提交 309dbbab 编写于 作者: R Russell King 提交者: Russell King

[ARM] dma: don't touch cache on dma_*_for_cpu()

As per the dma_unmap_* calls, we don't touch the cache when a DMA
buffer transitions from device to CPU ownership.  Presently, no
problems have been identified with speculative cache prefetching
which in itself is a new feature in later architectures.  We may
have to revisit the DMA API later for these architectures anyway.
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>
上级 0e18b5d7
......@@ -376,11 +376,7 @@ static inline void dma_sync_single_range_for_cpu(struct device *dev,
{
BUG_ON(!valid_dma_direction(dir));
if (!dmabounce_sync_for_cpu(dev, handle, offset, size, dir))
return;
if (!arch_is_coherent())
dma_cache_maint(dma_to_virt(dev, handle) + offset, size, dir);
dmabounce_sync_for_cpu(dev, handle, offset, size, dir);
}
static inline void dma_sync_single_range_for_device(struct device *dev,
......
......@@ -585,12 +585,8 @@ void dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg,
int i;
for_each_sg(sg, s, nents, i) {
if (!dmabounce_sync_for_cpu(dev, sg_dma_address(s), 0,
sg_dma_len(s), dir))
continue;
if (!arch_is_coherent())
dma_cache_maint(sg_virt(s), s->length, dir);
dmabounce_sync_for_cpu(dev, sg_dma_address(s), 0,
sg_dma_len(s), dir);
}
}
EXPORT_SYMBOL(dma_sync_sg_for_cpu);
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册