public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Imre Deak <imre.deak@intel.com>
To: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Cc: <intel-gfx@lists.freedesktop.org>, <intel-xe@lists.freedesktop.org>
Subject: Re: [PATCH 05/14] drm/i915/dp: Rework pipe joiner logic in compute_config
Date: Fri, 23 Jan 2026 21:06:36 +0200	[thread overview]
Message-ID: <aXPGvBQ29dzanJfR@ideak-desk> (raw)
In-Reply-To: <20260121035330.2793386-6-ankit.k.nautiyal@intel.com>

On Wed, Jan 21, 2026 at 09:23:21AM +0530, Ankit Nautiyal wrote:
> Currently, the number of joined pipes are determined early in the flow,
> which limits flexibility for accounting DSC slice overhead. To address
> this, recompute the joined pipe count during DSC configuration.
> 
> Refactor intel_dp_dsc_compute_config() to iterate over joiner candidates
> and select the minimal joiner configuration that satisfies the mode
> requirements. This prepares the logic for future changes that will
> consider DSC slice overhead.
> 
> Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
> ---
>  drivers/gpu/drm/i915/display/intel_dp.c | 102 +++++++++++++++++++-----
>  1 file changed, 84 insertions(+), 18 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> index 02381f84fa58..d96d9ac1e830 100644
> --- a/drivers/gpu/drm/i915/display/intel_dp.c
> +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> @@ -2790,33 +2790,20 @@ bool intel_dp_joiner_needs_dsc(struct intel_display *display,
>  }
>  
>  static int
> -intel_dp_compute_link_config(struct intel_encoder *encoder,
> -			     struct intel_crtc_state *pipe_config,
> -			     struct drm_connector_state *conn_state,
> -			     bool respect_downstream_limits)
> +_intel_dp_compute_link_config(struct intel_encoder *encoder,

intel_dp_compute_link_for_joined_pipes(), similarly to the MST
counterpart?

> +			      struct intel_crtc_state *pipe_config,
> +			      struct drm_connector_state *conn_state,
> +			      bool respect_downstream_limits)
>  {
>  	struct intel_display *display = to_intel_display(encoder);
> -	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
> +	int num_joined_pipes = intel_crtc_num_joined_pipes(pipe_config);
>  	struct intel_connector *connector =
>  		to_intel_connector(conn_state->connector);
> -	const struct drm_display_mode *adjusted_mode =
> -		&pipe_config->hw.adjusted_mode;
>  	struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
>  	struct link_config_limits limits;
>  	bool dsc_needed, joiner_needs_dsc;
> -	int num_joined_pipes;
>  	int ret = 0;
>  
> -	if (pipe_config->fec_enable &&
> -	    !intel_dp_supports_fec(intel_dp, connector, pipe_config))
> -		return -EINVAL;
> -
> -	num_joined_pipes = intel_dp_num_joined_pipes(intel_dp, connector,
> -						     adjusted_mode->crtc_hdisplay,
> -						     adjusted_mode->crtc_clock);
> -	if (num_joined_pipes > 1)
> -		pipe_config->joiner_pipes = GENMASK(crtc->pipe + num_joined_pipes - 1, crtc->pipe);
> -
>  	joiner_needs_dsc = intel_dp_joiner_needs_dsc(display, num_joined_pipes);
>  
>  	dsc_needed = joiner_needs_dsc || intel_dp->force_dsc_en ||
> @@ -2879,6 +2866,85 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
>  	return 0;
>  }
>  
> +static int
> +intel_dp_compute_link_config(struct intel_encoder *encoder,
> +			     struct intel_crtc_state *crtc_state,
> +			     struct drm_connector_state *conn_state,
> +			     bool respect_downstream_limits)
> +{
> +	struct intel_display *display = to_intel_display(encoder);
> +	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
> +	struct intel_connector *connector =
> +		to_intel_connector(conn_state->connector);
> +	const struct drm_display_mode *adjusted_mode =
> +		&crtc_state->hw.adjusted_mode;
> +	struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
> +	int num_joined_pipes;
> +	int ret = 0;
> +	int i;
> +
> +	if (crtc_state->fec_enable &&
> +	    !intel_dp_supports_fec(intel_dp, connector, crtc_state))
> +		return -EINVAL;
> +
> +	for (i = 0; i < ARRAY_SIZE(joiner_candidates); i++) {
> +		enum joiner_type joiner = joiner_candidates[i];
> +		int max_dotclk = display->cdclk.max_dotclk_freq;
> +
> +		if (joiner == FORCED_JOINER) {
> +			if (!connector->force_joined_pipes)
> +				continue;
> +			num_joined_pipes = connector->force_joined_pipes;
> +		} else {
> +			num_joined_pipes = 1 << joiner;
> +		}
> +
> +		if ((joiner >= NO_JOINER && !intel_dp_has_joiner(intel_dp)) ||
> +		    (joiner == BIG_JOINER && !HAS_BIGJOINER(display)) ||
> +		    (joiner == ULTRA_JOINER && !HAS_ULTRAJOINER(display))) {
> +			ret = -EINVAL;
> +			break;

A simple loop over the number of possible joined pipes would be simpler
here as well, as in the mode_valid() function.

