microblaze: use generic dma_noncoherent_ops

Switch to the generic noncoherent direct mapping implementation.

This removes the direction-based optimizations in
sync_{single,sg}_for_{cpu,device} which were marked untestested and
do not match the usually very well tested {un,}map_{single,sg}
implementations.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
This commit is contained in:
Christoph Hellwig
2018-07-19 05:54:39 -07:00
committed by Michal Simek
parent bd05a58ddf
commit 5411ad27db
6 changed files with 22 additions and 166 deletions

View File

@@ -553,8 +553,6 @@ void __init *early_get_page(void);
extern unsigned long ioremap_bot, ioremap_base;
void *consistent_alloc(gfp_t gfp, size_t size, dma_addr_t *dma_handle);
void consistent_free(size_t size, void *vaddr);
void consistent_sync(void *vaddr, size_t size, int direction);
void consistent_sync_page(struct page *page, unsigned long offset,
size_t size, int direction);