NVDEC (H264): Use separate contexts per channel and decode frames in DTS order (#2671)

* Use separate NVDEC contexts per channel (for FFMPEG)

* Remove NVDEC -> VIC frame override hack

* Add missing bottom_field_pic_order_in_frame_present_flag

* Make FFMPEG logging static

* nit: Remove empty lines

* New FFMPEG decoding approach -- call h264_decode_frame directly, trim surface cache to reduce memory usage

* Fix case

* Silence warnings

* PR feedback

* Per-decoder rather than per-codec ownership of surfaces on the cache
This commit is contained in:
gdkchan 2021-09-28 19:43:40 -03:00 committed by GitHub
parent 0d23504e30
commit f4f496cb48
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
18 changed files with 358 additions and 200 deletions

View file

@ -22,21 +22,6 @@ namespace Ryujinx.HLE.HOS.Services.Nv
var vic = new VicDevice(Smmu);
Host1x.RegisterDevice(ClassId.Nvdec, nvdec);
Host1x.RegisterDevice(ClassId.Vic, vic);
nvdec.FrameDecoded += (FrameDecodedEventArgs e) =>
{
// FIXME:
// Figure out what is causing frame ordering issues on H264.
// For now this is needed as workaround.
if (e.CodecId == CodecId.H264)
{
vic.SetSurfaceOverride(e.LumaOffset, e.ChromaOffset, 0);
}
else
{
vic.DisableSurfaceOverride();
}
};
}
public void Dispose()