Jonathan Lennox
2016-Jun-30 16:35 UTC
[opus] Bug fix for configure output for Neon --disable-asm --enable-intrinsics
Tim pointed out to me last night that the configure output in the --disable-asm
--enable-intrinsics case was wrong with my aarch64 patches, producing
Run-time CPU detection: ........ no (NEON Intrinsics)
I had a patch in my aarch64 patch series that was supposed to fix this, but it
looks like it was buggy. The following patch applies over Tim’s aarch64v8
staging branch
https://git.xiph.org/?p=users/tterribe/opus.git;a=shortlog;h=refs/heads/aarch64v8
:
It should be squashed with the “Clean up formatting of configure output” patch
(e59b0f6) for the final commit.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.xiph.org/pipermail/opus/attachments/20160630/cd8c8028/attachment.html>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 0001-Fixed-version-of-configure-cleanup.patch.txt
URL:
<http://lists.xiph.org/pipermail/opus/attachments/20160630/cd8c8028/attachment.txt>
Maybe Matching Threads
- Opus floating-point NEON jump table question
- [RFC PATCHv1] cover: celt_pitch_xcorr: Introduce ARM neon intrinsics
- [RFC PATCHv1] cover: celt_pitch_xcorr: Introduce ARM neon intrinsics
- [PATCH 1/3] Add configure check for Aarch64-specific Neon intrinsics.
- [RFC PATCHv1] cover: celt_pitch_xcorr: Introduce ARM neon intrinsics