t; + i915_gem_shrinker_lock(_priv->drm, >unlock))
> + break;
> +
> schedule_timeout_killable(1);
> if (fatal_signal_pending(current))
> return false;
> +
> if (--timeout == 0) {
>
gt;
> subdir-ccflags-$(CONFIG_DRM_I915_WERROR) := -Werror
> +subdir-ccflags-$(CONFIG_DRM_I915_KCOV) := $(CFLAGS_KCOV)
> subdir-ccflags-y += \
> $(call as-instr,movntdqa (%eax)$(comma)%xmm0,-DCONFIG_AS_MOVNTDQA)
>
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
28,7 @@ void i915_gem_request_retire_upto(struct
> drm_i915_gem_request *req)
> struct drm_i915_gem_request *tmp;
>
> lockdep_assert_held(>i915->drm.struct_mutex);
> -
> - if (list_empty(>link))
> - return;
> + GEM_BUG_ON(list_empty(>link)
On to, 2016-08-04 at 12:34 +0100, Chris Wilson wrote:
> On Thu, Aug 04, 2016 at 02:17:22PM +0300, Joonas Lahtinen wrote:
> >
> > >
> > > --- a/drivers/gpu/drm/i915/intel_display.c
> > > +++ b/drivers/gpu/drm/i915/intel_display.c
> > > @@ -2465,9 +2
On to, 2016-08-04 at 11:42 +0100, Chris Wilson wrote:
> On Thu, Aug 04, 2016 at 01:36:24PM +0300, Joonas Lahtinen wrote:
> >
> > On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> > >
> > > We don't need to incur the overhead of checking whether the object
ier[0] == I915_FORMAT_MOD_X_TILED)) {
> DRM_DEBUG("tiling_mode doesn't match fb modifier\n");
> return -EINVAL;
> }
> } else {
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On ke, 2016-08-03 at 14:43 +0100, Chris Wilson wrote:
> On Wed, Aug 03, 2016 at 04:30:35PM +0300, Joonas Lahtinen wrote:
> >
> > On ke, 2016-08-03 at 13:04 +0100, Chris Wilson wrote:
> > >
> > > On Wed, Aug 03, 2016 at 12:56:39PM +0100, Chris Wilson wrote:
> &
On to, 2016-08-04 at 08:30 +0100, Chris Wilson wrote:
> On Thu, Aug 04, 2016 at 10:25:42AM +0300, Joonas Lahtinen wrote:
> >
> > On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> > >
> > > If we make the observation that mmap-offsets are only released whe
On to, 2016-08-04 at 12:41 +0100, Chris Wilson wrote:
> On Thu, Aug 04, 2016 at 02:17:22PM +0300, Joonas Lahtinen wrote:
> >
> > On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> > >
> > > --- a/drivers/gpu/drm/i915/i915_drv.h
> > > +++ b/drivers
m/i915: Support for pread/pwrite ...")
> Reported-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
> Cc: Ankitprasad Sharma <ankitprasad
after we lookup the engine's id, we double check that
This double check is nowhere to be seen, time to update this comment
too?
The code itself is quite OK, but the comments mislead my understanding
of the code again.
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Cor
ol over the flags
> field was too noisy for a simple patch. Note, that we can use the lower
s/, / /,s/can/could/
With that,
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
importantly, accessing
> dev_priv->mm.interruptible not under any controlling lock. That takes
> passing around bool interruptible and suddenly we have a bigger patch.
> :|
Not sure what to do with this information, will you send a new
revision?
Regards, Joonas
&
On to, 2016-08-04 at 07:30 +0100, Chris Wilson wrote:
> On Thu, Aug 04, 2016 at 09:12:07AM +0300, Joonas Lahtinen wrote:
> >
> > On ke, 2016-08-03 at 20:38 +0100, Chris Wilson wrote:
> > >
> > > In the next merge, we can build support for kcov at the indiv
On ke, 2016-08-03 at 14:49 +0100, Chris Wilson wrote:
> On Wed, Aug 03, 2016 at 04:43:38PM +0300, Joonas Lahtinen wrote:
> >
> > On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> > >
> > > Inside the kthread context, we can't be interrupted by signals so
&g
On to, 2016-08-04 at 11:02 +0100, Chris Wilson wrote:
> On Thu, Aug 04, 2016 at 11:26:04AM +0300, Joonas Lahtinen wrote:
> >
> > On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> > >
> > > With a bit of care (and leniency) we can iterate over the o
). This brings us into line with the other
> delayed checks (and madvise in general).
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Don't necessarily agree with all the "== 0", because it's in the
minority but;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@lin
+ goto out_unlocked;
> +
> + intel_runtime_pm_get(dev_priv);
>
As discussed in IRC, pread_ioctl does not take RPM for the fallback
path, it should.
> + ret = i915_mutex_lock_interruptible(dev);
> + if (ret)
> + goto out_rpm;
> +
&g
+ to_rps_client(file));
Long line.
This and explanation touched up,
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -3853,11 +3853,6 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void
> *data,
> goto unlock;
> }
>
> - if (i915_gem_obj_is_pinned(obj)) {
> - ret = -EINVAL;
> - goto out;
>
On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> After removing the user of this wart, we can remove the wart entirely.
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
--
Joonas Lahtinen
O
gt;struct_mutex);
> + } else {
> + ret = 0;
> +unref:
No, nope, nein, ei, njet, inte, nack; this shall not pass.
Most inappropriate use of goto I've seen shortly.
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
a kthread context and can't be interrupted */
> WARN_ON(i915_gem_object_unbind(obj));
> WARN_ON(i915_gem_object_put_pages(obj));
> -
> - dev_priv->mm.interruptible = was_interruptible;
> }
>
> i915_gem_object_put(obj);
--
J
On ke, 2016-08-03 at 14:36 +0100, Chris Wilson wrote:
> On Wed, Aug 03, 2016 at 04:23:16PM +0300, Joonas Lahtinen wrote:
> >
> > On ma, 2016-08-01 at 19:22 +0100, Chris Wilson wrote:
> > >
> > > /**
> > > @@ -1647,6 +1629,15 @@ int i915_gem_fault
that exceed the wakelock (such as legacy ringbuffers)
> and short lived pin that do live within the wakelock (such as execlist
> ringbuffers).
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
--
Jo
on <ch...@chris-wilson.co.uk>
> Reviewed-by: Matthew Auld <matthew.a...@intel.com>
Hmm, did not get my e-mail?
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
> ---
> drivers/gpu/drm/i915/intel_engine_cs.c | 50
> +
&g
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/li
all hoops.
With this delta,
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freed
Those functions look like they will get nuked in the future.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Int
;pid upon error instead. Note that we take the context's
> creator's pid rather than the file's pid in order to better track fd
> passed over sockets.
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com&g
h page) or very short lived
> (such as execbuf) and not part of the precious GGTT.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
> ---
> drivers/gpu/drm/i915/i915_debugfs.c | 100
> +
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/li
lockdep_assert_held(>vm->dev->struct_mutex);
> + i915_gem_object_put(vma->obj);
> +}
As noted in IRC, the locking asymmetry is kinda meh, to be consolidated
in the future I hope.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas La
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/li
Only commit title could now talk about moving domains instead of
clflush, then;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx m
on <ch...@chris-wilson.co.uk>
> Reviewed-by: Matthew Auld <matthew.a...@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
hg or cmpxchg) i.e. atomic_fetch_and_zero() and
> similar to llist.
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
> Cc: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
> Cc: Daniel Vetter <daniel.vet...@ffwll.ch>
> ---
> drivers/gpu/drm/i915/i91
d change function name
approppriately and we're good;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx
== -EACCES) /* unhandled chained batch */
>- return NULL;
>- else
>- return ERR_PTR(ret);
>+ return vma;
> }
You made it the success path too, so just 'out'.
