aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/block
diff options
context:
space:
mode:
authorMichał Mirosław <mirq-linux@rere.qmqm.pl>2010-12-17 02:56:44 -0500
committerJens Axboe <jaxboe@fusionio.com>2010-12-17 02:56:44 -0500
commita8adbe378b56acd5945df70753c7e8f6fe223304 (patch)
treeafbaae1cb71be95f342f6638bfe2d3f1e22dd9a4 /drivers/block
parente4ea0c16a85d221ebcc3a21f32e321440459e0fc (diff)
fs/splice: Pull buf->ops->confirm() from splice_from_pipe actors
This patch pulls calls to buf->ops->confirm() from all actors passed (also indirectly) to splice_from_pipe_feed(). Is avoiding the call to buf->ops->confirm() while splice()ing to /dev/null is an intentional optimization? No other user does that and this will remove this special case. Against current linux.git 6313e3c21743cc88bb5bd8aa72948ee1e83937b6. Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Diffstat (limited to 'drivers/block')
-rw-r--r--drivers/block/loop.c4
1 files changed, 0 insertions, 4 deletions
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index 7ea0bea2f7e3..c87b0840d2cb 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -397,10 +397,6 @@ lo_splice_actor(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
397 sector_t IV; 397 sector_t IV;
398 int size, ret; 398 int size, ret;
399 399
400 ret = buf->ops->confirm(pipe, buf);
401 if (unlikely(ret))
402 return ret;
403
404 IV = ((sector_t) page->index << (PAGE_CACHE_SHIFT - 9)) + 400 IV = ((sector_t) page->index << (PAGE_CACHE_SHIFT - 9)) +
405 (buf->offset >> 9); 401 (buf->offset >> 9);
406 size = sd->len; 402 size = sd->len;