From c8ad620638f97bdb7c8ef8cc788f08a04b14eadc Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Mon, 29 Aug 2011 09:52:23 -0600 Subject: writeback: show raw dirtied_when in trace writeback_single_inode Save inode->dirtied_when in the raw trace output for reliable scripting, and to also show in formatted output the relative age in seconds for easy human reading. CC: Jan Kara Acked-by: Christoph Hellwig Signed-off-by: Wu Fengguang diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index 6bca4cc..5f17270 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -298,7 +298,7 @@ DECLARE_EVENT_CLASS(writeback_single_inode_template, __array(char, name, 32) __field(unsigned long, ino) __field(unsigned long, state) - __field(unsigned long, age) + __field(unsigned long, dirtied_when) __field(unsigned long, writeback_index) __field(long, nr_to_write) __field(unsigned long, wrote) @@ -309,19 +309,19 @@ DECLARE_EVENT_CLASS(writeback_single_inode_template, dev_name(inode->i_mapping->backing_dev_info->dev), 32); __entry->ino = inode->i_ino; __entry->state = inode->i_state; - __entry->age = (jiffies - inode->dirtied_when) * - 1000 / HZ; + __entry->dirtied_when = inode->dirtied_when; __entry->writeback_index = inode->i_mapping->writeback_index; __entry->nr_to_write = nr_to_write; __entry->wrote = nr_to_write - wbc->nr_to_write; ), - TP_printk("bdi %s: ino=%lu state=%s age=%lu " + TP_printk("bdi %s: ino=%lu state=%s dirtied_when=%lu age=%lu " "index=%lu to_write=%ld wrote=%lu", __entry->name, __entry->ino, show_inode_state(__entry->state), - __entry->age, + __entry->dirtied_when, + (jiffies - __entry->dirtied_when) / HZ, __entry->writeback_index, __entry->nr_to_write, __entry->wrote -- cgit v0.10.2 From 4cc122ac913dbd662403a4a173c41e0d59ab5bdd Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Thu, 15 Sep 2011 17:21:15 -0300 Subject: spi/imx: Fix spi-imx when the hardware SPI chipselects are used commit 22a85e4cd51 (spi/imx: add device tree probe support) broke spi-imx usage when the SPI chipselect is the one internal to the controller. On a mx31pdk board the following error is seen: Registering mxc_nand as whole device ------------[ cut here ]------------ WARNING: at drivers/gpio/gpiolib.c:101 gpio_ensure_requested+0x4c/0xf4() autorequest GPIO-0 Modules linked in: [] (unwind_backtrace+0x0/0xf4) from [] (warn_slowpath_common+0x4c/0x64) [] (warn_slowpath_common+0x4c/0x64) from [] (warn_slowpath_fmt+0x30/0x40) [] (warn_slowpath_fmt+0x30/0x40) from [] (gpio_ensure_requested+0x4c/0xf4) [] (gpio_ensure_requested+0x4c/0xf4) from [] (gpio_direction_output+0xa0/0x138) [] (gpio_direction_output+0xa0/0x138) from [] (spi_imx_setup+0x38/0x4c) [] (spi_imx_setup+0x38/0x4c) from [] (spi_setup+0x38/0x50) [] (spi_setup+0x38/0x50) from [] (spi_add_device+0x94/0x124) [] (spi_add_device+0x94/0x124) from [] (spi_new_device+0x74/0xac) [] (spi_new_device+0x74/0xac) from [] (spi_match_master_to_boardinfo+0x20/0x40) [] (spi_match_master_to_boardinfo+0x20/0x40) from [] (spi_register_master+0xb0/0x104) [] (spi_register_master+0xb0/0x104) from [] (spi_bitbang_start+0x104/0x17c) [] (spi_bitbang_start+0x104/0x17c) from [] (spi_imx_probe+0x2fc/0x404) [] (spi_imx_probe+0x2fc/0x404) from [] (platform_drv_probe+0x18/0x1c) [] (platform_drv_probe+0x18/0x1c) from [] (driver_probe_device+0x78/0x174) [] (driver_probe_device+0x78/0x174) from [] (__driver_attach+0x8c/0x90) [] (__driver_attach+0x8c/0x90) from [] (bus_for_each_dev+0x60/0x8c) [] (bus_for_each_dev+0x60/0x8c) from [] (bus_add_driver+0xa0/0x288) [] (bus_add_driver+0xa0/0x288) from [] (driver_register+0x78/0x18c) [] (driver_register+0x78/0x18c) from [] (do_one_initcall+0x34/0x178) [] (do_one_initcall+0x34/0x178) from [] (kernel_init+0x74/0x118) [] (kernel_init+0x74/0x118) from [] (kernel_thread_exit+0x0/0x8) ---[ end trace 759f924b30fd5a44 ]--- Fix this issue by using the original chip select logic and make spi-imx to work again. Tested on a mx31pdk that uses the hardware SPI chipselect pins and also on a mx27pdk that uses GPIO as SPI chipselect. Signed-off-by: Fabio Estevam Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c index 8ac6542..fa594d6 100644 --- a/drivers/spi/spi-imx.c +++ b/drivers/spi/spi-imx.c @@ -786,9 +786,11 @@ static int __devinit spi_imx_probe(struct platform_device *pdev) int cs_gpio = of_get_named_gpio(np, "cs-gpios", i); if (cs_gpio < 0) cs_gpio = mxc_platform_info->chipselect[i]; + + spi_imx->chipselect[i] = cs_gpio; if (cs_gpio < 0) continue; - spi_imx->chipselect[i] = cs_gpio; + ret = gpio_request(spi_imx->chipselect[i], DRIVER_NAME); if (ret) { while (i > 0) { -- cgit v0.10.2 From 4f6c7e159361b330d767b107226bcc6938abad16 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 15 Sep 2011 11:04:56 +0800 Subject: ASoC: bf5xx-ad73311: Fix prototype for bf5xx_probe Fix below build warning: sound/soc/blackfin/bf5xx-ad73311.c: warning: initialization from incompatible pointer type Signed-off-by: Axel Lin Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/blackfin/bf5xx-ad73311.c b/sound/soc/blackfin/bf5xx-ad73311.c index 732a247..b94eb7e 100644 --- a/sound/soc/blackfin/bf5xx-ad73311.c +++ b/sound/soc/blackfin/bf5xx-ad73311.c @@ -128,7 +128,7 @@ static int snd_ad73311_configure(void) return 0; } -static int bf5xx_probe(struct platform_device *pdev) +static int bf5xx_probe(struct snd_soc_card *card) { int err; if (gpio_request(GPIO_SE, "AD73311_SE")) { -- cgit v0.10.2 From c0d5f9db1c7d1b8a9e2f217706e8ea233bac2754 Mon Sep 17 00:00:00 2001 From: Jim Schutt Date: Fri, 16 Sep 2011 08:27:31 -0600 Subject: libceph: initialize ack_stamp to avoid unnecessary connection reset Commit 4cf9d544631c recorded when an outgoing ceph message was ACKed, in order to avoid unnecessary connection resets when an OSD is busy. However, ack_stamp is uninitialized, so there is a window between when the message is sent and when it is ACKed in which handle_timeout() interprets the unitialized value as an expired timeout, and resets the connection unnecessarily. Close the window by initializing ack_stamp. Signed-off-by: Jim Schutt Signed-off-by: Sage Weil diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c index c340e2e..9918e9e 100644 --- a/net/ceph/messenger.c +++ b/net/ceph/messenger.c @@ -2307,6 +2307,7 @@ struct ceph_msg *ceph_msg_new(int type, int front_len, gfp_t flags) m->front_max = front_len; m->front_is_vmalloc = false; m->more_to_follow = false; + m->ack_stamp = 0; m->pool = NULL; /* middle */ -- cgit v0.10.2 From 1cad78932a0d139dceff78e68808e160a224d57a Mon Sep 17 00:00:00 2001 From: Noah Watkins Date: Mon, 12 Sep 2011 14:51:53 -0700 Subject: libceph: fix parse options memory leak ceph_destroy_options does not free opt->mon_addr that is allocated in ceph_parse_options. Signed-off-by: Noah Watkins Signed-off-by: Sage Weil diff --git a/net/ceph/ceph_common.c b/net/ceph/ceph_common.c index 132963a..2883ea0 100644 --- a/net/ceph/ceph_common.c +++ b/net/ceph/ceph_common.c @@ -232,6 +232,7 @@ void ceph_destroy_options(struct ceph_options *opt) ceph_crypto_key_destroy(opt->key); kfree(opt->key); } + kfree(opt->mon_addr); kfree(opt); } EXPORT_SYMBOL(ceph_destroy_options); -- cgit v0.10.2 From 935b639a049053d0ccbcf7422f2f9cd221642f58 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 16 Sep 2011 11:13:17 -0700 Subject: libceph: fix linger request requeuing The r_req_lru_item list node moves between several lists, and that cycle is not directly related (and does not begin) with __register_request(). Initialize it in the request constructor, not __register_request(). This fixes later badness (below) when OSDs restart underneath an rbd mount. Crashes we've seen due to this include: [ 213.974288] kernel BUG at net/ceph/messenger.c:2193! and [ 144.035274] BUG: unable to handle kernel NULL pointer dereference at 0000000000000048 [ 144.035278] IP: [] con_work+0x1463/0x2ce0 [libceph] Signed-off-by: Sage Weil diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c index 16836a7..88ad8a2 100644 --- a/net/ceph/osd_client.c +++ b/net/ceph/osd_client.c @@ -217,6 +217,7 @@ struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc, INIT_LIST_HEAD(&req->r_unsafe_item); INIT_LIST_HEAD(&req->r_linger_item); INIT_LIST_HEAD(&req->r_linger_osd); + INIT_LIST_HEAD(&req->r_req_lru_item); req->r_flags = flags; WARN_ON((flags & (CEPH_OSD_FLAG_READ|CEPH_OSD_FLAG_WRITE)) == 0); @@ -816,13 +817,10 @@ static void __register_request(struct ceph_osd_client *osdc, { req->r_tid = ++osdc->last_tid; req->r_request->hdr.tid = cpu_to_le64(req->r_tid); - INIT_LIST_HEAD(&req->r_req_lru_item); - dout("__register_request %p tid %lld\n", req, req->r_tid); __insert_request(osdc, req); ceph_osdc_get_request(req); osdc->num_requests++; - if (osdc->num_requests == 1) { dout(" first request, scheduling timeout\n"); __schedule_osd_timeout(osdc); -- cgit v0.10.2 From 08cefa9fa7e5b3ddaefb8b7bfa408d148429c08d Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Sun, 11 Sep 2011 10:42:20 -0700 Subject: sparc64: Future proof Niagara cpu detection. Recognize T4 and T5 chips. Treating them both as "T2 plus other stuff" should be extremely safe and make sure distributions will work when those chips actually ship to customers. Signed-off-by: David S. Miller diff --git a/arch/sparc/include/asm/spitfire.h b/arch/sparc/include/asm/spitfire.h index 55a17c6..d06a2660 100644 --- a/arch/sparc/include/asm/spitfire.h +++ b/arch/sparc/include/asm/spitfire.h @@ -43,6 +43,8 @@ #define SUN4V_CHIP_NIAGARA1 0x01 #define SUN4V_CHIP_NIAGARA2 0x02 #define SUN4V_CHIP_NIAGARA3 0x03 +#define SUN4V_CHIP_NIAGARA4 0x04 +#define SUN4V_CHIP_NIAGARA5 0x05 #define SUN4V_CHIP_UNKNOWN 0xff #ifndef __ASSEMBLY__ diff --git a/arch/sparc/include/asm/xor_64.h b/arch/sparc/include/asm/xor_64.h index 9ed6ff6..ee8edc6 100644 --- a/arch/sparc/include/asm/xor_64.h +++ b/arch/sparc/include/asm/xor_64.h @@ -66,6 +66,8 @@ static struct xor_block_template xor_block_niagara = { ((tlb_type == hypervisor && \ (sun4v_chip_type == SUN4V_CHIP_NIAGARA1 || \ sun4v_chip_type == SUN4V_CHIP_NIAGARA2 || \ - sun4v_chip_type == SUN4V_CHIP_NIAGARA3)) ? \ + sun4v_chip_type == SUN4V_CHIP_NIAGARA3 || \ + sun4v_chip_type == SUN4V_CHIP_NIAGARA4 || \ + sun4v_chip_type == SUN4V_CHIP_NIAGARA5)) ? \ &xor_block_niagara : \ &xor_block_VIS) diff --git a/arch/sparc/kernel/cpu.c b/arch/sparc/kernel/cpu.c index 9810fd8..ba9b1ce 100644 --- a/arch/sparc/kernel/cpu.c +++ b/arch/sparc/kernel/cpu.c @@ -481,6 +481,18 @@ static void __init sun4v_cpu_probe(void) sparc_pmu_type = "niagara3"; break; + case SUN4V_CHIP_NIAGARA4: + sparc_cpu_type = "UltraSparc T4 (Niagara4)"; + sparc_fpu_type = "UltraSparc T4 integrated FPU"; + sparc_pmu_type = "niagara4"; + break; + + case SUN4V_CHIP_NIAGARA5: + sparc_cpu_type = "UltraSparc T5 (Niagara5)"; + sparc_fpu_type = "UltraSparc T5 integrated FPU"; + sparc_pmu_type = "niagara5"; + break; + default: printk(KERN_WARNING "CPU: Unknown sun4v cpu type [%s]\n", prom_cpu_compatible); diff --git a/arch/sparc/kernel/cpumap.c b/arch/sparc/kernel/cpumap.c index 4197e8d..9323eaf 100644 --- a/arch/sparc/kernel/cpumap.c +++ b/arch/sparc/kernel/cpumap.c @@ -325,6 +325,8 @@ static int iterate_cpu(struct cpuinfo_tree *t, unsigned int root_index) case SUN4V_CHIP_NIAGARA1: case SUN4V_CHIP_NIAGARA2: case SUN4V_CHIP_NIAGARA3: + case SUN4V_CHIP_NIAGARA4: + case SUN4V_CHIP_NIAGARA5: rover_inc_table = niagara_iterate_method; break; default: diff --git a/arch/sparc/kernel/head_64.S b/arch/sparc/kernel/head_64.S index 0eac1b2..0d810c2 100644 --- a/arch/sparc/kernel/head_64.S +++ b/arch/sparc/kernel/head_64.S @@ -133,7 +133,7 @@ prom_sun4v_name: prom_niagara_prefix: .asciz "SUNW,UltraSPARC-T" prom_sparc_prefix: - .asciz "SPARC-T" + .asciz "SPARC-" .align 4 prom_root_compatible: .skip 64 @@ -396,7 +396,7 @@ sun4v_chip_type: or %g1, %lo(prom_cpu_compatible), %g1 sethi %hi(prom_sparc_prefix), %g7 or %g7, %lo(prom_sparc_prefix), %g7 - mov 7, %g3 + mov 6, %g3 90: ldub [%g7], %g2 ldub [%g1], %g4 cmp %g2, %g4 @@ -408,10 +408,23 @@ sun4v_chip_type: sethi %hi(prom_cpu_compatible), %g1 or %g1, %lo(prom_cpu_compatible), %g1 - ldub [%g1 + 7], %g2 + ldub [%g1 + 6], %g2 + cmp %g2, 'T' + be,pt %xcc, 70f + cmp %g2, 'M' + bne,pn %xcc, 4f + nop + +70: ldub [%g1 + 7], %g2 cmp %g2, '3' be,pt %xcc, 5f mov SUN4V_CHIP_NIAGARA3, %g4 + cmp %g2, '4' + be,pt %xcc, 5f + mov SUN4V_CHIP_NIAGARA4, %g4 + cmp %g2, '5' + be,pt %xcc, 5f + mov SUN4V_CHIP_NIAGARA5, %g4 ba,pt %xcc, 4f nop @@ -545,6 +558,12 @@ niagara_tlb_fixup: cmp %g1, SUN4V_CHIP_NIAGARA3 be,pt %xcc, niagara2_patch nop + cmp %g1, SUN4V_CHIP_NIAGARA4 + be,pt %xcc, niagara2_patch + nop + cmp %g1, SUN4V_CHIP_NIAGARA5 + be,pt %xcc, niagara2_patch + nop call generic_patch_copyops nop diff --git a/arch/sparc/kernel/setup_64.c b/arch/sparc/kernel/setup_64.c index 3c5bb78..4139289 100644 --- a/arch/sparc/kernel/setup_64.c +++ b/arch/sparc/kernel/setup_64.c @@ -425,10 +425,14 @@ static void __init init_sparc64_elf_hwcap(void) else if (tlb_type == hypervisor) { if (sun4v_chip_type == SUN4V_CHIP_NIAGARA1 || sun4v_chip_type == SUN4V_CHIP_NIAGARA2 || - sun4v_chip_type == SUN4V_CHIP_NIAGARA3) + sun4v_chip_type == SUN4V_CHIP_NIAGARA3 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA4 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA5) cap |= HWCAP_SPARC_BLKINIT; if (sun4v_chip_type == SUN4V_CHIP_NIAGARA2 || - sun4v_chip_type == SUN4V_CHIP_NIAGARA3) + sun4v_chip_type == SUN4V_CHIP_NIAGARA3 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA4 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA5) cap |= HWCAP_SPARC_N2; } @@ -452,11 +456,15 @@ static void __init init_sparc64_elf_hwcap(void) if (sun4v_chip_type == SUN4V_CHIP_NIAGARA1) cap |= AV_SPARC_ASI_BLK_INIT; if (sun4v_chip_type == SUN4V_CHIP_NIAGARA2 || - sun4v_chip_type == SUN4V_CHIP_NIAGARA3) + sun4v_chip_type == SUN4V_CHIP_NIAGARA3 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA4 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA5) cap |= (AV_SPARC_VIS | AV_SPARC_VIS2 | AV_SPARC_ASI_BLK_INIT | AV_SPARC_POPC); - if (sun4v_chip_type == SUN4V_CHIP_NIAGARA3) + if (sun4v_chip_type == SUN4V_CHIP_NIAGARA3 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA4 || + sun4v_chip_type == SUN4V_CHIP_NIAGARA5) cap |= (AV_SPARC_VIS3 | AV_SPARC_HPC | AV_SPARC_FMAF); } -- cgit v0.10.2 From f630c1bdfbf8fe423325beaf60027cfc7fd7c610 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Thu, 15 Sep 2011 11:45:15 +0100 Subject: ARM: 7091/1: errata: D-cache line maintenance operation by MVA may not succeed This patch implements a workaround for erratum 764369 affecting Cortex-A9 MPCore with two or more processors (all current revisions). Under certain timing circumstances, a data cache line maintenance operation by MVA targeting an Inner Shareable memory region may fail to proceed up to either the Point of Coherency or to the Point of Unification of the system. This workaround adds a DSB instruction before the relevant cache maintenance functions and sets a specific bit in the diagnostic control register of the SCU. Cc: Signed-off-by: Will Deacon Signed-off-by: Catalin Marinas Tested-by: Mark Rutland Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 3269576..3146ed3 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1283,6 +1283,20 @@ config ARM_ERRATA_364296 processor into full low interrupt latency mode. ARM11MPCore is not affected. +config ARM_ERRATA_764369 + bool "ARM errata: Data cache line maintenance operation by MVA may not succeed" + depends on CPU_V7 && SMP + help + This option enables the workaround for erratum 764369 + affecting Cortex-A9 MPCore with two or more processors (all + current revisions). Under certain timing circumstances, a data + cache line maintenance operation by MVA targeting an Inner + Shareable memory region may fail to proceed up to either the + Point of Coherency or to the Point of Unification of the + system. This workaround adds a DSB instruction before the + relevant cache maintenance functions and sets a specific bit + in the diagnostic control register of the SCU. + endmenu source "arch/arm/common/Kconfig" diff --git a/arch/arm/kernel/smp_scu.c b/arch/arm/kernel/smp_scu.c index 79ed5e7..7fcddb7 100644 --- a/arch/arm/kernel/smp_scu.c +++ b/arch/arm/kernel/smp_scu.c @@ -13,6 +13,7 @@ #include #include +#include #define SCU_CTRL 0x00 #define SCU_CONFIG 0x04 @@ -37,6 +38,15 @@ void __init scu_enable(void __iomem *scu_base) { u32 scu_ctrl; +#ifdef CONFIG_ARM_ERRATA_764369 + /* Cortex-A9 only */ + if ((read_cpuid(CPUID_ID) & 0xff0ffff0) == 0x410fc090) { + scu_ctrl = __raw_readl(scu_base + 0x30); + if (!(scu_ctrl & 1)) + __raw_writel(scu_ctrl | 0x1, scu_base + 0x30); + } +#endif + scu_ctrl = __raw_readl(scu_base + SCU_CTRL); /* already enabled? */ if (scu_ctrl & 1) diff --git a/arch/arm/mm/cache-v7.S b/arch/arm/mm/cache-v7.S index 3b24bfa..07c4bc8 100644 --- a/arch/arm/mm/cache-v7.S +++ b/arch/arm/mm/cache-v7.S @@ -174,6 +174,10 @@ ENTRY(v7_coherent_user_range) dcache_line_size r2, r3 sub r3, r2, #1 bic r12, r0, r3 +#ifdef CONFIG_ARM_ERRATA_764369 + ALT_SMP(W(dsb)) + ALT_UP(W(nop)) +#endif 1: USER( mcr p15, 0, r12, c7, c11, 1 ) @ clean D line to the point of unification add r12, r12, r2 @@ -223,6 +227,10 @@ ENTRY(v7_flush_kern_dcache_area) add r1, r0, r1 sub r3, r2, #1 bic r0, r0, r3 +#ifdef CONFIG_ARM_ERRATA_764369 + ALT_SMP(W(dsb)) + ALT_UP(W(nop)) +#endif 1: mcr p15, 0, r0, c7, c14, 1 @ clean & invalidate D line / unified line add r0, r0, r2 @@ -247,6 +255,10 @@ v7_dma_inv_range: sub r3, r2, #1 tst r0, r3 bic r0, r0, r3 +#ifdef CONFIG_ARM_ERRATA_764369 + ALT_SMP(W(dsb)) + ALT_UP(W(nop)) +#endif mcrne p15, 0, r0, c7, c14, 1 @ clean & invalidate D / U line tst r1, r3 @@ -270,6 +282,10 @@ v7_dma_clean_range: dcache_line_size r2, r3 sub r3, r2, #1 bic r0, r0, r3 +#ifdef CONFIG_ARM_ERRATA_764369 + ALT_SMP(W(dsb)) + ALT_UP(W(nop)) +#endif 1: mcr p15, 0, r0, c7, c10, 1 @ clean D / U line add r0, r0, r2 @@ -288,6 +304,10 @@ ENTRY(v7_dma_flush_range) dcache_line_size r2, r3 sub r3, r2, #1 bic r0, r0, r3 +#ifdef CONFIG_ARM_ERRATA_764369 + ALT_SMP(W(dsb)) + ALT_UP(W(nop)) +#endif 1: mcr p15, 0, r0, c7, c14, 1 @ clean & invalidate D / U line add r0, r0, r2 -- cgit v0.10.2 From 4dfa33868d9f8595717047eb058322e0999691cc Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 17 Sep 2011 12:50:27 +0100 Subject: ARM: nommu: fix warning with checksyscalls.sh CALL scripts/checksyscalls.sh :46:1: warning: "__IGNORE_migrate_pages" redefined In file included from :2: arch/arm/include/asm/unistd.h:482:1: warning: this is the location of the previous definition This is caused because we define __IGNORE_migrate_pages to be 1, but in the case of nommu, it's defined to be empty. Fix this by just defining the __IGNORE_ symbols to be empty. Signed-off-by: Russell King diff --git a/arch/arm/include/asm/unistd.h b/arch/arm/include/asm/unistd.h index 2c04ed5..c60a294 100644 --- a/arch/arm/include/asm/unistd.h +++ b/arch/arm/include/asm/unistd.h @@ -478,8 +478,8 @@ /* * Unimplemented (or alternatively implemented) syscalls */ -#define __IGNORE_fadvise64_64 1 -#define __IGNORE_migrate_pages 1 +#define __IGNORE_fadvise64_64 +#define __IGNORE_migrate_pages #endif /* __KERNEL__ */ #endif /* __ASM_ARM_UNISTD_H */ -- cgit v0.10.2 From 3be209a8e22cedafc1b6945608b7bb8d9887ab61 Mon Sep 17 00:00:00 2001 From: Shawn Bohrer Date: Mon, 12 Sep 2011 09:28:04 -0500 Subject: sched/rt: Migrate equal priority tasks to available CPUs Commit 43fa5460fe60dea5c610490a1d263415419c60f6 ("sched: Try not to migrate higher priority RT tasks") also introduced a change in behavior which keeps RT tasks on the same CPU if there is an equal priority RT task currently running even if there are empty CPUs available. This can cause unnecessary wakeup latencies, and can prevent the scheduler from balancing all RT tasks across available CPUs. This change causes an RT task to search for a new CPU if an equal priority RT task is already running on wakeup. Lower priority tasks will still have to wait on higher priority tasks, but the system should still balance out because there is always the possibility that if there are both a high and low priority RT tasks on a given CPU that the high priority task could wakeup while the low priority task is running and force it to search for a better runqueue. Signed-off-by: Shawn Bohrer Acked-by: Steven Rostedt Tested-by: Steven Rostedt Signed-off-by: Peter Zijlstra Cc: stable@kernel.org # 37+ Link: http://lkml.kernel.org/r/1315837684-18733-1-git-send-email-sbohrer@rgmadvisors.com Signed-off-by: Ingo Molnar diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index 97540f0..af11778 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -1050,7 +1050,7 @@ select_task_rq_rt(struct task_struct *p, int sd_flag, int flags) */ if (curr && unlikely(rt_task(curr)) && (curr->rt.nr_cpus_allowed < 2 || - curr->prio < p->prio) && + curr->prio <= p->prio) && (p->rt.nr_cpus_allowed > 1)) { int target = find_lowest_rq(p); @@ -1581,7 +1581,7 @@ static void task_woken_rt(struct rq *rq, struct task_struct *p) p->rt.nr_cpus_allowed > 1 && rt_task(rq->curr) && (rq->curr->rt.nr_cpus_allowed < 2 || - rq->curr->prio < p->prio)) + rq->curr->prio <= p->prio)) push_rt_tasks(rq); } -- cgit v0.10.2 From 7cd015a0a37265e23bbdce7ce8f7e00507abe9ed Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Mon, 5 Sep 2011 11:53:21 -0700 Subject: drm/i915: Enable dither whenever display bpc < frame buffer bpc We want to enable dithering on any pipe where the frame buffer has more color resolution than the output device. The previous code was incorrectly clamping the frame buffer bpc to the display bpc, effectively disabling dithering all of the time as the computed frame buffer bpc would never be larger than the display bpc. Signed-off-by: Keith Packard Reported-by: Oliver Hartkopp Tested-by: Oliver Hartkopp diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 56a8554..9fb4a40 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -4687,13 +4687,13 @@ static bool intel_choose_pipe_bpp_dither(struct drm_crtc *crtc, bpc = 6; /* min is 18bpp */ break; case 24: - bpc = min((unsigned int)8, display_bpc); + bpc = 8; break; case 30: - bpc = min((unsigned int)10, display_bpc); + bpc = 10; break; case 48: - bpc = min((unsigned int)12, display_bpc); + bpc = 12; break; default: DRM_DEBUG("unsupported depth, assuming 24 bits\n"); @@ -4701,10 +4701,12 @@ static bool intel_choose_pipe_bpp_dither(struct drm_crtc *crtc, break; } + display_bpc = min(display_bpc, bpc); + DRM_DEBUG_DRIVER("setting pipe bpc to %d (max display bpc %d)\n", bpc, display_bpc); - *pipe_bpp = bpc * 3; + *pipe_bpp = display_bpc * 3; return display_bpc != bpc; } -- cgit v0.10.2 From 5bd078dda4d4fbdb4bd138a6bd5b6e274c019ed2 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Wed, 14 Sep 2011 11:31:36 -0500 Subject: irq: Add declaration of irq_domain_simple_ops to irqdomain.h irq_domain_simple_ops is exported, but is not declared in irqdomain.h, so add it. Signed-off-by: Rob Herring Cc: Grant Likely Cc: marc.zyngier@arm.com Cc: thomas.abraham@linaro.org Cc: jamie@jamieiles.com Cc: b-cousson@ti.com Cc: shawn.guo@linaro.org Cc: linux-arm-kernel@lists.infradead.org Cc: devicetree-discuss@lists.ozlabs.org Link: http://lkml.kernel.org/r/1316017900-19918-2-git-send-email-robherring2@gmail.com Signed-off-by: Thomas Gleixner diff --git a/include/linux/irqdomain.h b/include/linux/irqdomain.h index e807ad6..3ad553e 100644 --- a/include/linux/irqdomain.h +++ b/include/linux/irqdomain.h @@ -80,6 +80,7 @@ extern void irq_domain_del(struct irq_domain *domain); #endif /* CONFIG_IRQ_DOMAIN */ #if defined(CONFIG_IRQ_DOMAIN) && defined(CONFIG_OF_IRQ) +extern struct irq_domain_ops irq_domain_simple_ops; extern void irq_domain_add_simple(struct device_node *controller, int irq_base); extern void irq_domain_generate_simple(const struct of_device_id *match, u64 phys_base, unsigned int irq_start); -- cgit v0.10.2 From eef24afb28561a5a9f4be8f8da97735b7e6a826f Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Wed, 14 Sep 2011 11:31:37 -0500 Subject: irq: Fix check for already initialized irq_domain in irq_domain_add The sanity check in irq_domain_add() tests desc->irq_data != NULL or irq_data->domain != NULL. This prevents adding an irq_domain to a irq descriptor when irq_data exists, which true when the irq descriptor exists. This went unnoticed so far as the simple domain code did not enter this code path because domain->nr_irqs is always 0 for the simple domains. Split the check for irq_data == NULL out and have a separate warning for it. [ tglx: Made the check for irq_data == NULL separate ] Signed-off-by: Rob Herring Cc: Grant Likely Cc: marc.zyngier@arm.com Cc: thomas.abraham@linaro.org Cc: jamie@jamieiles.com Cc: b-cousson@ti.com Cc: shawn.guo@linaro.org Cc: linux-arm-kernel@lists.infradead.org Cc: devicetree-discuss@lists.ozlabs.org Link: http://lkml.kernel.org/r/1316017900-19918-3-git-send-email-robherring2@gmail.com Signed-off-by: Thomas Gleixner diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c index d5828da..b57a377 100644 --- a/kernel/irq/irqdomain.c +++ b/kernel/irq/irqdomain.c @@ -29,7 +29,11 @@ void irq_domain_add(struct irq_domain *domain) */ for (hwirq = 0; hwirq < domain->nr_irq; hwirq++) { d = irq_get_irq_data(irq_domain_to_irq(domain, hwirq)); - if (d || d->domain) { + if (!d) { + WARN(1, "error: assigning domain to non existant irq_desc"); + return; + } + if (d->domain) { /* things are broken; just report, don't clean up */ WARN(1, "error: irq_desc already assigned to a domain"); return; -- cgit v0.10.2 From f93dc4b6c975baeef9267a62451b370fbc586f3f Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 19 Sep 2011 23:33:35 +0100 Subject: ASoC: Remove bitrotted wm8962_resume() This functionality is now subsumed within the bias management, using the standard cache management functionality, without assuming the cache type. Signed-off-by: Mark Brown Acked-by: Liam Girdwood diff --git a/sound/soc/codecs/wm8962.c b/sound/soc/codecs/wm8962.c index 1725550..d2c315f 100644 --- a/sound/soc/codecs/wm8962.c +++ b/sound/soc/codecs/wm8962.c @@ -3479,31 +3479,6 @@ int wm8962_mic_detect(struct snd_soc_codec *codec, struct snd_soc_jack *jack) } EXPORT_SYMBOL_GPL(wm8962_mic_detect); -#ifdef CONFIG_PM -static int wm8962_resume(struct snd_soc_codec *codec) -{ - u16 *reg_cache = codec->reg_cache; - int i; - - /* Restore the registers */ - for (i = 1; i < codec->driver->reg_cache_size; i++) { - switch (i) { - case WM8962_SOFTWARE_RESET: - continue; - default: - break; - } - - if (reg_cache[i] != wm8962_reg[i]) - snd_soc_write(codec, i, reg_cache[i]); - } - - return 0; -} -#else -#define wm8962_resume NULL -#endif - #if defined(CONFIG_INPUT) || defined(CONFIG_INPUT_MODULE) static int beep_rates[] = { 500, 1000, 2000, 4000, @@ -4015,7 +3990,6 @@ static int wm8962_remove(struct snd_soc_codec *codec) static struct snd_soc_codec_driver soc_codec_dev_wm8962 = { .probe = wm8962_probe, .remove = wm8962_remove, - .resume = wm8962_resume, .set_bias_level = wm8962_set_bias_level, .reg_cache_size = WM8962_MAX_REGISTER + 1, .reg_word_size = sizeof(u16), -- cgit v0.10.2 From 480e5926ce3bb61ec229be2dab08bdce8abb8d2e Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Tue, 20 Sep 2011 17:07:28 +0200 Subject: [S390] kvm: fix address mode switching 598841ca9919d008b520114d8a4378c4ce4e40a1 ([S390] use gmap address spaces for kvm guest images) changed kvm to use a separate address space for kvm guests. This address space was switched in __vcpu_run In some cases (preemption, page fault) there is the possibility that this address space switch is lost. The typical symptom was a huge amount of validity intercepts or random guest addressing exceptions. Fix this by doing the switch in sie_loop and sie_exit and saving the address space in the gmap structure itself. Also use the preempt notifier. Signed-off-by: Christian Borntraeger Acked-by: Avi Kivity Signed-off-by: Heiko Carstens diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 519eb5f..c0cb794 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -658,12 +658,14 @@ static inline void pgste_set_pte(pte_t *ptep, pgste_t pgste) * struct gmap_struct - guest address space * @mm: pointer to the parent mm_struct * @table: pointer to the page directory + * @asce: address space control element for gmap page table * @crst_list: list of all crst tables used in the guest address space */ struct gmap { struct list_head list; struct mm_struct *mm; unsigned long *table; + unsigned long asce; struct list_head crst_list; }; diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c index 532fd43..2b45591 100644 --- a/arch/s390/kernel/asm-offsets.c +++ b/arch/s390/kernel/asm-offsets.c @@ -10,6 +10,7 @@ #include #include #include +#include /* * Make sure that the compiler is new enough. We want a compiler that @@ -126,6 +127,7 @@ int main(void) DEFINE(__LC_KERNEL_STACK, offsetof(struct _lowcore, kernel_stack)); DEFINE(__LC_ASYNC_STACK, offsetof(struct _lowcore, async_stack)); DEFINE(__LC_PANIC_STACK, offsetof(struct _lowcore, panic_stack)); + DEFINE(__LC_USER_ASCE, offsetof(struct _lowcore, user_asce)); DEFINE(__LC_INT_CLOCK, offsetof(struct _lowcore, int_clock)); DEFINE(__LC_MCCK_CLOCK, offsetof(struct _lowcore, mcck_clock)); DEFINE(__LC_MACHINE_FLAGS, offsetof(struct _lowcore, machine_flags)); @@ -151,6 +153,7 @@ int main(void) DEFINE(__LC_VDSO_PER_CPU, offsetof(struct _lowcore, vdso_per_cpu_data)); DEFINE(__LC_GMAP, offsetof(struct _lowcore, gmap)); DEFINE(__LC_CMF_HPP, offsetof(struct _lowcore, cmf_hpp)); + DEFINE(__GMAP_ASCE, offsetof(struct gmap, asce)); #endif /* CONFIG_32BIT */ return 0; } diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index 5f729d6..713da07 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -1076,6 +1076,11 @@ sie_loop: lg %r14,__LC_THREAD_INFO # pointer thread_info struct tm __TI_flags+7(%r14),_TIF_EXIT_SIE jnz sie_exit + lg %r14,__LC_GMAP # get gmap pointer + ltgr %r14,%r14 + jz sie_gmap + lctlg %c1,%c1,__GMAP_ASCE(%r14) # load primary asce +sie_gmap: lg %r14,__SF_EMPTY(%r15) # get control block pointer SPP __SF_EMPTY(%r15) # set guest id sie 0(%r14) @@ -1083,6 +1088,7 @@ sie_done: SPP __LC_CMF_HPP # set host id lg %r14,__LC_THREAD_INFO # pointer thread_info struct sie_exit: + lctlg %c1,%c1,__LC_USER_ASCE # load primary asce ni __TI_flags+6(%r14),255-(_TIF_SIE>>8) lg %r14,__SF_EMPTY+8(%r15) # load guest register save area stmg %r0,%r13,0(%r14) # save guest gprs 0-13 diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index f17296e..b4eced13 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -263,10 +263,12 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) vcpu->arch.guest_fpregs.fpc &= FPC_VALID_MASK; restore_fp_regs(&vcpu->arch.guest_fpregs); restore_access_regs(vcpu->arch.guest_acrs); + gmap_enable(vcpu->arch.gmap); } void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) { + gmap_disable(vcpu->arch.gmap); save_fp_regs(&vcpu->arch.guest_fpregs); save_access_regs(vcpu->arch.guest_acrs); restore_fp_regs(&vcpu->arch.host_fpregs); @@ -461,7 +463,6 @@ static void __vcpu_run(struct kvm_vcpu *vcpu) local_irq_disable(); kvm_guest_enter(); local_irq_enable(); - gmap_enable(vcpu->arch.gmap); VCPU_EVENT(vcpu, 6, "entering sie flags %x", atomic_read(&vcpu->arch.sie_block->cpuflags)); if (sie64a(vcpu->arch.sie_block, vcpu->arch.guest_gprs)) { @@ -470,7 +471,6 @@ static void __vcpu_run(struct kvm_vcpu *vcpu) } VCPU_EVENT(vcpu, 6, "exit sie icptcode %d", vcpu->arch.sie_block->icptcode); - gmap_disable(vcpu->arch.gmap); local_irq_disable(); kvm_guest_exit(); local_irq_enable(); diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 4d1f2bc..f69ff3c 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -160,6 +160,8 @@ struct gmap *gmap_alloc(struct mm_struct *mm) table = (unsigned long *) page_to_phys(page); crst_table_init(table, _REGION1_ENTRY_EMPTY); gmap->table = table; + gmap->asce = _ASCE_TYPE_REGION1 | _ASCE_TABLE_LENGTH | + _ASCE_USER_BITS | __pa(table); list_add(&gmap->list, &mm->context.gmap_list); return gmap; @@ -240,10 +242,6 @@ EXPORT_SYMBOL_GPL(gmap_free); */ void gmap_enable(struct gmap *gmap) { - /* Load primary space page table origin. */ - S390_lowcore.user_asce = _ASCE_TYPE_REGION1 | _ASCE_TABLE_LENGTH | - _ASCE_USER_BITS | __pa(gmap->table); - asm volatile("lctlg 1,1,%0\n" : : "m" (S390_lowcore.user_asce) ); S390_lowcore.gmap = (unsigned long) gmap; } EXPORT_SYMBOL_GPL(gmap_enable); @@ -254,10 +252,6 @@ EXPORT_SYMBOL_GPL(gmap_enable); */ void gmap_disable(struct gmap *gmap) { - /* Load primary space page table origin. */ - S390_lowcore.user_asce = - gmap->mm->context.asce_bits | __pa(gmap->mm->pgd); - asm volatile("lctlg 1,1,%0\n" : : "m" (S390_lowcore.user_asce) ); S390_lowcore.gmap = 0UL; } EXPORT_SYMBOL_GPL(gmap_disable); -- cgit v0.10.2 From b6cf8788a3382c2000743a0e393bcc8aeb0601cb Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Tue, 20 Sep 2011 17:07:29 +0200 Subject: [S390] kvm: extension capability for new address space layout 598841ca9919d008b520114d8a4378c4ce4e40a1 ([S390] use gmap address spaces for kvm guest images) changed kvm on s390 to use a separate address space for kvm guests. We can now put KVM guests anywhere in the user address mode with a size up to 8PB - as long as the memory is 1MB-aligned. This change was done without KVM extension capability bit. The change was added after 3.0, but we still have a chance to add a feature bit before 3.1 (keeping the releases in a sane state). We use number 71 to avoid collisions with other pending kvm patches as requested by Alexander Graf. Signed-off-by: Christian Borntraeger Acked-by: Avi Kivity Cc: Alexander Graf Signed-off-by: Heiko Carstens diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index b4eced13..dc2b580 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -123,6 +123,7 @@ int kvm_dev_ioctl_check_extension(long ext) switch (ext) { case KVM_CAP_S390_PSW: + case KVM_CAP_S390_GMAP: r = 1; break; default: diff --git a/include/linux/kvm.h b/include/linux/kvm.h index 2c366b5..aace6b8 100644 --- a/include/linux/kvm.h +++ b/include/linux/kvm.h @@ -553,6 +553,7 @@ struct kvm_ppc_pvinfo { #define KVM_CAP_SPAPR_TCE 63 #define KVM_CAP_PPC_SMT 64 #define KVM_CAP_PPC_RMA 65 +#define KVM_CAP_S390_GMAP 71 #ifdef KVM_CAP_IRQ_ROUTING -- cgit v0.10.2 From 6760b109603c794e4bd281c0014fef069c019b6a Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 20 Sep 2011 23:35:15 +0100 Subject: ARM: fix vmlinux.lds.S discarding sections We are seeing linker errors caused by sections being discarded, despite the linker script trying to keep them. The result is (eg): `.exit.text' referenced in section `.alt.smp.init' of drivers/built-in.o: defined in discarded section `.exit.text' of drivers/built-in.o `.exit.text' referenced in section `.alt.smp.init' of net/built-in.o: defined in discarded section `.exit.text' of net/built-in.o This is the relevent part of the linker script (reformatted to make it clearer): | SECTIONS | { | /* | * unwind exit sections must be discarded before the rest of the | * unwind sections get included. | */ | /DISCARD/ : { | *(.ARM.exidx.exit.text) | *(.ARM.extab.exit.text) | } | ... | .exit.text : { | *(.exit.text) | *(.memexit.text) | } | ... | /DISCARD/ : { | *(.exit.text) | *(.memexit.text) | *(.exit.data) | *(.memexit.data) | *(.memexit.rodata) | *(.exitcall.exit) | *(.discard) | *(.discard.*) | } | } Now, this is what the linker manual says about discarded output sections: | The special output section name `/DISCARD/' may be used to discard | input sections. Any input sections which are assigned to an output | section named `/DISCARD/' are not included in the output file. No questions, no exceptions. It doesn't say "unless they are listed before the /DISCARD/ section." Now, this is what asn-generic/vmlinux.lds.S says: | /* | * Default discarded sections. | * | * Some archs want to discard exit text/data at runtime rather than | * link time due to cross-section references such as alt instructions, | * bug table, eh_frame, etc. DISCARDS must be the last of output | * section definitions so that such archs put those in earlier section | * definitions. | */ And guess what - the list _always_ includes .exit.text etc. Now, what's actually happening is that the linker is reading the script, and it finds the first /DISCARD/ output section at the beginning of the script. It continues reading the script, and finds the 'DISCARD' macro at the end, which having been postprocessed results in another /DISCARD/ output section. As the linker already contains the earlier /DISCARD/ output section, it adds it to that existing section, so it effectively is placed at the start. This can be seen by using the -M option to ld: | Linker script and memory map | | 0xc037c080 jiffies = jiffies_64 | | /DISCARD/ | *(.ARM.exidx.exit.text) | *(.ARM.extab.exit.text) | *(.exit.text) | *(.memexit.text) | *(.exit.data) | *(.memexit.data) | *(.memexit.rodata) | *(.exitcall.exit) | *(.discard) | *(.discard.*) | | 0xc0008000 . = 0xc0008000 | | .head.text 0xc0008000 0x1d0 | 0xc0008000 _text = . | *(.head.text) | .head.text 0xc0008000 0x1d0 arch/arm/kernel/head.o | 0xc0008000 stext | | .text 0xc0008200 0x2d78d0 | 0xc0008200 _stext = . | 0xc0008200 __exception_text_start = . | *(.exception.text) | .exception.text | ... As you can see, all the discarded sections are grouped together - and as a result of it being the first output section, they all appear before any other section. The result is that not only is the unwind information discarded (as intended), but also the .exit.text, despite us wanting to have the .exit.text preserved. We can't move the unwind information elsewhere, because it'll then be included even when we do actually discard the .exit.text (and similar) sections. So, work around this by avoiding the generic DISCARDS macro, and instead conditionalize the sections to be discarded ourselves. This avoids the ambiguity in how the linker assigns input sections to output sections, making our script less dependent on undocumented linker behaviour. Reported-by: Rob Herring Tested-by: Mark Brown Signed-off-by: Russell King diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index bf977f8..4e66f62 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -23,8 +23,10 @@ #if defined(CONFIG_SMP_ON_UP) && !defined(CONFIG_DEBUG_SPINLOCK) #define ARM_EXIT_KEEP(x) x +#define ARM_EXIT_DISCARD(x) #else #define ARM_EXIT_KEEP(x) +#define ARM_EXIT_DISCARD(x) x #endif OUTPUT_ARCH(arm) @@ -39,6 +41,11 @@ jiffies = jiffies_64 + 4; SECTIONS { /* + * XXX: The linker does not define how output sections are + * assigned to input sections when there are multiple statements + * matching the same input section name. There is no documented + * order of matching. + * * unwind exit sections must be discarded before the rest of the * unwind sections get included. */ @@ -47,6 +54,9 @@ SECTIONS *(.ARM.extab.exit.text) ARM_CPU_DISCARD(*(.ARM.exidx.cpuexit.text)) ARM_CPU_DISCARD(*(.ARM.extab.cpuexit.text)) + ARM_EXIT_DISCARD(EXIT_TEXT) + ARM_EXIT_DISCARD(EXIT_DATA) + EXIT_CALL #ifndef CONFIG_HOTPLUG *(.ARM.exidx.devexit.text) *(.ARM.extab.devexit.text) @@ -58,6 +68,8 @@ SECTIONS #ifndef CONFIG_SMP_ON_UP *(.alt.smp.init) #endif + *(.discard) + *(.discard.*) } #ifdef CONFIG_XIP_KERNEL @@ -279,9 +291,6 @@ SECTIONS STABS_DEBUG .comment 0 : { *(.comment) } - - /* Default discards */ - DISCARDS } /* -- cgit v0.10.2 From 01f96c0a9922cd9919baf9d16febdf7016177a12 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 21 Sep 2011 15:30:20 +1000 Subject: md: Avoid waking up a thread after it has been freed. Two related problems: 1/ some error paths call "md_unregister_thread(mddev->thread)" without subsequently clearing ->thread. A subsequent call to mddev_unlock will try to wake the thread, and crash. 