Remove obsolete MSVC 10.0 workaround for std::floor

MSVC 14.x and 15.x handle -0 correctly unless /fp:fast is used. /fp:precise
is the default.


bug: v8:3477, v8:8912
Change-Id: I242a1dfd845f750cab7c56f13107612259d44d23
Reviewed-on: https://chromium-review.googlesource.com/c/1487414Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#59849}

src/compiler/machine-operator-reducer.cc

@@ -1461,7 +1461,7 @@ Reduction MachineOperatorReducer::ReduceFloat64RoundDown(Node* node) {
   DCHECK_EQ(IrOpcode::kFloat64RoundDown, node->opcode());
   Float64Matcher m(node->InputAt(0));
   if (m.HasValue()) {
-    return ReplaceFloat64(Floor(m.Value()));
+    return ReplaceFloat64(std::floor(m.Value()));
   }
   return NoChange();
 }

src/objects/js-objects.cc

@@ -5575,7 +5575,7 @@ double JSDate::CurrentTimeValue(Isolate* isolate) {
   // the number in a Date object representing a particular instant in
   // time is milliseconds. Therefore, we floor the result of getting
   // the OS time.
-  return Floor(V8::GetCurrentPlatform()->CurrentClockTimeMillis());
+  return std::floor(V8::GetCurrentPlatform()->CurrentClockTimeMillis());
 }
 
 // static

src/utils.h

@@ -220,14 +220,6 @@ T Nabs(T a) {
   return a < 0 ? a : -a;
 }
 
-// Floor(-0.0) == 0.0
-inline double Floor(double x) {
-#if V8_CC_MSVC
-  if (x == 0) return x;  // Fix for issue 3477.
-#endif
-  return std::floor(x);
-}
-
 inline double Modulo(double x, double y) {
 #if defined(V8_OS_WIN)
   // Workaround MS fmod bugs. ECMA-262 says:

test/unittests/compiler/machine-operator-reducer-unittest.cc

@@ -2227,7 +2227,7 @@ TEST_F(MachineOperatorReducerTest, Float64RoundDownWithConstant) {
     Reduction r = Reduce(graph()->NewNode(
         machine()->Float64RoundDown().placeholder(), Float64Constant(x)));
     ASSERT_TRUE(r.Changed());
-    EXPECT_THAT(r.replacement(), IsFloat64Constant(Floor(x)));
+    EXPECT_THAT(r.replacement(), IsFloat64Constant(std::floor(x)));
   }
 }