diff --git a/crypto/async_tx/Kconfig b/crypto/async_tx/Kconfig index e5aeb2b79e6f..e28e276ac611 100644 --- a/crypto/async_tx/Kconfig +++ b/crypto/async_tx/Kconfig @@ -23,3 +23,8 @@ config ASYNC_RAID6_RECOV select ASYNC_CORE select ASYNC_PQ +config ASYNC_TX_DISABLE_PQ_VAL_DMA + bool + +config ASYNC_TX_DISABLE_XOR_VAL_DMA + bool diff --git a/crypto/async_tx/async_pq.c b/crypto/async_tx/async_pq.c index 6b5cc4fba59f..ec87f53d5059 100644 --- a/crypto/async_tx/async_pq.c +++ b/crypto/async_tx/async_pq.c @@ -240,6 +240,16 @@ async_gen_syndrome(struct page **blocks, unsigned int offset, int disks, } EXPORT_SYMBOL_GPL(async_gen_syndrome); +static inline struct dma_chan * +pq_val_chan(struct async_submit_ctl *submit, struct page **blocks, int disks, size_t len) +{ + #ifdef CONFIG_ASYNC_TX_DISABLE_PQ_VAL_DMA + return NULL; + #endif + return async_tx_find_channel(submit, DMA_PQ_VAL, NULL, 0, blocks, + disks, len); +} + /** * async_syndrome_val - asynchronously validate a raid6 syndrome * @blocks: source blocks from idx 0..disks-3, P @ disks-2 and Q @ disks-1 @@ -260,9 +270,7 @@ async_syndrome_val(struct page **blocks, unsigned int offset, int disks, size_t len, enum sum_check_flags *pqres, struct page *spare, struct async_submit_ctl *submit) { - struct dma_chan *chan = async_tx_find_channel(submit, DMA_PQ_VAL, - NULL, 0, blocks, disks, - len); + struct dma_chan *chan = pq_val_chan(submit, blocks, disks, len); struct dma_device *device = chan ? chan->device : NULL; struct dma_async_tx_descriptor *tx; unsigned char coefs[disks-2]; diff --git a/crypto/async_tx/async_xor.c b/crypto/async_tx/async_xor.c index 79182dcb91b7..079ae8ca590b 100644 --- a/crypto/async_tx/async_xor.c +++ b/crypto/async_tx/async_xor.c @@ -234,6 +234,17 @@ static int page_is_zero(struct page *p, unsigned int offset, size_t len) memcmp(a, a + 4, len - 4) == 0); } +static inline struct dma_chan * +xor_val_chan(struct async_submit_ctl *submit, struct page *dest, + struct page **src_list, int src_cnt, size_t len) +{ + #ifdef CONFIG_ASYNC_TX_DISABLE_XOR_VAL_DMA + return NULL; + #endif + return async_tx_find_channel(submit, DMA_XOR_VAL, &dest, 1, src_list, + src_cnt, len); +} + /** * async_xor_val - attempt a xor parity check with a dma engine. * @dest: destination page used if the xor is performed synchronously @@ -255,9 +266,7 @@ async_xor_val(struct page *dest, struct page **src_list, unsigned int offset, int src_cnt, size_t len, enum sum_check_flags *result, struct async_submit_ctl *submit) { - struct dma_chan *chan = async_tx_find_channel(submit, DMA_XOR_VAL, - &dest, 1, src_list, - src_cnt, len); + struct dma_chan *chan = xor_val_chan(submit, dest, src_list, src_cnt, len); struct dma_device *device = chan ? chan->device : NULL; struct dma_async_tx_descriptor *tx = NULL; dma_addr_t *dma_src = NULL; diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig index 5903a88351bf..b401dadad4a8 100644 --- a/drivers/dma/Kconfig +++ b/drivers/dma/Kconfig @@ -26,6 +26,8 @@ config INTEL_IOATDMA select DMA_ENGINE select DCA select ASYNC_TX_DISABLE_CHANNEL_SWITCH + select ASYNC_TX_DISABLE_PQ_VAL_DMA + select ASYNC_TX_DISABLE_XOR_VAL_DMA help Enable support for the Intel(R) I/OAT DMA engine present in recent Intel Xeon chipsets. diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index bd0b248de2cf..8f99354082ce 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -632,11 +632,21 @@ static bool device_has_all_tx_types(struct dma_device *device) #if defined(CONFIG_ASYNC_XOR) || defined(CONFIG_ASYNC_XOR_MODULE) if (!dma_has_cap(DMA_XOR, device->cap_mask)) return false; + + #ifndef CONFIG_ASYNC_TX_DISABLE_XOR_VAL_DMA + if (!dma_has_cap(DMA_XOR_VAL, device->cap_mask)) + return false; + #endif #endif #if defined(CONFIG_ASYNC_PQ) || defined(CONFIG_ASYNC_PQ_MODULE) if (!dma_has_cap(DMA_PQ, device->cap_mask)) return false; + + #ifndef