Revision: 18325
Author: [email protected]
Date: Tue Dec 17 12:42:34 2013 UTC
Log: Merged r17432, r17526, r17545, r17620, r17747 into 3.20 branch.
Do not remove HAdd with zero if the other operand is a double.
Age ICs on context dispose notification.
Fix y-umlaut to uppercase.
Fix potential assertion failure in ConvertCase.
Fix duplicate check in DependentCode::Insert.
BUG=309723,v8:2984,318454
[email protected]
Review URL: https://chromiumcodereview.appspot.com/109213005
http://code.google.com/p/v8/source/detail?r=18325
Added:
/branches/3.20/test/mjsunit/regress/regress-2984.js
/branches/3.20/test/mjsunit/regress/regress-add-minus-zero.js
Modified:
/branches/3.20/src/compiler.cc
/branches/3.20/src/heap.cc
/branches/3.20/src/hydrogen-instructions.cc
/branches/3.20/src/objects.cc
/branches/3.20/src/runtime.cc
/branches/3.20/src/unicode.h
/branches/3.20/src/version.cc
=======================================
--- /dev/null
+++ /branches/3.20/test/mjsunit/regress/regress-2984.js Tue Dec 17 12:42:34
2013 UTC
@@ -0,0 +1,34 @@
+// Copyright 2013 the V8 project authors. All rights reserved.
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+// * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+// * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following
+// disclaimer in the documentation and/or other materials provided
+// with the distribution.
+// * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived
+// from this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+assertEquals("\u0178", "\xff".toUpperCase());
+assertEquals("abcdefghijklmn\xffopq",
+ ("ABCDEFGHIJKL" + "MN\u0178OPQ").toLowerCase());
+assertEquals("\xff", "\u0178".toLowerCase());
+assertEquals("ABCDEFGHIJKLMN\u0178OPQ",
+ ("abcdefghijk" + "lmn\xffopq").toUpperCase());
+
=======================================
--- /dev/null
+++ /branches/3.20/test/mjsunit/regress/regress-add-minus-zero.js Tue Dec
17 12:42:34 2013 UTC
@@ -0,0 +1,38 @@
+// Copyright 2013 the V8 project authors. All rights reserved.
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+// * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+// * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following
+// disclaimer in the documentation and/or other materials provided
+// with the distribution.
+// * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived
+// from this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+// Flags: --allow-natives-syntax
+
+var o = { a: 0 };
+
+function f(x) { return -o.a + 0; };
+
+assertEquals("Infinity", String(1/f()));
+assertEquals("Infinity", String(1/f()));
+%OptimizeFunctionOnNextCall(f);
+assertEquals("Infinity", String(1/f()));
+
=======================================
--- /branches/3.20/src/compiler.cc Tue Aug 13 17:09:37 2013 UTC
+++ /branches/3.20/src/compiler.cc Tue Dec 17 12:42:34 2013 UTC
@@ -264,8 +264,11 @@
void OptimizingCompiler::RecordOptimizationStats() {
Handle<JSFunction> function = info()->closure();
- int opt_count = function->shared()->opt_count();
- function->shared()->set_opt_count(opt_count + 1);
+ if (!function->IsOptimized()) {
+ // Concurrent recompilation and OSR may race. Increment only once.
+ int opt_count = function->shared()->opt_count();
+ function->shared()->set_opt_count(opt_count + 1);
+ }
double ms_creategraph =
static_cast<double>(time_taken_to_create_graph_) / 1000;
double ms_optimize = static_cast<double>(time_taken_to_optimize_) / 1000;
=======================================
--- /branches/3.20/src/heap.cc Thu Sep 26 13:29:44 2013 UTC
+++ /branches/3.20/src/heap.cc Tue Dec 17 12:42:34 2013 UTC
@@ -709,6 +709,7 @@
isolate()->optimizing_compiler_thread()->Flush();
}
flush_monomorphic_ics_ = true;
+ AgeInlineCaches();
return ++contexts_disposed_;
}
@@ -1095,8 +1096,6 @@
isolate_->counters()->objs_since_last_full()->Set(0);
- contexts_disposed_ = 0;
-
flush_monomorphic_ics_ = false;
}
@@ -4000,13 +3999,12 @@
return result;
}
- Object* result;
+ SeqTwoByteString* result;
{ MaybeObject* maybe_result = AllocateRawTwoByteString(1);
- if (!maybe_result->ToObject(&result)) return maybe_result;
+ if (!maybe_result->To<SeqTwoByteString>(&result)) return maybe_result;
}
- String* answer = String::cast(result);
- answer->Set(0, code);
- return answer;
+ result->SeqTwoByteStringSet(0, code);
+ return result;
}
@@ -5957,12 +5955,7 @@
size_factor * IncrementalMarking::kAllocatedThreshold;
if (contexts_disposed_ > 0) {
- if (hint >= kMaxHint) {
- // The embedder is requesting a lot of GC work after context
disposal,
- // we age inline caches so that they don't keep objects from
- // the old context alive.
