Re: [PATCH 09/25] drm/i915: enable power wells on Haswell init

[Daniel Vetter <daniel@ffwll.ch>]

On Wed, May 09, 2012 at 03:37:16PM -0300, Eugeni Dodonov wrote:
> This attempts to enable all the available power wells during the
> initialization.
> 
> Those power wells can be enabled in parallel or on-demand, and disabled
> when no longer needed, but this is out of scope of this initial
> enablement. Proper tracking of who uses which power well will require
> a considerable rework of our display handling, so we just leave them all
> enabled when the driver is loaded for now.
> 
> v2: use more generic and future-proof code
> 
> Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_pm.c |   36 ++++++++++++++++++++++++++++++++++++
>  1 file changed, 36 insertions(+)
> 
> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> index 21587f8..bd9549d 100644
> --- a/drivers/gpu/drm/i915/intel_pm.c
> +++ b/drivers/gpu/drm/i915/intel_pm.c
> @@ -3559,6 +3559,37 @@ void intel_sanitize_pm(struct drm_device *dev)
>  		dev_priv->display.sanitize_pm(dev);
>  }
>  
> +/* Starting with Haswell, we have different power wells for
> + * different parts of the GPU. This attempts to enable them all.
> + */
> +void intel_init_power_wells(struct drm_device *dev)
> +{
> +	struct drm_i915_private *dev_priv = dev->dev_private;
> +	unsigned long power_wells[] = {
> +		HSW_PWR_WELL_CTL1,
> +		HSW_PWR_WELL_CTL2,
> +		HSW_PWR_WELL_CTL4
> +	};
> +	int i;
> +
> +	if (!IS_HASWELL(dev))
> +		return;
> +
> +	mutex_lock(&dev->struct_mutex);
> +
> +	for (i = 0; i < ARRAY_SIZE(power_wells); i++) {
> +		int well = I915_READ(power_wells[i]);
> +
> +		if ((well & HSW_PWR_WELL_STATE) == 0) {
> +			I915_WRITE(power_wells[i], well & HSW_PWR_WELL_ENABLE);
> +			if (wait_for(I915_READ(power_wells[i] & HSW_PWR_WELL_STATE), 20))
> +				DRM_ERROR("Error enabling power well %lx\n", power_wells[i]);
> +		}
> +	}
> +
> +	mutex_unlock(&dev->struct_mutex);
> +}
> +
>  /* Set up chip specific power management-related functions */
>  void intel_init_pm(struct drm_device *dev)
>  {
> @@ -3707,5 +3738,10 @@ void intel_init_pm(struct drm_device *dev)
>  		else
>  			dev_priv->display.get_fifo_size = i830_get_fifo_size;
>  	}
> +
> +	/* We attempt to init the necessary power wells early in the initialization
> +	 * time, so the subsystems that expect power to be enabled can work.
> +	 */
> +	intel_init_power_wells(dev);

Setting up hw registers in init_pm is a no-go, this function is not called
on resume (and after hw reset). Also note that I've just merged a patch
from Chris that moves around our hw init stuff a bit (it's now consistent
between driver load, gpu reset and resume). So if this needs to run before
we can set up the rings, it'll fail on latest dinq.

-Daniel
-- 
Daniel Vetter
Mail: daniel@ffwll.ch
Mobile: +41 (0)79 365 57 48