Re: [Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-31 Thread Daniel Vetter
On Mon, Jan 30, 2012 at 04:55:49PM +0100, Daniel Vetter wrote: > Chris Wilson reported that with a bunch of patches to no longer force > batchbuffer (in most cases at least) into the mappable part of gtt > that his Pineview died while prefetching the last page of the gtt. > > Turns out that since

Re: [Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-31 Thread Daniel Vetter
On Tue, Jan 31, 2012 at 02:24:36PM +0100, Daniel Vetter wrote: > Ok, I've botched up the testing on my snb, shame on me. > > Without this patch my snb also dies with gem_cs_prefetch, but like Chris' > pnv, survives without a hitch. So it looks like I've lucked out. This is getting funny: I've mea

Re: [Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-31 Thread Daniel Vetter
On Mon, Jan 30, 2012 at 08:22:29PM +0100, Daniel Vetter wrote: > On Mon, Jan 30, 2012 at 08:14:21PM +0100, Daniel Vetter wrote: > > On Mon, Jan 30, 2012 at 11:09:01AM -0800, Eric Anholt wrote: > > > On Mon, 30 Jan 2012 16:55:49 +0100, Daniel Vetter > > > wrote: > > > > Chris Wilson reported that

Re: [Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-30 Thread Daniel Vetter
On Mon, Jan 30, 2012 at 08:14:21PM +0100, Daniel Vetter wrote: > On Mon, Jan 30, 2012 at 11:09:01AM -0800, Eric Anholt wrote: > > On Mon, 30 Jan 2012 16:55:49 +0100, Daniel Vetter > > wrote: > > > Chris Wilson reported that with a bunch of patches to no longer force > > > batchbuffer (in most cas

Re: [Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-30 Thread Daniel Vetter
On Mon, Jan 30, 2012 at 11:09:01AM -0800, Eric Anholt wrote: > On Mon, 30 Jan 2012 16:55:49 +0100, Daniel Vetter > wrote: > > Chris Wilson reported that with a bunch of patches to no longer force > > batchbuffer (in most cases at least) into the mappable part of gtt > > that his Pineview died whi

Re: [Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-30 Thread Eric Anholt
On Mon, 30 Jan 2012 16:55:49 +0100, Daniel Vetter wrote: > Chris Wilson reported that with a bunch of patches to no longer force > batchbuffer (in most cases at least) into the mappable part of gtt > that his Pineview died while prefetching the last page of the gtt. > > Turns out that since my i

[Intel-gfx] [PATCH 2/2] drm/i915: drop the guard page at the end of the gtt

2012-01-30 Thread Daniel Vetter
Chris Wilson reported that with a bunch of patches to no longer force batchbuffer (in most cases at least) into the mappable part of gtt that his Pineview died while prefetching the last page of the gtt. Turns out that since my intel-gtt rewrite we don't actually put a dummy pte in that last page