cachestat: do not flush stats in recency check

syzbot detects that cachestat() is flushing stats, which can sleep, in its
RCU read section (see [1]).  This is done in the workingset_test_recent()
step (which checks if the folio's eviction is recent).

Move the stat flushing step to before the RCU read section of cachestat,
and skip stat flushing during the recency check.

[1]: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/

Link: https://lkml.kernel.org/r/20240627201737.3506959-1-nphamcs@gmail.com
Fixes: b00684722262 ("mm: workingset: move the stats flush into workingset_test_recent()")
Signed-off-by: Nhat Pham <nphamcs@gmail.com>
Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/
Debugged-by: Johannes Weiner <hannes@cmpxchg.org>
Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: David Hildenbrand <david@redhat.com>
Cc: "Huang, Ying" <ying.huang@intel.com>
Cc: Kairui Song <kasong@tencent.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Cc: <stable@vger.kernel.org>	[6.8+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

2 files changed, 15 insertions, 4 deletions

diff --git a/mm/filemap.c b/mm/filemap.c
index fedefb10d947..298485d4b992 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -4248,6 +4248,9 @@ static void filemap_cachestat(struct address_space *mapping,
 	XA_STATE(xas, &mapping->i_pages, first_index);
 	struct folio *folio;
 
+	/* Flush stats (and potentially sleep) outside the RCU read section. */
+	mem_cgroup_flush_stats_ratelimited(NULL);
+
 	rcu_read_lock();
 	xas_for_each(&xas, folio, last_index) {
 		int order;
@@ -4311,7 +4314,7 @@ static void filemap_cachestat(struct address_space *mapping,
 					goto resched;
 			}
 #endif
-			if (workingset_test_recent(shadow, true, &workingset))
+			if (workingset_test_recent(shadow, true, &workingset, false))
 				cs->nr_recently_evicted += nr_pages;
 
 			goto resched;
diff --git a/mm/workingset.c b/mm/workingset.c
index c22adb93622a..a2b28e356e68 100644
--- a/mm/workingset.c
+++ b/mm/workingset.c
@@ -412,10 +412,12 @@ void *workingset_eviction(struct folio *folio, struct mem_cgroup *target_memcg)
  * @file: whether the corresponding folio is from the file lru.
  * @workingset: where the workingset value unpacked from shadow should
  * be stored.
+ * @flush: whether to flush cgroup rstat.
  *
  * Return: true if the shadow is for a recently evicted folio; false otherwise.
  */
-bool workingset_test_recent(void *shadow, bool file, bool *workingset)
+bool workingset_test_recent(void *shadow, bool file, bool *workingset,
+				bool flush)
 {
 	struct mem_cgroup *eviction_memcg;
 	struct lruvec *eviction_lruvec;
@@ -467,10 +469,16 @@ bool workingset_test_recent(void *shadow, bool file, bool *workingset)
 
 	/*
 	 * Flush stats (and potentially sleep) outside the RCU read section.
+	 *
+	 * Note that workingset_test_recent() itself might be called in RCU read
+	 * section (for e.g, in cachestat) - these callers need to skip flushing
+	 * stats (via the flush argument).
+	 *
 	 * XXX: With per-memcg flushing and thresholding, is ratelimiting
 	 * still needed here?
 	 */
-	mem_cgroup_flush_stats_ratelimited(eviction_memcg);
+	if (flush)
+		mem_cgroup_flush_stats_ratelimited(eviction_memcg);
 
 	eviction_lruvec = mem_cgroup_lruvec(eviction_memcg, pgdat);
 	refault = atomic_long_read(&eviction_lruvec->nonresident_age);
@@ -558,7 +566,7 @@ void workingset_refault(struct folio *folio, void *shadow)
 
 	mod_lruvec_state(lruvec, WORKINGSET_REFAULT_BASE + file, nr);
 
-	if (!workingset_test_recent(shadow, file, &workingset))
+	if (!workingset_test_recent(shadow, file, &workingset, true))
 		return;
 
 	folio_set_active(folio);