Re: [PATCH v3 08/10] drm/i915/display: Implement Wa_14021768792 for BMG DP for link_m/n ratio > 10

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] drm/i915/dp: state leak of bypass_m_n_ratio_limit disables DRRS
--

commit 0e4fb34450fd70c4ab5e5e995c5d7cab890c8cac
Author: Ankit Nautiyal <ankit.k.nautiyal@intel.com>

drm/i915/display: Implement Wa_14021768792 for BMG DP for link_m/n ratio > 10

Handle the bypass logic for the M/N ratio limit for DP. Calculate the
M/N ratio, check if it can bypass the limit, and set the appropriate
flags for the workaround to restrict DRRS and seamless M/N fastsets.

> diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> index 8de51582d18c8..3911942de8e53 100644
> --- a/drivers/gpu/drm/i915/display/intel_dp.c
> +++ b/drivers/gpu/drm/i915/display/intel_dp.c

[ ... ]

> @@ -3405,10 +3409,21 @@ void intel_dp_check_m_n_ratio(struct intel_crtc_state *crtc_state,
>  
>  	m_n_ratio = DIV_ROUND_UP(m_n->link_m, m_n->link_n);
>  
> -	if (m_n_ratio > max_m_n_ratio)
> -		drm_WARN(display->drm, 1,
> -			 "Link M/N ratio (%d) exceeds max allowed (%d)\n",
> -			 m_n_ratio, max_m_n_ratio);
> +	if (m_n_ratio <= max_m_n_ratio)
> +		return;

[Severity: Medium]
If a previous iteration or modeset evaluated an M/N ratio > 10 and set the
bypass_m_n_ratio_limit flag to true, does returning early here leave the flag
incorrectly set?

For example, when this function is called during the MST BPP fallback loop:

drivers/gpu/drm/i915/display/intel_dp_mst.c:intel_dp_mtp_tu_compute_config() {
    ...
		intel_dp_mst_compute_m_n(crtc_state,
					 local_bw_overhead,
					 link_bpp_x16,
					 &crtc_state->dp_m_n);

		intel_dp_check_m_n_ratio(crtc_state, &crtc_state->dp_m_n);
    ...
}

The crtc_state->dp_m_n struct is reused across fallback iterations. If a
rejected high-ratio configuration sets bypass_m_n_ratio_limit to true, it
appears a subsequent accepted lower-ratio configuration will return early
without clearing it. This would permanently disable DRRS for the CRTC.
Should we set m_n->bypass_m_n_ratio_limit = false before returning?

> +
> +	if (intel_dp_is_uhbr(crtc_state) &&
> +	    intel_display_can_bypass_m_n_limit(display, m_n_ratio, crtc->pipe)) {
> +		m_n->bypass_m_n_ratio_limit = true;
> +		drm_dbg_kms(display->drm,
> +			    "Bypassing Link_m/Link_n ratio limit (ratio %d)\n",
> +			    m_n_ratio);
> +		return;
> +	}
> +
> +	drm_WARN(display->drm, 1,
> +		 "Link M/N ratio (%d) exceeds max allowed (%d)\n",
> +		 m_n_ratio, max_m_n_ratio);
>  }

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260916090946.677564-1-suraj.kandpal@intel.com?part=8