated hardwall? */ struct hardwall_info *hardwall; diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c index 88be49e..c37b6e3 100644 --- a/arch/tile/kernel/process.c +++ b/arch/tile/kernel/process.c @@ -408,6 +408,15 @@ static void save_arch_state(struct thread_struct *t) #if CHIP_HAS_PROC_STATUS_SPR() t->proc_status = __insn_mfspr(SPR_PROC_STATUS); #endif +#if !CHIP_HAS_FIXED_INTVEC_BASE() + t->interrupt_vector_base = __insn_mfspr(SPR_INTERRUPT_VECTOR_BASE_0); +#endif +#if CHIP_HAS_TILE_RTF_HWM() + t->tile_rtf_hwm = __insn_mfspr(SPR_TILE_RTF_HWM); +#endif +#if CHIP_HAS_DSTREAM_PF() + t->dstream_pf = __insn_mfspr(SPR_DSTREAM_PF); +#endif } static void restore_arch_state(const struct thread_struct *t) @@ -428,14 +437,14 @@ static void restore_arch_state(const struct thread_struct *t) #if CHIP_HAS_PROC_STATUS_SPR() __insn_mtspr(SPR_PROC_STATUS, t->proc_status); #endif +#if !CHIP_HAS_FIXED_INTVEC_BASE() + __insn_mtspr(SPR_INTERRUPT_VECTOR_BASE_0, t->interrupt_vector_base); +#endif #if CHIP_HAS_TILE_RTF_HWM() - /* - * Clear this whenever we switch back to a process in case - * the previous process was monkeying with it. Even if enabled - * in CBOX_MSR1 via TILE_RTF_HWM_MIN, it's still just a - * performance hint, so isn't worth a full save/restore. - */ - __insn_mtspr(SPR_TILE_RTF_HWM, 0); + __insn_mtspr(SPR_TILE_RTF_HWM, t->tile_rtf_hwm); +#endif +#if CHIP_HAS_DSTREAM_PF() + __insn_mtspr(SPR_DSTREAM_PF, t->dstream_pf); #endif } -- cgit v0.10.2 From 0fab59e5ddb4e0557825e9ab8e41b66b5f336941 Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Wed, 15 Sep 2010 11:17:04 -0400 Subject: arch/tile: fix memcpy_fromio()/memcpy_toio() signatures This tripped up a driver (not yet committed to git). Fix it now. Signed-off-by: Chris Metcalf diff --git a/arch/tile/include/asm/io.h b/arch/tile/include/asm/io.h index 8c95bef..ee43328 100644 --- a/arch/tile/include/asm/io.h +++ b/arch/tile/include/asm/io.h @@ -164,22 +164,22 @@ static inline void _tile_writeq(u64 val, unsigned long addr) #define iowrite32 writel #define iowrite64 writeq -static inline void *memcpy_fromio(void *dst, void *src, int len) +static inline void memcpy_fromio(void *dst, const volatile void __iomem *src, + size_t len) { int x; BUG_ON((unsigned long)src & 0x3); for (x = 0; x < len; x += 4) *(u32 *)(dst + x) = readl(src + x); - return dst; } -static inline void *memcpy_toio(void *dst, void *src, int len) +static inline void memcpy_toio(volatile void __iomem *dst, const void *src, + size_t len) { int x; BUG_ON((unsigned long)dst & 0x3); for (x = 0; x < len; x += 4) writel(*(u32 *)(src + x), dst + x); - return dst; } /* -- cgit v0.10.2 From 7040dea4d2a0609241c7a98a944b7c432c69db2e Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Wed, 15 Sep 2010 11:17:05 -0400 Subject: arch/tile: fix formatting bug in register dumps This cut-and-paste bug was caused by rewriting the register dump code to use only a single printk per line of output. Signed-off-by: Chris Metcalf diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c index c37b6e3..84c2911 100644 --- a/arch/tile/kernel/process.c +++ b/arch/tile/kernel/process.c @@ -667,7 +667,7 @@ void show_regs(struct pt_regs *regs) regs->regs[51], regs->regs[52], regs->tp); pr_err(" sp : "REGFMT" lr : "REGFMT"\n", regs->sp, regs->lr); #else - for (i = 0; i < 52; i += 3) + for (i = 0; i < 52; i += 4) pr_err(" r%-2d: "REGFMT" r%-2d: "REGFMT " r%-2d: "REGFMT" r%-2d: "REGFMT"\n", i, regs->regs[i], i+1, regs->regs[i+1], -- cgit v0.10.2 From c494bc6c534c78fac2d308ad38073b9226448b0d Mon Sep 17 00:00:00 2001 From: Dominik Brodowski Date: Mon, 30 Aug 2010 08:18:54 +0200 Subject: pcmcia serial_cs.c: fix multifunction card handling We shouldn't overwrite pre-set values, and we should also set the port address to the beginning, and not the end of the 8-port range. CC: linux-serial@vger.kernel.org Reported-by: Komuro Hardware-supplied-by: Jochen Frieling Tested-by: Wolfram Sang Signed-off-by: Dominik Brodowski diff --git a/drivers/serial/serial_cs.c b/drivers/serial/serial_cs.c index 141c695..7d475b2 100644 --- a/drivers/serial/serial_cs.c +++ b/drivers/serial/serial_cs.c @@ -335,8 +335,6 @@ static int serial_probe(struct pcmcia_device *link) info->p_dev = link; link->priv = info; - link->resource[0]->flags |= IO_DATA_PATH_WIDTH_8; - link->resource[0]->end = 8; link->conf.Attributes = CONF_ENABLE_IRQ; if (do_sound) { link->conf.Attributes |= CONF_ENABLE_SPKR; @@ -411,6 +409,27 @@ static int setup_serial(struct pcmcia_device *handle, struct serial_info * info, /*====================================================================*/ +static int pfc_config(struct pcmcia_device *p_dev) +{ + unsigned int port = 0; + struct serial_info *info = p_dev->priv; + + if ((p_dev->resource[1]->end != 0) && + (resource_size(p_dev->resource[1]) == 8)) { + port = p_dev->resource[1]->start; + info->slave = 1; + } else if ((info->manfid == MANFID_OSITECH) && + (resource_size(p_dev->resource[0]) == 0x40)) { + port = p_dev->resource[0]->start + 0x28; + info->slave = 1; + } + if (info->slave) + return setup_serial(p_dev, info, port, p_dev->irq); + + dev_warn(&p_dev->dev, "no usable port range found, giving up\n"); + return -ENODEV; +} + static int simple_config_check(struct pcmcia_device *p_dev, cistpl_cftable_entry_t *cf, cistpl_cftable_entry_t *dflt, @@ -461,23 +480,8 @@ static int simple_config(struct pcmcia_device *link) struct serial_info *info = link->priv; int i = -ENODEV, try; - /* If the card is already configured, look up the port and irq */ - if (link->function_config) { - unsigned int port = 0; - if ((link->resource[1]->end != 0) && - (resource_size(link->resource[1]) == 8)) { - port = link->resource[1]->end; - info->slave = 1; - } else if ((info->manfid == MANFID_OSITECH) && - (resource_size(link->resource[0]) == 0x40)) { - port = link->resource[0]->start + 0x28; - info->slave = 1; - } - if (info->slave) { - return setup_serial(link, info, port, - link->irq); - } - } + link->resource[0]->flags |= IO_DATA_PATH_WIDTH_8; + link->resource[0]->end = 8; /* First pass: look for a config entry that looks normal. * Two tries: without IO aliases, then with aliases */ @@ -491,8 +495,7 @@ static int simple_config(struct pcmcia_device *link) if (!pcmcia_loop_config(link, simple_config_check_notpicky, NULL)) goto found_port; - printk(KERN_NOTICE - "serial_cs: no usable port range found, giving up\n"); + dev_warn(&link->dev, "no usable port range found, giving up\n"); return -1; found_port: @@ -558,6 +561,7 @@ static int multi_config(struct pcmcia_device *link) int i, base2 = 0; /* First, look for a generic full-sized window */ + link->resource[0]->flags |= IO_DATA_PATH_WIDTH_8; link->resource[0]->end = info->multi * 8; if (pcmcia_loop_config(link, multi_config_check, &base2)) { /* If that didn't work, look for two windows */ @@ -565,15 +569,14 @@ static int multi_config(struct pcmcia_device *link) info->multi = 2; if (pcmcia_loop_config(link, multi_config_check_notpicky, &base2)) { - printk(KERN_NOTICE "serial_cs: no usable port range" + dev_warn(&link->dev, "no usable port range " "found, giving up\n"); return -ENODEV; } } if (!link->irq) - dev_warn(&link->dev, - "serial_cs: no usable IRQ found, continuing...\n"); + dev_warn(&link->dev, "no usable IRQ found, continuing...\n"); /* * Apply any configuration quirks. @@ -675,6 +678,7 @@ static int serial_config(struct pcmcia_device * link) multifunction cards that ask for appropriate IO port ranges */ if ((info->multi == 0) && (link->has_func_id) && + (link->socket->pcmcia_pfc == 0) && ((link->func_id == CISTPL_FUNCID_MULTI) || (link->func_id == CISTPL_FUNCID_SERIAL))) pcmcia_loop_config(link, serial_check_for_multi, info); @@ -685,7 +689,13 @@ static int serial_config(struct pcmcia_device * link) if (info->quirk && info->quirk->multi != -1) info->multi = info->quirk->multi; - if (info->multi > 1) + dev_info(&link->dev, + "trying to set up [0x%04x:0x%04x] (pfc: %d, multi: %d, quirk: %p)\n", + link->manf_id, link->card_id, + link->socket->pcmcia_pfc, info->multi, info->quirk); + if (link->socket->pcmcia_pfc) + i = pfc_config(link); + else if (info->multi > 1) i = multi_config(link); else i = simple_config(link); @@ -704,7 +714,7 @@ static int serial_config(struct pcmcia_device * link) return 0; failed: - dev_warn(&link->dev, "serial_cs: failed to initialize\n"); + dev_warn(&link->dev, "failed to initialize\n"); serial_remove(link); return -ENODEV; } -- cgit v0.10.2 From eb838fe109b8f51ba590802761753a2631c3f7f0 Mon Sep 17 00:00:00 2001 From: Dominik Brodowski Date: Mon, 13 Sep 2010 16:51:36 +0200 Subject: pcmcia: per-device, not per-socket debug messages As the iomem / ioport setup differs per device, it is much better to print out the device instead of the socket. Tested-by: Wolfram Sang Signed-off-by: Dominik Brodowski diff --git a/drivers/pcmcia/pcmcia_resource.c b/drivers/pcmcia/pcmcia_resource.c index 54aa1c2..a5c1765 100644 --- a/drivers/pcmcia/pcmcia_resource.c +++ b/drivers/pcmcia/pcmcia_resource.c @@ -163,7 +163,7 @@ static int pcmcia_access_config(struct pcmcia_device *p_dev, c = p_dev->function_config; if (!(c->state & CONFIG_LOCKED)) { - dev_dbg(&s->dev, "Configuration isnt't locked\n"); + dev_dbg(&p_dev->dev, "Configuration isnt't locked\n"); mutex_unlock(&s->ops_mutex); return -EACCES; } @@ -220,7 +220,7 @@ int pcmcia_map_mem_page(struct pcmcia_device *p_dev, window_handle_t wh, s->win[w].card_start = offset; ret = s->ops->set_mem_map(s, &s->win[w]); if (ret) - dev_warn(&s->dev, "failed to set_mem_map\n"); + dev_warn(&p_dev->dev, "failed to set_mem_map\n"); mutex_unlock(&s->ops_mutex); return ret; } /* pcmcia_map_mem_page */ @@ -244,18 +244,18 @@ int pcmcia_modify_configuration(struct pcmcia_device *p_dev, c = p_dev->function_config; if (!(s->state & SOCKET_PRESENT)) { - dev_dbg(&s->dev, "No card present\n"); + dev_dbg(&p_dev->dev, "No card present\n"); ret = -ENODEV; goto unlock; } if (!(c->state & CONFIG_LOCKED)) { - dev_dbg(&s->dev, "Configuration isnt't locked\n"); + dev_dbg(&p_dev->dev, "Configuration isnt't locked\n"); ret = -EACCES; goto unlock; } if (mod->Attributes & (CONF_IRQ_CHANGE_VALID | CONF_VCC_CHANGE_VALID)) { - dev_dbg(&s->dev, + dev_dbg(&p_dev->dev, "changing Vcc or IRQ is not allowed at this time\n"); ret = -EINVAL; goto unlock; @@ -265,20 +265,22 @@ int pcmcia_modify_configuration(struct pcmcia_device *p_dev, if ((mod->Attributes & CONF_VPP1_CHANGE_VALID) && (mod->Attributes & CONF_VPP2_CHANGE_VALID)) { if (mod->Vpp1 != mod->Vpp2) { - dev_dbg(&s->dev, "Vpp1 and Vpp2 must be the same\n"); + dev_dbg(&p_dev->dev, + "Vpp1 and Vpp2 must be the same\n"); ret = -EINVAL; goto unlock; } s->socket.Vpp = mod->Vpp1; if (s->ops->set_socket(s, &s->socket)) { - dev_printk(KERN_WARNING, &s->dev, + dev_printk(KERN_WARNING, &p_dev->dev, "Unable to set VPP\n"); ret = -EIO; goto unlock; } } else if ((mod->Attributes & CONF_VPP1_CHANGE_VALID) || (mod->Attributes & CONF_VPP2_CHANGE_VALID)) { - dev_dbg(&s->dev, "changing Vcc is not allowed at this time\n"); + dev_dbg(&p_dev->dev, + "changing Vcc is not allowed at this time\n"); ret = -EINVAL; goto unlock; } @@ -401,7 +403,7 @@ int pcmcia_release_window(struct pcmcia_device *p_dev, struct resource *res) win = &s->win[w]; if (!(p_dev->_win & CLIENT_WIN_REQ(w))) { - dev_dbg(&s->dev, "not releasing unknown window\n"); + dev_dbg(&p_dev->dev, "not releasing unknown window\n"); mutex_unlock(&s->ops_mutex); return -EINVAL; } @@ -439,7 +441,7 @@ int pcmcia_request_configuration(struct pcmcia_device *p_dev, return -ENODEV; if (req->IntType & INT_CARDBUS) { - dev_dbg(&s->dev, "IntType may not be INT_CARDBUS\n"); + dev_dbg(&p_dev->dev, "IntType may not be INT_CARDBUS\n"); return -EINVAL; } @@ -447,7 +449,7 @@ int pcmcia_request_configuration(struct pcmcia_device *p_dev, c = p_dev->function_config; if (c->state & CONFIG_LOCKED) { mutex_unlock(&s->ops_mutex); - dev_dbg(&s->dev, "Configuration is locked\n"); + dev_dbg(&p_dev->dev, "Configuration is locked\n"); return -EACCES; } @@ -455,7 +457,7 @@ int pcmcia_request_configuration(struct pcmcia_device *p_dev, s->socket.Vpp = req->Vpp; if (s->ops->set_socket(s, &s->socket)) { mutex_unlock(&s->ops_mutex); - dev_printk(KERN_WARNING, &s->dev, + dev_printk(KERN_WARNING, &p_dev->dev, "Unable to set socket state\n"); return -EINVAL; } @@ -569,19 +571,20 @@ int pcmcia_request_io(struct pcmcia_device *p_dev) int ret = -EINVAL; mutex_lock(&s->ops_mutex); - dev_dbg(&s->dev, "pcmcia_request_io: %pR , %pR", &c->io[0], &c->io[1]); + dev_dbg(&p_dev->dev, "pcmcia_request_io: %pR , %pR", + &c->io[0], &c->io[1]); if (!(s->state & SOCKET_PRESENT)) { - dev_dbg(&s->dev, "pcmcia_request_io: No card present\n"); + dev_dbg(&p_dev->dev, "pcmcia_request_io: No card present\n"); goto out; } if (c->state & CONFIG_LOCKED) { - dev_dbg(&s->dev, "Configuration is locked\n"); + dev_dbg(&p_dev->dev, "Configuration is locked\n"); goto out; } if (c->state & CONFIG_IO_REQ) { - dev_dbg(&s->dev, "IO already configured\n"); + dev_dbg(&p_dev->dev, "IO already configured\n"); goto out; } @@ -601,7 +604,7 @@ int pcmcia_request_io(struct pcmcia_device *p_dev) c->state |= CONFIG_IO_REQ; p_dev->_io = 1; - dev_dbg(&s->dev, "pcmcia_request_io succeeded: %pR , %pR", + dev_dbg(&p_dev->dev, "pcmcia_request_io succeeded: %pR , %pR", &c->io[0], &c->io[1]); out: mutex_unlock(&s->ops_mutex); @@ -800,7 +803,7 @@ int pcmcia_request_window(struct pcmcia_device *p_dev, win_req_t *req, window_ha int w; if (!(s->state & SOCKET_PRESENT)) { - dev_dbg(&s->dev, "No card present\n"); + dev_dbg(&p_dev->dev, "No card present\n"); return -ENODEV; } @@ -809,12 +812,12 @@ int pcmcia_request_window(struct pcmcia_device *p_dev, win_req_t *req, window_ha req->Size = s->map_size; align = (s->features & SS_CAP_MEM_ALIGN) ? req->Size : s->map_size; if (req->Size & (s->map_size-1)) { - dev_dbg(&s->dev, "invalid map size\n"); + dev_dbg(&p_dev->dev, "invalid map size\n"); return -EINVAL; } if ((req->Base && (s->features & SS_CAP_STATIC_MAP)) || (req->Base & (align-1))) { - dev_dbg(&s->dev, "invalid base address\n"); + dev_dbg(&p_dev->dev, "invalid base address\n"); return -EINVAL; } if (req->Base) @@ -826,7 +829,7 @@ int pcmcia_request_window(struct pcmcia_device *p_dev, win_req_t *req, window_ha if (!(s->state & SOCKET_WIN_REQ(w))) break; if (w == MAX_WIN) { - dev_dbg(&s->dev, "all windows are used already\n"); + dev_dbg(&p_dev->dev, "all windows are used already\n"); mutex_unlock(&s->ops_mutex); return -EINVAL; } @@ -837,7 +840,7 @@ int pcmcia_request_window(struct pcmcia_device *p_dev, win_req_t *req, window_ha win->res = pcmcia_find_mem_region(req->Base, req->Size, align, 0, s); if (!win->res) { - dev_dbg(&s->dev, "allocating mem region failed\n"); + dev_dbg(&p_dev->dev, "allocating mem region failed\n"); mutex_unlock(&s->ops_mutex); return -EINVAL; } @@ -851,7 +854,7 @@ int pcmcia_request_window(struct pcmcia_device *p_dev, win_req_t *req, window_ha win->card_start = 0; if (s->ops->set_mem_map(s, win) != 0) { - dev_dbg(&s->dev, "failed to set memory mapping\n"); + dev_dbg(&p_dev->dev, "failed to set memory mapping\n"); mutex_unlock(&s->ops_mutex); return -EIO; } @@ -874,7 +877,7 @@ int pcmcia_request_window(struct pcmcia_device *p_dev, win_req_t *req, window_ha if (win->res) request_resource(&iomem_resource, res); - dev_dbg(&s->dev, "request_window results in %pR\n", res); + dev_dbg(&p_dev->dev, "request_window results in %pR\n", res); mutex_unlock(&s->ops_mutex); *wh = res; -- cgit v0.10.2 From b76dc0546709aef18f123847680108c2fd33f203 Mon Sep 17 00:00:00 2001 From: Dominik Brodowski Date: Mon, 13 Sep 2010 20:23:12 +0200 Subject: pcmcia pcnet_cs: try setting io_lines to 16 if card setup fails Some pcnet_cs compatible cards require an exact 16-lines match of the ioport areas specified in CIS, but set the "iolines" value in the CIS incorrectly. We can easily work around this issue -- same as we do in serial_cs -- by first trying setting iolines to the CIS-specified value, and then trying a 16-line match. Reported-and-tested-by: Wolfram Sang Hardware-supplied-by: Jochen Frieling CC: netdev@vger.kernel.org Signed-off-by: Dominik Brodowski diff --git a/drivers/net/pcmcia/pcnet_cs.c b/drivers/net/pcmcia/pcnet_cs.c index 49279b0..f9b509a 100644 --- a/drivers/net/pcmcia/pcnet_cs.c +++ b/drivers/net/pcmcia/pcnet_cs.c @@ -508,7 +508,8 @@ static int pcnet_confcheck(struct pcmcia_device *p_dev, unsigned int vcc, void *priv_data) { - int *has_shmem = priv_data; + int *priv = priv_data; + int try = (*priv & 0x1); int i; cistpl_io_t *io = &cfg->io; @@ -525,77 +526,103 @@ static int pcnet_confcheck(struct pcmcia_device *p_dev, i = p_dev->resource[1]->end = 0; } - *has_shmem = ((cfg->mem.nwin == 1) && - (cfg->mem.win[0].len >= 0x4000)); + *priv &= ((cfg->mem.nwin == 1) && + (cfg->mem.win[0].len >= 0x4000)) ? 0x10 : ~0x10; + p_dev->resource[0]->start = io->win[i].base; p_dev->resource[0]->end = io->win[i].len; - p_dev->io_lines = io->flags & CISTPL_IO_LINES_MASK; + if (!try) + p_dev->io_lines = io->flags & CISTPL_IO_LINES_MASK; + else + p_dev->io_lines = 16; if (p_dev->resource[0]->end + p_dev->resource[1]->end >= 32) return try_io_port(p_dev); - return 0; + return -EINVAL; +} + +static hw_info_t *pcnet_try_config(struct pcmcia_device *link, + int *has_shmem, int try) +{ + struct net_device *dev = link->priv; + hw_info_t *local_hw_info; + pcnet_dev_t *info = PRIV(dev); + int priv = try; + int ret; + + ret = pcmcia_loop_config(link, pcnet_confcheck, &priv); + if (ret) { + dev_warn(&link->dev, "no useable port range found\n"); + return NULL; + } + *has_shmem = (priv & 0x10); + + if (!link->irq) + return NULL; + + if (resource_size(link->resource[1]) == 8) { + link->conf.Attributes |= CONF_ENABLE_SPKR; + link->conf.Status = CCSR_AUDIO_ENA; + } + if ((link->manf_id == MANFID_IBM) && + (link->card_id == PRODID_IBM_HOME_AND_AWAY)) + link->conf.ConfigIndex |= 0x10; + + ret = pcmcia_request_configuration(link, &link->conf); + if (ret) + return NULL; + + dev->irq = link->irq; + dev->base_addr = link->resource[0]->start; + + if (info->flags & HAS_MISC_REG) { + if ((if_port == 1) || (if_port == 2)) + dev->if_port = if_port; + else + dev_notice(&link->dev, "invalid if_port requested\n"); + } else + dev->if_port = 0; + + if ((link->conf.ConfigBase == 0x03c0) && + (link->manf_id == 0x149) && (link->card_id == 0xc1ab)) { + dev_info(&link->dev, + "this is an AX88190 card - use axnet_cs instead.\n"); + return NULL; + } + + local_hw_info = get_hwinfo(link); + if (!local_hw_info) + local_hw_info = get_prom(link); + if (!local_hw_info) + local_hw_info = get_dl10019(link); + if (!local_hw_info) + local_hw_info = get_ax88190(link); + if (!local_hw_info) + local_hw_info = get_hwired(link); + + return local_hw_info; } static int pcnet_config(struct pcmcia_device *link) { struct net_device *dev = link->priv; pcnet_dev_t *info = PRIV(dev); - int ret, start_pg, stop_pg, cm_offset; + int start_pg, stop_pg, cm_offset; int has_shmem = 0; hw_info_t *local_hw_info; dev_dbg(&link->dev, "pcnet_config\n"); - ret = pcmcia_loop_config(link, pcnet_confcheck, &has_shmem); - if (ret) - goto failed; - - if (!link->irq) - goto failed; - - if (resource_size(link->resource[1]) == 8) { - link->conf.Attributes |= CONF_ENABLE_SPKR; - link->conf.Status = CCSR_AUDIO_ENA; - } - if ((link->manf_id == MANFID_IBM) && - (link->card_id == PRODID_IBM_HOME_AND_AWAY)) - link->conf.ConfigIndex |= 0x10; - - ret = pcmcia_request_configuration(link, &link->conf); - if (ret) - goto failed; - dev->irq = link->irq; - dev->base_addr = link->resource[0]->start; - if (info->flags & HAS_MISC_REG) { - if ((if_port == 1) || (if_port == 2)) - dev->if_port = if_port; - else - printk(KERN_NOTICE "pcnet_cs: invalid if_port requested\n"); - } else { - dev->if_port = 0; - } - - if ((link->conf.ConfigBase == 0x03c0) && - (link->manf_id == 0x149) && (link->card_id == 0xc1ab)) { - printk(KERN_INFO "pcnet_cs: this is an AX88190 card!\n"); - printk(KERN_INFO "pcnet_cs: use axnet_cs instead.\n"); - goto failed; - } - - local_hw_info = get_hwinfo(link); - if (local_hw_info == NULL) - local_hw_info = get_prom(link); - if (local_hw_info == NULL) - local_hw_info = get_dl10019(link); - if (local_hw_info == NULL) - local_hw_info = get_ax88190(link); - if (local_hw_info == NULL) - local_hw_info = get_hwired(link); - - if (local_hw_info == NULL) { - printk(KERN_NOTICE "pcnet_cs: unable to read hardware net" - " address for io base %#3lx\n", dev->base_addr); - goto failed; + local_hw_info = pcnet_try_config(link, &has_shmem, 0); + if (!local_hw_info) { + /* check whether forcing io_lines to 16 helps... */ + pcmcia_disable_device(link); + local_hw_info = pcnet_try_config(link, &has_shmem, 1); + if (local_hw_info == NULL) { + dev_notice(&link->dev, "unable to read hardware net" + " address for io base %#3lx\n", dev->base_addr); + goto failed; + } } info->flags = local_hw_info->flags; -- cgit v0.10.2 From 359f64f7b3997e94ee71039b5fcdc1278b9b77c4 Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Wed, 15 Sep 2010 10:18:51 -0700 Subject: omap: Fix compile dependency to LEDS_CLASS If we LEDS_CLASS is not selected, we will get undefined reference to `led_classdev_register'. Signed-off-by: Tony Lindgren diff --git a/arch/arm/plat-omap/Kconfig b/arch/arm/plat-omap/Kconfig index e39a417..a92cb49 100644 --- a/arch/arm/plat-omap/Kconfig +++ b/arch/arm/plat-omap/Kconfig @@ -33,7 +33,7 @@ config OMAP_DEBUG_DEVICES config OMAP_DEBUG_LEDS bool depends on OMAP_DEBUG_DEVICES - default y if LEDS + default y if LEDS_CLASS config OMAP_RESET_CLOCKS bool "Reset unused clocks during boot" -- cgit v0.10.2 From ae44855ae8b36e4194a0a43eec6351e81f880955 Mon Sep 17 00:00:00 2001 From: Akinobu Mita Date: Sat, 21 Aug 2010 18:27:50 +0900 Subject: watchdog: sb_wdog: release irq and reboot notifier in error path and module_exit() irq and reboot notifier are acquired in module_init() but never released. They should be released correctly, otherwise reloading the module or error during module_init() will cause a problem. Signed-off-by: Akinobu Mita Cc: Andrew Sharp Signed-off-by: Wim Van Sebroeck diff --git a/drivers/watchdog/sb_wdog.c b/drivers/watchdog/sb_wdog.c index 88c83aa..f31493e 100644 --- a/drivers/watchdog/sb_wdog.c +++ b/drivers/watchdog/sb_wdog.c @@ -305,7 +305,7 @@ static int __init sbwdog_init(void) if (ret) { printk(KERN_ERR "%s: failed to request irq 1 - %d\n", ident.identity, ret); - return ret; + goto out; } ret = misc_register(&sbwdog_miscdev); @@ -313,14 +313,20 @@ static int __init sbwdog_init(void) printk(KERN_INFO "%s: timeout is %ld.%ld secs\n", ident.identity, timeout / 1000000, (timeout / 100000) % 10); - } else - free_irq(1, (void *)user_dog); + return 0; + } + free_irq(1, (void *)user_dog); +out: + unregister_reboot_notifier(&sbwdog_notifier); + return ret; } static void __exit sbwdog_exit(void) { misc_deregister(&sbwdog_miscdev); + free_irq(1, (void *)user_dog); + unregister_reboot_notifier(&sbwdog_notifier); } module_init(sbwdog_init); -- cgit v0.10.2 From 0e901bed4e053098f1c8411dcbf21324b7f61775 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Sun, 29 Aug 2010 13:53:14 +0300 Subject: watchdog: ts72xx_wdt: disable watchdog at probe Since it may be already enabled by bootloader or some other utility. This patch makes sure that the watchdog is disabled before any userspace daemon opens the device. It is also required by the watchdog API. Signed-off-by: Mika Westerberg Signed-off-by: Wim Van Sebroeck diff --git a/drivers/watchdog/ts72xx_wdt.c b/drivers/watchdog/ts72xx_wdt.c index 458c499..18cdeb4 100644 --- a/drivers/watchdog/ts72xx_wdt.c +++ b/drivers/watchdog/ts72xx_wdt.c @@ -449,6 +449,9 @@ static __devinit int ts72xx_wdt_probe(struct platform_device *pdev) wdt->pdev = pdev; mutex_init(&wdt->lock); + /* make sure that the watchdog is disabled */ + ts72xx_wdt_stop(wdt); + error = misc_register(&ts72xx_wdt_miscdev); if (error) { dev_err(&pdev->dev, "failed to register miscdev\n"); -- cgit v0.10.2 From 0a18e15598274b79ce14342ce0bfb76a87dadb45 Mon Sep 17 00:00:00 2001 From: Kevin Wells Date: Tue, 17 Aug 2010 17:45:28 -0700 Subject: watchdog: Enable NXP LPC32XX support in Kconfig (resend) The NXP LPC32XX processor use the same watchdog as the Philips PNX4008 processor. Signed-off-by: Kevin Wells Tested-by: Wolfram Sang Signed-off-by: Wim Van Sebroeck diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig index b036677..24efd8e 100644 --- a/drivers/watchdog/Kconfig +++ b/drivers/watchdog/Kconfig @@ -213,11 +213,11 @@ config OMAP_WATCHDOG here to enable the OMAP1610/OMAP1710/OMAP2420/OMAP3430/OMAP4430 watchdog timer. config PNX4008_WATCHDOG - tristate "PNX4008 Watchdog" - depends on ARCH_PNX4008 + tristate "PNX4008 and LPC32XX Watchdog" + depends on ARCH_PNX4008 || ARCH_LPC32XX help Say Y here if to include support for the watchdog timer - in the PNX4008 processor. + in the PNX4008 or LPC32XX processor. This driver can be built as a module by choosing M. The module will be called pnx4008_wdt. -- cgit v0.10.2 From 2d2b6901649a62977452be85df53eda2412def24 Mon Sep 17 00:00:00 2001 From: Petr Tesarik Date: Wed, 15 Sep 2010 15:35:48 -0700 Subject: [IA64] Optimize ticket spinlocks in fsys_rt_sigprocmask Tony's fix (f574c843191728d9407b766a027f779dcd27b272) has a small bug, it incorrectly uses "r3" as a scratch register in the first of the two unlock paths ... it is also inefficient. Optimize the fast path again. Signed-off-by: Petr Tesarik Signed-off-by: Tony Luck diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S index 471a1e7..331d42b 100644 --- a/arch/ia64/kernel/fsys.S +++ b/arch/ia64/kernel/fsys.S @@ -420,34 +420,31 @@ EX(.fail_efault, ld8 r14=[r33]) // r14 <- *set ;; RSM_PSR_I(p0, r18, r19) // mask interrupt delivery - mov ar.ccv=0 andcm r14=r14,r17 // filter out SIGKILL & SIGSTOP + mov r8=EINVAL // default to EINVAL #ifdef CONFIG_SMP // __ticket_spin_trylock(r31) ld4 r17=[r31] - mov r8=EINVAL // default to EINVAL - ;; - extr r9=r17,17,15 ;; - xor r18=r17,r9 + mov.m ar.ccv=r17 + extr.u r9=r17,17,15 adds r19=1,r17 + extr.u r18=r17,0,15 ;; - extr.u r18=r18,0,15 + cmp.eq p6,p7=r9,r18 ;; - cmp.eq p0,p7=0,r18 +(p6) cmpxchg4.acq r9=[r31],r19,ar.ccv +(p6) dep.z r20=r19,1,15 // next serving ticket for unlock (p7) br.cond.spnt.many .lock_contention - mov.m ar.ccv=r17 - ;; - cmpxchg4.acq r9=[r31],r19,ar.ccv ;; cmp4.eq p0,p7=r9,r17 + adds r31=2,r31 (p7) br.cond.spnt.many .lock_contention ld8 r3=[r2] // re-read current->blocked now that we hold the lock ;; #else ld8 r3=[r2] // re-read current->blocked now that we hold the lock - mov r8=EINVAL // default to EINVAL #endif add r18=IA64_TASK_PENDING_OFFSET+IA64_SIGPENDING_SIGNAL_OFFSET,r16 add r19=IA64_TASK_SIGNAL_OFFSET,r16 @@ -503,16 +500,8 @@ EX(.fail_efault, ld8 r14=[r33]) // r14 <- *set #ifdef CONFIG_SMP // __ticket_spin_unlock(r31) - adds r31=2,r31 - ;; - ld2.bias r2=[r31] - mov r3=65534 - ;; - adds r2=2,r2 - ;; - and r3=r3,r2 - ;; - st2.rel [r31]=r3 + st2.rel [r31]=r20 + mov r20=0 // i must not leak kernel bits... #endif SSM_PSR_I(p0, p9, r31) ;; @@ -535,16 +524,7 @@ EX(.fail_efault, (p15) st8 [r34]=r3) .sig_pending: #ifdef CONFIG_SMP // __ticket_spin_unlock(r31) - adds r31=2,r31 - ;; - ld2.bias r2=[r31] - mov r3=65534 - ;; - adds r2=2,r2 - ;; - and r3=r3,r2 - ;; - st2.rel [r31]=r3 + st2.rel [r31]=r20 // release the lock #endif SSM_PSR_I(p0, p9, r17) ;; -- cgit v0.10.2 From b4aaa78f4c2f9cde2f335b14f4ca30b01f9651ca Mon Sep 17 00:00:00 2001 From: Dan Rosenberg Date: Wed, 15 Sep 2010 19:08:24 -0400 Subject: drivers/video/via/ioctl.c: prevent reading uninitialized stack memory The VIAFB_GET_INFO device ioctl allows unprivileged users to read 246 bytes of uninitialized stack memory, because the "reserved" member of the viafb_ioctl_info struct declared on the stack is not altered or zeroed before being copied back to the user. This patch takes care of it. Signed-off-by: Dan Rosenberg Signed-off-by: Florian Tobias Schandinat diff --git a/drivers/video/via/ioctl.c b/drivers/video/via/ioctl.c index da03c07..4d553d0 100644 --- a/drivers/video/via/ioctl.c +++ b/drivers/video/via/ioctl.c @@ -25,6 +25,8 @@ int viafb_ioctl_get_viafb_info(u_long arg) { struct viafb_ioctl_info viainfo; + memset(&viainfo, 0, sizeof(struct viafb_ioctl_info)); + viainfo.viafb_id = VIAID; viainfo.vendor_id = PCI_VIA_VENDOR_ID; -- cgit v0.10.2 From 145a902bfeb1f89a41165bd2d1e633ce070bcb73 Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Thu, 16 Sep 2010 10:07:53 +0200 Subject: ALSA: HDA: Enable internal speaker on Dell M101z BugLink: http://launchpad.net/bugs/640254 In some cases a magic processing coefficient is needed to enable the internal speaker on Dell M101z. According to Realtek, this processing coefficient is only present on ALC269vb. Cc: stable@kernel.org Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index bcbf916..a1312a6 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -14453,6 +14453,7 @@ static void alc269_auto_init(struct hda_codec *codec) enum { ALC269_FIXUP_SONY_VAIO, + ALC269_FIXUP_DELL_M101Z, }; static const struct hda_verb alc269_sony_vaio_fixup_verbs[] = { @@ -14464,11 +14465,20 @@ static const struct alc_fixup alc269_fixups[] = { [ALC269_FIXUP_SONY_VAIO] = { .verbs = alc269_sony_vaio_fixup_verbs }, + [ALC269_FIXUP_DELL_M101Z] = { + .verbs = (const struct hda_verb[]) { + /* Enables internal speaker */ + {0x20, AC_VERB_SET_COEF_INDEX, 13}, + {0x20, AC_VERB_SET_PROC_COEF, 0x4040}, + {} + } + }, }; static struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x104d, 0x9071, "Sony VAIO", ALC269_FIXUP_SONY_VAIO), SND_PCI_QUIRK(0x104d, 0x9077, "Sony VAIO", ALC269_FIXUP_SONY_VAIO), + SND_PCI_QUIRK(0x1028, 0x0470, "Dell M101z", ALC269_FIXUP_DELL_M101Z), {} }; -- cgit v0.10.2 From c33f543d320843e1732534c3931da4bbd18e6c14 Mon Sep 17 00:00:00 2001 From: Patrick Simmons Date: Wed, 8 Sep 2010 10:34:28 -0400 Subject: oprofile: Add Support for Intel CPU Family 6 / Model 22 (Intel Celeron 540) This patch adds CPU type detection for the Intel Celeron 540, which is part of the Core 2 family according to Wikipedia; the family and ID pair is absent from the Volume 3B table referenced in the source code comments. I have tested this patch on an Intel Celeron 540 machine reporting itself as Family 6 Model 22, and OProfile runs on the machine without issue. Spec: http://download.intel.com/design/mobile/SPECUPDT/317667.pdf Signed-off-by: Patrick Simmons Acked-by: Andi Kleen Acked-by: Arnd Bergmann Cc: stable@kernel.org Signed-off-by: Robert Richter diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index cfe4faa..009b819 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -671,7 +671,9 @@ static int __init ppro_init(char **cpu_type) case 14: *cpu_type = "i386/core"; break; - case 15: case 23: + case 0x0f: + case 0x16: + case 0x17: *cpu_type = "i386/core_2"; break; case 0x1a: -- cgit v0.10.2 From 37880c909c6a22674d3c0f83f2737264b4e60fe1 Mon Sep 17 00:00:00 2001 From: christophe leroy Date: Thu, 16 Sep 2010 09:05:25 +0200 Subject: spi/mpc8xxx: fix buffer overrun on large transfers It fixes an issue when sending-only or receiving-only more than PAGE_SIZE bytes. Signed-off-by: christophe leroy Acked-by: Joakim Tjernlund Signed-off-by: Grant Likely diff --git a/drivers/spi/spi_mpc8xxx.c b/drivers/spi/spi_mpc8xxx.c index d31b57f..1dd86b8 100644 --- a/drivers/spi/spi_mpc8xxx.c +++ b/drivers/spi/spi_mpc8xxx.c @@ -408,11 +408,17 @@ static void mpc8xxx_spi_cpm_bufs_start(struct mpc8xxx_spi *mspi) xfer_ofs = mspi->xfer_in_progress->len - mspi->count; - out_be32(&rx_bd->cbd_bufaddr, mspi->rx_dma + xfer_ofs); + if (mspi->rx_dma == mspi->dma_dummy_rx) + out_be32(&rx_bd->cbd_bufaddr, mspi->rx_dma); + else + out_be32(&rx_bd->cbd_bufaddr, mspi->rx_dma + xfer_ofs); out_be16(&rx_bd->cbd_datlen, 0); out_be16(&rx_bd->cbd_sc, BD_SC_EMPTY | BD_SC_INTRPT | BD_SC_WRAP); - out_be32(&tx_bd->cbd_bufaddr, mspi->tx_dma + xfer_ofs); + if (mspi->tx_dma == mspi->dma_dummy_tx) + out_be32(&tx_bd->cbd_bufaddr, mspi->tx_dma); + else + out_be32(&tx_bd->cbd_bufaddr, mspi->tx_dma + xfer_ofs); out_be16(&tx_bd->cbd_datlen, xfer_len); out_be16(&tx_bd->cbd_sc, BD_SC_READY | BD_SC_INTRPT | BD_SC_WRAP | BD_SC_LAST); -- cgit v0.10.2 From 8699a0b657b43fa6401537dfe345bee7aa8115ec Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 16 Sep 2010 22:52:32 +0200 Subject: ALSA: pcm - Fix unbalanced pm_qos_request The pm_qos_request isn't freed properly when OSS PCM emulation is used because it skips snd_pcm_hw_free() call but directly releases the stream. This resulted in Oops later. Tested-by: Simon Kirby Signed-off-by: Takashi Iwai diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c index 134fc6c..d4eb2ef 100644 --- a/sound/core/pcm_native.c +++ b/sound/core/pcm_native.c @@ -1992,6 +1992,8 @@ void snd_pcm_release_substream(struct snd_pcm_substream *substream) substream->ops->close(substream); substream->hw_opened = 0; } + if (pm_qos_request_active(&substream->latency_pm_qos_req)) + pm_qos_remove_request(&substream->latency_pm_qos_req); if (substream->pcm_release) { substream->pcm_release(substream); substream->pcm_release = NULL; -- cgit v0.10.2 From 901d46d5a8eb821b03ca9e8cf005beb0c92f31ea Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 16 Sep 2010 23:06:50 +0200 Subject: ALSA: pcm - Fix race with proc files The PCM proc files may open a race against substream close, which can end up with an Oops. Use the open_mutex to protect for it. Signed-off-by: Takashi Iwai diff --git a/sound/core/pcm.c b/sound/core/pcm.c index 204af48..ac242a3 100644 --- a/sound/core/pcm.c +++ b/sound/core/pcm.c @@ -372,14 +372,17 @@ static void snd_pcm_substream_proc_hw_params_read(struct snd_info_entry *entry, struct snd_info_buffer *buffer) { struct snd_pcm_substream *substream = entry->private_data; - struct snd_pcm_runtime *runtime = substream->runtime; + struct snd_pcm_runtime *runtime; + + mutex_lock(&substream->pcm->open_mutex); + runtime = substream->runtime; if (!runtime) { snd_iprintf(buffer, "closed\n"); - return; + goto unlock; } if (runtime->status->state == SNDRV_PCM_STATE_OPEN) { snd_iprintf(buffer, "no setup\n"); - return; + goto unlock; } snd_iprintf(buffer, "access: %s\n", snd_pcm_access_name(runtime->access)); snd_iprintf(buffer, "format: %s\n", snd_pcm_format_name(runtime->format)); @@ -398,20 +401,25 @@ static void snd_pcm_substream_proc_hw_params_read(struct snd_info_entry *entry, snd_iprintf(buffer, "OSS period frames: %lu\n", (unsigned long)runtime->oss.period_frames); } #endif + unlock: + mutex_unlock(&substream->pcm->open_mutex); } static void snd_pcm_substream_proc_sw_params_read(struct snd_info_entry *entry, struct snd_info_buffer *buffer) { struct snd_pcm_substream *substream = entry->private_data; - struct snd_pcm_runtime *runtime = substream->runtime; + struct snd_pcm_runtime *runtime; + + mutex_lock(&substream->pcm->open_mutex); + runtime = substream->runtime; if (!runtime) { snd_iprintf(buffer, "closed\n"); - return; + goto unlock; } if (runtime->status->state == SNDRV_PCM_STATE_OPEN) { snd_iprintf(buffer, "no setup\n"); - return; + goto unlock; } snd_iprintf(buffer, "tstamp_mode: %s\n", snd_pcm_tstamp_mode_name(runtime->tstamp_mode)); snd_iprintf(buffer, "period_step: %u\n", runtime->period_step); @@ -421,24 +429,29 @@ static void snd_pcm_substream_proc_sw_params_read(struct snd_info_entry *entry, snd_iprintf(buffer, "silence_threshold: %lu\n", runtime->silence_threshold); snd_iprintf(buffer, "silence_size: %lu\n", runtime->silence_size); snd_iprintf(buffer, "boundary: %lu\n", runtime->boundary); + unlock: + mutex_unlock(&substream->pcm->open_mutex); } static void snd_pcm_substream_proc_status_read(struct snd_info_entry *entry, struct snd_info_buffer *buffer) { struct snd_pcm_substream *substream = entry->private_data; - struct snd_pcm_runtime *runtime = substream->runtime; + struct snd_pcm_runtime *runtime; struct snd_pcm_status status; int err; + + mutex_lock(&substream->pcm->open_mutex); + runtime = substream->runtime; if (!runtime) { snd_iprintf(buffer, "closed\n"); - return; + goto unlock; } memset(&status, 0, sizeof(status)); err = snd_pcm_status(substream, &status); if (err < 0) { snd_iprintf(buffer, "error %d\n", err); - return; + goto unlock; } snd_iprintf(buffer, "state: %s\n", snd_pcm_state_name(status.state)); snd_iprintf(buffer, "owner_pid : %d\n", pid_vnr(substream->pid)); @@ -452,6 +465,8 @@ static void snd_pcm_substream_proc_status_read(struct snd_info_entry *entry, snd_iprintf(buffer, "-----\n"); snd_iprintf(buffer, "hw_ptr : %ld\n", runtime->status->hw_ptr); snd_iprintf(buffer, "appl_ptr : %ld\n", runtime->control->appl_ptr); + unlock: + mutex_unlock(&substream->pcm->open_mutex); } #ifdef CONFIG_SND_PCM_XRUN_DEBUG -- cgit v0.10.2 From 8702d33aa6e6d753ef99163afe48aba1323374ef Mon Sep 17 00:00:00 2001 From: Stefan Richter Date: Wed, 15 Sep 2010 13:02:44 +0200 Subject: firewire: nosy: fix build when CONFIG_FIREWIRE=N drivers/firewire/nosy* is a stand-alone driver that does not depend on CONFIG_FIREWIRE. Hence let make descend into drivers/firewire/ also if that option is off. The stand-alone driver drivers/ieee1394/init_ohci1394_dma* will soon be moved into drivers/firewire/ too and will require the same makefile fix. Side effect: As mentioned in https://bugzilla.novell.com/show_bug.cgi?id=586172#c24 this influences the order in which either firewire-ohci or ohci1394 is going to be bound to an OHCI-1394 controller in case of a modular build of both drivers if no modprobe blacklist entries are configured. However, a user of such a setup cannot expect deterministic behavior anyway. The Kconfig help and the migration guide at ieee1394.wiki.kernel.org recommend blacklist entries when a dual IEEE 1394 stack build is being used. (The coexistence period of the two stacks is planned to end soon.) Cc: Michal Marek Signed-off-by: Stefan Richter diff --git a/drivers/Makefile b/drivers/Makefile index 91874e0..0bbb456 100644 --- a/drivers/Makefile +++ b/drivers/Makefile @@ -50,7 +50,7 @@ obj-$(CONFIG_SPI) += spi/ obj-y += net/ obj-$(CONFIG_ATM) += atm/ obj-$(CONFIG_FUSION) += message/ -obj-$(CONFIG_FIREWIRE) += firewire/ +obj-y += firewire/ obj-y += ieee1394/ obj-$(CONFIG_UIO) += uio/ obj-y += cdrom/ -- cgit v0.10.2 From ae00d4f37f4df56821331deb1028748110dd6dc9 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 16 Sep 2010 16:26:51 -0700 Subject: ceph: fix cap_snap and realm split The cap_snap creation/queueing relies on both the current i_head_snapc _and_ the i_snap_realm pointers being correct, so that the new cap_snap can properly reference the old context and the new i_head_snapc can be updated to reference the new snaprealm's context. To fix this, we: - move inodes completely to the new (split) realm so that i_snap_realm is correct, and - generate the new snapc's _before_ queueing the cap_snaps in ceph_update_snap_trace(). Signed-off-by: Sage Weil diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 50461b8..efbc604 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -411,8 +411,8 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) if (i_size < page_off + len) len = i_size - page_off; - dout("writepage %p page %p index %lu on %llu~%u\n", - inode, page, page->index, page_off, len); + dout("writepage %p page %p index %lu on %llu~%u snapc %p\n", + inode, page, page->index, page_off, len, snapc); writeback_stat = atomic_long_inc_return(&client->writeback_count); if (writeback_stat > diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 9e836af..9e6eef1 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -119,6 +119,7 @@ static struct ceph_snap_realm *ceph_create_snap_realm( INIT_LIST_HEAD(&realm->children); INIT_LIST_HEAD(&realm->child_item); INIT_LIST_HEAD(&realm->empty_item); + INIT_LIST_HEAD(&realm->dirty_item); INIT_LIST_HEAD(&realm->inodes_with_caps); spin_lock_init(&realm->inodes_with_caps_lock); __insert_snap_realm(&mdsc->snap_realms, realm); @@ -604,6 +605,7 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc, struct ceph_snap_realm *realm; int invalidate = 0; int err = -ENOMEM; + LIST_HEAD(dirty_realms); dout("update_snap_trace deletion=%d\n", deletion); more: @@ -626,24 +628,6 @@ more: } } - if (le64_to_cpu(ri->seq) > realm->seq) { - dout("update_snap_trace updating %llx %p %lld -> %lld\n", - realm->ino, realm, realm->seq, le64_to_cpu(ri->seq)); - /* - * if the realm seq has changed, queue a cap_snap for every - * inode with open caps. we do this _before_ we update - * the realm info so that we prepare for writeback under the - * _previous_ snap context. - * - * ...unless it's a snap deletion! - */ - if (!deletion) - queue_realm_cap_snaps(realm); - } else { - dout("update_snap_trace %llx %p seq %lld unchanged\n", - realm->ino, realm, realm->seq); - } - /* ensure the parent is correct */ err = adjust_snap_realm_parent(mdsc, realm, le64_to_cpu(ri->parent)); if (err < 0) @@ -651,6 +635,8 @@ more: invalidate += err; if (le64_to_cpu(ri->seq) > realm->seq) { + dout("update_snap_trace updating %llx %p %lld -> %lld\n", + realm->ino, realm, realm->seq, le64_to_cpu(ri->seq)); /* update realm parameters, snap lists */ realm->seq = le64_to_cpu(ri->seq); realm->created = le64_to_cpu(ri->created); @@ -668,9 +654,17 @@ more: if (err < 0) goto fail; + /* queue realm for cap_snap creation */ + list_add(&realm->dirty_item, &dirty_realms); + invalidate = 1; } else if (!realm->cached_context) { + dout("update_snap_trace %llx %p seq %lld new\n", + realm->ino, realm, realm->seq); invalidate = 1; + } else { + dout("update_snap_trace %llx %p seq %lld unchanged\n", + realm->ino, realm, realm->seq); } dout("done with %llx %p, invalidated=%d, %p %p\n", realm->ino, @@ -683,6 +677,14 @@ more: if (invalidate) rebuild_snap_realms(realm); + /* + * queue cap snaps _after_ we've built the new snap contexts, + * so that i_head_snapc can be set appropriately. + */ + list_for_each_entry(realm, &dirty_realms, dirty_item) { + queue_realm_cap_snaps(realm); + } + __cleanup_empty_realms(mdsc); return 0; @@ -816,6 +818,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc, }; struct inode *inode = ceph_find_inode(sb, vino); struct ceph_inode_info *ci; + struct ceph_snap_realm *oldrealm; if (!inode) continue; @@ -841,18 +844,19 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc, dout(" will move %p to split realm %llx %p\n", inode, realm->ino, realm); /* - * Remove the inode from the realm's inode - * list, but don't add it to the new realm - * yet. We don't want the cap_snap to be - * queued (again) by ceph_update_snap_trace() - * below. Queue it _now_, under the old context. + * Move the inode to the new realm */ spin_lock(&realm->inodes_with_caps_lock); list_del_init(&ci->i_snap_realm_item); + list_add(&ci->i_snap_realm_item, + &realm->inodes_with_caps); + oldrealm = ci->i_snap_realm; + ci->i_snap_realm = realm; spin_unlock(&realm->inodes_with_caps_lock); spin_unlock(&inode->i_lock); - ceph_queue_cap_snap(ci); + ceph_get_snap_realm(mdsc, realm); + ceph_put_snap_realm(mdsc, oldrealm); iput(inode); continue; @@ -880,43 +884,9 @@ skip_inode: ceph_update_snap_trace(mdsc, p, e, op == CEPH_SNAP_OP_DESTROY); - if (op == CEPH_SNAP_OP_SPLIT) { - /* - * ok, _now_ add the inodes into the new realm. - */ - for (i = 0; i < num_split_inos; i++) { - struct ceph_vino vino = { - .ino = le64_to_cpu(split_inos[i]), - .snap = CEPH_NOSNAP, - }; - struct inode *inode = ceph_find_inode(sb, vino); - struct ceph_inode_info *ci; - - if (!inode) - continue; - ci = ceph_inode(inode); - spin_lock(&inode->i_lock); - if (list_empty(&ci->i_snap_realm_item)) { - struct ceph_snap_realm *oldrealm = - ci->i_snap_realm; - - dout(" moving %p to split realm %llx %p\n", - inode, realm->ino, realm); - spin_lock(&realm->inodes_with_caps_lock); - list_add(&ci->i_snap_realm_item, - &realm->inodes_with_caps); - ci->i_snap_realm = realm; - spin_unlock(&realm->inodes_with_caps_lock); - ceph_get_snap_realm(mdsc, realm); - ceph_put_snap_realm(mdsc, oldrealm); - } - spin_unlock(&inode->i_lock); - iput(inode); - } - + if (op == CEPH_SNAP_OP_SPLIT) /* we took a reference when we created the realm, above */ ceph_put_snap_realm(mdsc, realm); - } __cleanup_empty_realms(mdsc); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index c33897a..c80bfbe 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -690,6 +690,8 @@ struct ceph_snap_realm { struct list_head empty_item; /* if i have ref==0 */ + struct list_head dirty_item; /* if realm needs new context */ + /* the current set of snaps for this realm */ struct ceph_snap_context *cached_context; -- cgit v0.10.2 From 89e45aac42d40426c97e6901811309bf49c4993f Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 17 Sep 2010 03:24:13 +0200 Subject: x86: Fix instruction breakpoint encoding Lengths and types of breakpoints are encoded in a half byte into CPU registers. However when we extract these values and store them, we add a high half byte part to them: 0x40 to the length and 0x80 to the type. When that gets reloaded to the CPU registers, the high part is masked. While making the instruction breakpoints available for perf, I zapped that high part on instruction breakpoint encoding and that broke the arch -> generic translation used by ptrace instruction breakpoints. Writing dr7 to set an inst breakpoint was then failing. There is no apparent reason for these high parts so we could get rid of them altogether. That's an invasive change though so let's do that later and for now fix the problem by restoring that inst breakpoint high part encoding in this sole patch. Reported-by: Kelvie Wong Signed-off-by: Frederic Weisbecker Cc: Prasad Cc: Mahesh Salgaonkar Cc: Will Deacon diff --git a/arch/x86/include/asm/hw_breakpoint.h b/arch/x86/include/asm/hw_breakpoint.h index 528a11e..824ca07 100644 --- a/arch/x86/include/asm/hw_breakpoint.h +++ b/arch/x86/include/asm/hw_breakpoint.h @@ -20,7 +20,7 @@ struct arch_hw_breakpoint { #include /* Available HW breakpoint length encodings */ -#define X86_BREAKPOINT_LEN_X 0x00 +#define X86_BREAKPOINT_LEN_X 0x40 #define X86_BREAKPOINT_LEN_1 0x40 #define X86_BREAKPOINT_LEN_2 0x44 #define X86_BREAKPOINT_LEN_4 0x4c diff --git a/arch/x86/kernel/hw_breakpoint.c b/arch/x86/kernel/hw_breakpoint.c index a474ec3..ff15c9d 100644 --- a/arch/x86/kernel/hw_breakpoint.c +++ b/arch/x86/kernel/hw_breakpoint.c @@ -206,11 +206,27 @@ int arch_check_bp_in_kernelspace(struct perf_event *bp) int arch_bp_generic_fields(int x86_len, int x86_type, int *gen_len, int *gen_type) { - /* Len */ - switch (x86_len) { - case X86_BREAKPOINT_LEN_X: + /* Type */ + switch (x86_type) { + case X86_BREAKPOINT_EXECUTE: + if (x86_len != X86_BREAKPOINT_LEN_X) + return -EINVAL; + + *gen_type = HW_BREAKPOINT_X; *gen_len = sizeof(long); + return 0; + case X86_BREAKPOINT_WRITE: + *gen_type = HW_BREAKPOINT_W; break; + case X86_BREAKPOINT_RW: + *gen_type = HW_BREAKPOINT_W | HW_BREAKPOINT_R; + break; + default: + return -EINVAL; + } + + /* Len */ + switch (x86_len) { case X86_BREAKPOINT_LEN_1: *gen_len = HW_BREAKPOINT_LEN_1; break; @@ -229,21 +245,6 @@ int arch_bp_generic_fields(int x86_len, int x86_type, return -EINVAL; } - /* Type */ - switch (x86_type) { - case X86_BREAKPOINT_EXECUTE: - *gen_type = HW_BREAKPOINT_X; - break; - case X86_BREAKPOINT_WRITE: - *gen_type = HW_BREAKPOINT_W; - break; - case X86_BREAKPOINT_RW: - *gen_type = HW_BREAKPOINT_W | HW_BREAKPOINT_R; - break; - default: - return -EINVAL; - } - return 0; } @@ -316,9 +317,6 @@ int arch_validate_hwbkpt_settings(struct perf_event *bp) ret = -EINVAL; switch (info->len) { - case X86_BREAKPOINT_LEN_X: - align = sizeof(long) -1; - break; case X86_BREAKPOINT_LEN_1: align = 0; break; -- cgit v0.10.2 From 068e35eee9ef98eb4cab55181977e24995d273be Mon Sep 17 00:00:00 2001 From: Matt Helsley Date: Mon, 13 Sep 2010 13:01:18 -0700 Subject: hw breakpoints: Fix pid namespace bug Hardware breakpoints can't be registered within pid namespaces because tsk->pid is passed rather than the pid in the current namespace. (See https://bugzilla.kernel.org/show_bug.cgi?id=17281 ) This is a quick fix demonstrating the problem but is not the best method of solving the problem since passing pids internally is not the best way to avoid pid namespace bugs. Subsequent patches will show a better solution. Much thanks to Frederic Weisbecker for doing the bulk of the work finding this bug. Reported-by: Robin Green Signed-off-by: Matt Helsley Signed-off-by: Peter Zijlstra Cc: Prasad Cc: Arnaldo Carvalho de Melo Cc: Steven Rostedt Cc: Will Deacon Cc: Mahesh Salgaonkar Cc: 2.6.33-2.6.35 LKML-Reference: Signed-off-by: Ingo Molnar Signed-off-by: Frederic Weisbecker diff --git a/kernel/hw_breakpoint.c b/kernel/hw_breakpoint.c index d71a987..c7c2aed 100644 --- a/kernel/hw_breakpoint.c +++ b/kernel/hw_breakpoint.c @@ -433,7 +433,8 @@ register_user_hw_breakpoint(struct perf_event_attr *attr, perf_overflow_handler_t triggered, struct task_struct *tsk) { - return perf_event_create_kernel_counter(attr, -1, tsk->pid, triggered); + return perf_event_create_kernel_counter(attr, -1, task_pid_vnr(tsk), + triggered); } EXPORT_SYMBOL_GPL(register_user_hw_breakpoint); -- cgit v0.10.2 From 126925c090155f13e90b9e7e8c4010e96027c00a Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 7 Sep 2010 17:02:47 +1000 Subject: md: call md_update_sb even for 'external' metadata arrays. Now that we depend on md_update_sb to clear variable bits in mddev->flags (rather than trying not to set them) it is important to always call md_update_sb when appropriate. md_check_recovery has this job but explicitly avoids it for ->external metadata arrays. This is not longer appropraite, or needed. However we do want to avoid taking the mddev lock if only MD_CHANGE_PENDING is set as that is not cleared by md_update_sb for external-metadata arrays. Reported-by: "Kwolek, Adam" Signed-off-by: NeilBrown diff --git a/drivers/md/md.c b/drivers/md/md.c index 43cf9cc..bdd9bba 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -7069,7 +7069,7 @@ void md_check_recovery(mddev_t *mddev) if (mddev->ro && !test_bit(MD_RECOVERY_NEEDED, &mddev->recovery)) return; if ( ! ( - (mddev->flags && !mddev->external) || + (mddev->flags & ~ (1<recovery) || test_bit(MD_RECOVERY_DONE, &mddev->recovery) || (mddev->external == 0 && mddev->safemode == 1) || -- cgit v0.10.2 From ddcf3522cf03a147c867a2e0155761652dbd156a Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 8 Sep 2010 16:48:17 +1000 Subject: md: fix v1.x metadata update when a disk is missing. If an array with 1.x metadata is assembled with the last disk missing, md doesn't properly record the fact that the disk was missing. This is unlikely to cause a real problem as the event count will be different to the count on the missing disk so it won't be included in the array. However it could still cause confusion. So make sure we clear all the relevant slots, not just the early ones. Signed-off-by: NeilBrown diff --git a/drivers/md/md.c b/drivers/md/md.c index bdd9bba..f20d13e 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -1643,7 +1643,9 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev) bmask = queue_logical_block_size(rdev->bdev->bd_disk->queue)-1; if (rdev->sb_size & bmask) rdev->sb_size = (rdev->sb_size | bmask) + 1; - } + } else + max_dev = le32_to_cpu(sb->max_dev); + for (i=0; idev_roles[i] = cpu_to_le16(0xfffe); -- cgit v0.10.2 From 79077319d7c7844d5d836e52099a7a1bcadf9b04 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Sun, 12 Sep 2010 19:58:04 +0100 Subject: drm/i915/crt: Downgrade warnings for hotplug failures These are not fatal errors, so do not alarm the user by filling the logs with *** ERROR ***. Especially as we know that g4x CRT detection is a little sticky. On the one hand the errors are valid since they are warning us of a stall -- we poll the register whilst holding the mode lock so not even the mouse will update. On the other hand, those stalls were already present yet nobody complained. Reported-by: Andi Kleen Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=18332 Signed-off-by: Chris Wilson diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 4b77351..8f6f38c 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -188,7 +188,7 @@ static bool intel_ironlake_crt_detect_hotplug(struct drm_connector *connector) if (wait_for((I915_READ(PCH_ADPA) & ADPA_CRT_HOTPLUG_FORCE_TRIGGER) == 0, 1000, 1)) - DRM_ERROR("timed out waiting for FORCE_TRIGGER"); + DRM_DEBUG_KMS("timed out waiting for FORCE_TRIGGER"); if (turn_off_dac) { I915_WRITE(PCH_ADPA, temp); @@ -245,7 +245,7 @@ static bool intel_crt_detect_hotplug(struct drm_connector *connector) if (wait_for((I915_READ(PORT_HOTPLUG_EN) & CRT_HOTPLUG_FORCE_DETECT) == 0, 1000, 1)) - DRM_ERROR("timed out waiting for FORCE_DETECT to go off"); + DRM_DEBUG_KMS("timed out waiting for FORCE_DETECT to go off"); } stat = I915_READ(PORT_HOTPLUG_STAT); -- cgit v0.10.2 From e259befd9013e212648c3bd4f6f1fbf92d0dd51d Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 17 Sep 2010 00:32:02 +0100 Subject: drm/i915: Fix Sandybridge fence registers With 5 places to update when adding handling for fence registers, it is easy to overlook one or two. Correct that oversight, but fence management should be improved before a new set of registers is added. Bugzilla: https://bugs.freedesktop.org/show_bug?id=30199 Original patch by: Yuanhan Liu Signed-off-by: Chris Wilson Cc: stable@kernel.org diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 16fca1d..cf4ffbe 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2351,14 +2351,21 @@ i915_gem_object_get_fence_reg(struct drm_gem_object *obj) reg->obj = obj; - if (IS_GEN6(dev)) + switch (INTEL_INFO(dev)->gen) { + case 6: sandybridge_write_fence_reg(reg); - else if (IS_I965G(dev)) + break; + case 5: + case 4: i965_write_fence_reg(reg); - else if (IS_I9XX(dev)) + break; + case 3: i915_write_fence_reg(reg); - else + break; + case 2: i830_write_fence_reg(reg); + break; + } trace_i915_gem_object_get_fence(obj, obj_priv->fence_reg, obj_priv->tiling_mode); @@ -2381,22 +2388,26 @@ i915_gem_clear_fence_reg(struct drm_gem_object *obj) struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct drm_i915_fence_reg *reg = &dev_priv->fence_regs[obj_priv->fence_reg]; + uint32_t fence_reg; - if (IS_GEN6(dev)) { + switch (INTEL_INFO(dev)->gen) { + case 6: I915_WRITE64(FENCE_REG_SANDYBRIDGE_0 + (obj_priv->fence_reg * 8), 0); - } else if (IS_I965G(dev)) { + break; + case 5: + case 4: I915_WRITE64(FENCE_REG_965_0 + (obj_priv->fence_reg * 8), 0); - } else { - uint32_t fence_reg; - - if (obj_priv->fence_reg < 8) - fence_reg = FENCE_REG_830_0 + obj_priv->fence_reg * 4; + break; + case 3: + if (obj_priv->fence_reg > 8) + fence_reg = FENCE_REG_945_8 + (obj_priv->fence_reg - 8) * 4; else - fence_reg = FENCE_REG_945_8 + (obj_priv->fence_reg - - 8) * 4; + case 2: + fence_reg = FENCE_REG_830_0 + obj_priv->fence_reg * 4; I915_WRITE(fence_reg, 0); + break; } reg->obj = NULL; diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c index 2c6b98f..31f0858 100644 --- a/drivers/gpu/drm/i915/i915_suspend.c +++ b/drivers/gpu/drm/i915/i915_suspend.c @@ -789,16 +789,25 @@ int i915_save_state(struct drm_device *dev) dev_priv->saveSWF2[i] = I915_READ(SWF30 + (i << 2)); /* Fences */ - if (IS_I965G(dev)) { + switch (INTEL_INFO(dev)->gen) { + case 6: + for (i = 0; i < 16; i++) + dev_priv->saveFENCE[i] = I915_READ64(FENCE_REG_SANDYBRIDGE_0 + (i * 8)); + break; + case 5: + case 4: for (i = 0; i < 16; i++) dev_priv->saveFENCE[i] = I915_READ64(FENCE_REG_965_0 + (i * 8)); - } else { - for (i = 0; i < 8; i++) - dev_priv->saveFENCE[i] = I915_READ(FENCE_REG_830_0 + (i * 4)); - + break; + case 3: if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev)) for (i = 0; i < 8; i++) dev_priv->saveFENCE[i+8] = I915_READ(FENCE_REG_945_8 + (i * 4)); + case 2: + for (i = 0; i < 8; i++) + dev_priv->saveFENCE[i] = I915_READ(FENCE_REG_830_0 + (i * 4)); + break; + } return 0; @@ -815,15 +824,24 @@ int i915_restore_state(struct drm_device *dev) I915_WRITE(HWS_PGA, dev_priv->saveHWS); /* Fences */ - if (IS_I965G(dev)) { + switch (INTEL_INFO(dev)->gen) { + case 6: + for (i = 0; i < 16; i++) + I915_WRITE64(FENCE_REG_SANDYBRIDGE_0 + (i * 8), dev_priv->saveFENCE[i]); + break; + case 5: + case 4: for (i = 0; i < 16; i++) I915_WRITE64(FENCE_REG_965_0 + (i * 8), dev_priv->saveFENCE[i]); - } else { - for (i = 0; i < 8; i++) - I915_WRITE(FENCE_REG_830_0 + (i * 4), dev_priv->saveFENCE[i]); + break; + case 3: + case 2: if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev)) for (i = 0; i < 8; i++) I915_WRITE(FENCE_REG_945_8 + (i * 4), dev_priv->saveFENCE[i+8]); + for (i = 0; i < 8; i++) + I915_WRITE(FENCE_REG_830_0 + (i * 4), dev_priv->saveFENCE[i]); + break; } i915_restore_display(dev); -- cgit v0.10.2 From 41a51428916ab04587bacee2dda61c4a0c4fc02f Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 17 Sep 2010 08:22:30 +0100 Subject: drm/i915,agp/intel: Add second set of PCI-IDs for B43 There is a second revision of B43 (a desktop gen4 part) floating around, functionally equivalent to the original B43, so simply add the new PCI-IDs. Bugzilla: https://bugs.freedesktop.org/show_bugs.cgi?id=30221 Signed-off-by: Chris Wilson Cc: stable@kernel.org diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c index eab58db..cd18493 100644 --- a/drivers/char/agp/intel-agp.c +++ b/drivers/char/agp/intel-agp.c @@ -806,6 +806,8 @@ static const struct intel_driver_description { "G45/G43", NULL, &intel_i965_driver }, { PCI_DEVICE_ID_INTEL_B43_HB, PCI_DEVICE_ID_INTEL_B43_IG, "B43", NULL, &intel_i965_driver }, + { PCI_DEVICE_ID_INTEL_B43_1_HB, PCI_DEVICE_ID_INTEL_B43_1_IG, + "B43", NULL, &intel_i965_driver }, { PCI_DEVICE_ID_INTEL_G41_HB, PCI_DEVICE_ID_INTEL_G41_IG, "G41", NULL, &intel_i965_driver }, { PCI_DEVICE_ID_INTEL_IRONLAKE_D_HB, PCI_DEVICE_ID_INTEL_IRONLAKE_D_IG, diff --git a/drivers/char/agp/intel-agp.h b/drivers/char/agp/intel-agp.h index ee189c7..d09b1ab 100644 --- a/drivers/char/agp/intel-agp.h +++ b/drivers/char/agp/intel-agp.h @@ -186,6 +186,8 @@ #define PCI_DEVICE_ID_INTEL_Q33_IG 0x29D2 #define PCI_DEVICE_ID_INTEL_B43_HB 0x2E40 #define PCI_DEVICE_ID_INTEL_B43_IG 0x2E42 +#define PCI_DEVICE_ID_INTEL_B43_1_HB 0x2E90 +#define PCI_DEVICE_ID_INTEL_B43_1_IG 0x2E92 #define PCI_DEVICE_ID_INTEL_GM45_HB 0x2A40 #define PCI_DEVICE_ID_INTEL_GM45_IG 0x2A42 #define PCI_DEVICE_ID_INTEL_EAGLELAKE_HB 0x2E00 diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 216deb5..6dbe14c 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -170,6 +170,7 @@ static const struct pci_device_id pciidlist[] = { /* aka */ INTEL_VGA_DEVICE(0x2e22, &intel_g45_info), /* G45_G */ INTEL_VGA_DEVICE(0x2e32, &intel_g45_info), /* G41_G */ INTEL_VGA_DEVICE(0x2e42, &intel_g45_info), /* B43_G */ + INTEL_VGA_DEVICE(0x2e92, &intel_g45_info), /* B43_G.1 */ INTEL_VGA_DEVICE(0xa001, &intel_pineview_info), INTEL_VGA_DEVICE(0xa011, &intel_pineview_info), INTEL_VGA_DEVICE(0x0042, &intel_ironlake_d_info), -- cgit v0.10.2 From 99f76891a33d130776da3b01935d978b1e75fe68 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Mon, 13 Sep 2010 13:38:55 +0100 Subject: ARM: 6375/1: plat-nomadik: MTU timer trivial bug fix timer0 to 3 are all on mtu block 0, so don't calculate the clock event rate based upon mtu block 1's clock speed. Acked-by: Alessandro Rubini Signed-off-by: Jonas Aaberg Signed-off-by: Linus Walleij Signed-off-by: Russell King diff --git a/arch/arm/plat-nomadik/timer.c b/arch/arm/plat-nomadik/timer.c index ea3ca86..d673888 100644 --- a/arch/arm/plat-nomadik/timer.c +++ b/arch/arm/plat-nomadik/timer.c @@ -131,17 +131,12 @@ void __init nmdk_timer_init(void) { unsigned long rate; struct clk *clk0; - struct clk *clk1; u32 cr; clk0 = clk_get_sys("mtu0", NULL); BUG_ON(IS_ERR(clk0)); - clk1 = clk_get_sys("mtu1", NULL); - BUG_ON(IS_ERR(clk1)); - clk_enable(clk0); - clk_enable(clk1); /* * Tick rate is 2.4MHz for Nomadik and 110MHz for ux500: @@ -170,15 +165,8 @@ void __init nmdk_timer_init(void) pr_err("timer: failed to initialize clock source %s\n", nmdk_clksrc.name); - /* Timer 1 is used for events, fix according to rate */ - cr = MTU_CRn_32BITS; - rate = clk_get_rate(clk1); - if (rate > 16 << 20) { - rate /= 16; - cr |= MTU_CRn_PRESCALE_16; - } else { - cr |= MTU_CRn_PRESCALE_1; - } + /* Timer 1 is used for events */ + clockevents_calc_mult_shift(&nmdk_clkevt, rate, MTU_MIN_RANGE); writel(cr | MTU_CRn_ONESHOT, mtu_base + MTU_CR(1)); /* off, currently */ -- cgit v0.10.2 From a0719f52d90aed5b82ecf2f3ed14bc4ced3d1a8a Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Mon, 13 Sep 2010 13:40:04 +0100 Subject: ARM: 6376/1: plat-nomadik: MTU: Change prescaler limit and comment updates The prescaler 16 is now used only when the timer runs at 32 MHz or more. Some comment updates as well. Acked-by: Alessandro Rubini Signed-off-by: Jonas Aaberg Signed-off-by: Linus Walleij Signed-off-by: Russell King diff --git a/arch/arm/plat-nomadik/timer.c b/arch/arm/plat-nomadik/timer.c index d673888..aedf9c1 100644 --- a/arch/arm/plat-nomadik/timer.c +++ b/arch/arm/plat-nomadik/timer.c @@ -1,5 +1,5 @@ /* - * linux/arch/arm/mach-nomadik/timer.c + * linux/arch/arm/plat-nomadik/timer.c * * Copyright (C) 2008 STMicroelectronics * Copyright (C) 2010 Alessandro Rubini @@ -75,7 +75,7 @@ static void nmdk_clkevt_mode(enum clock_event_mode mode, cr = readl(mtu_base + MTU_CR(1)); writel(0, mtu_base + MTU_LR(1)); writel(cr | MTU_CRn_ENA, mtu_base + MTU_CR(1)); - writel(0x2, mtu_base + MTU_IMSC); + writel(1 << 1, mtu_base + MTU_IMSC); break; case CLOCK_EVT_MODE_SHUTDOWN: case CLOCK_EVT_MODE_UNUSED: @@ -131,7 +131,7 @@ void __init nmdk_timer_init(void) { unsigned long rate; struct clk *clk0; - u32 cr; + u32 cr = MTU_CRn_32BITS; clk0 = clk_get_sys("mtu0", NULL); BUG_ON(IS_ERR(clk0)); @@ -139,12 +139,15 @@ void __init nmdk_timer_init(void) clk_enable(clk0); /* - * Tick rate is 2.4MHz for Nomadik and 110MHz for ux500: - * use a divide-by-16 counter if it's more than 16MHz + * Tick rate is 2.4MHz for Nomadik and 2.4Mhz, 100MHz or 133 MHz + * for ux500. + * Use a divide-by-16 counter if the tick rate is more than 32MHz. + * At 32 MHz, the timer (with 32 bit counter) can be programmed + * to wake-up at a max 127s a head in time. Dividing a 2.4 MHz timer + * with 16 gives too low timer resolution. */ - cr = MTU_CRn_32BITS;; rate = clk_get_rate(clk0); - if (rate > 16 << 20) { + if (rate > 32000000) { rate /= 16; cr |= MTU_CRn_PRESCALE_16; } else { -- cgit v0.10.2 From 63f469324f999a28e67b90eb27a5fe0e379b7064 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Mon, 13 Sep 2010 13:44:47 +0100 Subject: ARM: 6377/1: supply _cansleep gpio function to U300 We have to use _cansleep gpio accessors in the MMCI driver so as to avoid slowpath warnings, now U300 has MMCI but doesn't have these functions in place to siply wrap the existing non-sleeping functions into sleepable variants. Signed-off-by: Linus Walleij Signed-off-by: Russell King diff --git a/arch/arm/mach-u300/include/mach/gpio.h b/arch/arm/mach-u300/include/mach/gpio.h index 7b1fc98..d5a71ab 100644 --- a/arch/arm/mach-u300/include/mach/gpio.h +++ b/arch/arm/mach-u300/include/mach/gpio.h @@ -273,6 +273,9 @@ extern void gpio_pullup(unsigned gpio, int value); extern int gpio_get_value(unsigned gpio); extern void gpio_set_value(unsigned gpio, int value); +#define gpio_get_value_cansleep gpio_get_value +#define gpio_set_value_cansleep gpio_set_value + /* wrappers to sleep-enable the previous two functions */ static inline unsigned gpio_to_irq(unsigned gpio) { -- cgit v0.10.2 From 6491848d1ab246f6d243ddef25085fc1d836ff2c Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Tue, 14 Sep 2010 09:50:03 +0100 Subject: ARM: 6387/1: errata: check primary part ID in proc-v7.S Kconfig doesn't have any knowledge of specific v7 cores, so it is possible to select errata workarounds that may cause inadvertent behaviour when executed on a core other than those targetted by the fix. This patch improves the variant and revision checking in proc-v7.S so that the primary part number is also considered when applying errata workarounds. Acked-by: Catalin Marinas Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 6a8506d..1f16f9e 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -204,8 +204,13 @@ __v7_setup: bne 2f and r5, r0, #0x00f00000 @ variant and r6, r0, #0x0000000f @ revision - orr r0, r6, r5, lsr #20-4 @ combine variant and revision + orr r6, r6, r5, lsr #20-4 @ combine variant and revision + ubfx r0, r0, #4, #12 @ primary part number + /* Cortex-A8 Errata */ + ldr r10, =0x00000c08 @ Cortex-A8 primary part number + teq r0, r10 + bne 2f #ifdef CONFIG_ARM_ERRATA_430973 teq r5, #0x00100000 @ only present in r1p* mrceq p15, 0, r10, c1, c0, 1 @ read aux control register @@ -213,14 +218,14 @@ __v7_setup: mcreq p15, 0, r10, c1, c0, 1 @ write aux control register #endif #ifdef CONFIG_ARM_ERRATA_458693 - teq r0, #0x20 @ only present in r2p0 + teq r6, #0x20 @ only present in r2p0 mrceq p15, 0, r10, c1, c0, 1 @ read aux control register orreq r10, r10, #(1 << 5) @ set L1NEON to 1 orreq r10, r10, #(1 << 9) @ set PLDNOP to 1 mcreq p15, 0, r10, c1, c0, 1 @ write aux control register #endif #ifdef CONFIG_ARM_ERRATA_460075 - teq r0, #0x20 @ only present in r2p0 + teq r6, #0x20 @ only present in r2p0 mrceq p15, 1, r10, c9, c0, 2 @ read L2 cache aux ctrl register tsteq r10, #1 << 22 orreq r10, r10, #(1 << 22) @ set the Write Allocate disable bit -- cgit v0.10.2 From 9f05027c7cb3cfe56a31892bd83391138d41a667 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Tue, 14 Sep 2010 09:51:43 +0100 Subject: ARM: 6388/1: errata: DMB operation may be faulty On versions of the Cortex-A9 up to and including r2p2, under rare circumstances, a DMB instruction between 2 write operations may not ensure the correct visibility ordering of the 2 writes. This workaround sets a bit in the diagnostic register of the Cortex-A9, causing the DMB instruction to behave like a DSB, which functions correctly on the affected cores. Acked-by: Catalin Marinas Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 553b7cf..ab14c0a 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1051,6 +1051,18 @@ config ARM_ERRATA_460075 ACTLR register. Note that setting specific bits in the ACTLR register may not be available in non-secure mode. +config ARM_ERRATA_742230 + bool "ARM errata: DMB operation may be faulty" + depends on CPU_V7 && SMP + help + This option enables the workaround for the 742230 Cortex-A9 + (r1p0..r2p2) erratum. Under rare circumstances, a DMB instruction + between two write operations may not ensure the correct visibility + ordering of the two writes. This workaround sets a specific bit in + the diagnostic register of the Cortex-A9 which causes the DMB + instruction to behave as a DSB, ensuring the correct behaviour of + the two writes. + config PL310_ERRATA_588369 bool "Clean & Invalidate maintenance operations do not invalidate clean lines" depends on CACHE_L2X0 && ARCH_OMAP4 diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 1f16f9e..945f363 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -201,7 +201,7 @@ __v7_setup: mrc p15, 0, r0, c0, c0, 0 @ read main ID register and r10, r0, #0xff000000 @ ARM? teq r10, #0x41000000 - bne 2f + bne 3f and r5, r0, #0x00f00000 @ variant and r6, r0, #0x0000000f @ revision orr r6, r6, r5, lsr #20-4 @ combine variant and revision @@ -231,8 +231,20 @@ __v7_setup: orreq r10, r10, #(1 << 22) @ set the Write Allocate disable bit mcreq p15, 1, r10, c9, c0, 2 @ write the L2 cache aux ctrl register #endif + b 3f + + /* Cortex-A9 Errata */ +2: ldr r10, =0x00000c09 @ Cortex-A9 primary part number + teq r0, r10 + bne 3f +#ifdef CONFIG_ARM_ERRATA_742230 + cmp r6, #0x22 @ only present up to r2p2 + mrcle p15, 0, r10, c15, c0, 1 @ read diagnostic register + orrle r10, r10, #1 << 4 @ set bit #4 + mcrle p15, 0, r10, c15, c0, 1 @ write diagnostic register +#endif -2: mov r10, #0 +3: mov r10, #0 #ifdef HARVARD_CACHE mcr p15, 0, r10, c7, c5, 0 @ I+BTB cache invalidate #endif -- cgit v0.10.2 From a672e99b129e286df2e2697a1b603d82321117f3 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Tue, 14 Sep 2010 09:53:02 +0100 Subject: ARM: 6389/1: errata: incorrect hazard handling in the SCU may lead to data corruption On the r2p0, r2p1 and r2p2 versions of the Cortex-A9, data corruption can occur if a shared cache line is replaced on one CPU as another CPU is accessing it. This workaround sets two bits in the diagnostic register of the Cortex-A9, reducing the linefill issuing capabilities of the processor and avoiding the erroneous behaviour. Acked-by: Catalin Marinas Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index ab14c0a..0f89335 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1063,6 +1063,20 @@ config ARM_ERRATA_742230 instruction to behave as a DSB, ensuring the correct behaviour of the two writes. +config ARM_ERRATA_742231 + bool "ARM errata: Incorrect hazard handling in the SCU may lead to data corruption" + depends on CPU_V7 && SMP + help + This option enables the workaround for the 742231 Cortex-A9 + (r2p0..r2p2) erratum. Under certain conditions, specific to the + Cortex-A9 MPCore micro-architecture, two CPUs working in SMP mode, + accessing some data located in the same cache line, may get corrupted + data due to bad handling of the address hazard when the line gets + replaced from one of the CPUs at the same time as another CPU is + accessing it. This workaround sets specific bits in the diagnostic + register of the Cortex-A9 which reduces the linefill issuing + capabilities of the processor. + config PL310_ERRATA_588369 bool "Clean & Invalidate maintenance operations do not invalidate clean lines" depends on CACHE_L2X0 && ARCH_OMAP4 diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 945f363..0801292 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -243,6 +243,15 @@ __v7_setup: orrle r10, r10, #1 << 4 @ set bit #4 mcrle p15, 0, r10, c15, c0, 1 @ write diagnostic register #endif +#ifdef CONFIG_ARM_ERRATA_742231 + teq r6, #0x20 @ present in r2p0 + teqne r6, #0x21 @ present in r2p1 + teqne r6, #0x22 @ present in r2p2 + mrceq p15, 0, r10, c15, c0, 1 @ read diagnostic register + orreq r10, r10, #1 << 12 @ set bit #12 + orreq r10, r10, #1 << 22 @ set bit #22 + mcreq p15, 0, r10, c15, c0, 1 @ write diagnostic register +#endif 3: mov r10, #0 #ifdef HARVARD_CACHE -- cgit v0.10.2 From 1a8e41cd672f894bbd74874eac601e6cedf838fb Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Thu, 16 Sep 2010 17:57:17 +0100 Subject: ARM: 6395/1: VExpress: Set bit 22 in the PL310 (cache controller) AuxCtlr register Clearing bit 22 in the PL310 Auxiliary Control register (shared attribute override enable) has the side effect of transforming Normal Shared Non-cacheable reads into Cacheable no-allocate reads. Coherent DMA buffers in Linux always have a Cacheable alias via the kernel linear mapping and the processor can speculatively load cache lines into the PL310 controller. With bit 22 cleared, Non-cacheable reads would unexpectedly hit such cache lines leading to buffer corruption. Cc: Nicolas Pitre Cc: Signed-off-by: Catalin Marinas Signed-off-by: Russell King diff --git a/arch/arm/mach-vexpress/ct-ca9x4.c b/arch/arm/mach-vexpress/ct-ca9x4.c index 577df6c..1c9c13e 100644 --- a/arch/arm/mach-vexpress/ct-ca9x4.c +++ b/arch/arm/mach-vexpress/ct-ca9x4.c @@ -227,7 +227,7 @@ static void ct_ca9x4_init(void) int i; #ifdef CONFIG_CACHE_L2X0 - l2x0_init(MMIO_P2V(CT_CA9X4_L2CC), 0x00000000, 0xfe0fffff); + l2x0_init(MMIO_P2V(CT_CA9X4_L2CC), 0x00400000, 0xfe0fffff); #endif clkdev_add_table(lookups, ARRAY_SIZE(lookups)); -- cgit v0.10.2 From 5f4874903df3562b9d5649fc1cf7b8c6bb238e42 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Thu, 9 Sep 2010 14:45:00 +0100 Subject: GFS2: gfs2_logd should be using interruptible waits Looks like this crept in, in a recent update. Reported-by: Krzysztof Urbaniak Signed-off-by: Steven Whitehouse diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index cde1248..ac750bd 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -932,7 +932,7 @@ int gfs2_logd(void *data) do { prepare_to_wait(&sdp->sd_logd_waitq, &wait, - TASK_UNINTERRUPTIBLE); + TASK_INTERRUPTIBLE); if (!gfs2_ail_flush_reqd(sdp) && !gfs2_jrnl_flush_reqd(sdp) && !kthread_should_stop()) -- cgit v0.10.2 From b2b163bb82b12bae2504a5b31399c37d099ad3cc Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 17 Sep 2010 14:56:16 +0100 Subject: ARM: prevent multiple syscall restarts Al Viro reports that calling "sys_sigsuspend(-ERESTARTNOHAND, 0, 0)" with two signals coming and being handled in kernel space results in the syscall restart being done twice. Avoid this by clearing the 'why' flag when we call the signal handling code to prevent further syscall restarts after the first. Acked-by: Al Viro Signed-off-by: Russell King diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index f05a35a..4a560d3 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -48,6 +48,8 @@ work_pending: beq no_work_pending mov r0, sp @ 'regs' mov r2, why @ 'syscall' + tst r1, #_TIF_SIGPENDING @ delivering a signal? + movne why, #0 @ prevent further restarts bl do_notify_resume b ret_slow_syscall @ Check work again -- cgit v0.10.2 From e835124c2be289515b918f2688ced4249e2de566 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 17 Sep 2010 08:03:08 -0700 Subject: ceph: only send one flushsnap per cap_snap per mds session Sending multiple flushsnap messages is problematic because we ignore the response if the tid doesn't match, and the server may only respond to each one once. It's also a waste. So, skip cap_snaps that are already on the flushing list, unless the caller tells us to resend (because we are reconnecting). Signed-off-by: Sage Weil diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 9fbe901..b01c316 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1195,10 +1195,14 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, * asynchronously back to the MDS once sync writes complete and dirty * data is written out. * + * Unless @again is true, skip cap_snaps that were already sent to + * the MDS (i.e., during this session). + * * Called under i_lock. Takes s_mutex as needed. */ void __ceph_flush_snaps(struct ceph_inode_info *ci, - struct ceph_mds_session **psession) + struct ceph_mds_session **psession, + int again) __releases(ci->vfs_inode->i_lock) __acquires(ci->vfs_inode->i_lock) { @@ -1240,6 +1244,13 @@ retry: dout("no auth cap (migrating?), doing nothing\n"); goto out; } + + /* only flush each capsnap once */ + if (!again && !list_empty(&capsnap->flushing_item)) { + dout("already flushed %p, skipping\n", capsnap); + continue; + } + mds = ci->i_auth_cap->session->s_mds; mseq = ci->i_auth_cap->mseq; @@ -1314,7 +1325,7 @@ static void ceph_flush_snaps(struct ceph_inode_info *ci) struct inode *inode = &ci->vfs_inode; spin_lock(&inode->i_lock); - __ceph_flush_snaps(ci, NULL); + __ceph_flush_snaps(ci, NULL, 0); spin_unlock(&inode->i_lock); } @@ -1477,7 +1488,7 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags, /* flush snaps first time around only */ if (!list_empty(&ci->i_cap_snaps)) - __ceph_flush_snaps(ci, &session); + __ceph_flush_snaps(ci, &session, 0); goto retry_locked; retry: spin_lock(&inode->i_lock); @@ -1894,7 +1905,7 @@ static void kick_flushing_capsnaps(struct ceph_mds_client *mdsc, if (cap && cap->session == session) { dout("kick_flushing_caps %p cap %p capsnap %p\n", inode, cap, capsnap); - __ceph_flush_snaps(ci, &session); + __ceph_flush_snaps(ci, &session, 1); } else { pr_err("%p auth cap %p not mds%d ???