Carlos Eduardo Seo
20be4c3c3e
Add asm implementation for chacha20 using vector instructions on ppc64le. Below, the difference using the new code: name old speed new speed delta ChaCha20/32-16 167MB/s ± 0% 129MB/s ± 0% -22.60% (p=0.008 n=5+5) ChaCha20/63-16 308MB/s ± 0% 249MB/s ± 0% -19.00% (p=0.008 n=5+5) ChaCha20/64-16 357MB/s ± 0% 251MB/s ± 0% -29.57% (p=0.008 n=5+5) ChaCha20/256-16 398MB/s ± 0% 1199MB/s ± 0% +201.20% (p=0.008 n=5+5) ChaCha20/1024-16 413MB/s ± 0% 1398MB/s ± 0% +238.67% (p=0.008 n=5+5) ChaCha20/1350-16 395MB/s ± 0% 1189MB/s ± 0% +200.71% (p=0.008 n=5+5) ChaCha20/65536-16 420MB/s ± 0% 1489MB/s ± 0% +254.10% (p=0.008 n=5+5) Small sizes are slower due the fact that it always calculates using 256 bytes of key stream. This change was originally from Paulo Flabiano Smorigo <pfsmorigo@linux.vnet.ibm.com> and started as CL 108999 (https://go-review.googlesource.com/c/crypto/+/108999). Fixes golang/go#25051 Change-Id: Ie510494249b227379e23d993467256b3d4088035 Reviewed-on: https://go-review.googlesource.com/c/crypto/+/172177 Run-TryBot: Carlos Eduardo Seo <cseo@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
52 lines
1.6 KiB
Go
52 lines
1.6 KiB
Go
// Copyright 2019 The Go Authors. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style
|
|
// license that can be found in the LICENSE file.
|
|
|
|
// +build ppc64le,!gccgo,!appengine
|
|
|
|
package chacha20
|
|
|
|
import "encoding/binary"
|
|
|
|
var haveAsm = true
|
|
|
|
const bufSize = 256
|
|
|
|
//go:noescape
|
|
func chaCha20_ctr32_vmx(out, inp *byte, len int, key *[8]uint32, counter *uint32)
|
|
|
|
func (c *Cipher) xorKeyStreamAsm(dst, src []byte) {
|
|
if len(src) >= bufSize {
|
|
chaCha20_ctr32_vmx(&dst[0], &src[0], len(src)-len(src)%bufSize, &c.key, &c.counter)
|
|
}
|
|
if len(src)%bufSize != 0 {
|
|
chaCha20_ctr32_vmx(&c.buf[0], &c.buf[0], bufSize, &c.key, &c.counter)
|
|
start := len(src) - len(src)%bufSize
|
|
ts, td, tb := src[start:], dst[start:], c.buf[:]
|
|
// Unroll loop to XOR 32 bytes per iteration.
|
|
for i := 0; i < len(ts)-32; i += 32 {
|
|
td, tb = td[:len(ts)], tb[:len(ts)] // bounds check elimination
|
|
s0 := binary.LittleEndian.Uint64(ts[0:8])
|
|
s1 := binary.LittleEndian.Uint64(ts[8:16])
|
|
s2 := binary.LittleEndian.Uint64(ts[16:24])
|
|
s3 := binary.LittleEndian.Uint64(ts[24:32])
|
|
b0 := binary.LittleEndian.Uint64(tb[0:8])
|
|
b1 := binary.LittleEndian.Uint64(tb[8:16])
|
|
b2 := binary.LittleEndian.Uint64(tb[16:24])
|
|
b3 := binary.LittleEndian.Uint64(tb[24:32])
|
|
binary.LittleEndian.PutUint64(td[0:8], s0^b0)
|
|
binary.LittleEndian.PutUint64(td[8:16], s1^b1)
|
|
binary.LittleEndian.PutUint64(td[16:24], s2^b2)
|
|
binary.LittleEndian.PutUint64(td[24:32], s3^b3)
|
|
ts, td, tb = ts[32:], td[32:], tb[32:]
|
|
}
|
|
td, tb = td[:len(ts)], tb[:len(ts)] // bounds check elimination
|
|
for i, v := range ts {
|
|
td[i] = tb[i] ^ v
|
|
}
|
|
c.len = bufSize - (len(src) % bufSize)
|
|
|
|
}
|
|
|
|
}
|