Blue Screen twice - Any ideas why?

pdsnickles · Apr 2, 2009

I got a blue screen on the 1st day I was up and running with Windows Vista 64 (pre-loaded on my new HP Pavilion computer).

Everything was working beautifully, I was very happy that it all went well and I had no major problems getting set up.

Then out of the blue, nothing was going on, I went to the START menu and as soon as I touched the Start button it crashed into blue screen.

WHen it restarted I got this message:
Problem signature
Problem Event Name: BlueScreen
OS Version: 6.0.6001.2.1.0.768.3
Locale ID: 1033

Files that help describe the problem
Mini033109-01.dmp
sysdata.xml
Version.txt

View a temporary copy of these files
Warning: If a virus or other security threat caused the problem, opening a copy of the files could harm your computer.

Extra information about the problem
BCCode: 124
BCP1: 0000000000000000
BCP2: FFFFFA8009D58030
BCP3: 00000000B4002000
BCP4: 00000000C0000145
OS Version: 6_0_6001
Service Pack: 1_0
Product: 768_1

I have NO idea what that means. Does ANYone??

Then, the next day, this time I was playing an mp3 via WMP that came loaded with Vista, and I then I opened Firefox and before it could open it went to Blue Screen and restarted.

This is the latest Firefox that I downloaded from the Firefox site, and the same one I had been using with XP (but this was a new download).

I ran a check on Hardware and the Diagnoses all said "No problem".

Today it has worked fine all day. I've run Firefox, I've run WMP, I've run my Quintessential mp3 player, etc. with no problems.

If anyone has any ideas what to do to diagnose this or fix it I'd appreciate it. I called HP and the techs there in India seemed to know less than me. Man, that's scary! I did get through very quickly though. :confused:

SCSIraidGURU · Apr 2, 2009

What are your specs? Are you overclocking? Did you install clean or over another OS?

pdsnickles · Apr 2, 2009

Specs...
Pavilion a6750y Desktop
AMD Phenom X4 9650 quad core processor
8gigs PC2 6400 DDR2 SDRAM
750 gig SATA Hard Drive (7200 rpm)
ATI Radeon HD 3200 graphics with 256mb memory
Is that what you need?

No, I'm not overclocking. I assume that has to do with gaming? I don't know what it is but pretty sure I'm not doing it!

No, Vista 64bit was pre-loaded on my new HP computer.

H2SO4 · Apr 3, 2009

Bugcheck 0x124 is a hardware error, always. The hardware is reporting an unrecoverable event to the OS, and the message is merely being passed along. It's possible to dig out more info about the hardware error report, but since the machine is new and presumably under warranty you _really_ don't want to be fixing the hardware yourself.

Take it back to HP and tell them you deserve better than a machine which is registering machine check exceptions on its second day

pdsnickles · Apr 4, 2009

H2SO4 said:
Bugcheck 0x124 is a hardware error, always. The hardware is reporting an unrecoverable event to the OS, and the message is merely being passed along. It's possible to dig out more info about the hardware error report, but since the machine is new and presumably under warranty you _really_ don't want to be fixing the hardware yourself.

Take it back to HP and tell them you deserve better than a machine which is registering machine check exceptions on its second day

Hi, thank you for that.
Is it certain that the blue screen refers to the HP computer and not to some other hardware - the monitor, the USB hard drive connected, or ?

I really hate to take it back after all the trouble I've gone through to set it up both physically and with software...

But I certainly will consider doing so if you are SURE it's an HP hard disk error.

H2SO4 · Apr 4, 2009

In your \windows\minidump folder there will likely be a few DMP files - one corresponding to each bluescreen. If you copy them to another location, then zip them up and upload the zip here, it may be possible to give you more info regarding the specifics of the hardware error event.

pdsnickles · Apr 5, 2009

Okay, thanks. Here are the 2 .dmp files I found. (attached here)

I appreciate the help.

I will of course request a new computer if this one is defective - but I am hoping it is not because I'd hate to go through that hassle!

By the way it has not happened since the last time, April 1, and I've been using my computer all day every day.

