Ack: [PATCH] [QUANTAL] Revert "drm/i915: GFX_MODE Flush TLB Invalidate Mode must be '1' for scanline waits"

[Brad Figg]

On 04/23/2013 07:56 AM, Steve Conklin wrote:
> This reverts commit 4c443ec9afe7f6fab41106c617136aca44d69a7b.
> 
> This commit has been shown to cause GPU hangs on Sandy Bridge GPUs
> ---
>  drivers/gpu/drm/i915/intel_ringbuffer.c |    5 -----
>  1 file changed, 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 83eaa1d..102b6a7 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -435,11 +435,6 @@ static int init_render_ring(struct intel_ring_buffer *ring)
>  	if (INTEL_INFO(dev)->gen >= 6)
>  		I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
>  
> -	/* Required for the hardware to program scanline values for waiting */
> -	if (INTEL_INFO(dev)->gen == 6)
> -		I915_WRITE(GFX_MODE,
> -			   _MASKED_BIT_ENABLE(GFX_TLB_INVALIDATE_ALWAYS));
> -
>  	if (IS_GEN7(dev))
>  		I915_WRITE(GFX_MODE_GEN7,
>  			   _MASKED_BIT_DISABLE(GFX_TLB_INVALIDATE_ALWAYS) |
> 

This is causing a lot of people grief right now.

-- 
Brad Figg brad.figg at canonical.com http://www.canonical.com