9fans - fans of the OS Plan 9 from Bell Labs
 help / color / mirror / Atom feed
From: erik quanstrom <quanstro@quanstro.net>
To: 9fans@9fans.net
Subject: Re: [9fans] very high interrupt rates
Date: Thu, 25 Aug 2011 09:27:08 -0400	[thread overview]
Message-ID: <cde03641e3bc42a521ebcf7d368d3e90@ladd.quanstro.net> (raw)
In-Reply-To: <742f85caa9c3458f33cfb43a9c9cb233@quintile.net>

> My new desktop machine reports very high interrupt rates in stats,
> the bargraph on stats is stock at 100%, a count of 176000 (ints/sec ?)
>
> I added erik's interrupt counter code and see that these are occuring
> on irq 10, which my hardware ties to the SATA interface and usbuhci
> (a combined Intel 82801EB/ER chipset).
>
> the other difference between this system and my previous one is
> I now have a USB keyboard and mouse (rather than PS2).
>
> Is there a problem with my USB or is it a case of "they all do that"?
>
> -Steve
>
> -----------------
> larch% grep '10|clock' /dev/irqalloc
>          50          18           1196593585 clock
>         105          10         167041275929 sdF (ata)
>         105          10         167041275930 sdE (ata)
>         105          10         167041275932 usbuhci

if you're using the standard HZ of 100, then your machine
has been up ~138/nmach days?  if that's correct, then nevermind.
otherwise, is it possible that you are counting cycles and not
interrupts?

in any event, if you use a 9atom kernel, you may be able to
get the ata drives on a different vectors (assuming they support msi)
and then you could tell more precicely what's going on.
sadly usb is incompatable with msi; broken by design.

as a wild guess, i think you could fix your problem either by
using the 9atom ide driver or telling the motherboard to
present the drives as ahci.

- erik

p.s  the format i've been using for irqalloc on 386 has been

          Reads from irqalloc return the enabled interrupts, one line
          per interrupt.  Each line contains three fields separated by
          white space: the trap [sic] number, the IRQ it is assigned to, and
          the name of the device using it.

minooka; cat /dev/irqalloc
          3           0                    0                    0 trap     debugpt
          7           0                    0                    0 trap     mathemu
          8           0                    0                    0 trap     doublefault
          9           0                    0                    0 trap     mathover
         14           0                    0                    0 trap     fault386
         15           0                    0                    0 trap     unexpected
         16           0                    0                    0 trap     matherror
         50          18           9052758077       20828942019320 lapic    clock
         51          19                    0                    0 lapic    lapicerror
         63          31                    0                    0 lapic    lapicspurious
         65           1                    6              4231380 ioapic   kbd
         73          15             76377411         457040194416 msi      ether0
         81          15               520404           4488965428 msi      ether1
         89           5                    0                    0 msi      ether2
         97           4                  408              7531904 ioapic   COM1
        105          11                    3                51732 msi      sdE (ahci)

see http://www.quanstro.net/magic/man2html/3/arch



      reply	other threads:[~2011-08-25 13:27 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-08-25  9:59 Steve Simon
2011-08-25 13:27 ` erik quanstrom [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=cde03641e3bc42a521ebcf7d368d3e90@ladd.quanstro.net \
    --to=quanstro@quanstro.net \
    --cc=9fans@9fans.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).