[PATCH 1/2] drm/i915: Fix Sink CRC

[Daniel Vetter]

On Mon, Sep 15, 2014 at 07:21:50PM -0400, Rodrigo Vivi wrote:
> In some cases like when PSR just got enabled the panel need more vblank
> times to calculate CRC. I figured that out with the new PSR test cases
> facing some cases that I had a green screen but a blank CRC. Even with
> 2 vblank waits on kernel + 2 vblank waits on test case.
> 
> So let's give up to 6 vblank wait time. However we now check for
> TEST_CRC_COUNT that shows when panel finished to calculate CRC and
> has it ready.
> 
> v2: Jani pointed out attempts decrements was wrong and should never reach
> the error condition. And Daniel pointed out that EIO is more appropriated than
> EGAIN. Also I realized that I have to read test_crc_count after setting
> test_sink
> 
> Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
> Cc: Jani Nikula <jani.nikula at intel.com>
> Signed-off-by: Rodrigo Vivi <rodrigo.vivi at intel.com>
> ---
>  drivers/gpu/drm/i915/intel_dp.c | 21 +++++++++++++++------
>  include/drm/drm_dp_helper.h     |  5 +++--
>  2 files changed, 18 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index f79473b..fae0fae 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -3468,21 +3468,30 @@ int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc)
>  	struct drm_device *dev = intel_dig_port->base.base.dev;
>  	struct intel_crtc *intel_crtc =
>  		to_intel_crtc(intel_dig_port->base.base.crtc);
> -	u8 buf[1];
> +	u8 buf;
> +	int test_crc_count;
> +	int attempts = 6;
>  
> -	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, buf) < 0)
> +	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, &buf) < 0)
>  		return -EAGAIN;
>  
> -	if (!(buf[0] & DP_TEST_CRC_SUPPORTED))
> +	if (!(buf & DP_TEST_CRC_SUPPORTED))
>  		return -ENOTTY;
>  
>  	if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK,
>  			       DP_TEST_SINK_START) < 0)
>  		return -EAGAIN;
>  
> -	/* Wait 2 vblanks to be sure we will have the correct CRC value */
> -	intel_wait_for_vblank(dev, intel_crtc->pipe);
> -	intel_wait_for_vblank(dev, intel_crtc->pipe);
> +	drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, &buf);
> +	test_crc_count = buf & DP_TEST_COUNT_MASK;
> +
> +	do {
> +		drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, &buf);
> +		intel_wait_for_vblank(dev, intel_crtc->pipe);
> +	} while (--attempts && (buf & DP_TEST_COUNT_MASK) == test_crc_count);
> +
> +	if (attempts == 0)
> +		return -EIO;

I still think that some debug output here would be useful - after all
we've seen this happen now with your panel, so it could happen at other
places too. If there's nothing else I'll do that when applying.
-Daniel

>  
>  	if (drm_dp_dpcd_read(&intel_dp->aux, DP_TEST_CRC_R_CR, crc, 6) < 0)
>  		return -EAGAIN;
> diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h
> index 9305c71..8edeed0 100644
> --- a/include/drm/drm_dp_helper.h
> +++ b/include/drm/drm_dp_helper.h
> @@ -303,7 +303,8 @@
>  #define DP_TEST_CRC_B_CB		    0x244
>  
>  #define DP_TEST_SINK_MISC		    0x246
> -#define DP_TEST_CRC_SUPPORTED		    (1 << 5)
> +# define DP_TEST_CRC_SUPPORTED		    (1 << 5)
> +# define DP_TEST_COUNT_MASK		    0x7
>  
>  #define DP_TEST_RESPONSE		    0x260
>  # define DP_TEST_ACK			    (1 << 0)
> @@ -313,7 +314,7 @@
>  #define DP_TEST_EDID_CHECKSUM		    0x261
>  
>  #define DP_TEST_SINK			    0x270
> -#define DP_TEST_SINK_START	    (1 << 0)
> +# define DP_TEST_SINK_START		    (1 << 0)
>  
>  #define DP_PAYLOAD_TABLE_UPDATE_STATUS      0x2c0   /* 1.2 MST */
>  # define DP_PAYLOAD_TABLE_UPDATED           (1 << 0)
> -- 
> 1.9.3
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch