[turbofan] Add backend support for float32 operations.

test/unittests/compiler/ia32/instruction-selector-ia32-unittest.cc

Index: test/unittests/compiler/ia32/instruction-selector-ia32-unittest.cc
diff --git a/test/unittests/compiler/ia32/instruction-selector-ia32-unittest.cc b/test/unittests/compiler/ia32/instruction-selector-ia32-unittest.cc
index 9138ab2ca671503f3f43ece746ead4c028533c74..6c4ca9cbf6740b1a5c06bec88e8162630e5e9e90 100644
--- a/test/unittests/compiler/ia32/instruction-selector-ia32-unittest.cc
+++ b/test/unittests/compiler/ia32/instruction-selector-ia32-unittest.cc
@@ -91,7 +91,7 @@ TEST_F(InstructionSelectorTest, ChangeFloat32ToFloat64WithParameter) {
   m.Return(m.ChangeFloat32ToFloat64(m.Parameter(0)));
   Stream s = m.Build();
   ASSERT_EQ(1U, s.size());
-  EXPECT_EQ(kSSECvtss2sd, s[0]->arch_opcode());
+  EXPECT_EQ(kSSEFloat32ToFloat64, s[0]->arch_opcode());
   EXPECT_EQ(1U, s[0]->InputCount());
   EXPECT_EQ(1U, s[0]->OutputCount());
 }
@@ -102,7 +102,7 @@ TEST_F(InstructionSelectorTest, TruncateFloat64ToFloat32WithParameter) {
   m.Return(m.TruncateFloat64ToFloat32(m.Parameter(0)));
   Stream s = m.Build();
   ASSERT_EQ(1U, s.size());
-  EXPECT_EQ(kSSECvtsd2ss, s[0]->arch_opcode());
+  EXPECT_EQ(kSSEFloat64ToFloat32, s[0]->arch_opcode());
   EXPECT_EQ(1U, s[0]->InputCount());
   EXPECT_EQ(1U, s[0]->OutputCount());
 }