irregular BSOD Bugcheck 124 (0x00000124) - hal.dll win7 64bit

[gazzawazza]

hi all

I'm struggling with an irregular, intermittent bugcheck 124. I would be VERY appreciative of any help.

These are my full specs:

Windows 7 Home Premium 64-bit SP1 - all current / appropriate patches
Intel Core i5 4670K @ 3.40 – Intel Turbo mode ON
Noctua NH-U12s
Gigabyte Z87-D3HP – BIOS F6
2 x 4GB Corsair Dual-Channel DDR3 @ 799MHz (9-9-9-24) - "CML8GX3M2A1600C9"
1279MB NVIDIA GeForce GTX 570 (MSI)
1863GB Seagate ST2000DM001-1CH164 ATA Device (SATA) - firmware CC29
2794GB Seagate ST3000DM001-1CH166 ATA Device (SATA) - firmware CC29
TSSTcorp CDDVDW SH-224DB ATA Device
Realtek (onboard) High Definition Audio

EDIT: oops forgot these:

BeQuiet BN194 Straight Power E9 700W
Edimax N600 Dual Band Wireless Adapter PCIe (EW-7722PnD)
TSSTcorp CDDVDW SH-224DB ATA Device (optical drive)


Bluescreenviewer reports that likely fault points are ntoskrnl.exe and hal.dll.

I've had 12 crashes of exactly the same nature, since Feb 2014. This PC build has been in operation for only a wee bit longer than that i.e. I've had the crashes since pretty much the beginning. I've had no other STOP errors.

I'm an ex IT engineer and am reasonably comfortable with dealing with problems. This time however I simply can't find a pattern of incidence.

I've had BSODs within 12 hours of each other, yet the last crash was over a month since the previous one.

The only crashes I've physically witnessed have been when I've been watching a streamed video in my browser (that's 3 out of the 12). I tried actually disabling hardware acceleration within Flash, in case it was my gfx card being awkward but have had a crash since doing this. Also, I've never had a crash while gaming.

On one or more occasions I've simply woken up to find the PC restarted at a bugcheck 124 in the event log - I'm pretty sure that the only thing that was running on the PC before I went to sleep was my BOINC client. I can't say for certain whether I'd suspended GPU crunching (I normally do this, so my gfx card fan doesn't disturb me).

IMO my system, CPU & GPU temps seem in order. My max temps in stress tests have been 71 (in prime95) and 73 (reported in the Intel CPU tester).

I'm using Open Hardware Monitor to keep an eye on temps.

Can't tell you offhand what my GPU hits but I think it's around 80 degrees when using MSI Kombustor.


Things I've tried to pinpoint the problem:


PRIME 95
I stress tested for maybe 3 hours each on “blend”, “In-place large FFTs” and “small FFTs”. No problems.

I used v27.9 build 1.

I have recently discovered that v28 has come out.

********************************
- I re-ran using v28.5 build 2 a couple of days ago.
- Setting was “custom” with all normal settings (8-4096K FFT) but using a total of 90% of my available memory (5.5GB).
- I ran for 5 hours with no problems.
- Maximum Core temp was 71 degrees.
- Only odd thing was that the fastest worker (core 1) was 8 minutes in front of slowest core (core 4)
********************************


DISK CHECKS
I've done full chkdsks on both my drives (the long check - verifying free used space). No problems there.


MEMORY CHECKS
I ran the native windows memory tester and also MemTest 86. No errors. However, I did not run test for very long (maybe an hour).

I re-ran using MemTest 86 5.1.0 last night. It ran for nearly 4 hours and made approx 7.5 passes (without error) before I exited. All tests run and I set all CPU cores to run in parallel.


INTEL CPU DIAGNOSTIC
I have run the intel processor diagnostic tool (I've been speaking with Intel support, in case it's the CPU).

I let the tool run for the default one pass (the burn-in option runs for 120mins). It passed, although I got a warning that my system was overclocked (which I knew). I have the default Intel turbo mode enabled.


DUST LEVELS
I've got dust filters that I clean every one to two weeks and I probably every month to 6 weeks have the case-side off and give the system a quick check / blow-out with compressed air. I always check the CPU cooler heatsink and fan. I also always give the graphics card a blow-out.


RESEATING
I've reseated, I'm pretty sure, the SATA cables. I've definitely fully removed my graphics card (and disconnected the power cables). I've reseated the RAM sticks too. I've sadly had crashes since.


DRIVERS
All drivers should be current – I have used drivers via Intel’s chipset device software / INF update utility. Other drivers have been sourced from e.g. Nvidia, Realtek, Intel or windows / Microsoft.

