Re: [PATCH] drm/i915: compute the target_clock for edp directly

[Chris Wilson <chris@chris-wilson.co.uk>]

On Wed, 30 May 2012 15:34:20 +0200, Daniel Vetter <daniel.vetter@ffwll.ch> wrote:
> Instead of abusing into mode->clock, because we should touch that
> one at all. First prep step to constify the mode argument to the
> intel_dp_mode_fixup function.
> 
> The next patch will stop us from modifying mode->clock.
> 
> Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> ---
>  drivers/gpu/drm/i915/intel_display.c |   16 ++++++++--------
>  drivers/gpu/drm/i915/intel_dp.c      |   12 ++++++++++++
>  drivers/gpu/drm/i915/intel_drv.h     |    2 ++
>  3 files changed, 22 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index 9147894..a5d06ed 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -4431,16 +4431,8 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
>  	/* CPU eDP doesn't require FDI link, so just set DP M/N
>  	   according to current link config */
>  	if (is_cpu_edp) {
> -		target_clock = mode->clock;
>  		intel_edp_link_config(edp_encoder, &lane, &link_bw);
>  	} else {
> -		/* [e]DP over FDI requires target mode clock
> -		   instead of link clock */
> -		if (is_dp)
> -			target_clock = mode->clock;
> -		else
> -			target_clock = adjusted_mode->clock;
> -
>  		/* FDI is a binary signal running at ~2.7GHz, encoding
>  		 * each output octet as 10 bits. The actual frequency
>  		 * is stored as a divider into a 100MHz clock, and the
> @@ -4451,6 +4443,14 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
>  		link_bw = intel_fdi_link_freq(dev) * MHz(100)/KHz(1)/10;
>  	}
>  
> +	/* [e]DP over FDI requires target mode clock instead of link clock. */
> +	if (edp_encoder)
> +		target_clock = intel_edp_target_clock(edp_encoder, mode);

Given the edp_encoder, we know that adjusted_mode contains either the
fixed_panel clock or the original clock depending on whether a fixed
mode was found. So the layering violation could be dropped here in
favour of target_clock = adjusted_mode->clok.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre