From 40bc2f5f84f9fcd882995168c988b7d18809ddcb Mon Sep 17 00:00:00 2001
From: Sun Yimin <emmansun@users.noreply.github.com>
Date: Tue, 3 Sep 2024 15:20:10 +0800
Subject: [PATCH] debug 1

---
 sm3/sm3blocks_s390x.s       | 32 +++++++++++++-------------------
 sm3/sm3blocks_s390x_test.go |  5 ++++-
 2 files changed, 17 insertions(+), 20 deletions(-)

diff --git a/sm3/sm3blocks_s390x.s b/sm3/sm3blocks_s390x.s
index 114d798..b9a532b 100644
--- a/sm3/sm3blocks_s390x.s
+++ b/sm3/sm3blocks_s390x.s
@@ -5,6 +5,7 @@
 //go:build !purego
 
 #include "textflag.h"
+#include "go_asm.h"
 
 DATA mask<>+0x00(SB)/8, $0x0001020310111213
 DATA mask<>+0x08(SB)/8, $0x0405060714151617
@@ -31,34 +32,27 @@ TEXT ·transposeMatrix(SB),NOSPLIT,$0
 	MOVD	dig+0(FP), R1
 
 	MOVD 	(R1), R2
-	VL 0(R2), V0
-	VL 16(R2), V4
+	VLM (R2), V0, V1
 	MOVD 	8(R1), R2
-	VL 0(R2), V1
-	VL 16(R2), V5
+	VLM (R2), V2, V3
 	MOVD 	16(R1), R2
-	VL 0(R2), V2
-	VL 16(R2), V6
+	VLM (R2), V4, V5
 	MOVD 	24(R1), R2
-	VL 0(R2), V3
-	VL 16(R2), V7
+	VLM (R2), V6, V7
 
 	MOVD $mask<>+0x00(SB), R2
-	VLM 0(R2), V8, V11
+	VLM (R2), V8, V11
 
-	TRANSPOSE_MATRIX(V0, V1, V2, V3, V8, V9, V10, V11, V12, V13, V14, V15)
-	TRANSPOSE_MATRIX(V4, V5, V6, V7, V8, V9, V10, V11, V12, V13, V14, V15)
+	TRANSPOSE_MATRIX(V0, V2, V4, V6, V8, V9, V10, V11, V12, V13, V14, V15)
+	TRANSPOSE_MATRIX(V1, V3, V5, V7, V8, V9, V10, V11, V12, V13, V14, V15)
 
 	MOVD 	(R1), R2
-	VST V0, 0(R2)
-	VST V4, 16(R2)
+	VSTM V0, V1, (R2)
 	MOVD 	8(R1), R2
-	VST V1, 0(R2)
-	VST V5, 16(R2)
+	VSTM V2, V3, (R2)
 	MOVD 	16(R1), R2
-	VST V2, 0(R2)
-	VST V6, 16(R2)
+	VSTM V4, V5, (R2)
 	MOVD 	24(R1), R2
-	VST V3, 0(R2)
-	VST V7, 16(R2)
+	VSTM V6, V7, (R2)
+
 	RET
diff --git a/sm3/sm3blocks_s390x_test.go b/sm3/sm3blocks_s390x_test.go
index eda869e..08a29ff 100644
--- a/sm3/sm3blocks_s390x_test.go
+++ b/sm3/sm3blocks_s390x_test.go
@@ -9,15 +9,18 @@ import (
 
 func TestTransposeMatrix(t *testing.T) {
 	var m [4][8]uint32
+	var k uint32 = 0
 	for i := 0; i < 4; i++ {
 		for j := 0; j < 8; j++ {
-			m[i][j] = uint32(i*4 + j)
+			m[i][j] = k
+			k++
 			fmt.Printf("%04x ", m[i][j])
 		}
 		fmt.Println()
 	}
 	input := [4]*[8]uint32{&m[0], &m[1], &m[2], &m[3]}
 	transposeMatrix(&input[0])
+	fmt.Println()
 	for i := 0; i < 4; i++ {
 		for j := 0; j < 8; j++ {
 			m[i][j] = uint32(i*4 + j)