[FFmpeg-devel] [PATCH] x86/lpc: write a new Welch windowing function

[Lynne]

Old one was written with the assumption only even inputs would be given.
This very messy replacement supports even and odd inputs, and supports
AVX2 for extra speed. The buffers given are usually quite big (4k samples),
so the speedup is worth it.
The new SSE version is still faster than the old inline asm version.

The only place where this function mismatches the C version is with very
small odd buffers (3 samples when using AVX2). I haven't found
such scenarios possible.

Also checkasm is provided to make sure this monstrosity works.

This fixes some FATE tests.

Patch attached.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-x86-lpc-write-a-new-Welch-windowing-function.patch
Type: text/x-diff
Size: 17985 bytes
Desc: not available
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20220920/b22efccc/attachment.patch>