[PATCH] drm/i915: Add missing ring_mask to Pineview

[PATCH] drm/i915: Add missing ring_mask to Pineview

[Chris Wilson]

It appears that we never told Pineview it has a RENDER_RING. This was
all fine until we started using the ring_mask for determining all the
available rings to initialise for legacy ringbuffer submission in commit
88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
latent bug since the ring_mask inception in commit 73ae478cdf6a
("drm/i915: Replace has_bsd/blt/vebox with a mask").

To prevent similar mishaps in future, add a WARN_ON() if we find
ourselves with a device without any rings.

Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Ben Widawsky <ben@bwidawsk.net>
---
 drivers/gpu/drm/i915/i915_pci.c        | 1 +
 drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
index 949c01686a66..2587b1bd41f4 100644
--- a/drivers/gpu/drm/i915/i915_pci.c
+++ b/drivers/gpu/drm/i915/i915_pci.c
@@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
 	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_overlay = 1,
+	.ring_mask = RENDER_RING,
 	GEN_DEFAULT_PIPEOFFSETS,
 	CURSOR_OFFSETS,
 };
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 31d43dfa7469..e6422aac2919 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
 	unsigned int i;
 	int ret;
 
+	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
 	WARN_ON(INTEL_INFO(dev_priv)->ring_mask &
 		GENMASK(sizeof(mask) * BITS_PER_BYTE - 1, I915_NUM_ENGINES));
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Ro.CI.BAT: failure for drm/i915: Add missing ring_mask to Pineview

[Patchwork]

== Series Details ==

Series: drm/i915: Add missing ring_mask to Pineview
URL   : https://patchwork.freedesktop.org/series/10368/
State : failure

== Summary ==

Series 10368v1 drm/i915: Add missing ring_mask to Pineview
http://patchwork.freedesktop.org/api/1.0/series/10368/revisions/1/mbox

Test drv_getparams_basic:
        Subgroup basic-eu-total:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-subslice-total:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test drv_hangman:
        Subgroup error-state-basic:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test drv_module_reload_basic:
                skip       -> PASS       (ro-ivb-i7-3770)
                pass       -> FAIL       (fi-hsw-i7-4770k)
                skip       -> PASS       (ro-skl3-i5-6260u)
Test gem_basic:
        Subgroup bad-close:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup create-close:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup create-fd-close:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_busy:
        Subgroup basic-blt:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-bsd:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-parallel-blt:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-parallel-bsd:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-parallel-render:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-parallel-vebox:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-render:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-vebox:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_close_race:
        Subgroup basic-process:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-threads:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_cs_tlb:
        Subgroup basic-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_ctx_exec:
        Subgroup basic:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_ctx_param:
        Subgroup basic:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_exec_basic:
        Subgroup basic-blt:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-bsd:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-render:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-vebox:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup gtt-blt:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup gtt-bsd:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup gtt-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup gtt-render:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup gtt-vebox:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup readonly-blt:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup readonly-bsd:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup readonly-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup readonly-render:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup readonly-vebox:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_exec_create:
        Subgroup basic:
                pass       -> SKIP       (fi-hsw-i7-4770k)
Test gem_exec_flush:
        Subgroup basic-batch-kernel-default-cmd:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-batch-kernel-default-uc:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-batch-kernel-default-wb:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-uc-pro-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
        Subgroup basic-uc-prw-default:
                pass       -> SKIP       (fi-hsw-i7-4770k)
WARNING: Long output truncated

Results at /archive/results/CI_IGT_test/RO_Patchwork_1636/

cb7629d drm-intel-nightly: 2016y-07m-28d-11h-02m-33s UTC integration manifest
28c795d drm/i915: Add missing ring_mask to Pineview

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Joonas Lahtinen]

On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> It appears that we never told Pineview it has a RENDER_RING. This was
> all fine until we started using the ring_mask for determining all the
> available rings to initialise for legacy ringbuffer submission in commit
> 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> latent bug since the ring_mask inception in commit 73ae478cdf6a
> ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> 
> To prevent similar mishaps in future, add a WARN_ON() if we find
> ourselves with a device without any rings.
> 
> Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> Cc: Ben Widawsky <ben@bwidawsk.net>
> ---
>  drivers/gpu/drm/i915/i915_pci.c        | 1 +
>  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
>  2 files changed, 2 insertions(+)
> 
> diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> index 949c01686a66..2587b1bd41f4 100644
> --- a/drivers/gpu/drm/i915/i915_pci.c
> +++ b/drivers/gpu/drm/i915/i915_pci.c
> @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
>  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
>  	.need_gfx_hws = 1, .has_hotplug = 1,
>  	.has_overlay = 1,
> +	.ring_mask = RENDER_RING,
>  	GEN_DEFAULT_PIPEOFFSETS,
>  	CURSOR_OFFSETS,
>  };
> diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> index 31d43dfa7469..e6422aac2919 100644
> --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
>  	unsigned int i;
>  	int ret;
>  
> +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

