Messages in this thread |  | | | Date | Fri, 10 Dec 1999 16:37:51 +0100 (CET) | | From | Andrea Arcangeli <> | | Subject | Re: timer_bh behaviour incorrect for 2.2.13? |
| |
On Fri, 10 Dec 1999, Ingo Molnar wrote:
>what i ment was that i think that this is physically impossible. Since >local interrupts are disabled, even with the worst cache-miss and TLB-miss >scenario possible, it's impossible for the above big chunk of code to >execute just inbetween those two others.
What about your NMI in 2.3.x?
I agree that looks physically impossible in 2.2.x IA32 but it doesn't look impossible in 2.3.x IA32 without disabling the NMI watchdog.
More in general I don't write SMP code assuming that both cpus runs at the same speed. I assume that every CPU can stop for some time at any time for any reason. I consdier depending on the timings as a race. The common code doesn't run only IA32. Of course I agree that the single __cli is fine for testing in 2.2.13.
Andrea
- 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/
|  |