- AgeInlineCaches();
- }
+ contexts_disposed_ = 0;
int mark_sweep_time = Min(TimeMarkSweepWouldTakeInMs(), 1000);
if (hint >= mark_sweep_time && !FLAG_expose_gc &&
incremental_marking()->IsStopped()) {
@@ -5971,8 +5964,8 @@
"idle notification: contexts disposed");
} else {
AdvanceIdleIncrementalMarking(step_size);
- contexts_disposed_ = 0;
}
+
// After context disposal there is likely a lot of garbage remaining,
reset
// the idle notification counters in order to trigger more incremental
GCs
// on subsequent idle notifications.
=======================================
--- /branches/3.20/src/hydrogen-instructions.cc Fri Sep 27 09:50:47 2013 UTC
+++ /branches/3.20/src/hydrogen-instructions.cc Tue Dec 17 12:42:34 2013 UTC
@@ -1206,8 +1206,15 @@
HValue* HAdd::Canonicalize() {
- if (IsIdentityOperation(left(), right(), 0)) return left();
- if (IsIdentityOperation(right(), left(), 0)) return right();
+ // Adding 0 is an identity operation except in case of -0: -0 + 0 = +0
+ if (IsIdentityOperation(left(), right(), 0) &&
+ !left()->representation().IsDouble()) { // Left could be -0.
+ return left();
+ }
+ if (IsIdentityOperation(right(), left(), 0) &&
+ !left()->representation().IsDouble()) { // Right could be -0.
+ return right();
+ }
return this;
}
=======================================
--- /branches/3.20/src/objects.cc Tue Dec 17 10:02:26 2013 UTC
+++ /branches/3.20/src/objects.cc Tue Dec 17 12:42:34 2013 UTC
@@ -9321,7 +9321,7 @@
if (value->IsSmi()) {
// No optimized code map.
ASSERT_EQ(0, Smi::cast(value)->value());
- // Crate 3 entries per context {context, code, literals}.
+ // Create 3 entries per context {context, code, literals}.
MaybeObject* maybe = heap->AllocateFixedArray(kInitialLength);
if (!maybe->To(&new_code_map)) return maybe;
new_code_map->set(kEntriesStart + 0, native_context);
@@ -11243,11 +11243,9 @@
int start = starts.at(group);
int end = starts.at(group + 1);
int number_of_entries = starts.number_of_entries();
- if (start < end && entries->object_at(end - 1) == *object) {
- // Do not append the compilation info if it is already in the array.
- // It is sufficient to just check only the last element because
- // we process embedded maps of an optimized code in one batch.
- return entries;
+ // Check for existing entry to avoid duplicates.
+ for (int i = start; i < end; i++) {
+ if (entries->object_at(i) == *object) return entries;
}
if (entries->length() < kCodesStartIndex + number_of_entries + 1) {
Factory* factory = entries->GetIsolate()->factory();
=======================================
--- /branches/3.20/src/runtime.cc Tue Dec 17 11:02:35 2013 UTC
+++ /branches/3.20/src/runtime.cc Tue Dec 17 12:42:34 2013 UTC
@@ -6120,6 +6120,7 @@
MUST_USE_RESULT static MaybeObject* ConvertCaseHelper(
Isolate* isolate,
String* s,
+ String::Encoding result_encoding,
int length,
int input_string_length,
unibrow::Mapping<Converter, 128>* mapping) {
@@ -6135,7 +6136,7 @@
// might break in the future if we implement more context and locale
// dependent upper/lower conversions.