\n", inode, cap, session->s_mds); diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 9e6eef1..190b6c4 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -717,7 +717,7 @@ static void flush_snaps(struct ceph_mds_client *mdsc) igrab(inode); spin_unlock(&mdsc->snap_flush_lock); spin_lock(&inode->i_lock); - __ceph_flush_snaps(ci, &session); + __ceph_flush_snaps(ci, &session, 0); spin_unlock(&inode->i_lock); iput(inode); spin_lock(&mdsc->snap_flush_lock); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index c80bfbe..b87638e 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -828,7 +828,8 @@ extern void ceph_put_cap_refs(struct ceph_inode_info *ci, int had); extern void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, struct ceph_snap_context *snapc); extern void __ceph_flush_snaps(struct ceph_inode_info *ci, - struct ceph_mds_session **psession); + struct ceph_mds_session **psession, + int again); extern void ceph_check_caps(struct ceph_inode_info *ci, int flags, struct ceph_mds_session *session); extern void ceph_check_delayed_caps(struct ceph_mds_client *mdsc); -- cgit v0.10.2 From 5facb097137e7509a1b73448fe20226a4fbfe8cb Mon Sep 17 00:00:00 2001 From: Kuninori Morimoto Date: Fri, 17 Sep 2010 17:24:10 +0200 Subject: hwmon: (lis3lv02d) Prevent NULL pointer dereference If CONFIG_PM was selected and lis3lv02d_platform_data was NULL, the kernel will be panic when halt command run. Reported-by: Yusuke Goda Signed-off-by: Kuninori Morimoto Acked-by: Samu Onkalo Sigend-off-by: Jean Delvare diff --git a/drivers/hwmon/lis3lv02d_i2c.c b/drivers/hwmon/lis3lv02d_i2c.c index dc1f540..8e5933b 100644 --- a/drivers/hwmon/lis3lv02d_i2c.c +++ b/drivers/hwmon/lis3lv02d_i2c.c @@ -121,7 +121,7 @@ static int lis3lv02d_i2c_suspend(struct i2c_client *client, pm_message_t mesg) { struct lis3lv02d *lis3 = i2c_get_clientdata(client); - if (!lis3->pdata->wakeup_flags) + if (!lis3->pdata || !lis3->pdata->wakeup_flags) lis3lv02d_poweroff(lis3); return 0; } @@ -130,7 +130,7 @@ static int lis3lv02d_i2c_resume(struct i2c_client *client) { struct lis3lv02d *lis3 = i2c_get_clientdata(client); - if (!lis3->pdata->wakeup_flags) + if (!lis3->pdata || !lis3->pdata->wakeup_flags) lis3lv02d_poweron(lis3); return 0; } diff --git a/drivers/hwmon/lis3lv02d_spi.c b/drivers/hwmon/lis3lv02d_spi.c index 82b1680..b9be5e3 100644 --- a/drivers/hwmon/lis3lv02d_spi.c +++ b/drivers/hwmon/lis3lv02d_spi.c @@ -92,7 +92,7 @@ static int lis3lv02d_spi_suspend(struct spi_device *spi, pm_message_t mesg) { struct lis3lv02d *lis3 = spi_get_drvdata(spi); - if (!lis3->pdata->wakeup_flags) + if (!lis3->pdata || !lis3->pdata->wakeup_flags) lis3lv02d_poweroff(&lis3_dev); return 0; @@ -102,7 +102,7 @@ static int lis3lv02d_spi_resume(struct spi_device *spi) { struct lis3lv02d *lis3 = spi_get_drvdata(spi); - if (!lis3->pdata->wakeup_flags) + if (!lis3->pdata || !lis3->pdata->wakeup_flags) lis3lv02d_poweron(lis3); return 0; -- cgit v0.10.2 From 9e012c1acc1fd617a708c00ff1fdaa847f9faf67 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 17 Sep 2010 17:24:11 +0200 Subject: hwmon: New subsystem maintainers Guenter Roeck volunteered to adopt the hwmon subsystem as long as he wasn't the only maintainer. As this was also my own condition, we can add the two of us as co-maintainers of the hwmon subsystem. Signed-off-by: Jean Delvare Acked-by: Guenter Roeck diff --git a/MAINTAINERS b/MAINTAINERS index e7c528f..8540dea 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2657,9 +2657,12 @@ S: Maintained F: drivers/media/video/gspca/ HARDWARE MONITORING +M: Jean Delvare +M: Guenter Roeck L: lm-sensors@lm-sensors.org W: http://www.lm-sensors.org/ -S: Orphan +T: quilt kernel.org/pub/linux/kernel/people/jdelvare/linux-2.6/jdelvare-hwmon/ +S: Maintained F: Documentation/hwmon/ F: drivers/hwmon/ F: include/linux/hwmon*.h -- cgit v0.10.2 From 96f3640894012be7dd15a384566bfdc18297bc6c Mon Sep 17 00:00:00 2001 From: Guillem Jover Date: Fri, 17 Sep 2010 17:24:11 +0200 Subject: hwmon: (f75375s) Shift control mode to the correct bit position The spec notes that fan0 and fan1 control mode bits are located in bits 7-6 and 5-4 respectively, but the FAN_CTRL_MODE macro was making the bits shift by 5 instead of by 4. Signed-off-by: Guillem Jover Cc: Riku Voipio Cc: stable@kernel.org Signed-off-by: Jean Delvare diff --git a/drivers/hwmon/f75375s.c b/drivers/hwmon/f75375s.c index 0f58ecc..e5828c0 100644 --- a/drivers/hwmon/f75375s.c +++ b/drivers/hwmon/f75375s.c @@ -79,7 +79,7 @@ enum chips { f75373, f75375 }; #define F75375_REG_PWM2_DROP_DUTY 0x6C #define FAN_CTRL_LINEAR(nr) (4 + nr) -#define FAN_CTRL_MODE(nr) (5 + ((nr) * 2)) +#define FAN_CTRL_MODE(nr) (4 + ((nr) * 2)) /* * Data structures and manipulation thereof -- cgit v0.10.2 From c3b327d60bbba3f5ff8fd87d1efc0e95eb6c121b Mon Sep 17 00:00:00 2001 From: Guillem Jover Date: Fri, 17 Sep 2010 17:24:12 +0200 Subject: hwmon: (f75375s) Do not overwrite values read from registers All bits in the values read from registers to be used for the next write were getting overwritten, avoid doing so to not mess with the current configuration. Signed-off-by: Guillem Jover Cc: Riku Voipio Cc: stable@kernel.org Signed-off-by: Jean Delvare diff --git a/drivers/hwmon/f75375s.c b/drivers/hwmon/f75375s.c index e5828c0..9638d58 100644 --- a/drivers/hwmon/f75375s.c +++ b/drivers/hwmon/f75375s.c @@ -298,7 +298,7 @@ static int set_pwm_enable_direct(struct i2c_client *client, int nr, int val) return -EINVAL; fanmode = f75375_read8(client, F75375_REG_FAN_TIMER); - fanmode = ~(3 << FAN_CTRL_MODE(nr)); + fanmode &= ~(3 << FAN_CTRL_MODE(nr)); switch (val) { case 0: /* Full speed */ @@ -350,7 +350,7 @@ static ssize_t set_pwm_mode(struct device *dev, struct device_attribute *attr, mutex_lock(&data->update_lock); conf = f75375_read8(client, F75375_REG_CONFIG1); - conf = ~(1 << FAN_CTRL_LINEAR(nr)); + conf &= ~(1 << FAN_CTRL_LINEAR(nr)); if (val == 0) conf |= (1 << FAN_CTRL_LINEAR(nr)) ; -- cgit v0.10.2 From f17c811d1433aa1966f9c5a744841427e9a97ecf Mon Sep 17 00:00:00 2001 From: Yong Wang Date: Fri, 17 Sep 2010 17:24:12 +0200 Subject: hwmon: (emc1403) Remove unnecessary hwmon_device_unregister It is unnecessary and wrong to call hwmon_device_unregister in error handling before hwmon_device_register is called. Signed-off-by: Yong Wang Reviewed-by: Guenter Roeck Cc: stable@kernel.org Signed-off-by: Jean Delvare diff --git a/drivers/hwmon/emc1403.c b/drivers/hwmon/emc1403.c index 5b58b20..8dee3f3 100644 --- a/drivers/hwmon/emc1403.c +++ b/drivers/hwmon/emc1403.c @@ -308,7 +308,6 @@ static int emc1403_probe(struct i2c_client *client, res = sysfs_create_group(&client->dev.kobj, &m_thermal_gr); if (res) { dev_warn(&client->dev, "create group failed\n"); - hwmon_device_unregister(data->hwmon_dev); goto thermal_error1; } data->hwmon_dev = hwmon_device_register(&client->dev); -- cgit v0.10.2 From 022b75a3df2b5aeeb70c5d51bc1fe55722fdd759 Mon Sep 17 00:00:00 2001 From: Jonas Jonsson Date: Fri, 17 Sep 2010 17:24:13 +0200 Subject: hwmon: (w83627ehf) Use proper exit sequence According to the datasheet for Winbond W83627DHG the proper way to exit the Extended Function Mode is to write 0xaa to the EFER(0x2e or 0x4e). Signed-off-by: Jonas Jonsson Signed-off-by: Jean Delvare diff --git a/drivers/hwmon/w83627ehf.c b/drivers/hwmon/w83627ehf.c index e96e69d..072c580 100644 --- a/drivers/hwmon/w83627ehf.c +++ b/drivers/hwmon/w83627ehf.c @@ -127,6 +127,7 @@ superio_enter(int ioreg) static inline void superio_exit(int ioreg) { + outb(0xaa, ioreg); outb(0x02, ioreg); outb(0x02, ioreg + 1); } -- cgit v0.10.2 From a51b9944a1aaca34c9061d3973663fee54e9d1c1 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Fri, 17 Sep 2010 17:24:14 +0200 Subject: hwmon: (adm1031) Replace update_rate sysfs attribute with update_interval The attribute reflects an interval, not a rate. Signed-off-by: Guenter Roeck Acked-by: Ira W. Snyder Signed-off-by: Jean Delvare diff --git a/Documentation/hwmon/sysfs-interface b/Documentation/hwmon/sysfs-interface index ff45d1f..48ceabe 100644 --- a/Documentation/hwmon/sysfs-interface +++ b/Documentation/hwmon/sysfs-interface @@ -91,12 +91,11 @@ name The chip name. I2C devices get this attribute created automatically. RO -update_rate The rate at which the chip will update readings. +update_interval The interval at which the chip will update readings. Unit: millisecond RW - Some devices have a variable update rate. This attribute - can be used to change the update rate to the desired - frequency. + Some devices have a variable update rate or interval. + This attribute can be used to change it to the desired value. ************ diff --git a/drivers/hwmon/adm1031.c b/drivers/hwmon/adm1031.c index 15c1a96..0683e6b 100644 --- a/drivers/hwmon/adm1031.c +++ b/drivers/hwmon/adm1031.c @@ -79,7 +79,7 @@ struct adm1031_data { int chip_type; char valid; /* !=0 if following fields are valid */ unsigned long last_updated; /* In jiffies */ - unsigned int update_rate; /* In milliseconds */ + unsigned int update_interval; /* In milliseconds */ /* The chan_select_table contains the possible configurations for * auto fan control. */ @@ -743,23 +743,23 @@ static SENSOR_DEVICE_ATTR(temp3_crit_alarm, S_IRUGO, show_alarm, NULL, 12); static SENSOR_DEVICE_ATTR(temp3_fault, S_IRUGO, show_alarm, NULL, 13); static SENSOR_DEVICE_ATTR(temp1_crit_alarm, S_IRUGO, show_alarm, NULL, 14); -/* Update Rate */ -static const unsigned int update_rates[] = { +/* Update Interval */ +static const unsigned int update_intervals[] = { 16000, 8000, 4000, 2000, 1000, 500, 250, 125, }; -static ssize_t show_update_rate(struct device *dev, - struct device_attribute *attr, char *buf) +static ssize_t show_update_interval(struct device *dev, + struct device_attribute *attr, char *buf) { struct i2c_client *client = to_i2c_client(dev); struct adm1031_data *data = i2c_get_clientdata(client); - return sprintf(buf, "%u\n", data->update_rate); + return sprintf(buf, "%u\n", data->update_interval); } -static ssize_t set_update_rate(struct device *dev, - struct device_attribute *attr, - const char *buf, size_t count) +static ssize_t set_update_interval(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) { struct i2c_client *client = to_i2c_client(dev); struct adm1031_data *data = i2c_get_clientdata(client); @@ -771,12 +771,15 @@ static ssize_t set_update_rate(struct device *dev, if (err) return err; - /* find the nearest update rate from the table */ - for (i = 0; i < ARRAY_SIZE(update_rates) - 1; i++) { - if (val >= update_rates[i]) + /* + * Find the nearest update interval from the table. + * Use it to determine the matching update rate. + */ + for (i = 0; i < ARRAY_SIZE(update_intervals) - 1; i++) { + if (val >= update_intervals[i]) break; } - /* if not found, we point to the last entry (lowest update rate) */ + /* if not found, we point to the last entry (lowest update interval) */ /* set the new update rate while preserving other settings */ reg = adm1031_read_value(client, ADM1031_REG_FAN_FILTER); @@ -785,14 +788,14 @@ static ssize_t set_update_rate(struct device *dev, adm1031_write_value(client, ADM1031_REG_FAN_FILTER, reg); mutex_lock(&data->update_lock); - data->update_rate = update_rates[i]; + data->update_interval = update_intervals[i]; mutex_unlock(&data->update_lock); return count; } -static DEVICE_ATTR(update_rate, S_IRUGO | S_IWUSR, show_update_rate, - set_update_rate); +static DEVICE_ATTR(update_interval, S_IRUGO | S_IWUSR, show_update_interval, + set_update_interval); static struct attribute *adm1031_attributes[] = { &sensor_dev_attr_fan1_input.dev_attr.attr, @@ -830,7 +833,7 @@ static struct attribute *adm1031_attributes[] = { &sensor_dev_attr_auto_fan1_min_pwm.dev_attr.attr, - &dev_attr_update_rate.attr, + &dev_attr_update_interval.attr, &dev_attr_alarms.attr, NULL @@ -981,7 +984,8 @@ static void adm1031_init_client(struct i2c_client *client) mask = ADM1031_UPDATE_RATE_MASK; read_val = adm1031_read_value(client, ADM1031_REG_FAN_FILTER); i = (read_val & mask) >> ADM1031_UPDATE_RATE_SHIFT; - data->update_rate = update_rates[i]; + /* Save it as update interval */ + data->update_interval = update_intervals[i]; } static struct adm1031_data *adm1031_update_device(struct device *dev) @@ -993,7 +997,8 @@ static struct adm1031_data *adm1031_update_device(struct device *dev) mutex_lock(&data->update_lock); - next_update = data->last_updated + msecs_to_jiffies(data->update_rate); + next_update = data->last_updated + + msecs_to_jiffies(data->update_interval); if (time_after(jiffies, next_update) || !data->valid) { dev_dbg(&client->dev, "Starting adm1031 update\n"); -- cgit v0.10.2 From bc482bf0ce918b39a1fa60b9341f1add9318d833 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Fri, 17 Sep 2010 17:24:15 +0200 Subject: hwmon: (lm95241) Replace rate sysfs attribute with update_interval update_interval is the matching attribute defined in the hwmon sysfs ABI. Use it. Signed-off-by: Guenter Roeck Signed-off-by: Jean Delvare diff --git a/drivers/hwmon/lm95241.c b/drivers/hwmon/lm95241.c index 94741d4..464340f 100644 --- a/drivers/hwmon/lm95241.c +++ b/drivers/hwmon/lm95241.c @@ -91,7 +91,7 @@ static struct lm95241_data *lm95241_update_device(struct device *dev); struct lm95241_data { struct device *hwmon_dev; struct mutex update_lock; - unsigned long last_updated, rate; /* in jiffies */ + unsigned long last_updated, interval; /* in jiffies */ char valid; /* zero until following fields are valid */ /* registers values */ u8 local_h, local_l; /* local */ @@ -114,23 +114,23 @@ show_temp(local); show_temp(remote1); show_temp(remote2); -static ssize_t show_rate(struct device *dev, struct device_attribute *attr, +static ssize_t show_interval(struct device *dev, struct device_attribute *attr, char *buf) { struct lm95241_data *data = lm95241_update_device(dev); - snprintf(buf, PAGE_SIZE - 1, "%lu\n", 1000 * data->rate / HZ); + snprintf(buf, PAGE_SIZE - 1, "%lu\n", 1000 * data->interval / HZ); return strlen(buf); } -static ssize_t set_rate(struct device *dev, struct device_attribute *attr, +static ssize_t set_interval(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { struct i2c_client *client = to_i2c_client(dev); struct lm95241_data *data = i2c_get_clientdata(client); - strict_strtol(buf, 10, &data->rate); - data->rate = data->rate * HZ / 1000; + strict_strtol(buf, 10, &data->interval); + data->interval = data->interval * HZ / 1000; return count; } @@ -286,7 +286,8 @@ static DEVICE_ATTR(temp2_min, S_IWUSR | S_IRUGO, show_min1, set_min1); static DEVICE_ATTR(temp3_min, S_IWUSR | S_IRUGO, show_min2, set_min2); static DEVICE_ATTR(temp2_max, S_IWUSR | S_IRUGO, show_max1, set_max1); static DEVICE_ATTR(temp3_max, S_IWUSR | S_IRUGO, show_max2, set_max2); -static DEVICE_ATTR(rate, S_IWUSR | S_IRUGO, show_rate, set_rate); +static DEVICE_ATTR(update_interval, S_IWUSR | S_IRUGO, show_interval, + set_interval); static struct attribute *lm95241_attributes[] = { &dev_attr_temp1_input.attr, @@ -298,7 +299,7 @@ static struct attribute *lm95241_attributes[] = { &dev_attr_temp3_min.attr, &dev_attr_temp2_max.attr, &dev_attr_temp3_max.attr, - &dev_attr_rate.attr, + &dev_attr_update_interval.attr, NULL }; @@ -376,7 +377,7 @@ static void lm95241_init_client(struct i2c_client *client) { struct lm95241_data *data = i2c_get_clientdata(client); - data->rate = HZ; /* 1 sec default */ + data->interval = HZ; /* 1 sec default */ data->valid = 0; data->config = CFG_CR0076; data->model = 0; @@ -410,7 +411,7 @@ static struct lm95241_data *lm95241_update_device(struct device *dev) mutex_lock(&data->update_lock); - if (time_after(jiffies, data->last_updated + data->rate) || + if (time_after(jiffies, data->last_updated + data->interval) || !data->valid) { dev_dbg(&client->dev, "Updating lm95241 data.\n"); data->local_h = -- cgit v0.10.2 From 14eff1812679c76564b775aa95cdd378965f6cfb Mon Sep 17 00:00:00 2001 From: Daniel Walker Date: Fri, 17 Sep 2010 16:42:10 +0100 Subject: ARM: 6398/1: add proc info for ARM11MPCore/Cortex-A9 from ARM Setting of these bits can cause issues on other SMP SoC's not produced by ARM. Acked-by: Catalin Marinas Signed-off-by: Daniel Walker Signed-off-by: Russell King diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 0801292..7563ff0 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -186,13 +186,14 @@ cpu_v7_name: * It is assumed that: * - cache type register is implemented */ -__v7_setup: +__v7_ca9mp_setup: #ifdef CONFIG_SMP mrc p15, 0, r0, c1, c0, 1 tst r0, #(1 << 6) @ SMP/nAMP mode enabled? orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting #endif +__v7_setup: adr r12, __v7_setup_stack @ the local stack stmia r12, {r0-r5, r7, r9, r11, lr} bl v7_flush_dcache_all @@ -349,6 +350,29 @@ cpu_elf_name: .section ".proc.info.init", #alloc, #execinstr + .type __v7_ca9mp_proc_info, #object +__v7_ca9mp_proc_info: + .long 0x410fc090 @ Required ID value + .long 0xff0ffff0 @ Mask for ID + .long PMD_TYPE_SECT | \ + PMD_SECT_AP_WRITE | \ + PMD_SECT_AP_READ | \ + PMD_FLAGS + .long PMD_TYPE_SECT | \ + PMD_SECT_XN | \ + PMD_SECT_AP_WRITE | \ + PMD_SECT_AP_READ + b __v7_ca9mp_setup + .long cpu_arch_name + .long cpu_elf_name + .long HWCAP_SWP|HWCAP_HALF|HWCAP_THUMB|HWCAP_FAST_MULT|HWCAP_EDSP + .long cpu_v7_name + .long v7_processor_functions + .long v7wbi_tlb_fns + .long v6_user_fns + .long v7_cache_fns + .size __v7_ca9mp_proc_info, . - __v7_ca9mp_proc_info + /* * Match any ARMv7 processor core. */ -- cgit v0.10.2 From a43fb73101eaf6db0b33d22c152b338ab8b3edbb Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 17 Sep 2010 09:54:08 -0700 Subject: ceph: check mapping to determine if FILE_CACHE cap is used See if the i_data mapping has any pages to determine if the FILE_CACHE capability is currently in use, instead of assuming it is any time the rdcache_gen value is set (i.e., issued -> used). This allows the MDS RECALL_STATE process work for inodes that have cached pages. Signed-off-by: Sage Weil diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index b01c316..73c1530 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -814,7 +814,7 @@ int __ceph_caps_used(struct ceph_inode_info *ci) used |= CEPH_CAP_PIN; if (ci->i_rd_ref) used |= CEPH_CAP_FILE_RD; - if (ci->i_rdcache_ref || ci->i_rdcache_gen) + if (ci->i_rdcache_ref || ci->vfs_inode.i_data.nrpages) used |= CEPH_CAP_FILE_CACHE; if (ci->i_wr_ref) used |= CEPH_CAP_FILE_WR; -- cgit v0.10.2 From 653d48b22166db2d8b1515ebe6f9f0f7c95dfc86 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 17 Sep 2010 14:34:39 +0100 Subject: arm: fix really nasty sigreturn bug If a signal hits us outside of a syscall and another gets delivered when we are in sigreturn (e.g. because it had been in sa_mask for the first one and got sent to us while we'd been in the first handler), we have a chance of returning from the second handler to location one insn prior to where we ought to return. If r0 happens to contain -513 (-ERESTARTNOINTR), sigreturn will get confused into doing restart syscall song and dance. Incredible joy to debug, since it manifests as random, infrequent and very hard to reproduce double execution of instructions in userland code... The fix is simple - mark it "don't bother with restarts" in wrapper, i.e. set r8 to 0 in sys_sigreturn and sys_rt_sigreturn wrappers, suppressing the syscall restart handling on return from these guys. They can't legitimately return a restart-worthy error anyway. Testcase: #include #include #include #include #include void f(int n) { __asm__ __volatile__( "ldr r0, [%0]\n" "b 1f\n" "b 2f\n" "1:b .\n" "2:\n" : : "r"(&n)); } void handler1(int sig) { } void handler2(int sig) { raise(1); } void handler3(int sig) { exit(0); } main() { struct sigaction s = {.sa_handler = handler2}; struct itimerval t1 = { .it_value = {1} }; struct itimerval t2 = { .it_value = {2} }; signal(1, handler1); sigemptyset(&s.sa_mask); sigaddset(&s.sa_mask, 1); sigaction(SIGALRM, &s, NULL); signal(SIGVTALRM, handler3); setitimer(ITIMER_REAL, &t1, NULL); setitimer(ITIMER_VIRTUAL, &t2, NULL); f(-513); /* -ERESTARTNOINTR */ write(1, "buggered\n", 9); return 1; } Signed-off-by: Al Viro Acked-by: Russell King Cc: stable@kernel.org Signed-off-by: Linus Torvalds diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index f05a35a..1b560825 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -418,11 +418,13 @@ ENDPROC(sys_clone_wrapper) sys_sigreturn_wrapper: add r0, sp, #S_OFF + mov why, #0 @ prevent syscall restart handling b sys_sigreturn ENDPROC(sys_sigreturn_wrapper) sys_rt_sigreturn_wrapper: add r0, sp, #S_OFF + mov why, #0 @ prevent syscall restart handling b sys_rt_sigreturn ENDPROC(sys_rt_sigreturn_wrapper) -- cgit v0.10.2 From be4f104dfd3b5e3ae262bff607965cfc38027dec Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 17 Sep 2010 12:30:31 -0700 Subject: ceph: select CRYPTO We select CRYPTO_AES, but not CRYPTO. Signed-off-by: Sage Weil diff --git a/fs/ceph/Kconfig b/fs/ceph/Kconfig index bc87b9c..0fcd264 100644 --- a/fs/ceph/Kconfig +++ b/fs/ceph/Kconfig @@ -3,6 +3,7 @@ config CEPH_FS depends on INET && EXPERIMENTAL select LIBCRC32C select CRYPTO_AES + select CRYPTO help Choose Y or M here to include support for mounting the experimental Ceph distributed file system. Ceph is an extremely -- cgit v0.10.2 From 5343795fda222183c1f77d0f5870f2027713bb0d Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Sat, 18 Sep 2010 09:54:38 +0900 Subject: ARM: S3C64XX: Prototype SPI devices Avoids build warnings due to the undeclared non-statics. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s3c64xx/dev-spi.c b/arch/arm/mach-s3c64xx/dev-spi.c index bfa3ef7..405e621 100644 --- a/arch/arm/mach-s3c64xx/dev-spi.c +++ b/arch/arm/mach-s3c64xx/dev-spi.c @@ -22,6 +22,7 @@ #include #include +#include static char *spi_src_clks[] = { [S3C64XX_SPI_SRCCLK_PCLK] = "pclk", -- cgit v0.10.2 From 591cd25ee35986184870c447ff5c12e78bc80699 Mon Sep 17 00:00:00 2001 From: Darius Augulis Date: Thu, 9 Sep 2010 21:40:22 +0900 Subject: ARM: S3C64XX: Fix coding style errors on mach-real6410 Fix errors reported by checkpatch.pl script Signed-off-by: Darius Augulis [kgene.kim@samsung.com: minor title fix] Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s3c64xx/mach-real6410.c b/arch/arm/mach-s3c64xx/mach-real6410.c index 5c07d01..9f79e6b 100644 --- a/arch/arm/mach-s3c64xx/mach-real6410.c +++ b/arch/arm/mach-s3c64xx/mach-real6410.c @@ -30,73 +30,73 @@ #include #include -#define UCON S3C2410_UCON_DEFAULT | S3C2410_UCON_UCLK -#define ULCON S3C2410_LCON_CS8 | S3C2410_LCON_PNONE | S3C2410_LCON_STOPB -#define UFCON S3C2410_UFCON_RXTRIG8 | S3C2410_UFCON_FIFOMODE +#define UCON (S3C2410_UCON_DEFAULT | S3C2410_UCON_UCLK) +#define ULCON (S3C2410_LCON_CS8 | S3C2410_LCON_PNONE | S3C2410_LCON_STOPB) +#define UFCON (S3C2410_UFCON_RXTRIG8 | S3C2410_UFCON_FIFOMODE) static struct s3c2410_uartcfg real6410_uartcfgs[] __initdata = { [0] = { - .hwport = 0, - .flags = 0, - .ucon = UCON, - .ulcon = ULCON, - .ufcon = UFCON, + .hwport = 0, + .flags = 0, + .ucon = UCON, + .ulcon = ULCON, + .ufcon = UFCON, }, [1] = { - .hwport = 1, - .flags = 0, - .ucon = UCON, - .ulcon = ULCON, - .ufcon = UFCON, + .hwport = 1, + .flags = 0, + .ucon = UCON, + .ulcon = ULCON, + .ufcon = UFCON, }, [2] = { - .hwport = 2, - .flags = 0, - .ucon = UCON, - .ulcon = ULCON, - .ufcon = UFCON, + .hwport = 2, + .flags = 0, + .ucon = UCON, + .ulcon = ULCON, + .ufcon = UFCON, }, [3] = { - .hwport = 3, - .flags = 0, - .ucon = UCON, - .ulcon = ULCON, - .ufcon = UFCON, + .hwport = 3, + .flags = 0, + .ucon = UCON, + .ulcon = ULCON, + .ufcon = UFCON, }, }; /* DM9000AEP 10/100 ethernet controller */ static struct resource real6410_dm9k_resource[] = { - [0] = { - .start = S3C64XX_PA_XM0CSN1, - .end = S3C64XX_PA_XM0CSN1 + 1, - .flags = IORESOURCE_MEM - }, - [1] = { - .start = S3C64XX_PA_XM0CSN1 + 4, - .end = S3C64XX_PA_XM0CSN1 + 5, - .flags = IORESOURCE_MEM - }, - [2] = { - .start = S3C_EINT(7), - .end = S3C_EINT(7), - .flags = IORESOURCE_IRQ, - } + [0] = { + .start = S3C64XX_PA_XM0CSN1, + .end = S3C64XX_PA_XM0CSN1 + 1, + .flags = IORESOURCE_MEM + }, + [1] = { + .start = S3C64XX_PA_XM0CSN1 + 4, + .end = S3C64XX_PA_XM0CSN1 + 5, + .flags = IORESOURCE_MEM + }, + [2] = { + .start = S3C_EINT(7), + .end = S3C_EINT(7), + .flags = IORESOURCE_IRQ, + } }; static struct dm9000_plat_data real6410_dm9k_pdata = { - .flags = (DM9000_PLATF_16BITONLY | DM9000_PLATF_NO_EEPROM), + .flags = (DM9000_PLATF_16BITONLY | DM9000_PLATF_NO_EEPROM), }; static struct platform_device real6410_device_eth = { - .name = "dm9000", - .id = -1, - .num_resources = ARRAY_SIZE(real6410_dm9k_resource), - .resource = real6410_dm9k_resource, - .dev = { - .platform_data = &real6410_dm9k_pdata, - }, + .name = "dm9000", + .id = -1, + .num_resources = ARRAY_SIZE(real6410_dm9k_resource), + .resource = real6410_dm9k_resource, + .dev = { + .platform_data = &real6410_dm9k_pdata, + }, }; static struct platform_device *real6410_devices[] __initdata = { @@ -129,12 +129,12 @@ static void __init real6410_machine_init(void) /* set timing for nCS1 suitable for ethernet chip */ __raw_writel((0 << S3C64XX_SROM_BCX__PMC__SHIFT) | - (6 << S3C64XX_SROM_BCX__TACP__SHIFT) | - (4 << S3C64XX_SROM_BCX__TCAH__SHIFT) | - (1 << S3C64XX_SROM_BCX__TCOH__SHIFT) | - (13 << S3C64XX_SROM_BCX__TACC__SHIFT) | - (4 << S3C64XX_SROM_BCX__TCOS__SHIFT) | - (0 << S3C64XX_SROM_BCX__TACS__SHIFT), S3C64XX_SROM_BC1); + (6 << S3C64XX_SROM_BCX__TACP__SHIFT) | + (4 << S3C64XX_SROM_BCX__TCAH__SHIFT) | + (1 << S3C64XX_SROM_BCX__TCOH__SHIFT) | + (13 << S3C64XX_SROM_BCX__TACC__SHIFT) | + (4 << S3C64XX_SROM_BCX__TCOS__SHIFT) | + (0 << S3C64XX_SROM_BCX__TACS__SHIFT), S3C64XX_SROM_BC1); platform_add_devices(real6410_devices, ARRAY_SIZE(real6410_devices)); } -- cgit v0.10.2 From 4d89ecaae9c145e60b920acb1120f1de3d6eac6e Mon Sep 17 00:00:00 2001 From: Darius Augulis Date: Thu, 9 Sep 2010 21:41:31 +0900 Subject: ARM: S3C64XX: Add IORESOURCE_IRQ_HIGHLEVEL flag to dm9000 on mach-real6410 Add IORESOURCE_IRQ_HIGHLEVEL irq flag to dm9000 driver platform data in board mach-real6410. Signed-off-by: Darius Augulis [kgene.kim@samsung.com: minor title fix] Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s3c64xx/mach-real6410.c b/arch/arm/mach-s3c64xx/mach-real6410.c index 9f79e6b..e130379 100644 --- a/arch/arm/mach-s3c64xx/mach-real6410.c +++ b/arch/arm/mach-s3c64xx/mach-real6410.c @@ -81,7 +81,7 @@ static struct resource real6410_dm9k_resource[] = { [2] = { .start = S3C_EINT(7), .end = S3C_EINT(7), - .flags = IORESOURCE_IRQ, + .flags = IORESOURCE_IRQ | IORESOURCE_IRQ_HIGHLEVEL } }; -- cgit v0.10.2 From 7cdffc86528ec9c55c83c649b6d64cadeb558136 Mon Sep 17 00:00:00 2001 From: Dominik Brodowski Date: Sat, 18 Sep 2010 10:19:13 +0200 Subject: pcmcia: preserve configuration information if request_io fails partly If pcmcia_request_io() only fails partly -- for the second of two requested resources -- preserve the configuration settings for the first one. Signed-off-by: Dominik Brodowski diff --git a/drivers/pcmcia/pcmcia_resource.c b/drivers/pcmcia/pcmcia_resource.c index a5c1765..9ba4dad 100644 --- a/drivers/pcmcia/pcmcia_resource.c +++ b/drivers/pcmcia/pcmcia_resource.c @@ -595,7 +595,13 @@ int pcmcia_request_io(struct pcmcia_device *p_dev) if (c->io[1].end) { ret = alloc_io_space(s, &c->io[1], p_dev->io_lines); if (ret) { + struct resource tmp = c->io[0]; + /* release the previously allocated resource */ release_io_space(s, &c->io[0]); + /* but preserve the settings, for they worked... */ + c->io[0].end = resource_size(&tmp); + c->io[0].start = tmp.start; + c->io[0].flags = tmp.flags; goto out; } } else -- cgit v0.10.2 From 50aff040363d31f87e94f38f1710973d99489951 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Sat, 21 Aug 2010 14:40:20 +0800 Subject: ocfs2/net: fix uninitialized ret in o2net_send_message_vec() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit mmotm/fs/ocfs2/cluster/tcp.c: In function ‘o2net_send_message_vec’: mmotm/fs/ocfs2/cluster/tcp.c:980:6: warning: ‘ret’ may be used uninitialized in this function It seems a real bug introduced by commit 9af0b38ff3 (ocfs2/net: Use wait_event() in o2net_send_message_vec()). cc: Sunil Mushran Signed-off-by: Wu Fengguang Signed-off-by: Joel Becker diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 1361997..cbe2f05 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -977,7 +977,7 @@ static int o2net_tx_can_proceed(struct o2net_node *nn, int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec, size_t caller_veclen, u8 target_node, int *status) { - int ret; + int ret = 0; struct o2net_msg *msg = NULL; size_t veclen, caller_bytes = 0; struct kvec *vec = NULL; -- cgit v0.10.2 From 31019075f41c785eff7f38c62e4c700af019fdb7 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 14 Sep 2010 04:23:47 -0400 Subject: alpha: Use static const char * const where possible Acked-by: Richard Henderson Signed-off-by: Joe Perches Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/err_ev6.c b/arch/alpha/kernel/err_ev6.c index 8ca6345..253cf1a 100644 --- a/arch/alpha/kernel/err_ev6.c +++ b/arch/alpha/kernel/err_ev6.c @@ -90,11 +90,13 @@ static int ev6_parse_cbox(u64 c_addr, u64 c1_syn, u64 c2_syn, u64 c_stat, u64 c_sts, int print) { - char *sourcename[] = { "UNKNOWN", "UNKNOWN", "UNKNOWN", - "MEMORY", "BCACHE", "DCACHE", - "BCACHE PROBE", "BCACHE PROBE" }; - char *streamname[] = { "D", "I" }; - char *bitsname[] = { "SINGLE", "DOUBLE" }; + static const char * const sourcename[] = { + "UNKNOWN", "UNKNOWN", "UNKNOWN", + "MEMORY", "BCACHE", "DCACHE", + "BCACHE PROBE", "BCACHE PROBE" + }; + static const char * const streamname[] = { "D", "I" }; + static const char * const bitsname[] = { "SINGLE", "DOUBLE" }; int status = MCHK_DISPOSITION_REPORT; int source = -1, stream = -1, bits = -1; diff --git a/arch/alpha/kernel/err_marvel.c b/arch/alpha/kernel/err_marvel.c index 5c905aa..648ae88 100644 --- a/arch/alpha/kernel/err_marvel.c +++ b/arch/alpha/kernel/err_marvel.c @@ -589,22 +589,23 @@ marvel_print_pox_spl_cmplt(u64 spl_cmplt) static void marvel_print_pox_trans_sum(u64 trans_sum) { - char *pcix_cmd[] = { "Interrupt Acknowledge", - "Special Cycle", - "I/O Read", - "I/O Write", - "Reserved", - "Reserved / Device ID Message", - "Memory Read", - "Memory Write", - "Reserved / Alias to Memory Read Block", - "Reserved / Alias to Memory Write Block", - "Configuration Read", - "Configuration Write", - "Memory Read Multiple / Split Completion", - "Dual Address Cycle", - "Memory Read Line / Memory Read Block", - "Memory Write and Invalidate / Memory Write Block" + static const char * const pcix_cmd[] = { + "Interrupt Acknowledge", + "Special Cycle", + "I/O Read", + "I/O Write", + "Reserved", + "Reserved / Device ID Message", + "Memory Read", + "Memory Write", + "Reserved / Alias to Memory Read Block", + "Reserved / Alias to Memory Write Block", + "Configuration Read", + "Configuration Write", + "Memory Read Multiple / Split Completion", + "Dual Address Cycle", + "Memory Read Line / Memory Read Block", + "Memory Write and Invalidate / Memory Write Block" }; #define IO7__POX_TRANSUM__PCI_ADDR__S (0) diff --git a/arch/alpha/kernel/err_titan.c b/arch/alpha/kernel/err_titan.c index f7ed97c..c3b3781 100644 --- a/arch/alpha/kernel/err_titan.c +++ b/arch/alpha/kernel/err_titan.c @@ -75,8 +75,12 @@ titan_parse_p_serror(int which, u64 serror, int print) int status = MCHK_DISPOSITION_REPORT; #ifdef CONFIG_VERBOSE_MCHECK - char *serror_src[] = {"GPCI", "APCI", "AGP HP", "AGP LP"}; - char *serror_cmd[] = {"DMA Read", "DMA RMW", "SGTE Read", "Reserved"}; + static const char * const serror_src[] = { + "GPCI", "APCI", "AGP HP", "AGP LP" + }; + static const char * const serror_cmd[] = { + "DMA Read", "DMA RMW", "SGTE Read", "Reserved" + }; #endif /* CONFIG_VERBOSE_MCHECK */ #define TITAN__PCHIP_SERROR__LOST_UECC (1UL << 0) @@ -140,14 +144,15 @@ titan_parse_p_perror(int which, int port, u64 perror, int print) int status = MCHK_DISPOSITION_REPORT; #ifdef CONFIG_VERBOSE_MCHECK - char *perror_cmd[] = { "Interrupt Acknowledge", "Special Cycle", - "I/O Read", "I/O Write", - "Reserved", "Reserved", - "Memory Read", "Memory Write", - "Reserved", "Reserved", - "Configuration Read", "Configuration Write", - "Memory Read Multiple", "Dual Address Cycle", - "Memory Read Line","Memory Write and Invalidate" + static const char * const perror_cmd[] = { + "Interrupt Acknowledge", "Special Cycle", + "I/O Read", "I/O Write", + "Reserved", "Reserved", + "Memory Read", "Memory Write", + "Reserved", "Reserved", + "Configuration Read", "Configuration Write", + "Memory Read Multiple", "Dual Address Cycle", + "Memory Read Line", "Memory Write and Invalidate" }; #endif /* CONFIG_VERBOSE_MCHECK */ @@ -273,11 +278,11 @@ titan_parse_p_agperror(int which, u64 agperror, int print) int cmd, len; unsigned long addr; - char *agperror_cmd[] = { "Read (low-priority)", "Read (high-priority)", - "Write (low-priority)", - "Write (high-priority)", - "Reserved", "Reserved", - "Flush", "Fence" + static const char * const agperror_cmd[] = { + "Read (low-priority)", "Read (high-priority)", + "Write (low-priority)", "Write (high-priority)", + "Reserved", "Reserved", + "Flush", "Fence" }; #endif /* CONFIG_VERBOSE_MCHECK */ diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c index 5d1e6d6..16bc1f5 100644 --- a/arch/alpha/kernel/osf_sys.c +++ b/arch/alpha/kernel/osf_sys.c @@ -594,7 +594,7 @@ SYSCALL_DEFINE2(osf_sigstack, struct sigstack __user *, uss, SYSCALL_DEFINE3(osf_sysinfo, int, command, char __user *, buf, long, count) { - char *sysinfo_table[] = { + const char *sysinfo_table[] = { utsname()->sysname, utsname()->nodename, utsname()->release, @@ -606,7 +606,7 @@ SYSCALL_DEFINE3(osf_sysinfo, int, command, char __user *, buf, long, count) "dummy", /* secure RPC domain */ }; unsigned long offset; - char *res; + const char *res; long len, err = -EINVAL; offset = command-1; -- cgit v0.10.2 From af96f8a340525f24abecc25848a07fff7636a53e Mon Sep 17 00:00:00 2001 From: matt mooney Date: Tue, 14 Sep 2010 05:27:39 -0400 Subject: alpha: remove unnecessary cast from void* in assignment. Acked-by: Jan-Benedict Glaw Signed-off-by: matt mooney Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/pci-sysfs.c b/arch/alpha/kernel/pci-sysfs.c index 738fc82..b899e95 100644 --- a/arch/alpha/kernel/pci-sysfs.c +++ b/arch/alpha/kernel/pci-sysfs.c @@ -66,7 +66,7 @@ static int pci_mmap_resource(struct kobject *kobj, { struct pci_dev *pdev = to_pci_dev(container_of(kobj, struct device, kobj)); - struct resource *res = (struct resource *)attr->private; + struct resource *res = attr->private; enum pci_mmap_state mmap_type; struct pci_bus_region bar; int i; diff --git a/arch/alpha/kernel/srm_env.c b/arch/alpha/kernel/srm_env.c index 4afc1a1..f0df3fb 100644 --- a/arch/alpha/kernel/srm_env.c +++ b/arch/alpha/kernel/srm_env.c @@ -87,7 +87,7 @@ static int srm_env_proc_show(struct seq_file *m, void *v) srm_env_t *entry; char *page; - entry = (srm_env_t *)m->private; + entry = m->private; page = (char *)__get_free_page(GFP_USER); if (!page) return -ENOMEM; -- cgit v0.10.2 From b97f897d60c34adb9a5a37a14033e6a867e86f48 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 14 Sep 2010 09:00:22 -0400 Subject: alpha: fix build breakage in asm/cacheflush.h Alpha SMP flush_icache_user_range() is implemented as an inline function inside include/asm/cacheflush.h. It dereferences @current but doesn't include linux/sched.h and thus causes build failure if linux/sched.h wasn't included previously. Fix it by including the needed header file explicitly. Signed-off-by: Tejun Heo Reported-by: Stephen Rothwell Signed-off-by: Matt Turner diff --git a/arch/alpha/include/asm/cacheflush.h b/arch/alpha/include/asm/cacheflush.h index 01d71e1..012f124 100644 --- a/arch/alpha/include/asm/cacheflush.h +++ b/arch/alpha/include/asm/cacheflush.h @@ -43,6 +43,8 @@ extern void smp_imb(void); /* ??? Ought to use this in arch/alpha/kernel/signal.c too. */ #ifndef CONFIG_SMP +#include + extern void __load_new_mm_context(struct mm_struct *); static inline void flush_icache_user_range(struct vm_area_struct *vma, struct page *page, -- cgit v0.10.2 From 12e750d956eec8b1778679aff705f617095c46c8 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 14 Sep 2010 19:34:56 -0400 Subject: alpha: kill big kernel lock All uses of the BKL on alpha are totally bogus, nothing is really protected by this. Remove the remaining users so we don't have to mark alpha as 'depends on BKL'. Signed-off-by: Arnd Bergmann Cc: Richard Henderson Cc: Ivan Kokshaysky Cc: linux-alpha@vger.kernel.org Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c index 16bc1f5..547e8b8 100644 --- a/arch/alpha/kernel/osf_sys.c +++ b/arch/alpha/kernel/osf_sys.