On Mon, 6 Nov 2006, Jason Pool wrote:
> I'm having problems with my IOgear USB KVM switch. I am running on a
> Athlon XP with an Abit NF7 nForce2 motherboard. Using Kubuntu 6.10 with
> stock kernel 2.6.18.1 and Ubuntu kernel 2.6.17-10-generic with the KVM
> plugged in, I get this in /var/log/messages every few seconds:
>
> Oct 29 02:53:28 zanzibar kernel: [ 105.489309] usb 1-1.1: reset low
> speed USB device using ohci_hcd and address 7
>
> This causes my USB keyboard to temporarily become unresponsive, and also
> makes the last key typed when being reset stick leading to "tyinnnnng
> lkeee thssssssss". The problem goes away with the keyboard plugged
> directly into the machine without the KVM, and my USB mouse seems
> unaffected either way.
>
> I tried installing Fedora Core 6, and the installer and installation
> gave me the same trouble, as well as the gparted 0.3.1 boot cd. This
> has only been a problem recently, Ubuntu 6.06 gave me no problems (a
> version of 2.6.15), so I reverted to a Debian 2.6.16 kernel to see how
> that went, and it works fine.
>
> Some change in 2.6.17 (or possibly later in 2.6.16) seems to have
> started this problem for me. I've seen a thread last month regarding a
> similar problem with another KVM, but I wasn't able to find any
> resolution. If there's a patch to regress back to the old behavior
> (which seems to be simply ignoring the warnings that it now acts on), it
> would be appreciated, or any help where I should modify the source to
> change this.
>
> Any help would be appreciated. I'd be more than happy to provide any
> additional information if needed.
Below is a patch for 2.6.18 which causes these errors to be ignored.
Unfortunately the standard kernel can't afford to ignore the errors; some
devices really do need to be reset when these sorts of errors occur.
In any case it's clear that your KVM is the source of the problem. From
time to time it fails to send responses from the keyboard back to the
computer, causing the computer to think the keyboard is malfunctioning and
needs to be reset.
Alan Stern
Index: 2.6.18/drivers/usb/input/hid-core.c
===================================================================
--- 2.6.18.orig/drivers/usb/input/hid-core.c
+++ 2.6.18/drivers/usb/input/hid-core.c
@@ -1037,9 +1037,7 @@ static void hid_irq_in(struct urb *urb,
case -EILSEQ: /* protocol error or unplug */
case -EPROTO: /* protocol error or unplug */
case -ETIMEDOUT: /* NAK */
- clear_bit(HID_IN_RUNNING, &hid->iofl);
- hid_io_error(hid);
- return;
+ break;
default: /* error */
warn("input irq status %d received", urb->status);
}
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Linux-usb-users@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users
