在回发中选择保持单选按钮(keep radio button selected across postback)

 问题似乎是调用了managed_shared_memory的析构函数，但是在try catch块之后你继续依赖它所持有的状态或分配。 
 
managed_shared_memory segment;
 
 回想一下，析构函数总是在作用域的末尾调用（基本上是任何结束} ），这也适用于try块。 
 
 如果这不是``main`函数，那么让异常转义给调用者（没有try catch）或者重新抛出它（或者另一个异常）可能会更好： 
 
    try {
        managed_shared_memory segment(open_only, "MySharedMemory");
        res=segment.find<MyType>("MyType instance");
        cout<<"Child: Segment of length "<<res.second<<" is found at "<<res.first<<endl;
        cout<<"Child: "<<res.first->first<<", "<<res.first->second<<endl;

    } catch (interprocess_exception &e) {
        cerr<<"Error while opening the segment"<<endl;
        throw;
    }
 
 确实，看看文档 
 
 
 
 托管内存段最重要的服务是： 
 
 
 
  
 动态分配内存部分的段。 
 
  
 在内存段中构造C ++对象。 这些对象可以是匿名的，也可以将名称与它们关联起来。 
 
  
 搜索命名对象的功能。 
 
  
 自定义许多功能：内存分配算法，索引类型或字符类型。 
 
  
 原子构造和析构，以便如果在两个进程之间共享段，则无法创建两个对象 
 与同名相关联，简化了同步。 
 
 
 

The problem seems to be that the destructor of managed_shared_memory is called, but you continue to rely on state or allocation held by it after the try catch block.
 
managed_shared_memory segment;
 
Recall that destructors are always called at the end of a scope (basically any closing }), and this goes for try blocks as well.
 
If this was not the ``main` function, it would likely be better to let the exception escape to the caller (no try catch), or re-throw it (or another exception):
 
    try {
        managed_shared_memory segment(open_only, "MySharedMemory");
        res=segment.find<MyType>("MyType instance");
        cout<<"Child: Segment of length "<<res.second<<" is found at "<<res.first<<endl;
        cout<<"Child: "<<res.first->first<<", "<<res.first->second<<endl;

    } catch (interprocess_exception &e) {
        cerr<<"Error while opening the segment"<<endl;
        throw;
    }
 
Indeed, have a look at the documentation
 
 
 
The most important services of a managed memory segment are:
 
 
 
  
Dynamic allocation of portions of a memory the segment.
 
  
Construction of C++ objects in the memory segment. These objects can be anonymous or we can associate a name to them.
 
  
Searching capabilities for named objects.
 
  
Customization of many features: memory allocation algorithm, index types or character types.
 
  
Atomic constructions and destructions so that if the segment is shared between two processes it's impossible to create two objects
 associated with the same name, simplifying synchronization.