From: majianpeng Date: Tue, 3 Jul 2012 02:11:54 +0000 (+1000) Subject: md/raid5: In ops_run_io, inc nr_pending before calling md_wait_for_blocked_rdev X-Git-Tag: v3.4.5~57 X-Git-Url: https://git.karo-electronics.de/?a=commitdiff_plain;h=8d9369807370331cebf3e237b95ecce068af80f1;p=karo-tx-linux.git md/raid5: In ops_run_io, inc nr_pending before calling md_wait_for_blocked_rdev commit 1850753d2e6d9ca7856581ca5d3cf09521e6a5d7 upstream. In ops_run_io(), the call to md_wait_for_blocked_rdev will decrement nr_pending so we lose the reference we hold on the rdev. So atomic_inc it first to maintain the reference. This bug was introduced by commit 73e92e51b7969ef5477d md/raid5. Don't write to known bad block on doubtful devices. which appeared in 3.0, so patch is suitable for stable kernels since then. Signed-off-by: majianpeng Signed-off-by: NeilBrown Signed-off-by: Greg Kroah-Hartman --- diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index f351422938e0..8fc591877dc1 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -583,6 +583,12 @@ static void ops_run_io(struct stripe_head *sh, struct stripe_head_state *s) * a chance*/ md_check_recovery(conf->mddev); } + /* + * Because md_wait_for_blocked_rdev + * will dec nr_pending, we must + * increment it first. + */ + atomic_inc(&rdev->nr_pending); md_wait_for_blocked_rdev(rdev, conf->mddev); } else { /* Acknowledged bad block - skip the write */