Skip to content

[AArch64] use isTRNMask to calculate shuffle costs #171524

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
davemgreen merged 7 commits into from
Dec 15, 2025
Merged

[AArch64] use isTRNMask to calculate shuffle costs #171524

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
657317c
[AArch64] add slp vectorizer test for new isTRNMask
ginsbach Nov 23, 2025
17d3d39
[AArch64] use isTRNMask to calculate shuffle costs
ginsbach Nov 14, 2025
b320449
alternative implementation directly in getShuffleCost
ginsbach Dec 12, 2025
bba438d
add cost model tests for trn with flipped inputs
ginsbach Dec 12, 2025
dbe240a
detect flipped trn miscategorised as SK_InsertSubvector
ginsbach Dec 12, 2025
ae3f0fa
Merge branch 'main' into isTRNMask-for-shuffle-cost
MacDue Dec 13, 2025
be5f7c6
explanatory comment for SK_InsertSubvector
ginsbach Dec 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 6 additions & 1 deletion llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6134,8 +6134,13 @@ AArch64TTIImpl::getShuffleCost(TTI::ShuffleKind Kind, VectorType *DstTy,
unsigned Unused;
if (LT.second.isFixedLengthVector() &&
LT.second.getVectorNumElements() == Mask.size() &&
(Kind == TTI::SK_PermuteTwoSrc || Kind == TTI::SK_PermuteSingleSrc) &&
(Kind == TTI::SK_PermuteTwoSrc || Kind == TTI::SK_PermuteSingleSrc ||
// Discrepancies between isTRNMask and ShuffleVectorInst::isTransposeMask
// mean that we can end up with shuffles that satisfy isTRNMask, but end
// up labelled as TTI::SK_InsertSubvector. (e.g. {2, 0}).
Kind == TTI::SK_InsertSubvector) &&
Copy link
Member

@MacDue MacDue Dec 14, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe leave a comment about why SK_InsertSubvector is included here?

Copy link
Contributor Author

@ginsbach ginsbach Dec 14, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Commit 7 adds such a comment.

(isZIPMask(Mask, LT.second.getVectorNumElements(), Unused, Unused) ||
isTRNMask(Mask, LT.second.getVectorNumElements(), Unused, Unused) ||
isUZPMask(Mask, LT.second.getVectorNumElements(), Unused) ||
isREVMask(Mask, LT.second.getScalarSizeInBits(),
LT.second.getVectorNumElements(), 16) ||
Expand Down
Loading