[PatternMatch, InstSimplify] allow undef elements when matching vector -0.0

[spatel]

Summary

Our undef simplification with FP binops - including constant folding - is incomplete and doesn't match the DAG.
But I think it's ok to add the FP equivalent of D42818 and use it for the few cases in InstSimplify of -0.0 folds (that's the only current use of m_NegZero() that I see).

[spatel]

Ping.

[spatel]

Ping * 2.

[RKSimon]

LGTM - thanks.

llvm/trunk/include/llvm/IR/PatternMatch.h

template <typename ITy> bool match(ITy *V) {
     return false;
   }
 };
 
 /// Match an arbitrary zero/null constant. This includes
 /// zero_initializer for vectors and ConstantPointerNull for pointers.
 inline match_zero m_Zero() { return match_zero(); }
 
-struct match_neg_zero {
-  template <typename ITy> bool match(ITy *V) {
-    if (const auto *C = dyn_cast<Constant>(V))
-      return C->isNegativeZeroValue();
-    return false;
-  }
-};
-
-/// Match an arbitrary zero/null constant. This includes
-/// zero_initializer for vectors and ConstantPointerNull for pointers. For
-/// floating point constants, this will match negative zero but not positive
-/// zero
-inline match_neg_zero m_NegZero() { return match_neg_zero(); }
-
 struct match_any_zero {
   template <typename ITy> bool match(ITy *V) {
     if (const auto *C = dyn_cast<Constant>(V))
       return C->isZeroValue();
     return false;
   }
 };
 
template <int64_t Val> struct constantint_match {
   }
 };
 
 /// Match a ConstantInt with a specific value.
 template <int64_t Val> inline constantint_match<Val> m_ConstantInt() {
   return constantint_match<Val>();
 }
 
-/// This helper class is used to match scalar and vector constants that satisfy
-/// a specified predicate. For vector constants, undefined elements are ignored.
+/// This helper class is used to match scalar and vector integer constants that
+/// satisfy a specified predicate.
+/// For vector constants, undefined elements are ignored.
 template <typename Predicate> struct cst_pred_ty : public Predicate {
   template <typename ITy> bool match(ITy *V) {
     if (const auto *CI = dyn_cast<ConstantInt>(V))
       return this->isValue(CI->getValue());
     if (V->getType()->isVectorTy()) {
       if (const auto *C = dyn_cast<Constant>(V)) {
         if (const auto *CI = dyn_cast_or_null<ConstantInt>(C->getSplatValue()))
           return this->isValue(CI->getValue());
if (V->getType()->isVectorTy())
             Res = &CI->getValue();
             return true;
           }
 
     return false;
   }
 };
 
+/// This helper class is used to match scalar and vector floating-point
+/// constants that satisfy a specified predicate.
+/// For vector constants, undefined elements are ignored.
+template <typename Predicate> struct cstfp_pred_ty : public Predicate {
+  template <typename ITy> bool match(ITy *V) {
+    if (const auto *CF = dyn_cast<ConstantFP>(V))
+      return this->isValue(CF->getValueAPF());
+    if (V->getType()->isVectorTy()) {
+      if (const auto *C = dyn_cast<Constant>(V)) {
+        if (const auto *CF = dyn_cast_or_null<ConstantFP>(C->getSplatValue()))
+          return this->isValue(CF->getValueAPF());
+
+        // Non-splat vector constant: check each element for a match.
+        unsigned NumElts = V->getType()->getVectorNumElements();
+        assert(NumElts != 0 && "Constant vector with no elements?");
+        for (unsigned i = 0; i != NumElts; ++i) {
+          Constant *Elt = C->getAggregateElement(i);
+          if (!Elt)
+            return false;
+          if (isa<UndefValue>(Elt))
+            continue;
+          auto *CF = dyn_cast<ConstantFP>(Elt);
+          if (!CF || !this->isValue(CF->getValueAPF()))
+            return false;
+        }
+        return true;
+      }
+    }
+    return false;
+  }
+};
+
 ///////////////////////////////////////////////////////////////////////////////
 //
 // Encapsulate constant value queries for use in templated predicate matchers.
 // This allows checking if constants match using compound predicates and works
 // with vector constants, possibly with relaxed constraints. For example, ignore
 // undef values.
 //
 ///////////////////////////////////////////////////////////////////////////////

 struct is_sign_mask {
   bool isValue(const APInt &C) { return C.isSignMask(); }
 };
 /// Match an integer or vector with only the sign bit(s) set.
 inline cst_pred_ty<is_sign_mask> m_SignMask() {
   return cst_pred_ty<is_sign_mask>();
 }
 
