Re: [PATCH 2/2] drm/i915: Use Write-Through cacheing for the display plane on Iris

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Jul 30, 2013 at 07:10:03PM +0100, Chris Wilson wrote:
> On Tue, Jul 30, 2013 at 09:01:22PM +0300, Ville Syrjälä wrote:
> > Hmm, and what about CPU access in general? CPU doesn't know about the
> > WT policy, so I'm assuming we'd need to flush after all CPU writes.
> 
> Hmm, I had assumed the opposite. I've not yet seen signs of cache dirt.
> Certainly GTT access should fine, and CPU reads. It is just those CPU
> writes we need to worry about, and now I do indeed worry.

Ok, didn't take too long to force a few updates through CPU mappings to
a WT display, and yes the CPU writes are incoherent. So we need to be
much more careful with the flushes here.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx





[Index of Archives]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]
  Powered by Linux