Re: [PATCH] drm/i915: get runtime PM reference around GEM set_caching IOCTL

From: Imre Deak <imre.deak@intel.com>
To: Daniel Vetter <daniel@ffwll.ch>, Paulo Zanoni <przanoni@gmail.com>
Cc: Intel Graphics Development <intel-gfx@lists.freedesktop.org>
Subject: Re: [PATCH] drm/i915: get runtime PM reference around GEM set_caching IOCTL
Date: Tue, 17 Nov 2015 21:22:10 +0200	[thread overview]
Message-ID: <1447788130.9159.12.camel@intel.com> (raw)
In-Reply-To: <20151117190055.GX16848@phenom.ffwll.local>

On ti, 2015-11-17 at 20:00 +0100, Daniel Vetter wrote:
> On Wed, Nov 04, 2015 at 05:47:12PM -0200, Paulo Zanoni wrote:
> > 2015-11-04 17:25 GMT-02:00 Imre Deak <imre.deak@intel.com>:
> > > After Damien's D3 fix I started to get runtime suspend residency
> > > for the
> > > first time and that revealed a breakage on the set_caching IOCTL
> > > path
> > > that accesses the HW but doesn't take an RPM ref. Fix this up.
> > 
> > Oh, well, that's the the RPM problem that prevents me from sleeping
> > at
> > night: how can we be 100% sure we wrapped every single possible
> > driver
> > entry point? At least we shout errors and warns when we detect the
> > problem.
> > 
> > By the way, we already have some tests for specific IOCTLs on
> > igt/pm_rpm, so adding a new one for this IOCTL wouldn't hurt :)
> 
> Yeah, rpm subtests for set_caching and set_tiling are indeed missing.
> Imre, can you please add them?

Ok, will add them.

--Imre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx