[Freedreno] [RFC PATCH] drm/msm/dpu: Move TE setup to prepare_for_kickoff()

[Dmitry Baryshkov]

On 08/02/2023 23:37, Jessica Zhang wrote:
> Currently, DPU will enable TE during prepare_commit(). However, this
> will cause issues when trying to read/write to register in
> get_autorefresh_config(), because the core clock rates aren't set at
> that time.
> 
> This used to work because phys_enc->hw_pp is only initialized in mode
> set [1], so the first prepare_commit() will return before any register
> read/write as hw_pp would be NULL.
> 
> However, when we try to implement support for INTF TE, we will run into
> the clock issue described above as hw_intf will *not* be NULL on the
> first prepare_commit(). This is because the initialization of
> dpu_enc->hw_intf has been moved to dpu_encoder_setup() [2].
> 
> To avoid this issue, let's enable TE during prepare_for_kickoff()
> instead as the core clock rates are guaranteed to be set then.
> 
> Depends on: "Implement tearcheck support on INTF block" [3]
> 
> [1] https://gitlab.freedesktop.org/drm/msm/-/blob/msm-next/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c#L1109
> [2] https://gitlab.freedesktop.org/drm/msm/-/blob/msm-next/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c#L2339
> [3] https://patchwork.freedesktop.org/series/112332/
> 
> Signed-off-by: Jessica Zhang <quic_jesszhan at quicinc.com>
> ---
>   .../drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c  | 78 ++++++++++---------
>   1 file changed, 43 insertions(+), 35 deletions(-)
> 
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c
> index 279a0b7015ce..746250bce3d1 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c
> @@ -587,39 +587,6 @@ static void dpu_encoder_phys_cmd_destroy(struct dpu_encoder_phys *phys_enc)
>   	kfree(cmd_enc);
>   }
>   
> -static void dpu_encoder_phys_cmd_prepare_for_kickoff(
> -		struct dpu_encoder_phys *phys_enc)
> -{
> -	struct dpu_encoder_phys_cmd *cmd_enc =
> -			to_dpu_encoder_phys_cmd(phys_enc);
> -	int ret;
> -
> -	if (!phys_enc->hw_pp) {
> -		DPU_ERROR("invalid encoder\n");
> -		return;
> -	}
> -	DRM_DEBUG_KMS("id:%u pp:%d pending_cnt:%d\n", DRMID(phys_enc->parent),
> -		      phys_enc->hw_pp->idx - PINGPONG_0,
> -		      atomic_read(&phys_enc->pending_kickoff_cnt));
> -
> -	/*
> -	 * Mark kickoff request as outstanding. If there are more than one,
> -	 * outstanding, then we have to wait for the previous one to complete
> -	 */
> -	ret = _dpu_encoder_phys_cmd_wait_for_idle(phys_enc);
> -	if (ret) {
> -		/* force pending_kickoff_cnt 0 to discard failed kickoff */
> -		atomic_set(&phys_enc->pending_kickoff_cnt, 0);
> -		DRM_ERROR("failed wait_for_idle: id:%u ret:%d pp:%d\n",
> -			  DRMID(phys_enc->parent), ret,
> -			  phys_enc->hw_pp->idx - PINGPONG_0);
> -	}
> -
> -	DPU_DEBUG_CMDENC(cmd_enc, "pp:%d pending_cnt %d\n",
> -			phys_enc->hw_pp->idx - PINGPONG_0,
> -			atomic_read(&phys_enc->pending_kickoff_cnt));
> -}
> -
>   static bool dpu_encoder_phys_cmd_is_ongoing_pptx(
>   		struct dpu_encoder_phys *phys_enc)
>   {
> @@ -645,8 +612,7 @@ static bool dpu_encoder_phys_cmd_is_ongoing_pptx(
>   	return false;
>   }
>   
> -static void dpu_encoder_phys_cmd_prepare_commit(
> -		struct dpu_encoder_phys *phys_enc)
> +static void dpu_encoder_phys_cmd_enable_te(struct dpu_encoder_phys *phys_enc)
>   {
>   	struct dpu_encoder_phys_cmd *cmd_enc =
>   		to_dpu_encoder_phys_cmd(phys_enc);
> @@ -704,6 +670,48 @@ static void dpu_encoder_phys_cmd_prepare_commit(
>   			 "disabled autorefresh\n");
>   }
>   
> +static void dpu_encoder_phys_cmd_prepare_for_kickoff(
> +		struct dpu_encoder_phys *phys_enc)

Could you please move the function back to the place, so that we can see 
the actual difference?

> +{
> +	struct dpu_encoder_phys_cmd *cmd_enc =
> +			to_dpu_encoder_phys_cmd(phys_enc);
> +	int ret;
> +
> +	if (!phys_enc->hw_pp) {
> +		DPU_ERROR("invalid encoder\n");
> +		return;
> +	}
> +
> +
> +	DRM_DEBUG_KMS("id:%u pp:%d pending_cnt:%d\n", DRMID(phys_enc->parent),
> +		      phys_enc->hw_pp->idx - PINGPONG_0,
> +		      atomic_read(&phys_enc->pending_kickoff_cnt));
> +
> +	/*
> +	 * Mark kickoff request as outstanding. If there are more than one,
> +	 * outstanding, then we have to wait for the previous one to complete
> +	 */
> +	ret = _dpu_encoder_phys_cmd_wait_for_idle(phys_enc);
> +	if (ret) {
> +		/* force pending_kickoff_cnt 0 to discard failed kickoff */
> +		atomic_set(&phys_enc->pending_kickoff_cnt, 0);
> +		DRM_ERROR("failed wait_for_idle: id:%u ret:%d pp:%d\n",
> +			  DRMID(phys_enc->parent), ret,
> +			  phys_enc->hw_pp->idx - PINGPONG_0);
> +	}
> +
> +	dpu_encoder_phys_cmd_enable_te(phys_enc);
> +
> +	DPU_DEBUG_CMDENC(cmd_enc, "pp:%d pending_cnt %d\n",
> +			phys_enc->hw_pp->idx - PINGPONG_0,
> +			atomic_read(&phys_enc->pending_kickoff_cnt));
> +}
> +
> +static void dpu_encoder_phys_cmd_prepare_commit(
> +		struct dpu_encoder_phys *phys_enc)
> +{
> +}

There is no need to have the empty callback, you can skip it completely. 
Actually, if it is not needed anymore for the cmd encoders, you can drop 
the .prepare_commit from struct dpu_encoder_phys_ops. And then, by 
extension, dpu_encoder_prepare_commit(), dpu_kms_prepare_commit(). This 
sounds like a nice second patch for this rfc.

> +
>   static int _dpu_encoder_phys_cmd_wait_for_ctl_start(
>   		struct dpu_encoder_phys *phys_enc)
>   {

-- 
With best wishes
Dmitry