>  	WARN_ON(INTEL_INFO(dev_priv)->ring_mask &
>  		GENMASK(sizeof(mask) * BITS_PER_BYTE - 1, I915_NUM_ENGINES));
>  
-- 
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: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Daniel Vetter]

On Fri, Jul 29, 2016 at 11:42:24AM +0300, Joonas Lahtinen wrote:
> On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> > It appears that we never told Pineview it has a RENDER_RING. This was
> > all fine until we started using the ring_mask for determining all the
> > available rings to initialise for legacy ringbuffer submission in commit
> > 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> > latent bug since the ring_mask inception in commit 73ae478cdf6a
> > ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> > 
> > To prevent similar mishaps in future, add a WARN_ON() if we find
> > ourselves with a device without any rings.
> > 
> > Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> > Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > Cc: Ben Widawsky <ben@bwidawsk.net>
> > ---
> >  drivers/gpu/drm/i915/i915_pci.c        | 1 +
> >  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
> >  2 files changed, 2 insertions(+)
> > 
> > diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> > index 949c01686a66..2587b1bd41f4 100644
> > --- a/drivers/gpu/drm/i915/i915_pci.c
> > +++ b/drivers/gpu/drm/i915/i915_pci.c
> > @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
> >  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
> >  	.need_gfx_hws = 1, .has_hotplug = 1,
> >  	.has_overlay = 1,
> > +	.ring_mask = RENDER_RING,
> >  	GEN_DEFAULT_PIPEOFFSETS,
> >  	CURSOR_OFFSETS,
> >  };
> > diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> > index 31d43dfa7469..e6422aac2919 100644
> > --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> > +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> > @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
> >  	unsigned int i;
> >  	int ret;
> >  
> > +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
> 
> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

dim says this also needs a

Cc: drm-intel-fixes@lists.freedesktop.org

> 
> >  	WARN_ON(INTEL_INFO(dev_priv)->ring_mask &
> >  		GENMASK(sizeof(mask) * BITS_PER_BYTE - 1, I915_NUM_ENGINES));
> >  
> -- 
> Joonas Lahtinen
> Open Source Technology Center
> Intel Corporation

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Chris Wilson]

On Fri, Jul 29, 2016 at 10:57:49AM +0200, Daniel Vetter wrote:
> On Fri, Jul 29, 2016 at 11:42:24AM +0300, Joonas Lahtinen wrote:
> > On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> > > It appears that we never told Pineview it has a RENDER_RING. This was
> > > all fine until we started using the ring_mask for determining all the
> > > available rings to initialise for legacy ringbuffer submission in commit
> > > 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> > > latent bug since the ring_mask inception in commit 73ae478cdf6a
> > > ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> > > 
> > > To prevent similar mishaps in future, add a WARN_ON() if we find
> > > ourselves with a device without any rings.
> > > 
> > > Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> > > Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > Cc: Ben Widawsky <ben@bwidawsk.net>
> > > ---
> > >  drivers/gpu/drm/i915/i915_pci.c        | 1 +
> > >  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
> > >  2 files changed, 2 insertions(+)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> > > index 949c01686a66..2587b1bd41f4 100644
> > > --- a/drivers/gpu/drm/i915/i915_pci.c
> > > +++ b/drivers/gpu/drm/i915/i915_pci.c
> > > @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
> > >  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
> > >  	.need_gfx_hws = 1, .has_hotplug = 1,
> > >  	.has_overlay = 1,
> > > +	.ring_mask = RENDER_RING,
> > >  	GEN_DEFAULT_PIPEOFFSETS,
> > >  	CURSOR_OFFSETS,
> > >  };
> > > diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > index 31d43dfa7469..e6422aac2919 100644
> > > --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> > > +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
> > >  	unsigned int i;
> > >  	int ret;
> > >  
> > > +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
> > 
> > Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> 
> dim says this also needs a
> 
> Cc: drm-intel-fixes@lists.freedesktop.org

dim says nothing of the sort for me. The bug is only relevant from
88d2ba2e95c8
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: ✗ Ro.CI.BAT: failure for drm/i915: Add missing ring_mask to Pineview

