[FFmpeg-devel] [PATCH 3/4] avcodec/pixlet: Avoid signed integer overflow in scaling in filterfn()

Michael Niedermayer michael at niedermayer.cc
Tue Oct 27 18:21:18 EET 2020


Fixes: signed integer overflow: 11494 * 1073741824000000 cannot be represented in type 'long'
Fixes: 26586/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_PIXLET_fuzzer-5752633970917376

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
---
 libavcodec/pixlet.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/libavcodec/pixlet.c b/libavcodec/pixlet.c
index 42acd683fa..fd5e2ef804 100644
--- a/libavcodec/pixlet.c
+++ b/libavcodec/pixlet.c
@@ -405,7 +405,7 @@ static void filterfn(int16_t *dest, int16_t *tmp, unsigned size, int64_t scale)
                 (int64_t) low [i - 1] * -INT64_C(325392907)  +
                 (int64_t) high[i + 0] *  INT64_C(1518500249) +
                 (int64_t) high[i - 1] *  INT64_C(1518500249);
-        dest[i * 2] = av_clip_int16(((value >> 32) * scale) >> 32);
+        dest[i * 2] = av_clip_int16(((value >> 32) * (uint64_t)scale) >> 32);
     }
 
     for (i = 0; i < hsize; i++) {
@@ -416,7 +416,7 @@ static void filterfn(int16_t *dest, int16_t *tmp, unsigned size, int64_t scale)
                 (int64_t) high[i + 1] *  INT64_C(303700064)  +
                 (int64_t) high[i + 0] * -INT64_C(3644400640) +
                 (int64_t) high[i - 1] *  INT64_C(303700064);
-        dest[i * 2 + 1] = av_clip_int16(((value >> 32) * scale) >> 32);
+        dest[i * 2 + 1] = av_clip_int16(((value >> 32) * (uint64_t)scale) >> 32);
     }
 }
 
-- 
2.17.1



More information about the ffmpeg-devel mailing list