Messages in this thread | | | Date | Mon, 16 Aug 1999 10:27:07 -0700 (PDT) | From | Linus Torvalds <> | Subject | Re: [patch] possible SMP races all over the place in wait_event interface |
| |
On Mon, 16 Aug 1999, Andrea Arcangeli wrote: > > Maybe I am dreaming, but it seems to me that the wait_even interface has > SMP races. To make it SMP safe we should enforce read and write ordering > in the wait_event code.
The spinlocks and atomic operations implied by "wake_up()" and company make this a non-issue: a spinlock has to guarantee write ordering anyway, otherwise it wouldn't be an effective spinlock..
So about half of your patch is unnecessary, but the latter half looks valid.
Linus
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |