Re: [PATCH] drm/i915/bdw: Cleanup pre prod workarounds

From: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
To: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH] drm/i915/bdw: Cleanup pre prod workarounds
Date: Fri, 19 Sep 2014 16:55:56 +0300	[thread overview]
Message-ID: <20140919135556.GN12416@intel.com> (raw)
In-Reply-To: <1411131908-29771-1-git-send-email-mika.kuoppala@intel.com>

On Fri, Sep 19, 2014 at 04:05:08PM +0300, Mika Kuoppala wrote:
> as these have been fixed in production hw and hurt performance
> if applied.
> 
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83482
> Tested-by: zhoujian <jianx.zhou@intel.com>
> Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 13 +------------
>  1 file changed, 1 insertion(+), 12 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 681ea86..dfb3bc6 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -722,19 +722,8 @@ static int bdw_init_workarounds(struct intel_engine_cs *ring)
>  	intel_ring_emit_wa(ring, GEN7_ROW_CHICKEN2,
>  			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
>  
> -	/*
> -	 * This GEN8_CENTROID_PIXEL_OPT_DIS W/A is only needed for
> -	 * pre-production hardware
> -	 */
>  	intel_ring_emit_wa(ring, HALF_SLICE_CHICKEN3,
> -			   _MASKED_BIT_ENABLE(GEN8_CENTROID_PIXEL_OPT_DIS
> -					      | GEN8_SAMPLER_POWER_BYPASS_DIS));
> -
> -	intel_ring_emit_wa(ring, GEN7_HALF_SLICE_CHICKEN1,
> -			   _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE));
> -
> -	intel_ring_emit_wa(ring, COMMON_SLICE_CHICKEN2,
> -			   _MASKED_BIT_ENABLE(GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE));
> +			   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));

You should adjust the requested ring space too. Looks like this will
leave the number of intel_ring_emit_wa() calls even so no need to
worry about the QW tail padding quite yet.

>  
>  	/* Use Force Non-Coherent whenever executing a 3D context. This is a
>  	 * workaround for for a possible hang in the unlikely event a TLB
> -- 
> 1.9.1
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Ville Syrjälä
Intel OTC