diff options
67 files changed, 890 insertions, 635 deletions
diff --git a/Documentation/filesystems/ubifs.txt b/Documentation/filesystems/ubifs.txt index 84da2a4ba25a..12fedb7834c6 100644 --- a/Documentation/filesystems/ubifs.txt +++ b/Documentation/filesystems/ubifs.txt @@ -79,13 +79,6 @@ Mount options (*) == default. -norm_unmount (*) commit on unmount; the journal is committed - when the file-system is unmounted so that the - next mount does not have to replay the journal - and it becomes very fast; -fast_unmount do not commit on unmount; this option makes - unmount faster, but the next mount slower - because of the need to replay the journal. bulk_read read more in one go to take advantage of flash media that read faster sequentially no_bulk_read (*) do not bulk-read diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 77b047475539..85040cfeb5e5 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -650,6 +650,7 @@ ENTRY(fp_enter) no_fp: mov pc, lr __und_usr_unknown: + enable_irq mov r0, sp adr lr, ret_from_exception b do_undefinstr diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 06269ea375c5..49a6ba926c2b 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -136,7 +136,7 @@ ENTRY(mcount) ldmia sp!, {r0-r3, pc} trace: - ldr r1, [fp, #-4] + ldr r1, [fp, #-4] @ lr of instrumented routine mov r0, lr sub r0, r0, #MCOUNT_INSN_SIZE mov lr, pc diff --git a/arch/arm/kernel/irq.c b/arch/arm/kernel/irq.c index 7141cee1fab7..363db186cb93 100644 --- a/arch/arm/kernel/irq.c +++ b/arch/arm/kernel/irq.c @@ -101,7 +101,7 @@ unlock: /* Handle bad interrupts */ static struct irq_desc bad_irq_desc = { .handle_irq = handle_bad_irq, - .lock = SPIN_LOCK_UNLOCKED + .lock = __SPIN_LOCK_UNLOCKED(bad_irq_desc.lock), }; /* diff --git a/arch/arm/mach-msm/board-halibut.c b/arch/arm/mach-msm/board-halibut.c index c2a96e3965a6..e61967dde9a1 100644 --- a/arch/arm/mach-msm/board-halibut.c +++ b/arch/arm/mach-msm/board-halibut.c @@ -27,6 +27,7 @@ #include <asm/mach/map.h> #include <asm/mach/flash.h> +#include <mach/irqs.h> #include <mach/board.h> #include <mach/msm_iomap.h> diff --git a/arch/arm/mach-omap1/devices.c b/arch/arm/mach-omap1/devices.c index 77382d8b6b2f..ba5d7c08dc17 100644 --- a/arch/arm/mach-omap1/devices.c +++ b/arch/arm/mach-omap1/devices.c @@ -181,7 +181,7 @@ void __init omap1_init_mmc(struct omap_mmc_platform_data **mmc_data, } size = OMAP1_MMC_SIZE; - omap_mmc_add(i, base, size, irq, mmc_data[i]); + omap_mmc_add("mmci-omap", i, base, size, irq, mmc_data[i]); }; } diff --git a/arch/arm/mach-omap1/mcbsp.c b/arch/arm/mach-omap1/mcbsp.c index ca7a0cc1707c..575ba31295cf 100644 --- a/arch/arm/mach-omap1/mcbsp.c +++ b/arch/arm/mach-omap1/mcbsp.c @@ -28,81 +28,8 @@ #define DPS_RSTCT2_PER_EN (1 << 0) #define DSP_RSTCT2_WD_PER_EN (1 << 1) -struct mcbsp_internal_clk { - struct clk clk; - struct clk **childs; - int n_childs; -}; - #if defined(CONFIG_ARCH_OMAP15XX) || defined(CONFIG_ARCH_OMAP16XX) -static void omap_mcbsp_clk_init(struct mcbsp_internal_clk *mclk) -{ - const char *clk_names[] = { "dsp_ck", "api_ck", "dspxor_ck" }; - int i; - - mclk->n_childs = ARRAY_SIZE(clk_names); - mclk->childs = kzalloc(mclk->n_childs * sizeof(struct clk *), - GFP_KERNEL); - - for (i = 0; i < mclk->n_childs; i++) { - /* We fake a platform device to get correct device id */ - struct platform_device pdev; - - pdev.dev.bus = &platform_bus_type; - pdev.id = mclk->clk.id; - mclk->childs[i] = clk_get(&pdev.dev, clk_names[i]); - if (IS_ERR(mclk->childs[i])) - printk(KERN_ERR "Could not get clock %s (%d).\n", - clk_names[i], mclk->clk.id); - } -} - -static int omap_mcbsp_clk_enable(struct clk *clk) -{ - struct mcbsp_internal_clk *mclk = container_of(clk, - struct mcbsp_internal_clk, clk); - int i; - - for (i = 0; i < mclk->n_childs; i++) - clk_enable(mclk->childs[i]); - return 0; -} - -static void omap_mcbsp_clk_disable(struct clk *clk) -{ - struct mcbsp_internal_clk *mclk = container_of(clk, - struct mcbsp_internal_clk, clk); - int i; - - for (i = 0; i < mclk->n_childs; i++) - clk_disable(mclk->childs[i]); -} - -static struct mcbsp_internal_clk omap_mcbsp_clks[] = { - { - .clk = { - .name = "mcbsp_clk", - .id = 1, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, - { - .clk = { - .name = "mcbsp_clk", - .id = 3, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, -}; - -#define omap_mcbsp_clks_size ARRAY_SIZE(omap_mcbsp_clks) -#else -#define omap_mcbsp_clks_size 0 -static struct mcbsp_internal_clk __initdata *omap_mcbsp_clks; -static inline void omap_mcbsp_clk_init(struct mcbsp_internal_clk *mclk) -{ } +const char *clk_names[] = { "dsp_ck", "api_ck", "dspxor_ck" }; #endif static void omap1_mcbsp_request(unsigned int id) @@ -167,8 +94,9 @@ static struct omap_mcbsp_platform_data omap15xx_mcbsp_pdata[] = { .rx_irq = INT_McBSP1RX, .tx_irq = INT_McBSP1TX, .ops = &omap1_mcbsp_ops, - .clk_name = "mcbsp_clk", - }, + .clk_names = clk_names, + .num_clks = 3, + }, { .phys_base = OMAP1510_MCBSP2_BASE, .dma_rx_sync = OMAP_DMA_MCBSP2_RX, @@ -184,7 +112,8 @@ static struct omap_mcbsp_platform_data omap15xx_mcbsp_pdata[] = { .rx_irq = INT_McBSP3RX, .tx_irq = INT_McBSP3TX, .ops = &omap1_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 3, }, }; #define OMAP15XX_MCBSP_PDATA_SZ ARRAY_SIZE(omap15xx_mcbsp_pdata) @@ -202,7 +131,8 @@ static struct omap_mcbsp_platform_data omap16xx_mcbsp_pdata[] = { .rx_irq = INT_McBSP1RX, .tx_irq = INT_McBSP1TX, .ops = &omap1_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 3, }, { .phys_base = OMAP1610_MCBSP2_BASE, @@ -219,7 +149,8 @@ static struct omap_mcbsp_platform_data omap16xx_mcbsp_pdata[] = { .rx_irq = INT_McBSP3RX, .tx_irq = INT_McBSP3TX, .ops = &omap1_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 3, }, }; #define OMAP16XX_MCBSP_PDATA_SZ ARRAY_SIZE(omap16xx_mcbsp_pdata) @@ -230,15 +161,6 @@ static struct omap_mcbsp_platform_data omap16xx_mcbsp_pdata[] = { int __init omap1_mcbsp_init(void) { - int i; - - for (i = 0; i < omap_mcbsp_clks_size; i++) { - if (cpu_is_omap15xx() || cpu_is_omap16xx()) { - omap_mcbsp_clk_init(&omap_mcbsp_clks[i]); - clk_register(&omap_mcbsp_clks[i].clk); - } - } - if (cpu_is_omap730()) omap_mcbsp_count = OMAP730_MCBSP_PDATA_SZ; if (cpu_is_omap15xx()) diff --git a/arch/arm/mach-omap2/devices.c b/arch/arm/mach-omap2/devices.c index 9d7216ff6c9f..ce03fa750775 100644 --- a/arch/arm/mach-omap2/devices.c +++ b/arch/arm/mach-omap2/devices.c @@ -421,6 +421,7 @@ void __init omap2_init_mmc(struct omap_mmc_platform_data **mmc_data, int nr_controllers) { int i; + char *name; for (i = 0; i < nr_controllers; i++) { unsigned long base, size; @@ -450,12 +451,14 @@ void __init omap2_init_mmc(struct omap_mmc_platform_data **mmc_data, continue; } - if (cpu_is_omap2420()) + if (cpu_is_omap2420()) { size = OMAP2420_MMC_SIZE; - else + name = "mmci-omap"; + } else { size = HSMMC_SIZE; - - omap_mmc_add(i, base, size, irq, mmc_data[i]); + name = "mmci-omap-hs"; + } + omap_mmc_add(name, i, base, size, irq, mmc_data[i]); }; } diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c index b0f8e7d62798..b52a02fc7cd6 100644 --- a/arch/arm/mach-omap2/id.c +++ b/arch/arm/mach-omap2/id.c @@ -172,9 +172,13 @@ void __init omap34xx_check_revision(void) omap_revision = OMAP3430_REV_ES3_0; rev_name = "ES3.0"; break; + case 4: + omap_revision = OMAP3430_REV_ES3_1; + rev_name = "ES3.1"; + break; default: /* Use the latest known revision as default */ - omap_revision = OMAP3430_REV_ES3_0; + omap_revision = OMAP3430_REV_ES3_1; rev_name = "Unknown revision\n"; } } diff --git a/arch/arm/mach-omap2/irq.c b/arch/arm/mach-omap2/irq.c index 636e2821af7d..9ba20d985dda 100644 --- a/arch/arm/mach-omap2/irq.c +++ b/arch/arm/mach-omap2/irq.c @@ -134,6 +134,7 @@ static struct irq_chip omap_irq_chip = { .ack = omap_mask_ack_irq, .mask = omap_mask_irq, .unmask = omap_unmask_irq, + .disable = omap_mask_irq, }; static void __init omap_irq_bank_init_one(struct omap_irq_bank *bank) diff --git a/arch/arm/mach-omap2/mcbsp.c b/arch/arm/mach-omap2/mcbsp.c index e20023c9d15d..a9e631fc1134 100644 --- a/arch/arm/mach-omap2/mcbsp.c +++ b/arch/arm/mach-omap2/mcbsp.c @@ -24,106 +24,7 @@ #include <mach/cpu.h> #include <mach/mcbsp.h> -struct mcbsp_internal_clk { - struct clk clk; - struct clk **childs; - int n_childs; -}; - -#if defined(CONFIG_ARCH_OMAP24XX) || defined(CONFIG_ARCH_OMAP34XX) -static void omap_mcbsp_clk_init(struct mcbsp_internal_clk *mclk) -{ - const char *clk_names[] = { "mcbsp_ick", "mcbsp_fck" }; - int i; - - mclk->n_childs = ARRAY_SIZE(clk_names); - mclk->childs = kzalloc(mclk->n_childs * sizeof(struct clk *), - GFP_KERNEL); - - for (i = 0; i < mclk->n_childs; i++) { - /* We fake a platform device to get correct device id */ - struct platform_device pdev; - - pdev.dev.bus = &platform_bus_type; - pdev.id = mclk->clk.id; - mclk->childs[i] = clk_get(&pdev.dev, clk_names[i]); - if (IS_ERR(mclk->childs[i])) - printk(KERN_ERR "Could not get clock %s (%d).\n", - clk_names[i], mclk->clk.id); - } -} - -static int omap_mcbsp_clk_enable(struct clk *clk) -{ - struct mcbsp_internal_clk *mclk = container_of(clk, - struct mcbsp_internal_clk, clk); - int i; - - for (i = 0; i < mclk->n_childs; i++) - clk_enable(mclk->childs[i]); - return 0; -} - -static void omap_mcbsp_clk_disable(struct clk *clk) -{ - struct mcbsp_internal_clk *mclk = container_of(clk, - struct mcbsp_internal_clk, clk); - int i; - - for (i = 0; i < mclk->n_childs; i++) - clk_disable(mclk->childs[i]); -} - -static struct mcbsp_internal_clk omap_mcbsp_clks[] = { - { - .clk = { - .name = "mcbsp_clk", - .id = 1, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, - { - .clk = { - .name = "mcbsp_clk", - .id = 2, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, - { - .clk = { - .name = "mcbsp_clk", - .id = 3, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, - { - .clk = { - .name = "mcbsp_clk", - .id = 4, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, - { - .clk = { - .name = "mcbsp_clk", - .id = 5, - .enable = omap_mcbsp_clk_enable, - .disable = omap_mcbsp_clk_disable, - }, - }, -}; - -#define omap_mcbsp_clks_size ARRAY_SIZE(omap_mcbsp_clks) -#else -#define omap_mcbsp_clks_size 0 -static struct mcbsp_internal_clk __initdata *omap_mcbsp_clks; -static inline void omap_mcbsp_clk_init(struct clk *clk) -{ } -#endif +const char *clk_names[] = { "mcbsp_ick", "mcbsp_fck" }; static void omap2_mcbsp2_mux_setup(void) { @@ -156,7 +57,8 @@ static struct omap_mcbsp_platform_data omap2420_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP1_IRQ_RX, .tx_irq = INT_24XX_MCBSP1_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP24XX_MCBSP2_BASE, @@ -165,7 +67,8 @@ static struct omap_mcbsp_platform_data omap2420_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP2_IRQ_RX, .tx_irq = INT_24XX_MCBSP2_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, }; #define OMAP2420_MCBSP_PDATA_SZ ARRAY_SIZE(omap2420_mcbsp_pdata) @@ -183,7 +86,8 @@ static struct omap_mcbsp_platform_data omap2430_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP1_IRQ_RX, .tx_irq = INT_24XX_MCBSP1_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP24XX_MCBSP2_BASE, @@ -192,7 +96,8 @@ static struct omap_mcbsp_platform_data omap2430_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP2_IRQ_RX, .tx_irq = INT_24XX_MCBSP2_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP2430_MCBSP3_BASE, @@ -201,7 +106,8 @@ static struct omap_mcbsp_platform_data omap2430_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP3_IRQ_RX, .tx_irq = INT_24XX_MCBSP3_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP2430_MCBSP4_BASE, @@ -210,7 +116,8 @@ static struct omap_mcbsp_platform_data omap2430_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP4_IRQ_RX, .tx_irq = INT_24XX_MCBSP4_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP2430_MCBSP5_BASE, @@ -219,7 +126,8 @@ static struct omap_mcbsp_platform_data omap2430_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP5_IRQ_RX, .tx_irq = INT_24XX_MCBSP5_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, }; #define OMAP2430_MCBSP_PDATA_SZ ARRAY_SIZE(omap2430_mcbsp_pdata) @@ -237,7 +145,8 @@ static struct omap_mcbsp_platform_data omap34xx_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP1_IRQ_RX, .tx_irq = INT_24XX_MCBSP1_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP34XX_MCBSP2_BASE, @@ -246,7 +155,8 @@ static struct omap_mcbsp_platform_data omap34xx_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP2_IRQ_RX, .tx_irq = INT_24XX_MCBSP2_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP34XX_MCBSP3_BASE, @@ -255,7 +165,8 @@ static struct omap_mcbsp_platform_data omap34xx_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP3_IRQ_RX, .tx_irq = INT_24XX_MCBSP3_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP34XX_MCBSP4_BASE, @@ -264,7 +175,8 @@ static struct omap_mcbsp_platform_data omap34xx_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP4_IRQ_RX, .tx_irq = INT_24XX_MCBSP4_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, { .phys_base = OMAP34XX_MCBSP5_BASE, @@ -273,7 +185,8 @@ static struct omap_mcbsp_platform_data omap34xx_mcbsp_pdata[] = { .rx_irq = INT_24XX_MCBSP5_IRQ_RX, .tx_irq = INT_24XX_MCBSP5_IRQ_TX, .ops = &omap2_mcbsp_ops, - .clk_name = "mcbsp_clk", + .clk_names = clk_names, + .num_clks = 2, }, }; #define OMAP34XX_MCBSP_PDATA_SZ ARRAY_SIZE(omap34xx_mcbsp_pdata) @@ -284,14 +197,6 @@ static struct omap_mcbsp_platform_data omap34xx_mcbsp_pdata[] = { static int __init omap2_mcbsp_init(void) { - int i; - - for (i = 0; i < omap_mcbsp_clks_size; i++) { - /* Once we call clk_get inside init, we do not register it */ - omap_mcbsp_clk_init(&omap_mcbsp_clks[i]); - clk_register(&omap_mcbsp_clks[i].clk); - } - if (cpu_is_omap2420()) omap_mcbsp_count = OMAP2420_MCBSP_PDATA_SZ; if (cpu_is_omap2430()) diff --git a/arch/arm/mach-omap2/sleep24xx.S b/arch/arm/mach-omap2/sleep24xx.S index 43336b93b21c..bf9e96105e11 100644 --- a/arch/arm/mach-omap2/sleep24xx.S +++ b/arch/arm/mach-omap2/sleep24xx.S @@ -93,9 +93,8 @@ ENTRY(omap24xx_cpu_suspend) orr r4, r4, #0x40 @ enable self