abrown opened PR #2958 from fix-avx512-flags
to main
:
Previously, the multiple flags for certain AVX512 instructions were
checked usingOR
: e.g., if the CPU has AVX512VLOR
AVX512DQ,
emitVPMULLQ
. This is incorrect--the logic should beAND
. The Intel
Software Developer Manual, vol. 1, sec. 15.4, has more information on
this (notable there is the suggestion to check withXGETBV
that the OS
is allowing the use of the XMM registers--but that is a separate issue).
This change switches toAND
logic in the new backend.<!--
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.
-->
abrown requested cfallin for a review on PR #2958.
abrown has marked PR #2958 as ready for review.
cfallin submitted PR review.
cfallin merged PR #2958.
Last updated: Nov 22 2024 at 17:03 UTC