Bug 316801

Summary: khubd consumes 100% CPU and fills system log
Product: [Fedora] Fedora Reporter: Richard Körber <redhat>
Component: kernelAssignee: Kernel Maintainer List <kernel-maint>
Status: CLOSED INSUFFICIENT_DATA QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: low Docs Contact:
Priority: low    
Version: 7CC: chris.brown
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2008-02-16 02:44:16 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 Richard Körber 2007-10-03 13:25:21 UTC
Description of problem:
khubd suddenly starts to consume 100% CPU time, and /var/log/messages is floated
with error messages.

It starts with this message:

Oct  3 14:29:34 monique kernel: hub 5-8:1.0: activate --> -110

Which is continued by this message:

Oct  3 14:29:35 monique kernel: hub 5-8:1.0: activate --> -108
Oct  3 14:29:35 monique kernel: hub 5-8:1.0: hub_port_status failed (err = -108)

The last two lines are repeated some HUNDRED times per SECOND. After 20 minutes,
the /var/log/message grew 60 MByte large.

Killing khubd showed no effect (even signal 9 did not impress the process).

Shutting down the system seems to be impossible, too. Those "hub_port_status"
messages are shown on the text terminal. After a few minutes waiting for the
system to shut down, I finally pressed the reset button.

Version-Release number of selected component (if applicable):
kernel-2.6.22.9-91.fc7

Note: This is the first 2.6.22 kernel I could test due to bug 249652. This bug
did not occur with the 2.6.21-1.3228 kernel.

How reproducible:
Sometimes. The system may work for hours or even days before showing this effect.

Steps to Reproduce:
None found yet. Just wait for it to happpen.

Actual results:
khubd consumes 100% CPU. /var/log/messages is floated with error messages.
Reboot/hardware reset is required.

Expected results:
Everything works normally.

Additional info:

lsusb:
Bus 005 Device 005: ID 050d:0234 Belkin Components F5U234 USB 2.0 4-Port Hub
Bus 005 Device 001: ID 0000:0000  
Bus 002 Device 002: ID 046d:c311 Logitech, Inc. (Internet Pro Keyboard)
Bus 002 Device 001: ID 0000:0000  
Bus 004 Device 004: ID 04a9:220d Canon, Inc. CanoScan N670U/N676U/LiDE 20
Bus 004 Device 001: ID 0000:0000  
Bus 001 Device 003: ID 046d:c50e Logitech, Inc. MX-1000 Cordless Mouse Receiver
Bus 001 Device 001: ID 0000:0000  
Bus 003 Device 001: ID 0000:0000  

I did not connect or disconnect any USB devices when the bug started.

Also see bug 246713, which may be related to this bug. I have booted kernel
2.6.22.9-91 with the usbcore.autosuspend=0 option.

Comment 1 Chuck Ebbert 2007-10-03 23:45:32 UTC
(In reply to comment #0)
> Also see bug 246713, which may be related to this bug. I have booted kernel
> 2.6.22.9-91 with the usbcore.autosuspend=0 option.

Zero means suspend with no delay. To disable, use usbcore.autosuspend=-1



Comment 2 Richard Körber 2007-10-09 23:03:50 UTC
Thanks! The system is now running stable since I used usbcore.autosuspend=-1.

Is -1 the default value? (I.e. do I need to put this boot option into grub.conf,
anyways?)

Comment 3 Chuck Ebbert 2007-10-09 23:06:42 UTC
(In reply to comment #2)
> Thanks! The system is now running stable since I used usbcore.autosuspend=-1.
> 
> Is -1 the default value? (I.e. do I need to put this boot option into grub.conf,
> anyways?)

-1 is not the default.

Comment 4 Pete Zaitcev 2007-10-10 07:57:32 UTC
Would it be possible to try Rawhide kernel? I want to know if the new
"hub-only" autosuspend in it works ok.

The kernel can be found in
http://download.fedora.redhat.com/pub/fedora/linux/development/x86_64/os/Packages/
(or i386/ if you have a 32-bit system). Keep in mind though, it would prevent
normal updates with yum from taking place, because it's "newer" than fc7.
So just do rpm -i, test, rpm -e.

Comment 5 Chuck Ebbert 2007-10-10 16:45:16 UTC
I misspoke in comment #3: usbcore.autosuspend=-1 is the default in Fedora 7. So
this problem was caused by setting autosuspend to 0 (which we should just reject
anyway, it makes no sense to set it to that.)

Comment 6 Richard Körber 2007-10-12 06:40:09 UTC
(In reply to comment #4)
Pete, I have tried kernel-2.6.23-0.224.rc9.git6.fc8 (i686) from Rawhide, without
the autosuspend option being set. After about 30 hours of uptime, the error
appeared again.

First:

Oct 12 05:55:43 monique kernel: hub 5-8:1.0: activate --> -110
Oct 12 05:55:43 monique kernel: hub 5-8:1.0: hub_port_status failed (err = -108)
Oct 12 05:55:43 monique last message repeated 3 times

Then repeatedly until reboot (about every two seconds):

Oct 12 05:55:46 monique kernel: hub 5-8:1.0: activate --> -108
Oct 12 05:55:46 monique kernel: hub 5-8:1.0: hub_port_status failed (err = -108)
Oct 12 05:55:46 monique last message repeated 3 times

This weekend I will test the latest F7 production kernel without the autosuspend
option. According to Chuck in comment #5, it should work as reliable as with
autosuspend=-1.

Comment 7 Richard Körber 2007-10-17 07:28:42 UTC
kernel-2.6.22.9-91.fc7 is also working stable without autosuspend option. Which
means that the bug occured because I was messing with the kernel options. I am
sorry for the noise.

Comment 8 Richard Körber 2007-11-10 09:53:42 UTC
The bug reoccured today with kernel 2.6.23.1-10.fc7 without autosuspend option.
It happened again after about 30 hours uptime.

Nov 10 04:17:47 monique kernel: hub 5-8:1.0: activate --> -110
Nov 10 04:17:47 monique kernel: hub 5-8:1.0: hub_port_status failed (err = -108)
Nov 10 04:17:47 monique last message repeated 3 times

Then every two seconds:

Nov 10 04:17:50 monique kernel: hub 5-8:1.0: activate --> -108
Nov 10 04:17:50 monique kernel: hub 5-8:1.0: hub_port_status failed (err = -108)
Nov 10 04:17:50 monique last message repeated 3 times

What is the reason for this error? And is there a way to fix it for good?

Comment 9 Chuck Ebbert 2007-11-12 21:13:49 UTC
Unfortunately the Fedora 8 autosuspend changes leaked into the Fedora 7 kernel.
This controls autosuspend by device classes, and apparently doesn't work. Looks
like it should go back to being disabled completely by default.

Comment 10 Christopher Brown 2008-01-14 18:18:30 UTC
Hello,

I'm reviewing this bug as part of the kernel bug triage project, an attempt to
isolate current bugs in the Fedora kernel.

http://fedoraproject.org/wiki/KernelBugTriage

I am CC'ing myself to this bug and will try and assist you in resolving it if I can.

There hasn't been much activity on this bug for a while. Could you tell me if
you are still having problems with the latest kernel?

If the problem no longer exists then please close this bug or I'll do so in a
few days if there is no additional information lodged.

Comment 11 Christopher Brown 2008-02-16 02:44:16 UTC
Closing as per previous comment. Please re-open if this is still an issue for you.