[Chris Wilson]

On Fri, Jul 29, 2016 at 05:51:43AM -0000, Patchwork wrote:
> == Series Details ==
> 
> Series: drm/i915: Add missing ring_mask to Pineview
> URL   : https://patchwork.freedesktop.org/series/10368/
> State : failure
> 
> == Summary ==
> 
> Series 10368v1 drm/i915: Add missing ring_mask to Pineview
> http://patchwork.freedesktop.org/api/1.0/series/10368/revisions/1/mbox
> 
> Test drv_getparams_basic:
>         Subgroup basic-eu-total:
>                 pass       -> SKIP       (fi-hsw-i7-4770k)
>         Subgroup basic-subslice-total:
>                 pass       -> SKIP       (fi-hsw-i7-4770k)
> Test drv_hangman:
>         Subgroup error-state-basic:
>                 pass       -> SKIP       (fi-hsw-i7-4770k)
> Test drv_module_reload_basic:
>                 skip       -> PASS       (ro-ivb-i7-3770)
>                 pass       -> FAIL       (fi-hsw-i7-4770k)
>                 skip       -> PASS       (ro-skl3-i5-6260u)
...
Got a sane result from trybot
https://patchwork.freedesktop.org/series/10390/
and pushed.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Ben Widawsky]

On Fri, Jul 29, 2016 at 12:10:30PM +0100, Chris Wilson wrote:
> On Fri, Jul 29, 2016 at 10:57:49AM +0200, Daniel Vetter wrote:
> > On Fri, Jul 29, 2016 at 11:42:24AM +0300, Joonas Lahtinen wrote:
> > > On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> > > > It appears that we never told Pineview it has a RENDER_RING. This was
> > > > all fine until we started using the ring_mask for determining all the
> > > > available rings to initialise for legacy ringbuffer submission in commit
> > > > 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> > > > latent bug since the ring_mask inception in commit 73ae478cdf6a
> > > > ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> > > > 
> > > > To prevent similar mishaps in future, add a WARN_ON() if we find
> > > > ourselves with a device without any rings.
> > > > 
> > > > Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> > > > Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> > > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > > > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > Cc: Ben Widawsky <ben@bwidawsk.net>
> > > > ---
> > > >  drivers/gpu/drm/i915/i915_pci.c        | 1 +
> > > >  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
> > > >  2 files changed, 2 insertions(+)
> > > > 
> > > > diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> > > > index 949c01686a66..2587b1bd41f4 100644
> > > > --- a/drivers/gpu/drm/i915/i915_pci.c
> > > > +++ b/drivers/gpu/drm/i915/i915_pci.c
> > > > @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
> > > >  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
> > > >  	.need_gfx_hws = 1, .has_hotplug = 1,
> > > >  	.has_overlay = 1,
> > > > +	.ring_mask = RENDER_RING,
> > > >  	GEN_DEFAULT_PIPEOFFSETS,
> > > >  	CURSOR_OFFSETS,
> > > >  };
> > > > diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > index 31d43dfa7469..e6422aac2919 100644
> > > > --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
> > > >  	unsigned int i;
> > > >  	int ret;
> > > >  
> > > > +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
> > > 
> > > Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> > 
> > dim says this also needs a
> > 
> > Cc: drm-intel-fixes@lists.freedesktop.org
> 
> dim says nothing of the sort for me. The bug is only relevant from
> 88d2ba2e95c8
> -Chris
> 

I was about to say I can't imagine that a bug like this would have existed for
that long without anybody yelling about it.

