2001-10-18 02:10:52

by Jens Gecius

Subject: /proc/interrupts on 2.4.13-1

Hi folks!

/proc/interrupts shows

CPU0 CPU1
0: 6700068 6617178 IO-APIC-edge timer
1: 15716 15044 IO-APIC-edge keyboard
2: 0 0 XT-PIC cascade
3: 0 1 IO-APIC-edge serial
8: 1 1 IO-APIC-edge rtc
9: 788995 790370 IO-APIC-level eth0
10: 93412 93466 IO-APIC-level aic7xxx, Ensoniq AudioPCI
11: 3476219 3482580 IO-APIC-level nvidia
12: 31650 31277 IO-APIC-level usb-uhci, usb-uhci
14: 505063 501971 IO-APIC-edge ide0
15: 10 0 IO-APIC-edge ide1
NMI: 0 0
LOC: 13315157 13315147
ERR: 0
MIS: 3690

Especially the last makes me curious: does MIS mean "missed"? And:
might this be related to esd running here having some trouble to keep
playing without short little "breaks"? The count on MIS is actually
growing steadily, the uptime is currently only 37 hours.

The box is used most time with two niced seti-processes. Other than
that, cpu stat is hardly at 10%.

--
Tschoe, http://gecius.de/gpg-key.txt - Fingerprint:
Jens 1AAB 67A2 1068 77CA 6B0A 41A4 18D4 A89B 28D0 F097

2001-10-18 04:06:10

by Mark Hahn

Subject: Re: /proc/interrupts on 2.4.13-1

> MIS: 3690
>
> Especially the last makes me curious: does MIS mean "missed"? And:

no. see the excellent comment in arch/i386/kernel/io_apic.c near irq_mis_count

2001-10-18 09:56:33

by Jens Gecius

Subject: Re: /proc/interrupts on 2.4.13-1

Mark Hahn <[email protected]> writes:

> > MIS: 3690
> >
> > Especially the last makes me curious: does MIS mean "missed"? And:
>
> no. see the excellent comment in arch/i386/kernel/io_apic.c near irq_mis_count

That IS an excellent comment. Thanks for pointing it out.

--
Tschoe, http://gecius.de/gpg-key.txt - Fingerprint:
Jens 1AAB 67A2 1068 77CA 6B0A 41A4 18D4 A89B 28D0 F097