ardour/ardour
ardour/ardour
13
1
Fork 0
Code Pull Requests Releases Activity

Fix Intel/AMD FMA mix_buffers_with_gain()

Browse Source 
_mm256_fmadd_ps (a, b, c) performs (a * b + c)
This commit is contained in:
Robin Gareus 2021-01-18 15:28:24 +01:00
parent 097e5a7145
commit 1fc8b55314
Signed by: rgareus
GPG Key ID: A090BCE02CF57F04
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
libs/ardour/x86_functions_fma.cc
View File

@ -1,5 +1,6 @@
/*
* Copyright (C) 2020 Ayan Shafqat <ayan.x.shafqat@gmail.com>
* Copyright (C) 2021 Robin Gareus <robin@gareus.org>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@ -82,8 +83,8 @@ x86_fma_mix_buffers_with_gain(
d1 = _mm256_load_ps(dst + 8);
// dst = dst + (src * gain)
d0 = _mm256_fmadd_ps(g0, d0, s0);
d1 = _mm256_fmadd_ps(g0, d1, s1);
d0 = _mm256_fmadd_ps(g0, s0, d0);
d1 = _mm256_fmadd_ps(g0, s1, d1);
// Store result
_mm256_store_ps(dst + 0, d0);