v8/src/top.cc

924 lines
30 KiB
C++
Raw Normal View History

// Copyright 2006-2008 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.
#include "v8.h"
#include "api.h"
#include "bootstrapper.h"
#include "debug.h"
#include "execution.h"
#include "string-stream.h"
#include "platform.h"
namespace v8 { namespace internal {
ThreadLocalTop Top::thread_local_;
Mutex* Top::break_access_ = OS::CreateMutex();
NoAllocationStringAllocator* preallocated_message_space = NULL;
Address top_addresses[] = {
#define C(name) reinterpret_cast<Address>(Top::name()),
TOP_ADDRESS_LIST(C)
#undef C
NULL
};
Address Top::get_address_from_id(Top::AddressId id) {
return top_addresses[id];
}
char* Top::Iterate(ObjectVisitor* v, char* thread_storage) {
ThreadLocalTop* thread = reinterpret_cast<ThreadLocalTop*>(thread_storage);
Iterate(v, thread);
return thread_storage + sizeof(ThreadLocalTop);
}
void Top::Iterate(ObjectVisitor* v, ThreadLocalTop* thread) {
v->VisitPointer(&(thread->pending_exception_));
v->VisitPointer(&(thread->pending_message_obj_));
v->VisitPointer(
bit_cast<Object**, Script**>(&(thread->pending_message_script_)));
v->VisitPointer(bit_cast<Object**, Context**>(&(thread->context_)));
v->VisitPointer(&(thread->scheduled_exception_));
for (v8::TryCatch* block = thread->try_catch_handler_;
block != NULL;
block = block->next_) {
v->VisitPointer(bit_cast<Object**, void**>(&(block->exception_)));
v->VisitPointer(bit_cast<Object**, void**>(&(block->message_)));
}
// Iterate over pointers on native execution stack.
for (StackFrameIterator it(thread); !it.done(); it.Advance()) {
it.frame()->Iterate(v);
}
}
void Top::Iterate(ObjectVisitor* v) {
ThreadLocalTop* current_t = &thread_local_;
Iterate(v, current_t);
}
void Top::InitializeThreadLocal() {
thread_local_.c_entry_fp_ = 0;
thread_local_.handler_ = 0;
thread_local_.stack_is_cooked_ = false;
thread_local_.try_catch_handler_ = NULL;
thread_local_.context_ = NULL;
thread_local_.external_caught_exception_ = false;
thread_local_.failed_access_check_callback_ = NULL;
clear_pending_exception();
clear_pending_message();
clear_scheduled_exception();
thread_local_.save_context_ = NULL;
thread_local_.catcher_ = NULL;
}
// Create a dummy thread that will wait forever on a semaphore. The only
// purpose for this thread is to have some stack area to save essential data
// into for use by a stacks only core dump (aka minidump).
class PreallocatedMemoryThread: public Thread {
public:
PreallocatedMemoryThread() : keep_running_(true) {
wait_for_ever_semaphore_ = OS::CreateSemaphore(0);
data_ready_semaphore_ = OS::CreateSemaphore(0);
}
// When the thread starts running it will allocate a fixed number of bytes
// on the stack and publish the location of this memory for others to use.
void Run() {
EmbeddedVector<char, 15 * 1024> local_buffer;
// Initialize the buffer with a known good value.
OS::StrNCpy(local_buffer, "Trace data was not generated.\n",
local_buffer.length());
// Publish the local buffer and signal its availability.
data_ = local_buffer.start();
length_ = local_buffer.length();
data_ready_semaphore_->Signal();
while (keep_running_) {
// This thread will wait here until the end of time.
wait_for_ever_semaphore_->Wait();
}
// Make sure we access the buffer after the wait to remove all possibility
// of it being optimized away.
