diff options
| author | Jack Lloyd <[email protected]> | 2018-11-04 02:34:23 -0500 |
|---|---|---|
| committer | Jack Lloyd <[email protected]> | 2018-11-04 04:19:02 -0500 |
| commit | a30ca750648bdb28f7d78ad8b6a17821e8917eb2 (patch) | |
| tree | 7cd43c0c24ce203b761f643e5aaea3bf924d13ce /src/build-data/policy/bsi.txt | |
| parent | eee2f37464b3de5760a8bdd26fd8deefd1643147 (diff) | |
Add ChaCha using SIMD_4x32
This allows supporting SSE2, NEON and AltiVec in a single codebase,
so drop the NEON and SSE2 code.
This new impl avoids having to do shuffles with every round and so
is about 10% faster on Skylake.
Also, fix bugs in both baseline and AVX2 implementations when the
low counter overflowed. The SSE2 and NEON code were also buggy here.
Diffstat (limited to 'src/build-data/policy/bsi.txt')
| -rw-r--r-- | src/build-data/policy/bsi.txt | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/src/build-data/policy/bsi.txt b/src/build-data/policy/bsi.txt index f49ab79b0..763ca69e1 100644 --- a/src/build-data/policy/bsi.txt +++ b/src/build-data/policy/bsi.txt @@ -121,7 +121,8 @@ cfb # stream chacha -chacha_sse2 +chacha_simd32 +chacha_avx2 ofb rc4 salsa20 |