I did download a new driver for my Dell Monitor because I saw some strange display occurences, so I'm hoping maybe that was the problem? (The Dell was a monitor I had and used with XP, and is LCD but about 2 years old.)

My Hardware Diagnostic Test showed Passed on the hard drive surface test, etc.:
Volume Used Space: 158.71 GB Random Seek Test Test Started
4/3/2009 12:30:40 PM Test Finished: Passed4/3/2009 12:31:36 PM Funnel Seek Test Test Started
4/3/2009 12:31:36 PM Test Finished: Passed4/3/2009 12:32:58 PM Surface Scan Test Test Started
4/3/2009 12:35:15 PM Test Finished: Passed4/3/2009 12:43:45 PM Surface Scan Test 2 Test Started
4/3/2009 12:47:54 PM Test Finished: Passed4/3/2009 12:56:15 PM

Pavilion a6750y Desktop
AMD Phenom X4 9650 quad core processor
8gigs PC2 6400 DDR2 SDRAM
750 gig SATA Hard Drive (7200 rpm)
ATI Radeon HD 3200 graphics with 256mb memory
2MB L2 _ 2MB shared L3 Cache memory
Windows Vista Home Premium 64bit

H2SO4 · Apr 6, 2009

Here's the standard Microsoft blurb on the type of crash suffered by your new machine on "Wed Apr 1 14:23:53.802 2009 (GMT+10)":

WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
WHEA_ERROR_RECORD structure that describes the error conditon.
Arguments:
Arg1: 0000000000000000, Machine Check Exception
Arg2: fffffa8009d58030, Address of the WHEA_ERROR_RECORD structure.
Arg3: 00000000b4002000, High order 32-bits of the MCi_STATUS value.
Arg4: 00000000c0000145, Low order 32-bits of the MCi_STATUS value.

In other words, your AMD hardware notified the OS of an unrecoverable hardware fault using a mechanism known as a "Machine Check Exception" (MCE), and it passed along two 32-bit numbers. Each of those 64 bits has a specific meaning, like a bank of 64 toggle switches on an electronic control board. To understand what they mean, first convert the numbers to binary so that individual bits become visible:

3: kd> .formats b4002000
Binary:
10110100 00000000 00100000 00000000

3: kd> .formats c0000145
Binary:
11000000 00000000 00000001 01000101

More info on each of the 64 bits can be found on page 213 onwards of AMD's "BIOS and kernel development" doc:

http://vivian.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/41256.pdf

The end result will not be something friendly and practical like "disconnect that $19.95 SD card reader and It Will Be All Good". Instead, MCEs deliver messages such as this one, corresponding to bit 45 (highlighted red in your bitmask above):

"UECC: Uncorrectable ECC Error. Read-write; set or cleared by hardware. 1=The error was an
uncorrectable ECC error."

ECC is a memory thing. Your hardware appears to be reporting memory problems, and if you decoded the other 63 bits it would probably get down to the specific memory bank that's responsible.

In your situation, I would presonally take the machine back because it's new and I wouldn't want to run the risk of further headaches, even though the root cause may be no more than a stray toenail clipping from a RAM factory technician

You may of course be more adventurous than I am.

pdsnickles · Apr 6, 2009

Thank you very much for the detailed and expert reply.
If I may, can I just ask another question or two?

So this is a RAM error. You are saying it could be caused by many things, and it's hard to say what, right?

Am I correct in thinking it COULD have been caused by my Dell monitor not having the proper driver? (This has been my theory from the beginning, and I'm not sure why other than the fact it gave me problems when I first hooked it up to my old XP machine (a Dell!), but then I worked it out somehow and it never caused problems again. And the fact is, I did get a message from MS Update saying I should update the driver, which I did after this RAM error occured.

And am I correct that this would usually be covered under my 1-year warranty? (I've passed the easy 14 day warranty within which Best Buy allows me to just do a quick switch.)

If I am under warranty then I think I'll give it another month or so to see if I get ANY more blue screens.

Meanwhile I'm going to TRY to get the above DMP files to HP and see if anyone there is as knowledgeable as you about them and can advise me one way or another, from their point of view. At least I'll have the info on record, then.

