diff --git a/llvm/lib/Analysis/ValueTracking.cpp b/llvm/lib/Analysis/ValueTracking.cpp
--- a/llvm/lib/Analysis/ValueTracking.cpp
+++ b/llvm/lib/Analysis/ValueTracking.cpp
@@ -3397,9 +3397,6 @@
 /// NOTE: Do not check 'nsz' here because that fast-math-flag does not guarantee
 ///       that a value is not -0.0. It only guarantees that -0.0 may be treated
 ///       the same as +0.0 in floating-point ops.
-///
-/// NOTE: this function will need to be revisited when we support non-default
-/// rounding modes!
 bool llvm::CannotBeNegativeZero(const Value *V, const TargetLibraryInfo *TLI,
                                 unsigned Depth) {
   if (auto *CFP = dyn_cast<ConstantFP>(V))
@@ -3429,9 +3426,21 @@
     case Intrinsic::sqrt:
     case Intrinsic::canonicalize:
       return CannotBeNegativeZero(Call->getArgOperand(0), TLI, Depth + 1);
+    case Intrinsic::experimental_constrained_sqrt: {
+      // NOTE: This rounding mode restriction may be too strict.
+      const auto *CI = cast<ConstrainedFPIntrinsic>(Call);
+      if (CI->getRoundingMode() == RoundingMode::NearestTiesToEven)
+        return CannotBeNegativeZero(Call->getArgOperand(0), TLI, Depth + 1);
+      else
+        return false;
+    }
     // fabs(x) != -0.0
     case Intrinsic::fabs:
       return true;
+    // sitofp and uitofp turn into +0.0 for zero.
+    case Intrinsic::experimental_constrained_sitofp:
+    case Intrinsic::experimental_constrained_uitofp:
+      return true;
     }
   }
 
diff --git a/llvm/test/Transforms/InstSimplify/strictfp-fsub.ll b/llvm/test/Transforms/InstSimplify/strictfp-fsub.ll
--- a/llvm/test/Transforms/InstSimplify/strictfp-fsub.ll
+++ b/llvm/test/Transforms/InstSimplify/strictfp-fsub.ll
@@ -216,12 +216,10 @@
   ret float %sub
 }
 
-; TODO: This can't fire currently due to missing constrained support:
 define float @fold_fsub_sitofp_x_n0_defaultenv(i32 %a) #0 {
 ; CHECK-LABEL: @fold_fsub_sitofp_x_n0_defaultenv(
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
-; CHECK-NEXT:    [[SUB:%.*]] = call float @llvm.experimental.constrained.fsub.f32(float [[FPA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[FPA]]
 ;
   %fpa = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.ignore") #0
   %sub = call float @llvm.experimental.constrained.fsub.f32(float %fpa, float -0.0, metadata !"round.tonearest", metadata !"fpexcept.ignore") #0
diff --git a/llvm/test/Transforms/InstSimplify/strictfp-sqrt-nonneg.ll b/llvm/test/Transforms/InstSimplify/strictfp-sqrt-nonneg.ll
--- a/llvm/test/Transforms/InstSimplify/strictfp-sqrt-nonneg.ll
+++ b/llvm/test/Transforms/InstSimplify/strictfp-sqrt-nonneg.ll
@@ -9,8 +9,7 @@
 ; CHECK-LABEL: @nonneg_u_defaultenv(
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0:[0-9]+]]
 ; CHECK-NEXT:    [[SQRA:%.*]] = call float @llvm.experimental.constrained.sqrt.f32(float [[FPA]], metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
-; CHECK-NEXT:    [[SUB:%.*]] = call float @llvm.experimental.constrained.fsub.f32(float [[SQRA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[SQRA]]
 ;
   %fpa = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.ignore") #0
   %sqra = call float @llvm.experimental.constrained.sqrt.f32(float %fpa, metadata !"round.tonearest", metadata !"fpexcept.ignore") #0
@@ -22,8 +21,7 @@
 ; CHECK-LABEL: @nonneg_s_defaultenv(
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
 ; CHECK-NEXT:    [[SQRA:%.*]] = call float @llvm.experimental.constrained.sqrt.f32(float [[FPA]], metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
-; CHECK-NEXT:    [[SUB:%.*]] = call float @llvm.experimental.constrained.fsub.f32(float [[SQRA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.ignore") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[SQRA]]
 ;
   %fpa = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.ignore") #0
   %sqra = call float @llvm.experimental.constrained.sqrt.f32(float %fpa, metadata !"round.tonearest", metadata !"fpexcept.ignore") #0
@@ -35,8 +33,7 @@
 ; CHECK-LABEL: @nonneg_u_maytrap(
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.maytrap") #[[ATTR0]]
 ; CHECK-NEXT:    [[SQRA:%.*]] = call float @llvm.experimental.constrained.sqrt.f32(float [[FPA]], metadata !"round.tonearest", metadata !"fpexcept.maytrap") #[[ATTR0]]
-; CHECK-NEXT:    [[SUB:%.*]] = call nnan float @llvm.experimental.constrained.fsub.f32(float [[SQRA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.maytrap") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[SQRA]]
 ;
   %fpa = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.maytrap") #0
   %sqra = call float @llvm.experimental.constrained.sqrt.f32(float %fpa, metadata !"round.tonearest", metadata !"fpexcept.maytrap") #0
@@ -48,8 +45,7 @@
 ; CHECK-LABEL: @nonneg_s_maytrap(
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.maytrap") #[[ATTR0]]
 ; CHECK-NEXT:    [[SQRA:%.*]] = call float @llvm.experimental.constrained.sqrt.f32(float [[FPA]], metadata !"round.tonearest", metadata !"fpexcept.maytrap") #[[ATTR0]]
-; CHECK-NEXT:    [[SUB:%.*]] = call nnan float @llvm.experimental.constrained.fsub.f32(float [[SQRA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.maytrap") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[SQRA]]
 ;
   %fpa = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.maytrap") #0
   %sqra = call float @llvm.experimental.constrained.sqrt.f32(float %fpa, metadata !"round.tonearest", metadata !"fpexcept.maytrap") #0
@@ -63,7 +59,7 @@
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR0]]
 ; CHECK-NEXT:    [[SQRA:%.*]] = call float @llvm.experimental.constrained.sqrt.f32(float [[FPA]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR0]]
 ; CHECK-NEXT:    [[SUB:%.*]] = call nnan float @llvm.experimental.constrained.fsub.f32(float [[SQRA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[SQRA]]
 ;
   %fpa = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.strict") #0
   %sqra = call float @llvm.experimental.constrained.sqrt.f32(float %fpa, metadata !"round.tonearest", metadata !"fpexcept.strict") #0
@@ -77,7 +73,7 @@
 ; CHECK-NEXT:    [[FPA:%.*]] = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 [[A:%.*]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR0]]
 ; CHECK-NEXT:    [[SQRA:%.*]] = call float @llvm.experimental.constrained.sqrt.f32(float [[FPA]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR0]]
 ; CHECK-NEXT:    [[SUB:%.*]] = call nnan float @llvm.experimental.constrained.fsub.f32(float [[SQRA]], float -0.000000e+00, metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR0]]
-; CHECK-NEXT:    ret float [[SUB]]
+; CHECK-NEXT:    ret float [[SQRA]]
 ;
   %fpa = call float @llvm.experimental.constrained.sitofp.f32.i32(i32 %a, metadata !"round.tonearest", metadata !"fpexcept.strict") #0
   %sqra = call float @llvm.experimental.constrained.sqrt.f32(float %fpa, metadata !"round.tonearest", metadata !"fpexcept.strict") #0