aboutsummaryrefslogtreecommitdiff
path: root/src/Compilation.zig
diff options
context:
space:
mode:
authorFrank Denis <github@pureftpd.org>2020-10-23 16:18:35 +0200
committerAndrew Kelley <andrew@ziglang.org>2020-10-25 21:13:14 -0400
commit72064eba233fd3fd8e3dc0ab6f5fc73c6bdf1922 (patch)
treed4f479fb0ee636c75f23810e751190a6050eae2e /src/Compilation.zig
parent1b4ab749cffeff146b21fb0f6d30a5b8d2186765 (diff)
downloadzig-72064eba233fd3fd8e3dc0ab6f5fc73c6bdf1922.tar.gz
zig-72064eba233fd3fd8e3dc0ab6f5fc73c6bdf1922.zip
std/crypto: vectorize BLAKE3
Gives a ~40% speedup on x86_64. However, the generic code remains faster on aarch64. This is still processing only one block at a time for now. I'm pretty confident that processing more blocks per round will eventually give a substantial performance improvement on all platforms with vector units.
Diffstat (limited to 'src/Compilation.zig')
0 files changed, 0 insertions, 0 deletions