math: add ARM64 implementation of frexp #47589
Conversation
google-cla
bot
added
the
cla: yes
|
This PR (HEAD: 9e4a3d8) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
|
This PR (HEAD: bf0f19f) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
src/math/frexp_arm64.s
Outdated
| SUB $(52), R1, R1 // set R1 to -52 | ||
| FMOVD F0, R0 | ||
| // tmp = int((x>>shift)&mask) | ||
| UBFX $52, R0, $11, R0 |
There was a problem hiding this comment.
I am not familiar with it but is it a wrong indent?
There was a problem hiding this comment.
Thank you for pointing out. This style twist was causing by space. I have replaced it by tab. Thank you.
|
This PR (HEAD: 9d8f0be) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
|
This PR (HEAD: 2b942c1) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
|
This PR (HEAD: 24f4cf6) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
|
This PR (HEAD: 544bfa2) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
howjmay
changed the title
The implementation of Frexp on ARM64 has been implemented in this PR. The following benchmark was run on Apple Silicon M1 chips. BenchmarkArchFrexp-8 1.581 ns/op 0 B/op 0 allocs/op BenchmarkFrexpGO-8 2.079 ns/op 0 B/op 0 allocs/op The time spent on each iteration drop 23.95%.
|
This PR (HEAD: 92fe95f) has been imported to Gerrit for code review. Please visit https://go-review.googlesource.com/c/go/+/340629 to see it. Tip: You can toggle comments from me using the |
The implementation of Frexp on ARM64 has been implemented in this PR.
The following benchmark was run on Apple Silicon M1 chips.
BenchmarkArchFrexp-8 1.581 ns/op 0 B/op 0 allocs/op
BenchmarkFrexpGO-8 2.079 ns/op 0 B/op 0 allocs/op
The time spent on each iteration drop 23.95%.