From bdd43cb39f1b1e897c5d7992d05d0b9f0dd786d1 Mon Sep 17 00:00:00 2001 From: Michal Nazarewicz Date: Wed, 5 Sep 2012 07:50:41 +0200 Subject: drivers: dma-contiguous: refactor dma_alloc_from_contiguous() The dma_alloc_from_contiguous() function returns either a valid pointer to a page structure or NULL, the error code set when pageno >= cma->count is not used at all and can be safely removed. This commit also changes the function to avoid goto and have only one exit path and one place where mutex is unlocked. Signed-off-by: Michal Nazarewicz [fixed compilation break caused by missing semicolon] Signed-off-by: Marek Szyprowski --- drivers/base/dma-contiguous.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-) diff --git a/drivers/base/dma-contiguous.c b/drivers/base/dma-contiguous.c index 34d94c762a1..9a1469474f5 100644 --- a/drivers/base/dma-contiguous.c +++ b/drivers/base/dma-contiguous.c @@ -315,6 +315,7 @@ struct page *dma_alloc_from_contiguous(struct device *dev, int count, { unsigned long mask, pfn, pageno, start = 0; struct cma *cma = dev_get_cma_area(dev); + struct page *page = NULL; int ret; if (!cma || !cma->count) @@ -336,18 +337,17 @@ struct page *dma_alloc_from_contiguous(struct device *dev, int count, for (;;) { pageno = bitmap_find_next_zero_area(cma->bitmap, cma->count, start, count, mask); - if (pageno >= cma->count) { - ret = -ENOMEM; - goto error; - } + if (pageno >= cma->count) + break; pfn = cma->base_pfn + pageno; ret = alloc_contig_range(pfn, pfn + count, MIGRATE_CMA); if (ret == 0) { bitmap_set(cma->bitmap, pageno, count); + page = pfn_to_page(pfn); break; } else if (ret != -EBUSY) { - goto error; + break; } pr_debug("%s(): memory range at %p is busy, retrying\n", __func__, pfn_to_page(pfn)); @@ -356,12 +356,8 @@ struct page *dma_alloc_from_contiguous(struct device *dev, int count, } mutex_unlock(&cma_mutex); - - pr_debug("%s(): returned %p\n", __func__, pfn_to_page(pfn)); - return pfn_to_page(pfn); -error: - mutex_unlock(&cma_mutex); - return NULL; + pr_debug("%s(): returned %p\n", __func__, page); + return page; } /** -- cgit v1.2.3-18-g5258 From 5a796eeb7bbacc628044b12a2d000167ddbabec8 Mon Sep 17 00:00:00 2001 From: Hiroshi Doyu Date: Tue, 11 Sep 2012 07:39:39 +0200 Subject: ARM: dma-mapping: Small logical clean up Skip unnecessary operations if order == 0. A little bit easier to read. Signed-off-by: Hiroshi Doyu Signed-off-by: Marek Szyprowski --- arch/arm/mm/dma-mapping.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 13f555d6249..c390cea6a25 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1012,11 +1012,12 @@ static struct page **__iommu_alloc_buffer(struct device *dev, size_t size, gfp_t if (!pages[i]) goto error; - if (order) + if (order) { split_page(pages[i], order); - j = 1 << order; - while (--j) - pages[i + j] = pages[i] + j; + j = 1 << order; + while (--j) + pages[i + j] = pages[i] + j; + } __dma_clear_buffer(pages[i], PAGE_SIZE << order); i += 1 << order; -- cgit v1.2.3-18-g5258 From 75c5971614932ca53009cbbdfd6c8f96eab9e46f Mon Sep 17 00:00:00 2001 From: Hiroshi Doyu Date: Tue, 11 Sep 2012 07:39:48 +0200 Subject: ARM: dma-mapping: Refrain noisy console message With many IOMMU'able devices, console gets noisy. Tegra30 has a few dozen of IOMMU'able devices. Signed-off-by: Hiroshi Doyu Signed-off-by: Marek Szyprowski --- arch/arm/mm/dma-mapping.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index c390cea6a25..3819f029a40 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1666,7 +1666,7 @@ int arm_iommu_attach_device(struct device *dev, dev->archdata.mapping = mapping; set_dma_ops(dev, &iommu_ops); - pr_info("Attached IOMMU controller to %s device.\n", dev_name(dev)); + pr_debug("Attached IOMMU controller to %s device.