[flang-commits] [flang] [flang][NFC] Use tablegen to create LoopVersioning constructor (PR #90037)
via flang-commits
flang-commits at lists.llvm.org
Thu Apr 25 03:25:51 PDT 2024
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-flang-fir-hlfir
Author: Tom Eccles (tblah)
<details>
<summary>Changes</summary>
The pass is currently defined as only considering function arguments as candidates for the optimization. I would prefer to generalise the pass for other top level operations only when there is a concrete use case before making too many assumptions about the current set of top level operations. Therefore I have not adapted this pass to run on all top level operations.
---
Full diff: https://github.com/llvm/llvm-project/pull/90037.diff
4 Files Affected:
- (modified) flang/include/flang/Optimizer/Transforms/Passes.h (+1-1)
- (modified) flang/include/flang/Optimizer/Transforms/Passes.td (-1)
- (modified) flang/include/flang/Tools/CLOptions.inc (+1-1)
- (modified) flang/lib/Optimizer/Transforms/LoopVersioning.cpp (-4)
``````````diff
diff --git a/flang/include/flang/Optimizer/Transforms/Passes.h b/flang/include/flang/Optimizer/Transforms/Passes.h
index fd7a4a3883c996..613e606954652a 100644
--- a/flang/include/flang/Optimizer/Transforms/Passes.h
+++ b/flang/include/flang/Optimizer/Transforms/Passes.h
@@ -47,6 +47,7 @@ namespace fir {
#define GEN_PASS_DECL_POLYMORPHICOPCONVERSION
#define GEN_PASS_DECL_OPENACCDATAOPERANDCONVERSION
#define GEN_PASS_DECL_ADDDEBUGINFO
+#define GEN_PASS_DECL_LOOPVERSIONING
#include "flang/Optimizer/Transforms/Passes.h.inc"
std::unique_ptr<mlir::Pass> createAffineDemotionPass();
@@ -62,7 +63,6 @@ std::unique_ptr<mlir::Pass> createStackArraysPass();
std::unique_ptr<mlir::Pass> createAliasTagsPass();
std::unique_ptr<mlir::Pass>
createAddDebugInfoPass(fir::AddDebugInfoOptions options = {});
-std::unique_ptr<mlir::Pass> createLoopVersioningPass();
std::unique_ptr<mlir::Pass>
createMemoryAllocationPass(bool dynOnHeap, std::size_t maxStackSize);
diff --git a/flang/include/flang/Optimizer/Transforms/Passes.td b/flang/include/flang/Optimizer/Transforms/Passes.td
index c3d5c336af40ba..383b1628b7153c 100644
--- a/flang/include/flang/Optimizer/Transforms/Passes.td
+++ b/flang/include/flang/Optimizer/Transforms/Passes.td
@@ -321,7 +321,6 @@ def LoopVersioning : Pass<"loop-versioning", "mlir::func::FuncOp"> {
an array has element sized stride. The element sizes stride allows some
loops to be vectorized as well as other loop optimizations.
}];
- let constructor = "::fir::createLoopVersioningPass()";
let dependentDialects = [ "fir::FIROpsDialect" ];
}
diff --git a/flang/include/flang/Tools/CLOptions.inc b/flang/include/flang/Tools/CLOptions.inc
index f24716333d9acf..0b031abe0e6026 100644
--- a/flang/include/flang/Tools/CLOptions.inc
+++ b/flang/include/flang/Tools/CLOptions.inc
@@ -253,7 +253,7 @@ inline void createDefaultFIROptimizerPassPipeline(
}
if (pc.LoopVersioning)
- pm.addPass(fir::createLoopVersioningPass());
+ pm.addPass(fir::createLoopVersioning());
pm.addPass(mlir::createCSEPass());
diff --git a/flang/lib/Optimizer/Transforms/LoopVersioning.cpp b/flang/lib/Optimizer/Transforms/LoopVersioning.cpp
index 7cbd2dd1f897a5..0afc9c24b45b0f 100644
--- a/flang/lib/Optimizer/Transforms/LoopVersioning.cpp
+++ b/flang/lib/Optimizer/Transforms/LoopVersioning.cpp
@@ -556,7 +556,3 @@ void LoopVersioningPass::runOnOperation() {
LLVM_DEBUG(llvm::dbgs() << "=== End " DEBUG_TYPE " ===\n");
}
-
-std::unique_ptr<mlir::Pass> fir::createLoopVersioningPass() {
- return std::make_unique<LoopVersioningPass>();
-}
``````````
</details>
https://github.com/llvm/llvm-project/pull/90037
More information about the flang-commits
mailing list