lkml.org 
[lkml]   [2021]   [Dec]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: [PATCH] kasan: fix quarantine conflicting with init_on_free
On Mon, Dec 20, 2021 at 6:07 PM Marco Elver <elver@google.com> wrote:
>
> On Mon, 20 Dec 2021 at 17:37, <andrey.konovalov@linux.dev> wrote:
> >
> > From: Andrey Konovalov <andreyknvl@google.com>
> >
> > KASAN's quarantine might save its metadata inside freed objects. As
> > this happens after the memory is zeroed by the slab allocator when
> > init_on_free is enabled, the memory coming out of quarantine is not
> > properly zeroed.
> >
> > This causes lib/test_meminit.c tests to fail with Generic KASAN.
> >
> > Zero the metadata when the object is removed from quarantine.
> >
> > Fixes: 6471384af2a6 ("mm: security: introduce init_on_alloc=1 and init_on_free=1 boot options")
> > Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
> > ---
> > mm/kasan/quarantine.c | 11 +++++++++++
> > 1 file changed, 11 insertions(+)
> >
> > diff --git a/mm/kasan/quarantine.c b/mm/kasan/quarantine.c
> > index 587da8995f2d..2e50869fd8e2 100644
> > --- a/mm/kasan/quarantine.c
> > +++ b/mm/kasan/quarantine.c
> > @@ -132,11 +132,22 @@ static void *qlink_to_object(struct qlist_node *qlink, struct kmem_cache *cache)
> > static void qlink_free(struct qlist_node *qlink, struct kmem_cache *cache)
> > {
> > void *object = qlink_to_object(qlink, cache);
> > + struct kasan_free_meta *meta = kasan_get_free_meta(cache, object);
> > unsigned long flags;
> >
> > if (IS_ENABLED(CONFIG_SLAB))
> > local_irq_save(flags);
> >
> > + /*
> > + * If init_on_free is enabled and KASAN's free metadata is stored in
> > + * the object, zero the metadata. Otherwise, the object's memory will
> > + * not be properly zeroed, as KASAN saves the metadata after the slab
> > + * allocator zeroes the object.
> > + */
> > + if (slab_want_init_on_free(cache) &&
> > + cache->kasan_info.free_meta_offset == 0)
> > + memset(meta, 0, sizeof(*meta));
>
> memzero_explicit()
>
> although in this case it probably doesn't matter much, because AFAIK
> memzero_explicit() only exists to prevent the compiler from eliding
> the zeroing. Up to you.

I've thought about using memzero_explicit(), but the rest of
init_on_alloc/free code uses memset(0) so I decided to use it as well.
If we decide to switch to memzero_explicit(), it makes sense to do it
everywhere.

Thanks!

\
 
 \ /
  Last update: 2021-12-20 18:16    [W:2.051 / U:0.340 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site