Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Nautiyal, Ankit K" <ankit.k.nautiyal@intel.com>
To: Suraj Kandpal <suraj.kandpal@intel.com>,
	<intel-gfx@lists.freedesktop.org>
Subject: Re: [Intel-gfx] [PATCH v3 1/3] drm/i915/dsc: Move rc param calculation for native_420
Date: Tue, 4 Jul 2023 13:55:02 +0530	[thread overview]
Message-ID: <c2060069-478b-590c-9169-50ddafa5f314@intel.com> (raw)
In-Reply-To: <20230703101244.2489790-2-suraj.kandpal@intel.com>


On 7/3/2023 3:42 PM, Suraj Kandpal wrote:
> Move rc_param calculation for native_420 into calculate_rc_parameter.
> second_line_bpg_offset and second_line_offset_adj are both rc params
> and it would be better to have these calculated where all the other
> rc parameters are calculated.
>
> --v2
> -Add the reason for commit in commit message [Jani]
>
> --v3
> -Move nsl_second_line_bpg_offset with the other 420 calculation
> in calculate_rc_param [Ankit]
>
> Cc: Jani Nikula <jani.nikula@linux.intel.com>
> Cc: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
> Signed-off-by: Suraj Kandpal <suraj.kandpal@intel.com>
> ---
>   drivers/gpu/drm/i915/display/intel_vdsc.c | 45 ++++++++++++-----------
>   1 file changed, 24 insertions(+), 21 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_vdsc.c b/drivers/gpu/drm/i915/display/intel_vdsc.c
> index bd9116d2cd76..7d0edb440ca6 100644
> --- a/drivers/gpu/drm/i915/display/intel_vdsc.c
> +++ b/drivers/gpu/drm/i915/display/intel_vdsc.c
> @@ -78,6 +78,27 @@ calculate_rc_params(struct drm_dsc_config *vdsc_cfg)
>   	else
>   		vdsc_cfg->first_line_bpg_offset = 2 * (vdsc_cfg->slice_height - 1);
>   
> +	/* According to DSC 1.2 specs in Section 4.1 if native_420 is set:

Start comment from next line, as per comment format.

Otherwise LGTM.

Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>

> +	 * -second_line_bpg_offset is 12 in general and equal to 2*(slice_height-1) if slice
> +	 * height < 8.
> +	 * -second_line_offset_adj is 512 as shown by emperical values to yield best chroma
> +	 * preservation in second line.
> +	 * -nsl_bpg_offset is calculated as second_line_offset/slice_height -1 then rounded
> +	 * up to 16 fractional bits, we left shift second line offset by 11 to preserve 11
> +	 * fractional bits.
> +	 */
> +	if (vdsc_cfg->native_420) {
> +		if (vdsc_cfg->slice_height >= 8)
> +			vdsc_cfg->second_line_bpg_offset = 12;
> +		else
> +			vdsc_cfg->second_line_bpg_offset =
> +				2 * (vdsc_cfg->slice_height - 1);
> +
> +		vdsc_cfg->second_line_offset_adj = 512;
> +		vdsc_cfg->nsl_bpg_offset = DIV_ROUND_UP(vdsc_cfg->second_line_bpg_offset << 11,
> +							vdsc_cfg->slice_height - 1);
> +	}
> +
>   	/* Our hw supports only 444 modes as of today */
>   	if (bpp >= 12)
>   		vdsc_cfg->initial_offset = 2048;
> @@ -190,30 +211,12 @@ int intel_dsc_compute_params(struct intel_crtc_state *pipe_config)
>   	vdsc_cfg->bits_per_pixel = compressed_bpp << 4;
>   
>   	/*
> -	 * According to DSC 1.2 specs in Section 4.1 if native_420 is set:
> -	 * -We need to double the current bpp.
> -	 * -second_line_bpg_offset is 12 in general and equal to 2*(slice_height-1) if slice
> -	 * height < 8.
> -	 * -second_line_offset_adj is 512 as shown by emperical values to yeild best chroma
> -	 * preservation in second line.
> -	 * -nsl_bpg_offset is calculated as second_line_offset/slice_height -1 then rounded
> -	 * up to 16 fractional bits, we left shift second line offset by 11 to preserve 11
> -	 * fractional bits.
> +	 * According to DSC 1.2 specs in Section 4.1 if native_420 is set
> +	 * we need to double the current bpp.
>   	 */
> -	if (vdsc_cfg->native_420) {
> +	if (vdsc_cfg->native_420)
>   		vdsc_cfg->bits_per_pixel <<= 1;
>   
> -		if (vdsc_cfg->slice_height >= 8)
> -			vdsc_cfg->second_line_bpg_offset = 12;
> -		else
> -			vdsc_cfg->second_line_bpg_offset =
> -				2 * (vdsc_cfg->slice_height - 1);
> -
> -		vdsc_cfg->second_line_offset_adj = 512;
> -		vdsc_cfg->nsl_bpg_offset = DIV_ROUND_UP(vdsc_cfg->second_line_bpg_offset << 11,
> -							vdsc_cfg->slice_height - 1);
> -	}
> -
>   	vdsc_cfg->bits_per_component = pipe_config->pipe_bpp / 3;
>   
>   	drm_dsc_set_rc_buf_thresh(vdsc_cfg);

  reply	other threads:[~2023-07-04  8:25 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-03 10:12 [Intel-gfx] [PATCH v3 0/3] Add rc_range_params for YUV420 Suraj Kandpal
2023-07-03 10:12 ` [Intel-gfx] [PATCH v3 1/3] drm/i915/dsc: Move rc param calculation for native_420 Suraj Kandpal
2023-07-04  8:25   ` Nautiyal, Ankit K [this message]
2023-07-03 10:12 ` [Intel-gfx] [PATCH v3 2/3] drm/i915/drm: Fix comment for YUV420 qp table declaration Suraj Kandpal
2023-07-04  8:11   ` Nautiyal, Ankit K
2023-07-03 10:12 ` [Intel-gfx] [PATCH v3 3/3] drm/i915/dsc: Add rc_range_parameter calculation for YCBCR420 Suraj Kandpal
2023-07-04 13:39   ` Nautiyal, Ankit K
2023-07-03 10:38 ` [Intel-gfx] ✗ Fi.CI.SPARSE: warning for Add rc_range_params for YUV420 Patchwork
2023-07-03 10:48 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2023-07-03 12:00 ` [Intel-gfx] ✓ Fi.CI.IGT: " 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=c2060069-478b-590c-9169-50ddafa5f314@intel.com \
    --to=ankit.k.nautiyal@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=suraj.kandpal@intel.com \
    /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