mm, slab: initialize object alignment on cache creation
authorDavid Rientjes <rientjes@google.com>
Thu, 25 Sep 2014 23:05:20 +0000 (16:05 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 26 Sep 2014 15:10:35 +0000 (08:10 -0700)
Since commit 4590685546a3 ("mm/sl[aou]b: Common alignment code"), the
"ralign" automatic variable in __kmem_cache_create() may be used as
uninitialized.

The proper alignment defaults to BYTES_PER_WORD and can be overridden by
SLAB_RED_ZONE or the alignment specified by the caller.

This fixes https://bugzilla.kernel.org/show_bug.cgi?id=85031

Signed-off-by: David Rientjes <rientjes@google.com>
Reported-by: Andrei Elovikov <a.elovikov@gmail.com>
Acked-by: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/slab.c

index a467b308c682334254c53fdf12a6114aad1ce1fd..b8b619bc84adab2535dd81e170030171fcfab1c4 100644 (file)
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -2124,7 +2124,8 @@ static int __init_refok setup_cpu_cache(struct kmem_cache *cachep, gfp_t gfp)
 int
 __kmem_cache_create (struct kmem_cache *cachep, unsigned long flags)
 {
-       size_t left_over, freelist_size, ralign;
+       size_t left_over, freelist_size;
+       size_t ralign = BYTES_PER_WORD;
        gfp_t gfp;
        int err;
        size_t size = cachep->size;
@@ -2157,14 +2158,6 @@ __kmem_cache_create (struct kmem_cache *cachep, unsigned long flags)
                size &= ~(BYTES_PER_WORD - 1);
        }
 
-       /*
-        * Redzoning and user store require word alignment or possibly larger.
-        * Note this will be overridden by architecture or caller mandated
-        * alignment if either is greater than BYTES_PER_WORD.
-        */
-       if (flags & SLAB_STORE_USER)
-               ralign = BYTES_PER_WORD;
-
        if (flags & SLAB_RED_ZONE) {
                ralign = REDZONE_ALIGN;
                /* If redzoning, ensure that the second redzone is suitably