I actually did call them and was - foolishly - hoping that I could get someone to analyze the logs, as you did, and figure out what the blue screen was about.

Instead I got a dumb newbie Indian girl whose very first words about my problem were "We can do a System Restore". I said, "Well, I know that. I called you because I am looking to find out WHY my computer crashed and perhaps prevent it from happening again." Then I eventually got to a Supervisor (She told me I HAD to let her try some things first.)(She had me boot into Safe mode and then run WMP to check if it was working properly - only she quickly found that WMP does not run in Safe mode! - Great suggestion!)

So then the Supe came on the line and he also had no clue as to what my problem was, but blamed in on software that I had added, even though I assured him that it was all software I'd been using with XP and that is said to work with Vista64bit.

So, I don't have much hope for getting this problem recognized by HP but I'll write them and send them the files along with your explanation if you don't mind. I won't use your handle just say you were an expert on a support forum. Youre certainly way more of an expert than the average HP tech support employee, I can assure you!

Thanks again.

H2SO4 · Apr 6, 2009

pdsnickles said:
Am I correct in thinking it COULD have been caused by my Dell monitor not having the proper driver? (This has been my theory from the beginning, and I'm not sure why other than the fact it gave me problems when I first hooked it up to my old XP machine (a Dell!)

This wasn't a driver (software) problem. Let me try an analogy: you drive your car and fill it with gas, check its tyre pressures, keep the coolant topped up - all the normal things. When the level of liquid in the tank falls, you get a warning light on the dash. If you leave your lights on, the car may beep at you. Let's call these "software" warnings. Now, if you're real unlucky, one morning the car completely refuses to start and it has to be towed. The mechanic hooks up one of those new-fangled electoronic diagnosis units to the special socket, and it tells him that the problem is error code ABC123XYZ, which he looks up in his little booklet provided by the car manufacturer to find out that the manifold pressure has dipped below some threshold of safety. That would be a "hardware" error report.

A Machine Check Exception (MCE) is in this latter category. Literally nothing that you could do to the machine, short of outright abuse like poking the motherboard with a stick, would by itself produce an MCE on otherwise healthy hardware. You could find the ugliest OS ever written, fill it with the all-time-buggiest apps, use nothing but the oldest drivers, sprinkle liberally with malware - and it STILL would never produce an MCE because all of that would be at the software layer and an MCE is hardware

pdsnickles said:
And am I correct that this would usually be covered under my 1-year warranty? (I've passed the easy 14 day warranty within which Best Buy allows me to just do a quick switch.)

Yes, absolutely. That dump constitutes proof positive of a hardware error report, if you can get it to the right tech who understands the MCE mechanism and the implications.

pdsnickles said:
So, I don't have much hope for getting this problem recognized by HP but I'll write them and send them the files along with your explanation if you don't mind. I won't use your handle just say you were an expert on a support forum. Youre certainly way more of an expert than the average HP tech support employee, I can assure you!

Thanks again.

No problem

They probably won't give a flying fig what some "expert" said, but this Wikipedia link might be useful to you in your attempts to get through to a more senior tech:

http://en.wikipedia.org/wiki/Machine_Check_Exception

"A Machine Check Exception (MCE) is a type of computer hardware error that occurs when a computer's central processing unit detects an unrecoverable hardware problem."

As a suggestion, check out that "mcat" tool from the link at the bottom of the page. It apparently does an automated analysis of the MCE report data which you'd have in your event logs. Another bit of evidence to use with the senior HP techs, should you need it.

pdsnickles · Apr 7, 2009

H2SO4 said:
No problem

They probably won't give a flying fig what some "expert" said, but this Wikipedia link might be useful to you in your attempts to get through to a more senior tech:
Machine Check Exception - Wikipedia, the free encyclopedia

"A Machine Check Exception (MCE) is a type of computer hardware error that occurs when a computer's central processing unit detects an unrecoverable hardware problem."

As a suggestion, check out that "mcat" tool from the link at the bottom of the page. It apparently does an automated analysis of the MCE report data which you'd have in your event logs. Another bit of evidence to use with the senior HP techs, should you need it.