refresh on idle req mov r5, #0x2000 @ set delay (DPLL relock + DLL relock) str r4, [r2] @ make it so - mov r2, #0 nop - mcr p15, 0, r2, c7, c0, 4 @ wait for interrupt + mcr p15, 0, r3, c7, c0, 4 @ wait for interrupt nop loop: subs r5, r5, #0x1 @ awake, wait just a bit diff --git a/arch/arm/mach-omap2/timer-gp.c b/arch/arm/mach-omap2/timer-gp.c index ae6036300f60..9fc13a2cc3f4 100644 --- a/arch/arm/mach-omap2/timer-gp.c +++ b/arch/arm/mach-omap2/timer-gp.c @@ -118,7 +118,8 @@ static void __init omap2_gp_clockevent_init(void) clockevent_gpt.max_delta_ns = clockevent_delta2ns(0xffffffff, &clockevent_gpt); clockevent_gpt.min_delta_ns = - clockevent_delta2ns(1, &clockevent_gpt); + clockevent_delta2ns(3, &clockevent_gpt); + /* Timer internal resynch latency. */ clockevent_gpt.cpumask = cpumask_of(0); clockevents_register_device(&clockevent_gpt); diff --git a/arch/arm/mach-sa1100/generic.c b/arch/arm/mach-sa1100/generic.c index c1fbd5b5f9c4..23cfdd593954 100644 --- a/arch/arm/mach-sa1100/generic.c +++ b/arch/arm/mach-sa1100/generic.c @@ -289,7 +289,7 @@ static struct platform_device sa11x0pcmcia_device = { }; static struct platform_device sa11x0mtd_device = { - .name = "flash", + .name = "sa1100-mtd", .id = -1, }; diff --git a/arch/arm/mm/fault-armv.c b/arch/arm/mm/fault-armv.c index 81d0b8772de3..bc0099d5ae85 100644 --- a/arch/arm/mm/fault-armv.c +++ b/arch/arm/mm/fault-armv.c @@ -66,7 +66,10 @@ static int adjust_pte(struct vm_area_struct *vma, unsigned long address) * fault (ie, is old), we can safely ignore any issues. */ if (ret && (pte_val(entry) & L_PTE_MT_MASK) != shared_pte_mask) { - flush_cache_page(vma, address, pte_pfn(entry)); + unsigned long pfn = pte_pfn(entry); + flush_cache_page(vma, address, pfn); + outer_flush_range((pfn << PAGE_SHIFT), + (pfn << PAGE_SHIFT) + PAGE_SIZE); pte_val(entry) &= ~L_PTE_MT_MASK; pte_val(entry) |= shared_pte_mask; set_pte_at(vma->vm_mm, address, pte, entry); diff --git a/arch/arm/plat-omap/devices.c b/arch/arm/plat-omap/devices.c index ac15c23fd5da..208dbb121f47 100644 --- a/arch/arm/plat-omap/devices.c +++ b/arch/arm/plat-omap/devices.c @@ -200,14 +200,15 @@ void omap_mcbsp_register_board_cfg(struct omap_mcbsp_platform_data *config, /* * Register MMC devices. Called from mach-omap1 and mach-omap2 device init. */ -int __init omap_mmc_add(int id, unsigned long base, unsigned long size, - unsigned int irq, struct omap_mmc_platform_data *data) +int __init omap_mmc_add(const char *name, int id, unsigned long base, + unsigned long size, unsigned int irq, + struct omap_mmc_platform_data *data) { struct platform_device *pdev; struct resource res[OMAP_MMC_NR_RES]; int ret; - pdev = platform_device_alloc("mmci-omap", id); + pdev = platform_device_alloc(name, id); if (!pdev) return -ENOMEM; diff --git a/arch/arm/plat-omap/dma.c b/arch/arm/plat-omap/dma.c index e77373c39f8c..47ec77af4ccb 100644 --- a/arch/arm/plat-omap/dma.c +++ b/arch/arm/plat-omap/dma.c @@ -709,6 +709,7 @@ int omap_request_dma(int dev_id, const char *dev_name, chan->dev_name = dev_name; chan->callback = callback; chan->data = data; + chan->flags = 0; #ifndef CONFIG_ARCH_OMAP1 if (cpu_class_is_omap2()) { @@ -1888,11 +1889,11 @@ static int omap2_dma_handle_ch(int ch) status = dma_read(CSR(ch)); } + dma_write(status, CSR(ch)); + if (likely(dma_chan[ch].callback != NULL)) dma_chan[ch].callback(ch, status, dma_chan[ch].data); - dma_write(status, CSR(ch)); - return 0; } diff --git a/arch/arm/plat-omap/include/mach/cpu.h b/arch/arm/plat-omap/include/mach/cpu.h index b2062f1175de..a8e1178a9468 100644 --- a/arch/arm/plat-omap/include/mach/cpu.h +++ b/arch/arm/plat-omap/include/mach/cpu.h @@ -339,6 +339,7 @@ IS_OMAP_TYPE(3430, 0x3430) #define OMAP3430_REV_ES2_0 0x34301034 #define OMAP3430_REV_ES2_1 0x34302034 #define OMAP3430_REV_ES3_0 0x34303034 +#define OMAP3430_REV_ES3_1 0x34304034 /* * omap_chip bits diff --git a/arch/arm/plat-omap/include/mach/mcbsp.h b/arch/arm/plat-omap/include/mach/mcbsp.h index eef873db3d48..113c2466c86a 100644 --- a/arch/arm/plat-omap/include/mach/mcbsp.h +++ b/arch/arm/plat-omap/include/mach/mcbsp.h @@ -344,7 +344,8 @@ struct omap_mcbsp_platform_data { u8 dma_rx_sync, dma_tx_sync; u16 rx_irq, tx_irq; struct omap_mcbsp_ops *ops; - char const *clk_name; + char const **clk_names; + int num_clks; }; struct omap_mcbsp { @@ -376,7 +377,8 @@ struct omap_mcbsp { /* Protect the field .free, while checking if the mcbsp is in use */ spinlock_t lock; struct omap_mcbsp_platform_data *pdata; - struct clk *clk; + struct clk **clks; + int num_clks; }; extern struct omap_mcbsp **mcbsp_ptr; extern int omap_mcbsp_count; diff --git a/arch/arm/plat-omap/include/mach/mmc.h b/arch/arm/plat-omap/include/mach/mmc.h index 031250f02805..73a9e15031b1 100644 --- a/arch/arm/plat-omap/include/mach/mmc.h +++ b/arch/arm/plat-omap/include/mach/mmc.h @@ -115,8 +115,9 @@ void omap1_init_mmc(struct omap_mmc_platform_data **mmc_data, int nr_controllers); void omap2_init_mmc(struct omap_mmc_platform_data **mmc_data, int nr_controllers); -int omap_mmc_add(int id, unsigned long base, unsigned long size, - unsigned int irq, struct omap_mmc_platform_data *data); +int omap_mmc_add(const char *name, int id, unsigned long base, + unsigned long size, unsigned int irq, + struct omap_mmc_platform_data *data); #else static inline void omap1_init_mmc(struct omap_mmc_platform_data **mmc_data, int nr_controllers) @@ -126,8 +127,9 @@ static inline void omap2_init_mmc(struct omap_mmc_platform_data **mmc_data, int nr_controllers) { } -static inline int omap_mmc_add(int id, unsigned long base, unsigned long size, - unsigned int irq, struct omap_mmc_platform_data *data) +static inline int omap_mmc_add(const char *name, int id, unsigned long base, + unsigned long size, unsigned int irq, + struct omap_mmc_platform_data *data) { return 0; } diff --git a/arch/arm/plat-omap/mcbsp.c b/arch/arm/plat-omap/mcbsp.c index f2401a831f99..e5842e30e534 100644 --- a/arch/arm/plat-omap/mcbsp.c +++ b/arch/arm/plat-omap/mcbsp.c @@ -214,6 +214,7 @@ EXPORT_SYMBOL(omap_mcbsp_set_io_type); int omap_mcbsp_request(unsigned int id) { struct omap_mcbsp *mcbsp; + int i; int err; if (!omap_mcbsp_check_valid_id(id)) { @@ -225,7 +226,8 @@ int omap_mcbsp_request(unsigned int id) if (mcbsp->pdata && mcbsp->pdata->ops && mcbsp->pdata->ops->request) mcbsp->pdata->ops->request(id); - clk_enable(mcbsp->clk); + for (i = 0; i < mcbsp->num_clks; i++) + clk_enable(mcbsp->clks[i]); spin_lock(&mcbsp->lock); if (!mcbsp->free) { @@ -276,6 +278,7 @@ EXPORT_SYMBOL(omap_mcbsp_request); void omap_mcbsp_free(unsigned int id) { struct omap_mcbsp *mcbsp; + int i; if (!omap_mcbsp_check_valid_id(id)) { printk(KERN_ERR "%s: Invalid id (%d)\n", __func__, id + 1); @@ -286,7 +289,8 @@ void omap_mcbsp_free(unsigned int id) if (mcbsp->pdata && mcbsp->pdata->ops && mcbsp->pdata->ops->free) mcbsp->pdata->ops->free(id); - clk_disable(mcbsp->clk); + for (i = mcbsp->num_clks - 1; i >= 0; i--) + clk_disable(mcbsp->clks[i]); spin_lock(&mcbsp->lock); if (mcbsp->free) { @@ -872,6 +876,7 @@ static int __devinit omap_mcbsp_probe(struct platform_device *pdev) struct omap_mcbsp_platform_data *pdata = pdev->dev.platform_data; struct omap_mcbsp *mcbsp; int id = pdev->id - 1; + int i; int ret = 0; if (!pdata) { @@ -916,14 +921,25 @@ static int __devinit omap_mcbsp_probe(struct platform_device *pdev) mcbsp->dma_rx_sync = pdata->dma_rx_sync; mcbsp->dma_tx_sync = pdata->dma_tx_sync; - if (pdata->clk_name) - mcbsp->clk = clk_get(&pdev->dev, pdata->clk_name); - if (IS_ERR(mcbsp->clk)) { - dev_err(&pdev->dev, - "Invalid clock configuration for McBSP%d.\n", - mcbsp->id); - ret = PTR_ERR(mcbsp->clk); - goto err_clk; + if (pdata->num_clks) { + mcbsp->num_clks = pdata->num_clks; + mcbsp->clks = kzalloc(mcbsp->num_clks * sizeof(struct clk *), + GFP_KERNEL); + if (!mcbsp->clks) { + ret = -ENOMEM; + goto exit; + } + for (i = 0; i < mcbsp->num_clks; i++) { + mcbsp->clks[i] = clk_get(&pdev->dev, pdata->clk_names[i]); + if (IS_ERR(mcbsp->clks[i])) { + dev_err(&pdev->dev, + "Invalid %s configuration for McBSP%d.\n", + pdata->clk_names[i], mcbsp->id); + ret = PTR_ERR(mcbsp->clks[i]); + goto err_clk; + } + } + } mcbsp->pdata = pdata; @@ -932,6 +948,9 @@ static int __devinit omap_mcbsp_probe(struct platform_device *pdev) return 0; err_clk: + while (i--) + clk_put(mcbsp->clks[i]); + kfree(mcbsp->clks); iounmap(mcbsp->io_base); err_ioremap: mcbsp->free = 0; @@ -942,6 +961,7 @@ exit: static int __devexit omap_mcbsp_remove(struct platform_device *pdev) { struct omap_mcbsp *mcbsp = platform_get_drvdata(pdev); + int i; platform_set_drvdata(pdev, NULL); if (mcbsp) { @@ -950,12 +970,18 @@ static int __devexit omap_mcbsp_remove(struct platform_device *pdev) mcbsp->pdata->ops->free) mcbsp->pdata->ops->free(mcbsp->id); - clk_disable(mcbsp->clk); - clk_put(mcbsp->clk); + for (i = mcbsp->num_clks - 1; i >= 0; i--) { + clk_disable(mcbsp->clks[i]); + clk_put(mcbsp->clks[i]); + } iounmap(mcbsp->io_base); - mcbsp->clk = NULL; + if (mcbsp->num_clks) { + kfree(mcbsp->clks); + mcbsp->clks = NULL; + mcbsp->num_clks = 0; + } mcbsp->free = 0; mcbsp->dev = NULL; } diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 77bba4c083cb..a603bbf9b1b7 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -61,9 +61,14 @@ #define EM_MSG_LED_VALUE_ON 0x00010000 static int ahci_skip_host_reset; +static int ahci_ignore_sss; + module_param_named(skip_host_reset, ahci_skip_host_reset, int, 0444); MODULE_PARM_DESC(skip_host_reset, "skip global host reset (0=don't skip, 1=skip)"); +module_param_named(ignore_sss, ahci_ignore_sss, int, 0444); +MODULE_PARM_DESC(ignore_sss, "Ignore staggered spinup flag (0=don't ignore, 1=ignore)"); + static int ahci_enable_alpm(struct ata_port *ap, enum link_pm policy); static void ahci_disable_alpm(struct ata_port *ap); @@ -2692,8 +2697,10 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) host->iomap = pcim_iomap_table(pdev); host->private_data = hpriv; - if (!(hpriv->cap & HOST_CAP_SSS)) + if (!(hpriv->cap & HOST_CAP_SSS) || ahci_ignore_sss) host->flags |= ATA_HOST_PARALLEL_SCAN; + else + printk(KERN_INFO "ahci: SSS flag set, parallel bus scan disabled\n"); if (pi.flags & ATA_FLAG_EM) ahci_reset_em(host); diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 88c242856dae..9fbf0595f3d4 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -164,6 +164,11 @@ MODULE_LICENSE("GPL"); MODULE_VERSION(DRV_VERSION); +static bool ata_sstatus_online(u32 sstatus) +{ + return (sstatus & 0xf) == 0x3; +} + /** * ata_link_next - link iteration helper * @link: the previous link, NULL to start @@ -1015,18 +1020,6 @@ static const char *sata_spd_string(unsigned int spd) return spd_str[spd - 1]; } -void ata_dev_disable(struct ata_device *dev) -{ - if (ata_dev_enabled(dev)) { - if (ata_msg_drv(dev->link->ap)) - ata_dev_printk(dev, KERN_WARNING, "disabled\n"); - ata_acpi_on_disable(dev); - ata_down_xfermask_limit(dev, ATA_DNXFER_FORCE_PIO0 | - ATA_DNXFER_QUIET); - dev->class++; - } -} - static int ata_dev_set_dipm(struct ata_device *dev, enum link_pm policy) { struct ata_link *link = dev->link; @@ -2239,6 +2232,40 @@ retry: return rc; } +static int ata_do_link_spd_horkage(struct ata_device *dev) +{ + struct ata_link *plink = ata_dev_phys_link(dev); + u32 target, target_limit; + + if (!sata_scr_valid(plink)) + return 0; + + if (dev->horkage & ATA_HORKAGE_1_5_GBPS) + target = 1; + else + return 0; + + target_limit = (1 << target) - 1; + + /* if already on stricter limit, no need to push further */ + if (plink->sata_spd_limit <= target_limit) + return 0; + + plink->sata_spd_limit = target_limit; + + /* Request another EH round by returning -EAGAIN if link is + * going faster than the target speed. Forward progress is + * guaranteed by setting sata_spd_limit to target_limit above. + */ + if (plink->sata_spd > target) { + ata_dev_printk(dev, KERN_INFO, + "applying link speed limit horkage to %s\n", + sata_spd_string(target)); + return -EAGAIN; + } + return 0; +} + static inline u8 ata_dev_knobble(struct ata_device *dev) { struct ata_port *ap = dev->link->ap; @@ -2329,6 +2356,10 @@ int ata_dev_configure(struct ata_device *dev) return 0; } + rc = ata_do_link_spd_horkage(dev); + if (rc) + return rc; + /* let ACPI work its magic */ rc = ata_acpi_on_devcfg(dev); if (rc) @@ -2784,7 +2815,7 @@ int ata_bus_probe(struct ata_port *ap) /* This is the last chance, better to slow * down than lose it. */ - sata_down_spd_limit(&ap->link); + sata_down_spd_limit(&ap->link, 0); ata_down_xfermask_limit(dev, ATA_DNXFER_PIO); } } @@ -2880,21 +2911,27 @@ void ata_port_disable(struct ata_port *ap) /** * sata_down_spd_limit - adjust SATA spd limit downward * @link: Link to adjust SATA spd limit for + * @spd_limit: Additional limit * * Adjust SATA spd limit of @link downward. Note that this * function only adjusts the limit. The change must be applied * using sata_set_spd(). * + * If @spd_limit is non-zero, the speed is limited to equal to or + * lower than @spd_limit if such speed is supported. If + * @spd_limit is slower than any supported speed, only the lowest + * supported speed is allowed. + * * LOCKING: * Inherited from caller. * * RETURNS: * 0 on success, negative errno on failure */ -int sata_down_spd_limit(struct ata_link *link) +int sata_down_spd_limit(struct ata_link *link, u32 spd_limit) { u32 sstatus, spd, mask; - int rc, highbit; + int rc, bit; if (!sata_scr_valid(link)) return -EOPNOTSUPP; @@ -2903,7 +2940,7 @@ int sata_down_spd_limit(struct ata_link *link) * If not, use cached value in link->sata_spd. */ rc = sata_scr_read(link, SCR_STATUS, &sstatus); - if (rc == 0) + if (rc == 0 && ata_sstatus_online(sstatus)) spd = (sstatus >> 4) & 0xf; else spd = link->sata_spd; @@ -2913,8 +2950,8 @@ int sata_down_spd_limit(struct ata_link *link) return -EINVAL; /* unconditionally mask off the highest bit */ - highbit = fls(mask) - 1; - mask &= ~(1 << highbit); + bit = fls(mask) - 1; + mask &= ~(1 << bit); /* Mask off all speeds higher than or equal to the current * one. Force 1.5Gbps if current SPD is not available. @@ -2928,6 +2965,15 @@ int