Diagnosing first vs subsequent performance

[Ilya Anfimov]

On Tue, Jan 19, 2016 at 12:54:15PM -0700, Lloyd Brown wrote:
> Sure.  I've generated them using something like this:
> 
> > $ for i in {0..3}; do echo "Testing display $i"; for j in {1..2}; do
> > echo "Instance $j"; DISPLAY=:${i}.0 glxinfo >
> > /tmp/glxinfo.display${i}.instance${j}; echo "sleeping 30s"; sleep 30;
> > echo "---"; done; echo "sleeping 5s"; sleep 5; echo "------"; done
> 
> Note that I did add the delays on purpose.  For some reason, when I ran
> the loop without them, it would hang up the second Xorg display
> (DISPLAY=:1.0, in this case), when it got to the second glxinfo.  I
> don't entirely know what was going on, but the delays seemed to help it
> complete.
> 
> I admit that I'm curious to know what you're looking for in the attached
> documents, but if you're trying to verify that the first and second
> calls to glxinfo are the same, they certainly are.  Here are the MD5SUMs:

 I expected that, probably, server would
fail to initialize NV-GLX direct rendering extension and this would
result in indirect glx rendering.
 
 My expectations was wrong. Probably.

> 
> > $ md5sum glxinfo.display*
> > 97379d9a7c3332e660ec71a045d6ea1f  glxinfo.display0.instance1
> > 97379d9a7c3332e660ec71a045d6ea1f  glxinfo.display0.instance2
> > d6484c73ca120a43f4b1abc5b508c168  glxinfo.display1.instance1
> > d6484c73ca120a43f4b1abc5b508c168  glxinfo.display1.instance2
> > ba989769089853b4cd44bb6c68a53a80  glxinfo.display2.instance1
> > ba989769089853b4cd44bb6c68a53a80  glxinfo.display2.instance2
> > 11be7dc4379f37cbeeef346df799014b  glxinfo.display3.instance1
> > 11be7dc4379f37cbeeef346df799014b  glxinfo.display3.instance2
> 
> Note that I did upgrade to the RHEL xorg-x11-server-Xorg RPM to
> v1.15.0-36.  It did not have any effect; the same behavior occurred
> either way.
> 
> 
> Lloyd
> 
> 
> On 01/19/2016 12:05 PM, Ilya Anfimov wrote:
> > Good day. Could you provide glxinfo output, first and second run on
> > different displays?
> 
> -- 
> Lloyd Brown
> Systems Administrator
> Fulton Supercomputing Lab
> Brigham Young University
> http://marylou.byu.edu
> 


> _______________________________________________
> xorg at lists.x.org: X.Org support
> Archives: http://lists.freedesktop.org/archives/xorg
> Info: http://lists.x.org/mailman/listinfo/xorg
> Your subscription address: %(user_address)s