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include @@ -69,7 +68,6 @@ SYSCALL_DEFINE4(osf_set_program_attributes, unsigned long, text_start, { struct mm_struct *mm; - lock_kernel(); mm = current->mm; mm->end_code = bss_start + bss_len; mm->start_brk = bss_start + bss_len; @@ -78,7 +76,6 @@ SYSCALL_DEFINE4(osf_set_program_attributes, unsigned long, text_start, printk("set_program_attributes(%lx %lx %lx %lx)\n", text_start, text_len, bss_start, bss_len); #endif - unlock_kernel(); return 0; } @@ -517,7 +514,6 @@ SYSCALL_DEFINE2(osf_proplist_syscall, enum pl_code, code, long error; int __user *min_buf_size_ptr; - lock_kernel(); switch (code) { case PL_SET: if (get_user(error, &args->set.nbytes)) @@ -547,7 +543,6 @@ SYSCALL_DEFINE2(osf_proplist_syscall, enum pl_code, code, error = -EOPNOTSUPP; break; }; - unlock_kernel(); return error; } diff --git a/arch/alpha/kernel/traps.c b/arch/alpha/kernel/traps.c index b14f015..0414e02 100644 --- a/arch/alpha/kernel/traps.c +++ b/arch/alpha/kernel/traps.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include @@ -623,7 +622,6 @@ do_entUna(void * va, unsigned long opcode, unsigned long reg, return; } - lock_kernel(); printk("Bad unaligned kernel access at %016lx: %p %lx %lu\n", pc, va, opcode, reg); do_exit(SIGSEGV); @@ -646,7 +644,6 @@ got_exception: * Yikes! No one to forward the exception to. * Since the registers are in a weird format, dump them ourselves. */ - lock_kernel(); printk("%s(%d): unhandled unaligned exception\n", current->comm, task_pid_nr(current)); -- cgit v0.10.2 From 531f0474bf889d489b28ea4b15e87b7cbb77f580 Mon Sep 17 00:00:00 2001 From: Mikael Pettersson Date: Thu, 16 Sep 2010 14:12:55 -0400 Subject: alpha: wire up fanotify and prlimit64 syscalls The 2.6.36-rc kernel added three new system calls: fanotify_init, fanotify_mark, and prlimit64. This patch wires them up on Alpha. Built and booted on an XP900. Untested beyond that. Signed-off-by: Mikael Pettersson Signed-off-by: Matt Turner diff --git a/arch/alpha/include/asm/unistd.h b/arch/alpha/include/asm/unistd.h index 804e531..9a9672f 100644 --- a/arch/alpha/include/asm/unistd.h +++ b/arch/alpha/include/asm/unistd.h @@ -449,10 +449,13 @@ #define __NR_pwritev 491 #define __NR_rt_tgsigqueueinfo 492 #define __NR_perf_event_open 493 +#define __NR_fanotify_init 494 +#define __NR_fanotify_mark 495 +#define __NR_prlimit64 496 #ifdef __KERNEL__ -#define NR_SYSCALLS 494 +#define NR_SYSCALLS 497 #define __ARCH_WANT_IPC_PARSE_VERSION #define __ARCH_WANT_OLD_READDIR diff --git a/arch/alpha/kernel/systbls.S b/arch/alpha/kernel/systbls.S index 09acb78..ce594ef 100644 --- a/arch/alpha/kernel/systbls.S +++ b/arch/alpha/kernel/systbls.S @@ -512,6 +512,9 @@ sys_call_table: .quad sys_pwritev .quad sys_rt_tgsigqueueinfo .quad sys_perf_event_open + .quad sys_fanotify_init + .quad sys_fanotify_mark /* 495 */ + .quad sys_prlimit64 .size sys_call_table, . - sys_call_table .type sys_call_table, @object -- cgit v0.10.2 From bdc8b8914b17cf97b53e5f7af8168d68cfb089d6 Mon Sep 17 00:00:00 2001 From: Michael Cree Date: Sun, 19 Sep 2010 02:05:40 -0400 Subject: alpha: Shift perf event pending work earlier in timer interrupt Pending work from the performance event subsystem is executed in the timer interrupt. This patch shifts the call to perf_event_do_pending() before the call to update_process_times() as the latter may call back into the perf event subsystem and it is prudent to have the pending work executed first. Signed-off-by: Michael Cree Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/time.c b/arch/alpha/kernel/time.c index eacceb2..396af17 100644 --- a/arch/alpha/kernel/time.c +++ b/arch/alpha/kernel/time.c @@ -191,16 +191,16 @@ irqreturn_t timer_interrupt(int irq, void *dev) write_sequnlock(&xtime_lock); -#ifndef CONFIG_SMP - while (nticks--) - update_process_times(user_mode(get_irq_regs())); -#endif - if (test_perf_event_pending()) { clear_perf_event_pending(); perf_event_do_pending(); } +#ifndef CONFIG_SMP + while (nticks--) + update_process_times(user_mode(get_irq_regs())); +#endif + return IRQ_HANDLED; } -- cgit v0.10.2 From 2deba1bd7126aadb5750beb927c878a6490065e6 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 18 Sep 2010 08:38:47 -0400 Subject: alpha: belated ERESTART_RESTARTBLOCK race fix same thing as had been done on other targets back in 2003 - move setting ->restart_block.fn into {rt_,}sigreturn(). Tested-by: Michael Cree Signed-off-by: Al Viro Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/signal.c b/arch/alpha/kernel/signal.c index 0932dbb..e5b1b3c 100644 --- a/arch/alpha/kernel/signal.c +++ b/arch/alpha/kernel/signal.c @@ -239,6 +239,8 @@ restore_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, unsigned long usp; long i, err = __get_user(regs->pc, &sc->sc_pc); + current_thread_info()->restart_block.fn = do_no_restart_syscall; + sw->r26 = (unsigned long) ret_from_sys_call; err |= __get_user(regs->r0, sc->sc_regs+0); @@ -591,7 +593,6 @@ syscall_restart(unsigned long r0, unsigned long r19, regs->pc -= 4; break; case ERESTART_RESTARTBLOCK: - current_thread_info()->restart_block.fn = do_no_restart_syscall; regs->r0 = EINTR; break; } -- cgit v0.10.2 From 392fb6e35400edbee183baba24b34a0fa2053813 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 18 Sep 2010 08:40:07 -0400 Subject: alpha: unb0rk sigsuspend() and rt_sigsuspend() Old code used to set regs->r0 and regs->r19 to force the right return value. Leaving that after switch to ERESTARTNOHAND was a Bad Idea(tm), since now that screws the restart - if we hit the case when get_signal_to_deliver() returns 0, we will step back to syscall insn, with v0 set to EINTR and a3 to 1. The latter won't matter, since EINTR is 4, aka __NR_write. Testcase: #include #define _GNU_SOURCE #include #include main() { sigset_t mask; sigemptyset(&mask); sigaddset(&mask, SIGCONT); sigprocmask(SIG_SETMASK, &mask, NULL); kill(0, SIGCONT); syscall(__NR_sigsuspend, 1, "b0rken\n", 7); } results on alpha in immediate message to stdout... Fix is obvious; moreover, since we don't need regs anymore, we can switch to normal prototypes for these guys and lose the wrappers. Even better, rt_sigsuspend() is identical to generic version in kernel/signal.c now. Tested-by: Michael Cree Signed-off-by: Al Viro Signed-off-by: Matt Turner diff --git a/arch/alpha/include/asm/unistd.h b/arch/alpha/include/asm/unistd.h index 9a9672f..058937b 100644 --- a/arch/alpha/include/asm/unistd.h +++ b/arch/alpha/include/asm/unistd.h @@ -466,6 +466,7 @@ #define __ARCH_WANT_SYS_OLD_GETRLIMIT #define __ARCH_WANT_SYS_OLDUMOUNT #define __ARCH_WANT_SYS_SIGPENDING +#define __ARCH_WANT_SYS_RT_SIGSUSPEND /* "Conditional" syscalls. What we want is diff --git a/arch/alpha/kernel/entry.S b/arch/alpha/kernel/entry.S index b45d913..bceb325 100644 --- a/arch/alpha/kernel/entry.S +++ b/arch/alpha/kernel/entry.S @@ -779,38 +779,6 @@ sys_rt_sigreturn: .end sys_rt_sigreturn .align 4 - .globl sys_sigsuspend - .ent sys_sigsuspend -sys_sigsuspend: - .prologue 0 - mov $sp, $17 - br $1, do_switch_stack - mov $sp, $18 - subq $sp, 16, $sp - stq $26, 0($sp) - jsr $26, do_sigsuspend - ldq $26, 0($sp) - lda $sp, SWITCH_STACK_SIZE+16($sp) - ret -.end sys_sigsuspend - - .align 4 - .globl sys_rt_sigsuspend - .ent sys_rt_sigsuspend -sys_rt_sigsuspend: - .prologue 0 - mov $sp, $18 - br $1, do_switch_stack - mov $sp, $19 - subq $sp, 16, $sp - stq $26, 0($sp) - jsr $26, do_rt_sigsuspend - ldq $26, 0($sp) - lda $sp, SWITCH_STACK_SIZE+16($sp) - ret -.end sys_rt_sigsuspend - - .align 4 .globl sys_sethae .ent sys_sethae sys_sethae: diff --git a/arch/alpha/kernel/signal.c b/arch/alpha/kernel/signal.c index e5b1b3c..0f6b51a 100644 --- a/arch/alpha/kernel/signal.c +++ b/arch/alpha/kernel/signal.c @@ -144,8 +144,7 @@ SYSCALL_DEFINE5(rt_sigaction, int, sig, const struct sigaction __user *, act, /* * Atomically swap in the new signal mask, and wait for a signal. */ -asmlinkage int -do_sigsuspend(old_sigset_t mask, struct pt_regs *regs, struct switch_stack *sw) +SYSCALL_DEFINE1(sigsuspend, old_sigset_t, mask) { mask &= _BLOCKABLE; spin_lock_irq(¤t->sighand->siglock); @@ -154,41 +153,6 @@ do_sigsuspend(old_sigset_t mask, struct pt_regs *regs, struct switch_stack *sw) recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); - /* Indicate EINTR on return from any possible signal handler, - which will not come back through here, but via sigreturn. */ - regs->r0 = EINTR; - regs->r19 = 1; - - current->state = TASK_INTERRUPTIBLE; - schedule(); - set_thread_flag(TIF_RESTORE_SIGMASK); - return -ERESTARTNOHAND; -} - -asmlinkage int -do_rt_sigsuspend(sigset_t __user *uset, size_t sigsetsize, - struct pt_regs *regs, struct switch_stack *sw) -{ - sigset_t set; - - /* XXX: Don't preclude handling different sized sigset_t's. */ - if (sigsetsize != sizeof(sigset_t)) - return -EINVAL; - if (copy_from_user(&set, uset, sizeof(set))) - return -EFAULT; - - sigdelsetmask(&set, ~_BLOCKABLE); - spin_lock_irq(¤t->sighand->siglock); - current->saved_sigmask = current->blocked; - current->blocked = set; - recalc_sigpending(); - spin_unlock_irq(¤t->sighand->siglock); - - /* Indicate EINTR on return from any possible signal handler, - which will not come back through here, but via sigreturn. */ - regs->r0 = EINTR; - regs->r19 = 1; - current->state = TASK_INTERRUPTIBLE; schedule(); set_thread_flag(TIF_RESTORE_SIGMASK); -- cgit v0.10.2 From 53293638618f1a8b0b182dfedaab08b28930f992 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 18 Sep 2010 08:41:16 -0400 Subject: alpha: fix a 14 years old bug in sigreturn tracing The way sigreturn() is implemented on alpha breaks PTRACE_SYSCALL, all way back to 1.3.95 when alpha has grown PTRACE_SYSCALL support. What happens is direct return to ret_from_syscall, in order to bypass mangling of a3 (error indicator) and prevent other mutilations of registers (e.g. by syscall restart). That's fine, but... the entire TIF_SYSCALL_TRACE codepath is kept separate on alpha and post-syscall stopping/notifying the tracer is after the syscall. And the normal path we are forcibly switching to doesn't have it. So we end up with *one* stop in traced sigreturn() vs. two in other syscalls. And yes, strace is visibly broken by that; try to strace the following #include #include void f(int sig) {} main() { signal(SIGHUP, f); raise(SIGHUP); write(1, "eeeek\n", 6); } and watch the show. The close(1) = 405 in the end of strace output is coming from return value of write() (6 == __NR_close on alpha) and syscall number of exit_group() (__NR_exit_group == 405 there). The fix is fairly simple - the only thing we end up missing is the call of syscall_trace() and we can tell whether we'd been called from the SYSCALL_TRACE path by checking ra value. Since we are setting the switch_stack up (that's what sys_sigreturn() does), we have the right environment for calling syscall_trace() - just before we call undo_switch_stack() and return. Since undo_switch_stack() will overwrite s0 anyway, we can use it to store the result of "has it been called from SYSCALL_TRACE path?" check. The same thing applies in rt_sigreturn(). Tested-by: Michael Cree Signed-off-by: Al Viro Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/entry.S b/arch/alpha/kernel/entry.S index bceb325..a3e9cd8 100644 --- a/arch/alpha/kernel/entry.S +++ b/arch/alpha/kernel/entry.S @@ -430,6 +430,7 @@ strace: beq $1, 1f ldq $27, 0($2) 1: jsr $26, ($27), sys_gettimeofday +ret_from_straced: ldgp $gp, 0($26) /* check return.. */ @@ -757,11 +758,15 @@ sys_vfork: .ent sys_sigreturn sys_sigreturn: .prologue 0 + lda $9, ret_from_straced + cmpult $26, $9, $9 mov $sp, $17 lda $18, -SWITCH_STACK_SIZE($sp) lda $sp, -SWITCH_STACK_SIZE($sp) jsr $26, do_sigreturn - br $1, undo_switch_stack + bne $9, 1f + jsr $26, syscall_trace +1: br $1, undo_switch_stack br ret_from_sys_call .end sys_sigreturn @@ -770,11 +775,15 @@ sys_sigreturn: .ent sys_rt_sigreturn sys_rt_sigreturn: .prologue 0 + lda $9, ret_from_straced + cmpult $26, $9, $9 mov $sp, $17 lda $18, -SWITCH_STACK_SIZE($sp) lda $sp, -SWITCH_STACK_SIZE($sp) jsr $26, do_rt_sigreturn - br $1, undo_switch_stack + bne $9, 1f + jsr $26, syscall_trace +1: br $1, undo_switch_stack br ret_from_sys_call .end sys_rt_sigreturn -- cgit v0.10.2 From 494486a1d2697f2153199b6501ab5b4d6e15a2bb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 18 Sep 2010 08:42:27 -0400 Subject: alpha: deal with multiple simultaneously pending signals Unlike the other targets, alpha sets _one_ sigframe and buggers off until the next syscall/interrupt, even if more signals are pending. It leads to quite a few unpleasant inconsistencies, starting with SIGSEGV potentially arriving not where it should and including e.g. mess with sigsuspend(); consider two pending signals blocked until sigsuspend() unblocks them. We pick the first one; then, if we are hit by interrupt while in the handler, we process the second one as well. If we are not, and if no syscalls had been made, we get out of the first handler and leave the second signal pending; normally sigreturn() would've picked it anyway, but here it starts with restoring the original mask and voila - the second signal is blocked again. On everything else we get both delivered consistently. It's actually easy to fix; the only thing to watch out for is prevention of double syscall restart. Fortunately, the idea I've nicked from arm fix by rmk works just fine... Testcase demonstrating the behaviour in question; on alpha we get one or both flags set (usually one), on everything else both are always set. #include #include int had1, had2; void f1(int sig) { had1 = 1; } void f2(int sig) { had2 = 1; } main() { sigset_t set1, set2; sigemptyset(&set1); sigemptyset(&set2); sigaddset(&set2, 1); sigaddset(&set2, 2); signal(1, f1); signal(2, f2); sigprocmask(SIG_SETMASK, &set2, NULL); raise(1); raise(2); sigsuspend(&set1); printf("had1:%d had2:%d\n", had1, had2); } Tested-by: Michael Cree Signed-off-by: Al Viro Signed-off-by: Matt Turner diff --git a/arch/alpha/kernel/entry.S b/arch/alpha/kernel/entry.S index a3e9cd8..ab1ee0a 100644 --- a/arch/alpha/kernel/entry.S +++ b/arch/alpha/kernel/entry.S @@ -317,14 +317,14 @@ ret_from_sys_call: ldq $0, SP_OFF($sp) and $0, 8, $0 beq $0, restore_all -ret_from_reschedule: +ret_to_user: /* Make sure need_resched and sigpending don't change between sampling and the rti. */ lda $16, 7 call_pal PAL_swpipl ldl $5, TI_FLAGS($8) and $5, _TIF_WORK_MASK, $2 - bne $5, work_pending + bne $2, work_pending restore_all: RESTORE_ALL call_pal PAL_rti @@ -363,7 +363,7 @@ $ret_success: * $8: current. * $19: The old syscall number, or zero if this is not a return * from a syscall that errored and is possibly restartable. - * $20: Error indication. + * $20: The old a3 value */ .align 4 @@ -392,12 +392,18 @@ $work_resched: $work_notifysig: mov $sp, $16 - br $1, do_switch_stack + bsr $1, do_switch_stack mov $sp, $17 mov $5, $18 + mov $19, $9 /* save old syscall number */ + mov $20, $10 /* save old a3 */ + and $5, _TIF_SIGPENDING, $2 + cmovne $2, 0, $9 /* we don't want double syscall restarts */ jsr $26, do_notify_resume + mov $9, $19 + mov $10, $20 bsr $1, undo_switch_stack - br restore_all + br ret_to_user .end work_pending /* -- cgit v0.10.2 From 79e27dc0677b969e2d53b76fa0fa58467cce946a Mon Sep 17 00:00:00 2001 From: Peter Korsgaard Date: Sat, 18 Sep 2010 19:55:10 +0100 Subject: ARM: 6400/1: at91: fix arch_gettimeoffset fallout 5cfc8ee0bb51 (ARM: convert arm to arch_gettimeoffset()) marked all of at91 AND at91x40 as needing ARCH_USES_GETTIMEOFFSET, and hence no high res timer support / accurate clock_gettime() - But only at91x40 needs it. Cc: stable@kernel.org Signed-off-by: Peter Korsgaard Acked-by: John Stultz Acked-by: Jean-Christophe PLAGNIOL-VILLARD Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 0f89335..88c97bc 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -271,7 +271,6 @@ config ARCH_AT91 bool "Atmel AT91" select ARCH_REQUIRE_GPIOLIB select HAVE_CLK - select ARCH_USES_GETTIMEOFFSET help This enables support for systems based on the Atmel AT91RM9200, AT91SAM9 and AT91CAP9 processors. -- cgit v0.10.2 From d907387c42e9e39261629890e45a08ef4c3ed3fe Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Mon, 13 Sep 2010 16:01:24 +0100 Subject: ARM: 6383/1: Implement phys_mem_access_prot() to avoid attributes aliasing ARMv7 onwards requires that there are no aliases to the same physical location using different memory types (i.e. Normal vs Strongly Ordered). Access to SO mappings when the unaligned accesses are handled in hardware is also Unpredictable (pgprot_noncached() mappings in user space). The /dev/mem driver requires uncached mappings with O_SYNC. The patch implements the phys_mem_access_prot() function which generates Strongly Ordered memory attributes if !pfn_valid() (independent of O_SYNC) and Normal Noncacheable (writecombine) if O_SYNC. Signed-off-by: Catalin Marinas Signed-off-by: Russell King diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h index ab68cf1..e90b167 100644 --- a/arch/arm/include/asm/pgtable.h +++ b/arch/arm/include/asm/pgtable.h @@ -317,6 +317,10 @@ static inline pte_t pte_mkspecial(pte_t pte) { return pte; } #ifdef CONFIG_ARM_DMA_MEM_BUFFERABLE #define pgprot_dmacoherent(prot) \ __pgprot_modify(prot, L_PTE_MT_MASK|L_PTE_EXEC, L_PTE_MT_BUFFERABLE) +#define __HAVE_PHYS_MEM_ACCESS_PROT +struct file; +extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, + unsigned long size, pgprot_t vma_prot); #else #define pgprot_dmacoherent(prot) \ __pgprot_modify(prot, L_PTE_MT_MASK|L_PTE_EXEC, L_PTE_MT_UNCACHED) diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c index 6e1c4f6..a486bd0 100644 --- a/arch/arm/mm/mmu.c +++ b/arch/arm/mm/mmu.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -498,6 +499,19 @@ static void __init build_mem_type_table(void) } } +#ifdef CONFIG_ARM_DMA_MEM_BUFFERABLE +pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, + unsigned long size, pgprot_t vma_prot) +{ + if (!pfn_valid(pfn)) + return pgprot_noncached(vma_prot); + else if (file->f_flags & O_SYNC) + return pgprot_writecombine(vma_prot); + return vma_prot; +} +EXPORT_SYMBOL(phys_mem_access_prot); +#endif + #define vectors_base() (vectors_high() ? 0xffff0000 : 0) static void __init *early_alloc(unsigned long sz) -- cgit v0.10.2 From cbfa5184cc5f58627f08c7fad225424f565b439d Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Fri, 17 Sep 2010 12:30:11 +0200 Subject: ASoC: fix clkdev API usage in sh/migor.c The clkdev API doesn't use .name and .id members of struct clk for clock lookup. Instead clocks should be added to a lookup list. Without this patch audio om the Migo-R board fails silently. Signed-off-by: Guennadi Liakhovetski Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/sh/migor.c b/sound/soc/sh/migor.c index b823a5c..87e2b7f 100644 --- a/sound/soc/sh/migor.c +++ b/sound/soc/sh/migor.c @@ -12,6 +12,7 @@ #include #include +#include #include #include @@ -40,12 +41,12 @@ static struct clk_ops siumckb_clk_ops = { }; static struct clk siumckb_clk = { - .name = "siumckb_clk", - .id = -1, .ops = &siumckb_clk_ops, .rate = 0, /* initialised at run-time */ }; +static struct clk_lookup *siumckb_lookup; + static int migor_hw_params(struct snd_pcm_substream *substream, struct snd_pcm_hw_params *params) { @@ -180,6 +181,13 @@ static int __init migor_init(void) if (ret < 0) return ret; + siumckb_lookup = clkdev_alloc(&siumckb_clk, "siumckb_clk", NULL); + if (!siumckb_lookup) { + ret = -ENOMEM; + goto eclkdevalloc; + } + clkdev_add(siumckb_lookup); + /* Port number used on this machine: port B */ migor_snd_device = platform_device_alloc("soc-audio", 1); if (!migor_snd_device) { @@ -200,12 +208,15 @@ static int __init migor_init(void) epdevadd: platform_device_put(migor_snd_device); epdevalloc: + clkdev_drop(siumckb_lookup); +eclkdevalloc: clk_unregister(&siumckb_clk); return ret; } static void __exit migor_exit(void) { + clkdev_drop(siumckb_lookup); clk_unregister(&siumckb_clk); platform_device_unregister(migor_snd_device); } -- cgit v0.10.2 From d93c333dc867d04111c9dcebdbcaa9ae8b2d5c2d Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 19 Sep 2010 16:16:58 +0100 Subject: ARM: Fix build error when using KCONFIG_CONFIG Jonathan Cameron reports that when using the environment variable KCONFIG_CONFIG, he encounters this error: make[2]: *** No rule to make target `.config', needed by `arch/arm/boot/compressed/vmlinux.lds' Reported-by: Jonathan Cameron Signed-off-by: Russell King diff --git a/arch/arm/boot/compressed/Makefile b/arch/arm/boot/compressed/Makefile index b23f6bc..65a7c1c 100644 --- a/arch/arm/boot/compressed/Makefile +++ b/arch/arm/boot/compressed/Makefile @@ -116,5 +116,5 @@ CFLAGS_font.o := -Dstatic= $(obj)/font.c: $(FONTC) $(call cmd,shipped) -$(obj)/vmlinux.lds: $(obj)/vmlinux.lds.in arch/arm/boot/Makefile .config +$(obj)/vmlinux.lds: $(obj)/vmlinux.lds.in arch/arm/boot/Makefile $(KCONFIG_CONFIG) @sed "$(SEDFLAGS)" < $< > $@ -- cgit v0.10.2 From 112d421df2fddc0278584b084f4fcfedd144c5f4 Mon Sep 17 00:00:00 2001 From: Jan Harkes Date: Fri, 17 Sep 2010 23:26:01 -0400 Subject: Coda: mount hangs because of missed REQ_WRITE rename Coda's REQ_* defines were renamed to avoid clashes with the block layer (commit 4aeefdc69f7b: "coda: fixup clash with block layer REQ_* defines"). However one was missed and response messages are no longer matched with requests and waiting threads are no longer woken up. This patch fixes this. Signed-off-by: Jan Harkes [ Also fixed up whitespace while at it -Linus ] Signed-off-by: Linus Torvalds diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index de89645..116af75 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -184,8 +184,8 @@ static ssize_t coda_psdev_write(struct file *file, const char __user *buf, } /* adjust outsize. is this useful ?? */ - req->uc_outSize = nbytes; - req->uc_flags |= REQ_WRITE; + req->uc_outSize = nbytes; + req->uc_flags |= CODA_REQ_WRITE; count = nbytes; /* Convert filedescriptor into a file handle */ -- cgit v0.10.2 From e4ff1c39ee1122198e8355069da59297038e55bb Mon Sep 17 00:00:00 2001 From: Arnaud Patard Date: Sun, 22 Aug 2010 22:49:46 +0200 Subject: ARM: kirkwood: Unbreak PCIe I/O port The support for the 2 pcie port of the 6282 has broken i/o port by switching *_IO_PHYS_BASE and *_IO_BUS_BASE. In fact, the patches reintroduced the same bug solved by commit 35f029e2514be209eb0e88c7d927f3bcc42a5cc2. So, I'm adding back *_IO_BUS_BASE in resource declaration and fix definition of KIRKWOOD_PCIE1_IO_BUS_BASE. With this change, the xgi card on my t5325 is working again. Signed-off-by: Arnaud Patard Acked-by: Saeed Bishara Signed-off-by: Nicolas Pitre Cc: stable@kernel.org diff --git a/arch/arm/mach-kirkwood/include/mach/kirkwood.h b/arch/arm/mach-kirkwood/include/mach/kirkwood.h index 93fc2ec..6e924b3 100644 --- a/arch/arm/mach-kirkwood/include/mach/kirkwood.h +++ b/arch/arm/mach-kirkwood/include/mach/kirkwood.h @@ -38,7 +38,7 @@ #define KIRKWOOD_PCIE1_IO_PHYS_BASE 0xf3000000 #define KIRKWOOD_PCIE1_IO_VIRT_BASE 0xfef00000 -#define KIRKWOOD_PCIE1_IO_BUS_BASE 0x00000000 +#define KIRKWOOD_PCIE1_IO_BUS_BASE 0x00100000 #define KIRKWOOD_PCIE1_IO_SIZE SZ_1M #define KIRKWOOD_PCIE_IO_PHYS_BASE 0xf2000000 diff --git a/arch/arm/mach-kirkwood/pcie.c b/arch/arm/mach-kirkwood/pcie.c index 55e7f00..513ad31 100644 --- a/arch/arm/mach-kirkwood/pcie.c +++ b/arch/arm/mach-kirkwood/pcie.c @@ -117,7 +117,7 @@ static void __init pcie0_ioresources_init(struct pcie_port *pp) * IORESOURCE_IO */ pp->res[0].name = "PCIe 0 I/O Space"; - pp->res[0].start = KIRKWOOD_PCIE_IO_PHYS_BASE; + pp->res[0].start = KIRKWOOD_PCIE_IO_BUS_BASE; pp->res[0].end = pp->res[0].start + KIRKWOOD_PCIE_IO_SIZE - 1; pp->res[0].flags = IORESOURCE_IO; @@ -139,7 +139,7 @@ static void __init pcie1_ioresources_init(struct pcie_port *pp) * IORESOURCE_IO */ pp->res[0].name = "PCIe 1 I/O Space"; - pp->res[0].start = KIRKWOOD_PCIE1_IO_PHYS_BASE; + pp->res[0].start = KIRKWOOD_PCIE1_IO_BUS_BASE; pp->res[0].end = pp->res[0].start + KIRKWOOD_PCIE1_IO_SIZE - 1; pp->res[0].flags = IORESOURCE_IO; -- cgit v0.10.2 From 863636828f1fcd9fdc15e24d620aa53cf18b432f Mon Sep 17 00:00:00 2001 From: Simon Guinot Date: Fri, 17 Sep 2010 23:33:51 +0200 Subject: dmaengine: fix interrupt clearing for mv_xor When using simultaneously the two DMA channels on a same engine, some transfers are never completed. For example, an endless lock can occur while writing heavily on a RAID5 array (with async-tx offload support enabled). Note that this issue can also be reproduced by using the DMA test client. On a same engine, the interrupt cause register is shared between two DMA channels. This patch make sure that the cause bit is only cleared for the requested channel. Signed-off-by: Simon Guinot Tested-by: Luc Saillard Acked-by: Saeed Bishara Signed-off-by: Nicolas Pitre diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c index 86c5ae9..411d5bf 100644 --- a/drivers/dma/mv_xor.c +++ b/drivers/dma/mv_xor.c @@ -162,7 +162,7 @@ static int mv_is_err_intr(u32 intr_cause) static void mv_xor_device_clear_eoc_cause(struct mv_xor_chan *chan) { - u32 val = (1 << (1 + (chan->idx * 16))); + u32 val = ~(1 << (chan->idx * 16)); dev_dbg(chan->device->common.dev, "%s, val 0x%08x\n", __func__, val); __raw_writel(val, XOR_INTR_CAUSE(chan)); } -- cgit v0.10.2 From c4a90588fa07ea47df7a67fd6cb03d6bc0f99634 Mon Sep 17 00:00:00 2001 From: Eric Miao Date: Tue, 14 Sep 2010 16:14:15 +0800 Subject: ARM: dove: fix __io() definition to use bus based offset Signed-off-by: Eric Miao Acked-by: Saeed Bishara Signed-off-by: Nicolas Pitre diff --git a/arch/arm/mach-dove/include/mach/io.h b/arch/arm/mach-dove/include/mach/io.h index 3b3e472..eb4936f 100644 --- a/arch/arm/mach-dove/include/mach/io.h +++ b/arch/arm/mach-dove/include/mach/io.h @@ -13,8 +13,8 @@ #define IO_SPACE_LIMIT 0xffffffff -#define __io(a) ((void __iomem *)(((a) - DOVE_PCIE0_IO_PHYS_BASE) +\ - DOVE_PCIE0_IO_VIRT_BASE)) -#define __mem_pci(a) (a) +#define __io(a) ((void __iomem *)(((a) - DOVE_PCIE0_IO_BUS_BASE) + \ + DOVE_PCIE0_IO_VIRT_BASE)) +#define __mem_pci(a) (a) #endif -- cgit v0.10.2 From f539dfedbd169e5ed47912bb517c75976ab556f3 Mon Sep 17 00:00:00 2001 From: Simon Guinot Date: Sun, 19 Sep 2010 15:30:59 +0200 Subject: leds: leds-ns2: fix locking This patch replace all the lock functions with the irq safe variant. The ns2_led_{set,get}_mode() functions must be safe in all context. For example, the trigger timer call led_set_brightness() in a softirq context. Signed-off-by: Simon Guinot Signed-off-by: Nicolas Pitre diff --git a/drivers/leds/leds-ns2.c b/drivers/leds/leds-ns2.c index 74dce4b..350eb34 100644 --- a/drivers/leds/leds-ns2.c +++ b/drivers/leds/leds-ns2.c @@ -81,7 +81,7 @@ static int ns2_led_get_mode(struct ns2_led_data *led_dat, int cmd_level; int slow_level; - read_lock(&led_dat->rw_lock); + read_lock_irq(&led_dat->rw_lock); cmd_level = gpio_get_value(led_dat->cmd); slow_level = gpio_get_value(led_dat->slow); @@ -95,7 +95,7 @@ static int ns2_led_get_mode(struct ns2_led_data *led_dat, } } - read_unlock(&led_dat->rw_lock); + read_unlock_irq(&led_dat->rw_lock); return ret; } @@ -104,8 +104,9 @@ static void ns2_led_set_mode(struct ns2_led_data *led_dat, enum ns2_led_modes mode) { int i; + unsigned long flags; - write_lock(&led_dat->rw_lock); + write_lock_irqsave(&led_dat->rw_lock, flags); for (i = 0; i < ARRAY_SIZE(ns2_led_modval); i++) { if (mode == ns2_led_modval[i].mode) { @@ -116,7 +117,7 @@ static void ns2_led_set_mode(struct ns2_led_data *led_dat, } } - write_unlock(&led_dat->rw_lock); + write_unlock_irqrestore(&led_dat->rw_lock, flags); } static void ns2_led_set(struct led_classdev *led_cdev, -- cgit v0.10.2 From 31c4a3d3a0f84a5847665f8aa0552d188389f791 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sun, 19 Sep 2010 19:40:22 -0700 Subject: mm: further fix swapin race condition Commit 4969c1192d15 ("mm: fix swapin race condition") is now agreed to be incomplete. There's a race, not very much less likely than the original race envisaged, in which it is further necessary to check that the swapcache page's swap has not changed. Here's the reasoning: cast in terms of reuse_swap_page(), but probably could be reformulated to rely on try_to_free_swap() instead, or on swapoff+swapon. A, faults into do_swap_page(): does page1 = lookup_swap_cache(swap1) and comes through the lock_page(page1). B, a racing thread of the same process, faults on the same address: does page1 = lookup_swap_cache(swap1) and now waits in lock_page(page1), but for whatever reason is unlucky not to get the lock any time soon. A carries on through do_swap_page(), a write fault, but cannot reuse the swap page1 (another reference to swap1). Unlocks the page1 (but B doesn't get it yet), does COW in do_wp_page(), page2 now in that pte. C, perhaps the parent of A+B, comes in and write faults the same swap page1 into its mm, reuse_swap_page() succeeds this time, swap1 is freed. kswapd comes in after some time (B still unlucky) and swaps out some pages from A+B and C: it allocates the original swap1 to page2 in A+B, and some other swap2 to the original page1 now in C. But does not immediately free page1 (actually it couldn't: B holds a reference), leaving it in swap cache for now. B at last gets the lock on page1, hooray! Is PageSwapCache(page1)? Yes. Is pte_same(*page_table, orig_pte)? Yes, because page2 has now been given the swap1 which page1 used to have. So B proceeds to insert page1 into A+B's page_table, though its content now belongs to C, quite different from what A wrote there. B ought to have checked that page1's swap was still swap1. Signed-off-by: Hugh Dickins Reviewed-by: Rik van Riel Cc: stable@kernel.org Signed-off-by: Linus Torvalds diff --git a/mm/memory.c b/mm/memory.c index 71b161b..0e18b4d 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -2680,10 +2680,12 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, delayacct_clear_flag(DELAYACCT_PF_SWAPIN); /* - * Make sure try_to_free_swap didn't release the swapcache - * from under us. The page pin isn't enough to prevent that. + * Make sure try_to_free_swap or reuse_swap_page or swapoff did not + * release the swapcache from under us. The page pin, and pte_same + * test below, are not enough to exclude that. Even if it is still + * swapcache, we need to check that the page's swap has not changed. */ - if (unlikely(!PageSwapCache(page))) + if (unlikely(!PageSwapCache(page) || page_private(page) != entry.val)) goto out_page; if (ksm_might_need_to_copy(page, vma, address)) { -- cgit v0.10.2 From 20cd514d0f3d288d968217028ca67b70e707d896 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2010 15:13:04 +0100 Subject: frv: restart_block.fn needs to be reset on sigreturn Reset restart_block.fn on executing a sigreturn such that any currently pending system call restarts will be forced to return -EINTR. Signed-off-by: Al Viro Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c index 0974c0e..7fc2961 100644 --- a/arch/frv/kernel/signal.c +++ b/arch/frv/kernel/signal.c @@ -121,6 +121,9 @@ static int restore_sigcontext(struct sigcontext __user *sc, int *_gr8) struct user_context *user = current->thread.user; unsigned long tbr, psr; + /* Always make any pending restarted system calls return -EINTR */ + current_thread_info()->restart_block.fn = do_no_restart_syscall; + tbr = user->i.tbr; psr = user->i.psr; if (copy_from_user(user, &sc->sc_context, sizeof(sc->sc_context))) -- cgit v0.10.2 From 5f4ad04a1e805d14de080ff9d5384b4d20518a9a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2010 15:13:09 +0100 Subject: frv: fix address verification holes in setup_frame/setup_rt_frame a) sa_handler might be maliciously set to point to kernel memory; blindly dereferencing it in FDPIC case is a Bad Idea(tm). b) I'm not sure you need that set_fs(USER_DS) there at all, but if you do, you'd better do it *before* checking the frame you've decided to use with access_ok(), lest sigaltstack() becomes a convenient roothole. Signed-off-by: Al Viro Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c index 7fc2961..5fb2d06 100644 --- a/arch/frv/kernel/signal.c +++ b/arch/frv/kernel/signal.c @@ -253,6 +253,8 @@ static int setup_frame(int sig, struct k_sigaction *ka, sigset_t *set) struct sigframe __user *frame; int rsig; + set_fs(USER_DS); + frame = get_sigframe(ka, sizeof(*frame)); if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) @@ -296,22 +298,23 @@ static int setup_frame(int