sata_down_spd_limit(struct ata_link *link) if (!mask) return -EINVAL; + if (spd_limit) { + if (mask & ((1 << spd_limit) - 1)) + mask &= (1 << spd_limit) - 1; + else { + bit = ffs(mask) - 1; + mask = 1 << bit; + } + } + link->sata_spd_limit = mask; ata_link_printk(link, KERN_WARNING, "limiting SATA link speed to %s\n", @@ -4215,6 +4261,9 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = { /* Devices that do not need bridging limits applied */ { "MTRON MSP-SATA*", NULL, ATA_HORKAGE_BRIDGE_OK, }, + /* Devices which aren't very happy with higher link speeds */ + { "WD My Book", NULL, ATA_HORKAGE_1_5_GBPS, }, + /* End Marker */ { } }; @@ -4709,8 +4758,7 @@ void swap_buf_le16(u16 *buf, unsigned int buf_words) /** * ata_qc_new - Request an available ATA command, for queueing - * @ap: Port associated with device @dev - * @dev: Device from whom we request an available command structure + * @ap: target port * * LOCKING: * None. @@ -5175,7 +5223,7 @@ bool ata_phys_link_online(struct ata_link *link) u32 sstatus; if (sata_scr_read(link, SCR_STATUS, &sstatus) == 0 && - (sstatus & 0xf) == 0x3) + ata_sstatus_online(sstatus)) return true; return false; } @@ -5199,7 +5247,7 @@ bool ata_phys_link_offline(struct ata_link *link) u32 sstatus; if (sata_scr_read(link, SCR_STATUS, &sstatus) == 0 && - (sstatus & 0xf) != 0x3) + !ata_sstatus_online(sstatus)) return true; return false; } @@ -5412,8 +5460,8 @@ void ata_dev_init(struct ata_device *dev) dev->horkage = 0; spin_unlock_irqrestore(ap->lock, flags); - memset((void *)dev + ATA_DEVICE_CLEAR_OFFSET, 0, - sizeof(*dev) - ATA_DEVICE_CLEAR_OFFSET); + memset((void *)dev + ATA_DEVICE_CLEAR_BEGIN, 0, + ATA_DEVICE_CLEAR_END - ATA_DEVICE_CLEAR_BEGIN); dev->pio_mask = UINT_MAX; dev->mwdma_mask = UINT_MAX; dev->udma_mask = UINT_MAX; diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index 8147a8386370..ce2ef0475339 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -82,6 +82,10 @@ enum { ATA_EH_FASTDRAIN_INTERVAL = 3000, ATA_EH_UA_TRIES = 5, + + /* probe speed down parameters, see ata_eh_schedule_probe() */ + ATA_EH_PROBE_TRIAL_INTERVAL = 60000, /* 1 min */ + ATA_EH_PROBE_TRIALS = 2, }; /* The following table determines how we sequence resets. Each entry @@ -1176,6 +1180,32 @@ void ata_eh_qc_retry(struct ata_queued_cmd *qc) } /** + * ata_dev_disable - disable ATA device + * @dev: ATA device to disable + * + * Disable @dev. + * + * Locking: + * EH context. + */ +void ata_dev_disable(struct ata_device *dev) +{ + if (!ata_dev_enabled(dev)) + return; + + if (ata_msg_drv(dev->link->ap)) + ata_dev_printk(dev, KERN_WARNING, "disabled\n"); + ata_acpi_on_disable(dev); + ata_down_xfermask_limit(dev, ATA_DNXFER_FORCE_PIO0 | ATA_DNXFER_QUIET); + dev->class++; + + /* From now till the next successful probe, ering is used to + * track probe failures. Clear accumulated device error info. + */ + ata_ering_clear(&dev->ering); +} + +/** * ata_eh_detach_dev - detach ATA device * @dev: ATA device to detach * @@ -1849,7 +1879,7 @@ static unsigned int ata_eh_speed_down(struct ata_device *dev, /* speed down? */ if (verdict & ATA_EH_SPDN_SPEED_DOWN) { /* speed down SATA link speed if possible */ - if (sata_down_spd_limit(link) == 0) { + if (sata_down_spd_limit(link, 0) == 0) { action |= ATA_EH_RESET; goto done; } @@ -2601,11 +2631,11 @@ int ata_eh_reset(struct ata_link *link, int classify, } if (try == max_tries - 1) { - sata_down_spd_limit(link); + sata_down_spd_limit(link, 0); if (slave) - sata_down_spd_limit(slave); + sata_down_spd_limit(slave, 0); } else if (rc == -EPIPE) - sata_down_spd_limit(failed_link); + sata_down_spd_limit(failed_link, 0); if (hardreset) reset = hardreset; @@ -2744,6 +2774,8 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, readid_flags, dev->id); switch (rc) { case 0: + /* clear error info accumulated during probe */ + ata_ering_clear(&dev->ering); new_mask |= 1 << dev->devno; break; case -ENOENT: @@ -2947,9 +2979,24 @@ static int ata_eh_skip_recovery(struct ata_link *link) return 1; } +static int ata_count_probe_trials_cb(struct ata_ering_entry *ent, void *void_arg) +{ + u64 interval = msecs_to_jiffies(ATA_EH_PROBE_TRIAL_INTERVAL); + u64 now = get_jiffies_64(); + int *trials = void_arg; + + if (ent->timestamp < now - min(now, interval)) + return -1; + + (*trials)++; + return 0; +} + static int ata_eh_schedule_probe(struct ata_device *dev) { struct ata_eh_context *ehc = &dev->link->eh_context; + struct ata_link *link = ata_dev_phys_link(dev); + int trials = 0; if (!(ehc->i.probe_mask & (1 << dev->devno)) || (ehc->did_probe_mask & (1 << dev->devno))) @@ -2962,6 +3009,25 @@ static int ata_eh_schedule_probe(struct ata_device *dev) ehc->saved_xfer_mode[dev->devno] = 0; ehc->saved_ncq_enabled &= ~(1 << dev->devno); + /* Record and count probe trials on the ering. The specific + * error mask used is irrelevant. Because a successful device + * detection clears the ering, this count accumulates only if + * there are consecutive failed probes. + * + * If the count is equal to or higher than ATA_EH_PROBE_TRIALS + * in the last ATA_EH_PROBE_TRIAL_INTERVAL, link speed is + * forced to 1.5Gbps. + * + * This is to work around cases where failed link speed + * negotiation results in device misdetection leading to + * infinite DEVXCHG or PHRDY CHG events. + */ + ata_ering_record(&dev->ering, 0, AC_ERR_OTHER); + ata_ering_map(&dev->ering, ata_count_probe_trials_cb, &trials); + + if (trials > ATA_EH_PROBE_TRIALS) + sata_down_spd_limit(link, 1); + return 1; } @@ -2969,7 +3035,11 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) { struct ata_eh_context *ehc = &dev->link->eh_context; - ehc->tries[dev->devno]--; + /* -EAGAIN from EH routine indicates retry without prejudice. + * The requester is responsible for ensuring forward progress. + */ + if (err != -EAGAIN) + ehc->tries[dev->devno]--; switch (err) { case -ENODEV: @@ -2979,12 +3049,13 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) /* give it just one more chance */ ehc->tries[dev->devno] = min(ehc->tries[dev->devno], 1); case -EIO: - if (ehc->tries[dev->devno] == 1 && dev->pio_mode > XFER_PIO_0) { + if (ehc->tries[dev->devno] == 1) { /* This is the last chance, better to slow * down than lose it. */ - sata_down_spd_limit(ata_dev_phys_link(dev)); - ata_down_xfermask_limit(dev, ATA_DNXFER_PIO); + sata_down_spd_limit(ata_dev_phys_link(dev), 0); + if (dev->pio_mode > XFER_PIO_0) + ata_down_xfermask_limit(dev, ATA_DNXFER_PIO); } } diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c index 98ca07a2db87..619f2c33950e 100644 --- a/drivers/ata/libata-pmp.c +++ b/drivers/ata/libata-pmp.c @@ -729,7 +729,7 @@ static int sata_pmp_eh_recover_pmp(struct ata_port *ap, if (tries) { /* consecutive revalidation failures? speed down */ if (reval_failed) - sata_down_spd_limit(link); + sata_down_spd_limit(link, 0); else reval_failed = 1; diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c index 3c4c5ae277ba..b9747fa59e54 100644 --- a/drivers/ata/libata-scsi.c +++ b/drivers/ata/libata-scsi.c @@ -415,6 +415,7 @@ int ata_std_bios_param(struct scsi_device *sdev, struct block_device *bdev, /** * ata_get_identity - Handler for HDIO_GET_IDENTITY ioctl + * @ap: target port * @sdev: SCSI device to get identify data for * @arg: User buffer area for identify data * diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h index fe2839e58774..cea8014cd87e 100644 --- a/drivers/ata/libata.h +++ b/drivers/ata/libata.h @@ -79,7 +79,6 @@ extern int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev, u64 block, u32 n_block, unsigned int tf_flags, unsigned int tag); extern u64 ata_tf_read_block(struct ata_taskfile *tf, struct ata_device *dev); -extern void ata_dev_disable(struct ata_device *dev); extern void ata_pio_queue_task(struct ata_port *ap, void *data, unsigned long delay); extern void ata_port_flush_task(struct ata_port *ap); @@ -100,7 +99,7 @@ extern int ata_dev_reread_id(struct ata_device *dev, unsigned int readid_flags); extern int ata_dev_revalidate(struct ata_device *dev, unsigned int new_class, unsigned int readid_flags); extern int ata_dev_configure(struct ata_device *dev); -extern int sata_down_spd_limit(struct ata_link *link); +extern int sata_down_spd_limit(struct ata_link *link, u32 spd_limit); extern int ata_down_xfermask_limit(struct ata_device *dev, unsigned int sel); extern void ata_sg_clean(struct ata_queued_cmd *qc); extern void ata_qc_free(struct ata_queued_cmd *qc); @@ -160,6 +159,7 @@ extern void ata_scsi_error(struct Scsi_Host *host); extern void ata_port_wait_eh(struct ata_port *ap); extern void ata_eh_fastdrain_timerfn(unsigned long arg); extern void ata_qc_schedule_eh(struct ata_queued_cmd *qc); +extern void ata_dev_disable(struct ata_device *dev); extern void ata_eh_detach_dev(struct ata_device *dev); extern void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev, unsigned int action); diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c index 3080f371222c..f1b26f7c8e4d 100644 --- a/drivers/ata/pata_qdi.c +++ b/drivers/ata/pata_qdi.c @@ -12,7 +12,7 @@ * * Probe code based on drivers/ide/legacy/qd65xx.c * Rewritten from the work of Colten Edwards <pje120@cs.usask.ca> by - * Samuel Thibault <samuel.thibault@fnac.net> + * Samuel Thibault <samuel.thibault@ens-lyon.org> */ #include <linux/kernel.h> diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c index f2d8a020ea53..4ae1a4138b47 100644 --- a/drivers/ata/sata_mv.c +++ b/drivers/ata/sata_mv.c @@ -663,8 +663,8 @@ static const struct pci_device_id mv_pci_tbl[] = { { PCI_VDEVICE(MARVELL, 0x5081), chip_508x }, /* RocketRAID 1720/174x have different identifiers */ { PCI_VDEVICE(TTI, 0x1720), chip_6042 }, - { PCI_VDEVICE(TTI, 0x1740), chip_508x }, - { PCI_VDEVICE(TTI, 0x1742), chip_508x }, + { PCI_VDEVICE(TTI, 0x1740), chip_6042 }, + { PCI_VDEVICE(TTI, 0x1742), chip_6042 }, { PCI_VDEVICE(MARVELL, 0x6040), chip_604x }, { PCI_VDEVICE(MARVELL, 0x6041), chip_604x }, diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c index c49ad0e61b6f..444af0415ca1 100644 --- a/drivers/ata/sata_nv.c +++ b/drivers/ata/sata_nv.c @@ -436,11 +436,16 @@ static struct ata_port_operations nv_nf2_ops = { .hardreset = nv_noclassify_hardreset, }; -/* CK804 finally gets hardreset right */ +/* For initial probing after boot and hot plugging, hardreset mostly + * works fine on CK804 but curiously, reprobing on the initial port by + * rescanning or rmmod/insmod fails to acquire the initial D2H Reg FIS + * in somewhat undeterministic way. Use noclassify hardreset. + */ static struct ata_port_operations nv_ck804_ops = { .inherits = &nv_common_ops, .freeze = nv_ck804_freeze, .thaw = nv_ck804_thaw, + .hardreset = nv_noclassify_hardreset, .host_stop = nv_ck804_host_stop, }; diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c index 9f029595f454..d0091609e210 100644 --- a/drivers/ata/sata_sil.c +++ b/drivers/ata/sata_sil.c @@ -324,7 +324,7 @@ static void sil_fill_sg(struct ata_queued_cmd *qc) prd->addr = cpu_to_le32(addr); prd->flags_len = cpu_to_le32(sg_len); - VPRINTK("PRD[%u] = (0x%X, 0x%X)\n", pi, addr, sg_len); + VPRINTK("PRD[%u] = (0x%X, 0x%X)\n", si, addr, sg_len); last_prd = prd; prd++; diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index f5be8081cd81..735bbe2be51a 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -761,7 +761,7 @@ source "drivers/char/hw_random/Kconfig" config NVRAM tristate "/dev/nvram support" - depends on ATARI || X86 || ARM || GENERIC_NVRAM + depends on ATARI || X86 || (ARM && RTC_DRV_CMOS) || GENERIC_NVRAM ---help--- If you say Y here and create a character special file /dev/nvram with major number 10 and minor number 144 using mknod ("man mknod"), diff --git a/drivers/ide/qd65xx.c b/drivers/ide/qd65xx.c index 5b2e3af43c4b..08c4fa35e9b1 100644 --- a/drivers/ide/qd65xx.c +++ b/drivers/ide/qd65xx.c @@ -16,7 +16,7 @@ /* * Rewritten from the work of Colten Edwards <pje120@cs.usask.ca> by - * Samuel Thibault <samuel.thibault@fnac.net> + * Samuel Thibault <samuel.thibault@ens-lyon.org> */ #include <linux/module.h> diff --git a/drivers/ide/qd65xx.h b/drivers/ide/qd65xx.h index 6636f9665d16..d7e67a1a1dcc 100644 --- a/drivers/ide/qd65xx.h +++ b/drivers/ide/qd65xx.h @@ -4,7 +4,7 @@ /* * Authors: Petr Soucek <petr@ryston.cz> - * Samuel Thibault <samuel.thibault@fnac.net> + * Samuel Thibault <samuel.thibault@ens-lyon.org> */ /* truncates a in [b,c] */ diff --git a/drivers/mfd/pcf50633-core.c b/drivers/mfd/pcf50633-core.c index 24508e28e3fb..ea9488e7ad6d 100644 --- a/drivers/mfd/pcf50633-core.c +++ b/drivers/mfd/pcf50633-core.c @@ -626,7 +626,6 @@ static int __devinit pcf50633_probe(struct i2c_client *client, } if (client->irq) { - set_irq_handler(client->irq, handle_level_irq); ret = request_irq(client->irq, pcf50633_irq, IRQF_TRIGGER_LOW, "pcf50633", pcf); diff --git a/drivers/mtd/maps/sa1100-flash.c b/drivers/mtd/maps/sa1100-flash.c index 7df6bbf0e4d9..6f6a0f6dafd6 100644 --- a/drivers/mtd/maps/sa1100-flash.c +++ b/drivers/mtd/maps/sa1100-flash.c @@ -453,7 +453,7 @@ static struct platform_driver sa1100_mtd_driver = { .resume = sa1100_mtd_resume, .shutdown = sa1100_mtd_shutdown, .driver = { - .name = "flash", + .name = "sa1100-mtd", .owner = THIS_MODULE, }, }; @@ -474,4 +474,4 @@ module_exit(sa1100_mtd_exit); MODULE_AUTHOR("Nicolas Pitre"); MODULE_DESCRIPTION("SA1100 CFI map driver"); MODULE_LICENSE("GPL"); -MODULE_ALIAS("platform:flash"); +MODULE_ALIAS("platform:sa1100-mtd"); diff --git a/drivers/net/arm/etherh.c b/drivers/net/arm/etherh.c index d15d8b79d8e5..54b52e5b1821 100644 --- a/drivers/net/arm/etherh.c +++ b/drivers/net/arm/etherh.c @@ -646,7 +646,7 @@ static const struct net_device_ops etherh_netdev_ops = { .ndo_get_stats = ei_get_stats, .ndo_set_multicast_list = ei_set_multicast_list, .ndo_validate_addr = eth_validate_addr, - .ndo_set_mac_address = eth_set_mac_addr, + .ndo_set_mac_address = eth_mac_addr, .ndo_change_mtu = eth_change_mtu, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = ei_poll, diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 8e93341f3e82..9c2358391147 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -553,12 +553,24 @@ static void detach_groups(struct config_group *group) child = sd->s_dentry; + /* + * Note: we hide this from lockdep since we have no way + * to teach lockdep about recursive + * I_MUTEX_PARENT -> I_MUTEX_CHILD patterns along a path + * in an inode tree, which are valid as soon as + * I_MUTEX_PARENT -> I_MUTEX_CHILD is valid from a + * parent inode to one of its children. + */ + lockdep_off(); mutex_lock(&child->d_inode->i_mutex); + lockdep_on(); configfs_detach_group(sd->s_element); child->d_inode->i_flags |= S_DEAD; + lockdep_off(); mutex_unlock(&child->d_inode->i_mutex); + lockdep_on(); d_delete(child); dput(child); @@ -748,11 +760,22 @@ static int configfs_attach_item(struct config_item *parent_item, * We are going to remove an inode and its dentry but * the VFS may already have hit and used them. Thus, * we must lock them as rmdir() would. + * + * Note: we hide this from lockdep since we have no way + * to teach lockdep about recursive + * I_MUTEX_PARENT -> I_MUTEX_CHILD patterns along a path + * in an inode tree, which are valid as soon as + * I_MUTEX_PARENT -> I_MUTEX_CHILD is valid from a + * parent inode to one of its children. */ + lockdep_off(); mutex_lock(&dentry->d_inode->i_mutex); + lockdep_on(); configfs_remove_dir(item); dentry->d_inode->i_flags |= S_DEAD; + lockdep_off(); mutex_unlock(&dentry->d_inode->i_mutex); + lockdep_on(); d_delete(dentry); } } @@ -787,14 +810,25 @@ static int configfs_attach_group(struct config_item *parent_item, * * We must also lock the inode to remove it safely in case of * error, as rmdir() would. + * + * Note: we hide this from lockdep since we have no way + * to teach lockdep about recursive + * I_MUTEX_PARENT -> I_MUTEX_CHILD patterns along a path + * in an inode tree, which are valid as soon as + * I_MUTEX_PARENT -> I_MUTEX_CHILD is valid from a + * parent inode to one of its children. */ + lockdep_off(); mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_CHILD); + lockdep_on(); ret = populate_groups(to_config_group(item)); if (ret) { configfs_detach_item(item); dentry->d_inode->i_flags |= S_DEAD; } + lockdep_off(); mutex_unlock(&dentry->d_inode->i_mutex); + lockdep_on(); if (ret) d_delete(dentry); } @@ -956,7 +990,17 @@ static int configfs_depend_prep(struct dentry *origin, BUG_ON(!origin || !sd); /* Lock this guy on the way down */ + /* + * Note: we hide this from lockdep since we have no way + * to teach lockdep about recursive + * I_MUTEX_PARENT -> I_MUTEX_CHILD patterns along a path + * in an inode tree, which are valid as soon as + * I_MUTEX_PARENT -> I_MUTEX_CHILD is valid from a + * parent inode to one of its children. + */ + lockdep_off(); mutex_lock(&sd->s_dentry->d_inode->i_mutex); + lockdep_on(); if (sd->s_element == target) /* Boo-yah */ goto out; @@ -970,7 +1014,9 @@ static int configfs_depend_prep(struct dentry *origin, } /* We looped all our children and didn't find target */ + lockdep_off(); mutex_unlock(&sd->s_dentry->d_inode->i_mutex); + lockdep_on(); ret = -ENOENT; out: @@ -990,11 +1036,16 @@ static void configfs_depend_rollback(struct dentry *origin, struct dentry *dentry = item->ci_dentry; while (dentry != origin) { + /* See comments in configfs_depend_prep() */ + lockdep_off(); mutex_unlock(&dentry->d_inode->i_mutex); + lockdep_on(); dentry = dentry->d_parent; } + lockdep_off(); mutex_unlock(&origin->d_inode->i_mutex); + lockdep_on(); } int configfs_depend_item(struct configfs_subsystem *subsys, @@ -1329,8 +1380,16 @@ static int configfs_rmdir(struct inode *dir, struct dentry *dentry) } /* Wait until the racing operation terminates */ + /* + * Note: we hide this from lockdep since we are locked + * with subclass I_MUTEX_NORMAL from vfs_rmdir() (why + * not I_MUTEX_CHILD?), and I_MUTEX_XATTR or + * I_MUTEX_QUOTA are not relevant for the locked inode. + */ + lockdep_off(); mutex_lock(wait_mutex); mutex_unlock(wait_mutex); + lockdep_on(); } } while (ret == -EAGAIN); diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index d861096c9d81..60fe74035db5 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -5390,6 +5390,9 @@ int ocfs2_remove_btree_range(struct inode *inode, goto out; } + vfs_dq_free_space_nodirty(inode, + ocfs2_clusters_to_bytes(inode->i_sb, len)); + ret = ocfs2_remove_extent(inode, et, cpos, len, handle, meta_ac, dealloc); if (ret) { diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index b1cc7c381e88..e9d7c2038c0f 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -38,6 +38,7 @@ #include "dlmglue.h" #include "file.h" #include "inode.h" +#include "super.h" static int ocfs2_dentry_revalidate(struct dentry *dentry, @@ -294,6 +295,34 @@ out_attach: return ret; } +static DEFINE_SPINLOCK(dentry_list_lock); + +/* We limit the number of dentry locks to drop in one go. We have + * this limit so that we don't starve other users of ocfs2_wq. */ +#define DL_INODE_DROP_COUNT 64 + +/* Drop inode references from dentry locks */ +void ocfs2_drop_dl_inodes(struct work_struct *work) +{ + struct ocfs2_super *osb = container_of(work, struct ocfs2_super, + dentry_lock_work); + struct ocfs2_dentry_lock *dl; + int drop_count = DL_INODE_DROP_COUNT; + + spin_lock(&dentry_list_lock); + while (osb->dentry_lock_list && drop_count--) { + dl = osb->dentry_lock_list; + osb->dentry_lock_list = dl->dl_next; + spin_unlock(&dentry_list_lock); + iput(dl->dl_inode); + kfree(dl); + spin_lock(&dentry_list_lock); + } + if (osb->dentry_lock_list) + queue_work(ocfs2_wq, &osb->dentry_lock_work); + spin_unlock(&dentry_list_lock); +} + /* * ocfs2_dentry_iput() and friends. * @@ -318,16 +347,23 @@ out_attach: static void ocfs2_drop_dentry_lock(struct ocfs2_super *osb, struct ocfs2_dentry_lock *dl) { - iput(dl->dl_inode); ocfs2_simple_drop_lockres(osb, &dl->dl_lockres); ocfs2_lock_res_free(&dl->dl_lockres); - kfree(dl); + + /* We leave dropping of inode reference to ocfs2_wq as that can + * possibly lead to inode deletion which gets tricky */ + spin_lock(&dentry_list_lock); + if (!osb->dentry_lock_list) + queue_work(ocfs2_wq, &osb->dentry_lock_work); + dl->dl_next = osb->dentry_lock_list; + osb->dentry_lock_list = dl; + spin_unlock(&dentry_list_lock); } void ocfs2_dentry_lock_put(struct ocfs2_super *osb, struct ocfs2_dentry_lock *dl) { - int unlock = 0; + int unlock; BUG_ON(dl->dl_count == 0); diff --git a/fs/ocfs2/dcache.h b/fs/ocfs2/dcache.h index c091c34d9883..d06e16c06640 100644 --- a/fs/ocfs2/dcache.h +++ b/fs/ocfs2/dcache.h @@ -29,8 +29,13 @@ extern struct dentry_operations ocfs2_dentry_ops; struct ocfs2_dentry_lock { + /* Use count of dentry lock */ unsigned int dl_count; - u64 dl_parent_blkno; + union { + /* Linked list of dentry locks to release */ + struct ocfs2_dentry_lock *dl_next; + u64 dl_parent_blkno; + }; /* * The ocfs2_dentry_lock keeps an inode reference until @@ -47,6 +52,8 @@ int ocfs2_dentry_attach_lock(struct dentry *dentry, struct inode *inode, void ocfs2_dentry_lock_put(struct ocfs2_super *osb, struct ocfs2_dentry_lock *dl); +void ocfs2_drop_dl_inodes(struct work_struct *work); + struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno, int skip_unhashed); diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index b0c4cadd4c45..206a2370876a 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2860,6 +2860,10 @@ static void ocfs2_unlock_ast(void *opaque, int error) case OCFS2_UNLOCK_CANCEL_CONVERT: mlog(0, "Cancel convert success for %s\n", lockres->l_name); lockres->l_action = OCFS2_AST_INVALID; + /* Downconvert thread may have requeued this lock, we + * need to wake it. */ + if (lockres->l_flags & OCFS2_LOCK_BLOCKED) + ocfs2_wake_downconvert_thread(ocfs2_get_lockres_osb(lockres)); break; case OCFS2_UNLOCK_DROP_LOCK: lockres->l_level = DLM_LOCK_IV; diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index ad5c24a29edd..077384135f4e 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -210,6 +210,7 @@ struct ocfs2_journal; struct ocfs2_slot_info; struct ocfs2_recovery_map; struct ocfs2_quota_recovery; +struct ocfs2_dentry_lock; struct ocfs2_super { struct task_struct *commit_task; @@ -325,6 +326,11 @@ struct ocfs2_super struct list_head blocked_lock_list; unsigned long blocked_lock_count; + /* List of dentry locks to release. Anyone can add locks to + * the list, ocfs2_wq processes the list */ + struct ocfs2_dentry_lock *dentry_lock_list; + struct work_struct dentry_lock_work; + wait_queue_head_t osb_mount_event; /* Truncate log info */ diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index f4efa89baee5..1ed0f7c86869 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -754,7 +754,9 @@ static int ocfs2_mark_dquot_dirty(struct dquot *dquot) if (dquot->dq_flags & mask) sync = 1; spin_unlock(&dq_data_lock); - if (!sync) { + /* This is a slight hack but we can't afford getting global quota + * lock if we already have a transaction started. */ + if (!sync || journal_current_handle()) { status = ocfs2_write_dquot(dquot); goto out; } diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 43ed11345b59..b1cb38fbe807 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1887,6 +1887,9 @@ static int ocfs2_initialize_super(struct super_block *sb, INIT_WORK(&journal->j_recovery_work, ocfs2_complete_recovery); journal->j_state = OCFS2_JOURNAL_FREE; + INIT_WORK(&osb->dentry_lock_work, ocfs2_drop_dl_inodes); + osb->dentry_lock_list = NULL; + /* get some pseudo constants for clustersize bits */ osb->s_clustersize_bits = le32_to_cpu(di->id2.i_super.s_clustersize_bits); diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index e1d638af6ac3..915039fffe6e 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -4729,13 +4729,6 @@ static int ocfs2_xattr_bucket_value_truncate(struct inode *inode, vb.vb_xv = (struct ocfs2_xattr_value_root *) (vb.vb_bh->b_data + offset % blocksize); - ret = ocfs2_xattr_bucket_journal_access(ctxt->handle, bucket, - OCFS2_JOURNAL_ACCESS_WRITE); - if (ret) { - mlog_errno(ret); - goto out; - } - /* * From here on out we have to dirty the bucket. The generic * value calls only modify one of the bucket's bhs, but we need @@ -4748,12 +4741,18 @@ static int ocfs2_xattr_bucket_value_truncate(struct inode *inode, ret = ocfs2_xattr_value_truncate(inode, &vb, len, ctxt); if (ret) { mlog_errno(ret); - goto out_dirty; + goto out; + } + + ret = ocfs2_xattr_bucket_journal_access(ctxt->handle, bucket, + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret) { + mlog_errno(ret); + goto out; } xe->xe_value_size = cpu_to_le64(len); -out_dirty: ocfs2_xattr_bucket_journal_dirty(ctxt->handle, bucket); out: diff --git a/fs/ubifs/budget.c b/fs/ubifs/budget.c index 175f9c590b77..f393620890ee 100644 --- a/fs/ubifs/budget.c +++ b/fs/ubifs/budget.c @@ -689,7 +689,7 @@ long long ubifs_reported_space(const struct ubifs_info *c, long long free) } /** - * ubifs_get_free_space - return amount of free space. + * ubifs_get_free_space_nolock - return amount of free space. * @c: UBIFS file-system description object * * This function calculates amount of free space to report to user-space. @@ -704,16 +704,14 @@ long long ubifs_reported_space(const struct ubifs_info *c, long long free) * traditional file-systems, because they have way less overhead than UBIFS. * So, to keep users happy, UBIFS tries to take the overhead into account. */ -long long ubifs_get_free_space(struct ubifs_info *c) +long long ubifs_get_free_space_nolock(struct ubifs_info *c) { - int min_idx_lebs, rsvd_idx_lebs, lebs; + int rsvd_idx_lebs, lebs; long long available, outstanding, free; - spin_lock(&c->space_lock); - min_idx_lebs = c->min_idx_lebs; - ubifs_assert(min_idx_lebs == ubifs_calc_min_idx_lebs(c)); + ubifs_assert(c->min_idx_lebs == ubifs_calc_min_idx_lebs(c)); outstanding = c->budg_data_growth + c->budg_dd_growth; - available = ubifs_calc_available(c, min_idx_lebs); + available = ubifs_calc_available(c, c->min_idx_lebs); /* * When reporting free space to user-space, UBIFS guarantees that it is @@ -726,15 +724,14 @@ long long ubifs_get_free_space(struct ubifs_info *c) * Note, the calculations below are similar to what we have in * 'do_budget_space()', so refer there for comments. */ - if (min_idx_lebs > c->lst.idx_lebs) - rsvd_idx_lebs = min_idx_lebs - c->lst.idx_lebs; + if (c->min_idx_lebs > c->lst.idx_lebs) + rsvd_idx_lebs = c->min_idx_lebs - c->lst.idx_lebs; else rsvd_idx_lebs = 0; lebs = c->lst.empty_lebs + c->freeable_cnt + c->idx_gc_cnt - c->lst.taken_empty_lebs; lebs -= rsvd_idx_lebs; available += lebs * (c->dark_wm - c->leb_overhead); - spin_unlock(&c->space_lock); if (available > outstanding) free = ubifs_reported_space(c, available - outstanding); @@ -742,3 +739,21 @@ long long ubifs_get_free_space(struct ubifs_info *c) free = 0; return free; } + +/** + * ubifs_get_free_space - return amount of free space. + * @c: UBIFS file-system description object + * + * This function calculates and retuns amount of free space to report to + * user-space. + */ +long long ubifs_get_free_space(struct ubifs_info *c) +{ + long long free; + + spin_lock(&c->space_lock); + free = ubifs_get_free_space_nolock(c); + spin_unlock(&c->space_lock); + + return free; +} diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index 792c5a16c182..e975bd82f38b 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -620,9 +620,11 @@ void dbg_dump_budg(struct ubifs_info *c) c->dark_wm, c->dead_wm, c->max_idx_node_sz); printk(KERN_DEBUG "\tgc_lnum %d, ihead_lnum %d\n", c->gc_lnum, c->ihead_lnum); - for (i = 0; i < c->jhead_cnt; i++) - printk(KERN_DEBUG "\tjhead %d\t LEB %d\n", - c->jheads[i].wbuf.jhead, c->jheads[i].wbuf.lnum); + /* If we are in R/O mode, journal heads do not exist */ + if (c->jheads) + for (i = 0; i < c->jhead_cnt; i++) + printk(KERN_DEBUG "\tjhead %d\t LEB %d\n", + c->jheads[i].wbuf.jhead, c->jheads[i].wbuf.lnum); for (rb = rb_first(&c->buds); rb; rb = rb_next(rb)) { bud = rb_entry(rb, struct ubifs_bud, rb); printk(KERN_DEBUG "\tbud LEB %d\n", bud->lnum); @@ -637,10 +639,7 @@ void dbg_dump_budg(struct ubifs_info *c) /* Print budgeting predictions */ available = ubifs_calc_available(c, c->min_idx_lebs); outstanding = c->budg_data_growth + c->budg_dd_growth; - if (available > outstanding) - free = ubifs_reported_space(c, available - outstanding); - else - free = 0; + free = ubifs_get_free_space_nolock(c); printk(KERN_DEBUG "Budgeting predictions:\n"); printk(KERN_DEBUG "\tavailable: %lld, outstanding %lld, free %lld\n", available, outstanding, free); @@ -861,6 +860,65 @@ void dbg_dump_index(struct ubifs_info *c) } /** + * dbg_save_space_info - save information about flash space. + * @c: UBIFS file-system description object + * + * This function saves information about UBIFS free space, dirty space, etc, in + * order to check it later. + */ +void dbg_save_space_info(struct ubifs_info *c) +{ + struct ubifs_debug_info *d = c->dbg; + + ubifs_get_lp_stats(c, &d->saved_lst); + + spin_lock(&c->space_lock); + d->saved_free = ubifs_get_free_space_nolock(c); + spin_unlock(&c->space_lock); +} + +/** + * dbg_check_space_info - check flash space information. + * @c: UBIFS file-system description object + * + * This function compares current flash space information with the information + * which was saved when the 'dbg_save_space_info()' function was called. + * Returns zero if the information has not changed, and %-EINVAL it it has + * changed. + */ +int dbg_check_space_info(struct ubifs_info *c) +{ + struct ubifs_debug_info *d = c->dbg; + struct ubifs_lp_stats lst; + long long avail, free; + + spin_lock(&c->space_lock); + avail = ubifs_calc_available(c, c->min_idx_lebs); + spin_unlock(&c->space_lock); + free = ubifs_get_free_space(c); + + if (free != d->saved_free) { + ubifs_err("free space changed from %lld to %lld", + d->saved_free, free); + goto out; + } + + return 0; + +out: + ubifs_msg("saved lprops statistics dump"); + dbg_dump_lstats(&d->saved_lst); + ubifs_get_lp_stats(c, &lst); + ubifs_msg("current lprops statistics dump"); + dbg_dump_lstats(&d->saved_lst); + spin_lock(&c->space_lock); + dbg_dump_budg(c); + spin_unlock(&c->space_lock); + dump_stack(); + return -EINVAL; +} + +/** * dbg_check_synced_i_size - check synchronized inode size. * @inode: inode to check * @@ -1349,7 +1407,7 @@ int dbg_check_tnc(struct ubifs_info *c, int extra) * @c: UBIFS file-system description object * @leaf_cb: called for each leaf node * @znode_cb: called for each indexing node - * @priv: private date which is passed to callbacks + * @priv: private data which is passed to callbacks * * This function walks the UBIFS index and calls the @leaf_cb for each leaf * node and @znode_cb for each indexing node. Returns zero in case of success @@ -2409,7 +2467,7 @@ void ubifs_debugging_exit(struct ubifs_info *c) * Root directory for UBIFS stuff in debugfs. Contains sub-directories which * contain the stuff specific to particular file-system mounts. */ -static struct dentry *debugfs_rootdir; +static struct dentry *dfs_rootdir; /** * dbg_debugfs_init - initialize debugfs file-system. @@ -2421,9 +2479,9 @@ static struct dentry *debugfs_rootdir; */ int dbg_debugfs_init(void) { - debugfs_rootdir = debugfs_create_dir("ubifs", NULL); - if (IS_ERR(debugfs_rootdir)) { - int err = PTR_ERR(debugfs_rootdir); + dfs_rootdir = debugfs_create_dir("ubifs", NULL); + if (IS_ERR(dfs_rootdir)) { + int err = PTR_ERR(dfs_rootdir); ubifs_err("cannot create \"ubifs\" debugfs directory, " "error %d\n", err); return err; @@ -2437,7 +2495,7 @@ int dbg_debugfs_init(void) */ void dbg_debugfs_exit(void) { - debugfs_remove(debugfs_rootdir); + debugfs_remove(dfs_rootdir); } static int open_debugfs_file(struct inode *inode, struct file *file) @@ -2452,13 +2510,13 @@ static ssize_t write_debugfs_file(struct file *file, const char __user *buf, struct ubifs_info *c = file->private_data; struct ubifs_debug_info *d = c->dbg; - if (file->f_path.dentry == d->dump_lprops) + if (file->f_path.dentry == d->dfs_dump_lprops) dbg_dump_lprops(c); - else if (file->f_path.dentry == d->dump_budg) { + else if (file->f_path.dentry == d->dfs_dump_budg) { spin_lock(&c->space_lock); dbg_dump_budg(c); spin_unlock(&c->space_lock); - } else if (file->f_path.dentry == d->dump_tnc) { + } else if (file->f_path.dentry == d->dfs_dump_tnc) { mutex_lock(&c->tnc_mutex); dbg_dump_tnc(c); mutex_unlock(&c->tnc_mutex); @@ -2469,7 +2527,7 @@ static ssize_t write_debugfs_file(struct file *file, const char __user *buf, return count; } -static const struct file_operations debugfs_fops = { +static const struct file_operations dfs_fops = { .open = open_debugfs_file, .write = write_debugfs_file, .owner = THIS_MODULE, @@ -2494,36 +2552,32 @@ int dbg_debugfs_init_fs(struct ubifs_info *c) struct dentry *dent; struct ubifs_debug_info *d = c->dbg; - sprintf(d->debugfs_dir_name, "ubi%d_%d", c->vi.ubi_num, c->vi.vol_id); - d->debugfs_dir = debugfs_create_dir(d->debugfs_dir_name, - debugfs_rootdir); - if (IS_ERR(d->debugfs_dir)) { - err = PTR_ERR(d->debugfs_dir); + sprintf(d->dfs_dir_name, "ubi%d_%d", c->vi.ubi_num, c->vi.vol_id); + d->dfs_dir = debugfs_create_dir(d->dfs_dir_name, dfs_rootdir); + if (IS_ERR(d->dfs_dir)) { + err = PTR_ERR(d->dfs_dir); ubifs_err("cannot create \"%s\" debugfs directory, error %d\n", - d->debugfs_dir_name, err); + d->dfs_dir_name, err); goto out; } fname = "dump_lprops"; - dent = debugfs_create_file(fname, S_IWUGO, d->debugfs_dir, c, - &debugfs_fops); + dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops); if (IS_ERR(dent)) goto out_remove; - d->dump_lprops = dent; + d->dfs_dump_lprops = dent; fname = "dump_budg"; - dent = debugfs_create_file(fname, S_IWUGO, d->debugfs_dir, c, - &debugfs_fops); + dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops); if (IS_ERR(dent)) goto out_remove; - d->dump_budg = dent; + d->dfs_dump_budg = dent; fname = "dump_tnc"; - dent = debugfs_create_file(fname, S_IWUGO, d->debugfs_dir, c, - &debugfs_fops); + dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops); if (IS_ERR(dent)) goto out_remove; - d->dump_tnc = dent; + d->dfs_dump_tnc = dent; return 0; @@ -2531,7 +2585,7 @@ out_remove: err = PTR_ERR(dent); ubifs_err("cannot create \"%s\" debugfs directory, error %d\n", fname, err); - debugfs_remove_recursive(d->debugfs_dir); + debugfs_remove_recursive(d->dfs_dir); out: return err; } @@ -2542,7 +2596,7 @@ out: */ void dbg_debugfs_exit_fs(struct ubifs_info *c) { - debugfs_remove_recursive(c->dbg->debugfs_dir); + debugfs_remove_recursive(c->dbg->dfs_dir); } #endif /* CONFIG_UBIFS_FS_DEBUG */ diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h index 9820d6999f7e..c1cd73b2e06e 100644 --- a/fs/ubifs/debug.h +++ b/fs/ubifs/debug.h @@ -41,15 +41,17 @@ * @chk_lpt_wastage: used by LPT tree size checker * @chk_lpt_lebs: used by LPT tree size checker * @new_nhead_offs: used by LPT tree size checker - * @new_ihead_lnum: used by debugging to check ihead_lnum - * @new_ihead_offs: used by debugging to check ihead_offs + * @new_ihead_lnum: used by debugging to check @c->ihead_lnum + * @new_ihead_offs: used by debugging to check @c->ihead_offs * - * debugfs_dir_name: name of debugfs directory containing this file-system's - * files - * debugfs_dir: direntry object of the file-system debugfs directory - * dump_lprops: "dump lprops" debugfs knob - * dump_budg: "dump budgeting information" debugfs knob - * dump_tnc: "dump TNC" debugfs knob + * @saved_lst: saved lprops statistics (used by 'dbg_save_space_info()') + * @saved_free: saved free space (used by 'dbg_save_space_info()') + * + * dfs_dir_name: name of debugfs directory containing this file-system's files + * dfs_dir: direntry object of the file-system debugfs directory + * dfs_dump_lprops: "dump lprops" debugfs knob + * dfs_dump_budg: "dump budgeting information" debugfs knob + * dfs_dump_tnc: "dump TNC" debugfs knob */ struct ubifs_debug_info { void *buf; @@ -69,11 +71,14 @@ struct ubifs_debug_info { int new_ihead_lnum; int new_ihead_offs; - char debugfs_dir_name[100]; - struct dentry *debugfs_dir; - struct dentry *dump_lprops; - struct dentry *dump_budg; - struct dentry *dump_tnc; + struct ubifs_lp_stats saved_lst; + long long saved_free; + + char dfs_dir_name[100]; + struct dentry *dfs_dir; + struct dentry *dfs_dump_lprops; + struct dentry *dfs_dump_budg; + struct dentry *dfs_dump_tnc; }; #define ubifs_assert(expr) do { \ @@ -297,7 +302,8 @@ int dbg_walk_index(struct ubifs_info *c, dbg_leaf_callback leaf_cb, dbg_znode_callback znode_cb, void *priv); /* Checking functions */ - +void dbg_save_space_info(struct ubifs_info *c); +int dbg_check_space_info(struct ubifs_info *c); int dbg_check_lprops(struct ubifs_info *c); int dbg_old_index_check_init(struct ubifs_info *c, struct ubifs_zbranch *zroot); int dbg_check_old_index(struct ubifs_info *c, struct ubifs_zbranch *zroot); @@ -439,6 +445,8 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c); #define dbg_walk_index(c, leaf_cb, znode_cb, priv) 0 #define dbg_old_index_check_init(c, zroot) 0 +#define dbg_save_space_info(c) ({}) +#define dbg_check_space_info(c) 0 #define dbg_check_old_index(c, zroot) 0 #define dbg_check_cats(c) 0 #define dbg_check_ltab(c) 0 diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index f448ab1f9c38..f55d523c52bb 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -482,30 +482,29 @@ static int ubifs_dir_release(struct inode *dir, struct file *file) } /** - * lock_2_inodes - lock two UBIFS inodes. + * lock_2_inodes - a wrapper for locking two UBIFS inodes. * @inode1: first inode * @inode2: second inode + * + * We do not implement any tricks to guarantee strict lock ordering, because + * VFS has already done it for us on the @i_mutex. So this is just a simple + * wrapper function. */ static void lock_2_inodes(struct inode *inode1, struct inode *inode2) { - if (inode1->i_ino < inode2->i_ino) { - mutex_lock_nested(&ubifs_inode(inode1)->ui_mutex, WB_MUTEX_2); - mutex_lock_nested(&ubifs_inode(inode2)->ui_mutex, WB_MUTEX_3); - } else { - mutex_lock_nested(&ubifs_inode(inode2)->ui_mutex, WB_MUTEX_2); - mutex_lock_nested(&ubifs_inode(inode1)->ui_mutex, WB_MUTEX_3); - } + mutex_lock_nested(&ubifs_inode(inode1)->ui_mutex, WB_MUTEX_1); + mutex_lock_nested(&ubifs_inode(inode2)->ui_mutex, WB_MUTEX_2); } /** - * unlock_2_inodes - unlock two UBIFS inodes inodes. + * unlock_2_inodes - a wrapper for unlocking two UBIFS inodes. * @inode1: first inode * @inode2: second inode */ static void unlock_2_inodes(struct inode *inode1, struct inode *inode2) { - mutex_unlock(&ubifs_inode(inode1)->ui_mutex); mutex_unlock(&ubifs_inode(inode2)->ui_mutex); + mutex_unlock(&ubifs_inode(inode1)->ui_mutex); } static int ubifs_link(struct dentry *old_dentry, struct inode *dir, @@ -527,6 +526,8 @@ static int ubifs_link(struct dentry *old_dentry, struct inode *dir, dbg_gen("dent '%.*s' to ino %lu (nlink %d) in dir ino %lu", dentry->d_name.len, dentry->d_name.name, inode->i_ino, inode->i_nlink, dir->i_ino); + ubifs_assert(mutex_is_locked(&dir->i_mutex)); + ubifs_assert(mutex_is_locked(&inode->i_mutex)); err = dbg_check_synced_i_size(inode); if (err) return err; @@ -580,6 +581,8 @@ static int ubifs_unlink(struct inode *dir, struct dentry *dentry) dbg_gen("dent '%.*s' from ino %lu (nlink %d) in dir ino %lu", dentry->d_name.len, dentry->d_name.name, inode->i_ino, inode->i_nlink, dir->i_ino); + ubifs_assert(mutex_is_locked(&dir->i_mutex)); + ubifs_assert(mutex_is_locked(&inode->i_mutex)); err = dbg_check_synced_i_size(inode); if (err) return err; @@ -667,7 +670,8 @@ static int ubifs_rmdir(struct inode *dir, struct dentry *dentry) dbg_gen("directory '%.*s', ino %lu in dir ino %lu", dentry->d_name.len, dentry->d_name.name, inode->i_ino, dir->i_ino); - + ubifs_assert(mutex_is_locked(&dir->i_mutex)); + ubifs_assert(mutex_is_locked(&inode->i_mutex)); err = check_dir_empty(c, dentry->d_inode); if (err) return err; @@ -922,59 +926,30 @@ out_budg: } /** - * lock_3_inodes - lock three UBIFS inodes for rename. + * lock_3_inodes - a wrapper for locking three UBIFS inodes. * @inode1: first inode * @inode2: second inode * @inode3: third inode * - * For 'ubifs_rename()', @inode1 may be the same as @inode2 whereas @inode3 may - * be null. + * This function is used for 'ubifs_rename()' and @inode1 may be the same as + * @inode2 whereas @inode3 may be %NULL. + * + * We do not implement any tricks to guarantee strict lock ordering, because + * VFS has already done it for us on the @i_mutex. So this is just a simple + * wrapper function. */ static void lock_3_inodes(struct inode *inode1, struct inode *inode2, struct inode *inode3) { - struct inode *i1, *i2, *i3; - - if (!inode3) { - if (inode1 != inode2) { - lock_2_inodes(inode1, inode2); - return; - } - mutex_lock_nested(&ubifs_inode(inode1)->ui_mutex, WB_MUTEX_1); - return; - } - - if (inode1 == inode2) { - lock_2_inodes(inode1, inode3); - return; - } - - /* 3 different inodes */ - if (inode1 < inode2) { - i3 = inode2; - if (inode1 < inode3) { - i1 = inode1; - i2 = inode3; - } else { - i1 = inode3; - i2 = inode1; - } - } else { - i3 = inode1; - if (inode2 < inode3) { - i1 = inode2; - i2 = inode3; - } else { - i1 = inode3; - i2 = inode2; - } - } - mutex_lock_nested(&ubifs_inode(i1)->ui_mutex, WB_MUTEX_1); - lock_2_inodes(i2, i3); + mutex_lock_nested(&ubifs_inode(inode1)->ui_mutex, WB_MUTEX_1); + if (inode2 != inode1) + mutex_lock_nested(&ubifs_inode(inode2)->ui_mutex, WB_MUTEX_2); + if (inode3) + mutex_lock_nested(&ubifs_inode(inode3)->ui_mutex, WB_MUTEX_3); } /** - * unlock_3_inodes - unlock three UBIFS inodes for rename. + * unlock_3_inodes - a wrapper for unlocking three UBIFS inodes for rename. * @inode1: first inode * @inode2: second inode * @inode3: third inode @@ -982,11 +957,11 @@ static void lock_3_inodes(struct inode *inode1, struct inode *inode2, static void unlock_3_inodes(struct inode *inode1, struct inode *inode2, struct inode *inode3) { - mutex_unlock(&ubifs_inode(inode1)->ui_mutex); - if (inode1 != inode2) - mutex_unlock(&ubifs_inode(inode2)->ui_mutex); if (inode3) mutex_unlock(&ubifs_inode(inode3)->ui_mutex); + if (inode1 != inode2) + mutex_unlock(&ubifs_inode(inode2)->ui_mutex); + mutex_unlock(&ubifs_inode(inode1)->ui_mutex); } static int ubifs_rename(struct inode *old_dir, struct dentry *old_dentry, @@ -1020,6 +995,11 @@ static int ubifs_rename(struct inode *old_dir, struct dentry *old_dentry, "dir ino %lu", old_dentry->d_name.len, old_dentry->d_name.name, old_inode->i_ino, old_dir->i_ino, new_dentry->d_name.len, new_dentry->d_name.name, new_dir->i_ino); + ubifs_assert(mutex_is_locked(&old_dir->i_mutex)); + ubifs_assert(mutex_is_locked(&new_dir->i_mutex)); + if (unlink) + ubifs_assert(mutex_is_locked(&new_inode->i_mutex)); + if (unlink && is_dir) { err = check_dir_empty(c, new_inode); @@ -1199,7 +1179,7 @@ int ubifs_getattr(struct vfsmount *mnt, struct dentry *dentry, return 0; } -struct inode_operations ubifs_dir_inode_operations = { +const struct inode_operations ubifs_dir_inode_operations = { .lookup = ubifs_lookup, .create = ubifs_create, .link = ubifs_link, @@ -1219,7 +1199,7 @@ struct inode_operations ubifs_dir_inode_operations = { #endif }; -struct file_operations ubifs_dir_operations = { +const struct file_operations ubifs_dir_operations = { .llseek = ubifs_dir_llseek, .release = ubifs_dir_release, .read = generic_read_dir, diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index bf37374567fa..93b6de51f261 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -432,7 +432,6 @@ static int ubifs_write_begin(struct file *file, struct address_space *mapping, int uninitialized_var(err), appending = !!