2/ Most calls to md_wakeup_thread are protected against the thread disappeared either by: - holding the ->mutex - having an active request, so something else must be keeping the array active. However mddev_unlock calls md_wakeup_thread after dropping the mutex and without any certainty of an active request, so the ->thread could theoretically disappear. So we need a spinlock to provide some protections. So change md_unregister_thread to take a pointer to the thread pointer, and ensure that it always does the required locking, and clears the pointer properly. Reported-by: "Moshe Melnikov" Signed-off-by: NeilBrown cc: stable@kernel.org diff --git a/drivers/md/md.c b/drivers/md/md.c index 5404b22..5c95ccb 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -61,6 +61,11 @@ static void autostart_arrays(int part); #endif +/* pers_list is a list of registered personalities protected + * by pers_lock. + * pers_lock does extra service to protect accesses to + * mddev->thread when the mutex cannot be held. + */ static LIST_HEAD(pers_list); static DEFINE_SPINLOCK(pers_lock); @@ -739,7 +744,12 @@ static void mddev_unlock(mddev_t * mddev) } else mutex_unlock(&mddev->reconfig_mutex); + /* was we've dropped the mutex we need a spinlock to + * make sur the thread doesn't disappear + */ + spin_lock(&pers_lock); md_wakeup_thread(mddev->thread); + spin_unlock(&pers_lock); } static mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr) @@ -6429,11 +6439,18 @@ mdk_thread_t *md_register_thread(void (*run) (mddev_t *), mddev_t *mddev, return thread; } -void md_unregister_thread(mdk_thread_t *thread) +void md_unregister_thread(mdk_thread_t **threadp) { + mdk_thread_t *thread = *threadp; if (!thread) return; dprintk("interrupting MD-thread pid %d\n", task_pid_nr(thread->tsk)); + /* Locking ensures that mddev_unlock does not wake_up a + * non-existent thread + */ + spin_lock(&pers_lock); + *threadp = NULL; + spin_unlock(&pers_lock); kthread_stop(thread->tsk); kfree(thread); @@ -7340,8 +7357,7 @@ static void reap_sync_thread(mddev_t *mddev) mdk_rdev_t *rdev; /* resync has finished, collect result */ - md_unregister_thread(mddev->sync_thread); - mddev->sync_thread = NULL; + md_unregister_thread(&mddev->sync_thread); if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery) && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) { /* success...*/ diff --git a/drivers/md/md.h b/drivers/md/md.h index 1e586bb..0a309dc 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -560,7 +560,7 @@ extern int register_md_personality(struct mdk_personality *p); extern int unregister_md_personality(struct mdk_personality *p); extern mdk_thread_t * md_register_thread(void (*run) (mddev_t *mddev), mddev_t *mddev, const char *name); -extern void md_unregister_thread(mdk_thread_t *thread); +extern void md_unregister_thread(mdk_thread_t **threadp); extern void md_wakeup_thread(mdk_thread_t *thread); extern void md_check_recovery(mddev_t *mddev); extern void md_write_start(mddev_t *mddev, struct bio *bi); diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c index 3535c23..d5b5fb3 100644 --- a/drivers/md/multipath.c +++ b/drivers/md/multipath.c @@ -514,8 +514,7 @@ static int multipath_stop (mddev_t *mddev) { multipath_conf_t *conf = mddev->private; - md_unregister_thread(mddev->thread); - mddev->thread = NULL; + md_unregister_thread(&mddev->thread); blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/ mempool_destroy(conf->pool); kfree(conf->multipaths); diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index f4622dd..d9587df 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -2562,8 +2562,7 @@ static int stop(mddev_t *mddev) raise_barrier(conf); lower_barrier(conf); - md_unregister_thread(mddev->thread); - mddev->thread = NULL; + md_unregister_thread(&mddev->thread); if (conf->r1bio_pool) mempool_destroy(conf->r1bio_pool); kfree(conf->mirrors); diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index d7a8468..0cd9672 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -2955,7 +2955,7 @@ static int run(mddev_t *mddev) return 0; out_free_conf: - md_unregister_thread(mddev->thread); + md_unregister_thread(&mddev->thread); if (conf->r10bio_pool) mempool_destroy(conf->r10bio_pool); safe_put_page(conf->tmppage); @@ -2973,8 +2973,7 @@ static int stop(mddev_t *mddev) raise_barrier(conf, 0); lower_barrier(conf); - md_unregister_thread(mddev->thread); - mddev->thread = NULL; + md_unregister_thread(&mddev->thread); blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/ if (conf->r10bio_pool) mempool_destroy(conf->r10bio_pool); diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 43709fa..ac5e8b5 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -4941,8 +4941,7 @@ static int run(mddev_t *mddev) return 0; abort: - md_unregister_thread(mddev->thread); - mddev->thread = NULL; + md_unregister_thread(&mddev->thread); if (conf) { print_raid5_conf(conf); free_conf(conf); @@ -4956,8 +4955,7 @@ static int stop(mddev_t *mddev) { raid5_conf_t *conf = mddev->private; - md_unregister_thread(mddev->thread); - mddev->thread = NULL; + md_unregister_thread(&mddev->thread); if (mddev->queue) mddev->queue->backing_dev_info.congested_fn = NULL; free_conf(conf); -- cgit v0.10.2 From ddd53bf11172822bc02b32b876a491010257d1b0 Mon Sep 17 00:00:00 2001 From: Mathias Krause Date: Tue, 20 Sep 2011 14:22:59 -0700 Subject: sparc, exec: remove redundant addr_limit assignment The address limit is already set in flush_old_exec() so this assignment of USER_DS is redundant. Signed-off-by: Mathias Krause Cc: David Miller Signed-off-by: Andrew Morton Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/process_32.c b/arch/sparc/kernel/process_32.c index c8cc461..f793742 100644 --- a/arch/sparc/kernel/process_32.c +++ b/arch/sparc/kernel/process_32.c @@ -380,8 +380,7 @@ void flush_thread(void) #endif } - /* Now, this task is no longer a kernel thread. */ - current->thread.current_ds = USER_DS; + /* This task is no longer a kernel thread. */ if (current->thread.flags & SPARC_FLAG_KTHREAD) { current->thread.flags &= ~SPARC_FLAG_KTHREAD; diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c index c158a95..d959cd0 100644 --- a/arch/sparc/kernel/process_64.c +++ b/arch/sparc/kernel/process_64.c @@ -368,9 +368,6 @@ void flush_thread(void) /* Clear FPU register state. */ t->fpsaved[0] = 0; - - if (get_thread_current_ds() != ASI_AIUS) - set_fs(USER_DS); } /* It's a bit more tricky when 64-bit tasks are involved... */ -- cgit v0.10.2 From 6fcb8a3a3a542ca078dc739fe0ea97191bc59e74 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Mon, 19 Sep 2011 16:17:55 +0200 Subject: mfd: Fix generic irq chip ack function name for jz4740-adc In commit 659fb32d1b6("genirq: replace irq_gc_ack() with {set,clr}_bit variants"), irq_gc_ack was renamed to irq_gc_ack_set_bit. The jz4740-adc driver still uses the old name which results in a compile error. Signed-off-by: Lars-Peter Clausen Signed-off-by: Samuel Ortiz diff --git a/drivers/mfd/jz4740-adc.c b/drivers/mfd/jz4740-adc.c index 21131c7..563654c 100644 --- a/drivers/mfd/jz4740-adc.c +++ b/drivers/mfd/jz4740-adc.c @@ -273,7 +273,7 @@ static int __devinit jz4740_adc_probe(struct platform_device *pdev) ct->regs.ack = JZ_REG_ADC_STATUS; ct->chip.irq_mask = irq_gc_mask_set_bit; ct->chip.irq_unmask = irq_gc_mask_clr_bit; - ct->chip.irq_ack = irq_gc_ack; + ct->chip.irq_ack = irq_gc_ack_set_bit; irq_setup_generic_chip(gc, IRQ_MSK(5), 0, 0, IRQ_NOPROBE | IRQ_LEVEL); -- cgit v0.10.2 From f0e8ed858edb327802ee65fd695cc1538286226f Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 20 Sep 2011 11:41:54 +0100 Subject: ASoC: Ensure we generate a driver name Commit 873bd4c (ASoC: Don't set invalid name string to snd_card->driver field) broke generation of a driver name for all ASoC cards relying on the automatic generation of one. Fix this by using the old default with spaces replaced by underscores. Signed-off-by: Mark Brown Acked-by: Takashi Iwai diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index d2ef014..ef69f5a 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include @@ -1434,9 +1435,20 @@ static void snd_soc_instantiate_card(struct snd_soc_card *card) "%s", card->name); snprintf(card->snd_card->longname, sizeof(card->snd_card->longname), "%s", card->long_name ? card->long_name : card->name); - if (card->driver_name) - strlcpy(card->snd_card->driver, card->driver_name, - sizeof(card->snd_card->driver)); + snprintf(card->snd_card->driver, sizeof(card->snd_card->driver), + "%s", card->driver_name ? card->driver_name : card->name); + for (i = 0; i < ARRAY_SIZE(card->snd_card->driver); i++) { + switch (card->snd_card->driver[i]) { + case '_': + case '-': + case '\0': + break; + default: + if (!isalnum(card->snd_card->driver[i])) + card->snd_card->driver[i] = '_'; + break; + } + } if (card->late_probe) { ret = card->late_probe(card); -- cgit v0.10.2 From 47997d756aa2a84ab577e1b0383cc12d582fc69c Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Wed, 21 Sep 2011 16:08:03 +0200 Subject: x86/rtc: Don't recursively acquire rtc_lock A deadlock was introduced on x86 in commit ef68c8f87ed1 ("x86: Serialize EFI time accesses on rtc_lock") because efi_get_time() and friends can be called with rtc_lock already held by read_persistent_time(), e.g.: timekeeping_init() read_persistent_clock() <-- acquire rtc_lock efi_get_time() phys_efi_get_time() <-- acquire rtc_lock To fix this let's push the locking down into the get_wallclock() and set_wallclock() implementations. Only the clock implementations that access the x86 RTC directly need to acquire rtc_lock, so it makes sense to push the locking down into the rtc, vrtc and efi code. The virtualization implementations don't require rtc_lock to be held because they provide their own serialization. Signed-off-by: Matt Fleming Acked-by: Jan Beulich Acked-by: Avi Kivity [for the virtualization aspect] Cc: "H. Peter Anvin" Cc: Zhang Rui Cc: Josh Boyer Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index 3f2ad26..ccdbc16 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -42,8 +42,11 @@ int mach_set_rtc_mmss(unsigned long nowtime) { int real_seconds, real_minutes, cmos_minutes; unsigned char save_control, save_freq_select; + unsigned long flags; int retval = 0; + spin_lock_irqsave(&rtc_lock, flags); + /* tell the clock it's being set */ save_control = CMOS_READ(RTC_CONTROL); CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); @@ -93,12 +96,17 @@ int mach_set_rtc_mmss(unsigned long nowtime) CMOS_WRITE(save_control, RTC_CONTROL); CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); + spin_unlock_irqrestore(&rtc_lock, flags); + return retval; } unsigned long mach_get_cmos_time(void) { unsigned int status, year, mon, day, hour, min, sec, century = 0; + unsigned long flags; + + spin_lock_irqsave(&rtc_lock, flags); /* * If UIP is clear, then we have >= 244 microseconds before @@ -125,6 +133,8 @@ unsigned long mach_get_cmos_time(void) status = CMOS_READ(RTC_CONTROL); WARN_ON_ONCE(RTC_ALWAYS_BCD && (status & RTC_DM_BINARY)); + spin_unlock_irqrestore(&rtc_lock, flags); + if (RTC_ALWAYS_BCD || !(status & RTC_DM_BINARY)) { sec = bcd2bin(sec); min = bcd2bin(min); @@ -169,24 +179,15 @@ EXPORT_SYMBOL(rtc_cmos_write); int update_persistent_clock(struct timespec now) { - unsigned long flags; - int retval; - - spin_lock_irqsave(&rtc_lock, flags); - retval = x86_platform.set_wallclock(now.tv_sec); - spin_unlock_irqrestore(&rtc_lock, flags); - - return retval; + return x86_platform.set_wallclock(now.tv_sec); } /* not static: needed by APM */ void read_persistent_clock(struct timespec *ts) { - unsigned long retval, flags; + unsigned long retval; - spin_lock_irqsave(&rtc_lock, flags); retval = x86_platform.get_wallclock(); - spin_unlock_irqrestore(&rtc_lock, flags); ts->tv_sec = retval; ts->tv_nsec = 0; diff --git a/arch/x86/platform/mrst/vrtc.c b/arch/x86/platform/mrst/vrtc.c index 73d70d6..6d5dbcd 100644 --- a/arch/x86/platform/mrst/vrtc.c +++ b/arch/x86/platform/mrst/vrtc.c @@ -58,8 +58,11 @@ EXPORT_SYMBOL_GPL(vrtc_cmos_write); unsigned long vrtc_get_time(void) { u8 sec, min, hour, mday, mon; + unsigned long flags; u32 year; + spin_lock_irqsave(&rtc_lock, flags); + while ((vrtc_cmos_read(RTC_FREQ_SELECT) & RTC_UIP)) cpu_relax(); @@ -70,6 +73,8 @@ unsigned long vrtc_get_time(void) mon = vrtc_cmos_read(RTC_MONTH); year = vrtc_cmos_read(RTC_YEAR); + spin_unlock_irqrestore(&rtc_lock, flags); + /* vRTC YEAR reg contains the offset to 1960 */ year += 1960; @@ -83,8 +88,10 @@ unsigned long vrtc_get_time(void) int vrtc_set_mmss(unsigned long nowtime) { int real_sec, real_min; + unsigned long flags; int vrtc_min; + spin_lock_irqsave(&rtc_lock, flags); vrtc_min = vrtc_cmos_read(RTC_MINUTES); real_sec = nowtime % 60; @@ -95,6 +102,8 @@ int vrtc_set_mmss(unsigned long nowtime) vrtc_cmos_write(real_sec, RTC_SECONDS); vrtc_cmos_write(real_min, RTC_MINUTES); + spin_unlock_irqrestore(&rtc_lock, flags); + return 0; } -- cgit v0.10.2 From 11032c17bd3f96a4319e4415c07fc791525b79fd Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 21 Sep 2011 12:48:06 -0700 Subject: sparc: Make '-p' boot option meaningful again. If "-p" is given on the command line, clear the CON_BOOT flag for the initial early boot PROM console. This is necessary to try and see crash messages that occur between the registry of the VT console and the probing of the first framebuffer or serial console. During this time no console messages are emitted because the VT console registry (even if no backend is registered to it) removes the early console if CON_BOOT is set. Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/setup_32.c b/arch/sparc/kernel/setup_32.c index d26e1f6..3e3e291 100644 --- a/arch/sparc/kernel/setup_32.c +++ b/arch/sparc/kernel/setup_32.c @@ -137,7 +137,7 @@ static void __init process_switch(char c) prom_halt(); break; case 'p': - /* Just ignore, this behavior is now the default. */ + prom_early_console.flags &= ~CON_BOOT; break; default: printk("Unknown boot switch (-%c)\n", c); diff --git a/arch/sparc/kernel/setup_64.c b/arch/sparc/kernel/setup_64.c index 4139289..c965595a 100644 --- a/arch/sparc/kernel/setup_64.c +++ b/arch/sparc/kernel/setup_64.c @@ -106,7 +106,7 @@ static void __init process_switch(char c) prom_halt(); break; case 'p': - /* Just ignore, this behavior is now the default. */ + prom_early_console.flags &= ~CON_BOOT; break; case 'P': /* Force UltraSPARC-III P-Cache on. */ -- cgit v0.10.2 From cc68c81aed7d892deaf12d720d5455208e94cd0a Mon Sep 17 00:00:00 2001 From: Simon Farnsworth Date: Wed, 21 Sep 2011 17:13:30 +0100 Subject: drm/i915: Enable SDVO hotplug interrupts for HDMI and DVI I was seeing a nasty 5 frame glitch every 10 seconds, caused by the poll for connection on DVI attached by SDVO. As my SDVO DVI supports hotplug detect interrupts, the fix is to enable them, and hook them in to the various bits of driver infrastructure so that they work reliably. Note that this is only tested on single-function DVI-D SDVOs, on two platforms (965GME and 945GSE), and has not been checked against a specification document. With lots of help from Adam Jackson on IRC. Signed-off-by: Simon Farnsworth Reviewed-by: Keith Packard Signed-off-by: Keith Packard diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 0b2ee9d..fe1099d 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -337,9 +337,6 @@ extern void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder, struct drm_connector *connector, struct intel_load_detect_pipe *old); -extern struct drm_connector* intel_sdvo_find(struct drm_device *dev, int sdvoB); -extern int intel_sdvo_supports_hotplug(struct drm_connector *connector); -extern void intel_sdvo_set_hotplug(struct drm_connector *connector, int enable); extern void intelfb_restore(void); extern void intel_crtc_fb_gamma_set(struct drm_crtc *crtc, u16 red, u16 green, u16 blue, int regno); diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 30fe554..6348c49 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -92,6 +92,11 @@ struct intel_sdvo { */ uint16_t attached_output; + /* + * Hotplug activation bits for this device + */ + uint8_t hotplug_active[2]; + /** * This is used to select the color range of RBG outputs in HDMI mode. * It is only valid when using TMDS encoding and 8 bit per color mode. @@ -1208,74 +1213,20 @@ static bool intel_sdvo_get_capabilities(struct intel_sdvo *intel_sdvo, struct in return true; } -/* No use! */ -#if 0 -struct drm_connector* intel_sdvo_find(struct drm_device *dev, int sdvoB) -{ - struct drm_connector *connector = NULL; - struct intel_sdvo *iout = NULL; - struct intel_sdvo *sdvo; - - /* find the sdvo connector */ - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - iout = to_intel_sdvo(connector); - - if (iout->type != INTEL_OUTPUT_SDVO) - continue; - - sdvo = iout->dev_priv; - - if (sdvo->sdvo_reg == SDVOB && sdvoB) - return connector; - - if (sdvo->sdvo_reg == SDVOC && !sdvoB) - return connector; - - } - - return NULL; -} - -int intel_sdvo_supports_hotplug(struct drm_connector *connector) +static int intel_sdvo_supports_hotplug(struct intel_sdvo *intel_sdvo) { u8 response[2]; - u8 status; - struct intel_sdvo *intel_sdvo; - DRM_DEBUG_KMS("\n"); - - if (!connector) - return 0; - - intel_sdvo = to_intel_sdvo(connector); return intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_HOT_PLUG_SUPPORT, &response, 2) && response[0]; } -void intel_sdvo_set_hotplug(struct drm_connector *connector, int on) +static void intel_sdvo_enable_hotplug(struct intel_encoder *encoder) { - u8 response[2]; - u8 status; - struct intel_sdvo *intel_sdvo = to_intel_sdvo(connector); - - intel_sdvo_write_cmd(intel_sdvo, SDVO_CMD_GET_ACTIVE_HOT_PLUG, NULL, 0); - intel_sdvo_read_response(intel_sdvo, &response, 2); - - if (on) { - intel_sdvo_write_cmd(intel_sdvo, SDVO_CMD_GET_HOT_PLUG_SUPPORT, NULL, 0); - status = intel_sdvo_read_response(intel_sdvo, &response, 2); - - intel_sdvo_write_cmd(intel_sdvo, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &response, 2); - } else { - response[0] = 0; - response[1] = 0; - intel_sdvo_write_cmd(intel_sdvo, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &response, 2); - } + struct intel_sdvo *intel_sdvo = to_intel_sdvo(&encoder->base); - intel_sdvo_write_cmd(intel_sdvo, SDVO_CMD_GET_ACTIVE_HOT_PLUG, NULL, 0); - intel_sdvo_read_response(intel_sdvo, &response, 2); + intel_sdvo_write_cmd(intel_sdvo, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &intel_sdvo->hotplug_active, 2); } -#endif static bool intel_sdvo_multifunc_encoder(struct intel_sdvo *intel_sdvo) @@ -2045,6 +1996,7 @@ intel_sdvo_dvi_init(struct intel_sdvo *intel_sdvo, int device) { struct drm_encoder *encoder = &intel_sdvo->base.base; struct drm_connector *connector; + struct intel_encoder *intel_encoder = to_intel_encoder(encoder); struct intel_connector *intel_connector; struct intel_sdvo_connector *intel_sdvo_connector; @@ -2062,7 +2014,17 @@ intel_sdvo_dvi_init(struct intel_sdvo *intel_sdvo, int device) intel_connector = &intel_sdvo_connector->base; connector = &intel_connector->base; - connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; + if (intel_sdvo_supports_hotplug(intel_sdvo) & (1 << device)) { + connector->polled = DRM_CONNECTOR_POLL_HPD; + intel_sdvo->hotplug_active[0] |= 1 << device; + /* Some SDVO devices have one-shot hotplug interrupts. + * Ensure that they get re-enabled when an interrupt happens. + */ + intel_encoder->hot_plug = intel_sdvo_enable_hotplug; + intel_sdvo_enable_hotplug(intel_encoder); + } + else + connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; encoder->encoder_type = DRM_MODE_ENCODER_TMDS; connector->connector_type = DRM_MODE_CONNECTOR_DVID; @@ -2569,6 +2531,14 @@ bool intel_sdvo_init(struct drm_device *dev, int sdvo_reg) if (!intel_sdvo_get_capabilities(intel_sdvo, &intel_sdvo->caps)) goto err; + /* Set up hotplug command - note paranoia about contents of reply. + * We assume that the hardware is in a sane state, and only touch + * the bits we think we understand. + */ + intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_ACTIVE_HOT_PLUG, + &intel_sdvo->hotplug_active, 2); + intel_sdvo->hotplug_active[0] &= ~0x3; + if (intel_sdvo_output_setup(intel_sdvo, intel_sdvo->caps.output_flags) != true) { DRM_DEBUG_KMS("SDVO output failed to setup on SDVO%c\n", -- cgit v0.10.2 From cd0de039bff32ee314046c0e4c047c38aa696f84 Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Mon, 19 Sep 2011 21:34:19 -0700 Subject: drm/i915: FBC off for ironlake and older, otherwise on by default Make the default FBC behaviour chipset specific, allowing us to turn it on by default for Ironlake and older where it has been seen to cause trouble with screen updates. Signed-off-by: Keith Packard Tested-by: Francis Moreau diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index ce045a8..f07e425 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -67,11 +67,11 @@ module_param_named(i915_enable_rc6, i915_enable_rc6, int, 0600); MODULE_PARM_DESC(i915_enable_rc6, "Enable power-saving render C-state 6 (default: true)"); -unsigned int i915_enable_fbc __read_mostly = 1; +unsigned int i915_enable_fbc __read_mostly = -1; module_param_named(i915_enable_fbc, i915_enable_fbc, int, 0600); MODULE_PARM_DESC(i915_enable_fbc, "Enable frame buffer compression for power savings " - "(default: false)"); + "(default: -1 (use per-chip default))"); unsigned int i915_lvds_downclock __read_mostly = 0; module_param_named(lvds_downclock, i915_lvds_downclock, int, 0400); diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9fb4a40..04411ad 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -1799,6 +1799,7 @@ static void intel_update_fbc(struct drm_device *dev) struct drm_framebuffer *fb; struct intel_framebuffer *intel_fb; struct drm_i915_gem_object *obj; + int enable_fbc; DRM_DEBUG_KMS("\n"); @@ -1839,8 +1840,15 @@ static void intel_update_fbc(struct drm_device *dev) intel_fb = to_intel_framebuffer(fb); obj = intel_fb->obj; - if (!i915_enable_fbc) { - DRM_DEBUG_KMS("fbc disabled per module param (default off)\n"); + enable_fbc = i915_enable_fbc; + if (enable_fbc < 0) { + DRM_DEBUG_KMS("fbc set to per-chip default\n"); + enable_fbc = 1; + if (INTEL_INFO(dev)->gen <= 5) + enable_fbc = 0; + } + if (!enable_fbc) { + DRM_DEBUG_KMS("fbc disabled per module param\n"); dev_priv->no_fbc_reason = FBC_MODULE_PARAM; goto out_disable; } -- cgit v0.10.2 From 6bf9e9b09c3abb5447bbbf16c2d0cbe721e44f3f Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 16 Sep 2011 21:21:43 +0200 Subject: hwmon: (coretemp) Drop duplicate function get_pkg_tjmax Function get_pkg_tjmax is a simplified copy of get_tjmax. Drop it and always use get_tjmax, result is the same and this avoids code duplication. Also make get_tjmax less verbose: don't warn about MSR read failure when failure was expected, and don't report TjMax in the logs unless debugging is enabled. Signed-off-by: Jean Delvare Cc: Fenghua Yu Cc: Guenter Roeck Cc: Durgadoss R Acked-by: Fenghua Yu Signed-off-by: Guenter Roeck diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 4112576..256f708 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -374,7 +374,6 @@ static int adjust_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev) static int get_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev) { - /* The 100C is default for both mobile and non mobile CPUs */ int err; u32 eax, edx; u32 val; @@ -385,7 +384,8 @@ static int get_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev) */ err = rdmsr_safe_on_cpu(id, MSR_IA32_TEMPERATURE_TARGET, &eax, &edx); if (err) { - dev_warn(dev, "Unable to read TjMax from CPU.\n"); + if (c->x86_model > 0xe && c->x86_model != 0x1c) + dev_warn(dev, "Unable to read TjMax from CPU %u\n", id); } else { val = (eax >> 16) & 0xff; /* @@ -393,7 +393,7 @@ static int get_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev) * will be used */ if (val) { - dev_info(dev, "TjMax is %d C.\n", val); + dev_dbg(dev, "TjMax is %d degrees C\n", val); return val * 1000; } } @@ -414,21 +414,6 @@ static void __devinit get_ucode_rev_on_cpu(void *edx) rdmsr(MSR_IA32_UCODE_REV, eax, *(u32 *)edx); } -static int get_pkg_tjmax(unsigned int cpu, struct device *dev) -{ - int err; - u32 eax, edx, val; - - err = rdmsr_safe_on_cpu(cpu, MSR_IA32_TEMPERATURE_TARGET, &eax, &edx); - if (!err) { - val = (eax >> 16) & 0xff; - if (val) - return val * 1000; - } - dev_warn(dev, "Unable to read Pkg-TjMax from CPU:%u\n", cpu); - return 100000; /* Default TjMax: 100 degree celsius */ -} - static int create_name_attr(struct platform_data *pdata, struct device *dev) { sysfs_attr_init(&pdata->name_attr.attr); @@ -588,10 +573,7 @@ static int create_core_data(struct platform_data *pdata, goto exit_free; /* We can access status register. Get Critical Temperature */ - if (pkg_flag) - tdata->tjmax = get_pkg_tjmax(pdev->id, &pdev->dev); - else - tdata->tjmax = get_tjmax(c, cpu, &pdev->dev); + tdata->tjmax = get_tjmax(c, cpu, &pdev->dev); /* * Test if we can access the intrpt register. If so, increase the -- cgit v0.10.2 From a45a8c8571c0be6a6bd72ae5a14255c26b14b504 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 16 Sep 2011 21:24:02 +0200 Subject: hwmon: (coretemp) Let the user force TjMax On old CPUs (and even some recent Atom CPUs) TjMax can't be read from the CPU registers, so it is guessed by the driver using a complex heuristic which isn't reliable. So let users who know their CPU's TjMax pass it as a module parameter. Signed-off-by: Jean Delvare Cc: Fenghua Yu Cc: "R, Durgadoss" Cc: Guenter Roeck Cc: Alexander Stein Acked-by: Fenghua Yu Signed-off-by: Guenter Roeck diff --git a/Documentation/hwmon/coretemp b/Documentation/hwmon/coretemp index fa8776a..49c0d42 100644 --- a/Documentation/hwmon/coretemp +++ b/Documentation/hwmon/coretemp @@ -49,9 +49,10 @@ tempX_label - Contains string "Core X", where X is processor number. For Package temp, this will be "Physical id Y", where Y is the package number. -The TjMax temperature is set to 85 degrees C if undocumented model specific -register (UMSR) 0xee has bit 30 set. If not the TjMax is 100 degrees C as -(sometimes) documented in processor datasheet. +On CPU models which support it, TjMax is read from a model-specific register. +On other models, it is set to an arbitrary value based on weak heuristics. +If these heuristics don't work for you, you can pass the correct TjMax value +as a module parameter (tjmax). Appendix A. Known TjMax lists (TBD): Some information comes from ark.intel.com diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 256f708..5a41e9d 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -36,11 +36,20 @@ #include #include #include +#include #include #include #define DRVNAME "coretemp" +/* + * force_tjmax only matters when TjMax can't be read from the CPU itself. + * When set, it replaces the driver's suboptimal heuristic. + */ +static int force_tjmax; +module_param_named(tjmax, force_tjmax, int, 0444); +MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius"); + #define BASE_SYSFS_ATTR_NO 2 /* Sysfs Base attr no for coretemp */ #define NUM_REAL_CORES 16 /* Number of Real cores per cpu */ #define CORETEMP_NAME_LENGTH 17 /* String Length of attrs */ @@ -398,6 +407,12 @@ static int get_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev) } } + if (force_tjmax) { + dev_notice(dev, "TjMax forced to %d degrees C by user\n", + force_tjmax); + return force_tjmax * 1000; + } + /* * An assumption is made for early CPUs and unreadable MSR. * NOTE: the calculated value may not be correct. -- cgit v0.10.2 From f4af6fd6e21792ca4deca3d29c113a575594078e Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 19 Sep 2011 21:41:16 -0700 Subject: hwmon: (coretemp) Don't use threshold registers for tempX_max With commit c814a4c7c4aad795835583344353963a0a673eb0, the meaning of tempX_max was changed. It no longer returns the value of bits 8:15 of MSR_IA32_TEMPERATURE_TARGET, but instead returns the value of CPU threshold register T1. tempX_max_hyst was added to reflect the value of temperature threshold register T0. As it turns out, T0 and T1 are used on some systems, presumably by the BIOS. Also, T0 and T1 don't have a well defined meaning. The thresholds may be used as upper or lower limits, and it is not guaranteed that T0 <= T1. Thus, the new attribute mapping does not reflect the actual usage of the threshold registers. Also, register contents are changed during runtime by an entity other than the hwmon driver, meaning the values cached by the driver do not reflect actual register contents. Revert most of c814a4c7c4aad795835583344353963a0a673eb0 to address the problem. Support for T0 and T1 will be added back in with a separate commit, using new attribute names. Signed-off-by: Guenter Roeck Cc: Fenghua Yu Cc: Durgadoss R Acked-by: Jean Delvare diff --git a/Documentation/hwmon/coretemp b/Documentation/hwmon/coretemp index 49c0d42..84d46c0 100644 --- a/Documentation/hwmon/coretemp +++ b/Documentation/hwmon/coretemp @@ -35,13 +35,6 @@ the Out-Of-Spec bit. Following table summarizes the exported sysfs files: All Sysfs entries are named with their core_id (represented here by 'X'). tempX_input - Core temperature (in millidegrees Celsius). tempX_max - All cooling devices should be turned on (on Core2). - Initialized with IA32_THERM_INTERRUPT. When the CPU - temperature reaches this temperature, an interrupt is - generated and tempX_max_alarm is set. -tempX_max_hyst - If the CPU temperature falls below than temperature, - an interrupt is generated and tempX_max_alarm is reset. -tempX_max_alarm - Set if the temperature reaches or exceeds tempX_max. - Reset if the temperature drops to or below tempX_max_hyst. tempX_crit - Maximum junction temperature (in millidegrees Celsius). tempX_crit_alarm - Set when Out-of-spec bit is set, never clears. Correct CPU operation is no longer guaranteed. diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 5a41e9d..4736415 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -54,8 +54,7 @@ MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius"); #define NUM_REAL_CORES 16 /* Number of Real cores per cpu */ #define CORETEMP_NAME_LENGTH 17 /* String Length of attrs */ #define MAX_CORE_ATTRS 4 /* Maximum no of basic attrs */ -#define MAX_THRESH_ATTRS 3 /* Maximum no of Threshold attrs */ -#define TOTAL_ATTRS (MAX_CORE_ATTRS + MAX_THRESH_ATTRS) +#define TOTAL_ATTRS (MAX_CORE_ATTRS + 1) #define MAX_CORE_DATA (NUM_REAL_CORES + BASE_SYSFS_ATTR_NO) #ifdef CONFIG_SMP @@ -78,8 +77,6 @@ MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius"); * This value is passed as "id" field to rdmsr/wrmsr functions. * @status_reg: One of IA32_THERM_STATUS or IA32_PACKAGE_THERM_STATUS, * from where the temperature values should be read. - * @intrpt_reg: One of IA32_THERM_INTERRUPT or IA32_PACKAGE_THERM_INTERRUPT, - * from where the thresholds are read. * @attr_size: Total number of pre-core attrs displayed in the sysfs. * @is_pkg_data: If this is 1, the temp_data holds pkgtemp data. * Otherwise, temp_data holds coretemp data. @@ -88,13 +85,11 @@ MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius"); struct temp_data { int temp; int ttarget; - int tmin; int tjmax; unsigned long last_updated; unsigned int cpu; u32 cpu_core_id; u32 status_reg; - u32 intrpt_reg; int attr_size; bool is_pkg_data; bool valid; @@ -152,19 +147,6 @@ static ssize_t show_crit_alarm(struct device *dev, return sprintf(buf, "%d\n", (eax >> 5) & 1); } -static ssize_t show_max_alarm(struct device *dev, - struct device_attribute *devattr, char *buf) -{ - u32 eax, edx; - struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); - struct platform_data *pdata = dev_get_drvdata(dev); - struct temp_data *tdata = pdata->core_data[attr->index]; - - rdmsr_on_cpu(tdata->cpu, tdata->status_reg, &eax, &edx); - - return sprintf(buf, "%d\n", !!(eax & THERM_STATUS_THRESHOLD1)); -} - static ssize_t show_tjmax(struct device *dev, struct device_attribute *devattr, char *buf) { @@ -183,83 +165,6 @@ static ssize_t show_ttarget(struct device *dev, return sprintf(buf, "%d\n", pdata->core_data[attr->index]->ttarget); } -static ssize_t store_ttarget(struct device *dev, - struct device_attribute *devattr, - const char *buf, size_t count) -{ - struct platform_data *pdata = dev_get_drvdata(dev); - struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); - struct temp_data *tdata = pdata->core_data[attr->index]; - u32 eax, edx; - unsigned long val; - int diff; - - if (strict_strtoul(buf, 10, &val)) - return -EINVAL; - - /* - * THERM_MASK_THRESHOLD1 is 7 bits wide. Values are entered in terms - * of milli degree celsius. Hence don't accept val > (127 * 1000) - */ - if (val > tdata->tjmax || val > 127000) - return -EINVAL; - - diff = (tdata->tjmax - val) / 1000; - - mutex_lock(&tdata->update_lock); - rdmsr_on_cpu(tdata->cpu, tdata->intrpt_reg, &eax, &edx); - eax = (eax & ~THERM_MASK_THRESHOLD1) | - (diff << THERM_SHIFT_THRESHOLD1); - wrmsr_on_cpu(tdata->cpu, tdata->intrpt_reg, eax, edx); - tdata->ttarget = val; - mutex_unlock(&tdata->update_lock); - - return count; -} - -static ssize_t show_tmin(struct device *dev, - struct device_attribute *devattr, char *buf) -{ - struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); - struct platform_data *pdata = dev_get_drvdata(dev); - - return sprintf(buf, "%d\n", pdata->core_data[attr->index]->tmin); -} - -static ssize_t store_tmin(struct device *dev, - struct device_attribute *devattr, - const char *buf, size_t count) -{ - struct platform_data *pdata = dev_get_drvdata(dev); - struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); - struct temp_data *tdata = pdata->core_data[attr->index]; - u32 eax, edx; - unsigned long val; - int diff; - - if (strict_strtoul(buf, 10, &val)) - return -EINVAL; - - /* - * THERM_MASK_THRESHOLD0 is 7 bits wide. Values are entered in terms - * of milli degree celsius. Hence don't accept val > (127 * 1000) - */ - if (val > tdata->tjmax || val > 127000) - return -EINVAL; - - diff = (tdata->tjmax - val) / 1000; - - mutex_lock(&tdata->update_lock); - rdmsr_on_cpu(tdata->cpu, tdata->intrpt_reg, &eax, &edx); - eax = (eax & ~THERM_MASK_THRESHOLD0) | - (diff << THERM_SHIFT_THRESHOLD0); - wrmsr_on_cpu(tdata->cpu, tdata->intrpt_reg, eax, edx); - tdata->tmin = val; - mutex_unlock(&tdata->update_lock); - - return count; -} - static ssize_t show_temp(struct device *dev, struct device_attribute *devattr, char *buf) { @@ -445,16 +350,11 @@ static int create_core_attrs(struct temp_data *tdata, struct device *dev, static ssize_t (*rd_ptr[TOTAL_ATTRS]) (struct device *dev, struct device_attribute *devattr, char *buf) = { show_label, show_crit_alarm, show_temp, show_tjmax, - show_max_alarm, show_ttarget, show_tmin }; - static ssize_t (*rw_ptr[TOTAL_ATTRS]) (struct device *dev, - struct device_attribute *devattr, const char *buf, - size_t count) = { NULL, NULL, NULL, NULL, NULL, - store_ttarget, store_tmin }; + show_ttarget }; static const char *names[TOTAL_ATTRS] = { "temp%d_label", "temp%d_crit_alarm", "temp%d_input", "temp%d_crit", - "temp%d_max_alarm", "temp%d_max", - "temp%d_max_hyst" }; + "temp%d_max" }; for (i = 0; i < tdata->attr_size; i++) { snprintf(tdata->attr_name[i], CORETEMP_NAME_LENGTH, names[i], @@ -462,10 +362,6 @@ static int create_core_attrs(struct temp_data *tdata, struct device *dev, sysfs_attr_init(&tdata->sd_attrs[i].dev_attr.attr); tdata->sd_attrs[i].dev_attr.attr.name = tdata->attr_name[i]; tdata->sd_attrs[i].dev_attr.attr.mode = S_IRUGO; - if (rw_ptr[i]) { - tdata->sd_attrs[i].dev_attr.attr.mode |= S_IWUSR; - tdata->sd_attrs[i].dev_attr.store = rw_ptr[i]; - } tdata->sd_attrs[i].dev_attr.show = rd_ptr[i]; tdata->sd_attrs[i].index = attr_no; err = device_create_file(dev, &tdata->sd_attrs[i].dev_attr); @@ -538,8 +434,6 @@ static struct temp_data *init_temp_data(unsigned int cpu, int pkg_flag) tdata->status_reg = pkg_flag ? MSR_IA32_PACKAGE_THERM_STATUS : MSR_IA32_THERM_STATUS; - tdata->intrpt_reg = pkg_flag ? MSR_IA32_PACKAGE_THERM_INTERRUPT : - MSR_IA32_THERM_INTERRUPT; tdata->is_pkg_data = pkg_flag; tdata->cpu = cpu; tdata->cpu_core_id = TO_CORE_ID(cpu); @@ -591,19 +485,18 @@ static int create_core_data(struct platform_data *pdata, tdata->tjmax = get_tjmax(c, cpu, &pdev->dev); /* - * Test if we can access the intrpt register. If so, increase the - * 'size' enough to have ttarget/tmin/max_alarm interfaces. - * Initialize ttarget with bits 16:22 of MSR_IA32_THERM_INTERRUPT + * Read the still undocumented bits 8:15 of IA32_TEMPERATURE_TARGET. + * The target temperature is available on older CPUs but not in this + * register. Atoms don't have the register at all. */ - err = rdmsr_safe_on_cpu(cpu, tdata->intrpt_reg, &eax, &edx); - if (!err) { - tdata->attr_size += MAX_THRESH_ATTRS; - tdata->tmin = tdata->tjmax - - ((eax & THERM_MASK_THRESHOLD0) >> - THERM_SHIFT_THRESHOLD0) * 1000; - tdata->ttarget = tdata->tjmax - - ((eax & THERM_MASK_THRESHOLD1) >> - THERM_SHIFT_THRESHOLD1) * 1000; + if (c->x86_model > 0xe && c->x86_model != 0x1c) { + err = rdmsr_safe_on_cpu(cpu, MSR_IA32_TEMPERATURE_TARGET, + &eax, &edx); + if (!err) { + tdata->ttarget + = tdata->tjmax - ((eax >> 8) & 0xff) * 1000; + tdata->attr_size++; + } } pdata->core_data[attr_no] = tdata; -- cgit v0.10.2 From 7cbe1ceeb6ab2e4bc334b3dab3254d4414b72a23 Mon Sep 17 00:00:00 2001 From: Chris Peterson Date: Wed, 21 Sep 2011 06:52:09 -0700 Subject: hwmon: (w83791d) rename prototype parameter from 'register' to 'reg' gcc -Wextra warns "register is not at beginning of declaration" because the compiler thinks the parameter has been marked as a 'register' variable, but the function prototype intended to name the parameter "register" (which is a reserved keyword). Signed-off-by: Chris Peterson Acked-by: Marc Hulsman Signed-off-by: Guenter Roeck diff --git a/drivers/hwmon/w83791d.c b/drivers/hwmon/w83791d.c index 17cf1ab..8c2844e 100644 --- a/drivers/hwmon/w83791d.c +++ b/drivers/hwmon/w83791d.c @@ -329,8 +329,8 @@ static int w83791d_detect(struct i2c_client *client, struct i2c_board_info *info); static int w83791d_remove(struct i2c_client *client); -static int w83791d_read(struct i2c_client *client, u8 register); -static int w83791d_write(struct i2c_client *client, u8 register, u8 value); +static int w83791d_read(struct i2c_client *client, u8 reg); +static int w83791d_write(struct i2c_client *client, u8 reg, u8 value); static struct w83791d_data *w83791d_update_device(struct device *dev); #ifdef DEBUG -- cgit v0.10.2 From cc41d586e8b4d76164fe7731c1c49be6cc5fc7e6 Mon Sep 17 00:00:00 2001 From: Roland Stigge Date: Wed, 21 Sep 2011 13:06:11 -0400 Subject: hwmon: (ds620) Fix handling of negative temperatures Signed (negative) temperatures were not handled correctly. Signed-off-by: Roland Stigge Signed-off-by: Guenter Roeck Cc: stable@kernel.org # v2.6.38+ diff --git a/drivers/hwmon/ds620.c b/drivers/hwmon/ds620.c index 257957c..4f7c3fc 100644 --- a/drivers/hwmon/ds620.c +++ b/drivers/hwmon/ds620.c @@ -72,7 +72,7 @@ struct ds620_data { char valid; /* !