ironlake wierdness bisection result

ironlake wierdness bisection result

[Dave Airlie]

Okay I spent today bisecting this and although I'm not 100% sure I got
it right the whole way down, I think where I landed seems sane, so I
said I'd share

This seems to cause the blank rendering zones people have been reporting.

Dave.

commit c07496fa61f4c5cb2addd1c57f6b22fcaeea2eeb
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Fri Apr 13 15:51:51 2012 +0200

drm/i915: don't pwrite tiled objects through the gtt

    ... we will botch up the bit17 swizzling. Furthermore tiled pwrite is
    a (now) unused slowpath, so no one really cares.

    This fixes the last swizzling issues I have with i-g-t on my bit17
    swizzling i915G. No regression, it's been broken since the dawn of
    gem, but it's nice for regression tracking when really _all_ i-g-t
    tests work.

    Actually this is not true, Chris Wilson noticed while reviewing this
    patch that the commit

    commit d9e86c0ee60f323e890484628f351bf50fa9a15d
    Author: Chris Wilson <chris@chris-wilson.co.uk>
    Date:   Wed Nov 10 16:40:20 2010 +0000

        drm/i915: Pipelined fencing [infrastructure]
contained a functional change that broke things.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Re: ironlake wierdness bisection result

[Dave Airlie]

On Tue, Sep 18, 2012 at 4:43 PM, Dave Airlie <airlied@gmail.com> wrote:
> Okay I spent today bisecting this and although I'm not 100% sure I got
> it right the whole way down, I think where I landed seems sane, so I
> said I'd share
>
> This seems to cause the blank rendering zones people have been reporting.

Though I'm of course having trouble reproducing the pain now, so it
might all be lies,

but hey its a better starting point than none, any way it could make
sense to be this commit?

I'll do another round of bisecton tomorrow and see if I can find a better test.

Dave.

Re: ironlake wierdness bisection result

[Daniel Vetter]

On Tue, Sep 18, 2012 at 04:58:52PM +1000, Dave Airlie wrote:
> On Tue, Sep 18, 2012 at 4:43 PM, Dave Airlie <airlied@gmail.com> wrote:
> > Okay I spent today bisecting this and although I'm not 100% sure I got
> > it right the whole way down, I think where I landed seems sane, so I
> > said I'd share
> >
> > This seems to cause the blank rendering zones people have been reporting.
> 
> Though I'm of course having trouble reproducing the pain now, so it
> might all be lies,
> 
> but hey its a better starting point than none, any way it could make
> sense to be this commit?

Well, if we presume that something went wrong with relocs, the only thing
this patch might affect is gtt placement. But userspace shouldn't take
this path at all (at least if you're running anything remotely new). So I
really can't see what's going on here and how this changes things ...
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch