diff options
| author | David Sherwood <david.sherwood@arm.com> | 2021-11-22 11:38:06 +0000 |
|---|---|---|
| committer | David Sherwood <david.sherwood@arm.com> | 2022-01-13 09:43:07 +0000 |
| commit | 31009f0b5afb504fc1f30769c038e1b7be6ea45b (patch) | |
| tree | 452570f6868b7ac33debfbbf2c618b3d7ebff6e3 /llvm/lib/CodeGen/CodeGenPrepare.cpp | |
| parent | 7ce48be0fd83fb4fe3d0104f324bbbcfcc82983c (diff) | |
| download | llvm-31009f0b5afb504fc1f30769c038e1b7be6ea45b.zip llvm-31009f0b5afb504fc1f30769c038e1b7be6ea45b.tar.gz llvm-31009f0b5afb504fc1f30769c038e1b7be6ea45b.tar.bz2 | |
[CodeGen][AArch64] Ensure isSExtCheaperThanZExt returns true for negative constants
When we know the value we're extending is a negative constant then it
makes sense to use SIGN_EXTEND because this may improve code quality in
some cases, particularly when doing a constant splat of an unpacked vector
type. For example, for SVE when splatting the value -1 into all elements
of a vector of type <vscale x 2 x i32> the element type will get promoted
from i32 -> i64. In this case we want the splat value to sign-extend from
(i32 -1) -> (i64 -1), whereas currently it zero-extends from
(i32 -1) -> (i64 0xFFFFFFFF). Sign-extending the constant means we can use
a single mov immediate instruction.
New tests added here:
CodeGen/AArch64/sve-vector-splat.ll
I believe we see some code quality improvements in these existing
tests too:
CodeGen/AArch64/dag-numsignbits.ll
CodeGen/AArch64/reduce-and.ll
CodeGen/AArch64/unfold-masked-merge-vector-variablemask.ll
The apparent regressions in CodeGen/AArch64/fast-isel-cmp-vec.ll only
occur because the test disables codegen prepare and branch folding.
Differential Revision: https://reviews.llvm.org/D114357
Diffstat (limited to 'llvm/lib/CodeGen/CodeGenPrepare.cpp')
| -rw-r--r-- | llvm/lib/CodeGen/CodeGenPrepare.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp index 747f4e4..8053f1d 100644 --- a/llvm/lib/CodeGen/CodeGenPrepare.cpp +++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp @@ -7004,7 +7004,7 @@ bool CodeGenPrepare::optimizeSwitchInst(SwitchInst *SI) { // matching the argument extension instead. Instruction::CastOps ExtType = Instruction::ZExt; // Some targets prefer SExt over ZExt. - if (TLI->isSExtCheaperThanZExt(OldVT, RegType)) + if (TLI->isSExtCheaperThanZExt(OldVT, RegType, SDValue())) ExtType = Instruction::SExt; if (auto *Arg = dyn_cast<Argument>(Cond)) { |