Crash in [@ js::MicroTaskQueueSet::empty]
Categories
(Core :: Machine Learning: On Device, defect)
Tracking
()
People
(Reporter: afranchuk, Unassigned)
References
Details
(Keywords: crash)
Crash Data
Crash report: https://crash-stats.mozilla.org/report/index/2946356c-dd7c-49bd-ba5a-3bb270260219
Reason:
SIGSEGV / SEGV_MAPERR
Top 10 frames:
0 libxul.so js::MicroTaskQueueSet::empty() /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/js/src/vm/JSContext.h:199
0 libxul.so JS::HasAnyMicroTasks(JSContext*) /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/js/src/vm/JSContext.cpp:1153
0 libxul.so mozilla::CycleCollectedJSContext::PerformMicroTaskCheckPoint(bool) /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/xpcom/base/CycleCollectedJSContext.cpp:1155
1 libxul.so mozilla::CycleCollectedJSContext::BeforeProcessTask(bool) /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/xpcom/base/CycleCollectedJSContext.cpp:725
2 libxul.so nsThread::ProcessNextEvent(bool, bool*) /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/xpcom/threads/nsThread.cpp:1096
3 libxul.so NS_ProcessNextEvent(nsIThread*, bool) /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/xpcom/threads/nsThreadUtils.cpp:461
4 libxul.so SpinEventLoopUntil<> /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/xpcom/threads/SpinEventLoopUntil.h:176
4 libxul.so nsThread::Shutdown() /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/xpcom/threads/nsThread.cpp:908
5 libxul.so mozilla::dom::LlamaStreamSource::~LlamaStreamSource() /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/toolkit/components/ml/backends/llama/LlamaRunner.cpp:473
6 libxul.so mozilla::dom::LlamaStreamSource::~LlamaStreamSource() /builddir/build/BUILD/firefox-147.0.2-build/firefox-147.0.2/toolkit/components/ml/backends/llama/LlamaRunner.cpp:476
This is a new crash in 147, which I'll call a "near-top-crasher" for macos content processes in release (ranked 11th in crash pings): https://crash-pings.mozilla.org/#a0RUdL.
This is far underrepresented in crash reports: there's only 1 report!
Of note is that this is only occurring on x86_64, not arm64 macs.
This may be contributing to an overall increasing trend of crashes on macOS that we're seeing.
Comment 1•6 months ago
|
||
This crash was addressed in Bug 1996263 which wasn't uplifted -- moving to appropriate component for retriage tho.
Comment 3•6 months ago
|
||
Copying crash signatures from duplicate bugs.
Comment 4•6 months ago
|
||
We're in rc week, isn't it too late?
| Reporter | ||
Comment 5•6 months ago
|
||
:padenot, IIUC, this is a dup of bug 1996263 (the reason it was not seen was that there was no associated crash signature on that bug).
Comment 6•6 months ago
|
||
It's too late for Fx148, but I added a comment on Bug 1996263 about a release uplift request for a dot release.
| Reporter | ||
Updated•4 months ago
|
Description
•