That's pretty much all I found by scrolling through it;
Reviewed-by: Joonas Lahtinen <joo
s Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http
static bool i915_vma_fence_prepare(struct i915_vma *vma, int tiling_mode)
> +{
> + struct drm_i915_private *dev_priv = to_i915(vma->vm->dev);
> + u32 size;
> +
> + if (!i915_vma_is_map_and_fenceable(vma))
> + return true;
> +
> + if (INTEL_GEN(
i915_gem_object_to_ggtt(obj, );
>
> - if (view.type == I915_GGTT_VIEW_NORMAL)
> + if (i915_vma_is_map_and_fenceable(vma))
> i915_gem_object_unpin_fence(obj);
>
> - vma = i915_gem_object_to_ggtt(obj, );
> i915_gem_object_unpin_from_display_plane(vma);
T
ng part of the mapping or never
> + * having accessed it before (at this partials' range).
> + */
> + const struct i915_ggtt_view *view = >ggtt_view;
I now see why you did the rename. Do not have a better idea really, so;
Reviewed-by: Joonas Lah
le *pages;
> void __iomem *iomap;
> - u64 size;
> + u64 size, display_alignment;
Unrelated variables, better off their own lines.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporatio
On ma, 2016-08-15 at 10:25 +0100, Chris Wilson wrote:
> On Mon, Aug 15, 2016 at 12:18:20PM +0300, Joonas Lahtinen wrote:
> >
> > On pe, 2016-08-12 at 11:28 +0100, Chris Wilson wrote:
> > >
> > > + if (1) {
> > Umm? At least ought to have TODO: / FIXME: or s
o use fences for unmappable scanouts.
> v4: Rebase to skip over some local patches
> v5: Rebase to defer until after we have unmappable GTT fault support
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
if (!i915_vma_is_ggtt(vma))
> + continue;
> +
> + if (i915_vma_is_active(vma))
> + continue;
Could combine these two to one if.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahti
r semantically correct.
>
> An unfortunate side-effect is we lose the wondrously named
> i915_gem_object_finish_gtt(), not to be confused with
> i915_gem_gtt_finish_object()!
>
Does what it promises.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Rega
On pe, 2016-08-12 at 12:13 +0100, Chris Wilson wrote:
> On Fri, Aug 12, 2016 at 01:50:56PM +0300, Joonas Lahtinen wrote:
> >
> > On pe, 2016-08-12 at 11:28 +0100, Chris Wilson wrote:
> > >
> > > @@ -1715,10 +1716,10 @@ int i915_gem_fault(struct vm_area_struct *
On ma, 2016-08-15 at 11:12 +0100, Chris Wilson wrote:
> On Mon, Aug 15, 2016 at 12:59:09PM +0300, Joonas Lahtinen wrote:
> >
> > >
> > > When we bump the MRU access tracking on set-to-gtt, we need to not only
> > > bump the primary GGTT VMA but all
t; + else
> + return -EAGAIN;
> + }
>
> /* Let the mmu-notifier know that we have begun and need cancellation */
> ret = __i915_gem_userptr_set_active(obj, true);
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
ting context id. This is a one way transition; SVM
> + * contexts can not be downgraded into PPGTT contexts once converted.
> + */
> +#define I915_GEM_CONTEXT_ENABLE_SVM (1<<0)
> +#define I915_GEM_CONTEXT_FULL_PPGTT (1<<1)
BIT()
With the above addressed;
heck against busyness before reporting the request, so if
> we claim the GPU had hung but this engine hadn't we could potential skip
> the request cleanup - triggering the self-check BUG.
>
Could document the splat here.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
GuC usage
> + * @vma: gem obj to be released
The GEM object is kinda created implicitly, so not sure how to include
it in wording, improvise.
With the minor tweaks to comments,
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas
hris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
ht
is not relevant to this piece of code anymore?
Looks like more proper RCU code;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
> - rcu_read_lock();
> for_each_engine(engine, i915)
> if (unlikely(intel_engine_wakeup(engine))
ng a barrier or two.
I already provided review feedback for this series, so after that is
applied I'll proceed to review. Don't forget to CC me.