\n", dev_name(dev)); return 0; } -- cgit v1.2.3-18-g5258 From dd37e9405a8e85be49a60b2530efeb5f06bcb753 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Tue, 21 Aug 2012 12:20:17 +0200 Subject: ARM: add coherent dma ops arch_is_coherent is problematic as it is a global symbol. This doesn't work for multi-platform kernels or platforms which can support per device coherent DMA. This adds arm_coherent_dma_ops to be used for devices which connected coherently (i.e. to the ACP port on Cortex-A9 or A15). The arm_dma_ops are modified at boot when arch_is_coherent is true. Signed-off-by: Rob Herring Cc: Russell King Cc: Marek Szyprowski Signed-off-by: Marek Szyprowski --- arch/arm/include/asm/dma-mapping.h | 1 + arch/arm/mm/dma-mapping.c | 71 +++++++++++++++++++++++++++++++------- 2 files changed, 60 insertions(+), 12 deletions(-) diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h index 5c44dcb0987..23004847bb0 100644 --- a/arch/arm/include/asm/dma-mapping.h +++ b/arch/arm/include/asm/dma-mapping.h @@ -13,6 +13,7 @@ #define DMA_ERROR_CODE (~0) extern struct dma_map_ops arm_dma_ops; +extern struct dma_map_ops arm_coherent_dma_ops; static inline struct dma_map_ops *get_dma_ops(struct device *dev) { diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 13f555d6249..7d772c0a93f 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -73,11 +73,18 @@ static dma_addr_t arm_dma_map_page(struct device *dev, struct page *page, unsigned long offset, size_t size, enum dma_data_direction dir, struct dma_attrs *attrs) { - if (!arch_is_coherent() && !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) + if (!dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) __dma_page_cpu_to_dev(page, offset, size, dir); return pfn_to_dma(dev, page_to_pfn(page)) + offset; } +static dma_addr_t arm_coherent_dma_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + return pfn_to_dma(dev, page_to_pfn(page)) + offset; +} + /** * arm_dma_unmap_page - unmap a buffer previously mapped through dma_map_page() * @dev: valid struct device pointer, or NULL for ISA and EISA-like devices @@ -96,7 +103,7 @@ static void arm_dma_unmap_page(struct device *dev, dma_addr_t handle, size_t size, enum dma_data_direction dir, struct dma_attrs *attrs) { - if (!arch_is_coherent() && !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) + if (!dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) __dma_page_dev_to_cpu(pfn_to_page(dma_to_pfn(dev, handle)), handle & ~PAGE_MASK, size, dir); } @@ -106,8 +113,7 @@ static void arm_dma_sync_single_for_cpu(struct device *dev, { unsigned int offset = handle & (PAGE_SIZE - 1); struct page *page = pfn_to_page(dma_to_pfn(dev, handle-offset)); - if (!arch_is_coherent()) - __dma_page_dev_to_cpu(page, offset, size, dir); + __dma_page_dev_to_cpu(page, offset, size, dir); } static void arm_dma_sync_single_for_device(struct device *dev, @@ -115,8 +121,7 @@ static void arm_dma_sync_single_for_device(struct device *dev, { unsigned int offset = handle & (PAGE_SIZE - 1); struct page *page = pfn_to_page(dma_to_pfn(dev, handle-offset)); - if (!arch_is_coherent()) - __dma_page_cpu_to_dev(page, offset, size, dir); + __dma_page_cpu_to_dev(page, offset, size, dir); } static int arm_dma_set_mask(struct device *dev, u64 dma_mask); @@ -138,6 +143,22 @@ struct dma_map_ops arm_dma_ops = { }; EXPORT_SYMBOL(arm_dma_ops); +static void *arm_coherent_dma_alloc(struct device *dev, size_t size, + dma_addr_t *handle, gfp_t gfp, struct dma_attrs *attrs); +static void arm_coherent_dma_free(struct device *dev, size_t size, void *cpu_addr, + dma_addr_t handle, struct dma_attrs *attrs); + +struct dma_map_ops arm_coherent_dma_ops = { + .alloc = arm_coherent_dma_alloc, + .free = arm_coherent_dma_free, + .mmap = arm_dma_mmap, + .get_sgtable = arm_dma_get_sgtable, + .map_page = arm_coherent_dma_map_page, + .map_sg = arm_dma_map_sg, + .set_dma_mask = arm_dma_set_mask, +}; +EXPORT_SYMBOL(arm_coherent_dma_ops); + static u64 get_coherent_dma_mask(struct device *dev) { u64 mask = (u64)arm_dma_limit; @@ -586,7 +607,7 @@ static void *__alloc_simple_buffer(struct device *dev, size_t size, gfp_t gfp, static void *__dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, - gfp_t gfp, pgprot_t prot, const void *caller) + gfp_t gfp, pgprot_t prot, bool is_coherent, const void *caller) { u64 mask = get_coherent_dma_mask(dev); struct page *page; @@ -619,7 +640,7 @@ static void *__dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, *handle = DMA_ERROR_CODE; size = PAGE_ALIGN(size); - if (arch_is_coherent() || nommu()) + if (is_coherent || nommu()) addr = __alloc_simple_buffer(dev, size, gfp, &page); else if (gfp & GFP_ATOMIC) addr = __alloc_from_pool(size, &page); @@ -647,7 +668,20 @@ void *arm_dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, if (dma_alloc_from_coherent(dev, size, handle, &memory)) return memory; - return __dma_alloc(dev, size, handle, gfp, prot, + return __dma_alloc(dev, size, handle, gfp, prot, false, + __builtin_return_address(0)); +} + +static void *arm_coherent_dma_alloc(struct device *dev, size_t size, + dma_addr_t *handle, gfp_t gfp, struct dma_attrs *attrs) +{ + pgprot_t prot = __get_dma_pgprot(attrs, pgprot_kernel); + void *memory; + + if (dma_alloc_from_coherent(dev, size, handle, &memory)) + return memory; + + return __dma_alloc(dev, size, handle, gfp, prot, true, __builtin_return_address(0)); } @@ -684,8 +718,9 @@ int arm_dma_mmap(struct device *dev, struct vm_area_struct *vma, /* * Free a buffer as defined by the above mapping. */ -void arm_dma_free(struct device *dev, size_t size, void *cpu_addr, - dma_addr_t handle, struct dma_attrs *attrs) +static void __arm_dma_free(struct device *dev, size_t size, void *cpu_addr, + dma_addr_t handle, struct dma_attrs *attrs, + bool is_coherent) { struct page *page = pfn_to_page(dma_to_pfn(dev, handle)); @@ -694,7 +729,7 @@ void arm_dma_free(struct device *dev, size_t size, void *cpu_addr, size = PAGE_ALIGN(size); - if (arch_is_coherent() || nommu()) { + if (is_coherent || nommu()) { __dma_free_buffer(page, size); } else if (__free_from_pool(cpu_addr, size)) { return; @@ -710,6 +745,18 @@ void arm_dma_free(struct device *dev, size_t size, void *cpu_addr, } } +void arm_dma_free(struct device *dev, size_t size, void *cpu_addr, + dma_addr_t handle, struct dma_attrs *attrs) +{ + __arm_dma_free(dev, size, cpu_addr, handle, attrs, false); +} + +static void arm_coherent_dma_free(struct device *dev, size_t size, void *cpu_addr, + dma_addr_t handle, struct dma_attrs *attrs) +{ + __arm_dma_free(dev, size, cpu_addr, handle, attrs, true); +} + int arm_dma_get_sgtable(struct device *dev, struct sg_table *sgt, void *cpu_addr, dma_addr_t handle, size_t size, struct dma_attrs *attrs) -- cgit v1.2.3-18-g5258 From 0fa478df444f5837336d7e2fd0b41643c8d704c2 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Tue, 21 Aug 2012 12:23:23 +0200 Subject: ARM: add coherent iommu dma ops Remove arch_is_coherent() from iommu dma ops and implement separate coherent ops functions. Signed-off-by: Rob Herring Cc: Russell King Cc: Marek Szyprowski Signed-off-by: Marek Szyprowski --- arch/arm/mm/dma-mapping.c | 183 ++++++++++++++++++++++++++++++++++++---------- 1 file changed, 143 insertions(+), 40 deletions(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 7d772c0a93f..5d5a4c4a5db 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1350,7 +1350,8 @@ static int arm_iommu_get_sgtable(struct device *dev, struct sg_table *sgt, */ static int __map_sg_chunk(struct device *dev, struct scatterlist *sg, size_t size, dma_addr_t *handle, - enum dma_data_direction dir, struct dma_attrs *attrs) + enum dma_data_direction dir, struct dma_attrs *attrs, + bool is_coherent) { struct dma_iommu_mapping *mapping = dev->archdata.mapping; dma_addr_t iova, iova_base; @@ -1369,8 +1370,8 @@ static int __map_sg_chunk(struct device *dev, struct scatterlist *sg, phys_addr_t phys = page_to_phys(sg_page(s)); unsigned int len = PAGE_ALIGN(s->offset + s->length); - if (!arch_is_coherent() && - !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) + if (!is_coherent && + !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) __dma_page_cpu_to_dev(sg_page(s), s->offset, s->length, dir); ret = iommu_map(mapping->domain, iova, phys, len, 0); @@ -1388,20 +1389,9 @@ fail: return ret; } -/** - * arm_iommu_map_sg - map a set of SG buffers for streaming mode DMA - * @dev: valid struct device pointer - * @sg: list of buffers - * @nents: number of buffers to map - * @dir: DMA transfer direction - * - * Map a set of buffers described by scatterlist in streaming mode for DMA. - * The scatter gather list elements are merged together (if possible) and - * tagged with the appropriate dma address and length. They are obtained via - * sg_dma_{address,length}. - */ -int arm_iommu_map_sg(struct device *dev, struct scatterlist *sg, int nents, - enum dma_data_direction dir, struct dma_attrs *attrs) +static int __iommu_map_sg(struct device *dev, struct scatterlist *sg, int nents, + enum dma_data_direction dir, struct dma_attrs *attrs, + bool is_coherent) { struct scatterlist *s = sg, *dma = sg, *start = sg; int i, count = 0; @@ -1417,7 +1407,7 @@ int arm_iommu_map_sg(struct device *dev, struct scatterlist *sg, int nents, if (s->offset || (size & ~PAGE_MASK) || size + s->length > max) { if (__map_sg_chunk(dev, start, size, &dma->dma_address, - dir, attrs) < 0) + dir, attrs, is_coherent) < 0) goto bad_mapping; dma->dma_address += offset; @@ -1430,7 +1420,8 @@ int arm_iommu_map_sg(struct device *dev, struct scatterlist *sg, int nents, } size += s->length; } - if (__map_sg_chunk(dev, start, size, &dma->dma_address, dir, attrs) < 0) + if (__map_sg_chunk(dev, start, size, &dma->dma_address, dir, attrs, + is_coherent) < 0) goto bad_mapping; dma->dma_address += offset; @@ -1445,17 +1436,44 @@ bad_mapping: } /** - * arm_iommu_unmap_sg - unmap a set of SG buffers mapped by dma_map_sg + * arm_coherent_iommu_map_sg - map a set of SG buffers for streaming mode DMA * @dev: valid struct device pointer * @sg: list of buffers - * @nents: number of buffers to unmap (same as was passed to dma_map_sg) - * @dir: DMA transfer direction (same as was passed to dma_map_sg) + * @nents: number of buffers to map + * @dir: DMA transfer direction * - * Unmap a set of streaming mode DMA translations. Again, CPU access - * rules concerning calls here are the same as for dma_unmap_single(). + * Map a set of i/o coherent buffers described by scatterlist in streaming + * mode for DMA. The scatter gather list elements are merged together (if + * possible) and tagged with the appropriate dma address and length. They are + * obtained via sg_dma_{address,length}. */ -void arm_iommu_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, - enum dma_data_direction dir, struct dma_attrs *attrs) +int arm_coherent_iommu_map_sg(struct device *dev, struct scatterlist *sg, + int nents, enum dma_data_direction dir, struct dma_attrs *attrs) +{ + return __iommu_map_sg(dev, sg, nents, dir, attrs, true); +} + +/** + * arm_iommu_map_sg - map a set of SG buffers for streaming mode DMA + * @dev: valid struct device pointer + * @sg: list of buffers + * @nents: number of buffers to map + * @dir: DMA transfer direction + * + * Map a set of buffers described by scatterlist in streaming mode for DMA. + * The scatter gather list elements are merged together (if possible) and + * tagged with the appropriate dma