Good day,

I'm testing a USB2 hard drive box under 2.4.19-pre10 with mixed results,
mostly lousy.  I'm getting hard lockups (keyboard/touchpad
nonresponsive, external access to running SSHD futile) in various
circumstances which all seem to inhibit logging.

So how should I collect debug info to describe this to the list?

Context: Sony Vaio Z505LS, Adaptec USB 2.0 cardbus card.  Generic USB
drivebox.  

  http://www.directron.com/usbbox20.html

(Box is well constructed but there evidently have been some problems
since Directron gives return instructions right in the description!)

Debian Woody for init scripts including hotplug.  In-kernel PCMCIA
compiled as modules, USB as modules.  There is a memory stick slot in
the machine that shows up as a usb-storage device on the Vaio's built-in
USB1.1.  Hotplug detects everything nicely.  /proc/bus/usb drivers and
devices is uninteresting.  The 1.1 ports talk to everything I throw at
them, including the hard drive box.


Scenario #1: Hard lock if a partition on the USB drive is mounted while
launching X; mounting after X is running is ok.  The Z505LS has an ATI
Rage Mobility chipset if that is relevant.

Scenario #2: Hard lock at shutdown if a partition was mounted, then
unmounted before shutting down.  (Could there be a wrong order in
umount, ehci unload, and pcmcia unload in the Debian scripts?)

Scenario #3: Hard lock during cardctl eject unless there was never a
drive plugged into the card.  No difference if partitions are unmounted
or mounted, or if rmmod ehci and rmmod usb-storage.  In one case the
lockup left the PCMCIA eject beep running; never heard a machine scream
before!

Another negative, possibly related to SCSI modules, is that
unmount/unplug drive and replug drive sometimes gives a different drive
letter and sometimes does not.


I can get one scenario to work --

1. Boot without the card inserted.
2. Launch X.
3. Insert the card.
4. Plug in the drive.
5. Mount a partition.
6. Do whatever.
7. Shut down X.
8. Leave the partition mounted and the card inserted, and poweroff.


Interestingly, I don't have any problems with the same PC card and an
external USB2 CD-RW drive.  So if there is a hardware issue I must
suspect the external hard drive box.  I've tried three different hard
drives in it; currently a UDMA2 western digital drive seems to be okay,
and a UDMA5 (ATA-100) drive was least stable.  I compiled a kernel on
the UDMA2 drive.  9 megabytes/sec hdparm is not bad (and very adequate
for my purposes), about what I'd get from the drive inside a desktop
unit.


Or does anyone have a specific USB2 drivebox to recommend?

Any suggestions about debug approach?  Appreciate your help.

-- thanks, SP



-------------------------------------------------------
This sf.net email is sponsored by: Jabber - The world's fastest growing 
real-time communications platform! Don't just IM. Build it in! 
http://www.jabber.com/osdn/xim
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to