[turbofan] Add support for overflow add/sub to the MachineOperatorReducer.

test/cctest/compiler/test-run-machops.cc

Index: test/cctest/compiler/test-run-machops.cc
diff --git a/test/cctest/compiler/test-run-machops.cc b/test/cctest/compiler/test-run-machops.cc
index 4841bbb1b6954d98c7b23efd1194ae66e4b012e0..4c3ed4db8d0adf8b23958ae6d2eb3809e3212981 100644
--- a/test/cctest/compiler/test-run-machops.cc
+++ b/test/cctest/compiler/test-run-machops.cc
@@ -4016,22 +4016,6 @@ TEST(RunSpillLotsOfThingsWithCall) {
 #endif  // MACHINE_ASSEMBLER_SUPPORTS_CALL_C
 
 
-static bool sadd_overflow(int32_t x, int32_t y, int32_t* val) {
-  int32_t v =
-      static_cast<int32_t>(static_cast<uint32_t>(x) + static_cast<uint32_t>(y));
-  *val = v;
-  return (((v ^ x) & (v ^ y)) >> 31) & 1;
-}
-
-
-static bool ssub_overflow(int32_t x, int32_t y, int32_t* val) {
-  int32_t v =
-      static_cast<int32_t>(static_cast<uint32_t>(x) - static_cast<uint32_t>(y));
-  *val = v;
-  return (((v ^ x) & (v ^ ~y)) >> 31) & 1;
-}
-
-
 TEST(RunInt32AddWithOverflowP) {
   int32_t actual_val = -1;
   RawMachineAssemblerTester<int32_t> m;
@@ -4044,7 +4028,7 @@ TEST(RunInt32AddWithOverflowP) {
   FOR_INT32_INPUTS(i) {
     FOR_INT32_INPUTS(j) {
       int32_t expected_val;
-      int expected_ovf = sadd_overflow(*i, *j, &expected_val);
+      int expected_ovf = bits::SignedAddOverflow32(*i, *j, &expected_val);
       CHECK_EQ(expected_ovf, bt.call(*i, *j));
       CHECK_EQ(expected_val, actual_val);
     }
@@ -4063,7 +4047,7 @@ TEST(RunInt32AddWithOverflowImm) {
       m.StoreToPointer(&actual_val, kMachInt32, val);
       m.Return(ovf);
       FOR_INT32_INPUTS(j) {
-        int expected_ovf = sadd_overflow(*i, *j, &expected_val);
+        int expected_ovf = bits::SignedAddOverflow32(*i, *j, &expected_val);
         CHECK_EQ(expected_ovf, m.Call(*j));
         CHECK_EQ(expected_val, actual_val);
       }
@@ -4076,7 +4060,7 @@ TEST(RunInt32AddWithOverflowImm) {
       m.StoreToPointer(&actual_val, kMachInt32, val);
       m.Return(ovf);
       FOR_INT32_INPUTS(j) {
-        int expected_ovf = sadd_overflow(*i, *j, &expected_val);
+        int expected_ovf = bits::SignedAddOverflow32(*i, *j, &expected_val);
         CHECK_EQ(expected_ovf, m.Call(*j));
         CHECK_EQ(expected_val, actual_val);
       }
@@ -4089,7 +4073,7 @@ TEST(RunInt32AddWithOverflowImm) {
       Node* ovf = m.Projection(1, add);
       m.StoreToPointer(&actual_val, kMachInt32, val);
       m.Return(ovf);
-      int expected_ovf = sadd_overflow(*i, *j, &expected_val);
+      int expected_ovf = bits::SignedAddOverflow32(*i, *j, &expected_val);
       CHECK_EQ(expected_ovf, m.Call());
       CHECK_EQ(expected_val, actual_val);
     }
@@ -4113,7 +4097,7 @@ TEST(RunInt32AddWithOverflowInBranchP) {
   FOR_INT32_INPUTS(i) {
     FOR_INT32_INPUTS(j) {
       int32_t expected;
-      if (sadd_overflow(*i, *j, &expected)) expected = constant;
+      if (bits::SignedAddOverflow32(*i, *j, &expected)) expected = constant;
       CHECK_EQ(expected, bt.call(*i, *j));
     }
   }
@@ -4132,7 +4116,7 @@ TEST(RunInt32SubWithOverflowP) {
   FOR_INT32_INPUTS(i) {
     FOR_INT32_INPUTS(j) {
       int32_t expected_val;
-      int expected_ovf = ssub_overflow(*i, *j, &expected_val);
+      int expected_ovf = bits::SignedSubOverflow32(*i, *j, &expected_val);
       CHECK_EQ(expected_ovf, bt.call(*i, *j));
       CHECK_EQ(expected_val, actual_val);
     }
@@ -4151,7 +4135,7 @@ TEST(RunInt32SubWithOverflowImm) {
       m.StoreToPointer(&actual_val, kMachInt32, val);
       m.Return(ovf);
       FOR_INT32_INPUTS(j) {
-        int expected_ovf = ssub_overflow(*i, *j, &expected_val);
+        int expected_ovf = bits::SignedSubOverflow32(*i, *j, &expected_val);
         CHECK_EQ(expected_ovf, m.Call(*j));
         CHECK_EQ(expected_val, actual_val);
       }
@@ -4164,7 +4148,7 @@ TEST(RunInt32SubWithOverflowImm) {
       m.StoreToPointer(&actual_val, kMachInt32, val);
       m.Return(ovf);
       FOR_INT32_INPUTS(j) {
-        int expected_ovf = ssub_overflow(*j, *i, &expected_val);
+        int expected_ovf = bits::SignedSubOverflow32(*j, *i, &expected_val);
         CHECK_EQ(expected_ovf, m.Call(*j));
         CHECK_EQ(expected_val, actual_val);
       }
@@ -4177,7 +4161,7 @@ TEST(RunInt32SubWithOverflowImm) {
       Node* ovf = m.Projection(1, add);
       m.StoreToPointer(&actual_val, kMachInt32, val);
       m.Return(ovf);
-      int expected_ovf = ssub_overflow(*i, *j, &expected_val);
+      int expected_ovf = bits::SignedSubOverflow32(*i, *j, &expected_val);
       CHECK_EQ(expected_ovf, m.Call());
       CHECK_EQ(expected_val, actual_val);
     }
@@ -4201,7 +4185,7 @@ TEST(RunInt32SubWithOverflowInBranchP) {
   FOR_INT32_INPUTS(i) {
     FOR_INT32_INPUTS(j) {
       int32_t expected;
-      if (ssub_overflow(*i, *j, &expected)) expected = constant;
+      if (bits::SignedSubOverflow32(*i, *j, &expected)) expected = constant;
       CHECK_EQ(expected, bt.call(*i, *j));
     }
   }