[SOLVED] BSOD SYSTEM_SERVICE_EXCEPTION Ntoskrnl.exe

[cloudropis_1]

For a long while I've been having BSODs like this on my machine. All generic BSODs like this one, or FAULTY_PAGE_CORRUPTED_HARDWARE, or others I've forgotten but that googling always lead to the same results: basically, they were so generic they are attributable to everything that can go wrong in a PC: corrupted OS, bad drives, bad RAM, outdated drivers etc. Now I'm finally bruteforcing this issue in order to fix it, as it made playing games (the instances where I get these BSODs) kind of a pain.

• OS should be fine, I've run /sfc and dism multiple times
• drives should be fine, I control them with CrystalDiskInfo pretty often. One of my HDDs is three years old, but the other one is brand new and I still get the same problems with stuff installed on it
• temperatures are great as I live in a well climated region and the PC gets amazing airflow

Now I'm gonna update every driver I can, use DDU to clean my GPU drivers and reinstall, and run memtest/the Windows diagnostic equivalent tool. In the meanwhile, this is the .dmp of the latest crash. Googling this issue I've seen some users in this forum examining them in-depth through a debugger, that would be really appreciated.
https://1drv.ms/u/s!AszPI4pVTqe0gUFLMjWtvSUJbRm3?e=WRcbur

I've had these BSODs since building this PC three years ago, they were never so frequent I had to pay them much attention, but now it's getting a bit ridicolous

 

[PC Tailor]

I have ran a debugger on your DMP file and you can view the full report here: https://pste.eu/p/lBua.html

Summary of Findings:

BugCheck 3B, {c0000005, fffff80264d5c4e2, fffff08516337b40, 0}
Probably caused by : win32kfull.sys ( win32kfull!Win32FreePoolImpl+52 )

SYSTEM_SERVICE_EXCEPTION (3b)
An exception happened while executing a system service routine.

Arguments:
Arg1: 00000000c0000005, Exception code that caused the bugcheck
Arg2: fffff80264d5c4e2, Address of the instruction which caused the bugcheck
Arg3: fffff08516337b40, Address of the context record for the exception that caused the bugcheck
Arg4: 0000000000000000, zero.

DEFAULT_BUCKET_ID: WIN8_DRIVER_FAULT
PROCESS_NAME: WWE2K19_x64.ex
MODULE_NAME: win32kfull
IMAGE_NAME: win32kfull.sys

A memory access violation occurred. (Parameter 4 of the bug check is the address that the driver attempted to access.)

System Service can often be driver related, however you say you have had other BSODs? Various BSODs can be indicate of hardware, especially if you have hardware based bugchecks occurs, such as CORRUPTED HARDWARE or WHEA errors.

Do the BSODs occur under certain circumstances? Do they occur in safe mode?

From the report, just consider the following:

• The Scarlet Crush driver has caused multiple BSOD problems (many of which myself and other colleagues have debugged) - I would remove this. You can also use Autoruns to disable it.
• Virtual boxes can cause various problems
• Your BIOS is out of date - the latest official version is F9 - GB BIOS Link Page
• Depending on your responses to above, as with any memory corruption, it is best to run memtest to verify the integrity of your memory modules.

 

[cloudropis_1]

I have ran a debugger on your DMP file and you can view the full report here: https://pste.eu/p/lBua.html

Summary of Findings:


System Service can often be driver related, however you say you have had other BSODs? Various BSODs can be indicate of hardware, especially if you have hardware based bugchecks occurs, such as CORRUPTED HARDWARE or WHEA errors.

Do the BSODs occur under certain circumstances? Do they occur in safe mode?

From the report, just consider the following:

• The Scarlet Crush driver has caused multiple BSOD problems (many of which myself and other colleagues have debugged) - I would remove this. You can also use Autoruns to disable it.
• Virtual boxes can cause various problems
• Your BIOS is out of date - the latest official version is F9 - GB BIOS Link Page
• Depending on your responses to above, as with any memory corruption, it is best to run memtest to verify the integrity of your memory modules.

Thanks a lot of the help so far.

Yes I've got plenty of other BSODs over the years. The first kind is the "while playing games" kind. They were pretty inconsistent, sometime I wouldn't get one for weeks and weeks, others I'd get three in a week. I can't remember the error messages, but a lot of them were about ntoskrnl.exe when examined with BSV. "Video memory management internal" was an old classic, I've actually found old threads of mine about this from another forum:

Spoiler

https://www.eightforums.com/members/cloudropis.73543/#recent-content

The second kind is "sometimes when I wake up the machine from Sleep mode it crashes after thirty seconds of normal use", you can actually see a thread about this from this forum if you check my user history. From my research that's just W10 acting up.

