linux-kernel - [PATCH stable v4.19] mm: memcg: fix memcg reclaim soft lockup

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>] [day] [month] [year] [list]

Message-Id: <20200921180610.62242-1-jpitti@cisco.com>
Date:   Mon, 21 Sep 2020 11:06:10 -0700
From:   Julius Hemanth Pitti <jpitti@...co.com>
To:     gregkh@...uxfoundation.org, akpm@...ux-foundation.org,
        xlpang@...ux.alibaba.com, mhocko@...e.com, vdavydov.dev@...il.com,
        ktkhai@...tuozzo.com, hannes@...xchg.org
Cc:     stable@...r.kernel.org, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org, xe-linux-external@...co.com,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Julius Hemanth Pitti <jpitti@...co.com>
Subject: [PATCH stable v4.19] mm: memcg: fix memcg reclaim soft lockup

From: Xunlei Pang <xlpang@...ux.alibaba.com>

commit e3336cab2579012b1e72b5265adf98e2d6e244ad upstream

We've met softlockup with "CONFIG_PREEMPT_NONE=y", when the target memcg
doesn't have any reclaimable memory.

It can be easily reproduced as below:

  watchdog: BUG: soft lockup - CPU#0 stuck for 111s![memcg_test:2204]
  CPU: 0 PID: 2204 Comm: memcg_test Not tainted 5.9.0-rc2+ #12
  Call Trace:
    shrink_lruvec+0x49f/0x640
    shrink_node+0x2a6/0x6f0
    do_try_to_free_pages+0xe9/0x3e0
    try_to_free_mem_cgroup_pages+0xef/0x1f0
    try_charge+0x2c1/0x750
    mem_cgroup_charge+0xd7/0x240
    __add_to_page_cache_locked+0x2fd/0x370
    add_to_page_cache_lru+0x4a/0xc0
    pagecache_get_page+0x10b/0x2f0
    filemap_fault+0x661/0xad0
    ext4_filemap_fault+0x2c/0x40
    __do_fault+0x4d/0xf9
    handle_mm_fault+0x1080/0x1790

It only happens on our 1-vcpu instances, because there's no chance for
oom reaper to run to reclaim the to-be-killed process.

Add a cond_resched() at the upper shrink_node_memcgs() to solve this
issue, this will mean that we will get a scheduling point for each memcg
in the reclaimed hierarchy without any dependency on the reclaimable
memory in that memcg thus making it more predictable.

[jpitti@...co.com:
   - backported to v4.19.y
   - Upstream patch applies fix in shrink_node_memcgs(), which
     is not present to v4.19.y. Appled to shrink_node()]

Suggested-by: Michal Hocko <mhocko@...e.com>
Signed-off-by: Xunlei Pang <xlpang@...ux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@...ux-foundation.org>
Acked-by: Chris Down <chris@...isdown.name>
Acked-by: Michal Hocko <mhocko@...e.com>
Acked-by: Johannes Weiner <hannes@...xchg.org>
Link: http://lkml.kernel.org/r/1598495549-67324-1-git-send-email-xlpang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@...ux-foundation.org>
Fixes: b0dedc49a2da ("mm/vmscan.c: iterate only over charged shrinkers during memcg shrink_slab()")
Cc: stable@...r.kernel.org
Signed-off-by: Julius Hemanth Pitti <jpitti@...co.com>
---
 mm/vmscan.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/mm/vmscan.c b/mm/vmscan.c
index bc2ecd43251a..b93dc8fc6007 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2708,6 +2708,14 @@ static bool shrink_node(pg_data_t *pgdat, struct scan_control *sc)
 			unsigned long reclaimed;
 			unsigned long scanned;
 
+			/*
+			 * This loop can become CPU-bound when target memcgs
+			 * aren't eligible for reclaim - either because they
+			 * don't have any reclaimable pages, or because their
+			 * memory is explicitly protected. Avoid soft lockups.
+			 */
+			cond_resched();
+
 			switch (mem_cgroup_protected(root, memcg)) {
 			case MEMCG_PROT_MIN:
 				/*
-- 
2.17.1