[Intel-gfx] [PATCH 25/26] drm/i915: Round to closest in M/N calculations

[Jani Nikula]

On Tue, 03 May 2022, Ville Syrjala <ville.syrjala at linux.intel.com> wrote:
> From: Ville Syrjälä <ville.syrjala at linux.intel.com>
>
> Rounding to nearest is what we do for most other clock calculations
> so should probably do that for M/N too.
>
> TODO: GOP seems to truncate instead so fastboot is going to be
> a PITA to get right. Not sure what to do about it yet.

Meh. Damned if you do, damned if you don't.

Acked-by: Jani Nikula <jani.nikula at intel.com>

>
> Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> ---
>  drivers/gpu/drm/i915/display/intel_display.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> index 89a7c8c1be28..c4257630a3fe 100644
> --- a/drivers/gpu/drm/i915/display/intel_display.c
> +++ b/drivers/gpu/drm/i915/display/intel_display.c
> @@ -2856,7 +2856,7 @@ static void compute_m_n(u32 *ret_m, u32 *ret_n,
>  	else
>  		*ret_n = min_t(unsigned int, roundup_pow_of_two(n), DATA_LINK_N_MAX);
>  
> -	*ret_m = div_u64(mul_u32_u32(m, *ret_n), n);
> +	*ret_m = DIV_ROUND_CLOSEST_ULL(mul_u32_u32(m, *ret_n), n);
>  	intel_reduce_m_n_ratio(ret_m, ret_n);
>  }
>  
> @@ -4602,7 +4602,8 @@ int intel_dotclock_calculate(int link_freq,
>  	if (!m_n->link_n)
>  		return 0;
>  
> -	return div_u64(mul_u32_u32(m_n->link_m, link_freq), m_n->link_n);
> +	return DIV_ROUND_CLOSEST_ULL(mul_u32_u32(m_n->link_m, link_freq),
> +				     m_n->link_n);
>  }
>  
>  /* Returns the currently programmed mode of the given encoder. */

-- 
Jani Nikula, Intel Open Source Graphics Center