linux-kernel - [PATCH mm-unstable v2 3/5] mm: increment gen # before restarting traversal

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

Message-ID: <20240813204716.842811-4-kinseyho@google.com>
Date: Tue, 13 Aug 2024 20:47:13 +0000
From: Kinsey Ho <kinseyho@...gle.com>
To: Andrew Morton <akpm@...ux-foundation.org>
Cc: linux-mm@...ck.org, linux-kernel@...r.kernel.org, cgroups@...r.kernel.org, 
	Yosry Ahmed <yosryahmed@...gle.com>, Roman Gushchin <roman.gushchin@...ux.dev>, 
	Johannes Weiner <hannes@...xchg.org>, Michal Hocko <mhocko@...nel.org>, 
	Shakeel Butt <shakeel.butt@...ux.dev>, Muchun Song <muchun.song@...ux.dev>, 
	Tejun Heo <tj@...nel.org>, Zefan Li <lizefan.x@...edance.com>, mkoutny@...e.com, 
	Kinsey Ho <kinseyho@...gle.com>
Subject: [PATCH mm-unstable v2 3/5] mm: increment gen # before restarting traversal

The generation number in struct mem_cgroup_reclaim_iter should be
incremented on every round-trip. Currently, it is possible for a
concurrent reclaimer to jump in at the end of the hierarchy, causing a
traversal restart (resetting the iteration position) without
incrementing the generation number.

By resetting the position without incrementing the generation, it's
possible for another ongoing mem_cgroup_iter() thread to walk the tree
twice.

Move the traversal restart such that the generation number is
incremented before the restart.

Signed-off-by: Kinsey Ho <kinseyho@...gle.com>
---
 mm/memcontrol.c | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 1688aae3b1b4..937b7efc41ca 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1036,7 +1036,7 @@ struct mem_cgroup *mem_cgroup_iter(struct mem_cgroup *root,
 		root = root_mem_cgroup;
 
 	rcu_read_lock();
-
+restart:
 	if (reclaim) {
 		struct mem_cgroup_per_node *mz;
 
@@ -1063,14 +1063,6 @@ struct mem_cgroup *mem_cgroup_iter(struct mem_cgroup *root,
 	for (;;) {
 		css = css_next_descendant_pre(css, &root->css);
 		if (!css) {
-			/*
-			 * Reclaimers share the hierarchy walk, and a
-			 * new one might jump in right at the end of
-			 * the hierarchy - make sure they see at least
-			 * one group and restart from the beginning.
-			 */
-			if (!prev)
-				continue;
 			break;
 		}
 
@@ -1093,8 +1085,18 @@ struct mem_cgroup *mem_cgroup_iter(struct mem_cgroup *root,
 		 */
 		(void)cmpxchg(&iter->position, pos, memcg);
 
-		if (!memcg)
+		if (!memcg) {
 			iter->generation++;
+
+			/*
+			 * Reclaimers share the hierarchy walk, and a
+			 * new one might jump in right at the end of
+			 * the hierarchy - make sure they see at least
+			 * one group and restart from the beginning.
+			 */
+			if (!prev)
+				goto restart;
+		}
 	}
 
 out_unlock:
-- 
2.46.0.76.ge559c4bf1a-goog