Crash after removing USB to Ethernet Adapter while hiberna..

[Guest]

Archived from groups: microsoft.public.windowsxp.general (More info?)

Problem
We have been testing some USB to Ethernet adapters and are encountering blue
screen system crashes that are 100% reproducible in certain circumstances
where the USB to Ethernet adapter is removed while hibernating.

The system crash screen contains the following information:
BAD_SPOOL_CALLER
STOP : 0x0000007E (0xC0000005, 0xF8850371, 0xF8AB3C10, 0xF8AB390C)
usbhub.sys

The specific steps that reproduce the problem are:
1) Power up Desktop PC or Laptop.
2) Connect USB to Ethernet Adapter.
3) Induce Desktop PC or Laptop to hibernate (not standby!).
4) Disconnect USB to Ethernet Adapter.
5) Wake Desktop PC or Laptop from hibernation. -> blue screen crash.

The problem was observed with the following USB to Ethernet Adapters:
Hawking HUF11 – tried following drivers
Provider: RealTek, File:RTL8150.SYS, Version: 5.119.0801.2003
Provider: SMC2208USB/ETH, File SMC2208.SYS, Version 5.115.0430.2002
Linksys USB200M – with driver (this dev
Provider: Linksys, File: USB200M.sys, Version 1.0.0.4
ADMtek ADM8511 USB To Fast Ethernet Converter – with driver
Provider: ADMtek, File ADM8511.SYS, Version: 2.04.2001.0719

We tried all the latest associated drivers we could find with no success.

The problem was observed with the following Desktops and laptops:
DELL Latitude D610, Windows XP Professional SP2
DELL Latitude D600, Windows XP Professional SP2
Panisonic Tough Book, Windows XP Professional SP2
DELL Optiplex GX260, Windows XP Professional SP1 -- occurs with Hawking
device only

Note:
The problem is not reproducible if the same steps are performed using stand
by mode. Once the PC has entered and recovered from stand by mode, the
problem is no longer reproducible even in hibernation mode until a clean
reboot.

Can anyone suggest a fix or work around. Though the scenario is not typical
in home use, it is expected to be a common scenario in our application.

 

[Guest]

Archived from groups: microsoft.public.windowsxp.general (More info?)

> We have been testing some USB to Ethernet adapters and are encountering
> blue
> screen system crashes that are 100% reproducible in certain circumstances
> where the USB to Ethernet adapter is removed while hibernating.

Why would this be a surprise? Hibernation absolutely depends on the state
of the machine to be EXACTLY as it was when hibernation occurred. If you
want to be able to unplug something, use Standby which refreshes the System
State/Configuration upon restart.

--
..__
(__
__ )teve www.steve.shattuck.net steve@shattuck.net

 

[Guest]

Archived from groups: microsoft.public.windowsxp.general (More info?)

Ar you suggesting that Microsoft designed XP to crash if you unplug cables
from a laptop while hibernating? We have 1000s of technicians that are going
to be using laptops in the near future and it is very likely that they will
occassionally leave the laptops and connected test equipment long enough for
the computer to enter hibernation not by choice (default is 2hrs on most
systems) . If it happens to be the end of the day when they return to the
laptop, they will assume the laptop is off, disconnect cables and pack things
up for the day. The next morning the system will crash when powered up.

I have searched every where for issues related to system crashed when USB
devices are unplugged and none offer solutions and most appear to be
intermittent and difficult to reproduce. I thought, since this scenario is
reproducable 100% of the time on the 4 different computers we tested using 4
different adapters with unique device drivers (some digitally signed), that
Microsoft might be interested in investigating this. A highly reproducable
bug should be an easy fix and perhaps it may solve some of the other
seemingly intermittent scenarios as well. I called Microsoft technical
support and the only avenue they offered for reporting possible bugs was the
newsgroups. The way I see it, either Windows XP has a bug or device driver
developers are collectively making the same mistake. It took weeks of
testing to isloate the scenarios that reproduce the problem and if Microsoft
is not interested in taking advantage of this information I'll be very
disappointed.

"Steve Shattuck" wrote:

>
> Why would this be a surprise? Hibernation absolutely depends on the state
> of the machine to be EXACTLY as it was when hibernation occurred. If you
> want to be able to unplug something, use Standby which refreshes the System
> State/Configuration upon restart.
>
> ..__
> (__
> __ )teve www.steve.shattuck.net steve@shattuck.net