[FFmpeg-devel] [PATCH 5/9] avformat/mov: Check extend and base offset

[Rémi Denis-Courmont]

Le 17 juin 2024 01:08:27 GMT+02:00, Michael Niedermayer <michael at niedermayer.cc> a écrit :
>Fixes: signed integer overflow: 2314885530818453536 + 9151314442816847872 cannot be represented in type 'long'
>Fixes: 68359/clusterfuzz-testcase-minimized-ffmpeg_dem_MOV_fuzzer-6571950311800832
>
>Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
>Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
>---
> libavformat/mov.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
>diff --git a/libavformat/mov.c b/libavformat/mov.c
>index 9016cd5ad08..46cbce98040 100644
>--- a/libavformat/mov.c
>+++ b/libavformat/mov.c
>@@ -8131,7 +8131,9 @@ static int mov_read_iloc(MOVContext *c, AVIOContext *pb, MOVAtom atom)
>         }
>         for (int j = 0; j < extent_count; j++) {
>             if (rb_size(pb, &extent_offset, offset_size) < 0 ||
>-                rb_size(pb, &extent_length, length_size) < 0)
>+                rb_size(pb, &extent_length, length_size) < 0 ||
>+                base_offset < 0 || extent_offset < 0 ||
>+                base_offset + (uint64_t)extent_offset > INT64_MAX)

Can we please stop with the bespoke arithmetic overflow checks and add dedicated helpers instead, similar to what GCC and C23 have?

>                 return AVERROR_INVALIDDATA;
>             if (offset_type == 1)
>                 c->heif_item[i].is_idat_relative = 1;