> +		}
> +
> +		if (adjusted_mode->hdisplay > num_joined_pipes * intel_dp_hdisplay_limit(display))
> +			continue;
> +
> +		/*
> +		 * NOTE:
> +		 * The crtc_state->joiner_pipes should have been set at the end
> +		 * only if all the conditions are met. However that would mean
> +		 * that num_joined_pipes is passed around to all helpers and
> +		 * make them use it instead of using crtc_state->joiner_pipes
> +		 * directly or indirectly (via intel_crtc_num_joined_pipes()).
> +		 *
> +		 * For now, setting crtc_state->joiner_pipes to the candidate
> +		 * value to avoid the above churn and resetting it to 0, in case
> +		 * no joiner candidate is found to be suitable for the given
> +		 * configuration.
> +		 */
> +		if (num_joined_pipes > 1)
> +			crtc_state->joiner_pipes = GENMASK(crtc->pipe + num_joined_pipes - 1,
> +							   crtc->pipe);
> +
> +		ret = _intel_dp_compute_link_config(encoder, crtc_state, conn_state,
> +						    respect_downstream_limits);
> +		if (ret)
> +			continue;
> +
> +		max_dotclk *= num_joined_pipes;
> +
> +		if (adjusted_mode->crtc_clock <= max_dotclk) {

There seems to be a difference in the maximum dotclock for the DSC and
non-DSC modes, introduced later in the patchset. Since the non-DSC max
dotclock may be lower than the DSC one (oddly), there should be a
fallback to DSC mode for a given number of joined pipes if the non-DSC
mode is not possible with the same number of joined pipes because of the
mode specific dot clock limit. So the above check should be moved to
intel_dp_compute_link_for_joined_pipes() and performed for both the DSC
and non-DSC modes.

> +			ret = 0;
> +			break;
> +		}
> +	}
> +
> +	if (ret < 0)
> +		crtc_state->joiner_pipes = 0;
> +
> +	return ret;
> +}
> +
>  bool intel_dp_limited_color_range(const struct intel_crtc_state *crtc_state,
>  				  const struct drm_connector_state *conn_state)
>  {
> -- 
> 2.45.2
> 

  reply	other threads:[~2026-01-23 19:06 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-21  3:53 [PATCH 00/14] Account for DSC bubble overhead for horizontal slices Ankit Nautiyal
2026-01-21  3:53 ` [PATCH 01/14] drm/i915/dp: Early reject bad hdisplay in intel_dp_mode_valid Ankit Nautiyal
2026-01-23 18:06   ` Imre Deak
2026-01-21  3:53 ` [PATCH 02/14] drm/i915/dp: Move num_joined_pipes and related checks together Ankit Nautiyal
2026-01-23 18:25   ` Imre Deak
2026-01-21  3:53 ` [PATCH 03/14] drm/i915/dp: Extract helper to get the hdisplay limit Ankit Nautiyal
2026-01-23 18:28   ` Imre Deak
2026-01-21  3:53 ` [PATCH 04/14] drm/i915/dp: Rework pipe joiner logic in mode_valid Ankit Nautiyal
2026-01-21 10:54   ` Jani Nikula
2026-01-21 11:19     ` Nautiyal, Ankit K
2026-01-23 18:44       ` Imre Deak
2026-01-21  3:53 ` [PATCH 05/14] drm/i915/dp: Rework pipe joiner logic in compute_config Ankit Nautiyal
2026-01-23 19:06   ` Imre Deak [this message]
2026-01-21  3:53 ` [PATCH 06/14] drm/i915/dp_mst: Move the check for dotclock at the end Ankit Nautiyal
2026-01-23 19:11   ` Imre Deak
2026-01-21  3:53 ` [PATCH 07/14] drm/i915/dp_mst: Move the joiner dependent code together Ankit Nautiyal
2026-01-23 19:16   ` Imre Deak
2026-01-21  3:53 ` [PATCH 08/14] drm/i915/dp_mst: Rework pipe joiner logic in mode_valid Ankit Nautiyal
2026-01-21  3:53 ` [PATCH 09/14] drm/i915/dp_mst: Extract helper to compute link for given joiner config Ankit Nautiyal
2026-01-23 19:23   ` Imre Deak
2026-01-21  3:53 ` [PATCH 10/14] drm/i915/dp_mst: Rework pipe joiner logic in compute_config Ankit Nautiyal
2026-01-21  3:53 ` [PATCH 11/14] drm/i915/dp: Introduce helper to check pixel rate against dotclock limits Ankit Nautiyal
2026-01-23 19:30   ` Imre Deak
2026-01-21  3:53 ` [PATCH 12/14] drm/i915/dp: Refactor dsc_slice_count handling in intel_dp_mode_valid() Ankit Nautiyal
2026-01-21  3:53 ` [PATCH 13/14] drm/i915/dp: Account for DSC slice overhead Ankit Nautiyal
2026-01-21  3:53 ` [PATCH 14/14] drm/i915/display: Add upper limit check for pixel clock Ankit Nautiyal
2026-01-23 19:48   ` Imre Deak
2026-01-21  5:59 ` ✗ i915.CI.BAT: failure for Account for DSC bubble overhead for horizontal slices (rev3) Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=aXPGvBQ29dzanJfR@ideak-desk \
    --to=imre.deak@intel.com \
    --cc=ankit.k.nautiyal@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=intel-xe@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox