Don't decrement bi_size in bio_endio
authorNeilBrown <neilb@suse.de>
Thu, 27 Sep 2007 10:46:13 +0000 (12:46 +0200)
committerJens Axboe <axboe@carl.home.kernel.dk>
Wed, 10 Oct 2007 07:25:57 +0000 (09:25 +0200)
commit5bb23a688b2de23d7765a1dd439d89c038378978
tree15e5e07c28b5aacd19d76ccdd4f7b2b75804d347
parent9cc54d40b8ca01fcefc9151044b6996565061d90
Don't decrement bi_size in bio_endio

The only caller of bio_endio that does not pass the full bi_size
is end_that_request_first.  Also, no ->bi_end_io method is really
interested in bi_size being decremented.

So move the decrement and related code into ll_rw_blk and merge it
with order_bio_endio to form req_bio_endio which does endio functionality
specific to request completion.

As some ->bi_end_io methods do check bi_size of 0, we set it thus for
now, but that will go in the next patch.

Signed-off-by: Neil Brown <neilb@suse.de>
### Diffstat output
 ./block/ll_rw_blk.c |   42 +++++++++++++++++++++++++++---------------
 ./fs/bio.c          |   23 +++++++++++------------
 2 files changed, 38 insertions(+), 27 deletions(-)

diff .prev/block/ll_rw_blk.c ./block/ll_rw_blk.c
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
block/ll_rw_blk.c
fs/bio.c