// For each thread, we have a ThreadData that stores all tracking info generated
// on this thread. This prevents the need for locking as data accumulates.

该文件用到了chromium之ThreadLocalStorage

看看头文件

class ThreadData {
public:
typedef std::map<Location, Births*> BirthMap;
typedef std::map<const Births*, DeathData> DeathMap; ThreadData(); // Using Thread Local Store, find the current instance for collecting data.
// If an instance does not exist, construct one (and remember it for use on
// this thread.
// If shutdown has already started, and we don't yet have an instance, then
// return null.
static ThreadData* current(); // For a given about:objects URL, develop resulting HTML, and append to
// output.
static void WriteHTML(const std::string& query, std::string* output); // For a given accumulated array of results, use the comparator to sort and
// subtotal, writing the results to the output.
static void WriteHTMLTotalAndSubtotals(
const DataCollector::Collection& match_array,
const Comparator& comparator, std::string* output); // In this thread's data, find a place to record a new birth.
Births* FindLifetime(const Location& location); // Find a place to record a death on this thread.
void TallyADeath(const Births& lifetimes, const base::TimeDelta& duration); // (Thread safe) Get start of list of instances.
static ThreadData* first();
// Iterate through the null terminated list of instances.
ThreadData* next() const { return next_; } MessageLoop* message_loop() const { return message_loop_; }
const std::string ThreadName() const; // Using our lock, make a copy of the specified maps. These calls may arrive
// from non-local threads.
void SnapshotBirthMap(BirthMap *output) const;
void SnapshotDeathMap(DeathMap *output) const; static void RunOnAllThreads(void (*Func)()); // Set internal status_ to either become ACTIVE, or later, to be SHUTDOWN,
// based on argument being true or false respectively.
// IF tracking is not compiled in, this function will return false.
static bool StartTracking(bool status);
static bool IsActive(); #ifdef OS_WIN
// WARNING: ONLY call this function when all MessageLoops are still intact for
// all registered threads. IF you call it later, you will crash.
// Note: You don't need to call it at all, and you can wait till you are
// single threaded (again) to do the cleanup via
// ShutdownSingleThreadedCleanup().
// Start the teardown (shutdown) process in a multi-thread mode by disabling
// further additions to thread database on all threads. First it makes a
// local (locked) change to prevent any more threads from registering. Then
// it Posts a Task to all registered threads to be sure they are aware that no
// more accumulation can take place.
static void ShutdownMultiThreadTracking();
#endif // WARNING: ONLY call this function when you are running single threaded
// (again) and all message loops and threads have terminated. Until that
// point some threads may still attempt to write into our data structures.
// Delete recursively all data structures, starting with the list of
// ThreadData instances.
static void ShutdownSingleThreadedCleanup();

大致想想怎么用

1. 开始

开始运行,

StartTracking(true)

2. 结束,并输出

StartTracking(false)
WriteHTML(...)

有个测试用例,可以看看
TEST_F(TrackedObjectsTest, MinimalStartupShutdown) {
// Minimal test doesn't even create any tasks.
if (!ThreadData::StartTracking(true))
return; EXPECT_FALSE(ThreadData::first()); // No activity even on this thread.
ThreadData* data = ThreadData::current();
EXPECT_TRUE(ThreadData::first()); // Now class was constructed.
EXPECT_TRUE(data);
EXPECT_TRUE(!data->next());
EXPECT_EQ(data, ThreadData::current());
ThreadData::BirthMap birth_map;
data->SnapshotBirthMap(&birth_map); // Get all birth data
EXPECT_EQ(0u, birth_map.size());
ThreadData::DeathMap death_map;
data->SnapshotDeathMap(&death_map);
EXPECT_EQ(0u, death_map.size());
ThreadData::ShutdownSingleThreadedCleanup(); // Do it again, just to be sure we reset state completely.
ThreadData::StartTracking(true);
EXPECT_FALSE(ThreadData::first()); // No activity even on this thread.
data = ThreadData::current();
EXPECT_TRUE(ThreadData::first()); // Now class was constructed.
EXPECT_TRUE(data);
EXPECT_TRUE(!data->next());
EXPECT_EQ(data, ThreadData::current());
birth_map.clear();
data->SnapshotBirthMap(&birth_map);
EXPECT_EQ(0u, birth_map.size());
death_map.clear();
data->SnapshotDeathMap(&death_map);
EXPECT_EQ(0u, death_map.size());
ThreadData::ShutdownSingleThreadedCleanup();
}

chromium之tracked_objects的更多相关文章

  1. chromium之tracked

    //------------------------------------------------------------------------------ // Tracked is the b ...

