diff options
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | 2025-05-20 09:33:28 +0100 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-05-20 09:33:28 +0100 |
| commit | 57aa7fa6ed64394ecd7d68fe34984b89778f319b (patch) | |
| tree | 07d862ae8e3835258eac5e5ae2ae1a525a9af462 /flang/lib/Frontend/CompilerInvocation.cpp | |
| parent | b3e15baf7c649a12e71ff281b83e9213abae2977 (diff) | |
| download | llvm-57aa7fa6ed64394ecd7d68fe34984b89778f319b.zip llvm-57aa7fa6ed64394ecd7d68fe34984b89778f319b.tar.gz llvm-57aa7fa6ed64394ecd7d68fe34984b89778f319b.tar.bz2 | |
[X86] SimplifyDemandedVectorEltsForTargetNode - replace packed fcmp node with scalar fcmp node if only element0 is demanded (#140563)
These unnecessary vectorisation can appear due to fplogic opcodes only being available for 128-bit types - which can prevent folds that only work on the scalar source types and also lead to fcmp of garbage data in the upper elements.
Fixes #140534
Diffstat (limited to 'flang/lib/Frontend/CompilerInvocation.cpp')
0 files changed, 0 insertions, 0 deletions