Re: [PATCH v2 0/2] MR882: Use atomic write for releasing threadsafe_queue spin lock.

[Piotr Caban (＠piotr)]

Piotr Caban (@piotr) commented about dlls/msvcp90/details.c:
>      queue->tail = p;
>      if(!queue->head)
>          queue->head = p;
> -
>    queue->lock = 0;
> -
>    WriteRelease(&queue->lock, 0);

Isn't it better to make the `lock` member volatile?