[llvm] [AMDGPU] expand-fp: Change frem expansion criterion (PR #158285)

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Tue Sep 16 07:38:12 PDT 2025


================
@@ -74,11 +74,63 @@ class FRemExpander {
   /// Constant 1 of type \p ExTy.
   Value *One;
 
+  /// The frem argument/return types that can be expanded by this class.
+  // TODO The expansion could work for other floating point types
+  // as well, but this would require additional testing.
+  inline static const SmallVector<MVT, 3> ExpandableTypes{MVT::f16, MVT::f32,
+                                                          MVT::f64};
+
+  /// Libcalls for frem instructions of the type at the corresponding
+  /// positions of ExpandableTypes.
+  inline static const SmallVector<RTLIB::Libcall, 3> FremLibcalls{
+      RTLIB::REM_F32, RTLIB::REM_F32, RTLIB::REM_F64};
+
+  /// Return the Libcall for frem instructions of expandable type \p VT or
+  /// std::nullopt if \p VT is not expandable.
+  static std::optional<RTLIB::Libcall> getFremLibcallForType(EVT VT) {
+    auto *It = find(ExpandableTypes, VT.getSimpleVT());
+    if (It == ExpandableTypes.end())
+      return {};
+
+    return FremLibcalls[It - ExpandableTypes.begin()];
+  };
+
 public:
   static bool canExpandType(Type *Ty) {
-    // TODO The expansion should work for other floating point types
-    // as well, but this would require additional testing.
-    return Ty->isIEEELikeFPTy() && !Ty->isBFloatTy() && !Ty->isFP128Ty();
+    EVT VT = EVT::getEVT(Ty);
+    assert(VT.isSimple() && "Can expand only simple types");
+
+    return find(ExpandableTypes, VT.getSimpleVT());
+  }
+
+  /// Return true if the pass should expand a frem instruction of the
+  /// given \p Ty for the target represented by \p TLI. Expansion
+  /// should happen if the legalization for the scalar type uses a
+  /// non-existing libcall.
+  static bool shouldExpandFremType(const TargetLowering &TLI, EVT VT) {
+    TargetLowering::LegalizeAction LA = TLI.getOperationAction(ISD::FREM, VT);
----------------
arsenm wrote:

This probably mishandles vectors, but I doubt any target does anything other than scalarize this 

https://github.com/llvm/llvm-project/pull/158285


More information about the llvm-commits mailing list