[Intel-gfx] [PATCH] drm/i915: Forcewake Register Range changes for CHV

[Ville Syrjälä]

On Thu, Dec 11, 2014 at 12:48:41PM +0530, deepak.s at linux.intel.com wrote:
> From: Deepak S <deepak.s at linux.intel.com>
> 
> According to updated BSpec, Render/Common Wells register range changed.
> Updating the same to match the spec and avoid extra forcewake for none
> forcewake range.
> 
> Signed-off-by: Deepak S <deepak.s at linux.intel.com>
> ---
>  drivers/gpu/drm/i915/intel_uncore.c | 9 +++------
>  1 file changed, 3 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
> index 46de8d7..dd36f9b 100644
> --- a/drivers/gpu/drm/i915/intel_uncore.c
> +++ b/drivers/gpu/drm/i915/intel_uncore.c
> @@ -647,9 +647,9 @@ void assert_force_wake_inactive(struct drm_i915_private *dev_priv)
>  
>  #define FORCEWAKE_CHV_RENDER_RANGE_OFFSET(reg) \
>  	(REG_RANGE((reg), 0x2000, 0x4000) || \
> -	 REG_RANGE((reg), 0x5000, 0x8000) || \
> +	 REG_RANGE((reg), 0x5200, 0x8000) || \
>  	 REG_RANGE((reg), 0x8300, 0x8500) || \
> -	 REG_RANGE((reg), 0xB000, 0xC000) || \
> +	 REG_RANGE((reg), 0xB000, 0xB480) || \
>  	 REG_RANGE((reg), 0xE000, 0xE800))
>  
>  #define FORCEWAKE_CHV_MEDIA_RANGE_OFFSET(reg) \
> @@ -665,10 +665,7 @@ void assert_force_wake_inactive(struct drm_i915_private *dev_priv)
>  	 REG_RANGE((reg), 0x8000, 0x8300) || \
>  	 REG_RANGE((reg), 0x8500, 0x8600) || \
>  	 REG_RANGE((reg), 0x9000, 0xB000) || \
> -	 REG_RANGE((reg), 0xC000, 0xC800) || \
> -	 REG_RANGE((reg), 0xF000, 0x10000) || \
> -	 REG_RANGE((reg), 0x14000, 0x14400) || \
> -	 REG_RANGE((reg), 0x22000, 0x24000))
> +	 REG_RANGE((reg), 0xF000, 0x10000))

Looks correct.

It looks like the media offsets could use a small adjustment as well:
- REG_RANGE((reg), 0x30000, 0x40000)
+ REG_RANGE((reg), 0x30000, 0x38000)

So with that changed 
Reviewed-by: Ville Syrjälä <ville.syrjala at linux.intel.com>

One thing I don't understand though; Why is the [0x100000,0x180000[ range
marked as GWK in the spreadsheet? That's where the forcewake req/ack
registers live as well as some other GTFIFO stuff etc. How can you take
forcewake if the forcewake req/ack itself would need forcewake. I think
this must be an error in the spreadsheet and that range should be GNW.

>  
>  #define FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg) \
>  	REG_RANGE((reg), 0xB00,  0x2000)
> -- 
> 1.9.1

-- 
Ville Syrjälä
Intel OTC