address and length. They are obtained via + * sg_dma_{address,length}. + */ +int arm_iommu_map_sg(struct device *dev, struct scatterlist *sg, + int nents, enum dma_data_direction dir, struct dma_attrs *attrs) +{ + return __iommu_map_sg(dev, sg, nents, dir, attrs, false); +} + +static void __iommu_unmap_sg(struct device *dev, struct scatterlist *sg, + int nents, enum dma_data_direction dir, struct dma_attrs *attrs, + bool is_coherent) { struct scatterlist *s; int i; @@ -1464,13 +1482,45 @@ void arm_iommu_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, if (sg_dma_len(s)) __iommu_remove_mapping(dev, sg_dma_address(s), sg_dma_len(s)); - if (!arch_is_coherent() && + if (!is_coherent && !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) __dma_page_dev_to_cpu(sg_page(s), s->offset, s->length, dir); } } +/** + * arm_coherent_iommu_unmap_sg - unmap a set of SG buffers mapped by dma_map_sg + * @dev: valid struct device pointer + * @sg: list of buffers + * @nents: number of buffers to unmap (same as was passed to dma_map_sg) + * @dir: DMA transfer direction (same as was passed to dma_map_sg) + * + * Unmap a set of streaming mode DMA translations. Again, CPU access + * rules concerning calls here are the same as for dma_unmap_single(). + */ +void arm_coherent_iommu_unmap_sg(struct device *dev, struct scatterlist *sg, + int nents, enum dma_data_direction dir, struct dma_attrs *attrs) +{ + __iommu_unmap_sg(dev, sg, nents, dir, attrs, true); +} + +/** + * arm_iommu_unmap_sg - unmap a set of SG buffers mapped by dma_map_sg + * @dev: valid struct device pointer + * @sg: list of buffers + * @nents: number of buffers to unmap (same as was passed to dma_map_sg) + * @dir: DMA transfer direction (same as was passed to dma_map_sg) + * + * Unmap a set of streaming mode DMA translations. Again, CPU access + * rules concerning calls here are the same as for dma_unmap_single(). + */ +void arm_iommu_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, + enum dma_data_direction dir, struct dma_attrs *attrs) +{ + __iommu_unmap_sg(dev, sg, nents, dir, attrs, false); +} + /** * arm_iommu_sync_sg_for_cpu * @dev: valid struct device pointer @@ -1485,8 +1535,7 @@ void arm_iommu_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int i; for_each_sg(sg, s, nents, i) - if (!arch_is_coherent()) - __dma_page_dev_to_cpu(sg_page(s), s->offset, s->length, dir); + __dma_page_dev_to_cpu(sg_page(s), s->offset, s->length, dir); } @@ -1504,22 +1553,21 @@ void arm_iommu_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int i; for_each_sg(sg, s, nents, i) - if (!arch_is_coherent()) - __dma_page_cpu_to_dev(sg_page(s), s->offset, s->length, dir); + __dma_page_cpu_to_dev(sg_page(s), s->offset, s->length, dir); } /** - * arm_iommu_map_page + * arm_coherent_iommu_map_page * @dev: valid struct device pointer * @page: page that buffer resides in * @offset: offset into page for start of buffer * @size: size of buffer to map * @dir: DMA transfer direction * - * IOMMU aware version of arm_dma_map_page() + * Coherent IOMMU aware version of arm_dma_map_page() */ -static dma_addr_t arm_iommu_map_page(struct device *dev, struct page *page, +static dma_addr_t arm_coherent_iommu_map_page(struct device *dev, struct page *page, unsigned long offset, size_t size, enum dma_data_direction dir, struct dma_attrs *attrs) { @@ -1527,9 +1575,6 @@ static dma_addr_t arm_iommu_map_page(struct device *dev, struct page *page, dma_addr_t dma_addr; int ret, len = PAGE_ALIGN(size + offset); - if (!arch_is_coherent() && !