From: Andrei Vagin <avagin@xxxxxxxxxx> commit 432dc0654c612457285a5dcf9bb13968ac6f0804 upstream. The inc_rlimit_get_ucounts() increments the specified rlimit counter and then checks its limit. If the value exceeds the limit, the function returns an error without decrementing the counter. Link: https://lkml.kernel.org/r/20241101191940.3211128-1-roman.gushchin@xxxxxxxxx Fixes: 15bc01effefe ("ucounts: Fix signal ucount refcounting") Signed-off-by: Andrei Vagin <avagin@xxxxxxxxxx> Co-developed-by: Roman Gushchin <roman.gushchin@xxxxxxxxx> Signed-off-by: Roman Gushchin <roman.gushchin@xxxxxxxxx> Tested-by: Roman Gushchin <roman.gushchin@xxxxxxxxx> Acked-by: Alexey Gladkov <legion@xxxxxxxxxx> Cc: Kees Cook <kees@xxxxxxxxxx> Cc: Andrei Vagin <avagin@xxxxxxxxxx> Cc: "Eric W. Biederman" <ebiederm@xxxxxxxxxxxx> Cc: Alexey Gladkov <legion@xxxxxxxxxx> Cc: Oleg Nesterov <oleg@xxxxxxxxxx> Cc: <stable@xxxxxxxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- kernel/ucount.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) --- a/kernel/ucount.c +++ b/kernel/ucount.c @@ -318,7 +318,7 @@ long inc_rlimit_get_ucounts(struct ucoun for (iter = ucounts; iter; iter = iter->ns->ucounts) { long new = atomic_long_add_return(1, &iter->rlimit[type]); if (new < 0 || new > max) - goto unwind; + goto dec_unwind; if (iter == ucounts) ret = new; if (!override_rlimit) @@ -336,7 +336,6 @@ long inc_rlimit_get_ucounts(struct ucoun dec_unwind: dec = atomic_long_sub_return(1, &iter->rlimit[type]); WARN_ON_ONCE(dec < 0); -unwind: do_dec_rlimit_put_ucounts(ucounts, iter, type); return 0; } Patches currently in stable-queue which might be from avagin@xxxxxxxxxx are queue-6.11/signal-restore-the-override_rlimit-logic.patch queue-6.11/ucounts-fix-counter-leak-in-inc_rlimit_get_ucounts.patch
