Stop 0x50 w/ Verifier & WinDBG loaded

G

Guest

Dell D600 M Laptop
512 MB RAM
1.7 GHz Centrino
BugCheck 50, {ff909000, 0, 804e422b, 0}

I have been fighting a stop 0x50 error for several months now. Dell has
replaced motherboard, memory, processor, and HD. 2 other people in my area
with the exact same configuration have the same problem, and it started
happening for all of us around March. I went to SP2 to hopefully fix this,
but it didn't help and possibly made it worse. I am in SP1 right now.

I have multiple mini and kernel dump files. The first set was prior to
turning on the verifier. It shows a random mem address in upper memory, but
always the same source address which shows up in the NT kernel itself
804e422b.

Turning on the verifier switched all of the blue screens to 0xC4_090
errors instead. I used SiSoft Sandra to exercise the system, and got
consistent failures during the HD test with verifier on. I could not even
run that test without crashinng.

The first post-verifier dump identified SAVRT.SYS, so I disabled all
Symantec Antivirus services (9.0 Corporate) and processes and tried again.
Next, it reported a fault back in ntoskrnl.exe. I backed out one of the
following hotfixes to put NTOSKRNL.EXE back prior to March 2005
(890859,826939, 885835). These hotfixes werre pushed down by a corporate
auto-update tool, which I suspect is the reason why this started happening
all of a sudden.

With an October 2004 version of NTOSKRNL.EXE resinstalled, I ran the
SiSoft disk test twice in a row with no problems. I don't have enough run
time to prove this 100%, but if this fixes it then I can't upgrade to SP2 or
accept further security updates on these three Dell systems.

Dell is trying to troubleshoot this, as is our IT department. But, we
clearly need Microsoft's help in taking it to the next level. WinDBG is
helpful, but I am a novice in using it. I am a firmware programmer by trade,
but am not skilled in Windows kernel troubleshooting.

I can forward whichever dumps and logs you want.

Regards,
Ron Garnett
 
D

David Candy

http://oca.microsoft.com and upload there. Or raise a support incident with
MS direct (ie telephone).

As you have Windbg try a full dump (they are too big to send to anyone) as
each increase in size stores more information about the crash.

Here we really just do what you can do for yourself (windbg), identify the
third party driver, and get you to disable it.

You don't mention if you tried Clean Boot Troubleshooting. Type it in Help
to see how (I suppose you've already tried this).
 
G

Guest

Ron Garnett said:
Dell D600 M Laptop
512 MB RAM
1.7 GHz Centrino
BugCheck 50, {ff909000, 0, 804e422b, 0}

I have been fighting a stop 0x50 error for several months now. Dell has
replaced motherboard, memory, processor, and HD. 2 other people in my area
with the exact same configuration have the same problem, and it started
happening for all of us around March. I went to SP2 to hopefully fix this,
but it didn't help and possibly made it worse. I am in SP1 right now.

I have multiple mini and kernel dump files. The first set was prior to
turning on the verifier. It shows a random mem address in upper memory, but
always the same source address which shows up in the NT kernel itself
804e422b.

Turning on the verifier switched all of the blue screens to 0xC4_090
errors instead. I used SiSoft Sandra to exercise the system, and got
consistent failures during the HD test with verifier on. I could not even
run that test without crashinng.

The first post-verifier dump identified SAVRT.SYS, so I disabled all
Symantec Antivirus services (9.0 Corporate) and processes and tried again.
Next, it reported a fault back in ntoskrnl.exe. I backed out one of the
following hotfixes to put NTOSKRNL.EXE back prior to March 2005
(890859,826939, 885835). These hotfixes werre pushed down by a corporate
auto-update tool, which I suspect is the reason why this started happening
all of a sudden.

With an October 2004 version of NTOSKRNL.EXE resinstalled, I ran the
SiSoft disk test twice in a row with no problems. I don't have enough run
time to prove this 100%, but if this fixes it then I can't upgrade to SP2 or
accept further security updates on these three Dell systems.

Dell is trying to troubleshoot this, as is our IT department. But, we
clearly need Microsoft's help in taking it to the next level. WinDBG is
helpful, but I am a novice in using it. I am a firmware programmer by trade,
but am not skilled in Windows kernel troubleshooting.

I can forward whichever dumps and logs you want.

Regards,
Ron Garnett
Here's one referal to the 0x50

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/
ddtools/hh/ddtools/BCCodes_afaf4889-08de-40ea-b335-729d8213f3b5
..xml.asp

also take a look here
http://support.microsoft.com/?kbid=894278&sd=RMVP#kb1

the search will go on.
 
G

Guest

David,

1) I couldn't seem to find anywhere to upload to on oca.microsoft.com. I
also saw no evidence of any reports I had sent, so I may a browser config
problem, or am on crack.

2) I hadn't tried Clean Boot exactly, but had disabled as may services as
I could. Verifier certainly makes things less stable, so maybe I can enable
verifier for everything not MS, then turn off uncertified drivers to see if
that stabilizes things. The least stable drivers with verifier are
Microsoft's, because I can't even boot with all MS drivers enabled. That
seems counter-intuitive if the certification process is supposed to wring out
such things, but I am not naive enough to believe it's a fix-all. In any
case, I'll spend more time in this area.

3) I set the full dump boot on, so will try this again later.

4) I am having trouble getting the WinDBG symbols to load correctly
"xpsp1sym_x86_chk.exe," so that might be impdeding things a bit. All I have
figured out so far is the "!analyze -v" thing, but that is clearly spelled
out. I'm not sure what else to do in the program relative to this bug.
 
G

Guest

Unrepper,

Thanks for the links. Neither of the virus-related links panned out. I
had seen those already, but none of the signature files exist.

I turned off verifier for a bit, but still got a crash. OCA told me it
was related to a video driver, although none of the dumps showed that. I
tried turning off all acceleration to see if that helps. Time will tell.
 
G

Guest

My last reply seemed to go away, so I apologize if this is a dupe.

Thanks for the links. The viral things didn't pan out. I had seen those
and tried already, but no signatures appeared.

I actually crashed while trying to get to the first link, but it was before
I had finished the cut and paste operation. OCA told me that the video
driver was hung, but the crash dump was the old stop 0x50 with no details on
which driver failed. I turned off acceleration anyway, just in case.

Ron
 
C

Chris Swinney

Did you manage to resolve this issue. We have a similar error with a
PAGE_FAULT_IN_NONPAGED_AREA blue screen.



Ours seems to be profile related not hardware related. The stops appear to
be random but we can force the issue by simply searching the registry for a
text string that would not exist (eg "Testest"). This WILL ALWAYS cause a
blue screen stop 50 error at the end of the registry search in certain
profiles, yet in others, the error will not happen. However, we have run
certain stress tests and mrmory test with no problems.



Chris





Kelly said:
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top