avcodec/qpeg: Optimize long runs in qpeg_decode_intra() not spanning a full row
authorMichael Niedermayer <michael@niedermayer.cc>
Mon, 7 Jan 2019 01:44:50 +0000 (02:44 +0100)
committerMichael Niedermayer <michael@niedermayer.cc>
Mon, 7 Jan 2019 22:21:47 +0000 (23:21 +0100)
Fixes: Timeout
Fixes: 11354/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_QPEG_fuzzer-5766275943366656

Before: Executed clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_QPEG_fuzzer-5766275943366656 in 9470 ms
After : Executed clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_QPEG_fuzzer-5766275943366656 in  134 ms

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
libavcodec/qpeg.c

index cb45262..654fd99 100644 (file)
@@ -80,7 +80,10 @@ static void qpeg_decode_intra(QpegContext *qctx, uint8_t *dst,
 
             p = bytestream2_get_byte(&qctx->buffer);
             for(i = 0; i < run; i++) {
-                dst[filled++] = p;
+                int step = FFMIN(run - i, width - filled);
+                memset(dst+filled, p, step);
+                filled += step;
+                i      += step - 1;
                 if (filled >= width) {
                     filled = 0;
                     dst -= stride;