
Signal is active with 1 mentions in latest observed window
Recommended action window: Monitor and triage in normal cycle
NVD description
It is possible to cause an use-after-free write in SANM decoding with a carefully crafted animation using subversion <2. When a STOR chunk is present, a subsequent FOBJ chunk will be saved in ctx->stored_frame. Stored frames can later be referenced by FTCH chunks. For files using subversion < 2, the undecoded frame is stored, and decoded again when the FTCH chunks are parsed. However, in process_frame_obj if the frame has an invalid size, there’s an early return, with a value of 0. This causes the code in decode_frame to still store the raw frame buffer into ctx->stored_frame. Leaving ctx->has_dimensions set to false. A subsequent chunk with type FTCH would call process_ftch and decode that frame obj again, adding to the top/left values and calling process_frame_obj again. Given that we never set ctx->have_dimensions before, this time we set the dimensions, calling init_buffers, which can reallocate the buffer in ctx->stored_frame, freeing the previous one. However, the GetByteContext object gb still holds a reference to the old buffer. Finally, when the code tries to decode the frame, codecs that accept a GetByteContext as a parameter will trigger a use-after-free read when using gb. GetByteContext is only used for reading bytes, so at most one could read invalid data. There are no heap allocations between the free and when the object is accessed. However, upon returning to process_ftch, the code restores the original values for top/left in stored_frame, writing 4 bytes to the freed data at offset 6, potentially corrupting the allocator’s metadata. This issue can be triggered just by probing whether a file has the sanm format. We recommend upgrading to version 8.0 or beyond.
Priority
LOW
Exploitation
NONE
PoC
NONE
Patch
NONE
Momentum
NONE

1) Google has also funded and built tools that produce fixes 2) The FFmpeg dispute was in 2025. Google changed its disclosure policy on November 13, 2026. 3) Google’s Big Sleep report for CVE-2025-59734 documented a use-after-free write in FFmpeg’s SANM decoder. It included a testcase generator, build and execution instructions, and an AddressSanitizer trace. FFmpeg’s team fixed it in 48 hours. 4) Marcus is spot on about the high level point though. Automated discovery can create more remediation work than volunteer maintainers can absorb. That’s a fact. Once an open source project becomes critical infrastructure the big companies using it need to start throwing down money to support the team or suffer the consequences.