OS::StrNCpy(local_buffer, "PreallocatedMemoryThread shutting down.\n",
local_buffer.length());
}
static char* data() {
if (data_ready_semaphore_ != NULL) {
// Initial access is guarded until the data has been published.
data_ready_semaphore_->Wait();
delete data_ready_semaphore_;
data_ready_semaphore_ = NULL;
}
return data_;
}
static unsigned length() {
if (data_ready_semaphore_ != NULL) {
// Initial access is guarded until the data has been published.
data_ready_semaphore_->Wait();
delete data_ready_semaphore_;
data_ready_semaphore_ = NULL;
}
return length_;
}
static void StartThread() {
if (the_thread_ != NULL) return;
the_thread_ = new PreallocatedMemoryThread();
the_thread_->Start();
}
// Stop the PreallocatedMemoryThread and release its resources.
static void StopThread() {
if (the_thread_ == NULL) return;
the_thread_->keep_running_ = false;
wait_for_ever_semaphore_->Signal();
// Wait for the thread to terminate.
the_thread_->Join();
if (data_ready_semaphore_ != NULL) {
delete data_ready_semaphore_;
data_ready_semaphore_ = NULL;
}
delete wait_for_ever_semaphore_;
wait_for_ever_semaphore_ = NULL;
// Done with the thread entirely.
delete the_thread_;
the_thread_ = NULL;
}
private:
// Used to make sure that the thread keeps looping even for spurious wakeups.
bool keep_running_;
// The preallocated memory thread singleton.
static PreallocatedMemoryThread* the_thread_;
// This semaphore is used by the PreallocatedMemoryThread to wait for ever.
static Semaphore* wait_for_ever_semaphore_;
// Semaphore to signal that the data has been initialized.
static Semaphore* data_ready_semaphore_;
// Location and size of the preallocated memory block.
static char* data_;
static unsigned length_;
DISALLOW_COPY_AND_ASSIGN(PreallocatedMemoryThread);
};
PreallocatedMemoryThread* PreallocatedMemoryThread::the_thread_ = NULL;
Semaphore* PreallocatedMemoryThread::wait_for_ever_semaphore_ = NULL;
Semaphore* PreallocatedMemoryThread::data_ready_semaphore_ = NULL;
char* PreallocatedMemoryThread::data_ = NULL;
unsigned PreallocatedMemoryThread::length_ = 0;
static bool initialized = false;
void Top::Initialize() {
CHECK(!initialized);
InitializeThreadLocal();
// Only preallocate on the first initialization.
if (FLAG_preallocate_message_memory && (preallocated_message_space == NULL)) {
// Start the thread which will set aside some memory.
PreallocatedMemoryThread::StartThread();
preallocated_message_space =
new NoAllocationStringAllocator(PreallocatedMemoryThread::data(),
PreallocatedMemoryThread::length());
PreallocatedStorage::Init(PreallocatedMemoryThread::length() / 4);
}
initialized = true;
}
void Top::TearDown() {
if (initialized) {
// Remove the external reference to the preallocated stack memory.
if (preallocated_message_space != NULL) {
delete preallocated_message_space;
preallocated_message_space = NULL;
}
PreallocatedMemoryThread::StopThread();
initialized = false;
}
}
// There are cases where the C stack is separated from JS stack (ARM simulator).
// To figure out the order of top-most JS try-catch handler and the top-most C
// try-catch handler, the C try-catch handler keeps a reference to the top-most
// JS try_catch handler when it was created.
//
// Here is a picture to explain the idea:
// Top::thread_local_.handler_ Top::thread_local_.try_catch_handler_
//
// | |
// v v
//
// | JS handler | | C try_catch handler |
// | next |--+ +-------- | js_handler_ |
// | | | next_ |--+
// | | |
// | JS handler |--+ <---------+ |
// | next |
//
// If the top-most JS try-catch handler is not equal to
// Top::thread_local_.try_catch_handler_.js_handler_, it means the JS handler
// is on the top. Otherwise, it means the C try-catch handler is on the top.
//
void Top::RegisterTryCatchHandler(v8::TryCatch* that) {
StackHandler* handler =
reinterpret_cast<StackHandler*>(thread_local_.handler_);
// Find the top-most try-catch handler.
while (handler != NULL && !handler->is_try_catch()) {
handler = handler->next();
}
that->js_handler_ = handler; // casted to void*
thread_local_.try_catch_handler_ = that;
}
void Top::UnregisterTryCatchHandler(v8::TryCatch* that) {
ASSERT(thread_local_.try_catch_handler_ == that);
thread_local_.try_catch_handler_ = that->next_;
thread_local_.catcher_ = NULL;
}
void Top::MarkCompactPrologue(bool is_compacting) {
MarkCompactPrologue(is_compacting, &thread_local_);
}
void Top::MarkCompactPrologue(bool is_compacting, char* data) {
MarkCompactPrologue(is_compacting, reinterpret_cast<ThreadLocalTop*>(data));
}
void Top::MarkCompactPrologue(bool is_compacting, ThreadLocalTop* thread) {
if (is_compacting) {
StackFrame::CookFramesForThread(thread);
}
}
void Top::MarkCompactEpilogue(bool is_compacting, char* data) {
MarkCompactEpilogue(is_compacting, reinterpret_cast<ThreadLocalTop*>(data));
}
void Top::MarkCompactEpilogue(bool is_compacting) {
MarkCompactEpilogue(is_compacting, &thread_local_);
}
void Top::MarkCompactEpilogue(bool is_compacting, ThreadLocalTop* thread) {
if (is_compacting) {
StackFrame::UncookFramesForThread(thread);
}
}
static int stack_trace_nesting_level = 0;
static StringStream* incomplete_message = NULL;
Handle<String> Top::StackTrace() {
if (stack_trace_nesting_level == 0) {
stack_trace_nesting_level++;
HeapStringAllocator allocator;
StringStream::ClearMentionedObjectCache();
StringStream accumulator(&allocator);
incomplete_message = &accumulator;
PrintStack(&accumulator);
Handle<String> stack_trace = accumulator.ToString();
incomplete_message = NULL;
stack_trace_nesting_level = 0;
return stack_trace;
} else if (stack_trace_nesting_level == 1) {
stack_trace_nesting_level++;
OS::PrintError(
"\n\nAttempt to print stack while printing stack (double fault)\n");
OS::PrintError(
"If you are lucky you may find a partial stack dump on stdout.\n\n");
incomplete_message->OutputToStdOut();
return Factory::empty_symbol();
} else {
OS::Abort();
// Unreachable
return Factory::empty_symbol();
}
}
void Top::PrintStack() {
if (stack_trace_nesting_level == 0) {
stack_trace_nesting_level++;
StringAllocator* allocator;
if (preallocated_message_space == NULL) {
allocator = new HeapStringAllocator();
} else {
allocator = preallocated_message_space;
}
NativeAllocationChecker allocation_checker(
!FLAG_preallocate_message_memory ?
NativeAllocationChecker::ALLOW :
NativeAllocationChecker::DISALLOW);
StringStream::ClearMentionedObjectCache();
StringStream accumulator(allocator);
incomplete_message = &accumulator;
PrintStack(&accumulator);
accumulator.OutputToStdOut();
accumulator.Log();
incomplete_message = NULL;
stack_trace_nesting_level = 0;
if (preallocated_message_space == NULL) {
// Remove the HeapStringAllocator created above.
delete allocator;
}
} else if (stack_trace_nesting_level == 1) {
stack_trace_nesting_level++;
OS::PrintError(
"\n\nAttempt to print stack while printing stack (double fault)\n");
OS::PrintError(
"If you are lucky you may find a partial stack dump on stdout.\n\n");
incomplete_message->OutputToStdOut();
}
}
static void PrintFrames(StringStream* accumulator,
StackFrame::PrintMode mode) {
StackFrameIterator it;
for (int i = 0; !it.done(); it.Advance()) {
it.frame()->Print(accumulator, mode, i++);
}
}
void Top::PrintStack(StringStream* accumulator) {
// The MentionedObjectCache is not GC-proof at the moment.
