diff options
| author | Sander de Smalen <sander.desmalen@arm.com> | 2021-05-18 07:37:31 +0100 |
|---|---|---|
| committer | Sander de Smalen <sander.desmalen@arm.com> | 2021-05-18 08:03:48 +0100 |
| commit | 81fdc73e5dc74bfe7cda1e16f01f23836406d709 (patch) | |
| tree | ffbb5980b9a028061a8b3d0a78a61f2dcd5359f6 /llvm/lib/CodeGen/ReachingDefAnalysis.cpp | |
| parent | 00a1007545ba8ad2105876a41364b35235749e2a (diff) | |
| download | llvm-81fdc73e5dc74bfe7cda1e16f01f23836406d709.zip llvm-81fdc73e5dc74bfe7cda1e16f01f23836406d709.tar.gz llvm-81fdc73e5dc74bfe7cda1e16f01f23836406d709.tar.bz2 | |
[LV] Return both fixed and scalable Max VF from computeMaxVF.
This patch introduces a new class, MaxVFCandidates, that holds the
maximum vectorization factors that have been computed for both scalable
and fixed-width vectors.
This patch is intended to be NFC for fixed-width vectors, although
considering a scalable max VF (which is disabled by default) pessimises
tail-loop elimination, since it can no longer determine if any chosen VF
(less than fixed/scalable MaxVFs) is guaranteed to handle all vector
iterations if the trip-count is known. This issue will be addressed in
a future patch.
Reviewed By: fhahn, david-arm
Differential Revision: https://reviews.llvm.org/D98721
Diffstat (limited to 'llvm/lib/CodeGen/ReachingDefAnalysis.cpp')
0 files changed, 0 insertions, 0 deletions