feat(plex): HW HDR→SDR tonemap for 10-bit/HDR transcode (OpenCL on the iGPU)
Upgrade mode-3's 10-bit/HDR branch from "route to software libx264" (BUG 1's behaviour) to a full-GPU pipeline: VAAPI decode → OpenCL tonemap (hable, HDR10/HLG → bt709 SDR nv12) → h264_vaapi. Gen 9.5 (UHD 630) can't tonemap in the VAAPI VPP (tonemap_vaapi fails), so OpenCL is the only HW route — proven on the prod iGPU at ~4.2x realtime, ~1 CPU core. - stream.py: `_select_encode` returns (hw, tonemap); a 10-bit/HDR HW-decodable source picks the tonemap graph when the OpenCL runtime is present, else drops to software. `_opencl_present` gates on a registered ICD. `_venc_opts` / `_hwaccel_in_opts` gain the tonemap variant (named vaapi+opencl devices, `-filter_hw_device ocl`, hwmap→tonemap_opencl→hwmap reverse). `_want_hw` kept as the hw-half wrapper. The runtime HW→SW fallback still backstops a session that emits no segment. - Dockerfile: add the Intel NEO OpenCL runtime (intel-opencl-icd). It was removed from Debian trixie, so it's pulled from bookworm (22.43 still supports Gen 9.5; the .deb runs on the trixie base — verified live). Harmless without /dev/dri. - route + FE: thread `tonemap` through the session response; the stats overlay mode row reads "transcode · GPU · HDR→SDR" for the tonemap path. - tests: tonemap command branch, the (hw, tonemap) selection matrix (auto/on × 8-bit/10-bit × opencl), start_session tonemap wiring, and the fmtMode tag.
This commit is contained in:
@@ -267,10 +267,36 @@ class TestTranscodeCommand:
|
||||
)
|
||||
assert cmd[cmd.index("-vf") + 1] == "scale_vaapi=format=nv12" # not a software scale
|
||||
|
||||
def test_hdr_tonemap_uses_the_opencl_graph(self):
|
||||
# The HDR/10-bit HW path: VAAPI decode → OpenCL tonemap → VAAPI encode. Gen 9.5 can't tonemap in
|
||||
# the VAAPI VPP, so this is the only HW route for a 10-bit/HDR source.
|
||||
cmd = stream._ffmpeg_cmd(
|
||||
self.SRC, 0.0, self.OUT, None, transcode=True, hw=True, tonemap=True
|
||||
)
|
||||
assert _venc(cmd) == "h264_vaapi"
|
||||
assert "libx264" not in cmd
|
||||
# Named devices + an OpenCL device DERIVED from the VAAPI one, all before -i.
|
||||
assert cmd[cmd.index("-init_hw_device") + 1] == f"vaapi=va:{stream._VAAPI_DEVICE}"
|
||||
assert "opencl=ocl@va" in cmd
|
||||
assert cmd[cmd.index("-hwaccel_device") + 1] == "va"
|
||||
assert cmd.index("-init_hw_device") < cmd.index("-i")
|
||||
# The filter graph runs on the derived OpenCL device and tone-maps HDR→SDR to 8-bit nv12.
|
||||
assert cmd[cmd.index("-filter_hw_device") + 1] == "ocl"
|
||||
vf = cmd[cmd.index("-vf") + 1]
|
||||
assert "hwmap=derive_device=opencl" in vf
|
||||
assert "tonemap_opencl=" in vf and "format=nv12" in vf
|
||||
assert "hwmap=derive_device=vaapi:reverse=1" in vf
|
||||
assert "scale_vaapi" not in vf # the plain 8-bit path's filter must NOT be used here
|
||||
assert (
|
||||
cmd[cmd.index("-force_key_frames") + 1] == f"expr:gte(t,n_forced*{stream._SEG_SECONDS})"
|
||||
)
|
||||
|
||||
|
||||
class TestWantHw:
|
||||
"""VAAPI is used only where the render node was passed in AND (in auto) the source is HW-decodable
|
||||
AND 8-bit SDR, with an admin `plex_transcode_hw` policy (auto/on/off) on top."""
|
||||
"""VAAPI is used only where the render node was passed in AND (in auto) the source is HW-decodable,
|
||||
with an admin `plex_transcode_hw` policy (auto/on/off) on top. A HW session on a 10-bit/HDR source
|
||||
tone-maps on the GPU via OpenCL (tonemap=True) when the OpenCL runtime is present, else drops to
|
||||
software — see the `_select_encode` cases below."""