(pos + len > inode->i_size); struct page *page; - ubifs_assert(ubifs_inode(inode)->ui_size == inode->i_size); if (unlikely(c->ro_media)) @@ -1541,7 +1540,7 @@ static int ubifs_file_mmap(struct file *file, struct vm_area_struct *vma) return 0; } -struct address_space_operations ubifs_file_address_operations = { +const struct address_space_operations ubifs_file_address_operations = { .readpage = ubifs_readpage, .writepage = ubifs_writepage, .write_begin = ubifs_write_begin, @@ -1551,7 +1550,7 @@ struct address_space_operations ubifs_file_address_operations = { .releasepage = ubifs_releasepage, }; -struct inode_operations ubifs_file_inode_operations = { +const struct inode_operations ubifs_file_inode_operations = { .setattr = ubifs_setattr, .getattr = ubifs_getattr, #ifdef CONFIG_UBIFS_FS_XATTR @@ -1562,14 +1561,14 @@ struct inode_operations ubifs_file_inode_operations = { #endif }; -struct inode_operations ubifs_symlink_inode_operations = { +const struct inode_operations ubifs_symlink_inode_operations = { .readlink = generic_readlink, .follow_link = ubifs_follow_link, .setattr = ubifs_setattr, .getattr = ubifs_getattr, }; -struct file_operations ubifs_file_operations = { +const struct file_operations ubifs_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, .write = do_sync_write, diff --git a/fs/ubifs/gc.c b/fs/ubifs/gc.c index 9832f9abe28e..a711d33b3d3e 100644 --- a/fs/ubifs/gc.c +++ b/fs/ubifs/gc.c @@ -31,6 +31,26 @@ * to be reused. Garbage collection will cause the number of dirty index nodes * to grow, however sufficient space is reserved for the index to ensure the * commit will never run out of space. + * + * Notes about dead watermark. At current UBIFS implementation we assume that + * LEBs which have less than @c->dead_wm bytes of free + dirty space are full + * and not worth garbage-collecting. The dead watermark is one min. I/O unit + * size, or min. UBIFS node size, depending on what is greater. Indeed, UBIFS + * Garbage Collector has to synchronize the GC head's write buffer before + * returning, so this is about wasting one min. I/O unit. However, UBIFS GC can + * actually reclaim even very small pieces of dirty space by garbage collecting + * enough dirty LEBs, but we do not bother doing this at this implementation. + * + * Notes about dark watermark. The results of GC work depends on how big are + * the UBIFS nodes GC deals with. Large nodes make GC waste more space. Indeed, + * if GC move data from LEB A to LEB B and nodes in LEB A are large, GC would + * have to waste large pieces of free space at the end of LEB B, because nodes + * from LEB A would not fit. And the worst situation is when all nodes are of + * maximum size. So dark watermark is the amount of free + dirty space in LEB + * which are guaranteed to be reclaimable. If LEB has less space, the GC migh + * be unable to reclaim it. So, LEBs with free + dirty greater than dark + * watermark are "good" LEBs from GC's point of few. The other LEBs are not so + * good, and GC takes extra care when moving them. */ #include <linux/pagemap.h> @@ -381,7 +401,7 @@ int ubifs_garbage_collect_leb(struct ubifs_info *c, struct ubifs_lprops *lp) /* * Don't release the LEB until after the next commit, because - * it may contain date which is needed for recovery. So + * it may contain data which is needed for recovery. So * although we freed this LEB, it will become usable only after * the commit. */ @@ -810,8 +830,9 @@ out: * ubifs_destroy_idx_gc - destroy idx_gc list. * @c: UBIFS file-system description object * - * This function destroys the idx_gc list. It is called when unmounting or - * remounting read-only so locks are not needed. + * This function destroys the @c->idx_gc list. It is called when unmounting + * so locks are not needed. Returns zero in case of success and a negative + * error code in case of failure. */ void ubifs_destroy_idx_gc(struct ubifs_info *c) { @@ -824,7 +845,6 @@ void ubifs_destroy_idx_gc(struct ubifs_info *c) list_del(&idx_gc->list); kfree(idx_gc); } - } /** diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 01682713af69..e8e632a1dcdf 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -29,7 +29,7 @@ * would have been wasted for padding to the nearest minimal I/O unit boundary. * Instead, data first goes to the write-buffer and is flushed when the * buffer is full or when it is not used for some time (by timer). This is - * similarto the mechanism is used by JFFS2. + * similar to the mechanism is used by JFFS2. * * Write-buffers are defined by 'struct ubifs_wbuf' objects and protected by * mutexes defined inside these objects. Since sometimes upper-level code @@ -75,7 +75,7 @@ void ubifs_ro_mode(struct ubifs_info *c, int err) * @lnum: logical eraseblock number * @offs: offset within the logical eraseblock * @quiet: print no messages - * @chk_crc: indicates whether to always check the CRC + * @must_chk_crc: indicates whether to always check the CRC * * This function checks node magic number and CRC checksum. This function also * validates node length to prevent UBIFS from becoming crazy when an attacker @@ -83,11 +83,17 @@ void ubifs_ro_mode(struct ubifs_info *c, int err) * node length in the common header could cause UBIFS to read memory outside of * allocated buffer when checking the CRC checksum. * - * This function returns zero in case of success %-EUCLEAN in case of bad CRC - * or magic. + * This function may skip data nodes CRC checking if @c->no_chk_data_crc is + * true, which is controlled by corresponding UBIFS mount option. However, if + * @must_chk_crc is true, then @c->no_chk_data_crc is ignored and CRC is + * checked. Similarly, if @c->always_chk_crc is true, @c->no_chk_data_crc is + * ignored and CRC is checked. + * + * This function returns zero in case of success and %-EUCLEAN in case of bad + * CRC or magic. */ int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, - int offs, int quiet, int chk_crc) + int offs, int quiet, int must_chk_crc) { int err = -EINVAL, type, node_len; uint32_t crc, node_crc, magic; @@ -123,9 +129,9 @@ int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, node_len > c->ranges[type].max_len) goto out_len; - if (!chk_crc && type == UBIFS_DATA_NODE && !c->always_chk_crc) - if (c->no_chk_data_crc) - return 0; + if (!must_chk_crc && type == UBIFS_DATA_NODE && !c->always_chk_crc && + c->no_chk_data_crc) + return 0; crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8); node_crc = le32_to_cpu(ch->crc); diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index 9b7c54e0cd2a..a11ca0958a23 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -208,7 +208,7 @@ again: offs = 0; out: - err = ubifs_wbuf_seek_nolock(wbuf, lnum, offs, UBI_SHORTTERM); + err = ubifs_wbuf_seek_nolock(wbuf, lnum, offs, wbuf->dtype); if (err) goto out_unlock; diff --git a/fs/ubifs/lprops.c b/fs/ubifs/lprops.c index dfd2bcece27a..4cdd284dea56 100644 --- a/fs/ubifs/lprops.c +++ b/fs/ubifs/lprops.c @@ -635,10 +635,10 @@ const struct ubifs_lprops *ubifs_change_lp(struct ubifs_info *c, * @c: UBIFS file-system description object * @st: return statistics */ -void ubifs_get_lp_stats(struct ubifs_info *c, struct ubifs_lp_stats *st) +void ubifs_get_lp_stats(struct ubifs_info *c, struct ubifs_lp_stats *lst) { spin_lock(&c->space_lock); - memcpy(st, &c->lst, sizeof(struct ubifs_lp_stats)); + memcpy(lst, &c->lst, sizeof(struct ubifs_lp_stats)); spin_unlock(&c->space_lock); } @@ -678,6 +678,9 @@ int ubifs_change_one_lp(struct ubifs_info *c, int lnum, int free, int dirty, out: ubifs_release_lprops(c); + if (err) + ubifs_err("cannot change properties of LEB %d, error %d", + lnum, err); return err; } @@ -714,6 +717,9 @@ int ubifs_update_one_lp(struct ubifs_info *c, int lnum, int free, int dirty, out: ubifs_release_lprops(c); + if (err) + ubifs_err("cannot update properties of LEB %d, error %d", + lnum, err); return err; } @@ -737,6 +743,8 @@ int ubifs_read_one_lp(struct ubifs_info *c, int lnum, struct ubifs_lprops *lp) lpp = ubifs_lpt_lookup(c, lnum); if (IS_ERR(lpp)) { err = PTR_ERR(lpp); + ubifs_err("cannot read properties of LEB %d, error %d", + lnum, err); goto out; } diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index 96ca95707175..3216a1f277f8 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -556,23 +556,23 @@ no_space: } /** - * next_pnode - find next pnode. + * next_pnode_to_dirty - find next pnode to dirty. * @c: UBIFS file-system description object * @pnode: pnode * - * This function returns the next pnode or %NULL if there are no more pnodes. + * This function returns the next pnode to dirty or %NULL if there are no more + * pnodes. Note that pnodes that have never been written (lnum == 0) are + * skipped. */ -static struct ubifs_pnode *next_pnode(struct ubifs_info *c, - struct ubifs_pnode *pnode) +static struct ubifs_pnode *next_pnode_to_dirty(struct ubifs_info *c, + struct ubifs_pnode *pnode) { struct ubifs_nnode *nnode; int iip; /* Try to go right */ nnode = pnode->parent; - iip = pnode->iip + 1; - if (iip < UBIFS_LPT_FANOUT) { - /* We assume here that LEB zero is never an LPT LEB */ + for (iip = pnode->iip + 1; iip < UBIFS_LPT_FANOUT; iip++) { if (nnode->nbranch[iip].lnum) return ubifs_get_pnode(c, nnode, iip); } @@ -583,8 +583,11 @@ static struct ubifs_pnode *next_pnode(struct ubifs_info *c, nnode = nnode->parent; if (!nnode) return NULL; - /* We assume here that LEB zero is never an LPT LEB */ - } while (iip >= UBIFS_LPT_FANOUT || !nnode->nbranch[iip].lnum); + for (; iip < UBIFS_LPT_FANOUT; iip++) { + if (nnode->nbranch[iip].lnum) + break; + } + } while (iip >= UBIFS_LPT_FANOUT); /* Go right */ nnode = ubifs_get_nnode(c, nnode, iip); @@ -593,12 +596,29 @@ static struct ubifs_pnode *next_pnode(struct ubifs_info *c, /* Go down to level 1 */ while (nnode->level > 1) { - nnode = ubifs_get_nnode(c, nnode, 0); + for (iip = 0; iip < UBIFS_LPT_FANOUT; iip++) { + if (nnode->nbranch[iip].lnum) + break; + } + if (iip >= UBIFS_LPT_FANOUT) { + /* + * Should not happen, but we need to keep going + * if it does. + */ + iip = 0; + } + nnode = ubifs_get_nnode(c, nnode, iip); if (IS_ERR(nnode)) return (void *)nnode; } - return ubifs_get_pnode(c, nnode, 0); + for (iip = 0; iip < UBIFS_LPT_FANOUT; iip++) + if (nnode->nbranch[iip].lnum) + break; + if (iip >= UBIFS_LPT_FANOUT) + /* Should not happen, but we need to keep going if it does */ + iip = 0; + return ubifs_get_pnode(c, nnode, iip); } /** @@ -688,7 +708,7 @@ static int make_tree_dirty(struct ubifs_info *c) pnode = pnode_lookup(c, 0); while (pnode) { do_make_pnode_dirty(c, pnode); - pnode = next_pnode(c, pnode); + pnode = next_pnode_to_dirty(c, pnode); if (IS_ERR(pnode)) return PTR_ERR(pnode); } diff --git a/fs/ubifs/master.c b/fs/ubifs/master.c index 71d5493bf565..a88f33801b98 100644 --- a/fs/ubifs/master.c +++ b/fs/ubifs/master.c @@ -354,7 +354,7 @@ int ubifs_write_master(struct ubifs_info *c) int err, lnum, offs, len; if (c->ro_media) - return -EINVAL; + return -EROFS; lnum = UBIFS_MST_LNUM; offs = c->mst_offs + c->mst_node_alsz; diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 9e6f403f170e..152a7b34a141 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -46,7 +46,7 @@ * Orphans are accumulated in a rb-tree. When an inode's link count drops to * zero, the inode number is added to the rb-tree. It is removed from the tree * when the inode is deleted. Any new orphans that are in the orphan tree when - * the commit is run, are written to the orphan area in 1 or more orph nodes. + * the commit is run, are written to the orphan area in 1 or more orphan nodes. * If the orphan area is full, it is consolidated to make space. There is * always enough space because validation prevents the user from creating more * than the maximum number of orphans allowed. @@ -231,7 +231,7 @@ static int tot_avail_orphs(struct ubifs_info *c) } /** - * do_write_orph_node - write a node + * do_write_orph_node - write a node to the orphan head. * @c: UBIFS file-system description object * @len: length of node * @atomic: write atomically @@ -264,11 +264,11 @@ static int do_write_orph_node(struct ubifs_info *c, int len, int atomic) } /** - * write_orph_node - write an orph node + * write_orph_node - write an orphan node. * @c: UBIFS file-system description object * @atomic: write atomically * - * This function builds an orph node from the cnext list and writes it to the + * This function builds an orphan node from the cnext list and writes it to the * orphan head. On success, %0 is returned, otherwise a negative error code * is returned. */ @@ -326,11 +326,11 @@ static int write_orph_node(struct ubifs_info *c, int atomic) } /** - * write_orph_nodes - write orph nodes until there are no