[Intel-gfx] [PATCH v1 0/2] disable rings before sync flush

Chris Wilson chris at chris-wilson.co.uk
Wed Mar 12 09:12:05 CET 2014


On Wed, Mar 12, 2014 at 08:06:26AM +0000, Chris Wilson wrote:
> On Wed, Mar 12, 2014 at 12:20:09PM +0530, naresh.kumar.kachhi at intel.com wrote:
> > From: Naresh Kumar Kachhi <naresh.kumar.kachhi at intel.com>
> > 
> > initial version: stop rings before sync flush
> > 
> > v1: address the comments from Chris Wilson on
> > http://www.spinics.net/lists/intel-gfx/msg41025.html.
> > Rather than adding stop and start ring, we are moving
> > ring disable before hw_status_page_setup (sync flush).
> > 
> > Naresh Kumar Kachhi (2):
> >   drm/i915: disable rings before HW status page setup
> >   drm/i915: warn if ring is active before sync flush
> 
> That looks much safer in general, thanks.

And it raises the spectre of what if the rings didn't stop
immediately... Maybe we should throw in a wait_for(ring_idle(ring));
That could explain some of our mysterious problems when trying to start
rings and the random retry behaviour we resort to...
Make that patch 1.5 :)
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre



More information about the Intel-gfx mailing list