Regards, Joonas
> -Chris
>
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Inte
On ke, 2016-08-10 at 14:00 +0300, Joonas Lahtinen wrote:
>
> Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
>
> I still think it's fragile, though. But lets see once the dust settles
> if we can make improvements.
>
Daniel pointed out that the engine_id
for (j = 0; j < ee->num_requests; j++) {
> - err_printf(m, " seqno 0x%08x, emitted %ld,
> tail 0x%08x\n",
> + err_printf(m, " seqno 0x%08x, emitted %ld,
> head 0x%08x tail 0x%08x\n",
Ma
er-intuitive for it describes other VMAs
rather than our new VMA...
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On ke, 2016-08-10 at 08:38 +0100, Chris Wilson wrote:
> On Wed, Aug 10, 2016 at 10:29:59AM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > >
> > > if (obj->pin_display) {
> > > -
/*
> * Clear this page out of any CPU caches for coherent swap-in/out.
> */
Move/update the comment too?
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On ke, 2016-08-10 at 08:15 +0100, Chris Wilson wrote:
> On Wed, Aug 10, 2016 at 10:04:16AM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > >
> > > - if (i) {
> > > + active_bo = NULL;
> > Could be i
et);
> + return ERR_CAST(vma);
> }
> + ring->vma = vma;
> + if (HAS_LLC(engine->i915) && !vma->obj->stolen)
> + ring->vmap = true;
use_vmap/need_vmap or something? 'vmap' sounds like the actual mapping.
> ret
node.start + 2*CACHELINE_BYTES;
While correcting formatting; add spaces around *
With those two tweaks;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On ke, 2016-08-10 at 10:02 +0100, Chris Wilson wrote:
> On Wed, Aug 10, 2016 at 11:41:39AM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > >
> > > @@ -771,6 +771,13 @@ static int do_rcs_switch(struct
On ke, 2016-08-10 at 09:36 +0100, Chris Wilson wrote:
> On Wed, Aug 10, 2016 at 11:07:46AM +0300, Joonas Lahtinen wrote:
> >
> > On ke, 2016-08-10 at 08:15 +0100, Chris Wilson wrote:
> > >
> > > On Wed, Aug 10, 2016 at 10:04:16AM +0300, Joonas Lahtinen wrote
On ke, 2016-08-10 at 09:25 +0100, Chris Wilson wrote:
> On Wed, Aug 10, 2016 at 11:03:39AM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > >
> > > - if (!i915_gem_obj_ggtt_bound(ctx_obj))
> > > -
On ti, 2016-08-09 at 12:53 +0100, Chris Wilson wrote:
> On Tue, Aug 09, 2016 at 02:44:41PM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > > @@ -446,15 +458,7 @@ int i915_error_state_to_str(struct
> >
On ma, 2016-08-08 at 10:09 +0100, Chris Wilson wrote:
> On Mon, Aug 08, 2016 at 12:01:07PM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > > @@ -390
On ke, 2016-08-10 at 12:13 +0200, Daniel Vetter wrote:
> On Wed, Aug 10, 2016 at 12:12:37PM +0200, Daniel Vetter wrote:
> >
> > On Tue, Aug 09, 2016 at 10:05:30AM +0100, Chris Wilson wrote:
> > >
> > > On Tue, Aug 09, 2016 at 11:48:56AM +0300, Joonas Lahtinen wr
erence handling of the request->batch_obj next to
> where the active tracking is set up for the request.
>
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Sou
++pin_count;
variables names to form pin_display_*
> + seq_printf(m, "%u mapped objects, %llu bytes\n",
> + mapped_count, mapped_size);
> + seq_printf(m, "%u pinned objects, %llu bytes\n",
"display pinned objects"
With those;
R
5_gem_object_unpin_map(ce->state->obj);
> +unpin_vma:
> + __i915_vma_unpin(ce->state);
err_vma while at it?
> @@ -2161,7 +2162,7 @@ static int execlists_context_deferred_alloc(struct
> i915_gem_context *ctx,
> }
>
> ce->ring = ring;
> - ce->state = ctx_obj;
> + ce->state = vma;
Maybe the member name could be just ce->vma too?