sig, struct k_sigaction *ka, sigset_t *set) (unsigned long) (frame->retcode + 2)); } - /* set up registers for signal handler */ - __frame->sp = (unsigned long) frame; - __frame->lr = (unsigned long) &frame->retcode; - __frame->gr8 = sig; - + /* Set up registers for the signal handler */ if (current->personality & FDPIC_FUNCPTRS) { struct fdpic_func_descriptor __user *funcptr = (struct fdpic_func_descriptor __user *) ka->sa.sa_handler; - __get_user(__frame->pc, &funcptr->text); - __get_user(__frame->gr15, &funcptr->GOT); + struct fdpic_func_descriptor desc; + if (copy_from_user(&desc, funcptr, sizeof(desc))) + goto give_sigsegv; + __frame->pc = desc.text; + __frame->gr15 = desc.GOT; } else { __frame->pc = (unsigned long) ka->sa.sa_handler; __frame->gr15 = 0; } - set_fs(USER_DS); + __frame->sp = (unsigned long) frame; + __frame->lr = (unsigned long) &frame->retcode; + __frame->gr8 = sig; /* the tracer may want to single-step inside the handler */ if (test_thread_flag(TIF_SINGLESTEP)) @@ -341,6 +344,8 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, struct rt_sigframe __user *frame; int rsig; + set_fs(USER_DS); + frame = get_sigframe(ka, sizeof(*frame)); if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) @@ -395,22 +400,23 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, } /* Set up registers for signal handler */ - __frame->sp = (unsigned long) frame; - __frame->lr = (unsigned long) &frame->retcode; - __frame->gr8 = sig; - __frame->gr9 = (unsigned long) &frame->info; - if (current->personality & FDPIC_FUNCPTRS) { struct fdpic_func_descriptor __user *funcptr = (struct fdpic_func_descriptor __user *) ka->sa.sa_handler; - __get_user(__frame->pc, &funcptr->text); - __get_user(__frame->gr15, &funcptr->GOT); + struct fdpic_func_descriptor desc; + if (copy_from_user(&desc, funcptr, sizeof(desc))) + goto give_sigsegv; + __frame->pc = desc.text; + __frame->gr15 = desc.GOT; } else { __frame->pc = (unsigned long) ka->sa.sa_handler; __frame->gr15 = 0; } - set_fs(USER_DS); + __frame->sp = (unsigned long) frame; + __frame->lr = (unsigned long) &frame->retcode; + __frame->gr8 = sig; + __frame->gr9 = (unsigned long) &frame->info; /* the tracer may want to single-step inside the handler */ if (test_thread_flag(TIF_SINGLESTEP)) -- cgit v0.10.2 From ad0acab4557a91a60251b3ffe3478fad572d6f61 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2010 15:13:14 +0100 Subject: frv: avoid infinite loop of SIGSEGV delivery Use force_sigsegv() rather than force_sig(SIGSEGV, ...) as the former resets the SEGV handler pointer which will kill the process, rather than leaving it open to an infinite loop if the SEGV handler itself caused a SEGV signal. Signed-off-by: Al Viro Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c index 5fb2d06..48203c6 100644 --- a/arch/frv/kernel/signal.c +++ b/arch/frv/kernel/signal.c @@ -329,7 +329,7 @@ static int setup_frame(int sig, struct k_sigaction *ka, sigset_t *set) return 0; give_sigsegv: - force_sig(SIGSEGV, current); + force_sigsegv(sig, current); return -EFAULT; } /* end setup_frame() */ @@ -431,7 +431,7 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, return 0; give_sigsegv: - force_sig(SIGSEGV, current); + force_sigsegv(sig, current); return -EFAULT; } /* end setup_rt_frame() */ -- cgit v0.10.2 From 44c7afffa429a7cdcd1755019ab76566ff41e66d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2010 15:13:19 +0100 Subject: frv: handling of restart into restart_syscall is fscked do_signal() should place the syscall number in gr7, not gr8 when handling ERESTART_WOULDBLOCK. Signed-off-by: Al Viro Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c index 48203c6..bd13b57 100644 --- a/arch/frv/kernel/signal.c +++ b/arch/frv/kernel/signal.c @@ -547,7 +547,7 @@ no_signal: break; case -ERESTART_RESTARTBLOCK: - __frame->gr8 = __NR_restart_syscall; + __frame->gr7 = __NR_restart_syscall; __frame->pc -= 4; break; } -- cgit v0.10.2 From ed1cde68365c1251b920900c51e73675ff38dc66 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2010 15:13:25 +0100 Subject: frv: double syscall restarts, syscall restart in sigreturn() We need to make sure that only the first do_signal() to be handled on the way out syscall will bother with syscall restarts; additionally, the check on the "signal has user handler" path had been wrong - compare with restart prevention in sigreturn()... Signed-off-by: Al Viro Signed-off-by: David Howells Signed-off-by: Linus Torvalds diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c index bd13b57..bab0129 100644 --- a/arch/frv/kernel/signal.c +++ b/arch/frv/kernel/signal.c @@ -446,7 +446,7 @@ static int handle_signal(unsigned long sig, siginfo_t *info, int ret; /* Are we from a system call? */ - if (in_syscall(__frame)) { + if (__frame->syscallno != -1) { /* If so, check system call restarting.. */ switch (__frame->gr8) { case -ERESTART_RESTARTBLOCK: @@ -465,6 +465,7 @@ static int handle_signal(unsigned long sig, siginfo_t *info, __frame->gr8 = __frame->orig_gr8; __frame->pc -= 4; } + __frame->syscallno = -1; } /* Set up the stack frame */ @@ -551,6 +552,7 @@ no_signal: __frame->pc -= 4; break; } + __frame->syscallno = -1; } /* if there's no signal to deliver, we just put the saved sigmask -- cgit v0.10.2 From af6261031317f646d22f994c0b467521e47aa49f Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 20 Sep 2010 10:31:40 +0100 Subject: drm/i915: Hold a reference to the object whilst unbinding the eviction list During heavy aperture thrashing we may be forced to wait upon several active objects during eviction. The active list may be the last reference to these objects and so the action of waiting upon one of them may cause another to be freed (and itself unbound). To prevent the object disappearing underneath us, we need to acquire and hold a reference whilst unbinding. This should fix the reported page refcount OOPS: kernel BUG at drivers/gpu/drm/i915/i915_gem.c:1444! ... RIP: 0010:[] [] i915_gem_object_put_pages+0x25/0xf5 [i915] Call Trace: [] i915_gem_object_unbind+0xc5/0x1a7 [i915] [] i915_gem_evict_something+0x3bd/0x409 [i915] [] ? drm_gem_object_lookup+0x27/0x57 [drm] [] i915_gem_object_bind_to_gtt+0x1d3/0x279 [i915] [] i915_gem_object_pin+0xa3/0x146 [i915] [] ? drm_gem_object_lookup+0x4c/0x57 [drm] [] i915_gem_do_execbuffer+0x50d/0xe32 [i915] Reported-by: Shawn Starr Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=18902 Signed-off-by: Chris Wilson diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index 72cae3c..e85246e 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -79,6 +79,7 @@ mark_free(struct drm_i915_gem_object *obj_priv, struct list_head *unwind) { list_add(&obj_priv->evict_list, unwind); + drm_gem_object_reference(&obj_priv->base); return drm_mm_scan_add_block(obj_priv->gtt_space); } @@ -165,6 +166,7 @@ i915_gem_evict_something(struct drm_device *dev, int min_size, unsigned alignmen list_for_each_entry(obj_priv, &unwind_list, evict_list) { ret = drm_mm_scan_remove_block(obj_priv->gtt_space); BUG_ON(ret); + drm_gem_object_unreference(&obj_priv->base); } /* We expect the caller to unpin, evict all and try again, or give up. @@ -181,18 +183,21 @@ found: * scanning, therefore store to be evicted objects on a * temporary list. */ list_move(&obj_priv->evict_list, &eviction_list); - } + } else + drm_gem_object_unreference(&obj_priv->base); } /* Unbinding will emit any required flushes */ list_for_each_entry_safe(obj_priv, tmp_obj_priv, &eviction_list, evict_list) { #if WATCH_LRU - DRM_INFO("%s: evicting %p\n", __func__, obj); + DRM_INFO("%s: evicting %p\n", __func__, &obj_priv->base); #endif ret = i915_gem_object_unbind(&obj_priv->base); if (ret) return ret; + + drm_gem_object_unreference(&obj_priv->base); } /* The just created free hole should be on the top of the free stack -- cgit v0.10.2 From 024cfa5943a7e89565c60b612d698c2bfb3da66a Mon Sep 17 00:00:00 2001 From: Mathias Nyman Date: Mon, 6 Sep 2010 13:52:01 +0300 Subject: usb: musb_debugfs: don't use the struct file private_data field with seq_files seq_files use the private_data field of a file struct for storing a seq_file structure, data should be stored in seq_file's own private field (e.g. file->private_data->private) Otherwise seq_release() will free the private data when the file is closed. Signed-off-by: Mathias Nyman Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/musb/musb_debugfs.c b/drivers/usb/musb/musb_debugfs.c index c79a5e3..9e8639d 100644 --- a/drivers/usb/musb/musb_debugfs.c +++ b/drivers/usb/musb/musb_debugfs.c @@ -195,15 +195,14 @@ static const struct file_operations musb_regdump_fops = { static int musb_test_mode_open(struct inode *inode, struct file *file) { - file->private_data = inode->i_private; - return single_open(file, musb_test_mode_show, inode->i_private); } static ssize_t musb_test_mode_write(struct file *file, const char __user *ubuf, size_t count, loff_t *ppos) { - struct musb *musb = file->private_data; + struct seq_file *s = file->private_data; + struct musb *musb = s->private; u8 test = 0; char buf[18]; -- cgit v0.10.2 From fc9282506114d4be188a464af2d373db31dd781c Mon Sep 17 00:00:00 2001 From: Alek Du Date: Mon, 6 Sep 2010 14:50:57 +0100 Subject: USB: EHCI: Disable langwell/penwell LPM capability We have to do so due to HW limitation. Signed-off-by: Alek Du Signed-off-by: Alan Cox Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/ehci-pci.c b/drivers/usb/host/ehci-pci.c index 58b72d7..a1e8d27 100644 --- a/drivers/usb/host/ehci-pci.c +++ b/drivers/usb/host/ehci-pci.c @@ -119,6 +119,11 @@ static int ehci_pci_setup(struct usb_hcd *hcd) ehci->broken_periodic = 1; ehci_info(ehci, "using broken periodic workaround\n"); } + if (pdev->device == 0x0806 || pdev->device == 0x0811 + || pdev->device == 0x0829) { + ehci_info(ehci, "disable lpm for langwell/penwell\n"); + ehci->has_lpm = 0; + } break; case PCI_VENDOR_ID_TDI: if (pdev->device == PCI_DEVICE_ID_TDI_EHCI) { -- cgit v0.10.2 From fc8f2a7608d855b911e35a33e771e6358c705c43 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Mon, 6 Sep 2010 23:27:09 +0800 Subject: USB: otg: twl4030: fix phy initialization(v1) Commit 461c317705eca5cac09a360f488715927fd0a927(into 2.6.36-v3) is put forward to power down phy if no usb cable is connected, but does introduce the two issues below: 1), phy is not into work state if usb cable is connected with PC during poweron, so musb device mode is not usable in such case, follows the reasons: -twl4030_phy_resume is not called, so regulators are not enabled i2c access are not enabled usb mode not configurated 2), The kernel warings[1] of regulators 'unbalanced disables' is caused if poweron without usb cable connected with PC or b-device. This patch fixes the two issues above: -power down phy only if no usb cable is connected with PC and b-device -do phy initialization(via __twl4030_phy_resume) if usb cable is connected with PC(vbus event) or another b-device(ID event) in twl4030_usb_probe. This patch also doesn't put VUSB3V1 LDO into active mode in twl4030_usb_ldo_init until VBUS/ID change detected, so we can save more power consumption than before. This patch is verified OK on Beagle board either connected with usb cable or not when poweron. [1]. warnings of 'unbalanced disables' of regulators. [root@OMAP3EVM /]# dmesg ------------[ cut here ]------------ WARNING: at drivers/regulator/core.c:1357 _regulator_disable+0x38/0x128() unbalanced disables for VUSB1V8 Modules linked in: Backtrace: [] (dump_backtrace+0x0/0x110) from [] (dump_stack+0x18/0x1c) r7:c78179d8 r6:c01ed6b8 r5:c0410822 r4:0000054d [] (dump_stack+0x0/0x1c) from [] (warn_slowpath_common+0x54/0x6c) [] (warn_slowpath_common+0x0/0x6c) from [] (warn_slowpath_fmt+0x38/0x40) r9:00000000 r8:00000000 r7:c78e6608 r6:00000000 r5:fffffffb r4:c78e6c00 [] (warn_slowpath_fmt+0x0/0x40) from [] (_regulator_disable+0x38/0x128) r3:c0410e53 r2:c0410ad5 [] (_regulator_disable+0x0/0x128) from [] (regulator_disable+0x24/0x38) r7:c78e6608 r6:00000000 r5:c78e6c40 r4:c78e6c00 [] (regulator_disable+0x0/0x38) from [] (twl4030_phy_power+0x15c/0x17c) r5:c78595c0 r4:00000000 [] (twl4030_phy_power+0x0/0x17c) from [] (twl4030_phy_suspend+0x20/0x2c) r6:00000000 r5:c78595c0 r4:c78595c0 [] (twl4030_phy_suspend+0x0/0x2c) from [] (twl4030_usb_irq+0x11c/0x16c) r5:c78595c0 r4:00000040 [] (twl4030_usb_irq+0x0/0x16c) from [] (twl4030_usb_probe+0x2c4/0x32c) r6:00000000 r5:00000000 r4:c78595c0 [] (twl4030_usb_probe+0x0/0x32c) from [] (platform_drv_probe+0x20/0x24) r7:00000000 r6:c047d49c r5:c78e6608 r4:c047d49c [] (platform_drv_probe+0x0/0x24) from [] (driver_probe_device+0xd0/0x190) [] (driver_probe_device+0x0/0x190) from [] (__device_attach+0x44/0x48) r7:00000000 r6:c78e6608 r5:c78e6608 r4:c047d49c [] (__device_attach+0x0/0x48) from [] (bus_for_each_drv+0x50/0x90) r5:c0214390 r4:00000000 [] (bus_for_each_drv+0x0/0x90) from [] (device_attach+0x70/0x94) r6:c78e663c r5:c78e6608 r4:c78e6608 [] (device_attach+0x0/0x94) from [] (bus_probe_device+0x2c/0x48) r7:00000000 r6:00000002 r5:c78e6608 r4:c78e6600 [] (bus_probe_device+0x0/0x48) from [] (device_add+0x340/0x4b4) [] (device_add+0x0/0x4b4) from [] (platform_device_add+0x110/0x16c) [] (platform_device_add+0x0/0x16c) from [] (add_numbered_child+0xd8/0x118) r7:00000000 r6:c045f15c r5:c78e6600 r4:00000000 [] (add_numbered_child+0x0/0x118) from [] (twl_probe+0x3a4/0x72c) [] (twl_probe+0x0/0x72c) from [] (i2c_device_probe+0x7c/0xa4) [] (i2c_device_probe+0x0/0xa4) from [] (driver_probe_device+0xd0/0x190) r5:c7856e20 r4:c047c860 [] (driver_probe_device+0x0/0x190) from [] (__device_attach+0x44/0x48) r7:c7856e04 r6:c7856e20 r5:c7856e20 r4:c047c860 [] (__device_attach+0x0/0x48) from [] (bus_for_each_drv+0x50/0x90) r5:c0214390 r4:00000000 [] (bus_for_each_drv+0x0/0x90) from [] (device_attach+0x70/0x94) r6:c7856e54 r5:c7856e20 r4:c7856e20 [] (device_attach+0x0/0x94) from [] (bus_probe_device+0x2c/0x48) r7:c7856e04 r6:c78fd048 r5:c7856e20 r4:c7856e20 [] (bus_probe_device+0x0/0x48) from [] (device_add+0x340/0x4b4) [] (device_add+0x0/0x4b4) from [] (device_register+0x1c/0x20) [] (device_register+0x0/0x20) from [] (i2c_new_device+0xec/0x150) r5:c7856e00 r4:c7856e20 [] (i2c_new_device+0x0/0x150) from [] (i2c_register_adapter+0xa0/0x1c4) r7:00000000 r6:c78fd078 r5:c78fd048 r4:c781d5c0 [] (i2c_register_adapter+0x0/0x1c4) from [] (i2c_add_numbered_adapter+0x9c/0xb4) r7:00000a28 r6:c04600a8 r5:c78fd048 r4:00000000 [] (i2c_add_numbered_adapter+0x0/0xb4) from [] (omap_i2c_probe+0x324/0x3e8) r5:00000000 r4:c78fd000 [] (omap_i2c_probe+0x0/0x3e8) from [] (platform_drv_probe+0x20/0x24) [] (platform_drv_probe+0x0/0x24) from [] (driver_probe_device+0xd0/0x190) [] (driver_probe_device+0x0/0x190) from [] (__driver_attach+0x68/0x8c) r7:c78b2140 r6:c047e214 r5:c04600e4 r4:c04600b0 [] (__driver_attach+0x0/0x8c) from [] (bus_for_each_dev+0x50/0x84) r7:c78b2140 r6:c047e214 r5:c0214304 r4:00000000 [] (bus_for_each_dev+0x0/0x84) from [] (driver_attach+0x20/0x28) r6:c047e214 r5:c047e214 r4:c00270d0 [] (driver_attach+0x0/0x28) from [] (bus_add_driver+0xa8/0x228) [] (bus_add_driver+0x0/0x228) from [] (driver_register+0xb0/0x13c) [] (driver_register+0x0/0x13c) from [] (platform_driver_register+0x4c/0x60) r9:00000000 r8:c001f688 r7:00000013 r6:c005b6fc r5:c00083dc r4:c00270d0 [] (platform_driver_register+0x0/0x60) from [] (omap_i2c_init_driver+0x14/0x1c) [] (omap_i2c_init_driver+0x0/0x1c) from [] (do_one_initcall+0xd0/0x1a4) [] (do_one_initcall+0x0/0x1a4) from [] (kernel_init+0x9c/0x154) [] (kernel_init+0x0/0x154) from [] (do_exit+0x0/0x688) r5:c00083dc r4:00000000 ---[ end trace 1b75b31a2719ed1d ]--- Signed-off-by: Ming Lei Cc: David Brownell Cc: Felipe Balbi Cc: Anand Gadiyar Cc: Mike Frysinger Cc: Sergei Shtylyov Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/otg/twl4030-usb.c b/drivers/usb/otg/twl4030-usb.c index 05aaac1..0bc9769 100644 --- a/drivers/usb/otg/twl4030-usb.c +++ b/drivers/usb/otg/twl4030-usb.c @@ -347,11 +347,20 @@ static void twl4030_i2c_access(struct twl4030_usb *twl, int on) } } -static void twl4030_phy_power(struct twl4030_usb *twl, int on) +static void __twl4030_phy_power(struct twl4030_usb *twl, int on) { - u8 pwr; + u8 pwr = twl4030_usb_read(twl, PHY_PWR_CTRL); + + if (on) + pwr &= ~PHY_PWR_PHYPWD; + else + pwr |= PHY_PWR_PHYPWD; - pwr = twl4030_usb_read(twl, PHY_PWR_CTRL); + WARN_ON(twl4030_usb_write_verify(twl, PHY_PWR_CTRL, pwr) < 0); +} + +static void twl4030_phy_power(struct twl4030_usb *twl, int on) +{ if (on) { regulator_enable(twl->usb3v1); regulator_enable(twl->usb1v8); @@ -365,15 +374,13 @@ static void twl4030_phy_power(struct twl4030_usb *twl, int on) twl_i2c_write_u8(TWL4030_MODULE_PM_RECEIVER, 0, VUSB_DEDICATED2); regulator_enable(twl->usb1v5); - pwr &= ~PHY_PWR_PHYPWD; - WARN_ON(twl4030_usb_write_verify(twl, PHY_PWR_CTRL, pwr) < 0); + __twl4030_phy_power(twl, 1); twl4030_usb_write(twl, PHY_CLK_CTRL, twl4030_usb_read(twl, PHY_CLK_CTRL) | (PHY_CLK_CTRL_CLOCKGATING_EN | PHY_CLK_CTRL_CLK32K_EN)); - } else { - pwr |= PHY_PWR_PHYPWD; - WARN_ON(twl4030_usb_write_verify(twl, PHY_PWR_CTRL, pwr) < 0); + } else { + __twl4030_phy_power(twl, 0); regulator_disable(twl->usb1v5); regulator_disable(twl->usb1v8); regulator_disable(twl->usb3v1); @@ -387,19 +394,25 @@ static void twl4030_phy_suspend(struct twl4030_usb *twl, int controller_off) twl4030_phy_power(twl, 0); twl->asleep = 1; + dev_dbg(twl->dev, "%s\n", __func__); } -static void twl4030_phy_resume(struct twl4030_usb *twl) +static void __twl4030_phy_resume(struct twl4030_usb *twl) { - if (!twl->asleep) - return; - twl4030_phy_power(twl, 1); twl4030_i2c_access(twl, 1); twl4030_usb_set_mode(twl, twl->usb_mode); if (twl->usb_mode == T2_USB_MODE_ULPI) twl4030_i2c_access(twl, 0); +} + +static void twl4030_phy_resume(struct twl4030_usb *twl) +{ + if (!twl->asleep) + return; + __twl4030_phy_resume(twl); twl->asleep = 0; + dev_dbg(twl->dev, "%s\n", __func__); } static int twl4030_usb_ldo_init(struct twl4030_usb *twl) @@ -408,8 +421,8 @@ static int twl4030_usb_ldo_init(struct twl4030_usb *twl) twl_i2c_write_u8(TWL4030_MODULE_PM_MASTER, 0xC0, PROTECT_KEY); twl_i2c_write_u8(TWL4030_MODULE_PM_MASTER, 0x0C, PROTECT_KEY); - /* put VUSB3V1 LDO in active state */ - twl_i2c_write_u8(TWL4030_MODULE_PM_RECEIVER, 0, VUSB_DEDICATED2); + /* Keep VUSB3V1 LDO in sleep state until VBUS/ID change detected*/ + /*twl_i2c_write_u8(TWL4030_MODULE_PM_RECEIVER, 0, VUSB_DEDICATED2);*/ /* input to VUSB3V1 LDO is from VBAT, not VBUS */ twl_i2c_write_u8(TWL4030_MODULE_PM_RECEIVER, 0x14, VUSB_DEDICATED1); @@ -502,6 +515,26 @@ static irqreturn_t twl4030_usb_irq(int irq, void *_twl) return IRQ_HANDLED; } +static void twl4030_usb_phy_init(struct twl4030_usb *twl) +{ + int status; + + status = twl4030_usb_linkstat(twl); + if (status >= 0) { + if (status == USB_EVENT_NONE) { + __twl4030_phy_power(twl, 0); + twl->asleep = 1; + } else { + __twl4030_phy_resume(twl); + twl->asleep = 0; + } + + blocking_notifier_call_chain(&twl->otg.notifier, status, + twl->otg.gadget); + } + sysfs_notify(&twl->dev->kobj, NULL, "vbus"); +} + static int twl4030_set_suspend(struct otg_transceiver *x, int suspend) { struct twl4030_usb *twl = xceiv_to_twl(x); @@ -550,7 +583,6 @@ static int __devinit twl4030_usb_probe(struct platform_device *pdev) struct twl4030_usb_data *pdata = pdev->dev.platform_data; struct twl4030_usb *twl; int status, err; - u8 pwr; if (!pdata) { dev_dbg(&pdev->dev, "platform_data not available\n"); @@ -569,10 +601,7 @@ static int __devinit twl4030_usb_probe(struct platform_device *pdev) twl->otg.set_peripheral = twl4030_set_peripheral; twl->otg.set_suspend = twl4030_set_suspend; twl->usb_mode = pdata->usb_mode; - - pwr = twl4030_usb_read(twl, PHY_PWR_CTRL); - - twl->asleep = (pwr & PHY_PWR_PHYPWD); + twl->asleep = 1; /* init spinlock for workqueue */ spin_lock_init(&twl->lock); @@ -610,15 +639,10 @@ static int __devinit twl4030_usb_probe(struct platform_device *pdev) return status; } - /* The IRQ handler just handles changes from the previous states - * of the ID and VBUS pins ... in probe() we must initialize that - * previous state. The easy way: fake an IRQ. - * - * REVISIT: a real IRQ might have happened already, if PREEMPT is - * enabled. Else the IRQ may not yet be configured or enabled, - * because of scheduling delays. + /* Power down phy or make it work according to + * current link state. */ - twl4030_usb_irq(twl->irq, twl); + twl4030_usb_phy_init(twl); dev_info(&pdev->dev, "Initialized TWL4030 USB module\n"); return 0; -- cgit v0.10.2 From a0846f1868b11cd827bdfeaf4527d8b1b1c0b098 Mon Sep 17 00:00:00 2001 From: Dan Rosenberg Date: Wed, 15 Sep 2010 17:44:16 -0400 Subject: USB: serial/mos*: prevent reading uninitialized stack memory The TIOCGICOUNT device ioctl in both mos7720.c and mos7840.c allows unprivileged users to read uninitialized stack memory, because the "reserved" member of the serial_icounter_struct struct declared on the stack is not altered or zeroed before being copied back to the user. This patch takes care of it. Signed-off-by: Dan Rosenberg Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/mos7720.c b/drivers/usb/serial/mos7720.c index 30922a7..aa66581 100644 --- a/drivers/usb/serial/mos7720.c +++ b/drivers/usb/serial/mos7720.c @@ -2024,6 +2024,9 @@ static int mos7720_ioctl(struct tty_struct *tty, struct file *file, case TIOCGICOUNT: cnow = mos7720_port->icount; + + memset(&icount, 0, sizeof(struct serial_icounter_struct)); + icount.cts = cnow.cts; icount.dsr = cnow.dsr; icount.rng = cnow.rng; diff --git a/drivers/usb/serial/mos7840.c b/drivers/usb/serial/mos7840.c index 1c9b6e9..1a42bc2 100644 --- a/drivers/usb/serial/mos7840.c +++ b/drivers/usb/serial/mos7840.c @@ -2285,6 +2285,9 @@ static int mos7840_ioctl(struct tty_struct *tty, struct file *file, case TIOCGICOUNT: cnow = mos7840_port->icount; smp_rmb(); + + memset(&icount, 0, sizeof(struct serial_icounter_struct)); + icount.cts = cnow.cts; icount.dsr = cnow.dsr; icount.rng = cnow.rng; -- cgit v0.10.2 From f299470a15ab3057afbf598cec59246a90ade449 Mon Sep 17 00:00:00 2001 From: Felipe Balbi Date: Thu, 9 Sep 2010 09:04:25 +0300 Subject: USB: musb: MAINTAINERS: Fix my mail address If we don't, contributors to musb and any USB OMAP code will be sending mails to an unexistent inbox. Signed-off-by: Felipe Balbi Signed-off-by: Greg Kroah-Hartman diff --git a/MAINTAINERS b/MAINTAINERS index 411b0d0..50b8148 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3942,7 +3942,7 @@ F: drivers/char/isicom.c F: include/linux/isicom.h MUSB MULTIPOINT HIGH SPEED DUAL-ROLE CONTROLLER -M: Felipe Balbi +M: Felipe Balbi L: linux-usb@vger.kernel.org T: git git://gitorious.org/usb/usb.git S: Maintained @@ -4240,7 +4240,7 @@ S: Maintained F: drivers/char/hw_random/omap-rng.c OMAP USB SUPPORT -M: Felipe Balbi +M: Felipe Balbi M: David Brownell L: linux-usb@vger.kernel.org L: linux-omap@vger.kernel.org -- cgit v0.10.2 From 476f771cb9b6cd4845dcd18f16a2f03a89ee63fc Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Sat, 4 Sep 2010 10:23:23 +0300 Subject: serial: amba-pl010: fix set_ldisc Commit d87d9b7d1 ("tty: serial - fix tty referencing in set_ldisc") changed set_ldisc to take ldisc number as parameter. This patch fixes AMBA PL010 driver according the new prototype. Signed-off-by: Mika Westerberg Cc: Alan Cox Cc: Russell King Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/serial/amba-pl010.c b/drivers/serial/amba-pl010.c index 50441ff..2904aa0 100644 --- a/drivers/serial/amba-pl010.c +++ b/drivers/serial/amba-pl010.c @@ -472,14 +472,9 @@ pl010_set_termios(struct uart_port *port, struct ktermios *termios, spin_unlock_irqrestore(&uap->port.lock, flags); } -static void pl010_set_ldisc(struct uart_port *port) +static void pl010_set_ldisc(struct uart_port *port, int new) { - int line = port->line; - - if (line >= port->state->port.tty->driver->num) - return; - - if (port->state->port.tty->ldisc->ops->num == N_PPS) { + if (new == N_PPS) { port->flags |= UPF_HARDPPS_CD; pl010_enable_ms(port); } else -- cgit v0.10.2 From e3671ac429fe50cf0c1b4f1dc4b7237207f1d956 Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Mon, 6 Sep 2010 13:41:02 +0100 Subject: serial: mfd: fix bug in serial_hsu_remove() Medfield HSU driver deal with 4 pci devices(3 uart ports + 1 dma controller), so in pci remove func, we need handle them differently Signed-off-by: Feng Tang Signed-off-by: Alan Cox Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/serial/mfd.c b/drivers/serial/mfd.c index bc9af50..324c385 100644 --- a/drivers/serial/mfd.c +++ b/drivers/serial/mfd.c @@ -1423,7 +1423,6 @@ static void hsu_global_init(void) } phsu = hsu; - hsu_debugfs_init(hsu); return; @@ -1435,18 +1434,20 @@ err_free_region: static void serial_hsu_remove(struct pci_dev *pdev) { - struct hsu_port *hsu; - int i; + void *priv = pci_get_drvdata(pdev); + struct uart_hsu_port *up; - hsu = pci_get_drvdata(pdev); - if (!hsu) + if (!priv) return; - for (i = 0; i < 3; i++) - uart_remove_one_port(&serial_hsu_reg, &hsu->port[i].port); + /* For port 0/1/2, priv is the address of uart_hsu_port */ + if (pdev->device != 0x081E) { + up = priv; + uart_remove_one_port(&serial_hsu_reg, &up->port); + } pci_set_drvdata(pdev, NULL); - free_irq(hsu->irq, hsu); + free_irq(pdev->irq, priv); pci_disable_device(pdev); } -- cgit v0.10.2 From 350aede603f7db7a9b4c1a340fbe89ccae6523a2 Mon Sep 17 00:00:00 2001 From: Sven Eckelmann Date: Sun, 5 Sep 2010 01:58:18 +0200 Subject: Revert: "Staging: batman-adv: Adding netfilter-bridge hooks" This reverts commit 96d592ed599434d2d5f339a1d282871bc6377d2c. The netfilter hook seems to be misused and may leak skbs in situations when NF_HOOK returns NF_STOLEN. It may not filter everything as expected. Also the ethernet bridge tables are not yet capable to understand batman-adv packet correctly. It was only added for testing purposes and can be removed again. Reported-by: Vasiliy Kulikov Signed-off-by: Sven Eckelmann Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/staging/batman-adv/hard-interface.c b/drivers/staging/batman-adv/hard-interface.c index baa8b05..6e973a7 100644 --- a/drivers/staging/batman-adv/hard-interface.c +++ b/drivers/staging/batman-adv/hard-interface.c @@ -30,7 +30,6 @@ #include "hash.h" #include -#include #define MIN(x, y) ((x) < (y) ? (x) : (y)) @@ -431,11 +430,6 @@ out: return NOTIFY_DONE; } -static int batman_skb_recv_finish(struct sk_buff *skb) -{ - return NF_ACCEPT; -} - /* receive a packet with the batman ethertype coming on a hard * interface */ int batman_skb_recv(struct sk_buff *skb, struct net_device *dev, @@ -456,13 +450,6 @@ int batman_skb_recv(struct sk_buff *skb, struct net_device *dev, if (atomic_read(&module_state) != MODULE_ACTIVE) goto err_free; - /* if netfilter/ebtables wants to block incoming batman - * packets then give them a chance to do so here */ - ret = NF_HOOK(PF_BRIDGE, NF_BR_LOCAL_IN, skb, dev, NULL, - batman_skb_recv_finish); - if (ret != 1) - goto err_out; - /* packet should hold at least type and version */ if (unlikely(skb_headlen(skb) < 2)) goto err_free; diff --git a/drivers/staging/batman-adv/send.c b/drivers/staging/batman-adv/send.c index 055edee..da3c82e4 100644 --- a/drivers/staging/batman-adv/send.c +++ b/drivers/staging/batman-adv/send.c @@ -29,7 +29,6 @@ #include "vis.h" #include "aggregation.h" -#include static void send_outstanding_bcast_packet(struct work_struct *work); @@ -92,12 +91,9 @@ int send_skb_packet(struct sk_buff *skb, /* dev_queue_xmit() returns a negative result on error. However on * congestion and traffic shaping, it drops and returns NET_XMIT_DROP - * (which is > 0). This will not be treated as an error. - * Also, if netfilter/ebtables wants to block outgoing batman - * packets then giving them a chance to do so here */ + * (which is > 0). This will not be treated as an error. */ - return NF_HOOK(PF_BRIDGE, NF_BR_LOCAL_OUT, skb, NULL, skb->dev, - dev_queue_xmit); + return dev_queue_xmit(skb); send_skb_err: kfree_skb(skb); return NET_XMIT_DROP; -- cgit v0.10.2 From dd173abfead903c7df54e977535973f3312cd307 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 6 Sep 2010 14:32:30 +0200 Subject: Staging: vt6655: fix buffer overflow "param->u.wpa_associate.wpa_ie_len" comes from the user. We should check it so that the copy_from_user() doesn't overflow the buffer. Also further down in the function, we assume that if "param->u.wpa_associate.wpa_ie_len" is set then "abyWPAIE[0]" is initialized. To make that work, I changed the test here to say that if "wpa_ie_len" is set then "wpa_ie" has to be a valid pointer or we return -EINVAL. Oddly, we only use the first element of the abyWPAIE[] array. So I suspect there may be some other issues in this function. Signed-off-by: Dan Carpenter Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/staging/vt6655/wpactl.c b/drivers/staging/vt6655/wpactl.c index 0142338b..4bdb836 100644 --- a/drivers/staging/vt6655/wpactl.c +++ b/drivers/staging/vt6655/wpactl.c @@ -766,9 +766,14 @@ static int wpa_set_associate(PSDevice pDevice, DBG_PRT(MSG_LEVEL_DEBUG, KERN_INFO "wpa_ie_len = %d\n", param->u.wpa_associate.wpa_ie_len); - if (param->u.wpa_associate.wpa_ie && - copy_from_user(&abyWPAIE[0], param->u.wpa_associate.wpa_ie, param->u.wpa_associate.wpa_ie_len)) - return -EINVAL; + if (param->u.wpa_associate.wpa_ie_len) { + if (!param->u.wpa_associate.wpa_ie) + return -EINVAL; + if (param->u.wpa_associate.wpa_ie_len > sizeof(abyWPAIE)) + return -EINVAL; + if (copy_from_user(&abyWPAIE[0], param->u.wpa_associate.wpa_ie, param->u.wpa_associate.wpa_ie_len)) + return -EFAULT; + } if (param->u.wpa_associate.mode == 1) pMgmt->eConfigMode = WMAC_CONFIG_IBSS_STA; -- cgit v0.10.2 From b30a3f6257ed2105259b404d419b4964e363928c Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 20 Sep 2010 16:56:53 -0700 Subject: Linux 2.6.36-rc5 diff --git a/Makefile b/Makefile index 92ab33f..3133a57 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 2 PATCHLEVEL = 6 SUBLEVEL = 36 -EXTRAVERSION = -rc4 +EXTRAVERSION = -rc5 NAME = Sheep on Meth # *DOCUMENTATION* -- cgit v0.10.2 From 6df7aadcd9290807c464675098b5dd2dc9da5075 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Thu, 16 Sep 2010 14:43:08 +0530 Subject: virtio: console: Fix poll blocking even though there is data to read I found this while working on a Linux agent for spice, the symptom I was seeing was select blocking on the spice vdagent virtio serial port even though there were messages queued up there. virtio_console's port_fops_poll checks port->inbuf != NULL to determine if read won't block. However if an application reads enough bytes from inbuf through port_fops_read, to empty the current port->inbuf, port->inbuf will be NULL even though there may be buffers left in the virtqueue. This causes poll() to block even though there is data to be read, this patch fixes this by using will_read_block(port) instead of the port->inbuf != NULL check. Signed-off-By: Hans de Goede Signed-off-by: Amit Shah Signed-off-by: Rusty Russell Cc: stable@kernel.org diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index 942a982..2f2e31b 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -642,7 +642,7 @@ static unsigned int port_fops_poll(struct file *filp, poll_table *wait) poll_wait(filp, &port->waitqueue, wait); ret = 0; - if (port->inbuf) + if (!will_read_block(port)) ret |= POLLIN | POLLRDNORM; if (!will_write_block(port)) ret |= POLLOUT; -- cgit v0.10.2 From 65745422a898741ee0e7068ef06624ab06e8aefa Mon Sep 17 00:00:00 2001 From: Amit Shah Date: Tue, 14 Sep 2010 13:26:16 +0530 Subject: virtio: console: Prevent userspace from submitting NULL buffers A userspace could submit a buffer with 0 length to be written to the host. Prevent such a situation. This was not needed previously, but recent changes in the way write() works exposed this condition to trigger a virtqueue event to the host, causing a NULL buffer to be sent across. Signed-off-by: Amit Shah Signed-off-by: Rusty Russell CC: stable@kernel.org diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index 2f2e31b..c810481 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -596,6 +596,10 @@ static ssize_t port_fops_write(struct file *filp, const char __user *ubuf, ssize_t ret; bool nonblock; + /* Userspace could be out to fool us */ + if (!count) + return 0; + port = filp->private_data; nonblock = filp->f_flags & O_NONBLOCK; -- cgit v0.10.2 From 9b6efcd2e2275e13403700913b7a1da92cf11ad2 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Tue, 21 Sep 2010 10:54:01 -0600 Subject: lguest: update comments to reflect LHCALL_LOAD_GDT_ENTRY. We used to have a hypercall which reloaded the entire GDT, then we switched to one which loaded a single entry (to match the IDT code). Some comments were not updated, so fix them. Signed-off-by: Rusty Russell Reported by: Eviatar Khen diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 9257510..9d5f558 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -324,9 +324,8 @@ static void lguest_load_gdt(const struct desc_ptr *desc) } /* - * For a single GDT entry which changes, we do the lazy thing: alter our GDT, - * then tell the Host to reload the entire thing. This operation is so rare - * that this naive implementation is reasonable. + * For a single GDT entry which changes, we simply change our copy and + * then tell the host about it. */ static void lguest_write_gdt_entry(struct desc_struct *dt, int entrynum, const void *desc, int type) @@ -338,9 +337,13 @@ static void lguest_write_gdt_entry(struct desc_struct *dt, int entrynum, } /* - * OK, I lied. There are three "thread local storage" GDT entries which change + * There are three "thread local storage" GDT entries which change * on every context switch (these three entries are how glibc implements - * __thread variables). So we have a hypercall specifically for this case. + * __thread variables). As an optimization, we have a hypercall + * specifically for this case. + * + * Wouldn't it be nicer to have a general LOAD_GDT_ENTRIES hypercall + * which took a range of entries? */ static void lguest_load_tls(struct thread_struct *t, unsigned int cpu) { -- cgit v0.10.2 From 46b30ea9bc3698bc1d1e6fd726c9601d46fa0a91 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 21 Sep 2010 07:57:19 +0200 Subject: percpu: fix pcpu_last_unit_cpu pcpu_first/last_unit_cpu are used to track which cpu has the first and last units assigned. This in turn is used to determine the span of a chunk for man/unmap cache flushes and whether an address belongs to the first chunk or not in per_cpu_ptr_to_phys(). When the number of possible CPUs isn't power of two, a chunk may contain unassigned units towards the end of a chunk. The logic to determine pcpu_last_unit_cpu was incorrect when there was an unused unit at the end of a chunk. It failed to ignore the unused unit and assigned the unused marker NR_CPUS to pcpu_last_unit_cpu. This was discovered through kdump failure which was caused by malfunctioning per_cpu_ptr_to_phys() on a kvm setup with 50 possible CPUs by CAI Qian. Signed-off-by: Tejun Heo Reported-by: CAI Qian Cc: stable@kernel.org diff --git a/mm/percpu.c b/mm/percpu.c index 58c572b..c76ef38 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1401,9 +1401,9 @@ int __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai, if (pcpu_first_unit_cpu == NR_CPUS) pcpu_first_unit_cpu = cpu; + pcpu_last_unit_cpu = cpu; } } - pcpu_last_unit_cpu = cpu; pcpu_nr_units = unit; for_each_possible_cpu(cpu) -- cgit v0.10.2 From 05c5e7698bdc54b3079a3517d86077f49ebcc788 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 20 Sep 2010 23:24:52 -0700 Subject: sparc64: Fix race in signal instruction flushing. If another cpu does a very wide munmap() on the signal frame area, it can tear down the page table hierarchy from underneath us. Borrow an idea from the 64-bit fault path's get_user_insn(), and disable cross call interrupts during the page table traversal to lock them in place while we operate. Reported-by: Al Viro Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/signal32.c b/arch/sparc/kernel/signal32.c index ea22cd3..76b67c4 100644 --- a/arch/sparc/kernel/signal32.c +++ b/arch/sparc/kernel/signal32.c @@ -453,6 +453,64 @@ static int save_fpu_state32(struct pt_regs *regs, __siginfo_fpu_t __user *fpu) return err; } +/* The I-cache flush instruction only works in the primary ASI, which + * right now is the nucleus, aka. kernel space. + * + * Therefore we have to kick the instructions out using the kernel + * side linear mapping of the physical address backing the user + * instructions. + */ +static void flush_signal_insns(unsigned long address) +{ + unsigned long pstate, paddr; + pte_t *ptep, pte; + pgd_t *pgdp; + pud_t *pudp; + pmd_t *pmdp; + + /* Commit all stores of the instructions we are about to flush. */ + wmb(); + + /* Disable cross-call reception. In this way even a very wide + * munmap() on another cpu can't tear down the page table + * hierarchy from underneath us, since that can't complete + * until the IPI tlb flush returns. + */ + + __asm__ __volatile__("rdpr %%pstate, %0" : "=r" (pstate)); + __asm__ __volatile__("wrpr %0, %1, %%pstate" + : : "r" (pstate), "i" (PSTATE_IE)); + + pgdp = pgd_offset(current->mm, address); + if (pgd_none(*pgdp)) + goto out_irqs_on; + pudp = pud_offset(pgdp, address); + if (pud_none(*pudp)) + goto out_irqs_on; + pmdp = pmd_offset(pudp, address); + if (pmd_none(*pmdp)) + goto out_irqs_on; + + ptep = pte_offset_map(pmdp, address); + pte = *ptep; + if (!pte_present(pte)) + goto out_unmap; + + paddr = (unsigned long) page_address(pte_page(pte)); + + __asm__ __volatile__("flush %0 + %1" + : /* no outputs */ + : "r" (paddr), + "r" (address & (PAGE_SIZE - 1)) + : "memory"); + +out_unmap: + pte_unmap(ptep); +out_irqs_on: + __asm__ __volatile__("wrpr %0, 0x0, %%pstate" : : "r" (pstate)); + +} + static void setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, int signo, sigset_t *oldset) { @@ -547,13 +605,7 @@ static void setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, if (ka->ka_restorer) { regs->u_regs[UREG_I7] = (unsigned long)ka->ka_restorer; } else { - /* Flush instruction space. */ unsigned long address = ((unsigned long)&(sf->insns[0])); - pgd_t *pgdp = pgd_offset(current->mm, address); - pud_t *pudp = pud_offset(pgdp, address); - pmd_t *pmdp = pmd_offset(pudp, address); - pte_t *ptep; - pte_t pte; regs->u_regs[UREG_I7] = (unsigned long) (&(sf->insns[0]) - 2); @@ -562,22 +614,7 @@ static void setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, if (err) goto sigsegv; - preempt_disable(); - ptep = pte_offset_map(pmdp, address); - pte = *ptep; - if (pte_present(pte)) { - unsigned long page = (unsigned long) - page_address(pte_page(pte)); - - wmb(); - __asm__ __volatile__("flush %0 + %1" - : /* no outputs */ - : "r" (page), - "r" (address & (PAGE_SIZE - 1)) - : "memory"); - } - pte_unmap(ptep); - preempt_enable(); + flush_signal_insns(address); } return; @@ -687,12 +724,7 @@ static void setup_rt_frame32(struct k_sigaction *ka, struct pt_regs *regs, if (ka->ka_restorer) regs->u_regs[UREG_I7] = (unsigned long)ka->ka_restorer; else { - /* Flush instruction space. */ unsigned long address = ((unsigned long)&(sf->insns[0])); - pgd_t *pgdp = pgd_offset(current->mm, address); - pud_t *pudp = pud_offset(pgdp, address); - pmd_t *pmdp = pmd_offset(pudp, address); - pte_t *ptep; regs->u_regs[UREG_I7] = (unsigned long) (&(sf->insns[0]) - 2); @@ -704,21 +736,7 @@ static void setup_rt_frame32(struct k_sigaction *ka, struct pt_regs *regs, if (err) goto sigsegv; - preempt_disable(); - ptep = pte_offset_map(pmdp, address); - if (pte_present(*ptep)) { - unsigned long page = (unsigned long) - page_address(pte_page(*ptep)); - - wmb(); - __asm__ __volatile__("flush %0 + %1" - : /* no outputs */ - : "r" (page), - "r" (address & (PAGE_SIZE - 1)) - : "memory"); - } - pte_unmap(ptep); - preempt_enable(); + flush_signal_insns(address); } return; -- cgit v0.10.2 From 0f9f1ee9d1412d45a22bfd69dfd4d4324b506e9e Mon Sep 17 00:00:00 2001 From: Luke Yelavich Date: Tue, 21 Sep 2010 17:05:46 +1000 Subject: ALSA: hda - Add Dell Latitude E6400 model quirk BugLink: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/643891 Set the Dell Latitude E6400 (1028:0233) SSID to use AD1984_DELL_DESKTOP Cc: stable@kernel.org Signed-off-by: Luke Yelavich Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_analog.c b/sound/pci/hda/patch_analog.c index b697fd2..10bbbaf 100644 --- a/sound/pci/hda/patch_analog.c +++ b/sound/pci/hda/patch_analog.c @@ -3641,6 +3641,7 @@ static struct snd_pci_quirk ad1984_cfg_tbl[] = { /* Lenovo Thinkpad T61/X61 */ SND_PCI_QUIRK_VENDOR(0x17aa, "Lenovo Thinkpad", AD1984_THINKPAD), SND_PCI_QUIRK(0x1028, 0x0214, "Dell T3400", AD1984_DELL_DESKTOP), + SND_PCI_QUIRK(0x1028, 0x0233, "Dell Latitude E6400", AD1984_DELL_DESKTOP), {} }; -- cgit v0.10.2 From b0722cb1ac84863f57471d2b254457c100319300 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 13 Sep 2010 14:09:33 +0200 Subject: cciss: freeing uninitialized data on error path The "h->scatter_list" is allocated inside a for loop. If any of those allocations fail, then the rest of the list is uninitialized data. When we free it we should start from the top and free backwards so that we don't call kfree() on uninitialized pointers. Also if the allocation for "h->scatter_list" fails then we would get an Oops here. I should have noticed this when I send: 4ee69851c "cciss: handle allocation failure." but I didn't. Sorry about that. Signed-off-by: Dan Carpenter Signed-off-by: Jens Axboe diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 6124c2f..5e4fadc 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -4792,7 +4792,7 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, clean4: kfree(h->cmd_pool_bits); /* Free up sg elements */ - for (k = 0; k < h->nr_cmds; k++) + for (k-- ; k >= 0; k--) kfree(h->scatter_list[k]); kfree(h->scatter_list); cciss_free_sg_chain_blocks(h->cmd_sg_list, h->nr_cmds); -- cgit v0.10.2 From a45dc2d2b8d1afa57c91dcfac224e50ffcd3f805 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Mon, 13 Sep 2010 21:32:19 +0200 Subject: block: fix blk_rq_map_kern bio direction flag This bug was introduced in 7b6d91daee5cac6402186ff224c3af39d79f4a0e "block: unify flags for struct bio and struct request" Cc: Boaz Harrosh Signed-off-by: Benny Halevy Signed-off-by: Jens Axboe diff --git a/block/blk-map.c b/block/blk-map.c index c65d759..ade0a08 100644 --- a/block/blk-map.c +++ b/block/blk-map.c @@ -307,7 +307,7 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf, return PTR_ERR(bio); if (rq_data_dir(rq) == WRITE) - bio->bi_rw |= (1 << REQ_WRITE); + bio->bi_rw |= REQ_WRITE; if (do_copy) rq->cmd_flags |= REQ_COPY_USER; -- cgit v0.10.2 From 180be2a0420a0d680285e9352c0db85016b9ed2a Mon Sep 17 00:00:00 2001 From: Vivek Goyal Date: Tue, 14 Sep 2010 08:47:11 +0200 Subject: cfq-iosched: fix a kernel OOPs when usb key is inserted Mike reported a kernel crash when a usb key hotplug is performed while all kernel thrads are not in a root cgroup and are running in one of the child cgroups of blkio controller. BUG: unable to handle kernel NULL pointer dereference at 0000002c IP: [] cfq_get_queue+0x232/0x412 *pde = 00000000 Oops: 0000 [#1] PREEMPT last sysfs file: /sys/devices/pci0000:00/0000:00:1d.7/usb2/2-1/2-1:1.0/host3/scsi_host/host3/uevent [..] Pid: 30039, comm: scsi_scan_3 Not tainted 2.6.35.2-fg.roam #1 Volvi2 /Aspire 4315 EIP: 0060:[] EFLAGS: 00010086 CPU: 0 EIP is at cfq_get_queue+0x232/0x412 EAX: f705f9c0 EBX: e977abac ECX: 00000000 EDX: 00000000 ESI: f00da400 EDI: f00da4ec EBP: e977a800 ESP: dff8fd00 DS: 007b ES: 007b FS: 0000 GS: 0000 SS: 0068 Process scsi_scan_3 (pid: 30039, ti=dff8e000 task=f6b6c9a0 task.ti=dff8e000) Stack: 00000000 00000000 00000001 01ff0000 f00da508 00000000 f00da524 f00da540 <0> e7994940 dd631750 f705f9c0 e977a820 e977ac44 f00da4d0 00000001 f6b6c9a0 <0> 00000010 00008010 0000000b 00000000 00000001 e977a800 dd76fac0 00000246 Call Trace: [] ? cfq_set_request+0x228/0x34c [] ? cfq_set_request+0x0/0x34c [] ? elv_set_request+0xf/0x1c [] ? get_request+0x1ad/0x22f [] ? get_request_wait+0x1f/0x11a [] ? kvasprintf+0x33/0x3b [] ? scsi_execute+0x1d/0x103 [] ? scsi_execute_req+0x58/0x83 [] ? scsi_probe_and_add_lun+0x188/0x7c2 [] ? attribute_container_add_device+0x15/0xfa [] ? kobject_get+0xf/0x13 [] ? get_device+0x10/0x14 [] ? scsi_alloc_target+0x217/0x24d [] ? __scsi_scan_target+0x95/0x480 [] ? dequeue_entity+0x14/0x1fe [] ? update_curr+0x165/0x1ab [] ? update_curr+0x165/0x1ab [] ? scsi_scan_channel+0x4a/0x76 [] ? scsi_scan_host_selected+0x77/0xad [] ? do_scan_async+0x0/0x11a [] ? do_scsi_scan_host+0x51/0x56 [] ? do_scan_async+0x0/0x11a [] ? do_scan_async+0xe/0x11a [] ? do_scan_async+0x0/0x11a [] ? kthread+0x5e/0x63 [] ? kthread+0x0/0x63 [] ? kernel_thread_helper+0x6/0x10 Code: 44 24 1c 54 83 44 24 18 54 83 fa 03 75 94 8b 06 c7 86 64 02 00 00 01 00 00 00 83 e0 03 09 f0 89 06 8b 44 24 28 8b 90 58 01 00 00 <8b> 42 2c 85 c0 75 03 8b 42 08 8d 54 24 48 52 8d 4c 24 50 51 68 EIP: [] cfq_get_queue+0x232/0x412 SS:ESP 0068:dff8fd00 CR2: 000000000000002c ---[ end trace 9a88306573f69b12 ]--- The problem here is that we don't have bdi->dev information available when thread does some IO. Hence when dev_name() tries to access bdi->dev, it crashes. This problem does not happen if kernel threads are in root group as root group is statically allocated at device initialization time and we don't hit this piece of code. Fix it by delaying the filling of major and minor number information of device in blk_group. Initially a blk_group is created with 0 as device information and this information is filled later once some more IO comes in from same group. Reported-by: Mike Kazantsev Signed-off-by: Vivek Goyal Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index f65c6f0..9eba291 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1019,10 +1019,20 @@ cfq_find_alloc_cfqg(struct cfq_data *cfqd, struct cgroup *cgroup, int create) */ atomic_set(&cfqg->ref, 1); - /* Add group onto cgroup list */ - sscanf(dev_name(bdi->dev), "%u:%u", &major, &minor); - cfq_blkiocg_add_blkio_group(blkcg, &cfqg->blkg, (void *)cfqd, + /* + * Add group onto cgroup list. It might happen that bdi->dev is + * not initiliazed yet. Initialize this new group without major + * and minor info and this info will be filled in once a new thread + * comes for IO. See code above. + */ + if (bdi->dev) { + sscanf(dev_name(bdi->dev), "%u:%u", &major, &minor); + cfq_blkiocg_add_blkio_group(blkcg, &cfqg->blkg, (void *)cfqd, MKDEV(major, minor)); + } else + cfq_blkiocg_add_blkio_group(blkcg, &cfqg->blkg, (void *)cfqd, + 0); + cfqg->weight = blkcg_get_weight(blkcg, cfqg->blkg.dev); /* Add group on cfqd list */ -- cgit v0.10.2 From bb7ab785ad05a97a2c9ffb3a06547ed39f3133e8 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Tue, 21 Sep 2010 03:26:35 -0400 Subject: oprofile: Add Support for Intel CPU Family 6 / Model 29 This patch adds CPU type detection for dunnington processor (Family 6 / Model 29) to be identified as core 2 family cpu type (wikipedia source). I tested oprofile on Intel(R) Xeon(R) CPU E7440 reporting itself as model 29, and it runs without an issue. Spec: http://www.intel.com/Assets/en_US/PDF/specupdate/320336.pdf Signed-off-by: Jiri Olsa Acked-by: Andi Kleen Cc: stable@kernel.org Signed-off-by: Robert Richter diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index 009b819..f1575c9 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -674,6 +674,7 @@ static int __init ppro_init(char **cpu_type) case 0x0f: case 0x16: case 0x17: + case 0x1d: *cpu_type = "i386/core_2"; break; case 0x1a: -- cgit v0.10.2 From f6c3f1686e7ec1dd8725a9a3dcb857dfd0c7a5bf Mon Sep 17 00:00:00 2001 From: Suresh Siddha Date: Mon, 13 Sep 2010 11:02:21 -0700 Subject: sched: Fix nohz balance kick There's a situation where the nohz balancer will try to wake itself: cpu-x is idle which is also ilb_cpu got a scheduler tick during idle and the nohz_kick_needed() in trigger_load_balance() checks for rq_x->nr_running which might not be zero (because of someone waking a task on this rq etc) and this leads to the situation of the cpu-x sending a kick to itself. And this can cause a lockup. Avoid this by not marking ourself eligible for kicking. Signed-off-by: Suresh Siddha Signed-off-by: Peter Zijlstra LKML-Reference: <1284400941.2684.19.camel@sbsiddha-MOBL3.sc.intel.com> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index a171138..db3f674 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -3630,7 +3630,7 @@ static inline int nohz_kick_needed(struct rq *rq, int cpu) if (time_before(now, nohz.next_balance)) return 0; - if (!rq->nr_running) + if (rq->idle_at_tick) return 0; first_pick_cpu = atomic_read(&nohz.first_pick_cpu); -- cgit v0.10.2 From 83d9f65bdae6f6b34d75282c6618d3f50846849a Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sun, 29 Aug 2010 18:12:24 -0700 Subject: x86, setup: Fix earlyprintk=serial,ttyS0,115200 Torsten reported that there is garbage output, after commit 8fee13a48e4879fba57725f6d9513df4bfa8e9f3 (x86, setup: enable early console output from the decompressor) It turns out we missed the offset for that case. Reported-by: Torsten Kaiser Signed-off-by: Yinghai Lu LKML-Reference: <4C7B0578.8090807@kernel.org> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/boot/early_serial_console.c b/arch/x86/boot/early_serial_console.c index 030f4b9..407a8e2 100644 --- a/arch/x86/boot/early_serial_console.c +++ b/arch/x86/boot/early_serial_console.c @@ -58,7 +58,7 @@ static void parse_earlyprintk(void) if (arg[pos] == ',') pos++; - if (!strncmp(arg, "ttyS", 4)) { + if (!strncmp(arg + pos, "ttyS", 4)) { static const int bases[] = { 0x3f8, 0x2f8 }; int idx = 0; -- cgit v0.10.2 From 74b3c444a963ba55aef89b33a1bcaada9a4c206f Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sun, 29 Aug 2010 18:13:10 -0700 Subject: x86, setup: Fix earlyprintk=serial,0x3f8,115200 earlyprintk can take and I/O port, so we need to handle this case in the setup code too, otherwise 0x3f8 will be treated as a baud rate. Signed-off-by: Yinghai Lu LKML-Reference: <4C7B05A6.4010801@kernel.org> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/boot/early_serial_console.c b/arch/x86/boot/early_serial_console.c index 407a8e2..5df2869 100644 --- a/arch/x86/boot/early_serial_console.c +++ b/arch/x86/boot/early_serial_console.c @@ -58,7 +58,19 @@ static void parse_earlyprintk(void) if (arg[pos] == ',') pos++; - if (!strncmp(arg + pos, "ttyS", 4)) { + /* + * make sure we have + * "serial,0x3f8,115200" + * "serial,ttyS0,115200" + * "ttyS0,115200" + */ + if (pos == 7 && !strncmp(arg + pos, "0x", 2)) { + port = simple_strtoull(arg + pos, &e, 16); + if (port == 0 || arg + pos == e) + port = DEFAULT_SERIAL_PORT; + else + pos = e - arg; + } else if (!strncmp(arg + pos, "ttyS", 4)) { static const int bases[] = { 0x3f8, 0x2f8 }; int idx = 0; -- cgit v0.10.2 From 9eecabcb9a924f1e11ba670365fd4babe423045c Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Tue, 21 Sep 2010 22:28:23 +0100 Subject: intel-iommu: Abort IOMMU setup for igfx if BIOS gave no shadow GTT space Yet another BIOS bug; Lenovo this time (X201). Red Hat bug #593516. Signed-off-by: David Woodhouse diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index c3ceebb..dee88c6c 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -3761,6 +3761,23 @@ static void __devinit quirk_iommu_rwbf(struct pci_dev *dev) DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2a40, quirk_iommu_rwbf); +static void __devinit quirk_calpella_no_shadow_gtt(struct pci_dev *dev) +{ + unsigned short ggc; + + if (pci_read_config_word(dev, 0x52, &ggc)) + return; + + if (!(ggc & 0x800)) { + printk(KERN_INFO "DMAR: BIOS has allocated no shadow GTT; disabling IOMMU for graphics\n"); + dmar_map_gfx = 0; + } +} +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x0040, quirk_calpella_no_shadow_gtt); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x0044, quirk_calpella_no_shadow_gtt); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x0062, quirk_calpella_no_shadow_gtt); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x006a, quirk_calpella_no_shadow_gtt); + /* On Tylersburg chipsets, some BIOSes have been known to enable the ISOCH DMAR unit for the Azalia sound device, but not give it any TLB entries, which causes it to deadlock. Check for that. We do -- cgit v0.10.2 From eecfd57f6429d9d8e10be186566ef99fced55163 Mon Sep 17 00:00:00 2001 From: Adam Jackson Date: Wed, 25 Aug 2010 21:17:34 +0100 Subject: intel-iommu: Use symbolic values instead of magic numbers in Lenovo w/a Commit 9eecabcb9a924f1e11ba670365fd4babe423045c ("intel-iommu: Abort IOMMU setup for igfx if BIOS gave no shadow GTT space") uses a bunch of magic numbers. Provide #defines for those to make it look slightly saner. Signed-off-by: Adam Jackson Signed-off-by: David Woodhouse diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index dee88c6c..4dfce3d 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -3761,14 +3761,24 @@ static void __devinit quirk_iommu_rwbf(struct pci_dev *dev) DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2a40, quirk_iommu_rwbf); +#define GGC 0x52 +#define GGC_MEMORY_SIZE_MASK (0xf << 8) +#define GGC_MEMORY_SIZE_NONE (0x0 << 8) +#define GGC_MEMORY_SIZE_1M (0x1 << 8) +#define GGC_MEMORY_SIZE_2M (0x3 << 8) +#define GGC_MEMORY_VT_ENABLED (0x8 << 8) +#define GGC_MEMORY_SIZE_2M_VT (0x9 << 8) +#define GGC_MEMORY_SIZE_3M_VT (0xa << 8) +#define GGC_MEMORY_SIZE_4M_VT (0xb << 8) + static void __devinit quirk_calpella_no_shadow_gtt(struct pci_dev *dev) { unsigned short ggc; - if (pci_read_config_word(dev, 0x52, &ggc)) + if (pci_read_config_word(dev, GGC, &ggc)) return; - if (!(ggc & 0x800)) { + if (!(ggc & GGC_MEMORY_VT_ENABLED)) { printk(KERN_INFO "DMAR: BIOS has allocated no shadow GTT; disabling IOMMU for graphics\n"); dmar_map_gfx = 0; } -- cgit v0.10.2 From 5c64eb26ed5c5550fbabd345e573af3fc6a7f775 Mon Sep 17 00:00:00 2001 From: Mathias Nyman Date: Thu, 26 Aug 2010 07:36:44 +0000 Subject: i2c-omap: Make sure i2c bus is free before setting it to idle If the i2c bus receives an interrupt with both BB (bus busy) and ARDY (register access ready) statuses set during the tranfer of the last message the bus was put to idle while still busy. This caused bus to timeout. Signed-off-by: Mathias Nyman Acked-by: Tony Lindgren Signed-off-by: Ben Dooks diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c index 7674efb..b33c785 100644 --- a/drivers/i2c/busses/i2c-omap.c +++ b/drivers/i2c/busses/i2c-omap.c @@ -680,6 +680,8 @@ omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num) if (r == 0) r = num; + + omap_i2c_wait_for_bb(dev); out: omap_i2c_idle(dev); return r; -- cgit v0.10.2 From 8b15575cae7a93a784c3005c42b069edd9ba64dd Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 21 Sep 2010 14:35:37 -0700 Subject: fs: {lock,unlock}_flocks() stubs to prepare for BKL removal The lock structs are currently protected by the BKL, but are accessed by code in fs/locks.c and misc file system and DLM code. These stubs will allow all users to switch to the new interface before the implementation is changed to a spinlock. Acked-by: Arnd Bergmann Signed-off-by: Sage Weil Signed-off-by: Linus Torvalds diff --git a/include/linux/fs.h b/include/linux/fs.h index 76041b6..63d069b 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1093,6 +1093,10 @@ struct file_lock { #include +/* temporary stubs for BKL removal */ +#define lock_flocks() lock_kernel() +#define unlock_flocks() unlock_kernel() + extern void send_sigio(struct fown_struct *fown, int fd, int band); #ifdef CONFIG_FILE_LOCKING -- cgit v0.10.2 From 5222454cf941b09205d502242dd1e44fd62aa551 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 22 Sep 2010 12:31:46 +1000 Subject: drm/radeon: don't allow device to be opened if powered down If the switcheroo has switched the device off, don't let X open it. Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 5eee3c4..8fbbe1c 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -203,6 +203,10 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) */ int radeon_driver_firstopen_kms(struct drm_device *dev) { + struct radeon_device *rdev = dev->dev_private; + + if (rdev->powered_down) + return -EINVAL; return 0; } -- cgit v0.10.2 From 0fbecd400dd0a82d465b3086f209681e8c54cb0f Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Tue, 21 Sep 2010 02:15:15 +0200 Subject: drm/ttm: Clear the ghost cpu_writers flag on ttm_buffer_object_transfer. It makes sense for a BO to move after a process has requested exclusive RW access on it (e.g. because the BO used to be located in unmappable VRAM and we intercepted the CPU access from the fault handler). If we let the ghost object inherit cpu_writers from the original object, ttm_bo_release_list() will raise a kernel BUG when the ghost object is destroyed. This can be reproduced with the nouveau driver on nv5x. Reported-by: Marcin Slusarz Reviewed-by: Jerome Glisse Tested-by: Marcin Slusarz Signed-off-by: Francisco Jerez Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c index 7cffb3e..3451a82 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_util.c +++ b/drivers/gpu/drm/ttm/ttm_bo_util.c @@ -351,6 +351,7 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo, INIT_LIST_HEAD(&fbo->lru); INIT_LIST_HEAD(&fbo->swap); fbo->vm_node = NULL; + atomic_set(&fbo->cpu_writers, 0); fbo->sync_obj = driver->sync_obj_ref(bo->sync_obj); kref_init(&fbo->list_kref); -- cgit v0.10.2 From 392c21802ee3aa85cee0e703105f797a8a7b9416 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 21 Sep 2010 21:41:12 -0700 Subject: sparc: Don't mask signal when we can't setup signal frame. Don't invoke the signal handler tracehook in that situation either. Reported-by: Al Viro Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/signal32.c b/arch/sparc/kernel/signal32.c index 76b67c4..643a354 100644 --- a/arch/sparc/kernel/signal32.c +++ b/arch/sparc/kernel/signal32.c @@ -511,8 +511,8 @@ out_irqs_on: } -static void setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, - int signo, sigset_t *oldset) +static int setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, + int signo, sigset_t *oldset) { struct signal_frame32 __user *sf; int sigframe_size; @@ -620,13 +620,16 @@ static void setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, sigill: do_exit(SIGILL); + return -EINVAL; + sigsegv: force_sigsegv(signo, current); + return -EFAULT; } -static void setup_rt_frame32(struct k_sigaction *ka, struct pt_regs *regs, - unsigned long signr, sigset_t *oldset, - siginfo_t *info) +static int setup_rt_frame32(struct k_sigaction *ka, struct pt_regs *regs, + unsigned long signr, sigset_t *oldset, + siginfo_t *info) { struct rt_signal_frame32 __user *sf; int sigframe_size; @@ -738,22 +741,30 @@ static void setup_rt_frame32(struct k_sigaction *ka, struct pt_regs *regs, flush_signal_insns(address); } - return; + return 0; sigill: do_exit(SIGILL); + return -EINVAL; + sigsegv: force_sigsegv(signr, current); + return -EFAULT; } -static inline void handle_signal32(unsigned long signr, struct k_sigaction *ka, - siginfo_t *info, - sigset_t *oldset, struct pt_regs *regs) +static inline int handle_signal32(unsigned long signr, struct k_sigaction *ka, + siginfo_t *info, + sigset_t *oldset, struct pt_regs *regs) { + int err; + if (ka->sa.sa_flags & SA_SIGINFO) - setup_rt_frame32(ka, regs, signr, oldset, info); + err = setup_rt_frame32(ka, regs, signr, oldset, info); else - setup_frame32(ka, regs, signr, oldset); + err = setup_frame32(ka, regs, signr, oldset); + + if (err) + return err; spin_lock_irq(¤t->sighand->siglock); sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask); @@ -761,6 +772,10 @@ static inline void handle_signal32(unsigned long signr, struct k_sigaction *ka, sigaddset(¤t->blocked,signr); recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); + + tracehook_signal_handler(signr, info, ka, regs, 0); + + return 0; } static inline void syscall_restart32(unsigned long orig_i0, struct pt_regs *regs, @@ -807,16 +822,14 @@ void do_signal32(sigset_t *oldset, struct pt_regs * regs, if (signr > 0) { if (restart_syscall) syscall_restart32(orig_i0, regs, &ka.sa); - handle_signal32(signr, &ka, &info, oldset, regs); - - /* A signal was successfully delivered; the saved - * sigmask will have been stored in the signal frame, - * and will be restored by sigreturn, so we can simply - * clear the TS_RESTORE_SIGMASK flag. - */ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; - - tracehook_signal_handler(signr, &info, &ka, regs, 0); + if (handle_signal32(signr, &ka, &info, oldset, regs) == 0) { + /* A signal was successfully delivered; the saved + * sigmask will have been stored in the signal frame, + * and will be restored by sigreturn, so we can simply + * clear the TS_RESTORE_SIGMASK flag. + */ + current_thread_info()->status &= ~TS_RESTORE_SIGMASK; + } return; } if (restart_syscall && diff --git a/arch/sparc/kernel/signal_32.c b/arch/sparc/kernel/signal_32.c index 9882df9..99c85e9 100644 --- a/arch/sparc/kernel/signal_32.c +++ b/arch/sparc/kernel/signal_32.c @@ -315,8 +315,8 @@ save_fpu_state(struct pt_regs *regs, __siginfo_fpu_t __user *fpu) return err; } -static void setup_frame(struct k_sigaction *ka, struct pt_regs *regs, - int signo, sigset_t *oldset) +static int setup_frame(struct k_sigaction *ka, struct pt_regs *regs, + int signo, sigset_t *oldset) { struct signal_frame __user *sf; int sigframe_size, err; @@ -384,16 +384,19 @@ static void setup_frame(struct k_sigaction *ka, struct pt_regs *regs, /* Flush instruction space. */ flush_sig_insns(current->mm, (unsigned long) &(sf->insns[0])); } - return; + return 0; sigill_and_return: do_exit(SIGILL); + return -EINVAL; + sigsegv: force_sigsegv(signo, current); + return -EFAULT; } -static void setup_rt_frame(struct k_sigaction *ka, struct pt_regs *regs, - int signo, sigset_t *oldset, siginfo_t *info) +static int setup_rt_frame(struct k_sigaction *ka, struct pt_regs *regs, + int signo, sigset_t *oldset, siginfo_t *info) { struct rt_signal_frame __user *sf; int sigframe_size; @@ -466,22 +469,30 @@ static void setup_rt_frame(struct k_sigaction *ka, struct pt_regs *regs, /* Flush instruction space. */ flush_sig_insns(current->mm, (unsigned long) &(sf->insns[0])); } - return; + return 0; sigill: do_exit(SIGILL); + return -EINVAL; + sigsegv: force_sigsegv(signo, current); + return -EFAULT; } -static inline void +static inline int handle_signal(unsigned long signr, struct k_sigaction *ka, siginfo_t *info, sigset_t *oldset, struct pt_regs *regs) { + int err; + if (ka->sa.sa_flags & SA_SIGINFO) - setup_rt_frame(ka, regs, signr, oldset, info); + err = setup_rt_frame(ka, regs, signr, oldset, info); else - setup_frame(ka, regs, signr, oldset); + err = setup_frame(ka, regs, signr, oldset); + + if (err) + return err; spin_lock_irq(¤t->sighand->siglock); sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask); @@ -489,6 +500,10 @@ handle_signal(unsigned long signr, struct k_sigaction *ka, sigaddset(¤t->blocked, signr); recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); + + tracehook_signal_handler(signr, info, ka, regs, 0); + + return 0; } static inline void syscall_restart(unsigned long orig_i0, struct pt_regs *regs, @@ -546,17 +561,15 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) if (signr > 0) { if (restart_syscall) syscall_restart(orig_i0, regs, &ka.sa); - handle_signal(signr, &ka, &info, oldset, regs); - - /* a signal was successfully delivered; the saved - * sigmask will have been stored in the signal frame, - * and will be restored by sigreturn, so we can simply - * clear the TIF_RESTORE_SIGMASK flag. - */ - if (test_thread_flag(TIF_RESTORE_SIGMASK)) - clear_thread_flag(TIF_RESTORE_SIGMASK); - - tracehook_signal_handler(signr, &info, &ka, regs, 0); + if (handle_signal(signr, &ka, &info, oldset, regs) == 0) { + /* a signal was successfully delivered; the saved + * sigmask will have been stored in the signal frame, + * and will be restored by sigreturn, so we can simply + * clear the TIF_RESTORE_SIGMASK flag. + */ + if (test_thread_flag(TIF_RESTORE_SIGMASK)) + clear_thread_flag(TIF_RESTORE_SIGMASK); + } return; } if (restart_syscall && diff --git a/arch/sparc/kernel/signal_64.c b/arch/sparc/kernel/signal_64.c index 9fa48c3..3f19e67 100644 --- a/arch/sparc/kernel/signal_64.c +++ b/arch/sparc/kernel/signal_64.c @@ -409,7 +409,7 @@ static inline void __user *get_sigframe(struct k_sigaction *ka, struct pt_regs * return (void __user *) sp; } -static inline void +static inline int setup_rt_frame(struct k_sigaction *ka, struct pt_regs *regs, int signo, sigset_t *oldset, siginfo_t *info) { @@ -483,26 +483,37 @@ setup_rt_frame(struct k_sigaction *ka, struct pt_regs *regs, } /* 4. return to kernel instructions */ regs->u_regs[UREG_I7] = (unsigned long)ka->ka_restorer; - return; + return 0; sigill: do_exit(SIGILL); + return -EINVAL; + sigsegv: force_sigsegv(signo, current); + return -EFAULT; } -static inline void handle_signal(unsigned long signr, struct k_sigaction *ka, - siginfo_t *info, - sigset_t *oldset, struct pt_regs *regs) +static inline int handle_signal(unsigned long signr, struct k_sigaction *ka, + siginfo_t *info, + sigset_t *oldset, struct pt_regs *regs) { - setup_rt_frame(ka, regs, signr, oldset, - (ka->sa.sa_flags & SA_SIGINFO) ? info : NULL); + int err; + + err = setup_rt_frame(ka, regs, signr, oldset, + (ka->sa.sa_flags & SA_SIGINFO) ? info : NULL); + if (err) + return err; spin_lock_irq(¤t->sighand->siglock); sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask); if (!(ka->sa.sa_flags & SA_NOMASK)) sigaddset(¤t->blocked,signr); recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); + + tracehook_signal_handler(signr, info, ka, regs, 0); + + return 0; } static inline void syscall_restart(unsigned long orig_i0, struct pt_regs *regs, @@ -571,16 +582,14 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) if (signr > 0) { if (restart_syscall) syscall_restart(orig_i0, regs, &ka.sa); - handle_signal(signr, &ka, &info, oldset, regs); - - /* A signal was successfully delivered; the saved - * sigmask will have been stored in the signal frame, - * and will be restored by sigreturn, so we can simply - * clear the TS_RESTORE_SIGMASK flag. - */ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; - - tracehook_signal_handler(signr, &info, &ka, regs, 0); + if (handle_signal(signr, &ka, &info, oldset, regs) == 0) { + /* A signal was successfully delivered; the saved + * sigmask will have been stored in the signal frame, + * and will be restored by sigreturn, so we can simply + * clear the TS_RESTORE_SIGMASK flag. + */ + current_thread_info()->status &= ~TS_RESTORE_SIGMASK; + } return; } if (restart_syscall && -- cgit v0.10.2 From c27852597829128a9c9d96d79ec454a83c6b0da5 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 21 Sep 2010 22:30:13 -0700 Subject: sparc: Prevent no-handler signal syscall restart recursion. Explicitly clear the "in-syscall" bit when we have no signal handler and back up the program counters to back up the system call. Reported-by: Al Viro Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/signal32.c b/arch/sparc/kernel/signal32.c index 643a354..75fad42 100644 --- a/arch/sparc/kernel/signal32.c +++ b/arch/sparc/kernel/signal32.c @@ -616,7 +616,7 @@ static int setup_frame32(struct k_sigaction *ka, struct pt_regs *regs, flush_signal_insns(address); } - return; + return 0; sigill: do_exit(SIGILL); @@ -840,12 +840,14 @@ void do_signal32(sigset_t *oldset, struct pt_regs * regs, regs->u_regs[UREG_I0] = orig_i0; regs->tpc -= 4; regs->tnpc -= 4; + pt_regs_clear_syscall(regs); } if (restart_syscall && regs->u_regs[UREG_I0] == ERESTART_RESTARTBLOCK) { regs->u_regs[UREG_G1] = __NR_restart_syscall; regs->tpc -= 4; regs->tnpc -= 4; + pt_regs_clear_syscall(regs); } /* If there's no signal to deliver, we just put the saved sigmask diff --git a/arch/sparc/kernel/signal_32.c b/arch/sparc/kernel/signal_32.c index 99c85e9..5e5c5fd 100644 --- a/arch/sparc/kernel/signal_32.c +++ b/arch/sparc/kernel/signal_32.c @@ -580,12 +580,14 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) regs->u_regs[UREG_I0] = orig_i0; regs->pc -= 4; regs->npc -= 4; + pt_regs_clear_syscall(regs); } if (restart_syscall && regs->u_regs[UREG_I0] == ERESTART_RESTARTBLOCK) { regs->u_regs[UREG_G1] = __NR_restart_syscall; regs->pc -= 4; regs->npc -= 4; + pt_regs_clear_syscall(regs); } /* if there's no signal to deliver, we just put the saved sigmask diff --git a/arch/sparc/kernel/signal_64.c b/arch/sparc/kernel/signal_64.c index 3f19e67..006fe45 100644 --- a/arch/sparc/kernel/signal_64.c +++ b/arch/sparc/kernel/signal_64.c @@ -600,12 +600,14 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) regs->u_regs[UREG_I0] = orig_i0; regs->tpc -= 4; regs->tnpc -= 4; + pt_regs_clear_syscall(regs); } if (restart_syscall && regs->u_regs[UREG_I0] == ERESTART_RESTARTBLOCK) { regs->u_regs[UREG_G1] = __NR_restart_syscall; regs->tpc -= 4; regs->tnpc -= 4; + pt_regs_clear_syscall(regs); } /* If there's no signal to deliver, we just put the saved sigmask -- cgit v0.10.2 From 976e48f8a5b02fc33f3e5cad87fb3fcea041a49c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 21 Sep 2010 11:48:55 +0200 Subject: bdi: Initialize noop_backing_dev_info properly Properly initialize this backing dev info so that writeback code does not barf when getting to it e.g. via sb->s_bdi. Cc: stable@kernel.org Signed-off-by: Jan Kara Signed-off-by: Jens Axboe diff --git a/mm/backing-dev.c b/mm/backing-dev.c index c2bf86f..65d4204 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -30,6 +30,7 @@ EXPORT_SYMBOL_GPL(default_backing_dev_info); struct backing_dev_info noop_backing_dev_info = { .name = "noop", + .capabilities = BDI_CAP_NO_ACCT_AND_WRITEBACK, }; EXPORT_SYMBOL_GPL(noop_backing_dev_info); @@ -243,6 +244,7 @@ static int __init default_bdi_init(void) err = bdi_init(&default_backing_dev_info); if (!err) bdi_register(&default_backing_dev_info, NULL, "default"); + err = bdi_init(&noop_backing_dev_info); return err; } -- cgit v0.10.2 From 371d217ee1ff8b418b8f73fb2a34990f951ec2d4 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 21 Sep 2010 11:49:01 +0200 Subject: char: Mark /dev/zero and /dev/kmem as not capable of writeback These devices don't do any writeback but their device inodes still can get dirty so mark bdi appropriately so that bdi code does the right thing and files inodes to lists of bdi carrying the device inodes. Cc: stable@kernel.org Signed-off-by: Jan Kara Signed-off-by: Jens Axboe diff --git a/drivers/char/mem.c b/drivers/char/mem.c index a398ecd..1f528fa 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -788,10 +788,11 @@ static const struct file_operations zero_fops = { /* * capabilities for /dev/zero * - permits private mappings, "copies" are taken of the source of zeros + * - no writeback happens */ static struct backing_dev_info zero_bdi = { .name = "char/mem", - .capabilities = BDI_CAP_MAP_COPY, + .capabilities = BDI_CAP_MAP_COPY | BDI_CAP_NO_ACCT_AND_WRITEBACK, }; static const struct file_operations full_fops = { diff --git a/fs/char_dev.c b/fs/char_dev.c index f80a4f2..143d393 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -40,7 +40,9 @@ struct backing_dev_info directly_mappable_cdev_bdi = { #endif /* permit direct mmap, for read, write or exec */ BDI_CAP_MAP_DIRECT | - BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP), + BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP | + /* no writeback happens */ + BDI_CAP_NO_ACCT_AND_WRITEBACK), }; static struct kobj_map *cdev_map; -- cgit v0.10.2 From 692ebd17c2905313fff3c504c249c6a0faad16ec Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 21 Sep 2010 11:51:01 +0200 Subject: bdi: Fix warnings in __mark_inode_dirty for /dev/zero and friends Inodes of devices such as /dev/zero can get dirty for example via utime(2) syscall or due to atime update. Backing device of such inodes (zero_bdi, etc.) is however unable to handle dirty inodes and thus __mark_inode_dirty complains. In fact, inode should be rather dirtied against backing device of the filesystem holding it. This is generally a good rule except for filesystems such as 'bdev' or 'mtd_inodefs'. Inodes in these pseudofilesystems are referenced