diff options
| author | Archana R <aravind5@in.ibm.com> | 2022-10-31 11:47:17 -0500 |
|---|---|---|
| committer | Lynn Boger <laboger@linux.vnet.ibm.com> | 2023-02-03 19:01:06 +0000 |
| commit | cd1fc871568e9947e84377f82c8d7a4882a07067 (patch) | |
| tree | ef410ca867d20d4dd970ad6bbc01ac062e6da54a /test/codegen | |
| parent | a96487613e2703d1eb7e12c51647e36973df9fd0 (diff) | |
| download | go-cd1fc871568e9947e84377f82c8d7a4882a07067.tar.xz | |
cmd/compile: intrinsify math/bits/ReverseBytes{16|32|64} for ppc64/power10
This change intrinsifies ReverseBytes{16|32|64} by generating the
corresponding new instructions in Power10: brh, brd and brw and
adds a verification test for the same.
On Power 9 and 8, the .go code performs optimally as it is.
Performance improvement seen on Power10:
ReverseBytes32 1.38ns ± 0% 1.18ns ± 0% -14.2
ReverseBytes64 1.52ns ± 0% 1.11ns ± 0% -26.87
ReverseBytes16 1.41ns ± 1% 1.18ns ± 0% -16.47
Change-Id: I88f127f3ab9ba24a772becc21ad90acfba324b37
Reviewed-on: https://go-review.googlesource.com/c/go/+/446675
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Diffstat (limited to 'test/codegen')
| -rw-r--r-- | test/codegen/mathbits.go | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/test/codegen/mathbits.go b/test/codegen/mathbits.go index 09aa5a136e..0eed27a619 100644 --- a/test/codegen/mathbits.go +++ b/test/codegen/mathbits.go @@ -198,6 +198,7 @@ func ReverseBytes64(n uint64) uint64 { // amd64:"BSWAPQ" // s390x:"MOVDBR" // arm64:"REV" + // ppc64x/power10: "BRD" return bits.ReverseBytes64(n) } @@ -205,6 +206,7 @@ func ReverseBytes32(n uint32) uint32 { // amd64:"BSWAPL" // s390x:"MOVWBR" // arm64:"REVW" + // ppc64x/power10: "BRW" return bits.ReverseBytes32(n) } @@ -214,6 +216,7 @@ func ReverseBytes16(n uint16) uint16 { // arm/5:"SLL","SRL","ORR" // arm/6:"REV16" // arm/7:"REV16" + // ppc64x/power10: "BRH" return bits.ReverseBytes16(n) } |