enne2/R36SHack
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

docs: record NV12 benchmark results and next-step analysis

Browse Source 
Add benchmark log table to development-status.md comparing:
  - a201594: extract_dup+from_buffer_copy (2 copies, 6MB/frame) → 36.5ms, 87.6% budget
  - da02e74: buffer.map+memmove into reusable ctypes array (1 copy, 3MB/frame) → 33.6ms, 80.5% budget

Note that the 3.1MB memmove is now the remaining bottleneck and further
reduction would require DMA-buf zero-copy via kernel VPU driver support.

Update next actions: profile SDL upload overhead, explore dmabuf fd path,
and consider 720p downscale option if stutter appears under combined load.
This commit is contained in:
Matteo Benedetto
2026-03-24 01:14:04 +01:00
parent da02e7446f
commit eaffa0026d
1 changed files with 21 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/development-status.md
+21 -4
View File
@@ -56,6 +56,23 @@ Milestone 3 — SDL Video Viewport, HUD, and Wayland Compatibility
- **Rockchip MPP hardware decode now deployed**: `librockchip-mpp` and `gst-mpp` compiled from source via Docker QEMU (arm64v8/ubuntu:focal), installed on device, `mppvideodec` confirmed visible to GStreamer. The `gstreamer_backend.py` probe auto-boosts `mppvideodec` rank if `/dev/vpu_service` is accessible.
- **Pre-built .so files bundled** in `deploy/arkos/mpp-libs/`; `setup_hw_decode.sh` installs them automatically without network access.
## NV12 Render Path Benchmark Log
All runs performed on the physical R36S (RK3326, 4× A35 @ 1.3 GHz, 1 GB RAM) over SSH.
Stream: 1920×1080 H.264 MKV @ 24fps via MiniDLNA over LAN. Frame budget: 41.7 ms.
| Commit | Copy strategy | Copy mean | Copy % budget | FPS | Dropped | A/V drift |
|--------|--------------|-----------|---------------|-----|---------|-----------|
| `a201594` | `extract_dup` → bytes + `from_buffer_copy` → ctypes (2 copies, 6 MB/frame) | 36,499 µs | 87.6% | 24.01 | 1 | −42.8 ms |
| `da02e74` | `buffer.map(READ)` + `memmove` into reusable ctypes array (1 copy, 3 MB/frame) | 33,551 µs | 80.5% | 23.98 | 0 | −38.0 ms |
**Key observations (`da02e74`):**
- 1 dropped frame eliminated (0 vs 1)
- Jitter improved: stdev 2.6 ms vs 3.6 ms
- A/V drift tighter: −38 ms vs −43 ms
- Copy cost still 80.5% of frame budget — the 3.1 MB `memmove` on each frame is the remaining bottleneck
- Further reduction requires DMA-buf zero-copy (kernel VPU→SDL import without CPU memcpy), which depends on device driver support not currently available through gst-mpp's appsink path
## Blockers Or Open Questions
- `SDL2_ttf` system library needed for text rendering (`sudo dnf install SDL2_ttf` on Fedora, `sudo apt install libsdl2-ttf-2.0-0` on Debian/Ubuntu). The app handles its absence gracefully but will show no text.
@@ -77,8 +94,8 @@ Milestone 3 — SDL Video Viewport, HUD, and Wayland Compatibility
## Next Recommended Actions
1. Run a real H.264 video end-to-end via app and observe whether `mppvideodec` is selected (logs will show "HW decode: boosted rank of mppvideodec") and whether playback is smooth.
2. Measure CPU/load on RK3326 hardware during HW-accelerated video playback vs. `avdec_h264` software path.
3. Run a visual playback smoke test and confirm SDL-rendered video plus HUD eliminates flashing on host playback.
4. If RGBA upload cost is too high, add a follow-up YUV texture upload path using `SDL_UpdateYUVTexture`.
1. Consider profiling what the remaining 19.5% of frame budget (≈8 ms) consists of — likely SDL_UpdateNVTexture upload + render call overhead + Python GIL churn. If SDL upload is the bottleneck, try `SDL_LockTexture` for direct write instead.
2. Investigate DMA-buf / dmabuf fd import as a future zero-copy path: gst-mpp may expose DRM DMA-buf fds that SDL's KMSDRM backend can import directly, eliminating the CPU memmove entirely.
3. Run a visual playback smoke test on device directly via the app launcher (MatHacks.sh) to confirm HUD and video render correctly together under KMSDRM at the current 80.5% copy load.
4. If 80.5% copy cost causes visible stutter under load (UI overhead competing for A35 cycles), the next option is to reduce resolution at the appsink by inserting a `videoscale` element to 1280×720 before the appsink, cutting memmove to ~1.3 MB/frame (≈35% budget).
5. `avdec_hevc` is still missing (HEVC decoders not in system apt `gstreamer1.0-libav 1.16.1`); `mppvideodec` covers H.264/H.265/VP8/VP9 via HW so this is less critical now.