rusty_h264-v0.17.0

Other

• overflow-checks cost 22.5% of decode — moved to release-checked
• audit pass 2 — 6% to 89% of in-scope gates, every unit >=85%
• close H-03/04/05/07/13/14/15/16/31/39 + threat model
• use-protection-please first pass — 41 gates x 8 units, depth=deep

rusty_h264-encoder-v0.17.0

Fixed

• the profile census compiles again, and the decoder builds no_std (#20)

Other

• make main green — fix the root cause behind each failing gate
• narrow the motion grids to (i16,i16) - 2.90 GB of pool clear removed
• four more, from the encoder and the PPS-gated grids
• narrow the reference-index grids i32 -> i8 (-103 MB of memset)
• overflow-checks cost 22.5% of decode — moved to release-checked
• audit pass 2 — 6% to 89% of in-scope gates, every unit >=85%
• close H-03/04/05/07/13/14/15/16/31/39 + threat model
• use-protection-please first pass — 41 gates x 8 units, depth=deep

rusty_h264-decoder-v0.17.0

Fixed

• the profile census compiles again, and the decoder builds no_std (#20)

Other

• Merge remote-tracking branch 'origin/main' into fix16
• gate the field-picture refusal too (#16)
• Decode frame pictures of streams that permit field coding
• 87% of every allocation was one Box::new that skipped its own pool
• the plane recycle pool was never fed - 4.34 GB of wasted zeroing
• narrow the motion grids to (i16,i16) - 2.90 GB of pool clear removed
• the List-1 grids are B-only, so stop clearing them without B (-36%)
• four more, from the encoder and the PPS-gated grids
• the mvd cache stores what the context can read, not what was parsed
• second pass of the copy hunt — the census's own two blind spots
• carry the POC lookup table, not its per-block expansion (-37 MB)
• narrow the reference-index grids i32 -> i8 (-103 MB of memset)
• overflow-checks cost 22.5% of decode — moved to release-checked
• audit pass 2 — 6% to 89% of in-scope gates, every unit >=85%
• close H-03/04/05/07/13/14/15/16/31/39 + threat model
• use-protection-please first pass — 41 gates x 8 units, depth=deep
• no production panic sites — 47 to 0 (H-18)
• put the 8x8-transform flag in the intra table, so the filter stops asking
• move the derivation's output pointer into a register argument
• build the packed record where it lives (spills -72%)
• mark single-partition B macroblocks uniform (mb_uniform now -86..-96%)
• coded flags a word at a time, and a provable kind row
• pack a uniform macroblock from one block, not sixteen
• mark the B classes whose motion the decoder already knows is uniform
• one motion test per macroblock edge when both sides are uniform
• take the uniformity answer from mb_kind instead of recomputing it
• derive boundary strengths at u8, not i32 (-223 instrs on the decoder's derivation path)

rusty_h264-common-v0.17.0

Other

• (miri) make the H-23 gate finish — fix one real test bug, skip the sweeps
• make main green — fix the root cause behind each failing gate
• bump rusty_alloc-api to =2.2.6 (#19)
• bump rusty_alloc-api to =2.2.5 (#17)
• 87% of every allocation was one Box::new that skipped its own pool
• narrow the motion grids to (i16,i16) - 2.90 GB of pool clear removed
• narrow the reference-index grids i32 -> i8 (-103 MB of memset)
• overflow-checks cost 22.5% of decode — moved to release-checked
• audit pass 2 — 6% to 89% of in-scope gates, every unit >=85%
• close H-03/04/05/07/13/14/15/16/31/39 + threat model
• use-protection-please first pass — 41 gates x 8 units, depth=deep
• put the 8x8-transform flag in the intra table, so the filter stops asking
• record one more refutation on the derivation's argument list
• move the derivation's output pointer into a register argument
• build the packed record where it lives (spills -72%)
• add a spill column to the row-hook census, and record what it settled
• outline the per-lane edge arms, now the minority case
• the filter keeps boundary strengths at byte width
• an intra record is one boolean, and the filter stops touching a TLS
• coded flags a word at a time, and a provable kind row
• pack a uniform macroblock from one block, not sixteen
• one motion test per macroblock edge when both sides are uniform
• take the uniformity answer from mb_kind instead of recomputing it
• fold the paired motion tests, share the edge nnz mask, split the hot prefix
• derive boundary strengths at u8, not i32 (-223 instrs on the decoder's derivation path)

rusty_h264-accel-v0.17.0

Other

• overflow-checks cost 22.5% of decode — moved to release-checked
• audit pass 2 — 6% to 89% of in-scope gates, every unit >=85%
• close H-03/04/05/07/13/14/15/16/31/39 + threat model
• use-protection-please first pass — 41 gates x 8 units, depth=deep
• an intra record is one boolean, and the filter stops touching a TLS
• pack a uniform macroblock from one block, not sixteen
• one motion test per macroblock edge when both sides are uniform