[FFmpeg-devel] [PATCH 2/3] avcodec/agm: Check for too many too short codes in make_new_tree()
Paul B Mahol
onemda at gmail.com
Sun Apr 21 12:36:00 EEST 2019
On 4/21/19, Michael Niedermayer <michael at niedermayer.cc> wrote:
> Fixes: SEGV on unknown address
> Fixes:
> 14198/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_AGM_fuzzer-5723579234123776
>
> untested due to lack of valid samples using this codepath
>
> 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/agm.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/libavcodec/agm.c b/libavcodec/agm.c
> index f3d81bf163..c9d7be5521 100644
> --- a/libavcodec/agm.c
> +++ b/libavcodec/agm.c
> @@ -919,7 +919,7 @@ static void get_tree_codes(uint32_t *codes, Node *nodes,
> int idx, uint32_t pfx,
> }
> }
>
> -static void make_new_tree(const uint8_t *bitlens, uint32_t *codes)
> +static int make_new_tree(const uint8_t *bitlens, uint32_t *codes)
> {
> int zlcount = 0, curlen, idx, nindex, last, llast;
> int blcounts[32] = { 0 };
> @@ -959,6 +959,9 @@ static void make_new_tree(const uint8_t *bitlens,
> uint32_t *codes)
> int p = node_idx[nindex - 1 + 512];
> int ch = syms[256 * curlen + i];
>
> + if (nindex <= 0)
> + return AVERROR_INVALIDDATA;
> +
> if (nodes[p].child[0] == -1) {
> nodes[p].child[0] = ch;
> } else {
> @@ -998,6 +1001,7 @@ static void make_new_tree(const uint8_t *bitlens,
> uint32_t *codes)
> next:
>
> get_tree_codes(codes, nodes, 256, 0, 0);
> + return 0;
> }
>
> static int build_huff(const uint8_t *bitlen, VLC *vlc)
> @@ -1008,7 +1012,9 @@ static int build_huff(const uint8_t *bitlen, VLC *vlc)
> uint32_t codes[256];
> int nb_codes = 0;
>
> - make_new_tree(bitlen, new_codes);
> + int ret = make_new_tree(bitlen, new_codes);
> + if (ret < 0)
> + return ret;
>
> for (int i = 0; i < 256; i++) {
> if (bitlen[i]) {
> --
> 2.21.0
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".
LGTM
More information about the ffmpeg-devel
mailing list