Bug 7715 - uhci_hcd: host controller process error
Summary: uhci_hcd: host controller process error
Status: REJECTED INVALID
Alias: None
Product: Drivers
Classification: Unclassified
Component: USB (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: Alan Stern
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-12-19 15:24 UTC by Christoph Auer
Modified: 2006-12-30 09:04 UTC (History)
1 user (show)

See Also:
Kernel Version: 2.6.19.1
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
contains dmesg/syslog output for 2.6.17.14, 2.6.18.6 and 2.6.19.1 (15.79 KB, application/gzip)
2006-12-19 15:30 UTC, Christoph Auer
Details
bluetooth module (hci_usb) not loaded (14.97 KB, text/plain)
2006-12-20 09:48 UTC, Christoph Auer
Details

Description Christoph Auer 2006-12-19 15:24:48 UTC
Most recent kernel where this bug did *NOT* occur:
Distribution: Debian Etch
Hardware Environment: Intel Pentium M processor 1.60GHz  (HP Notebook nc6120)
Software Environment: 
Problem Description:
I'm getting errors like this:

uhci_hcd 0000:00:1d.1: host controller process error, something bad happened!
uhci_hcd 0000:00:1d.1: host controller halted, very bad!

Sometimes already while booting up, sometimes after some minutes... I was not
able to find out what causes this.
As I read in an similar bugreport (bug 6809) I recompiled the kernel with 
CONFIG_USB_DEBUG turned on and loaded the uhci-hcd module with debug=2
I will attach files with dmesg and syslog output(at the moment I don't see how).

To see if it is only a 2.6.19 problem I compiled also the latest 2.6.17 and 2.6.18.
Same problem on 2.6.17.14 and 2.6.18.6(happens not so frequently). 
Is this maybe a hardware issue or should I try also 2.6.16.35?

After getting this error messages (Controller halted) I'm not able to switch the
WLAN(ipw2200) on (kill switch). 

Steps to reproduce:
Unable to see conditions with causes halting of controller.
Comment 1 Christoph Auer 2006-12-19 15:30:05 UTC
Created attachment 9890 [details]
contains dmesg/syslog output for 2.6.17.14, 2.6.18.6 and 2.6.19.1
Comment 2 Andrew Morton 2006-12-19 16:05:01 UTC
On Tue, 19 Dec 2006 15:29:24 -0800
bugme-daemon@bugzilla.kernel.org wrote:

> http://bugzilla.kernel.org/show_bug.cgi?id=7715
> 
>            Summary: uhci_hcd: host controller process error
>     Kernel Version: 2.6.19.1
>             Status: NEW
>           Severity: normal
>              Owner: greg@kroah.com
>          Submitter: c_a@gmx.net
> 
> 
> Most recent kernel where this bug did *NOT* occur:
> Distribution: Debian Etch
> Hardware Environment: Intel Pentium M processor 1.60GHz  (HP Notebook nc6120)
> Software Environment: 
> Problem Description:
> I'm getting errors like this:
> 
> uhci_hcd 0000:00:1d.1: host controller process error, something bad happened!
> uhci_hcd 0000:00:1d.1: host controller halted, very bad!
> 
> Sometimes already while booting up, sometimes after some minutes... I was not
> able to find out what causes this.
> As I read in an similar bugreport (bug 6809) I recompiled the kernel with 
> CONFIG_USB_DEBUG turned on and loaded the uhci-hcd module with debug=2
> I will attach files with dmesg and syslog output(at the moment I don't see how).
> 
> To see if it is only a 2.6.19 problem I compiled also the latest 2.6.17 and 2.6.18.
> Same problem on 2.6.17.14 and 2.6.18.6(happens not so frequently). 
> Is this maybe a hardware issue or should I try also 2.6.16.35?
> 
> After getting this error messages (Controller halted) I'm not able to switch the
> WLAN(ipw2200) on (kill switch). 
> 
> Steps to reproduce:
> Unable to see conditions with causes halting of controller.
> 

Comment 3 Alan Stern 2006-12-20 07:33:25 UTC
I don't see any indication in your logs of a software problem.  In fact, the
dmesg log from 2.6.17.14 shows that the controller stopped spontaneously at a
time when it had nothing to do!

Considering the other messages about errors communicating with the Bluetooth
module, this certainly looks like a hardware failure.  Can you plug the
Bluetooth device into a different USB port?
Comment 4 Christoph Auer 2006-12-20 07:56:50 UTC
The Bluetooth and WLAN devices are integrated in the notebook, so I'm unable to
do that. All I can do is to switch it off - in the BIOS or via the kill switch
on which the wifi is also connected.
Comment 5 Christoph Auer 2006-12-20 08:41:25 UTC
OK, if I deactivate the integrated Bluetooth module I don't get any errors any
more, so the bluetooth hardware is broken?

btw: the wifi led now isn't working any more if I deaktivate the bluetooth but
this isn't really a problem ;)

Wifi and kill switch still working...

Sorry for the noise!
Comment 6 Alan Stern 2006-12-20 08:49:41 UTC
In theory, a broken Bluetooth module should not be able to cause the errors you
saw.  Only a broken UHCI controller (or a software bug in the driver) could.

If you're happy with the current state of affairs, you can mark this bug report
REJECTED.  If you want to experiment a little more, try this: Rename the
Bluetooth driver file so that it can't be loaded automatically, turn the module
back on in the BIOS, and see what happens.  With no driver available, the system
shouldn't use the Bluetooth hardware -- so any remaining problems will be caused
by the USB hardware or driver.
Comment 7 Christoph Auer 2006-12-20 09:46:23 UTC
Thanks for giving me this option.

I made the following things:
1. reactivated the bluetooth in the BIOS
2. /lib/modules/2.6.19.1rev4/kernel/drivers# mv bluetooth/ bluetoOoth
3. rebooted the machine
Now the module for my bluetooth device (hci_usb) isn't loaded any more.

I am still getting errors... see dmesg output (2.6.19.1_dmesg3).
Comment 8 Christoph Auer 2006-12-20 09:48:10 UTC
Created attachment 9899 [details]
bluetooth module (hci_usb) not loaded
Comment 9 Alan Stern 2006-12-20 20:06:44 UTC
It seems clear that the problem lies in the USB controller hardware.  It wasn't
doing anything, the driver wasn't doing anything -- and the controller just
decided to stop working.
Comment 10 Christoph Auer 2006-12-29 11:32:57 UTC
The problem is getting heavier. Sometimes when booting up I get a Kernel Oops
with call trace output. The only thing I can do after this is to press the
powerbutton... 
I'm asking me if I should remove the hole
/lib/modules/`kernelversion`/kernel/drivers/usb/ directory to resolve this
problem? It would be bad if my ipw2200 is internaly connected to an usb controller..

As it is quite shure now that this bugreport is based on a hardware problem you
can mark this bug as closed (if this is the usual procedure).
Comment 11 Alan Stern 2006-12-30 09:04:09 UTC
Okay, I'll close the bug report.

You ought to be able to turn off the USB and Bluetooth controllers in the BIOS.
 Then there won't be any need to remove driver files from the module directory.

Note You need to log in before you can comment on or make changes to this bug.