linux-kernel - [PATCH v2] blk-cgroup: don't clear stat in blkcg_reset

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20240821020756.786000-1-lilingfeng@huaweicloud.com>
Date: Wed, 21 Aug 2024 10:07:56 +0800
From: Li Lingfeng <lilingfeng@...weicloud.com>
To: tj@...nel.org,
	josef@...icpanda.com,
	hch@....de,
	axboe@...nel.dk,
	longman@...hat.com,
	ming.lei@...hat.com
Cc: cgroups@...r.kernel.org,
	linux-block@...r.kernel.org,
	linux-kernel@...r.kernel.org,
	yukuai1@...weicloud.com,
	houtao1@...wei.com,
	yi.zhang@...wei.com,
	yangerkun@...wei.com,
	yukuai3@...wei.com,
	lilingfeng@...weicloud.com,
	lilingfeng3@...wei.com
Subject: [PATCH v2] blk-cgroup: don't clear stat in blkcg_reset_stats()

From: Li Lingfeng <lilingfeng3@...wei.com>

The list corruption described in commit 6da668063279 ("blk-cgroup: fix
list corruption from resetting io stat") has no effect. It's unnecessary
to fix it.

As for cgroup v1, it does not use iostat any more after commit
ad7c3b41e86b("blk-throttle: Fix io statistics for cgroup v1"), so using
memset to clear iostat has no real effect.
As for cgroup v2, it will not call blkcg_reset_stats() to corrupt the
list.

The list of root cgroup can be used by both cgroup v1 and v2 while
non-root cgroup can't since it must be removed before switch between
cgroup v1 and v2.
So it may has effect if the list of root used by cgroup v2 was corrupted
after switching to cgroup v1, and switch back to cgroup v2 to use the
corrupted list again.
However, the root cgroup will not use the list any more after commit
ef45fe470e1e("blk-cgroup: show global disk stats in root cgroup io.stat").

Although this has no negative effect, it is not necessary. Remove the
related code for cleanup. No function change.

Signed-off-by: Li Lingfeng <lilingfeng3@...wei.com>
---
  v1->v2: remove the fix tag and mark this as cleanup.
 block/blk-cgroup.c | 24 ------------------------
 1 file changed, 24 deletions(-)

diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c
index 69e70964398c..6d091fa55b8c 100644
--- a/block/blk-cgroup.c
+++ b/block/blk-cgroup.c
@@ -629,29 +629,6 @@ static void blkg_iostat_set(struct blkg_iostat *dst, struct blkg_iostat *src)
 	}
 }

-static void __blkg_clear_stat(struct blkg_iostat_set *bis)
-{
-	struct blkg_iostat cur = {0};
-	unsigned long flags;
-
-	flags = u64_stats_update_begin_irqsave(&bis->sync);
-	blkg_iostat_set(&bis->cur, &cur);
-	blkg_iostat_set(&bis->last, &cur);
-	u64_stats_update_end_irqrestore(&bis->sync, flags);
-}
-
-static void blkg_clear_stat(struct blkcg_gq *blkg)
-{
-	int cpu;
-
-	for_each_possible_cpu(cpu) {
-		struct blkg_iostat_set *s = per_cpu_ptr(blkg->iostat_cpu, cpu);
-
-		__blkg_clear_stat(s);
-	}
-	__blkg_clear_stat(&blkg->iostat);
-}
-
 static int blkcg_reset_stats(struct cgroup_subsys_state *css,
 			     struct cftype *cftype, u64 val)
 {
@@ -668,7 +645,6 @@ static int blkcg_reset_stats(struct cgroup_subsys_state *css,
 	 * anyway.  If you get hit by a race, retry.
 	 */
 	hlist_for_each_entry(blkg, &blkcg->blkg_list, blkcg_node) {
-		blkg_clear_stat(blkg);
 		for (i = 0; i < BLKCG_MAX_POLS; i++) {
 			struct blkcg_policy *pol = blkcg_policy[i];

-- 
2.39.2