AssertNoAllocation nogc;
ASSERT(StringStream::IsMentionedObjectCacheClear());
// Avoid printing anything if there are no frames.
if (c_entry_fp(GetCurrentThread()) == 0) return;
accumulator->Add(
"\n==== Stack trace ============================================\n\n");
PrintFrames(accumulator, StackFrame::OVERVIEW);
accumulator->Add(
"\n==== Details ================================================\n\n");
PrintFrames(accumulator, StackFrame::DETAILS);
accumulator->PrintMentionedObjectCache();
accumulator->Add("=====================\n\n");
}
void Top::SetFailedAccessCheckCallback(v8::FailedAccessCheckCallback callback) {
ASSERT(thread_local_.failed_access_check_callback_ == NULL);
thread_local_.failed_access_check_callback_ = callback;
}
void Top::ReportFailedAccessCheck(JSObject* receiver, v8::AccessType type) {
if (!thread_local_.failed_access_check_callback_) return;
ASSERT(receiver->IsAccessCheckNeeded());
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
ASSERT(Top::context());
// The callers of this method are not expecting a GC.
AssertNoAllocation no_gc;
// Get the data object from access check info.
JSFunction* constructor = JSFunction::cast(receiver->map()->constructor());
Object* info = constructor->shared()->function_data();
if (info == Heap::undefined_value()) return;
Object* data_obj = FunctionTemplateInfo::cast(info)->access_check_info();
if (data_obj == Heap::undefined_value()) return;
HandleScope scope;
Handle<JSObject> receiver_handle(receiver);
Handle<Object> data(AccessCheckInfo::cast(data_obj)->data());
thread_local_.failed_access_check_callback_(
v8::Utils::ToLocal(receiver_handle),
type,
v8::Utils::ToLocal(data));
}
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
enum MayAccessDecision {
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
YES, NO, UNKNOWN
};
static MayAccessDecision MayAccessPreCheck(JSObject* receiver,
v8::AccessType type) {
// During bootstrapping, callback functions are not enabled yet.
if (Bootstrapper::IsActive()) return YES;
if (receiver->IsJSGlobalProxy()) {
Object* receiver_context = JSGlobalProxy::cast(receiver)->context();
if (!receiver_context->IsContext()) return NO;
// Get the global context of current top context.
// avoid using Top::global_context() because it uses Handle.
Context* global_context = Top::context()->global()->global_context();
if (receiver_context == global_context) return YES;
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
if (Context::cast(receiver_context)->security_token() ==
global_context->security_token())
return YES;
}
return UNKNOWN;
}
bool Top::MayNamedAccess(JSObject* receiver, Object* key, v8::AccessType type) {
ASSERT(receiver->IsAccessCheckNeeded());
// Check for compatibility between the security tokens in the
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
// current lexical context and the accessed object.
ASSERT(Top::context());
// The callers of this method are not expecting a GC.
AssertNoAllocation no_gc;
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
MayAccessDecision decision = MayAccessPreCheck(receiver, type);
if (decision != UNKNOWN) return decision == YES;
// Get named access check callback
JSFunction* constructor = JSFunction::cast(receiver->map()->constructor());
Object* info = constructor->shared()->function_data();
if (info == Heap::undefined_value()) return false;
Object* data_obj = FunctionTemplateInfo::cast(info)->access_check_info();
if (data_obj == Heap::undefined_value()) return false;
Object* fun_obj = AccessCheckInfo::cast(data_obj)->named_callback();
v8::NamedSecurityCallback callback =
v8::ToCData<v8::NamedSecurityCallback>(fun_obj);
if (!callback) return false;
HandleScope scope;
Handle<JSObject> receiver_handle(receiver);
Handle<Object> key_handle(key);
Handle<Object> data(AccessCheckInfo::cast(data_obj)->data());
LOG(ApiNamedSecurityCheck(key));
bool result = false;
{
// Leaving JavaScript.
