Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Rasmus Villemoes <linux@rasmusvillemoes.dk>
To:     Andrew Morton <akpm@linux-foundation.org>,
        Alexander Lobakin <alobakin@pm.me>,
        Mel Gorman <mgorman@techsingularity.net>,
        Vlastimil Babka <vbabka@suse.cz>
Cc:     Rasmus Villemoes <linux@rasmusvillemoes.dk>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org
Subject: [PATCH] mm/page_alloc: __alloc_pages_bulk(): do bounds check before accessing array
Date:   Fri,  7 May 2021 08:45:03 +0200
Message-Id: <20210507064504.1712559-1-linux@rasmusvillemoes.dk>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

In the event that somebody would call this with an already fully
populated page_array, the last loop iteration would do an access
beyond the end of page_array.

It's of course extremely unlikely that would ever be done, but this
triggers my internal static analyzer. Also, if it really is not
supposed to be invoked this way (i.e., with no NULL entries in
page_array), the nr_populated<nr_pages check could simply be removed
instead.

Fixes: 0f87d9d30f21 (mm/page_alloc: add an array-based interface to the bulk page allocator)
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
---
 mm/page_alloc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index bcdc0c6f21f1..66785946eb28 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -5053,7 +5053,7 @@ unsigned long __alloc_pages_bulk(gfp_t gfp, int preferred_nid,
 	 * Skip populated array elements to determine if any pages need
 	 * to be allocated before disabling IRQs.
 	 */
-	while (page_array && page_array[nr_populated] && nr_populated < nr_pages)
+	while (page_array && nr_populated < nr_pages && page_array[nr_populated])
 		nr_populated++;
 
 	/* Use the single page allocator for one page. */
-- 
2.29.2