Repository navigation
Speed up the TCQ q1 decide in the NEON st8 loop and the q1 rate C code - #5493
Merged
urvangjoshi merged 4 commits intoOct 15, 2026
Merged
Conversation
jjustiss-apple
changed the title
Contributor
Author
|
This PR depends on #5489 to land first as that PR creates |
Contributor
|
there'll be merge conflict with #5489 right? Can you merge that and resolve conflicts before I review? Thanks. |
Contributor
Author
Yes, there'll definitely be merge conflicts. I'll ping you once the merge is clean and the PR is ready for review. |
jjustiss-apple
marked this pull request as draft
October 8, 2026 16:11
jjustiss-apple
force-pushed
the
jjustiss/tcq-q1-neon
branch
from
October 8, 2026 16:43
9e80ca4 to
4f5ffea
Compare
jjustiss-apple
marked this pull request as ready for review
October 9, 2026 16:40
Contributor
Author
|
@jianj-g This one is rebased and ready for review |
jianj-g
reviewed
Oct 13, 2026
jjustiss-apple
changed the title
jianj-g
approved these changes
Oct 15, 2026
urvangjoshi
force-pushed
the
jjustiss/tcq-q1-neon
branch
from
October 15, 2026 15:57
8cc1b7a to
003b2e9
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Make the q1 state decision in the NEON st8 loop branchless (compare, blend
and store in NEON registers). The old per-state branches mispredict on real
content. av2_decide_states_q1_neon is a thin RTCD wrapper around the same
impl for unit testing, as in AVX2.
The C q1 rate functions read the packed cost tables directly instead of the
generic cost path.
CTC (RA, cpu-used 1, 33 frames, Apple Silicon):