mirror of
https://github.com/klzgrad/naiveproxy.git
synced 2024-11-28 08:16:09 +03:00
243 lines
7.1 KiB
C++
243 lines
7.1 KiB
C++
// Copyright (c) 2013 The Chromium Authors. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style license that can be
|
|
// found in the LICENSE file.
|
|
|
|
#include "tools/gn/scheduler.h"
|
|
|
|
#include <algorithm>
|
|
|
|
#include "base/bind.h"
|
|
#include "base/command_line.h"
|
|
#include "base/single_thread_task_runner.h"
|
|
#include "base/strings/string_number_conversions.h"
|
|
#include "build/build_config.h"
|
|
#include "tools/gn/standard_out.h"
|
|
#include "tools/gn/switches.h"
|
|
#include "tools/gn/target.h"
|
|
|
|
#if defined(OS_WIN)
|
|
#include <windows.h>
|
|
#else
|
|
#include <unistd.h>
|
|
#endif
|
|
|
|
Scheduler* g_scheduler = nullptr;
|
|
|
|
namespace {
|
|
|
|
#if defined(OS_WIN)
|
|
int GetCPUCount() {
|
|
SYSTEM_INFO sysinfo;
|
|
::GetSystemInfo(&sysinfo);
|
|
return sysinfo.dwNumberOfProcessors;
|
|
}
|
|
#else
|
|
int GetCPUCount() {
|
|
return static_cast<int>(sysconf(_SC_NPROCESSORS_ONLN));
|
|
}
|
|
#endif
|
|
|
|
int GetThreadCount() {
|
|
std::string thread_count =
|
|
base::CommandLine::ForCurrentProcess()->GetSwitchValueASCII(
|
|
switches::kThreads);
|
|
|
|
// See if an override was specified on the command line.
|
|
int result;
|
|
if (!thread_count.empty() && base::StringToInt(thread_count, &result))
|
|
return result;
|
|
|
|
// Base the default number of worker threads on number of cores in the
|
|
// system. When building large projects, the speed can be limited by how fast
|
|
// the main thread can dispatch work and connect the dependency graph. If
|
|
// there are too many worker threads, the main thread can be starved and it
|
|
// will run slower overall.
|
|
//
|
|
// One less worker thread than the number of physical CPUs seems to be a
|
|
// good value, both theoretically and experimentally. But always use at
|
|
// least some workers to prevent us from being too sensitive to I/O latency
|
|
// on low-end systems.
|
|
//
|
|
// The minimum thread count is based on measuring the optimal threads for the
|
|
// Chrome build on a several-year-old 4-core MacBook.
|
|
int num_cores = GetCPUCount() / 2; // Almost all CPUs now are hyperthreaded.
|
|
return std::max(num_cores - 1, 8);
|
|
}
|
|
|
|
} // namespace
|
|
|
|
Scheduler::Scheduler()
|
|
: pool_(new base::SequencedWorkerPool(GetThreadCount(),
|
|
"worker_",
|
|
base::TaskPriority::USER_VISIBLE)),
|
|
input_file_manager_(new InputFileManager),
|
|
verbose_logging_(false),
|
|
work_count_(0),
|
|
is_failed_(false),
|
|
has_been_shutdown_(false) {
|
|
g_scheduler = this;
|
|
}
|
|
|
|
Scheduler::~Scheduler() {
|
|
if (!has_been_shutdown_)
|
|
pool_->Shutdown();
|
|
g_scheduler = nullptr;
|
|
}
|
|
|
|
bool Scheduler::Run() {
|
|
runner_.Run();
|
|
bool local_is_failed;
|
|
{
|
|
base::AutoLock lock(lock_);
|
|
local_is_failed = is_failed();
|
|
has_been_shutdown_ = true;
|
|
}
|
|
// Don't do this inside the lock since it will block on the workers, which
|
|
// may be in turn waiting on the lock.
|
|
pool_->Shutdown();
|
|
return !local_is_failed;
|
|
}
|
|
|
|
void Scheduler::Log(const std::string& verb, const std::string& msg) {
|
|
if (task_runner()->BelongsToCurrentThread()) {
|
|
LogOnMainThread(verb, msg);
|
|
} else {
|
|
// The run loop always joins on the sub threads, so the lifetime of this
|
|
// object outlives the invocations of this function, hence "unretained".
|
|
task_runner()->PostTask(FROM_HERE,
|
|
base::Bind(&Scheduler::LogOnMainThread,
|
|
base::Unretained(this), verb, msg));
|
|
}
|
|
}
|
|
|
|
void Scheduler::FailWithError(const Err& err) {
|
|
DCHECK(err.has_error());
|
|
{
|
|
base::AutoLock lock(lock_);
|
|
|
|
if (is_failed_ || has_been_shutdown_)
|
|
return; // Ignore errors once we see one.
