2.4.0-test10 problems

2.4.0-test10 problems

[Richard Polton]

(usb people, issue 1 is partly relevant and issue 3 is definitely USB.)

I have been testing my test10 installation and have come up with
a few old problems, all of which have been reported before.

1. Warm reboot fails to restart, i.e. hangs after displaying 'Restarting

    system'. In this particular scenario, the power switch is disabled
    too and the only way in which the machine responds is by switching
    off at the wall and pulling the battery. Note that this scenario has
been
    observed only if I boot, get the login prompt (optionally log in)
and then
    ctrl-alt-del. A similar scenario occurs after an amount of time
using the
    machine and then rebooting. In this case, the machine restarts
successfully
    but hangs when it tries to initialise (right word?) the UHCI
controller.

    Here is the exchange Randy Dunlap and I had about this some
(internet) time ago:



> Not absolutely sure about the numbers, but I think that the
> UHCI controller
>
> is Intel 82371AB USB Host Controller (PIIX4) (rev. 0x01).
>
> The uhci hcd is built into the kernel. There are not any usb
> modules loaded
>
> on reboot. When I say 'locks up ...' I do indeed mean that shutdown
> succeeds
> as does some portion up until the HCD initialises of startup.
>
> Thanks,
>
> Richard
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Richard,

I have that same USB host controller in my PC (not laptop).

I have built usb-uhci and uhci-alt (which Richard is not using
BTW) into the kernel (2 separate kernels of course) and
rebooted with both of them successfully.

Can you provide a kernel log up to the point of the failure?

Maybe there's something other than USB that's locking up,
but it appears that it's the USB UHCI HCD that's the problem
because that's the latest info on the console.
Or does usb-uhci actively complain about a problem?

My experiments didn't work, so I'll try just thinking
about it... The usb Makefile and module inits did change
recently.  Maybe that has something to do with it.
I'll keep thinking/checking on it.

~Randy

> "Dunlap, Randy" wrote:
>
> > > I am running 2.4.0-test8 on my laptop with a UHCI controller. When

> > > performing a
> > > warm reboot the machine locks up on initialisation of the
> controller.
> > > This did not occur under 2.4.0-test5. Any thoughts?
> >
> > Need more info, such as:
> >
> > Which uhci host controller driver (HCD)?
> > Is the uhci HCD built as a module or in-kernel?
> >
> > If module:
> > Is the uhci HCD loaded when you do the warm reboot?
> >
> > What other USB drivers are loaded when you do the reboot?
> >
> > And you say: "locks up on initialisation of the controller".
> > Does this mean that the shutdown succeeds and some portion
> > of the following init succeeds, up until the HCD is trying
> > to initialize?
> >
> > ~Randy


    I thought that it had been fixed, but I was wrong.


2. Parallel port cdrom fails to mount disks:
    modprobe friq
    modprobe pcd
    mount /dev/pcd0 /mnt/cdrom    fails with a return code of 32.
    Additionally, lsmod after the aborted mount shows that the following
modules
    have been unexpectedly loaded: nls_iso8859-1, nls_cp437, vfat and
fat. Note
    that the cd is an iso9660 formatted disk.

    Here is the relevant section of my log file:

Nov  7 20:21:52 turbocharged kernel: paride: version 1.05 installed
Nov  7 20:21:52 turbocharged kernel: paride: friq registered as protocol
0
Nov  7 20:22:07 turbocharged kernel: pcd: pcd version 1.07, major 46,
nice 0
Nov  7 20:22:07 turbocharged kernel: pcd0: friq 1.01, Freecom IQ ASIC-2
adapter at 0x378, mode 0 (4-bit), delay 1
Nov  7 20:22:09 turbocharged kernel: pcd0: Master: R/RW 2x2x24
Nov  7 20:22:09 turbocharged kernel: pcd0: mode sense capabilities
completion: alt=0x53 stat=0x51 err=0x60 loop=0 phase=3
Nov  7 20:22:09 turbocharged kernel: pcd0: mode sense capabilities:
Sense key: 6, ASC: 29, ASQ: 0
Nov  7 20:22:09 turbocharged kernel: Uniform CD-ROM driver Revision:
3.11
Nov  7 20:22:57 turbocharged kernel: cdrom: open failed.
Nov  7 20:22:57 turbocharged kernel: VFS: Disk change detected on device
pcd(46,0)
Nov  7 20:23:17 turbocharged kernel: cdrom: open failed.
Nov  7 20:23:17 turbocharged kernel: VFS: Disk change detected on device
pcd(46,0)
Nov  7 20:23:25 turbocharged kernel: cdrom: open failed.
Nov  7 20:23:25 turbocharged kernel: VFS: Disk change detected on device
pcd(46,0)


