Re: [Intel-gfx] [PATCH] drm/i915/ppgtt: Defer request freeing on reset

On Tue, Dec 31, 2013 at 11:47:13PM -0800, Ben Widawsky wrote: > drivers/gpu/drm/i915/i915_gem.c | 18 ++ > 1 file changed, 10 insertions(+), 8 deletions(-) > > diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c > index 7ba12b4..4bd7057 100644 > --- a/dr

[Intel-gfx] [PATCH 3/3] drm/i915: Remove redundant list_empty(eb->vmas) tests in execbuffer

Part of the pre-validation for an execbuffer call is that there is at least one object in the execlist. As we bail if we fail to lookup any object, we can be sure that after the eb_lookup_vma() there is at least one object in the vma list and so we do not need to assert. Signed-off-by: Chris Wilso

[Intel-gfx] [PATCH 1/3] drm/i915: Only bind each object rather than for every execbuffer

One side-effect of the introduction of ppgtt was that we needed to rebind the object into the appropriate vm (and global gtt in some peculiar cases). For simplicity this was done twice for every object on every call to execbuffer. However, that adds a tremendous amount of CPU overhead (rewriting al

[Intel-gfx] [PATCH 2/3] drm/i915: Fix obj->map_and_fenceable for ppgtt

With ppgtt, it is no longer correct to mark an object as map_and_fenceable if we simply unbind it from the global gtt. This has consequences during execbuffer where we simply use obj->map_and_fenceable in use_cpu_reloc() to decide which access method to use for writing into the object. Now for a pp

Re: [Intel-gfx] [PATCH] drm/i915: set ctx->initialized only after RCS

On Tue, Dec 31, 2013 at 11:26:17AM +, Chris Wilson wrote: > On Mon, Dec 30, 2013 at 01:34:46PM -0800, Ben Widawsky wrote: > > On Sun, Dec 29, 2013 at 09:59:26AM +, Chris Wilson wrote: > > > On Sat, Dec 28, 2013 at 01:31:49PM -0800, Ben Widawsky wrote: > > > > The initialized flag is used to

[Intel-gfx] [PATCH] [v2] drm/i915/ppgtt: Defer request freeing on reset

We need to defer the free request until the object/vma is capable of being freed - or else we have a problem when we try to destroy the context. The exact same issue is described and fixed here: commit e20780439b26ba95aeb29d3e27cd8cc32bc82a4c Author: Ben Widawsky Date: Fri Dec 6 14:11:22 2013

[Intel-gfx] [PULL] drm-intel-fixes

Hi Dave, Jani has asked me to merge the one revert for fixes. It has been in my repo and undergone our QA nightly testing over the last few days. I am not able to verify this patch myself, as I do not have any of the effected machines. If you prefer to wait for Daniel's return from vacation, I be

Re: [Intel-gfx] [PATCH 2/2] drm/i915: Avoid dereference past end of page array in gen8_ppgtt_insert_entries()

On Tue, Dec 31, 2013 at 03:50:31PM +, Chris Wilson wrote: > The bug from gen6_ppgtt_insert_entries() was replicated into > gen8_ppgtt_insert_entries(). This applies the fix for the OOPS from the > previous patch to the gen8 routine. > > Signed-off-by: Chris Wilson > Cc: Daniel Vetter > Cc: B

Re: [Intel-gfx] [PATCH 1/3] drm/i915: Only bind each object rather than for every execbuffer

On Wed, Jan 01, 2014 at 02:00:54PM +, Chris Wilson wrote: > One side-effect of the introduction of ppgtt was that we needed to > rebind the object into the appropriate vm (and global gtt in some > peculiar cases). For simplicity this was done twice for every object on > every call to execbuffer

[Intel-gfx] -nightly and -fixes status

Hi Daniel, and welcome back. Tomorrow I go on vacation, and since it's more or less the end of the day for anyone still submitting or reviewing patches, I figured now is as good a time as any to do this. http://cgit.freedesktop.org/~bwidawsk/drm-intel/log/?h=drm-intel-fixes got one patch, which I