elliottt opened PR #4729 from trevor/x64-bitcast-fabs-fneg
to main
:
This PR doesn't need to be reviewed until #4722 is merged
Migrate the
bitcast
,fabs
, andfneg
instructions to ISLE
<!--Please ensure that the following steps are all taken care of before submitting
the PR.
[ ] This has been discussed in issue #..., or if not, please tell us why
here.[ ] A short description of what this does, why it is needed; if the
description becomes long, the matter should probably be discussed in an issue
first.[ ] This PR contains test cases, if meaningful.
- [ ] A reviewer from the core maintainer team has been assigned for this PR.
If you don't know who could review this, please indicate so. The list of
suggested reviewers on the right can help you.Please ensure all communication adheres to the code of conduct.
-->
cfallin created PR review comment:
I wonder, would it be better to use a VCodeConstant here to get the appropriate MSBs-only-in-each-lane value?
cfallin submitted PR review.
cfallin submitted PR review.
elliottt has marked PR #4729 as ready for review.
elliottt edited PR #4729 from trevor/x64-bitcast-fabs-fneg
to main
:
Migrate the
bitcast
,fabs
, andfneg
instructions to ISLE
<!--Please ensure that the following steps are all taken care of before submitting
the PR.
[ ] This has been discussed in issue #..., or if not, please tell us why
here.[ ] A short description of what this does, why it is needed; if the
description becomes long, the matter should probably be discussed in an issue
first.[ ] This PR contains test cases, if meaningful.
- [ ] A reviewer from the core maintainer team has been assigned for this PR.
If you don't know who could review this, please indicate so. The list of
suggested reviewers on the right can help you.Please ensure all communication adheres to the code of conduct.
-->
elliottt updated PR #4729 from trevor/x64-bitcast-fabs-fneg
to main
.
elliottt updated PR #4729 from trevor/x64-bitcast-fabs-fneg
to main
.
elliottt requested cfallin for a review on PR #4729.
elliottt submitted PR review.
elliottt created PR review comment:
Would that be a savings over the two instructions?
cfallin submitted PR review.
cfallin created PR review comment:
Possibly, though looking at
xorps
I see the basic SSE version is notreg, reg/mem
but justreg, reg
so it would still be a separate load; one load vs. dependent chain of two 1-cycle ops, the latter likely wins. Anyway it's an unimportant enough question that I don't care too much to investigate further :-)
cfallin submitted PR review.
cfallin merged PR #4729.
Last updated: Dec 23 2024 at 12:05 UTC