Crash during "make check"

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Crash during "make check"

mmuetzel
Octave (hg id 6310bb807752) crashed during "make check" for which I managed to save core dump.

I am attaching gdb's output for "thread apply all bt".

IIUC, all threads for which there are symbols were waiting when Octave crashed. One thread seems to execute functions in libgomp. Some other threads seem to be JVM related.
To be honest, I am not sure what to make out of this. And I don't know if this is any helpful.

Any hints on what to do with this would be appreciated.

Markus

octave_crash_bt.log (73K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Crash during "make check"

mmuetzel
Am 18. Juli 2020 um 15:13 Uhr schrieb "Markus M├╝tzel":
> Octave (hg id 6310bb807752) crashed during "make check" for which I managed to save core dump.
>
> I am attaching gdb's output for "thread apply all bt".
>
> IIUC, all threads for which there are symbols were waiting when Octave crashed. One thread seems to execute functions in libgomp. Some other threads seem to be JVM related.
> To be honest, I am not sure what to make out of this. And I don't know if this is any helpful.
>
> Any hints on what to do with this would be appreciated.
>

Forgot to mention the reason for the crash according to gdb:
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x00007fb514f7a82c in ?? ()

That is Thread 1 (a JVM thread?) in the backtrace of the previous message.

And the last lines appearing in the shell running "make check" were:
  specfun/perms.m ................................................ pass   30/30  
  specfun/pow2.m .................................................fatal: caught signal Segmentation fault -- stopping myself...
/bin/bash: line 1: 27619 Segmentation fault      (core dumped) /bin/bash ../run-octave --no-init-file --silent --no-history -p /home/osboxes/Documents/Repositories/Octave/octave-1/.build/test/mex /home/osboxes/Documents/Repositories/Octave/octave-1/.build/../test/fntests.m /home/osboxes/Documents/Repositories/Octave/octave-1/.build/../test
make[3]: *** [Makefile:31674: check-local] Error 139
make[3]: Leaving directory '/home/osboxes/Documents/Repositories/Octave/octave-1/.build'
make[2]: *** [Makefile:27804: check-am] Error 2
make[2]: Leaving directory '/home/osboxes/Documents/Repositories/Octave/octave-1/.build'
make[1]: *** [Makefile:27512: check-recursive] Error 1
make[1]: Leaving directory '/home/osboxes/Documents/Repositories/Octave/octave-1/.build'
make: *** [Makefile:27806: check] Error 2


I'm wondering why there were so many threads that apparently belong to the JVM at that point of the test suite. (I am counting 17.)
Are they hanging around until Octave closes once they were started? Is that expected? Or are we doing something (wrong) that prevents them from ever finishing.

Markus