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
into a zlib function?
> @@ -629,6 +680,18 @@ i915_error_object_create(struct drm_i915_private *i915,
>
> dst->gtt_offset = vma->node.start;
> dst->page_count = 0;
> + dst->unused = 0;
> +
> + memset(, 0, sizeof(zstream));
> + zstream.workspace
On ke, 2016-08-10 at 11:52 +0100, Chris Wilson wrote:
> On Wed, Aug 10, 2016 at 01:32:29PM +0300, Joonas Lahtinen wrote:
> >
> > On su, 2016-08-07 at 15:45 +0100, Chris Wilson wrote:
> > >
> > > @@ -309,12 +310,30 @@ void i915_error_printf(struct
> > >
On ma, 2016-06-20 at 08:17 -0400, Zhi Wang wrote:
> The access right of kernel param "i915.enable_gvt" should be read-only as
> it only applies during module load (and is not *runtime* writable).
>
> Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Reviewed-by: J
ain. (We had to wait until any outstanding retire worker
> finally ran and cleared the active status.)
>
> Fixes: 67d97da34917 ("drm/i915: Only start retire worker when idle")
> Testcase: igt/pm_rps/reset
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co
EN(dev_priv);
> + so->ggtt_offset = 0;
Previousy not done, does it address a bug? It's going to get
overwritten or the render_state has failed to initialize and is
forgotten, no? If it fixes bug, I think the site fondling uninitialized
object should be fixed.
Reviewed-by: Joonas Lahtinen &l
On ke, 2016-07-20 at 14:12 +0100, Chris Wilson wrote:
> Both the ->dispatch_execbuffer and ->emit_bb_start callbacks do exactly
> the same thing, add MI_BATCHBUFFER_START to the request's ringbuffer -
> we need only one vfunc.
>
Some ranting below,
Reviewed-by: Joonas Laht
ne->emit_request = gen6_emit_request;
Not sure if I would prefer the assignment here still. If overrides
happen from all around the codebase, it'll be harder to come up with
what are the possible values for a vfunc, right?
Regards, Joonas
> engine->submit_request = i9xx_sub
i915_gem_request *req)
> intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
> intel_ring_emit(ring, req->fence.seqno);
> intel_ring_emit(ring, MI_USER_INTERRUPT);
> - __intel_engine_submit(req->engine);
> + intel_ring_advance
ATION_0 | IGT_ROTATION_180;
> + test_bad_prop_value();
> + }
> +
> igt_subtest_f("primary-rotation-180") {
> data.rotation = IGT_ROTATION_180;
> test_plane_rotation(, IGT_PLANE_PRIMARY);
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
re confident this is pure mechanical work.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.fre
as a
mechanical change with no hidden functional changes. Then,
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@
ctx_obj;
> ce->initialised = engine->init_context == NULL;
>
> return 0;
>
> error_ringbuf:
All the same to change the label too, then;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
> - intel_ringbuffer_free(ringbuf);
> + intel_ringbuffer_free(ring);
> error_deref_obj:
> i915_gem_object_put(ctx_obj);
> - ce->ringbuf = NULL;
> + ce->ring = NULL;
> ce->state = NULL;
> return ret;
> }
> @@ -2407,7 +2404,7 @@ void intel_lr_context_reset(struct drm_i915_private
> *dev_priv,
>
> i915_gem_object_unpin_map(ctx_obj);
>
> - ce->ringbuf->head = 0;
> - ce->ringbuf->tail = 0;
> + ce->ring->head = 0;
> + ce->ring->tail = 0;
> }
> }
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
bit when the rotate bit is
> set, so we'll have to reject the 180+X case.
>
> Signed-off-by: Ville Syrjälä <ville.syrj...@linux.intel.com>
Seems to match spec. Maybe add your own T-b, as you have obviously
tested on hardware.
Reviewed-by: Joonas Lahtinen <joonas.
