Re: [PATCH 3/3] drm/i915: go back to switch for VLV mem freq detection

[Ben Widawsky <ben@bwidawsk.net>]

On Wed, May 01, 2013 at 04:09:22PM -0700, Jesse Barnes wrote:
> We need to catch the invalid case and override it.
> 
> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
> ---
>  drivers/gpu/drm/i915/intel_pm.c |   16 +++++++++++++++-
>  1 file changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> index 0f4b46e..0a3b0b3 100644
> --- a/drivers/gpu/drm/i915/intel_pm.c
> +++ b/drivers/gpu/drm/i915/intel_pm.c
> @@ -2902,7 +2902,21 @@ static void valleyview_enable_rps(struct drm_device *dev)
>  		   GEN7_RC_CTL_TO_MODE);
>  
>  	valleyview_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS, &val);
> -	dev_priv->mem_freq = 800 + (266 * (val >> 6) & 3);
> +	switch ((val >> 6) & 3) {
> +	case 0:
> +		dev_priv->mem_freq = 800;
> +		break;
> +	case 1:
> +		dev_priv->mem_freq = 1066;
> +		break;
> +	case 2:
> +		dev_priv->mem_freq = 1333;
> +		break;
> +	case 3:
> +		DRM_ERROR("invalid mem freq, assuming 800MHz\n");
> +		dev_priv->mem_freq = 800;
> +		break;
> +	}
>  	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
>  
>  	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no");


I guess it doesn't handle the last case, but:
dev_priv->mem_freq = 800 + (266 * (val))) + val/2

or

u32 freqs[] = {800,1066,1333,800};
dev_priv->mem_freq = freqs[(val >> 6) & 3];

Are two ways I would have used before making a switch block :P

Just a thought, but perhaps we don't want to enable RPS if we can't
reliably figure out the memory freq. (case 3)

Reviewed-by: Ben Widawsky <ben@bwidawsk.net>

-- 
Ben Widawsky, Intel Open Source Technology Center