I'm never sure whether it's best to continue things like this on the 
mailing list so everyone can follow along or in the Bugzilla entry...


On Sat, 5 Mar 2005, Andrew Morton wrote:

> Begin forwarded message:
> 
> Date: Sat, 5 Mar 2005 13:06:01 -0800
> From: [EMAIL PROTECTED]
> To: [EMAIL PROTECTED]
> Subject: [Bug 4294] kexec : on reboot, kernel hangs when reading 
> /proc/bus/usb/devices
> 
> 
> http://bugme.osdl.org/show_bug.cgi?id=4294
> 
> 
> 
> 
> 
> ------- Additional Comments From [EMAIL PROTECTED]  2005-03-05 13:06 -------
> Created an attachment (id=4676)
>  --> (http://bugme.osdl.org/attachment.cgi?id=4676&action=view)
> sysrq-t output
> 
> 
> I hacked /etc/rc.rc/rc.sysinit to comment out the lines which read
> /proc/usb/devices, so I could boot to initlevel 1. Here is the scenario :
> 
> # #do not unload uhci-hcd
> # kexec --load --initrd=/boot/initrd-2.6.11-mm1.img
> '--command-line=root=/dev/hda6 init 1' /boot/vmlinuz-2.6.11-mm1
> # reboot
> [ wait for the command prompt ]
> # /etc/init.d/syslog start 
> # cat /proc/usb/devices &
> [ press sysrq-t ]
> 
> One can see 2 processus in 'D' state : 
> - khubd in wait_for_completion
> - cat in __down/__down_failed/.text.lock.usb/usb_device_read
> 
> also; note the message before an after Sysrq-t : 
>  new low speed USB device using uhci_hcd and address 4
>  device not accepting address 4, error -110
> 
> 
> FYI :
> # cat /proc/interrupts 
>          CPU0       
>   0:  1535222          XT-PIC  timer
>   1:     6982          XT-PIC  i8042
>   2:        0          XT-PIC  cascade
>   5:     1644          XT-PIC  uhci_hcd:usb1, uhci_hcd:usb2, eth0, Ensoniq
> AudioPCI
>   7:        0          XT-PIC  parport0
>   9:        0          XT-PIC  acpi
>  10:        2          XT-PIC  ohci1394
>  12:    21150          XT-PIC  i8042
>  14:    11182          XT-PIC  ide0
>  15:    36076          XT-PIC  ide1
> NMI:        0 
> LOC:        0 
> ERR:        0
> MIS:        0

What's happening is that the system is failing to initialize the mouse 
correctly.  There are multiple retries (and multiple _layers_ of retries!) 
and the timeout is set to 5 seconds.  Hence the delays apparent in your 
syslog.  During that time the /proc/bus/usb/devices file is basically 
locked, so the process attempting to read it gets blocked.

I assume nothing like this happens during a normal bootup, only when you 
use kexec?  Actually I'm not at all familiar with how kexec works; can 
someone provide a quick explanation?

If you rmmod uhci-hcd before the kexec and then modprobe it afterward, 
do detection and initialization of the mouse work okay?

Assuming yes, it appears that uhci-hcd isn't reacting well to kexec.  How 
does kexec impinge on the driver, and what is it supposed to do?

Alan Stern

P.S.: For a possibly related bug report, see

http://bugme.osdl.org/show_bug.cgi?id=4276



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to