[2/2] dma-mapping: deprecate dma_zalloc_coherent

Commit Message

Christoph Hellwig Dec. 14, 2018, 8:25 a.m. UTC

We now always return zeroed memory from dma_alloc_coherent.  Note that
simply passing GFP_ZERO to dma_alloc_coherent wasn't always doing the
right thing to start with given that various allocators are not backed
by the page allocator and thus would ignore GFP_ZERO.

Signed-off-by: Christoph Hellwig <[e-mail]>
---
 Documentation/DMA-API.txt   | 9 ---------
 include/linux/dma-mapping.h | 7 ++++---
 2 files changed, 4 insertions(+), 12 deletions(-)

Patch

diff --git a/Documentation/DMA-API.txt b/Documentation/DMA-API.txt
index 016eb6909b8a..e133ccd60228 100644
--- a/Documentation/DMA-API.txt
+++ b/Documentation/DMA-API.txt
@@ -58,15 +58,6 @@  specify the ``GFP_`` flags (see kmalloc()) for the allocation (the
 implementation may choose to ignore flags that affect the location of
 the returned memory, like GFP_DMA).
 
-::
-
-	void *
-	dma_zalloc_coherent(struct device *dev, size_t size,
-			    dma_addr_t *dma_handle, gfp_t flag)
-
-Wraps dma_alloc_coherent() and also zeroes the returned memory if the
-allocation attempt succeeded.
-
 ::
 
 	void
diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index f422aec0f53c..a52c6409bdc2 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -644,12 +644,13 @@  static inline unsigned long dma_max_pfn(struct device *dev)
 }
 #endif
 
+/*
+ * Please always use dma_alloc_coherent instead as it already zeroes the memory!
+ */
 static inline void *dma_zalloc_coherent(struct device *dev, size_t size,
 					dma_addr_t *dma_handle, gfp_t flag)
 {
-	void *ret = dma_alloc_coherent(dev, size, dma_handle,
-				       flag | __GFP_ZERO);
-	return ret;
+	return dma_alloc_coherent(dev, size, dma_handle, flag);
 }
 
 static inline int dma_get_cache_alignment(void)