I've also used the native driver verifier. I ran it for 48 hours without seeming issue, as per this forum's recommendations (omitting the following: Low Resource Simulation, Force Pending I/O Requests & IRP Logging). I ran it on all non-microsoft drivers (including unknowns).


BIOS
I'm using the latest stable BIOS (F6). There's a beta 7c published but I don't like the idea of using a beta BIOS (unless instructed to and/or absolutely necessary).

Everything in the BIOS is pretty much set to auto, except the memory profile, which is set to XMP. I have had my crashes with both normal and XMP profile selected.

I had actually hoped that the XMP profile might slip a wee bit extra voltage through, just in case the RAM was hungry - I do know btw that there's a clear problem with gigabyte boards and using all 4 DIMMs. I've only got 2 sticks active.

I'm pretty sure I pulled the CMOS battery after flashing the last BIOS, so we should be working with 'clean' data.


EVENT LOG
I'm not an expert but I couldn't really see any pattern to any problems immediately prior to the crashes.


O/S FILES
I've, literally in the last few days, run SFC. No problems reported.


Things I've not tried:

• Running in safe-mode - I didn't feel it feasible to do this, since the crashes can be well over a month apart.
• RAM compatibility - at the suggestion of my parts reseller, I checked gigabyte for RAM compatibility on my mobo. My RAM is not listed but (A) the list is very out-of-date (June 2013) (B) the list openly acknowledges that there are only so much RAM that can be tested (which seems fair enough) (C) I've had a little experience with RAM failure and these crashes don't really match up with my previous experiences. It would be great if it was - common problem, as I understand it, and easily fixable.

************************************

EDIT:

Other problems with system

I've had some key LED failures on my Ducky keyboard.

I know it's perfectly possible for components to fail like this (bathtub effect) but for a premium product, I'm a little surprised. Not that Ducky would admit to the contrary but they claim it's extremely unusual for this to happen.

I had 1 key LED fail within 3 months, got a new keyboard from reseller, then have had another 2 failures on the new (identical) keyboard within 2.5 months, so RMA'd the keyboard to Ducky, got it repaired, only for another one to faile very recently.

I only mention this, just in the off-chance this is a pointer to the problem. For example, the only way I could think of the LEDs failing would be some kind of power surge but would more not go in this scenario? So could it be the PSU? Or maybe some bizarre fluctuation in mobo power distribution? Or perhaps just a single dodgy USB port? Or maybe a duff batch of LEDs?

Equally well, it could just be a crappy keyboard. Maybe it's just coincidence but you can't buy this keyboard any more (at least with the blue LEDs). Maybe this is just due to them not selling well, which I find strange though cos the LED colour looks a bit like colours under UV - sorta an intense mid blue colour (pretty cool I think lol).

************************************

I think that's about it for now.

I've used your dm log app but wasn't sure how to 'present' the output?

Thanks for anything you can do for me.



Regards,

Gary

 

[HarriePateman]

This a very nice descriptive post, thank you. Its not often we see these!

However please can you post your DMP Files, as this will help us furthermore to sort our you problem.

http://www.sevenforums.com/bsod-help-support/96879-blue-screen-death-bsod-posting-instructions.html

 

[gazzawazza]

This a very nice descriptive post, thank you. Its not often we see these!

However please can you post your DMP Files, as this will help us furthermore to sort our you problem.

http://www.sevenforums.com/bsod-help-support/96879-blue-screen-death-bsod-posting-instructions.html

Thanks very much Harrie!

As said, I've got an IT background and try to be methodical - glad you feel it was constructive.

Uploading logs as we speak.



Regards,

Gary

 

[HarriePateman]

Brilliant i will look into this for you.

 

[gazzawazza]

Brilliant i will look into this for you.

Thank you very much.


Gary

 

[HarriePateman]

Okay so we have a hardware error: 0x124

I was check your CPU and i noticed this:

3: kd> !cpuid
[U][B][COLOR="Red"]Unable to get information for processor 0
Unable to get information for processor 1
Unable to get information for processor 2[/COLOR][/B][/U]
CP  F/M/S  Manufacturer     MHz
 3  6,60,3  GenuineIntel    3400

You mention you have Turbo mode active, which is fine and it shouldn't cause any damage otherwise Intel wouldn't place it there!

However it is strange that your DMP Files are only detecting one of your four cores.

We know that the temps are fine, so thats not an issue!

Can i recommend you disable Turbo mode, and we get another BSOD then we know that Turbo mode is not the cause here.

Im going to continue looking into your DMP Files further for you.

 

[gazzawazza]

Okay so we have a hardware error: 0x124

I was check your CPU and i noticed this:

3: kd> !cpuid
[U][B][COLOR=Red]Unable to get information for processor 0
Unable to get information for processor 1
Unable to get information for processor 2[/COLOR][/B][/U]
CP  F/M/S  Manufacturer     MHz
 3  6,60,3  GenuineIntel    3400

You mention you have Turbo mode active, which is fine and it shouldn't cause any damage otherwise Intel wouldn't place it there!

However it is strange that your DMP Files are only detecting one of your four cores.

We know that the temps are fine, so thats not an issue!

Can i recommend you disable Turbo mode, and we get another BSOD then we know that Turbo mode is not the cause here.

Im going to continue looking into your DMP Files further for you.

Fascinating but also bizarre.

I'll do a reboot and set Intel Turbo accordingly.

All the system spec software I've used all detect the cores. Having said that, they might fill in the detail on basis of CPU name or perhaps the CPU name contains the core info - stuff like speccy reports all 4 cores individually, open hardware monitor does too, as does your seven forum's SystemInfo app.

I assume that .dmp files normally correctly report CPU cores?


Regards,

Gary

 

[HarriePateman]

Yes normally, DMP are quite accurate.

 

[gazzawazza]

Yes normally, DMP are quite accurate.

Any idea where the DMP files draw their info from? Could it be that the mobo isn't reporting the info properly?

 

[HarriePateman]

When a BSOD happens, the information is written directly into a DMP. I would think its highly doubtful that your DMPS are incorrect, as its not that kind of STOP code.

Hopefully we shouldn't receive another BSOD but if we do then we know its not Turbo Boost causing the problem.

 

[gazzawazza]

When a BSOD happens, the information is written directly into a DMP. I would think its highly doubtful that your DMPS are incorrect, as its not that kind of STOP code.

Hopefully we shouldn't receive another BSOD but if we do then we know its not Turbo Boost causing the problem.

okey dokey.

Just confirming that open hardware monitor reports all cores at 3400MHz


Ta.

Gaz

 

[HarriePateman]

Brilliant.

 

[HarriePateman]

Okay i have asked some of the other guys to come and take a look for you, as most of my knowledge goes in to BSOD's and Developing.

 

[gazzawazza]

Okay i have asked some of the other guys to come and take a look for you, as most of my knowledge goes in to BSOD's and Developing.

Ok. Np.

I take it then that nothing else has leapt out at you?

Did you see my initial edited comment about my Ducky keyboard having repeated key LED failures?

I know it seems completely unrelated but it's the only other thing happening that seems unusual enough to be unexplained.


Thanks,

Gaz

 

[HarriePateman]

Okay i have asked some of the other guys to come and take a look for you, as most of my knowledge goes in to BSOD's and Developing.

Ok. Np.

I take it then that nothing else has leapt out at you?

Did you see my initial edited comment about my Ducky keyboard having repeated key LED failures?

I know it seems completely unrelated but it's the only other thing happening that seems unusual enough to be unexplained.


Thanks,

Gaz

Ahhh brilliant you should mention that! im going to try an explain this as easy as I can.

When Turbo mode is active it Causes an unstable system (in your case)

Judging by your specs, you have quite a beast there!

That's a lot of power your system is pulling.

When a system is low on power things start to become unstable, the same as if you are overclocking.

The LED'S on your keyboard were flickering because there was a shortage of power to power them.

The same for your CPU, its lacking in power, that's why we received a "Fatal hardware Error"

There is nothing wrong with your CPU, however they May be power dips causing it to become unstable.

Resulting in a BSOD.

Unfortunately PSU's are so hard to diagnose...

 

[gazzawazza]

Okay i have asked some of the other guys to come and take a look for you, as most of my knowledge goes in to BSOD's and Developing.

Ok. Np.

I take it then that nothing else has leapt out at you?

Did you see my initial edited comment about my Ducky keyboard having repeated key LED failures?

I know it seems completely unrelated but it's the only other thing happening that seems unusual enough to be unexplained.


Thanks,

Gaz

Ahhh brilliant you should mention that! im going to try an explain this as easy as I can.

When Turbo mode is active it Causes an unstable system (in your case)

Judging by your specs, you have quite a beast there!

That's a lot of power your system is pulling.

When a system is low on power things start to become unstable, the same as if you are overclocking.

The LED'S on your keyboard were flickering because there was a shortage of power to power them.

The same for your CPU, its lacking in power, that's why we received a "Fatal hardware Error"

There is nothing wrong with your CPU, however they May be power dips causing it to become unstable.

