2025-05-26 - 2026-05-26
Overview
37 Pull requests merged by 1 user
Merged
#37 v3d_runner: SPV path search + bench preflight — RETRACTS PR #36's headline
Merged
#36 bench: H.264 primitive bench now measures both substrates + comparison table
Merged
#35 h264: V3D shaders for the 4 bS=4 intra deblock variants — deblock QPU complete
Merged
#34 h264: V3D shaders for all 15 avg_ qpel positions — qpel QPU complete
Merged
#33 h264: V3D shaders for the 8 diagonal qpel positions
Merged
#32 h264: V3D shaders for the 4 single-axis quarter-pel qpel variants
Merged
#31 h264: V3D shader for qpel mc22 (2D half-pel 'j' position)
Merged
#30 h264: V3D shader for qpel mc02 (vertical half-pel)
Merged
#29 h264: V3D shaders for chroma deblock V + H (4:2:0)
Merged
#28 h264: V3D shader for deblock_luma_h — first QPU port since cycle 9
Merged
#27 h264: promote remaining intra prediction modes (17) to public API
Merged
#26 h264: promote Intra_4x4 luma prediction (9 modes) to public API
Merged
#25 h264: expose chroma DC 2x2 Hadamard as public API
Merged
#24 bench: H.264 primitives NEON CPU baseline (1080p budget projection)
Merged
#23 h264: chroma DC 2x2 Hadamard pre-pass primitive
Merged
#22 h264: Intra_8x8 luma — 6 directional modes (DDL/DDR/VR/HD/VL/HU)
Merged
#21 h264: Intra_8x8 luma (High profile) — pre-filter + 3 modes (V/H/DC)
Merged
#20 h264: qpel avg — 12 remaining variants (closes the matrix)
Merged
#19 h264: qpel avg anchors (avg_mc20/02/22, biprediction support)
Merged
#18 h264: qpel diagonals — 8 positions (mc11/12/13/21/23/31/32/33)
Merged
#17 h264: qpel single-axis quarter-pel — mc10/mc30/mc01/mc03 (CPU/NEON)
Merged
#16 h264: qpel mc22 (2D half-pel, CPU/NEON)
Merged
#15 h264: qpel mc02 (vertical half-pel, CPU/NEON)
Merged
#14 h264: Intra_8x8 chroma prediction — 4-mode C reference + spec gates
Merged
#13 h264: Intra_16x16 luma prediction — 4-mode C reference + spec gates
Merged
#12 h264: Intra_4x4 luma prediction — 9-mode C reference + spec gates
Merged
#11 h264: deblock bS=4 intra variants (luma + chroma, V + H)
Merged
#10 h264: deblock chroma_v + chroma_h (CPU/NEON, bS<4)
Merged
#9 h264: deblock_luma_h (CPU/NEON via vendored ff_h264_h_loop_filter)
Merged
#8 cycle 9: V3D shader for H.264 luma qpel mc20 — closes 9/9 QPU coverage
Merged
#7 QPU is default substrate: recipe table + ctx env-var override
Merged
#6 v3d_runner: buffer pool for QPU dispatch hot path
Merged
#4 docs: architecture backlog — correct fleet hardware mapping
Merged
#5 CMakeLists: make daedalus-fourier.pc relocatable via ${pcfiledir}
Merged
#3 docs: architecture backlog for multi-SoC daedalus generalization
Merged
#2 Phase 8c: H.264 luma qpel mc20 through public API
Merged
#1 CMakeLists: install rules + pkg-config for daedalus_core