Object* o;
- { MaybeObject* maybe_o = s->IsOneByteRepresentation()
+ { MaybeObject* maybe_o = result_encoding == String::ONE_BYTE_ENCODING
? isolate->heap()->AllocateRawOneByteString(length)
: isolate->heap()->AllocateRawTwoByteString(length);
if (!maybe_o->ToObject(&o)) return maybe_o;
@@ -6143,6 +6144,8 @@
String* result = String::cast(o);
bool has_changed_character = false;
+ DisallowHeapAllocation no_gc;
+
// Convert all characters to upper case, assuming that they will fit
// in the buffer
Access<ConsStringIteratorOp> op(
@@ -6151,6 +6154,10 @@
unibrow::uchar chars[Converter::kMaxWidth];
// We can assume that the string is not empty
uc32 current = stream.GetNext();
+ // y with umlauts is the only character that stops fitting into one-byte
+ // when converting to uppercase.
+ static const uc32 yuml_code = 0xff;
+ bool ignore_yuml = result->IsSeqTwoByteString() || Converter::kIsToLower;
for (int i = 0; i < length;) {
bool has_next = stream.HasMore();
uc32 next = has_next ? stream.GetNext() : 0;
@@ -6159,13 +6166,14 @@
// The case conversion of this character is the character itself.
result->Set(i, current);
i++;
- } else if (char_length == 1) {
+ } else if (char_length == 1 && (ignore_yuml || current != yuml_code)) {
// Common case: converting the letter resulted in one character.
ASSERT(static_cast<uc32>(chars[0]) != current);
result->Set(i, chars[0]);
has_changed_character = true;
i++;
} else if (length == input_string_length) {
+ bool found_yuml = (current == yuml_code);
// We've assumed that the result would be as long as the
// input but here is a character that converts to several
// characters. No matter, we calculate the exact length
@@ -6185,6 +6193,7 @@
int current_length = i + char_length + next_length;
while (stream.HasMore()) {
current = stream.GetNext();
+ found_yuml |= (current == yuml_code);
// NOTE: we use 0 as the next character here because, while
// the next character may affect what a character converts to,
// it does not in any case affect the length of what it convert
@@ -6197,8 +6206,10 @@
return Failure::OutOfMemoryException(0x13);
}
}
- // Try again with the real length.
- return Smi::FromInt(current_length);
+ // Try again with the real length. Return signed if we need
+ // to allocate a two-byte string for y-umlaut to uppercase.
+ return (found_yuml && !ignore_yuml) ? Smi::FromInt(-current_length)
+ : Smi::FromInt(current_length);
} else {
for (int j = 0; j < char_length; j++) {
result->Set(i, chars[j]);
@@ -6244,121 +6255,107 @@
}
-enum AsciiCaseConversion {
- ASCII_TO_LOWER,
- ASCII_TO_UPPER
-};
-
-
-template <AsciiCaseConversion dir>
-struct FastAsciiConverter {
- static bool Convert(char* dst, char* src, int length, bool* changed_out)
{
+template<class Converter>
+static bool FastAsciiConvert(char* dst,
+ char* src,
+ int length,
+ bool* changed_out) {
#ifdef DEBUG
char* saved_dst = dst;
char* saved_src = src;
#endif
- // We rely on the distance between upper and lower case letters
- // being a known power of 2.
- ASSERT('a' - 'A' == (1 << 5));
- // Boundaries for the range of input characters than require
conversion.
- const char lo = (dir == ASCII_TO_LOWER) ? 'A' - 1 : 'a' - 1;
- const char hi = (dir == ASCII_TO_LOWER) ? 'Z' + 1 : 'z' + 1;
- bool changed = false;
- uintptr_t or_acc = 0;
- char* const limit = src + length;
+ DisallowHeapAllocation no_gc;
+ // We rely on the distance between upper and lower case letters
+ // being a known power of 2.
+ ASSERT('a' - 'A' == (1 << 5));
+ // Boundaries for the range of input characters than require conversion.
+ static const char lo = Converter::kIsToLower ? 'A' - 1 : 'a' - 1;
+ static const char hi = Converter::kIsToLower ? 'Z' + 1 : 'z' + 1;
+ bool changed = false;
+ uintptr_t or_acc = 0;
+ char* const limit = src + length;
#ifdef V8_HOST_CAN_READ_UNALIGNED
- // Process the prefix of the input that requires no conversion one
- // (machine) word at a time.
- while (src <= limit - sizeof(uintptr_t)) {
- uintptr_t w = *reinterpret_cast<uintptr_t*>(src);
- or_acc |= w;
- if (AsciiRangeMask(w, lo, hi) != 0) {
- changed = true;
- break;
- }
- *reinterpret_cast<uintptr_t*>(dst) = w;
- src += sizeof(uintptr_t);
- dst += sizeof(uintptr_t);
+ // Process the prefix of the input that requires no conversion one
+ // (machine) word at a time.
+ while (src <= limit - sizeof(uintptr_t)) {
+ uintptr_t w = *reinterpret_cast<uintptr_t*>(src);
+ or_acc |= w;
+ if (AsciiRangeMask(w, lo, hi) != 0) {
+ changed = true;
+ break;
}
- // Process the remainder of the input performing conversion when
- // required one word at a time.
- while (src <= limit - sizeof(uintptr_t)) {
- uintptr_t w = *reinterpret_cast<uintptr_t*>(src);
- or_acc |= w;
- uintptr_t m = AsciiRangeMask(w, lo, hi);
- // The mask has high (7th) bit set in every byte that needs
- // conversion and we know that the distance between cases is
- // 1 << 5.
- *reinterpret_cast<uintptr_t*>(dst) = w ^ (m >> 2);
- src += sizeof(uintptr_t);
- dst += sizeof(uintptr_t);
- }
+ *reinterpret_cast<uintptr_t*>(dst) = w;
+ src += sizeof(uintptr_t);
+ dst += sizeof(uintptr_t);
+ }
+ // Process the remainder of the input performing conversion when
+ // required one word at a time.
+ while (src <= limit - sizeof(uintptr_t)) {
+ uintptr_t w = *reinterpret_cast<uintptr_t*>(src);
+ or_acc |= w;
+ uintptr_t m = AsciiRangeMask(w, lo, hi);
+ // The mask has high (7th) bit set in every byte that needs
+ // conversion and we know that the distance between cases is
+ // 1 << 5.
+ *reinterpret_cast<uintptr_t*>(dst) = w ^ (m >> 2);
+ src += sizeof(uintptr_t);
+ dst += sizeof(uintptr_t);
+ }
#endif
- // Process the last few bytes of the input (or the whole input if
- // unaligned access is not supported).
- while (src < limit) {
- char c = *src;
- or_acc |= c;
- if (lo < c && c < hi) {
- c ^= (1 << 5);
- changed = true;
- }
- *dst = c;
- ++src;
- ++dst;
+ // Process the last few bytes of the input (or the whole input if
+ // unaligned access is not supported).
+ while (src < limit) {
+ char c = *src;
+ or_acc |= c;
+ if (lo < c && c < hi) {
+ c ^= (1 << 5);
+ changed = true;
}
- if ((or_acc & kAsciiMask) != 0) {
- return false;
- }
-#ifdef DEBUG
- CheckConvert(saved_dst, saved_src, length, changed);
-#endif
- *changed_out = changed;
- return true;
+ *dst = c;
+ ++src;
+ ++dst;
+ }
+ if ((or_acc & kAsciiMask) != 0) {
+ return false;
}
+
+ ASSERT(CheckFastAsciiConvert(
+ saved_dst, saved_src, length, changed,
Converter::kIsToLower));
+
+ *changed_out = changed;
+ return true;
+}
#ifdef DEBUG
- static void CheckConvert(char* dst, char* src, int length, bool changed)
{
- bool expected_changed = false;
- for (int i = 0; i < length; i++) {
- if (dst[i] == src[i]) continue;
- expected_changed = true;
- if (dir == ASCII_TO_LOWER) {
- ASSERT('A' <= src[i] && src[i] <= 'Z');
- ASSERT(dst[i] == src[i] + ('a' - 'A'));
- } else {
- ASSERT(dir == ASCII_TO_UPPER);
- ASSERT('a' <= src[i] && src[i] <= 'z');
- ASSERT(dst[i] == src[i] - ('a' - 'A'));
- }
+static bool CheckFastAsciiConvert(char* dst,
+ char* src,
+ int length,
+ bool changed,
+ bool is_to_lower) {
+ bool expected_changed = false;
+ for (int i = 0; i < length; i++) {
+ if (dst[i] == src[i]) continue;
+ expected_changed = true;
+ if (is_to_lower) {
+ ASSERT('A' <= src[i] && src[i] <= 'Z');
+ ASSERT(dst[i] == src[i] + ('a' - 'A'));
+ } else {
+ ASSERT('a' <= src[i] && src[i] <= 'z');
+ ASSERT(dst[i] == src[i] - ('a' - 'A'));
}
- ASSERT(expected_changed == changed);
}
+ return (expected_changed == changed);
+}
#endif
-};
-
-
-struct ToLowerTraits {
- typedef unibrow::ToLowercase UnibrowConverter;
-
- typedef FastAsciiConverter<ASCII_TO_LOWER> AsciiConverter;
-};
-
-
-struct ToUpperTraits {
- typedef unibrow::ToUppercase UnibrowConverter;
-
- typedef FastAsciiConverter<ASCII_TO_UPPER> AsciiConverter;
-};
} // namespace
-template <typename ConvertTraits>
+template <class Converter>
MUST_USE_RESULT static MaybeObject* ConvertCase(
Arguments args,
Isolate* isolate,
- unibrow::Mapping<typename ConvertTraits::UnibrowConverter, 128>*
mapping) {
+ unibrow::Mapping<Converter, 128>* mapping) {
SealHandleScope shs(isolate);
CONVERT_ARG_CHECKED(String, s, 0);
s = s->TryFlattenGetString();
@@ -6380,7 +6377,7 @@
}
SeqOneByteString* result = SeqOneByteString::cast(o);
bool has_changed_character;
- bool is_ascii = ConvertTraits::AsciiConverter::Convert(
+ bool is_ascii = FastAsciiConvert<Converter>(
reinterpret_cast<char*>(result->GetChars()),
reinterpret_cast<char*>(SeqOneByteString::cast(s)->GetChars()),
length,
@@ -6391,31 +6388,35 @@
}
}
+ String::Encoding result_encoding = s->IsOneByteRepresentation()
+ ? String::ONE_BYTE_ENCODING : String::TWO_BYTE_ENCODING;
Object* answer;
- { MaybeObject* maybe_answer =
- ConvertCaseHelper(isolate, s, length, length, mapping);
+ { MaybeObject* maybe_answer = ConvertCaseHelper(
+ isolate, s, result_encoding, length, length, mapping);
if (!maybe_answer->ToObject(&answer)) return maybe_answer;
}
if (answer->IsSmi()) {
- // Retry with correct length.
- { MaybeObject* maybe_answer =
- ConvertCaseHelper(isolate,
- s, Smi::cast(answer)->value(), length,
mapping);
- if (!maybe_answer->ToObject(&answer)) return maybe_answer;
+ int new_length = Smi::cast(answer)->value();
+ if (new_length < 0) {
+ result_encoding = String::TWO_BYTE_ENCODING;
+ new_length = -new_length;
}
+ MaybeObject* maybe_answer = ConvertCaseHelper(
+ isolate, s, result_encoding, new_length, length, mapping);
+ if (!maybe_answer->ToObject(&answer)) return maybe_answer;
}
return answer;
}
RUNTIME_FUNCTION(MaybeObject*, Runtime_StringToLowerCase) {
- return ConvertCase<ToLowerTraits>(
+ return ConvertCase(
args, isolate, isolate->runtime_state()->to_lower_mapping());
}
RUNTIME_FUNCTION(MaybeObject*, Runtime_StringToUpperCase) {
- return ConvertCase<ToUpperTraits>(
+ return ConvertCase(
args, isolate, isolate->runtime_state()->to_upper_mapping());
}
=======================================
--- /branches/3.20/src/unicode.h Fri Mar 15 12:39:50 2013 UTC
+++ /branches/3.20/src/unicode.h Tue Dec 17 12:42:34 2013 UTC
@@ -235,6 +235,7 @@
};
struct ToLowercase {
static const int kMaxWidth = 3;
+ static const bool kIsToLower = true;
static int Convert(uchar c,
uchar n,
uchar* result,
@@ -242,6 +243,7 @@
};
struct ToUppercase {
static const int kMaxWidth = 3;
+ static const bool kIsToLower = false;
static int Convert(uchar c,
uchar n,
uchar* result,
=======================================
--- /branches/3.20/src/version.cc Tue Dec 17 11:02:35 2013 UTC
+++ /branches/3.20/src/version.cc Tue Dec 17 12:42:34 2013 UTC
@@ -35,7 +35,7 @@
#define MAJOR_VERSION 3
#define MINOR_VERSION 20
#define BUILD_NUMBER 17
-#define PATCH_LEVEL 21
+#define PATCH_LEVEL 22
// Use 1 for candidates and 0 otherwise.
// (Boolean macro values are not supported by all preprocessors.)
#define IS_CANDIDATE_VERSION 0
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
