The patch titled
Subject: mm: memcg: use READ_ONCE()/WRITE_ONCE() to access stock->cached
has been added to the -mm mm-unstable branch. Its filename is
mm-memcg-use-read_once-write_once-to-access-stock-cached.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-memcg-use-read_once-write_once-to-access-stock-cached.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Roman Gushchin <roman.gushchin@xxxxxxxxx>
Subject: mm: memcg: use READ_ONCE()/WRITE_ONCE() to access stock->cached
Date: Tue, 2 May 2023 09:08:39 -0700
A memcg pointer in the percpu stock can be accessed by drain_all_stock()
from another cpu in a lockless way. In theory it might lead to an issue,
similar to the one which has been discovered with stock->cached_objcg,
where the pointer was zeroed between the check for being NULL and
dereferencing. In this case the issue is unlikely a real problem, but to
make it bulletproof and similar to stock->cached_objcg, let's annotate all
accesses to stock->cached with READ_ONCE()/WTRITE_ONCE().
Link: https://lkml.kernel.org/r/20230502160839.361544-2-roman.gushchin@xxxxxxxxx
Signed-off-by: Roman Gushchin <roman.gushchin@xxxxxxxxx>
Cc: Dmitry Vyukov <dvyukov@xxxxxxxxxx>
Cc: Yosry Ahmed <yosryahmed@xxxxxxxxxx>
Cc: Shakeel Butt <shakeelb@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/memcontrol.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
--- a/mm/memcontrol.c~mm-memcg-use-read_once-write_once-to-access-stock-cached
+++ a/mm/memcontrol.c
@@ -2275,7 +2275,7 @@ static bool consume_stock(struct mem_cgr
local_lock_irqsave(&memcg_stock.stock_lock, flags);
stock = this_cpu_ptr(&memcg_stock);
- if (memcg == stock->cached && stock->nr_pages >= nr_pages) {
+ if (memcg == READ_ONCE(stock->cached) && stock->nr_pages >= nr_pages) {
stock->nr_pages -= nr_pages;
ret = true;
}
@@ -2290,7 +2290,7 @@ static bool consume_stock(struct mem_cgr
*/
static void drain_stock(struct memcg_stock_pcp *stock)
{
- struct mem_cgroup *old = stock->cached;
+ struct mem_cgroup *old = READ_ONCE(stock->cached);
if (!old)
return;
@@ -2303,7 +2303,7 @@ static void drain_stock(struct memcg_sto
}
css_put(&old->css);
- stock->cached = NULL;
+ WRITE_ONCE(stock->cached, NULL);
}
static void drain_local_stock(struct work_struct *dummy)
@@ -2338,10 +2338,10 @@ static void __refill_stock(struct mem_cg
struct memcg_stock_pcp *stock;
stock = this_cpu_ptr(&memcg_stock);
- if (stock->cached != memcg) { /* reset if necessary */
+ if (READ_ONCE(stock->cached) != memcg) { /* reset if necessary */
drain_stock(stock);
css_get(&memcg->css);
- stock->cached = memcg;
+ WRITE_ONCE(stock->cached, memcg);
}
stock->nr_pages += nr_pages;
@@ -2383,7 +2383,7 @@ static void drain_all_stock(struct mem_c
bool flush = false;
rcu_read_lock();
- memcg = stock->cached;
+ memcg = READ_ONCE(stock->cached);
if (memcg && stock->nr_pages &&
mem_cgroup_is_descendant(memcg, root_memcg))
flush = true;
_
Patches currently in -mm which might be from roman.gushchin@xxxxxxxxx are
mm-kmem-fix-a-null-pointer-dereference-in-obj_stock_flush_required.patch
mm-memcg-use-read_once-write_once-to-access-stock-cached.patch