from ordinary filesystem inodes and carry mapping with real data of the device. Thus for these inodes we have to use inode->i_mapping->backing_dev_info as we did so far. We distinguish these filesystems by checking whether sb->s_bdi points to a non-trivial backing device or not. Example: Assume we have an ext3 filesystem on /dev/sda1 mounted on /. There's a device inode A described by a path "/dev/sdb" on this filesystem. This inode will be dirtied against backing device "8:0" after this patch. bdev filesystem contains block device inode B coupled with our inode A. When someone modifies a page of /dev/sdb, it's B that gets dirtied and the dirtying happens against the backing device "8:16". Thus both inodes get filed to a correct bdi list. Cc: stable@kernel.org Signed-off-by: Jan Kara Signed-off-by: Jens Axboe diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 81e086d..5581122 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -52,8 +52,6 @@ struct wb_writeback_work { #define CREATE_TRACE_POINTS #include -#define inode_to_bdi(inode) ((inode)->i_mapping->backing_dev_info) - /* * We don't actually have pdflush, but this one is exported though /proc... */ @@ -71,6 +69,27 @@ int writeback_in_progress(struct backing_dev_info *bdi) return test_bit(BDI_writeback_running, &bdi->state); } +static inline struct backing_dev_info *inode_to_bdi(struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + struct backing_dev_info *bdi = inode->i_mapping->backing_dev_info; + + /* + * For inodes on standard filesystems, we use superblock's bdi. For + * inodes on virtual filesystems, we want to use inode mapping's bdi + * because they can possibly point to something useful (think about + * block_dev filesystem). + */ + if (sb->s_bdi && sb->s_bdi != &noop_backing_dev_info) { + /* Some device inodes could play dirty tricks. Catch them... */ + WARN(bdi != sb->s_bdi && bdi_cap_writeback_dirty(bdi), + "Dirtiable inode bdi %s != sb bdi %s\n", + bdi->name, sb->s_bdi->name); + return sb->s_bdi; + } + return bdi; +} + static void bdi_queue_work(struct backing_dev_info *bdi, struct wb_writeback_work *work) { -- cgit v0.10.2 From 0873a5ae747847ee55a63db409dff3476e45bcd9 Mon Sep 17 00:00:00 2001 From: "Erik J. Staab" Date: Wed, 22 Sep 2010 11:07:41 +0200 Subject: ALSA: oxygen: fix analog capture on Claro halo cards On the HT-Omega Claro halo card, the ADC data must be captured from the second I2S input. Using the default first input, which isn't connected to anything, would result in silence. Signed-off-by: Erik J. Staab Signed-off-by: Clemens Ladisch Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/oxygen/oxygen.c b/sound/pci/oxygen/oxygen.c index 289cb4d..6c0a11a 100644 --- a/sound/pci/oxygen/oxygen.c +++ b/sound/pci/oxygen/oxygen.c @@ -543,6 +543,10 @@ static int __devinit get_oxygen_model(struct oxygen *chip, chip->model.suspend = claro_suspend; chip->model.resume = claro_resume; chip->model.set_adc_params = set_ak5385_params; + chip->model.device_config = PLAYBACK_0_TO_I2S | + PLAYBACK_1_TO_SPDIF | + CAPTURE_0_FROM_I2S_2 | + CAPTURE_1_FROM_SPDIF; break; } if (id->driver_data == MODEL_MERIDIAN || -- cgit v0.10.2 From d47372e852391d0c6553dfbc7c4c56b89b527e13 Mon Sep 17 00:00:00 2001 From: Dimitris Papastamos Date: Tue, 21 Sep 2010 15:03:26 +0100 Subject: ASoC: Fix soc-cache buffer overflow bug Make sure we stay within the cache boundaries when updating the register cache. Signed-off-by: Dimitris Papastamos Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/soc-cache.c b/sound/soc/soc-cache.c index adbc68c..15d2779 100644 --- a/sound/soc/soc-cache.c +++ b/sound/soc/soc-cache.c @@ -203,8 +203,9 @@ static int snd_soc_8_16_write(struct snd_soc_codec *codec, unsigned int reg, data[1] = (value >> 8) & 0xff; data[2] = value & 0xff; - if (!snd_soc_codec_volatile_register(codec, reg)) - reg_cache[reg] = value; + if (!snd_soc_codec_volatile_register(codec, reg) + && reg < codec->driver->reg_cache_size) + reg_cache[reg] = value; if (codec->cache_only) { codec->cache_sync = 1; -- cgit v0.10.2 From 9a81c16b527528ad307843be5571111aa8d35a80 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2010 21:48:57 +0100 Subject: powerpc: fix double syscall restarts Make sigreturn zero regs->trap, make do_signal() do the same on all paths. As it is, signal interrupting e.g. read() from fd 512 (== ERESTARTSYS) with another signal getting unblocked when the first handler finishes will lead to restart one insn earlier than it ought to. Same for multiple signals with in-kernel handlers interrupting that sucker at the same time. Same for multiple signals of any kind interrupting that sucker on 64bit... Signed-off-by: Al Viro Acked-by: Paul Mackerras Signed-off-by: Linus Torvalds diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c index 7109f5b..2300426 100644 --- a/arch/powerpc/kernel/signal.c +++ b/arch/powerpc/kernel/signal.c @@ -138,6 +138,7 @@ static int do_signal_pending(sigset_t *oldset, struct pt_regs *regs) ti->local_flags &= ~_TLF_RESTORE_SIGMASK; sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); } + regs->trap = 0; return 0; /* no signals delivered */ } @@ -164,6 +165,7 @@ static int do_signal_pending(sigset_t *oldset, struct pt_regs *regs) ret = handle_rt_signal64(signr, &ka, &info, oldset, regs); } + regs->trap = 0; if (ret) { spin_lock_irq(¤t->sighand->siglock); sigorsets(¤t->blocked, ¤t->blocked, diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c index 2666101..b96a3a0 100644 --- a/arch/powerpc/kernel/signal_32.c +++ b/arch/powerpc/kernel/signal_32.c @@ -511,6 +511,7 @@ static long restore_user_regs(struct pt_regs *regs, if (!sig) save_r2 = (unsigned int)regs->gpr[2]; err = restore_general_regs(regs, sr); + regs->trap = 0; err |= __get_user(msr, &sr->mc_gregs[PT_MSR]); if (!sig) regs->gpr[2] = (unsigned long) save_r2; @@ -884,7 +885,6 @@ int handle_rt_signal32(unsigned long sig, struct k_sigaction *ka, regs->nip = (unsigned long) ka->sa.sa_handler; /* enter the signal handler in big-endian mode */ regs->msr &= ~MSR_LE; - regs->trap = 0; return 1; badframe: @@ -1228,7 +1228,6 @@ int handle_signal32(unsigned long sig, struct k_sigaction *ka, regs->nip = (unsigned long) ka->sa.sa_handler; /* enter the signal handler in big-endian mode */ regs->msr &= ~MSR_LE; - regs->trap = 0; return 1; diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c index 2fe6fc6..27c4a45 100644 --- a/arch/powerpc/kernel/signal_64.c +++ b/arch/powerpc/kernel/signal_64.c @@ -178,7 +178,7 @@ static long restore_sigcontext(struct pt_regs *regs, sigset_t *set, int sig, err |= __get_user(regs->xer, &sc->gp_regs[PT_XER]); err |= __get_user(regs->ccr, &sc->gp_regs[PT_CCR]); /* skip SOFTE */ - err |= __get_user(regs->trap, &sc->gp_regs[PT_TRAP]); + regs->trap = 0; err |= __get_user(regs->dar, &sc->gp_regs[PT_DAR]); err |= __get_user(regs->dsisr, &sc->gp_regs[PT_DSISR]); err |= __get_user(regs->result, &sc->gp_regs[PT_RESULT]); -- cgit v0.10.2 From 0077ca0b5c986477e33451b797b6e7dc92a8bbc0 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 22 Sep 2010 18:47:40 +0100 Subject: ASoC: Fix multi-componentism Spot the build testing. Signed-off-by: Mark Brown diff --git a/sound/soc/soc-cache.c b/sound/soc/soc-cache.c index 15d2779..f6b0d28 100644 --- a/sound/soc/soc-cache.c +++ b/sound/soc/soc-cache.c @@ -204,7 +204,7 @@ static int snd_soc_8_16_write(struct snd_soc_codec *codec, unsigned int reg, data[2] = value & 0xff; if (!snd_soc_codec_volatile_register(codec, reg) - && reg < codec->driver->reg_cache_size) + && reg < codec->reg_cache_size) reg_cache[reg] = value; if (codec->cache_only) { -- cgit v0.10.2 From 9f9ff20d46c6728b092f34b6a642e1e81ab5e254 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 14 Aug 2010 11:01:45 +0200 Subject: dma/shdma: move dereference below the NULL check "param" can be NULL here, so only dereference it after the check. Signed-off-by: Dan Carpenter Signed-off-by: Dan Williams diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c index fb64cf3..eb6b54d 100644 --- a/drivers/dma/shdma.c +++ b/drivers/dma/shdma.c @@ -580,7 +580,6 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_slave_sg( sh_chan = to_sh_chan(chan); param = chan->private; - slave_addr = param->config->addr; /* Someone calling slave DMA on a public channel? */ if (!param || !sg_len) { @@ -589,6 +588,8 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_slave_sg( return NULL; } + slave_addr = param->config->addr; + /* * if (param != NULL), this is a successfully requested slave channel, * therefore param->config != NULL too. -- cgit v0.10.2 From d3f3cf859db17cc5f8156c5bfcd032413e44483b Mon Sep 17 00:00:00 2001 From: Mathieu Lacage Date: Sat, 14 Aug 2010 15:02:44 +0200 Subject: missing inline keyword for static function in linux/dmaengine.h Add a missing inline keyword for static function in linux/dmaengine.h to avoid duplicate symbol definitions. Signed-off-by: Mathieu Lacage Signed-off-by: Dan Williams diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index c61d4ca..e210649 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -548,7 +548,7 @@ static inline bool dma_dev_has_pq_continue(struct dma_device *dma) return (dma->max_pq & DMA_HAS_PQ_CONTINUE) == DMA_HAS_PQ_CONTINUE; } -static unsigned short dma_dev_to_maxpq(struct dma_device *dma) +static inline unsigned short dma_dev_to_maxpq(struct dma_device *dma) { return dma->max_pq & ~DMA_HAS_PQ_CONTINUE; } -- cgit v0.10.2 From 767b68e96993e29e3480d7ecdd9c4b84667c5762 Mon Sep 17 00:00:00 2001 From: Dan Rosenberg Date: Wed, 22 Sep 2010 14:32:56 -0400 Subject: Prevent freeing uninitialized pointer in compat_do_readv_writev In 32-bit compatibility mode, the error handling for compat_do_readv_writev() may free an uninitialized pointer, potentially leading to all sorts of ugly memory corruption. This is reliably triggerable by unprivileged users by invoking the readv()/writev() syscalls with an invalid iovec pointer. The below patch fixes this to emulate the non-compat version. Introduced by commit b83733639a49 ("compat: factor out compat_rw_copy_check_uvector from compat_do_readv_writev") Signed-off-by: Dan Rosenberg Cc: stable@kernel.org (2.6.35) Cc: Al Viro Signed-off-by: Linus Torvalds diff --git a/fs/compat.c b/fs/compat.c index 718c706..0644a15 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1153,7 +1153,7 @@ static ssize_t compat_do_readv_writev(int type, struct file *file, { compat_ssize_t tot_len; struct iovec iovstack[UIO_FASTIOV]; - struct iovec *iov; + struct iovec *iov = iovstack; ssize_t ret; io_fn_t fn; iov_fn_t fnv; -- cgit v0.10.2 From f19e8aa11afa24036c6273428da51949b5acf30c Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 22 Sep 2010 13:04:52 -0700 Subject: oom: always return a badness score of non-zero for eligible tasks A task's badness score is roughly a proportion of its rss and swap compared to the system's capacity. The scale ranges from 0 to 1000 with the highest score chosen for kill. Thus, this scale operates on a resolution of 0.1% of RAM + swap. Admin tasks are also given a 3% bonus, so the badness score of an admin task using 3% of memory, for example, would still be 0. It's possible that an exceptionally large number of tasks will combine to exhaust all resources but never have a single task that uses more than 0.1% of RAM and swap (or 3.0% for admin tasks). This patch ensures that the badness score of any eligible task is never 0 so the machine doesn't unnecessarily panic because it cannot find a task to kill. Signed-off-by: David Rientjes Cc: Dave Hansen Cc: Nitin Gupta Cc: Pekka Enberg Cc: Minchan Kim Cc: KAMEZAWA Hiroyuki Cc: KOSAKI Motohiro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/oom_kill.c b/mm/oom_kill.c index fc81cb2..859250c 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -208,8 +208,13 @@ unsigned int oom_badness(struct task_struct *p, struct mem_cgroup *mem, */ points += p->signal->oom_score_adj; - if (points < 0) - return 0; + /* + * Never return 0 for an eligible task that may be killed since it's + * possible that no single user task uses more than 0.1% of memory and + * no single admin tasks uses more than 3.0%. + */ + if (points <= 0) + return 1; return (points < 1000) ? points : 1000; } -- cgit v0.10.2 From a9e31765e7d528858e1b0c202b823cf4df7577ca Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 22 Sep 2010 13:04:53 -0700 Subject: ipmi: fix acpi probe print After d9e1b6c45059ccf ("ipmi: fix ACPI detection with regspacing") we get [ 11.026326] ipmi_si: probing via ACPI [ 11.030019] ipmi_si 00:09: (null) regsize 1 spacing 1 irq 0 [ 11.035594] ipmi_si: Adding ACPI-specified kcs state machine on an old system with only one range for ipmi kcs range. Try to fix it by adding another res pointer. Signed-off-by: Yinghai Lu Signed-off-by: Corey Minyard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 3822b4f..2be457a 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -2126,7 +2126,7 @@ static int __devinit ipmi_pnp_probe(struct pnp_dev *dev, { struct acpi_device *acpi_dev; struct smi_info *info; - struct resource *res; + struct resource *res, *res_second; acpi_handle handle; acpi_status status; unsigned long long tmp; @@ -2182,13 +2182,13 @@ static int __devinit ipmi_pnp_probe(struct pnp_dev *dev, info->io.addr_data = res->start; info->io.regspacing = DEFAULT_REGSPACING; - res = pnp_get_resource(dev, + res_second = pnp_get_resource(dev, (info->io.addr_type == IPMI_IO_ADDR_SPACE) ? IORESOURCE_IO : IORESOURCE_MEM, 1); - if (res) { - if (res->start > info->io.addr_data) - info->io.regspacing = res->start - info->io.addr_data; + if (res_second) { + if (res_second->start > info->io.addr_data) + info->io.regspacing = res_second->start - info->io.addr_data; } info->io.regsize = DEFAULT_REGSPACING; info->io.regshift = 0; -- cgit v0.10.2 From c227e69028473c7c7994a9b0a2cc0034f3f7e0fe Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 22 Sep 2010 13:04:54 -0700 Subject: /proc/vmcore: fix seeking Commit 73296bc611 ("procfs: Use generic_file_llseek in /proc/vmcore") broke seeking on /proc/vmcore. This changes it back to use default_llseek in order to restore the original behaviour. The problem with generic_file_llseek is that it only allows seeks up to inode->i_sb->s_maxbytes, which is zero on procfs and some other virtual file systems. We should merge generic_file_llseek and default_llseek some day and clean this up in a proper way, but for 2.6.35/36, reverting vmcore is the safer solution. Signed-off-by: Arnd Bergmann Cc: Frederic Weisbecker Reported-by: CAI Qian Tested-by: CAI Qian Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 91c817f..2367fb3 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -163,7 +163,7 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer, static const struct file_operations proc_vmcore_operations = { .read = read_vmcore, - .llseek = generic_file_llseek, + .llseek = default_llseek, }; static struct vmcore* __init get_new_element(void) -- cgit v0.10.2 From 710224fa2750cf449c02dd115548acebfdd2c86a Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Wed, 22 Sep 2010 13:04:55 -0700 Subject: arm: fix "arm: fix pci_set_consistent_dma_mask for dmabounce devices" This fixes the regression caused by the commit 6fee48cd330c68 ("dma-mapping: arm: use generic pci_set_dma_mask and pci_set_consistent_dma_mask"). ARM needs to clip the dma coherent mask for dmabounce devices. This restores the old trick. Note that strictly speaking, the DMA API doesn't allow architectures to do such but I'm not sure it's worth adding the new API to set the dma mask that allows architectures to clip it. Reported-by: Krzysztof Halasa Signed-off-by: FUJITA Tomonori Acked-by: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/arm/common/it8152.c b/arch/arm/common/it8152.c index 7974baa..1bec96e 100644 --- a/arch/arm/common/it8152.c +++ b/arch/arm/common/it8152.c @@ -271,6 +271,14 @@ int dma_needs_bounce(struct device *dev, dma_addr_t dma_addr, size_t size) ((dma_addr + size - PHYS_OFFSET) >= SZ_64M); } +int dma_set_coherent_mask(struct device *dev, u64 mask) +{ + if (mask >= PHYS_OFFSET + SZ_64M - 1) + return 0; + + return -EIO; +} + int __init it8152_pci_setup(int nr, struct pci_sys_data *sys) { it8152_io.start = IT8152_IO_BASE + 0x12000; diff --git a/arch/arm/mach-ixp4xx/common-pci.c b/arch/arm/mach-ixp4xx/common-pci.c index 61cd4d6..24498a9 100644 --- a/arch/arm/mach-ixp4xx/common-pci.c +++ b/arch/arm/mach-ixp4xx/common-pci.c @@ -503,6 +503,14 @@ struct pci_bus * __devinit ixp4xx_scan_bus(int nr, struct pci_sys_data *sys) return pci_scan_bus(sys->busnr, &ixp4xx_ops, sys); } +int dma_set_coherent_mask(struct device *dev, u64 mask) +{ + if (mask >= SZ_64M - 1) + return 0; + + return -EIO; +} + EXPORT_SYMBOL(ixp4xx_pci_read); EXPORT_SYMBOL(ixp4xx_pci_write); diff --git a/arch/arm/mach-ixp4xx/include/mach/hardware.h b/arch/arm/mach-ixp4xx/include/mach/hardware.h index f91ca6d..8138371 100644 --- a/arch/arm/mach-ixp4xx/include/mach/hardware.h +++ b/arch/arm/mach-ixp4xx/include/mach/hardware.h @@ -26,6 +26,8 @@ #define PCIBIOS_MAX_MEM 0x4BFFFFFF #endif +#define ARCH_HAS_DMA_SET_COHERENT_MASK + #define pcibios_assign_all_busses() 1 /* Register locations and bits */ diff --git a/arch/arm/mach-pxa/include/mach/hardware.h b/arch/arm/mach-pxa/include/mach/hardware.h index 7f64d24..428cc7b 100644 --- a/arch/arm/mach-pxa/include/mach/hardware.h +++ b/arch/arm/mach-pxa/include/mach/hardware.h @@ -309,7 +309,7 @@ extern unsigned long get_clock_tick_rate(void); #define PCIBIOS_MIN_IO 0 #define PCIBIOS_MIN_MEM 0 #define pcibios_assign_all_busses() 1 +#define ARCH_HAS_DMA_SET_COHERENT_MASK #endif - #endif /* _ASM_ARCH_HARDWARE_H */ diff --git a/arch/arm/mach-pxa/include/mach/io.h b/arch/arm/mach-pxa/include/mach/io.h index 262691f..fdca3be 100644 --- a/arch/arm/mach-pxa/include/mach/io.h +++ b/arch/arm/mach-pxa/include/mach/io.h @@ -6,6 +6,8 @@ #ifndef __ASM_ARM_ARCH_IO_H #define __ASM_ARM_ARCH_IO_H +#include + #define IO_SPACE_LIMIT 0xffffffff /* diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index ce29b81..ba8319a 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -102,6 +102,9 @@ static inline u64 dma_get_mask(struct device *dev) return DMA_BIT_MASK(32); } +#ifdef ARCH_HAS_DMA_SET_COHERENT_MASK +int dma_set_coherent_mask(struct device *dev, u64 mask); +#else static inline int dma_set_coherent_mask(struct device *dev, u64 mask) { if (!dma_supported(dev, mask)) @@ -109,6 +112,7 @@ static inline int dma_set_coherent_mask(struct device *dev, u64 mask) dev->coherent_dma_mask = mask; return 0; } +#endif extern u64 dma_get_required_mask(struct device *dev); -- cgit v0.10.2 From d544b7a40ad3423676b8876aad64fc5f87296b2d Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 22 Sep 2010 13:04:57 -0700 Subject: vmware balloon: rename module In an effort to minimize customer confusion we want to unify naming convention for VMware-provided kernel modules. This change renames the balloon driver from vmware_ballon to vmw_balloon. We expect to follow this naming convention (vmw_) for all modules that are part of mainline kernel and/or being distributed by VMware, with the sole exception of vmxnet3 driver (since the name of mainline driver happens to match with the name used in VMware Tools). Signed-off-by: Dmitry Torokhov Acked-by: Bhavesh Davda Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig index 0b591b6..b743312 100644 --- a/drivers/misc/Kconfig +++ b/drivers/misc/Kconfig @@ -368,7 +368,7 @@ config VMWARE_BALLOON If unsure, say N. To compile this driver as a module, choose M here: the - module will be called vmware_balloon. + module will be called vmw_balloon. config ARM_CHARLCD bool "ARM Ltd. Character LCD Driver" diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile index 255a80d..42eab95 100644 --- a/drivers/misc/Makefile +++ b/drivers/misc/Makefile @@ -33,5 +33,5 @@ obj-$(CONFIG_IWMC3200TOP) += iwmc3200top/ obj-$(CONFIG_HMC6352) += hmc6352.o obj-y += eeprom/ obj-y += cb710/ -obj-$(CONFIG_VMWARE_BALLOON) += vmware_balloon.o +obj-$(CONFIG_VMWARE_BALLOON) += vmw_balloon.o obj-$(CONFIG_ARM_CHARLCD) += arm-charlcd.o diff --git a/drivers/misc/vmw_balloon.c b/drivers/misc/vmw_balloon.c new file mode 100644 index 0000000..2a1e804 --- /dev/null +++ b/drivers/misc/vmw_balloon.c @@ -0,0 +1,844 @@ +/* + * VMware Balloon driver. + * + * Copyright (C) 2000-2010, VMware, Inc. All Rights Reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; version 2 of the License and no later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. + * + * Maintained by: Dmitry Torokhov + */ + +/* + * This is VMware physical memory management driver for Linux. The driver + * acts like a "balloon" that can be inflated to reclaim physical pages by + * reserving them in the guest and invalidating them in the monitor, + * freeing up the underlying machine pages so they can be allocated to + * other guests. The balloon can also be deflated to allow the guest to + * use more physical memory. Higher level policies can control the sizes + * of balloons in VMs in order to manage physical memory resources. + */ + +//#define DEBUG +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +MODULE_AUTHOR("VMware, Inc."); +MODULE_DESCRIPTION("VMware Memory Control (Balloon) Driver"); +MODULE_VERSION("1.2.1.1-k"); +MODULE_ALIAS("dmi:*:svnVMware*:*"); +MODULE_ALIAS("vmware_vmmemctl"); +MODULE_LICENSE("GPL"); + +/* + * Various constants controlling rate of inflaint/deflating balloon, + * measured in pages. + */ + +/* + * Rate of allocating memory when there is no memory pressure + * (driver performs non-sleeping allocations). + */ +#define VMW_BALLOON_NOSLEEP_ALLOC_MAX 16384U + +/* + * Rates of memory allocaton when guest experiences memory pressure + * (driver performs sleeping allocations). + */ +#define VMW_BALLOON_RATE_ALLOC_MIN 512U +#define VMW_BALLOON_RATE_ALLOC_MAX 2048U +#define VMW_BALLOON_RATE_ALLOC_INC 16U + +/* + * Rates for releasing pages while deflating balloon. + */ +#define VMW_BALLOON_RATE_FREE_MIN 512U +#define VMW_BALLOON_RATE_FREE_MAX 16384U +#define VMW_BALLOON_RATE_FREE_INC 16U + +/* + * When guest is under memory pressure, use a reduced page allocation + * rate for next several cycles. + */ +#define VMW_BALLOON_SLOW_CYCLES 4 + +/* + * Use __GFP_HIGHMEM to allow pages from HIGHMEM zone. We don't + * allow wait (__GFP_WAIT) for NOSLEEP page allocations. Use + * __GFP_NOWARN, to suppress page allocation failure warnings. + */ +#define VMW_PAGE_ALLOC_NOSLEEP (__GFP_HIGHMEM|__GFP_NOWARN) + +/* + * Use GFP_HIGHUSER when executing in a separate kernel thread + * context and allocation can sleep. This is less stressful to + * the guest memory system, since it allows the thread to block + * while memory is reclaimed, and won't take pages from emergency + * low-memory pools. + */ +#define VMW_PAGE_ALLOC_CANSLEEP (GFP_HIGHUSER) + +/* Maximum number of page allocations without yielding processor */ +#define VMW_BALLOON_YIELD_THRESHOLD 1024 + +/* Maximum number of refused pages we accumulate during inflation cycle */ +#define VMW_BALLOON_MAX_REFUSED 16 + +/* + * Hypervisor communication port definitions. + */ +#define VMW_BALLOON_HV_PORT 0x5670 +#define VMW_BALLOON_HV_MAGIC 0x456c6d6f +#define VMW_BALLOON_PROTOCOL_VERSION 2 +#define VMW_BALLOON_GUEST_ID 1 /* Linux */ + +#define VMW_BALLOON_CMD_START 0 +#define VMW_BALLOON_CMD_GET_TARGET 1 +#define VMW_BALLOON_CMD_LOCK 2 +#define VMW_BALLOON_CMD_UNLOCK 3 +#define VMW_BALLOON_CMD_GUEST_ID 4 + +/* error codes */ +#define VMW_BALLOON_SUCCESS 0 +#define VMW_BALLOON_FAILURE -1 +#define VMW_BALLOON_ERROR_CMD_INVALID 1 +#define VMW_BALLOON_ERROR_PPN_INVALID 2 +#define VMW_BALLOON_ERROR_PPN_LOCKED 3 +#define VMW_BALLOON_ERROR_PPN_UNLOCKED 4 +#define VMW_BALLOON_ERROR_PPN_PINNED 5 +#define VMW_BALLOON_ERROR_PPN_NOTNEEDED 6 +#define VMW_BALLOON_ERROR_RESET 7 +#define VMW_BALLOON_ERROR_BUSY 8 + +#define VMWARE_BALLOON_CMD(cmd, data, result) \ +({ \ + unsigned long __stat, __dummy1, __dummy2; \ + __asm__ __volatile__ ("inl (%%dx)" : \ + "=a"(__stat), \ + "=c"(__dummy1), \ + "=d"(__dummy2), \ + "=b"(result) : \ + "0"(VMW_BALLOON_HV_MAGIC), \ + "1"(VMW_BALLOON_CMD_##cmd), \ + "2"(VMW_BALLOON_HV_PORT), \ + "3"(data) : \ + "memory"); \ + result &= -1UL; \ + __stat & -1UL; \ +}) + +#ifdef CONFIG_DEBUG_FS +struct vmballoon_stats { + unsigned int timer; + + /* allocation statustics */ + unsigned int alloc; + unsigned int alloc_fail; + unsigned int sleep_alloc; + unsigned int sleep_alloc_fail; + unsigned int refused_alloc; + unsigned int refused_free; + unsigned int free; + + /* monitor operations */ + unsigned int lock; + unsigned int lock_fail; + unsigned int unlock; + unsigned int unlock_fail; + unsigned int target; + unsigned int target_fail; + unsigned int start; + unsigned int start_fail; + unsigned int guest_type; + unsigned int guest_type_fail; +}; + +#define STATS_INC(stat) (stat)++ +#else +#define STATS_INC(stat) +#endif + +struct vmballoon { + + /* list of reserved physical pages */ + struct list_head pages; + + /* transient list of non-balloonable pages */ + struct list_head refused_pages; + unsigned int n_refused_pages; + + /* balloon size in pages */ + unsigned int size; + unsigned int target; + + /* reset flag */ + bool reset_required; + + /* adjustment rates (pages per second) */ + unsigned int rate_alloc; + unsigned int rate_free; + + /* slowdown page allocations for next few cycles */ + unsigned int slow_allocation_cycles; + +#ifdef CONFIG_DEBUG_FS + /* statistics */ + struct vmballoon_stats stats; + + /* debugfs file exporting statistics */ + struct dentry *dbg_entry; +#endif + + struct sysinfo sysinfo; + + struct delayed_work dwork; +}; + +static struct vmballoon balloon; +static struct workqueue_struct *vmballoon_wq; + +/* + * Send "start" command to the host, communicating supported version + * of the protocol. + */ +static bool vmballoon_send_start(struct vmballoon *b) +{ + unsigned long status, dummy; + + STATS_INC(b->stats.start); + + status = VMWARE_BALLOON_CMD(START, VMW_BALLOON_PROTOCOL_VERSION, dummy); + if (status == VMW_BALLOON_SUCCESS) + return true; + + pr_debug("%s - failed, hv returns %ld\n", __func__, status); + STATS_INC(b->stats.start_fail); + return false; +} + +static bool vmballoon_check_status(struct vmballoon *b, unsigned long status) +{ + switch (status) { + case VMW_BALLOON_SUCCESS: + return true; + + case VMW_BALLOON_ERROR_RESET: + b->reset_required = true; + /* fall through */ + + default: + return false; + } +} + +/* + * Communicate guest type to the host so that it can adjust ballooning + * algorithm to the one most appropriate for the guest. This command + * is normally issued after sending "start" command and is part of + * standard reset sequence. + */ +static bool vmballoon_send_guest_id(struct vmballoon *b) +{ + unsigned long status, dummy; + + status = VMWARE_BALLOON_CMD(GUEST_ID, VMW_BALLOON_GUEST_ID, dummy); + + STATS_INC(b->stats.guest_type); + + if (vmballoon_check_status(b, status)) + return true; + + pr_debug("%s - failed, hv returns %ld\n", __func__, status); + STATS_INC(b->stats.guest_type_fail); + return false; +} + +/* + * Retrieve desired balloon size from the host. + */ +static bool vmballoon_send_get_target(struct vmballoon *b, u32 *new_target) +{ + unsigned long status; + unsigned long target; + unsigned long limit; + u32 limit32; + + /* + * si_meminfo() is cheap. Moreover, we want to provide dynamic + * max balloon size later. So let us call si_meminfo() every + * iteration. + */ + si_meminfo(&b->sysinfo); + limit = b->sysinfo.totalram; + + /* Ensure limit fits in 32-bits */ + limit32 = (u32)limit; + if (limit != limit32) + return false; + + /* update stats */ + STATS_INC(b->stats.target); + + status = VMWARE_BALLOON_CMD(GET_TARGET, limit, target); + if (vmballoon_check_status(b, status)) { + *new_target = target; + return true; + } + + pr_debug("%s - failed, hv returns %ld\n", __func__, status); + STATS_INC(b->stats.target_fail); + return false; +} + +/* + * Notify the host about allocated page so that host can use it without + * fear that guest will need it. Host may reject some pages, we need to + * check the return value and maybe submit a different page. + */ +static bool vmballoon_send_lock_page(struct vmballoon *b, unsigned long pfn) +{ + unsigned long status, dummy; + u32 pfn32; + + pfn32 = (u32)pfn; + if (pfn32 != pfn) + return false; + + STATS_INC(b->stats.lock); + + status = VMWARE_BALLOON_CMD(LOCK, pfn, dummy); + if (vmballoon_check_status(b, status)) + return true; + + pr_debug("%s - ppn %lx, hv returns %ld\n", __func__, pfn, status); + STATS_INC(b->stats.lock_fail); + return false; +} + +/* + * Notify the host that guest intends to release given page back into + * the pool of available (to the guest) pages. + */ +static bool vmballoon_send_unlock_page(struct vmballoon *b, unsigned long pfn) +{ + unsigned long status, dummy; + u32 pfn32; + + pfn32 = (u32)pfn; + if (pfn32 != pfn) + return false; + + STATS_INC(b->stats.unlock); + + status = VMWARE_BALLOON_CMD(UNLOCK, pfn, dummy); + if (vmballoon_check_status(b, status)) + return true; + + pr_debug("%s - ppn %lx, hv returns %ld\n", __func__, pfn, status); + STATS_INC(b->stats.unlock_fail); + return false; +} + +/* + * Quickly release all pages allocated for the balloon. This function is + * called when host decides to "reset" balloon for one reason or another. + * Unlike normal "deflate" we do not (shall not) notify host of the pages + * being released. + */ +static void vmballoon_pop(struct vmballoon *b) +{ + struct page *page, *next; + unsigned int count = 0; + + list_for_each_entry_safe(page, next, &b->pages, lru) { + list_del(&page->lru); + __free_page(page); + STATS_INC(b->stats.free); + b->size--; + + if (++count >= b->rate_free) { + count = 0; + cond_resched(); + } + } +} + +/* + * Perform standard reset sequence by popping the balloon (in case it + * is not empty) and then restarting protocol. This operation normally + * happens when host responds with VMW_BALLOON_ERROR_RESET to a command. + */ +static void vmballoon_reset(struct vmballoon *b) +{ + /* free all pages, skipping monitor unlock */ + vmballoon_pop(b); + + if (vmballoon_send_start(b)) { + b->reset_required = false; + if (!vmballoon_send_guest_id(b)) + pr_err("failed to send guest ID to the host\n"); + } +} + +/* + * Allocate (or reserve) a page for the balloon and notify the host. If host + * refuses the page put it on "refuse" list and allocate another one until host + * is satisfied. "Refused" pages are released at the end of inflation cycle + * (when we allocate b->rate_alloc pages). + */ +static int vmballoon_reserve_page(struct vmballoon *b, bool can_sleep) +{ + struct page *page; + gfp_t flags; + bool locked = false; + + do { + if (!can_sleep) + STATS_INC(b->stats.alloc); + else + STATS_INC(b->stats.sleep_alloc); + + flags = can_sleep ? VMW_PAGE_ALLOC_CANSLEEP : VMW_PAGE_ALLOC_NOSLEEP; + page = alloc_page(flags); + if (!page) { + if (!can_sleep) + STATS_INC(b->stats.alloc_fail); + else + STATS_INC(b->stats.sleep_alloc_fail); + return -ENOMEM; + } + + /* inform monitor */ + locked = vmballoon_send_lock_page(b, page_to_pfn(page)); + if (!locked) { + STATS_INC(b->stats.refused_alloc); + + if (b->reset_required) { + __free_page(page); + return -EIO; + } + + /* + * Place page on the list of non-balloonable pages + * and retry allocation, unless we already accumulated + * too many of them, in which case take a breather. + */ + list_add(&page->lru, &b->refused_pages); + if (++b->n_refused_pages >= VMW_BALLOON_MAX_REFUSED) + return -EIO; + } + } while (!locked); + + /* track allocated page */ + list_add(&page->lru, &b->pages); + + /* update balloon size */ + b->size++; + + return 0; +} + +/* + * Release the page allocated for the balloon. Note that we first notify + * the host so it can make sure the page will be available for the guest + * to use, if needed. + */ +static int vmballoon_release_page(struct vmballoon *b, struct page *page) +{ + if (!vmballoon_send_unlock_page(b, page_to_pfn(page))) + return -EIO; + + list_del(&page->lru); + + /* deallocate page */ + __free_page(page); + STATS_INC(b->stats.free); + + /* update balloon size */ + b->size--; + + return 0; +} + +/* + * Release pages that were allocated while attempting to inflate the + * balloon but were refused by the host for one reason or another. + */ +static void vmballoon_release_refused_pages(struct vmballoon *b) +{ + struct page *page, *next; + + list_for_each_entry_safe(page, next, &b->refused_pages, lru) { + list_del(&page->lru); + __free_page(page); + STATS_INC(b->stats.refused_free); + } + + b->n_refused_pages = 0; +} + +/* + * Inflate the balloon towards its target size. Note that we try to limit + * the rate of allocation to make sure we are not choking the rest of the + * system. + */ +static void vmballoon_inflate(struct vmballoon *b) +{ + unsigned int goal; + unsigned int rate; + unsigned int i; + unsigned int allocations = 0; + int error = 0; + bool alloc_can_sleep = false; + + pr_debug("%s - size: %d, target %d\n", __func__, b->size, b->target); + + /* + * First try NOSLEEP page allocations to inflate balloon. + * + * If we do not throttle nosleep allocations, we can drain all + * free pages in the guest quickly (if the balloon target is high). + * As a side-effect, draining free pages helps to inform (force) + * the guest to start swapping if balloon target is not met yet, + * which is a desired behavior. However, balloon driver can consume + * all available CPU cycles if too many pages are allocated in a + * second. Therefore, we throttle nosleep allocations even when + * the guest is not under memory pressure. OTOH, if we have already + * predicted that the guest is under memory pressure, then we + * slowdown page allocations considerably. + */ + + goal = b->target - b->size; + /* + * Start with no sleep allocation rate which may be higher + * than sleeping allocation rate. + */ + rate = b->slow_allocation_cycles ? + b->rate_alloc : VMW_BALLOON_NOSLEEP_ALLOC_MAX; + + pr_debug("%s - goal: %d, no-sleep rate: %d, sleep rate: %d\n", + __func__, goal, rate, b->rate_alloc); + + for (i = 0; i < goal; i++) { + + error = vmballoon_reserve_page(b, alloc_can_sleep); + if (error) { + if (error != -ENOMEM) { + /* + * Not a page allocation failure, stop this + * cycle. Maybe we'll get new target from + * the host soon. + */ + break; + } + + if (alloc_can_sleep) { + /* + * CANSLEEP page allocation failed, so guest + * is under severe memory pressure. Quickly + * decrease allocation rate. + */ + b->rate_alloc = max(b->rate_alloc / 2, + VMW_BALLOON_RATE_ALLOC_MIN); + break; + } + + /* + * NOSLEEP page allocation failed, so the guest is + * under memory pressure. Let us slow down page + * allocations for next few cycles so that the guest + * gets out of memory pressure. Also, if we already + * allocated b->rate_alloc pages, let's pause, + * otherwise switch to sleeping allocations. + */ + b->slow_allocation_cycles = VMW_BALLOON_SLOW_CYCLES; + + if (i >= b->rate_alloc) + break; + + alloc_can_sleep = true; + /* Lower rate for sleeping allocations. */ + rate = b->rate_alloc; + } + + if (++allocations > VMW_BALLOON_YIELD_THRESHOLD) { + cond_resched(); + allocations = 0; + } + + if (i >= rate) { + /* We allocated enough pages, let's take a break. */ + break; + } + } + + /* + * We reached our goal without failures so try increasing + * allocation rate. + */ + if (error == 0 && i >= b->rate_alloc) { + unsigned int mult = i / b->rate_alloc; + + b->rate_alloc = + min(b->rate_alloc + mult * VMW_BALLOON_RATE_ALLOC_INC, + VMW_BALLOON_RATE_ALLOC_MAX); + } + + vmballoon_release_refused_pages(b); +} + +/* + * Decrease the size of the balloon allowing guest to use more memory. + */ +static void vmballoon_deflate(struct vmballoon *b) +{ + struct page *page, *next; + unsigned int i = 0; + unsigned int goal; + int error; + + pr_debug("%s - size: %d, target %d\n", __func__, b->size, b->target); + + /* limit deallocation rate */ + goal = min(b->size - b->target, b->rate_free); + + pr_debug("%s - goal: %d, rate: %d\n", __func__, goal, b->rate_free); + + /* free pages to reach target */ + list_for_each_entry_safe(page, next, &b->pages, lru) { + error = vmballoon_release_page(b, page); + if (error) { + /* quickly decrease rate in case of error */ + b->rate_free = max(b->rate_free / 2, + VMW_BALLOON_RATE_FREE_MIN); + return; + } + + if (++i >= goal) + break; + } + + /* slowly increase rate if there were no errors */ + b->rate_free = min(b->rate_free + VMW_BALLOON_RATE_FREE_INC, + VMW_BALLOON_RATE_FREE_MAX); +} + +/* + * Balloon work function: reset protocol, if needed, get the new size and + * adjust balloon as needed. Repeat in 1 sec. + */ +static void vmballoon_work(struct work_struct *work) +{ + struct delayed_work *dwork = to_delayed_work(work); + struct vmballoon *b = container_of(dwork, struct vmballoon, dwork); + unsigned int target; + + STATS_INC(b->stats.timer); + + if (b->reset_required) + vmballoon_reset(b); + + if (b->slow_allocation_cycles > 0) + b->slow_allocation_cycles--; + + if (vmballoon_send_get_target(b, &target)) { + /* update target, adjust size */ + b->target = target; + + if (b->size < target) + vmballoon_inflate(b); + else if (b->size > target) + vmballoon_deflate(b); + } + + queue_delayed_work(vmballoon_wq, dwork, round_jiffies_relative(HZ)); +} + +/* + * DEBUGFS Interface + */ +#ifdef CONFIG_DEBUG_FS + +static int vmballoon_debug_show(struct seq_file *f, void *offset) +{ + struct vmballoon *b = f->private; + struct vmballoon_stats *stats = &b->stats; + + /* format size info */ + seq_printf(f, + "target: %8d pages\n" + "current: %8d pages\n", + b->target, b->size); + + /* format rate info */ + seq_printf(f, + "rateNoSleepAlloc: %8d pages/sec\n" + "rateSleepAlloc: %8d pages/sec\n" + "rateFree: %8d pages/sec\n", + VMW_BALLOON_NOSLEEP_ALLOC_MAX, + b->rate_alloc, b->rate_free); + + seq_printf(f, + "\n" + "timer: %8u\n" + "start: %8u (%4u failed)\n" + "guestType: %8u (%4u failed)\n" + "lock: %8u (%4u failed)\n" + "unlock: %8u (%4u failed)\n" + "target: %8u (%4u failed)\n" + "primNoSleepAlloc: %8u (%4u failed)\n" + "primCanSleepAlloc: %8u (%4u failed)\n" + "primFree: %8u\n" + "errAlloc: %8u\n" + "errFree: %8u\n", + stats->timer, + stats->start, stats->start_fail, + stats->guest_type, stats->guest_type_fail, + stats->lock, stats->lock_fail, + stats->unlock, stats->unlock_fail, + stats->target, stats->target_fail, + stats->alloc, stats->alloc_fail, + stats->sleep_alloc, stats->sleep_alloc_fail, + stats->free, + stats->refused_alloc, stats->refused_free); + + return 0; +} + +static int vmballoon_debug_open(struct inode *inode, struct file *file) +{ + return single_open(file, vmballoon_debug_show, inode->i_private); +} + +static const struct file_operations vmballoon_debug_fops = { + .owner = THIS_MODULE, + .open = vmballoon_debug_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init vmballoon_debugfs_init(struct vmballoon *b) +{ + int error; + + b->dbg_entry = debugfs_create_file("vmmemctl", S_IRUGO, NULL, b, + &vmballoon_debug_fops); + if (IS_ERR(b->dbg_entry)) { + error = PTR_ERR(b->dbg_entry); + pr_err("failed to create debugfs entry, error: %d\n", error); + return error; + } + + return 0; +} + +static void __exit vmballoon_debugfs_exit(struct vmballoon *b) +{ + debugfs_remove(b->dbg_entry); +} + +#else + +static inline int vmballoon_debugfs_init(struct vmballoon *b) +{ + return 0; +} + +static inline void vmballoon_debugfs_exit(struct vmballoon *b) +{ +} + +#endif /* CONFIG_DEBUG_FS */ + +static int __init vmballoon_init(void) +{ + int error; + + /* + * Check if we are running on VMware's hypervisor and bail out + * if we are not. + */ + if (x86_hyper != &x86_hyper_vmware) + return -ENODEV; + + vmballoon_wq = create_freezeable_workqueue("vmmemctl"); + if (!vmballoon_wq) { + pr_err("failed to create workqueue\n"); + return -ENOMEM; + } + + INIT_LIST_HEAD(&balloon.pages); + INIT_LIST_HEAD(&balloon.refused_pages); + + /* initialize rates */ + balloon.rate_alloc = VMW_BALLOON_RATE_ALLOC_MAX; + balloon.rate_free = VMW_BALLOON_RATE_FREE_MAX; + + INIT_DELAYED_WORK(&balloon.dwork, vmballoon_work); + + /* + * Start balloon. + */ + if (!vmballoon_send_start(&balloon)) { + pr_err("failed to send start command to the host\n"); + error = -EIO; + goto fail; + } + + if (!vmballoon_send_guest_id(&balloon)) { + pr_err("failed to send guest ID to the host\n"); + error = -EIO; + goto fail; + } + + error = vmballoon_debugfs_init(&balloon); + if (error) + goto fail; + + queue_delayed_work(vmballoon_wq, &balloon.dwork, 0); + + return 0; + +fail: + destroy_workqueue(vmballoon_wq); + return error; +} +module_init(vmballoon_init); + +static void __exit vmballoon_exit(void) +{ + cancel_delayed_work_sync(&balloon.dwork); + destroy_workqueue(vmballoon_wq); + + vmballoon_debugfs_exit(&balloon); + + /* + * Deallocate all reserved memory, and reset connection with monitor. + * Reset connection before deallocating memory to avoid potential for + * additional spurious resets from guest touching deallocated pages. + */ + vmballoon_send_start(&balloon); + vmballoon_pop(&balloon); +} +module_exit(vmballoon_exit); diff --git a/drivers/misc/vmware_balloon.c b/drivers/misc/vmware_balloon.c deleted file mode 100644 index 2a1e804..0000000 --- a/drivers/misc/vmware_balloon.c +++ /dev/null @@ -1,844 +0,0 @@ -/* - * VMware Balloon driver. - * - * Copyright (C) 2000-2010, VMware, Inc. All Rights Reserved. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms of the GNU General Public License as published by the - * Free Software Foundation; version 2 of the License and no later version. - * - * This program is distributed in the hope that it will be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or - * NON INFRINGEMENT. See the GNU General Public License for more - * details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. - * - * Maintained by: Dmitry Torokhov - */ - -/* - * This is VMware physical memory management driver for Linux. The driver - * acts like a "balloon" that can be inflated to reclaim physical pages by - * reserving them in the guest and invalidating them in the monitor, - * freeing up the underlying machine pages so they can be allocated to - * other guests. The balloon can also be deflated to allow the guest to - * use more physical memory. Higher level policies can control the sizes - * of balloons in VMs in order to manage physical memory resources. - */ - -//#define DEBUG -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt - -#include -#include -#include -#include -#include -#include -#include -#include -#include - -MODULE_AUTHOR("VMware, Inc."); -MODULE_DESCRIPTION("VMware Memory Control (Balloon) Driver"); -MODULE_VERSION("1.2.1.1-k"); -MODULE_ALIAS("dmi:*:svnVMware*:*"); -MODULE_ALIAS("vmware_vmmemctl"); -MODULE_LICENSE("GPL"); - -/* - * Various constants controlling rate of inflaint/deflating balloon, - * measured in pages. - */ - -/* - * Rate of allocating memory when there is no memory pressure - * (driver performs non-sleeping allocations). - */ -#define VMW_BALLOON_NOSLEEP_ALLOC_MAX 16384U - -/* - * Rates of memory allocaton when guest experiences memory pressure - * (driver performs sleeping allocations). - */ -#define VMW_BALLOON_RATE_ALLOC_MIN 512U -#define VMW_BALLOON_RATE_ALLOC_MAX 2048U -#define VMW_BALLOON_RATE_ALLOC_INC 16U - -/* - * Rates for releasing pages while deflating balloon. - */ -#define VMW_BALLOON_RATE_FREE_MIN 512U -#define VMW_BALLOON_RATE_FREE_MAX 16384U -#define VMW_BALLOON_RATE_FREE_INC 16U - -/* - * When guest is under memory pressure, use a reduced page allocation - * rate for next several cycles. - */ -#define VMW_BALLOON_SLOW_CYCLES 4 - -/* - * Use __GFP_HIGHMEM to allow pages from HIGHMEM zone. We don't - * allow wait (__GFP_WAIT) for NOSLEEP page allocations. Use - * __GFP_NOWARN, to suppress page allocation failure warnings. - */ -#define VMW_PAGE_ALLOC_NOSLEEP (__GFP_HIGHMEM|__GFP_NOWARN) - -/* - * Use GFP_HIGHUSER when executing in a separate kernel thread - * context and allocation can sleep. This is less stressful to - * the guest memory system, since it allows the thread to block - * while memory is reclaimed, and won't take pages from emergency - * low-memory pools. - */ -#define VMW_PAGE_ALLOC_CANSLEEP (GFP_HIGHUSER) - -/* Maximum number of page allocations without yielding processor */ -#define VMW_BALLOON_YIELD_THRESHOLD 1024 - -/* Maximum number of refused pages we accumulate during inflation cycle */ -#define VMW_BALLOON_MAX_REFUSED 16 - -/* - * Hypervisor communication port definitions. - */ -#define VMW_BALLOON_HV_PORT 0x5670 -#define VMW_BALLOON_HV_MAGIC 0x456c6d6f -#define VMW_BALLOON_PROTOCOL_VERSION 2 -#define VMW_BALLOON_GUEST_ID 1 /* Linux */ - -#define VMW_BALLOON_CMD_START 0 -#define VMW_BALLOON_CMD_GET_TARGET 1 -#define VMW_BALLOON_CMD_LOCK 2 -#define VMW_BALLOON_CMD_UNLOCK 3 -#define VMW_BALLOON_CMD_GUEST_ID 4 - -/* error codes */ -#define VMW_BALLOON_SUCCESS 0 -#define VMW_BALLOON_FAILURE -1 -#define VMW_BALLOON_ERROR_CMD_INVALID 1 -#define VMW_BALLOON_ERROR_PPN_INVALID 2 -#define VMW_BALLOON_ERROR_PPN_LOCKED 3 -#define VMW_BALLOON_ERROR_PPN_UNLOCKED 4 -#define VMW_BALLOON_ERROR_PPN_PINNED 5 -#define VMW_BALLOON_ERROR_PPN_NOTNEEDED 6 -#define VMW_BALLOON_ERROR_RESET 7 -#define VMW_BALLOON_ERROR_BUSY 8 - -#define VMWARE_BALLOON_CMD(cmd, data, result) \ -({ \ - unsigned long __stat, __dummy1, __dummy2; \ - __asm__ __volatile__ ("inl (%%dx)" : \ - "=a"(__stat), \ - "=c"(__dummy1), \ - "=d"(__dummy2), \ - "=b"(result) : \ - "0"(VMW_BALLOON_HV_MAGIC), \ - "1"(VMW_BALLOON_CMD_##cmd), \ - "2"(VMW_BALLOON_HV_PORT), \ - "3"(data) : \ - "memory"); \ - result &= -1UL; \ - __stat & -1UL; \ -}) - -#ifdef CONFIG_DEBUG_FS -struct vmballoon_stats { - unsigned int timer; - - /* allocation statustics */ - unsigned int alloc; - unsigned int alloc_fail; - unsigned int sleep_alloc; - unsigned int sleep_alloc_fail; - unsigned int refused_alloc; - unsigned int refused_free; - unsigned int free; - - /* monitor operations */ - unsigned int lock; - unsigned int lock_fail; - unsigned int unlock; - unsigned int unlock_fail; - unsigned int target; - unsigned int target_fail; - unsigned int start; - unsigned int start_fail; - unsigned int guest_type; - unsigned int guest_type_fail; -}; - -#define STATS_INC(stat) (stat)++ -#else -#define STATS_INC(stat) -#endif - -struct vmballoon { - - /* list of reserved physical pages */ - struct list_head pages; - - /* transient list of non-balloonable pages */ - struct list_head refused_pages; - unsigned int n_refused_pages; - - /* balloon size in pages */ - unsigned int size; - unsigned int target; - - /* reset flag */ - bool reset_required; - - /* adjustment rates (pages per second) */ - unsigned int rate_alloc; - unsigned int rate_free; - - /* slowdown page allocations for next few cycles */ - unsigned int slow_allocation_cycles; - -#ifdef CONFIG_DEBUG_FS - /* statistics */ - struct vmballoon_stats stats; - - /* debugfs file exporting statistics */ - struct dentry *dbg_entry; -#endif - - struct sysinfo sysinfo; - - struct delayed_work dwork; -}; - -static struct vmballoon balloon; -static struct workqueue_struct *vmballoon_wq; - -/* - * Send "start" command to the host, communicating supported version - * of the protocol. - */ -static bool vmballoon_send_start(struct vmballoon *b) -{ - unsigned long status, dummy; - - STATS_INC(b->stats.start); - - status = VMWARE_BALLOON_CMD(START, VMW_BALLOON_PROTOCOL_VERSION, dummy); - if (status == VMW_BALLOON_SUCCESS) - return true; - - pr_debug("%s - failed, hv returns %ld\n", __func__, status); - STATS_INC(b->stats.start_fail); - return false; -} - -static bool vmballoon_check_status(struct vmballoon *b, unsigned long status) -{ - switch (status) { - case VMW_BALLOON_SUCCESS: - return true; - - case VMW_BALLOON_ERROR_RESET: - b->reset_required = true; - /* fall through */ - - default: - return false; - } -} - -/* - * Communicate guest type to the host so that it can adjust ballooning - * algorithm to the one most appropriate for the guest. This command - * is normally issued after sending "start" command and is part of - * standard reset sequence. - */ -static bool vmballoon_send_guest_id(struct vmballoon *b) -{ - unsigned long status, dummy; - - status = VMWARE_BALLOON_CMD(GUEST_ID, VMW_BALLOON_GUEST_ID, dummy); - - STATS_INC(b->stats.guest_type); - - if (vmballoon_check_status(b, status)) - return true; - - pr_debug("%s - failed, hv returns %ld\n", __func__, status); - STATS_INC(b->stats.guest_type_fail); - return false; -} - -/* - * Retrieve desired balloon size from the host. - */ -static bool vmballoon_send_get_target(struct vmballoon *b, u32 *new_target) -{ - unsigned long status; - unsigned long target; - unsigned long limit; - u32 limit32; - - /* - * si_meminfo() is cheap. Moreover, we want to provide dynamic - * max balloon size later. So let us call si_meminfo() every - * iteration. - */ - si_meminfo(&b->sysinfo); - limit = b->sysinfo.totalram; - - /* Ensure limit fits in 32-bits */ - limit32 = (u32)limit; - if (limit != limit32) - return false; - - /* update stats */ - STATS_INC(b->stats.target); - - status = VMWARE_BALLOON_CMD(GET_TARGET, limit, target); - if (vmballoon_check_status(b, status)) { - *new_target = target; - return true; - } - - pr_debug("%s - failed, hv returns %ld\n", __func__, status); - STATS_INC(b->stats.target_fail); - return false; -} - -/* - * Notify the host about allocated page so that host can use it without - * fear that guest will need it. Host may reject some pages, we need to - * check the return value and maybe submit a different page. - */ -static bool vmballoon_send_lock_page(struct vmballoon *b, unsigned long pfn) -{ - unsigned long status, dummy; - u32 pfn32; - - pfn32 = (u32)pfn; - if (pfn32 != pfn) - return false; - - STATS_INC(b->stats.lock); - - status = VMWARE_BALLOON_CMD(LOCK, pfn, dummy); - if (vmballoon_check_status(b, status)) - return true; - - pr_debug("%s - ppn %lx, hv returns %ld\n", __func__, pfn, status); - STATS_INC(b->stats.lock_fail); - return false; -} - -/* - * Notify the host that guest intends to release given page back into - * the pool of available (to the guest) pages. - */ -static bool vmballoon_send_unlock_page(struct vmballoon *b, unsigned long pfn) -{ - unsigned long status, dummy; - u32 pfn32; - - pfn32 = (u32)pfn; - if (pfn32 != pfn) - return false; - - STATS_INC(b->stats.unlock); - - status = VMWARE_BALLOON_CMD(UNLOCK, pfn, dummy); - if (vmballoon_check_status(b, status)) - return true; - - pr_debug("%s - ppn %lx, hv returns %ld\n", __func__, pfn, status); - STATS_INC(b->stats.unlock_fail); - return false; -} - -/* - * Quickly release all pages allocated for the balloon. This function is - * called when host decides to "reset" balloon for one reason or another. - * Unlike normal "deflate" we do not (shall not) notify host of the pages - * being released. - */ -static void vmballoon_pop(struct vmballoon *b) -{ - struct page *page, *next; - unsigned int count = 0; - - list_for_each_entry_safe(page, next, &b->pages, lru) { - list_del(&page->lru); - __free_page(page); - STATS_INC(b->stats.free); - b->size--; - - if (++count >= b->rate_free) { - count = 0; - cond_resched(); - } - } -} - -/* - * Perform standard reset sequence by popping the balloon (in case it - * is not empty) and then restarting protocol. This operation normally - * happens when host responds with VMW_BALLOON_ERROR_RESET to a command. - */ -static void vmballoon_reset(struct vmballoon *b) -{ - /* free all pages, skipping monitor unlock */ - vmballoon_pop(b); - - if (vmballoon_send_start(b)) { - b->reset_required = false; - if (!vmballoon_send_guest_id(b)) - pr_err("failed to send guest ID to the host\n"); - } -} - -/* - * Allocate (or reserve) a page for the balloon and notify the host. If host - * refuses the page put it on "refuse" list and allocate another one until host - * is satisfied. "Refused" pages are released at the end of inflation cycle - * (when we allocate b->rate_alloc pages). - */ -static int vmballoon_reserve_page(struct vmballoon *b, bool can_sleep) -{ - struct page *page; - gfp_t flags; - bool locked = false; - - do { - if (!can_sleep) - STATS_INC(b->stats.alloc); - else - STATS_INC(b->stats.sleep_alloc); - - flags = can_sleep ? VMW_PAGE_ALLOC_CANSLEEP : VMW_PAGE_ALLOC_NOSLEEP; - page = alloc_page(flags); - if (!page) { - if (!can_sleep) - STATS_INC(b->stats.alloc_fail); - else - STATS_INC(b->stats.sleep_alloc_fail); - return -ENOMEM; - } - - /* inform monitor */ - locked = vmballoon_send_lock_page(b, page_to_pfn(page)); - if (!locked) { - STATS_INC(b->stats.refused_alloc); - - if (b->reset_required) { - __free_page(page); - return -EIO; - } - - /* - * Place page on the list of non-balloonable pages - * and retry allocation, unless we already accumulated - * too many of them, in which case take a breather. - */ - list_add(&page->lru, &b->refused_pages); - if (++b->n_refused_pages >= VMW_BALLOON_MAX_REFUSED) - return -EIO; - } - } while (!locked); - - /* track allocated page */ - list_add(&page->lru, &b->pages); - - /* update balloon size */ - b->size++; - - return 0; -} - -/* - * Release the page allocated for the balloon. Note that we first notify - * the host so it can make sure the page will be available for the guest - * to use, if needed. - */ -static int vmballoon_release_page(struct vmballoon *b, struct page *page) -{ - if (!vmballoon_send_unlock_page(b, page_to_pfn(page))) - return -EIO; - - list_del(&page->lru); - - /* deallocate page */ - __free_page(page); - STATS_INC(b->stats.free); - - /* update balloon size */ - b->size--; - - return 0; -} - -/* - * Release pages that were allocated while attempting to inflate the - * balloon but were refused by the host for one reason or another. - */ -static void vmballoon_release_refused_pages(struct vmballoon *b) -{ - struct page *page, *next; - - list_for_each_entry_safe(page, next, &b->refused_pages, lru) { - list_del(&page->lru); - __free_page(page); - STATS_INC(b->stats.refused_free); - } - - b->n_refused_pages = 0; -} - -/* - * Inflate the balloon towards its target size. Note that we try to limit - * the rate of allocation to make sure we are not choking the rest of the - * system. - */ -static void vmballoon_inflate(struct vmballoon *b) -{ - unsigned int goal; - unsigned int rate; - unsigned int i; - unsigned int allocations = 0; - int error = 0; - bool alloc_can_sleep = false; - - pr_debug("%s - size: %d, target %d\n", __func__, b->size, b->target); - - /* - * First try NOSLEEP page allocations to inflate balloon. - * - * If we do not throttle nosleep allocations, we can drain all - * free pages in the guest quickly (if the balloon target is high). - * As a side-effect, draining free pages helps to inform (force) - * the guest to start swapping if balloon target is not met yet, - * which is a desired behavior. However, balloon driver can consume - * all available CPU cycles if too many pages are allocated in a - * second. Therefore, we throttle nosleep allocations even when - * the guest is not under memory pressure. OTOH, if we have already - * predicted that the guest is under memory pressure, then we - * slowdown page allocations considerably. - */ - - goal = b->target - b->size; - /* - * Start with no sleep allocation rate which may be higher - * than sleeping allocation rate. - */ - rate = b->slow_allocation_cycles ? - b->rate_alloc : VMW_BALLOON_NOSLEEP_ALLOC_MAX; - - pr_debug("%s - goal: %d, no-sleep rate: %d, sleep rate: %d\n", - __func__, goal, rate, b->rate_alloc); - - for (i = 0; i < goal; i++) { - - error = vmballoon_reserve_page(b, alloc_can_sleep); - if (error) { - if (error != -ENOMEM) { - /* - * Not a page allocation failure, stop this - * cycle. Maybe we'll get new target from - * the host soon. - */ - break; - } - - if (alloc_can_sleep) { - /* - * CANSLEEP page allocation failed, so guest - * is under severe memory pressure. Quickly - * decrease allocation rate. - */ - b->rate_alloc = max(b->rate_alloc / 2, - VMW_BALLOON_RATE_ALLOC_MIN); - break; - } - - /* - * NOSLEEP page allocation failed, so the guest is - * under memory pressure. Let us slow down page - * allocations for next few cycles so that the guest - * gets out of memory pressure. Also, if we already - * allocated b->rate_alloc pages, let's pause, - * otherwise switch to sleeping allocations. - */ - b->slow_allocation_cycles = VMW_BALLOON_SLOW_CYCLES; - - if (i >= b->rate_alloc) - break; - - alloc_can_sleep = true; - /* Lower rate for sleeping allocations. */ - rate = b->rate_alloc; - } - - if (++allocations > VMW_BALLOON_YIELD_THRESHOLD) { - cond_resched(); - allocations = 0; - } - - if (i >= rate) { - /* We allocated enough pages, let's take a break. */ - break; - } - } - - /* - * We reached our goal without failures so try increasing - * allocation rate. - */ - if (error == 0 && i >= b->rate_alloc) { - unsigned int mult = i / b->rate_alloc; - - b->rate_alloc = - min(b->rate_alloc + mult * VMW_BALLOON_RATE_ALLOC_INC, - VMW_BALLOON_RATE_ALLOC_MAX); - } - - vmballoon_release_refused_pages(b); -} - -/* - * Decrease the size of the balloon allowing guest to use more memory. - */ -static void vmballoon_deflate(struct vmballoon *b) -{ - struct page *page, *next; - unsigned int i = 0; - unsigned int goal; - int error; - - pr_debug("%s - size: %d, target %d\n", __func__, b->size, b->target); - - /* limit deallocation rate */ - goal = min(b->size - b->target, b->rate_free); - - pr_debug("%s - goal: %d, rate: %d\n", __func__, goal, b->rate_free); - - /* free pages to reach target */ - list_for_each_entry_safe(page, next, &b->pages, lru) { - error = vmballoon_release_page(b, page); - if (error) { - /* quickly decrease rate in case of error */ - b->rate_free = max(b->rate_free / 2, - VMW_BALLOON_RATE_FREE_MIN); - return; - } - - if (++i >= goal) - break; - } - - /* slowly increase rate if there were no errors */ - b->rate_free = min(b->rate_free + VMW_BALLOON_RATE_FREE_INC, - VMW_BALLOON_RATE_FREE_MAX); -} - -/* - * Balloon work function: reset protocol, if needed, get the new size and - * adjust balloon as needed. Repeat in 1 sec. - */ -static void vmballoon_work(struct work_struct *work) -{ - struct delayed_work *dwork = to_delayed_work(work); - struct vmballoon *b = container_of(dwork, struct vmballoon, dwork); - unsigned int target; - - STATS_INC(b->stats.timer); - - if (b->reset_required) - vmballoon_reset(b); - - if (b->slow_allocation_cycles > 0) - b->slow_allocation_cycles--; - - if (vmballoon_send_get_target(b, &target)) { - /* update target, adjust size */ - b->target = target; - - if (b->size < target) - vmballoon_inflate(b); - else if (b->size > target) - vmballoon_deflate(b); - } - - queue_delayed_work(vmballoon_wq, dwork, round_jiffies_relative(HZ)); -} - -/* - * DEBUGFS Interface - */ -#ifdef CONFIG_DEBUG_FS - -static int vmballoon_debug_show(struct seq_file *f, void *offset) -{ - struct vmballoon *b = f->private; - struct vmballoon_stats *stats = &b->stats; - - /* format size info */ - seq_printf(f, - "target: %8d pages\n" - "current: %8d pages\n", - b->target, b->size); - - /* format rate info */ - seq_printf(f, - "rateNoSleepAlloc: %8d pages/sec\n" - "rateSleepAlloc: %8d pages/sec\n" - "rateFree: %8d pages/sec\n", - VMW_BALLOON_NOSLEEP_ALLOC_MAX, - b->rate_alloc, b->rate_free); - - seq_printf(f, - "\n" - "timer: %8u\n" - "start: %8u (%4u failed)\n" - "guestType: %8u (%4u failed)\n" - "lock: %8u (%4u failed)\n" - "unlock: %8u (%4u failed)\n" - "target: %8u (%4u failed)\n" - "primNoSleepAlloc: %8u (%4u failed)\n" - "primCanSleepAlloc: %8u (%4u failed)\n" - "primFree: %8u\n" - "errAlloc: %8u\n" - "errFree: %8u\n", - stats->timer, - stats->start, stats->start_fail, - stats->guest_type, stats->guest_type_fail, - stats->lock, stats->lock_fail, - stats->unlock, stats->unlock_fail, - stats->target, stats->target_fail, - stats->alloc, stats->alloc_fail, - stats->sleep_alloc, stats->sleep_alloc_fail, - stats->free, - stats->refused_alloc, stats->refused_free); - - return 0; -} - -static int vmballoon_debug_open(struct inode *inode, struct file *file) -{ - return single_open(file, vmballoon_debug_show, inode->i_private); -} - -static const struct file_operations vmballoon_debug_fops = { - .owner = THIS_MODULE, - .open = vmballoon_debug_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static int __init vmballoon_debugfs_init(struct vmballoon *b) -{ - int error; - - b->dbg_entry = debugfs_create_file("vmmemctl", S_IRUGO, NULL, b, - &vmballoon_debug_fops); - if (IS_ERR(b->dbg_entry)) { - error = PTR_ERR(b->dbg_entry); - pr_err("failed to create debugfs entry, error: %d\n", error); - return error; - } - - return 0; -} - -static void __exit vmballoon_debugfs_exit(struct vmballoon *b) -{ - debugfs_remove(b->dbg_entry); -} - -#else - -static inline int vmballoon_debugfs_init(struct vmballoon *b) -{ - return 0; -} - -static inline void vmballoon_debugfs_exit(struct vmballoon *b) -{ -} - -#endif /* CONFIG_DEBUG_FS */ - -static int __init vmballoon_init(void) -{ - int error; - - /* - * Check if we are running on VMware's hypervisor and bail out - * if we are not. - */ - if (x86_hyper != &x86_hyper_vmware) - return -ENODEV; - - vmballoon_wq = create_freezeable_workqueue("vmmemctl"); - if (!vmballoon_wq) { - pr_err("failed to create workqueue\n"); - return -ENOMEM; - } - - INIT_LIST_HEAD(&balloon.pages); - INIT_LIST_HEAD(&balloon.refused_pages); - - /* initialize rates */ - balloon.rate_alloc = VMW_BALLOON_RATE_ALLOC_MAX; - balloon.rate_free = VMW_BALLOON_RATE_FREE_MAX; - - INIT_DELAYED_WORK(&balloon.dwork, vmballoon_work); - - /* - * Start balloon. - */ - if (!vmballoon_send_start(&balloon)) { - pr_err("failed to send start command to the host\n"); - error = -EIO; - goto fail; - } - - if (!vmballoon_send_guest_id(&balloon)) { - pr_err("failed to send guest ID to the host\n"); - error = -EIO; - goto fail; - } - - error = vmballoon_debugfs_init(&balloon); - if (error) - goto fail; - - queue_delayed_work(vmballoon_wq, &balloon.dwork, 0); - - return 0; - -fail: - destroy_workqueue(vmballoon_wq); - return error; -} -module_init(vmballoon_init); - -static void __exit vmballoon_exit(void) -{ - cancel_delayed_work_sync(&balloon.dwork); - destroy_workqueue(vmballoon_wq); - - vmballoon_debugfs_exit(&balloon); - - /* - * Deallocate all reserved memory, and reset connection with monitor. - * Reset connection before deallocating memory to avoid potential for - * additional spurious resets from guest touching deallocated pages. - */ - vmballoon_send_start(&balloon); - vmballoon_pop(&balloon); -} -module_exit(vmballoon_exit); -- cgit v0.10.2 From f5665518c20c01e9045314872878f5788cb3fff2 Mon Sep 17 00:00:00 2001 From: Hans-Christian Egtvedt Date: Wed, 22 Sep 2010 13:04:58 -0700 Subject: MAINTAINERS: change AVR32 and AT32AP maintainer Alter the maintainer of the AVR32 architecture and the AVR32/AT32AP machine support to me. Haavard is moving on to new challenges, and we've found it better to transfer the maintainer part to me. I will have good contact with Haavard anyway. Signed-off-by: Hans-Christian Egtvedt Acked-by: Haavard Skinnemoen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 50b8148..726433a1 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1220,7 +1220,7 @@ F: drivers/auxdisplay/ F: include/linux/cfag12864b.h AVR32 ARCHITECTURE -M: Haavard Skinnemoen +M: Hans-Christian Egtvedt W: http://www.atmel.com/products/AVR32/ W: http://avr32linux.org/ W: http://avrfreaks.net/ @@ -1228,7 +1228,7 @@ S: Supported F: arch/avr32/ AVR32/AT32AP MACHINE SUPPORT -M: Haavard Skinnemoen +M: Hans-Christian Egtvedt S: Supported F: arch/avr32/mach-at32ap/ -- cgit v0.10.2 From eba93fcc34d6c4387ce8fbb53bb7b685f91f3343 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 22 Sep 2010 13:04:59 -0700 Subject: drivers/rtc/rtc-ab3100.c: add missing platform_set_drvdata() in ab3100_rtc_probe() Otherwise, calling platform_get_drvdata() in ab3100_rtc_remove() returns NULL. Signed-off-by: Axel Lin Acked-by:Wan ZongShun Acked-by: Linus Walleij Cc: Alessandro Zummo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-ab3100.c b/drivers/rtc/rtc-ab3100.c index d26780e..261a07e 100644 --- a/drivers/rtc/rtc-ab3100.c +++ b/drivers/rtc/rtc-ab3100.c @@ -235,6 +235,7 @@ static int __init ab3100_rtc_probe(struct platform_device *pdev) err = PTR_ERR(rtc); return err; } + platform_set_drvdata(pdev, rtc); return 0; } @@ -244,6 +245,7 @@ static int __exit ab3100_rtc_remove(struct platform_device *pdev) struct rtc_device *rtc = platform_get_drvdata(pdev); rtc_device_unregister(rtc); + platform_set_drvdata(pdev, NULL); return 0; } -- cgit v0.10.2 From d1908362ae0b97374eb8328fbb471576332f9fb1 Mon Sep 17 00:00:00 2001 From: Minchan Kim Date: Wed, 22 Sep 2010 13:05:01 -0700 Subject: vmscan: check all_unreclaimable in direct reclaim path M. Vefa Bicakci reported 2.6.35 kernel hang up when hibernation on his 32bit 3GB mem machine. (https://bugzilla.kernel.org/show_bug.cgi?id=16771). Also he bisected the regression to commit bb21c7ce18eff8e6e7877ca1d06c6db719376e3c Author: KOSAKI Motohiro Date: Fri Jun 4 14:15:05 2010 -0700 vmscan: fix do_try_to_free_pages() return value when priority==0 reclaim failure At first impression, this seemed very strange because the above commit only chenged function return value and hibernate_preallocate_memory() ignore return value of shrink_all_memory(). But it's related. Now, page allocation from hibernation code may enter infinite loop if the system has highmem. The reasons are that vmscan don't care enough OOM case when oom_killer_disabled. The problem sequence is following as. 1. hibernation 2. oom_disable 3. alloc_pages 4. do_try_to_free_pages if (scanning_global_lru(sc) && !all_unreclaimable) return 1; If kswapd is not freozen, it would set zone->all_unreclaimable to 1 and then shrink_zones maybe return true(ie, all_unreclaimable is true). So at last, alloc_pages could go to _nopage_. If it is, it should have no problem. This patch adds all_unreclaimable check to protect in direct reclaim path, too. It can care of hibernation OOM case and help bailout all_unreclaimable case slightly. Signed-off-by: KOSAKI Motohiro Signed-off-by: Minchan Kim Reported-by: M. Vefa Bicakci Reported-by: Reviewed-by: Johannes Weiner Tested-by: Acked-by: Rafael J. Wysocki Acked-by: Rik van Riel Acked-by: KAMEZAWA Hiroyuki Cc: Balbir Singh Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/vmscan.c b/mm/vmscan.c index c391c32..c5dfabf 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -1804,12 +1804,11 @@ static void shrink_zone(int priority, struct zone *zone, * If a zone is deemed to be full of pinned pages then just give it a light * scan then give up on it. */ -static bool shrink_zones(int priority, struct zonelist *zonelist, +static void shrink_zones(int priority, struct zonelist *zonelist, struct scan_control *sc) { struct zoneref *z; struct zone *zone; - bool all_unreclaimable = true; for_each_zone_zonelist_nodemask(zone, z, zonelist, gfp_zone(sc->gfp_mask), sc->nodemask) { @@ -1827,8 +1826,38 @@ static bool shrink_zones(int priority, struct zonelist *zonelist, } shrink_zone(priority, zone, sc); - all_unreclaimable = false; } +} + +static bool zone_reclaimable(struct zone *zone) +{ + return zone->pages_scanned < zone_reclaimable_pages(zone) * 6; +} + +/* + * As hibernation is going on, kswapd is freezed so that it can't mark + * the zone into all_unreclaimable. It can't handle OOM during hibernation. + * So let's check zone's unreclaimable in direct reclaim as well as kswapd. + */ +static bool all_unreclaimable(struct zonelist *zonelist, + struct scan_control *sc) +{ + struct zoneref *z; + struct zone *zone; + bool all_unreclaimable = true; + + for_each_zone_zonelist_nodemask(zone, z, zonelist, + gfp_zone(sc->gfp_mask), sc->nodemask) { + if (!populated_zone(zone)) + continue; + if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) + continue; + if (zone_reclaimable(zone)) { + all_unreclaimable = false; + break; + } + } + return all_unreclaimable; } @@ -1852,7 +1881,6 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist, struct scan_control *sc) { int priority; - bool all_unreclaimable; unsigned long total_scanned = 0; struct reclaim_state *reclaim_state = current->reclaim_state; struct zoneref *z; @@ -1869,7 +1897,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist, sc->nr_scanned = 0; if (!priority) disable_swap_token(); - all_unreclaimable = shrink_zones(priority, zonelist, sc); + shrink_zones(priority, zonelist, sc); /* * Don't shrink slabs when reclaiming memory from * over limit cgroups @@ -1931,7 +1959,7 @@ out: return sc->nr_reclaimed; /* top priority shrink_zones still had more to do? don't OOM, then */ - if (scanning_global_lru(sc) && !all_unreclaimable) + if (scanning_global_lru(sc) && !all_unreclaimable(zonelist, sc)) return 1; return 0; @@ -2197,8 +2225,7 @@ loop_again: total_scanned += sc.nr_scanned; if (zone->all_unreclaimable) continue; - if (nr_slab == 0 && - zone->pages_scanned >= (zone_reclaimable_pages(zone) * 6)) + if (nr_slab == 0 && !zone_reclaimable(zone)) zone->all_unreclaimable = 1; /* * If we've done a decent amount of scanning and -- cgit v0.10.2 From a0c42bac79731276c9b2f28d54f9e658fcf843a2 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Sep 2010 13:05:03 -0700 Subject: aio: do not return ERESTARTSYS as a result of AIO OCFS2 can return ERESTARTSYS from its write function when the process is signalled while waiting for a cluster lock (and the filesystem is mounted with intr mount option). Generally, it seems reasonable to allow filesystems to return this error code from its IO functions. As we must not leak ERESTARTSYS (and similar error codes) to userspace as a result of an AIO operation, we have to properly convert it to EINTR inside AIO code (restarting the syscall isn't really an option because other AIO could have been already submitted by the same io_submit syscall). Signed-off-by: Jan Kara Reviewed-by: Jeff Moyer Cc: Christoph Hellwig Cc: Zach Brown Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/aio.c b/fs/aio.c index 1320b2a..250b0a7 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -712,8 +712,16 @@ static ssize_t aio_run_iocb(struct kiocb *iocb) */ ret = retry(iocb); - if (ret != -EIOCBRETRY && ret != -EIOCBQUEUED) + if (ret != -EIOCBRETRY && ret != -EIOCBQUEUED) { + /* + * There's no easy way to restart the syscall since other AIO's + * may be already running. Just fail this IO with EINTR. + */ + if (unlikely(ret == -ERESTARTSYS || ret == -ERESTARTNOINTR || + ret == -ERESTARTNOHAND || ret == -ERESTART_RESTARTBLOCK)) + ret = -EINTR; aio_complete(iocb, ret, 0); + } out: spin_lock_irq(&ctx->ctx_lock); -- cgit v0.10.2 From 85a00d9bbfb4704fbf368944b1cb9fed8f1598c5 Mon Sep 17 00:00:00 2001 From: Peter Jones Date: Wed, 22 Sep 2010 13:05:04 -0700 Subject: efifb: check that the base address is plausible on pci systems Some Apple machines have identical DMI data but different memory configurations for the video. Given that, check that the address in our table is actually within the range of a PCI BAR on a VGA device in the machine. This also fixes up the return value from set_system(), which has always been wrong, but never resulted in bad behavior since there's only ever been one matching entry in the dmi table. The patch 1) stops people's machines from crashing when we get their display wrong, which seems to be unfortunately inevitable, 2) allows us to support identical dmi data with differing video memory configurations This also adds me as the efifb maintainer, since I've effectively been acting as such for quite some time. Signed-off-by: Peter Jones Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 726433a1..4d4881d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2199,6 +2199,12 @@ W: http://acpi4asus.sf.net S: Maintained F: drivers/platform/x86/eeepc-laptop.c +EFIFB FRAMEBUFFER DRIVER +L: linux-fbdev@vger.kernel.org +M: Peter Jones +S: Maintained +F: drivers/video/efifb.c + EFS FILESYSTEM W: http://aeschi.ch.eu.org/efs/ S: Orphan diff --git a/drivers/video/efifb.c b/drivers/video/efifb.c index 815f84b..c082b61 100644 --- a/drivers/video/efifb.c +++ b/drivers/video/efifb.c @@ -13,7 +13,7 @@ #include #include #include - +#include #include