|
||||
|
||||
SRC = Path("/media/movie.mkv")
|
||||
|
||||
@@ -283,6 +309,11 @@ class TestWantHw:
|
||||
# Default the bit-depth probe so the auto-decodable tests don't shell out to ffprobe.
|
||||
monkeypatch.setattr(stream, "_probe_video_hwsafe", lambda src: safe)
|
||||
|
||||
@staticmethod
|
||||
def _opencl(monkeypatch, present):
|
||||
# Pin the OpenCL-runtime probe so the 10-bit routing doesn't depend on the CI box's /etc/OpenCL.
|
||||
monkeypatch.setattr(stream, "_opencl_present", lambda: present)
|
||||
|
||||
def test_no_hw_when_not_transcoding(self, monkeypatch):
|
||||
self._policy(monkeypatch, "auto")
|
||||
monkeypatch.setattr(stream, "_render_node_present", lambda: True)
|
||||
@@ -322,15 +353,38 @@ class TestWantHw:
|
||||
is False
|
||||
)
|
||||
|
||||
def test_auto_falls_back_to_software_for_a_10bit_or_hdr_source(self, monkeypatch):
|
||||
# BUG 1: a decodable HEVC that is 10-bit/HDR still breaks scale_vaapi=format=nv12 → software.
|
||||
def test_auto_10bit_source_falls_back_to_software_without_opencl(self, monkeypatch):
|
||||
# A decodable HEVC that is 10-bit/HDR can't use scale_vaapi=format=nv12 (Gen 9.5), and with no
|
||||
# OpenCL runtime to tonemap on the GPU it drops to software libx264 (BUG 1's behaviour).
|
||||
self._policy(monkeypatch, "auto")
|
||||
self._hwsafe(monkeypatch, False)
|
||||
self._opencl(monkeypatch, False)
|
||||
monkeypatch.setattr(stream, "_render_node_present", lambda: True)
|
||||
assert (
|
||||
stream._want_hw(None, _FakeItem(codec_video="hevc"), transcode=True, src=self.SRC)
|
||||
is False
|
||||
)
|
||||
assert stream._select_encode(
|
||||
None, _FakeItem(codec_video="hevc"), transcode=True, src=self.SRC
|
||||
) == (False, False)
|
||||
|
||||
def test_auto_10bit_source_uses_hw_tonemap_with_opencl(self, monkeypatch):
|
||||
# With the OpenCL runtime present, the same 10-bit/HDR HEVC tone-maps on the GPU (VAAPI decode →
|
||||
# OpenCL tonemap → VAAPI encode) instead of dropping to CPU — the HW HDR path.
|
||||
self._policy(monkeypatch, "auto")
|
||||
self._hwsafe(monkeypatch, False)
|
||||
self._opencl(monkeypatch, True)
|
||||
monkeypatch.setattr(stream, "_render_node_present", lambda: True)
|
||||
assert stream._select_encode(
|
||||
None, _FakeItem(codec_video="hevc"), transcode=True, src=self.SRC
|
||||
) == (True, True)
|
||||
|
||||
def test_auto_8bit_source_uses_plain_hw_not_tonemap(self, monkeypatch):
|
||||
# An 8-bit SDR source takes the plain full-GPU scale path — never the tonemap graph — even where
|
||||
# OpenCL is available.
|
||||
self._policy(monkeypatch, "auto")
|
||||
self._hwsafe(monkeypatch, True)
|
||||
self._opencl(monkeypatch, True)
|
||||
monkeypatch.setattr(stream, "_render_node_present", lambda: True)
|
||||
assert stream._select_encode(
|
||||
None, _FakeItem(codec_video="hevc"), transcode=True, src=self.SRC
|
||||
) == (True, False)
|
||||
|
||||
def test_off_forces_software_even_for_a_decodable_source(self, monkeypatch):
|
||||
self._policy(monkeypatch, "off")
|
||||
@@ -356,6 +410,17 @@ class TestWantHw:
|
||||
is False
|
||||
)
|
||||
|
||||
def test_on_still_picks_the_tonemap_graph_for_a_10bit_source(self, monkeypatch):
