cmd/compile/internal: optimize multiplication use new operation 'ADDshiftLLV' on loong64

goos: linux
goarch: loong64
pkg: cmd/compile/internal/test
cpu: Loongson-3A6000-HV @ 2500.00MHz
                  │     old      │                 new                  │
                  │    sec/op    │    sec/op     vs base                │
MulconstI32/3       0.8004n ± 0%   0.4247n ± 2%  -46.94% (p=0.000 n=10)
MulconstI32/5       0.8005n ± 0%   0.4256n ± 1%  -46.83% (p=0.000 n=10)
MulconstI32/12      1.2010n ± 0%   0.8005n ± 0%  -33.35% (p=0.000 n=10)
MulconstI32/120     0.8090n ± 0%   0.8067n ± 0%   -0.28% (p=0.007 n=10)
MulconstI32/-120    0.8109n ± 0%   0.8072n ± 0%   -0.47% (p=0.000 n=10)
MulconstI32/65537   0.8004n ± 0%   0.8004n ± 0%        ~ (p=1.000 n=10)
MulconstI32/65538   0.8005n ± 0%   0.8005n ± 0%        ~ (p=0.265 n=10)
MulconstI64/3       0.8005n ± 0%   0.4241n ± 1%  -47.02% (p=0.000 n=10)
MulconstI64/5       0.8004n ± 0%   0.4249n ± 1%  -46.91% (p=0.000 n=10)
MulconstI64/12      1.2010n ± 0%   0.8004n ± 0%  -33.36% (p=0.000 n=10)
MulconstI64/120     0.8005n ± 0%   0.8005n ± 0%        ~ (p=0.635 n=10)
MulconstI64/-120    0.8005n ± 0%   0.8005n ± 0%        ~ (p=0.837 n=10)
MulconstI64/65537   0.8005n ± 0%   0.8005n ± 0%        ~ (p=0.837 n=10)
MulconstI64/65538   0.8096n ± 0%   0.8004n ± 0%   -1.14% (p=0.000 n=10)
MulconstU32/3       0.8004n ± 0%   0.4263n ± 1%  -46.75% (p=0.000 n=10)
MulconstU32/5       0.8005n ± 0%   0.4262n ± 1%  -46.76% (p=0.000 n=10)
MulconstU32/12      1.2010n ± 0%   0.8005n ± 0%  -33.35% (p=0.000 n=10)
MulconstU32/120     0.8105n ± 0%   0.8096n ± 0%        ~ (p=0.183 n=10)
MulconstU32/65537   0.8004n ± 0%   0.8004n ± 0%        ~ (p=1.000 n=10)
MulconstU32/65538   0.8005n ± 0%   0.8005n ± 0%        ~ (p=1.000 n=10)
MulconstU64/3       0.8004n ± 0%   0.4265n ± 4%  -46.71% (p=0.000 n=10)
MulconstU64/5       0.8004n ± 0%   0.4256n ± 0%  -46.82% (p=0.000 n=10)
MulconstU64/12      1.2010n ± 0%   0.8004n ± 0%  -33.36% (p=0.000 n=10)
MulconstU64/120     0.8005n ± 0%   0.8005n ± 0%        ~ (p=0.387 n=10)
MulconstU64/65537   0.8005n ± 0%   0.8005n ± 0%        ~ (p=0.265 n=10)
MulconstU64/65538   0.8080n ± 0%   0.8004n ± 0%   -0.93% (p=0.000 n=10)
geomean             0.8539n        0.6597n       -22.74%

Change-Id: Ie33e88985d7639f481bbba540bc917b9f185c357
Reviewed-on: https://go-review.googlesource.com/c/go/+/693855
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: sophie zhao <zhaoxiaolin@loongson.cn>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>

2 files changed, 49 insertions, 49 deletions

diff --git a/test/codegen/arithmetic.go b/test/codegen/arithmetic.go
index 39a7986c7b..67adb50fa5 100644
--- a/test/codegen/arithmetic.go
+++ b/test/codegen/arithmetic.go
@@ -257,7 +257,7 @@ func Mul_96(n int) int {
 	// 386:`SHLL\t[$]5`,`LEAL\t\(.*\)\(.*\*2\),`,-`IMULL`
 	// arm64:`LSL\t[$]5`,`ADD\sR[0-9]+<<1,\sR[0-9]+`,-`MUL`
 	// arm:`SLL\t[$]5`,`ADD\sR[0-9]+<<1,\sR[0-9]+`,-`MUL`
-	// loong64:"ADDVU","SLLV\t[$]5",-"MULV"
+	// loong64:"SLLV\t[$]5","ALSLV\t[$]1,"
 	// s390x:`SLD\t[$]5`,`SLD\t[$]6`,-`MULLD`
 	return n * 96
 }
@@ -317,7 +317,7 @@ func MergeMuls5(a, n int) int {
 // Multiplications folded negation
 
 func FoldNegMul(a int) int {
-	// loong64:"MULV","MOVV\t[$]-11",-"SUBVU\tR[0-9], R0,"
+	// loong64:"SUBVU","ALSLV\t[$]2","ALSLV\t[$]1"
 	return (-a) * 11
 }
 
diff --git a/test/codegen/multiply.go b/test/codegen/multiply.go
index bb22d1a2b1..dc2910dab7 100644
--- a/test/codegen/multiply.go
+++ b/test/codegen/multiply.go
@@ -24,7 +24,7 @@ func m2(x int64) int64 {
 func m3(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2"
 	// arm64: "ADD\tR[0-9]+<<1,"
-	// loong64: "ADDVU","ADDVU"
+	// loong64: "ALSLV\t[$]1,"
 	return x * 3
 }
 func m4(x int64) int64 {
@@ -36,19 +36,19 @@ func m4(x int64) int64 {
 func m5(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]4"
 	// arm64: "ADD\tR[0-9]+<<2,"
-	// loong64: "SLLV\t[$]2,","ADDVU"
+	// loong64: "ALSLV\t[$]2,"
 	return x * 5
 }
 func m6(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]1", "LEAQ\t.*[*]2"
 	// arm64: "ADD\tR[0-9]+,", "ADD\tR[0-9]+<<1,"
-	// loong64: "ADDVU","ADDVU","ADDVU"
+	// loong64: "ADDVU", "ADDVU", "ADDVU"
 	return x * 6
 }
 func m7(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2"
 	// arm64: "LSL\t[$]3,", "SUB\tR[0-9]+,"
-	// loong64: "SLLV\t[$]3,","SUBVU"
+	// loong64: "ALSLV\t[$]1,", "ALSLV\t[$]1,"
 	return x * 7
 }
 func m8(x int64) int64 {
@@ -60,43 +60,43 @@ func m8(x int64) int64 {
 func m9(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]8"
 	// arm64: "ADD\tR[0-9]+<<3,"
-	// loong64: "SLLV\t[$]3,","ADDVU"
+	// loong64: "ALSLV\t[$]3,"
 	return x * 9
 }
 func m10(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]1", "LEAQ\t.*[*]4"
 	// arm64: "ADD\tR[0-9]+,", "ADD\tR[0-9]+<<2,"
-	// loong64: "ADDVU","SLLV\t[$]3,","ADDVU"
+	// loong64: "ADDVU", "ALSLV\t[$]2,"
 	return x * 10
 }
 func m11(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]4", "LEAQ\t.*[*]2"
 	// arm64: "MOVD\t[$]11,", "MUL"
-	// loong64: "MOVV\t[$]11,", "MULV"
+	// loong64: "ALSLV\t[$]2,", "ALSLV\t[$]1,"
 	return x * 11
 }
 func m12(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2", "SHLQ\t[$]2,"
 	// arm64: "LSL\t[$]2,", "ADD\tR[0-9]+<<1,"
-	// loong64: "ADDVU","ADDVU","SLLV\t[$]2,"
+	// loong64: "SLLV", "ALSLV\t[$]1,"
 	return x * 12
 }
 func m13(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2", "LEAQ\t.*[*]4"
 	// arm64: "MOVD\t[$]13,", "MUL"
-	// loong64: "MOVV\t[$]13,","MULV"
+	// loong64: "ALSLV\t[$]1,", "ALSLV\t[$]2,"
 	return x * 13
 }
 func m14(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]14,"
 	// arm64: "LSL\t[$]4,", "SUB\tR[0-9]+<<1,"
