Re: [PATCH 03/11] blk-throttle: ignore cgroup without io queued in blk_throtl_cancel_bios
From: Tejun Heo <tj@kernel.org>
Date: 2022-11-23 18:16:13
Also in:
cgroups, lkml
On Wed, Nov 23, 2022 at 02:03:53PM +0800, Kemeng Shi wrote:
quoted hunk ↗ jump to hunk
Ignore cgroup without io queued in blk_throtl_cancel_bios for two reasons: 1. Save cpu cycle for trying to dispatch cgroup which is no io queued. 2. Avoid non-consistent state that cgroup is inserted to service queue without THROTL_TG_PENDING set as tg_update_disptime will unconditional re-insert cgroup to service queue. If we are on the default hierarchy, IO dispatched from child in tg_dispatch_one_bio will trigger inserting cgroup to service queue without erase first and ruin the tree. Signed-off-by: Kemeng Shi <redacted> --- block/blk-throttle.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)diff --git a/block/blk-throttle.c b/block/blk-throttle.c index b33bcf53b36e..acfac916ed99 100644 --- a/block/blk-throttle.c +++ b/block/blk-throttle.c@@ -1728,7 +1728,12 @@ void blk_throtl_cancel_bios(struct gendisk *disk) * Set the flag to make sure throtl_pending_timer_fn() won't * stop until all throttled bios are dispatched. */ - blkg_to_tg(blkg)->flags |= THROTL_TG_CANCELING; + tg->flags |= THROTL_TG_CANCELING; + + /* No need to dispatch cgroup without io queued */
It's probably more important to mention that putting this in without PENDING set can corrupt the tree in the future through double insertion.
+ if (!(tg->flags & THROTL_TG_PENDING)) + continue; +
Other than than the above comment, Acked-by: Tejun Heo [off-list ref] Thanks. -- tejun