[PATCH] kasan: fix bad call to unpoison_slab_object

[PATCH] kasan: fix bad call to unpoison_slab_object

[andrey.konovalov]

From: Andrey Konovalov <andreyknvl@gmail.com>

Commit 29d7355a9d05 ("kasan: save alloc stack traces for mempool") messed
up one of the calls to unpoison_slab_object: the last two arguments are
supposed to be GFP flags and whether to init the object memory.

Fix the call.

Without this fix, unpoison_slab_object provides the object's size as
GFP flags to unpoison_slab_object, which can cause LOCKDEP reports
(and probably other issues).

Fixes: 29d7355a9d05 ("kasan: save alloc stack traces for mempool")
Reported-by: Brad Spengler <spender@grsecurity.net>
Signed-off-by: Andrey Konovalov <andreyknvl@gmail.com>
---
 mm/kasan/common.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/kasan/common.c b/mm/kasan/common.c
index e7c9a4dc89f8..85e7c6b4575c 100644
--- a/mm/kasan/common.c
+++ b/mm/kasan/common.c
@@ -532,7 +532,7 @@ void __kasan_mempool_unpoison_object(void *ptr, size_t size, unsigned long ip)
 		return;
 
 	/* Unpoison the object and save alloc info for non-kmalloc() allocations. */
-	unpoison_slab_object(slab->slab_cache, ptr, size, flags);
+	unpoison_slab_object(slab->slab_cache, ptr, flags, false);
 
 	/* Poison the redzone and save alloc info for kmalloc() allocations. */
 	if (is_kmalloc_cache(slab->slab_cache))
-- 
2.25.1

Re: [PATCH] kasan: fix bad call to unpoison_slab_object

[Marco Elver]

On Fri, 14 Jun 2024 at 16:16, <andrey.konovalov@linux.dev> wrote:
>
> From: Andrey Konovalov <andreyknvl@gmail.com>
>
> Commit 29d7355a9d05 ("kasan: save alloc stack traces for mempool") messed
> up one of the calls to unpoison_slab_object: the last two arguments are
> supposed to be GFP flags and whether to init the object memory.
>
> Fix the call.
>
> Without this fix, unpoison_slab_object provides the object's size as
> GFP flags to unpoison_slab_object, which can cause LOCKDEP reports
> (and probably other issues).
>
> Fixes: 29d7355a9d05 ("kasan: save alloc stack traces for mempool")
> Reported-by: Brad Spengler <spender@grsecurity.net>
> Signed-off-by: Andrey Konovalov <andreyknvl@gmail.com>

Oof.

Acked-by: Marco Elver <elver@google.com>

mm needs explicit Cc: stable, right? If so, we better add Cc: stable as well.

> ---
>  mm/kasan/common.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/kasan/common.c b/mm/kasan/common.c
> index e7c9a4dc89f8..85e7c6b4575c 100644
> --- a/mm/kasan/common.c
> +++ b/mm/kasan/common.c
> @@ -532,7 +532,7 @@ void __kasan_mempool_unpoison_object(void *ptr, size_t size, unsigned long ip)
>                 return;
>
>         /* Unpoison the object and save alloc info for non-kmalloc() allocations. */
> -       unpoison_slab_object(slab->slab_cache, ptr, size, flags);
> +       unpoison_slab_object(slab->slab_cache, ptr, flags, false);
>
>         /* Poison the redzone and save alloc info for kmalloc() allocations. */
>         if (is_kmalloc_cache(slab->slab_cache))
> --
> 2.25.1
>

Re: [PATCH] kasan: fix bad call to unpoison_slab_object

[Andrey Konovalov]

On Fri, Jun 14, 2024 at 4:29 PM Marco Elver <elver@google.com> wrote:
>
> On Fri, 14 Jun 2024 at 16:16, <andrey.konovalov@linux.dev> wrote:
> >
> > From: Andrey Konovalov <andreyknvl@gmail.com>
> >
> > Commit 29d7355a9d05 ("kasan: save alloc stack traces for mempool") messed
> > up one of the calls to unpoison_slab_object: the last two arguments are
> > supposed to be GFP flags and whether to init the object memory.
> >
> > Fix the call.
> >
> > Without this fix, unpoison_slab_object provides the object's size as
> > GFP flags to unpoison_slab_object, which can cause LOCKDEP reports
> > (and probably other issues).
> >
> > Fixes: 29d7355a9d05 ("kasan: save alloc stack traces for mempool")
> > Reported-by: Brad Spengler <spender@grsecurity.net>
> > Signed-off-by: Andrey Konovalov <andreyknvl@gmail.com>
>
> Oof.
>
> Acked-by: Marco Elver <elver@google.com>
>
> mm needs explicit Cc: stable, right? If so, we better add Cc: stable as well.

Makes sense, sent v2 with CC stable and a commit message fix.

Thanks!