-	// loong64: "ADDVU","SLLV\t[$]4,","SUBVU"
+	// loong64: "ADDVU", "ALSLV\t[$]1", "ALSLV\t[$]2"
 	return x * 14
 }
 func m15(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2", "LEAQ\t.*[*]4"
 	// arm64: "LSL\t[$]4,", "SUB\tR[0-9]+,"
-	// loong64: "SLLV\t[$]4,","SUBVU"
+	// loong64: "ALSLV\t[$]1,", "ALSLV\t[$]2,"
 	return x * 15
 }
 func m16(x int64) int64 {
@@ -108,79 +108,79 @@ func m16(x int64) int64 {
 func m17(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]1", "LEAQ\t.*[*]8"
 	// arm64: "ADD\tR[0-9]+<<4,"
-	// loong64: "SLLV\t[$]4,","ADDVU"
+	// loong64: "ALSLV\t[$]"
 	return x * 17
 }
 func m18(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]1", "LEAQ\t.*[*]8"
 	// arm64: "ADD\tR[0-9]+,", "ADD\tR[0-9]+<<3,"
-	// loong64: "ADDVU","SLLV\t[$]4,","ADDVU"
+	// loong64: "ADDVU", "ALSLV\t[$]3,"
 	return x * 18
 }
 func m19(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]8", "LEAQ\t.*[*]2"
 	// arm64: "MOVD\t[$]19,", "MUL"
-	// loong64: "MOVV\t[$]19,","MULV"
+	// loong64: "ALSLV\t[$]3,", "ALSLV\t[$]1,"
 	return x * 19
 }
 func m20(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]4", "SHLQ\t[$]2,"
 	// arm64: "LSL\t[$]2,", "ADD\tR[0-9]+<<2,"
-	// loong64: "SLLV\t[$]2,","SLLV\t[$]4,","ADDVU"
+	// loong64: "SLLV\t[$]2,", "ALSLV\t[$]2," 
 	return x * 20
 }
 func m21(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]4", "LEAQ\t.*[*]4"
 	// arm64: "MOVD\t[$]21,", "MUL"
-	// loong64: "MOVV\t[$]21,","MULV"
+	// loong64: "ALSLV\t[$]2,", "ALSLV\t[$]2,"
 	return x * 21
 }
 func m22(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]22,"
 	// arm64: "MOVD\t[$]22,", "MUL"
-	// loong64: "MOVV\t[$]22,","MULV"
+	// loong64: "ADDVU", "ALSLV\t[$]2,", "ALSLV\t[$]2,"
 	return x * 22
 }
 func m23(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]23,"
 	// arm64: "MOVD\t[$]23,", "MUL"
-	// loong64: "MOVV\t[$]23,","MULV"
+	// loong64: "ALSLV\t[$]1,", "SUBVU", "ALSLV\t[$]3,"
 	return x * 23
 }
 func m24(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2", "SHLQ\t[$]3,"
 	// arm64: "LSL\t[$]3,", "ADD\tR[0-9]+<<1,"
-	// loong64: "ADDVU","ADDVU","SLLV\t[$]3,"
+	// loong64: "SLLV\t[$]3", "ALSLV\t[$]1,"
 	return x * 24
 }
 func m25(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]4", "LEAQ\t.*[*]4"
 	// arm64: "MOVD\t[$]25,", "MUL"
-	// loong64: "MOVV\t[$]25,","MULV"
+	// loong64: "ALSLV\t[$]2,", "ALSLV\t[$]2,"
 	return x * 25
 }
 func m26(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]26,"
 	// arm64: "MOVD\t[$]26,", "MUL"
-	// loong64: "MOVV\t[$]26,","MULV"
+	// loong64: "ADDVU", "ALSLV\t[$]1,", "ALSLV\t[$]3,"
 	return x * 26
 }
 func m27(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]2", "LEAQ\t.*[*]8"
 	// arm64: "MOVD\t[$]27,", "MUL"
-	// loong64: "MOVV\t[$]27,","MULV"
+	// loong64: "ALSLV\t[$]1,", "ALSLV\t[$]3,"
 	return x * 27
 }
