Repository navigation
Conversation
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
Thalley
force-pushed
the
optimizations
branch
from
October 20, 2026 13:23
65a0c7e to
04cb2a7
Compare
Thalley
marked this pull request as ready for review
October 20, 2026 13:26
added 4 commits
October 22, 2026 11:48
copysignf is a branchless optimization for when *xs can be positive or negative. Estimated optimization at ~1.3%. Signed-off-by: Emil Gydesen <thalley@thalley.com>
Instead of computing it for nc1 and nc2 individually, compute it once before and apply to both. Estimated optimization at ~0.04% Signed-off-by: Emil Gydesen <thalley@thalley.com>
This commit optimizes the loop in fft_bf2 by avoiding duplicated computations, and instead compute some results once instead. Estimated optimization at ~2.07% Signed-off-by: Emil Gydesen <thalley@thalley.com>
By caching the results in a stack allocated xq, some duplicated computations could be avoided. This is a tradeoff between stack usage and cycles. Estimated optimziation at ~10.55%. Signed-off-by: Emil Gydesen <thalley@thalley.com>
Thalley
force-pushed
the
optimizations
branch
from
October 22, 2026 09:53
04cb2a7 to
9a70d4d
Compare
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Some various minor optimizations. Tested with
perfon x86 to be around ~13-14% total optimization for the encoder, and ~2-3% for the decoder.For ARM Cortex-M33 we are seeing