CONFIG_ASYNC_TX_DISABLE_PQ_VAL_DMA + if (!dma_has_cap(DMA_PQ_VAL, device->cap_mask)) + return false; + #endif #endif return true; diff --git a/drivers/dma/ioat/dca.c b/drivers/dma/ioat/dca.c index 69d02615c4d6..abd9038e06b1 100644 --- a/drivers/dma/ioat/dca.c +++ b/drivers/dma/ioat/dca.c @@ -98,17 +98,17 @@ static int dca_enabled_in_bios(struct pci_dev *pdev) cpuid_level_9 = cpuid_eax(9); res = test_bit(0, &cpuid_level_9); if (!res) - dev_err(&pdev->dev, "DCA is disabled in BIOS\n"); + dev_dbg(&pdev->dev, "DCA is disabled in BIOS\n"); return res; } -static int system_has_dca_enabled(struct pci_dev *pdev) +int system_has_dca_enabled(struct pci_dev *pdev) { if (boot_cpu_has(X86_FEATURE_DCA)) return dca_enabled_in_bios(pdev); - dev_err(&pdev->dev, "boot cpu doesn't have X86_FEATURE_DCA\n"); + dev_dbg(&pdev->dev, "boot cpu doesn't have X86_FEATURE_DCA\n"); return 0; } diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h index c14fdfeb7f33..45edde996480 100644 --- a/drivers/dma/ioat/dma.h +++ b/drivers/dma/ioat/dma.h @@ -297,9 +297,7 @@ static inline bool is_ioat_suspended(unsigned long status) /* channel was fatally programmed */ static inline bool is_ioat_bug(unsigned long err) { - return !!(err & (IOAT_CHANERR_SRC_ADDR_ERR|IOAT_CHANERR_DEST_ADDR_ERR| - IOAT_CHANERR_NEXT_ADDR_ERR|IOAT_CHANERR_CONTROL_ERR| - IOAT_CHANERR_LENGTH_ERR)); + return !!err; } static inline void ioat_unmap(struct pci_dev *pdev, dma_addr_t addr, size_t len, diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c index 96ffab7d37a7..8f1f7f05deaa 100644 --- a/drivers/dma/ioat/dma_v2.c +++ b/drivers/dma/ioat/dma_v2.c @@ -279,6 +279,8 @@ void ioat2_timer_event(unsigned long data) u32 chanerr; chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET); + dev_err(to_dev(chan), "%s: Channel halted (%x)\n", + __func__, chanerr); BUG_ON(is_ioat_bug(chanerr)); } diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c index 35d1e33afd5b..42f6f10fb0cc 100644 --- a/drivers/dma/ioat/dma_v3.c +++ b/drivers/dma/ioat/dma_v3.c @@ -378,6 +378,8 @@ static void ioat3_timer_event(unsigned long data) u32 chanerr; chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET); + dev_err(to_dev(chan), "%s: Channel halted (%x)\n", + __func__, chanerr); BUG_ON(is_ioat_bug(chanerr)); } @@ -569,7 +571,7 @@ __ioat3_prep_xor_lock(struct dma_chan *c, enum sum_check_flags *result, dump_desc_dbg(ioat, compl_desc); /* we leave the channel locked to ensure in order submission */ - return &desc->txd; + return &compl_desc->txd; } static struct dma_async_tx_descriptor * @@ -728,7 +730,7 @@ __ioat3_prep_pq_lock(struct dma_chan *c, enum sum_check_flags *result, dump_desc_dbg(ioat, compl_desc); /* we leave the channel locked to ensure in order submission */ - return &desc->txd; + return &compl_desc->txd; } static struct dma_async_tx_descriptor * @@ -736,10 +738,16 @@ ioat3_prep_pq(struct dma_chan *chan, dma_addr_t *dst, dma_addr_t *src, unsigned int src_cnt, const unsigned char *scf, size_t len, unsigned long flags) { + /* specify valid address for disabled result */ + if (flags & DMA_PREP_PQ_DISABLE_P) + dst[0] = dst[1]; + if (flags & DMA_PREP_PQ_DISABLE_Q) + dst[1] = dst[0]; + /* handle the single source multiply case from the raid6 * recovery path */ - if (unlikely((flags & DMA_PREP_PQ_DISABLE_P) && src_cnt == 1)) { + if ((flags & DMA_PREP_PQ_DISABLE_P) && src_cnt == 1) { dma_addr_t single_source[2]; unsigned char single_source_coef[2]; @@ -761,6 +769,12 @@ ioat3_prep_pq_val(struct dma_chan *chan, dma_addr_t *pq, dma_addr_t *src, unsigned int src_cnt, const unsigned char *scf, size_t len, enum sum_check_flags *pqres, unsigned long flags) { + /* specify valid address for disabled result */ + if (flags & DMA_PREP_PQ_DISABLE_P) + pq[0] = pq[1]; + if (flags & DMA_PREP_PQ_DISABLE_Q) + pq[1] = pq[0]; + /* the cleanup routine only sets bits on