 func m28(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]28,"
 	// arm64: "LSL\t[$]5, "SUB\tR[0-9]+<<2,"
-	// loong64: "SLLV\t[$]5,","SLLV\t[$]2,","SUBVU"
+	// loong64: "ALSLV\t[$]1,","SLLV\t[$]2,","ALSLV\t[$]3,"
 	return x * 28
 }
 func m29(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]29,"
 	// arm64: "MOVD\t[$]29,", "MUL"
-	// loong64: "MOVV\t[$]29,","MULV"
+	// loong64: "ALSLV\t[$]1,","SLLV\t[$]5,","SUBVU"
 	return x * 29
 }
 func m30(x int64) int64 {
@@ -204,49 +204,49 @@ func m32(x int64) int64 {
 func m33(x int64) int64 {
 	// amd64: "SHLQ\t[$]2,", "LEAQ\t.*[*]8"
 	// arm64: "ADD\tR[0-9]+<<5,"
-	// loong64: "SLLV\t[$]5,","ADDVU"
+	// loong64: "ADDVU", "ALSLV\t[$]4,"
 	return x * 33
 }
 func m34(x int64) int64 {
 	// amd64: "SHLQ\t[$]5,", "LEAQ\t.*[*]2"
 	// arm64: "ADD\tR[0-9]+,", "ADD\tR[0-9]+<<4,"
-	// loong64: "ADDVU","SLLV\t[$]5,","ADDVU"
+	// loong64: "ADDVU", "ALSLV\t[$]4,"
 	return x * 34
 }
 func m35(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]35,"
 	// arm64: "MOVD\t[$]35,", "MUL"
-	// loong64: "MOVV\t[$]35,","MULV"
+	// loong64: "ALSLV\t[$]4,", "ALSLV\t[$]1,"
 	return x * 35
 }
 func m36(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]8", "SHLQ\t[$]2,"
 	// arm64: "LSL\t[$]2,", "ADD\tR[0-9]+<<3,"
-	// loong64: "SLLV\t[$]2,","SLLV\t[$]5,","ADDVU"
+	// loong64: "SLLV\t[$]2,", "ALSLV\t[$]3,"
 	return x * 36
 }
 func m37(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]8", "LEAQ\t.*[*]4"
 	// arm64: "MOVD\t[$]37,", "MUL"
-	// loong64: "MOVV\t[$]37,","MULV"
+	// loong64: "ALSLV\t[$]3,", "ALSLV\t[$]2,"
 	return x * 37
 }
 func m38(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]38,"
 	// arm64: "MOVD\t[$]38,", "MUL"
-	// loong64: "MOVV\t[$]38,","MULV"
+	// loong64: "ALSLV\t[$]3,", "ALSLV\t[$]2,"
 	return x * 38
 }
 func m39(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]39,"
 	// arm64: "MOVD\t[$]39,", "MUL"
-	// loong64: "MOVV\t[$]39,", "MULV"
+	// loong64: "ALSLV\t[$]2,", "SUBVU", "ALSLV\t[$]3,"
 	return x * 39
 }
 func m40(x int64) int64 {
 	// amd64: "LEAQ\t.*[*]4", "SHLQ\t[$]3,"
 	// arm64: "LSL\t[$]3,", "ADD\tR[0-9]+<<2,"
-	// loong64: "SLLV\t[$]3,","SLLV\t[$]5,","ADDVU"
+	// loong64: "SLLV\t[$]3,", "ALSLV\t[$]2,"
 	return x * 40
 }
 
@@ -265,7 +265,7 @@ func mn2(x int64) int64 {
 func mn3(x int64) int64 {
 	// amd64: "NEGQ", "LEAQ\t.*[*]2"
 	// arm64: "SUB\tR[0-9]+<<2,"
-	// loong64: "SLLV\t[$]2,","SUBVU"
+	// loong64: "SUBVU", "ALSLV\t[$]1,"
 	return x * -3
 }
 func mn4(x int64) int64 {
@@ -277,19 +277,19 @@ func mn4(x int64) int64 {
 func mn5(x int64) int64 {
 	// amd64: "NEGQ", "LEAQ\t.*[*]4"
 	// arm64: "NEG\tR[0-9]+,", "ADD\tR[0-9]+<<2,"
-	// loong64: "SUBVU\tR[0-9], R0,","SLLV\t[$]2,","SUBVU"
+	// loong64: "SUBVU", "ALSLV\t[$]2,"
 	return x * -5
 }
 func mn6(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-6,"
 	// arm64: "ADD\tR[0-9]+,", "SUB\tR[0-9]+<<2,"
-	// loong64: "ADDVU","SLLV\t[$]3,","SUBVU"
+	// loong64: "ADDVU", "SUBVU", "ALSLV\t[$]3,"
 	return x * -6
 }
 func mn7(x int64) int64 {
 	// amd64: "NEGQ", "LEAQ\t.*[*]8"
 	// arm64: "SUB\tR[0-9]+<<3,"
-	// loong64: "SLLV\t[$]3","SUBVU"
+	// loong64: "SUBVU", "ALSLV\t[$]3,"
 	return x * -7
 }
 func mn8(x int64) int64 {
@@ -301,43 +301,43 @@ func mn8(x int64) int64 {
 func mn9(x int64) int64 {
 	// amd64: "NEGQ", "LEAQ\t.*[*]8"
 	// arm64: "NEG\tR[0-9]+,", "ADD\tR[0-9]+<<3,"
-	// loong64: "SUBVU\tR[0-9], R0,","SLLV\t[$]3","SUBVU"
+	// loong64: "SUBVU", "ALSLV\t[$]3,"
 	return x * -9
 }
 func mn10(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-10,"
 	// arm64: "MOVD\t[$]-10,", "MUL"
-	// loong64: "MOVV\t[$]-10","MULV"
+	// loong64: "ADDVU", "ALSLV\t[$]3", "SUBVU"
 	return x * -10
 }
 func mn11(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-11,"
 	// arm64: "MOVD\t[$]-11,", "MUL"
-	// loong64: "MOVV\t[$]-11","MULV"
+	// loong64: "ALSLV\t[$]2,", "SUBVU", "ALSLV\t[$]4,"
 	return x * -11
 }
 func mn12(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-12,"
 	// arm64: "LSL\t[$]2,", "SUB\tR[0-9]+<<2,"
-	// loong64: "SLLV\t[$]2,","SLLV\t[$]4,","SUBVU"
+	// loong64: "SUBVU", "SLLV\t[$]2,", "ALSLV\t[$]4,"
 	return x * -12
 }
 func mn13(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-13,"
 	// arm64: "MOVD\t[$]-13,", "MUL"
-	// loong64: "MOVV\t[$]-13","MULV"
+	// loong64: "ALSLV\t[$]4,", "SLLV\t[$]2, ", "SUBVU"
 	return x * -13
 }
 func mn14(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-14,"
 	// arm64: "ADD\tR[0-9]+,", "SUB\tR[0-9]+<<3,"
-	// loong64: "ADDVU","SLLV\t[$]4,","SUBVU"
+	// loong64: "ADDVU", "SUBVU", "ALSLV\t[$]4,"
 	return x * -14
 }
 func mn15(x int64) int64 {
 	// amd64: "SHLQ\t[$]4,", "SUBQ"
 	// arm64: "SUB\tR[0-9]+<<4,"
-	// loong64: "SLLV\t[$]4,","SUBVU"
+	// loong64: "SUBVU", "ALSLV\t[$]4,"
 	return x * -15
 }
 func mn16(x int64) int64 {
@@ -349,24 +349,24 @@ func mn16(x int64) int64 {
 func mn17(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-17,"
 	// arm64: "NEG\tR[0-9]+,", "ADD\tR[0-9]+<<4,"
-	// loong64: "SUBVU\tR[0-9], R0,","SLLV\t[$]4,","SUBVU"
+	// loong64: "SUBVU", "ALSLV\t[$]4,"
 	return x * -17
 }
 func mn18(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-18,"
 	// arm64: "MOVD\t[$]-18,", "MUL"
-	// loong64: "MOVV\t[$]-18","MULV"
+	// loong64: "ADDVU", "ALSLV\t[$]4,", "SUBVU"
 	return x * -18
 }
 func mn19(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-19,"
 	// arm64: "MOVD\t[$]-19,", "MUL"
-	// loong64: "MOVV\t[$]-19","MULV"
+	// loong64: "ALSLV\t[$]1,", "ALSLV\t[$]4,", "SUBVU"
 	return x * -19
 }
 func mn20(x int64) int64 {
 	// amd64: "IMUL3Q\t[$]-20,"
 	// arm64: "MOVD\t[$]-20,", "MUL"
-	// loong64: "MOVV\t[$]-20","MULV"
+	// loong64: "SLLV\t[$]2,", "ALSLV\t[$]4,", "SUBVU"
 	return x * -20
 }