-- 
Ben Widawsky, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Daniel Vetter]

On Fri, Jul 29, 2016 at 12:10:30PM +0100, Chris Wilson wrote:
> On Fri, Jul 29, 2016 at 10:57:49AM +0200, Daniel Vetter wrote:
> > On Fri, Jul 29, 2016 at 11:42:24AM +0300, Joonas Lahtinen wrote:
> > > On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> > > > It appears that we never told Pineview it has a RENDER_RING. This was
> > > > all fine until we started using the ring_mask for determining all the
> > > > available rings to initialise for legacy ringbuffer submission in commit
> > > > 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> > > > latent bug since the ring_mask inception in commit 73ae478cdf6a
> > > > ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> > > > 
> > > > To prevent similar mishaps in future, add a WARN_ON() if we find
> > > > ourselves with a device without any rings.
> > > > 
> > > > Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> > > > Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> > > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > > > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > Cc: Ben Widawsky <ben@bwidawsk.net>
> > > > ---
> > > >  drivers/gpu/drm/i915/i915_pci.c        | 1 +
> > > >  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
> > > >  2 files changed, 2 insertions(+)
> > > > 
> > > > diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> > > > index 949c01686a66..2587b1bd41f4 100644
> > > > --- a/drivers/gpu/drm/i915/i915_pci.c
> > > > +++ b/drivers/gpu/drm/i915/i915_pci.c
> > > > @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
> > > >  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
> > > >  	.need_gfx_hws = 1, .has_hotplug = 1,
> > > >  	.has_overlay = 1,
> > > > +	.ring_mask = RENDER_RING,
> > > >  	GEN_DEFAULT_PIPEOFFSETS,
> > > >  	CURSOR_OFFSETS,
> > > >  };
> > > > diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > index 31d43dfa7469..e6422aac2919 100644
> > > > --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
> > > >  	unsigned int i;
> > > >  	int ret;
> > > >  
> > > > +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
> > > 
> > > Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> > 
> > dim says this also needs a
> > 
> > Cc: drm-intel-fixes@lists.freedesktop.org
> 
> dim says nothing of the sort for me. The bug is only relevant from
> 88d2ba2e95c8

$ dim fixes 88d2ba2e95c8
Fixes: 88d2ba2e95c8 ("drm/i915: Move common engine setup into intel_engine_cs.c")
Cc: drm-intel-fixes@lists.freedesktop.org

Cc: -fixes is for the current -rc, which the above seems to have escaped
into.

Cc: -stable is for when it also needs to hit the stable kernels (which in
this case is indeed not needed).

So yeah I think it needs Cc: -fixes, for 4.8.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Chris Wilson]

On Tue, Aug 02, 2016 at 04:34:25PM +0200, Daniel Vetter wrote:
> On Fri, Jul 29, 2016 at 12:10:30PM +0100, Chris Wilson wrote:
> > On Fri, Jul 29, 2016 at 10:57:49AM +0200, Daniel Vetter wrote:
> > > On Fri, Jul 29, 2016 at 11:42:24AM +0300, Joonas Lahtinen wrote:
> > > > On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> > > > > It appears that we never told Pineview it has a RENDER_RING. This was
> > > > > all fine until we started using the ring_mask for determining all the
> > > > > available rings to initialise for legacy ringbuffer submission in commit
> > > > > 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> > > > > latent bug since the ring_mask inception in commit 73ae478cdf6a
> > > > > ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> > > > > 
> > > > > To prevent similar mishaps in future, add a WARN_ON() if we find
> > > > > ourselves with a device without any rings.
> > > > > 
> > > > > Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> > > > > Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> > > > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > > > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > > > > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > Cc: Ben Widawsky <ben@bwidawsk.net>
> > > > > ---
> > > > >  drivers/gpu/drm/i915/i915_pci.c        | 1 +
> > > > >  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
> > > > >  2 files changed, 2 insertions(+)
> > > > > 
> > > > > diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> > > > > index 949c01686a66..2587b1bd41f4 100644
> > > > > --- a/drivers/gpu/drm/i915/i915_pci.c
> > > > > +++ b/drivers/gpu/drm/i915/i915_pci.c
> > > > > @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
> > > > >  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
> > > > >  	.need_gfx_hws = 1, .has_hotplug = 1,
> > > > >  	.has_overlay = 1,
> > > > > +	.ring_mask = RENDER_RING,
> > > > >  	GEN_DEFAULT_PIPEOFFSETS,
> > > > >  	CURSOR_OFFSETS,
> > > > >  };
> > > > > diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > > index 31d43dfa7469..e6422aac2919 100644
> > > > > --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > > +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > > @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
> > > > >  	unsigned int i;
> > > > >  	int ret;
> > > > >  
> > > > > +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
> > > > 
> > > > Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> > > 
> > > dim says this also needs a
> > > 
> > > Cc: drm-intel-fixes@lists.freedesktop.org
> > 
> > dim says nothing of the sort for me. The bug is only relevant from
> > 88d2ba2e95c8
> 
> $ dim fixes 88d2ba2e95c8

