Message-ID: <e9b238af-28dd-f38e-24e3-4065c39d4e25@linaro.org>
Date:   Mon, 30 Jan 2023 20:36:34 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.6.0
Subject: Re: [v1 2/3] drm/msm/disp/dpu1: add dspps into reservation if there
 is a ctm request
Content-Language: en-GB
To:     Kalyan Thota <quic_kalyant@quicinc.com>,
        dri-devel@lists.freedesktop.org, linux-arm-msm@vger.kernel.org,
        freedreno@lists.freedesktop.org, devicetree@vger.kernel.org
Cc:     linux-kernel@vger.kernel.org, robdclark@chromium.org,
        dianders@chromium.org, swboyd@chromium.org,
        quic_vpolimer@quicinc.com, quic_abhinavk@quicinc.com,
        marijn.suijten@somainline.org
References: <1675092092-26412-1-git-send-email-quic_kalyant@quicinc.com>
 <1675092092-26412-3-git-send-email-quic_kalyant@quicinc.com>
From:   Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
In-Reply-To: <1675092092-26412-3-git-send-email-quic_kalyant@quicinc.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 30/01/2023 17:21, Kalyan Thota wrote:
> Add dspp blocks into the topology for reservation, if there is a ctm
> request for that composition.
> 
> Changes in v1:
> - Minor nits (Dmitry)
> 
> Signed-off-by: Kalyan Thota <quic_kalyant@quicinc.com>
> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
> ---
>   drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c | 13 ++++++-------
>   1 file changed, 6 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> index 9c6817b..3bd46b4 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> @@ -545,7 +545,8 @@ bool dpu_encoder_use_dsc_merge(struct drm_encoder *drm_enc)
>   static struct msm_display_topology dpu_encoder_get_topology(
>   			struct dpu_encoder_virt *dpu_enc,
>   			struct dpu_kms *dpu_kms,
> -			struct drm_display_mode *mode)
> +			struct drm_display_mode *mode,
> +			struct drm_crtc_state *crtc_state)
>   {
>   	struct msm_display_topology topology = {0};
>   	int i, intf_count = 0;
> @@ -573,11 +574,9 @@ static struct msm_display_topology dpu_encoder_get_topology(
>   	else
>   		topology.num_lm = (mode->hdisplay > MAX_HDISPLAY_SPLIT) ? 2 : 1;
>   
> -	if (dpu_enc->disp_info.intf_type == DRM_MODE_ENCODER_DSI) {
> -		if (dpu_kms->catalog->dspp &&
> -			(dpu_kms->catalog->dspp_count >= topology.num_lm))
> -			topology.num_dspp = topology.num_lm;
> -	}
> +	if (dpu_kms->catalog->dspp &&
> +	    crtc_state->ctm && (dpu_kms->catalog->dspp_count >= topology.num_lm))

This condition doesn't look correct anymore for the following reasons:
- if there are no DSPPs we will completely ignore the ctm property
- if there are not enough DSPPs, the CTM property will be ignore

I think, this should be just:

if (crtc_state->ctm)
     topology.num_dspp = topology.num_lm;


> +		topology.num_dspp = topology.num_lm;
>   
>   	topology.num_enc = 0;
>   	topology.num_intf = intf_count;
> @@ -643,7 +642,7 @@ static int dpu_encoder_virt_atomic_check(
>   		}
>   	}
>   
> -	topology = dpu_encoder_get_topology(dpu_enc, dpu_kms, adj_mode);
> +	topology = dpu_encoder_get_topology(dpu_enc, dpu_kms, adj_mode, crtc_state);
>   
>   	/* Reserve dynamic resources now. */
>   	if (!ret) {

-- 
With best wishes
Dmitry