[cairo] Cairo performance testing with X11 forwarding

[Patrick Wilbur]

I have been manually (as in, not using * on the command-line) testing
traces in the benchmark/ directory, one-by-one, to see what I can observe.

For some of the tests (including swfdec-youtube.trace,
firefox-scrolling.trace, gnome-system-monitor.trace, and
chromium-tabs.trace) everything works like a charm when run over ssh -X.
 These I can use as several benchmarks, which is great.

For others, I see a spike in CPU usage followed by CPU usage dropping out
(back to my normal idle usage).  At this point, the test hangs and
cairo-perf-trace displays no further output (than up to the test name just
before printing values).

Seems quite strange, no?  While I feel the traces that are working make
good benchmarks, it would be nice to be able to get all traces working.

(I still have yet to upgrade the underlying host system from 11.10 to
12.04, which would match the VM's OS, but I'm not certain that would make a
difference anyway.)

Pat

--
Patrick F. Wilbur
Researcher, Consultant, Educator,
Computer Science Graduate at Clarkson University

patrick.wilbur at gmail.com
wilburpf at clarkson.edu

Check out our book: http://runningxen.com
My website: http://pdub.net
Consulting: http://pfwilbur.com





On Wed, Jun 13, 2012 at 1:00 PM, Chris Wilson <chris at chris-wilson.co.uk>wrote:

> On Tue, 12 Jun 2012 09:44:04 -0400, Patrick Wilbur <
> patrick.wilbur at gmail.com> wrote:
> > First, thank you very much for your reply.
> >
> > Hmm. I am using ssh -X to access a KVM virtual machine from the host
> > system. Could Ubuntu 11.10 Desktop (my host system) have a different
> > library version than Ubuntu 12.04 Server (my VM)? My VM also doesn't
> have X
> > installed, but it has various libraries installed due to installing
> several
> > GUI apps--if lacking a library, would that affect anything and would the
> > benchmarks fail silently?
>
> The benchmarks won't run at all unless all of its prequisites are
> installed, so it would seem like that isn't the issue. Look at what is
> happening during the runs to see whether the test is CPU bound or if the
> communication has been disrupted. Hopefully, it is CPU bound as then
> finding the offending bottleneck will be a lot easier; otherwise there
> is a bug in the stack causing the requests to go missing.
> -Chris
>
> --
> Chris Wilson, Intel Open Source Technology Centre
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cairographics.org/archives/cairo/attachments/20120616/83f7a623/attachment.html>