[PATCH] D156344: Disable call to fma for soft-float

[Kishan Parmar via Phabricator via cfe-commits]

long5hot updated this revision to Diff 544432.
long5hot added a comment.
Herald added subscribers: cfe-commits, MaskRay.
Herald added a project: clang.

After clang release/16.x there is a regression that -mabi=elfv1
or -mabi=elfv2 are being unused and throws warning. But clang-trunk
throws error for -mabi=elfv2 or elfv1. Intent of this patch to accept
elfv1 or elfv2 for -mabi.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D156344/new/

https://reviews.llvm.org/D156344

Files:
  clang/lib/Driver/ToolChains/Clang.cpp
  llvm/lib/Target/PowerPC/PPCISelLowering.cpp
  llvm/test/CodeGen/PowerPC/ppcsoftops.ll


Index: llvm/test/CodeGen/PowerPC/ppcsoftops.ll
===================================================================
--- llvm/test/CodeGen/PowerPC/ppcsoftops.ll
+++ llvm/test/CodeGen/PowerPC/ppcsoftops.ll
@@ -51,4 +51,25 @@
   ; CHECK-LABEL:      __divdf3
 }
 
+; Function Attrs: noinline nounwind optnone uwtable
+define dso_local zeroext i32 @func(double noundef %0, double noundef %1) #0 {
+  %3 = alloca double, align 8
+  %4 = alloca double, align 8
+  store double %0, ptr %3, align 8
+  store double %1, ptr %4, align 8
+  %5 = load double, ptr %3, align 8
+  %6 = load double, ptr %4, align 8
+  %7 = fneg double %6
+  %8 = call double @llvm.fmuladd.f64(double %7, double 0x41F0000000000000, double %5)
+  %9 = fptoui double %8 to i32
+  ret i32 %9
+
+  ; CHECK-LABEL:      __muldf3
+  ; CHECK-LABEL:      __adddf3
+}
+
+; Function Attrs: nocallback nofree nosync nounwind speculatable willreturn memory(none)
+declare double @llvm.fmuladd.f64(double, double, double) #1
+
 attributes #0 = {"use-soft-float"="true" }
+attributes #1 = { nocallback nofree nosync nounwind speculatable willreturn memory(none) }
Index: llvm/lib/Target/PowerPC/PPCISelLowering.cpp
===================================================================
--- llvm/lib/Target/PowerPC/PPCISelLowering.cpp
+++ llvm/lib/Target/PowerPC/PPCISelLowering.cpp
@@ -17225,7 +17225,7 @@
 
 bool PPCTargetLowering::isFMAFasterThanFMulAndFAdd(const Function &F,
                                                    Type *Ty) const {
-  if (Subtarget.hasSPE())
+  if (Subtarget.hasSPE() || Subtarget.useSoftFloat())
     return false;
   switch (Ty->getScalarType()->getTypeID()) {
   case Type::FloatTyID:
Index: clang/lib/Driver/ToolChains/Clang.cpp
===================================================================
--- clang/lib/Driver/ToolChains/Clang.cpp
+++ clang/lib/Driver/ToolChains/Clang.cpp
@@ -2068,6 +2068,12 @@
     } else if (V == "vec-extabi") {
       VecExtabi = true;
       A->claim();
+    } else if (V == "elfv1") {
+      ABIName = "elfv1";
+      A->claim();
+    } else if (V == "elfv2") {
+      ABIName = "elfv2";
+      A->claim();
     } else if (V != "altivec")
       // The ppc64 linux abis are all "altivec" abis by default. Accept and ignore
       // the option if given as we don't have backend support for any targets


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D156344.544432.patch
Type: text/x-patch
Size: 2320 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20230726/ac3f6cb5/attachment.bin>