more to commit + * write_orph_nodes - write orphan nodes until there are no more to commit. * @c: UBIFS file-system description object * @atomic: write atomically * - * This function writes orph nodes for all the orphans to commit. On success, + * This function writes orphan nodes for all the orphans to commit. On success, * %0 is returned, otherwise a negative error code is returned. */ static int write_orph_nodes(struct ubifs_info *c, int atomic) @@ -478,14 +478,14 @@ int ubifs_orphan_end_commit(struct ubifs_info *c) } /** - * clear_orphans - erase all LEBs used for orphans. + * ubifs_clear_orphans - erase all LEBs used for orphans. * @c: UBIFS file-system description object * * If recovery is not required, then the orphans from the previous session * are not needed. This function locates the LEBs used to record * orphans, and un-maps them. */ -static int clear_orphans(struct ubifs_info *c) +int ubifs_clear_orphans(struct ubifs_info *c) { int lnum, err; @@ -547,9 +547,9 @@ static int insert_dead_orphan(struct ubifs_info *c, ino_t inum) * do_kill_orphans - remove orphan inodes from the index. * @c: UBIFS file-system description object * @sleb: scanned LEB - * @last_cmt_no: cmt_no of last orph node read is passed and returned here + * @last_cmt_no: cmt_no of last orphan node read is passed and returned here * @outofdate: whether the LEB is out of date is returned here - * @last_flagged: whether the end orph node is encountered + * @last_flagged: whether the end orphan node is encountered * * This function is a helper to the 'kill_orphans()' function. It goes through * every orphan node in a LEB and for every inode number recorded, removes @@ -580,8 +580,8 @@ static int do_kill_orphans(struct ubifs_info *c, struct ubifs_scan_leb *sleb, /* * The commit number on the master node may be less, because * of a failed commit. If there are several failed commits in a - * row, the commit number written on orph nodes will continue to - * increase (because the commit number is adjusted here) even + * row, the commit number written on orphan nodes will continue + * to increase (because the commit number is adjusted here) even * though the commit number on the master node stays the same * because the master node has not been re-written. */ @@ -589,9 +589,9 @@ static int do_kill_orphans(struct ubifs_info *c, struct ubifs_scan_leb *sleb, c->cmt_no = cmt_no; if (cmt_no < *last_cmt_no && *last_flagged) { /* - * The last orph node had a higher commit number and was - * flagged as the last written for that commit number. - * That makes this orph node, out of date. + * The last orphan node had a higher commit number and + * was flagged as the last written for that commit + * number. That makes this orphan node, out of date. */ if (!first) { ubifs_err("out of order commit number %llu in " @@ -658,10 +658,10 @@ static int kill_orphans(struct ubifs_info *c) /* * Orph nodes always start at c->orph_first and are written to each * successive LEB in turn. Generally unused LEBs will have been unmapped - * but may contain out of date orph nodes if the unmap didn't go - * through. In addition, the last orph node written for each commit is + * but may contain out of date orphan nodes if the unmap didn't go + * through. In addition, the last orphan node written for each commit is * marked (top bit of orph->cmt_no is set to 1). It is possible that - * there are orph nodes from the next commit (i.e. the commit did not + * there are orphan nodes from the next commit (i.e. the commit did not * complete successfully). In that case, no orphans will have been lost * due to the way that orphans are written, and any orphans added will * be valid orphans anyway and so can be deleted. @@ -718,7 +718,7 @@ int ubifs_mount_orphans(struct ubifs_info *c, int unclean, int read_only) if (unclean) err = kill_orphans(c); else if (!read_only) - err = clear_orphans(c); + err = ubifs_clear_orphans(c); return err; } diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 89556ee72518..1182b66a5491 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -397,6 +397,7 @@ static int ubifs_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_namelen = UBIFS_MAX_NLEN; buf->f_fsid.val[0] = le32_to_cpu(uuid[0]) ^ le32_to_cpu(uuid[2]); buf->f_fsid.val[1] = le32_to_cpu(uuid[1]) ^ le32_to_cpu(uuid[3]); + ubifs_assert(buf->f_bfree <= c->block_cnt); return 0; } @@ -432,33 +433,24 @@ static int ubifs_sync_fs(struct super_block *sb, int wait) int i, err; struct ubifs_info *c = sb->s_fs_info; struct writeback_control wbc = { - .sync_mode = wait ? WB_SYNC_ALL : WB_SYNC_NONE, + .sync_mode = WB_SYNC_ALL, .range_start = 0, .range_end = LLONG_MAX, .nr_to_write = LONG_MAX, }; /* - * Note by akpm about WB_SYNC_NONE used above: zero @wait is just an - * advisory thing to help the file system shove lots of data into the - * queues. If some gets missed then it'll be picked up on the second + * Zero @wait is just an advisory thing to help the file system shove + * lots of data into the queues, and there will be the second * '->sync_fs()' call, with non-zero @wait. */ + if (!wait) + return 0; if (sb->s_flags & MS_RDONLY) return 0; /* - * Synchronize write buffers, because 'ubifs_run_commit()' does not - * do this if it waits for an already running commit. - */ - for (i = 0; i < c->jhead_cnt; i++) { - err = ubifs_wbuf_sync(&c->jheads[i].wbuf); - if (err) - return err; - } - - /* * VFS calls '->sync_fs()' before synchronizing all dirty inodes and * pages, so synchronize them first, then commit the journal. Strictly * speaking, it is not necessary to commit the journal here, @@ -469,6 +461,16 @@ static int ubifs_sync_fs(struct super_block *sb, int wait) */ generic_sync_sb_inodes(sb, &wbc); + /* + * Synchronize write buffers, because 'ubifs_run_commit()' does not + * do this if it waits for an already running commit. + */ + for (i = 0; i < c->jhead_cnt; i++) { + err = ubifs_wbuf_sync(&c->jheads[i].wbuf); + if (err) + return err; + } + err = ubifs_run_commit(c); if (err) return err; @@ -572,15 +574,8 @@ static int init_constants_early(struct ubifs_info *c) c->ranges[UBIFS_IDX_NODE].max_len = INT_MAX; /* - * Initialize dead and dark LEB space watermarks. - * - * Dead space is the space which cannot be used. Its watermark is - * equivalent to min. I/O unit or minimum node size if it is greater - * then min. I/O unit. - * - * Dark space is the space which might be used, or might not, depending - * on which node should be written to the LEB. Its watermark is - * equivalent to maximum UBIFS node size. + * Initialize dead and dark LEB space watermarks. See gc.c for comments + * about these values. */ c->dead_wm = ALIGN(MIN_WRITE_SZ, c->min_io_size); c->dark_wm = ALIGN(UBIFS_MAX_NODE_SZ, c->min_io_size); @@ -741,12 +736,12 @@ static void init_constants_master(struct ubifs_info *c) * take_gc_lnum - reserve GC LEB. * @c: UBIFS file-system description object * - * This function ensures that the LEB reserved for garbage collection is - * unmapped and is marked as "taken" in lprops. We also have to set free space - * to LEB size and dirty space to zero, because lprops may contain out-of-date - * information if the file-system was un-mounted before it has been committed. - * This function returns zero in case of success and a negative error code in - * case of failure. + * This function ensures that the LEB reserved for garbage collection is marked + * as "taken" in lprops. We also have to set free space to LEB size and dirty + * space to zero, because lprops may contain out-of-date information if the + * file-system was un-mounted before it has been committed. This function + * returns zero in case of success and a negative error code in case of + * failure. */ static int take_gc_lnum(struct ubifs_info *c) { @@ -757,10 +752,6 @@ static int take_gc_lnum(struct ubifs_info *c) return -EINVAL; } - err = ubifs_leb_unmap(c, c->gc_lnum); - if (err) - return err; - /* And we have to tell lprops that this LEB is taken */ err = ubifs_change_one_lp(c, c->gc_lnum, c->leb_size, 0, LPROPS_TAKEN, 0, 0); @@ -966,13 +957,16 @@ static int ubifs_parse_options(struct ubifs_info *c, char *options, token = match_token(p, tokens, args); switch (token) { + /* + * %Opt_fast_unmount and %Opt_norm_unmount options are ignored. + * We accepte them in order to be backware-compatible. But this + * should be removed at some point. + */ case Opt_fast_unmount: c->mount_opts.unmount_mode = 2; - c->fast_unmount = 1; break; case Opt_norm_unmount: c->mount_opts.unmount_mode = 1; - c->fast_unmount = 0; break; case Opt_bulk_read: c->mount_opts.bulk_read = 2; @@ -1094,12 +1088,7 @@ static int check_free_space(struct ubifs_info *c) ubifs_err("insufficient free space to mount in read/write mode"); dbg_dump_budg(c); dbg_dump_lprops(c); - /* - * We return %-EINVAL instead of %-ENOSPC because it seems to - * be the closest error code mentioned in the mount function - * documentation. - */ - return -EINVAL; + return -ENOSPC; } return 0; } @@ -1286,10 +1275,19 @@ static int mount_ubifs(struct ubifs_info *c) if (err) goto out_orphans; err = ubifs_rcvry_gc_commit(c); - } else + } else { err = take_gc_lnum(c); - if (err) - goto out_orphans; + if (err) + goto out_orphans; + + /* + * GC LEB may contain garbage if there was an unclean + * reboot, and it should be un-mapped. + */ + err = ubifs_leb_unmap(c, c->gc_lnum); + if (err) + return err; + } err = dbg_check_lprops(c); if (err) @@ -1298,6 +1296,16 @@ static int mount_ubifs(struct ubifs_info *c) err = ubifs_recover_size(c); if (err) goto out_orphans; + } else { + /* + * Even if we mount read-only, we have to set space in GC LEB + * to proper value because this affects UBIFS free space + * reporting. We do not want to have a situation when + * re-mounting from R/O to R/W changes amount of free space. + */ + err = take_gc_lnum(c); + if (err) + goto out_orphans; } spin_lock(&ubifs_infos_lock); @@ -1310,14 +1318,17 @@ static int mount_ubifs(struct ubifs_info *c) else { c->need_recovery = 0; ubifs_msg("recovery completed"); + /* GC LEB has to be empty and taken at this point */ + ubifs_assert(c->lst.taken_empty_lebs == 1); } - } + } else + ubifs_assert(c->lst.taken_empty_lebs == 1); - err = dbg_debugfs_init_fs(c); + err = dbg_check_filesystem(c); if (err) goto out_infos; - err = dbg_check_filesystem(c); + err = dbg_debugfs_init_fs(c); if (err) goto out_infos; @@ -1351,7 +1362,6 @@ static int mount_ubifs(struct ubifs_info *c) c->uuid[4], c->uuid[5], c->uuid[6], c->uuid[7], c->uuid[8], c->uuid[9], c->uuid[10], c->uuid[11], c->uuid[12], c->uuid[13], c->uuid[14], c->uuid[15]); - dbg_msg("fast unmount: %d", c->fast_unmount); dbg_msg("big_lpt %d", c->big_lpt); dbg_msg("log LEBs: %d (%d - %d)", c->log_lebs, UBIFS_LOG_LNUM, c->log_last); @@ -1475,10 +1485,8 @@ static int ubifs_remount_rw(struct ubifs_info *c) { int err, lnum; - if (c->ro_media) - return -EINVAL; - mutex_lock(&c->umount_mutex); + dbg_save_space_info(c); c->remounting_rw = 1; c->always_chk_crc = 1; @@ -1514,6 +1522,12 @@ static int ubifs_remount_rw(struct ubifs_info *c) err = ubifs_recover_inl_heads(c, c->sbuf); if (err) goto out; + } else { + /* A readonly mount is not allowed to have orphans */ + ubifs_assert(c->tot_orphans == 0); + err = ubifs_clear_orphans(c); + if (err) + goto out; } if (!(c->mst_node->flags & cpu_to_le32(UBIFS_MST_DIRTY))) { @@ -1569,7 +1583,7 @@ static int ubifs_remount_rw(struct ubifs_info *c) if (c->need_recovery) err = ubifs_rcvry_gc_commit(c); else - err = take_gc_lnum(c); + err = ubifs_leb_unmap(c, c->gc_lnum); if (err) goto out; @@ -1582,8 +1596,9 @@ static int ubifs_remount_rw(struct ubifs_info *c) c->vfs_sb->s_flags &= ~MS_RDONLY; c->remounting_rw = 0; c->always_chk_crc = 0; + err = dbg_check_space_info(c); mutex_unlock(&c->umount_mutex); - return 0; + return err; out: vfree(c->orph_buf); @@ -1603,43 +1618,18 @@ out: } /** - * commit_on_unmount - commit the journal when un-mounting. - * @c: UBIFS file-system description object - * - * This function is called during un-mounting and re-mounting, and it commits - * the journal unless the "fast unmount" mode is enabled. - */ -static void commit_on_unmount(struct ubifs_info *c) -{ - struct super_block *sb = c->vfs_sb; - long long bud_bytes; - - /* - * This function is called before the background thread is stopped, so - * we may race with ongoing commit, which means we have to take - * @c->bud_lock to access @c->bud_bytes. - */ - spin_lock(&c->buds_lock); - bud_bytes = c->bud_bytes; - spin_unlock(&c->buds_lock); - - if (!c->fast_unmount && !(sb->s_flags & MS_RDONLY) && bud_bytes) - ubifs_run_commit(c); -} - -/** * ubifs_remount_ro - re-mount in read-only mode. * @c: UBIFS file-system description object * - * We rely on VFS to have stopped writing. Possibly the background thread could - * be running a commit, however kthread_stop will wait in that case. + * We assume VFS has stopped writing. Possibly the background thread could be + * running a commit, however kthread_stop will wait in that case. */ static void ubifs_remount_ro(struct ubifs_info *c) { int i, err; ubifs_assert(!c->need_recovery); - commit_on_unmount(c); + ubifs_assert(!(c->vfs_sb->s_flags & MS_RDONLY)); mutex_lock(&c->umount_mutex); if (c->bgt) { @@ -1647,27 +1637,29 @@ static void ubifs_remount_ro(struct ubifs_info *c) c->bgt = NULL; } + dbg_save_space_info(c); + for (i = 0; i < c->jhead_cnt; i++) { ubifs_wbuf_sync(&c->jheads[i].wbuf); del_timer_sync(&c->jheads[i].wbuf.timer); } - if (!c->ro_media) { - c->mst_node->flags &= ~cpu_to_le32(UBIFS_MST_DIRTY); - c->mst_node->flags |= cpu_to_le32(UBIFS_MST_NO_ORPHS); - c->mst_node->gc_lnum = cpu_to_le32(c->gc_lnum); - err = ubifs_write_master(c); - if (err) - ubifs_ro_mode(c, err); - } + c->mst_node->flags &= ~cpu_to_le32(UBIFS_MST_DIRTY); + c->mst_node->flags |= cpu_to_le32(UBIFS_MST_NO_ORPHS); + c->mst_node->gc_lnum = cpu_to_le32(c->gc_lnum); + err = ubifs_write_master(c); + if (err) + ubifs_ro_mode(c, err); - ubifs_destroy_idx_gc(c); free_wbufs(c); vfree(c->orph_buf); c->orph_buf = NULL; vfree(c->ileb_buf); c->ileb_buf = NULL; ubifs_lpt_free(c, 1); + err = dbg_check_space_info(c); + if (err) + ubifs_ro_mode(c, err); mutex_unlock(&c->umount_mutex); } @@ -1760,11 +1752,20 @@ static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data) } if ((sb->s_flags & MS_RDONLY) && !