<div class="gmail_quote">On Fri, Nov 4, 2011 at 14:31, Chris Wilson <span dir="ltr"><<a href="mailto:chris@chris-wilson.co.uk">chris@chris-wilson.co.uk</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im">On Fri, 4 Nov 2011 14:23:19 -0200, Eugeni Dodonov <<a href="mailto:eugeni@dodonov.net">eugeni@dodonov.net</a>> wrote:<br>
> Yep, it works for me. I can disable and enable rings at will by echo [01] ><br>
> /sys/kernel/debug/dri/0/i915_ring_stop<br>
><br>
> I was using it for working further on that i915_error_state heuristics<br>
> idea, outside of intel_decode. It is *very* handy to discover specific<br>
> patterns in error_state's.<br>
<br>
</div>intel_gpu_dump or perchance reading the code ;-)<br></blockquote><div><br>I mean, I am using it in a combination with the intel_gpu_dump, based on the original idea which came from <a href="http://comments.gmane.org/gmane.comp.freedesktop.xorg.drivers.intel/6408">http://comments.gmane.org/gmane.comp.freedesktop.xorg.drivers.intel/6408</a>, only outside of the real intel_decode routine.<br>
<br>But yes, reading the code is the best way, the overall idea here is to at least indicate which code to read in the first place (kernel or ddx or mesa or va or....) :).<br><br></div></div>-- <br>Eugeni Dodonov<a href="http://eugeni.dodonov.net/" target="_blank"><br>
</a><br>