What state does dim fixes depend upon? If did not report anything when I
ran it against the patch before pushing.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Add missing ring_mask to Pineview

[Daniel Vetter]

On Tue, Aug 02, 2016 at 03:47:29PM +0100, Chris Wilson wrote:
> On Tue, Aug 02, 2016 at 04:34:25PM +0200, Daniel Vetter wrote:
> > On Fri, Jul 29, 2016 at 12:10:30PM +0100, Chris Wilson wrote:
> > > On Fri, Jul 29, 2016 at 10:57:49AM +0200, Daniel Vetter wrote:
> > > > On Fri, Jul 29, 2016 at 11:42:24AM +0300, Joonas Lahtinen wrote:
> > > > > On pe, 2016-07-29 at 00:45 +0100, Chris Wilson wrote:
> > > > > > It appears that we never told Pineview it has a RENDER_RING. This was
> > > > > > all fine until we started using the ring_mask for determining all the
> > > > > > available rings to initialise for legacy ringbuffer submission in commit
> > > > > > 88d2ba2e95c8 ("drm/i915: Unify engine init loop"). Though really it is a
> > > > > > latent bug since the ring_mask inception in commit 73ae478cdf6a
> > > > > > ("drm/i915: Replace has_bsd/blt/vebox with a mask").
> > > > > > 
> > > > > > To prevent similar mishaps in future, add a WARN_ON() if we find
> > > > > > ourselves with a device without any rings.
> > > > > > 
> > > > > > Fixes: 73ae478cdf6a ("drm/i915: Replace has_bsd/blt/vebox with a mask")
> > > > > > Fixes: 88d2ba2e95c8 ("drm/i915: Unify engine init loop")
> > > > > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > > > > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > > > > > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > > Cc: Ben Widawsky <ben@bwidawsk.net>
> > > > > > ---
> > > > > >  drivers/gpu/drm/i915/i915_pci.c        | 1 +
> > > > > >  drivers/gpu/drm/i915/intel_engine_cs.c | 1 +
> > > > > >  2 files changed, 2 insertions(+)
> > > > > > 
> > > > > > diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> > > > > > index 949c01686a66..2587b1bd41f4 100644
> > > > > > --- a/drivers/gpu/drm/i915/i915_pci.c
> > > > > > +++ b/drivers/gpu/drm/i915/i915_pci.c
> > > > > > @@ -173,6 +173,7 @@ static const struct intel_device_info intel_pineview_info = {
> > > > > >  	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
> > > > > >  	.need_gfx_hws = 1, .has_hotplug = 1,
> > > > > >  	.has_overlay = 1,
> > > > > > +	.ring_mask = RENDER_RING,
> > > > > >  	GEN_DEFAULT_PIPEOFFSETS,
> > > > > >  	CURSOR_OFFSETS,
> > > > > >  };
> > > > > > diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > > > index 31d43dfa7469..e6422aac2919 100644
> > > > > > --- a/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > > > +++ b/drivers/gpu/drm/i915/intel_engine_cs.c
> > > > > > @@ -117,6 +117,7 @@ int intel_engines_init(struct drm_device *dev)
> > > > > >  	unsigned int i;
> > > > > >  	int ret;
> > > > > >  
> > > > > > +	WARN_ON(INTEL_INFO(dev_priv)->ring_mask == 0);
> > > > > 
> > > > > Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> > > > 
> > > > dim says this also needs a
> > > > 
> > > > Cc: drm-intel-fixes@lists.freedesktop.org
> > > 
> > > dim says nothing of the sort for me. The bug is only relevant from
> > > 88d2ba2e95c8
> > 
> > $ dim fixes 88d2ba2e95c8
> 
> What state does dim fixes depend upon? If did not report anything when I
> ran it against the patch before pushing.

3 cases for where the offending commit is:
- It's in a release tag -> cc: stable
- It's in an -rc tag -> cc: fixes
- It tries to make an educate guess for when stuff needs to be
  cherry-picked to dinf.

It's not entirely correct. tbh I'm not sure whether dim fixes is a good
idea, or whether we should just pimp the script to figure this out at
cherry-pick time. This is kinda still open and Jani&me are trying to
figure out what is best. Suggestions welcome.

In the end I simply did another scan for Fixes: when cherry-picking to
make sure I didn't miss anything.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx