B
Ben Harris
I've been seeing a strange problem with EWF running on XP Embedded. I've
Googled the newsgroup archive, but haven't found anything that seems to
help. First, let me explain a bit of background information:
- I've installed Q823025 and Q832662, to update the EWF files to the latest
versions.
- The image I'm using has always worked fine previously without any EWF
components included.
- I'm using a RAM EWF, and using the standard Target Designer interface to
implement it (leaving a 0.5MB empty space at the end of the disk). I'm not
using Slobodan's component yet, but I intend to try that today to see if it
resolves the strange behaviour I'm seeing.
- I've tried two different hardware setups (identical hardware), and both
setups have the same problem.
Initially, after installing the RAM EWF, everything worked fine. I could
enable it, and disable it using EWFMGR, and it worked as it was supposed to.
Then, I saw a one-off BSOD on boot-up (I didn't make a note of the STOP code
at the time). After that single BSOD, it all seemed to work fine
afterwards - and then I got another BSOD! Not 100% happy that EWF was
working well enough to release this build to production, I wrote a simple
batch file which would toggle EWF enablen/disable, depeneding on the state
of it at the last boot, and then issue a "SHUTDOWN -R -T 0". I also wrote a
program which would wait a random amount of time (up to 60 seconds) before
calling the batch file, and made a shortcut to it in the Startup folder so
it would run every boot. The machine then seemed to work fine for a couple
of hours, continually toggling EWF,and rebooting; and then I got another
BSOD. This time, I made a note, and it was a STOP 0x00000050 :
PAGE_FAULT_IN_NONPAGED_AREA. Also, the machine didn't recover after a reset,
and continued to BSOD with the same error on every subsequent boot. I
repeated exactly the same test on another board with a different set of
identical hardware, although this time I was using a hard disk, and after a
while the same thing happened again.
As I didn't know whether EWF was enabled or disabled at the time it BSOD'ed
on me, I reinstalled XPe on both boards, and set one to EWF enabled, and the
other to EWF disabled. I then used the same program I had used before, but
disabled the toggle enabled/disabled function, so that the board was simply
rebooting after a random amount of time. I left this running over the
weekend, to see the results today (Monday morning). I was expecting to find
that the board with EWF enabled would have a BSOD, and the one with it
disabled was running fine. Instead, the opposite happened - the one with EWF
enabled was still randomly rebooting (as it was supposed to be), and the one
with it disabled was sitting with exactly the same STOP error on the screen.
Has anyone got any thoughts or suggestions? I'll try out Slobodan's EWF
component, and see if that makes any differece.
Ben Harris
Design Engineer
Arcom,
Clifton Road,
Cambridge CB1 7EA
United Kingdom
E-Mail: bharris "at" arcom "dot" com
Web: http://www.arcom.com
Googled the newsgroup archive, but haven't found anything that seems to
help. First, let me explain a bit of background information:
- I've installed Q823025 and Q832662, to update the EWF files to the latest
versions.
- The image I'm using has always worked fine previously without any EWF
components included.
- I'm using a RAM EWF, and using the standard Target Designer interface to
implement it (leaving a 0.5MB empty space at the end of the disk). I'm not
using Slobodan's component yet, but I intend to try that today to see if it
resolves the strange behaviour I'm seeing.
- I've tried two different hardware setups (identical hardware), and both
setups have the same problem.
Initially, after installing the RAM EWF, everything worked fine. I could
enable it, and disable it using EWFMGR, and it worked as it was supposed to.
Then, I saw a one-off BSOD on boot-up (I didn't make a note of the STOP code
at the time). After that single BSOD, it all seemed to work fine
afterwards - and then I got another BSOD! Not 100% happy that EWF was
working well enough to release this build to production, I wrote a simple
batch file which would toggle EWF enablen/disable, depeneding on the state
of it at the last boot, and then issue a "SHUTDOWN -R -T 0". I also wrote a
program which would wait a random amount of time (up to 60 seconds) before
calling the batch file, and made a shortcut to it in the Startup folder so
it would run every boot. The machine then seemed to work fine for a couple
of hours, continually toggling EWF,and rebooting; and then I got another
BSOD. This time, I made a note, and it was a STOP 0x00000050 :
PAGE_FAULT_IN_NONPAGED_AREA. Also, the machine didn't recover after a reset,
and continued to BSOD with the same error on every subsequent boot. I
repeated exactly the same test on another board with a different set of
identical hardware, although this time I was using a hard disk, and after a
while the same thing happened again.
As I didn't know whether EWF was enabled or disabled at the time it BSOD'ed
on me, I reinstalled XPe on both boards, and set one to EWF enabled, and the
other to EWF disabled. I then used the same program I had used before, but
disabled the toggle enabled/disabled function, so that the board was simply
rebooting after a random amount of time. I left this running over the
weekend, to see the results today (Monday morning). I was expecting to find
that the board with EWF enabled would have a BSOD, and the one with it
disabled was running fine. Instead, the opposite happened - the one with EWF
enabled was still randomly rebooting (as it was supposed to be), and the one
with it disabled was sitting with exactly the same STOP error on the screen.
Has anyone got any thoughts or suggestions? I'll try out Slobodan's EWF
component, and see if that makes any differece.
Ben Harris
Design Engineer
Arcom,
Clifton Road,
Cambridge CB1 7EA
United Kingdom
E-Mail: bharris "at" arcom "dot" com
Web: http://www.arcom.com