Random Reboots - Bugcheck 0x1000007f

Problem: I have constant FULL reboots. (3-4 reboots a day) No blue screen shows. I UNCHECKED automatic reboot. Still reboots without blue screen. (Does this mean I have hardware error?)

Here is my system:

Dimension XPS P4 3.0 GHZ
1GB Dual Channel 400MHz DDR SDRAM
Purchased: 5/23/03
12MB Radeon 9800 Graphics Card
Built in Intel PRO/100 network card
Sound Blaster Audigy Card

BIOS Version: A03
NVRAM: Cleared It

ERRORS:

Here are my main errors in Event Viewer:

All of these are SYSTEM REBOOTED FROM BUGCHECK Event ID: 1001

80% of reboots error: Bugcheck 0x1000007f Paramenter: 0x0000008

10% of reboots error: Bugcheck 0x100000d1

10% of reboots error: 0x1000008e Parameter: 0xc00001d

HARDWARE SOLUTIONS TRIED:

1. Took out every PCI board in system. Adaptec, Sound, Graphics card, network cards "“ Still rebooted with error
2. Took out Radeon graphics card. Put in old ATI card. Still rebooted
2. Checked memory with dell diagnostics for 16 hours "“ NO errors
3. Checked system board with dell diagnostics "“ no errors
4. Took out memory. Went from 1024MB to 512MB. Both sets of 512MB modules used. Still rebooted from both sets.
5. Disabled networking totally. Went into BIOS and turn off network card. Still rebooted.

Questions:

What else can I try? I am out of options.

Should I try BRAND NEW CRUCIAL MEMORY?

Could this be a software problem? Should I reformat? I do not want to.

Here is link to four dumps. They are labeled 7f, d1, 8e, Blue Screen.

http://www.sudberry.com/dump/

I just got a blue screen with Stop 0x00000000 (0x804FD313, 0xC0000000)

Hope you can decipher something.

 

Of course many things can cause such a thing as you have reported, RAM being the most common, or any faulty hardware, and even at one time a virus called Sasser.

If i was given this PC to fix i would try these things first considering the information you have provided.

Try completely different DDR chips, preferably 256MB chips. (i have seen many more faulty 512 chips than 256s)

Try another motherboard.

Re-install all drivers for motherboard and devices in system when booted.

All else fails a repair or full re-install of Windows. You can do this without changing or losing your current setup if you have access to another drive you can use to install windows. With just Windows and drivers installed, see if the PC behaves differently. It would take about 25 minutes to do if you have the 2nd drive to play with, and it will tell you a whole lot.

I would also consider it may be overheating. Pentium 4s over 2.8Ghz, especially Prescott P4s run extra hot. My Pentium 4 2.8 Prescott runs 10 degrees celcius hotter than the Standard 2.8Ghz P4 i had before it.

Try running the PC with the side off and in a cool environment for a while and see if it helps. Make sure also your CPU fan is spinning ok.

Make sure your power supply is blowing air out the back you can feel too.

This would go a long way to solving it or at least isolationg your problem more than you have already.

You already did the first things i would do before all this, which was take out all non essential cards from PC and also replace things like RAM and video card.

 

no need to run those dumps through the tool, i was able to get them from your site ok.

2. Checked memory with dell diagnostics for 16 hours "“ NO errors
3. Checked system board with dell diagnostics "“ no errors
These are not conslusive steps with intermittant errors.


Bugcheck-7f.dmp - bit errors transitioning from kernel to user mode (probable corrupt memory sent it silly)
Bugcheck-d1.dmp - corrupted memory
Bugcheck-8e.dmp - corrupted memory


Based on your troubleshooting thus far, i think you need to consider swapping mainboard and CPU. this is not a software error.

 

I have ordered 2 512MB crucial memory cards.

Probably dumb idea but I can return them.

If you had to guess do you think it is:

A. Motherboard
B. Processor

I should have bought the Dell service contract - ****.

I have always gambled and looks like my dice roll was bad this time.

thx for looking at the debugging.

 

If out of all the things i "guessed" above, i would put my money on memory being your problem. 2nd Choice motherboard. CPUs are very rarely faulty and in 6 years of building my own PCs for clients, i have only sent back 2 Pentium CPUs. One of them was blown by a power surge that took out the mobo as well, and not the CPUs fault. So i wouldnt worry about the CPU unless the fan isnt spinning.

Funny that the last post even mentioned corrupted memory a few times then he suggests a CPU/mobo change?

But it is possible of course that the motherboard is making the memory play up if it isnt handling it properly, overheating or has faulty sockets.

My money is on your RAM, let me know if i win

 

my guess is based on the odds that since it does it with either stick of ram, its less likely that they are both shot. *shrug* Its all a guess, i freely admit.

I would agree that its more likely a motherboard than a CPU. I have had quite a few CPUs go bad on me, most likely the L2 cache went bad on them.

Your guess is as good as mine. Let us know how it turns out.

 

True its less likely that 2 chips would fail. But the errors are so memory based and symptomatic of ram errors, thats why i lean that way, but good chance it is motherboard. Maybe the mobo is making the ram play up, so maybe any ram will behave the same. One thing for sure, it is hardware.

And i havent had many Pentiums go faulty but until i stopped buying them i had returns on about 20% of AMD chips. They jsut arent made as well, even thought thier speeds are good.

 

Bad news:
Joe wins the bet.
Brand new crucial memory put in this monrning. 1 hour later - bamb - reboot.

Question:

Before I order a Dell motherboard can you guys reassure me it is not the hard drive. A local tech says he has seen hard drives cause errors. When your computer goes to a certain spot it crashes. But he freely admits he doesn't know event codes like you guys do.

It doesn't crash when I am writing to the disk. It can crash just being idle.

 

When your computer goes to a certain spot it crashes.

In that case, you could get to a cmd prompt and run chkdsk /r since that will do surface testing of the drive and should certainly hit the entire drive. A crash during the chkdsk would not tell you for certain that it was the drive since the crashes are sorta random. However, a full chkdsk with no crash would certainly tell you that the tech's guess wasn't the problem.

I'm not familiar with the diagnostic tools Dell provides but they likely have a utility that will test out the entire drive for you as well but chkdsk should do it as well using that switch.

 

The data i examined could only have occurred in a situation where data was in memory, and was corrupted on a later access. Hard drive doesnt enter into it. running a full scandisk/chkdsk whatever could identify if you have a flaky i/o subsystem, we have no data at this time that indicated that.

I will offer you a 100% money back guarantee for the fee you paid me if it turns out not to be the motherboard.

 

Just completed chkdsk /r. Passed fine. No reboots.

Dell motherboard on order.

I'll keep you all informed how it goes. Could be a few days till I get it.

Thx again for all the help.

 

New motherboard installed.

Success!!!!! Solid as a rock. No reboots in 48 hours.

Joe's guarantee holds up.

I never thought motherboards could go bad. I always thought they were tested to last 300 years. I guess I was wrong.

This forum blows away Dell's forum. They are worthless there.

Thanks for the competence.

 

outstanding, im glad we were able to help. Thanks for following up.

 

Like a bad virus that won't go away my random reboots are happening again. This is with my new motherboard installed in Dec.

This is my main error:

Bugcheck 0x1000007f Paramenter: 0x0000008

The reboots are spaced out - 1/12/05 2/4/05 2/17/05

Here are my minidumps:

http://www.sudberry.com/dump/

Can you guys look at theses dumps.

Where do I go from here. I have a new motherboard on order from Dell.
Could it be the processor?
Powersupply? (Dell told me probably not in that the reboots are so far apart)

Its not my memory stiicks because I tried brand new crucial on my last motherboard and it rebooted on crucial also.

What does it lead to if your reboots are so spaced apart time wise like mine are?

 

Thanks to a coworker who explained this phenomenon to me, we have this:

Stop 7F, 0x00000008 (double-fault) error occurs because of a single-bit error in the ESP register
http://support.microsoft.com/?kbid=842465

d> .formats esp^ebp&FFFF0000
valuate expression:
Hex: 00080000
Binary: 00000000 00001000 00000000 00000000

kd> !cpuinfo
CP F/M/S Manufacturer MHz PRCB Signature MSR 8B Signature Features
TargetInfo::ReadMsr is not available in the current debug session
0 15,2,9 GenuineIntel 2992 0000001700000000 80073fff



Your next step is to follow the instructions in that KB article, pull down that intel utility, confirm you have the wrong stepping, then call and ask for new bios or chips.

 

I think this KB article that Joe sites has got to be it. My values as seen below match exactly to the values that are trouble according to that article.

Question:

I have latest BIOS from DELL A03. There is no higher version. My system is out of warranty so I cannot demand a new processor.

What are my options:

1. Call Intel directly see if they can do something for me?

2. Bite the bullet and buy a new processor for $200.00?

3. Is there a BIOS I can buy that would fix this and get me to a revision of 24 or later?

4. Is there an Intel software program that can load and fix things with this processor?


The symptoms that are described in this article have been observed most frequently on Intel Xeon processors that have the CPU Family, CPU Model, and CPU Stepping values of 15, 2, and 9 respectively and that are installed on motherboards that use ServerWorks chipsets. (The CPU Family, CPU Model, and CPU Stepping hexadecimal values are F, 2, and 9 respectively.) These processors require a revision value of 0x18 or later to function correctly. (0x18 is equivalent to a decimal value of 24.)

Here is the Intel Utility Results for my computer:

Intel(R) Processor Identification Utility
Version: 1.5.20050202
Time Stamp: 2005/02/19 18:58:52
Number of processors in system: 1
Current processor: #1
Processor Name: Intel(R) Pentium(R) 4 CPU 3.00GHz
Type: 0
Family: F
Model: 2
Stepping: 9
Revision: 21
L1 Trace Cache: 12 Kµops
L1 Data Cache: 8 KB
L2 Cache: 512 KB
Packaging: FC-PGA2
EIST: No
MMX(TM): Yes
SIMD: Yes
SIMD2: Yes
SIMD3: No
Enhanced Halt State: No
Execute Disable Bit: No
Hyper-Threading Technology: No
Intel(R) Extended Memory 64 Technology: No
Expected Processor Frequency: 3.0 GHz
Reported Processor Frequency: 3.0 GHz
Expected System Bus Frequency: 800 MHz
Reported System Bus Frequency: 800 MHz

 

i'd call them anyways. Your chip needs an update, and they should provide that in the form of a bios update, or a new chip. They may be willing to work with you since its a documented errata. Barring that, you'd need to try to find a new processor, or look at this as a chance to upgrade

Intel will refer you to your vendor. The BIOS is written by your vendor, they have to rely on dell to supply that for you.

 

Joe:

1. People are saying that your KB article refers to XEON processors and not mine. (Dell said that in a message I posted)
2. Do you think it could be a power supply issue ( I am running mother board monitor 5 and nothing is unusual)
3. I have a full dump. Could you really get to the bottom of it then. I can send you a CD of a full dump. (I can private message you for address)

I just don’t want to spend $350 for a processor and have it be something else. Crashed again just an hour ago. (the computer from he1*)


Experts exchange comments:


Comment from connj
Date: 02/20/2005 06:12PM PST
Comment


The dumps refer to a faulty ntoskrnl and win32k.sys which would lead one to believe that it is a hardware issue. From his previous posts it looks as if you haven't ruled out a power supply issue. If the computer is 2 yrs old and used quite often I don't see why it couldn't be a simple $60 power supply fix rather than a $300 processor. Have you checked that all your cooling fans are running smoothly??

Comment from cpc2004

If it is a hardware error, you will get a lot of different bugcheck codes such as 0A, 50, 8E and etc. Your dumps all fails with bugcheck code 7F with double fault (000000008) which usually caused for stack overflow. From the stack trace the caller is 0x7c90eb94 and I believe that the caller used up most of stacks and cause ntoskrnl.exe and win32k.sys failed with stack overflow and finally trigger double check. Since this is a mindump and I don't know 0x7c90eb94 belongs to which process. If you can attach the full dump here, I'm confident to find out the culprit.

STACK_TEXT:
b1b71364 804e0f07 00000001 0de6d8e0 00000000 hal!KeAcquireQueuedSpinLock+0x3b
b1b71364 00000030 00000001 0de6d8e0 00000000 nt!KiTrap0E+0xd0
WARNING: Frame IP not in any known module. Following frames may be wrong.
00000081 00000000 00000000 00000000 00000000 0x30
b1b719a4 80570593 b1b71a60 b1b71a64 b1b71a34 nt!KiCallUserMode+0x4
b1b71a00 bf813d09 00000002 b1b71a44 00000018 nt!KeUserModeCallback+0x87
b1b71a84 bf813ea0 bc72ea08 00000020 0009061e win32k!SfnDWORD+0xa8
b1b71acc bf814092 0072ea08 00000020 0009061e win32k!xxxSendMessageToClient+0x176
b1b71b18 bf80f470 bc72ea08 00000020 0009061e win32k!xxxSendMessageTimeout+0x1a6
b1b71b3c bf81fc67 bc72ea08 00000020 0009061e win32k!xxxSendMessage+0x1b
b1b71b6c bf81d7a8 e1bec3f8 00000001 00000200 win32k!xxxMouseActivate+0x22f
b1b71c94 bf801e3d 00000200 b1b71d14 00000000 win32k!xxxScanSysQueue+0x85f
b1b71ce8 bf80365e b1b71d14 000020ca 00000200 win32k!xxxRealInternalGetMessage+0x335
b1b71d48 804ddf0f 309a0d60 00000000 00000200 win32k!NtUserPeekMessage+0x40
b1b71d48 7c90eb94 309a0d60 00000000 00000200 nt!KiFastCallEntry+0xfc
0012cfe8 00000000 00000000 00000000 00000000 0x7c90eb94 <---

Mini011205-01.dmp BugCheck 1000007F, {8, 80042000, 0, 0} nt!CcFreeActiveVacb+ee
Mini021705-01.dmp BugCheck 1000007F, {8, 80042000, 0, 0} win32k!pfsubGetFontSub+4d
Mini020405-01.dmp BugCheck 1000007F, {8, 80042000, 0, 0} nt!KiTrap0E+d0

Comment from dbeers99
Date: 02/20/2005 07:25PM PST
Author Comment

http://support.microsoft.com/?kbid=137539&sd=RMVP

Double faults are common to hardware faliures as well. You have already checked ram, motherboard, could be a faulty software driver someplace but it just seems so typical of a bad power supply or at least one that is on its way out. Is there a way to check the voltages your getting from the power supply in your BIOS. Anyways I'm off to bed. Good night and Good Luck.

Comment from cpc2004
Date: 02/21/2005 01:25AM PST
Comment


The analyze report of Mini020405-01.dmp shows an unusual messages
OVERLAPPED_MODULE: dpcnet5u

I serached google and find out that dpcnet5u.sys belongs to direcway drivers.

 

I am now familiar with the errata referenced in this KB article, this phenomenon is known to occur on different configurations, not just Xeons/Serverworks bridges, although with less frequency. You are experiencing this problem documented in that KB, i have 100% confidence in this. Note the ambigious language. This is intentional.

I would be very interested to know how someone can look at three dumps that all have a clear bitflip in ESP and think its a power supply problem. I dont think so.

This analysis in incomplete. The crash is the result of a stack overflow, because ESP no longer has a valid value, due to a bit error. (see my earlier notes for evidence of this).

3. I have a full dump. Could you really get to the bottom of it then. I can send you a CD of a full dump. (I can private message you for address)
This option would cost you more money than just trying a new proc It doesnt matter, there is no additional data in the dumps. The bit is flipped, there is nothing in the dump that will explain why it happened, since it happened internal to the chip.

did you call dell yet?