cmd/compile: wire up Bswap/ReverseBytes intrinsics for loong64

Micro-benchmark results on Loongson 3A5000 and 3A6000:

goos: linux
goarch: loong64
pkg: math/bits
cpu: Loongson-3A6000 @ 2500.00MHz
               |  bench.old   |              bench.new               |
               |    sec/op    |    sec/op     vs base                |
ReverseBytes     2.0020n ± 0%   0.4040n ± 0%  -79.82% (p=0.000 n=20)
ReverseBytes16   0.8866n ± 1%   0.8007n ± 0%   -9.69% (p=0.000 n=20)
ReverseBytes32   1.2195n ± 0%   0.8007n ± 0%  -34.34% (p=0.000 n=20)
ReverseBytes64   2.0705n ± 0%   0.8008n ± 0%  -61.32% (p=0.000 n=20)
geomean           1.455n        0.6749n       -53.62%

goos: linux
goarch: loong64
pkg: math/bits
cpu: Loongson-3A5000 @ 2500.00MHz
               |  bench.old   |              bench.new               |
               |    sec/op    |    sec/op     vs base                |
ReverseBytes     2.8040n ± 0%   0.5205n ± 0%  -81.44% (p=0.000 n=20)
ReverseBytes16   0.7066n ± 0%   0.8011n ± 0%  +13.37% (p=0.000 n=20)
ReverseBytes32   1.5500n ± 0%   0.8010n ± 0%  -48.32% (p=0.000 n=20)
ReverseBytes64   2.7665n ± 0%   0.8010n ± 0%  -71.05% (p=0.000 n=20)
geomean           1.707n        0.7192n       -57.87%

Updates #59120

This patch is a copy of CL 483357.
Co-authored-by: WANG Xuerui <git@xen0n.name>

Change-Id: If355354cd031533df91991fcc3392e5a6c314295
Reviewed-on: https://go-review.googlesource.com/c/go/+/624576
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>

1 files changed, 4 insertions, 0 deletions

diff --git a/test/codegen/mathbits.go b/test/codegen/mathbits.go
index 4519d8bd6c..715f67a3c8 100644
--- a/test/codegen/mathbits.go
+++ b/test/codegen/mathbits.go
@@ -208,6 +208,7 @@ func ReverseBytes(n uint) uint {
 	// 386:"BSWAPL"
 	// s390x:"MOVDBR"
 	// arm64:"REV"
+	// loong64:"REVBV"
 	return bits.ReverseBytes(n)
 }
 
@@ -217,6 +218,7 @@ func ReverseBytes64(n uint64) uint64 {
 	// s390x:"MOVDBR"
 	// arm64:"REV"
 	// ppc64x/power10: "BRD"
+	// loong64:"REVBV"
 	return bits.ReverseBytes64(n)
 }
 
@@ -225,6 +227,7 @@ func ReverseBytes32(n uint32) uint32 {
 	// 386:"BSWAPL"
 	// s390x:"MOVWBR"
 	// arm64:"REVW"
+	// loong64:"REVB2W"
 	// ppc64x/power10: "BRW"
 	return bits.ReverseBytes32(n)
 }
@@ -235,6 +238,7 @@ func ReverseBytes16(n uint16) uint16 {
 	// arm/5:"SLL","SRL","ORR"
 	// arm/6:"REV16"
 	// arm/7:"REV16"
+	// loong64:"REVB2H"
 	// ppc64x/power10: "BRH"
 	return bits.ReverseBytes16(n)
 }