=0 if following fields are valid */ unsigned long last_updated; /* In jiffies */ - u16 temp[3]; /* Register values, word */ + s16 temp[3]; /* Register values, word */ }; /* -- cgit v0.10.2 From 5ebbf72dc51bd3b481aa91fea37a7157da5fc548 Mon Sep 17 00:00:00 2001 From: archit taneja Date: Fri, 5 Aug 2011 04:19:21 -0300 Subject: [media] OMAP_VOUT: Fix build break caused by update_mode removal in DSS2 The DSS2 driver does not support the configuration of the update_mode of a panel anymore. Remove the setting of update_mode done in omap_vout_probe(). Ignore configuration of TE since omap_vout driver doesn't support manual update displays anyway. Signed-off-by: Archit Taneja Tested-by: Koen Kooi Signed-off-by: Vaibhav Hiremath Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/omap/omap_vout.c b/drivers/media/video/omap/omap_vout.c index b5ef362..b3a5ecd 100644 --- a/drivers/media/video/omap/omap_vout.c +++ b/drivers/media/video/omap/omap_vout.c @@ -2194,19 +2194,6 @@ static int __init omap_vout_probe(struct platform_device *pdev) "'%s' Display already enabled\n", def_display->name); } - /* set the update mode */ - if (def_display->caps & - OMAP_DSS_DISPLAY_CAP_MANUAL_UPDATE) { - if (dssdrv->enable_te) - dssdrv->enable_te(def_display, 0); - if (dssdrv->set_update_mode) - dssdrv->set_update_mode(def_display, - OMAP_DSS_UPDATE_MANUAL); - } else { - if (dssdrv->set_update_mode) - dssdrv->set_update_mode(def_display, - OMAP_DSS_UPDATE_AUTO); - } } } -- cgit v0.10.2 From dd182e5416e872e30d90cf071758eec1cf6340d5 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Sat, 16 Jul 2011 00:51:00 -0300 Subject: [media] uvcvideo: Set alternate setting 0 on resume if the bus has been reset If the bus has been reset on resume, set the alternate setting to 0. This should be the default value, but some devices crash or otherwise misbehave if they don't receive a SET_INTERFACE request before any other video control request. Microdia's 0c45:6437 camera has been found to require this change or it will stop sending video data after resume. uvc_video.c] Signed-off-by: Ming Lei Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/uvc/uvc_driver.c b/drivers/media/video/uvc/uvc_driver.c index d29f9c2..e4100b1 100644 --- a/drivers/media/video/uvc/uvc_driver.c +++ b/drivers/media/video/uvc/uvc_driver.c @@ -1961,7 +1961,7 @@ static int __uvc_resume(struct usb_interface *intf, int reset) list_for_each_entry(stream, &dev->streams, list) { if (stream->intf == intf) - return uvc_video_resume(stream); + return uvc_video_resume(stream, reset); } uvc_trace(UVC_TRACE_SUSPEND, "Resume: video streaming USB interface " diff --git a/drivers/media/video/uvc/uvc_video.c b/drivers/media/video/uvc/uvc_video.c index 8244167..ffd1158 100644 --- a/drivers/media/video/uvc/uvc_video.c +++ b/drivers/media/video/uvc/uvc_video.c @@ -1104,10 +1104,18 @@ int uvc_video_suspend(struct uvc_streaming *stream) * buffers, making sure userspace applications are notified of the problem * instead of waiting forever. */ -int uvc_video_resume(struct uvc_streaming *stream) +int uvc_video_resume(struct uvc_streaming *stream, int reset) { int ret; + /* If the bus has been reset on resume, set the alternate setting to 0. + * This should be the default value, but some devices crash or otherwise + * misbehave if they don't receive a SET_INTERFACE request before any + * other video control request. + */ + if (reset) + usb_set_interface(stream->dev->udev, stream->intfnum, 0); + stream->frozen = 0; ret = uvc_commit_video(stream, &stream->ctrl); diff --git a/drivers/media/video/uvc/uvcvideo.h b/drivers/media/video/uvc/uvcvideo.h index df32a43..cbdd49b 100644 --- a/drivers/media/video/uvc/uvcvideo.h +++ b/drivers/media/video/uvc/uvcvideo.h @@ -638,7 +638,7 @@ extern void uvc_mc_cleanup_entity(struct uvc_entity *entity); /* Video */ extern int uvc_video_init(struct uvc_streaming *stream); extern int uvc_video_suspend(struct uvc_streaming *stream); -extern int uvc_video_resume(struct uvc_streaming *stream); +extern int uvc_video_resume(struct uvc_streaming *stream, int reset); extern int uvc_video_enable(struct uvc_streaming *stream, int enable); extern int uvc_probe_video(struct uvc_streaming *stream, struct uvc_streaming_control *probe); -- cgit v0.10.2 From 8280b662df96f4172c4972b14a4aec0daf272b8f Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 6 Sep 2011 10:23:18 -0300 Subject: [media] v4l: Fix use-after-free case in v4l2_device_release Drivers that have no v4l2_device release callback might free the v4l2_device instance in the video_device release callback. Make sure we don't access the v4l2_device instance after it gets freed. Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/v4l2-dev.c b/drivers/media/video/v4l2-dev.c index 06f1400..d721565 100644 --- a/drivers/media/video/v4l2-dev.c +++ b/drivers/media/video/v4l2-dev.c @@ -173,6 +173,17 @@ static void v4l2_device_release(struct device *cd) media_device_unregister_entity(&vdev->entity); #endif + /* Do not call v4l2_device_put if there is no release callback set. + * Drivers that have no v4l2_device release callback might free the + * v4l2_dev instance in the video_device release callback below, so we + * must perform this check here. + * + * TODO: In the long run all drivers that use v4l2_device should use the + * v4l2_device release callback. This check will then be unnecessary. + */ + if (v4l2_dev->release == NULL) + v4l2_dev = NULL; + /* Release video_device and perform other cleanups as needed. */ vdev->release(vdev); -- cgit v0.10.2 From 236c5441d70312242b5b32d229259930046a80e5 Mon Sep 17 00:00:00 2001 From: Dave Young Date: Tue, 6 Sep 2011 09:08:08 -0300 Subject: [media] v4l: Make sure we hold a reference to the v4l2_device before using it v4l2_device_disconnect() calls dev_get_drvdata() and dev_set_drvdata() on the device it received in v4l2_device_register(). Get a reference to the device in v4l2_device_register() to make sure it won't disappear as long as we need it. Reported-by: Sitsofe Wheeler Signed-off-by: Dave Young Tested-by: Sitsofe Wheeler Acked-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/v4l2-device.c b/drivers/media/video/v4l2-device.c index c72856c..e6a2c3b 100644 --- a/drivers/media/video/v4l2-device.c +++ b/drivers/media/video/v4l2-device.c @@ -38,6 +38,7 @@ int v4l2_device_register(struct device *dev, struct v4l2_device *v4l2_dev) mutex_init(&v4l2_dev->ioctl_lock); v4l2_prio_init(&v4l2_dev->prio); kref_init(&v4l2_dev->ref); + get_device(dev); v4l2_dev->dev = dev; if (dev == NULL) { /* If dev == NULL, then name must be filled in by the caller */ @@ -93,6 +94,7 @@ void v4l2_device_disconnect(struct v4l2_device *v4l2_dev) if (dev_get_drvdata(v4l2_dev->dev) == v4l2_dev) dev_set_drvdata(v4l2_dev->dev, NULL); + put_device(v4l2_dev->dev); v4l2_dev->dev = NULL; } EXPORT_SYMBOL_GPL(v4l2_device_disconnect); -- cgit v0.10.2 From 4d9b2ebd335d83044b9e6656d0e604e8e1300334 Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Tue, 6 Sep 2011 19:16:18 -0300 Subject: [media] uvcvideo: Fix crash when linking entities The uvc_mc_register_entity() function wrongfully selects the media_entity associated with a UVC entity when creating links. This results in access to uninitialized media_entity structures and can hit a BUG_ON statement in media_entity_create_link(). Fix it. Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/uvc/uvc_entity.c b/drivers/media/video/uvc/uvc_entity.c index 48fea37..29e2399 100644 --- a/drivers/media/video/uvc/uvc_entity.c +++ b/drivers/media/video/uvc/uvc_entity.c @@ -49,7 +49,7 @@ static int uvc_mc_register_entity(struct uvc_video_chain *chain, if (remote == NULL) return -EINVAL; - source = (UVC_ENTITY_TYPE(remote) != UVC_TT_STREAMING) + source = (UVC_ENTITY_TYPE(remote) == UVC_TT_STREAMING) ? (remote->vdev ? &remote->vdev->entity : NULL) : &remote->subdev.entity; if (source == NULL) -- cgit v0.10.2 From e74d83aad3709a17d68f01481f2b5f240250b1c3 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Tue, 6 Sep 2011 11:02:15 -0300 Subject: [media] omap3isp: Fix build error in ispccdc.c The following build error occurs with 3.1-rc5: CC drivers/media/video/omap3isp/ispccdc.o This patch adds the missing 'linux/slab.h' include to fix the problem. Signed-off-by: Joerg Roedel Acked-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/omap3isp/ispccdc.c b/drivers/media/video/omap3isp/ispccdc.c index 9d3459d..80796eb 100644 --- a/drivers/media/video/omap3isp/ispccdc.c +++ b/drivers/media/video/omap3isp/ispccdc.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include "isp.h" -- cgit v0.10.2 From d0efab26f89506387a1bde898556660e06d7eb15 Mon Sep 17 00:00:00 2001 From: Vasily Averin Date: Fri, 2 Sep 2011 19:31:46 +0400 Subject: [SCSI] aacraid: reset should disable MSI interrupt scsi reset on hardware with enabled MSI interrupts generates WARNING message [11027.798722] aacraid: Host adapter abort request (0,0,0,0) [11027.798814] aacraid: Host adapter reset request. SCSI hang ? [11087.762237] aacraid: SCSI bus appears hung [11135.082543] ------------[ cut here ]------------ [11135.082646] WARNING: at drivers/pci/msi.c:658 pci_enable_msi_block+0x251/0x290() Signed-off-by: Vasily Averin Acked-by: Mark Salyzyn Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/scsi/aacraid/commsup.c b/drivers/scsi/aacraid/commsup.c index e7d0d47..e5f2d7d 100644 --- a/drivers/scsi/aacraid/commsup.c +++ b/drivers/scsi/aacraid/commsup.c @@ -1283,6 +1283,8 @@ static int _aac_reset_adapter(struct aac_dev *aac, int forced) kfree(aac->queues); aac->queues = NULL; free_irq(aac->pdev->irq, aac); + if (aac->msi) + pci_disable_msi(aac->pdev); kfree(aac->fsa_dev); aac->fsa_dev = NULL; quirks = aac_get_driver_ident(index)->quirks; -- cgit v0.10.2 From 24926dadc41cc566e974022b0e66231b82c6375f Mon Sep 17 00:00:00 2001 From: Mark Salyzyn Date: Thu, 1 Sep 2011 06:11:17 -0700 Subject: [SCSI] libsas: fix failure to revalidate domain for anything but the first expander child. In an enclosure model where there are chaining expanders to a large body of storage, it was discovered that libsas, responding to a broadcast event change, would only revalidate the domain of first child expander in the list. The issue is that the pointer value to the discovered source device was used to break out of the loop, rather than the content of the pointer. This still remains non-compliant as the revalidate domain code is supposed to loop through all child expanders, and not stop at the first one it finds that reports a change count. However, the design of this routine does not allow multiple device discoveries and that would be a more complicated set of patches reserved for another day. We are fixing the glaring bug rather than refactoring the code. Signed-off-by: Mark Salyzyn Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c index f84084b..c9e3dc0 100644 --- a/drivers/scsi/libsas/sas_expander.c +++ b/drivers/scsi/libsas/sas_expander.c @@ -1721,7 +1721,7 @@ static int sas_find_bcast_dev(struct domain_device *dev, list_for_each_entry(ch, &ex->children, siblings) { if (ch->dev_type == EDGE_DEV || ch->dev_type == FANOUT_DEV) { res = sas_find_bcast_dev(ch, src_dev); - if (src_dev) + if (*src_dev) return res; } } -- cgit v0.10.2 From 3538a001ea7db13fa1be2966b71f69d808acff01 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 31 Aug 2011 15:02:00 -0700 Subject: [SCSI] scsi: qla4xxx needs libiscsi.o qla4xxx driver needs to be linked with libiscsi.o to fix build errors. This happens when no other drivers that use libiscsi.o are enabled. ERROR: "iscsi_conn_stop" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_conn_get_addr_param" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_session_teardown" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_host_alloc" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_conn_start" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_conn_send_pdu" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_session_get_param" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_conn_get_param" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_set_param" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_session_failure" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_complete_pdu" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_session_setup" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_conn_bind" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_conn_setup" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! ERROR: "iscsi_itt_to_task" [drivers/scsi/qla4xxx/qla4xxx.ko] undefined! Signed-off-by: Randy Dunlap Reviewed-by: Mike Christie Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 3c08f53..6153a66 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -88,7 +88,7 @@ obj-$(CONFIG_SCSI_QLOGIC_FAS) += qlogicfas408.o qlogicfas.o obj-$(CONFIG_PCMCIA_QLOGIC) += qlogicfas408.o obj-$(CONFIG_SCSI_QLOGIC_1280) += qla1280.o obj-$(CONFIG_SCSI_QLA_FC) += qla2xxx/ -obj-$(CONFIG_SCSI_QLA_ISCSI) += qla4xxx/ +obj-$(CONFIG_SCSI_QLA_ISCSI) += libiscsi.o qla4xxx/ obj-$(CONFIG_SCSI_LPFC) += lpfc/ obj-$(CONFIG_SCSI_BFA_FC) += bfa/ obj-$(CONFIG_SCSI_PAS16) += pas16.o -- cgit v0.10.2 From 2ba34e43ba0469086d1ff81c13a9aa9070c1a0e1 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Thu, 22 Sep 2011 14:38:58 +0100 Subject: ALSA: fm801: Fix double free in case of error in tuner detection Commit 9676001559fce06e37c7dc230ab275f605556176 ("ALSA: fm801: add error handling if auto-detect fails") added incorrect error handling. Once we have successfully called snd_device_new(), the cleanup function fm801_free() will automatically be called by snd_card_free() and we must *not* also call fm801_free() directly. Reported-by: Hor Jiun Shyong References: http://bugs.debian.org/641946 Signed-off-by: Ben Hutchings Cc: stable@kernel.org [v3.0+] Signed-off-by: Takashi Iwai diff --git a/sound/pci/fm801.c b/sound/pci/fm801.c index f9123f0..a96456d 100644 --- a/sound/pci/fm801.c +++ b/sound/pci/fm801.c @@ -1236,7 +1236,6 @@ static int __devinit snd_fm801_create(struct snd_card *card, (tea575x_tuner & TUNER_TYPE_MASK) < 4) { if (snd_tea575x_init(&chip->tea)) { snd_printk(KERN_ERR "TEA575x radio not found\n"); - snd_fm801_free(chip); return -ENODEV; } } else if ((tea575x_tuner & TUNER_TYPE_MASK) == 0) { @@ -1251,7 +1250,6 @@ static int __devinit snd_fm801_create(struct snd_card *card, } if (tea575x_tuner == 4) { snd_printk(KERN_ERR "TEA575x radio not found\n"); - snd_fm801_free(chip); return -ENODEV; } } -- cgit v0.10.2 From c37279b92aba2893578f61076cd2eef5c5fa0e99 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Thu, 22 Sep 2011 14:39:52 +0100 Subject: ALSA: fm801: Gracefully handle failure of tuner auto-detect Commit 9676001559fce06e37c7dc230ab275f605556176 ("ALSA: fm801: add error handling if auto-detect fails") seems to break systems that were previously working without a tuner. As a bonus, this should fix init and cleanup for the case where the tuner is explicitly disabled. Reported-and-tested-by: Hor Jiun Shyong References: http://bugs.debian.org/641946 Signed-off-by: Ben Hutchings Cc: stable@kernel.org [v3.0+] Signed-off-by: Takashi Iwai diff --git a/sound/pci/fm801.c b/sound/pci/fm801.c index a96456d..32b02d9 100644 --- a/sound/pci/fm801.c +++ b/sound/pci/fm801.c @@ -68,6 +68,7 @@ MODULE_PARM_DESC(enable, "Enable FM801 soundcard."); module_param_array(tea575x_tuner, int, NULL, 0444); MODULE_PARM_DESC(tea575x_tuner, "TEA575x tuner access method (0 = auto, 1 = SF256-PCS, 2=SF256-PCP, 3=SF64-PCR, 8=disable, +16=tuner-only)."); +#define TUNER_DISABLED (1<<3) #define TUNER_ONLY (1<<4) #define TUNER_TYPE_MASK (~TUNER_ONLY & 0xFFFF) @@ -1150,7 +1151,8 @@ static int snd_fm801_free(struct fm801 *chip) __end_hw: #ifdef CONFIG_SND_FM801_TEA575X_BOOL - snd_tea575x_exit(&chip->tea); + if (!(chip->tea575x_tuner & TUNER_DISABLED)) + snd_tea575x_exit(&chip->tea); #endif if (chip->irq >= 0) free_irq(chip->irq, chip); @@ -1250,10 +1252,15 @@ static int __devinit snd_fm801_create(struct snd_card *card, } if (tea575x_tuner == 4) { snd_printk(KERN_ERR "TEA575x radio not found\n"); - return -ENODEV; + chip->tea575x_tuner = TUNER_DISABLED; } } - strlcpy(chip->tea.card, snd_fm801_tea575x_gpios[(tea575x_tuner & TUNER_TYPE_MASK) - 1].name, sizeof(chip->tea.card)); + if (!(chip->tea575x_tuner & TUNER_DISABLED)) { + strlcpy(chip->tea.card, + snd_fm801_tea575x_gpios[(tea575x_tuner & + TUNER_TYPE_MASK) - 1].name, + sizeof(chip->tea.card)); + } #endif *rchip = chip; -- cgit v0.10.2 From 8b267b312df9343fea3bd679c509b36214b5a854 Mon Sep 17 00:00:00 2001 From: Antonio Quartulli Date: Wed, 21 Sep 2011 16:06:42 +0200 Subject: batman-adv: do_bcast has to be true for broadcast packets only corrects a critical bug of the GW feature. This bug made all the unicast packets destined to a GW to be sent as broadcast. This bug is present even if the sender GW feature is configured as OFF. It's an urgent bug fix and should be committed as soon as possible. This was a regression introduced by 43676ab590c3f8686fd047d34c3e33803eef71f0 Signed-off-by: Antonio Quartulli Signed-off-by: Marek Lindner diff --git a/net/batman-adv/soft-interface.c b/net/batman-adv/soft-interface.c index 3e2f91f..05dd351 100644 --- a/net/batman-adv/soft-interface.c +++ b/net/batman-adv/soft-interface.c @@ -565,7 +565,7 @@ static int interface_tx(struct sk_buff *skb, struct net_device *soft_iface) struct orig_node *orig_node = NULL; int data_len = skb->len, ret; short vid = -1; - bool do_bcast = false; + bool do_bcast; if (atomic_read(&bat_priv->mesh_state) != MESH_ACTIVE) goto dropped; @@ -598,15 +598,15 @@ static int interface_tx(struct sk_buff *skb, struct net_device *soft_iface) tt_local_add(soft_iface, ethhdr->h_source); orig_node = transtable_search(bat_priv, ethhdr->h_dest); - if (is_multicast_ether_addr(ethhdr->h_dest) || - (orig_node && orig_node->gw_flags)) { + do_bcast = is_multicast_ether_addr(ethhdr->h_dest); + if (do_bcast || (orig_node && orig_node->gw_flags)) { ret = gw_is_target(bat_priv, skb, orig_node); if (ret < 0) goto dropped; - if (ret == 0) - do_bcast = true; + if (ret) + do_bcast = false; } /* ethernet packet should be broadcasted */ -- cgit v0.10.2 From a7f4d00a82feb5b311f765bf9522bc55bee0684f Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 22 Sep 2011 21:47:38 +0200 Subject: zorro: Defer device_register() until all devices have been identified MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As the Amiga Zorro II address space is limited to 8.5 MiB and Zorro devices can contain only one BAR, several Amiga Zorro II expansion boards (mainly graphics cards) contain multiple Zorro devices: a small one for the control registers and one (or more) for the graphics memory. The conversion of cirrusfb to the new driver framework introduced a regression: the driver contains a zorro_driver for the first Zorro device, and uses the (old) zorro_find_device() call to find the second Zorro device. However, as the Zorro core calls device_register() as soon as a Zorro device is identified, it may not have identified the second Zorro device belonging to the same physical Zorro expansion card. Hence cirrusfb could no longer find the second part of the Picasso II graphics card, causing a NULL pointer dereference. Defer the registration of Zorro devices with the driver framework until all Zorro devices have been identified to fix this. Note that the alternative solution (modifying cirrusfb to register a zorro_driver for all Zorro devices belonging to a graphics card, instead of only for the first one, and adding a synchronization mechanism to defer initialization until all have been found), is not an option, as on some cards one device may be optional (e.g. the second bank of 2 MiB of graphics memory on the Picasso IV in Zorro II mode). Reported-by: Ingo Jürgensmann Signed-off-by: Geert Uytterhoeven Cc: stable@kernel.org Signed-off-by: Linus Torvalds diff --git a/drivers/zorro/zorro.c b/drivers/zorro/zorro.c index e0c2807..181fa81 100644 --- a/drivers/zorro/zorro.c +++ b/drivers/zorro/zorro.c @@ -148,10 +148,10 @@ static int __init amiga_zorro_probe(struct platform_device *pdev) } platform_set_drvdata(pdev, bus); - /* Register all devices */ pr_info("Zorro: Probing AutoConfig expansion devices: %u device%s\n", zorro_num_autocon, zorro_num_autocon == 1 ? "" : "s"); + /* First identify all devices ... */ for (i = 0; i < zorro_num_autocon; i++) { z = &zorro_autocon[i]; z->id = (z->rom.er_Manufacturer<<16) | (z->rom.er_Product<<8); @@ -172,6 +172,11 @@ static int __init amiga_zorro_probe(struct platform_device *pdev) dev_set_name(&z->dev, "%02x", i); z->dev.parent = &bus->dev; z->dev.bus = &zorro_bus_type; + } + + /* ... then register them */ + for (i = 0; i < zorro_num_autocon; i++) { + z = &zorro_autocon[i]; error = device_register(&z->dev); if (error) { dev_err(&bus->dev, "Error registering device %s\n", -- cgit v0.10.2 From e369fde1af20c7c67acff3bc4a36e451f38ae7d0 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Thu, 22 Sep 2011 14:11:38 -0700 Subject: thp: fix khugepaged defrag tunable documentation Commit e27e6151b154 ("mm/thp: use conventional format for boolean attributes") changed /sys/kernel/mm/transparent_hugepage/khugepaged/defrag to be tuned by using 1 (enabled) or 0 (disabled) instead of "yes" and "no", respectively. Update the documentation. Signed-off-by: David Rientjes Signed-off-by: Linus Torvalds diff --git a/Documentation/vm/transhuge.txt b/Documentation/vm/transhuge.txt index 0924aac..29bdf62 100644 --- a/Documentation/vm/transhuge.txt +++ b/Documentation/vm/transhuge.txt @@ -123,10 +123,11 @@ be automatically shutdown if it's set to "never". khugepaged runs usually at low frequency so while one may not want to invoke defrag algorithms synchronously during the page faults, it should be worth invoking defrag at least in khugepaged. However it's -also possible to disable defrag in khugepaged: +also possible to disable defrag in khugepaged by writing 0 or enable +defrag in khugepaged by writing 1: -echo yes >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag -echo no >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag +echo 0 >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag +echo 1 >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag You can also control how many pages khugepaged should scan at each pass: -- cgit v0.10.2 From 5ce5ed35939ece8a9c7e0f10c7f3d5bc000c6b64 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Mon, 8 Aug 2011 13:08:19 +0200 Subject: TPM: TCG_ATMEL should depend on HAS_IOPORT MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On m68k, I get: drivers/char/tpm/tpm_atmel.h: In function ‘atmel_get_base_addr’: drivers/char/tpm/tpm_atmel.h:129: error: implicit declaration of function ‘ioport_map’ drivers/char/tpm/tpm_atmel.h:129: warning: return makes pointer from integer without a cast The code in tpm_atmel.h supports PPC64 (using the device tree and ioremap()) and "anything else" (using ioport_map()). However, ioportmap() is only available on platforms that set HAS_IOPORT. Although PC64 seems to have HAS_IOPORT, a "depends on HAS_IOPORT" should work, but I think it's better to expose the special PPC64 handling explicit using "depends on PPC64 || HAS_IOPORT". Signed-off-by: Geert Uytterhoeven Signed-off-by: Rajiv Andrade Signed-off-by: James Morris diff --git a/drivers/char/tpm/Kconfig b/drivers/char/tpm/Kconfig index f6595ab..fa567f1 100644 --- a/drivers/char/tpm/Kconfig +++ b/drivers/char/tpm/Kconfig @@ -43,6 +43,7 @@ config TCG_NSC config TCG_ATMEL tristate "Atmel TPM Interface" + depends on PPC64 || HAS_IOPORT ---help--- If you have a TPM security chip from Atmel say Yes and it will be accessible from within Linux. To compile this driver -- cgit v0.10.2 From de69113ec1896443c732e8b812e8005fb44eeeeb Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 4 Aug 2011 07:58:07 +0800 Subject: TPM: tpm_nsc: Fix a double free of pdev in cleanup_nsc platform_device_unregister() will release all resources and remove it from the subsystem, then drop reference count by calling platform_device_put(). We should not call kfree(pdev) after platform_device_unregister(pdev). Signed-off-by: Axel Lin Signed-off-by: Rajiv Andrade Signed-off-by: James Morris diff --git a/drivers/char/tpm/tpm_nsc.c b/drivers/char/tpm/tpm_nsc.c index 82facc9..4d24648 100644 --- a/drivers/char/tpm/tpm_nsc.c +++ b/drivers/char/tpm/tpm_nsc.c @@ -396,8 +396,6 @@ static void __exit cleanup_nsc(void) if (pdev) { tpm_nsc_remove(&pdev->dev); platform_device_unregister(pdev); - kfree(pdev); - pdev = NULL; } platform_driver_unregister(&nsc_drv); -- cgit v0.10.2 From 6b07d30aca7e52f2881b8c8c20c8a2cd28e8b3d3 Mon Sep 17 00:00:00 2001 From: Peter Huewe Date: Thu, 15 Sep 2011 14:37:43 -0300 Subject: TPM: Call tpm_transmit with correct size This patch changes the call of tpm_transmit by supplying the size of the userspace buffer instead of TPM_BUFSIZE. This got assigned CVE-2011-1161. [The first hunk didn't make sense given one could expect way less data than TPM_BUFSIZE, so added tpm_transmit boundary check over bufsiz instead The last parameter of tpm_transmit() reflects the amount of data expected from the device, and not the buffer size being supplied to it. It isn't ideal to parse it directly, so we just set it to the maximum the input buffer can handle and let the userspace API to do such job.] Signed-off-by: Rajiv Andrade Cc: Stable Kernel Signed-off-by: James Morris diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c index caf8012..1fe9793 100644 --- a/drivers/char/tpm/tpm.c +++ b/drivers/char/tpm/tpm.c @@ -383,6 +383,9 @@ static ssize_t tpm_transmit(struct tpm_chip *chip, const char *buf, u32 count, ordinal; unsigned long stop; + if (bufsiz > TPM_BUFSIZE) + bufsiz = TPM_BUFSIZE; + count = be32_to_cpu(*((__be32 *) (buf + 2))); ordinal = be32_to_cpu(*((__be32 *) (buf + 6))); if (count == 0) -- cgit v0.10.2 From 3321c07ae5068568cd61ac9f4ba749006a7185c9 Mon Sep 17 00:00:00 2001 From: Peter Huewe Date: Thu, 15 Sep 2011 14:47:42 -0300 Subject: TPM: Zero buffer after copying to userspace Since the buffer might contain security related data it might be a good idea to zero the buffer after we have copied it to userspace. This got assigned CVE-2011-1162. Signed-off-by: Rajiv Andrade Cc: Stable Kernel Signed-off-by: James Morris diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c index 1fe9793..9ca5c02 100644 --- a/drivers/char/tpm/tpm.c +++ b/drivers/char/tpm/tpm.c @@ -1105,6 +1105,7 @@ ssize_t tpm_read(struct file *file, char __user *buf, { struct tpm_chip *chip = file->private_data; ssize_t ret_size; + int rc; del_singleshot_timer_sync(&chip->user_read_timer); flush_work_sync(&chip->work); @@ -1115,8 +1116,11 @@ ssize_t tpm_read(struct file *file, char __user *buf, ret_size = size; mutex_lock(&chip->buffer_mutex); - if (copy_to_user(buf, chip->data_buffer, ret_size)) + rc = copy_to_user(buf, chip->data_buffer, ret_size); + memset(chip->data_buffer, 0, ret_size); + if (rc) ret_size = -EFAULT; + mutex_unlock(&chip->buffer_mutex); } -- cgit v0.10.2 From 34c869855a28af3c6031d29e2267b5a3cce0d67d Mon Sep 17 00:00:00 2001 From: Jarkko Nikula Date: Fri, 23 Sep 2011 11:19:13 +0300 Subject: ASoC: omap-mcbsp: Do not attempt to change DAI sysclk if stream is active Attempt to change McBSP CLKS source while another stream is active is not safe after commit d135865 ("OMAP: McBSP: implement functional clock switching via clock framework") in 2.6.37. CLKS parent clock switching using clock framework have to idle the McBSP before switching and then activate it again. This short break can cause a DMA transaction error to already running stream which halts and recovers only by closing and restarting the stream. This goes more fatal after commit e2fa61d ("OMAP3: l3: Introduce l3-interconnect error handling driver") in 2.6.39 where l3 driver detects a severe timeout error and does BUG_ON(). Fix this by not changing any configuration in omap_mcbsp_dai_set_dai_sysclk if the McBSP is already active. This test should have been here just from the beginning anyway. Signed-off-by: Jarkko Nikula Acked-by: Peter Ujfalusi Signed-off-by: Mark Brown Cc: stable@kernel.org diff --git a/sound/soc/omap/omap-mcbsp.c b/sound/soc/omap/omap-mcbsp.c index ebcc2d4..478d607 100644 --- a/sound/soc/omap/omap-mcbsp.c +++ b/sound/soc/omap/omap-mcbsp.c @@ -516,6 +516,12 @@ static int omap_mcbsp_dai_set_dai_sysclk(struct snd_soc_dai *cpu_dai, struct omap_mcbsp_reg_cfg *regs = &mcbsp_data->regs; int err = 0; + if (mcbsp_data->active) + if (freq == mcbsp_data->in_freq) + return 0; + else + return -EBUSY; + /* The McBSP signal muxing functions are only available on McBSP1 */ if (clk_id == OMAP_MCBSP_CLKR_SRC_CLKR || clk_id == OMAP_MCBSP_CLKR_SRC_CLKX || -- cgit v0.10.2 From d9ad77ebfd3a304306f9f18ff3f9e522da075e27 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Fri, 23 Sep 2011 14:00:54 +0100 Subject: Revert "drm/radeon/kms: fix typo in r100_blit_copy" This reverts commit 18b4fada275dd2b6dd9db904ddf70fe39e272222. This code was correct, apologies to anyone who noticed things broke. revert contents are different due to another commit in between. Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index 5b1837b..7fcdbbb 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -773,8 +773,8 @@ int r100_copy_blit(struct radeon_device *rdev, radeon_ring_write(rdev, (0x1fff) | (0x1fff << 16)); radeon_ring_write(rdev, 0); radeon_ring_write(rdev, (0x1fff) | (0x1fff << 16)); - radeon_ring_write(rdev, cur_pages); - radeon_ring_write(rdev, cur_pages); + radeon_ring_write(rdev, num_gpu_pages); + radeon_ring_write(rdev, num_gpu_pages); radeon_ring_write(rdev, cur_pages | (stride_pixels << 16)); } radeon_ring_write(rdev, PACKET0(RADEON_DSTCACHE_CTLSTAT, 0)); -- cgit v0.10.2 From fdfc61594e6de21c23f955818ef78bcab9bafe40 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 22 Sep 2011 10:47:23 -0400 Subject: drm/radeon/kms: fix DDIA enable on some rs690 systems DVOOutputControl checks the value of of bios scratch reg 3 on some tables and assumes the encoder is already enabled if the DFP2_ACTIVE bit is set. Clear that bit so the table sets the DDIA enable bit properly. Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 319d85d..13690f3 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -1507,7 +1507,14 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode) switch (mode) { case DRM_MODE_DPMS_ON: args.ucAction = ATOM_ENABLE; - atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); + /* workaround for DVOOutputControl on some RS690 systems */ + if (radeon_encoder->encoder_id == ENCODER_OBJECT_ID_INTERNAL_DDI) { + u32 reg = RREG32(RADEON_BIOS_3_SCRATCH); + WREG32(RADEON_BIOS_3_SCRATCH, reg & ~ATOM_S3_DFP2I_ACTIVE); + atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); + WREG32(RADEON_BIOS_3_SCRATCH, reg); + } else + atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) { args.ucAction = ATOM_LCD_BLON; atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); -- cgit v0.10.2 From 362e4e49abe53e89d87455dfcd7c1bbaf08a839d Mon Sep 17 00:00:00 2001 From: Thomas Pfaff Date: Thu, 22 Sep 2011 18:26:06 +0200 Subject: ALSA: usb-audio - clear chip->probing on error exit The Terratec Aureon 5.1 USB sound card support is broken since kernel 2.6.39. 2.6.39 introduced power management support for USB sound cards that added a probing flag in struct snd_usb_audio. During the probe of the card it gives following error message : usb 7-2: new full speed USB device number 2 using uhci_hcd cannot find UAC_HEADER snd-usb-audio: probe of 7-2:1.3 failed with error -5 input: USB Audio as /devices/pci0000:00/0000:00:1d.1/usb7/7-2/7-2:1.3/input/input6 generic-usb 0003:0CCD:0028.0001: input: USB HID v1.00 Device [USB Audio] on usb-0000:00:1d.1-2/input3 I can not comment about that "cannot find UAC_HEADER" error, but until 2.6.38 the card worked anyway. With 2.6.39 chip->probing remains 1 on error exit, and any later ioctl stops in snd_usb_autoresume with -ENODEV. Signed-off-by: Thomas Pfaff Cc: [2.6.39+] Signed-off-by: Takashi Iwai diff --git a/sound/usb/card.c b/sound/usb/card.c index 781d9e6..ed120ca 100644 --- a/sound/usb/card.c +++ b/sound/usb/card.c @@ -532,6 +532,7 @@ snd_usb_audio_probe(struct usb_device *dev, __error: if (chip && !chip->num_interfaces) snd_card_free(chip->card); + chip->probing = 0; mutex_unlock(®ister_mutex); __err_val: return NULL; -- cgit v0.10.2 From f66fedcb723a9f118170200e21dbabb305f8c702 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sat, 20 Aug 2011 14:39:23 +0900 Subject: perf probe: Fix regression of variable finder Fix to call convert_variable() if previous call does not fail. To call convert_variable, it ensures "ret" is 0. However, since "ret" has the return value of synthesize_perf_probe_arg() which always returns positive value if it succeeded, perf probe doesn't call convert_variable(). This will cause a SEGV when we add an event with arguments. This has to be fixed as it ensures "ret" is greater than 0 (or not negative). This regression has been introduced by my previous patch, f182e3e1. Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Paul Mackerras Cc: Pekka Enberg Cc: Peter Zijlstra Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20110820053922.3286.65805.stgit@fedora15 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 555fc38..5d73262 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -659,7 +659,7 @@ static int find_variable(Dwarf_Die *sc_die, struct probe_finder *pf) if (!die_find_variable_at(&pf->cu_die, pf->pvar->var, 0, &vr_die)) ret = -ENOENT; } - if (ret == 0) + if (ret >= 0) ret = convert_variable(&vr_die, pf); if (ret < 0) -- cgit v0.10.2 From adb091846318f86e4f46c7d6a7b40d2f478abdbe Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Wed, 24 Aug 2011 16:40:14 +1000 Subject: perf symbols: Fix ppc64 SEGV in dso__load_sym with debuginfo files 64bit PowerPC debuginfo files have an empty function descriptor section. I hit a SEGV when perf tried to use this section for symbol resolution. To fix this we need to check the section is valid and we can do this by checking for type SHT_PROGBITS. Cc: Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Eric B Munson Link: http://lkml.kernel.org/r/20110824065242.895239970@samba.org Signed-off-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 469c026..bb5d32f 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1111,6 +1111,8 @@ static int dso__load_sym(struct dso *dso, struct map *map, const char *name, } opdsec = elf_section_by_name(elf, &ehdr, &opdshdr, ".opd", &opdidx); + if (opdshdr.sh_type != SHT_PROGBITS) + opdsec = NULL; if (opdsec) opddata = elf_rawdata(opdsec, NULL); -- cgit v0.10.2 From 3f5a42722b9e78a434d5a4ee5e607dc33c69ac80 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Wed, 24 Aug 2011 16:40:15 +1000 Subject: perf symbols: /proc/kallsyms does not sort module symbols kallsyms__parse assumes that /proc/kallsyms is sorted and sets the end of the previous symbol to the start of the current one. Unfortunately module symbols are not sorted, eg: ffffffffa0081f30 t e1000_clean_rx_irq [e1000e] ffffffffa00817a0 t e1000_alloc_rx_buffers [e1000e] Some symbols end up with a negative length and others have a length larger than they should. This results in confusing perf output. We already have a function to fixup the end of zero length symbols so use that instead. Cc: Eric B Munson Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20110824065242.969681349@samba.org Signed-off-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index bb5d32f..f119e85 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -438,18 +438,11 @@ int kallsyms__parse(const char *filename, void *arg, char *line = NULL; size_t n; int err = -1; - u64 prev_start = 0; - char prev_symbol_type = 0; - char *prev_symbol_name; FILE *file = fopen(filename, "r"); if (file == NULL) goto out_failure; - prev_symbol_name = malloc(KSYM_NAME_LEN); - if (prev_symbol_name == NULL) - goto out_close; - err = 0; while (!feof(file)) { @@ -480,24 +473,18 @@ int kallsyms__parse(const char *filename, void *arg, break; } - if (prev_symbol_type) { - u64 end = start; - if (end != prev_start) - --end; - err = process_symbol(arg, prev_symbol_name, - prev_symbol_type, prev_start, end); - if (err) - break; - } - - memcpy(prev_symbol_name, symbol_name, len + 1); - prev_symbol_type = symbol_type; - prev_start = start; + /* + * module symbols are not sorted so we add all + * symbols with zero length and rely on + * symbols__fixup_end() to fix it up. + */ + err = process_symbol(arg, symbol_name, + symbol_type, start, start); + if (err) + break; } - free(prev_symbol_name); free(line); -out_close: fclose(file); return err; @@ -703,6 +690,8 @@ int dso__load_kallsyms(struct dso *dso, const char *filename, if (dso__load_all_kallsyms(dso, filename, map) < 0) return -1; + symbols__fixup_end(&dso->symbols[map->type]); + if (dso->kernel == DSO_TYPE_GUEST_KERNEL) dso->symtab_type = SYMTAB__GUEST_KALLSYMS; else -- cgit v0.10.2 From 318779086086f46127a249878eeebb3dc80578eb Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Wed, 24 Aug 2011 16:40:16 +1000 Subject: perf symbols: Preserve symbol scope when parsing /proc/kallsyms kallsyms__parse capitalises the symbol type, so every symbol is marked global. Remove this and fix symbol_type__is_a to handle both local and global symbols. Cc: Eric B Munson Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20110824065243.077125989@samba.org Signed-off-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index f119e85..0d94ddb 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -74,11 +74,13 @@ static void dso__set_sorted_by_name(struct dso *dso, enum map_type type) bool symbol_type__is_a(char symbol_type, enum map_type map_type) { + symbol_type = toupper(symbol_type); + switch (map_type) { case MAP__FUNCTION: return symbol_type == 'T' || symbol_type == 'W'; case MAP__VARIABLE: - return symbol_type == 'D' || symbol_type == 'd'; + return symbol_type == 'D'; default: return false; } @@ -463,7 +465,7 @@ int kallsyms__parse(const char *filename, void *arg, if (len + 2 >= line_len) continue; - symbol_type = toupper(line[len]); + symbol_type = line[len]; len += 2; symbol_name = line + len; len = line_len - len; -- cgit v0.10.2 From 694bf407b06113f5e0f71764756f11903126fec0 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Wed, 24 Aug 2011 16:40:17 +1000 Subject: perf symbols: Add some heuristics for choosing the best duplicate symbol Try and pick the best symbol based on a few heuristics: - Prefer a non weak symbol over a weak one - Prefer a global symbol over a non global one - Prefer a symbol with less underscores (idea taken from kallsyms.c) - If all else fails, choose the symbol with the longest name Cc: Eric B Munson Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20110824065243.161953371@samba.org Signed-off-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 0d94ddb..870b3b8 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -86,6 +86,92 @@ bool symbol_type__is_a(char symbol_type, enum map_type map_type) } } +static int prefix_underscores_count(const char *str) +{ + const char *tail = str; + + while (*tail == '_') + tail++; + + return tail - str; +} + +#define SYMBOL_A 0 +#define SYMBOL_B 1 + +static int choose_best_symbol(struct symbol *syma, struct symbol *symb) +{ + s64 a; + s64 b; + + /* Prefer a symbol with non zero length */ + a = syma->end - syma->start; + b = symb->end - symb->start; + if ((b == 0) && (a > 0)) + return SYMBOL_A; + else if ((a == 0) && (b > 0)) + return SYMBOL_B; + + /* Prefer a non weak symbol over a weak one */ + a = syma->binding == STB_WEAK; + b = symb->binding == STB_WEAK; + if (b && !a) + return SYMBOL_A; + if (a && !b) + return SYMBOL_B; + + /* Prefer a global symbol over a non global one */ + a = syma->binding == STB_GLOBAL; + b = symb->binding == STB_GLOBAL; + if (a && !b) + return SYMBOL_A; + if (b && !a) + return SYMBOL_B; + + /* Prefer a symbol with less underscores */ + a = prefix_underscores_count(syma->name); + b = prefix_underscores_count(symb->name); + if (b > a) + return SYMBOL_A; + else if (a > b) + return SYMBOL_B; + + /* If all else fails, choose the symbol with the longest name */ + if (strlen(syma->name) >= strlen(symb->name)) + return SYMBOL_A; + else + return SYMBOL_B; +} + +static void symbols__fixup_duplicate(struct rb_root *symbols) +{ + struct rb_node *nd; + struct symbol *curr, *next; + + nd = rb_first(symbols); + + while (nd) { + curr = rb_entry(nd, struct symbol, rb_node); +again: + nd = rb_next(&curr->rb_node); + next = rb_entry(nd, struct symbol, rb_node); + + if (!nd) + break; + + if (curr->start != next->start) + continue; + + if (choose_best_symbol(curr, next) == SYMBOL_A) { + rb_erase(&next->rb_node, symbols); + goto again; + } else { + nd = rb_next(&curr->rb_node); + rb_erase(&curr->rb_node, symbols); + } + } +} + static void symbols__fixup_end(struct rb_root *symbols) { struct rb_node *nd, *prevnd = rb_first(symbols); @@ -692,6 +778,7 @@ int dso__load_kallsyms(struct dso *dso, const char *filename, if (dso__load_all_kallsyms(dso, filename, map) < 0) return -1; + symbols__fixup_duplicate(&dso->symbols[map->type]); symbols__fixup_end(&dso->symbols[map->type]); if (dso->kernel == DSO_TYPE_GUEST_KERNEL) @@ -1269,6 +1356,7 @@ new_symbol: * For misannotated, zeroed, ASM function sizes. */ if (nr > 0) { + symbols__fixup_duplicate(&dso->symbols[map->type]); symbols__fixup_end(&dso->symbols[map->type]); if (kmap) { /* -- cgit v0.10.2 From 764e16a30a77a9c8346fbae6615e7c818ce9d00f Mon Sep 17 00:00:00 2001 From: David Ahern Date: Thu, 25 Aug 2011 10:17:55 -0600 Subject: perf record: Create events initially disabled and enable after init perf-record currently creates events enabled. When doing a system wide collection (-a arg) this causes data collection for perf's initialization activities -- eg., perf_event__synthesize_threads(). For some events (e.g., context switch S/W event or tracepoints like syscalls) perf's initialization causes a lot of events to be captured frequently generating "Check IO/CPU overload!" warnings on larger systems (e.g., 2 socket, quad core, hyperthreading). perf's initialization phase can be skipped by creating events disabled and then enabling them once the initialization is done. Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/1314289075-14706-1-git-send-email-dsahern@gmail.com Signed-off-by: David Ahern Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 6b0519f..f4c3fbe 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -161,6 +161,7 @@ static void config_attr(struct perf_evsel *evsel, struct perf_evlist *evlist) struct perf_event_attr *attr = &evsel->attr; int track = !evsel->idx; /* only the first counter needs these */ + attr->disabled = 1; attr->inherit = !no_inherit; attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED | PERF_FORMAT_TOTAL_TIME_RUNNING | @@ -671,6 +672,8 @@ static int __cmd_record(int argc, const char **argv) } } + perf_evlist__enable(evsel_list); + /* * Let the child rip */ diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index c12bd47..72e9f48 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -113,6 +113,19 @@ void perf_evlist__disable(struct perf_evlist *evlist) } } +void perf_evlist__enable(struct perf_evlist *evlist) +{ + int cpu, thread; + struct perf_evsel *pos; + + for (cpu = 0; cpu < evlist->cpus->nr; cpu++) { + list_for_each_entry(pos, &evlist->entries, node) { + for (thread = 0; thread < evlist->threads->nr; thread++) + ioctl(FD(pos, cpu, thread), PERF_EVENT_IOC_ENABLE); + } + } +} + int perf_evlist__alloc_pollfd(struct perf_evlist *evlist) { int nfds = evlist->cpus->nr * evlist->threads->nr * evlist->nr_entries; diff --git a/tools/perf/util/evlist.h b/tools/perf/util/evlist.h index ce85ae9..f349150 100644 --- a/tools/perf/util/evlist.h +++ b/tools/perf/util/evlist.h @@ -54,6 +54,7 @@ int perf_evlist__mmap(struct perf_evlist *evlist, int pages, bool overwrite); void perf_evlist__munmap(struct perf_evlist *evlist); void perf_evlist__disable(struct perf_evlist *evlist); +void perf_evlist__enable(struct perf_evlist *evlist); static inline void perf_evlist__set_maps(struct perf_evlist *evlist, struct cpu_map *cpus, -- cgit v0.10.2 From 6a0e55d85babfccfd976703852ec8bab388b3a10 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 30 Aug 2011 09:15:06 +1000 Subject: perf symbols: Synthesize anonymous mmap events perf_event__synthesize_mmap_events does not create anonymous mmap events even though the kernel does. As a result an already running application with dynamically created code will not get profiled - all samples end up in the unknown bucket. This patch skips any entries with '[' in the name to avoid adding events for special regions (eg the vsyscall page). All other executable mmaps are assumed to be anonymous and an event is synthesized. Acked-by: Pekka Enberg Cc: Eric B Munson Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Pekka Enberg Link: http://lkml.kernel.org/r/20110830091506.60b51fe8@kryten Signed-off-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 3c1b8a6..437f8ca 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -169,12 +169,17 @@ static int perf_event__synthesize_mmap_events(union perf_event *event, continue; pbf += n + 3; if (*pbf == 'x') { /* vm_exec */ + char anonstr[] = "//anon\n"; char *execname = strchr(bf, '/'); /* Catch VDSO */ if (execname == NULL) execname = strstr(bf, "[vdso]"); + /* Catch anonymous mmaps */ + if ((execname == NULL) && !strstr(bf, "[")) + execname = anonstr; + if (execname == NULL) continue; -- cgit v0.10.2 From 6bb8f311a870d6042e0310309eb3d607ae52fe3e Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Wed, 31 Aug 2011 11:51:45 +1000 Subject: perf sort: Fix symbol sort output by separating unresolved samples by type I took a profile that suggested 60% of total CPU time was in the hypervisor: ... 60.20% [H] 0x33d43c 4.43% [k] ._spin_lock_irqsave 1.07% [k] ._spin_lock Using perf stat to get the user/kernel/hypervisor breakdown contradicted this. The problem is we merge all unresolved samples into the one unknown bucket. If add a comparison by sample type to sort__sym_cmp we get the real picture: ... 57.11% [.] 0x80fbf63c 4.43% [k] ._spin_lock_irqsave 1.07% [k] ._spin_lock 0.65% [H] 0x33d43c So it was almost all userspace, not hypervisor as the initial profile suggested. I found another issue while adding this. Symbol sorting sometimes shows multiple entries for the unknown bucket: ... 16.65% [.] 0x6cd3a8 7.25% [.] 0x422460 5.37% [.] yylex 4.79% [.] malloc 4.78% [.] _int_malloc 4.03% [.] _int_free 3.95% [.] hash_source_code_string 2.82% [.] 0x532908 2.64% [.] 0x36b538 0.94% [H] 0x8000000000e132a4 0.82% [H] 0x800000000000e8b0 This happens because we aren't consistent with our sorting. On one hand we check to see if both symbols match and for two unresolved samples sym is NULL so we match: if (left->ms.sym == right->ms.sym) return 0; On the other hand we use sample IP for unresolved samples when comparing against a symbol: ip_l = left->ms.sym ? left->ms.sym->start : left->ip; ip_r = right->ms.sym ? right->ms.sym->start : right->ip; This means unresolved samples end up spread across the rbtree and we can't merge them all. If we use cmp_null all unresolved samples will end up in the one bucket and the output makes more sense: ... 39.12% [.] 0x36b538 5.37% [.] yylex 4.79% [.] malloc 4.78% [.] _int_malloc 4.03% [.] _int_free 3.95% [.] hash_source_code_string 2.26% [H] 0x800000000000e8b0 Acked-by: Eric B Munson Cc: Eric B Munson Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Ian Munsie Link: http://lkml.kernel.org/r/20110831115145.4f598ab2@kryten Signed-off-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c index 401e220..1ee8f1e 100644 --- a/tools/perf/util/sort.c +++ b/tools/perf/util/sort.c @@ -151,11 +151,17 @@ sort__sym_cmp(struct hist_entry *left, struct hist_entry *right) { u64 ip_l, ip_r; + if (!left->ms.sym && !right->ms.sym) + return right->level - left->level; + + if (!left->ms.sym || !right->ms.sym) + return cmp_null(left->ms.sym, right->ms.sym); + if (left->ms.sym == right->ms.sym) return 0; - ip_l = left->ms.sym ? left->ms.sym->start : left->ip; - ip_r = right->ms.sym ? right->ms.sym->start : right->ip; + ip_l = left->ms.sym->start; + ip_r = right->ms.sym->start; return (int64_t)(ip_r - ip_l); } -- cgit v0.10.2 From 936be50306a92356367f330ef9d44f1f62478d22 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Tue, 6 Sep 2011 09:12:26 -0600 Subject: perf tool: Fix endianness handling of u32 data in samples Currently, analyzing PPC data files on x86 the cpu field is always 0 and the tid and pid are backwards. For example, analyzing a PPC file on PPC the pid/tid fields show: rsyslogd 1210/1212 and analyzing the same PPC file using an x86 perf binary shows: rsyslogd 1212/1210 The problem is that the swap_op method for samples is perf_event__all64_swap which assumes all elements in the sample_data struct are u64s. cpu, tid and pid are u32s and need to be handled individually. Given that the swap is done before the sample is parsed, the simplest solution is to undo the 64-bit swap of those elements when the sample is parsed and do the proper swap. The RAW data field is generic and perf cannot have programmatic knowledge of how to treat that data. Instead a warning is given to the user. Thanks to Anton Blanchard for providing a data file for a mult-CPU PPC system so I could verify the fix for the CPU fields. v3 -> v4: - fixed use of WARN_ONCE v2 -> v3: - used WARN_ONCE for message regarding raw data - removed struct wrapper around union - fixed whitespace issues v1 -> v2: - added a union for undoing the byte-swap on u64 and redoing swap on u32's to address compiler errors (see git commit 65014ab3) Cc: Anton Blanchard Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/1315321946-16993-1-git-send-email-dsahern@gmail.com Signed-off-by: David Ahern Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-test.c b/tools/perf/builtin-test.c index 55f4c76..efe696f 100644 --- a/tools/perf/builtin-test.c +++ b/tools/perf/builtin-test.c @@ -561,7 +561,7 @@ static int test__basic_mmap(void) } err = perf_event__parse_sample(event, attr.sample_type, sample_size, - false, &sample); + false, &sample, false); if (err) { pr_err("Can't parse sample, err = %d\n", err); goto out_munmap; diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index 1d7f664..357a85b 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -186,6 +186,6 @@ const char *perf_event__name(unsigned int id); int perf_event__parse_sample(const union perf_event *event, u64 type, int sample_size, bool sample_id_all, - struct perf_sample *sample); + struct perf_sample *sample, bool swapped); #endif /* __PERF_RECORD_H */ diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index a03a36b..c5748c5 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -7,6 +7,8 @@ * Released under the GPL v2. (and only v2, not any later version) */ +#include +#include "asm/bug.h" #include "evsel.h" #include "evlist.h" #include "util.h" @@ -342,10 +344,20 @@ static bool sample_overlap(const union perf_event *event, int perf_event__parse_sample(const union perf_event *event, u64 type, int sample_size, bool sample_id_all, - struct perf_sample *data) + struct perf_sample *data, bool swapped) { const u64 *array; + /* + * used for cross-endian analysis. See git commit 65014ab3 + * for why this goofiness is needed. + */ + union { + u64 val64; + u32 val32[2]; + } u; + + data->cpu = data->pid = data->tid = -1; data->stream_id = data->id = data->time = -1ULL; @@ -366,9 +378,16 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, } if (type & PERF_SAMPLE_TID) { - u32 *p = (u32 *)array; - data->pid = p[0]; - data->tid = p[1]; + u.val64 = *array; + if (swapped) { + /* undo swap of u64, then swap on individual u32s */ + u.val64 = bswap_64(u.val64); + u.val32[0] = bswap_32(u.val32[0]); + u.val32[1] = bswap_32(u.val32[1]); + } + + data->pid = u.val32[0]; + data->tid = u.val32[1]; array++; } @@ -395,8 +414,15 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, } if (type & PERF_SAMPLE_CPU) { - u32 *p = (u32 *)array; - data->cpu = *p; + + u.val64 = *array; + if (swapped) { + /* undo swap of u64, then swap on individual u32s */ + u.val64 = bswap_64(u.val64); + u.val32[0] = bswap_32(u.val32[0]); + } + + data->cpu = u.val32[0]; array++; } @@ -423,18 +449,24 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, } if (type & PERF_SAMPLE_RAW) { - u32 *p = (u32 *)array; + u.val64 = *array; + if (WARN_ONCE(swapped, + "Endianness of raw data not corrected!\n")) { + /* undo swap of u64, then swap on individual u32s */ + u.val64 = bswap_64(u.val64); + u.val32[0] = bswap_32(u.val32[0]); + u.val32[1] = bswap_32(u.val32[1]); + } if (sample_overlap(event, array, sizeof(u32))) return -EFAULT; - data->raw_size = *p; - p++; + data->raw_size = u.val32[0]; - if (sample_overlap(event, p, data->raw_size)) + if (sample_overlap(event, &u.val32[1], data->raw_size)) return -EFAULT; - data->raw_data = p; + data->raw_data = &u.val32[1]; } return 0; diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h index 170601e..974d0cb 100644 --- a/tools/perf/util/session.h +++ b/tools/perf/util/session.h @@ -162,7 +162,8 @@ static inline int perf_session__parse_sample(struct perf_session *session, { return perf_event__parse_sample(event, session->sample_type, session->sample_size, - session->sample_id_all, sample); + session->sample_id_all, sample, + session->header.needs_swap); } struct perf_evsel *perf_session__find_first_evtype(struct perf_session *session, -- cgit v0.10.2 From be96ea8ffa788dccb1ba895cced29db6687c4911 Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Fri, 22 Oct 2010 17:25:01 +0200 Subject: perf symbols: Fix issue with binaries using 16-bytes buildids (v2) Buildid can vary in size. According to the man page of ld, buildid can be 160 bits (sha1) or 128 bits (md5, uuid). Perf assumes buildid size of 20 bytes (160 bits) regardless. When dealing with md5 buildids, it would thus read more than needed and that would cause mismatches and samples without symbols. This patch fixes this by taking into account the actual buildid size as encoded int he section header. The leftover bytes are also cleared. This second version fixes a minor issue with the memset() base position. Cc: David S. Miller Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Robert Richter Cc: Stephane Eranian Link: http://lkml.kernel.org/r/4cc1af3c.8ee7d80a.5a28.ffff868e@mx.google.com Signed-off-by: Stephane Eranian Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 870b3b8..40eeaf0 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1170,8 +1170,7 @@ static int dso__load_sym(struct dso *dso, struct map *map, const char *name, if (dso->has_build_id) { u8 build_id[BUILD_ID_SIZE]; - if (elf_read_build_id(elf, build_id, - BUILD_ID_SIZE) != BUILD_ID_SIZE) + if (elf_read_build_id(elf, build_id, BUILD_ID_SIZE) < 0) goto out_elf_end; if (!dso__build_id_equal(dso, build_id)) @@ -1443,8 +1442,8 @@ static int elf_read_build_id(Elf *elf, void *bf, size_t size) ptr = data->d_buf; while (ptr < (data->d_buf + data->d_size)) { GElf_Nhdr *nhdr = ptr; - int namesz = NOTE_ALIGN(nhdr->n_namesz), - descsz = NOTE_ALIGN(nhdr->n_descsz); + size_t namesz = NOTE_ALIGN(nhdr->n_namesz), + descsz = NOTE_ALIGN(nhdr->n_descsz); const char *name; ptr += sizeof(*nhdr); @@ -1453,8 +1452,10 @@ static int elf_read_build_id(Elf *elf, void *bf, size_t size) if (nhdr->n_type == NT_GNU_BUILD_ID && nhdr->n_namesz == sizeof("GNU")) { if (memcmp(name, "GNU", sizeof("GNU")) == 0) { - memcpy(bf, ptr, BUILD_ID_SIZE); - err = BUILD_ID_SIZE; + size_t sz = min(size, descsz); + memcpy(bf, ptr, sz); + memset(bf + sz, 0, size - sz); + err = descsz; break; } } @@ -1506,7 +1507,7 @@ int sysfs__read_build_id(const char *filename, void *build_id, size_t size) while (1) { char bf[BUFSIZ]; GElf_Nhdr nhdr; - int namesz, descsz; + size_t namesz, descsz; if (read(fd, &nhdr, sizeof(nhdr)) != sizeof(nhdr)) break; @@ -1515,15 +1516,16 @@ int sysfs__read_build_id(const char *filename, void *build_id, size_t size) descsz = NOTE_ALIGN(nhdr.n_descsz); if (nhdr.n_type == NT_GNU_BUILD_ID && nhdr.n_namesz == sizeof("GNU")) { - if (read(fd, bf, namesz) != namesz) + if (read(fd, bf, namesz) != (ssize_t)namesz) break; if (memcmp(bf, "GNU", sizeof("GNU")) == 0) { - if (read(fd, build_id, - BUILD_ID_SIZE) == BUILD_ID_SIZE) { + size_t sz = min(descsz, size); + if (read(fd, build_id, sz) == (ssize_t)sz) { + memset(build_id + sz, 0, size - sz); err = 0; break; } - } else if (read(fd, bf, descsz) != descsz) + } else if (read(fd, bf, descsz) != (ssize_t)descsz) break; } else { int n = namesz + descsz; -- cgit v0.10.2 From af52aafad26fe83edc3ff95d6f630c2fc98a0c4c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 14 Sep 2011 15:54:30 -0300 Subject: perf top: Fix userspace sample addr map offset The 'perf top' tool came from the kernel where we had each DSO (vmlinux, modules) loaded just once at a time. But userspace may have DSOs loaded in multiple addresses (shared libraries), requiring that we use the just resolved map instead of the first one found. Cc: David Ahern Cc: Frederic Weisbecker Cc: Mike Galbraith Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Stephane Eranian Link: http://lkml.kernel.org/n/tip-ag53wz0yllpgers0n2w7hchp@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index a43433f..d28013b 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -191,7 +191,8 @@ static void __zero_source_counters(struct sym_entry *syme) symbol__annotate_zero_histograms(sym); } -static void record_precise_ip(struct sym_entry *syme, int counter, u64 ip) +static void record_precise_ip(struct sym_entry *syme, struct map *map, + int counter, u64 ip) { struct annotation *notes; struct symbol *sym; @@ -205,8 +206,8 @@ static void record_precise_ip(struct sym_entry *syme, int counter, u64 ip) if (pthread_mutex_trylock(¬es->lock)) return; - ip = syme->map->map_ip(syme->map, ip); - symbol__inc_addr_samples(sym, syme->map, counter, ip); + ip = map->map_ip(map, ip); + symbol__inc_addr_samples(sym, map, counter, ip); pthread_mutex_unlock(¬es->lock); } @@ -810,7 +811,7 @@ static void perf_event__process_sample(const union perf_event *event, evsel = perf_evlist__id2evsel(top.evlist, sample->id); assert(evsel != NULL); syme->count[evsel->idx]++; - record_precise_ip(syme, evsel->idx, ip); + record_precise_ip(syme, al.map, evsel->idx, ip); pthread_mutex_lock(&top.active_symbols_lock); if (list_empty(&syme->node) || !syme->node.next) { static bool first = true; -- cgit v0.10.2 From 9e59e0995a57048c53773f1de99c6637ad12dd25 Mon Sep 17 00:00:00 2001 From: Darren Hart Date: Thu, 8 Sep 2011 13:42:39 -0700 Subject: perf tools: Add support for disabling -Werror via WERROR=0 GCC often introduces new warnings with lots of false positives - breaking -Werror builds. WERROR=0 allows one to build perf without much fuss - while still encouraging people to send patches to avoid the fuss of having to type WERROR=0. Bisecting back to commits that produce a (mostly harmless) warning on some compilers is more difficult. With WERROR=0 one could bisect without worrying about harmless warnings. Cc: Ingo Molnar Link: http://lkml.kernel.org/r/eac06c7cc4920e5d4830417d466161fb26c7359c.1315514559.git.dvhart@linux.intel.com Signed-off-by: Darren Hart Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 3b8f7b8..e9d5c27 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -30,6 +30,8 @@ endif # Define EXTRA_CFLAGS=-m64 or EXTRA_CFLAGS=-m32 as appropriate for cross-builds. # # Define NO_DWARF if you do not want debug-info analysis feature at all. +# +# Define WERROR=0 to disable treating any warnings as errors. $(OUTPUT)PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE @$(SHELL_PATH) util/PERF-VERSION-GEN $(OUTPUT) @@ -63,6 +65,11 @@ ifeq ($(ARCH),x86_64) endif endif +# Treat warnings as errors unless directed not to +ifneq ($(WERROR),0) + CFLAGS_WERROR := -Werror +endif + # # Include saner warnings here, which can catch bugs: # @@ -95,7 +102,7 @@ ifndef PERF_DEBUG CFLAGS_OPTIMIZE = -O6 endif -CFLAGS = -fno-omit-frame-pointer -ggdb3 -Wall -Wextra -std=gnu99 -Werror $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS) $(EXTRA_CFLAGS) +CFLAGS = -fno-omit-frame-pointer -ggdb3 -Wall -Wextra -std=gnu99 $(CFLAGS_WERROR) $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS) $(EXTRA_CFLAGS) EXTLIBS = -lpthread -lrt -lelf -lm ALL_CFLAGS = $(CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 ALL_LDFLAGS = $(LDFLAGS) -- cgit v0.10.2 From b3a242a6e4b8c09dbb466ab7a9d2c724e75faa67 Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Fri, 23 Sep 2011 06:35:00 -0400 Subject: hwmon: (coretemp) don't use kernel assigned CPU number as platform device ID ... as that has the potential to conflict with (particularly soft) CPU hot removal and re-adding. Signed-off-by: Jan Beulich [guenter.roeck@ericsson.com: use platform device ID as physical CPU id] Signed-off-by: Guenter Roeck diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 4736415..6420c3b 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -561,7 +561,7 @@ static int __devinit coretemp_probe(struct platform_device *pdev) if (err) goto exit_free; - pdata->phys_proc_id = TO_PHYS_ID(pdev->id); + pdata->phys_proc_id = pdev->id; platform_set_drvdata(pdev, pdata); pdata->hwmon_dev = hwmon_device_register(&pdev->dev); @@ -613,7 +613,7 @@ static int __cpuinit coretemp_device_add(unsigned int cpu) mutex_lock(&pdev_list_mutex); - pdev = platform_device_alloc(DRVNAME, cpu); + pdev = platform_device_alloc(DRVNAME, TO_PHYS_ID(cpu)); if (!pdev) { err = -ENOMEM; pr_err("Device allocation failed\n"); -- cgit v0.10.2 From e3204ed3a4a78ca4d10eee8b661b94429bd38da8 Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Fri, 23 Sep 2011 06:36:53 -0400 Subject: hwmon: (coretemp) constify static data These arrays won't ever be written to, so protect them from unintentional modification. Signed-off-by: Jan Beulich Signed-off-by: Guenter Roeck diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 6420c3b..ac1c81f 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -347,11 +347,11 @@ static int create_core_attrs(struct temp_data *tdata, struct device *dev, int attr_no) { int err, i; - static ssize_t (*rd_ptr[TOTAL_ATTRS]) (struct device *dev, + static ssize_t (*const rd_ptr[TOTAL_ATTRS]) (struct device *dev, struct device_attribute *devattr, char *buf) = { show_label, show_crit_alarm, show_temp, show_tjmax, show_ttarget }; - static const char *names[TOTAL_ATTRS] = { + static const char *const names[TOTAL_ATTRS] = { "temp%d_label", "temp%d_crit_alarm", "temp%d_input", "temp%d_crit", "temp%d_max" }; -- cgit v0.10.2 From 2f1c3db0a6adcfd12d556afa3605d4923658b307 Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Fri, 23 Sep 2011 06:40:08 -0400 Subject: hwmon: (coretemp) remove struct platform_data * parameter from create_core_data() The only caller of the function obtained the pointer solely for the purpose of passing it to this function, while it can be easily determined from the struct platform_device * parameter also passed. Signed-off-by: Jan Beulich Signed-off-by: Guenter Roeck diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index ac1c81f..44b2391 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -442,11 +442,11 @@ static struct temp_data *init_temp_data(unsigned int cpu, int pkg_flag) return tdata; } -static int create_core_data(struct platform_data *pdata, - struct platform_device *pdev, +static int create_core_data(struct platform_device *pdev, unsigned int cpu, int pkg_flag) { struct temp_data *tdata; + struct platform_data *pdata = platform_get_drvdata(pdev); struct cpuinfo_x86 *c = &cpu_data(cpu); u32 eax, edx; int err, attr_no; @@ -514,16 +514,13 @@ exit_free: static void coretemp_add_core(unsigned int cpu, int pkg_flag) { - struct platform_data *pdata; struct platform_device *pdev = coretemp_get_pdev(cpu); int err; if (!pdev) return; - pdata = platform_get_drvdata(pdev); - - err = create_core_data(pdata, pdev, cpu, pkg_flag); + err = create_core_data(pdev, cpu, pkg_flag); if (err) dev_err(&pdev->dev, "Adding Core %u failed\n", cpu); } -- cgit v0.10.2 From 2b022a82a01737012155b5ba462db74232ff1f2e Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 23 Sep 2011 15:38:53 -0300 Subject: perf python: Add missing perf_event__parse_sample 'swapped' parm Problem introduced in 936be50, that missed one perf_event__parse_sample user, the python binding. Reported-by: Linus Torvalds Cc: David Ahern Cc: Frederic Weisbecker Cc: Mike Galbraith Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Stephane Eranian Link: http://lkml.kernel.org/n/tip-ja4phms9618ggi657plyuch2@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/python.c b/tools/perf/util/python.c index cbc8f21..7624324 100644 --- a/tools/perf/util/python.c +++ b/tools/perf/util/python.c @@ -803,7 +803,7 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist, first = list_entry(evlist->entries.next, struct perf_evsel, node); err = perf_event__parse_sample(event, first->attr.sample_type, perf_evsel__sample_size(first), - sample_id_all, &pevent->sample); + sample_id_all, &pevent->sample, false); if (err) return PyErr_Format(PyExc_OSError, "perf: can't parse sample, err=%d", err); -- cgit v0.10.2 From 2b7fe39babcbd67458a0c4fbed5704089e0c8c1f Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Fri, 23 Sep 2011 15:43:54 -0700 Subject: scsi: SCSI_ISCI needs to select SCSI_SAS_HOST_SMP, fixes build error SCSI_ISCI needs to select SCSI_SAS_HOST_SMP to ensure that all needed symbols are available to it. Fixes this build error: ERROR: "try_test_sas_gpio_gp_bit" [drivers/scsi/isci/isci.ko] undefined! Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 8d9dae8..3878b73 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -837,6 +837,7 @@ config SCSI_ISCI # (temporary): known alpha quality driver depends on EXPERIMENTAL select SCSI_SAS_LIBSAS + select SCSI_SAS_HOST_SMP ---help--- This driver supports the 6Gb/s SAS capabilities of the storage control unit found in the Intel(R) C600 series chipset. -- cgit v0.10.2 From 8ec9c7fb15fd8edf6b34555f5b498033121b2173 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Fri, 23 Sep 2011 15:40:50 -0700 Subject: scsi: fix qla2xxx printk format warning sector_t can be different types, so cast it to its largest possible type. drivers/scsi/qla2xxx/qla_isr.c:1509:5: warning: format '%lx' expects type 'long unsigned int', but argument 5 has type 'sector_t' Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c index 646fc52..8a7591f 100644 --- a/drivers/scsi/qla2xxx/qla_isr.c +++ b/drivers/scsi/qla2xxx/qla_isr.c @@ -1507,8 +1507,8 @@ qla2x00_handle_dif_error(srb_t *sp, struct sts_entry_24xx *sts24) if (k != blocks_done) { qla_printk(KERN_WARNING, sp->fcport->vha->hw, - "unexpected tag values tag:lba=%x:%lx)\n", - e_ref_tag, lba_s); + "unexpected tag values tag:lba=%x:%llx)\n", + e_ref_tag, (unsigned long long)lba_s); return 1; } -- cgit v0.10.2 From 387719c2ec995019e0af0dc537ca17c9ebd5b64f Mon Sep 17 00:00:00 2001 From: Jeff Harris Date: Fri, 23 Sep 2011 11:49:36 -0400 Subject: spi: Fix WARN when removing spi-fsl-spi module If CPM mode is not used, the fsl_dummy_rx variable is never allocated. When the cleanup attempts to free it, the reference count is zero and a WARN is generated. The same CPM mode check used in the initialize is applied to the free as well. Tested on 2.6.33 with the previous spi_mpc8xxx driver. The renamed spi-fsl-spi driver looks to have the same problem. Signed-off-by: Jeff Harris Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-fsl-spi.c b/drivers/spi/spi-fsl-spi.c index d240755..24cacff 100644 --- a/drivers/spi/spi-fsl-spi.c +++ b/drivers/spi/spi-fsl-spi.c @@ -825,6 +825,9 @@ static void fsl_spi_cpm_free(struct mpc8xxx_spi *mspi) { struct device *dev = mspi->dev; + if (!(mspi->flags & SPI_CPM_MODE)) + return; + dma_unmap_single(dev, mspi->dma_dummy_rx, SPI_MRBLR, DMA_FROM_DEVICE); dma_unmap_single(dev, mspi->dma_dummy_tx, PAGE_SIZE, DMA_TO_DEVICE); cpm_muram_free(cpm_muram_offset(mspi->tx_bd)); -- cgit v0.10.2 From 6656b15d675c9c6a049db48d50994b3cd4e76bd6 Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Sat, 24 Sep 2011 08:30:44 +0200 Subject: ALSA: HDA: No power nids on 92HD93 This patch is necessary to make internal speakers work on this chip. Cc: stable@kernel.org BugLink: http://bugs.launchpad.net/bugs/854468 Tested-by: Alex Wolfson Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 1b7c114..987e3cf 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -5630,6 +5630,7 @@ again: switch (codec->vendor_id) { case 0x111d76d1: case 0x111d76d9: + case 0x111d76df: case 0x111d76e5: case 0x111d7666: case 0x111d7667: -- cgit v0.10.2 From 41bc3186b3c92a4ca05e2aa14bb6272fb491e679 Mon Sep 17 00:00:00 2001 From: Zhao Jin Date: Mon, 19 Sep 2011 12:19:51 +0800 Subject: KVM: MMU: fix incorrect return of spte __update_clear_spte_slow should return original spte while the current code returns low half of original spte combined with high half of new spte. Signed-off-by: Zhao Jin Reviewed-by: Xiao Guangrong Signed-off-by: Marcelo Tosatti diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 1c5b693..8e8da79 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -400,7 +400,8 @@ static u64 __update_clear_spte_slow(u64 *sptep, u64 spte) /* xchg acts as a barrier before the setting of the high bits */ orig.spte_low = xchg(&ssptep->spte_low, sspte.spte_low); - orig.spte_high = ssptep->spte_high = sspte.spte_high; + orig.spte_high = ssptep->spte_high; + ssptep->spte_high = sspte.spte_high; count_spte_clear(sptep, spte); return orig.spte; -- cgit v0.10.2 From 9be3be1f153e90ea4e1e5b6ed1d72a73d44318d1 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Tue, 13 Sep 2011 10:45:38 +0300 Subject: KVM: x86 emulator: fix Src2CL decode Src2CL decode (used for double width shifts) erronously decodes only bit 3 of %rcx, instead of bits 7:0. Fix by decoding %cl in its entirety. Signed-off-by: Avi Kivity Signed-off-by: Marcelo Tosatti diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index 6f08bc9..8b4cc5f 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -3603,7 +3603,7 @@ done_prefixes: break; case Src2CL: ctxt->src2.bytes = 1; - ctxt->src2.val = ctxt->regs[VCPU_REGS_RCX] & 0x8; + ctxt->src2.val = ctxt->regs[VCPU_REGS_RCX] & 0xff; break; case Src2ImmByte: rc = decode_imm(ctxt, &ctxt->src2, 1, true); -- cgit v0.10.2 From f9d81f61c84aca693bc353dfef4b8c36c2e5e1b5 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Sun, 25 Sep 2011 19:46:22 +0200 Subject: ptrace: PTRACE_LISTEN forgets to unlock ->siglock If PTRACE_LISTEN fails after lock_task_sighand() it doesn't drop ->siglock. Reported-by: Matt Fleming Signed-off-by: Oleg Nesterov Signed-off-by: Linus Torvalds diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 9de3ecf..a70d2a5 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -744,20 +744,17 @@ int ptrace_request(struct task_struct *child, long request, break; si = child->last_siginfo; - if (unlikely(!si || si->si_code >> 8 != PTRACE_EVENT_STOP)) - break; - - child->jobctl |= JOBCTL_LISTENING; - - /* - * If NOTIFY is set, it means event happened between start - * of this trap and now. Trigger re-trap immediately. - */ - if (child->jobctl & JOBCTL_TRAP_NOTIFY) - signal_wake_up(child, true); - + if (likely(si && (si->si_code >> 8) == PTRACE_EVENT_STOP)) { + child->jobctl |= JOBCTL_LISTENING; + /* + * If NOTIFY is set, it means event happened between + * start of this trap and now. Trigger re-trap. + */ + if (child->jobctl & JOBCTL_TRAP_NOTIFY) + signal_wake_up(child, true); + ret = 0; + } unlock_task_sighand(child, &flags); - ret = 0; break; case PTRACE_DETACH: /* detach a process that was attached. */ -- cgit v0.10.2 From 68e58a294fb26f692697179e3f3ecf88dd8cb97c Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Sun, 25 Sep 2011 23:26:15 +0100 Subject: dm: flakey fix corrupt_bio_byte error path If no arguments were provided to the corrupt_bio_byte feature an error should be returned immediately. Reported-by: Zdenek Kabelac Signed-off-by: Mike Snitzer Signed-off-by: Alasdair G Kergon diff --git a/drivers/md/dm-flakey.c b/drivers/md/dm-flakey.c index 89f73ca..f84c080 100644 --- a/drivers/md/dm-flakey.c +++ b/drivers/md/dm-flakey.c @@ -81,8 +81,10 @@ static int parse_features(struct dm_arg_set *as, struct flakey_c *fc, * corrupt_bio_byte */ if (!strcasecmp(arg_name, "corrupt_bio_byte")) { - if (!argc) + if (!argc) { ti->error = "Feature corrupt_bio_byte requires parameters"; + return -EINVAL; + } r = dm_read_arg(_args + 1, as, &fc->corrupt_bio_byte, &ti->error); if (r) -- cgit v0.10.2 From 876fbba1db4a377f050a2bb49b474c7527b2995d Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Sun, 25 Sep 2011 23:26:17 +0100 Subject: dm table: avoid crash if integrity profile changes Commit a63a5cf (dm: improve block integrity support) introduced a two-phase initialization of a DM device's integrity profile. This patch avoids dereferencing a NULL 'template_disk' pointer in blk_integrity_register() if there is an integrity profile mismatch in dm_table_set_integrity(). This can occur if the integrity profiles for stacked devices in a DM table are changed between the call to dm_table_prealloc_integrity() and dm_table_set_integrity(). Reported-by: Zdenek Kabelac Signed-off-by: Mike Snitzer Signed-off-by: Alasdair G Kergon Cc: stable@kernel.org # 2.6.39 diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index 986b875..3226698 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -1238,14 +1238,15 @@ static void dm_table_set_integrity(struct dm_table *t) return; template_disk = dm_table_get_integrity_disk(t, true); - if (!template_disk && - blk_integrity_is_initialized(dm_disk(t->md))) { + if (template_disk) + blk_integrity_register(dm_disk(t->md), + blk_get_integrity(template_disk)); + else if (blk_integrity_is_initialized(dm_disk(t->md))) DMWARN("%s: device no longer has a valid integrity profile", dm_device_name(t->md)); - return; - } - blk_integrity_register(dm_disk(t->md), - blk_get_integrity(template_disk)); + else + DMWARN("%s: unable to establish an integrity profile", + dm_device_name(t->md)); } static int device_flush_capable(struct dm_target *ti, struct dm_dev *dev, -- cgit v0.10.2 From 8232480944d173378082ebb2cac8a3207c08cf31 Mon Sep 17 00:00:00 2001 From: Jonthan Brassow Date: Sun, 25 Sep 2011 23:26:19 +0100 Subject: dm: raid fix write_mostly arg validation Fix off-by-one error in validation of write_mostly. The user-supplied value given for the 'write_mostly' argument must be an index starting at 0. The validation of the supplied argument failed to check for 'N' ('>' vs '>='), which would have caused an access beyond the end of the array. Reported-by: Doug Ledford Signed-off-by: Jonathan Brassow Signed-off-by: Alasdair G Kergon diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c index a002dd8..86df8b2 100644 --- a/drivers/md/dm-raid.c +++ b/drivers/md/dm-raid.c @@ -449,7 +449,7 @@ static int parse_raid_params(struct raid_set *rs, char **argv, rs->ti->error = "write_mostly option is only valid for RAID1"; return -EINVAL; } - if (value > rs->md.raid_disks) { + if (value >= rs->md.raid_disks) { rs->ti->error = "Invalid write_mostly drive index given"; return -EINVAL; } -- cgit v0.10.2 From 983c7db347db8ce2d8453fd1d89b7a4bb6920d56 Mon Sep 17 00:00:00 2001 From: Milan Broz Date: Sun, 25 Sep 2011 23:26:21 +0100 Subject: dm crypt: always disable discard_zeroes_data If optional discard support in dm-crypt is enabled, discards requests bypass the crypt queue and blocks of the underlying device are discarded. For the read path, discarded blocks are handled the same as normal ciphertext blocks, thus decrypted. So if the underlying device announces discarded regions return zeroes, dm-crypt must disable this flag because after decryption there is just random noise instead of zeroes. Signed-off-by: Milan Broz Signed-off-by: Alasdair G Kergon diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index 49da55c..8c2a000 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -1698,6 +1698,8 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) } ti->num_flush_requests = 1; + ti->discard_zeroes_data_unsupported = 1; + return 0; bad: diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index 3226698..bc04518 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -1283,6 +1283,22 @@ static bool dm_table_supports_flush(struct dm_table *t, unsigned flush) return 0; } +static bool dm_table_discard_zeroes_data(struct dm_table *t) +{ + struct dm_target *ti; + unsigned i = 0; + + /* Ensure that all targets supports discard_zeroes_data. */ + while (i < dm_table_get_num_targets(t)) { + ti = dm_table_get_target(t, i++); + + if (ti->discard_zeroes_data_unsupported) + return 0; + } + + return 1; +} + void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q, struct queue_limits *limits) { @@ -1305,6 +1321,9 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q, } blk_queue_flush(q, flush); + if (!dm_table_discard_zeroes_data(t)) + q->limits.discard_zeroes_data = 0; + dm_table_set_integrity(t); /* diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index 3fa1f3d..99e3e50 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -197,6 +197,11 @@ struct dm_target { * whether or not its underlying devices have support. */ unsigned discards_supported:1; + + /* + * Set if this target does not return zeroes on discarded blocks. + */ + unsigned discard_zeroes_data_unsupported:1; }; /* Each target can link one of these into the table */ -- cgit v0.10.2 From 4344646ff127eaea4d58266ccd8fdd2a38fb9672 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Mon, 26 Sep 2011 10:30:29 +0900 Subject: ARM: S3C2443: Fix bit-reset in setrate of clk_armdiv The changed statement should set the old armdiv bits to 0 and not everything else, before setting the new value. Signed-off-by: Heiko Stuebner Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-s3c2443/clock.c b/arch/arm/mach-s3c2443/clock.c index a1a7176..38058af 100644 --- a/arch/arm/mach-s3c2443/clock.c +++ b/arch/arm/mach-s3c2443/clock.c @@ -128,7 +128,7 @@ static int s3c2443_armclk_setrate(struct clk *clk, unsigned long rate) unsigned long clkcon0; clkcon0 = __raw_readl(S3C2443_CLKDIV0); - clkcon0 &= S3C2443_CLKDIV0_ARMDIV_MASK; + clkcon0 &= ~S3C2443_CLKDIV0_ARMDIV_MASK; clkcon0 |= val << S3C2443_CLKDIV0_ARMDIV_SHIFT; __raw_writel(clkcon0, S3C2443_CLKDIV0); } -- cgit v0.10.2 From b76f7cdca5af18fa7a59a5c8a9eb5e02418e827a Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Mon, 26 Sep 2011 13:16:45 +0900 Subject: ARM: S5P: fix incorrect loop iterator usage on gpio-interrupt Loop iterator value after terminating list_for_each_entry() is not NULL. This patch fixes incorrect iterator usage in GPIO interrupt code for SAMSUNG S5P platforms. Signed-off-by: Marek Szyprowski Signed-off-by: Kyungmin Park Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-s5p/irq-gpioint.c b/arch/arm/plat-s5p/irq-gpioint.c index f71078e..f88216d 100644 --- a/arch/arm/plat-s5p/irq-gpioint.c +++ b/arch/arm/plat-s5p/irq-gpioint.c @@ -114,17 +114,18 @@ static __init int s5p_gpioint_add(struct s3c_gpio_chip *chip) { static int used_gpioint_groups = 0; int group = chip->group; - struct s5p_gpioint_bank *bank = NULL; + struct s5p_gpioint_bank *b, *bank = NULL; struct irq_chip_generic *gc; struct irq_chip_type *ct; if (used_gpioint_groups >= S5P_GPIOINT_GROUP_COUNT) return -ENOMEM; - list_for_each_entry(bank, &banks, list) { - if (group >= bank->start && - group < bank->start + bank->nr_groups) + list_for_each_entry(b, &banks, list) { + if (group >= b->start && group < b->start + b->nr_groups) { + bank = b; break; + } } if (!bank) return -EINVAL; -- cgit v0.10.2 From d8e89b47e00ee80e920761145144640aac4cf71a Mon Sep 17 00:00:00 2001 From: Russell King Date: Thu, 22 Sep 2011 10:32:25 +0100 Subject: ARM: dma-mapping: free allocated page if unable to map If the attempt to map a page for DMA fails (eg, because we're out of mapping space) then we must not hold on to the page we allocated for DMA - doing so will result in a memory leak. Cc: Reported-by: Bryan Phillippe Tested-by: Bryan Phillippe Signed-off-by: Russell King diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 0a0a1e7..c3ff82f 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -324,6 +324,8 @@ __dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp, if (addr) *handle = pfn_to_dma(dev, page_to_pfn(page)); + else + __dma_free_buffer(page, size); return addr; } -- cgit v0.10.2 From 5fe6e0151dbd969f5fbcd94d05c968b76d76952b Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 26 Sep 2011 10:41:21 +0200 Subject: ALSA: hda/realtek - Avoid bogus HP-pin assignment When the headphone pin is assigned as primary output to line_out_pins[], the automatic HP-pin assignment by ASSID must be suppressed. Otherwise a wrong pin might be assigned to the headphone and breaks the auto-mute. Reference: https://bugzilla.novell.com/show_bug.cgi?id=716104 Signed-off-by: Takashi Iwai Cc: diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 0503c99..91a5381 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -1321,7 +1321,9 @@ do_sku: * 15 : 1 --> enable the function "Mute internal speaker * when the external headphone out jack is plugged" */ - if (!spec->autocfg.hp_pins[0]) { + if (!spec->autocfg.hp_pins[0] && + !(spec->autocfg.line_out_pins[0] && + spec->autocfg.line_out_type == AUTO_PIN_HP_OUT)) { hda_nid_t nid; tmp = (ass >> 11) & 0x3; /* HP to chassis */ if (tmp == 0) -- cgit v0.10.2 From 6ebbe7a07b3bc40b168d2afc569a6543c020d2e3 Mon Sep 17 00:00:00 2001 From: Simon Kirby Date: Thu, 22 Sep 2011 17:03:46 -0700 Subject: sched: Fix up wchan borkage Commit c259e01a1ec ("sched: Separate the scheduler entry for preemption") contained a boo-boo wrecking wchan output. It forgot to put the new schedule() function in the __sched section and thereby doesn't get properly ignored for things like wchan. Tested-by: Simon Kirby Cc: stable@kernel.org # 2.6.39+ Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/20110923000346.GA25425@hostway.ca Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index ec5f472..d249ea8 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4372,7 +4372,7 @@ static inline void sched_submit_work(struct task_struct *tsk) blk_schedule_flush_plug(tsk); } -asmlinkage void schedule(void) +asmlinkage void __sched schedule(void) { struct task_struct *tsk = current; -- cgit v0.10.2 From df77abcafc8dc881b6c9347548651777088e4b27 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 23 Sep 2011 14:34:12 +0100 Subject: ARM: 7099/1: futex: preserve oldval in SMP __futex_atomic_op The SMP implementation of __futex_atomic_op clobbers oldval with the status flag from the exclusive store. This causes it to always read as zero when performing the FUTEX_OP_CMP_* operation. This patch updates the ARM __futex_atomic_op implementations to take a tmp argument, allowing us to store the strex status flag without overwriting the register containing oldval. Cc: stable@kernel.org Reported-by: Minho Ban Reviewed-by: Nicolas Pitre Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/include/asm/futex.h b/arch/arm/include/asm/futex.h index 8c73900..253cc86 100644 --- a/arch/arm/include/asm/futex.h +++ b/arch/arm/include/asm/futex.h @@ -25,17 +25,17 @@ #ifdef CONFIG_SMP -#define __futex_atomic_op(insn, ret, oldval, uaddr, oparg) \ +#define __futex_atomic_op(insn, ret, oldval, tmp, uaddr, oparg) \ smp_mb(); \ __asm__ __volatile__( \ - "1: ldrex %1, [%2]\n" \ + "1: ldrex %1, [%3]\n" \ " " insn "\n" \ - "2: strex %1, %0, [%2]\n" \ - " teq %1, #0\n" \ + "2: strex %2, %0, [%3]\n" \ + " teq %2, #0\n" \ " bne 1b\n" \ " mov %0, #0\n" \ - __futex_atomic_ex_table("%4") \ - : "=&r" (ret), "=&r" (oldval) \ + __futex_atomic_ex_table("%5") \ + : "=&r" (ret), "=&r" (oldval), "=&r" (tmp) \ : "r" (uaddr), "r" (oparg), "Ir" (-EFAULT) \ : "cc", "memory") @@ -73,14 +73,14 @@ futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, #include #include -#define __futex_atomic_op(insn, ret, oldval, uaddr, oparg) \ +#define __futex_atomic_op(insn, ret, oldval, tmp, uaddr, oparg) \ __asm__ __volatile__( \ - "1: " T(ldr) " %1, [%2]\n" \ + "1: " T(ldr) " %1, [%3]\n" \ " " insn "\n" \ - "2: " T(str) " %0, [%2]\n" \ + "2: " T(str) " %0, [%3]\n" \ " mov %0, #0\n" \ - __futex_atomic_ex_table("%4") \ - : "=&r" (ret), "=&r" (oldval) \ + __futex_atomic_ex_table("%5") \ + : "=&r" (ret), "=&r" (oldval), "=&r" (tmp) \ : "r" (uaddr), "r" (oparg), "Ir" (-EFAULT) \ : "cc", "memory") @@ -117,7 +117,7 @@ futex_atomic_op_inuser (int encoded_op, u32 __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret; + int oldval = 0, ret, tmp; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; @@ -129,19 +129,19 @@ futex_atomic_op_inuser (int encoded_op, u32 __user *uaddr) switch (op) { case FUTEX_OP_SET: - __futex_atomic_op("mov %0, %3", ret, oldval, uaddr, oparg); + __futex_atomic_op("mov %0, %4", ret, oldval, tmp, uaddr, oparg); break; case FUTEX_OP_ADD: - __futex_atomic_op("add %0, %1, %3", ret, oldval, uaddr, oparg); + __futex_atomic_op("add %0, %1, %4", ret, oldval, tmp, uaddr, oparg); break; case FUTEX_OP_OR: - __futex_atomic_op("orr %0, %1, %3", ret, oldval, uaddr, oparg); + __futex_atomic_op("orr %0, %1, %4", ret, oldval, tmp, uaddr, oparg); break; case FUTEX_OP_ANDN: - __futex_atomic_op("and %0, %1, %3", ret, oldval, uaddr, ~oparg); + __futex_atomic_op("and %0, %1, %4", ret, oldval, tmp, uaddr, ~oparg); break; case FUTEX_OP_XOR: - __futex_atomic_op("eor %0, %1, %3", ret, oldval, uaddr, oparg); + __futex_atomic_op("eor %0, %1, %4", ret, oldval, tmp, uaddr, oparg); break; default: ret = -ENOSYS; -- cgit v0.10.2 From e0d32e335fc0e354db93a807736243f5035fbc67 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 26 Sep 2011 15:19:55 +0200 Subject: ALSA: hda/realtek - Don't detect LO jack when identical with HP The spec->autocfg.line_out_pins[] may contain the same pins as hp_pins[] depending on the configuration. When they are identical, detecting the line_jack_present flag screws up the auto-mute because alc_line_automute() is called unconditionally at initialization while it won't be triggered by unsol events, thus the old line_jack_present flag is kept for the whole run. For fixing this buggy behavior, the driver needs to check whether the line-outs are really individual, and skip if same as headphone jacks. Reference: https://bugzilla.novell.com/show_bug.cgi?id=716104 Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 91a5381..7a73621 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -578,6 +578,10 @@ static void alc_line_automute(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; + /* check LO jack only when it's different from HP */ + if (spec->autocfg.line_out_pins[0] == spec->autocfg.hp_pins[0]) + return; + spec->line_jack_present = detect_jacks(codec, ARRAY_SIZE(spec->autocfg.line_out_pins), spec->autocfg.line_out_pins); -- cgit v0.10.2 From 61a6a108d15213f5ee06332e1e7766d3860e4453 Mon Sep 17 00:00:00 2001 From: Thomas Pfaff Date: Mon, 26 Sep 2011 15:43:59 +0200 Subject: ALSA: usb-audio: Check for possible chip NULL pointer before clearing probing flag Before clearing the probing flag in the error exit path, check that the chip pointer is not NULL. Signed-off-by: Thomas Pfaff Cc: [2.6.39+] Signed-off-by: Takashi Iwai diff --git a/sound/usb/card.c b/sound/usb/card.c index ed120ca..d8f2bf4 100644 --- a/sound/usb/card.c +++ b/sound/usb/card.c @@ -530,9 +530,11 @@ snd_usb_audio_probe(struct usb_device *dev, return chip; __error: - if (chip && !chip->num_interfaces) - snd_card_free(chip->card); - chip->probing = 0; + if (chip) { + if (!chip->num_interfaces) + snd_card_free(chip->card); + chip->probing = 0; + } mutex_unlock(®ister_mutex); __err_val: return NULL; -- cgit v0.10.2 From e48f129c2f200dde8899f6ea5c6e7173674fc482 Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Tue, 6 Sep 2011 13:59:13 -0400 Subject: [SCSI] cxgb3i: convert cdev->l2opt to use rcu to prevent NULL dereference This oops was reported recently: d:mon> e cpu 0xd: Vector: 300 (Data Access) at [c0000000fd4c7120] pc: d00000000076f194: .t3_l2t_get+0x44/0x524 [cxgb3] lr: d000000000b02108: .init_act_open+0x150/0x3d4 [cxgb3i] sp: c0000000fd4c73a0 msr: 8000000000009032 dar: 0 dsisr: 40000000 current = 0xc0000000fd640d40 paca = 0xc00000000054ff80 pid = 5085, comm = iscsid d:mon> t [c0000000fd4c7450] d000000000b02108 .init_act_open+0x150/0x3d4 [cxgb3i] [c0000000fd4c7500] d000000000e45378 .cxgbi_ep_connect+0x784/0x8e8 [libcxgbi] [c0000000fd4c7650] d000000000db33f0 .iscsi_if_rx+0x71c/0xb18 [scsi_transport_iscsi2] [c0000000fd4c7740] c000000000370c9c .netlink_data_ready+0x40/0xa4 [c0000000fd4c77c0] c00000000036f010 .netlink_sendskb+0x4c/0x9c [c0000000fd4c7850] c000000000370c18 .netlink_sendmsg+0x358/0x39c [c0000000fd4c7950] c00000000033be24 .sock_sendmsg+0x114/0x1b8 [c0000000fd4c7b50] c00000000033d208 .sys_sendmsg+0x218/0x2ac [c0000000fd4c7d70] c00000000033f55c .sys_socketcall+0x228/0x27c [c0000000fd4c7e30] c0000000000086a4 syscall_exit+0x0/0x40 --- Exception: c01 (System Call) at 00000080da560cfc The root cause was an EEH error, which sent us down the offload_close path in the cxgb3 driver, which in turn sets cdev->l2opt to NULL, without regard for upper layer driver (like the cxgbi drivers) which might have execution contexts in the middle of its use. The result is the oops above, when t3_l2t_get attempts to dereference L2DATA(cdev)->nentries in arp_hash right after the EEH error handler sets it to NULL. The fix is to prevent the setting of the NULL pointer until after there are no further users of it. The t3cdev->l2opt pointer is now converted to be an rcu pointer and the L2DATA macro is now called under the protection of the rcu_read_lock(). When the EEH error path: t3_adapter_error->offload_close->cxgb3_offload_deactivate Is exectured, setting of that l2opt pointer to NULL, is now gated on an rcu quiescence point, preventing, allowing L2DATA callers to safely check for a NULL pointer without concern that the underlying data will be freeded before the pointer is dereferenced. This has been tested by the reporter and shown to fix the reproted oops [nhorman: fix up unitinialised variable reported by Dan Carpenter] Signed-off-by: Neil Horman Reviewed-by: Karen Xie Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c b/drivers/infiniband/hw/cxgb3/iwch_cm.c index 17bf9d9..6cd642a 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cm.c +++ b/drivers/infiniband/hw/cxgb3/iwch_cm.c @@ -287,7 +287,7 @@ void __free_ep(struct kref *kref) if (test_bit(RELEASE_RESOURCES, &ep->com.flags)) { cxgb3_remove_tid(ep->com.tdev, (void *)ep, ep->hwtid); dst_release(ep->dst); - l2t_release(L2DATA(ep->com.tdev), ep->l2t); + l2t_release(ep->com.tdev, ep->l2t); } kfree(ep); } @@ -1178,7 +1178,7 @@ static int act_open_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) release_tid(ep->com.tdev, GET_TID(rpl), NULL); cxgb3_free_atid(ep->com.tdev, ep->atid); dst_release(ep->dst); - l2t_release(L2DATA(ep->com.tdev), ep->l2t); + l2t_release(ep->com.tdev, ep->l2t); put_ep(&ep->com); return CPL_RET_BUF_DONE; } @@ -1377,7 +1377,7 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) if (!child_ep) { printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n", __func__); - l2t_release(L2DATA(tdev), l2t); + l2t_release(tdev, l2t); dst_release(dst); goto reject; } @@ -1956,7 +1956,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (!err) goto out; - l2t_release(L2DATA(h->rdev.t3cdev_p), ep->l2t); + l2t_release(h->rdev.t3cdev_p, ep->l2t); fail4: dst_release(ep->dst); fail3: @@ -2127,7 +2127,7 @@ int iwch_ep_redirect(void *ctx, struct dst_entry *old, struct dst_entry *new, PDBG("%s ep %p redirect to dst %p l2t %p\n", __func__, ep, new, l2t); dst_hold(new); - l2t_release(L2DATA(ep->com.tdev), ep->l2t); + l2t_release(ep->com.tdev, ep->l2t); ep->l2t = l2t; dst_release(old); ep->dst = new; diff --git a/drivers/net/cxgb3/cxgb3_offload.c b/drivers/net/cxgb3/cxgb3_offload.c index 805076c..da5a5d9 100644 --- a/drivers/net/cxgb3/cxgb3_offload.c +++ b/drivers/net/cxgb3/cxgb3_offload.c @@ -1146,12 +1146,14 @@ static void cxgb_redirect(struct dst_entry *old, struct dst_entry *new) if (te && te->ctx && te->client && te->client->redirect) { update_tcb = te->client->redirect(te->ctx, old, new, e); if (update_tcb) { + rcu_read_lock(); l2t_hold(L2DATA(tdev), e); + rcu_read_unlock(); set_l2t_ix(tdev, tid, e); } } } - l2t_release(L2DATA(tdev), e); + l2t_release(tdev, e); } /* @@ -1264,7 +1266,7 @@ int cxgb3_offload_activate(struct adapter *adapter) goto out_free; err = -ENOMEM; - L2DATA(dev) = t3_init_l2t(l2t_capacity); + RCU_INIT_POINTER(dev->l2opt, t3_init_l2t(l2t_capacity)); if (!L2DATA(dev)) goto out_free; @@ -1298,16 +1300,24 @@ int cxgb3_offload_activate(struct adapter *adapter) out_free_l2t: t3_free_l2t(L2DATA(dev)); - L2DATA(dev) = NULL; + rcu_assign_pointer(dev->l2opt, NULL); out_free: kfree(t); return err; } +static void clean_l2_data(struct rcu_head *head) +{ + struct l2t_data *d = container_of(head, struct l2t_data, rcu_head); + t3_free_l2t(d); +} + + void cxgb3_offload_deactivate(struct adapter *adapter) { struct t3cdev *tdev = &adapter->tdev; struct t3c_data *t = T3C_DATA(tdev); + struct l2t_data *d; remove_adapter(adapter); if (list_empty(&adapter_list)) @@ -1315,8 +1325,11 @@ void cxgb3_offload_deactivate(struct adapter *adapter) free_tid_maps(&t->tid_maps); T3C_DATA(tdev) = NULL; - t3_free_l2t(L2DATA(tdev)); - L2DATA(tdev) = NULL; + rcu_read_lock(); + d = L2DATA(tdev); + rcu_read_unlock(); + rcu_assign_pointer(tdev->l2opt, NULL); + call_rcu(&d->rcu_head, clean_l2_data); if (t->nofail_skb) kfree_skb(t->nofail_skb); kfree(t); diff --git a/drivers/net/cxgb3/l2t.c b/drivers/net/cxgb3/l2t.c index f452c40..4154097 100644 --- a/drivers/net/cxgb3/l2t.c +++ b/drivers/net/cxgb3/l2t.c @@ -300,14 +300,21 @@ static inline void reuse_entry(struct l2t_entry *e, struct neighbour *neigh) struct l2t_entry *t3_l2t_get(struct t3cdev *cdev, struct neighbour *neigh, struct net_device *dev) { - struct l2t_entry *e; - struct l2t_data *d = L2DATA(cdev); + struct l2t_entry *e = NULL; + struct l2t_data *d; + int hash; u32 addr = *(u32 *) neigh->primary_key; int ifidx = neigh->dev->ifindex; - int hash = arp_hash(addr, ifidx, d); struct port_info *p = netdev_priv(dev); int smt_idx = p->port_id; + rcu_read_lock(); + d = L2DATA(cdev); + if (!d) + goto done_rcu; + + hash = arp_hash(addr, ifidx, d); + write_lock_bh(&d->lock); for (e = d->l2tab[hash].first; e; e = e->next) if (e->addr == addr && e->ifindex == ifidx && @@ -338,6 +345,8 @@ struct l2t_entry *t3_l2t_get(struct t3cdev *cdev, struct neighbour *neigh, } done: write_unlock_bh(&d->lock); +done_rcu: + rcu_read_unlock(); return e; } diff --git a/drivers/net/cxgb3/l2t.h b/drivers/net/cxgb3/l2t.h index 7a12d52..c5f5479 100644 --- a/drivers/net/cxgb3/l2t.h +++ b/drivers/net/cxgb3/l2t.h @@ -76,6 +76,7 @@ struct l2t_data { atomic_t nfree; /* number of free entries */ rwlock_t lock; struct l2t_entry l2tab[0]; + struct rcu_head rcu_head; /* to handle rcu cleanup */ }; typedef void (*arp_failure_handler_func)(struct t3cdev * dev, @@ -99,7 +100,7 @@ static inline void set_arp_failure_handler(struct sk_buff *skb, /* * Getting to the L2 data from an offload device. */ -#define L2DATA(dev) ((dev)->l2opt) +#define L2DATA(cdev) (rcu_dereference((cdev)->l2opt)) #define W_TCB_L2T_IX 0 #define S_TCB_L2T_IX 7 @@ -126,15 +127,22 @@ static inline int l2t_send(struct t3cdev *dev, struct sk_buff *skb, return t3_l2t_send_slow(dev, skb, e); } -static inline void l2t_release(struct l2t_data *d, struct l2t_entry *e) +static inline void l2t_release(struct t3cdev *t, struct l2t_entry *e) { - if (atomic_dec_and_test(&e->refcnt)) + struct l2t_data *d; + + rcu_read_lock(); + d = L2DATA(t); + + if (atomic_dec_and_test(&e->refcnt) && d) t3_l2e_free(d, e); + + rcu_read_unlock(); } static inline void l2t_hold(struct l2t_data *d, struct l2t_entry *e) { - if (atomic_add_return(1, &e->refcnt) == 1) /* 0 -> 1 transition */ + if (d && atomic_add_return(1, &e->refcnt) == 1) /* 0 -> 1 transition */ atomic_dec(&d->nfree); } diff --git a/drivers/scsi/cxgbi/cxgb3i/cxgb3i.c b/drivers/scsi/cxgbi/cxgb3i/cxgb3i.c index bd22041..f586448 100644 --- a/drivers/scsi/cxgbi/cxgb3i/cxgb3i.c +++ b/drivers/scsi/cxgbi/cxgb3i/cxgb3i.c @@ -913,7 +913,7 @@ static void l2t_put(struct cxgbi_sock *csk) struct t3cdev *t3dev = (struct t3cdev *)csk->cdev->lldev; if (csk->l2t) { - l2t_release(L2DATA(t3dev), csk->l2t); + l2t_release(t3dev, csk->l2t); csk->l2t = NULL; cxgbi_sock_put(csk); } -- cgit v0.10.2 From 96067723e46b0dd24ae7b934085ab4eff4d26a1b Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Sun, 18 Sep 2011 18:56:20 +0400 Subject: [SCSI] 3w-9xxx: fix iommu_iova leak Following reports on the list, it looks like the 3e-9xxx driver will leak dma mappings every time we get a transient queueing error back from the card. This is because it maps the sg list in the routine that sends the command, but doesn't unmap again in the transient failure path (even though the command is sent back to the block layer). Fix by unmapping before returning the status. Reported-by: Chris Boot Tested-by: Chris Boot Acked-by: Adam Radford Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/scsi/3w-9xxx.c b/drivers/scsi/3w-9xxx.c index b7bd5b0..3868ab2 100644 --- a/drivers/scsi/3w-9xxx.c +++ b/drivers/scsi/3w-9xxx.c @@ -1800,10 +1800,12 @@ static int twa_scsi_queue_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_ switch (retval) { case SCSI_MLQUEUE_HOST_BUSY: twa_free_request_id(tw_dev, request_id); + twa_unmap_scsi_data(tw_dev, request_id); break; case 1: tw_dev->state[request_id] = TW_S_COMPLETED; twa_free_request_id(tw_dev, request_id); + twa_unmap_scsi_data(tw_dev, request_id); SCpnt->result = (DID_ERROR << 16); done(SCpnt); retval = 0; -- cgit v0.10.2 From f9783ec862ea8ce0071f34a7fd028229d9fd98b4 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 26 Sep 2011 16:40:33 +0200 Subject: [S390] Do not clobber personality flags on exec Analog to git commit 59e4c3a2fe9cb1681bb2cff508ff79466f7585ba do not clear the additional personality flags on exec. We need to inherit the personality bits in PER_MASK across exec. Signed-off-by: Martin Schwidefsky diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index 64b61bf..547f1a6 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -188,7 +188,8 @@ extern char elf_platform[]; #define SET_PERSONALITY(ex) \ do { \ if (personality(current->personality) != PER_LINUX32) \ - set_personality(PER_LINUX); \ + set_personality(PER_LINUX | \ + (current->personality & ~PER_MASK)); \ if ((ex).e_ident[EI_CLASS] == ELFCLASS32) \ set_thread_flag(TIF_31BIT); \ else \ -- cgit v0.10.2 From 05873df981ca1dd32f398e7b4e19864de907e064 Mon Sep 17 00:00:00 2001 From: Carsten Otte Date: Mon, 26 Sep 2011 16:40:34 +0200 Subject: [S390] gmap: always up mmap_sem properly If gmap_unmap_segment figures that the segment was not mapped in the first place, it need to up mmap_sem on exit. Cc: Signed-off-by: Carsten Otte Signed-off-by: Martin Schwidefsky diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index f69ff3c..5d56c2b 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -303,15 +303,15 @@ int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len) /* Walk the guest addr space page table */ table = gmap->table + (((to + off) >> 53) & 0x7ff); if (*table & _REGION_ENTRY_INV) - return 0; + goto out; table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); table = table + (((to + off) >> 42) & 0x7ff); if (*table & _REGION_ENTRY_INV) - return 0; + goto out; table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); table = table + (((to + off) >> 31) & 0x7ff); if (*table & _REGION_ENTRY_INV) - return 0; + goto out; table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); table = table + (((to + off) >> 20) & 0x7ff); @@ -319,6 +319,7 @@ int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len) flush |= gmap_unlink_segment(gmap, table); *table = _SEGMENT_ENTRY_INV; } +out: up_read(&gmap->mm->mmap_sem); if (flush) gmap_flush_tlb(gmap); -- cgit v0.10.2 From a681887f7b0f8b7698e4b22f1191015424fa96e4 Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Mon, 26 Sep 2011 16:40:35 +0200 Subject: [S390] cio: fix cio_tpi ignoring adapter interrupts Ensure that adapter interrupts are correctly processed when they are retrieved using TEST PENDING INTERRUPTION. Signed-off-by: Peter Oberparleiter Signed-off-by: Martin Schwidefsky diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c index cbde448..eb3140e 100644 --- a/drivers/s390/cio/cio.c +++ b/drivers/s390/cio/cio.c @@ -654,8 +654,8 @@ static struct io_subchannel_private console_priv; static int console_subchannel_in_use; /* - * Use tpi to get a pending interrupt, call the interrupt handler and - * return a pointer to the subchannel structure. + * Use cio_tpi to get a pending interrupt and call the interrupt handler. + * Return non-zero if an interrupt was processed, zero otherwise. */ static int cio_tpi(void) { @@ -667,6 +667,10 @@ static int cio_tpi(void) tpi_info = (struct tpi_info *)&S390_lowcore.subchannel_id; if (tpi(NULL) != 1) return 0; + if (tpi_info->adapter_IO) { + do_adapter_IO(tpi_info->isc); + return 1; + } irb = (struct irb *)&S390_lowcore.irb; /* Store interrupt response block to lowcore. */ if (tsch(tpi_info->schid, irb) != 0) -- cgit v0.10.2 From e8b364b88cc4001b21c28c1ecf1e1e3ffbe162e6 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 26 Sep 2011 19:40:23 +0200 Subject: PM / Clocks: Do not acquire a mutex under a spinlock Commit b7ab83e (PM: Use spinlock instead of mutex in clock management functions) introduced a regression causing clocks_mutex to be acquired under a spinlock. This happens because pm_clk_suspend() and pm_clk_resume() call pm_clk_acquire() under pcd->lock, but pm_clk_acquire() executes clk_get() which causes clocks_mutex to be acquired. Similarly, __pm_clk_remove(), executed under pcd->lock, calls clk_put(), which also causes clocks_mutex to be acquired. To fix those problems make pm_clk_add() call pm_clk_acquire(), so that pm_clk_suspend() and pm_clk_resume() don't have to do that. Change pm_clk_remove() and pm_clk_destroy() to separate modifications of the pcd->clock_list list from the actual removal of PM clock entry objects done by __pm_clk_remove(). Reported-and-tested-by: Guennadi Liakhovetski Signed-off-by: Rafael J. Wysocki Acked-by: Russell King diff --git a/drivers/base/power/clock_ops.c b/drivers/base/power/clock_ops.c index 2c18d58..b97294e 100644 --- a/drivers/base/power/clock_ops.c +++ b/drivers/base/power/clock_ops.c @@ -42,6 +42,22 @@ static struct pm_clk_data *__to_pcd(struct device *dev) } /** + * pm_clk_acquire - Acquire a device clock. + * @dev: Device whose clock is to be acquired. + * @ce: PM clock entry corresponding to the clock. + */ +static void pm_clk_acquire(struct device *dev, struct pm_clock_entry *ce) +{ + ce->clk = clk_get(dev, ce->con_id); + if (IS_ERR(ce->clk)) { + ce->status = PCE_STATUS_ERROR; + } else { + ce->status = PCE_STATUS_ACQUIRED; + dev_dbg(dev, "Clock %s managed by runtime PM.\n", ce->con_id); + } +} + +/** * pm_clk_add - Start using a device clock for power management. * @dev: Device whose clock is going to be used for power management. * @con_id: Connection ID of the clock. @@ -73,6 +89,8 @@ int pm_clk_add(struct device *dev, const char *con_id) } } + pm_clk_acquire(dev, ce); + spin_lock_irq(&pcd->lock); list_add_tail(&ce->node, &pcd->clock_list); spin_unlock_irq(&pcd->lock); @@ -82,17 +100,12 @@ int pm_clk_add(struct device *dev, const char *con_id) /** * __pm_clk_remove - Destroy PM clock entry. * @ce: PM clock entry to destroy. - * - * This routine must be called under the spinlock protecting the PM list of - * clocks corresponding the the @ce's device. */ static void __pm_clk_remove(struct pm_clock_entry *ce) { if (!ce) return; - list_del(&ce->node); - if (ce->status < PCE_STATUS_ERROR) { if (ce->status == PCE_STATUS_ENABLED) clk_disable(ce->clk); @@ -126,18 +139,22 @@ void pm_clk_remove(struct device *dev, const char *con_id) spin_lock_irq(&pcd->lock); list_for_each_entry(ce, &pcd->clock_list, node) { - if (!con_id && !ce->con_id) { - __pm_clk_remove(ce); - break; - } else if (!con_id || !ce->con_id) { + if (!con_id && !ce->con_id) + goto remove; + else if (!con_id || !ce->con_id) continue; - } else if (!strcmp(con_id, ce->con_id)) { - __pm_clk_remove(ce); - break; - } + else if (!strcmp(con_id, ce->con_id)) + goto remove; } spin_unlock_irq(&pcd->lock); + return; + + remove: + list_del(&ce->node); + spin_unlock_irq(&pcd->lock); + + __pm_clk_remove(ce); } /** @@ -175,20 +192,27 @@ void pm_clk_destroy(struct device *dev) { struct pm_clk_data *pcd = __to_pcd(dev); struct pm_clock_entry *ce, *c; + struct list_head list; if (!pcd) return; dev->power.subsys_data = NULL; + INIT_LIST_HEAD(&list); spin_lock_irq(&pcd->lock); list_for_each_entry_safe_reverse(ce, c, &pcd->clock_list, node) - __pm_clk_remove(ce); + list_move(&ce->node, &list); spin_unlock_irq(&pcd->lock); kfree(pcd); + + list_for_each_entry_safe_reverse(ce, c, &list, node) { + list_del(&ce->node); + __pm_clk_remove(ce); + } } #endif /* CONFIG_PM */ @@ -196,23 +220,6 @@ void pm_clk_destroy(struct device *dev) #ifdef CONFIG_PM_RUNTIME /** - * pm_clk_acquire - Acquire a device clock. - * @dev: Device whose clock is to be acquired. - * @con_id: Connection ID of the clock. - */ -static void pm_clk_acquire(struct device *dev, - struct pm_clock_entry *ce) -{ - ce->clk = clk_get(dev, ce->con_id); - if (IS_ERR(ce->clk)) { - ce->status = PCE_STATUS_ERROR; - } else { - ce->status = PCE_STATUS_ACQUIRED; - dev_dbg(dev, "Clock %s managed by runtime PM.\n", ce->con_id); - } -} - -/** * pm_clk_suspend - Disable clocks in a device's PM clock list. * @dev: Device to disable the clocks for. */ @@ -230,9 +237,6 @@ int pm_clk_suspend(struct device *dev) spin_lock_irqsave(&pcd->lock, flags); list_for_each_entry_reverse(ce, &pcd->clock_list, node) { - if (ce->status == PCE_STATUS_NONE) - pm_clk_acquire(dev, ce); - if (ce->status < PCE_STATUS_ERROR) { clk_disable(ce->clk); ce->status = PCE_STATUS_ACQUIRED; @@ -262,9 +266,6 @@ int pm_clk_resume(struct device *dev) spin_lock_irqsave(&pcd->lock, flags); list_for_each_entry(ce, &pcd->clock_list, node) { - if (ce->status == PCE_STATUS_NONE) - pm_clk_acquire(dev, ce); - if (ce->status < PCE_STATUS_ERROR) { clk_enable(ce->clk); ce->status = PCE_STATUS_ENABLED; -- cgit v0.10.2 From 83427c23a44e34be2985f5b1440c645ea8577938 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Tue, 27 Sep 2011 07:00:53 +0900 Subject: ARM: S5PV210: Rename sclk_cam clocks for FIMC media driver The sclk_cam clocks are now controlled by the top level FIMC media device driver bound to "s5p-fimc-md" platform device. Rename sclk_cam clocks so they accessible by the corresponding driver. Signed-off-by: Sylwester Nawrocki 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 52a8e60..f5f8fa8 100644 --- a/arch/arm/mach-s5pv210/clock.c +++ b/arch/arm/mach-s5pv210/clock.c @@ -815,8 +815,7 @@ static struct clksrc_clk clksrcs[] = { .reg_div = { .reg = S5P_CLK_DIV3, .shift = 20, .size = 4 }, }, { .clk = { - .name = "sclk_cam", - .devname = "s5pv210-fimc.0", + .name = "sclk_cam0", .enable = s5pv210_clk_mask0_ctrl, .ctrlbit = (1 << 3), }, @@ -825,8 +824,7 @@ static struct clksrc_clk clksrcs[] = { .reg_div = { .reg = S5P_CLK_DIV1, .shift = 12, .size = 4 }, }, { .clk = { - .name = "sclk_cam", - .devname = "s5pv210-fimc.1", + .name = "sclk_cam1", .enable = s5pv210_clk_mask0_ctrl, .ctrlbit = (1 << 4), }, -- cgit v0.10.2 From 00aaad227aeb30800ac85b981f3ea4e20ae9bd9e Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Tue, 27 Sep 2011 07:00:59 +0900 Subject: ARM: EXYNOS4: Rename sclk_cam clocks for FIMC driver The sclk_cam clocks are now controlled by the top level FIMC media device driver bound to "s5p-fimc-md" platform device. Rename sclk_cam clocks so they accessible by the corresponding driver. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Kukjin Kim diff --git a/arch/arm/mach-exynos4/clock.c b/arch/arm/mach-exynos4/clock.c index 79d6cd0..86964d2 100644 --- a/arch/arm/mach-exynos4/clock.c +++ b/arch/arm/mach-exynos4/clock.c @@ -899,8 +899,7 @@ static struct clksrc_clk clksrcs[] = { .reg_div = { .reg = S5P_CLKDIV_CAM, .shift = 28, .size = 4 }, }, { .clk = { - .name = "sclk_cam", - .devname = "exynos4-fimc.0", + .name = "sclk_cam0", .enable = exynos4_clksrc_mask_cam_ctrl, .ctrlbit = (1 << 16), }, @@ -909,8 +908,7 @@ static struct clksrc_clk clksrcs[] = { .reg_div = { .reg = S5P_CLKDIV_CAM, .shift = 16, .size = 4 }, }, { .clk = { - .name = "sclk_cam", - .devname = "exynos4-fimc.1", + .name = "sclk_cam1", .enable = exynos4_clksrc_mask_cam_ctrl, .ctrlbit = (1 << 20), }, -- cgit v0.10.2 From d94c177beeb4469cd4f6e83354ab0223353e98ed Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 26 Sep 2011 17:44:55 -0700 Subject: vfs pathname lookup: Add LOOKUP_AUTOMOUNT flag Since we've now turned around and made LOOKUP_FOLLOW *not* force an automount, we want to add the ability to force an automount event on lookup even if we don't happen to have one of the other flags that force it implicitly (LOOKUP_OPEN, LOOKUP_DIRECTORY, LOOKUP_PARENT..) Most cases will never want to use this, since you'd normally want to delay automounting as long as possible, which usually implies LOOKUP_OPEN (when we open a file or directory, we really cannot avoid the automount any more). But Trond argued sufficiently forcefully that at a minimum bind mounting a file and quotactl will want to force the automount lookup. Some other cases (like nfs_follow_remote_path()) could use it too, although LOOKUP_DIRECTORY would work there as well. This commit just adds the flag and logic, no users yet, though. It also doesn't actually touch the LOOKUP_NO_AUTOMOUNT flag that is related, and was made irrelevant by the same change that made us not follow on LOOKUP_FOLLOW. Cc: Trond Myklebust Cc: Ian Kent Cc: Jeff Layton Cc: Miklos Szeredi Cc: David Howells Cc: Al Viro Cc: Greg KH Signed-off-by: Linus Torvalds diff --git a/fs/namei.c b/fs/namei.c index f478836..09606fd 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -739,7 +739,7 @@ static int follow_automount(struct path *path, unsigned flags, * of the daemon to instantiate them before they can be used. */ if (!(flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY | - LOOKUP_OPEN | LOOKUP_CREATE)) && + LOOKUP_OPEN | LOOKUP_CREATE | LOOKUP_AUTOMOUNT)) && path->dentry->d_inode) return -EISDIR; diff --git a/include/linux/namei.h b/include/linux/namei.h index 76fe2c6..e13dac7 100644 --- a/include/linux/namei.h +++ b/include/linux/namei.h @@ -48,6 +48,7 @@ enum {LAST_NORM, LAST_ROOT, LAST_DOT, LAST_DOTDOT, LAST_BIND}; */ #define LOOKUP_FOLLOW 0x0001 #define LOOKUP_DIRECTORY 0x0002 +#define LOOKUP_AUTOMOUNT 0x0004 #define LOOKUP_PARENT 0x0010 #define LOOKUP_REVAL 0x0020 -- cgit v0.10.2 From 815d405ceff0d6964683f033e18b9b23a88fba87 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Sep 2011 20:36:09 -0400 Subject: VFS: Fix the remaining automounter semantics regressions The concensus seems to be that system calls such as stat() etc should not trigger an automount. Neither should the l* versions. This patch therefore adds a LOOKUP_AUTOMOUNT flag to tag those lookups that _should_ trigger an automount on the last path element. Signed-off-by: Trond Myklebust [ Edited to leave out the cases that are already covered by LOOKUP_OPEN, LOOKUP_DIRECTORY and LOOKUP_CREATE - all of which also fundamentally force automounting for their own reasons - Linus ] Signed-off-by: Linus Torvalds diff --git a/fs/namespace.c b/fs/namespace.c index 22bfe82..b4febb2 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1757,7 +1757,7 @@ static int do_loopback(struct path *path, char *old_name, return err; if (!old_name || !*old_name) return -EINVAL; - err = kern_path(old_name, LOOKUP_FOLLOW, &old_path); + err = kern_path(old_name, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &old_path); if (err) return err; diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 9b7dd70..5b19b6a 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2798,7 +2798,7 @@ static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt, goto out_put_mnt_ns; ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt, - export_path, LOOKUP_FOLLOW, &path); + export_path, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &path); nfs_referral_loop_unprotect(); put_mnt_ns(ns_private); diff --git a/fs/quota/quota.c b/fs/quota/quota.c index b34bdb2..10b6be3 100644 --- a/fs/quota/quota.c +++ b/fs/quota/quota.c @@ -355,7 +355,7 @@ SYSCALL_DEFINE4(quotactl, unsigned int, cmd, const char __user *, special, * resolution (think about autofs) and thus deadlocks could arise. */ if (cmds == Q_QUOTAON) { - ret = user_path_at(AT_FDCWD, addr, LOOKUP_FOLLOW, &path); + ret = user_path_at(AT_FDCWD, addr, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &path); if (ret) pathp = ERR_PTR(ret); else -- cgit v0.10.2 From 9058020cd9ae3423d6fe7de591698dc96b6701aa Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Tue, 27 Sep 2011 11:08:46 +0200 Subject: ASoC: ssm2602: Re-enable oscillator after suspend Currently the the internal oscillator is powered down when entering BIAS_OFF state, but not re-enabled when going back to BIAS_STANDBY. As a result the CODEC will stop working after suspend if the internal oscillator is used to generate the sysclock signal. This patch fixes it by clearing the appropriate bit in the power down register when the CODEC is re-enabled. Signed-off-by: Lars-Peter Clausen Signed-off-by: Mark Brown Cc: stable@kernel.org diff --git a/sound/soc/codecs/ssm2602.c b/sound/soc/codecs/ssm2602.c index 84f4ad5..9801cd7 100644 --- a/sound/soc/codecs/ssm2602.c +++ b/sound/soc/codecs/ssm2602.c @@ -431,7 +431,8 @@ static int ssm2602_set_dai_fmt(struct snd_soc_dai *codec_dai, static int ssm2602_set_bias_level(struct snd_soc_codec *codec, enum snd_soc_bias_level level) { - u16 reg = snd_soc_read(codec, SSM2602_PWR) & 0xff7f; + u16 reg = snd_soc_read(codec, SSM2602_PWR); + reg &= ~(PWR_POWER_OFF | PWR_OSC_PDN); switch (level) { case SND_SOC_BIAS_ON: -- cgit v0.10.2 From b6c8069d3577481390b3f24a8434ad72a3235594 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 27 Sep 2011 08:12:33 -0700 Subject: vfs: remove LOOKUP_NO_AUTOMOUNT flag That flag no longer makes sense, since we don't look up automount points as eagerly any more. Additionally, it turns out that the NO_AUTOMOUNT handling was buggy to begin with: it would avoid automounting even for cases where we really *needed* to do the automount handling, and could return ENOENT for autofs entries that hadn't been instantiated yet. With our new non-eager automount semantics, one discussion has been about adding a AT_AUTOMOUNT flag to vfs_fstatat (and thus the newfstatat() and fstatat64() system calls), but it's probably not worth it: you can always force at least directory automounting by simply adding the final '/' to the filename, which works for *all* of the stat family system calls, old and new. So AT_NO_AUTOMOUNT (and thus LOOKUP_NO_AUTOMOUNT) really were just a result of our bad default behavior. Acked-by: Ian Kent Acked-by: Trond Myklebust Signed-off-by: Linus Torvalds diff --git a/fs/namei.c b/fs/namei.c index 09606fd..0b3138d 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -721,12 +721,6 @@ static int follow_automount(struct path *path, unsigned flags, if (!path->dentry->d_op || !path->dentry->d_op->d_automount) return -EREMOTE; - /* We don't want to mount if someone supplied AT_NO_AUTOMOUNT - * and this is the terminal part of the path. - */ - if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_PARENT)) - return -EISDIR; /* we actually want to stop here */ - /* We don't want to mount if someone's just doing a stat - * unless they're stat'ing a directory and appended a '/' to * the name. diff --git a/fs/stat.c b/fs/stat.c index ba5316f..78a3aa8 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -81,8 +81,6 @@ int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat, if (!(flag & AT_SYMLINK_NOFOLLOW)) lookup_flags |= LOOKUP_FOLLOW; - if (flag & AT_NO_AUTOMOUNT) - lookup_flags |= LOOKUP_NO_AUTOMOUNT; if (flag & AT_EMPTY_PATH) lookup_flags |= LOOKUP_EMPTY; diff --git a/include/linux/namei.h b/include/linux/namei.h index e13dac7..409328d 100644 --- a/include/linux/namei.h +++ b/include/linux/namei.h @@ -53,7 +53,7 @@ enum {LAST_NORM, LAST_ROOT, LAST_DOT, LAST_DOTDOT, LAST_BIND}; #define LOOKUP_PARENT 0x0010 #define LOOKUP_REVAL 0x0020 #define LOOKUP_RCU 0x0040 -#define LOOKUP_NO_AUTOMOUNT 0x0080 + /* * Intent data */ -- cgit v0.10.2 From e451e61b56b0401442f7a306cd309c3b0c56c285 Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Tue, 27 Sep 2011 13:26:27 -0400 Subject: net: fix a typo in Documentation/networking/scaling.txt Signed-off-by: Jason Wang Signed-off-by: David S. Miller diff --git a/Documentation/networking/scaling.txt b/Documentation/networking/scaling.txt index 58fd741..8ce7c30 100644 --- a/Documentation/networking/scaling.txt +++ b/Documentation/networking/scaling.txt @@ -243,7 +243,7 @@ configured. The number of entries in the global flow table is set through: The number of entries in the per-queue flow table are set through: - /sys/class/net//queues/tx-/rps_flow_cnt + /sys/class/net//queues/rx-/rps_flow_cnt == Suggested Configuration -- cgit v0.10.2 From f2eaeb58bf6995a979c413ea0cc73289533feacb Mon Sep 17 00:00:00 2001 From: Dmitry Kravkov Date: Thu, 22 Sep 2011 02:33:31 +0000 Subject: bnx2x: fix hw attention handling Use register name to initialize attention mask Signed-off-by: Dmitry Kravkov Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller diff --git a/drivers/net/bnx2x/bnx2x_main.c b/drivers/net/bnx2x/bnx2x_main.c index c027e93..6b46917 100644 --- a/drivers/net/bnx2x/bnx2x_main.c +++ b/drivers/net/bnx2x/bnx2x_main.c @@ -4943,7 +4943,7 @@ static void bnx2x_init_def_sb(struct bnx2x *bp) int igu_seg_id; int port = BP_PORT(bp); int func = BP_FUNC(bp); - int reg_offset; + int reg_offset, reg_offset_en5; u64 section; int index; struct hc_sp_status_block_data sp_sb_data; @@ -4966,6 +4966,8 @@ static void bnx2x_init_def_sb(struct bnx2x *bp) reg_offset = (port ? MISC_REG_AEU_ENABLE1_FUNC_1_OUT_0 : MISC_REG_AEU_ENABLE1_FUNC_0_OUT_0); + reg_offset_en5 = (port ? MISC_REG_AEU_ENABLE5_FUNC_1_OUT_0 : + MISC_REG_AEU_ENABLE5_FUNC_0_OUT_0); for (index = 0; index < MAX_DYNAMIC_ATTN_GRPS; index++) { int sindex; /* take care of sig[0]..sig[4] */ @@ -4980,7 +4982,7 @@ static void bnx2x_init_def_sb(struct bnx2x *bp) * and not 16 between the different groups */ bp->attn_group[index].sig[4] = REG_RD(bp, - reg_offset + 0x10 + 0x4*index); + reg_offset_en5 + 0x4*index); else bp->attn_group[index].sig[4] = 0; } diff --git a/drivers/net/bnx2x/bnx2x_reg.h b/drivers/net/bnx2x/bnx2x_reg.h index 750e844..fc7bd0f 100644 --- a/drivers/net/bnx2x/bnx2x_reg.h +++ b/drivers/net/bnx2x/bnx2x_reg.h @@ -1384,6 +1384,18 @@ Latched ump_tx_parity; [31] MCP Latched scpad_parity; */ #define MISC_REG_AEU_ENABLE4_PXP_0 0xa108 #define MISC_REG_AEU_ENABLE4_PXP_1 0xa1a8 +/* [RW 32] fifth 32b for enabling the output for function 0 output0. Mapped + * as follows: [0] PGLUE config_space; [1] PGLUE misc_flr; [2] PGLUE B RBC + * attention [3] PGLUE B RBC parity; [4] ATC attention; [5] ATC parity; [6] + * mstat0 attention; [7] mstat0 parity; [8] mstat1 attention; [9] mstat1 + * parity; [31-10] Reserved; */ +#define MISC_REG_AEU_ENABLE5_FUNC_0_OUT_0 0xa688 +/* [RW 32] Fifth 32b for enabling the output for function 1 output0. Mapped + * as follows: [0] PGLUE config_space; [1] PGLUE misc_flr; [2] PGLUE B RBC + * attention [3] PGLUE B RBC parity; [4] ATC attention; [5] ATC parity; [6] + * mstat0 attention; [7] mstat0 parity; [8] mstat1 attention; [9] mstat1 + * parity; [31-10] Reserved; */ +#define MISC_REG_AEU_ENABLE5_FUNC_1_OUT_0 0xa6b0 /* [RW 1] set/clr general attention 0; this will set/clr bit 94 in the aeu 128 bit vector */ #define MISC_REG_AEU_GENERAL_ATTN_0 0xa000 -- cgit v0.10.2 From f99779035730f8c3be8aeadd386cc9e77e086446 Mon Sep 17 00:00:00 2001 From: Dmitry Kravkov Date: Thu, 22 Sep 2011 02:33:32 +0000 Subject: bnx2x: fix WOL by enablement PME in config space Signed-off-by: Dmitry Kravkov Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller diff --git a/drivers/net/bnx2x/bnx2x_main.c b/drivers/net/bnx2x/bnx2x_main.c index 6b46917..15f8000 100644 --- a/drivers/net/bnx2x/bnx2x_main.c +++ b/drivers/net/bnx2x/bnx2x_main.c @@ -7627,8 +7627,11 @@ u32 bnx2x_send_unload_req(struct bnx2x *bp, int unload_mode) u32 emac_base = port ? GRCBASE_EMAC1 : GRCBASE_EMAC0; u8 *mac_addr = bp->dev->dev_addr; u32 val; + u16 pmc; + /* The mac address is written to entries 1-4 to - preserve entry 0 which is used by the PMF */ + * preserve entry 0 which is used by the PMF + */ u8 entry = (BP_VN(bp) + 1)*8; val = (mac_addr[0] << 8) | mac_addr[1]; @@ -7638,6 +7641,11 @@ u32 bnx2x_send_unload_req(struct bnx2x *bp, int unload_mode) (mac_addr[4] << 8) | mac_addr[5]; EMAC_WR(bp, EMAC_REG_EMAC_MAC_MATCH + entry + 4, val); + /* Enable the PME and clear the status */ + pci_read_config_word(bp->pdev, bp->pm_cap + PCI_PM_CTRL, &pmc); + pmc |= PCI_PM_CTRL_PME_ENABLE | PCI_PM_CTRL_PME_STATUS; + pci_write_config_word(bp->pdev, bp->pm_cap + PCI_PM_CTRL, pmc); + reset_code = DRV_MSG_CODE_UNLOAD_REQ_WOL_EN; } else -- cgit v0.10.2 From a0babc80eb375aad69dab4687c3b44c47d42f241 Mon Sep 17 00:00:00 2001 From: Shmulik Ravid Date: Thu, 22 Sep 2011 02:33:33 +0000 Subject: bnx2x: add missing break in bnx2x_dcbnl_get_cap Signed-off-by: Dmitry Kravkov Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller diff --git a/drivers/net/bnx2x/bnx2x_dcb.c b/drivers/net/bnx2x/bnx2x_dcb.c index a1e004a..0b4acf6 100644 --- a/drivers/net/bnx2x/bnx2x_dcb.c +++ b/drivers/net/bnx2x/bnx2x_dcb.c @@ -2120,6 +2120,7 @@ static u8 bnx2x_dcbnl_get_cap(struct net_device *netdev, int capid, u8 *cap) break; case DCB_CAP_ATTR_DCBX: *cap = BNX2X_DCBX_CAPS; + break; default: rval = -EINVAL; break; -- cgit v0.10.2 From 2015de5fe2a47086a3260802275932bfd810884e Mon Sep 17 00:00:00 2001 From: Ben Greear Date: Tue, 27 Sep 2011 15:16:08 -0400 Subject: ipv6-multicast: Fix memory leak in input path. Have to free the skb before returning if we fail the fib lookup. Signed-off-by: Ben Greear Signed-off-by: David S. Miller diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c index 705c828..825d02f 100644 --- a/net/ipv6/ip6mr.c +++ b/net/ipv6/ip6mr.c @@ -2052,8 +2052,10 @@ int ip6_mr_input(struct sk_buff *skb) int err; err = ip6mr_fib_lookup(net, &fl6, &mrt); - if (err < 0) + if (err < 0) { + kfree_skb(skb); return err; + } read_lock(&mrt_lock); cache = ip6mr_cache_find(mrt, -- cgit v0.10.2 From d4cae56219755ccf8acfc8e2c1927009ff29d8c6 Mon Sep 17 00:00:00 2001 From: Madalin Bucur Date: Mon, 26 Sep 2011 07:04:36 +0000 Subject: net: check return value for dst_alloc return value of dst_alloc must be checked before use Signed-off-by: Madalin Bucur Signed-off-by: David S. Miller diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index 94fdcc7..552df27 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -1349,14 +1349,16 @@ static inline struct xfrm_dst *xfrm_alloc_dst(struct net *net, int family) BUG(); } xdst = dst_alloc(dst_ops, NULL, 0, 0, 0); - memset(&xdst->u.rt6.rt6i_table, 0, sizeof(*xdst) - sizeof(struct dst_entry)); - xfrm_policy_put_afinfo(afinfo); - if (likely(xdst)) + if (likely(xdst)) { + memset(&xdst->u.rt6.rt6i_table, 0, + sizeof(*xdst) - sizeof(struct dst_entry)); xdst->flo.ops = &xfrm_bundle_fc_ops; - else + } else xdst = ERR_PTR(-ENOBUFS); + xfrm_policy_put_afinfo(afinfo); + return xdst; } -- cgit v0.10.2 From fbe58186901155c0cb5398dd343337be0c456c04 Mon Sep 17 00:00:00 2001 From: Madalin Bucur Date: Mon, 26 Sep 2011 07:04:56 +0000 Subject: ipv6: check return value for dst_alloc return value of dst_alloc must be checked before use Signed-off-by: Madalin Bucur Signed-off-by: David S. Miller diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 1250f90..fb545ed 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -244,7 +244,9 @@ static inline struct rt6_info *ip6_dst_alloc(struct dst_ops *ops, { struct rt6_info *rt = dst_alloc(ops, dev, 0, 0, flags); - memset(&rt->rt6i_table, 0, sizeof(*rt) - sizeof(struct dst_entry)); + if (rt != NULL) + memset(&rt->rt6i_table, 0, + sizeof(*rt) - sizeof(struct dst_entry)); return rt; } -- cgit v0.10.2 From 67928c4041606f02725f3c95c4c0404e4532df1b Mon Sep 17 00:00:00 2001 From: Ben Greear Date: Fri, 23 Sep 2011 13:11:01 +0000 Subject: ipv6-multicast: Fix memory leak in IPv6 multicast. If reg_vif_xmit cannot find a routing entry, be sure to free the skb before returning the error. Signed-off-by: Ben Greear Signed-off-by: David S. Miller diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c index 825d02f..def0538 100644 --- a/net/ipv6/ip6mr.c +++ b/net/ipv6/ip6mr.c @@ -696,8 +696,10 @@ static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, int err; err = ip6mr_fib_lookup(net, &fl6, &mrt); - if (err < 0) + if (err < 0) { + kfree_skb(skb); return err; + } read_lock(&mrt_lock); dev->stats.tx_bytes += skb->len; -- cgit v0.10.2 From a102a9ece5489e1718cd7543aa079082450ac3a2 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 27 Sep 2011 15:48:34 -0700 Subject: Linux 3.1-rc8 diff --git a/Makefile b/Makefile index 733dcba..73bb458 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 3 PATCHLEVEL = 1 SUBLEVEL = 0 -EXTRAVERSION = -rc7 +EXTRAVERSION = -rc8 NAME = "Divemaster Edition" # *DOCUMENTATION* -- cgit v0.10.2 From 777eb1bf15b8532c396821774bf6451e563438f5 Mon Sep 17 00:00:00 2001 From: Hannes Reinecke Date: Wed, 28 Sep 2011 08:07:01 -0600 Subject: block: Free queue resources at blk_release_queue() A kernel crash is observed when a mounted ext3/ext4 filesystem is physically removed. The problem is that blk_cleanup_queue() frees up some resources eg by calling elevator_exit(), which are not checked for in normal operation. So we should rather move these calls to the destructor function blk_release_queue() as at that point all remaining references are gone. However, in doing so we have to ensure that any externally supplied queue_lock is disconnected as the driver might free up the lock after the call of blk_cleanup_queue(), Signed-off-by: Hannes Reinecke Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index b2ed78a..d34433a 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -348,9 +348,10 @@ void blk_put_queue(struct request_queue *q) EXPORT_SYMBOL(blk_put_queue); /* - * Note: If a driver supplied the queue lock, it should not zap that lock - * unexpectedly as some queue cleanup components like elevator_exit() and - * blk_throtl_exit() need queue lock. + * Note: If a driver supplied the queue lock, it is disconnected + * by this function. The actual state of the lock doesn't matter + * here as the request_queue isn't accessible after this point + * (QUEUE_FLAG_DEAD is set) and no other requests will be queued. */ void blk_cleanup_queue(struct request_queue *q) { @@ -367,10 +368,8 @@ void blk_cleanup_queue(struct request_queue *q) queue_flag_set_unlocked(QUEUE_FLAG_DEAD, q); mutex_unlock(&q->sysfs_lock); - if (q->elevator) - elevator_exit(q->elevator); - - blk_throtl_exit(q); + if (q->queue_lock != &q->__queue_lock) + q->queue_lock = &q->__queue_lock; blk_put_queue(q); } diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index e681805..60fda88 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -479,6 +479,11 @@ static void blk_release_queue(struct kobject *kobj) blk_sync_queue(q); + if (q->elevator) + elevator_exit(q->elevator); + + blk_throtl_exit(q); + if (rl->rq_pool) mempool_destroy(rl->rq_pool); -- cgit v0.10.2 From 0eb9782ad9b1bd496ba61cd5ea27ccb8db21e885 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 28 Sep 2011 08:11:00 -0700 Subject: hwmon: (coretemp) Fixup platform device ID change With recent change "hwmon: (coretemp) don't use kernel assigned CPU number as platform device ID", the microcode check is now running on random CPU. Fix that by checking the microcode before creating the platform device rather than at probe time. Also avoid calling TO_PHYS_ID(cpu) twice in the same function, it's expensive. Signed-off-by: Jean Delvare Cc: Jan Beulich Cc: Guenter Roeck Signed-off-by: Guenter Roeck diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 44b2391..cf5b1de 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -377,9 +377,9 @@ exit_free: } -static int __devinit chk_ucode_version(struct platform_device *pdev) +static int __cpuinit chk_ucode_version(unsigned int cpu) { - struct cpuinfo_x86 *c = &cpu_data(pdev->id); + struct cpuinfo_x86 *c = &cpu_data(cpu); int err; u32 edx; @@ -390,17 +390,15 @@ static int __devinit chk_ucode_version(struct platform_device *pdev) */ if (c->x86_model == 0xe && c->x86_mask < 0xc) { /* check for microcode update */ - err = smp_call_function_single(pdev->id, get_ucode_rev_on_cpu, + err = smp_call_function_single(cpu, get_ucode_rev_on_cpu, &edx, 1); if (err) { - dev_err(&pdev->dev, - "Cannot determine microcode revision of " - "CPU#%u (%d)!