Crash in [@ _chkstk | profiler_suspend_and_sample_thread]
Categories
(Core :: Gecko Profiler, defect, P3)
Tracking
()
People
(Reporter: gsvelto, Unassigned)
Details
(Keywords: crash)
Crash Data
Crash report: https://crash-stats.mozilla.org/report/index/26df3f04-f2b6-4f61-b37d-ec9fa0210430
Reason: EXCEPTION_STACK_OVERFLOW
Top 10 frames of crashing thread:
0 xul.dll _chkstk
1 xul.dll profiler_suspend_and_sample_thread tools/profiler/core/platform.cpp:5910
2 xul.dll mozilla::ThreadStackHelper::GetStack toolkit/components/backgroundhangmonitor/ThreadStackHelper.cpp:145
3 xul.dll static mozilla::BackgroundHangManager::MonitorThread toolkit/components/backgroundhangmonitor/BackgroundHangMonitor.cpp:80
4 nss3.dll _PR_NativeRunThread nsprpub/pr/src/threads/combined/pruthr.c:399
5 nss3.dll pr_root nsprpub/pr/src/md/windows/w95thred.c:139
6 ucrtbase.dll thread_start<unsigned int , 1>
7 kernel32.dll BaseThreadInitThunk
8 ntdll.dll RtlUserThreadStart
9 kernelbase.dll TerminateProcessOnMemoryExhaustion
This is a low-volume stack overflow crash happening on Windows. It seems to be caused by an interaction between the background hang manager and the profiler though I'm unsure what's going on. The stack of the crashing thread isn't particularly large and we don't seem to be allocating a large object where the crash occurs but maybe I'm missing something.
It does seem strange... I would expect more stack overflows deeper in the profiler and stack-walker code, but profiler_suspend_and_sample_thread itself shouldn't take much stack space.
Florian, any ideas on the BHR side? (Please redirect NI as appropriate.)
|
||
Comment 2•3 years ago
|
||
(In reply to Gerald Squelart [:gerald] (he/him) from comment #1)
Florian, any ideas on the BHR side? (Please redirect NI as appropriate.)
Unfortunately, no idea. Doug might have some.
|
||
Comment 4•2 years ago
|
||
Closing because no crashes reported for 12 weeks.
Description
•