Thanks... I guess...

I guess it's better that it happen now than after I REALLY have it all set up with ALL my software and all my preferences and so on and so forth....

Tonight I spent 2-3 hours looking at my Windows Logs under System to see if I could see anything in there... I found TONS of errors and warnings... Not sure what most of them mean... but apparently I've got some TCP-IP issues, a Kerberos error, a DCOB issue and Norton keeps trying to start EVEN THOUGH I UNINSTALLED IT within 5 min. of being up and running the first time! So I've got tons of errors, over and over, in those logs. I'm not sure how much if any is related to the above hardware problem and how much is related to common Vista problems or etc... and I've never looked in logs like that before so I don't know what is normal but it looks pretty severe to me. (see attached)

Okay, so I guess tomorrow I'll have to call HP and see if they can get me a new computer... I wish I'd have hooked up my computer right away so it would still be under Best Buy's warranty but I am 4 days past that. Oh well.

I guess "s**t happens" and this is a case in point.
No point in sticking it out with bad hardware.
Thanks for sticking with it and convincing me that I need to do return it.

attached are some jpg's of various errors - just because I collected them and you might be curious...

H2SO4 · Apr 7, 2009

EventID 19 is the interesting one in this context - WHEA Logger. If you run that mcat utility and feed it those event logs, it should spit out its own analysis.

I'd suggest not getting too caught up in attempts to "diagnose" the hardware though. Once you make the decision that you're not willing to live with having to wonder whether each and every niggling fault from now on is the result of some lurking hardware condition (and it might be!), just try to push through to someone senior at HP and tell them the machine has logged multiple MCEs. It's there in your event logs, after all. You don't really care why. You just want a machine that doesn't do it.

pdsnickles · Apr 7, 2009

H2SO4 said:
EventID 19 is the interesting one in this context - WHEA Logger. If you run that mcat utility and feed it those event logs, it should spit out its own analysis.

I'd suggest not getting too caught up in attempts to "diagnose" the hardware though. Once you make the decision that you're not willing to live with having to wonder whether each and every niggling fault from now on is the result of some lurking hardware condition (and it might be!), just try to push through to someone senior at HP and tell them the machine has logged multiple MCEs. It's there in your event logs, after all. You don't really care why. You just want a machine that doesn't do it.

Okay I'm going to call HP a bit later on today.

But I ran that MCAT program I got off the wikipedia page you recommended, but the .pdf that came with the download says it doesn't work with my AMD "Phenom" processor; that is, it lists it working only with the Athlon and another one, not Phenom.

I tried the program anyway but once I got to the command prompts I couldn't figure out how it was telling me to proceed. I got the attached screenshot and didn't know how to proceed from there.

My question is : when I call HP what do I point them to? The .dmp files or the .evtx (event log) files?

And you originally said it was a RAM problem (as I recall) but later implied that it is a processor problem. Do you know which it is, so I can explain it the right way to HP ?

pdsnickles · Apr 7, 2009

Follow up:
The following exchange took place at approximately 7:50pm Tuesday.
I'd already spent 2 hours and 15 minutes (still counting) online with HP chat support.
I gave them all the info and finally referred them here so they could look at the diagnostics you did already.
After 2 hours this is the exchange that took place. Mind you, nothing was done except me doing a hardware diagnostic (the processor passed all tests).

"HP Help : We can try a few steps and if that does not resolve the issue we can look for other options.
Me: Okay, I have other things to do. Let's get to it.
HP Help: Thanks a lot.Sir, please open command promt and type in chkdsk /
HP Help: because the kind of log that you have sent me , I see that there are some issues in the hardware , so we should try to fix any conflicts and IRQ requests between the various hardware devices.
Me:It says that it cannot be run because the volume is in use by another process...???The hardware error relates to the PROCESSOR, not other hardware. That is what the man said on the site. Do you have other information?
HP Help: sir, the processor sends some IRQ requests etc to other hardware devices, and the log suggests me that theres a conflick in such IRQ requests."