VMState state(EXTERNAL);
result = callback(v8::Utils::ToLocal(receiver_handle),
v8::Utils::ToLocal(key_handle),
type,
v8::Utils::ToLocal(data));
}
return result;
}
bool Top::MayIndexedAccess(JSObject* receiver,
uint32_t index,
v8::AccessType type) {
ASSERT(receiver->IsAccessCheckNeeded());
// Check for compatibility between the security tokens in the
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
// current lexical context and the accessed object.
ASSERT(Top::context());
// The callers of this method are not expecting a GC.
AssertNoAllocation no_gc;
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
MayAccessDecision decision = MayAccessPreCheck(receiver, type);
if (decision != UNKNOWN) return decision == YES;
// Get indexed access check callback
JSFunction* constructor = JSFunction::cast(receiver->map()->constructor());
Object* info = constructor->shared()->function_data();
if (info == Heap::undefined_value()) return false;
Object* data_obj = FunctionTemplateInfo::cast(info)->access_check_info();
if (data_obj == Heap::undefined_value()) return false;
Object* fun_obj = AccessCheckInfo::cast(data_obj)->indexed_callback();
v8::IndexedSecurityCallback callback =
v8::ToCData<v8::IndexedSecurityCallback>(fun_obj);
if (!callback) return false;
HandleScope scope;
Handle<JSObject> receiver_handle(receiver);
Handle<Object> data(AccessCheckInfo::cast(data_obj)->data());
LOG(ApiIndexedSecurityCheck(index));
bool result = false;
{
// Leaving JavaScript.
VMState state(EXTERNAL);
result = callback(v8::Utils::ToLocal(receiver_handle),
index,
type,
v8::Utils::ToLocal(data));
}
return result;
}
const char* Top::kStackOverflowMessage =
"Uncaught RangeError: Maximum call stack size exceeded";
Failure* Top::StackOverflow() {
HandleScope scope;
Handle<String> key = Factory::stack_overflow_symbol();
Handle<JSObject> boilerplate =
Split window support from V8. Here is a description of the background and design of split window in Chrome and V8: https://docs.google.com/a/google.com/Doc?id=chhjkpg_47fwddxbfr This change list splits the window object into two parts: 1) an inner window object used as the global object of contexts; 2) an outer window object exposed to JavaScript and accessible by the name 'window'. Firefox did it awhile ago, here are some discussions: https://wiki.mozilla.org/Gecko:SplitWindow. One additional benefit of splitting window in Chrome is that accessing global variables don't need security checks anymore, it can improve applications that use many global variables. V8 support of split window: There are a small number of changes on V8 api to support split window: Security context is removed from V8, so does related API functions; A global object can be detached from its context and reused by a new context; Access checks on an object template can be turned on/off by default; An object can turn on its access checks later; V8 has a new object type, ApiGlobalObject, which is the outer window object type. The existing JSGlobalObject becomes the inner window object type. Security checks are moved from JSGlobalObject to ApiGlobalObject. ApiGlobalObject is the one exposed to JavaScript, it is accessible through Context::Global(). ApiGlobalObject's prototype is set to JSGlobalObject so that property lookups are forwarded to JSGlobalObject. ApiGlobalObject forwards all other property access requests to JSGlobalObject, such as SetProperty, DeleteProperty, etc. Security token is moved to a global context, and ApiGlobalObject has a reference to its global context. JSGlobalObject has a reference to its global context as well. When accessing properties on a global object in JavaScript, the domain security check is performed by comparing the security token of the lexical context (Top::global_context()) to the token of global object's context. The check is only needed when the receiver is a window object, such as 'window.document'. Accessing global variables, such as 'var foo = 3; foo' does not need checks because the receiver is the inner window object. When an outer window is detached from its global context (when a frame navigates away from a page), it is completely detached from the inner window. A new context is created for the new page, and the outer global object is reused. At this point, the access check on the DOMWindow wrapper of the old context is turned on. The code in old context is still able to access DOMWindow properties, but it has to go through domain security checks. It is debatable on how to implement the outer window object. Currently each property access function has to check if the receiver is ApiGlobalObject type. This approach might be error-prone that one may forget to check the receiver when adding new functions. It is unlikely a performance issue because accessing global variables are more common than 'window.foo' style coding. I am still working on the ARM port, and I'd like to hear comments and suggestions on the best way to support it in V8. Review URL: http://codereview.chromium.org/7366 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2008-10-21 19:07:58 +00:00
Handle<JSObject>::cast(GetProperty(Top::builtins(), key));
Handle<Object> exception = Copy(boilerplate);
// TODO(1240995): To avoid having to call JavaScript code to compute
// the message for stack overflow exceptions which is very likely to
// double fault with another stack overflow exception, we use a
// precomputed message. This is somewhat problematic in that it
// doesn't use ReportUncaughtException to determine the location
// from where the exception occurred. It should probably be
// reworked.
DoThrow(*exception, NULL, kStackOverflowMessage);
return Failure::Exception();
}
Failure* Top::Throw(Object* exception, MessageLocation* location) {
DoThrow(exception, location, NULL);
return Failure::Exception();
}
Failure* Top::ReThrow(Object* exception, MessageLocation* location) {
// Set the exception being re-thrown.
set_pending_exception(exception);
return Failure::Exception();
}
void Top::ScheduleThrow(Object* exception) {
// When scheduling a throw we first throw the exception to get the
// error reporting if it is uncaught before rescheduling it.
Throw(exception);
thread_local_.scheduled_exception_ = pending_exception();
thread_local_.external_caught_exception_ = false;
clear_pending_exception();
}
Object* Top::PromoteScheduledException() {
Object* thrown = scheduled_exception();
clear_scheduled_exception();
// Re-throw the exception to avoid getting repeated error reporting.
return ReThrow(thrown);
}
void Top::PrintCurrentStackTrace(FILE* out) {
StackTraceFrameIterator it;
while (!it.done()) {
HandleScope scope;
// Find code position if recorded in relocation info.
JavaScriptFrame* frame = it.frame();
int pos = frame->code()->SourcePosition(frame->pc());
Handle<Object> pos_obj(Smi::FromInt(pos));
// Fetch function and receiver.
Handle<JSFunction> fun(JSFunction::cast(frame->function()));
Handle<Object> recv(frame->receiver());
// Advance to the next JavaScript frame and determine if the
// current frame is the top-level frame.
it.Advance();
Handle<Object> is_top_level = it.done()
? Factory::true_value()
: Factory::false_value();
// Generate and print stack trace line.
Handle<String> line =
Execution::GetStackTraceLine(recv, fun, pos_obj, is_top_level);
if (line->length() > 0) {
line->PrintOn(out);
fprintf(out, "\n");
}
}
}
void Top::ComputeLocation(MessageLocation* target) {
*target = MessageLocation(empty_script(), -1, -1);
StackTraceFrameIterator it;
if (!it.done()) {
JavaScriptFrame* frame = it.frame();
JSFunction* fun = JSFunction::cast(frame->function());
Object* script = fun->shared()->script();
if (script->IsScript() &&
!(Script::cast(script)->source()->IsUndefined())) {
int pos = frame->code()->SourcePosition(frame->pc());
// Compute the location from the function and the reloc info.
