From 18b93562cfcf0fb728e6fc985855f3e74aac438d Mon Sep 17 00:00:00 2001
From: Matt Arsenault <Matthew.Arsenault@amd.com>
Date: Tue, 31 Jan 2023 20:27:51 -0400
Subject: [PATCH] DAG: Expand legalization of is.fpclass to fcmp for DAZ

Try to use a compare with 0 if DAZ is assumed.
FPClassTest really needs to be marked as a bimask enum, but the API
for that is currently broken.
---
 llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp | 33 +++++++++++++++---------
 llvm/test/CodeGen/X86/is_fpclass.ll              | 28 ++++++++++----------
 2 files changed, 35 insertions(+), 26 deletions(-)

diff --git a/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp b/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
index 6da772d..3d213d3 100644
--- a/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
+++ b/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
@@ -8054,6 +8054,19 @@ SDValue TargetLowering::expandFMINNUM_FMAXNUM(SDNode *Node,
   return SDValue();
 }
 
+/// If this FPClassTest can be performed with a fcmp to 0, return the test mask
+/// for the floating-point mode.
+static bool isFCmpEqualZero(FPClassTest Test, const fltSemantics &Semantics,
+                            const MachineFunction &MF) {
+  if (Test == fcZero &&
+      MF.getDenormalMode(Semantics).Input == DenormalMode::IEEE)
+    return true;
+  if (Test == (fcZero | fcSubnormal) &&
+      MF.getDenormalMode(Semantics).inputsAreZero())
+    return true;
+  return false;
+}
+
 SDValue TargetLowering::expandIS_FPCLASS(EVT ResultVT, SDValue Op,
                                          FPClassTest Test, SDNodeFlags Flags,
                                          const SDLoc &DL,
@@ -8062,7 +8075,7 @@ SDValue TargetLowering::expandIS_FPCLASS(EVT ResultVT, SDValue Op,
   assert(OperandVT.isFloatingPoint());
 
   // Degenerated cases.
-  if (Test == 0)
+  if (Test == fcNone)
     return DAG.getBoolConstant(false, DL, ResultVT, OperandVT);
   if ((Test & fcAllFlags) == fcAllFlags)
     return DAG.getBoolConstant(true, DL, ResultVT, OperandVT);
@@ -8092,18 +8105,14 @@ SDValue TargetLowering::expandIS_FPCLASS(EVT ResultVT, SDValue Op,
   // Some checks can be implemented using float comparisons, if floating point
   // exceptions are ignored.
   if (Flags.hasNoFPExcept() &&
+      // TODO: Should check isCondCodeLegal
       isOperationLegalOrCustom(ISD::SETCC, OperandVT.getScalarType())) {
-    if (Test == fcZero) {
-      DenormalMode Mode = DAG.getMachineFunction().getDenormalMode(Semantics);
-      if (Mode.Input == DenormalMode::IEEE) {
-        // If denormals could be implicitly treated as 0, this is not equivalent
-        // to a compare with 0 since it will also be true for denormals.
-        //
-        // TODO: With DAZ, check == fcZero | fcSubnormal
-        return DAG.getSetCC(DL, ResultVT, Op,
-                            DAG.getConstantFP(0.0, DL, OperandVT),
-                            IsInverted ? ISD::SETUNE : ISD::SETOEQ);
-      }
+    if (isFCmpEqualZero(Test, Semantics, DAG.getMachineFunction())) {
+      // If denormals could be implicitly treated as 0, this is not equivalent
+      // to a compare with 0 since it will also be true for denormals.
+      return DAG.getSetCC(DL, ResultVT, Op,
+                          DAG.getConstantFP(0.0, DL, OperandVT),
+                          IsInverted ? ISD::SETUNE : ISD::SETOEQ);
     }
 
     if (Test == fcNan)
diff --git a/llvm/test/CodeGen/X86/is_fpclass.ll b/llvm/test/CodeGen/X86/is_fpclass.ll
index 6a53181..fc3e5f7 100644
--- a/llvm/test/CodeGen/X86/is_fpclass.ll
+++ b/llvm/test/CodeGen/X86/is_fpclass.ll
@@ -837,24 +837,24 @@ entry:
 define i1 @issubnormal_or_zero_f_daz(float %x) #0 {
 ; CHECK-32-LABEL: issubnormal_or_zero_f_daz:
 ; CHECK-32:       # %bb.0: # %entry
-; CHECK-32-NEXT:    movl $2147483647, %eax # imm = 0x7FFFFFFF
-; CHECK-32-NEXT:    andl {{[0-9]+}}(%esp), %eax
-; CHECK-32-NEXT:    sete %cl
-; CHECK-32-NEXT:    decl %eax
-; CHECK-32-NEXT:    cmpl $8388607, %eax # imm = 0x7FFFFF
-; CHECK-32-NEXT:    setb %al
-; CHECK-32-NEXT:    orb %cl, %al
+; CHECK-32-NEXT:    flds {{[0-9]+}}(%esp)
+; CHECK-32-NEXT:    fldz
+; CHECK-32-NEXT:    fucompp
+; CHECK-32-NEXT:    fnstsw %ax
+; CHECK-32-NEXT:    # kill: def $ah killed $ah killed $ax
+; CHECK-32-NEXT:    sahf
+; CHECK-32-NEXT:    setnp %cl
+; CHECK-32-NEXT:    sete %al
+; CHECK-32-NEXT:    andb %cl, %al
 ; CHECK-32-NEXT:    retl
 ;
 ; CHECK-64-LABEL: issubnormal_or_zero_f_daz:
 ; CHECK-64:       # %bb.0: # %entry
-; CHECK-64-NEXT:    movd %xmm0, %eax
-; CHECK-64-NEXT:    andl $2147483647, %eax # imm = 0x7FFFFFFF
-; CHECK-64-NEXT:    sete %cl
-; CHECK-64-NEXT:    decl %eax
-; CHECK-64-NEXT:    cmpl $8388607, %eax # imm = 0x7FFFFF
-; CHECK-64-NEXT:    setb %al
-; CHECK-64-NEXT:    orb %cl, %al
+; CHECK-64-NEXT:    xorps %xmm1, %xmm1
+; CHECK-64-NEXT:    cmpeqss %xmm0, %xmm1
+; CHECK-64-NEXT:    movd %xmm1, %eax
+; CHECK-64-NEXT:    andl $1, %eax
+; CHECK-64-NEXT:    # kill: def $al killed $al killed $eax
 ; CHECK-64-NEXT:    retq
 entry:
   %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 240)  ; 0xf0 = "subnormal|zero"
-- 
2.7.4