3. usb-storage: I am using a LaCie USB hard disk which generally behaves
perfectly well.
    In one situation, though, I have problems. I unmounted all the
partitions on the USB
    device manually and then issued the 'halt' command. During shutdown,
towards the end
    where usb-storage is removed (or something similar - I forgot to
bring that bit of the log file )-8   )
    I see reams of

    usb-storage: *** thread sleeping

    messages which, in fact, never stop. The only recourse is the power
switch.

Thanks for listening,

Richard

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: [linux-usb-devel] 2.4.0-test10 problems

[Richard Polton]

I currently do not use either APM or ACPI. Initially I used ACPI
and removing it in test8 appeared to fix the problem (but I suspect
that was just 'appear' rather than 'fix' 8-). I moved to APM instead
of ACPI in test9 - no change, and indeed in test10 I use neither.

There is a flag in the BIOS for Plug n Play OS. I shall toggle it and
observe the results.

Brad Hards wrote:

> Richard Polton wrote:
> > I have been testing my test10 installation and have come up with
> > a few old problems, all of which have been reported before.
> Don't known about the second two, but maybe can shed some light on the
> first one.
>
> > 1. Warm reboot fails to restart, i.e. hangs after displaying 'Restarting
> >
> >     system'. In this particular scenario, the power switch is disabled
> >     too and the only way in which the machine responds is by switching
> >     off at the wall and pulling the battery. Note that this scenario has
> > been
> >     observed only if I boot, get the login prompt (optionally log in)
> > and then
> >     ctrl-alt-del. A similar scenario occurs after an amount of time
> > using the
> >     machine and then rebooting. In this case, the machine restarts
> > successfully
> >     but hangs when it tries to initialise (right word?) the UHCI
> > controller.
> I think that the problem is pci / power management related. I have seen
> similar problems with my laptop (VAio F430), especially when warm
> booting from Win98 into Linux.
>
> Things to try:
> 1. Look at PnP or similar options (could be named anything) in the BIOS,
> and try toggling them.
>
> 2. Try APM instead of ACPI, or turn off power management.
>
> 3. If it hangs, 'soft cycle' the power. This is effective about 90% of
> the time for me.
>
> If any of this helps, you might like to post the results. I intend to do
> some more testing with -test10 over the next week or so, and will also
> post results.
> _______________________________________________
> linux-usb-devel@lists.sourceforge.net
> To unsubscribe, use the last form field at:
> http://lists.sourceforge.net/mailman/listinfo/linux-usb-devel

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: [linux-usb-devel] 2.4.0-test10 problems (power-down problem)

[Dan Streetman]

On Wed, 8 Nov 2000, Richard Polton wrote:

>the power switch is disabled
>too and the only way in which the machine responds is by switching
>off at the wall and pulling the battery.

I have seen this with my IBM Thinkpad 600E several times.

Many (newer, at least) IBM machines I've seen will power down if you hold the
power button down for 2-3 seconds.  Try that instead of pulling the plug and
battery.

It may be true for other machines also.

-- 
Dan Streetman
ddstreet@us.ibm.com


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: [linux-usb-devel] 2.4.0-test10 problems (power-down problem)

[Richard Polton]

The power switch is totally unresponsive in this situation.

Richard

Dan Streetman wrote:

> On Wed, 8 Nov 2000, Richard Polton wrote:
>
> >the power switch is disabled
> >too and the only way in which the machine responds is by switching
> >off at the wall and pulling the battery.
>
> I have seen this with my IBM Thinkpad 600E several times.
>
> Many (newer, at least) IBM machines I've seen will power down if you hold the
> power button down for 2-3 seconds.  Try that instead of pulling the plug and
> battery.
>
> It may be true for other machines also.
>
> --
> Dan Streetman
> ddstreet@us.ibm.com

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/