|
|
is_failed_ = true;
|
|
}
|
|
|
|
if (task_runner()->BelongsToCurrentThread()) {
|
|
FailWithErrorOnMainThread(err);
|
|
} else {
|
|
// The run loop always joins on the sub threads, so the lifetime of this
|
|
// object outlives the invocations of this function, hence "unretained".
|
|
task_runner()->PostTask(FROM_HERE,
|
|
base::Bind(&Scheduler::FailWithErrorOnMainThread,
|
|
base::Unretained(this), err));
|
|
}
|
|
}
|
|
|
|
void Scheduler::ScheduleWork(const base::Closure& work) {
|
|
IncrementWorkCount();
|
|
pool_->PostWorkerTaskWithShutdownBehavior(
|
|
FROM_HERE, base::Bind(&Scheduler::DoWork,
|
|
base::Unretained(this), work),
|
|
base::SequencedWorkerPool::BLOCK_SHUTDOWN);
|
|
}
|
|
|
|
void Scheduler::AddGenDependency(const base::FilePath& file) {
|
|
base::AutoLock lock(lock_);
|
|
gen_dependencies_.push_back(file);
|
|
}
|
|
|
|
std::vector<base::FilePath> Scheduler::GetGenDependencies() const {
|
|
base::AutoLock lock(lock_);
|
|
return gen_dependencies_;
|
|
}
|
|
|
|
void Scheduler::AddWrittenFile(const SourceFile& file) {
|
|
base::AutoLock lock(lock_);
|
|
written_files_.push_back(file);
|
|
}
|
|
|
|
void Scheduler::AddUnknownGeneratedInput(const Target* target,
|
|
const SourceFile& file) {
|
|
base::AutoLock lock(lock_);
|
|
unknown_generated_inputs_.insert(std::make_pair(file, target));
|
|
}
|
|
|
|
void Scheduler::AddWriteRuntimeDepsTarget(const Target* target) {
|
|
base::AutoLock lock(lock_);
|
|
write_runtime_deps_targets_.push_back(target);
|
|
}
|
|
|
|
std::vector<const Target*> Scheduler::GetWriteRuntimeDepsTargets() const {
|
|
base::AutoLock lock(lock_);
|
|
return write_runtime_deps_targets_;
|
|
}
|
|
|
|
bool Scheduler::IsFileGeneratedByWriteRuntimeDeps(
|
|
const OutputFile& file) const {
|
|
base::AutoLock lock(lock_);
|
|
// Number of targets should be quite small, so brute-force search is fine.
|
|
for (const Target* target : write_runtime_deps_targets_) {
|
|
if (file == target->write_runtime_deps_output()) {
|
|
return true;
|
|
}
|
|
}
|
|
return false;
|
|
}
|
|
|
|
std::multimap<SourceFile, const Target*>
|
|
Scheduler::GetUnknownGeneratedInputs() const {
|
|
base::AutoLock lock(lock_);
|
|
|
|
// Remove all unknown inputs that were written files. These are OK as inputs
|
|
// to build steps since they were written as a side-effect of running GN.
|
|
//
|
|
// It's assumed that this function is called once during cleanup to check for
|
|
// errors, so performing this work in the lock doesn't matter.
|
|
std::multimap<SourceFile, const Target*> filtered = unknown_generated_inputs_;
|
|
for (const SourceFile& file : written_files_)
|
|
filtered.erase(file);
|
|
|
|
return filtered;
|
|
}
|
|
|
|
void Scheduler::ClearUnknownGeneratedInputsAndWrittenFiles() {
|
|
base::AutoLock lock(lock_);
|
|
unknown_generated_inputs_.clear();
|
|
written_files_.clear();
|
|
}
|
|
|
|
void Scheduler::IncrementWorkCount() {
|
|
work_count_.Increment();
|
|
}
|
|
|
|
void Scheduler::DecrementWorkCount() {
|
|
if (!work_count_.Decrement()) {
|
|
if (task_runner()->BelongsToCurrentThread()) {
|
|
OnComplete();
|
|
} else {
|
|
task_runner()->PostTask(FROM_HERE, base::Bind(&Scheduler::OnComplete,
|
|
base::Unretained(this)));
|
|
}
|
|
}
|
|
}
|
|
|
|
void Scheduler::LogOnMainThread(const std::string& verb,
|
|
const std::string& msg) {
|
|
OutputString(verb, DECORATION_YELLOW);
|
|
OutputString(" " + msg + "\n");
|
|
}
|
|
|
|
void Scheduler::FailWithErrorOnMainThread(const Err& err) {
|
|
err.PrintToStdout();
|
|
runner_.Quit();
|
|
}
|
|
|
|
void Scheduler::DoWork(const base::Closure& closure) {
|
|
closure.Run();
|
|
DecrementWorkCount();
|
|
}
|
|
|
|
void Scheduler::OnComplete() {
|
|
// Should be called on the main thread.
|
|
DCHECK(task_runner()->BelongsToCurrentThread());
|
|
runner_.Quit();
|
|
}
|