Copy documentation abstract about each DMA attribute from Documentation/DMA-attributes.txt to the place with declaration. Suggested-by: Christoph Hellwig <hch at infradead.org> Signed-off-by: Krzysztof Kozlowski <k.kozlowski at samsung.com> --- include/linux/dma-mapping.h | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index 6d013ba94213..f349df4cb009 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -14,14 +14,47 @@ /** * List of possible attributes associated with a DMA mapping. The semantics * of each attribute should be defined in Documentation/DMA-attributes.txt. + * + * DMA_ATTR_WRITE_BARRIER: DMA to a memory region with this attribute + * forces all pending DMA writes to complete. */ #define DMA_ATTR_WRITE_BARRIER (1UL << 1) +/* + * DMA_ATTR_WEAK_ORDERING: Specifies that reads and writes to the mapping + * may be weakly ordered, that is that reads and writes may pass each other. + */ #define DMA_ATTR_WEAK_ORDERING (1UL << 2) +/* + * DMA_ATTR_WRITE_COMBINE: Specifies that writes to the mapping may be + * buffered to improve performance. + */ #define DMA_ATTR_WRITE_COMBINE (1UL << 3) +/* + * DMA_ATTR_NON_CONSISTENT: Lets the platform to choose to return either + * consistent or non-consistent memory as it sees fit. + */ #define DMA_ATTR_NON_CONSISTENT (1UL << 4) +/* + * DMA_ATTR_NO_KERNEL_MAPPING: Lets the platform to avoid creating a kernel + * virtual mapping for the allocated buffer. + */ #define DMA_ATTR_NO_KERNEL_MAPPING (1UL << 5) +/* + * DMA_ATTR_SKIP_CPU_SYNC: Allows platform code to skip synchronization of + * the CPU cache for the given buffer assuming that it has been already + * transferred to 'device' domain. + */ #define DMA_ATTR_SKIP_CPU_SYNC (1UL << 6) +/* + * DMA_ATTR_FORCE_CONTIGUOUS: Forces contiguous allocation of the buffer + * in physical memory. + */ #define DMA_ATTR_FORCE_CONTIGUOUS (1UL << 7) +/* + * DMA_ATTR_ALLOC_SINGLE_PAGES: This is a hint to the DMA-mapping subsystem + * that it's probably not worth the time to try to allocate memory to in a way + * that gives better TLB efficiency. + */ #define DMA_ATTR_ALLOC_SINGLE_PAGES (1UL << 8) /* -- 1.9.1
