On Wed, Feb 09, 2011 at 04:34:17PM -0000, Brian Murray wrote:
> It occurred to me that glxinfo is from a package in universe and I was
> able to recreate the crash using the following steps:
>
> 8:28:59 - flash:[~/Documents/subtitles] sudo chmod -x '/usr/bin/glxinfo'
> [sudo] password for bdmurray:
> 8:29:10 - flash:[~/Documents/subtitles] ubuntu-bug xorg
> ERROR: hook /usr/share/apport/package-hooks//source_xorg.py crashed:
> Traceback (most recent call last):
> File "/usr/lib/python2.7/dist-packages/apport/report.py", line 637, in add_hooks_info
> symb['add_info'](self, ui)
> File "/usr/share/apport/package-hooks//source_xorg.py", line 450, in add_info
> attach_3d_info(report, ui)
> File "/usr/share/apport/package-hooks//source_xorg.py", line 332, in attach_3d_info
> report['glxinfo'] = command_output_quiet(['glxinfo'])
> File "/usr/lib/python2.7/dist-packages/problem_report.py", line 510, in __setitem__
> and (hasattr(v[0], 'isalnum') or hasattr(v[0], 'read'))))
> AssertionError
>
> The existence of glxinfo should be tested before trying to run it.
Yeah, it used to be that glxinfo was installed by default along with the
rest of mesa, but a release or two ago the mesa utilities including this
one got broken out into a separate package, which we don't install by
default.
However, historically glxinfo has provided minimal usefulness; we don't
really seem to miss it when it's not attached. I think we've mostly
moved beyond the class of problems that glxinfo used to help with,
especially now that we have the more precise instrumentation from the
compiz and unity folks.
So, I'm thinking we should just drop the call entirely. I'll
doublecheck with Jay and David first though.
On Wed, Feb 09, 2011 at 04:34:17PM -0000, Brian Murray wrote: ~/Documents/ subtitles] sudo chmod -x '/usr/bin/glxinfo' ~/Documents/ subtitles] ubuntu-bug xorg apport/ package- hooks// source_ xorg.py crashed: python2. 7/dist- packages/ apport/ report. py", line 637, in add_hooks_info info']( self, ui) apport/ package- hooks// source_ xorg.py" , line 450, in add_info 3d_info( report, ui) apport/ package- hooks// source_ xorg.py" , line 332, in attach_3d_info output_ quiet([ 'glxinfo' ]) python2. 7/dist- packages/ problem_ report. py", line 510, in __setitem__
> It occurred to me that glxinfo is from a package in universe and I was
> able to recreate the crash using the following steps:
>
> 8:28:59 - flash:[
> [sudo] password for bdmurray:
> 8:29:10 - flash:[
> ERROR: hook /usr/share/
> Traceback (most recent call last):
> File "/usr/lib/
> symb['add_
> File "/usr/share/
> attach_
> File "/usr/share/
> report['glxinfo'] = command_
> File "/usr/lib/
> and (hasattr(v[0], 'isalnum') or hasattr(v[0], 'read'))))
> AssertionError
>
> The existence of glxinfo should be tested before trying to run it.
Yeah, it used to be that glxinfo was installed by default along with the
rest of mesa, but a release or two ago the mesa utilities including this
one got broken out into a separate package, which we don't install by
default.
However, historically glxinfo has provided minimal usefulness; we don't
really seem to miss it when it's not attached. I think we've mostly
moved beyond the class of problems that glxinfo used to help with,
especially now that we have the more precise instrumentation from the
compiz and unity folks.
So, I'm thinking we should just drop the call entirely. I'll
doublecheck with Jay and David first though.
Bryce