(*flags & MS_RDONLY)) { + if (c->ro_media) { + ubifs_msg("cannot re-mount due to prior errors"); + return -EROFS; + } err = ubifs_remount_rw(c); if (err) return err; - } else if (!(sb->s_flags & MS_RDONLY) && (*flags & MS_RDONLY)) + } else if (!(sb->s_flags & MS_RDONLY) && (*flags & MS_RDONLY)) { + if (c->ro_media) { + ubifs_msg("cannot re-mount due to prior errors"); + return -EROFS; + } ubifs_remount_ro(c); + } if (c->bulk_read == 1) bu_init(c); @@ -1774,10 +1775,11 @@ static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data) c->bu.buf = NULL; } + ubifs_assert(c->lst.taken_empty_lebs == 1); return 0; } -struct super_operations ubifs_super_operations = { +const struct super_operations ubifs_super_operations = { .alloc_inode = ubifs_alloc_inode, .destroy_inode = ubifs_destroy_inode, .put_super = ubifs_put_super, @@ -2044,15 +2046,6 @@ out_close: static void ubifs_kill_sb(struct super_block *sb) { - struct ubifs_info *c = sb->s_fs_info; - - /* - * We do 'commit_on_unmount()' here instead of 'ubifs_put_super()' - * in order to be outside BKL. - */ - if (sb->s_root) - commit_on_unmount(c); - /* The un-mount routine is actually done in put_super() */ generic_shutdown_super(sb); } diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c index f7e36f545527..fa28a84c6a1b 100644 --- a/fs/ubifs/tnc.c +++ b/fs/ubifs/tnc.c @@ -443,6 +443,11 @@ static int tnc_read_node_nm(struct ubifs_info *c, struct ubifs_zbranch *zbr, * This function performs that same function as ubifs_read_node except that * it does not require that there is actually a node present and instead * the return code indicates if a node was read. + * + * Note, this function does not check CRC of data nodes if @c->no_chk_data_crc + * is true (it is controlled by corresponding mount option). However, if + * @c->always_chk_crc is true, @c->no_chk_data_crc is ignored and CRC is always + * checked. */ static int try_read_node(const struct ubifs_info *c, void *buf, int type, int len, int lnum, int offs) @@ -470,9 +475,8 @@ static int try_read_node(const struct ubifs_info *c, void *buf, int type, if (node_len != len) return 0; - if (type == UBIFS_DATA_NODE && !c->always_chk_crc) - if (c->no_chk_data_crc) - return 0; + if (type == UBIFS_DATA_NODE && !c->always_chk_crc && c->no_chk_data_crc) + return 1; crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8); node_crc = le32_to_cpu(ch->crc); @@ -1506,7 +1510,7 @@ out: * * Note, if the bulk-read buffer length (@bu->buf_len) is known, this function * makes sure bulk-read nodes fit the buffer. Otherwise, this function prepares - * maxumum possible amount of nodes for bulk-read. + * maximum possible amount of nodes for bulk-read. */ int ubifs_tnc_get_bu_keys(struct ubifs_info *c, struct bu_info *bu) { diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index fc2a4cc66d03..039a68bee29a 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -426,9 +426,9 @@ struct ubifs_unclean_leb { * LEB properties flags. * * LPROPS_UNCAT: not categorized - * LPROPS_DIRTY: dirty > 0, not index + * LPROPS_DIRTY: dirty > free, dirty >= @c->dead_wm, not index * LPROPS_DIRTY_IDX: dirty + free > @c->min_idx_node_sze and index - * LPROPS_FREE: free > 0, not empty, not index + * LPROPS_FREE: free > 0, dirty < @c->dead_wm, not empty, not index * LPROPS_HEAP_CNT: number of heaps used for storing categorized LEBs * LPROPS_EMPTY: LEB is empty, not taken * LPROPS_FREEABLE: free + dirty == leb_size, not index, not taken @@ -961,7 +961,6 @@ struct ubifs_debug_info; * @cs_lock: commit state lock * @cmt_wq: wait queue to sleep on if the log is full and a commit is running * - * @fast_unmount: do not run journal commit before un-mounting * @big_lpt: flag that LPT is too big to write whole during commit * @no_chk_data_crc: do not check CRCs when reading data nodes (except during * recovery) @@ -1202,7 +1201,6 @@ struct ubifs_info { spinlock_t cs_lock; wait_queue_head_t cmt_wq; - unsigned int fast_unmount:1; unsigned int big_lpt:1; unsigned int no_chk_data_crc:1; unsigned int bulk_read:1; @@ -1405,13 +1403,13 @@ extern struct list_head ubifs_infos; extern spinlock_t ubifs_infos_lock; extern atomic_long_t ubifs_clean_zn_cnt; extern struct kmem_cache *ubifs_inode_slab; -extern struct super_operations ubifs_super_operations; -extern struct address_space_operations ubifs_file_address_operations; -extern struct file_operations ubifs_file_operations; -extern struct inode_operations ubifs_file_inode_operations; -extern struct file_operations ubifs_dir_operations; -extern struct inode_operations ubifs_dir_inode_operations; -extern struct inode_operations ubifs_symlink_inode_operations; +extern const struct super_operations ubifs_super_operations; +extern const struct address_space_operations ubifs_file_address_operations; +extern const struct file_operations ubifs_file_operations; +extern const struct inode_operations ubifs_file_inode_operations; +extern const struct file_operations ubifs_dir_operations; +extern const struct inode_operations ubifs_dir_inode_operations; +extern const struct inode_operations ubifs_symlink_inode_operations; extern struct backing_dev_info ubifs_backing_dev_info; extern struct ubifs_compressor *ubifs_compressors[UBIFS_COMPR_TYPES_CNT]; @@ -1428,7 +1426,7 @@ int ubifs_read_node_wbuf(struct ubifs_wbuf *wbuf, void *buf, int type, int len, int ubifs_write_node(struct ubifs_info *c, void *node, int len, int lnum, int offs, int dtype); int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, - int offs, int quiet, int chk_crc); + int offs, int quiet, int must_chk_crc); void ubifs_prepare_node(struct ubifs_info *c, void *buf, int len, int pad); void ubifs_prep_grp_node(struct ubifs_info *c, void *node, int len, int last); int ubifs_io_init(struct ubifs_info *c); @@ -1495,6 +1493,7 @@ void ubifs_release_ino_dirty(struct ubifs_info *c, struct inode *inode, void ubifs_cancel_ino_op(struct ubifs_info *c, struct inode *inode, struct ubifs_budget_req *req); long long ubifs_get_free_space(struct ubifs_info *c); +long long ubifs_get_free_space_nolock(struct ubifs_info *c); int ubifs_calc_min_idx_lebs(struct ubifs_info *c); void ubifs_convert_page_budget(struct ubifs_info *c); long long ubifs_reported_space(const struct ubifs_info *c, long long free); @@ -1603,6 +1602,7 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum); int ubifs_orphan_start_commit(struct ubifs_info *c); int ubifs_orphan_end_commit(struct ubifs_info *c); int ubifs_mount_orphans(struct ubifs_info *c, int unclean, int read_only); +int ubifs_clear_orphans(struct ubifs_info *c); /* lpt.c */ int ubifs_calc_lpt_geom(struct ubifs_info *c); @@ -1646,7 +1646,7 @@ const struct ubifs_lprops *ubifs_change_lp(struct ubifs_info *c, const struct ubifs_lprops *lp, int free, int dirty, int flags, int idx_gc_cnt); -void ubifs_get_lp_stats(struct ubifs_info *c, struct ubifs_lp_stats *stats); +void ubifs_get_lp_stats(struct ubifs_info *c, struct ubifs_lp_stats *lst); void ubifs_add_to_cat(struct ubifs_info *c, struct ubifs_lprops *lprops, int cat); void ubifs_replace_cat(struct ubifs_info *c, struct ubifs_lprops *old_lprops, diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 2ed035354c26..a608e72fa405 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -371,7 +371,11 @@ xfs_quiesce_attr( /* flush inodes and push all remaining buffers out to disk */ xfs_quiesce_fs(mp); - ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0); + /* + * Just warn here till VFS can correctly support + * read-only remount without racing. + */ + WARN_ON(atomic_read(&mp->m_active_trans) != 0); /* Push the superblock and write an unmount record */ error = xfs_log_sbcount(mp, 1); diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index b4c1ee713492..f8278cfcc1d3 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -55,17 +55,11 @@ xfs_swapext( struct file *file, *target_file; int error = 0; - sxp = kmem_alloc(sizeof(xfs_swapext_t), KM_MAYFAIL); - if (!sxp) { - error = XFS_ERROR(ENOMEM); - goto out; - } - /* Pull information for the target fd */ file = fget((int)sxp->sx_fdtarget); if (!file) { error = XFS_ERROR(EINVAL); - goto out_free_sxp; + goto out; } if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND)) { @@ -109,8 +103,6 @@ xfs_swapext( fput(target_file); out_put_file: fput(file); - out_free_sxp: - kmem_free(sxp); out: return error; } diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 35cca98bd94c..b1047de2fffd 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -70,16 +70,21 @@ STATIC void xlog_recover_check_summary(xlog_t *); xfs_buf_t * xlog_get_bp( xlog_t *log, - int num_bblks) + int nbblks) { - ASSERT(num_bblks > 0); + if (nbblks <= 0 || nbblks > log->l_logBBsize) { + xlog_warn("XFS: Invalid block length (0x%x) given for buffer", nbblks); + XFS_ERROR_REPORT("xlog_get_bp(1)", + XFS_ERRLEVEL_HIGH, log->l_mp); + return NULL; + } if (log->l_sectbb_log) { - if (num_bblks > 1) - num_bblks += XLOG_SECTOR_ROUNDUP_BBCOUNT(log, 1); - num_bblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, num_bblks); + if (nbblks > 1) + nbblks += XLOG_SECTOR_ROUNDUP_BBCOUNT(log, 1); + nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks); } - return xfs_buf_get_noaddr(BBTOB(num_bblks), log->l_mp->m_logdev_targp); + return xfs_buf_get_noaddr(BBTOB(nbblks), log->l_mp->m_logdev_targp); } void @@ -102,6 +107,13 @@ xlog_bread( { int error; + if (nbblks <= 0 || nbblks > log->l_logBBsize) { + xlog_warn("XFS: Invalid block length (0x%x) given for buffer", nbblks); + XFS_ERROR_REPORT("xlog_bread(1)", + XFS_ERRLEVEL_HIGH, log->l_mp); + return EFSCORRUPTED; + } + if (log->l_sectbb_log) { blk_no = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, blk_no); nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks); @@ -139,6 +151,13 @@ xlog_bwrite( { int error; + if (nbblks <= 0 || nbblks > log->l_logBBsize) { + xlog_warn("XFS: Invalid block length (0x%x) given for buffer", nbblks); + XFS_ERROR_REPORT("xlog_bwrite(1)", + XFS_ERRLEVEL_HIGH, log->l_mp); + return EFSCORRUPTED; + } + if (log->l_sectbb_log) { blk_no = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, blk_no); nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks); diff --git a/include/linux/ata.h b/include/linux/ata.h index a53318b8cbd0..08a86d5cdf1b 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -731,12 +731,17 @@ static inline int ata_id_current_chs_valid(const u16 *id) static inline int ata_id_is_cfa(const u16 *id) { - if (id[ATA_ID_CONFIG] == 0x848A) /* Standard CF */ + if (id[ATA_ID_CONFIG] == 0x848A) /* Traditional CF */ return 1; - /* Could be CF hiding as standard ATA */ - if (ata_id_major_version(id) >= 3 && - id[ATA_ID_COMMAND_SET_1] != 0xFFFF && - (id[ATA_ID_COMMAND_SET_1] & (1 << 2))) + /* + * CF specs don't require specific value in the word 0 anymore and yet + * they forbid to report the ATA version in the word 80 and require the + * CFA feature set support to be indicated in the word 83 in this case. + * Unfortunately, some cards only follow either of this requirements, + * and while those that don't indicate CFA feature support need some + * sort of quirk list, it seems impractical for the ones that do... + */ + if ((id[ATA_ID_COMMAND_SET_2] & 0xC004) == 0x4004) return 1; return 0; } diff --git a/include/linux/libata.h b/include/linux/libata.h index bca3ba25f52a..5d87bc09a1f5 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -380,6 +380,7 @@ enum { ATA_HORKAGE_ATAPI_MOD16_DMA = (1 << 11), /* use ATAPI DMA for commands not multiple of 16 bytes */ ATA_HORKAGE_FIRMWARE_WARN = (1 << 12), /* firwmare update warning */ + ATA_HORKAGE_1_5_GBPS = (1 << 13), /* force 1.5 Gbps */ /* DMA mask for user DMA control: User visible values; DO NOT renumber */ @@ -580,7 +581,7 @@ struct ata_device { acpi_handle acpi_handle; union acpi_object *gtf_cache; #endif - /* n_sector is used as CLEAR_OFFSET, read comment above CLEAR_OFFSET */ + /* n_sector is CLEAR_BEGIN, read comment above CLEAR_BEGIN */ u64 n_sectors; /* size of device, if ATA */ unsigned int class; /* ATA_DEV_xxx */ unsigned long unpark_deadline; @@ -605,20 +606,22 @@ struct ata_device { u16 heads; /* Number of heads */ u16 sectors; /* Number of sectors per track */ - /* error history */ - int spdn_cnt; - struct ata_ering ering; - union { u16 id[ATA_ID_WORDS]; /* IDENTIFY xxx DEVICE data */ u32 gscr[SATA_PMP_GSCR_DWORDS]; /* PMP GSCR block */ }; + + /* error history */ + int spdn_cnt; + /* ering is CLEAR_END, read comment above CLEAR_END */ + struct ata_ering ering; }; -/* Offset into struct ata_device. Fields above it are maintained - * acress device init. Fields below are zeroed. +/* Fields between ATA_DEVICE_CLEAR_BEGIN and ATA_DEVICE_CLEAR_END are + * cleared to zero on ata_dev_init(). */ -#define ATA_DEVICE_CLEAR_OFFSET offsetof(struct ata_device, n_sectors) +#define ATA_DEVICE_CLEAR_BEGIN offsetof(struct ata_device, n_sectors) +#define ATA_DEVICE_CLEAR_END offsetof(struct ata_device, ering) struct ata_eh_info { struct ata_device *dev; /* offending device */ diff --git a/mm/page-writeback.c b/mm/page-writeback.c index b493db7841dc..dc32dae01e5f 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -1051,13 +1051,22 @@ continue_unlock: } } - if (wbc->sync_mode == WB_SYNC_NONE) { - wbc->nr_to_write--; - if (wbc->nr_to_write <= 0) { - done = 1; - break; - } + if (nr_to_write > 0) + nr_to_write--; + else if (wbc->sync_mode == WB_SYNC_NONE) { + /* + * We stop writing back only if we are not + * doing integrity sync. In case of integrity + * sync we have to keep going because someone + * may be concurrently dirtying pages, and we + * might have synced a lot of newly appeared + * dirty pages, but have not synced all of the + * old dirty pages. + */ + done = 1; + break; } + if (wbc->nonblocking && bdi_write_congested(bdi)) { wbc->encountered_congestion = 1; done = 1; |