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) - __dma_page_cpu_to_dev(page, offset, size, dir); - dma_addr = __alloc_iova(mapping, len); if (dma_addr == DMA_ERROR_CODE) return dma_addr; @@ -1544,6 +1589,52 @@ fail: return DMA_ERROR_CODE; } +/** + * arm_iommu_map_page + * @dev: valid struct device pointer + * @page: page that buffer resides in + * @offset: offset into page for start of buffer + * @size: size of buffer to map + * @dir: DMA transfer direction + * + * IOMMU aware version of arm_dma_map_page() + */ +static dma_addr_t arm_iommu_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + if (!dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) + __dma_page_cpu_to_dev(page, offset, size, dir); + + return arm_coherent_iommu_map_page(dev, page, offset, size, dir, attrs); +} + +/** + * arm_coherent_iommu_unmap_page + * @dev: valid struct device pointer + * @handle: DMA address of buffer + * @size: size of buffer (same as passed to dma_map_page) + * @dir: DMA transfer direction (same as passed to dma_map_page) + * + * Coherent IOMMU aware version of arm_dma_unmap_page() + */ +static void arm_coherent_iommu_unmap_page(struct device *dev, dma_addr_t handle, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + struct dma_iommu_mapping *mapping = dev->archdata.mapping; + dma_addr_t iova = handle & PAGE_MASK; + struct page *page = phys_to_page(iommu_iova_to_phys(mapping->domain, iova)); + int offset = handle & ~PAGE_MASK; + int len = PAGE_ALIGN(size + offset); + + if (!iova) + return; + + iommu_unmap(mapping->domain, iova, len); + __free_iova(mapping, iova, len); +} + /** * arm_iommu_unmap_page * @dev: valid struct device pointer @@ -1566,7 +1657,7 @@ static void arm_iommu_unmap_page(struct device *dev, dma_addr_t handle, if (!iova) return; - if (!arch_is_coherent() && !dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) + if (!dma_get_attr(DMA_ATTR_SKIP_CPU_SYNC, attrs)) __dma_page_dev_to_cpu(page, offset, size, dir); iommu_unmap(mapping->domain, iova, len); @@ -1584,8 +1675,7 @@ static void arm_iommu_sync_single_for_cpu(struct device *dev, if (!iova) return; - if (!arch_is_coherent()) - __dma_page_dev_to_cpu(page, offset, size, dir); + __dma_page_dev_to_cpu(page, offset, size, dir); } static void arm_iommu_sync_single_for_device(struct device *dev, @@ -1619,6 +1709,19 @@ struct dma_map_ops iommu_ops = { .sync_sg_for_device = arm_iommu_sync_sg_for_device, }; +struct dma_map_ops iommu_coherent_ops = { + .alloc = arm_iommu_alloc_attrs, + .free = arm_iommu_free_attrs, + .mmap = arm_iommu_mmap_attrs, + .get_sgtable = arm_iommu_get_sgtable, + + .map_page = arm_coherent_iommu_map_page, + .unmap_page = arm_coherent_iommu_unmap_page, + + .map_sg = arm_coherent_iommu_map_sg, + .unmap_sg = arm_coherent_iommu_unmap_sg, +}; + /** * arm_iommu_create_mapping * @bus: pointer to the bus holding the client device (for IOMMU calls) -- cgit v1.2.3-18-g5258 From 48aa820f1e3824e46dde6251db98e5961abf605d Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Tue, 21 Aug 2012 12:26:24 +0200 Subject: ARM: kill off arch_is_coherent With ixp2xxx removed, there are no platforms that define arch_is_coherent, so the last occurrences of arch_is_coherent can be removed. Any new platform with coherent i/o should use coherent dma mapping functions. Signed-off-by: Rob Herring Cc: Russell King Cc: Marek Szyprowski Signed-off-by: Marek Szyprowski --- arch/arm/include/asm/barrier.h | 7 +++---- arch/arm/include/asm/memory.h | 8 -------- arch/arm/mm/mmu.c | 17 +++-------------- 3 files changed, 6 insertions(+), 26 deletions(-) diff --git a/arch/arm/include/asm/barrier.h b/arch/arm/include/asm/barrier.h index 05112380dc5..8dcd9c702d9 100644 --- a/arch/arm/include/asm/barrier.h +++ b/arch/arm/include/asm/barrier.h @@ -44,10 +44,9 @@ #define rmb() dsb() #define wmb() mb() #else -#include -#define mb() do { if (arch_is_coherent()) dmb(); else barrier(); } while (0) -#define rmb() do { if (arch_is_coherent()) dmb(); else barrier(); } while (0) -#define wmb() do { if (arch_is_coherent()) dmb(); else barrier(); } while (0) +#define mb() barrier() +#define rmb() barrier() +#define wmb() barrier() #endif #ifndef CONFIG_SMP diff --git a/arch/arm/include/asm/memory.h b/arch/arm/include/asm/memory.h index 5f6ddcc5645..73cf03aa981 100644 --- a/arch/arm/include/asm/memory.h +++ b/arch/arm/include/asm/memory.h @@ -275,14 +275,6 @@ static inline __deprecated void *bus_to_virt(unsigned long x) #define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT) #define virt_addr_valid(kaddr) ((unsigned long)(kaddr) >= PAGE_OFFSET && (unsigned long)(kaddr) < (unsigned long)high_memory) -/* - * Optional coherency support. Currently used only by selected - * Intel XSC3-based systems. - */ -#ifndef arch_is_coherent -#define arch_is_coherent() 0 -#endif - #endif #include diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c index c2fa21d0103..8fd039929ae 100644 --- a/arch/arm/mm/mmu.c +++ b/arch/arm/mm/mmu.c @@ -216,7 +216,7 @@ static struct mem_type mem_types[] = { .prot_l1 = PMD_TYPE_TABLE, .prot_sect = PROT_SECT_DEVICE | PMD_SECT_WB, .domain = DOMAIN_IO, - }, + }, [MT_DEVICE_WC] = { /* ioremap_wc */ .prot_pte = PROT_PTE_DEVICE | L_PTE_MT_DEV_WC, .prot_l1 = PMD_TYPE_TABLE, @@ -421,17 +421,6 @@ static void __init build_mem_type_table(void) cp = &cache_policies[cachepolicy]; vecs_pgprot = kern_pgprot = user_pgprot = cp->pte; - /* - * Enable CPU-specific coherency if supported. - * (Only available on XSC3 at the moment.) - */ - if (arch_is_coherent() && cpu_is_xsc3()) { - mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S; - mem_types[MT_MEMORY].prot_pte |= L_PTE_SHARED; - mem_types[MT_MEMORY_DMA_READY].prot_pte |= L_PTE_SHARED; - mem_types[MT_MEMORY_NONCACHED].prot_sect |= PMD_SECT_S; - mem_types[MT_MEMORY_NONCACHED].prot_pte |= L_PTE_SHARED; - } /* * ARMv6 and above have extended page tables. */ @@ -777,8 +766,8 @@ void __init iotable_init(struct map_desc *io_desc, int nr) create_mapping(md); vm->addr = (void *)(md->virtual & PAGE_MASK); vm->size = PAGE_ALIGN(md->length + (md->virtual & ~PAGE_MASK)); - vm->phys_addr = __pfn_to_phys(md->pfn); - vm->flags = VM_IOREMAP | VM_ARM_STATIC_MAPPING; + vm->phys_addr = __pfn_to_phys(md->pfn); + vm->flags = VM_IOREMAP | VM_ARM_STATIC_MAPPING; vm->flags |= VM_ARM_MTYPE(md->type); vm->caller = iotable_init; vm_area_add_early(vm++); -- cgit v1.2.3-18-g5258 From 1dc737c44a11c66f23ebeb09796f52f4b60d59c7 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Tue, 21 Aug 2012 12:31:06 +0200 Subject: ARM: highbank: add coherent DMA setup Some highbank DMA masters can support coherent (ACP) or non-coherent DMA. This sets up dma_map_ops for masters which are configured for coherent DMA. Signed-off-by: Rob Herring Signed-off-by: Marek Szyprowski --- .../devicetree/bindings/ata/ahci-platform.txt | 3 ++ .../devicetree/bindings/dma/arm-pl330.txt | 3 ++ .../devicetree/bindings/net/calxeda-xgmac.txt | 3 ++ arch/arm/boot/dts/highbank.dts | 1 + arch/arm/mach-highbank/highbank.c | 52 ++++++++++++++++++++++ 5 files changed, 62 insertions(+) diff --git a/Documentation/devicetree/bindings/ata/ahci-platform.txt b/Documentation/devicetree/bindings/ata/ahci-platform.txt index 8bb8a76d42e..6c1ad01d27e 100644 --- a/Documentation/devicetree/bindings/ata/ahci-platform.txt +++ b/Documentation/devicetree/bindings/ata/ahci-platform.txt @@ -8,6 +8,9 @@ Required properties: - interrupts : - reg : +Optional properties: +- dma-coherent : Present if dma operations are coherent + Example: sata@ffe08000 { compatible = "calxeda,hb-ahci"; diff --git a/Documentation/devicetree/bindings/dma/arm-pl330.txt b/Documentation/devicetree/bindings/dma/arm-pl330.txt index a4cd273b2a6..36e27d54260 100644 --- a/Documentation/devicetree/bindings/dma/arm-pl330.txt +++ b/Documentation/devicetree/bindings/dma/arm-pl330.txt @@ -9,6 +9,9 @@ Required properties: region. - interrupts: interrupt number to the cpu. +Optional properties: +- dma-coherent : Present if dma operations are coherent + Example: pdma0: pdma@12680000 { diff --git a/Documentation/devicetree/bindings/net/calxeda-xgmac.txt b/Documentation/devicetree/bindings/net/calxeda-xgmac.txt index 411727a3f82..c8ae996bd8f 100644 --- a/Documentation/devicetree/bindings/net/calxeda-xgmac.txt +++ b/Documentation/devicetree/bindings/net/calxeda-xgmac.txt @@ -6,6 +6,9 @@ Required properties: - interrupts : Should contain 3 xgmac interrupts. The 1st is main interrupt. The 2nd is pwr mgt interrupt. The 3rd is low power state interrupt. +Optional properties: +- dma-coherent : Present if dma operations are coherent + Example: ethernet@fff50000 { diff --git a/arch/arm/boot/dts/highbank.dts b/arch/arm/boot/dts/highbank.dts index 9fecf1ae777..7414577c177 100644 --- a/arch/arm/boot/dts/highbank.dts +++ b/arch/arm/boot/dts/highbank.dts @@ -121,6 +121,7 @@ compatible = "calxeda,hb-ahci"; reg = <0xffe08000 0x10000>; interrupts = <0 83 4>; + dma-coherent; }; sdhci@ffe0e000 { diff --git a/arch/arm/mach-highbank/highbank.c b/arch/arm/mach-highbank/highbank.c index d75b0a78d88..93617d61f4b 100644 --- a/arch/arm/mach-highbank/highbank.c +++ b/arch/arm/mach-highbank/highbank.c @@ -15,6 +15,7 @@ */ #include #include +#include #include #include #include @@ -23,6 +24,7 @@ #include #include #include +#include #include #include @@ -149,10 +151,60 @@ static void highbank_power_off(void) cpu_do_idle(); } +static int highbank_platform_notifier(struct notifier_block *nb, + unsigned long event, void *__dev) +{ + struct resource *res; + int reg = -1; + struct device *dev = __dev; + + if (event != BUS_NOTIFY_ADD_DEVICE) + return NOTIFY_DONE; + + if (of_device_is_compatible(dev->of_node, "calxeda,hb-ahci")) + reg = 0xc; + else if (of_device_is_compatible(dev->of_node, "calxeda,hb-sdhci")) + reg = 0x18; + else if (of_device_is_compatible(dev->of_node, "arm,pl330")) + reg = 0x20; + else if (of_device_is_compatible(dev->of_node, "calxeda,hb-xgmac")) { + res = platform_get_resource(to_platform_device(dev), + IORESOURCE_MEM, 0); + if (res) { + if (res->start == 0xfff50000) + reg = 0; + else if (res->start == 0xfff51000) + reg = 4; + } + } + + if (reg < 0) + return NOTIFY_DONE; + + if (of_property_read_bool(dev->of_node, "dma-coherent")) { + writel(0xff31, sregs_base + reg); + set_dma_ops(dev, &arm_coherent_dma_ops); + } else + writel(0, sregs_base + reg); + + return NOTIFY_OK; +} + +static struct notifier_block highbank_amba_nb = { + .notifier_call = highbank_platform_notifier, +}; + +static struct notifier_block highbank_platform_nb = { + .notifier_call = highbank_platform_notifier, +}; + static void __init highbank_init(void) { pm_power_off = highbank_power_off; + bus_register_notifier(&platform_bus_type, &highbank_platform_nb); + bus_register_notifier(&amba_bustype, &highbank_amba_nb); + of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL); } -- cgit v1.2.3-18-g5258 From abebfb18ea4fe69e9a04187a77fed4e11d67dba4 Mon Sep 17 00:00:00 2001 From: Hiroshi Doyu Date: Tue, 11 Sep 2012 07:43:59 +0200 Subject: ARM: dma-mapping: Remove unsed var at arm_coherent_iommu_unmap_page Signed-off-by: Hiroshi Doyu Signed-off-by: Marek Szyprowski --- arch/arm/mm/dma-mapping.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 5d5a4c4a5db..de7215c399c 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1624,7 +1624,6 @@ static void arm_coherent_iommu_unmap_page(struct device *dev, dma_addr_t handle, { struct dma_iommu_mapping *mapping = dev->archdata.mapping; dma_addr_t iova = handle & PAGE_MASK; - struct page *page = phys_to_page(iommu_iova_to_phys(mapping->domain, iova)); int offset = handle & ~PAGE_MASK; int len = PAGE_ALIGN(size + offset); -- cgit v1.2.3-18-g5258