Commit 30c00eda authored by FUJITA Tomonori's avatar FUJITA Tomonori Committed by Jens Axboe

block: blk_rq_map_kern uses the bounce buffers for stack buffers

blk_rq_map_kern is used for kernel internal I/Os. Some callers use
this function with stack buffers but DMA to/from the stack buffers
leads to memory corruption on a non-coherent platform.

This patch make blk_rq_map_kern uses the bounce buffers if a caller
passes a stack buffer (on the all platforms for simplicity).
Signed-off-by: default avatarFUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Tejun Heo <htejun@gmail.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarJens Axboe <jens.axboe@oracle.com>
parent 27f8221a
...@@ -269,6 +269,7 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf, ...@@ -269,6 +269,7 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf,
int reading = rq_data_dir(rq) == READ; int reading = rq_data_dir(rq) == READ;
int do_copy = 0; int do_copy = 0;
struct bio *bio; struct bio *bio;
unsigned long stack_mask = ~(THREAD_SIZE - 1);
if (len > (q->max_hw_sectors << 9)) if (len > (q->max_hw_sectors << 9))
return -EINVAL; return -EINVAL;
...@@ -279,6 +280,10 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf, ...@@ -279,6 +280,10 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf,
alignment = queue_dma_alignment(q) | q->dma_pad_mask; alignment = queue_dma_alignment(q) | q->dma_pad_mask;
do_copy = ((kaddr & alignment) || (len & alignment)); do_copy = ((kaddr & alignment) || (len & alignment));
if (!((kaddr & stack_mask) ^
((unsigned long)current->stack & stack_mask)))
do_copy = 1;
if (do_copy) if (do_copy)
bio = bio_copy_kern(q, kbuf, len, gfp_mask, reading); bio = bio_copy_kern(q, kbuf, len, gfp_mask, reading);
else else
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment