Page MenuHomePhabricator
Differential D143278

[SME2][AArch64] Add multi-indexed multiply-add long long intrinsics
ClosedPublic
Actions

Authored by kmclaughlin on Feb 3 2023, 9:03 AM.

Details

Reviewers
sdesmalen
david-arm
hassnaa-arm
aemerson
Commits
rGa9df62705085: [SME2][AArch64] Add multi-indexed multiply-add long long intrinsics
Summary

Adds intrinsics for the following SME2 instructions (1, 2 & 4 vector):

  • smlall
  • umlall
  • smlsll
  • umlsll
  • sumlall
  • usmlall
NOTE: These intrinsics are still in development and are subject to future changes.

Diff Detail

Event Timeline

kmclaughlin created this revision.Feb 3 2023, 9:03 AM
Herald added a project: Restricted Project. · View Herald TranscriptFeb 3 2023, 9:03 AM
Herald added subscribers: arphaman, hiraditya, kristof.beyls. · View Herald Transcript
kmclaughlin requested review of this revision.Feb 3 2023, 9:03 AM
Herald added a project: Restricted Project. · View Herald TranscriptFeb 3 2023, 9:03 AM
Herald added a subscriber: llvm-commits. · View Herald Transcript
Harbormaster completed remote builds in B211748: Diff 494667.Feb 3 2023, 9:03 AM
david-arm added inline comments.Feb 9 2023, 5:58 AM
llvm/lib/Target/AArch64/SMEInstrFormats.td
2599

Shouldn't these be VectorIndexH32b_timm to match the type in sme2_mla_ll_array_index_64b?

kmclaughlin updated this revision to Diff 497709.Feb 15 2023, 9:37 AM
kmclaughlin marked an inline comment as done.
  • Changed sme2_mla_ll_array_index_64b to use VectorIndexH32b_timm
Harbormaster completed remote builds in B213921: Diff 497709.Feb 15 2023, 9:38 AM
david-arm accepted this revision.Feb 16 2023, 5:19 AM

LGTM!

This revision is now accepted and ready to land.Feb 16 2023, 5:19 AM
kmclaughlin updated this revision to Diff 499819.Thu, Feb 23, 5:46 AM

Fixed the immediates used in the *za64.lane.vg4x1 tests so that they are within the correct range for VectorIndexH32b.

Harbormaster completed remote builds in B215494: Diff 499819.Thu, Feb 23, 7:21 AM
david-arm accepted this revision.Mon, Feb 27, 3:36 AM

LGTM!

Closed by commit rGa9df62705085: [SME2][AArch64] Add multi-indexed multiply-add long long intrinsics (authored by kmclaughlin). · Explain WhyMon, Feb 27, 5:59 AM
This revision was automatically updated to reflect the committed changes.
kmclaughlin added a commit: rGa9df62705085: [SME2][AArch64] Add multi-indexed multiply-add long long intrinsics.

Revision Contents
Changeset List

PathSize
llvm/
include/
llvm/
IR/
12 lines
lib/
Target/
AArch64/
60 lines
64 lines
test/
CodeGen/
AArch64/
534 lines

Diff 500752

View Options

llvm/include/llvm/IR/IntrinsicsAArch64.td

Loading...
View Options

llvm/lib/Target/AArch64/AArch64SMEInstrInfo.td

Loading...
View Options

llvm/lib/Target/AArch64/SMEInstrFormats.td

Loading...
View Options

llvm/test/CodeGen/AArch64/sme2-intrinsics-mlall.ll

Loading...