msrle: Use FFABS to determine the frame size in msrle_decode_pal4

Message ID 20150530142933.500475DC8E@aruru.libav.org
State New
Headers show

Commit Message

Janne Grunau May 30, 2015, 2:29 p.m.
Module: libav
Branch: master
Commit: 5ecabd3c54b7c802522dc338838c9a4c2dc42948

Author:    Luca Barbato <lu_zero@gentoo.org>
Committer: Luca Barbato <lu_zero@gentoo.org>
Date:      Mon May 25 21:53:26 2015 +0200

msrle: Use FFABS to determine the frame size in msrle_decode_pal4

As done in msrle_decode_8_16_24_32.

Bug-Id: CVE-2015-3395
CC: libav-stable@libav.org

---

 libavcodec/msrledec.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Patch

diff --git a/libavcodec/msrledec.c b/libavcodec/msrledec.c
index af2a247..370d9bd 100644
--- a/libavcodec/msrledec.c
+++ b/libavcodec/msrledec.c
@@ -39,7 +39,7 @@  static int msrle_decode_pal4(AVCodecContext *avctx, AVPicture *pic,
     unsigned int pixel_ptr = 0;
     int row_dec = pic->linesize[0];
     int row_ptr = (avctx->height - 1) * row_dec;
-    int frame_size = row_dec * avctx->height;
+    int frame_size = FFABS(row_dec) * avctx->height;
     int i;
 
     while (row_ptr >= 0) {