[Intel-gfx] intel_gpu_top profiles

[Eric Anholt]

On Fri, 2009-10-30 at 12:29 +0000, Peter Clifton wrote:
> Hi guys..
> 
> intel_gpu_top seemed like it has been a well kept secret.. I discovered
> it today, and took some profiles of the circuit board layout package I'm
> working on.
> 
> Can anyone suggest whether there appears to be "room for improvement"? I
> couldn't spot any obvious bottle-neck. I'm surprised that the pixel
> shader is seeing any use, (even without compiz), as I'm just using
> boring fixed-function stuff with no texturing, no lighting, nothing
> fancy..

I'd been meaning to type this up for a while, and your mail convinced me
to do so.  Here's a new wiki page:

http://dri.freedesktop.org/wiki/IntelPerformanceTuning

Does this help?  Are there parts that are unclear and confusing?

> Screen-captures of the profiles available here:
> 
> http://www2.eng.cam.ac.uk/~pcjc2/geda/intel_gpu_top_profiles/
> http://www2.eng.cam.ac.uk/~pcjc2/geda/intel_gpu_top_profiles/intel_gpu_load_displaylist_no_compiz_PCB.png
> http://www2.eng.cam.ac.uk/~pcjc2/geda/intel_gpu_top_profiles/intel_gpu_top_displaylist_compiz_PCB.png
> http://www2.eng.cam.ac.uk/~pcjc2/geda/intel_gpu_top_profiles/intel_gpu_top_no_displaylist_compiz_PCB.png
> http://www2.eng.cam.ac.uk/~pcjc2/geda/intel_gpu_top_profiles/intel_gpu_top_no_displaylist_no_compiz_PCB.png

OK, so this shows that the GPU is almost entirely idle, time to pull out
perf and see what's going wrong.

> The "displaylist" versions are where I get the rendering code to build
> all its rendering into a single displaylist, then just draw the
> display-list from the expose event handler. (Prodded repeatedly by our
> "benchmark" action).
> 
> The "no_displaylist" versions show what happens with our geometry
> generation overheads in place - obviously there is room for improvement
> in that case!

If you're using VBOs, no_displaylist should beat displaylist.

Otherwise, what a display list performance win probably means is that
you're getting VBOs made for you.

> What is "MASM CS CR" and "CL CS"? Those seem to vary in the compiz /
> non-compiz case. Is there a guide with these names somewhere (other than
> the GPU PRM?)

If we know what the names mean, we should improve the names in the
tool :)  Those came straight out of the docs.

-- 
Eric Anholt
eric at anholt.net                         eric.anholt at intel.com


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20091030/ad0080e0/attachment.sig>