_init_engine(engine);
> }
>
> int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
> @@ -2951,7 +2951,7 @@ int intel_init_vebox_ring_buffer(struct intel_engine_cs
> *engine)
> engine->irq_disable = hsw_vebox_irq_disable;
> }
>
> - re
e renames here too, with those;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Regards, Joonas
> ---
> drivers/gpu/drm/i915/intel_ringbuffer.c | 66
> -
> drivers/gpu/drm/i915/intel_ringbuffer.h | 6 +--
> 2 files changed, 36 inserti
On ke, 2016-07-20 at 14:11 +0100, Chris Wilson wrote:
> For more consistent oop-naming, we would use intel_ring_verb, so pick
> intel_ring_pin() and intel_ring_unpin().
>
The done renames clearly here, then;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
> Si
hase_filter)
> {
> struct drm_i915_private *dev_priv = overlay->i915;
> - struct intel_engine_cs *engine = _priv->engine[RCS];
> struct drm_i915_gem_request *req;
> struct intel_ring *ring;
> u32 flip_addr = overlay->flip_addr;
> @@ -285,7 +29
on of buffer objects themselves.
>
You should list all the renames here so the poor rebasers get to live.
With those listed;
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Comment below.
> Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
> ---
>
On pe, 2016-07-15 at 12:49 +0100, Chris Wilson wrote:
> On Fri, Jul 15, 2016 at 12:08:05PM +0100, Chris Wilson wrote:
> >
> > On Fri, Jul 15, 2016 at 01:47:34PM +0300, Joonas Lahtinen wrote:
> > >
> > > On pe, 2016-07-15 at 11:11 +0100, Chris Wilson wrote:
>
On pe, 2016-07-15 at 11:11 +0100, Chris Wilson wrote:
> +#define NO_WAITBOOST ERR_PTR(-1)
This does not seem to be a very common idiom in kernel, just add flags
maybe.
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporat
++
> drivers/gpu/drm/i915/i915_gem_request.h | 238
> 5 files changed, 905 insertions(+), 856 deletions(-)
> create mode 100644 drivers/gpu/drm/i915/i915_gem_request.c
> create mode 100644 drivers/gpu/drm/i915/i915_gem_request.h
>
Assuming pure
lson.co.uk>
Maybe even add a comment above the assignment that _GTT is used in the
purpose of being coherent.
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
> ---
> drivers/gpu/drm/i915/i915_gem_dmabuf.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --gi
> + return 0;
> +
Ditto.
> i915_gem_object_flush_gtt_write_domain(obj);
>
> old_write_domain = obj->base.write_domain;
> @@ -4238,7 +4250,7 @@ void i915_gem_free_object(struct drm_gem_object
> *gem_obj)
> int ret;
>
> vma->pin_count = 0;
> - ret = i915_vma_unbind(vma);
Maybe add a comment/TODO/FIXME: about the potential WARN.
Regards, Joonas
> + ret = __i915_vma_unbind_no_wait(vma);
> if (WARN_ON(ret == -ERESTARTSYS)) {
> bool was_interruptible;
>
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
ction support for CHV.
>
> Signed-off-by: Ville Syrjälä <ville.syrj...@linux.intel.com>
Looks cleaner,
Reviewed-by: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
> ---
> drivers/gpu/drm/i915/intel_display.c | 28 ++--
|='s in order so that new gens add features,
making easier to get a picture of what was added at which point.
Regards, Joonas
> intel_plane->pipe = pipe;
> intel_plane->plane = plane;
> intel_plane->frontbuffer_bit = INTEL_FRONTBUFFER_SPRITE(pipe, plane);
> @@ -1143,7 +1153,7 @@ intel_plane_init(struct drm_device *dev, enum pipe
> pipe, int plane)
> if (ret)
> goto fail;
>
> - intel_create_rotation_property(dev, intel_plane);
> + intel_create_rotation_property(intel_plane, supported_rotations);
>
> drm_plane_helper_add(_plane->base, _plane_helper_funcs);
>
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
601 - 700 of 2892 matches
Mail list logo