英文:
Mutex using 2 variables
问题
我正在开发一个由两个独立设备组成的系统，它们通过“共享内存”交换数据。
为了同步它们，我想使用一个内存位置作为“状态变量”或互斥锁。其中，第一个设备会获取互斥锁，而第二个设备会等待它被释放，反之亦然。但现在我发现其中一个设备只能从相同位置读取或写入，不能同时进行两者操作。
所以现在我正试图想出一种使用两个变量来实现互斥锁的方法。一个用于第一个设备写入，第二个设备读取，另一个用于第一个设备读取，第二个设备写入。
这种方法是否不安全？因为我认为这可能会产生大量的读/写循环，可能会引发竞争条件。我一直看到互斥锁只使用一个变量来实现，所以对此的任何指导都将非常有帮助。
英文:
I am working on a system that consists of 2 Independent devices that exchange data through a "common memory".
To sync them, I wanted to use a memory location as a "state variable" or a mutex. Where the first device would take the mutex and the second would wait for it to be released and vice-versa. But now I have discovered that one of those devices can either read OR write from the same location. It cannot do both.
So now I am trying to think of a way to implement a mutex using 2 variables. One where the first device will write and second would read, and the other where the first would read and second would write.
Is this an unsafe approach? Because I think this would create a lot of read/write cycles and possibly race conditions. I've always seen mutexes being implemented with only one variable, so any guidance on this would be very very helpful.
答案1
得分: 1
你正在研发CSMA中的碰撞检测算法，这可能不是你想做的事情，但也许是成功的唯一选择。是否有特定原因你不创建每个设备可写入的内存位置，而另一个设备从该位置读取？这更像是发布-订阅模式，而不是真正的共享内存模型。
英文:
You are inventing the collision detection algorithim in CSMA, probably not what you want to do, but may be your only choice to be successful. Is there a reason you aren't creating a memory location for each device to write to, and have the other device read from that location? More of a pub-sub than a truly shared memory model.
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论