+struct is_neg_zero {
+  bool isValue(const APFloat &C) { return C.isNegZero(); }
+};
+/// Match an FP or FP vector with all -0.0 values.
+inline cstfp_pred_ty<is_neg_zero> m_NegZero() {
+  return cstfp_pred_ty<is_neg_zero>();
+}
+
 ///////////////////////////////////////////////////////////////////////////////
 
 template <typename Class> struct bind_ty {
   Class *&VR;
 
   bind_ty(Class *&V) : VR(V) {}
 
   template <typename ITy> bool match(ITy *V) {

llvm/trunk/test/Transforms/InstSimplify/floating-point-arithmetic.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -instsimplify -S | FileCheck %s
 
 define <2 x float> @fsub_negzero_vec_undef_elts(<2 x float> %x) {
 ; CHECK-LABEL: @fsub_negzero_vec_undef_elts(
-; CHECK-NEXT:    [[R:%.*]] = fsub nsz <2 x float> [[X:%.*]], <float undef, float -0.000000e+00>
-; CHECK-NEXT:    ret <2 x float> [[R]]
+; CHECK-NEXT:    ret <2 x float> [[X:%.*]]
 ;
   %r = fsub nsz <2 x float> %x, <float undef, float -0.0>
   ret <2 x float> %r
 }
 
 ; fsub -0.0, (fsub -0.0, X) ==> X
 define float @fsub_-0_-0_x(float %a) {
 ; CHECK-LABEL: @fsub_-0_-0_x(

 ;
   %t1 = fsub <2 x float> <float -0.0, float -0.0>, %a
   %ret = fsub <2 x float> <float -0.0, float -0.0>, %t1
   ret <2 x float> %ret
 }
 
 define <2 x float> @fsub_-0_-0_x_vec_undef_elts(<2 x float> %a) {
 ; CHECK-LABEL: @fsub_-0_-0_x_vec_undef_elts(
-; CHECK-NEXT:    [[T1:%.*]] = fsub <2 x float> <float undef, float -0.000000e+00>, [[A:%.*]]
-; CHECK-NEXT:    [[RET:%.*]] = fsub <2 x float> <float -0.000000e+00, float undef>, [[T1]]
-; CHECK-NEXT:    ret <2 x float> [[RET]]
+; CHECK-NEXT:    ret <2 x float> [[A:%.*]]
 ;
   %t1 = fsub <2 x float> <float undef, float -0.0>, %a
   %ret = fsub <2 x float> <float -0.0, float undef>, %t1
   ret <2 x float> %ret
 }
 
 ; fsub 0.0, (fsub -0.0, X) != X
 define float @fsub_0_-0_x(float %a) {

 ; CHECK-NEXT:    ret float [[A:%.*]]
 ;
   %ret = fadd float %a, -0.0
   ret float %ret
 }
 
 define <2 x float> @fadd_x_n0_vec_undef_elt(<2 x float> %a) {
 ; CHECK-LABEL: @fadd_x_n0_vec_undef_elt(
-; CHECK-NEXT:    [[RET:%.*]] = fadd <2 x float> [[A:%.*]], <float -0.000000e+00, float undef>
-; CHECK-NEXT:    ret <2 x float> [[RET]]
+; CHECK-NEXT:    ret <2 x float> [[A:%.*]]
 ;
   %ret = fadd <2 x float> %a, <float -0.0, float undef>
   ret <2 x float> %ret
 }
 
 ; fmul X, 1.0 ==> X
 define double @fmul_X_1(double %a) {
 ; CHECK-LABEL: @fmul_X_1(