diff options
| author | cinap_lenrek <cinap_lenrek@felloff.net> | 2017-11-20 00:10:35 +0100 |
|---|---|---|
| committer | cinap_lenrek <cinap_lenrek@felloff.net> | 2017-11-20 00:10:35 +0100 |
| commit | 077e719dfbf9bf2582bed80026251cc0d108c16e (patch) | |
| tree | 5e8fd7d6297f5d007ea21c85f8346eae0717ed7d /lib/unicode | |
| parent | 1eb373945455f1ba03fa1b221529d74ca2a778ad (diff) | |
| download | plan9front-077e719dfbf9bf2582bed80026251cc0d108c16e.tar.xz | |
libsec: write optimized _chachablock() function for amd64 / sse2
doing 4 quarterround's in parallel using 128-bit
vector registers. for second round shuffle the columns and
then shuffle back.
code is rather obvious. only trick here is for the first
quaterround PSHUFLW/PSHUFHW is used to swap the halfwords
for the <<<16 rotation.
Diffstat (limited to 'lib/unicode')
0 files changed, 0 insertions, 0 deletions