  2. 【Chromium中文文档】线程

    线程 转载请注明出处:https://ahangchen.gitbooks.io/chromium_doc_zh/content/zh//General_Architecture/Threading. ...

  3. chromium之task

    // A task is a generic runnable thingy, usually used for running code on a // different thread or fo ...

  4. chromium源码阅读

    linux下chromium的入口函数在文件:src/chrome/app/chrome_exe_main_aura.cc 中 int main(int argc, const char** argv ...

  5. Chromium多线程模型设计和实现分析

    Chromium除了远近闻名的多进程架构之外,它的多线程模型也相当引人注目的.Chromium的多进程架构是为了解决网页的稳定性问题,而多线程模型则是为了解决网页的卡顿问题.为了达到这个目的,Chro ...

  6. QT5利用chromium内核与HTML页面交互

    在QT5.4之前,做QT开发浏览器只能选择QWebkit,但是有过使用的都会发现,这个webkit不是出奇的慢,简直是慢的令人发指,Release模式下还行,debug下你就无语了,但是webkit毕 ...

  7. Google之Chromium浏览器源码学习——base公共通用库(一)

    Google的优秀C++开源项目繁多,其中的Chromium浏览器项目可以说是很具有代表性的,此外还包括其第三开发开源库或是自己的优秀开源库,可以根据需要抽取自己感兴趣的部分.在研究.学习该项目前的时 ...

  8. 如何在windows上编译Chromium (CEF3) 并加入MP3支持(二)

    时隔一年,再次编译cef3,独一无二的目的仍为加入mp3支持.新版本的编译环境和注意事项都已经发生了变化,于是再记录一下. 一.编译版本 cef版本号格式为X.YYYY.A.gHHHHHHH X为主版 ...

  9. 如何在Windows上从源码编译Chromium (CEF3) 加入mp3支持

    一.什么是CEF CEF即Chromium Embeded Framework,由谷歌的开源浏览器项目Chromium扩展而来,可方便地嵌入其它程序中以得到浏览器功能. CEF包括CEF1和CEF3两 ...

随机推荐

  1. 关于C#的Lock锁思考

    大家都知道多线程并发时候存在一个线程同步的问题,一般使用lock关键字来处理. lock关键字的结果如下: object locker=new object(); lock(locker) { ... ...

  2. Fragment中的方法findFragmentById(int id)的返回值探讨

    在学习<Android编程权威指南>P124页的时候,遇到了这样的代码: 引起了我的疑问if的判断条件是(fragment==null),那执行完上一句 Fragment Fragment ...

  3. python 之闭包

    原文 函数作为返回值 高阶函数除了可以接受函数作为参数外,还可以把函数作为结果值返回. 我们来实现一个可变参数的求和.通常情况下,求和的函数是这样定义的: def calc_sum(*args): a ...

  4. ThreadPoolExecutor实现原理

    转载:https://blog.csdn.net/yanyan19880509/article/details/52718497 前言 做java开发的,一般都避免不了要面对java线程池技术,像to ...

  5. Angular6 基础(数据绑定、生命周期、父子组件通讯、响应式编程)

    Angular相比于vue来说,更像一个完整的框架,本身就集成了很多模块,如路由,HTTP,服务等,而vue是需要另外引入比如(vuex,axios等).Angular引入了依赖注入.单元测试.类等后 ...

  6. mysql基本面试题

    1.MySQL的复制原理以及流程 基本原理流程,3个线程以及之间的关联: 1. 主:binlog线程——记录下所有改变了数据库数据的语句,放进master上的binlog中: 2. 从:io线程——在 ...

  7. 【阿里云产品公测】PTS压力测试服务器性能

    作者:阿里云用户xsnjxjj 在PTS服务之前,经常使用webbench来对服务器进行压力测试,在看到阿里云PTS服务的介绍以后,深深的被PTS强大的功能所吸引     非常感谢阿里云团队给予的测试 ...

  8. Resharper安装及激活--转载

    原文地址:ReSharper2018破解详细方法   1.先安装好Resharper: 2.下载完补丁后解压,以管理员身份运行Patch.cmd,如下图所示,即破解成功: 3.打开VS,打开ReSha ...

  9. 开通cnblogs博客

    开通博客,准备记录学习和开发过程

  10. collectd配置

    udp proxy - 192.168.48.112 cat > /etc/collectd_25801.conf << EOF Hostname "kvm-48-112& ...