Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-10-12 Thread Ville Syrjälä
:10 AM > To: Zanoni, Paulo R > Cc: intel-gfx@lists.freedesktop.org > Subject: Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on > SKL too > > On Wed, Sep 23, 2015 at 12:52:26PM -0300, Paulo Zanoni wrote: > > The trick is not strictly necessary on SKL because the offset &

Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-10-12 Thread Hindman, Gavin
ulo R Cc: intel-gfx@lists.freedesktop.org Subject: Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too On Wed, Sep 23, 2015 at 12:52:26PM -0300, Paulo Zanoni wrote: > The trick is not strictly necessary on SKL because the offset > registers allow more bits. But for

Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-09-24 Thread Ville Syrjälä
On Wed, Sep 23, 2015 at 12:52:26PM -0300, Paulo Zanoni wrote: > The trick is not strictly necessary on SKL because the offset > registers allow more bits. But for FBC, doing this changes how the > hardware tracking works - it starts at the surface address we provide > - so there's a higher chance t

Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-09-24 Thread Chris Wilson
On Thu, Sep 24, 2015 at 10:55:17AM +0100, Tvrtko Ursulin wrote: > > On 09/23/2015 06:03 PM, Chris Wilson wrote: > >On Wed, Sep 23, 2015 at 12:52:26PM -0300, Paulo Zanoni wrote: > >>The trick is not strictly necessary on SKL because the offset > >>registers allow more bits. But for FBC, doing this

Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-09-24 Thread Tvrtko Ursulin
On 09/23/2015 06:03 PM, Chris Wilson wrote: On Wed, Sep 23, 2015 at 12:52:26PM -0300, Paulo Zanoni wrote: The trick is not strictly necessary on SKL because the offset registers allow more bits. But for FBC, doing this changes how the hardware tracking works - it starts at the surface address w

Re: [Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-09-23 Thread Chris Wilson
On Wed, Sep 23, 2015 at 12:52:26PM -0300, Paulo Zanoni wrote: > The trick is not strictly necessary on SKL because the offset > registers allow more bits. But for FBC, doing this changes how the > hardware tracking works - it starts at the surface address we provide > - so there's a higher chance t

[Intel-gfx] [PATCH 6/7] drm/i915: use compute_page_offset() on SKL too

2015-09-23 Thread Paulo Zanoni
The trick is not strictly necessary on SKL because the offset registers allow more bits. But for FBC, doing this changes how the hardware tracking works - it starts at the surface address we provide - so there's a higher chance that the CRTC will be pointing to an area of the frontbuffer that is ac