Re: Bulldog UPS

From: Paul J Gans (gans_at_panix.com)
Date: 06/12/05 

Date: Sat, 11 Jun 2005 22:16:18 +0000 (UTC)

Harold Stevens <wookie@deuce.localdomain> wrote:
>In <d8die9$17d$1@reader1.panix.com> Paul J Gans:

>[Snip...]

>> Does linux recognize the USB plug-in at all?

>Oh yeah; it's fine until a powerline glitch hits it. This appears to reset
>the USB device by shutting down the current device, and restarting it on a
>new device. That works fine for a time; however, at some point, it appears
>to get confused by all the "new" device numbers, and goes nuts looking for
>an "available" number. The time/number to "frenzy" is very erratic, but it
>always ends up eventually running amok in /var/log/messages:

> Jan 26 04:15:24 jacks kernel: drivers/usb/input/hid-core.c: ctrl urb status
> -71 received
> Jan 26 04:15:24 jacks last message repeated 87 times
> Jan 26 04:15:24 jacks kernel: drivers/usb/input/hid-core.c: input irq status
> -84 received
> Jan 26 04:15:24 jacks kernel: drivers/usb/input/hid-core.c: ctrl urb status
> -71 received
> Jan 26 04:15:25 jacks last message repeated 79 times
> Jan 26 04:15:25 jacks kernel: drivers/usb/input/hid-core.c: input irq status
> -84 received
> Jan 26 04:15:25 jacks kernel: drivers/usb/input/hid-core.c: ctrl urb status
> -71 received
> Jan 26 04:15:25 jacks last message repeated 79 times
> Jan 26 04:15:25 jacks kernel: drivers/usb/input/hid-core.c: input irq status
> -84 received
> Jan 26 04:15:25 jacks kernel: drivers/usb/input/hid-core.c: ctrl urb status
> -71 received
> Jan 26 04:15:25 jacks last message repeated 87 times

>etc., ad nauseum.

>Of course CPU load skyrockets (apparently hogging syslogd or whatnot).

>Might not be a kernel problem--Belkin doesn't release source (AFAIK), so I
>can't begin to sort it out, even if I had time and wanted to. I just use a
>Belkin UPS with older tty interface instead, and get on with my life. :)

That's ugly. Since the computer is not experiencing any
power glitches it must be the UPS that is doing it, as you
say. That's evil behavior... Really evil. The UPS
circuitry that deals with the USB port should not be
feeling any glitches either unless they stupidly have
its power supply on the input side of the UPS...

I'd send Belkin a nastygram. Probably do no good but an
engineering type might see it and get it fixed for 2009
or so.

    --- Paul J. Gans

Relevant Pages

  • Re: Nut,usb and installation
    ... On Sun, 25 Sep 2005, david walcroft wrote: ... >> always be loaded as the same usb device depending on other usb devices ... >> You will probably have to write a rule for your ups. ... >> Tom Taylor ...
    (Fedora)
  • Re: Finding device filenames for USB devices
    ... >> Suppose I plug a USB device such as a camera or UPS or keyboard into a ... >> USB port. ... figure the device filename from it? ... S: Product=Belkin UPS ...
    (comp.os.linux.hardware)
  • Re: Want to use APC BackUPS XS 1000 with FreeBSD
    ... >If I hook up the UPS to my FreeBSD system using the USB cable that came ... This is a USB device only. ...
    (comp.unix.bsd.freebsd.misc)
  • Re: USB on Solaris 9 x86
    ... >> Depends on the USB device. ... then there's no Solaris driver that controls ... This whitepaper even uses an USB UPS as a sample and includes some ...
    (comp.unix.solaris)