This is a note to let you know that I've just added the patch titled
block: fix bio_clone_blkg_association() to associate with proper blkcg_gq
to the 5.10-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
block-fix-bio_clone_blkg_association-to-associate-with-proper-blkcg_gq.patch
and it can be found in the queue-5.10 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From foo@baz Mon Jun 6 08:05:28 PM CEST 2022
From: Jan Kara <jack@xxxxxxx>
Date: Mon, 6 Jun 2022 19:56:41 +0200
Subject: block: fix bio_clone_blkg_association() to associate with proper blkcg_gq
To: stable@xxxxxxxxxxxxxxx
Cc: <linux-block@xxxxxxxxxxxxxxx>, Paolo Valente <paolo.valente@xxxxxxxxxx>, Jens Axboe <axboe@xxxxxxxxx>, Jan Kara <jack@xxxxxxx>, Logan Gunthorpe <logang@xxxxxxxxxxxx>, Donald Buczek <buczek@xxxxxxxxxxxxx>, Christoph Hellwig <hch@xxxxxx>
Message-ID: <20220606175655.8993-6-jack@xxxxxxx>
From: Jan Kara <jack@xxxxxxx>
commit 22b106e5355d6e7a9c3b5cb5ed4ef22ae585ea94 upstream.
Commit d92c370a16cb ("block: really clone the block cgroup in
bio_clone_blkg_association") changed bio_clone_blkg_association() to
just clone bio->bi_blkg reference from source to destination bio. This
is however wrong if the source and destination bios are against
different block devices because struct blkcg_gq is different for each
bdev-blkcg pair. This will result in IOs being accounted (and throttled
as a result) multiple times against the same device (src bdev) while
throttling of the other device (dst bdev) is ignored. In case of BFQ the
inconsistency can even result in crashes in bfq_bic_update_cgroup().
Fix the problem by looking up correct blkcg_gq for the cloned bio.
Reported-by: Logan Gunthorpe <logang@xxxxxxxxxxxx>
Reported-and-tested-by: Donald Buczek <buczek@xxxxxxxxxxxxx>
Fixes: d92c370a16cb ("block: really clone the block cgroup in bio_clone_blkg_association")
CC: stable@xxxxxxxxxxxxxxx
Reviewed-by: Christoph Hellwig <hch@xxxxxx>
Signed-off-by: Jan Kara <jack@xxxxxxx>
Link: https://lore.kernel.org/r/20220602081242.7731-1-jack@xxxxxxx
Signed-off-by: Jens Axboe <axboe@xxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
block/blk-cgroup.c | 8 ++------
1 file changed, 2 insertions(+), 6 deletions(-)
--- a/block/blk-cgroup.c
+++ b/block/blk-cgroup.c
@@ -1892,12 +1892,8 @@ EXPORT_SYMBOL_GPL(bio_associate_blkg);
*/
void bio_clone_blkg_association(struct bio *dst, struct bio *src)
{
- if (src->bi_blkg) {
- if (dst->bi_blkg)
- blkg_put(dst->bi_blkg);
- blkg_get(src->bi_blkg);
- dst->bi_blkg = src->bi_blkg;
- }
+ if (src->bi_blkg)
+ bio_associate_blkg_from_css(dst, &bio_blkcg(src)->css);
}
EXPORT_SYMBOL_GPL(bio_clone_blkg_association);
Patches currently in stable-queue which might be from jack@xxxxxxx are
queue-5.10/bfq-split-shared-queues-on-move-between-cgroups.patch
queue-5.10/ext4-avoid-cycles-in-directory-h-tree.patch
queue-5.10/bfq-get-rid-of-__bio_blkcg-usage.patch
queue-5.10/ext4-fix-bug_on-in-__es_tree_search.patch
queue-5.10/bfq-track-whether-bfq_group-is-still-online.patch
queue-5.10/bfq-make-sure-bfqg-for-which-we-are-queueing-requests-is-online.patch
queue-5.10/bfq-drop-pointless-unlock-lock-pair.patch
queue-5.10/ext4-fix-bug_on-in-ext4_writepages.patch
queue-5.10/dax-fix-cache-flush-on-pmd-mapped-pages.patch
queue-5.10/fsnotify-fix-wrong-lockdep-annotations.patch
queue-5.10/bfq-avoid-merging-queues-with-different-parents.patch
queue-5.10/inotify-show-inotify-mask-flags-in-proc-fdinfo.patch
queue-5.10/ext4-fix-use-after-free-in-ext4_rename_dir_prepare.patch
queue-5.10/fs-writeback-writeback_sb_inodes-recalculate-wrote-according-skipped-pages.patch
queue-5.10/bfq-remove-pointless-bfq_init_rq-calls.patch
queue-5.10/ext4-verify-dir-block-before-splitting-it.patch
queue-5.10/block-fix-bio_clone_blkg_association-to-associate-with-proper-blkcg_gq.patch
queue-5.10/bfq-update-cgroup-information-before-merging-bio.patch
