fegger
6516b45b02
The receiver decoded H.264 to YUV420P, converted it to RGBA via a CPU-intensive swscale pass, then uploaded 4 bytes/pixel to an SDL texture — only for the GPU to convert back to RGB during rendering. This eliminated the swscale pass entirely (40-60% of receiver CPU at 1080p) and cut the texture upload by 62%. - DecodedFrame now carries three YUV420P planes with their strides instead of a packed RGBA buffer; the decoder copies the planes directly from the AVFrame (zero conversion for the common software path). Non-YUV420P decoder output (e.g. NV12 from v4l2m2m) is converted once to YUV420P. - The SDL renderer uploads via SDL_UpdateYUVTexture with SDL_PIXELFORMAT_IYUV; the GPU does the YUV→RGB conversion during rendering. - Decoder threading: slice-level with 4 threads (parallelizes within a frame, no added latency), not frame-level (which buffers multiple frames — the initial thread_count=0 broke the loopback test because the H.264 decoder introduced a multi-frame delay before producing output). - The round-trip test converts decoded YUV back to RGBA for pixel comparison via a test-local swscale call (the pipeline itself never converts). meson test 5/5 in both configurations, valgrind clean.
54 lines
1.4 KiB
C++
54 lines
1.4 KiB
C++
#pragma once
|
|
|
|
#include "screencast/codec/encoder.h"
|
|
#include "screencast/codec/error.h"
|
|
|
|
#include <cstdint>
|
|
#include <memory>
|
|
#include <string>
|
|
#include <vector>
|
|
|
|
namespace sc {
|
|
|
|
// Decoded video frame in YUV420P planar format (the native decoder output,
|
|
// passed to the renderer without any colorspace conversion).
|
|
struct DecodedFrame {
|
|
int width = 0;
|
|
int height = 0;
|
|
uint64_t capture_timestamp_ns = 0;
|
|
|
|
// YUV420P planes; each row is `stride` bytes wide and may be padded
|
|
// beyond the picture width.
|
|
std::vector<std::byte> plane_y;
|
|
std::vector<std::byte> plane_u;
|
|
std::vector<std::byte> plane_v;
|
|
int stride_y = 0;
|
|
int stride_u = 0;
|
|
int stride_v = 0;
|
|
};
|
|
|
|
struct DecoderConfig {
|
|
std::string codec_name = "h264";
|
|
int width = 0;
|
|
int height = 0;
|
|
std::vector<std::byte> extradata; // SPS/PPS for H.264
|
|
// Probe a hardware decoder (v4l2 mem2mem) first and fall back to the
|
|
// software decoder automatically. Disable with --swdecode.
|
|
bool hardware_accel = true;
|
|
};
|
|
|
|
class Decoder {
|
|
public:
|
|
virtual ~Decoder() = default;
|
|
|
|
// Feed one encoded frame. Returns decoded frames when available.
|
|
virtual CodecResult<std::vector<DecodedFrame>> decode(const EncodedFrame& frame) = 0;
|
|
};
|
|
|
|
class DecoderFactory {
|
|
public:
|
|
static CodecResult<std::unique_ptr<Decoder>> create(const DecoderConfig& config);
|
|
};
|
|
|
|
} // namespace sc
|