Handle<Script> casted_script(Script::cast(script));
*target = MessageLocation(casted_script, pos, pos + 1);
}
}
}
void Top::ReportUncaughtException(Handle<Object> exception,
MessageLocation* location,
Handle<String> stack_trace) {
Handle<Object> message =
MessageHandler::MakeMessageObject("uncaught_exception",
location,
HandleVector<Object>(&exception, 1),
stack_trace);
// Report the uncaught exception.
MessageHandler::ReportMessage(location, message);
}
bool Top::ShouldReportException(bool* is_caught_externally) {
// Find the top-most try-catch handler.
StackHandler* handler =
StackHandler::FromAddress(Top::handler(Top::GetCurrentThread()));
while (handler != NULL && !handler->is_try_catch()) {
handler = handler->next();
}
// Get the address of the external handler so we can compare the address to
// determine which one is closer to the top of the stack.
v8::TryCatch* try_catch = thread_local_.try_catch_handler_;
// The exception has been externally caught if and only if there is
// an external handler which is on top of the top-most try-catch
// handler.
//
// See comments in RegisterTryCatchHandler for details.
*is_caught_externally = try_catch != NULL &&
(handler == NULL || handler == try_catch->js_handler_);
if (*is_caught_externally) {
// Only report the exception if the external handler is verbose.
return thread_local_.try_catch_handler_->is_verbose_;
} else {
// Report the exception if it isn't caught by JavaScript code.
return handler == NULL;
}
}
void Top::DoThrow(Object* exception,
MessageLocation* location,
const char* message) {
ASSERT(!has_pending_exception());
HandleScope scope;
Handle<Object> exception_handle(exception);
// Determine reporting and whether the exception is caught externally.
bool is_caught_externally = false;
bool is_out_of_memory = exception == Failure::OutOfMemoryException();
bool should_return_exception = ShouldReportException(&is_caught_externally);
bool report_exception = !is_out_of_memory && should_return_exception;
#ifdef ENABLE_DEBUGGER_SUPPORT
// Notify debugger of exception.
Debugger::OnException(exception_handle, report_exception);
#endif
// Generate the message.
Handle<Object> message_obj;
MessageLocation potential_computed_location;
bool try_catch_needs_message =
is_caught_externally &&
thread_local_.try_catch_handler_->capture_message_;
if (report_exception || try_catch_needs_message) {
if (location == NULL) {
// If no location was specified we use a computed one instead
ComputeLocation(&potential_computed_location);
location = &potential_computed_location;
}
Handle<String> stack_trace;
if (FLAG_trace_exception) stack_trace = StackTrace();
message_obj = MessageHandler::MakeMessageObject("uncaught_exception",
location, HandleVector<Object>(&exception_handle, 1), stack_trace);
}
// Save the message for reporting if the the exception remains uncaught.
thread_local_.has_pending_message_ = report_exception;
thread_local_.pending_message_ = message;
if (!message_obj.is_null()) {
thread_local_.pending_message_obj_ = *message_obj;
if (location != NULL) {
thread_local_.pending_message_script_ = *location->script();
thread_local_.pending_message_start_pos_ = location->start_pos();
thread_local_.pending_message_end_pos_ = location->end_pos();
}
}
if (is_caught_externally) {
thread_local_.catcher_ = thread_local_.try_catch_handler_;
}
// NOTE: Notifying the debugger or generating the message
// may have caused new exceptions. For now, we just ignore
// that and set the pending exception to the original one.
set_pending_exception(*exception_handle);
}
void Top::ReportPendingMessages() {
ASSERT(has_pending_exception());
setup_external_caught();
// If the pending exception is OutOfMemoryException set out_of_memory in
// the global context. Note: We have to mark the global context here
// since the GenerateThrowOutOfMemory stub cannot make a RuntimeCall to
// set it.
HandleScope scope;
if (thread_local_.pending_exception_ == Failure::OutOfMemoryException()) {
context()->mark_out_of_memory();
} else {
Handle<Object> exception(pending_exception());
bool external_caught = thread_local_.external_caught_exception_;
thread_local_.external_caught_exception_ = false;
if (external_caught) {
thread_local_.try_catch_handler_->exception_ =
thread_local_.pending_exception_;
if (!thread_local_.pending_message_obj_->IsTheHole()) {
try_catch_handler()->message_ = thread_local_.pending_message_obj_;
}
}
if (thread_local_.has_pending_message_) {
thread_local_.has_pending_message_ = false;
if (thread_local_.pending_message_ != NULL) {
MessageHandler::ReportMessage(thread_local_.pending_message_);
} else if (!thread_local_.pending_message_obj_->IsTheHole()) {
Handle<Object> message_obj(thread_local_.pending_message_obj_);
if (thread_local_.pending_message_script_ != NULL) {
Handle<Script> script(thread_local_.pending_message_script_);
int start_pos = thread_local_.pending_message_start_pos_;
int end_pos = thread_local_.pending_message_end_pos_;
MessageLocation location(script, start_pos, end_pos);
MessageHandler::ReportMessage(&location, message_obj);
} else {
MessageHandler::ReportMessage(NULL, message_obj);
}
}
}
thread_local_.external_caught_exception_ = external_caught;
set_pending_exception(*exception);
}
clear_pending_message();
}
void Top::TraceException(bool flag) {
FLAG_trace_exception = flag;
}
bool Top::optional_reschedule_exception(bool is_bottom_call) {
// Allways reschedule out of memory exceptions.
if (!is_out_of_memory()) {
// Never reschedule the exception if this is the bottom call.
bool clear_exception = is_bottom_call;
// If the exception is externally caught, clear it if there are no
// JavaScript frames on the way to the C++ frame that has the
// external handler.
if (thread_local_.external_caught_exception_) {
ASSERT(thread_local_.try_catch_handler_ != NULL);
Address external_handler_address =
reinterpret_cast<Address>(thread_local_.try_catch_handler_);
JavaScriptFrameIterator it;
if (it.done() || (it.frame()->sp() > external_handler_address)) {
clear_exception = true;
}
}
// Clear the exception if needed.
if (clear_exception) {
thread_local_.external_caught_exception_ = false;
clear_pending_exception();
return false;
}
}
// Reschedule the exception.
thread_local_.scheduled_exception_ = pending_exception();
clear_pending_exception();
return true;
}
bool Top::is_out_of_memory() {
if (has_pending_exception()) {
Object* e = pending_exception();
if (e->IsFailure() && Failure::cast(e)->IsOutOfMemoryException()) {
return true;
}
}
if (has_scheduled_exception()) {
Object* e = scheduled_exception();
if (e->IsFailure() && Failure::cast(e)->IsOutOfMemoryException()) {
return true;
}
}
return false;
}
Handle<Context> Top::global_context() {
GlobalObject* global = thread_local_.context_->global();
return Handle<Context>(global->global_context());
}
Object* Top::LookupSpecialFunction(JSObject* receiver,
JSObject* prototype,
JSFunction* function) {
if (receiver->IsJSArray()) {
FixedArray* table = context()->global_context()->special_function_table();
for (int index = 0; index < table->length(); index +=3) {
if ((prototype == table->get(index)) &&
(function == table->get(index+1))) {
return table->get(index+2);
}
}
}
return Heap::undefined_value();
}
char* Top::ArchiveThread(char* to) {
memcpy(to, reinterpret_cast<char*>(&thread_local_), sizeof(thread_local_));
InitializeThreadLocal();
return to + sizeof(thread_local_);
}
char* Top::RestoreThread(char* from) {
memcpy(reinterpret_cast<char*>(&thread_local_), from, sizeof(thread_local_));
return from + sizeof(thread_local_);
}
ExecutionAccess::ExecutionAccess() {
Top::break_access_->Lock();
}
ExecutionAccess::~ExecutionAccess() {
Top::break_access_->Unlock();
}
} } // namespace v8::internal