| Message ID | 20221118210918.3169-1-michael@niedermayer.cc |
|---|---|
| State | New |
| Headers | show |
| Series | [FFmpeg-devel,1/4] avcodec/tiff: Ignore tile_count | expand |
| Context | Check | Description |
|---|---|---|
| andriy/make_x86 | success | Make finished |
| andriy/make_fate_x86 | success | Make fate finished |
Quoting Michael Niedermayer (2022-11-18 22:09:15) > Fixes: out of array access > Fixes: 52427/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_TIFF_fuzzer-4849108968144896 > > Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg > Signed-off-by: Michael Niedermayer <michael@niedermayer.cc> > --- > libavcodec/tiff.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/libavcodec/tiff.c b/libavcodec/tiff.c > index cde318d5e5..b962dce5a2 100644 > --- a/libavcodec/tiff.c > +++ b/libavcodec/tiff.c > @@ -994,7 +994,7 @@ static int dng_decode_tiles(AVCodecContext *avctx, AVFrame *frame, > tile_count_y = (s->height + s->tile_length - 1) / s->tile_length; > > /* Iterate over the number of tiles */ > - for (tile_idx = 0; tile_idx < s->tile_count; tile_idx++) { > + for (tile_idx = 0; tile_idx < tile_count_x * tile_count_y; tile_idx++) { Then why store tile_count at all? Seems to be it will just confuse future readers. If I'm reading the code correctly, the only other use of tile_count is in setting has_tile_bits, but that also checks is_tiled, which is set together with tile_count.
On Sat, Nov 19, 2022 at 01:40:06PM +0100, Anton Khirnov wrote: > Quoting Michael Niedermayer (2022-11-18 22:09:15) > > Fixes: out of array access > > Fixes: 52427/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_TIFF_fuzzer-4849108968144896 > > > > Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg > > Signed-off-by: Michael Niedermayer <michael@niedermayer.cc> > > --- > > libavcodec/tiff.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/libavcodec/tiff.c b/libavcodec/tiff.c > > index cde318d5e5..b962dce5a2 100644 > > --- a/libavcodec/tiff.c > > +++ b/libavcodec/tiff.c > > @@ -994,7 +994,7 @@ static int dng_decode_tiles(AVCodecContext *avctx, AVFrame *frame, > > tile_count_y = (s->height + s->tile_length - 1) / s->tile_length; > > > > /* Iterate over the number of tiles */ > > - for (tile_idx = 0; tile_idx < s->tile_count; tile_idx++) { > > + for (tile_idx = 0; tile_idx < tile_count_x * tile_count_y; tile_idx++) { > > Then why store tile_count at all? Seems to be it will just confuse > future readers. > > If I'm reading the code correctly, the only other use of tile_count is > in setting has_tile_bits, but that also checks is_tiled, which is set > together with tile_count. i tend to make minimal changes with the bugfixes but yes you are correct the field seems useless i will remove it in that patch thx [...]
diff --git a/libavcodec/tiff.c b/libavcodec/tiff.c index cde318d5e5..b962dce5a2 100644 --- a/libavcodec/tiff.c +++ b/libavcodec/tiff.c @@ -994,7 +994,7 @@ static int dng_decode_tiles(AVCodecContext *avctx, AVFrame *frame, tile_count_y = (s->height + s->tile_length - 1) / s->tile_length; /* Iterate over the number of tiles */ - for (tile_idx = 0; tile_idx < s->tile_count; tile_idx++) { + for (tile_idx = 0; tile_idx < tile_count_x * tile_count_y; tile_idx++) { tile_x = tile_idx % tile_count_x; tile_y = tile_idx / tile_count_x;
Fixes: out of array access Fixes: 52427/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_TIFF_fuzzer-4849108968144896 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc> --- libavcodec/tiff.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)