Messages in this thread | | | Date | Thu, 16 Nov 2023 16:13:45 +0100 | Subject | Re: [PATCH] mm: kmem: properly initialize local objcg variable in current_obj_cgroup() | From | Vlastimil Babka <> |
| |
On 11/16/23 15:56, Erhard Furtner wrote: > On Thu, 16 Nov 2023 08:04:18 +0100 > Vlastimil Babka <vbabka@suse.cz> wrote: > >> On 11/16/23 03:51, Roman Gushchin wrote: >> > Actually the problem is caused by uninitialized local variable in >> > current_obj_cgroup(). If the root memory cgroup is set as an active >> > memory cgroup for a charging scope (as in the trace, where systemd >> > tries to create the first non-root cgroup, so the parent cgroup is >> > the root cgroup), the "for" loop is skipped and uninitialized objcg is >> > returned, causing a panic down the accounting stack. >> > >> > The fix is trivial: initialize the objcg variable to NULL >> > unconditionally before the "for" loop. >> > >> > Fixes: e86828e5446d ("mm: kmem: scoped objcg protection") >> > Reported-by: Erhard Furtner <erhard_f@mailbox.org> >> > Closes: https://github.com/ClangBuiltLinux/linux/issues/1959 >> > Signed-off-by: Roman Gushchin (Cruise) <roman.gushchin@linux.dev> >> > Cc: Shakeel Butt <shakeelb@google.com> >> > Cc: Vlastimil Babka <vbabka@suse.cz> >> > Cc: David Rientjes <rientjes@google.com> >> > Cc: Dennis Zhou <dennis@kernel.org> >> > Cc: Johannes Weiner <hannes@cmpxchg.org> >> > Cc: Michal Hocko <mhocko@kernel.org> >> > Cc: Muchun Song <muchun.song@linux.dev> >> > Cc: stable@vger.kernel.org >> >> Acked-by: Vlastimil Babka <vbabka@suse.cz> >> >> We could also do this to make it less confusing? >> >> diff --git a/mm/memcontrol.c b/mm/memcontrol.c >> index 774bd6e21e27..a08bcec661b6 100644 >> --- a/mm/memcontrol.c >> +++ b/mm/memcontrol.c >> @@ -3175,7 +3175,6 @@ __always_inline struct obj_cgroup *current_obj_cgroup(void) >> objcg = rcu_dereference_check(memcg->objcg, 1); >> if (likely(objcg)) >> break; >> - objcg = NULL; >> } >> >> return objcg; >> >> > > I can confirm the 1st patch from Roman fixes the issue on my amd64 and on my i686 box.
Good.
> The 2nd patch from Vlastimil unfortunately does not (only tried on amd64).
Ah no, I meant mine as an additional related cleanup that's related enough that it can be part of Roman's fix. But it's not a fix on its own. > Regards, > Erhard
| |