具有函数作用域静态变量初始化的 threadprivate

Question

我正在测试 threadprivate 子句在 openMP 中的工作方式。我设置了以下简单的测试用例

#include <iostream>
#include <omp.h>

void func(){
   static int some_id = omp_get_thread_num();    
#pragma omp threadprivate(some_id)
#pragma omp critical
  {
      std::cout << "id:" << some_id << std::endl;
      std::cout << "in thread " << omp_get_thread_num() << std::endl;
  }   
}

int main() {  
omp_set_num_threads(4);
#pragma omp parallel
    func();
}

我用 gcc7.2 和运行编译了这段代码，我得到了

id:1
in thread 1
id:0
in thread 0
id:0
in thread 3
id:0
in thread 2

我知道 C++ 标准（至少从 C++03 标准开始）保证静态变量由一个且只有一个线程初始化，所有其他线程都被阻塞，直到初始化完成。话虽如此，我希望所有线程中 id 的本地副本都具有相同的值（负责初始化的线程的 ID）。为什么线程#1 中的 id 与其他线程中的不同？显然我的 threadprivate 心智模型是有缺陷的。你能解释一下这是怎么回事吗？

Answer 1

我认为这是实现中的错误。 OpenMP API 规范要求 some_id 变量由每个线程初始化：

Each copy of a block-scope threadprivate variable that has a dynamic initializer is initialized the first time its thread encounters its definition; if its thread does not encounter its definition, its initialization is unspecified.

（见https://www.openmp.org/spec-html/5.1/openmpsu114.html#x149-1630002.21.2）

我机器上的 GCC 9.3.0 具有相同的行为：

$ g++ --version
g++ (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
Copyright (C) 2019 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

$ g++ -O0 -g -fopenmp -o tp tp.cc
$ ./tp
id:0
in thread 0
id:0
in thread 3
id:0
in thread 1
id:0
in thread 2

虽然 clang 12 做了正确的事情：

$ clang --version
clang version 12.0.0 (/net/home/micha/projects/llvm/llvm-project/clang d28af7c654d8db0b68c175db5ce212d74fb5e9bc)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir: /net/software/x86_64/clang/12.0.0/bin
$ clang++ -g -O0 -fopenmp -o tp ./tp.cc
$ ./tp
id:2
in thread 2
id:0
in thread 0
id:3
in thread 3
id:1
in thread 1

具有函数作用域静态变量初始化的 threadprivate

threadprivate with initialization of function-scope static variables

c++

multithreading

openmp