|
||||
# "on" trusts the GPU but must still route a 10-bit/HDR source through the tonemap graph (plain
|
||||
# scale_vaapi would hit the VPP wall) when OpenCL is present.
|
||||
self._policy(monkeypatch, "on")
|
||||
self._hwsafe(monkeypatch, False)
|
||||
self._opencl(monkeypatch, True)
|
||||
monkeypatch.setattr(stream, "_render_node_present", lambda: True)
|
||||
assert stream._select_encode(
|
||||
None, _FakeItem(codec_video="hevc"), transcode=True, src=self.SRC
|
||||
) == (True, True)
|
||||
|
||||
|
||||
class TestColdStartWarmup:
|
||||
"""The cold-start buffer warm-up: before reporting a session ready, start_session waits for the
|
||||
@@ -450,7 +515,9 @@ class TestTranscodeStartSession:
|
||||
def test_software_transcode_threads_the_knobs_into_the_command(
|
||||
self, capturing_env, monkeypatch
|
||||
):
|
||||
monkeypatch.setattr(stream, "_want_hw", lambda db, item, transcode, src: False)
|
||||
monkeypatch.setattr(
|
||||
stream, "_select_encode", lambda db, item, transcode, src: (False, False)
|
||||
)
|
||||
s = stream.start_session(
|
||||
db=None, item=_FakeItem(codec_video="mpeg4"), start_s=0.0, transcode=True
|
||||
)
|
||||
@@ -465,11 +532,13 @@ class TestTranscodeStartSession:
|
||||
assert kwargs.get("preexec_fn") is stream._sw_child_preexec
|
||||
|
||||
def test_hardware_transcode_uses_the_gpu_and_never_nices(self, capturing_env, monkeypatch):
|
||||
monkeypatch.setattr(stream, "_want_hw", lambda db, item, transcode, src: True)
|
||||
monkeypatch.setattr(
|
||||
stream, "_select_encode", lambda db, item, transcode, src: (True, False)
|
||||
)
|
||||
s = stream.start_session(
|
||||
db=None, item=_FakeItem(codec_video="hevc"), start_s=0.0, transcode=True
|
||||
)
|
||||
assert s is not None and s.transcode is True and s.hw is True
|
||||
assert s is not None and s.transcode is True and s.hw is True and s.tonemap is False
|
||||
assert len(_CapturingProc.calls) == 1 # first segment arrived — no HW→SW fallback respawn
|
||||
(cmd, *_), kwargs = _CapturingProc.calls[0]
|
||||
assert _venc(cmd) == "h264_vaapi"
|
||||
@@ -479,13 +548,31 @@ class TestTranscodeStartSession:
|
||||
) # no SW cap
|
||||
assert kwargs.get("preexec_fn") is None # GPU encode is cheap — never niced
|
||||
|
||||
def test_hdr_tonemap_transcode_threads_the_opencl_graph_and_flags_the_session(
|
||||
self, capturing_env, monkeypatch
|
||||
):
|
||||
# A HW tonemap session (10-bit/HDR + OpenCL) must build the OpenCL graph in the argv and carry
|
||||
# tonemap=True on the session, still on the GPU (hw=True) and never niced.
|
||||
monkeypatch.setattr(stream, "_select_encode", lambda db, item, transcode, src: (True, True))
|
||||
s = stream.start_session(
|
||||
db=None, item=_FakeItem(codec_video="hevc"), start_s=0.0, transcode=True
|
||||
)
|
||||
assert s is not None and s.transcode is True and s.hw is True and s.tonemap is True
|
||||
(cmd, *_), kwargs = _CapturingProc.calls[0]
|
||||
assert _venc(cmd) == "h264_vaapi"
|
||||
assert "opencl=ocl@va" in cmd and cmd[cmd.index("-filter_hw_device") + 1] == "ocl"
|
||||
assert "tonemap_opencl=" in cmd[cmd.index("-vf") + 1]
|
||||
assert kwargs.get("preexec_fn") is None
|
||||
|
||||
def test_a_hw_session_that_produces_nothing_falls_back_to_software(
|
||||
self, capturing_env, monkeypatch
|
||||
):
|
||||
# BUG 1 backstop: _want_hw picked HW but the VAAPI session emits no first segment (a VAProfile
|
||||
# refusal the gates didn't catch). start_session must respawn on libx264 and report hw=False so
|
||||
# the item still plays and the route/player honestly show CPU.
|
||||
monkeypatch.setattr(stream, "_want_hw", lambda db, item, transcode, src: True)
|
||||
monkeypatch.setattr(
|
||||
stream, "_select_encode", lambda db, item, transcode, src: (True, False)
|
||||
)
|
||||
seen = {"n": 0}
|
||||
|
||||
def flaky_wait(path, timeout=20.0, proc=None):
|
||||
@@ -510,7 +597,9 @@ class TestTranscodeStartSession:
|
||||
def test_remux_reads_no_transcode_knobs_and_stream_copies(self, capturing_env, monkeypatch):
|
||||
# A remux start must stream-copy the video and spawn no niced child (the transcode knobs and
|
||||
# the preexec nice are gated behind `transcode=True`).
|
||||
monkeypatch.setattr(stream, "_want_hw", lambda db, item, transcode, src: False)
|
||||
monkeypatch.setattr(
|
||||
stream, "_select_encode", lambda db, item, transcode, src: (False, False)
|
||||
)
|
||||
s = stream.start_session(db=None, item=_FakeItem(), start_s=0.0) # transcode defaults False
|
||||
assert s is not None and s.transcode is False
|
||||
(cmd, *_), kwargs = _CapturingProc.calls[0]
|
||||
|
||||
Reference in New Issue
Block a user