Crash in [@ vk_optimusGetDeviceProcAddr]
Categories
(Core :: Audio/Video, defect, P2)
Tracking
()
People
(Reporter: gsvelto, Unassigned)
References
Details
(Keywords: crash, csectype-uaf, sec-high)
Crash Data
Crash report: https://crash-stats.mozilla.org/report/index/48a43ba8-9cfa-4af5-b0d5-0014b0260623
Reason:
SIGSEGV / SI_KERNEL
Top 10 frames:
0 libnvidia-glcore.so.610.43.02 vk_optimusGetDeviceProcAddr
1 libxul.so mozilla::FFmpegVideoDecoder<62>::FFmpegVulkanVideoDecoder::Cleanup() dom/media/platforms/ffmpeg/FFmpegVulkanVideoDecoder.cpp:72
2 libxul.so mozilla::FFmpegVideoDecoder<62>::ProcessShutdown() dom/media/platforms/ffmpeg/FFmpegVideoDecoder.cpp:2598
3 libxul.so mozilla::FFmpegDataDecoder<53>::Shutdown()::$_0::operator()() const dom/media/platforms/ffmpeg/FFmpegDataDecoder.cpp:294
3 libxul.so mozilla::detail::ProxyFunctionRunnable<mozilla::FFmpegDataDecoder<53>::Shutdo... xpcom/threads/MozPromise.h:1835
4 libxul.so mozilla::TaskQueue::Runner::Run() xpcom/threads/TaskQueue.cpp:297
5 libxul.so nsThreadPool::Run() xpcom/threads/nsThreadPool.cpp:442
6 libxul.so nsThread::ProcessNextEvent(bool, bool*) xpcom/threads/nsThread.cpp:1173
6 libxul.so NS_ProcessNextEvent(nsIThread*, bool) xpcom/threads/nsThreadUtils.cpp:472
7 libxul.so mozilla::ipc::MessagePumpForNonMainThreads::Run(base::MessagePump::Delegate*) ipc/glue/MessagePump.cpp:297
This is a UAF in Nvidia's proprietary drivers so unlikely to be actionable. There's something worth mentioning though: crash ping data shows that a handful of users are encountering this crash up to 100+ times per day, so it's probably some kind of crash loop on their machines.
Updated•4 months ago
|
Comment 2•4 months ago
|
||
This is crashing under FFmpegVideoDecoder::FFmpegVulkanVideoDecoder::Cleanup() like the spike in libradeon.so UAFs. Maybe it IS our bug, or a bug in the in-tree ffmpeg we're responsible for.
Comment 3•4 months ago
|
||
The UAF crashes under the Vulkcan Cleanup() were first reported on 2026-05-13 and the earliest buildID is 20260512212623. Given the various driver versions that strongly points to a regression on our side.
The vast majority of crashes are in 152 and 153 nightlies, but there are 15 in Release so it's not "nightly only" functionality.
Comment 4•4 months ago
|
||
This specific signature is a small fraction of the whole (<1%) and was first reported in 152.0b3, but it's 4 of the fifteen 152 Release crashes.
Updated•4 months ago
|
Comment 5•4 months ago
|
||
The severity field is not set for this bug.
:jimm, could you have a look please?
For more information, please visit BugBot documentation.
Updated•4 months ago
|
Updated•4 months ago
|
Updated•3 months ago
|
Updated•3 months ago
|
Updated•1 month ago
|
Description
•