So that's where I stand after all this time online with HP. I will fill in the end part of this call later... I knew this would be difficult but didn't expect it to be like this. I am very disappointed in HP. I knew a time when they stood behind their products and would replace one quickly if it was not running right within the warranty period.

ArbiterAlan · Apr 7, 2009

H2SO4 said:
3: kd> .formats b4002000
Binary:
10110100 00000000 00100000 00000000

3: kd> .formats c0000145
Binary:
11000000 00000000 00000001 01000101

Sorry for sidetracking here but I was just wondering how you did that binary conversion. I've only learned how to convert IP addresses like 10.1.10.254 to 00001010.00000001.00001010.11110101

H2SO4 · Apr 7, 2009

I think it's safe to say that the person making those comments does not have the background or training necessary to understand what is happening. I would suggest not spending too long with them on the phone because that particular person cannot help you, and that simply leads to frustration.

Personally, my inclination would be to take it back to the store where I bought it. When a Ford or a Toyota malfunctions during its warranty period, it should be taken back to the dealer where it was purchased, not directly to Mr.Ford or Mr.Toyota themselves

I'd also suggest not getting caught up in "troubleshooting" discussions with their staff over whether the processor or the memory, or whatever is responsible. (The actual info in your case suggests to me that it's the memory, but that's almost irrelevant. The Wikipedia sentence I quoted just makes it sound like the 'CPU' is the conduit for the bad news, which is technically true.) Otherwise, if you let them talk "tech" at you, chances are you won't be able to counter their assertions - mostly because they're made up of words they don't really understand in a sequence they thought would sound convincing, like the phrases you quoted above.

In your position, I'd walk back into the store where I bought the computer and say:

- This hardware has registered multiple MCE events during its first days of operation.

- MCE events come from hardware, not software.

- The evidence is there in the form of MCE information recorded in the system's event logs and the memory dump generated by Windows in response to the MCE event.

- I'd like a machine that doesn't do MCEs please. Feel free to troubleshoot the intricacies of the exact error condition once you've supplied me with replacement hardware so that I can get on with the actual activities for which I purchased this computer.

No negotiations, no pseudo-technical helpdesk BS

pdsnickles · Apr 7, 2009

H2SO4, thank you very much for sticking with me through this. You are right, it was VERY frustrating spending 3 and 1/2 hours with HP tech support and getting essentially nowhere except MAYBE it will get kicked up the ladder. (He said it would, but...?)

If ONLY I could take the computer back to Best Buy! THAT would be soooo much easier!

I called them tonight and was told that they are very strict on their 14 day return policy, which I am now 4 days beyond. And though it makes total sense to you and I, if I take a WORKING computer in there, they are going to say, "Huh? You mean it's working fine but you want for us to bend our strict 14 day policy to replace it???"

I will try going back in there tomorrow and talking with the manager but I don't expect to get anywhere, except MAD like I am right NOW.

So my next question is, how do I get to the right person at HP to recognize this problem or at least give me the benefit of the doubt and give me a new computer????

I now am on a mission but I don't know exactly how to proceed. I'm going to poke around the net and try to find someone at HP Corporate who might not want bad publicity and who can recognize that if they don't take care of me they are going to get just that, all over the net and by word of mouth!

H2SO4 · Apr 8, 2009

Let me get something out of the way first: I'm a private individual helping out on a web forum for the sheer personal enjoyment I derive from it. I can't give you insight into the inner workings of any company, nor legal advice, nor anything official along those lines. All I can do is to explain how I would react in the same situation, in the hope that it may help you in some way.

FACT: An MCE is a hardware error. Intel says so:

http://download.intel.com/design/processor/manuals/253668.pdf

"14.1 MACHINE-CHECK ARCHITECTURE
The Pentium 4, Intel Xeon, and P6 family processors implement a machine-check architecture that provides a mechanism for detecting and reporting hardware (machine) errors, such as: system bus errors, ECC errors, parity errors, cache errors, and TLB errors."

The machine in question registered multiple MCEs during its first month of operation. The observation that it appears to work normally most of the time is irrelevant - it was purchased with the expectation that its hardware would be 100% functional, not 99%.

Nothing has been repaired at a hardware level since the MCE events, and therefore there is nothing to stop the same underlying hardware error condition from manifesting itself in the future, either as further MCE instances, or in more subtle ways which might lead to data integrity problems while being difficult to detect.

The warranty covers the machine against hardware defects during a certain period from the time of purchase, and this machine has already logged several hardware error (MCE) events during the warranty period. I therefore feel entirely justified in asking for the hardware to be replaced so that I can get on with my work and put this massive and troublesome invonvenience behind me.

==================

If you're having trouble verbally convincing the friendly folks at the store where you bought the machine to give you a replacement, try perhaps writing something along the lines above. It's under warranty and you've got a right to a product that works without defects. You're not the one on thin ice here. They are.

H2SO4 · Apr 8, 2009

ArbiterAlan said:
H2SO4 said:

3: kd> .formats b4002000
Binary:
10110100 00000000 00100000 00000000

3: kd> .formats c0000145
Binary:
11000000 00000000 00000001 01000101

Click to expand...

Sorry for sidetracking here but I was just wondering how you did that binary conversion. I've only learned how to convert IP addresses like 10.1.10.254 to 00001010.00000001.00001010.11110101

Nothing wrong with asking, but you should start another thread in the future. This one is sufficiently verbose as is.

The answer to your question is "I typed '.formats b4002000' into a debugger, pressed enter, and it did the conversion for me"

Actually, hex-to-binary conversions are much simpler than decimal-to-binary, which you already know how to do. (That's a side effect of the fact that 2^4=16 while base10 numbers have no such affinity for any power-of-2 multiple.) Each digit in a hexadecimal number can be converted by itself, which becomes a trivial task, and the results are simply concatenated in the same order as the original number. Use 0xb4002000 as the example (that 0x prefix is a "radix" which signifies that the number which follows is hexadecimal):

Hex: B 4 0 0 2 0 0 0
Bin: 1011 0100 0000 0000 0010 0000 0000 0000

In other words, 0xB is 1011 in binary, 0x4 is 0100 in binary, 0 is always 0, and so on. Read that entire list of binary numbers above from left to right, and if I've done my text formatting right you'll see it's the same number that the debugger spat out at me when I asked it to convert 0xb4002000.

ArbiterAlan · Apr 8, 2009

Thanks H2 (I just noticed I did 245 instead of 254, DOH!)

PDSnickles - I've worked in technical support myself (Not for HP).
A few things that may help you:
- They will want to start troubleshooting right off the bat, it's what they're trained to do and it can take a minute to get them past that mode.
- If you stick to saying "The PC has logged multiple Machine Check Errors already and this indicates a hardware failure. I have not modified any factory hardware settings, and I am not running any stress testing software to cause this error and the machine needs to be replaced." (You haven't overclocked the system or installed any hardware tweaking software - If you have, that problem may be yours.)
- If they insist on continuing to troubleshoot ask to speak to a supervisor - The worst they can do is say no and you can hang up to try the next rep. Hopefully You'll get a rep that knows hardware failures on a new PC are not correctable over the phone (unless you've tweaked the system - then you should undo the changes) and they will setup a replacement for you, if not a supervisor will.
- If you are selecting prompts when you call in do not select the option for trouble with your system if they have an option for checking the status of a return, or a warrenty department - these departments may be more likely to set up a replacement for you right off the bat.
- Try not to get mad at the rep on the phone (yes even the really stupid ones), it will make them more resistant to your requests. They are just getting paid to do what they are doing, they are not the company you bought the PC from.
Good luck with them, I hope you don't have to spend too much longer on the phone with them.

Blue Screen twice - Any ideas why?

My Computer

New Member

My Computer

My Computer

A bit of a numpty

My Computer

My Computer

A bit of a numpty

My Computer

Attachments

My Computer

A bit of a numpty

My Computer

My Computer

A bit of a numpty

My Computer

Attachments

My Computer

A bit of a numpty

My Computer

Attachments

My Computer

My Computer

New Member

My Computer

A bit of a numpty

My Computer

My Computer

A bit of a numpty

My Computer

A bit of a numpty

My Computer

New Member

My Computer