Bug 244024

Summary: 2.6.9-55 breaks USB on W2100z so unusable
Product: Red Hat Enterprise Linux 4 Reporter: Andrew Gormanly <a.gormanly>
Component: kernelAssignee: Pete Zaitcev <zaitcev>
Status: CLOSED WORKSFORME QA Contact: Martin Jenner <mjenner>
Severity: medium Docs Contact:
Priority: low    
Version: 4.5CC: jbaron
Target Milestone: ---Keywords: Regression
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2007-09-19 00:20:18 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Andrew Gormanly 2007-06-13 13:06:25 UTC 
Description of problem:
The latest kernel, 2.6.9-55, breaks USB on Sun Java Workstation W2100z's.

Version-Release number of selected component (if applicable):
kernel-smp-2.6.9-55.EL.x86_64

How reproducible:
Every time

Steps to Reproduce:
1. Boot
2.
3.
  
Actual results:
Keyboard and mouse don't work (though kudzu does not detect the USB controllers
as missing).  The machine is therefore unusable.

Expected results:
Keyboard and mouse work, users are able to also.

Additional info:
I suspect this is due to a conflict with having 2 different USB hubs in the
machines:

[root@poisson ~]# lspci | grep USB
01:00.0 USB Controller: Advanced Micro Devices [AMD] AMD-8111 USB (rev 0b)
01:00.1 USB Controller: Advanced Micro Devices [AMD] AMD-8111 USB (rev 0b)
01:03.0 USB Controller: NEC Corporation USB (rev 43)
01:03.1 USB Controller: NEC Corporation USB (rev 43)
01:03.2 USB Controller: NEC Corporation USB 2.0 (rev 04)

It is possible to turn off the USB 2.0 controller in the BIOS setup (and then
the fifth output line above disappears), but not to turn off either all NEC or
all AMD controllers.

However, kernels up to 2.6.9-42.0.3.EL work fine, so the bug has been introduced
recently.  I did not try 2.6.9-42.0.10.EL yet, but can do so if it would help.
Comment 1 Jason Baron 2007-06-29 15:55:49 UTC 
Can you be more specific and tell us the exact version of the last working kernel?
Comment 2 Andrew Gormanly 2007-06-29 17:34:58 UTC 
Sure, but it will have to be Monday now as the machines have jobs running and
it's the weekend here.  I'll post when I have the results.
Comment 3 Andrew Gormanly 2007-07-03 13:59:58 UTC 
Well, 2.6.9-42.0.10.EL works - but so does 2.6.9-55.0.2.EL so consider this fixed.
Comment 4 Pete Zaitcev 2007-09-19 00:20:03 UTC 
Hmm, very odd. I wonder if it has something to do with with HCD load
order again.