多线程调用GlobalData::RegisterService会出现coredump，asan能查出heap-use-after-free

[neuskb]

Describe the bug 多线程调用GlobalData::RegisterService，且service填同一个，会出现coredump，必现。如下方式：

  std::vector<std::thread> threads;
  for (int i = 0; i < 20; ++i) {
    threads.emplace_back([]() {
      while(1) {
        apollo::cyber::common::GlobalData::RegisterService("testminisim/test_register_msg");
        usleep(100);
      }
    });
  }

To Reproduce Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior A clear and concise description of what you expected to happen.

Screenshots If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

Smartphone (please complete the following information):

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context Add any other context about the problem here.

[neuskb]

知道原因了。GlobalData::RegisterService写的就有问题

[neuskb]

多线程调用RegisterService的时候，如果service填的还是同一个，出现踩内存的场景，虽然service_id_map_是线程安全的，但是只是Get和Set接口线程安全，在通过Get拿到name后，这个name填的是二级指针，所以name本身就是service_id_map_的value，另一个线程可能会再去修改这个value，造成Get出来的name 会有多线程同时读写的风险，用我上面贴的代码去复现，基本必现。

[daohu527]

Will check and feedback soon. I tested it and it will indeed report an error.

[neuskb]

Will check and feedback soon. I tested it and it will indeed report an error.

修改方案：

uint64_t GlobalData::RegisterService(const std::string& service) {
  auto id = Hash(service);
  std::lock_guard<std::mutex> lock(service_id_map_mutex_);
  while (service_id_map_.Has(id)) {
    std::string* name = nullptr;
    service_id_map_.Get(id, &name);
    if (service == *name) {
      break;
    }
    ++id;
    AWARN << "Service name hash collision: " << service << " <=> " << *name;
  }
  service_id_map_.Set(id, service);
  return id;
}