\n", pdev->id, err); + pr_err("Cannot determine microcode revision of " + "CPU#%u (%d)!\n", cpu, err); return -ENODEV; } else if (edx < 0x39) { - dev_err(&pdev->dev, - "Errata AE18 not fixed, update BIOS or " - "microcode of the CPU!\n"); + pr_err("Errata AE18 not fixed, update BIOS or " + "microcode of the CPU!\n"); return -ENODEV; } } @@ -544,11 +542,6 @@ static int __devinit coretemp_probe(struct platform_device *pdev) struct platform_data *pdata; int err; - /* Check the microcode version of the CPU */ - err = chk_ucode_version(pdev); - if (err) - return err; - /* Initialize the per-package data structures */ pdata = kzalloc(sizeof(struct platform_data), GFP_KERNEL); if (!pdata) @@ -630,7 +623,7 @@ static int __cpuinit coretemp_device_add(unsigned int cpu) } pdev_entry->pdev = pdev; - pdev_entry->phys_proc_id = TO_PHYS_ID(cpu); + pdev_entry->phys_proc_id = pdev->id; list_add_tail(&pdev_entry->list, &pdev_list); mutex_unlock(&pdev_list_mutex); @@ -691,6 +684,10 @@ static void __cpuinit get_core_online(unsigned int cpu) return; if (!pdev) { + /* Check the microcode version of the CPU */ + if (chk_ucode_version(cpu)) + return; + /* * Alright, we have DTS support. * We are bringing the _first_ core in this pkg -- cgit v0.10.2 From 20ecb499f64a7e8e7fe03f6098ab25c71b7a6481 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Sat, 24 Sep 2011 15:27:04 -0700 Subject: hwmon: (coretemp) Avoid leaving around dangling pointer Storing the struct temp_data pointer allocated from create_core_data() when returning an error has the potential of leaving around a pointer to freed memory. Reset it to NULL for error returns. Reported-by: Jan Beulich Signed-off-by: Guenter Roeck Acked-by: Jean Delvare diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index cf5b1de..9323837 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -506,6 +506,7 @@ static int create_core_data(struct platform_device *pdev, return 0; exit_free: + pdata->core_data[attr_no] = NULL; kfree(tdata); return err; } -- cgit v0.10.2 From 782e182e91e97f529a1edb30fdece9f1bef90ecc Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 28 Sep 2011 10:08:27 -0700 Subject: libceph: fix pg_temp mapping calculation We need to apply the modulo pg_num calculation before looking up a pgid in the pg_temp mapping rbtree. This fixes pg_temp mappings, and fixes (some) misdirected requests that result in messages like [WRN] client4104 10.0.1.219:0/275025290 misdirected client4104.1:129 0.1 to osd0 not [1,0] in e11/11 on the server and stall make the client block without getting a reply (at least until the pg_temp mapping goes way, but that can take a long long time). Reorder calc_pg_raw() a bit to make more sense. Signed-off-by: Sage Weil diff --git a/net/ceph/osdmap.c b/net/ceph/osdmap.c index e97c358..eceb8d5 100644 --- a/net/ceph/osdmap.c +++ b/net/ceph/osdmap.c @@ -1046,10 +1046,25 @@ static int *calc_pg_raw(struct ceph_osdmap *osdmap, struct ceph_pg pgid, struct ceph_pg_mapping *pg; struct ceph_pg_pool_info *pool; int ruleno; - unsigned poolid, ps, pps; + unsigned poolid, ps, pps, t; int preferred; + poolid = le32_to_cpu(pgid.pool); + ps = le16_to_cpu(pgid.ps); + preferred = (s16)le16_to_cpu(pgid.preferred); + + pool = __lookup_pg_pool(&osdmap->pg_pools, poolid); + if (!pool) + return NULL; + /* pg_temp? */ + if (preferred >= 0) + t = ceph_stable_mod(ps, le32_to_cpu(pool->v.lpg_num), + pool->lpgp_num_mask); + else + t = ceph_stable_mod(ps, le32_to_cpu(pool->v.pg_num), + pool->pgp_num_mask); + pgid.ps = cpu_to_le16(t); pg = __lookup_pg_mapping(&osdmap->pg_temp, pgid); if (pg) { *num = pg->len; @@ -1057,18 +1072,6 @@ static int *calc_pg_raw(struct ceph_osdmap *osdmap, struct ceph_pg pgid, } /* crush */ - poolid = le32_to_cpu(pgid.pool); - ps = le16_to_cpu(pgid.ps); - preferred = (s16)le16_to_cpu(pgid.preferred); - - /* don't forcefeed bad device ids to crush */ - if (preferred >= osdmap->max_osd || - preferred >= osdmap->crush->max_devices) - preferred = -1; - - pool = __lookup_pg_pool(&osdmap->pg_pools, poolid); - if (!pool) - return NULL; ruleno = crush_find_rule(osdmap->crush, pool->v.crush_ruleset, pool->v.type, pool->v.size); if (ruleno < 0) { @@ -1078,6 +1081,11 @@ static int *calc_pg_raw(struct ceph_osdmap *osdmap, struct ceph_pg pgid, return NULL; } + /* don't forcefeed bad device ids to crush */ + if (preferred >= osdmap->max_osd || + preferred >= osdmap->crush->max_devices) + preferred = -1; + if (preferred >= 0) pps = ceph_stable_mod(ps, le32_to_cpu(pool->v.lpgp_num), -- cgit v0.10.2 From 8adc8b3d780363d5df0dd6ace10336e3d7e331a1 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 28 Sep 2011 10:11:04 -0700 Subject: libceph: fix pg_temp mapping update The incremental map updates have a record for each pg_temp mapping that is to be add/updated (len > 0) or removed (len == 0). The old code was written as if the updates were a complete enumeration; that was just wrong. Update the code to remove 0-length entries and drop the rbtree traversal. This avoids misdirected (and hung) requests that manifest as server errors like [WRN] client4104 10.0.1.219:0/275025290 misdirected client4104.1:129 0.1 to osd0 not [1,0] in e11/11 Signed-off-by: Sage Weil diff --git a/net/ceph/osdmap.c b/net/ceph/osdmap.c index eceb8d5..fd863fe 100644 --- a/net/ceph/osdmap.c +++ b/net/ceph/osdmap.c @@ -339,6 +339,7 @@ static int __insert_pg_mapping(struct ceph_pg_mapping *new, struct ceph_pg_mapping *pg = NULL; int c; + dout("__insert_pg_mapping %llx %p\n", *(u64 *)&new->pgid, new); while (*p) { parent = *p; pg = rb_entry(parent, struct ceph_pg_mapping, node); @@ -366,16 +367,33 @@ static struct ceph_pg_mapping *__lookup_pg_mapping(struct rb_root *root, while (n) { pg = rb_entry(n, struct ceph_pg_mapping, node); c = pgid_cmp(pgid, pg->pgid); - if (c < 0) + if (c < 0) { n = n->rb_left; - else if (c > 0) + } else if (c > 0) { n = n->rb_right; - else + } else { + dout("__lookup_pg_mapping %llx got %p\n", + *(u64 *)&pgid, pg); return pg; + } } return NULL; } +static int __remove_pg_mapping(struct rb_root *root, struct ceph_pg pgid) +{ + struct ceph_pg_mapping *pg = __lookup_pg_mapping(root, pgid); + + if (pg) { + dout("__remove_pg_mapping %llx %p\n", *(u64 *)&pgid, pg); + rb_erase(&pg->node, root); + kfree(pg); + return 0; + } + dout("__remove_pg_mapping %llx dne\n", *(u64 *)&pgid); + return -ENOENT; +} + /* * rbtree of pg pool info */ @@ -711,7 +729,6 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end, void *start = *p; int err = -EINVAL; u16 version; - struct rb_node *rbp; ceph_decode_16_safe(p, end, version, bad); if (version > CEPH_OSDMAP_INC_VERSION) { @@ -861,7 +878,6 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end, } /* new_pg_temp */ - rbp = rb_first(&map->pg_temp); ceph_decode_32_safe(p, end, len, bad); while (len--) { struct ceph_pg_mapping *pg; @@ -872,18 +888,6 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end, ceph_decode_copy(p, &pgid, sizeof(pgid)); pglen = ceph_decode_32(p); - /* remove any? */ - while (rbp && pgid_cmp(rb_entry(rbp, struct ceph_pg_mapping, - node)->pgid, pgid) <= 0) { - struct ceph_pg_mapping *cur = - rb_entry(rbp, struct ceph_pg_mapping, node); - - rbp = rb_next(rbp); - dout(" removed pg_temp %llx\n", *(u64 *)&cur->pgid); - rb_erase(&cur->node, &map->pg_temp); - kfree(cur); - } - if (pglen) { /* insert */ ceph_decode_need(p, end, pglen*sizeof(u32), bad); @@ -903,17 +907,11 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end, } dout(" added pg_temp %llx len %d\n", *(u64 *)&pgid, pglen); + } else { + /* remove */ + __remove_pg_mapping(&map->pg_temp, pgid); } } - while (rbp) { - struct ceph_pg_mapping *cur = - rb_entry(rbp, struct ceph_pg_mapping, node); - - rbp = rb_next(rbp); - dout(" removed pg_temp %llx\n", *(u64 *)&cur->pgid); - rb_erase(&cur->node, &map->pg_temp); - kfree(cur); - } /* ignore the rest */ *p = end; -- cgit v0.10.2 From d5767c53535ac79758084773418e0ad186aba4a2 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 28 Sep 2011 10:23:44 -0700 Subject: bootup: move 'usermodehelper_enable()' to the end of do_basic_setup() Doing it just before starting to call into cpu_idle() made a sick kind of sense only because the original bug we fixed (see commit 288d5abec831: "Boot up with usermodehelper disabled") was about problems with some scheduler data structures not being initialized, and they had better be initialized at that point. But it really didn't make any other conceptual sense, and doing it after the initial "schedule()" call for the idle thread actually opened up a race: what if the main initialization thread did everything without needing to sleep, and got all the way into user land too? Without actually having scheduled back to the idle thread? Now, in normal circumstances that doesn't ever happen, but it looks like Richard Cochran triggered exactly that on his ARM IXP4xx machines: "I have some ARM IXP4xx based machines that use the two on chip MAC ports (aka NPEs). The NPE needs a firmware in order to function. Ever since the following commit [that 288d5abec831 one], it is no longer possible to bring up the interfaces during the init scripts." with a call trace showing an ioctl coming from user space. Richard says: "The init is busybox, and the startup script does mount, syslogd, and then ifup, so that all can go by quickly." The fix is to move the usermodehelper_enable() into the main 'init' thread, and just put it after we've done all our initcalls. By then, everything really should be up, but we've obviously not actually started the user-mode portion of init yet. Reported-and-tested-by: Richard Cochran Signed-off-by: Linus Torvalds diff --git a/init/main.c b/init/main.c index 2a9b88a..23702bb 100644 --- a/init/main.c +++ b/init/main.c @@ -381,9 +381,6 @@ static noinline void __init_refok rest_init(void) preempt_enable_no_resched(); schedule(); - /* At this point, we can enable user mode helper functionality */ - usermodehelper_enable(); - /* Call into cpu_idle with preempt disabled */ preempt_disable(); cpu_idle(); @@ -734,6 +731,7 @@ static void __init do_basic_setup(void) init_irq_proc(); do_ctors(); do_initcalls(); + usermodehelper_enable(); } static void __init do_pre_smp_initcalls(void) -- cgit v0.10.2 From f75159e9936143177b442afc78150b7a7ad8aa07 Mon Sep 17 00:00:00 2001 From: Richard Cochran Date: Tue, 20 Sep 2011 01:25:41 +0000 Subject: ptp: fix L2 event message recognition The IEEE 1588 standard defines two kinds of messages, event and general messages. Event messages require time stamping, and general do not. When using UDP transport, two separate ports are used for the two message types. The BPF designed to recognize event messages incorrectly classifies L2 general messages as event messages. This commit fixes the issue by extending the filter to check the message type field for L2 PTP packets. Event messages are be distinguished from general messages by testing the "general" bit. Signed-off-by: Richard Cochran Cc: Signed-off-by: David S. Miller diff --git a/include/linux/ptp_classify.h b/include/linux/ptp_classify.h index e07e274..1dc420b 100644 --- a/include/linux/ptp_classify.h +++ b/include/linux/ptp_classify.h @@ -51,6 +51,7 @@ #define PTP_CLASS_V2_VLAN (PTP_CLASS_V2 | PTP_CLASS_VLAN) #define PTP_EV_PORT 319 +#define PTP_GEN_BIT 0x08 /* indicates general message, if set in message type */ #define OFF_ETYPE 12 #define OFF_IHL 14 @@ -116,14 +117,20 @@ static inline int ptp_filter_init(struct sock_filter *f, int len) {OP_OR, 0, 0, PTP_CLASS_IPV6 }, /* */ \ {OP_RETA, 0, 0, 0 }, /* */ \ /*L3x*/ {OP_RETK, 0, 0, PTP_CLASS_NONE }, /* */ \ -/*L40*/ {OP_JEQ, 0, 6, ETH_P_8021Q }, /* f goto L50 */ \ +/*L40*/ {OP_JEQ, 0, 9, ETH_P_8021Q }, /* f goto L50 */ \ {OP_LDH, 0, 0, OFF_ETYPE + 4 }, /* */ \ - {OP_JEQ, 0, 9, ETH_P_1588 }, /* f goto L60 */ \ + {OP_JEQ, 0, 15, ETH_P_1588 }, /* f goto L60 */ \ + {OP_LDB, 0, 0, ETH_HLEN + VLAN_HLEN }, /* */ \ + {OP_AND, 0, 0, PTP_GEN_BIT }, /* */ \ + {OP_JEQ, 0, 12, 0 }, /* f goto L6x */ \ {OP_LDH, 0, 0, ETH_HLEN + VLAN_HLEN }, /* */ \ {OP_AND, 0, 0, PTP_CLASS_VMASK }, /* */ \ {OP_OR, 0, 0, PTP_CLASS_VLAN }, /* */ \ {OP_RETA, 0, 0, 0 }, /* */ \ -/*L50*/ {OP_JEQ, 0, 4, ETH_P_1588 }, /* f goto L61 */ \ +/*L50*/ {OP_JEQ, 0, 7, ETH_P_1588 }, /* f goto L61 */ \ + {OP_LDB, 0, 0, ETH_HLEN }, /* */ \ + {OP_AND, 0, 0, PTP_GEN_BIT }, /* */ \ + {OP_JEQ, 0, 4, 0 }, /* f goto L6x */ \ {OP_LDH, 0, 0, ETH_HLEN }, /* */ \ {OP_AND, 0, 0, PTP_CLASS_VMASK }, /* */ \ {OP_OR, 0, 0, PTP_CLASS_L2 }, /* */ \ -- cgit v0.10.2 From 02715ed2e7b71569ed6e61f0b23b0736a509913d Mon Sep 17 00:00:00 2001 From: Richard Cochran Date: Tue, 20 Sep 2011 01:25:42 +0000 Subject: dp83640: reduce driver noise The driver has two warning messages that might be triggered by normal use cases. When they appear, the messages give the impression of a never ending series of errors. This commit changes them to debug messages instead. Signed-off-by: Richard Cochran Signed-off-by: David S. Miller diff --git a/drivers/net/phy/dp83640.c b/drivers/net/phy/dp83640.c index cb6e0b4..edd7304 100644 --- a/drivers/net/phy/dp83640.c +++ b/drivers/net/phy/dp83640.c @@ -589,7 +589,7 @@ static void decode_rxts(struct dp83640_private *dp83640, prune_rx_ts(dp83640); if (list_empty(&dp83640->rxpool)) { - pr_warning("dp83640: rx timestamp pool is empty\n"); + pr_debug("dp83640: rx timestamp pool is empty\n"); goto out; } rxts = list_first_entry(&dp83640->rxpool, struct rxts, list); @@ -612,7 +612,7 @@ static void decode_txts(struct dp83640_private *dp83640, skb = skb_dequeue(&dp83640->tx_queue); if (!skb) { - pr_warning("dp83640: have timestamp but tx_queue empty\n"); + pr_debug("dp83640: have timestamp but tx_queue empty\n"); return; } ns = phy2txts(phy_txts); -- cgit v0.10.2 From d3c52173be57ec920deda28923d80b68e5630594 Mon Sep 17 00:00:00 2001 From: Ian Campbell Date: Wed, 21 Sep 2011 22:08:26 +0000 Subject: MAINTAINERS: tehuti: Alexander Indenbaum's address bounces I got: Generating server: Tehuti.onmicrosoft.com baum@tehutinetworks.net #< #5.1.1 smtp;550 5.1.1 RESOLVER.ADR.RecipNotFound; not found> #SMTP# Signed-off-by: Ian Campbell Cc: Alexander Indenbaum Cc: Andy Gospodarek Cc: Andrew Morton Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller diff --git a/MAINTAINERS b/MAINTAINERS index ae8820e..ace8f9c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -6374,7 +6374,6 @@ S: Supported F: arch/arm/mach-tegra TEHUTI ETHERNET DRIVER -M: Alexander Indenbaum M: Andy Gospodarek L: netdev@vger.kernel.org S: Supported -- cgit v0.10.2 From aabdcb0b553b9c9547b1a506b34d55a764745870 Mon Sep 17 00:00:00 2001 From: Oliver Hartkopp Date: Fri, 23 Sep 2011 08:23:47 +0000 Subject: can bcm: fix tx_setup off-by-one errors This patch fixes two off-by-one errors that canceled each other out. Checking for the same condition two times in bcm_tx_timeout_tsklet() reduced the count of frames to be sent by one. This did not show up the first time tx_setup is invoked as an additional frame is sent due to TX_ANNONCE. Invoking a second tx_setup on the same item led to a reduced (by 1) number of sent frames. Reported-by: Andre Naujoks Signed-off-by: Oliver Hartkopp Signed-off-by: David S. Miller diff --git a/net/can/bcm.c b/net/can/bcm.c index d6c8ae5..c9cdb8d 100644 --- a/net/can/bcm.c +++ b/net/can/bcm.c @@ -365,9 +365,6 @@ static void bcm_tx_timeout_tsklet(unsigned long data) bcm_send_to_user(op, &msg_head, NULL, 0); } - } - - if (op->kt_ival1.tv64 && (op->count > 0)) { /* send (next) frame */ bcm_can_tx(op); @@ -970,8 +967,9 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, /* spec: send can_frame when starting timer */ op->flags |= TX_ANNOUNCE; - if (op->kt_ival1.tv64 && (op->count > 0)) { - /* op->count-- is done in bcm_tx_timeout_handler */ + /* only start timer when having more frames than sent below */ + if (op->kt_ival1.tv64 && (op->count > 1)) { + /* op->count-- is done in bcm_tx_timeout_tsklet */ hrtimer_start(&op->timer, op->kt_ival1, HRTIMER_MODE_REL); } else @@ -979,8 +977,11 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, HRTIMER_MODE_REL); } - if (op->flags & TX_ANNOUNCE) + if (op->flags & TX_ANNOUNCE) { bcm_can_tx(op); + if (op->kt_ival1.tv64 && (op->count > 0)) + op->count--; + } return msg_head->nframes * CFSIZ + MHSIZ; } -- cgit v0.10.2 From 6482aa7c120447858da1869197b48eff66435a31 Mon Sep 17 00:00:00 2001 From: Divy Le Ray Date: Sat, 24 Sep 2011 06:11:31 +0000 Subject: cxgb4: Fix EEH on IBM P7IOC Fix EEH recovery on new P Series platform by requesting fundamental reset. Signed-off-by: Divy Le Ray Signed-off-by: David S. Miller diff --git a/drivers/net/cxgb4/cxgb4_main.c b/drivers/net/cxgb4/cxgb4_main.c index c9957b7..b4efa29 100644 --- a/drivers/net/cxgb4/cxgb4_main.c +++ b/drivers/net/cxgb4/cxgb4_main.c @@ -3712,6 +3712,9 @@ static int __devinit init_one(struct pci_dev *pdev, setup_debugfs(adapter); } + /* PCIe EEH recovery on powerpc platforms needs fundamental reset */ + pdev->needs_freset = 1; + if (is_offload(adapter)) attach_ulds(adapter); -- cgit v0.10.2 From 676a1184e8afd4fed7948232df1ff91517400859 Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Sun, 25 Sep 2011 02:21:30 +0000 Subject: ipv6: nullify ipv6_ac_list and ipv6_fl_list when creating new socket ipv6_ac_list and ipv6_fl_list from listening socket are inadvertently shared with new socket created for connection. Signed-off-by: Zheng Yan Signed-off-by: David S. Miller diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 3c9fa61..79cc646 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1383,6 +1383,8 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, newtp->af_specific = &tcp_sock_ipv6_mapped_specific; #endif + newnp->ipv6_ac_list = NULL; + newnp->ipv6_fl_list = NULL; newnp->pktoptions = NULL; newnp->opt = NULL; newnp->mcast_oif = inet6_iif(skb); @@ -1447,6 +1449,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, First: no IPv4 options. */ newinet->inet_opt = NULL; + newnp->ipv6_ac_list = NULL; newnp->ipv6_fl_list = NULL; /* Clone RX bits */ -- cgit v0.10.2 From 95de86cf5162f78bc5aea80d1a9e5a248196ffaf Mon Sep 17 00:00:00 2001 From: Brian King Date: Wed, 28 Sep 2011 05:33:43 +0000 Subject: ibmveth: Fix oops on request_irq failure If request_irq fails, the ibmveth driver will overwrite the rc and end up returning a successful rc on its open function, resulting in an oops later when a packet gets sent and buffers are not allocated due to the failed open. Signed-off-by: Brian King Signed-off-by: David S. Miller diff --git a/drivers/net/ibmveth.c b/drivers/net/ibmveth.c index 8dd5fcc..d393f1e 100644 --- a/drivers/net/ibmveth.c +++ b/drivers/net/ibmveth.c @@ -636,8 +636,8 @@ static int ibmveth_open(struct net_device *netdev) netdev_err(netdev, "unable to request irq 0x%x, rc %d\n", netdev->irq, rc); do { - rc = h_free_logical_lan(adapter->vdev->unit_address); - } while (H_IS_LONG_BUSY(rc) || (rc == H_BUSY)); + lpar_rc = h_free_logical_lan(adapter->vdev->unit_address); + } while (H_IS_LONG_BUSY(lpar_rc) || (lpar_rc == H_BUSY)); goto err_out; } -- cgit v0.10.2 From 605b91c8f619047ed6d88e24c531d4bdddad46f4 Mon Sep 17 00:00:00 2001 From: "Roy.Li" Date: Wed, 28 Sep 2011 19:51:54 +0000 Subject: net: Documentation: Fix type of variables Signed-off-by: Roy.Li Signed-off-by: David S. Miller diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt index 8154699..ca5cdcd 100644 --- a/Documentation/networking/ip-sysctl.txt +++ b/Documentation/networking/ip-sysctl.txt @@ -1042,7 +1042,7 @@ conf/interface/*: The functional behaviour for certain settings is different depending on whether local forwarding is enabled or not. -accept_ra - BOOLEAN +accept_ra - INTEGER Accept Router Advertisements; autoconfigure using them. Possible values are: @@ -1106,7 +1106,7 @@ dad_transmits - INTEGER The amount of Duplicate Address Detection probes to send. Default: 1 -forwarding - BOOLEAN +forwarding - INTEGER Configure interface-specific Host/Router behaviour. Note: It is recommended to have the same setting on all -- cgit v0.10.2 From 85a64889492b45f931ddac87ec09d84aa7347ee1 Mon Sep 17 00:00:00 2001 From: Jonathan Lallinger Date: Thu, 29 Sep 2011 07:58:41 +0000 Subject: RDSRDMA: Fix cleanup of rds_iw_mr_pool In the rds_iw_mr_pool struct the free_pinned field keeps track of memory pinned by free MRs. While this field is incremented properly upon allocation, it is never decremented upon unmapping. This would cause the rds_rdma module to crash the kernel upon unloading, by triggering the BUG_ON in the rds_iw_destroy_mr_pool function. This change keeps track of the MRs that become unpinned, so that free_pinned can be decremented appropriately. Signed-off-by: Jonathan Lallinger Signed-off-by: Steve Wise Signed-off-by: David S. Miller diff --git a/net/rds/iw_rdma.c b/net/rds/iw_rdma.c index 8b77edb..4e1de17 100644 --- a/net/rds/iw_rdma.c +++ b/net/rds/iw_rdma.c @@ -84,7 +84,8 @@ static int rds_iw_map_fastreg(struct rds_iw_mr_pool *pool, static void rds_iw_free_fastreg(struct rds_iw_mr_pool *pool, struct rds_iw_mr *ibmr); static unsigned int rds_iw_unmap_fastreg_list(struct rds_iw_mr_pool *pool, struct list_head *unmap_list, - struct list_head *kill_list); + struct list_head *kill_list, + int *unpinned); static void rds_iw_destroy_fastreg(struct rds_iw_mr_pool *pool, struct rds_iw_mr *ibmr); static int rds_iw_get_device(struct rds_sock *rs, struct rds_iw_device **rds_iwdev, struct rdma_cm_id **cm_id) @@ -499,7 +500,7 @@ static int rds_iw_flush_mr_pool(struct rds_iw_mr_pool *pool, int free_all) LIST_HEAD(unmap_list); LIST_HEAD(kill_list); unsigned long flags; - unsigned int nfreed = 0, ncleaned = 0, free_goal; + unsigned int nfreed = 0, ncleaned = 0, unpinned = 0, free_goal; int ret = 0; rds_iw_stats_inc(s_iw_rdma_mr_pool_flush); @@ -524,7 +525,8 @@ static int rds_iw_flush_mr_pool(struct rds_iw_mr_pool *pool, int free_all) * will be destroyed by the unmap function. */ if (!list_empty(&unmap_list)) { - ncleaned = rds_iw_unmap_fastreg_list(pool, &unmap_list, &kill_list); + ncleaned = rds_iw_unmap_fastreg_list(pool, &unmap_list, + &kill_list, &unpinned); /* If we've been asked to destroy all MRs, move those * that were simply cleaned to the kill list */ if (free_all) @@ -548,6 +550,7 @@ static int rds_iw_flush_mr_pool(struct rds_iw_mr_pool *pool, int free_all) spin_unlock_irqrestore(&pool->list_lock, flags); } + atomic_sub(unpinned, &pool->free_pinned); atomic_sub(ncleaned, &pool->dirty_count); atomic_sub(nfreed, &pool->item_count); @@ -828,7 +831,8 @@ static void rds_iw_free_fastreg(struct rds_iw_mr_pool *pool, static unsigned int rds_iw_unmap_fastreg_list(struct rds_iw_mr_pool *pool, struct list_head *unmap_list, - struct list_head *kill_list) + struct list_head *kill_list, + int *unpinned) { struct rds_iw_mapping *mapping, *next; unsigned int ncleaned = 0; @@ -855,6 +859,7 @@ static unsigned int rds_iw_unmap_fastreg_list(struct rds_iw_mr_pool *pool, spin_lock_irqsave(&pool->list_lock, flags); list_for_each_entry_safe(mapping, next, unmap_list, m_list) { + *unpinned += mapping->m_sg.len; list_move(&mapping->m_list, &laundered); ncleaned++; } -- cgit v0.10.2 From f4142cba4e4065a416e78ade905bea29ff3930e6 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 29 Sep 2011 12:18:59 -0700 Subject: sparc64: Force the execute bit in OpenFirmware's translation entries. In the OF 'translations' property, the template TTEs in the mappings never specify the executable bit. This is the case even though some of these mappings are for OF's code segment. Therefore, we need to force the execute bit on in every mapping. This problem can only really trigger on Niagara/sun4v machines and the history behind this is a little complicated. Previous to sun4v, the sun4u TTE entries lacked a hardware execute permission bit. So OF didn't have to ever worry about setting anything to handle executable pages. Any valid TTE loaded into the I-TLB would be respected by the chip. But sun4v Niagara chips have a real hardware enforced executable bit in their TTEs. So it has to be set or else the I-TLB throws an instruction access exception with type code 6 (protection violation). We've been extremely fortunate to not get bitten by this in the past. The best I can tell is that the OF's mappings for it's executable code were mapped using permanent locked mappings on sun4v in the past. Therefore, the fact that we didn't have the exec bit set in the OF translations we would use did not matter in practice. Thanks to Greg Onufer for helping me track this down. Signed-off-by: David S. Miller diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c index 581531d..8e073d8 100644 --- a/arch/sparc/mm/init_64.c +++ b/arch/sparc/mm/init_64.c @@ -511,6 +511,11 @@ static void __init read_obp_translations(void) for (i = 0; i < prom_trans_ents; i++) prom_trans[i].data &= ~0x0003fe0000000000UL; } + + /* Force execute bit on. */ + for (i = 0; i < prom_trans_ents; i++) + prom_trans[i].data |= (tlb_type == hypervisor ? + _PAGE_EXEC_4V : _PAGE_EXEC_4U); } static void __init hypervisor_tlb_lock(unsigned long vaddr, -- cgit v0.10.2 From 8e303f20f4b3611615118a22a737fd2dc7c4ef81 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Thu, 29 Sep 2011 17:05:08 +0200 Subject: perf tools: Fix raw sample reading Wrong pointer is being passed for raw data sanity checking, when parsing sample event. This ends up with invalid event and perf record being stuck in __perf_session__process_events function during processing build IDs (process_buildids function). Following command hangs up in my setup: ./perf record -e raw_syscalls:sys_enter ls The fix is to use proper pointer to the raw data instead of the 'u' union. Reviewed-by: David Ahern Cc: David Ahern Cc: Eric Dumazet Cc: Ingo Molnar Cc: Neil Horman Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Steven Rostedt Link: http://lkml.kernel.org/r/1317308709-9474-2-git-send-email-jolsa@redhat.com Signed-off-by: Jiri Olsa Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index c5748c5..e389815 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -449,6 +449,8 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, } if (type & PERF_SAMPLE_RAW) { + const u64 *pdata; + u.val64 = *array; if (WARN_ONCE(swapped, "Endianness of raw data not corrected!\n")) { @@ -462,11 +464,12 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, return -EFAULT; data->raw_size = u.val32[0]; + pdata = (void *) array + sizeof(u32); - if (sample_overlap(event, &u.val32[1], data->raw_size)) + if (sample_overlap(event, pdata, data->raw_size)) return -EFAULT; - data->raw_data = &u.val32[1]; + data->raw_data = (void *) pdata; } return 0; -- cgit v0.10.2 From 12d0d0d3a7349daa95dbfd5d7df8146255bc7c67 Mon Sep 17 00:00:00 2001 From: Oliver Hartkopp Date: Thu, 29 Sep 2011 15:33:47 -0400 Subject: can bcm: fix incomplete tx_setup fix The commit aabdcb0b553b9c9547b1a506b34d55a764745870 ("can bcm: fix tx_setup off-by-one errors") fixed only a part of the original problem reported by Andre Naujoks. It turned out that the original code needed to be re-ordered to reduce complexity and to finally fix the reported frame counting issues. Signed-off-by: Oliver Hartkopp Signed-off-by: David S. Miller diff --git a/net/can/bcm.c b/net/can/bcm.c index c9cdb8d..c84963d 100644 --- a/net/can/bcm.c +++ b/net/can/bcm.c @@ -344,6 +344,18 @@ static void bcm_send_to_user(struct bcm_op *op, struct bcm_msg_head *head, } } +static void bcm_tx_start_timer(struct bcm_op *op) +{ + if (op->kt_ival1.tv64 && op->count) + hrtimer_start(&op->timer, + ktime_add(ktime_get(), op->kt_ival1), + HRTIMER_MODE_ABS); + else if (op->kt_ival2.tv64) + hrtimer_start(&op->timer, + ktime_add(ktime_get(), op->kt_ival2), + HRTIMER_MODE_ABS); +} + static void bcm_tx_timeout_tsklet(unsigned long data) { struct bcm_op *op = (struct bcm_op *)data; @@ -365,23 +377,12 @@ static void bcm_tx_timeout_tsklet(unsigned long data) bcm_send_to_user(op, &msg_head, NULL, 0); } - - /* send (next) frame */ bcm_can_tx(op); - hrtimer_start(&op->timer, - ktime_add(ktime_get(), op->kt_ival1), - HRTIMER_MODE_ABS); - } else { - if (op->kt_ival2.tv64) { + } else if (op->kt_ival2.tv64) + bcm_can_tx(op); - /* send (next) frame */ - bcm_can_tx(op); - hrtimer_start(&op->timer, - ktime_add(ktime_get(), op->kt_ival2), - HRTIMER_MODE_ABS); - } - } + bcm_tx_start_timer(op); } /* @@ -961,28 +962,21 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, hrtimer_cancel(&op->timer); } - if ((op->flags & STARTTIMER) && - ((op->kt_ival1.tv64 && op->count) || op->kt_ival2.tv64)) { - + if (op->flags & STARTTIMER) { + hrtimer_cancel(&op->timer); /* spec: send can_frame when starting timer */ op->flags |= TX_ANNOUNCE; - - /* only start timer when having more frames than sent below */ - if (op->kt_ival1.tv64 && (op->count > 1)) { - /* op->count-- is done in bcm_tx_timeout_tsklet */ - hrtimer_start(&op->timer, op->kt_ival1, - HRTIMER_MODE_REL); - } else - hrtimer_start(&op->timer, op->kt_ival2, - HRTIMER_MODE_REL); } if (op->flags & TX_ANNOUNCE) { bcm_can_tx(op); - if (op->kt_ival1.tv64 && (op->count > 0)) + if (op->count) op->count--; } + if (op->flags & STARTTIMER) + bcm_tx_start_timer(op); + return msg_head->nframes * CFSIZ + MHSIZ; } -- cgit v0.10.2 From b0f84374b6ab0dc9c47975df0b02d46165d558d4 Mon Sep 17 00:00:00 2001 From: wangyanqing Date: Thu, 29 Sep 2011 15:09:40 +0800 Subject: bootup: move 'usermodehelper_enable()' a little earlier Commit d5767c53535a ("bootup: move 'usermodehelper_enable()' to the end of do_basic_setup()") moved 'usermodehelper_enable()' to end of do_basic_setup() to after the initcalls. But then I get failed to let uvesafb work on my computer, and lose the splash boot. So maybe we could start usermodehelper_enable a little early to make some task work that need eary init with the help of user mode. [ I would *really* prefer that initcalls not call into user space - even the real 'init' hasn't been execve'd yet, after all! But for uvesafb it really does look like we don't have much choice. I considered doing this when we mount the root filesystem, but depending on config options that is in multiple places. We could do the usermode helper enable as a rootfs_initcall().. So I'm just using wang yanqing's trivial patch. It's not wonderful, but it's simple and should work. We should revisit this some day, though. - Linus ] Signed-off-by: Linus Torvalds diff --git a/init/main.c b/init/main.c index 23702bb..03b408d 100644 --- a/init/main.c +++ b/init/main.c @@ -730,8 +730,8 @@ static void __init do_basic_setup(void) driver_init(); init_irq_proc(); do_ctors(); - do_initcalls(); usermodehelper_enable(); + do_initcalls(); } static void __init do_pre_smp_initcalls(void) -- cgit v0.10.2 From 16fa42affd484bb500533a7e78e0c72687eddd58 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 29 Sep 2011 15:57:01 +1000 Subject: powerpc: Fix device-tree matching for Apple U4 bridge Apple Quad G5 has some oddity in it's device-tree which causes the new generic matching code to fail to relate nodes for PCI-E devices below U4 with their respective struct pci_dev. This breaks graphics on those machines among others. This fixes it using a quirk which copies the node pointer from the host bridge for the root complex, which makes the generic code work for the children afterward. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Linus Torvalds diff --git a/arch/powerpc/platforms/powermac/pci.c b/arch/powerpc/platforms/powermac/pci.c index 5cc8385..31a7d3a 100644 --- a/arch/powerpc/platforms/powermac/pci.c +++ b/arch/powerpc/platforms/powermac/pci.c @@ -561,6 +561,20 @@ static struct pci_ops u4_pcie_pci_ops = .write = u4_pcie_write_config, }; +static void __devinit pmac_pci_fixup_u4_of_node(struct pci_dev *dev) +{ + /* Apple's device-tree "hides" the root complex virtual P2P bridge + * on U4. However, Linux sees it, causing the PCI <-> OF matching + * code to fail to properly match devices below it. This works around + * it by setting the node of the bridge to point to the PHB node, + * which is not entirely correct but fixes the matching code and + * doesn't break anything else. It's also the simplest possible fix. + */ + if (dev->dev.of_node == NULL) + dev->dev.of_node = pcibios_get_phb_of_node(dev->bus); +} +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_APPLE, 0x5b, pmac_pci_fixup_u4_of_node); + #endif /* CONFIG_PPC64 */ #ifdef CONFIG_PPC32 -- cgit v0.10.2 From 47ea91b4052d9e94b9dca5d7a3d947fbebd07ba9 Mon Sep 17 00:00:00 2001 From: Ram Pai Date: Thu, 22 Sep 2011 15:48:58 +0800 Subject: Resource: fix wrong resource window calculation __find_resource() incorrectly returns a resource window which overlaps an existing allocated window. This happens when the parent's resource-window spans 0x00000000 to 0xffffffff and is entirely allocated to all its children resource-windows. __find_resource() looks for gaps in resource allocation among the children resource windows. When it encounters the last child window it blindly tries the range next to one allocated to the last child. Since the last child's window ends at 0xffffffff the calculation overflows, leading the algorithm to believe that any window in the range 0x0000000 to 0xfffffff is available for allocation. This leads to a conflicting window allocation. Michal Ludvig reported this issue seen on his platform. The following patch fixes the problem and has been verified by Michal. I believe this bug has been there for ages. It got exposed by git commit 2bbc6942273b ("PCI : ability to relocate assigned pci-resources") Signed-off-by: Ram Pai Tested-by: Michal Ludvig Signed-off-by: Linus Torvalds diff --git a/kernel/resource.c b/kernel/resource.c index 3b3cedc..c8dc249 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -419,6 +419,9 @@ static int __find_resource(struct resource *root, struct resource *old, else tmp.end = root->end; + if (tmp.end < tmp.start) + goto next; + resource_clip(&tmp, constraint->min, constraint->max); arch_remove_reservations(&tmp); @@ -436,8 +439,10 @@ static int __find_resource(struct resource *root, struct resource *old, return 0; } } - if (!this) + +next: if (!this || this->end == root->end) break; + if (this != old) tmp.start = this->end + 1; this = this->sibling; -- cgit v0.10.2 From 798cb7e897210dbf827f69a302c0967e0380ebac Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 30 Sep 2011 08:52:26 +0200 Subject: ALSA: hda - Fix a regression of the position-buffer check The commit a810364a0424c297242c6c66071a42f7675a5568 ALSA: hda - Handle -1 as invalid position, too caused a regression on some machines that require the position-buffer instead of LPIB, e.g. resulting in noises with mic recording with PulseAudio. This patch fixes the detection by delaying the test at the timing as same as 3.0, i.e. doing the position check only when requested in azx_position_ok(). Reported-and-tested-by: Rocko Requin Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index be69822..e9a2a87 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -1924,7 +1924,8 @@ static unsigned int azx_via_get_position(struct azx *chip, } static unsigned int azx_get_position(struct azx *chip, - struct azx_dev *azx_dev) + struct azx_dev *azx_dev, + bool with_check) { unsigned int pos; int stream = azx_dev->substream->stream; @@ -1940,7 +1941,7 @@ static unsigned int azx_get_position(struct azx *chip, default: /* use the position buffer */ pos = le32_to_cpu(*azx_dev->posbuf); - if (chip->position_fix[stream] == POS_FIX_AUTO) { + if (with_check && chip->position_fix[stream] == POS_FIX_AUTO) { if (!pos || pos == (u32)-1) { printk(KERN_WARNING "hda-intel: Invalid position buffer, " @@ -1964,7 +1965,7 @@ static snd_pcm_uframes_t azx_pcm_pointer(struct snd_pcm_substream *substream) struct azx *chip = apcm->chip; struct azx_dev *azx_dev = get_azx_dev(substream); return bytes_to_frames(substream->runtime, - azx_get_position(chip, azx_dev)); + azx_get_position(chip, azx_dev, false)); } /* @@ -1987,7 +1988,7 @@ static int azx_position_ok(struct azx *chip, struct azx_dev *azx_dev) return -1; /* bogus (too early) interrupt */ stream = azx_dev->substream->stream; - pos = azx_get_position(chip, azx_dev); + pos = azx_get_position(chip, azx_dev, true); if (WARN_ONCE(!azx_dev->period_bytes, "hda-intel: zero azx_dev->period_bytes")) -- cgit v0.10.2 From d670ec13178d0fd8680e6742a2bc6e04f28f87d8 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 1 Sep 2011 12:42:04 +0200 Subject: posix-cpu-timers: Cure SMP wobbles David reported: Attached below is a watered-down version of rt/tst-cpuclock2.c from GLIBC. Just build it with "gcc -o test test.c -lpthread -lrt" or similar. Run it several times, and you will see cases where the main thread will measure a process clock difference before and after the nanosleep which is smaller than the cpu-burner thread's individual thread clock difference. This doesn't make any sense since the cpu-burner thread is part of the top-level process's thread group. I've reproduced this on both x86-64 and sparc64 (using both 32-bit and 64-bit binaries). For example: [davem@boricha build-x86_64-linux]$ ./test process: before(0.001221967) after(0.498624371) diff(497402404) thread: before(0.000081692) after(0.498316431) diff(498234739) self: before(0.001223521) after(0.001240219) diff(16698) [davem@boricha build-x86_64-linux]$ The diff of 'process' should always be >= the diff of 'thread'. I make sure to wrap the 'thread' clock measurements the most tightly around the nanosleep() call, and that the 'process' clock measurements are the outer-most ones. --- #include #include #include #include #include #include #include #include static pthread_barrier_t barrier; static void *chew_cpu(void *arg) { pthread_barrier_wait(&barrier); while (1) __asm__ __volatile__("" : : : "memory"); return NULL; } int main(void) { clockid_t process_clock, my_thread_clock, th_clock; struct timespec process_before, process_after; struct timespec me_before, me_after; struct timespec th_before, th_after; struct timespec sleeptime; unsigned long diff; pthread_t th; int err; err = clock_getcpuclockid(0, &process_clock); if (err) return 1; err = pthread_getcpuclockid(pthread_self(), &my_thread_clock); if (err) return 1; pthread_barrier_init(&barrier, NULL, 2); err = pthread_create(&th, NULL, chew_cpu, NULL); if (err) return 1; err = pthread_getcpuclockid(th, &th_clock); if (err) return 1; pthread_barrier_wait(&barrier); err = clock_gettime(process_clock, &process_before); if (err) return 1; err = clock_gettime(my_thread_clock, &me_before); if (err) return 1; err = clock_gettime(th_clock, &th_before); if (err) return 1; sleeptime.tv_sec = 0; sleeptime.tv_nsec = 500000000; nanosleep(&sleeptime, NULL); err = clock_gettime(th_clock, &th_after); if (err) return 1; err = clock_gettime(my_thread_clock, &me_after); if (err) return 1; err = clock_gettime(process_clock, &process_after); if (err) return 1; diff = process_after.tv_nsec - process_before.tv_nsec; printf("process: before(%lu.%.9lu) after(%lu.%.9lu) diff(%lu)\n", process_before.tv_sec, process_before.tv_nsec, process_after.tv_sec, process_after.tv_nsec, diff); diff = th_after.tv_nsec - th_before.tv_nsec; printf("thread: before(%lu.%.9lu) after(%lu.%.9lu) diff(%lu)\n", th_before.tv_sec, th_before.tv_nsec, th_after.tv_sec, th_after.tv_nsec, diff); diff = me_after.tv_nsec - me_before.tv_nsec; printf("self: before(%lu.%.9lu) after(%lu.%.9lu) diff(%lu)\n", me_before.tv_sec, me_before.tv_nsec, me_after.tv_sec, me_after.tv_nsec, diff); return 0; } This is due to us using p->se.sum_exec_runtime in thread_group_cputime() where we iterate the thread group and sum all data. This does not take time since the last schedule operation (tick or otherwise) into account. We can cure this by using task_sched_runtime() at the cost of having to take locks. This also means we can (and must) do away with thread_group_sched_runtime() since the modified thread_group_cputime() is now more accurate and would deadlock when called from thread_group_sched_runtime(). Aside of that it makes the function safe on 32 bit systems. The old code added t->se.sum_exec_runtime unprotected. sum_exec_runtime is a 64bit value and could be changed on another cpu at the same time. Reported-by: David Miller Signed-off-by: Peter Zijlstra Cc: stable@kernel.org Link: http://lkml.kernel.org/r/1314874459.7945.22.camel@twins Tested-by: David Miller Signed-off-by: Thomas Gleixner diff --git a/include/linux/sched.h b/include/linux/sched.h index 4ac2c05..41d0237 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1956,7 +1956,6 @@ static inline void disable_sched_clock_irqtime(void) {} extern unsigned long long task_sched_runtime(struct task_struct *task); -extern unsigned long long thread_group_sched_runtime(struct task_struct *task); /* sched_exec is called by processes performing an exec */ #ifdef CONFIG_SMP diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 58f405b..c8008dd 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -250,7 +250,7 @@ void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times) do { times->utime = cputime_add(times->utime, t->utime); times->stime = cputime_add(times->stime, t->stime); - times->sum_exec_runtime += t->se.sum_exec_runtime; + times->sum_exec_runtime += task_sched_runtime(t); } while_each_thread(tsk, t); out: rcu_read_unlock(); @@ -312,7 +312,8 @@ static int cpu_clock_sample_group(const clockid_t which_clock, cpu->cpu = cputime.utime; break; case CPUCLOCK_SCHED: - cpu->sched = thread_group_sched_runtime(p); + thread_group_cputime(p, &cputime); + cpu->sched = cputime.sum_exec_runtime; break; } return 0; diff --git a/kernel/sched.c b/kernel/sched.c index d249ea8..b50b0f0 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -3725,30 +3725,6 @@ unsigned long long task_sched_runtime(struct task_struct *p) } /* - * Return sum_exec_runtime for the thread group. - * In case the task is currently running, return the sum plus current's - * pending runtime that have not been accounted yet. - * - * Note that the thread group might have other running tasks as well, - * so the return value not includes other pending runtime that other - * running tasks might have. - */ -unsigned long long thread_group_sched_runtime(struct task_struct *p) -{ - struct task_cputime totals; - unsigned long flags; - struct rq *rq; - u64 ns; - - rq = task_rq_lock(p, &flags); - thread_group_cputime(p, &totals); - ns = totals.sum_exec_runtime + do_task_delta_exec(p, rq); - task_rq_unlock(rq, p, &flags); - - return ns; -} - -/* * Account user cpu time to a process. * @p: the process that the cpu time gets accounted to * @cputime: the cpu time spent in user space since the last update -- cgit v0.10.2 From b6316429af7f365f307dfd2b6a7a42f2563aef19 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 30 Sep 2011 15:23:54 -0400 Subject: Btrfs: force a page fault if we have a shorty copy on a page boundary A user reported a problem where ceph was getting into 100% cpu usage while doing some writing. It turns out it's because we were doing a short write on a not uptodate page, which means we'd fall back at one page at a time and fault the page in. The problem is our position is on the page boundary, so our fault in logic wasn't actually reading the page, so we'd just spin forever or until the page got read in by somebody else. This will force a readpage if we end up doing a short copy. Alexandre could reproduce this easily with ceph and reports it fixes his problem. I also wrote a reproducer that no longer hangs my box with this patch. Thanks, Reported-and-tested-by: Alexandre Oliva Signed-off-by: Josef Bacik Signed-off-by: Chris Mason diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 98d95bb..e730510 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1036,11 +1036,13 @@ out: * on error we return an unlocked page and the error value * on success we return a locked page and 0 */ -static int prepare_uptodate_page(struct page *page, u64 pos) +static int prepare_uptodate_page(struct page *page, u64 pos, + bool force_uptodate) { int ret = 0; - if ((pos & (PAGE_CACHE_SIZE - 1)) && !PageUptodate(page)) { + if (((pos & (PAGE_CACHE_SIZE - 1)) || force_uptodate) && + !PageUptodate(page)) { ret = btrfs_readpage(NULL, page); if (ret) return ret; @@ -1061,7 +1063,7 @@ static int prepare_uptodate_page(struct page *page, u64 pos) static noinline int prepare_pages(struct btrfs_root *root, struct file *file, struct page **pages, size_t num_pages, loff_t pos, unsigned long first_index, - size_t write_bytes) + size_t write_bytes, bool force_uptodate) { struct extent_state *cached_state = NULL; int i; @@ -1086,10 +1088,11 @@ again: } if (i == 0) - err = prepare_uptodate_page(pages[i], pos); + err = prepare_uptodate_page(pages[i], pos, + force_uptodate); if (i == num_pages - 1) err = prepare_uptodate_page(pages[i], - pos + write_bytes); + pos + write_bytes, false); if (err) { page_cache_release(pages[i]); faili = i - 1; @@ -1158,6 +1161,7 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, size_t num_written = 0; int nrptrs; int ret = 0; + bool force_page_uptodate = false; nrptrs = min((iov_iter_count(i) + PAGE_CACHE_SIZE - 1) / PAGE_CACHE_SIZE, PAGE_CACHE_SIZE / @@ -1200,7 +1204,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, * contents of pages from loop to loop */ ret = prepare_pages(root, file, pages, num_pages, - pos, first_index, write_bytes); + pos, first_index, write_bytes, + force_page_uptodate); if (ret) { btrfs_delalloc_release_space(inode, num_pages << PAGE_CACHE_SHIFT); @@ -1217,12 +1222,15 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, if (copied < write_bytes) nrptrs = 1; - if (copied == 0) + if (copied == 0) { + force_page_uptodate = true; dirty_pages = 0; - else + } else { + force_page_uptodate = false; dirty_pages = (copied + offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + } /* * If we had a short copy we need to release the excess delaloc -- cgit v0.10.2 From 9bfacd01dc9b7519e1e6da12b01963550b9d09a2 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Thu, 22 Sep 2011 00:06:05 -0700 Subject: [SCSI] qla2xxx: Fix crash in qla2x00_abort_all_cmds() on unload I hit a crash in qla2x00_abort_all_cmds() if the qla2xxx module is unloaded right after it is loaded. I debugged this down to the abort handling improperly treating a command of type SRB_ADISC_CMD as if it had a bsg_job to complete when that command actually uses the iocb_cmd part of the union. (I guess to hit this one has to unload the module while the async FC initialization is still in progress) It seems we should only look for a bsg_job if type is SRB_ELS_CMD_RPT, SRB_ELS_CMD_HST or SRB_CT_CMD, so switch the test to make that explicit. Signed-off-by: Roland Dreier Acked-by: Chad Dupuis Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c index 4cace3f..1e69527 100644 --- a/drivers/scsi/qla2xxx/qla_os.c +++ b/drivers/scsi/qla2xxx/qla_os.c @@ -1328,10 +1328,9 @@ qla2x00_abort_all_cmds(scsi_qla_host_t *vha, int res) qla2x00_sp_compl(ha, sp); } else { ctx = sp->ctx; - if (ctx->type == SRB_LOGIN_CMD || - ctx->type == SRB_LOGOUT_CMD) { - ctx->u.iocb_cmd->free(sp); - } else { + if (ctx->type == SRB_ELS_CMD_RPT || + ctx->type == SRB_ELS_CMD_HST || + ctx->type == SRB_CT_CMD) { struct fc_bsg_job *bsg_job = ctx->u.bsg_job; if (bsg_job->request->msgcode @@ -1343,6 +1342,8 @@ qla2x00_abort_all_cmds(scsi_qla_host_t *vha, int res) kfree(sp->ctx); mempool_free(sp, ha->srb_mempool); + } else { + ctx->u.iocb_cmd->free(sp); } } } -- cgit v0.10.2 From eff919ac0fc7565e71ffa35657c333dd8cdc0520 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sat, 1 Oct 2011 22:03:34 +0200 Subject: ASoC: use a valid device for dev_err() in Zylonite A recent conversion has introduced references to &pdev->dev, which does not actually exist in all the contexts it's used in. Replace this with card->dev where necessary, in order to let the driver build again. Signed-off-by: Arnd Bergmann Signed-off-by: Mark Brown Cc: stable@kernel.org diff --git a/sound/soc/pxa/zylonite.c b/sound/soc/pxa/zylonite.c index b644575..2b8350b 100644 --- a/sound/soc/pxa/zylonite.c +++ b/sound/soc/pxa/zylonite.c @@ -196,20 +196,20 @@ static int zylonite_probe(struct snd_soc_card *card) if (clk_pout) { pout = clk_get(NULL, "CLK_POUT"); if (IS_ERR(pout)) { - dev_err(&pdev->dev, "Unable to obtain CLK_POUT: %ld\n", + dev_err(card->dev, "Unable to obtain CLK_POUT: %ld\n", PTR_ERR(pout)); return PTR_ERR(pout); } ret = clk_enable(pout); if (ret != 0) { - dev_err(&pdev->dev, "Unable to enable CLK_POUT: %d\n", + dev_err(card->dev, "Unable to enable CLK_POUT: %d\n", ret); clk_put(pout); return ret; } - dev_dbg(&pdev->dev, "MCLK enabled at %luHz\n", + dev_dbg(card->dev, "MCLK enabled at %luHz\n", clk_get_rate(pout)); } @@ -241,7 +241,7 @@ static int zylonite_resume_pre(struct snd_soc_card *card) if (clk_pout) { ret = clk_enable(pout); if (ret != 0) - dev_err(&pdev->dev, "Unable to enable CLK_POUT: %d\n", + dev_err(card->dev, "Unable to enable CLK_POUT: %d\n", ret); } -- cgit v0.10.2 From a73914c35b05d80f8ce78288e10056c91090b666 Mon Sep 17 00:00:00 2001 From: Mark Salyzyn Date: Thu, 22 Sep 2011 08:32:23 -0700 Subject: [SCSI] libsas: fix panic when single phy is disabled on a wide port When a wide port is being utilized to a target, if one disables only one of the phys, we get an OS crash: BUG: unable to handle kernel NULL pointer dereference at 0000000000000238 IP: [] mutex_lock+0x21/0x50 PGD 4103f5067 PUD 41dba9067 PMD 0 Oops: 0002 [#1] SMP last sysfs file: /sys/bus/pci/slots/5/address CPU 0 Modules linked in: pm8001(U) ses enclosure fuse nfsd exportfs autofs4 ipmi_devintf ipmi_si ipmi_msghandler nfs lockd fscache nfs_acl auth_rpcgss 8021q fcoe libfcoe garp libfc scsi_transport_fc stp scsi_tgt llc sunrpc cpufreq_ondemand acpi_cpufreq freq_table ipv6 sr_mod cdrom dm_mirror dm_region_hash dm_log uinput sg i2c_i801 i2c_core iTCO_wdt iTCO_vendor_support e1000e mlx4_ib ib_mad ib_core mlx4_en mlx4_core ext3 jbd mbcache sd_mod crc_t10dif usb_storage ata_generic pata_acpi ata_piix libsas(U) scsi_transport_sas dm_mod [last unloaded: pm8001] Modules linked in: pm8001(U) ses enclosure fuse nfsd exportfs autofs4 ipmi_devintf ipmi_si ipmi_msghandler nfs lockd fscache nfs_acl auth_rpcgss 8021q fcoe libfcoe garp libfc scsi_transport_fc stp scsi_tgt llc sunrpc cpufreq_ondemand acpi_cpufreq freq_table ipv6 sr_mod cdrom dm_mirror dm_region_hash dm_log uinput sg i2c_i801 i2c_core iTCO_wdt iTCO_vendor_support e1000e mlx4_ib ib_mad ib_core mlx4_en mlx4_core ext3 jbd mbcache sd_mod crc_t10dif usb_storage ata_generic pata_acpi ata_piix libsas(U) scsi_transport_sas dm_mod [last unloaded: pm8001] Pid: 5146, comm: scsi_wq_5 Not tainted 2.6.32-71.29.1.el6.lustre.7.x86_64 #1 Storage Server RIP: 0010:[] [] mutex_lock+0x21/0x50 RSP: 0018:ffff8803e4e33d30 EFLAGS: 00010246 RAX: 0000000000000000 RBX: 0000000000000238 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff8803e664c800 RDI: 0000000000000238 RBP: ffff8803e4e33d40 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000000 R13: 0000000000000238 R14: ffff88041acb7200 R15: ffff88041c51ada0 FS: 0000000000000000(0000) GS:ffff880028200000(0000) knlGS:0000000000000000 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 0000000000000238 CR3: 0000000410143000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process scsi_wq_5 (pid: 5146, threadinfo ffff8803e4e32000, task ffff8803e4e294a0) Stack: ffff8803e664c800 0000000000000000 ffff8803e4e33d70 ffffffffa001f06e <0> ffff8803e4e33d60 ffff88041c51ada0 ffff88041acb7200 ffff88041bc0aa00 <0> ffff8803e4e33d90 ffffffffa0032b6c 0000000000000014 ffff88041acb7200 Call Trace: [] sas_port_delete_phy+0x2e/0xa0 [scsi_transport_sas] [] sas_unregister_devs_sas_addr+0xac/0xe0 [libsas] [] sas_ex_revalidate_domain+0x204/0x330 [libsas] [] ? sas_revalidate_domain+0x0/0x90 [libsas] [] sas_revalidate_domain+0x65/0x90 [libsas] [] worker_thread+0x170/0x2a0 [] ? autoremove_wake_function+0x0/0x40 [] ? worker_thread+0x0/0x2a0 [] kthread+0x96/0xa0 [] child_rip+0xa/0x20 [] ? kthread+0x0/0xa0 [] ? child_rip+0x0/0x20 Code: ff ff 85 c0 75 ed eb d6 66 90 55 48 89 e5 48 83 ec 10 48 89 1c 24 4c 89 64 24 08 0f 1f 44 00 00 48 89 fb e8 92 f4 ff ff 48 89 df ff 0f 79 05 e8 25 00 00 00 65 48 8b 04 25 08 cc 00 00 48 2d RIP [] mutex_lock+0x21/0x50 RSP CR2: 0000000000000238 The following patch is admittedly a band-aid, and does not solve the root cause, but it still is a good candidate for hardening as a pointer check before reference. Signed-off-by: Mark Salyzyn Tested-by: Jack Wang Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c index c9e3dc0..16ad97d 100644 --- a/drivers/scsi/libsas/sas_expander.c +++ b/drivers/scsi/libsas/sas_expander.c @@ -1769,10 +1769,12 @@ static void sas_unregister_devs_sas_addr(struct domain_device *parent, sas_disable_routing(parent, phy->attached_sas_addr); } memset(phy->attached_sas_addr, 0, SAS_ADDR_SIZE); - sas_port_delete_phy(phy->port, phy->phy); - if (phy->port->num_phys == 0) - sas_port_delete(phy->port); - phy->port = NULL; + if (phy->port) { + sas_port_delete_phy(phy->port, phy->phy); + if (phy->port->num_phys == 0) + sas_port_delete(phy->port); + phy->port = NULL; + } } static int sas_discover_bfs_by_root_level(struct domain_device *root, -- cgit v0.10.2 From 21d17dd2a377ba894f26989915eb3c6e427a3656 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 2 Oct 2011 20:41:04 +0800 Subject: ASoC: Fix setting update bits for WM8753_LADC and WM8753_RADC Current code set update bits for WM8753_LDAC and WM8753_RDAC twice, but missed setting update bits for WM8753_LADC and WM8753_RADC. I think it is a copy-paste bug in commit 776065 "ASoC: codecs: wm8753: Fix register cache incoherency". Signed-off-by: Axel Lin Signed-off-by: Mark Brown Cc: stable@kernel.org diff --git a/sound/soc/codecs/wm8753.c b/sound/soc/codecs/wm8753.c index ffa2ffe..aa091a0 100644 --- a/sound/soc/codecs/wm8753.c +++ b/sound/soc/codecs/wm8753.c @@ -1454,8 +1454,8 @@ static int wm8753_probe(struct snd_soc_codec *codec) /* set the update bits */ snd_soc_update_bits(codec, WM8753_LDAC, 0x0100, 0x0100); snd_soc_update_bits(codec, WM8753_RDAC, 0x0100, 0x0100); - snd_soc_update_bits(codec, WM8753_LDAC, 0x0100, 0x0100); - snd_soc_update_bits(codec, WM8753_RDAC, 0x0100, 0x0100); + snd_soc_update_bits(codec, WM8753_LADC, 0x0100, 0x0100); + snd_soc_update_bits(codec, WM8753_RADC, 0x0100, 0x0100); snd_soc_update_bits(codec, WM8753_LOUT1V, 0x0100, 0x0100); snd_soc_update_bits(codec, WM8753_ROUT1V, 0x0100, 0x0100); snd_soc_update_bits(codec, WM8753_LOUT2V, 0x0100, 0x0100); -- cgit v0.10.2 From b5c49d49b9e175fd56cb4b5cf2c4fd972d15e013 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sun, 2 Oct 2011 16:45:31 +0200 Subject: ASoC: omap_mcpdm_remove cannot be __devexit omap_mcpdm_remove is used from asoc_mcpdm_probe, which is an initcall, and must not be discarded when HOTPLUG is disabled. Signed-off-by: Arnd Bergmann Signed-off-by: Mark Brown diff --git a/sound/soc/omap/mcpdm.c b/sound/soc/omap/mcpdm.c index 928f037..50e5919 100644 --- a/sound/soc/omap/mcpdm.c +++ b/sound/soc/omap/mcpdm.c @@ -449,7 +449,7 @@ exit: return ret; } -int __devexit omap_mcpdm_remove(struct platform_device *pdev) +int omap_mcpdm_remove(struct platform_device *pdev) { struct omap_mcpdm *mcpdm_ptr = platform_get_drvdata(pdev); diff --git a/sound/soc/omap/mcpdm.h b/sound/soc/omap/mcpdm.h index df3e16f..20c20a8 100644 --- a/sound/soc/omap/mcpdm.h +++ b/sound/soc/omap/mcpdm.h @@ -150,4 +150,4 @@ extern int omap_mcpdm_request(void); extern void omap_mcpdm_free(void); extern int omap_mcpdm_set_offset(int offset1, int offset2); int __devinit omap_mcpdm_probe(struct platform_device *pdev); -int __devexit omap_mcpdm_remove(struct platform_device *pdev); +int omap_mcpdm_remove(struct platform_device *pdev); -- cgit v0.10.2 From 109bc10d30f33e84f1d7289f0039e0c858ade82f Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 3 Oct 2011 09:13:45 -0400 Subject: drm/radeon/kms: fix regression in DP aux defer handling An incorrect ordering in the error checking code lead to DP aux defer being skipped in the aux native write path. Move the bytes transferred check (ret == 0) below the defer check. Tracked down by: Brad Campbell Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=41121 Signed-off-by: Alex Deucher Cc: Brad Campbell Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index 7ad43c6..f526fa7 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -158,14 +158,14 @@ static int radeon_dp_aux_native_read(struct radeon_connector *radeon_connector, while (1) { ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus, msg, msg_bytes, recv, recv_bytes, delay, &ack); - if (ret == 0) - return -EPROTO; if (ret < 0) return ret; if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) return ret; else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER) udelay(400); + else if (ret == 0) + return -EPROTO; else return -EIO; } -- cgit v0.10.2 From 6375bda073724ead7df08746866b724b1799a295 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 3 Oct 2011 09:13:46 -0400 Subject: drm/radeon/kms: add retry limits for native DP aux defer The previous code could potentially loop forever. Limit the number of DP aux defer retries to 4 for native aux transactions, same as i2c over aux transactions. Noticed by: Brad Campbell Signed-off-by: Alex Deucher Cc: Brad Campbell Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index f526fa7..4da2388 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -115,6 +115,7 @@ static int radeon_dp_aux_native_write(struct radeon_connector *radeon_connector, u8 msg[20]; int msg_bytes = send_bytes + 4; u8 ack; + unsigned retry; if (send_bytes > 16) return -1; @@ -125,20 +126,20 @@ static int radeon_dp_aux_native_write(struct radeon_connector *radeon_connector, msg[3] = (msg_bytes << 4) | (send_bytes - 1); memcpy(&msg[4], send, send_bytes); - while (1) { + for (retry = 0; retry < 4; retry++) { ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus, msg, msg_bytes, NULL, 0, delay, &ack); if (ret < 0) return ret; if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) - break; + return send_bytes; else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER) udelay(400); else return -EIO; } - return send_bytes; + return -EIO; } static int radeon_dp_aux_native_read(struct radeon_connector *radeon_connector, @@ -149,13 +150,14 @@ static int radeon_dp_aux_native_read(struct radeon_connector *radeon_connector, int msg_bytes = 4; u8 ack; int ret; + unsigned retry; msg[0] = address; msg[1] = address >> 8; msg[2] = AUX_NATIVE_READ << 4; msg[3] = (msg_bytes << 4) | (recv_bytes - 1); - while (1) { + for (retry = 0; retry < 4; retry++) { ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus, msg, msg_bytes, recv, recv_bytes, delay, &ack); if (ret < 0) @@ -169,6 +171,8 @@ static int radeon_dp_aux_native_read(struct radeon_connector *radeon_connector, else return -EIO; } + + return -EIO; } static void radeon_write_dpcd_reg(struct radeon_connector *radeon_connector, -- cgit v0.10.2 From 5ba7ddf81634bfdf32d09261d2959e3f5b7c4263 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 3 Oct 2011 08:37:33 -0400 Subject: drm/radeon/kms: Fix logic error in DP HPD handler Only disable the pipe if the monitor is physically disconnected. The previous logic also disabled the pipe if the link was trained. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=41248 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index c4b8741..bce63fd 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -68,11 +68,11 @@ void radeon_connector_hotplug(struct drm_connector *connector) if (connector->connector_type == DRM_MODE_CONNECTOR_DisplayPort) { int saved_dpms = connector->dpms; - if (radeon_hpd_sense(rdev, radeon_connector->hpd.hpd) && - radeon_dp_needs_link_train(radeon_connector)) - drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON); - else + /* Only turn off the display it it's physically disconnected */ + if (!radeon_hpd_sense(rdev, radeon_connector->hpd.hpd)) drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF); + else if (radeon_dp_needs_link_train(radeon_connector)) + drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON); connector->dpms = saved_dpms; } } -- cgit v0.10.2 From b356fe0afe9aa2cfdc7be0bae22ea1c93afc2451 Mon Sep 17 00:00:00 2001 From: Nicholas Miell Date: Thu, 29 Sep 2011 19:07:14 -0700 Subject: drm/radeon/kms: fix cursor image off-by-one error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The mouse cursor hotspot calculation when the cursor is partially off the top or left side of the screen was off by one. Fixes https://bugs.freedesktop.org/show_bug.cgi?id=41158 Signed-off-by: Nicholas Miell Reviewed-by: Michel Dänzer Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c index 3189a7e..c495575 100644 --- a/drivers/gpu/drm/radeon/radeon_cursor.c +++ b/drivers/gpu/drm/radeon/radeon_cursor.c @@ -209,9 +209,9 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, int w = radeon_crtc->cursor_width; if (x < 0) - xorigin = -x + 1; + xorigin = -x; if (y < 0) - yorigin = -y + 1; + yorigin = -y; if (xorigin >= CURSOR_WIDTH) xorigin = CURSOR_WIDTH - 1; if (yorigin >= CURSOR_HEIGHT) -- cgit v0.10.2 From 7d309529b4c1a6281549bf86681cdfb4fbed49ed Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Fri, 30 Sep 2011 17:16:51 +0200 Subject: drm/radeon: Simplify cursor x/yorigin calculation. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Michel Dänzer Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c index c495575..bac8ee7 100644 --- a/drivers/gpu/drm/radeon/radeon_cursor.c +++ b/drivers/gpu/drm/radeon/radeon_cursor.c @@ -209,13 +209,9 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, int w = radeon_crtc->cursor_width; if (x < 0) - xorigin = -x; + xorigin = min(-x, CURSOR_WIDTH - 1); if (y < 0) - yorigin = -y; - if (xorigin >= CURSOR_WIDTH) - xorigin = CURSOR_WIDTH - 1; - if (yorigin >= CURSOR_HEIGHT) - yorigin = CURSOR_HEIGHT - 1; + yorigin = min(-y, CURSOR_HEIGHT - 1); if (ASIC_IS_AVIVO(rdev)) { int i = 0; -- cgit v0.10.2 From b8aee294d89502469f2d80ae6afb93398d8227e0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Fri, 30 Sep 2011 17:16:52 +0200 Subject: drm/radeon: Update AVIVO cursor coordinate origin before x/yorigin calculation. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fixes cursor disappearing prematurely when moving off a top/left edge which is not located at the desktop top/left edge. Signed-off-by: Michel Dänzer Cc: stable@kernel.org Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c index bac8ee7..f1d871d 100644 --- a/drivers/gpu/drm/radeon/radeon_cursor.c +++ b/drivers/gpu/drm/radeon/radeon_cursor.c @@ -208,6 +208,13 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, int xorigin = 0, yorigin = 0; int w = radeon_crtc->cursor_width; + if (ASIC_IS_AVIVO(rdev)) { + /* avivo cursor are offset into the total surface */ + x += crtc->x; + y += crtc->y; + } + DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y); + if (x < 0) xorigin = min(-x, CURSOR_WIDTH - 1); if (y < 0) @@ -217,11 +224,6 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, int i = 0; struct drm_crtc *crtc_p; - /* avivo cursor are offset into the total surface */ - x += crtc->x; - y += crtc->y; - DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y); - /* avivo cursor image can't end on 128 pixel boundary or * go past the end of the frame if both crtcs are enabled */ -- cgit v0.10.2 From 02e6859eae4a8e511fa1a2c9a689ea25cdc6166a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Fri, 30 Sep 2011 17:16:53 +0200 Subject: drm/radeon: Set cursor x/y to 0 when x/yorigin > 0. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Apart from the obvious cleanup, this should make the line cursor_end = x - xorigin + w; correct now. Signed-off-by: Michel Dänzer Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c index f1d871d..fde25c0 100644 --- a/drivers/gpu/drm/radeon/radeon_cursor.c +++ b/drivers/gpu/drm/radeon/radeon_cursor.c @@ -215,10 +215,14 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, } DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y); - if (x < 0) + if (x < 0) { xorigin = min(-x, CURSOR_WIDTH - 1); - if (y < 0) + x = 0; + } + if (y < 0) { yorigin = min(-y, CURSOR_HEIGHT - 1); + y = 0; + } if (ASIC_IS_AVIVO(rdev)) { int i = 0; @@ -251,16 +255,12 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, radeon_lock_cursor(crtc, true); if (ASIC_IS_DCE4(rdev)) { - WREG32(EVERGREEN_CUR_POSITION + radeon_crtc->crtc_offset, - ((xorigin ? 0 : x) << 16) | - (yorigin ? 0 : y)); + WREG32(EVERGREEN_CUR_POSITION + radeon_crtc->crtc_offset, (x << 16) | y); WREG32(EVERGREEN_CUR_HOT_SPOT + radeon_crtc->crtc_offset, (xorigin << 16) | yorigin); WREG32(EVERGREEN_CUR_SIZE + radeon_crtc->crtc_offset, ((w - 1) << 16) | (radeon_crtc->cursor_height - 1)); } else if (ASIC_IS_AVIVO(rdev)) { - WREG32(AVIVO_D1CUR_POSITION + radeon_crtc->crtc_offset, - ((xorigin ? 0 : x) << 16) | - (yorigin ? 0 : y)); + WREG32(AVIVO_D1CUR_POSITION + radeon_crtc->crtc_offset, (x << 16) | y); WREG32(AVIVO_D1CUR_HOT_SPOT + radeon_crtc->crtc_offset, (xorigin << 16) | yorigin); WREG32(AVIVO_D1CUR_SIZE + radeon_crtc->crtc_offset, ((w - 1) << 16) | (radeon_crtc->cursor_height - 1)); @@ -274,8 +274,8 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc, | yorigin)); WREG32(RADEON_CUR_HORZ_VERT_POSN + radeon_crtc->crtc_offset, (RADEON_CUR_LOCK - | ((xorigin ? 0 : x) << 16) - | (yorigin ? 0 : y))); + | (x << 16) + | y)); /* offset is from DISP(2)_BASE_ADDRESS */ WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, (radeon_crtc->legacy_cursor_offset + (yorigin * 256))); -- cgit v0.10.2 From a0db2dad0935e798973bb79676e722b82f177206 Mon Sep 17 00:00:00 2001 From: Andy Gospodarek Date: Fri, 23 Sep 2011 10:53:34 +0000 Subject: bonding: properly stop queuing work when requested During a test where a pair of bonding interfaces using ARP monitoring were both brought up and torn down (with an rmmod) repeatedly, a panic in the timer code was noticed. I tracked this down and determined that any of the bonding functions that ran as workqueue handlers and requeued more work might not properly exit when the module was removed. There was a flag protected by the bond lock called kill_timers that is set when the interface goes down or the module is removed, but many of the functions that monitor link status now unlock the bond lock to take rtnl first. There is a chance that another CPU running the rmmod could get the lock and set kill_timers after the first check has passed. This patch does not allow any function to queue work that will make itself run unless kill_timers is not set. I also noticed while doing this work that bond_resend_igmp_join_requests did not have a check for kill_timers, so I added the needed call there as well. Signed-off-by: Andy Gospodarek Reported-by: Liang Zheng Signed-off-by: David S. Miller diff --git a/drivers/net/bonding/bond_3ad.c b/drivers/net/bonding/bond_3ad.c index a047eb9..47b928e 100644 --- a/drivers/net/bonding/bond_3ad.c +++ b/drivers/net/bonding/bond_3ad.c @@ -2168,7 +2168,8 @@ void bond_3ad_state_machine_handler(struct work_struct *work) } re_arm: - queue_delayed_work(bond->wq, &bond->ad_work, ad_delta_in_ticks); + if (!bond->kill_timers) + queue_delayed_work(bond->wq, &bond->ad_work, ad_delta_in_ticks); out: read_unlock(&bond->lock); } diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index 7f8b20a3..d4fbd2e 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -1440,7 +1440,8 @@ void bond_alb_monitor(struct work_struct *work) } re_arm: - queue_delayed_work(bond->wq, &bond->alb_work, alb_delta_in_ticks); + if (!bond->kill_timers) + queue_delayed_work(bond->wq, &bond->alb_work, alb_delta_in_ticks); out: read_unlock(&bond->lock); } diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 43f2ea5..6d79b78 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -777,6 +777,9 @@ static void bond_resend_igmp_join_requests(struct bonding *bond) read_lock(&bond->lock); + if (bond->kill_timers) + goto out; + /* rejoin all groups on bond device */ __bond_resend_igmp_join_requests(bond->dev); @@ -790,9 +793,9 @@ static void bond_resend_igmp_join_requests(struct bonding *bond) __bond_resend_igmp_join_requests(vlan_dev); } - if (--bond->igmp_retrans > 0) + if ((--bond->igmp_retrans > 0) && !bond->kill_timers) queue_delayed_work(bond->wq, &bond->mcast_work, HZ/5); - +out: read_unlock(&bond->lock); } @@ -2538,7 +2541,7 @@ void bond_mii_monitor(struct work_struct *work) } re_arm: - if (bond->params.miimon) + if (bond->params.miimon && !bond->kill_timers) queue_delayed_work(bond->wq, &bond->mii_work, msecs_to_jiffies(bond->params.miimon)); out: @@ -2886,7 +2889,7 @@ void bond_loadbalance_arp_mon(struct work_struct *work) } re_arm: - if (bond->params.arp_interval) + if (bond->params.arp_interval && !bond->kill_timers) queue_delayed_work(bond->wq, &bond->arp_work, delta_in_ticks); out: read_unlock(&bond->lock); @@ -3154,7 +3157,7 @@ void bond_activebackup_arp_mon(struct work_struct *work) bond_ab_arp_probe(bond); re_arm: - if (bond->params.arp_interval) + if (bond->params.arp_interval && !bond->kill_timers) queue_delayed_work(bond->wq, &bond->arp_work, delta_in_ticks); out: read_unlock(&bond->lock); -- cgit v0.10.2 From d0e5d83284dac15c015bb48115b6780f5a6413cd Mon Sep 17 00:00:00 2001 From: David Vrabel Date: Fri, 30 Sep 2011 06:37:51 +0000 Subject: net: xen-netback: correctly restart Tx after a VM restore/migrate If a VM is saved and restored (or migrated) the netback driver will no longer process any Tx packets from the frontend. xenvif_up() does not schedule the processing of any pending Tx requests from the front end because the carrier is off. Without this initial kick the frontend just adds Tx requests to the ring without raising an event (until the ring is full). This was caused by 47103041e91794acdbc6165da0ae288d844c820b (net: xen-netback: convert to hw_features) which reordered the calls to xenvif_up() and netif_carrier_on() in xenvif_connect(). Signed-off-by: David Vrabel Cc: Ian Campbell Acked-by: Ian Campbell Signed-off-by: David S. Miller diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c index 0ca86f9..1825629 100644 --- a/drivers/net/xen-netback/interface.c +++ b/drivers/net/xen-netback/interface.c @@ -327,12 +327,12 @@ int xenvif_connect(struct xenvif *vif, unsigned long tx_ring_ref, xenvif_get(vif); rtnl_lock(); - if (netif_running(vif->dev)) - xenvif_up(vif); if (!vif->can_sg && vif->dev->mtu > ETH_DATA_LEN) dev_set_mtu(vif->dev, ETH_DATA_LEN); netdev_update_features(vif->dev); netif_carrier_on(vif->dev); + if (netif_running(vif->dev)) + xenvif_up(vif); rtnl_unlock(); return 0; -- cgit v0.10.2 From 7091fbd82cd5686444ffe9935ed6a8190101fe9d Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Fri, 30 Sep 2011 10:38:28 +0000 Subject: make PACKET_STATISTICS getsockopt report consistently between ring and non-ring This is a minor change. Up until kernel 2.6.32, getsockopt(fd, SOL_PACKET, PACKET_STATISTICS, ...) would return total and dropped packets since its last invocation. The introduction of socket queue overflow reporting [1] changed drop rate calculation in the normal packet socket path, but not when using a packet ring. As a result, the getsockopt now returns different statistics depending on the reception method used. With a ring, it still returns the count since the last call, as counts are incremented in tpacket_rcv and reset in getsockopt. Without a ring, it returns 0 if no drops occurred since the last getsockopt and the total drops over the lifespan of the socket otherwise. The culprit is this line in packet_rcv, executed on a drop: drop_n_acct: po->stats.tp_drops = atomic_inc_return(&sk->sk_drops); As it shows, the new drop number it taken from the socket drop counter, which is not reset at getsockopt. I put together a small example that demonstrates the issue [2]. It runs for 10 seconds and overflows the queue/ring on every odd second. The reported drop rates are: ring: 16, 0, 16, 0, 16, ... non-ring: 0, 15, 0, 30, 0, 46, 0, 60, 0 , 74. Note how the even ring counts monotonically increase. Because the getsockopt adds tp_drops to tp_packets, total counts are similarly reported cumulatively. Long story short, reinstating the original code, as the below patch does, fixes the issue at the cost of additional per-packet cycles. Another solution that does not introduce per-packet overhead is be to keep the current data path, record the value of sk_drops at getsockopt() at call N in a new field in struct packetsock and subtract that when reporting at call N+1. I'll be happy to code that, instead, it's just more messy. [1] http://patchwork.ozlabs.org/patch/35665/ [2] http://kernel.googlecode.com/files/test-packetsock-getstatistics.c Signed-off-by: Willem de Bruijn Signed-off-by: David S. Miller diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index c698cec..fabb4fa 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -961,7 +961,10 @@ static int packet_rcv(struct sk_buff *skb, struct net_device *dev, return 0; drop_n_acct: - po->stats.tp_drops = atomic_inc_return(&sk->sk_drops); + spin_lock(&sk->sk_receive_queue.lock); + po->stats.tp_drops++; + atomic_inc(&sk->sk_drops); + spin_unlock(&sk->sk_receive_queue.lock); drop_n_restore: if (skb_head != skb->data && skb_shared(skb)) { -- cgit v0.10.2 From 5f3a11419099d5cc010cfbfc524ca10d8fb81f89 Mon Sep 17 00:00:00 2001 From: Toshiharu Okada Date: Sun, 25 Sep 2011 21:27:42 +0000 Subject: pch_gbe: Fixed the issue on which PC was frozen when link was downed. When a link was downed during network use, there is an issue on which PC freezes. This patch fixed this issue. Signed-off-by: Toshiharu Okada Signed-off-by: David S. Miller diff --git a/drivers/net/pch_gbe/pch_gbe_main.c b/drivers/net/pch_gbe/pch_gbe_main.c index 567ff10..5474189 100644 --- a/drivers/net/pch_gbe/pch_gbe_main.c +++ b/drivers/net/pch_gbe/pch_gbe_main.c @@ -2195,7 +2195,7 @@ static int pch_gbe_napi_poll(struct napi_struct *napi, int budget) /* If no Tx and not enough Rx work done, * exit the polling mode */ - if ((work_done < budget) || !netif_running(netdev)) + if (work_done < budget) poll_end_flag = true; } -- cgit v0.10.2 From 805e969f6151eda7bc1a57e9c737054230acc3cc Mon Sep 17 00:00:00 2001 From: Toshiharu Okada Date: Sun, 25 Sep 2011 21:27:43 +0000 Subject: pch_gbe: Fixed the issue on which a network freezes The pch_gbe driver has an issue which a network stops, when receiving traffic is high. In the case, The link down and up are necessary to return a network. This patch fixed this issue. Signed-off-by: Toshiharu Okada Signed-off-by: David S. Miller diff --git a/drivers/net/pch_gbe/pch_gbe_main.c b/drivers/net/pch_gbe/pch_gbe_main.c index 5474189..b8b4ba2 100644 --- a/drivers/net/pch_gbe/pch_gbe_main.c +++ b/drivers/net/pch_gbe/pch_gbe_main.c @@ -1199,6 +1199,8 @@ static irqreturn_t pch_gbe_intr(int irq, void *data) iowrite32((int_en & ~PCH_GBE_INT_RX_FIFO_ERR), &hw->reg->INT_EN); pch_gbe_stop_receive(adapter); + int_st |= ioread32(&hw->reg->INT_ST); + int_st = int_st & ioread32(&hw->reg->INT_EN); } if (int_st & PCH_GBE_INT_RX_DMA_ERR) adapter->stats.intr_rx_dma_err_count++; @@ -1218,14 +1220,11 @@ static irqreturn_t pch_gbe_intr(int irq, void *data) /* Set Pause packet */ pch_gbe_mac_set_pause_packet(hw); } - if ((int_en & (PCH_GBE_INT_RX_DMA_CMPLT | PCH_GBE_INT_TX_CMPLT)) - == 0) { - return IRQ_HANDLED; - } } /* When request status is Receive interruption */ - if ((int_st & (PCH_GBE_INT_RX_DMA_CMPLT | PCH_GBE_INT_TX_CMPLT))) { + if ((int_st & (PCH_GBE_INT_RX_DMA_CMPLT | PCH_GBE_INT_TX_CMPLT)) || + (adapter->rx_stop_flag == true)) { if (likely(napi_schedule_prep(&adapter->napi))) { /* Enable only Rx Descriptor empty */ atomic_inc(&adapter->irq_sem); @@ -1385,7 +1384,7 @@ pch_gbe_clean_tx(struct pch_gbe_adapter *adapter, struct sk_buff *skb; unsigned int i; unsigned int cleaned_count = 0; - bool cleaned = false; + bool cleaned = true; pr_debug("next_to_clean : %d\n", tx_ring->next_to_clean); @@ -1396,7 +1395,6 @@ pch_gbe_clean_tx(struct pch_gbe_adapter *adapter, while ((tx_desc->gbec_status & DSC_INIT16) == 0x0000) { pr_debug("gbec_status:0x%04x\n", tx_desc->gbec_status); - cleaned = true; buffer_info = &tx_ring->buffer_info[i]; skb = buffer_info->skb; @@ -1439,8 +1437,10 @@ pch_gbe_clean_tx(struct pch_gbe_adapter *adapter, tx_desc = PCH_GBE_TX_DESC(*tx_ring, i); /* weight of a sort for tx, to avoid endless transmit cleanup */ - if (cleaned_count++ == PCH_GBE_TX_WEIGHT) + if (cleaned_count++ == PCH_GBE_TX_WEIGHT) { + cleaned = false; break; + } } pr_debug("called pch_gbe_unmap_and_free_tx_resource() %d count\n", cleaned_count); @@ -2168,7 +2168,6 @@ static int pch_gbe_napi_poll(struct napi_struct *napi, int budget) { struct pch_gbe_adapter *adapter = container_of(napi, struct pch_gbe_adapter, napi); - struct net_device *netdev = adapter->netdev; int work_done = 0; bool poll_end_flag = false; bool cleaned = false; @@ -2176,33 +2175,32 @@ static int pch_gbe_napi_poll(struct napi_struct *napi, int budget) pr_debug("budget : %d\n", budget); - /* Keep link state information with original netdev */ - if (!netif_carrier_ok(netdev)) { + pch_gbe_clean_rx(adapter, adapter->rx_ring, &work_done, budget); + cleaned = pch_gbe_clean_tx(adapter, adapter->tx_ring); + + if (!cleaned) + work_done = budget; + /* If no Tx and not enough Rx work done, + * exit the polling mode + */ + if (work_done < budget) poll_end_flag = true; - } else { - pch_gbe_clean_rx(adapter, adapter->rx_ring, &work_done, budget); + + if (poll_end_flag) { + napi_complete(napi); + if (adapter->rx_stop_flag) { + adapter->rx_stop_flag = false; + pch_gbe_start_receive(&adapter->hw); + } + pch_gbe_irq_enable(adapter); + } else if (adapter->rx_stop_flag) { adapter->rx_stop_flag = false; pch_gbe_start_receive(&adapter->hw); int_en = ioread32(&adapter->hw.reg->INT_EN); iowrite32((int_en | PCH_GBE_INT_RX_FIFO_ERR), - &adapter->hw.reg->INT_EN); + &adapter->hw.reg->INT_EN); } - cleaned = pch_gbe_clean_tx(adapter, adapter->tx_ring); - - if (cleaned) - work_done = budget; - /* If no Tx and not enough Rx work done, - * exit the polling mode - */ - if (work_done < budget) - poll_end_flag = true; - } - - if (poll_end_flag) { - napi_complete(napi); - pch_gbe_irq_enable(adapter); - } pr_debug("poll_end_flag : %d work_done : %d budget : %d\n", poll_end_flag, work_done, budget); -- cgit v0.10.2 From 2c8fc867602e385fd2abe76da0b6bda8ed907547 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Mon, 3 Oct 2011 14:28:18 -0400 Subject: ide-disk: Fix request requeuing Simon Kirby reported that on his RAID setup with idedisk underneath the box OOMs after a couple of days of runtime. Running with CONFIG_DEBUG_KMEMLEAK pointed to idedisk_prep_fn() which unconditionally allocates an ide_cmd struct. However, ide_requeue_and_plug() can be called more than once per request, either from the request issue or the IRQ handler path and do blk_peek_request() ends up in idedisk_prep_fn() repeatedly, allocating a struct ide_cmd everytime and "forgetting" the previous pointer. Make sure the code reuses the old allocated chunk. Reported-and-tested-by: Simon Kirby Cc: [ 39.x, 3.0.x ] Link: http://marc.info/?l=linux-kernel&m=131667641517919 Link: http://lkml.kernel.org/r/20110922072643.GA27232@hostway.ca Signed-off-by: Borislav Petkov Signed-off-by: David S. Miller diff --git a/drivers/ide/ide-disk.c b/drivers/ide/ide-disk.c index 2747980..16f69be 100644 --- a/drivers/ide/ide-disk.c +++ b/drivers/ide/ide-disk.c @@ -435,7 +435,12 @@ static int idedisk_prep_fn(struct request_queue *q, struct request *rq) if (!(rq->cmd_flags & REQ_FLUSH)) return BLKPREP_OK; - cmd = kzalloc(sizeof(*cmd), GFP_ATOMIC); + if (rq->special) { + cmd = rq->special; + memset(cmd, 0, sizeof(*cmd)); + } else { + cmd = kzalloc(sizeof(*cmd), GFP_ATOMIC); + } /* FIXME: map struct ide_taskfile on rq->cmd[] */ BUG_ON(cmd == NULL); -- cgit v0.10.2 From 05faadcf59507e8eea57ffbeea9cbb14c9a2ab3d Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 3 Oct 2011 18:09:14 -0700 Subject: lis3: fix regression of HP DriveGuard with 8bit chip Commit 2a7fade7e03 ("hwmon: lis3: Power on corrections") caused a regression on HP laptops with 8bit chip. Writing CTRL2_BOOT_8B bit seems clearing the BIOS setup, and no proper interrupt for DriveGuard will be triggered any more. Since the init code there is basically only for embedded devices, put a pdata check so that the problematic initialization will be skipped for hp_accel stuff. Signed-off-by: Takashi Iwai Cc: Eric Piel Cc: Samu Onkalo Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/misc/lis3lv02d/lis3lv02d.c b/drivers/misc/lis3lv02d/lis3lv02d.c index b928bc1..8b51cd6 100644 --- a/drivers/misc/lis3lv02d/lis3lv02d.c +++ b/drivers/misc/lis3lv02d/lis3lv02d.c @@ -375,12 +375,14 @@ void lis3lv02d_poweron(struct lis3lv02d *lis3) * both have been read. So the value read will always be correct. * Set BOOT bit to refresh factory tuning values. */ - lis3->read(lis3, CTRL_REG2, ®); - if (lis3->whoami == WAI_12B) - reg |= CTRL2_BDU | CTRL2_BOOT; - else - reg |= CTRL2_BOOT_8B; - lis3->write(lis3, CTRL_REG2, reg); + if (lis3->pdata) { + lis3->read(lis3, CTRL_REG2, ®); + if (lis3->whoami == WAI_12B) + reg |= CTRL2_BDU | CTRL2_BOOT; + else + reg |= CTRL2_BOOT_8B; + lis3->write(lis3, CTRL_REG2, reg); + } /* LIS3 power on delay is quite long */ msleep(lis3->pwron_delay / lis3lv02d_get_odr()); -- cgit v0.10.2 From 27504be5c10bdc61761ec0126b3645428598ad91 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Tue, 6 Sep 2011 17:16:34 +0900 Subject: spi-topcliff-pch: add tx-memory clear after complete transmitting Currently, in case of reading date from SPI flash, command is sent twice. The cause is that tx-memory clear processing is missing . This patch adds the tx-momory clear processing. Signed-off-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c index 1d23f38..7a98caa 100644 --- a/drivers/spi/spi-topcliff-pch.c +++ b/drivers/spi/spi-topcliff-pch.c @@ -837,6 +837,11 @@ static void pch_spi_start_transfer(struct pch_spi_data *data) dma_sync_sg_for_cpu(&data->master->dev, dma->sg_rx_p, dma->nent, DMA_FROM_DEVICE); + + dma_sync_sg_for_cpu(&data->master->dev, dma->sg_tx_p, dma->nent, + DMA_FROM_DEVICE); + memset(data->dma.tx_buf_virt, 0, PAGE_SIZE); + async_tx_ack(dma->desc_rx); async_tx_ack(dma->desc_tx); kfree(dma->sg_tx_p); -- cgit v0.10.2 From 8b7aa961a815dad389525dcd1c36a44c249b1760 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Tue, 6 Sep 2011 17:16:35 +0900 Subject: spi-topcliff-pch: Fix SSN Control issue During processing 1 command/data series, SSN should keep LOW. However, currently, SSN becomes HIGH. This patch fixes the issue. Signed-off-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c index 7a98caa..c36d516 100644 --- a/drivers/spi/spi-topcliff-pch.c +++ b/drivers/spi/spi-topcliff-pch.c @@ -58,6 +58,7 @@ #define PCH_SLEEP_TIME 10 #define SSN_LOW 0x02U +#define SSN_HIGH 0x03U #define SSN_NO_CONTROL 0x00U #define PCH_MAX_CS 0xFF #define PCI_DEVICE_ID_GE_SPI 0x8816 @@ -756,10 +757,6 @@ static void pch_spi_set_ir(struct pch_spi_data *data) wait_event_interruptible(data->wait, data->transfer_complete); - pch_spi_writereg(data->master, PCH_SSNXCR, SSN_NO_CONTROL); - dev_dbg(&data->master->dev, - "%s:no more control over SSN-writing 0 to SSNXCR.", __func__); - /* clear all interrupts */ pch_spi_writereg(data->master, PCH_SPSR, pch_spi_readreg(data->master, PCH_SPSR)); @@ -848,9 +845,6 @@ static void pch_spi_start_transfer(struct pch_spi_data *data) kfree(dma->sg_rx_p); spin_lock_irqsave(&data->lock, flags); - pch_spi_writereg(data->master, PCH_SSNXCR, SSN_NO_CONTROL); - dev_dbg(&data->master->dev, - "%s:no more control over SSN-writing 0 to SSNXCR.", __func__); /* clear fifo threshold, disable interrupts, disable SPI transfer */ pch_spi_setclr_reg(data->master, PCH_SPCR, 0, @@ -1167,6 +1161,7 @@ static void pch_spi_process_messages(struct work_struct *pwork) if (data->use_dma) pch_spi_request_dma(data, data->current_msg->spi->bits_per_word); + pch_spi_writereg(data->master, PCH_SSNXCR, SSN_NO_CONTROL); do { /* If we are already processing a message get the next transfer structure from the message otherwise retrieve @@ -1227,6 +1222,7 @@ static void pch_spi_process_messages(struct work_struct *pwork) } while (data->cur_trans != NULL); + pch_spi_writereg(data->master, PCH_SSNXCR, SSN_HIGH); if (data->use_dma) pch_spi_release_dma(data); } -- cgit v0.10.2 From 373b0eb64b9c97a14720896dcfdebfa3268b1991 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Tue, 6 Sep 2011 17:16:36 +0900 Subject: spi-topcliff-pch: Fix CPU read complete condition issue We found Rx data sometimes drops.(with non-DMA transfer mode) The cause is read complete condition is not true. This patch fixes the issue. Signed-off-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c index c36d516..6cb2fdb 100644 --- a/drivers/spi/spi-topcliff-pch.c +++ b/drivers/spi/spi-topcliff-pch.c @@ -317,16 +317,19 @@ static void pch_spi_handler_sub(struct pch_spi_data *data, u32 reg_spsr_val, /* if transfer complete interrupt */ if (reg_spsr_val & SPSR_FI_BIT) { - if (tx_index < bpw_len) + if ((tx_index == bpw_len) && (rx_index == tx_index)) { + /* disable interrupts */ + pch_spi_setclr_reg(data->master, PCH_SPCR, 0, PCH_ALL); + + /* transfer is completed; + inform pch_spi_process_messages */ + data->transfer_complete = true; + data->transfer_active = false; + wake_up(&data->wait); + } else { dev_err(&data->master->dev, "%s : Transfer is not completed", __func__); - /* disable interrupts */ - pch_spi_setclr_reg(data->master, PCH_SPCR, 0, PCH_ALL); - - /* transfer is completed;inform pch_spi_process_messages */ - data->transfer_complete = true; - data->transfer_active = false; - wake_up(&data->wait); + } } } -- cgit v0.10.2 From 25e803f9c339a070cd7361e853aebde11b1f9217 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Tue, 6 Sep 2011 17:16:37 +0900 Subject: spi-topcliff-pch: Add recovery processing in case FIFO overrun error occurs Add recovery processing in case FIFO overrun error occurs with DMA transfer mode. Signed-off-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c index 6cb2fdb..c8efa75 100644 --- a/drivers/spi/spi-topcliff-pch.c +++ b/drivers/spi/spi-topcliff-pch.c @@ -352,16 +352,26 @@ static irqreturn_t pch_spi_handler(int irq, void *dev_id) "%s returning due to suspend\n", __func__); return IRQ_NONE; } - if (data->use_dma) - return IRQ_NONE; io_remap_addr = data->io_remap_addr; spsr = io_remap_addr + PCH_SPSR; reg_spsr_val = ioread32(spsr); - if (reg_spsr_val & SPSR_ORF_BIT) - dev_err(&board_dat->pdev->dev, "%s Over run error", __func__); + if (reg_spsr_val & SPSR_ORF_BIT) { + dev_err(&board_dat->pdev->dev, "%s Over run error\n", __func__); + if (data->current_msg->complete != 0) { + data->transfer_complete = true; + data->current_msg->status = -EIO; + data->current_msg->complete(data->current_msg->context); + data->bcurrent_msg_processing = false; + data->current_msg = NULL; + data->cur_trans = NULL; + } + } + + if (data->use_dma) + return IRQ_NONE; /* Check if the interrupt is for SPI device */ if (reg_spsr_val & (SPSR_FI_BIT | SPSR_RFI_BIT)) { @@ -815,10 +825,11 @@ static void pch_spi_copy_rx_data_for_dma(struct pch_spi_data *data, int bpw) } } -static void pch_spi_start_transfer(struct pch_spi_data *data) +static int pch_spi_start_transfer(struct pch_spi_data *data) { struct pch_spi_dma_ctrl *dma; unsigned long flags; + int rtn; dma = &data->dma; @@ -833,7 +844,9 @@ static void pch_spi_start_transfer(struct pch_spi_data *data) initiating the transfer. */ dev_dbg(&data->master->dev, "%s:waiting for transfer to get over\n", __func__); - wait_event_interruptible(data->wait, data->transfer_complete); + rtn = wait_event_interruptible_timeout(data->wait, + data->transfer_complete, + msecs_to_jiffies(2 * HZ)); dma_sync_sg_for_cpu(&data->master->dev, dma->sg_rx_p, dma->nent, DMA_FROM_DEVICE); @@ -860,6 +873,8 @@ static void pch_spi_start_transfer(struct pch_spi_data *data) pch_spi_clear_fifo(data->master); spin_unlock_irqrestore(&data->lock, flags); + + return rtn; } static void pch_dma_rx_complete(void *arg) @@ -1187,7 +1202,8 @@ static void pch_spi_process_messages(struct work_struct *pwork) if (data->use_dma) { pch_spi_handle_dma(data, &bpw); - pch_spi_start_transfer(data); + if (!pch_spi_start_transfer(data)) + goto out; pch_spi_copy_rx_data_for_dma(data, bpw); } else { pch_spi_set_tx(data, &bpw); @@ -1225,6 +1241,7 @@ static void pch_spi_process_messages(struct work_struct *pwork) } while (data->cur_trans != NULL); +out: pch_spi_writereg(data->master, PCH_SSNXCR, SSN_HIGH); if (data->use_dma) pch_spi_release_dma(data); -- cgit v0.10.2 From f3e03e2eb0bcd9266720695b82a523e6bc526ff7 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Tue, 6 Sep 2011 17:16:38 +0900 Subject: spi-topcliff-pch: Fix overrun issue We found that adding load, Rx data sometimes drops.(with DMA transfer mode) The cause is that before starting Rx-DMA processing, Tx-DMA processing starts. This causes FIFO overrun occurs. This patch fixes the issue by modifying FIFO tx-threshold and DMA descriptor size like below. Current this patch Rx-descriptor 4Byte+12Byte*341 --> 12Byte*340-4Byte-12Byte Rx-threshold (Not modified) Tx-descriptor 4Byte+12Byte*341 --> 16Byte-12Byte*340 Rx-threshold 12Byte --> 2Byte Signed-off-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c index c8efa75..6a80749 100644 --- a/drivers/spi/spi-topcliff-pch.c +++ b/drivers/spi/spi-topcliff-pch.c @@ -50,6 +50,8 @@ #define PCH_RX_THOLD 7 #define PCH_RX_THOLD_MAX 15 +#define PCH_TX_THOLD 2 + #define PCH_MAX_BAUDRATE 5000000 #define PCH_MAX_FIFO_DEPTH 16 @@ -1040,8 +1042,7 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw) /* set receive fifo threshold and transmit fifo threshold */ pch_spi_setclr_reg(data->master, PCH_SPCR, ((size - 1) << SPCR_RFIC_FIELD) | - ((PCH_MAX_FIFO_DEPTH - PCH_DMA_TRANS_SIZE) << - SPCR_TFIC_FIELD), + (PCH_TX_THOLD << SPCR_TFIC_FIELD), MASK_RFIC_SPCR_BITS | MASK_TFIC_SPCR_BITS); spin_unlock_irqrestore(&data->lock, flags); @@ -1052,13 +1053,20 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw) /* offset, length setting */ sg = dma->sg_rx_p; for (i = 0; i < num; i++, sg++) { - if (i == 0) { - sg->offset = 0; + if (i == (num - 2)) { + sg->offset = size * i; + sg->offset = sg->offset * (*bpw / 8); sg_set_page(sg, virt_to_page(dma->rx_buf_virt), rem, sg->offset); sg_dma_len(sg) = rem; + } else if (i == (num - 1)) { + sg->offset = size * (i - 1) + rem; + sg->offset = sg->offset * (*bpw / 8); + sg_set_page(sg, virt_to_page(dma->rx_buf_virt), size, + sg->offset); + sg_dma_len(sg) = size; } else { - sg->offset = rem + size * (i - 1); + sg->offset = size * i; sg->offset = sg->offset * (*bpw / 8); sg_set_page(sg, virt_to_page(dma->rx_buf_virt), size, sg->offset); @@ -1082,6 +1090,16 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw) dma->desc_rx = desc_rx; /* TX */ + if (data->bpw_len > PCH_DMA_TRANS_SIZE) { + num = data->bpw_len / PCH_DMA_TRANS_SIZE; + size = PCH_DMA_TRANS_SIZE; + rem = 16; + } else { + num = 1; + size = data->bpw_len; + rem = data->bpw_len; + } + dma->sg_tx_p = kzalloc(sizeof(struct scatterlist)*num, GFP_ATOMIC); sg_init_table(dma->sg_tx_p, num); /* Initialize SG table */ /* offset, length setting */ -- cgit v0.10.2 From 12d5180bd7e683a4ae80830b82ba67e7b7fac7b2 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 4 Oct 2011 10:46:34 -0400 Subject: drm/radeon/kms: fix channel_remap setup (v2) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Most asics just use the hw default value which requires no explicit programming. For those that need a different value, the vbios will program it properly. As such, there's no need to program these registers explicitly in the driver. Changing MC_SHARED_CHREMAP requires a reload of all data in vram otherwise its contents will be scambled. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=40103 v2: drop now unused channel_remap functions. Signed-off-by: Alex Deucher Reviewed-by: Michel Dänzer 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 e8a7467..c4ffa14f 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1590,48 +1590,6 @@ static u32 evergreen_get_tile_pipe_to_backend_map(struct radeon_device *rdev, return backend_map; } -static void evergreen_program_channel_remap(struct radeon_device *rdev) -{ - u32 tcp_chan_steer_lo, tcp_chan_steer_hi, mc_shared_chremap, tmp; - - tmp = RREG32(MC_SHARED_CHMAP); - switch ((tmp & NOOFCHAN_MASK) >> NOOFCHAN_SHIFT) { - case 0: - case 1: - case 2: - case 3: - default: - /* default mapping */ - mc_shared_chremap = 0x00fac688; - break; - } - - switch (rdev->family) { - case CHIP_HEMLOCK: - case CHIP_CYPRESS: - case CHIP_BARTS: - tcp_chan_steer_lo = 0x54763210; - tcp_chan_steer_hi = 0x0000ba98; - break; - case CHIP_JUNIPER: - case CHIP_REDWOOD: - case CHIP_CEDAR: - case CHIP_PALM: - case CHIP_SUMO: - case CHIP_SUMO2: - case CHIP_TURKS: - case CHIP_CAICOS: - default: - tcp_chan_steer_lo = 0x76543210; - tcp_chan_steer_hi = 0x0000ba98; - break; - } - - WREG32(TCP_CHAN_STEER_LO, tcp_chan_steer_lo); - WREG32(TCP_CHAN_STEER_HI, tcp_chan_steer_hi); - WREG32(MC_SHARED_CHREMAP, mc_shared_chremap); -} - static void evergreen_gpu_init(struct radeon_device *rdev) { u32 cc_rb_backend_disable = 0; @@ -2078,8 +2036,6 @@ static void evergreen_gpu_init(struct radeon_device *rdev) WREG32(DMIF_ADDR_CONFIG, gb_addr_config); WREG32(HDP_ADDR_CONFIG, gb_addr_config); - evergreen_program_channel_remap(rdev); - num_shader_engines = ((RREG32(GB_ADDR_CONFIG) & NUM_SHADER_ENGINES(3)) >> 12) + 1; grbm_gfx_index = INSTANCE_BROADCAST_WRITES; diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c index 99fbd79..8c79ca9 100644 --- a/drivers/gpu/drm/radeon/ni.c +++ b/drivers/gpu/drm/radeon/ni.c @@ -569,36 +569,6 @@ static u32 cayman_get_tile_pipe_to_backend_map(struct radeon_device *rdev, return backend_map; } -static void cayman_program_channel_remap(struct radeon_device *rdev) -{ - u32 tcp_chan_steer_lo, tcp_chan_steer_hi, mc_shared_chremap, tmp; - - tmp = RREG32(MC_SHARED_CHMAP); - switch ((tmp & NOOFCHAN_MASK) >> NOOFCHAN_SHIFT) { - case 0: - case 1: - case 2: - case 3: - default: - /* default mapping */ - mc_shared_chremap = 0x00fac688; - break; - } - - switch (rdev->family) { - case CHIP_CAYMAN: - default: - //tcp_chan_steer_lo = 0x54763210 - tcp_chan_steer_lo = 0x76543210; - tcp_chan_steer_hi = 0x0000ba98; - break; - } - - WREG32(TCP_CHAN_STEER_LO, tcp_chan_steer_lo); - WREG32(TCP_CHAN_STEER_HI, tcp_chan_steer_hi); - WREG32(MC_SHARED_CHREMAP, mc_shared_chremap); -} - static u32 cayman_get_disable_mask_per_asic(struct radeon_device *rdev, u32 disable_mask_per_se, u32 max_disable_mask_per_se, @@ -842,8 +812,6 @@ static void cayman_gpu_init(struct radeon_device *rdev) WREG32(DMIF_ADDR_CONFIG, gb_addr_config); WREG32(HDP_ADDR_CONFIG, gb_addr_config); - cayman_program_channel_remap(rdev); - /* primary versions */ WREG32(CC_RB_BACKEND_DISABLE, cc_rb_backend_disable); WREG32(CC_SYS_RB_BACKEND_DISABLE, cc_rb_backend_disable); diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index 4720d00..b13c2ee 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -536,55 +536,6 @@ static u32 r700_get_tile_pipe_to_backend_map(struct radeon_device *rdev, return backend_map; } -static void rv770_program_channel_remap(struct radeon_device *rdev) -{ - u32 tcp_chan_steer, mc_shared_chremap, tmp; - bool force_no_swizzle; - - switch (rdev->family) { - case CHIP_RV770: - case CHIP_RV730: - force_no_swizzle = false; - break; - case CHIP_RV710: - case CHIP_RV740: - default: - force_no_swizzle = true; - break; - } - - tmp = RREG32(MC_SHARED_CHMAP); - switch ((tmp & NOOFCHAN_MASK) >> NOOFCHAN_SHIFT) { - case 0: - case 1: - default: - /* default mapping */ - mc_shared_chremap = 0x00fac688; - break; - case 2: - case 3: - if (force_no_swizzle) - mc_shared_chremap = 0x00fac688; - else - mc_shared_chremap = 0x00bbc298; - break; - } - - if (rdev->family == CHIP_RV740) - tcp_chan_steer = 0x00ef2a60; - else - tcp_chan_steer = 0x00fac688; - - /* RV770 CE has special chremap setup */ - if (rdev->pdev->device == 0x944e) { - tcp_chan_steer = 0x00b08b08; - mc_shared_chremap = 0x00b08b08; - } - - WREG32(TCP_CHAN_STEER, tcp_chan_steer); - WREG32(MC_SHARED_CHREMAP, mc_shared_chremap); -} - static void rv770_gpu_init(struct radeon_device *rdev) { int i, j, num_qd_pipes; @@ -785,8 +736,6 @@ static void rv770_gpu_init(struct radeon_device *rdev) WREG32(DCP_TILING_CONFIG, (gb_tiling_config & 0xffff)); WREG32(HDP_TILING_CONFIG, (gb_tiling_config & 0xffff)); - rv770_program_channel_remap(rdev); - WREG32(CC_RB_BACKEND_DISABLE, cc_rb_backend_disable); WREG32(CC_GC_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config); WREG32(GC_USER_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config); -- cgit v0.10.2 From 5f39e6705faade2e89d119958a8c51b9b6e2c53c Mon Sep 17 00:00:00 2001 From: Jon Mason Date: Mon, 3 Oct 2011 09:50:20 -0500 Subject: PCI: Disable MPS configuration by default Add the ability to disable PCI-E MPS turning and using the BIOS configured MPS defaults. Due to the number of issues recently discovered on some x86 chipsets, make this the default behavior. Also, add the option for peer to peer DMA MPS configuration. Peer to peer DMA is outside the scope of this patch, but MPS configuration could prevent it from working by having the MPS on one root port different than the MPS on another. To work around this, simply make the system wide MPS the smallest possible value (128B). Signed-off-by: Jon Mason Acked-by: Benjamin Herrenschmidt Signed-off-by: Linus Torvalds diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 4e84fd4..e9651f0 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -77,7 +77,7 @@ unsigned long pci_cardbus_mem_size = DEFAULT_CARDBUS_MEM_SIZE; unsigned long pci_hotplug_io_size = DEFAULT_HOTPLUG_IO_SIZE; unsigned long pci_hotplug_mem_size = DEFAULT_HOTPLUG_MEM_SIZE; -enum pcie_bus_config_types pcie_bus_config = PCIE_BUS_SAFE; +enum pcie_bus_config_types pcie_bus_config = PCIE_BUS_TUNE_OFF; /* * The default CLS is used if arch didn't set CLS explicitly and not @@ -3568,10 +3568,14 @@ static int __init pci_setup(char *str) pci_hotplug_io_size = memparse(str + 9, &str); } else if (!strncmp(str, "hpmemsize=", 10)) { pci_hotplug_mem_size = memparse(str + 10, &str); + } else if (!strncmp(str, "pcie_bus_tune_off", 17)) { + pcie_bus_config = PCIE_BUS_TUNE_OFF; } else if (!strncmp(str, "pcie_bus_safe", 13)) { pcie_bus_config = PCIE_BUS_SAFE; } else if (!strncmp(str, "pcie_bus_perf", 13)) { pcie_bus_config = PCIE_BUS_PERFORMANCE; + } else if (!strncmp(str, "pcie_bus_peer2peer", 18)) { + pcie_bus_config = PCIE_BUS_PEER2PEER; } else { printk(KERN_ERR "PCI: Unknown option `%s'\n", str); diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index f3f94a5..6ab6bd3 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -1458,12 +1458,24 @@ static int pcie_bus_configure_set(struct pci_dev *dev, void *data) */ void pcie_bus_configure_settings(struct pci_bus *bus, u8 mpss) { - u8 smpss = mpss; + u8 smpss; if (!pci_is_pcie(bus->self)) return; + if (pcie_bus_config == PCIE_BUS_TUNE_OFF) + return; + + /* FIXME - Peer to peer DMA is possible, though the endpoint would need + * to be aware to the MPS of the destination. To work around this, + * simply force the MPS of the entire system to the smallest possible. + */ + if (pcie_bus_config == PCIE_BUS_PEER2PEER) + smpss = 0; + if (pcie_bus_config == PCIE_BUS_SAFE) { + smpss = mpss; + pcie_find_smpss(bus->self, &smpss); pci_walk_bus(bus, pcie_find_smpss, &smpss); } diff --git a/include/linux/pci.h b/include/linux/pci.h index 8c230cb..9fc0122 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -621,8 +621,9 @@ struct pci_driver { extern void pcie_bus_configure_settings(struct pci_bus *bus, u8 smpss); enum pcie_bus_config_types { - PCIE_BUS_PERFORMANCE, + PCIE_BUS_TUNE_OFF, PCIE_BUS_SAFE, + PCIE_BUS_PERFORMANCE, PCIE_BUS_PEER2PEER, }; -- cgit v0.10.2 From 976d167615b64e14bc1491ca51d424e2ba9a5e84 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 4 Oct 2011 18:11:50 -0700 Subject: Linux 3.1-rc9 diff --git a/Makefile b/Makefile index 73bb458..31f967c 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 3 PATCHLEVEL = 1 SUBLEVEL = 0 -EXTRAVERSION = -rc8 +EXTRAVERSION = -rc9 NAME = "Divemaster Edition" # *DOCUMENTATION* -- cgit v0.10.2 From cb2d0f3e968bff7c6d262aca3e3ab8d4184e69b2 Mon Sep 17 00:00:00 2001 From: David Ward Date: Sun, 18 Sep 2011 12:53:20 +0000 Subject: macvlan/macvtap: Fix unicast between macvtap interfaces in bridge mode Packets should always be forwarded to the lowerdev using dev_forward_skb. vlan->forward is for packets being forwarded directly to another macvlan/ macvtap device (used for multicast in bridge mode). Reported-and-tested-by: Shlomo Pongratz Signed-off-by: David Ward Signed-off-by: David S. Miller diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index 05172c3..376e3e9 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -239,7 +239,7 @@ static int macvlan_queue_xmit(struct sk_buff *skb, struct net_device *dev) dest = macvlan_hash_lookup(port, eth->h_dest); if (dest && dest->mode == MACVLAN_MODE_BRIDGE) { /* send to lowerdev first for its network taps */ - vlan->forward(vlan->lowerdev, skb); + dev_forward_skb(vlan->lowerdev, skb); return NET_XMIT_SUCCESS; } -- cgit v0.10.2 From 260fcbeb1ae9e768a44c9925338fbacb0d7e5ba9 Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Thu, 29 Sep 2011 17:10:10 +0000 Subject: tcp: properly handle md5sig_pool references tcp_v4_clear_md5_list() assumes that multiple tcp md5sig peers only hold one reference to md5sig_pool. but tcp_v4_md5_do_add() increases use count of md5sig_pool for each peer. This patch makes tcp_v4_md5_do_add() only increases use count for the first tcp md5sig peer. Signed-off-by: Zheng Yan Signed-off-by: David S. Miller diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index c34f015..7963e03 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -927,18 +927,21 @@ int tcp_v4_md5_do_add(struct sock *sk, __be32 addr, } sk_nocaps_add(sk, NETIF_F_GSO_MASK); } - if (tcp_alloc_md5sig_pool(sk) == NULL) { + + md5sig = tp->md5sig_info; + if (md5sig->entries4 == 0 && + tcp_alloc_md5sig_pool(sk) == NULL) { kfree(newkey); return -ENOMEM; } - md5sig = tp->md5sig_info; if (md5sig->alloced4 == md5sig->entries4) { keys = kmalloc((sizeof(*keys) * (md5sig->entries4 + 1)), GFP_ATOMIC); if (!keys) { kfree(newkey); - tcp_free_md5sig_pool(); + if (md5sig->entries4 == 0) + tcp_free_md5sig_pool(); return -ENOMEM; } @@ -982,6 +985,7 @@ int tcp_v4_md5_do_del(struct sock *sk, __be32 addr) kfree(tp->md5sig_info->keys4); tp->md5sig_info->keys4 = NULL; tp->md5sig_info->alloced4 = 0; + tcp_free_md5sig_pool(); } else if (tp->md5sig_info->entries4 != i) { /* Need to do some manipulation */ memmove(&tp->md5sig_info->keys4[i], @@ -989,7 +993,6 @@ int tcp_v4_md5_do_del(struct sock *sk, __be32 addr) (tp->md5sig_info->entries4 - i) * sizeof(struct tcp4_md5sig_key)); } - tcp_free_md5sig_pool(); return 0; } } diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 79cc646..7b8fc57 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -591,7 +591,8 @@ static int tcp_v6_md5_do_add(struct sock *sk, const struct in6_addr *peer, } sk_nocaps_add(sk, NETIF_F_GSO_MASK); } - if (tcp_alloc_md5sig_pool(sk) == NULL) { + if (tp->md5sig_info->entries6 == 0 && + tcp_alloc_md5sig_pool(sk) == NULL) { kfree(newkey); return -ENOMEM; } @@ -600,8 +601,9 @@ static int tcp_v6_md5_do_add(struct sock *sk, const struct in6_addr *peer, (tp->md5sig_info->entries6 + 1)), GFP_ATOMIC); if (!keys) { - tcp_free_md5sig_pool(); kfree(newkey); + if (tp->md5sig_info->entries6 == 0) + tcp_free_md5sig_pool(); return -ENOMEM; } @@ -647,6 +649,7 @@ static int tcp_v6_md5_do_del(struct sock *sk, const struct in6_addr *peer) kfree(tp->md5sig_info->keys6); tp->md5sig_info->keys6 = NULL; tp->md5sig_info->alloced6 = 0; + tcp_free_md5sig_pool(); } else { /* shrink the database */ if (tp->md5sig_info->entries6 != i) @@ -655,7 +658,6 @@ static int tcp_v6_md5_do_del(struct sock *sk, const struct in6_addr *peer) (tp->md5sig_info->entries6 - i) * sizeof (tp->md5sig_info->keys6[0])); } - tcp_free_md5sig_pool(); return 0; } } -- cgit v0.10.2 From 1e5289e121372a3494402b1b131b41bfe1cf9b7f Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Sun, 2 Oct 2011 04:21:50 +0000 Subject: tcp: properly update lost_cnt_hint during shifting lost_skb_hint is used by tcp_mark_head_lost() to mark the first unhandled skb. lost_cnt_hint is the number of packets or sacked packets before the lost_skb_hint; When shifting a skb that is before the lost_skb_hint, if tcp_is_fack() is ture, the skb has already been counted in the lost_cnt_hint; if tcp_is_fack() is false, tcp_sacktag_one() will increase the lost_cnt_hint. So tcp_shifted_skb() does not need to adjust the lost_cnt_hint by itself. When shifting a skb that is equal to lost_skb_hint, the shifted packets will not be counted by tcp_mark_head_lost(). So tcp_shifted_skb() should adjust the lost_cnt_hint even tcp_is_fack(tp) is true. Signed-off-by: Zheng Yan Signed-off-by: David S. Miller diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 21fab3e..d73aab3 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1389,9 +1389,7 @@ static int tcp_shifted_skb(struct sock *sk, struct sk_buff *skb, BUG_ON(!pcount); - /* Tweak before seqno plays */ - if (!tcp_is_fack(tp) && tcp_is_sack(tp) && tp->lost_skb_hint && - !before(TCP_SKB_CB(tp->lost_skb_hint)->seq, TCP_SKB_CB(skb)->seq)) + if (skb == tp->lost_skb_hint) tp->lost_cnt_hint += pcount; TCP_SKB_CB(prev)->end_seq += shifted; -- cgit v0.10.2 From d6069dae4700cf5290a75002c46b9be8ea7eef3c Mon Sep 17 00:00:00 2001 From: Jason Gerecke Date: Tue, 4 Oct 2011 22:50:45 -0700 Subject: Input: wacom - revert "Cintiq 21UX2 does not have menu strips" This reverts commit 71c86ce59791bcd67af937bbea719a508079d7c2. The 21UX2 does have touchstrips, but they are in a somewhat- hidden location. Signed-off-by: Jason Gerecke Acked-by: Ping Cheng Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/tablet/wacom_wac.c b/drivers/input/tablet/wacom_wac.c index 0dc97ec..9dea718 100644 --- a/drivers/input/tablet/wacom_wac.c +++ b/drivers/input/tablet/wacom_wac.c @@ -1124,11 +1124,8 @@ void wacom_setup_input_capabilities(struct input_dev *input_dev, for (i = 0; i < 8; i++) __set_bit(BTN_0 + i, input_dev->keybit); - if (wacom_wac->features.type != WACOM_21UX2) { - input_set_abs_params(input_dev, ABS_RX, 0, 4096, 0, 0); - input_set_abs_params(input_dev, ABS_RY, 0, 4096, 0, 0); - } - + input_set_abs_params(input_dev, ABS_RX, 0, 4096, 0, 0); + input_set_abs_params(input_dev, ABS_RY, 0, 4096, 0, 0); input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0); __set_bit(INPUT_PROP_DIRECT, input_dev->propbit); -- cgit v0.10.2 From 3458e21c0d384ca04b27a2ea24d9314c1b57530f Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Wed, 5 Oct 2011 03:24:43 +0000 Subject: netfilter: Use proper rwlock init function Replace the open coded initialization with the init function. Signed-off-by: Thomas Gleixner Acked-by: Hans Schillstrom Signed-off-by: David S. Miller diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 2b771dc..5290ac3 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -3679,7 +3679,7 @@ int __net_init ip_vs_control_net_init(struct net *net) int idx; struct netns_ipvs *ipvs = net_ipvs(net); - ipvs->rs_lock = __RW_LOCK_UNLOCKED(ipvs->rs_lock); + rwlock_init(&ipvs->rs_lock); /* Initialize rs_table */ for (idx = 0; idx < IP_VS_RTAB_SIZE; idx++) -- cgit v0.10.2 From b64b73d7d0c480f75684519c6134e79d50c1b341 Mon Sep 17 00:00:00 2001 From: stephen hemminger Date: Mon, 3 Oct 2011 18:14:45 +0000 Subject: bridge: leave carrier on for empty bridge This resolves a regression seen by some users of bridging. Some users use the bridge like a dummy device. They expect to be able to put an IPv6 address on the device with no ports attached. Although there are better ways of doing this, there is no reason to not allow it. Note: the bridge still will reflect the state of ports in the bridge if there are any added. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c index 32b8f9f..ff3ed60 100644 --- a/net/bridge/br_device.c +++ b/net/bridge/br_device.c @@ -91,7 +91,6 @@ static int br_dev_open(struct net_device *dev) { struct net_bridge *br = netdev_priv(dev); - netif_carrier_off(dev); netdev_update_features(dev); netif_start_queue(dev); br_stp_enable_bridge(br); @@ -108,8 +107,6 @@ static int br_dev_stop(struct net_device *dev) { struct net_bridge *br = netdev_priv(dev); - netif_carrier_off(dev); - br_stp_disable_bridge(br); br_multicast_stop(br); -- cgit v0.10.2 From 186c6bbced722cfeff041d2a1264c95f5d042050 Mon Sep 17 00:00:00 2001 From: Benjamin Poirier Date: Tue, 4 Oct 2011 04:00:30 +0000 Subject: net: fix typos in Documentation/networking/scaling.txt The second hunk fixes rps_sock_flow_table but has to re-wrap the paragraph. Signed-off-by: Benjamin Poirier Signed-off-by: David S. Miller diff --git a/Documentation/networking/scaling.txt b/Documentation/networking/scaling.txt index 8ce7c30..fe67b5c 100644 --- a/Documentation/networking/scaling.txt +++ b/Documentation/networking/scaling.txt @@ -27,7 +27,7 @@ applying a filter to each packet that assigns it to one of a small number of logical flows. Packets for each flow are steered to a separate receive queue, which in turn can be processed by separate CPUs. This mechanism is generally known as “Receive-side Scaling” (RSS). The goal of RSS and -the other scaling techniques to increase performance uniformly. +the other scaling techniques is to increase performance uniformly. Multi-queue distribution can also be used for traffic prioritization, but that is not the focus of these techniques. @@ -186,10 +186,10 @@ are steered using plain RPS. Multiple table entries may point to the same CPU. Indeed, with many flows and few CPUs, it is very likely that a single application thread handles flows with many different flow hashes. -rps_sock_table is a global flow table that contains the *desired* CPU for -flows: the CPU that is currently processing the flow in userspace. Each -table value is a CPU index that is updated during calls to recvmsg and -sendmsg (specifically, inet_recvmsg(), inet_sendmsg(), inet_sendpage() +rps_sock_flow_table is a global flow table that contains the *desired* CPU +for flows: the CPU that is currently processing the flow in userspace. +Each table value is a CPU index that is updated during calls to recvmsg +and sendmsg (specifically, inet_recvmsg(), inet_sendmsg(), inet_sendpage() and tcp_splice_read()). When the scheduler moves a thread to a new CPU while it has outstanding -- cgit v0.10.2 From 29cf7a30f8a0ce4af2406d93d5a332099be26923 Mon Sep 17 00:00:00 2001 From: Paul Menzel Date: Wed, 31 Aug 2011 17:07:10 +0200 Subject: x86/PCI: use host bridge _CRS info on ASUS M2V-MX SE In summary, this DMI quirk uses the _CRS info by default for the ASUS M2V-MX SE by turning on `pci=use_crs` and is similar to the quirk added by commit 2491762cfb47 ("x86/PCI: use host bridge _CRS info on ASRock ALiveSATA2-GLAN") whose commit message should be read for further information. Since commit 3e3da00c01d0 ("x86/pci: AMD one chain system to use pci read out res") Linux gives the following oops: parport0: PC-style at 0x378, irq 7 [PCSPP,TRISTATE] HDA Intel 0000:20:01.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 HDA Intel 0000:20:01.0: setting latency timer to 64 BUG: unable to handle kernel paging request at ffffc90011c08000 IP: [] azx_probe+0x3ad/0x86b [snd_hda_intel] PGD 13781a067 PUD 13781b067 PMD 1300ba067 PTE 800000fd00000173 Oops: 0009 [#1] SMP last sysfs file: /sys/module/snd_pcm/initstate CPU 0 Modules linked in: snd_hda_intel(+) snd_hda_codec snd_hwdep snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_midi snd_rawmidi snd_seq_midi_event tpm_tis tpm snd_seq tpm_bios psmouse parport_pc snd_timer snd_seq_device parport processor evdev snd i2c_viapro thermal_sys amd64_edac_mod k8temp i2c_core soundcore shpchp pcspkr serio_raw asus_atk0110 pci_hotplug edac_core button snd_page_alloc edac_mce_amd ext3 jbd mbcache sha256_generic cryptd aes_x86_64 aes_generic cbc dm_crypt dm_mod raid1 md_mod usbhid hid sg sd_mod crc_t10dif sr_mod cdrom ata_generic uhci_hcd sata_via pata_via libata ehci_hcd usbcore scsi_mod via_rhine mii nls_base [last unloaded: scsi_wait_scan] Pid: 1153, comm: work_for_cpu Not tainted 2.6.37-1-amd64 #1 M2V-MX SE/System Product Name RIP: 0010:[] [] azx_probe+0x3ad/0x86b [snd_hda_intel] RSP: 0018:ffff88013153fe50 EFLAGS: 00010286 RAX: ffffc90011c08000 RBX: ffff88013029ec00 RCX: 0000000000000006 RDX: 0000000000000000 RSI: 0000000000000246 RDI: 0000000000000246 RBP: ffff88013341d000 R08: 0000000000000000 R09: 0000000000000040 R10: 0000000000000286 R11: 0000000000003731 R12: ffff88013029c400 R13: 0000000000000000 R14: 0000000000000000 R15: ffff88013341d090 FS: 0000000000000000(0000) GS:ffff8800bfc00000(0000) knlGS:00000000f7610ab0 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: ffffc90011c08000 CR3: 0000000132f57000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process work_for_cpu (pid: 1153, threadinfo ffff88013153e000, task ffff8801303c86c0) Stack: 0000000000000005 ffffffff8123ad65 00000000000136c0 ffff88013029c400 ffff8801303c8998 ffff88013341d000 ffff88013341d090 ffff8801322d9dc8 ffff88013341d208 0000000000000000 0000000000000000 ffffffff811ad232 Call Trace: [] ? __pm_runtime_set_status+0x162/0x186 [] ? local_pci_probe+0x49/0x92 [] ? do_work_for_cpu+0x0/0x1b [] ? do_work_for_cpu+0x0/0x1b [] ? do_work_for_cpu+0xb/0x1b [] ? kthread+0x7a/0x82 [] ? kernel_thread_helper+0x4/0x10 [] ? kthread+0x0/0x82 [] ? kernel_thread_helper+0x0/0x10 Code: f4 01 00 00 ef 31 f6 48 89 df e8 29 dd ff ff 85 c0 0f 88 2b 03 00 00 48 89 ef e8 b4 39 c3 e0 8b 7b 40 e8 fc 9d b1 e0 48 8b 43 38 <66> 8b 10 66 89 14 24 8b 43 14 83 e8 03 83 f8 01 77 32 31 d2 be RIP [] azx_probe+0x3ad/0x86b [snd_hda_intel] RSP CR2: ffffc90011c08000 ---[ end trace 8d1f3ebc136437fd ]--- Trusting the ACPI _CRS information (`pci=use_crs`) fixes this problem. $ dmesg | grep -i crs # with the quirk PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug The match has to be against the DMI board entries though since the vendor entries are not populated. DMI: System manufacturer System Product Name/M2V-MX SE, BIOS 0304 10/30/2007 This quirk should be removed when `pci=use_crs` is enabled for machines from 2006 or earlier or some other solution is implemented. Using coreboot [1] with this board the problem does not exist but this quirk also does not affect it either. To be safe though the check is tightened to only take effect when the BIOS from American Megatrends is used. 15:13 < ruik> but coreboot does not need that 15:13 < ruik> because i have there only one root bus 15:13 < ruik> the audio is behind a bridge $ sudo dmidecode BIOS Information Vendor: American Megatrends Inc. Version: 0304 Release Date: 10/30/2007 [1] http://www.coreboot.org/ Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=30552 Cc: stable@kernel.org (2.6.34) Cc: Thomas Gleixner Cc: Ingo Molnar Cc: H. Peter Anvin Cc: x86@kernel.org Signed-off-by: Paul Menzel Signed-off-by: Bjorn Helgaas Acked-by: Jesse Barnes Signed-off-by: Linus Torvalds diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c index 039d913..404f21a 100644 --- a/arch/x86/pci/acpi.c +++ b/arch/x86/pci/acpi.c @@ -43,6 +43,17 @@ static const struct dmi_system_id pci_use_crs_table[] __initconst = { DMI_MATCH(DMI_PRODUCT_NAME, "ALiveSATA2-GLAN"), }, }, + /* https://bugzilla.kernel.org/show_bug.cgi?id=30552 */ + /* 2006 AMD HT/VIA system with two host bridges */ + { + .callback = set_use_crs, + .ident = "ASUS M2V-MX SE", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), + DMI_MATCH(DMI_BOARD_NAME, "M2V-MX SE"), + DMI_MATCH(DMI_BIOS_VENDOR, "American Megatrends Inc."), + }, + }, {} }; -- cgit v0.10.2