validate failure, it * does not clear bits on validate success... so clear it here */ @@ -778,9 +792,9 @@ ioat3_prep_pqxor(struct dma_chan *chan, dma_addr_t dst, dma_addr_t *src, dma_addr_t pq[2]; memset(scf, 0, src_cnt); - flags |= DMA_PREP_PQ_DISABLE_Q; pq[0] = dst; - pq[1] = ~0; + flags |= DMA_PREP_PQ_DISABLE_Q; + pq[1] = dst; /* specify valid address for disabled result */ return __ioat3_prep_pq_lock(chan, NULL, pq, src, src_cnt, scf, len, flags); @@ -800,9 +814,9 @@ ioat3_prep_pqxor_val(struct dma_chan *chan, dma_addr_t *src, *result = 0; memset(scf, 0, src_cnt); - flags |= DMA_PREP_PQ_DISABLE_Q; pq[0] = src[0]; - pq[1] = ~0; + flags |= DMA_PREP_PQ_DISABLE_Q; + pq[1] = pq[0]; /* specify valid address for disabled result */ return __ioat3_prep_pq_lock(chan, result, pq, &src[1], src_cnt - 1, scf, len, flags); @@ -1117,6 +1131,7 @@ static int __devinit ioat3_dma_self_test(struct ioatdma_device *device) int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca) { struct pci_dev *pdev = device->pdev; + int dca_en = system_has_dca_enabled(pdev); struct dma_device *dma; struct dma_chan *c; struct ioat_chan_common *chan; @@ -1137,6 +1152,11 @@ int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca) dma->device_prep_dma_interrupt = ioat3_prep_interrupt_lock; cap = readl(device->reg_base + IOAT_DMA_CAP_OFFSET); + + /* dca is incompatible with raid operations */ + if (dca_en && (cap & (IOAT_CAP_XOR|IOAT_CAP_PQ))) + cap &= ~(IOAT_CAP_XOR|IOAT_CAP_PQ); + if (cap & IOAT_CAP_XOR) { is_raid_device = true; dma->max_xor = 8; @@ -1186,6 +1206,16 @@ int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca) device->timer_fn = ioat2_timer_event; } + #ifdef CONFIG_ASYNC_TX_DISABLE_PQ_VAL_DMA + dma_cap_clear(DMA_PQ_VAL, dma->cap_mask); + dma->device_prep_dma_pq_val = NULL; + #endif + + #ifdef CONFIG_ASYNC_TX_DISABLE_XOR_VAL_DMA + dma_cap_clear(DMA_XOR_VAL, dma->cap_mask); + dma->device_prep_dma_xor_val = NULL; + #endif + /* -= IOAT ver.3 workarounds =- */ /* Write CHANERRMSK_INT with 3E07h to mask out the errors * that can cause stability issues for IOAT ver.3 diff --git a/drivers/dma/ioat/hw.h b/drivers/dma/ioat/hw.h index 99afb12bd409..60e675455b6a 100644 --- a/drivers/dma/ioat/hw.h +++ b/drivers/dma/ioat/hw.h @@ -39,6 +39,8 @@ #define IOAT_VER_3_0 0x30 /* Version 3.0 */ #define IOAT_VER_3_2 0x32 /* Version 3.2 */ +int system_has_dca_enabled(struct pci_dev *pdev); + struct ioat_dma_descriptor { uint32_t size; union { diff --git a/drivers/dma/ioat/registers.h b/drivers/dma/ioat/registers.h index 63038e18ab03..f015ec196700 100644 --- a/drivers/dma/ioat/registers.h +++ b/drivers/dma/ioat/registers.h @@ -92,9 +92,7 @@ #define IOAT_CHANCTRL_ERR_COMPLETION_EN 0x0004 #define IOAT_CHANCTRL_INT_REARM 0x0001 #define IOAT_CHANCTRL_RUN (IOAT_CHANCTRL_INT_REARM |\ - IOAT_CHANCTRL_ERR_COMPLETION_EN |\ - IOAT_CHANCTRL_ANY_ERR_ABORT_EN |\ - IOAT_CHANCTRL_ERR_INT_EN) + IOAT_CHANCTRL_ANY_ERR_ABORT_EN) #define IOAT_DMA_COMP_OFFSET 0x02 /* 16-bit DMA channel compatibility */ #define IOAT_DMA_COMP_V1 0x0001 /* Compatibility with DMA version 1 */ diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c index b3b065c4e5c1..034ecf0ace03 100644 --- a/drivers/dma/shdma.c +++ b/drivers/dma/shdma.c @@ -640,17 +640,16 @@ static int __init sh_dmae_probe(struct platform_device *pdev) #endif struct sh_dmae_device *shdev; + /* get platform data */ + if (!pdev->dev.platform_data) + return -ENODEV; + shdev = kzalloc(sizeof(struct sh_dmae_device), GFP_KERNEL); if (!shdev) { dev_err(&pdev->dev, "No enough memory\n"); - err = -ENOMEM; - goto shdev_err; + return -ENOMEM; } - /* get platform data */ - if (!pdev->dev.platform_data) - goto shdev_err; - /* platform data */ memcpy(&shdev->pdata, pdev->dev.platform_data, sizeof(struct sh_dmae_pdata)); @@ -722,7 +721,6 @@ eirq_err: rst_err: kfree(shdev); -shdev_err: return err; }