Re: [Intel-gfx] [PATCH] drm/i915: fix cursor handling when runtime suspended

[Daniel Vetter <daniel@ffwll.ch>]

On Wed, Jul 23, 2014 at 06:30:59PM -0300, Paulo Zanoni wrote:
> From: Paulo Zanoni <paulo.r.zanoni@intel.com>
> 
> If we're runtime suspended and try to use the cursor interfaces, we
> will get a lot of WARNs saying we did the wrong thing.
> 
> For intel_crtc_update_cursor(), all we need to do is return if the
> CRTC is not active, since writing the registers won't really have any
> effect if the screen is not visible, and we will write the registers
> later when enabling the screen.
> 
> For intel_crtc_cursor_set_obj(), we just get the proper power domain
> reference, since this function does a lot of stuff.
> 
> Testcase: igt/pm_rpm/cursor
> Testcase: igt/pm_rpm/cursor-dpms
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=81645
> Cc: stable@vger.kernel.org
> Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_display.c | 15 ++++++++++++---
>  1 file changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index d1e9570..c8f36b0 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -8151,6 +8151,9 @@ static void intel_crtc_update_cursor(struct drm_crtc *crtc,
>  	if (base == 0 && intel_crtc->cursor_base == 0)
>  		return;
>  
> +	if (!intel_crtc->active)
> +		return;

Don't we need the same trick in intel_crtc_cursor_set_obj? This gets
called if the cursor object changes (instead of just moving it around).
> +
>  	I915_WRITE(CURPOS(pipe), pos);
>  
>  	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev))
> @@ -8181,6 +8184,8 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
>  	uint32_t addr;
>  	int ret;
>  
> +	intel_display_power_get(dev_priv, POWER_DOMAIN_PIPE(pipe));

This shouldn't be needed - if intel_crtc->active is set, we do hold a
reference to the pipe power domain already. There's no way the pipe would
be able to run otherwise ;-)
-Daniel

> +
>  	/* if we want to turn off the cursor ignore width and height */
>  	if (!obj) {
>  		DRM_DEBUG_KMS("cursor off\n");
> @@ -8195,13 +8200,14 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
>  			(width == 128 && height == 128 && !IS_GEN2(dev)) ||
>  			(width == 256 && height == 256 && !IS_GEN2(dev)))) {
>  		DRM_DEBUG("Cursor dimension not supported\n");
> -		return -EINVAL;
> +		ret = -EINVAL;
> +		goto fail;
>  	}
>  
>  	if (obj->base.size < width * height * 4) {
>  		DRM_DEBUG_KMS("buffer is too small\n");
>  		ret = -ENOMEM;
> -		goto fail;
> +		goto fail_unref;
>  	}
>  
>  	/* we only need to pin inside GTT if cursor is non-phy */
> @@ -8275,13 +8281,16 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
>  
>  	intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_CURSOR(pipe));
>  
> +	intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE(pipe));
>  	return 0;
>  fail_unpin:
>  	i915_gem_object_unpin_from_display_plane(obj);
>  fail_locked:
>  	mutex_unlock(&dev->struct_mutex);
> -fail:
> +fail_unref:
>  	drm_gem_object_unreference_unlocked(&obj->base);
> +fail:
> +	intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE(pipe));
>  	return ret;
>  }
>  
> -- 
> 2.0.1
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch