Re: [PATCH 1/2] drm/i915: Fix modeset handling during gpu reset, v2.

["Ville Syrjälä" <ville.syrjala@linux.intel.com>]

On Mon, May 02, 2016 at 10:57:01AM +0200, Maarten Lankhorst wrote:
> This function would call drm_modeset_lock_all, while the suspend/resume
> functions already have their own locking. Fix this by factoring out
> __intel_display_resume, and calling the atomic helpers for duplicating
> atomic state and disabling all crtc's during suspend.
> 
> Changes since v1:
> - Deal with -EDEADLK right after lock_all and clean up calls
>   to hw readout.
> - Always take all modeset locks so updates during gpu reset are blocked.

Found this patch by accident. --in-reply-to would have helped a bit.

> 
> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
> Fixes: e2c8b8701e2d ("drm/i915: Use atomic helpers for suspend, v2.")
> Cc: drm-intel-fixes@lists.freedesktop.org
> ---
>  drivers/gpu/drm/i915/intel_display.c | 141 ++++++++++++++++++++++-------------
>  1 file changed, 89 insertions(+), 52 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index 2f36414702fe..4fb904f2bcf9 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -3144,27 +3144,83 @@ static void intel_update_primary_planes(struct drm_device *dev)
>  	}
>  }
>  
> +static int
> +__intel_display_resume(struct drm_device *dev,
> +		       struct drm_atomic_state *state)
> +{
> +	struct drm_crtc_state *crtc_state;
> +	struct drm_crtc *crtc;
> +	int i;
> +
> +	intel_modeset_setup_hw_state(dev);
> +	i915_redisable_vga(dev);
> +
> +	if (!state)
> +		return 0;
> +
> +	for_each_crtc_in_state(state, crtc, crtc_state, i) {
> +		/*
> +		 * Force recalculation even if we restore
> +		 * current state. With fast modeset this may not result
> +		 * in a modeset when the state is compatible.
> +		 */
> +		crtc_state->mode_changed = true;
> +	}
> +
> +	return drm_atomic_commit(state);

The -EDEADLK warn could be here so we don't have to duplicate it in two
places perhaps? Extracting __intel_display_reset() could also be a separate
patch to make this stuff a bit easier to review.

Oh and BTW resume is also broken on platforms that have the force pipe A
quirk. I do have some patches lined up to nuke that quirk for good, which
I should probably post sooner rather than later. But those have at least
a theoretical chance of regressing something, so in the meantime I think
we'll still need to fix this thing for the normal resume path as well.

> +}
> +
>  void intel_prepare_reset(struct drm_device *dev)
>  {
> +	struct drm_atomic_state *state;
> +	struct drm_modeset_acquire_ctx *pctx;
> +	struct drm_i915_private *dev_priv = to_i915(dev);
> +	int ret;
> +
>  	/* no reset support for gen2 */
> -	if (IS_GEN2(dev))
> +	if (IS_GEN2(dev_priv))
>  		return;
>  
> -	/* reset doesn't touch the display */
> -	if (INTEL_INFO(dev)->gen >= 5 || IS_G4X(dev))
> +	drm_modeset_lock_all(dev);

Isn't that going to clash with the locking in
intel_update_primary_planes() ?

> +
> +	/* reset doesn't touch the display, but flips might get nuked anyway, */
> +	if (INTEL_INFO(dev_priv)->gen >= 5 || IS_G4X(dev_priv))
>  		return;
>  
> -	drm_modeset_lock_all(dev);
> +	pctx = dev->mode_config.acquire_ctx;

Still looks like power context.

> +
>  	/*
>  	 * Disabling the crtcs gracefully seems nicer. Also the
>  	 * g33 docs say we should at least disable all the planes.
>  	 */
> -	intel_display_suspend(dev);
> +
> +	state = drm_atomic_helper_duplicate_state(dev, pctx);
> +	if (IS_ERR(state)) {
> +		ret = PTR_ERR(state);
> +		state = NULL;
> +		DRM_ERROR("Duplicating state failed with %i\n", ret);
> +		goto err;
> +	}
> +
> +	ret = drm_atomic_helper_disable_all(dev, pctx);
> +	if (ret) {
> +		DRM_ERROR("Suspending crtc's failed with %i\n", ret);
> +		goto err;
> +	}
> +
> +	dev_priv->modeset_restore_state = state;
> +	state->acquire_ctx = pctx;
> +	return;
> +
> +err:
> +	drm_atomic_state_free(state);
>  }
>  
>  void intel_finish_reset(struct drm_device *dev)
>  {
>  	struct drm_i915_private *dev_priv = to_i915(dev);
> +	struct drm_atomic_state *state = dev_priv->modeset_restore_state;
> +	int ret;
>  
>  	/*
>  	 * Flips in the rings will be nuked by the reset,
> @@ -3177,6 +3233,8 @@ void intel_finish_reset(struct drm_device *dev)
>  	if (IS_GEN2(dev))
>  		return;
>  
> +	dev_priv->modeset_restore_state = NULL;
> +
>  	/* reset doesn't touch the display */
>  	if (INTEL_INFO(dev)->gen >= 5 || IS_G4X(dev)) {
>  		/*
> @@ -3189,26 +3247,27 @@ void intel_finish_reset(struct drm_device *dev)
>  		 * CS-based flips (which might get lost in gpu resets) any more.
>  		 */
>  		intel_update_primary_planes(dev);
> -		return;
> -	}
> -
> -	/*
> -	 * The display has been reset as well,
> -	 * so need a full re-initialization.
> -	 */
> -	intel_runtime_pm_disable_interrupts(dev_priv);
> -	intel_runtime_pm_enable_interrupts(dev_priv);
> +	} else {
> +		/*
> +		 * The display has been reset as well,
> +		 * so need a full re-initialization.
> +		 */
> +		intel_runtime_pm_disable_interrupts(dev_priv);
> +		intel_runtime_pm_enable_interrupts(dev_priv);
>  
> -	intel_modeset_init_hw(dev);
> +		intel_modeset_init_hw(dev);
>  
> -	spin_lock_irq(&dev_priv->irq_lock);
> -	if (dev_priv->display.hpd_irq_setup)
> -		dev_priv->display.hpd_irq_setup(dev);
> -	spin_unlock_irq(&dev_priv->irq_lock);
> +		spin_lock_irq(&dev_priv->irq_lock);
> +		if (dev_priv->display.hpd_irq_setup)
> +			dev_priv->display.hpd_irq_setup(dev);
> +		spin_unlock_irq(&dev_priv->irq_lock);
>  
> -	intel_display_resume(dev);
> +		ret = __intel_display_resume(dev, state);
> +		if (ret)
> +			DRM_ERROR("Restoring old state failed with %i\n", ret);
>  
> -	intel_hpd_init(dev_priv);
> +		intel_hpd_init(dev_priv);
> +	}
>  
>  	drm_modeset_unlock_all(dev);
>  }
> @@ -15957,9 +16016,10 @@ void intel_display_resume(struct drm_device *dev)
>  	struct drm_atomic_state *state = dev_priv->modeset_restore_state;
>  	struct drm_modeset_acquire_ctx ctx;
>  	int ret;
> -	bool setup = false;
>  
>  	dev_priv->modeset_restore_state = NULL;
> +	if (state)
> +		state->acquire_ctx = &ctx;
>  
>  	/*
>  	 * This is a cludge because with real atomic modeset mode_config.mutex
> @@ -15970,40 +16030,17 @@ void intel_display_resume(struct drm_device *dev)
>  	mutex_lock(&dev->mode_config.mutex);
>  	drm_modeset_acquire_init(&ctx, 0);
>  
> -retry:
> -	ret = drm_modeset_lock_all_ctx(dev, &ctx);
> -
> -	if (ret == 0 && !setup) {
> -		setup = true;
> -
> -		intel_modeset_setup_hw_state(dev);
> -		i915_redisable_vga(dev);
> -	}
> -
> -	if (ret == 0 && state) {
> -		struct drm_crtc_state *crtc_state;
> -		struct drm_crtc *crtc;
> -		int i;
> -
> -		state->acquire_ctx = &ctx;
> -
> -		for_each_crtc_in_state(state, crtc, crtc_state, i) {
> -			/*
> -			 * Force recalculation even if we restore
> -			 * current state. With fast modeset this may not result
> -			 * in a modeset when the state is compatible.
> -			 */
> -			crtc_state->mode_changed = true;
> -		}
> -
> -		ret = drm_atomic_commit(state);
> -	}
> +	while (1) {
> +		ret = drm_modeset_lock_all_ctx(dev, &ctx);
> +		if (ret != -EDEADLK)
> +			break;
>  
> -	if (ret == -EDEADLK) {
>  		drm_modeset_backoff(&ctx);
> -		goto retry;
>  	}
>  
> +	ret = __intel_display_resume(dev, state);

Shouldn't we skip this call if the lock_all failed?

> +	WARN_ON(ret == -EDEADLK);
> +
>  	drm_modeset_drop_locks(&ctx);
>  	drm_modeset_acquire_fini(&ctx);
>  	mutex_unlock(&dev->mode_config.mutex);
> -- 
> 2.5.5
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx

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