Messages in this thread | | | Date | Fri, 24 Dec 2004 15:41:37 -0800 (PST) | From | Linus Torvalds <> | Subject | Re: VM fixes [4/4] |
| |
On Fri, 24 Dec 2004, Andrea Arcangeli wrote: > > If those old cpus really supported smp in linux, then fixing this bit is > trivial, just change it to short. Do they support short at least?
It's not even about SMP. "byte" and "short" are not IRQ-safe or even preemption-safe (although I guess alpha doesn't support CONFIG_PREEMPT right now anyway) on pre-byte-access alphas.
Just don't do it. Maybe we'll never see another chip try what alpha did (it was arguably the single biggest mistake the early alphas had, and caused tons of system design trouble), but just use an "int".
That said, I'd suggest putting it in the thread structure instead. We already have thread-safe flags there, just use one of the bits. Yes, you'll need to use locked accesses to set it, but hey, how often does something like this get set anyway? And then you just do ti _right_, using set_thread_flag/clear_thread_flag etc..
Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |