Port and integrate the irregexp engine from V8

runtime/vm/intermediate_language_arm.cc

Index: runtime/vm/intermediate_language_arm.cc
diff --git a/runtime/vm/intermediate_language_arm.cc b/runtime/vm/intermediate_language_arm.cc
index f3212d12ecfacd02bc187e2b5f28ae2dec5b98d2..98a3c0bb412beafc9b74eea5715f92467d9c9fc7 100644
--- a/runtime/vm/intermediate_language_arm.cc
+++ b/runtime/vm/intermediate_language_arm.cc
@@ -1361,6 +1361,99 @@ void LoadIndexedInstr::EmitNativeCode(FlowGraphCompiler* compiler) {
 }
 
 
+Representation LoadCodeUnitsInstr::representation() const {
+  switch (class_id()) {
+    case kOneByteStringCid:
+    case kExternalOneByteStringCid:
+      // TODO(jgruber): kUnboxedUint32 could be a better choice.
+      return element_count() < 4 ? kTagged : kUnboxedMint;
+    case kTwoByteStringCid:
+    case kExternalTwoByteStringCid:
+      return element_count() < 2 ? kTagged : kUnboxedMint;
+    default:
+      UNIMPLEMENTED();
+      return kTagged;
+  }
+}
+
+
+LocationSummary* LoadCodeUnitsInstr::MakeLocationSummary(Isolate* isolate,
+                                                         bool opt) const {
+  const intptr_t kNumInputs = 2;
+  const intptr_t kNumTemps = 0;
+  LocationSummary* locs = new(isolate) LocationSummary(
+      isolate, kNumInputs, kNumTemps, LocationSummary::kNoCall);
+  locs->set_in(0, Location::RequiresRegister());
+  locs->set_in(1, Location::RequiresRegister());
+
+  if (representation() == kUnboxedMint) {
+    locs->set_out(0, Location::Pair(Location::RequiresRegister(),
+                                    Location::RequiresRegister()));
+  } else {
+    ASSERT(representation() == kTagged);
+    locs->set_out(0, Location::RequiresRegister());
+  }
+
+  return locs;
+}
+
+
+void LoadCodeUnitsInstr::EmitNativeCode(FlowGraphCompiler* compiler) {
+  const Register array = locs()->in(0).reg();
+  const Location index = locs()->in(1);
+
+  Address element_address = __ ElementAddressForRegIndex(
+        true,  IsExternal(), class_id(), index_scale(), array, index.reg());
+  // Warning: element_address may use register IP as base.
+
+  if (representation() == kUnboxedMint) {
+    ASSERT(locs()->out(0).IsPairLocation());
+    PairLocation* result_pair = locs()->out(0).AsPairLocation();
+    Register result1 = result_pair->At(0).reg();
+    Register result2 = result_pair->At(1).reg();
+    switch (class_id()) {
+      case kOneByteStringCid:
+      case kExternalOneByteStringCid:
+        ASSERT(element_count() == 4);
+        __ ldr(result1, element_address);
+        __ eor(result2, result2, Operand(result2));
+        break;
+      case kTwoByteStringCid:
+      case kExternalTwoByteStringCid:
+        ASSERT(element_count() == 2);
+        __ ldr(result1, element_address);
+        __ eor(result2, result2, Operand(result2));
+        break;
+      default:
+        UNREACHABLE();
+    }
+  } else {
+    ASSERT(representation() == kTagged);
+    Register result = locs()->out(0).reg();
+    switch (class_id()) {
+      case kOneByteStringCid:
+      case kExternalOneByteStringCid:
+        switch (element_count()) {
+          case 1: __ ldrb(result, element_address); break;
+          case 2: __ ldrh(result, element_address); break;

[zerny-google, 2014/10/07 12:18:35]

Nit: the fact that case 3 is not missing would be a bit more clear with the
explicit type of code units.

[jgruber1, 2014/10/07 15:00:25]

Acknowledged.

+          default: UNREACHABLE();
+        }
+        __ SmiTag(result);
+        break;
+      case kTwoByteStringCid:
+      case kExternalTwoByteStringCid:
+        ASSERT(element_count() == 1);
+        __ ldrh(result, element_address);
+        __ SmiTag(result);
+        break;
+      default:
+        UNREACHABLE();
+        break;
+    }
+  }
+}
+
+
 Representation StoreIndexedInstr::RequiredInputRepresentation(
     intptr_t idx) const {
   // Array can be a Dart object or a pointer to external data.
@@ -5094,6 +5187,28 @@ void MathUnaryInstr::EmitNativeCode(FlowGraphCompiler* compiler) {
 }
 
 
+LocationSummary* CaseInsensitiveCompareUC16Instr::MakeLocationSummary(
+    Isolate* isolate, bool opt) const {
+  const intptr_t kNumTemps = 0;
+  LocationSummary* summary = new(isolate) LocationSummary(
+      isolate, InputCount(), kNumTemps, LocationSummary::kCall);
+  summary->set_in(0, Location::RegisterLocation(R0));
+  summary->set_in(1, Location::RegisterLocation(R1));
+  summary->set_in(2, Location::RegisterLocation(R2));
+  summary->set_in(3, Location::RegisterLocation(R3));
+  summary->set_out(0, Location::RegisterLocation(R0));
+  return summary;
+}
+
+
+void CaseInsensitiveCompareUC16Instr::EmitNativeCode(
+    FlowGraphCompiler* compiler) {
+
+  // Call the function.
+  __ CallRuntime(TargetFunction(), TargetFunction().argument_count());
+}
+
+
 LocationSummary* MathMinMaxInstr::MakeLocationSummary(Isolate* isolate,
                                                       bool opt) const {
   if (result_cid() == kDoubleCid) {
@@ -6870,6 +6985,38 @@ void GotoInstr::EmitNativeCode(FlowGraphCompiler* compiler) {
 }
 
 
+LocationSummary* IndirectGotoInstr::MakeLocationSummary(Isolate* isolate,
+                                                        bool opt) const {
+  const intptr_t kNumInputs = 1;
+  const intptr_t kNumTemps = 1;
+
+  LocationSummary* locs = new(isolate) LocationSummary(

[zerny-google, 2014/10/07 12:18:35]

Nit: Not sure if locs or summary is the preferred naming convention. I'd stick
with just one for your changes...

[jgruber1, 2014/10/07 15:00:25]

Seems to be split 1:2 (locs:summary), I went with the majority and changed all
occurrences to summary.

+        isolate, kNumInputs, kNumTemps, LocationSummary::kNoCall);
+
+  locs->set_in(0, Location::RequiresRegister());
+  locs->set_temp(0, Location::RequiresRegister());
+
+  return locs;
+}
+
+
+void IndirectGotoInstr::EmitNativeCode(FlowGraphCompiler* compiler) {
+  Register target_address_reg = locs()->temp_slot(0)->reg();
+
+  // Load from [current frame pointer] + kPcMarkerSlotFromFp.
+  __ ldr(target_address_reg, Address(FP, kPcMarkerSlotFromFp * kWordSize));
+
+  // Add the offset.
+  Register offset_reg = locs()->in(0).reg();
+  __ add(target_address_reg,
+         target_address_reg,
+         Operand(offset_reg, ASR, kSmiTagSize));
+
+  // Jump to the absolute address.
+  __ bx(target_address_reg);
+}
+
+
 LocationSummary* StrictCompareInstr::MakeLocationSummary(Isolate* isolate,
                                                          bool opt) const {
   const intptr_t kNumInputs = 2;