From 9394c1c65e61eb6f4c1c99f342b49e451ec337b6 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Mon, 11 Mar 2013 13:52:12 +0100 Subject: ARM: 7669/1: keep __my_cpu_offset consistent with generic one Commit 14318efb(ARM: 7587/1: implement optimized percpu variable access) introduces arm's __my_cpu_offset to optimize percpu vaiable access, which really works well on hackbench, but will cause __my_cpu_offset to return garbage value before it is initialized in cpu_init() called by setup_arch, so accessing percpu variable before setup_arch may cause kernel hang. But generic __my_cpu_offset always returns zero before percpu area is brought up, and won't hang kernel. So the patch tries to clear __my_cpu_offset on boot CPU early to avoid boot hang. At least now percpu variable is accessed by lockdep before setup_arch(), and enabling CONFIG_LOCK_STAT or CONFIG_DEBUG_LOCKDEP can trigger kernel hang. Signed-off-by: Ming Lei Signed-off-by: Russell King diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index 1522c7a..dd1c6aa 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -456,6 +456,13 @@ void __init smp_setup_processor_id(void) for (i = 1; i < nr_cpu_ids; ++i) cpu_logical_map(i) = i == cpu ? 0 : i; + /* + * clear __my_cpu_offset on boot CPU to avoid hang caused by + * using percpu variable early, for example, lockdep will + * access percpu variable inside lock_release + */ + set_my_cpu_offset(0); + printk(KERN_INFO "Booting Linux on physical CPU 0x%x\n", mpidr); } -- cgit v0.10.2 From 049f3e84d34bfc37f25c91715f2e24a90fb8665e Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 29 Apr 2013 16:06:10 +0100 Subject: ARM: 7705/1: use optimized do_div only for EABI In OABI configurations, some uses of the do_div function cause gcc to run out of registers. To work around that, we can force the use of the out-of-line version for configurations that build a OABI kernel. Without this patch, building netx_defconfig results in: net/core/pktgen.c: In function 'pktgen_if_show': net/core/pktgen.c:682:2775: error: can't find a register in class 'GENERAL_REGS' while reloading 'asm' net/core/pktgen.c:682:3153: error: can't find a register in class 'GENERAL_REGS' while reloading 'asm' net/core/pktgen.c:682:2775: error: 'asm' operand has impossible constraints net/core/pktgen.c:682:3153: error: 'asm' operand has impossible constraints Signed-off-by: Arnd Bergmann Signed-off-by: Russell King diff --git a/arch/arm/include/asm/div64.h b/arch/arm/include/asm/div64.h index fe92ccf..191ada6 100644 --- a/arch/arm/include/asm/div64.h +++ b/arch/arm/include/asm/div64.h @@ -46,7 +46,7 @@ __rem; \ }) -#if __GNUC__ < 4 +#if __GNUC__ < 4 || !defined(CONFIG_AEABI) /* * gcc versions earlier than 4.0 are simply too problematic for the -- cgit v0.10.2 From faefd550c45d8d314e8f260f21565320355c947f Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Wed, 15 May 2013 09:39:17 +0100 Subject: ARM: 7722/1: zImage: Convert 32bits memory size and address from ATAG to 64bits DTB When CONFIG_ARM_APPENDED_DTB is selected, if the bootloader provides an ATAG_MEM it replaces the memory size and the memory address in the memory node of the device tree. In the case of a system which can handle more than 4GB, the memory node cell size is 4: each data (memory size and memory address) are 64 bits and then use 2 cells. The current code in atags_to_fdt.c made the assumption of a cell size of 2 (one cell for the memory size and one cell for the memory address), this leads to an improper write of the data and ends with a boot hang. This patch writes the memory size and the memory address on the memory node in the device tree depending of the size of the memory node (32 bits or 64 bits). It has been tested in the 2 cases: - with a dtb using skeleton.dtsi - and with a dtb using skeleton64.dtsi Signed-off-by: Gregory CLEMENT Acked-by: Nicolas Pitre Signed-off-by: Russell King diff --git a/arch/arm/boot/compressed/atags_to_fdt.c b/arch/arm/boot/compressed/atags_to_fdt.c index aabc02a..d1153c8 100644 --- a/arch/arm/boot/compressed/atags_to_fdt.c +++ b/arch/arm/boot/compressed/atags_to_fdt.c @@ -53,6 +53,17 @@ static const void *getprop(const void *fdt, const char *node_path, return fdt_getprop(fdt, offset, property, len); } +static uint32_t get_cell_size(const void *fdt) +{ + int len; + uint32_t cell_size = 1; + const uint32_t *size_len = getprop(fdt, "/", "#size-cells", &len); + + if (size_len) + cell_size = fdt32_to_cpu(*size_len); + return cell_size; +} + static void merge_fdt_bootargs(void *fdt, const char *fdt_cmdline) { char cmdline[COMMAND_LINE_SIZE]; @@ -95,9 +106,11 @@ static void merge_fdt_bootargs(void *fdt, const char *fdt_cmdline) int atags_to_fdt(void *atag_list, void *fdt, int total_space) { struct tag *atag = atag_list; - uint32_t mem_reg_property[2 * NR_BANKS]; + /* In the case of 64 bits memory size, need to reserve 2 cells for + * address and size for each bank */ + uint32_t mem_reg_property[2 * 2 * NR_BANKS]; int memcount = 0; - int ret; + int ret, memsize; /* make sure we've got an aligned pointer */ if ((u32)atag_list & 0x3) @@ -137,8 +150,25 @@ int atags_to_fdt(void *atag_list, void *fdt, int total_space) continue; if (!atag->u.mem.size) continue; - mem_reg_property[memcount++] = cpu_to_fdt32(atag->u.mem.start); - mem_reg_property[memcount++] = cpu_to_fdt32(atag->u.mem.size); + memsize = get_cell_size(fdt); + + if (memsize == 2) { + /* if memsize is 2, that means that + * each data needs 2 cells of 32 bits, + * so the data are 64 bits */ + uint64_t *mem_reg_prop64 = + (uint64_t *)mem_reg_property; + mem_reg_prop64[memcount++] = + cpu_to_fdt64(atag->u.mem.start); + mem_reg_prop64[memcount++] = + cpu_to_fdt64(atag->u.mem.size); + } else { + mem_reg_property[memcount++] = + cpu_to_fdt32(atag->u.mem.start); + mem_reg_property[memcount++] = + cpu_to_fdt32(atag->u.mem.size); + } + } else if (atag->hdr.tag == ATAG_INITRD2) { uint32_t initrd_start, initrd_size; initrd_start = atag->u.initrd.start; @@ -150,8 +180,10 @@ int atags_to_fdt(void *atag_list, void *fdt, int total_space) } } - if (memcount) - setprop(fdt, "/memory", "reg", mem_reg_property, 4*memcount); + if (memcount) { + setprop(fdt, "/memory", "reg", mem_reg_property, + 4 * memcount * memsize); + } return fdt_pack(fdt); } -- cgit v0.10.2 From 3b656fed6ff65d6d268da9ed0760c2a58d125771 Mon Sep 17 00:00:00 2001 From: Christian Daudt Date: Thu, 9 May 2013 22:21:01 +0100 Subject: ARM: 7716/1: bcm281xx: Add L2 support for Rev A2 chips Rev A2 SoCs have an unorthodox memory re-mapping and this needs to be reflected in the cache operations. This patch adds new outer cache functions for the l2x0 driver to support this SoC revision. It also adds a new compatible value for the cache to enable this functionality. Updates from V1: - remove section 1 altogether and note that in comments - simplify section selection caused by section 1 removal - BUG_ON just in case section 1 shows up Signed-off-by: Christian Daudt Reviewed-by: Will Deacon Acked-by: Olof Johansson Signed-off-by: Russell King diff --git a/Documentation/devicetree/bindings/arm/l2cc.txt b/Documentation/devicetree/bindings/arm/l2cc.txt index cbef09b..69ddf9f 100644 --- a/Documentation/devicetree/bindings/arm/l2cc.txt +++ b/Documentation/devicetree/bindings/arm/l2cc.txt @@ -16,6 +16,9 @@ Required properties: performs the same operation). "marvell,"aurora-outer-cache: Marvell Controller designed to be compatible with the ARM one with outer cache mode. + "bcm,bcm11351-a2-pl310-cache": For Broadcom bcm11351 chipset where an + offset needs to be added to the address before passing down to the L2 + cache controller - cache-unified : Specifies the cache is a unified cache. - cache-level : Should be set to 2 for a level 2 cache. - reg : Physical base address and size of cache controller's memory mapped diff --git a/arch/arm/boot/dts/bcm11351.dtsi b/arch/arm/boot/dts/bcm11351.dtsi index 41b2c6c..5e48c85 100644 --- a/arch/arm/boot/dts/bcm11351.dtsi +++ b/arch/arm/boot/dts/bcm11351.dtsi @@ -47,10 +47,10 @@ }; L2: l2-cache { - compatible = "arm,pl310-cache"; - reg = <0x3ff20000 0x1000>; - cache-unified; - cache-level = <2>; + compatible = "bcm,bcm11351-a2-pl310-cache"; + reg = <0x3ff20000 0x1000>; + cache-unified; + cache-level = <2>; }; timer@35006000 { diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c index c465fac..d70e0ab 100644 --- a/arch/arm/mm/cache-l2x0.c +++ b/arch/arm/mm/cache-l2x0.c @@ -523,6 +523,147 @@ static void aurora_flush_range(unsigned long start, unsigned long end) } } +/* + * For certain Broadcom SoCs, depending on the address range, different offsets + * need to be added to the address before passing it to L2 for + * invalidation/clean/flush + * + * Section Address Range Offset EMI + * 1 0x00000000 - 0x3FFFFFFF 0x80000000 VC + * 2 0x40000000 - 0xBFFFFFFF 0x40000000 SYS + * 3 0xC0000000 - 0xFFFFFFFF 0x80000000 VC + * + * When the start and end addresses have crossed two different sections, we + * need to break the L2 operation into two, each within its own section. + * For example, if we need to invalidate addresses starts at 0xBFFF0000 and + * ends at 0xC0001000, we need do invalidate 1) 0xBFFF0000 - 0xBFFFFFFF and 2) + * 0xC0000000 - 0xC0001000 + * + * Note 1: + * By breaking a single L2 operation into two, we may potentially suffer some + * performance hit, but keep in mind the cross section case is very rare + * + * Note 2: + * We do not need to handle the case when the start address is in + * Section 1 and the end address is in Section 3, since it is not a valid use + * case + * + * Note 3: + * Section 1 in practical terms can no longer be used on rev A2. Because of + * that the code does not need to handle section 1 at all. + * + */ +#define BCM_SYS_EMI_START_ADDR 0x40000000UL +#define BCM_VC_EMI_SEC3_START_ADDR 0xC0000000UL + +#define BCM_SYS_EMI_OFFSET 0x40000000UL +#define BCM_VC_EMI_OFFSET 0x80000000UL + +static inline int bcm_addr_is_sys_emi(unsigned long addr) +{ + return (addr >= BCM_SYS_EMI_START_ADDR) && + (addr < BCM_VC_EMI_SEC3_START_ADDR); +} + +static inline unsigned long bcm_l2_phys_addr(unsigned long addr) +{ + if (bcm_addr_is_sys_emi(addr)) + return addr + BCM_SYS_EMI_OFFSET; + else + return addr + BCM_VC_EMI_OFFSET; +} + +static void bcm_inv_range(unsigned long start, unsigned long end) +{ + unsigned long new_start, new_end; + + BUG_ON(start < BCM_SYS_EMI_START_ADDR); + + if (unlikely(end <= start)) + return; + + new_start = bcm_l2_phys_addr(start); + new_end = bcm_l2_phys_addr(end); + + /* normal case, no cross section between start and end */ + if (likely(bcm_addr_is_sys_emi(end) || !bcm_addr_is_sys_emi(start))) { + l2x0_inv_range(new_start, new_end); + return; + } + + /* They cross sections, so it can only be a cross from section + * 2 to section 3 + */ + l2x0_inv_range(new_start, + bcm_l2_phys_addr(BCM_VC_EMI_SEC3_START_ADDR-1)); + l2x0_inv_range(bcm_l2_phys_addr(BCM_VC_EMI_SEC3_START_ADDR), + new_end); +} + +static void bcm_clean_range(unsigned long start, unsigned long end) +{ + unsigned long new_start, new_end; + + BUG_ON(start < BCM_SYS_EMI_START_ADDR); + + if (unlikely(end <= start)) + return; + + if ((end - start) >= l2x0_size) { + l2x0_clean_all(); + return; + } + + new_start = bcm_l2_phys_addr(start); + new_end = bcm_l2_phys_addr(end); + + /* normal case, no cross section between start and end */ + if (likely(bcm_addr_is_sys_emi(end) || !bcm_addr_is_sys_emi(start))) { + l2x0_clean_range(new_start, new_end); + return; + } + + /* They cross sections, so it can only be a cross from section + * 2 to section 3 + */ + l2x0_clean_range(new_start, + bcm_l2_phys_addr(BCM_VC_EMI_SEC3_START_ADDR-1)); + l2x0_clean_range(bcm_l2_phys_addr(BCM_VC_EMI_SEC3_START_ADDR), + new_end); +} + +static void bcm_flush_range(unsigned long start, unsigned long end) +{ + unsigned long new_start, new_end; + + BUG_ON(start < BCM_SYS_EMI_START_ADDR); + + if (unlikely(end <= start)) + return; + + if ((end - start) >= l2x0_size) { + l2x0_flush_all(); + return; + } + + new_start = bcm_l2_phys_addr(start); + new_end = bcm_l2_phys_addr(end); + + /* normal case, no cross section between start and end */ + if (likely(bcm_addr_is_sys_emi(end) || !bcm_addr_is_sys_emi(start))) { + l2x0_flush_range(new_start, new_end); + return; + } + + /* They cross sections, so it can only be a cross from section + * 2 to section 3 + */ + l2x0_flush_range(new_start, + bcm_l2_phys_addr(BCM_VC_EMI_SEC3_START_ADDR-1)); + l2x0_flush_range(bcm_l2_phys_addr(BCM_VC_EMI_SEC3_START_ADDR), + new_end); +} + static void __init l2x0_of_setup(const struct device_node *np, u32 *aux_val, u32 *aux_mask) { @@ -765,6 +906,21 @@ static const struct l2x0_of_data aurora_no_outer_data = { }, }; +static const struct l2x0_of_data bcm_l2x0_data = { + .setup = pl310_of_setup, + .save = pl310_save, + .outer_cache = { + .resume = pl310_resume, + .inv_range = bcm_inv_range, + .clean_range = bcm_clean_range, + .flush_range = bcm_flush_range, + .sync = l2x0_cache_sync, + .flush_all = l2x0_flush_all, + .inv_all = l2x0_inv_all, + .disable = l2x0_disable, + }, +}; + static const struct of_device_id l2x0_ids[] __initconst = { { .compatible = "arm,pl310-cache", .data = (void *)&pl310_data }, { .compatible = "arm,l220-cache", .data = (void *)&l2x0_data }, @@ -773,6 +929,8 @@ static const struct of_device_id l2x0_ids[] __initconst = { .data = (void *)&aurora_no_outer_data}, { .compatible = "marvell,aurora-outer-cache", .data = (void *)&aurora_with_outer_data}, + { .compatible = "bcm,bcm11351-a2-pl310-cache", + .data = (void *)&bcm_l2x0_data}, {} }; -- cgit v0.10.2 From 665ba56fdab1670ef686ec35569aa1de0ec4ef4d Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Mon, 13 May 2013 15:39:17 +0100 Subject: ARM: 7717/1: mmc: mmci: Use devm_clk_get API Converting to devm_clk_get simplifies error handling in probe and we can remove other corresponding calls to clk_put. Signed-off-by: Ulf Hansson Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index f4f3038..d57ce43 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -1362,16 +1362,15 @@ static int mmci_probe(struct amba_device *dev, dev_dbg(mmc_dev(mmc), "designer ID = 0x%02x\n", host->hw_designer); dev_dbg(mmc_dev(mmc), "revision = 0x%01x\n", host->hw_revision); - host->clk = clk_get(&dev->dev, NULL); + host->clk = devm_clk_get(&dev->dev, NULL); if (IS_ERR(host->clk)) { ret = PTR_ERR(host->clk); - host->clk = NULL; goto host_free; } ret = clk_prepare_enable(host->clk); if (ret) - goto clk_free; + goto host_free; host->plat = plat; host->variant = variant; @@ -1576,8 +1575,6 @@ static int mmci_probe(struct amba_device *dev, iounmap(host->base); clk_disable: clk_disable_unprepare(host->clk); - clk_free: - clk_put(host->clk); host_free: mmc_free_host(mmc); rel_regions: @@ -1623,7 +1620,6 @@ static int mmci_remove(struct amba_device *dev) iounmap(host->base); clk_disable_unprepare(host->clk); - clk_put(host->clk); mmc_free_host(mmc); -- cgit v0.10.2 From c58a85090c8fbf7274af5ed0fe30828320c40ad6 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Mon, 13 May 2013 15:40:03 +0100 Subject: ARM: 7718/1: mmc: mmci: Set actual clock for debug purpose Update cclk to the acutal used value and copy it to the actual_clock variable in the mmc host for debug purpose. Signed-off-by: Fredrik Soderstedt Signed-off-by: Ulf Hansson Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index d57ce43..5f53cf8 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -196,6 +196,9 @@ static void mmci_set_clkreg(struct mmci_host *host, unsigned int desired) struct variant_data *variant = host->variant; u32 clk = variant->clkreg; + /* Make sure cclk reflects the current calculated clock */ + host->cclk = 0; + if (desired) { if (desired >= host->mclk) { clk = MCI_CLK_BYPASS; @@ -230,6 +233,9 @@ static void mmci_set_clkreg(struct mmci_host *host, unsigned int desired) /* clk |= MCI_CLK_PWRSAVE; */ } + /* Set actual clock for debug */ + host->mmc->actual_clock = host->cclk; + if (host->mmc->ios.bus_width == MMC_BUS_WIDTH_4) clk |= MCI_4BIT_BUS; if (host->mmc->ios.bus_width == MMC_BUS_WIDTH_8) -- cgit v0.10.2 From 024629c62ffc25f267d57bf588cc10c96ccc0ce5 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Mon, 13 May 2013 15:40:56 +0100 Subject: ARM: 7719/1: mmc: mmci: Support for CMD23 Support added for transmission of CMD23 during multi block read or write. In order to activate this feature, MMC_CAP_CMD23 flag needs to be enabled in the capabilities field. Note that CMD23 support is mandatory to support features like reliable write, data tag, context ID, packed command. This patch is based upon a patch from Saugata Das. Signed-off-by: Ulf Hansson Acked-by: Linus Walleij Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index 5f53cf8..9df8b84 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -848,7 +848,7 @@ mmci_data_irq(struct mmci_host *host, struct mmc_data *data, /* The error clause is handled above, success! */ data->bytes_xfered = data->blksz * data->blocks; - if (!data->stop) { + if (!data->stop || host->mrq->sbc) { mmci_request_end(host, data->mrq); } else { mmci_start_command(host, data->stop, 0); @@ -861,6 +861,7 @@ mmci_cmd_irq(struct mmci_host *host, struct mmc_command *cmd, unsigned int status) { void __iomem *base = host->base; + bool sbc = (cmd == host->mrq->sbc); host->cmd = NULL; @@ -875,7 +876,7 @@ mmci_cmd_irq(struct mmci_host *host, struct mmc_command *cmd, cmd->resp[3] = readl(base + MMCIRESPONSE3); } - if (!cmd->data || cmd->error) { + if ((!sbc && !cmd->data) || cmd->error) { if (host->data) { /* Terminate the DMA transfer */ if (dma_inprogress(host)) { @@ -884,7 +885,9 @@ mmci_cmd_irq(struct mmci_host *host, struct mmc_command *cmd, } mmci_stop_data(host); } - mmci_request_end(host, cmd->mrq); + mmci_request_end(host, host->mrq); + } else if (sbc) { + mmci_start_command(host, host->mrq->cmd, 0); } else if (!(cmd->data->flags & MMC_DATA_READ)) { mmci_start_data(host, cmd->data); } @@ -1125,7 +1128,10 @@ static void mmci_request(struct mmc_host *mmc, struct mmc_request *mrq) if (mrq->data && mrq->data->flags & MMC_DATA_READ) mmci_start_data(host, mrq->data); - mmci_start_command(host, mrq->cmd, 0); + if (mrq->sbc) + mmci_start_command(host, mrq->sbc, 0); + else + mmci_start_command(host, mrq->cmd, 0); spin_unlock_irqrestore(&host->lock, flags); } -- cgit v0.10.2 From 1fd83f0ecf87e33ab560e8229842cf10f91552ee Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Fri, 3 May 2013 12:51:17 +0100 Subject: ARM: 7713/1: mmc: mmci: Allow MMCI to request channels with information acquired from DT Currently, if DMA information isn't passed from platform data, then DMA will not be used. This patch allows DMA information obtained though Device Tree to be used as well. Cc: Chris Ball Cc: linux-mmc@vger.kernel.org Signed-off-by: Lee Jones Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index 9df8b84..c6d8b62 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -310,10 +310,8 @@ static void mmci_dma_setup(struct mmci_host *host) const char *rxname, *txname; dma_cap_mask_t mask; - if (!plat || !plat->dma_filter) { - dev_info(mmc_dev(host->mmc), "no DMA platform data\n"); - return; - } + host->dma_rx_channel = dma_request_slave_channel(mmc_dev(host->mmc), "rx"); + host->dma_tx_channel = dma_request_slave_channel(mmc_dev(host->mmc), "tx"); /* initialize pre request cookie */ host->next_data.cookie = 1; @@ -322,30 +320,33 @@ static void mmci_dma_setup(struct mmci_host *host) dma_cap_zero(mask); dma_cap_set(DMA_SLAVE, mask); - /* - * If only an RX channel is specified, the driver will - * attempt to use it bidirectionally, however if it is - * is specified but cannot be located, DMA will be disabled. - */ - if (plat->dma_rx_param) { - host->dma_rx_channel = dma_request_channel(mask, + if (plat && plat->dma_filter) { + if (!host->dma_rx_channel && plat->dma_rx_param) { + host->dma_rx_channel = dma_request_channel(mask, plat->dma_filter, plat->dma_rx_param); - /* E.g if no DMA hardware is present */ - if (!host->dma_rx_channel) - dev_err(mmc_dev(host->mmc), "no RX DMA channel\n"); - } + /* E.g if no DMA hardware is present */ + if (!host->dma_rx_channel) + dev_err(mmc_dev(host->mmc), "no RX DMA channel\n"); + } - if (plat->dma_tx_param) { - host->dma_tx_channel = dma_request_channel(mask, + if (!host->dma_tx_channel && plat->dma_tx_param) { + host->dma_tx_channel = dma_request_channel(mask, plat->dma_filter, plat->dma_tx_param); - if (!host->dma_tx_channel) - dev_warn(mmc_dev(host->mmc), "no TX DMA channel\n"); - } else { - host->dma_tx_channel = host->dma_rx_channel; + if (!host->dma_tx_channel) + dev_warn(mmc_dev(host->mmc), "no TX DMA channel\n"); + } } + /* + * If only an RX channel is specified, the driver will + * attempt to use it bidirectionally, however if it is + * is specified but cannot be located, DMA will be disabled. + */ + if (host->dma_rx_channel && !host->dma_tx_channel) + host->dma_tx_channel = host->dma_rx_channel; + if (host->dma_rx_channel) rxname = dma_chan_name(host->dma_rx_channel); else -- cgit v0.10.2 From 7c0136ef773c206e242b9718740377a45747bd70 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Tue, 14 May 2013 13:53:10 +0100 Subject: ARM: 7721/1: mmc: mmci: Fixup regulator handling for vqmmc We can not rely on regulator_is_enabled to decide whether to enable|disable the regulator. It would mean that the reference counter for it is not balanced properly. Instead keep track of our internal state by using a new flag in the host struct, so we can take correct decisions. Signed-off-by: Ulf Hansson Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index c6d8b62..a8bbdd1 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -1156,9 +1156,10 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios) if (!IS_ERR(mmc->supply.vmmc)) mmc_regulator_set_ocr(mmc, mmc->supply.vmmc, 0); - if (!IS_ERR(mmc->supply.vqmmc) && - regulator_is_enabled(mmc->supply.vqmmc)) + if (!IS_ERR(mmc->supply.vqmmc) && host->vqmmc_enabled) { regulator_disable(mmc->supply.vqmmc); + host->vqmmc_enabled = false; + } break; case MMC_POWER_UP: @@ -1174,12 +1175,13 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios) break; case MMC_POWER_ON: - if (!IS_ERR(mmc->supply.vqmmc) && - !regulator_is_enabled(mmc->supply.vqmmc)) { + if (!IS_ERR(mmc->supply.vqmmc) && !host->vqmmc_enabled) { ret = regulator_enable(mmc->supply.vqmmc); if (ret < 0) dev_err(mmc_dev(mmc), "failed to enable vqmmc regulator\n"); + else + host->vqmmc_enabled = true; } pwr |= MCI_PWR_ON; diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h index 1f33ad5..1383c9c 100644 --- a/drivers/mmc/host/mmci.h +++ b/drivers/mmc/host/mmci.h @@ -183,6 +183,7 @@ struct mmci_host { unsigned int cclk; u32 pwr_reg; u32 clk_reg; + bool vqmmc_enabled; struct mmci_platform_data *plat; struct variant_data *variant; -- cgit v0.10.2 From 0f3ed7f75cf1a16df9309f3a9ffaf62a3fc1f0bb Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 15 May 2013 20:47:33 +0100 Subject: ARM: 7724/1: mmc: mmci: Support signal voltage switch for UHS cards Add .start_signal_voltage_switch callback to be able to support UHS cards. The voltage switch requires the optional vqmmc regulator to exist since the actual voltage switch will be performed directly on it. Signed-off-by: Ulf Hansson Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index a8bbdd1..cb9e562 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -1266,6 +1266,39 @@ static int mmci_get_cd(struct mmc_host *mmc) return status; } +static int mmci_sig_volt_switch(struct mmc_host *mmc, struct mmc_ios *ios) +{ + int ret = 0; + + if (!IS_ERR(mmc->supply.vqmmc)) { + + pm_runtime_get_sync(mmc_dev(mmc)); + + switch (ios->signal_voltage) { + case MMC_SIGNAL_VOLTAGE_330: + ret = regulator_set_voltage(mmc->supply.vqmmc, + 2700000, 3600000); + break; + case MMC_SIGNAL_VOLTAGE_180: + ret = regulator_set_voltage(mmc->supply.vqmmc, + 1700000, 1950000); + break; + case MMC_SIGNAL_VOLTAGE_120: + ret = regulator_set_voltage(mmc->supply.vqmmc, + 1100000, 1300000); + break; + } + + if (ret) + dev_warn(mmc_dev(mmc), "Voltage switch failed\n"); + + pm_runtime_mark_last_busy(mmc_dev(mmc)); + pm_runtime_put_autosuspend(mmc_dev(mmc)); + } + + return ret; +} + static irqreturn_t mmci_cd_irq(int irq, void *dev_id) { struct mmci_host *host = dev_id; @@ -1282,6 +1315,7 @@ static const struct mmc_host_ops mmci_ops = { .set_ios = mmci_set_ios, .get_ro = mmci_get_ro, .get_cd = mmci_get_cd, + .start_signal_voltage_switch = mmci_sig_volt_switch, }; #ifdef CONFIG_OF -- cgit v0.10.2 From 9cc639a20fdc0b935e55d4992f93963f95233ca4 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 15 May 2013 20:48:23 +0100 Subject: ARM: 7725/1: mmc: mmci: Cache MMCIDATACTRL register Add a cache variable in the host struct that reflects the current data in the MMCIDATACTRL register. This patch will not introduce any functional change but instead provide an easy option to keep specific bits in the register between each data transfer. Signed-off-by: Ulf Hansson Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index cb9e562..ccfe0bc 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -191,6 +191,17 @@ static void mmci_write_pwrreg(struct mmci_host *host, u32 pwr) /* * This must be called with host->lock held */ +static void mmci_write_datactrlreg(struct mmci_host *host, u32 datactrl) +{ + if (host->datactrl_reg != datactrl) { + host->datactrl_reg = datactrl; + writel(datactrl, host->base + MMCIDATACTRL); + } +} + +/* + * This must be called with host->lock held + */ static void mmci_set_clkreg(struct mmci_host *host, unsigned int desired) { struct variant_data *variant = host->variant; @@ -281,7 +292,7 @@ static void mmci_set_mask1(struct mmci_host *host, unsigned int mask) static void mmci_stop_data(struct mmci_host *host) { - writel(0, host->base + MMCIDATACTRL); + mmci_write_datactrlreg(host, 0); mmci_set_mask1(host, 0); host->data = NULL; } @@ -559,7 +570,7 @@ static int mmci_dma_start_data(struct mmci_host *host, unsigned int datactrl) datactrl |= MCI_DPSM_DMAENABLE; /* Trigger the DMA transfer */ - writel(datactrl, host->base + MMCIDATACTRL); + mmci_write_datactrlreg(host, datactrl); /* * Let the MMCI say when the data is ended and it's time @@ -757,7 +768,7 @@ static void mmci_start_data(struct mmci_host *host, struct mmc_data *data) irqmask = MCI_TXFIFOHALFEMPTYMASK; } - writel(datactrl, base + MMCIDATACTRL); + mmci_write_datactrlreg(host, datactrl); writel(readl(base + MMCIMASK0) & ~MCI_DATAENDMASK, base + MMCIMASK0); mmci_set_mask1(host, irqmask); } diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h index 1383c9c..0b6cc54 100644 --- a/drivers/mmc/host/mmci.h +++ b/drivers/mmc/host/mmci.h @@ -183,6 +183,7 @@ struct mmci_host { unsigned int cclk; u32 pwr_reg; u32 clk_reg; + u32 datactrl_reg; bool vqmmc_enabled; struct mmci_platform_data *plat; struct variant_data *variant; -- cgit v0.10.2 From 0125962000777cd1e2ce53deefb99779d5ee5199 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 15 May 2013 20:53:22 +0100 Subject: ARM: 7726/1: mmc: mmci: Add card_busy function to improve UHS card support To verify a signal voltage switch at initialization of UHS cards the .card_busy callback is used. For some of the ST-variants, card busy detection on the DAT0 pin is supported. We extend the variant struct with a busy_detect flag to indicate support for it. A corresponding busy detect function, which polls the busy status bit, is then set to the .card_busy callback. Signed-off-by: Ulf Hansson Signed-off-by: Russell King diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index ccfe0bc..c3785ed 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -61,6 +61,7 @@ static unsigned int fmax = 515633; * @pwrreg_powerup: power up value for MMCIPOWER register * @signal_direction: input/out direction of bus signals can be indicated * @pwrreg_clkgate: MMCIPOWER register must be used to gate the clock + * @busy_detect: true if busy detection on dat0 is supported */ struct variant_data { unsigned int clkreg; @@ -74,6 +75,7 @@ struct variant_data { u32 pwrreg_powerup; bool signal_direction; bool pwrreg_clkgate; + bool busy_detect; }; static struct variant_data variant_arm = { @@ -132,6 +134,7 @@ static struct variant_data variant_ux500 = { .pwrreg_powerup = MCI_PWR_ON, .signal_direction = true, .pwrreg_clkgate = true, + .busy_detect = true, }; static struct variant_data variant_ux500v2 = { @@ -146,8 +149,28 @@ static struct variant_data variant_ux500v2 = { .pwrreg_powerup = MCI_PWR_ON, .signal_direction = true, .pwrreg_clkgate = true, + .busy_detect = true, }; +static int mmci_card_busy(struct mmc_host *mmc) +{ + struct mmci_host *host = mmc_priv(mmc); + unsigned long flags; + int busy = 0; + + pm_runtime_get_sync(mmc_dev(mmc)); + + spin_lock_irqsave(&host->lock, flags); + if (readl(host->base + MMCISTATUS) & MCI_ST_CARDBUSY) + busy = 1; + spin_unlock_irqrestore(&host->lock, flags); + + pm_runtime_mark_last_busy(mmc_dev(mmc)); + pm_runtime_put_autosuspend(mmc_dev(mmc)); + + return busy; +} + /* * Validate mmc prerequisites */ @@ -193,6 +216,9 @@ static void mmci_write_pwrreg(struct mmci_host *host, u32 pwr) */ static void mmci_write_datactrlreg(struct mmci_host *host, u32 datactrl) { + /* Keep ST Micro busy mode if enabled */ + datactrl |= host->datactrl_reg & MCI_ST_DPSM_BUSYMODE; + if (host->datactrl_reg != datactrl) { host->datactrl_reg = datactrl; writel(datactrl, host->base + MMCIDATACTRL); @@ -1319,7 +1345,7 @@ static irqreturn_t mmci_cd_irq(int irq, void *dev_id) return IRQ_HANDLED; } -static const struct mmc_host_ops mmci_ops = { +static struct mmc_host_ops mmci_ops = { .request = mmci_request, .pre_req = mmci_pre_request, .post_req = mmci_post_request, @@ -1455,6 +1481,11 @@ static int mmci_probe(struct amba_device *dev, goto clk_disable; } + if (variant->busy_detect) { + mmci_ops.card_busy = mmci_card_busy; + mmci_write_datactrlreg(host, MCI_ST_DPSM_BUSYMODE); + } + mmc->ops = &mmci_ops; /* * The ARM and ST versions of the block have slightly different diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h index 0b6cc54..69080fa 100644 --- a/drivers/mmc/host/mmci.h +++ b/drivers/mmc/host/mmci.h @@ -94,6 +94,7 @@ /* Extended status bits for the ST Micro variants */ #define MCI_ST_SDIOIT (1 << 22) #define MCI_ST_CEATAEND (1 << 23) +#define MCI_ST_CARDBUSY (1 << 24) #define MMCICLEAR 0x038 #define MCI_CMDCRCFAILCLR (1 << 0) @@ -110,6 +111,7 @@ /* Extended status bits for the ST Micro variants */ #define MCI_ST_SDIOITC (1 << 22) #define MCI_ST_CEATAENDC (1 << 23) +#define MCI_ST_BUSYENDC (1 << 24) #define MMCIMASK0 0x03c #define MCI_CMDCRCFAILMASK (1 << 0) -- cgit v0.10.2 From 9b97173e785a54c5df0aa23d1e1f680f61e36e43 Mon Sep 17 00:00:00 2001 From: Laura Abbott Date: Thu, 16 May 2013 19:40:22 +0100 Subject: ARM: 7728/1: mm: Use phys_addr_t properly for ioremap functions Several of the ioremap functions use unsigned long in places resulting in truncation if physical addresses greater than 4G are passed in. Change the types of the functions and the callers accordingly. Cc: Krzysztof Halasa Cc: Arnd Bergmann Cc: Stephen Boyd Signed-off-by: Laura Abbott Signed-off-by: Russell King diff --git a/arch/arm/include/asm/io.h b/arch/arm/include/asm/io.h index 652b560..d070741 100644 --- a/arch/arm/include/asm/io.h +++ b/arch/arm/include/asm/io.h @@ -130,16 +130,16 @@ static inline u32 __raw_readl(const volatile void __iomem *addr) */ extern void __iomem *__arm_ioremap_pfn_caller(unsigned long, unsigned long, size_t, unsigned int, void *); -extern void __iomem *__arm_ioremap_caller(unsigned long, size_t, unsigned int, +extern void __iomem *__arm_ioremap_caller(phys_addr_t, size_t, unsigned int, void *); extern void __iomem *__arm_ioremap_pfn(unsigned long, unsigned long, size_t, unsigned int); -extern void __iomem *__arm_ioremap(unsigned long, size_t, unsigned int); -extern void __iomem *__arm_ioremap_exec(unsigned long, size_t, bool cached); +extern void __iomem *__arm_ioremap(phys_addr_t, size_t, unsigned int); +extern void __iomem *__arm_ioremap_exec(phys_addr_t, size_t, bool cached); extern void __iounmap(volatile void __iomem *addr); extern void __arm_iounmap(volatile void __iomem *addr); -extern void __iomem * (*arch_ioremap_caller)(unsigned long, size_t, +extern void __iomem * (*arch_ioremap_caller)(phys_addr_t, size_t, unsigned int, void *); extern void (*arch_iounmap)(volatile void __iomem *); diff --git a/arch/arm/mach-ebsa110/core.c b/arch/arm/mach-ebsa110/core.c index b13cc74..8a53f34 100644 --- a/arch/arm/mach-ebsa110/core.c +++ b/arch/arm/mach-ebsa110/core.c @@ -116,7 +116,7 @@ static void __init ebsa110_map_io(void) iotable_init(ebsa110_io_desc, ARRAY_SIZE(ebsa110_io_desc)); } -static void __iomem *ebsa110_ioremap_caller(unsigned long cookie, size_t size, +static void __iomem *ebsa110_ioremap_caller(phys_addr_t cookie, size_t size, unsigned int flags, void *caller) { return (void __iomem *)cookie; diff --git a/arch/arm/mach-imx/mm-imx3.c b/arch/arm/mach-imx/mm-imx3.c index e0e69a6..eed32ca 100644 --- a/arch/arm/mach-imx/mm-imx3.c +++ b/arch/arm/mach-imx/mm-imx3.c @@ -65,7 +65,7 @@ static void imx3_idle(void) : "=r" (reg)); } -static void __iomem *imx3_ioremap_caller(unsigned long phys_addr, size_t size, +static void __iomem *imx3_ioremap_caller(phys_addr_t phys_addr, size_t size, unsigned int mtype, void *caller) { if (mtype == MT_DEVICE) { diff --git a/arch/arm/mach-iop13xx/io.c b/arch/arm/mach-iop13xx/io.c index 183dc8b..faaf7d4 100644 --- a/arch/arm/mach-iop13xx/io.c +++ b/arch/arm/mach-iop13xx/io.c @@ -23,7 +23,7 @@ #include "pci.h" -static void __iomem *__iop13xx_ioremap_caller(unsigned long cookie, +static void __iomem *__iop13xx_ioremap_caller(phys_addr_t cookie, size_t size, unsigned int mtype, void *caller) { void __iomem * retval; diff --git a/arch/arm/mach-ixp4xx/common.c b/arch/arm/mach-ixp4xx/common.c index 6600cff..d7223b3 100644 --- a/arch/arm/mach-ixp4xx/common.c +++ b/arch/arm/mach-ixp4xx/common.c @@ -559,7 +559,7 @@ void ixp4xx_restart(char mode, const char *cmd) * fallback to the default. */ -static void __iomem *ixp4xx_ioremap_caller(unsigned long addr, size_t size, +static void __iomem *ixp4xx_ioremap_caller(phys_addr_t addr, size_t size, unsigned int mtype, void *caller) { if (!is_pci_memory(addr)) diff --git a/arch/arm/mach-msm/common.h b/arch/arm/mach-msm/common.h index ce8215a..421cf77 100644 --- a/arch/arm/mach-msm/common.h +++ b/arch/arm/mach-msm/common.h @@ -23,7 +23,7 @@ extern void msm_map_msm8x60_io(void); extern void msm_map_msm8960_io(void); extern void msm_map_qsd8x50_io(void); -extern void __iomem *__msm_ioremap_caller(unsigned long phys_addr, size_t size, +extern void __iomem *__msm_ioremap_caller(phys_addr_t phys_addr, size_t size, unsigned int mtype, void *caller); extern struct smp_operations msm_smp_ops; diff --git a/arch/arm/mach-msm/io.c b/arch/arm/mach-msm/io.c index 123ef9c..fd65b6d 100644 --- a/arch/arm/mach-msm/io.c +++ b/arch/arm/mach-msm/io.c @@ -172,7 +172,7 @@ void __init msm_map_msm7x30_io(void) } #endif /* CONFIG_ARCH_MSM7X30 */ -void __iomem *__msm_ioremap_caller(unsigned long phys_addr, size_t size, +void __iomem *__msm_ioremap_caller(phys_addr_t phys_addr, size_t size, unsigned int mtype, void *caller) { if (mtype == MT_DEVICE) { diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c index 04d9006..f123d6e 100644 --- a/arch/arm/mm/ioremap.c +++ b/arch/arm/mm/ioremap.c @@ -331,10 +331,10 @@ void __iomem * __arm_ioremap_pfn_caller(unsigned long pfn, return (void __iomem *) (offset + addr); } -void __iomem *__arm_ioremap_caller(unsigned long phys_addr, size_t size, +void __iomem *__arm_ioremap_caller(phys_addr_t phys_addr, size_t size, unsigned int mtype, void *caller) { - unsigned long last_addr; + phys_addr_t last_addr; unsigned long offset = phys_addr & ~PAGE_MASK; unsigned long pfn = __phys_to_pfn(phys_addr); @@ -367,12 +367,12 @@ __arm_ioremap_pfn(unsigned long pfn, unsigned long offset, size_t size, } EXPORT_SYMBOL(__arm_ioremap_pfn); -void __iomem * (*arch_ioremap_caller)(unsigned long, size_t, +void __iomem * (*arch_ioremap_caller)(phys_addr_t, size_t, unsigned int, void *) = __arm_ioremap_caller; void __iomem * -__arm_ioremap(unsigned long phys_addr, size_t size, unsigned int mtype) +__arm_ioremap(phys_addr_t phys_addr, size_t size, unsigned int mtype) { return arch_ioremap_caller(phys_addr, size, mtype, __builtin_return_address(0)); @@ -387,7 +387,7 @@ EXPORT_SYMBOL(__arm_ioremap); * CONFIG_GENERIC_ALLOCATOR for allocating external memory. */ void __iomem * -__arm_ioremap_exec(unsigned long phys_addr, size_t size, bool cached) +__arm_ioremap_exec(phys_addr_t phys_addr, size_t size, bool cached) { unsigned int mtype; diff --git a/arch/arm/mm/nommu.c b/arch/arm/mm/nommu.c index d51225f..31bd77b 100644 --- a/arch/arm/mm/nommu.c +++ b/arch/arm/mm/nommu.c @@ -81,16 +81,16 @@ void __iomem *__arm_ioremap_pfn_caller(unsigned long pfn, unsigned long offset, return __arm_ioremap_pfn(pfn, offset, size, mtype); } -void __iomem *__arm_ioremap(unsigned long phys_addr, size_t size, +void __iomem *__arm_ioremap(phys_addr_t phys_addr, size_t size, unsigned int mtype) { return (void __iomem *)phys_addr; } EXPORT_SYMBOL(__arm_ioremap); -void __iomem * (*arch_ioremap_caller)(unsigned long, size_t, unsigned int, void *); +void __iomem * (*arch_ioremap_caller)(phys_addr_t, size_t, unsigned int, void *); -void __iomem *__arm_ioremap_caller(unsigned long phys_addr, size_t size, +void __iomem *__arm_ioremap_caller(phys_addr_t phys_addr, size_t size, unsigned int mtype, void *caller) { return __arm_ioremap(phys_addr, size, mtype); -- cgit v0.10.2 From b2a234ed6417cf88bdbf4b22700e0ae6e08653ce Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Sat, 18 May 2013 11:21:36 +0100 Subject: ARM: 7730/1: DMA-mapping: mark all !DMA_TO_DEVICE pages in unmapping as clean It is common for one sg to include many pages, so mark all these pages as clean to avoid unnecessary flushing on them in set_pte_at() or update_mmu_cache(). The patch might improve loading performance of applciation code a bit. On the below test code to read file(~1GByte size) from usb mass storage disk to buffer created with mmap(PROT_READ | PROT_EXEC) on Pandaboard, average ~1% improvement can be observed with the patch on 10 times test. unsigned int sum = 0; static unsigned long tv_diff(struct timeval *tv1, struct timeval *tv2) { return (tv2->tv_sec - tv1->tv_sec) * 1000000 + (tv2->tv_usec - tv1->tv_usec); } int main(int argc, char *argv[]) { char *mbuffer; int fd; int i; unsigned long page_size, size; struct stat stat; struct timeval t1, t2; page_size = getpagesize(); fd = open(argv[1], O_RDONLY); assert(fd >= 0); fstat(fd, &stat); size = stat.st_size; printf("%s: file %s, file size %lu, page size %lun", argv[0], read_filename, size, page_size); gettimeofday(&t1, NULL); mbuffer = mmap(NULL, size, PROT_READ | PROT_EXEC, MAP_SHARED, fd, 0); for (i = 0 ; i < size ; i += page_size) sum += mbuffer[i]; munmap(mbuffer, page_size); gettimeofday(&t2, NULL); printf("tread mmaped time: %luusn", tv_diff(&t1, &t2)); close(fd); } Acked-by: Nicolas Pitre Cc: Catalin Marinas Cc: Marek Szyprowski Signed-off-by: Ming Lei Signed-off-by: Russell King diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index ef3e0f3..c038ec0 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -880,10 +880,24 @@ static void __dma_page_dev_to_cpu(struct page *page, unsigned long off, dma_cache_maint_page(page, off, size, dir, dmac_unmap_area); /* - * Mark the D-cache clean for this page to avoid extra flushing. + * Mark the D-cache clean for these pages to avoid extra flushing. */ - if (dir != DMA_TO_DEVICE && off == 0 && size >= PAGE_SIZE) - set_bit(PG_dcache_clean, &page->flags); + if (dir != DMA_TO_DEVICE && size >= PAGE_SIZE) { + unsigned long pfn; + size_t left = size; + + pfn = page_to_pfn(page) + off / PAGE_SIZE; + off %= PAGE_SIZE; + if (off) { + pfn++; + left -= PAGE_SIZE - off; + } + while (left >= PAGE_SIZE) { + page = pfn_to_page(pfn++); + set_bit(PG_dcache_clean, &page->flags); + left -= PAGE_SIZE; + } + } } /** -- cgit v0.10.2 From 81f28946a8b066d484ca8935ff545d94ce730aa3 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 5 Jun 2013 02:44:00 +0100 Subject: ARM: 7746/1: mm: lazy cache flushing on non-mapped pages Currently flush_dcache_page() thinks pages as non-mapped if mapping_mapped(mapping) return false. This approach is very coase: - mmap on part of file may cause all pages backed on the file being thought as mmaped - file-backed pages aren't mapped into user space actually if the memory mmaped on the file isn't accessed This patch uses page_mapped() to decide if the page has been mapped. From the attached test code, I find there is much performance improvement(>25%) when accessing page caches via read under this situations, so memcpy benefits a lot from not flushing cache under this situation. No. read time without the patch No. read time with the patch ================================================================ No. 0, time 22615636 us No. 0, time 22014717 us No. 1, time 4387851 us No. 1, time 3113184 us No. 2, time 4276535 us No. 2, time 3005244 us No. 3, time 4259821 us No. 3, time 3001565 us No. 4, time 4263811 us No. 4, time 3002748 us No. 5, time 4258486 us No. 5, time 3004104 us No. 6, time 4253009 us No. 6, time 3002188 us No. 7, time 4262809 us No. 7, time 2998196 us No. 8, time 4264525 us No. 8, time 3007255 us No. 9, time 4267795 us No. 9, time 3005094 us 1), No.0. is to read the file from storage device, and others are to read the file from page caches basically. 