search

matcatc / Test_Parser

parses and displays the output from test logs
http://matcatc.github.com/Test_Parser
GNU General Public License v3.0
1 stars 0 forks source link

rerun auto-expand: seg fault #81

Closed matcatc closed 14 years ago

matcatc commented 14 years ago

The introduction of auto-expand has led to an intermittent seg-fault.

It has something to do with PyQt (libqt.so) according to gdb backtrace.

Question is, what do we do? Its not exactly our fault, but we don't want to release code that seg-faults (see this also.)

matcatc commented 14 years ago

It is possible that rerun doesn't actually introduce the problem, but only exacerbates it enough that it becomes visible or leads to a seg-fault. Ie: rerun could stress PyQt in just the right way, that it leads to a problem.

matcatc commented 14 years ago

Try isolating what function is causing the segfault.

matcatc commented 14 years ago

It appears to occur after the end of QtView.reRun(), which is in code that we don't have any control over. It is possible that its in code generated by MainWindow.ui, but based on the stack trace mentioned above, I think that is even unlikely.

Try updating PyQt to most recent version.

matcatc commented 14 years ago

Updated to most recent version. Now /usr/lib/python3.1/dist-packages/sip.so is causing a segfault, so the program won't even start.

I'm considering dropping PyQt based development for the moment.

matcatc commented 14 years ago

Seg fault might be due to the threading. After disabling threading, (this problem)[] when away. Also right after reenabling threading, I had a seg-fault. But I was able to rerun MANY times without a seg-fault when threading was disabled.

Document and close

matcatc commented 14 years ago

added ViewFactory and threading information

closed by eb95cbf86ae6c1d2973b6abcdd61f7dcec403adf closed by eb95cbf86ae6c1d2973b6abcdd61f7dcec403adf closed by eb95cbf86ae6c1d2973b6abcdd61f7dcec403adf