Messages in this thread | | | Date | Thu, 6 May 1999 23:56:51 +0200 (CEST) | From | Andrea Arcangeli <> | Subject | Re: [patch] SMP race fix [was Re: SMP lockup & 3c509 on 2.2.x [aka.the Deadly 'ping -f']] |
| |
On Thu, 6 May 1999, M.H.VanLeeuwen wrote:
>I've been running your changes on 2.2.6 for the last hour or so >w/o any lockups.
If I remeber well you was able to reproduce trivially the lockup in some minute of ping -f, right?
> 10: 4848399 4817093 IO-APIC-edge eth0 ^^^^^^^^^^^^ fine ;)
>here is /var/log/kernel... one interesting item i've not seen before is >the Infinite loop message, so far it's only happened once and eth0 continued >to function normally.
The message are 3c509 related. nothing related to my fixes. The timeout is because the incoming queue was full.
>May 6 12:58:23 linux1a kernel: eth0: Infinite loop in interrupt, status 2091.
I guess (and yes, this is only a guess, I don't have signed the NDA for the 3c509 specs ;) that it can be that your machine was fast enough to handling more work than max_interrupt_work. Everything 3c509 related and I seen such message at once here too on UP some time ago if I remeber well. But it has nothing to do with irq tha reenter itself, it's just that after some time you are doing work in the irq, you then stop receiving data to not stall too much in the irq.
So it seems that the race I spotted and fixed is the bug that was harming you.
Andrea Arcangeli
- 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/
| |