diff options
| author | Tejun Heo <tj@kernel.org> | 2015-07-09 23:39:48 +0300 |
|---|---|---|
| committer | Jens Axboe <axboe@fb.com> | 2015-07-09 23:41:08 +0300 |
| commit | 144232b34258c1fc19729e077c6fb161e30da07b (patch) | |
| tree | 8f048fd73d864a8f41917068d6e610fc5b0e0f0b /block | |
| parent | 838f13bf4b6737d4aec508558e45f81798fc2677 (diff) | |
| download | linux-144232b34258c1fc19729e077c6fb161e30da07b.tar.xz | |
blkcg: blkcg_css_alloc() should grab blkcg_pol_mutex while iterating blkcg_policy[]
An entry in blkcg_policy[] is stable while there are non-bypassing
in-flight IOs on a request_queue which has the policy activated. This
is why most derefs of blkcg_policy[] don't need explicit locking;
however, blkcg_css_alloc() isn't invoked from IO path and thus doesn't
have this protection and may race policies being added and removed.
Fix it by adding explicit blkcg_pol_mutex protection around
blkcg_policy[] iteration in blkcg_css_alloc().
Signed-off-by: Tejun Heo <tj@kernel.org>
Fixes: e48453c386f3 ("block, cgroup: implement policy-specific per-blkcg data")
Cc: Vivek Goyal <vgoyal@redhat.com>
Cc: Arianna Avanzini <avanzini.arianna@gmail.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Diffstat (limited to 'block')
| -rw-r--r-- | block/blk-cgroup.c | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c index 2ff74ffcbb27..05b893de516b 100644 --- a/block/blk-cgroup.c +++ b/block/blk-cgroup.c @@ -844,6 +844,8 @@ blkcg_css_alloc(struct cgroup_subsys_state *parent_css) goto free_blkcg; } + mutex_lock(&blkcg_pol_mutex); + for (i = 0; i < BLKCG_MAX_POLS ; i++) { struct blkcg_policy *pol = blkcg_policy[i]; struct blkcg_policy_data *cpd; @@ -860,6 +862,7 @@ blkcg_css_alloc(struct cgroup_subsys_state *parent_css) BUG_ON(blkcg->pd[i]); cpd = kzalloc(pol->cpd_size, GFP_KERNEL); if (!cpd) { + mutex_unlock(&blkcg_pol_mutex); ret = ERR_PTR(-ENOMEM); goto free_pd_blkcg; } @@ -868,6 +871,7 @@ blkcg_css_alloc(struct cgroup_subsys_state *parent_css) pol->cpd_init_fn(blkcg); } + mutex_unlock(&blkcg_pol_mutex); done: spin_lock_init(&blkcg->lock); INIT_RADIX_TREE(&blkcg->blkg_tree, GFP_ATOMIC); |