libva-v4l2-request-fourier: 9898331 -> c1bb444 (PR #9 / issue #8 fix) #52

2026-05-20T18:22:06Z

claude-noether commented

2026-05-20 18:22:06 +00:00

Bumps the libva-v4l2-request-fourier pin to include PR #9 (issue #8 partial fix).

Summary

arch/libva-v4l2-request-fourier/PKGBUILD: _commit 9898331 → c1bb444, pkgver 1.0.0.r380.9898331 → 1.0.0.r382.c1bb444. Two upstream commits in the bump:

0791f8e h264: max_num_ref_frames fallback + libva-boundary instrumentation
c1bb444 Merge PR #9 → master

The PR explicitly defers profile_idc / level_idc to a follow-up (libva's VAPictureParameterBufferH264 omits both — verified higgs /usr/include/va/va.h:3571-3622 — so the real fix needs SPS-NAL parsing OR a daedalus wire-protocol pass-through; both are operator-design calls).

Same maintenance window as the upcoming follow-ups for the rest of issue #8 if/when that lands.

Test plan

Source-side build verified on higgs (Debian 13 trixie, gcc 14.2.0, libva 2.22.0): clean meson + ninja link of v4l2_request_drv_video.so; vainfo enumerates all 8 codec profiles; multi-device probe still wires rkvdec / rpi-hevc-dec / daedalus_v4l2 slots.
CI build via fermi LXC actrunner-aarch64-hertz (path=arch/libva-v4l2-request-fourier).
Install + smoke on the deploy hosts (ampere, fresnel, higgs) — operator-driven post-merge.

Bumps the libva-v4l2-request-fourier pin to include PR #9 (issue #8 partial fix). ## Summary `arch/libva-v4l2-request-fourier/PKGBUILD`: `_commit` `9898331` → `c1bb444`, `pkgver` `1.0.0.r380.9898331` → `1.0.0.r382.c1bb444`. Two upstream commits in the bump: - `0791f8e` h264: max_num_ref_frames fallback + libva-boundary instrumentation - `c1bb444` Merge PR #9 → master The PR explicitly defers `profile_idc` / `level_idc` to a follow-up (libva's `VAPictureParameterBufferH264` omits both — verified higgs `/usr/include/va/va.h:3571-3622` — so the real fix needs SPS-NAL parsing OR a daedalus wire-protocol pass-through; both are operator-design calls). Same maintenance window as the upcoming follow-ups for the rest of issue #8 if/when that lands. ## Test plan - [x] Source-side build verified on higgs (Debian 13 trixie, gcc 14.2.0, libva 2.22.0): clean meson + ninja link of `v4l2_request_drv_video.so`; `vainfo` enumerates all 8 codec profiles; multi-device probe still wires `rkvdec` / `rpi-hevc-dec` / `daedalus_v4l2` slots. - [ ] CI build via fermi LXC actrunner-aarch64-hertz (path=arch/libva-v4l2-request-fourier). - [ ] Install + smoke on the deploy hosts (ampere, fresnel, higgs) — operator-driven post-merge.

claude-noether added 1 commit 2026-05-20 18:22:06 +00:00

libva-v4l2-request-fourier: 9898331 -> c1bb444 (PR #9 / issue #8 fix) e540384f50

Bumps the libva backend pin to include marfrit/libva-v4l2-request-fourier
PR #9 — h264_set_controls fix for the bitstream-vs-session value drift
that breaks the daedalus_v4l2 strict-consumer path (issue #8):

  * max_num_ref_frames fallback when VAAPI client left it 0 (count
    valid DPB entries, then per-profile spec minimum)
  * one-line request_log at h264_set_controls entry dumping raw
    VAAPI bitfields for disambiguating remaining PPS-flag-zero
    portion of #8

The PR explicitly defers the deeper "profile_idc / level_idc from
bitstream" portion of #8 — VAAPI's VAPictureParameterBufferH264 omits
both fields, so a real fix needs SPS-NAL parsing or daedalus
wire-protocol pass-through. Not in this bump.

pkgver: 1.0.0.r380.9898331 -> 1.0.0.r382.c1bb444 (commit count 380->382)
pkgrel: 1 (fresh pkgver, no rebuild-only iteration)

Verified on higgs (Debian 13 trixie, gcc 14.2.0, libva 2.22.0):
clean meson build, vainfo enumerates all 8 codec profiles, multi-device
probe still wires rkvdec / rpi-hevc-dec / daedalus_v4l2.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>