From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mikulas Patocka Subject: [PATCH 2/2] use bio_endio instead of a call to bi_end_io Date: Tue, 24 Jun 2008 01:28:21 -0400 (EDT) Message-ID: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Return-path: Sender: linux-raid-owner@vger.kernel.org To: linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org Cc: axboe@kernel.dk, mingo@redhat.com List-Id: linux-raid.ids Hi The is sequel to my previous patch to avoid bio_endio recursion. raid5 bypasses bio_endio and calls the method directly (although it does the same as bio_endio). I changed bio_endio to avoid recursion, so it should be used instead of calling the method. Mikulas -- Turn calls to bi->bi_end_io() into bio_endio(). Apparently bio_endio does exactly the same error processing as is hardcoded at these places. bio_endio() avoids recursion, so it should be used. Signed-off-by: Mikulas Patocka Index: linux-2.6.26-rc7-devel/drivers/md/raid5.c =================================================================== --- linux-2.6.26-rc7-devel.orig/drivers/md/raid5.c 2008-06-24 07:28:15.000000000 +0200 +++ linux-2.6.26-rc7-devel/drivers/md/raid5.c 2008-06-24 07:28:53.000000000 +0200 @@ -115,9 +115,7 @@ return_bi = bi->bi_next; bi->bi_next = NULL; bi->bi_size = 0; - bi->bi_end_io(bi, - test_bit(BIO_UPTODATE, &bi->bi_flags) - ? 0 : -EIO); + bio_endio(bi, 0); bi = return_bi; } } @@ -3689,9 +3687,7 @@ if ( rw == WRITE ) md_write_end(mddev); - bi->bi_end_io(bi, - test_bit(BIO_UPTODATE, &bi->bi_flags) - ? 0 : -EIO); + bio_endio(bi, 0); } return 0; } @@ -3994,12 +3990,8 @@ spin_lock_irq(&conf->device_lock); remaining = --raid_bio->bi_phys_segments; spin_unlock_irq(&conf->device_lock); - if (remaining == 0) { - - raid_bio->bi_end_io(raid_bio, - test_bit(BIO_UPTODATE, &raid_bio->bi_flags) - ? 0 : -EIO); - } + if (remaining == 0) + bio_endio(raid_bio, 0); if (atomic_dec_and_test(&conf->active_aligned_reads)) wake_up(&conf->wait_for_stripe); return handled;