2), file size is 512M, and is on ext4 over usb mass storage. 3), the test is done on Pandaboard. unsigned int sum = 0; unsigned long sum_val = 0; static unsigned long tv_diff(struct timeval *tv1, struct timeval *tv2) { return (tv2->tv_sec - tv1->tv_sec) * 1000000 + (tv2->tv_usec - tv1->tv_usec); } int main(int argc, char *argv[]) { char *mbuf, fbuf; int fd; int i; unsigned long page_size, size; struct stat stat; struct timeval t1, t2; unsigned char *rbuf = malloc(32 * page_size); if (!rbuf) { printf(" %sn", "malloc failed"); exit(-1); } page_size = getpagesize(); fd = open(argv[1], O_RDWR); assert(fd >= 0); fstat(fd, &stat); size = stat.st_size; printf("%s: file %s, size %lu, page size %lun", argv[0], argv[1], size, page_size); gettimeofday(&t1, NULL); mbuf = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); if (!mbuf) { printf(" %sn", "mmap failed"); exit(-1); } for (i = 0 ; i < size ; i += (page_size * 32)) { int rcnt; lseek(fd, i, SEEK_SET); rcnt = read(fd, rbuf, page_size * 32); if (rcnt != page_size * 32) { printf("%s: read faildn", __func__); exit(-1); } } free(rbuf); munmap(mbuf, size); gettimeofday(&t2, NULL); printf("tread mmaped time: %luusn", tv_diff(&t1, &t2)); close(fd); } Cc: Michel Lespinasse Cc: Andrew Morton Cc: Nicolas Pitre Reviewed-by: Will Deacon Acked-by: Catalin Marinas Signed-off-by: Ming Lei Signed-off-by: Russell King diff --git a/arch/arm/mm/flush.c b/arch/arm/mm/flush.c index 0d473cc..2ff66eb 100644 --- a/arch/arm/mm/flush.c +++ b/arch/arm/mm/flush.c @@ -287,7 +287,7 @@ void flush_dcache_page(struct page *page) mapping = page_mapping(page); if (!cache_ops_need_broadcast() && - mapping && !mapping_mapped(mapping)) + mapping && !page_mapped(page)) clear_bit(PG_dcache_clean, &page->flags); else { __flush_dcache_page(mapping, page); -- cgit v0.10.2 From 0305ffd7536270befc3e772bb978a6e3c6d37150 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Fri, 31 May 2013 05:17:29 +0100 Subject: ARM: 7740/1: MCPM: prettify debug output The CPU and cluster numbers (MPIDR affinity levels 0 and 1) may have at most 8 bits each. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King diff --git a/arch/arm/common/mcpm_head.S b/arch/arm/common/mcpm_head.S index 8178705..80f033614 100644 --- a/arch/arm/common/mcpm_head.S +++ b/arch/arm/common/mcpm_head.S @@ -32,11 +32,11 @@ 1901: adr r0, 1902b bl printascii mov r0, r9 - bl printhex8 + bl printhex2 adr r0, 1903b bl printascii mov r0, r10 - bl printhex8 + bl printhex2 adr r0, 1904b bl printascii #endif -- cgit v0.10.2 From 6c93dd438aad39c60630dbe47fde8193c845f8b8 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Fri, 31 May 2013 05:24:03 +0100 Subject: ARM: 7741/1: mcpm_platsmp.c: remove empty smp_init_cpus method The smp_init_cpus method in the smp_operations structure is optional and can be omitted entirely. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King diff --git a/arch/arm/common/mcpm_platsmp.c b/arch/arm/common/mcpm_platsmp.c index 3caed0d..510e5b1 100644 --- a/arch/arm/common/mcpm_platsmp.c +++ b/arch/arm/common/mcpm_platsmp.c @@ -19,10 +19,6 @@ #include #include -static void __init simple_smp_init_cpus(void) -{ -} - static int __cpuinit mcpm_boot_secondary(unsigned int cpu, struct task_struct *idle) { unsigned int mpidr, pcpu, pcluster, ret; @@ -74,7 +70,6 @@ static void mcpm_cpu_die(unsigned int cpu) #endif static struct smp_operations __initdata mcpm_smp_ops = { - .smp_init_cpus = simple_smp_init_cpus, .smp_boot_secondary = mcpm_boot_secondary, .smp_secondary_init = mcpm_secondary_init, #ifdef CONFIG_HOTPLUG_CPU -- cgit v0.10.2 From 1aa2b3b7a6c4f3dbd3671171113a20e6a6190e3b Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 5 Jun 2013 11:25:13 +0100 Subject: ARM: 7748/1: oabi: handle faults when loading swi instruction from userspace Running an OABI_COMPAT kernel on an SMP platform can lead to fun and games with page aging. If one CPU issues a swi instruction immediately before another CPU decides to mkold the page containing the swi instruction, then we will fault attempting to load the instruction during the vector_swi handler in order to retrieve its immediate field. Since this fault is not currently dealt with by our exception tables, this results in a panic: Unable to handle kernel paging request at virtual address 4020841c pgd = c490c000 [4020841c] *pgd=84451831, *pte=bf05859d, *ppte=00000000 Internal error: Oops: 17 [#1] PREEMPT SMP ARM Modules linked in: hid_sony(O) CPU: 1 Tainted: G W O (3.4.0-perf-gf496dca-01162-gcbcc62b #1) PC is at vector_swi+0x28/0x88 LR is at 0x40208420 This patch wraps all of the swi instruction loads with the USER macro and provides a shared exception table entry which simply rewinds the saved user PC and returns from the system call (without setting tbl, so there's no worries with tracing or syscall restarting). Returning to userspace will re-enter the page fault handler, from where we will probably send SIGSEGV to the current task. Reported-by: Wang, Yalin Reviewed-by: Nicolas Pitre Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index bc5bc0a..4bc816a 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -362,6 +362,16 @@ ENTRY(vector_swi) str r0, [sp, #S_OLD_R0] @ Save OLD_R0 zero_fp +#ifdef CONFIG_ALIGNMENT_TRAP + ldr ip, __cr_alignment + ldr ip, [ip] + mcr p15, 0, ip, c1, c0 @ update control register +#endif + + enable_irq + ct_user_exit + get_thread_info tsk + /* * Get the system call number. */ @@ -375,9 +385,9 @@ ENTRY(vector_swi) #ifdef CONFIG_ARM_THUMB tst r8, #PSR_T_BIT movne r10, #0 @ no thumb OABI emulation - ldreq r10, [lr, #-4] @ get SWI instruction + USER( ldreq r10, [lr, #-4] ) @ get SWI instruction #else - ldr r10, [lr, #-4] @ get SWI instruction + USER( ldr r10, [lr, #-4] ) @ get SWI instruction #endif #ifdef CONFIG_CPU_ENDIAN_BE8 rev r10, r10 @ little endian instruction @@ -392,22 +402,13 @@ ENTRY(vector_swi) /* Legacy ABI only, possibly thumb mode. */ tst r8, #PSR_T_BIT @ this is SPSR from save_user_regs addne scno, r7, #__NR_SYSCALL_BASE @ put OS number in - ldreq scno, [lr, #-4] + USER( ldreq scno, [lr, #-4] ) #else /* Legacy ABI only. */ - ldr scno, [lr, #-4] @ get SWI instruction + USER( ldr scno, [lr, #-4] ) @ get SWI instruction #endif -#ifdef CONFIG_ALIGNMENT_TRAP - ldr ip, __cr_alignment - ldr ip, [ip] - mcr p15, 0, ip, c1, c0 @ update control register -#endif - enable_irq - ct_user_exit - - get_thread_info tsk adr tbl, sys_call_table @ load syscall table pointer #if defined(CONFIG_OABI_COMPAT) @@ -442,6 +443,21 @@ local_restart: eor r0, scno, #__NR_SYSCALL_BASE @ put OS number back bcs arm_syscall b sys_ni_syscall @ not private func + +#if defined(CONFIG_OABI_COMPAT) || !defined(CONFIG_AEABI) + /* + * We failed to handle a fault trying to access the page + * containing the swi instruction, but we're not really in a + * position to return -EFAULT. Instead, return back to the + * instruction and re-enter the user fault handling path trying + * to page it in. This will likely result in sending SEGV to the + * current task. + */ +9001: + sub lr, lr, #4 + str lr, [sp, #S_PC] + b ret_fast_syscall +#endif ENDPROC(vector_swi) /* -- cgit v0.10.2 From 15e7e5c1ebf556cd620c9b091e121091ac760f6d Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 5 Jun 2013 11:27:26 +0100 Subject: ARM: 7749/1: spinlock: retry trylock operation if strex fails on free lock An exclusive store instruction may fail for reasons other than lock contention (e.g. a cache eviction during the critical section) so, in line with other architectures using similar exclusive instructions (alpha, mips, powerpc), retry the trylock operation if the lock appears to be free but the strex reported failure. Reported-by: Tony Thompson Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/include/asm/spinlock.h b/arch/arm/include/asm/spinlock.h index 6220e9f..f8b8965 100644 --- a/arch/arm/include/asm/spinlock.h +++ b/arch/arm/include/asm/spinlock.h @@ -97,19 +97,22 @@ static inline void arch_spin_lock(arch_spinlock_t *lock) static inline int arch_spin_trylock(arch_spinlock_t *lock) { - unsigned long tmp; + unsigned long contended, res; u32 slock; - __asm__ __volatile__( -" ldrex %0, [%2]\n" -" subs %1, %0, %0, ror #16\n" -" addeq %0, %0, %3\n" -" strexeq %1, %0, [%2]" - : "=&r" (slock), "=&r" (tmp) - : "r" (&lock->slock), "I" (1 << TICKET_SHIFT) - : "cc"); - - if (tmp == 0) { + do { + __asm__ __volatile__( + " ldrex %0, [%3]\n" + " mov %2, #0\n" + " subs %1, %0, %0, ror #16\n" + " addeq %0, %0, %4\n" + " strexeq %2, %0, [%3]" + : "=&r" (slock), "=&r" (contended), "=r" (res) + : "r" (&lock->slock), "I" (1 << TICKET_SHIFT) + : "cc"); + } while (res); + + if (!contended) { smp_mb(); return 1; } else { -- cgit v0.10.2 From 2874865c1271cc8e8b663804e5de4bc0c36273e1 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Thu, 6 Jun 2013 05:13:48 +0100 Subject: ARM: 7751/1: zImage: don't overwrite ourself with a page table When zImage is loaded into RAM at a low address but TEXT_OFFSET is set higher, we risk overwriting ourself with the page table needed to turn on the cache as it is located relative to the relocation address. Let's defer the cache setup after relocation in that case. Signed-off-by: Nicolas Pitre Reported-by: Stephen Boyd Tested-by: Stephen Boyd Signed-off-by: Russell King diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S index fe4d9c3..8e0d0ad 100644 --- a/arch/arm/boot/compressed/head.S +++ b/arch/arm/boot/compressed/head.S @@ -182,7 +182,19 @@ not_angel: ldr r4, =zreladdr #endif - bl cache_on + /* + * Set up a page table only if it won't overwrite ourself. + * That means r4 < pc && r4 - 16k page directory > &_end. + * Given that r4 > &_end is most unfrequent, we add a rough + * additional 1MB of room for a possible appended DTB. + */ + mov r0, pc + cmp r0, r4 + ldrcc r0, LC0+32 + addcc r0, r0, pc + cmpcc r4, r0 + orrcc r4, r4, #1 @ remember we skipped cache_on + blcs cache_on restart: adr r0, LC0 ldmia r0, {r1, r2, r3, r6, r10, r11, r12} @@ -228,7 +240,7 @@ restart: adr r0, LC0 * r0 = delta * r2 = BSS start * r3 = BSS end - * r4 = final kernel address + * r4 = final kernel address (possibly with LSB set) * r5 = appended dtb size (still unknown) * r6 = _edata * r7 = architecture ID @@ -276,6 +288,7 @@ restart: adr r0, LC0 */ cmp r0, #1 sub r0, r4, #TEXT_OFFSET + bic r0, r0, #1 add r0, r0, #0x100 mov r1, r6 sub r2, sp, r6 @@ -322,12 +335,13 @@ dtb_check_done: /* * Check to see if we will overwrite ourselves. - * r4 = final kernel address + * r4 = final kernel address (possibly with LSB set) * r9 = size of decompressed image * r10 = end of this image, including bss/stack/malloc space if non XIP * We basically want: * r4 - 16k page directory >= r10 -> OK * r4 + image length <= address of wont_overwrite -> OK + * Note: the possible LSB in r4 is harmless here. */ add r10, r10, #16384 cmp r4, r10 @@ -389,7 +403,8 @@ dtb_check_done: add sp, sp, r6 #endif - bl cache_clean_flush + tst r4, #1 + bleq cache_clean_flush adr r0, BSYM(restart) add r0, r0, r6 @@ -401,7 +416,7 @@ wont_overwrite: * r0 = delta * r2 = BSS start * r3 = BSS end - * r4 = kernel execution address + * r4 = kernel execution address (possibly with LSB set) * r5 = appended dtb size (0 if not present) * r7 = architecture ID * r8 = atags pointer @@ -464,6 +479,15 @@ not_relocated: mov r0, #0 cmp r2, r3 blo 1b + /* + * Did we skip the cache setup earlier? + * That is indicated by the LSB in r4. + * Do it now if so. + */ + tst r4, #1 + bic r4, r4, #1 + blne cache_on + /* * The C runtime environment should now be setup sufficiently. * Set up some pointers, and start decompressing. @@ -512,6 +536,7 @@ LC0: .word LC0 @ r1 .word _got_start @ r11 .word _got_end @ ip .word .L_user_stack_end @ sp + .word _end - restart + 16384 + 1024*1024 .size LC0, . - LC0 #ifdef CONFIG_ARCH_RPC -- cgit v0.10.2 From 621a0147d5c921f4cc33636ccd0602ad5d7cbfbc Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 12 Jun 2013 12:25:56 +0100 Subject: ARM: 7757/1: mm: don't flush icache in switch_mm with hardware broadcasting When scheduling an mm on a CPU where it hasn't previously been used, we flush the icache on that CPU so that any code loaded previously on a different core can be safely executed. For cores with hardware broadcasting of cache maintenance operations, this is clearly unnecessary, since the inner-shareable invalidation in __sync_icache_dcache will affect all CPUs. This patch conditionalises the icache flush in switch_mm based on cache_ops_need_broadcast(). Acked-by: Catalin Marinas Reported-by: Albin Tonnerre Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/include/asm/mmu_context.h b/arch/arm/include/asm/mmu_context.h index a7b85e0..2a45c33 100644 --- a/arch/arm/include/asm/mmu_context.h +++ b/arch/arm/include/asm/mmu_context.h @@ -18,6 +18,7 @@ #include #include #include +#include #include void __check_vmalloc_seq(struct mm_struct *mm); @@ -98,12 +99,16 @@ switch_mm(struct mm_struct *prev, struct mm_struct *next, #ifdef CONFIG_MMU unsigned int cpu = smp_processor_id(); -#ifdef CONFIG_SMP - /* check for possible thread migration */ - if (!cpumask_empty(mm_cpumask(next)) && + /* + * __sync_icache_dcache doesn't broadcast the I-cache invalidation, + * so check for possible thread migration and invalidate the I-cache + * if we're new to this CPU. + */ + if (cache_ops_need_broadcast() && + !cpumask_empty(mm_cpumask(next)) && !cpumask_test_cpu(cpu, mm_cpumask(next))) __flush_icache_all(); -#endif + if (!cpumask_test_and_set_cpu(cpu, mm_cpumask(next)) || prev != next) { check_and_switch_context(next, tsk); if (cache_is_vivt()) -- cgit v0.10.2 From 4a1b573346ee0d64d95beb78d49a5bbb574e6c6a Mon Sep 17 00:00:00 2001 From: Eduardo Valentin Date: Thu, 13 Jun 2013 22:58:52 +0100 Subject: ARM: 7758/1: introduce config HAS_BANDGAP Bandgap is a device used to measure temperature on electronic equipments. It is widely used in digital integrated circuits. It is based on the dependency between silicon voltage and temperature. This patch introduce HAS_BANDGAP config entry. This config is a boolean value so that arch code can flag if they feature a bandgap device. This config entry follows the same idea behind ARCH_HAS_CPUFREQ. Cc: linux-arm-kernel@lists.infradead.org Cc: linux-omap@vger.kernel.org Cc: linux-kernel@vger.kernel.org Reviewed-by: Fabio Stevam Acked-by: Tony Lindgren Acked-by: Amit Daniel Kachhap Signed-off-by: Eduardo Valentin Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index d423d58..bcbdec9 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -174,6 +174,9 @@ config ARCH_HAS_CPUFREQ and that the relevant menu configurations are displayed for it. +config ARCH_HAS_BANDGAP + bool + config GENERIC_HWEIGHT bool default y diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig index f49cd51..8620ab5 100644 --- a/arch/arm/mach-omap2/Kconfig +++ b/arch/arm/mach-omap2/Kconfig @@ -4,6 +4,7 @@ config ARCH_OMAP config ARCH_OMAP2PLUS bool "TI OMAP2/3/4/5 SoCs with device tree support" if (ARCH_MULTI_V6 || ARCH_MULTI_V7) select ARCH_HAS_CPUFREQ + select ARCH_HAS_BANDGAP select ARCH_HAS_HOLES_MEMORYMODEL select ARCH_OMAP select ARCH_REQUIRE_GPIOLIB -- cgit v0.10.2 From a4780adeefd042482f624f5e0d577bf9cdcbb760 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Andr=C3=A9=20Hentschel?= Date: Tue, 18 Jun 2013 23:23:26 +0100 Subject: ARM: 7735/2: Preserve the user r/w register TPIDRURW on context switch and fork MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since commit 6a1c53124aa1 the user writeable TLS register was zeroed to prevent it from being used as a covert channel between two tasks. There are more and more applications coming to Windows RT, Wine could support them, but mostly they expect to have the thread environment block (TEB) in TPIDRURW. This patch preserves that register per thread instead of clearing it. Unlike the TPIDRURO, which is already switched, the TPIDRURW can be updated from userspace so needs careful treatment in the case that we modify TPIDRURW and call fork(). To avoid this we must always read TPIDRURW in copy_thread. Signed-off-by: André Hentschel Signed-off-by: Will Deacon Signed-off-by: Jonathan Austin Signed-off-by: Russell King diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index 1995d1a..214d415 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -58,7 +58,7 @@ struct thread_info { struct cpu_context_save cpu_context; /* cpu context */ __u32 syscall; /* syscall number */ __u8 used_cp[16]; /* thread used copro */ - unsigned long tp_value; + unsigned long tp_value[2]; /* TLS registers */ #ifdef CONFIG_CRUNCH struct crunch_state crunchstate; #endif diff --git a/arch/arm/include/asm/tls.h b/arch/arm/include/asm/tls.h index 73409e6..83259b8 100644 --- a/arch/arm/include/asm/tls.h +++ b/arch/arm/include/asm/tls.h @@ -2,27 +2,30 @@ #define __ASMARM_TLS_H #ifdef __ASSEMBLY__ - .macro set_tls_none, tp, tmp1, tmp2 +#include + .macro switch_tls_none, base, tp, tpuser, tmp1, tmp2 .endm - .macro set_tls_v6k, tp, tmp1, tmp2 + .macro switch_tls_v6k, base, tp, tpuser, tmp1, tmp2 + mrc p15, 0, \tmp2, c13, c0, 2 @ get the user r/w register mcr p15, 0, \tp, c13, c0, 3 @ set TLS register - mov \tmp1, #0 - mcr p15, 0, \tmp1, c13, c0, 2 @ clear user r/w TLS register + mcr p15, 0, \tpuser, c13, c0, 2 @ and the user r/w register + str \tmp2, [\base, #TI_TP_VALUE + 4] @ save it .endm - .macro set_tls_v6, tp, tmp1, tmp2 + .macro switch_tls_v6, base, tp, tpuser, tmp1, tmp2 ldr \tmp1, =elf_hwcap ldr \tmp1, [\tmp1, #0] mov \tmp2, #0xffff0fff tst \tmp1, #HWCAP_TLS @ hardware TLS available? - mcrne p15, 0, \tp, c13, c0, 3 @ yes, set TLS register - movne \tmp1, #0 - mcrne p15, 0, \tmp1, c13, c0, 2 @ clear user r/w TLS register streq \tp, [\tmp2, #-15] @ set TLS value at 0xffff0ff0 + mrcne p15, 0, \tmp2, c13, c0, 2 @ get the user r/w register + mcrne p15, 0, \tp, c13, c0, 3 @ yes, set TLS register + mcrne p15, 0, \tpuser, c13, c0, 2 @ set user r/w register + strne \tmp2, [\base, #TI_TP_VALUE + 4] @ save it .endm - .macro set_tls_software, tp, tmp1, tmp2 + .macro switch_tls_software, base, tp, tpuser, tmp1, tmp2 mov \tmp1, #0xffff0fff str \tp, [\tmp1, #-15] @ set TLS value at 0xffff0ff0 .endm @@ -31,19 +34,30 @@ #ifdef CONFIG_TLS_REG_EMUL #define tls_emu 1 #define has_tls_reg 1 -#define set_tls set_tls_none +#define switch_tls switch_tls_none #elif defined(CONFIG_CPU_V6) #define tls_emu 0 #define has_tls_reg (elf_hwcap & HWCAP_TLS) -#define set_tls set_tls_v6 +#define switch_tls switch_tls_v6 #elif defined(CONFIG_CPU_32v6K) #define tls_emu 0 #define has_tls_reg 1 -#define set_tls set_tls_v6k +#define switch_tls switch_tls_v6k #else #define tls_emu 0 #define has_tls_reg 0 -#define set_tls set_tls_software +#define switch_tls switch_tls_software #endif +#ifndef __ASSEMBLY__ +static inline unsigned long get_tpuser(void) +{ + unsigned long reg = 0; + + if (has_tls_reg && !tls_emu) + __asm__("mrc p15, 0, %0, c13, c0, 2" : "=r" (reg)); + + return reg; +} +#endif #endif /* __ASMARM_TLS_H */ diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 582b405..a39cfc2a1 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -685,15 +685,16 @@ ENTRY(__switch_to) UNWIND(.fnstart ) UNWIND(.cantunwind ) add ip, r1, #TI_CPU_SAVE - ldr r3, [r2, #TI_TP_VALUE] ARM( stmia ip!, {r4 - sl, fp, sp, lr} ) @ Store most regs on stack THUMB( stmia ip!, {r4 - sl, fp} ) @ Store most regs on stack THUMB( str sp, [ip], #4 ) THUMB( str lr, [ip], #4 ) + ldr r4, [r2, #TI_TP_VALUE] + ldr r5, [r2, #TI_TP_VALUE + 4] #ifdef CONFIG_CPU_USE_DOMAINS ldr r6, [r2, #TI_CPU_DOMAIN] #endif - set_tls r3, r4, r5 + switch_tls r1, r4, r5, r3, r7 #if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP) ldr r7, [r2, #TI_TASK] ldr r8, =__stack_chk_guard diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index f219703..0870641 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -39,6 +39,7 @@ #include #include #include +#include #ifdef CONFIG_CC_STACKPROTECTOR #include @@ -343,7 +344,8 @@ copy_thread(unsigned long clone_flags, unsigned long stack_start, clear_ptrace_hw_breakpoint(p); if (clone_flags & CLONE_SETTLS) - thread->tp_value = childregs->ARM_r3; + thread->tp_value[0] = childregs->ARM_r3; + thread->tp_value[1] = get_tpuser(); thread_notify(THREAD_NOTIFY_COPY, thread); diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index 03deeff..2bc1514 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -849,7 +849,7 @@ long arch_ptrace(struct task_struct *child, long request, #endif case PTRACE_GET_THREAD_AREA: - ret = put_user(task_thread_info(child)->tp_value, + ret = put_user(task_thread_info(child)->tp_value[0], datap); break; diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 18b32e8..517bfd4 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -581,7 +581,7 @@ asmlinkage int arm_syscall(int no, struct pt_regs *regs) return regs->ARM_r0; case NR(set_tls): - thread->tp_value = regs->ARM_r0; + thread->tp_value[0] = regs->ARM_r0; if (tls_emu) return 0; if (has_tls_reg) { @@ -699,7 +699,7 @@ static int get_tp_trap(struct pt_regs *regs, unsigned int instr) int reg = (instr >> 12) & 15; if (reg == 15) return 1; - regs->uregs[reg] = current_thread_info()->tp_value; + regs->uregs[reg] = current_thread_info()->tp_value[0]; regs->ARM_pc += 4; return 0; } -- cgit v0.10.2 From c5f927a6f62196226915f12194c9d0df4e2210d7 Mon Sep 17 00:00:00 2001 From: Jed Davis Date: Thu, 20 Jun 2013 10:16:29 +0100 Subject: ARM: 7765/1: perf: Record the user-mode PC in the call chain. With this change, we no longer lose the innermost entry in the user-mode part of the call chain. See also the x86 port, which includes the ip. It's possible to partially work around this problem by post-processing the data to use the PERF_SAMPLE_IP value, but this works only if the CPU wasn't in the kernel when the sample was taken. Cc: Signed-off-by: Jed Davis Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index 8c3094d..d9f5cd4 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -569,6 +569,7 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) return; } + perf_callchain_store(entry, regs->ARM_pc); tail = (struct frame_tail __user *)regs->ARM_fp - 1; while ((entry->nr < PERF_MAX_STACK_DEPTH) && -- cgit v0.10.2 From 8121cf312a1908f67173ec3773da1688e04437e7 Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Thu, 20 Jun 2013 14:02:21 +0100 Subject: ARM: 7766/1: versatile: don't mark pen as __INIT When booting fewer cores than are physically present on a versatile platform (e.g. when passing maxcpus=N on the command line), some secondary cores may remain in the holding pen, which is marked __INIT, as each CPU's gic cpumask is initialised to 0xff, and thus an IPI to any CPU will wake up *all* secondaries. This behaviour is crucial to the GIC cpumask self-discovery. Late in the boot process, the memory comprising the holding pen will be released to the kernel for more general use, and may be overwritten with arbitrary data, which can cause the held secondaries to start behaving unpredictably. This can lead to all manner of odd behaviour from the kernel. As preventing cpus from entering the pen would require invasive changes to the GIC driver and to existing dts used in the wild, we instead remove the __INIT marker from the pen, keeping it around and leaving the unused secondary CPUs dormant. Link: http://lists.infradead.org/pipermail/linux-arm-kernel/2013-June/175039.html Signed-off-by: Mark Rutland Acked-by: Pawel Moll Acked-by: Nicolas Pitre Cc: Lorenzo Pieralisi Signed-off-by: Russell King diff --git a/arch/arm/plat-versatile/headsmp.S b/arch/arm/plat-versatile/headsmp.S index b178d44..2677bc3 100644 --- a/arch/arm/plat-versatile/headsmp.S +++ b/arch/arm/plat-versatile/headsmp.S @@ -11,8 +11,6 @@ #include #include - __INIT - /* * Realview/Versatile Express specific entry point for secondary CPUs. * This provides a "holding pen" into which all secondary cores are held -- cgit v0.10.2 From ae120d9edfe96628f03d87634acda0bfa7110632 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 21 Jun 2013 12:06:19 +0100 Subject: ARM: 7767/1: let the ASID allocator handle suspended animation When a CPU is running a process, the ASID for that process is held in a per-CPU variable (the "active ASIDs" array). When the ASID allocator handles a rollover, it copies the active ASIDs into a "reserved ASIDs" array to ensure that a process currently running on another CPU will continue to run unaffected. The active array is zero-ed to indicate that a rollover occurred. Because of this mechanism, a reserved ASID is only remembered for a single rollover. A subsequent rollover will completely refill the reserved ASIDs array. In a severely oversubscribed environment where a CPU can be prevented from running for extended periods of time (think virtual machines), the above has a horrible side effect: [P{a} denotes process P running with ASID a] CPU-0 CPU-1 A{x} [active = ] [suspended] runs B{y} [active = ] [rollover: active = <0 0> reserved = ] runs B{y} [active = <0 y> reserved = ] [rollover: active = <0 0> reserved = <0 y>] runs C{x} [active = <0 x>] [resumes] runs A{x} At that stage, both A and C have the same ASID, with deadly consequences. The fix is to preserve reserved ASIDs across rollovers if the CPU doesn't have an active ASID when the rollover occurs. Cc: # 3.9 Acked-by: Will Deacon Acked-by: Catalin Carinas Signed-off-by: Marc Zyngier Signed-off-by: Russell King diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c index 2ac3737..8e12fcb 100644 --- a/arch/arm/mm/context.c +++ b/arch/arm/mm/context.c @@ -128,6 +128,15 @@ static void flush_context(unsigned int cpu) asid = 0; } else { asid = atomic64_xchg(&per_cpu(active_asids, i), 0); + /* + * If this CPU has already been through a + * rollover, but hasn't run another task in + * the meantime, we must preserve its reserved + * ASID, as this is the only trace we have of + * the process it is still running. + */ + if (asid == 0) + asid = per_cpu(reserved_asids, i); __set_bit(ASID_TO_IDX(asid), asid_map); } per_cpu(reserved_asids, i) = asid; -- cgit v0.10.2 From b8e4a4740fa2b17c0a447b3ab783b3dc10702e27 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 21 Jun 2013 12:06:55 +0100 Subject: ARM: 7768/1: prevent risks of out-of-bound access in ASID allocator On a CPU that never ran anything, both the active and reserved ASID fields are set to zero. In this case the ASID_TO_IDX() macro will return -1, which is not a very useful value to index a bitmap. Instead of trying to offset the ASID so that ASID #1 is actually bit 0 in the asid_map bitmap, just always ignore bit 0 and start the search from bit 1. This makes the code a bit more readable, and without risk of OoB access. Cc: # 3.9 Acked-by: Will Deacon Acked-by: Catalin Marinas Reported-by: Catalin Marinas Signed-off-by: Marc Zyngier Signed-off-by: Russell King diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c index 8e12fcb..83e0905 100644 --- a/arch/arm/mm/context.c +++ b/arch/arm/mm/context.c @@ -39,10 +39,7 @@ * non 64-bit operations. */ #define ASID_FIRST_VERSION (1ULL << ASID_BITS) -#define NUM_USER_ASIDS (ASID_FIRST_VERSION - 1) - -#define ASID_TO_IDX(asid) ((asid & ~ASID_MASK) - 1) -#define IDX_TO_ASID(idx) ((idx + 1) & ~ASID_MASK) +#define NUM_USER_ASIDS ASID_FIRST_VERSION static DEFINE_RAW_SPINLOCK(cpu_asid_lock); static atomic64_t asid_generation = ATOMIC64_INIT(ASID_FIRST_VERSION); @@ -137,7 +134,7 @@ static void flush_context(unsigned int cpu) */ if (asid == 0) asid = per_cpu(reserved_asids, i); - __set_bit(ASID_TO_IDX(asid), asid_map); + __set_bit(asid & ~ASID_MASK, asid_map); } per_cpu(reserved_asids, i) = asid; } @@ -176,17 +173,19 @@ static u64 new_context(struct mm_struct *mm, unsigned int cpu) /* * Allocate a free ASID. If we can't find one, take a * note of the currently active ASIDs and mark the TLBs - * as requiring flushes. + * as requiring flushes. We always count from ASID #1, + * as we reserve ASID #0 to switch via TTBR0 and indicate + * rollover events. */ - asid = find_first_zero_bit(asid_map, NUM_USER_ASIDS); + asid = find_next_zero_bit(asid_map, NUM_USER_ASIDS, 1); if (asid == NUM_USER_ASIDS) { generation = atomic64_add_return(ASID_FIRST_VERSION, &asid_generation); flush_context(cpu); - asid = find_first_zero_bit(asid_map, NUM_USER_ASIDS); + asid = find_next_zero_bit(asid_map, NUM_USER_ASIDS, 1); } __set_bit(asid, asid_map); - asid = generation | IDX_TO_ASID(asid); + asid |= generation; cpumask_clear(mm_cpumask(mm)); } -- cgit v0.10.2 From 0d0752bca1f9a91fb646647aa4abbb21156f316c Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 21 Jun 2013 12:07:27 +0100 Subject: ARM: 7769/1: Cortex-A15: fix erratum 798181 implementation Looking into the active_asids array is not enough, as we also need to look into the reserved_asids array (they both represent processes that are currently running). Also, not holding the ASID allocator lock is racy, as another CPU could schedule that process and trigger a rollover, making the erratum workaround miss an IPI. Exposing this outside of context.c is a little ugly on the side, so let's define a new entry point that the erratum workaround can call to obtain the cpumask. Cc: # 3.9 Acked-by: Will Deacon Acked-by: Catalin Marinas Signed-off-by: Marc Zyngier Signed-off-by: Russell King diff --git a/arch/arm/include/asm/mmu_context.h b/arch/arm/include/asm/mmu_context.h index 2a45c33..b5792b7 100644 --- a/arch/arm/include/asm/mmu_context.h +++ b/arch/arm/include/asm/mmu_context.h @@ -28,7 +28,15 @@ void __check_vmalloc_seq(struct mm_struct *mm); void check_and_switch_context(struct mm_struct *mm, struct task_struct *tsk); #define init_new_context(tsk,mm) ({ atomic64_set(&mm->context.id, 0); 0; }) -DECLARE_PER_CPU(atomic64_t, active_asids); +#ifdef CONFIG_ARM_ERRATA_798181 +void a15_erratum_get_cpumask(int this_cpu, struct mm_struct *mm, + cpumask_t *mask); +#else /* !CONFIG_ARM_ERRATA_798181 */ +static inline void a15_erratum_get_cpumask(int this_cpu, struct mm_struct *mm, + cpumask_t *mask) +{ +} +#endif /* CONFIG_ARM_ERRATA_798181 */ #else /* !CONFIG_CPU_HAS_ASID */ diff --git a/arch/arm/kernel/smp_tlb.c b/arch/arm/kernel/smp_tlb.c index 9a52a07..a98b62d 100644 --- a/arch/arm/kernel/smp_tlb.c +++ b/arch/arm/kernel/smp_tlb.c @@ -103,7 +103,7 @@ static void broadcast_tlb_a15_erratum(void) static void broadcast_tlb_mm_a15_erratum(struct mm_struct *mm) { - int cpu, this_cpu; + int this_cpu; cpumask_t mask = { CPU_BITS_NONE }; if (!erratum_a15_798181()) @@ -111,21 +111,7 @@ static void broadcast_tlb_mm_a15_erratum(struct mm_struct *mm) dummy_flush_tlb_a15_erratum(); this_cpu = get_cpu(); - for_each_online_cpu(cpu) { - if (cpu == this_cpu) - continue; - /* - * We only need to send an IPI if the other CPUs are running - * the same ASID as the one being invalidated. There is no - * need for locking around the active_asids check since the - * switch_mm() function has at least one dmb() (as required by - * this workaround) in case a context switch happens on - * another CPU after the condition below. - */ - if (atomic64_read(&mm->context.id) == - atomic64_read(&per_cpu(active_asids, cpu))) - cpumask_set_cpu(cpu, &mask); - } + a15_erratum_get_cpumask(this_cpu, mm, &mask); smp_call_function_many(&mask, ipi_flush_tlb_a15_erratum, NULL, 1); put_cpu(); } diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c index 83e0905..eeab06e 100644 --- a/arch/arm/mm/context.c +++ b/arch/arm/mm/context.c @@ -45,10 +45,37 @@ static DEFINE_RAW_SPINLOCK(cpu_asid_lock); static atomic64_t asid_generation = ATOMIC64_INIT(ASID_FIRST_VERSION); static DECLARE_BITMAP(asid_map, NUM_USER_ASIDS); -DEFINE_PER_CPU(atomic64_t, active_asids); +static DEFINE_PER_CPU(atomic64_t, active_asids); static DEFINE_PER_CPU(u64, reserved_asids); static cpumask_t tlb_flush_pending; +#ifdef CONFIG_ARM_ERRATA_798181 +void a15_erratum_get_cpumask(int this_cpu, struct mm_struct *mm, + cpumask_t *mask) +{ + int cpu; + unsigned long flags; + u64 context_id, asid; + + raw_spin_lock_irqsave(&cpu_asid_lock, flags); + context_id = mm->context.id.counter; + for_each_online_cpu(cpu) { + if (cpu == this_cpu) + continue; + /* + * We only need to send an IPI if the other CPUs are + * running the same ASID as the one being invalidated. + */ + asid = per_cpu(active_asids, cpu).counter; + if (asid == 0) + asid = per_cpu(reserved_asids, cpu); + if (context_id == asid) + cpumask_set_cpu(cpu, mask); + } + raw_spin_unlock_irqrestore(&cpu_asid_lock, flags); +} +#endif + #ifdef CONFIG_ARM_LPAE static void cpu_set_reserved_ttbr0(void) { -- cgit v0.10.2 From 52c08a9e399739979027ea1e463f5529476da104 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 21 Jun 2013 14:49:43 +0100 Subject: ARM: 7770/1: remove residual ARMv2 support from decompressor arm26 support in Linux is long gone, yet it left an interresting, fossilized trace in the decompressor. Remove it so people won't get confused about what teqp is actually doing here... Signed-off-by: Marc Zyngier Signed-off-by: Russell King diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S index 8e0d0ad..15f2f57 100644 --- a/arch/arm/boot/compressed/head.S +++ b/arch/arm/boot/compressed/head.S @@ -141,7 +141,6 @@ start: mov r7, r1 @ save architecture ID mov r8, r2 @ save atags pointer -#ifndef __ARM_ARCH_2__ /* * Booting from Angel - need to enter SVC mode and disable * FIQs/IRQs (numeric definitions from angel arm.h source). @@ -157,10 +156,6 @@ not_angel: safe_svcmode_maskall r0 msr spsr_cxsf, r9 @ Save the CPU boot mode in @ SPSR -#else - teqp pc, #0x0c000003 @ turn off interrupts -#endif - /* * Note that some cache flushing and other stuff may * be needed here - is there an Angel SWI call for this? -- cgit v0.10.2 From 81793bab5d7cc0d0962dae47dab3a63ae4bd5265 Mon Sep 17 00:00:00 2001 From: Douglas Anderson Date: Mon, 24 Jun 2013 22:03:50 +0100 Subject: ARM: 7774/1: Fix dtb dependency to use order-only prerequisites The %.dtb dependency is specified to depend on the PHONY "scripts". That means that it'll build every time even if the underlying dtb file hasn't been touched. Use an order-only prerequisites to fix this. Also mark "dtbs" as PHONY for correctness. This was broken in (70b0476 ARM: 7513/1: Make sure dtc is built before running it). Reported-by: Mike Frysinger Signed-off-by: Doug Anderson Acked-by: Olof Johansson Reviewed-by: David Brown Signed-off-by: Russell King diff --git a/arch/arm/Makefile b/arch/arm/Makefile index 4737408..10c0809 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -289,9 +289,10 @@ zImage Image xipImage bootpImage uImage: vmlinux zinstall uinstall install: vmlinux $(Q)$(MAKE) $(build)=$(boot) MACHINE=$(MACHINE) $@ -%.dtb: scripts +%.dtb: | scripts $(Q)$(MAKE) $(build)=$(boot)/dts MACHINE=$(MACHINE) $(boot)/dts/$@ +PHONY += dtbs dtbs: scripts $(Q)$(MAKE) $(build)=$(boot)/dts MACHINE=$(MACHINE) dtbs -- cgit v0.10.2 From 1b21376a737aeaa82320ae014b8e1c2c53cfc479 Mon Sep 17 00:00:00 2001 From: Douglas Anderson Date: Thu, 27 Jun 2013 04:32:06 +0100 Subject: ARM: 7777/1: Avoid extra calls to the C compiler Starting up the C compiler can be a slow operation on some systems. Though these calls don't individually take a lot of time, they add up. Rearrange the ARM Makefile a bit to avoid extra calls to the compiler when they can be easily avoided. When running with the Chrome OS ARM cross compiler "armv7a-cros-linux-gnueabi-", this shaved .55 seconds (from 5.31 seconds to 4.76 seconds) off an incremental build of the kernel: time make -j32 ARCH=arm CROSS_COMPILE=armv7a-cros-linux-gnueabi- Thanks to Mike Frysinger for the clean trick to make this work. Signed-off-by: Doug Anderson Acked-by: Mike Frysinger Signed-off-by: Russell King diff --git a/arch/arm/Makefile b/arch/arm/Makefile index 10c0809..20559a0 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -59,37 +59,43 @@ comma = , # Note that GCC does not numerically define an architecture version # macro, but instead defines a whole series of macros which makes # testing for a specific architecture or later rather impossible. -arch-$(CONFIG_CPU_32v7) :=-D__LINUX_ARM_ARCH__=7 $(call cc-option,-march=armv7-a,-march=armv5t -Wa$(comma)-march=armv7-a) -arch-$(CONFIG_CPU_32v6) :=-D__LINUX_ARM_ARCH__=6 $(call cc-option,-march=armv6,-march=armv5t -Wa$(comma)-march=armv6) +arch-$(CONFIG_CPU_32v7) =-D__LINUX_ARM_ARCH__=7 $(call cc-option,-march=armv7-a,-march=armv5t -Wa$(comma)-march=armv7-a) +arch-$(CONFIG_CPU_32v6) =-D__LINUX_ARM_ARCH__=6 $(call cc-option,-march=armv6,-march=armv5t -Wa$(comma)-march=armv6) # Only override the compiler option if ARMv6. The ARMv6K extensions are # always available in ARMv7 ifeq ($(CONFIG_CPU_32v6),y) -arch-$(CONFIG_CPU_32v6K) :=-D__LINUX_ARM_ARCH__=6 $(call cc-option,-march=armv6k,-march=armv5t -Wa$(comma)-march=armv6k) +arch-$(CONFIG_CPU_32v6K) =-D__LINUX_ARM_ARCH__=6 $(call cc-option,-march=armv6k,-march=armv5t -Wa$(comma)-march=armv6k) endif -arch-$(CONFIG_CPU_32v5) :=-D__LINUX_ARM_ARCH__=5 $(call cc-option,-march=armv5te,-march=armv4t) -arch-$(CONFIG_CPU_32v4T) :=-D__LINUX_ARM_ARCH__=4 -march=armv4t -arch-$(CONFIG_CPU_32v4) :=-D__LINUX_ARM_ARCH__=4 -march=armv4 -arch-$(CONFIG_CPU_32v3) :=-D__LINUX_ARM_ARCH__=3 -march=armv3 +arch-$(CONFIG_CPU_32v5) =-D__LINUX_ARM_ARCH__=5 $(call cc-option,-march=armv5te,-march=armv4t) +arch-$(CONFIG_CPU_32v4T) =-D__LINUX_ARM_ARCH__=4 -march=armv4t +arch-$(CONFIG_CPU_32v4) =-D__LINUX_ARM_ARCH__=4 -march=armv4 +arch-$(CONFIG_CPU_32v3) =-D__LINUX_ARM_ARCH__=3 -march=armv3 + +# Evaluate arch cc-option calls now +arch-y := $(arch-y) # This selects how we optimise for the processor. -tune-$(CONFIG_CPU_ARM7TDMI) :=-mtune=arm7tdmi -tune-$(CONFIG_CPU_ARM720T) :=-mtune=arm7tdmi -tune-$(CONFIG_CPU_ARM740T) :=-mtune=arm7tdmi -tune-$(CONFIG_CPU_ARM9TDMI) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_ARM940T) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_ARM946E) :=$(call cc-option,-mtune=arm9e,-mtune=arm9tdmi) -tune-$(CONFIG_CPU_ARM920T) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_ARM922T) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_ARM925T) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_ARM926T) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_FA526) :=-mtune=arm9tdmi -tune-$(CONFIG_CPU_SA110) :=-mtune=strongarm110 -tune-$(CONFIG_CPU_SA1100) :=-mtune=strongarm1100 -tune-$(CONFIG_CPU_XSCALE) :=$(call cc-option,-mtune=xscale,-mtune=strongarm110) -Wa,-mcpu=xscale -tune-$(CONFIG_CPU_XSC3) :=$(call cc-option,-mtune=xscale,-mtune=strongarm110) -Wa,-mcpu=xscale -tune-$(CONFIG_CPU_FEROCEON) :=$(call cc-option,-mtune=marvell-f,-mtune=xscale) -tune-$(CONFIG_CPU_V6) :=$(call cc-option,-mtune=arm1136j-s,-mtune=strongarm) -tune-$(CONFIG_CPU_V6K) :=$(call cc-option,-mtune=arm1136j-s,-mtune=strongarm) +tune-$(CONFIG_CPU_ARM7TDMI) =-mtune=arm7tdmi +tune-$(CONFIG_CPU_ARM720T) =-mtune=arm7tdmi +tune-$(CONFIG_CPU_ARM740T) =-mtune=arm7tdmi +tune-$(CONFIG_CPU_ARM9TDMI) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_ARM940T) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_ARM946E) =$(call cc-option,-mtune=arm9e,-mtune=arm9tdmi) +tune-$(CONFIG_CPU_ARM920T) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_ARM922T) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_ARM925T) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_ARM926T) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_FA526) =-mtune=arm9tdmi +tune-$(CONFIG_CPU_SA110) =-mtune=strongarm110 +tune-$(CONFIG_CPU_SA1100) =-mtune=strongarm1100 +tune-$(CONFIG_CPU_XSCALE) =$(call cc-option,-mtune=xscale,-mtune=strongarm110) -Wa,-mcpu=xscale +tune-$(CONFIG_CPU_XSC3) =$(call cc-option,-mtune=xscale,-mtune=strongarm110) -Wa,-mcpu=xscale +tune-$(CONFIG_CPU_FEROCEON) =$(call cc-option,-mtune=marvell-f,-mtune=xscale) +tune-$(CONFIG_CPU_V6) =$(call cc-option,-mtune=arm1136j-s,-mtune=strongarm) +tune-$(CONFIG_CPU_V6K) =$(call cc-option,-mtune=arm1136j-s,-mtune=strongarm) + +# Evaluate tune cc-option calls now +tune-y := $(tune-y) ifeq ($(CONFIG_AEABI),y) CFLAGS_ABI :=-mabi=aapcs-linux -mno-thumb-interwork -- cgit v0.10.2