Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
 help / color / mirror / Atom feed
From: Hendrik Leppkes <h.leppkes@gmail.com>
To: FFmpeg development discussions and patches <ffmpeg-devel@ffmpeg.org>
Subject: Re: [FFmpeg-devel] [PATCH] d3d11va: let user can create SRV from output
Date: Wed, 9 Mar 2022 08:20:46 +0100
Message-ID: <CA+anqdw=vq0osCn5W7jw1eA7QuaQ7RnCmVtjU5ejzFizgr37_A@mail.gmail.com> (raw)
In-Reply-To: <ME3PR01MB562442911EA9A24F5402E084DA0A9@ME3PR01MB5624.ausprd01.prod.outlook.com>

On Wed, Mar 9, 2022 at 7:56 AM Wang Chuan <ouchuanm@outlook.com> wrote:
>
> Starting from Windows 8, users can create SRV from video resource
> and bind it to shaders directly. This can avoid unnecessary memcpy
> (ID3D11DeviceContext::CopyResource, etc), so create texture with
> [D3D11_BIND_SHADER_RESOURCE] as decoder's output if possible.
>
> Signed-off-by: Wang Chuan <ouchuanm@outlook.com>
> ---
>   libavcodec/dxva2.c            |  2 +-
>   libavutil/hwcontext_d3d11va.c | 10 ++++++++++
>   2 files changed, 11 insertions(+), 1 deletion(-)
>
> diff --git a/libavcodec/dxva2.c b/libavcodec/dxva2.c
> index 568d686f39..15b25d793c 100644
> --- a/libavcodec/dxva2.c
> +++ b/libavcodec/dxva2.c
> @@ -645,7 +645,7 @@ int ff_dxva2_common_frame_params(AVCodecContext *avctx,
>       if (frames_ctx->format == AV_PIX_FMT_D3D11) {
>           AVD3D11VAFramesContext *frames_hwctx = frames_ctx->hwctx;
>   -        frames_hwctx->BindFlags |= D3D11_BIND_DECODER;
> +        frames_hwctx->BindFlags |= (D3D11_BIND_DECODER |
> D3D11_BIND_SHADER_RESOURCE);
>       }
>   #endif
>   diff --git a/libavutil/hwcontext_d3d11va.c b/libavutil/hwcontext_d3d11va.c
> index 8ab96bad25..97ffd745bd 100644
> --- a/libavutil/hwcontext_d3d11va.c
> +++ b/libavutil/hwcontext_d3d11va.c
> @@ -203,6 +203,11 @@ static AVBufferRef
> *d3d11va_alloc_single(AVHWFramesContext *ctx)
>       };
>        hr = ID3D11Device_CreateTexture2D(device_hwctx->device, &texDesc,
> NULL, &tex);
> +    if (FAILED(hr) && (texDesc.BindFlags & D3D11_BIND_SHADER_RESOURCE)) {
> +        av_log(ctx, AV_LOG_ERROR, "Could not create the texture with
> [D3D11_BIND_SHADER_RESOURCE] flag");
> +        texDesc.BindFlags &= ~D3D11_BIND_SHADER_RESOURCE;
> +        hr = ID3D11Device_CreateTexture2D(device_hwctx->device,
> &texDesc, NULL, &tex);
> +    }
>       if (FAILED(hr)) {
>           av_log(ctx, AV_LOG_ERROR, "Could not create the texture
> (%lx)\n", (long)hr);
>           return NULL;
> @@ -278,6 +283,11 @@ static int d3d11va_frames_init(AVHWFramesContext *ctx)
>           }
>       } else if (!(texDesc.BindFlags & D3D11_BIND_RENDER_TARGET) &&
> texDesc.ArraySize > 0) {
>           hr = ID3D11Device_CreateTexture2D(device_hwctx->device,
> &texDesc, NULL, &hwctx->texture);
> +        if (FAILED(hr) && (texDesc.BindFlags &
> D3D11_BIND_SHADER_RESOURCE)) {
> +            av_log(ctx, AV_LOG_ERROR, "Could not create the texture
> with [D3D11_BIND_SHADER_RESOURCE] flag");
> +            texDesc.BindFlags &= ~D3D11_BIND_SHADER_RESOURCE;
> +            hr = ID3D11Device_CreateTexture2D(device_hwctx->device,
> &texDesc, NULL, &hwctx->texture);
> +        }

I really don't like these fallbacks. If a caller requests a certain
set of flags, it should fullfill these, or fail, and not change them.
Especially special-casing this one flag just makes it extra iffy, and
I know why you added it, but that doesn't help.

IMHO, its also not entirely unreasonable that if you require anything
but the default flags, to let the user just manage their own frames
context, the effort required is not that high and there is never a set
of flags you can set that'll cover every use case (for example, I use
surface sharing, thats not set by default either).

- Hendrik
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

      reply	other threads:[~2022-03-09  7:21 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-09  6:56 Wang Chuan
2022-03-09  7:20 ` Hendrik Leppkes [this message]

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='CA+anqdw=vq0osCn5W7jw1eA7QuaQ7RnCmVtjU5ejzFizgr37_A@mail.gmail.com' \
    --to=h.leppkes@gmail.com \
    --cc=ffmpeg-devel@ffmpeg.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

Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
		ffmpegdev@gitmailbox.com
	public-inbox-index ffmpegdev

Example config snippet for mirrors.


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git