linux-kernel - Re: [PATCH v5 4/8] media: mediatek: vcodec: Add single core VP9 decoding support for MT8189

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <5834fe661f38462d65795a20b5a4f9d2a12d3b35.camel@collabora.com>
Date: Thu, 06 Nov 2025 09:29:12 -0500
From: Nicolas Dufresne <nicolas.dufresne@...labora.com>
To: Kyrie Wu <kyrie.wu@...iatek.com>, Tiffany Lin
 <tiffany.lin@...iatek.com>,  Andrew-CT Chen <andrew-ct.chen@...iatek.com>,
 Yunfei Dong <yunfei.dong@...iatek.com>, Mauro Carvalho Chehab	
 <mchehab@...nel.org>, Rob Herring <robh@...nel.org>, Krzysztof Kozlowski	
 <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>, Matthias Brugger	
 <matthias.bgg@...il.com>, AngeloGioacchino Del Regno	
 <angelogioacchino.delregno@...labora.com>, Hans Verkuil
 <hverkuil@...all.nl>,  Nathan Hebert <nhebert@...omium.org>, Arnd Bergmann
 <arnd@...db.de>, Irui Wang <irui.wang@...iatek.com>,  George Sun
 <george.sun@...iatek.com>, linux-media@...r.kernel.org,
 devicetree@...r.kernel.org, 	linux-kernel@...r.kernel.org,
 linux-arm-kernel@...ts.infradead.org, 	linux-mediatek@...ts.infradead.org,
 Yilong Zhou <yilong.zhou@...iatek.com>
Cc: Neil Armstrong <neil.armstrong@...aro.org>, Andrzej Pietrasiewicz
	 <andrzejtp2010@...il.com>
Subject: Re: [PATCH v5 4/8] media: mediatek: vcodec: Add single core VP9
 decoding support for MT8189

Le jeudi 06 novembre 2025 à 14:13 +0800, Kyrie Wu a écrit :
> 1. add config to support 4K prob size;
> 2. Previously, calling vdec_vp9_slice_setup_single_from_src_to_dst
> with v4l2_m2m_next_src_buf to obtain both buffers resulted in -EINVAL,
> interrupting the decoding process. To resolve this,
> the interface should be updated to set both src and dst buffers
> for metadata configuration.

Please avoid using bullets in commit messages. Everyone will assume that if you
have two bullets, you should have two patches. And I think they would be right
for this patch.

> 
> Signed-off-by: Kyrie Wu <kyrie.wu@...iatek.com>
> ---
>  .../vcodec/decoder/vdec/vdec_vp9_req_lat_if.c | 43 ++++++++++++++-----
>  1 file changed, 32 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_vp9_req_lat_if.c b/drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_vp9_req_lat_if.c
> index fa0f406f7726..d966914db4b9 100644
> --- a/drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_vp9_req_lat_if.c
> +++ b/drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_vp9_req_lat_if.c
> @@ -23,6 +23,7 @@
>  
>  #define VP9_TILE_BUF_SIZE 4096
>  #define VP9_PROB_BUF_SIZE 2560
> +#define VP9_PROB_BUF_4K_SIZE 3840
>  #define VP9_COUNTS_BUF_SIZE 16384
>  
>  #define HDR_FLAG(x) (!!((hdr)->flags & V4L2_VP9_FRAME_FLAG_##x))
> @@ -542,6 +543,23 @@ static int vdec_vp9_slice_init_default_frame_ctx(struct vdec_vp9_slice_instance
>  	return ret;
>  }
>  
> +static size_t mtk_vcodec_get_vp9_prob_size(enum mtk_vcodec_dec_chip_name chip_name)
> +{
> +	size_t prob_size;
> +
> +	switch (chip_name) {
> +	case MTK_VDEC_MT8189:
> +	case MTK_VDEC_MT8196:
> +		prob_size = VP9_PROB_BUF_4K_SIZE;
> +		break;
> +	default:
> +		prob_size = VP9_PROB_BUF_SIZE;
> +		break;
> +	}
> +
> +	return prob_size;
> +}

This is another example of something that could be in the static compatible
data.

> +
>  static int vdec_vp9_slice_alloc_working_buffer(struct vdec_vp9_slice_instance *instance,
>  					       struct vdec_vp9_slice_vsi *vsi)
>  {
> @@ -616,7 +634,9 @@ static int vdec_vp9_slice_alloc_working_buffer(struct vdec_vp9_slice_instance *i
>  	}
>  
>  	if (!instance->prob.va) {
> -		instance->prob.size = VP9_PROB_BUF_SIZE;
> +		instance->prob.size =
> +			mtk_vcodec_get_vp9_prob_size(ctx->dev->chip_name);
> +
>  		if (mtk_vcodec_mem_alloc(ctx, &instance->prob))
>  			goto err;
>  	}
> @@ -696,21 +716,22 @@ static int vdec_vp9_slice_tile_offset(int idx, int mi_num, int tile_log2)
>  	return min(offset, mi_num);
>  }
>  
> -static
> -int vdec_vp9_slice_setup_single_from_src_to_dst(struct vdec_vp9_slice_instance *instance)
> +static int vdec_vp9_slice_setup_single_from_src_to_dst(struct vdec_vp9_slice_instance *instance,
> +						       struct mtk_vcodec_mem *bs,
> +						       struct vdec_fb *fb)
>  {
> -	struct vb2_v4l2_buffer *src;
> -	struct vb2_v4l2_buffer *dst;
> +	struct mtk_video_dec_buf *src_buf_info;
> +	struct mtk_video_dec_buf *dst_buf_info;
>  
> -	src = v4l2_m2m_next_src_buf(instance->ctx->m2m_ctx);
> -	if (!src)
> +	src_buf_info = container_of(bs, struct mtk_video_dec_buf, bs_buffer);
> +	if (!src_buf_info)
>  		return -EINVAL;
>  
> -	dst = v4l2_m2m_next_dst_buf(instance->ctx->m2m_ctx);
> -	if (!dst)
> +	dst_buf_info = container_of(fb, struct mtk_video_dec_buf, frame_buffer);
> +	if (!dst_buf_info)
>  		return -EINVAL;
>  
> -	v4l2_m2m_buf_copy_metadata(src, dst, true);
> +	v4l2_m2m_buf_copy_metadata(&src_buf_info->m2m_buf.vb, &dst_buf_info->m2m_buf.vb, true);
>  
>  	return 0;
>  }
> @@ -1800,7 +1821,7 @@ static int vdec_vp9_slice_setup_single(struct vdec_vp9_slice_instance *instance,
>  	struct vdec_vp9_slice_vsi *vsi = &pfc->vsi;
>  	int ret;
>  
> -	ret = vdec_vp9_slice_setup_single_from_src_to_dst(instance);
> +	ret = vdec_vp9_slice_setup_single_from_src_to_dst(instance, bs, fb);

I don't see a direct correlation betwen the buffer size change and this change.
Please split. Changes looks good, the commit message isn't very clear though.

Nicolas

>  	if (ret)
>  		goto err;
>  

Download attachment "signature.asc" of type "application/pgp-signature" (229 bytes)