mirror/gmsm
1
0
Fork 0
mirror of https://github.com/emmansun/gmsm.git synced 2025-05-14 04:56:21 +08:00
Code Issues Packages Projects Releases Wiki Activity

kdf-sm3: optimize arm64

Browse Source
This commit is contained in:
Sun Yimin 2024-05-21 17:38:04 +08:00 committed by GitHub
parent a81432fbf8
commit 7c832f65e7
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 6 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
sm3/sm3blocks_arm64.s
View File

@ -197,7 +197,9 @@ TEXT ·blockMultBy4(SB), NOSPLIT, $0
MOVD (srcPtrPtr), srcPtr4 MOVD (srcPtrPtr), srcPtr4
loop: loop:
// reset wordPtr
MOVD wordStart, wordPtr MOVD wordStart, wordPtr
// load message block // load message block
prepare4Words prepare4Words
prepare4Words prepare4Words
@ -271,9 +273,8 @@ loop:
ROUND_16_63(62, T30, c, d, e, f, g, h, a, b) ROUND_16_63(62, T30, c, d, e, f, g, h, a, b)
ROUND_16_63(63, T31, b, c, d, e, f, g, h, a) ROUND_16_63(63, T31, b, c, d, e, f, g, h, a)
MOVD statePtr, R20 VLD1.P 64(statePtr), [V8.S4, V9.S4, V10.S4, V11.S4]
VLD1.P 64(R20), [V8.S4, V9.S4, V10.S4, V11.S4] VLD1 (statePtr), [V12.S4, V13.S4, V14.S4, V15.S4]
VLD1 (R20), [V12.S4, V13.S4, V14.S4, V15.S4]
VEOR a.B16, V8.B16, a.B16 VEOR a.B16, V8.B16, a.B16
VEOR b.B16, V9.B16, b.B16 VEOR b.B16, V9.B16, b.B16
VEOR c.B16, V10.B16, c.B16 VEOR c.B16, V10.B16, c.B16
@ -282,9 +283,8 @@ loop:
VEOR f.B16, V13.B16, f.B16 VEOR f.B16, V13.B16, f.B16
VEOR g.B16, V14.B16, g.B16 VEOR g.B16, V14.B16, g.B16
VEOR h.B16, V15.B16, h.B16 VEOR h.B16, V15.B16, h.B16
MOVD statePtr, R20 VST1.P [e.S4, f.S4, g.S4, h.S4], -64(statePtr)
VST1.P [a.S4, b.S4, c.S4, d.S4], 64(R20) VST1 [a.S4, b.S4, c.S4, d.S4], (statePtr)
VST1 [e.S4, f.S4, g.S4, h.S4], (R20)
SUB $1, blockCount SUB $1, blockCount
CBNZ blockCount, loop CBNZ blockCount, loop

8
sm3/sm3blocks_test.go
View File

@ -3,7 +3,6 @@
package sm3 package sm3
import ( import (
"encoding/binary"
"fmt" "fmt"
"testing" "testing"
) )
@ -54,13 +53,6 @@ func TestBlockMultBy4(t *testing.T) {
buffer := make([]byte, preallocSize) buffer := make([]byte, preallocSize)
blockMultBy4(&digs[0], &p[0], &buffer[0], 1) blockMultBy4(&digs[0], &p[0], &buffer[0], 1)
expected := "[66c7f0f4 62eeedd9 d1f2d46b dc10e4e2 4167c487 5cf2f7a2 297da02b 8f4ba8e0]" expected := "[66c7f0f4 62eeedd9 d1f2d46b dc10e4e2 4167c487 5cf2f7a2 297da02b 8f4ba8e0]"
for i := 0; i < 128+68*4*4; i += 64 {
fmt.Printf("%08x %08x %08x %08x ", binary.LittleEndian.Uint32(buffer[i:]), binary.LittleEndian.Uint32(buffer[i+4:]), binary.LittleEndian.Uint32(buffer[i+8:]), binary.LittleEndian.Uint32(buffer[i+12:]))
fmt.Printf("%08x %08x %08x %08x ", binary.LittleEndian.Uint32(buffer[i+16:]), binary.LittleEndian.Uint32(buffer[i+20:]), binary.LittleEndian.Uint32(buffer[i+24:]), binary.LittleEndian.Uint32(buffer[i+28:]))
fmt.Printf("%08x %08x %08x %08x ", binary.LittleEndian.Uint32(buffer[i+32:]), binary.LittleEndian.Uint32(buffer[i+36:]), binary.LittleEndian.Uint32(buffer[i+40:]), binary.LittleEndian.Uint32(buffer[i+44:]))
fmt.Printf("%08x %08x %08x %08x ", binary.LittleEndian.Uint32(buffer[i+48:]), binary.LittleEndian.Uint32(buffer[i+52:]), binary.LittleEndian.Uint32(buffer[i+56:]), binary.LittleEndian.Uint32(buffer[i+60:]))
fmt.Println()
}
s := fmt.Sprintf("%x", digs[0][:]) s := fmt.Sprintf("%x", digs[0][:])
if s != expected { if s != expected {
t.Errorf("digs[0] got %s", s) t.Errorf("digs[0] got %s", s)