Resulting in a BSOD.

Unfortunately PSU's are so hard to diagnose...

Sorry to say this but the LEDs actually failed (they weren't flickering and I never saw any behaviour of that sort).

Also, I really would have thought that a 700W PSU would be enough for my system.

I am content to be told otherwise.

After all, I'm only using one GPU - it's not as if I've got SLi or crossfire. Also, I'm only mildly overclocking and the overclocking isn't constant.

I've actually just run BeQuiet's PSU calculator and the top PSU it recommends for my setup is 680W, so I really should be ok.

The calculator allows for:

specific CPU
specific GPU
nos of HDs drives (good selection of types here)
nos of optical drives (good selection of types too)
USB devices (just nos of devices not types)
PCI devices (limited options here - for example, I couldn't select NIC)
fans (just nos and what sizes)

Other factors considered:

water cooling (not selected)
expandability (selected)
overclocking (selected)

EDIT:

*********************************

I've tried a couple of other PSU calculators (newegg and Outervision - a couple mentioned on a Tom's Hardware post).

The NewEgg one was basic but came in at 540W

The outervision, which allows for wattage on basis of specific overclocking, came in at 507W (100% load)

*********************************

I just can't t believe that I'm running out of power. However, this doesn't exclude mobo power management issues (in line with Intel Turbo management) or PSU hiccups or hardware related mobo issues.

3 of the crashes have happened when I've only been watching a movie - that really doesn't take much juice.

I've also never had a crash when gaming, which is when you'd expect high power usage (due to graphics card being used hard - I've noticed that my card, according to speccy, has 3 modes, which involve GPU frequency throttling, so again power consumption is variable due to power-saving measures).

I'd have thought high power usage scenarios (e.g. stress tests) would be more likely to provoke crashes?

I suppose if I ran a combined CPU and GPU stress test, that would simulate very heavy loading?




Regards,

Gary

 

[HarriePateman]

Psu's can also malfunction at a low Power...

As this can make them un-stable....

Have you got another PSU that you can swap out and try?

 

[gazzawazza]

Psu's can also malfunction at a low Power...

As this can make them un-stable....

Have you got another PSU that you can swap out and try?

Hi Harrie

I might have been able to use my old PSU (except it probably would have been underpowered, so would have created another set of problems).

Anyway, unfortunately I've had a disaster with my mobo BIOS and the mobo looks like it's bricked.

The most frustrating thing is that I think it was caused by the backup BIOS chip being defective.

Gigabyte's dualBIOS feature allows you to swap your main BIOS image with your backup image (in theory preserving both). I wanted to roll back to the last stable mobo BIOS, having tried their beta BIOS out (which was ok but I wanted to do a test with the stable BIOS on some settings).

You can't go back to an earlier BIOS image in the same way though as flashing a newer one - you are just told you can't do it by Q-Flash and @BIOS (gigabyte’s windows based bios update utility).

So, I thought I can use the gigabyte backup BIOS chip & its image. The swap of good/main BIOS image and older backup image seemed to go fine but on reboot, I couldn't get system to post. I couldn’t even get into the BIOS settings (continuous rebooting).

Nothing seemed to work except when I managed to trigger the backup BIOS image (which should have been the ok beta bios) to overwrite the main BIOS. This has been the only time any part of the BIOS seemed to work. However, after the copy process completed and the system restarted, it still didn't boot / went into a continuous boot loop.

So it seems like the initial image swap / write might have failed. Either that or the more likely explanation (to my mind) is that the backup BIOS chip is damaged, so both the old and the new image got corrupted when they were written to it (thus meaning that a corrupt BIOS image got copied to the main BIOS chip).

I've remembered to clear the CMOS (I've shorted the appropriate jumper pins AND pulled the battery). Still repeatedly attempts and fails to post.

I've tried pulling ALL peripherals off, using just the onboard graphics (in case the graphics card isn’t properly recognised), running on 1 stick of RAM, no drives, etc... in an attempt to get to a stable BIOS screen. Nothing works now.

I'm furious.

I've even had to use dualBIOS on an older gigabyte mobo and it worked fine - the whole point of literally swapping the 2 images over is so you don't get left with 2 duff ones... yet I think I've now got 2 corrupt images and no way to flash, as far as I know, a good image (that I've got on a USB stick) onto the BIOS.

I’ve flashed firmware plenty of times and it’s never gone wrong, until now. Just can’t think of anything else I can try.





Gary

 

[HarriePateman]

Ohhh no Have you pulled the CMOSS Battery?

 

[gazzawazza]

Ohhh no Have you pulled the CMOSS Battery?

yep.