1. Damn, Scarlet Crush is the driver that lets me use my Dualshock 3 on my PC. It is an older version though, from when W10 wasn't even out, I new one popped out when it was. Maybe I should uninstall the old one and update to the most recent one? It's pretty important for my PC, I'd avoid completely getting rid of it unless there are absolutely no other ways. Do you have any idea what the best way to uninstall it is? I've never had to manually unistall a driver.
2. I haven't used my virtual machines in months. In fact, I deleted all the ones I've had installed, right now it's just Virtual Box's lone executable.
3. Yes I never bothered updating it because, afaik, unless you really need the features like hardware support of newer version, BIOS can be left alone. Is that incorrect? BIOS updates are nerve wrecking for me
4. Is Windows's integrated memory check good enough or do you really recommend memtest?

 

[PC Tailor]

Damn, Scarlet Crush is the driver that lets me use my Dualshock 3 on my PC. It is an older version though, from when W10 wasn't even out, I new one popped out when it was. Maybe I should uninstall the old one and update to the most recent one? It's pretty important for my PC, I'd avoid completely getting rid of it unless there are absolutely no other ways. Do you have any idea what the best way to uninstall it is? I've never had to manually unistall a driver.

Use autoruns first and simply disable it, and see if the issues repeat. There is no sure fire way to resolve this, all I will say is we've had multiple people who have had to uninstall it entirely to resolve their issue.

I haven't used my virtual machines in months. In fact, I deleted all the ones I've had installed, right now it's just Virtual Box's lone executable.

Well you have modules for the VirtualBox running all the time, again you can disable these in autoruns.

Yes I never bothered updating it because, afaik, unless you really need the features like hardware support of newer version, BIOS can be left alone. Is that incorrect? BIOS updates are nerve wrecking for me

BIOS can be left alone IF you're not currently experiencing a problem. They don't just come with hardware support, they come with firmware patches, improvements, compatibility and improved stability of your hardware. Basically if you have no issues, you don't need to update it, if you have issues, it's one of the first things you should update.

There is nothing to worry about, the consequences if it goes wrong can be daunting, but the only real way it often goes wrong is if say you have a power cut during the update, which can corrupt the BIOS. It's a simple process, and as long as it is followed to the manual, there is no problem.

Is Windows's integrated memory check good enough or do you really recommend memtest?

It CAN be good enough, but memtest is UNDOUBTEDLY good enough. The last thing you would want to do is test it on Windows, to then test it on Memtest and find out memtest found issues that Windows Diagnostic did not, so it's always best to go for the latter. I've had modules pass Windows and fail Memtest, but I've yet (myself) to see it the other way around.

If you've had a multitude of BSOD over the years, it can point to hardware.

 

[cloudropis_1]

Use autoruns first and simply disable it, and see if the issues repeat. There is no sure fire way to resolve this, all I will say is we've had multiple people who have had to uninstall it entirely to resolve their issue.


Well you have modules for the VirtualBox running all the time, again you can disable these in autoruns.


BIOS can be left alone IF you're not currently experiencing a problem. They don't just come with hardware support, they come with firmware patches, improvements, compatibility and improved stability of your hardware. Basically if you have no issues, you don't need to update it, if you have issues, it's one of the first things you should update.

There is nothing to worry about, the consequences if it goes wrong can be daunting, but the only real way it often goes wrong is if say you have a power cut during the update, which can corrupt the BIOS. It's a simple process, and as long as it is followed to the manual, there is no problem.


It CAN be good enough, but memtest is UNDOUBTEDLY good enough. The last thing you would want to do is test it on Windows, to then test it on Memtest and find out memtest found issues that Windows Diagnostic did not, so it's always best to go for the latter. I've had modules pass Windows and fail Memtest, but I've yet (myself) to see it the other way around.

If you've had a multitude of BSOD over the years, it can point to hardware.

- I've never used autoruns, I guess those are the ones I should disable?

Spoiler

If I disable them will they remain in this menu, unticked, so I can quickly renable them by ticking them again? I mean, they won't disappear from this list or anything like that? Do I need to restart my PC or something when I disable them?
(shenkuimza doesn't even give results in google, the hell is that. Didn't say so in the OP, but I scan my system constantly and never download anything from nor visit shady sources)

- Had no idea VBox did this, I may as well uninstall it

- I see, I will take my time and do a BIOS update then, and use memtest (as soon as I fetch a bootable drive)

Thanks again for the help, you are great. A couple of questions more:

• out of desperation I also downloaded my mobo's drivers again, even though they should all be up to date already and Windows's Manage Devices doesn't give me any yellow warnings. Does it make sense to install them just in case?
• After running memtest, updating my bios and reinstalling my GPU's drivers with DDU, I'll try fiddling around with autoruns. Vbox's stuff is out, but as I've explained I really hope Scp's driver isn't causing this issue as I kinda need it. Dooes it make sense to do all the above but leave that one enabled until I get another BSOD again? I never used a debugger to read dump files, I'm not sure exactly how your black magic works, but I guess you assume it may be a problem due to past experiences, but didn't find definitive evidence in the dump file itself, right? In this case, it makes sense to leave it on when all the rest is fixed in order to test it right? This discovery kinda made my world crumble as I'm also finding google results of your colleagues saying the same. I really hope it's just a consequence of bad situations like mine rather than the product itself being defective, especially since I'd expect some kind of controversy to have already risen up if it really was that bad of a program, and that hopefully it won't cause problems by itself when I fix the rest.

 

[PC Tailor]

If I disable them will they remain in this menu, unticked, so I can quickly renable them by ticking them again? I mean, they won't disappear from this list or anything like that? Do I need to restart my PC or something when I disable them?

Correct. Bonus of autoruns, doesn't get rid of anything, just disables them.

out of desperation I also downloaded my mobo's drivers again, even though they should all be up to date already and Windows's Manage Devices doesn't give me any yellow warnings. Does it make sense to install them just in case?

It's always best to keep drivers up to date so I would say so yes.

After running memtest, updating my bios and reinstalling my GPU's drivers with DDU, I'll try fiddling around with autoruns

I would honestly try autoruns first - as it's quick, simple and easy. And yes the ones you highlighted were correct, just apart from the Virtual Bus one, I can't say I know the origin of that one, but i don't believe it's your VirtualBox.

but I guess you assume it may be a problem due to past experiences, but didn't find definitive evidence in the dump file itself, right?

Correct - the debugging isn't an exact science - and it almost never quite simply gives you the answer, it's a matter of getting to the right information and then applying experience to it. Sometimes it is obvious, such as a third party module being present in your memory stack (in your case, this wasn't the case). But most of the time, it's a matter of just applying deduction and trial and error.

It wasn't DEFINITIVE evidence, but the Scarlet crush driver was RUNNING at the time of the crash, basically you have lots of other third party modules on your system, but the ones in my report were the only ones running at the time of the crash - so applying past experience then says that I know that driver causes issues, so it would be my first port of call. As I said, that is hoping that the issue is purely software related, as they can equally be hardware.

Out of curiosity, why is it so bad that you have to diable your playstation controller? I wouldn't have thought it was world crumbling!

 

[cloudropis_1]

Correct. Bonus of autoruns, doesn't get rid of anything, just disables them.


It's always best to keep drivers up to date so I would say so yes.


I would honestly try autoruns first - as it's quick, simple and easy. And yes the ones you highlighted were correct, just apart from the Virtual Bus one, I can't say I know the origin of that one, but i don't believe it's your VirtualBox.


Correct - the debugging isn't an exact science - and it almost never quite simply gives you the answer, it's a matter of getting to the right information and then applying experience to it. Sometimes it is obvious, such as a third party module being present in your memory stack (in your case, this wasn't the case). But most of the time, it's a matter of just applying deduction and trial and error.

It wasn't DEFINITIVE evidence, but the Scarlet crush driver was RUNNING at the time of the crash, basically you have lots of other third party modules on your system, but the ones in my report were the only ones running at the time of the crash - so applying past experience then says that I know that driver causes issues, so it would be my first port of call. As I said, that is hoping that the issue is purely software related, as they can equally be hardware.

Out of curiosity, why is it so bad that you have to diable your playstation controller? I wouldn't have thought it was world crumbling!

About the third one: considering I don't get BSODs consistently I wouldn't want to, for example, disable those processes first thing first, go on with my life for who knows how many days or weeks before an other BSODs happens, with the realization I didn't actually fix anything and I only deprieved myself of days without my controller because I disabled the process.
I don't remember if I mentioned it already but I don't get those BSOD's that often, whatever fixes the problem won't be immediately apparent because it's not like it crashes the whole time. Plus, if flashing the BIOS/reinstalling the drivers/replacing the RAM/whatever fixes the issue, disabling the DS3 process straight away won't actually do anything to fix it further, if anything it will just make it impossible to use my controller for no reason.
About ScpVbus, in the old threads about Scarlet being the deifnite culprit, like https://forums.tomshardware.com/threads/getting-random-bsod-on-my-pc-i-just-built.2275624/ , they straight up mention ScpVbus being the problematic process. Is it or am I mistaken? Should I just aovid the first one at the top of the screenshot then?

About the importance of controller: the DS3 is my controller of choice for PC games where it's favored compared to keyboard. With SCP Driver the system recognizes it as an Xbox 360 controller, which means it has full native support unlike, for example, the DS4 which needs some fiddling to work. The Xbox controllers are recognized natvely, of course, but the d-pad placement and quality make it pretty suboptimal for games like DMC or pretty much all 2D games. Simply put, if I have to disable SCP I'll will have to switch to either a controller with worse native support or another one that is straight up not good for my needs - I really, really need the DS3 specifically. And it's not like there are other alternative drivers either, SCP is already the best one, the only alternatives are straight up malware like Motioninjoy . So yeah, it's "first world problems"

Thanks again for the help dude, really appreciate it. Between the next very few days I should do everything. Luckily there's a game where I get either crashes or straight BSODs consistently (first time I had them happen so often and conssitently, it's the reason I made this thread), so I'll be able to test my new BIOS/drivers/disabled processes pretty easily. Will report anything new like new BSODs or the memtest results.

 

[Colif]

Unless the new version of SCP is newer than 2014, they haven't released anything new in months and I have seen it trip up windows driver verifer before.

Speaking of which, perhaps follow steps in 2nd post here and it might stop the guesswork

https://forums.tomshardware.com/thr...nclude-in-blue-screen-of-death-posts.3468965/

 

[cloudropis_1]

Unless the new version of SCP is newer than 2014, they haven't released anything new in months and I have seen it trip up windows driver verifer before.

Speaking of which, perhaps follow steps in 2nd post here and it might stop the guesswork

https://forums.tomshardware.com/thr...nclude-in-blue-screen-of-death-posts.3468965/

Well to be fair I have a very old version of that software, which has been superseded by a slightly more recent version. AFAIK it's still based on the same faulty driver though, so I have no idea if it will actually improve anything (unless you are gonna tell me that all the past cases of trip ups were pre-2014, if they are more recent than that then yeah it's a lost cause)

I will use driver verifier as a last resort, the BSOD loop scares the crap out of me. Plus, I have only one USB device I'll have to use for everything (memtest, bios update, W10 live usb etc) so it's kind of a pain lol. Thanks for the headsup.

 

[Colif]

you can run this to see what date the scarlet crush is showing as - all it does is looks at running drivers
helps to go into view and set it to hide all microsoft drivers, makes listing much shorter
http://www.nirsoft.net/utils/driverview.html

 

[cloudropis_1]

you can run this to see what date the scarlet crush is showing as - all it does is looks at running drivers
helps to go into view and set it to hide all microsoft drivers, makes listing much shorter
http://www.nirsoft.net/utils/driverview.html

Can't I just use autoruns for that? In the spoilered screenshot a few posts above it says it's from march 2014

 

[Colif]

Yes, sure. I haven't ever run it myself so I wasn't sure if it showed dates in it.

I don't know what percentage of people who run Driver Verifier get Boot loops, it has happened enough for me to add those 2 precautions, but it might not happen at all. It might be a faster way to see what is causing BSOD than to wait for the driver to show itself instead of blame ntoskrnl or another part of windows.

 

[cloudropis_1]

Correct. Bonus of autoruns, doesn't get rid of anything, just disables them.


It's always best to keep drivers up to date so I would say so yes.


I would honestly try autoruns first - as it's quick, simple and easy. And yes the ones you highlighted were correct, just apart from the Virtual Bus one, I can't say I know the origin of that one, but i don't believe it's your VirtualBox.


Correct - the debugging isn't an exact science - and it almost never quite simply gives you the answer, it's a matter of getting to the right information and then applying experience to it. Sometimes it is obvious, such as a third party module being present in your memory stack (in your case, this wasn't the case). But most of the time, it's a matter of just applying deduction and trial and error.

It wasn't DEFINITIVE evidence, but the Scarlet crush driver was RUNNING at the time of the crash, basically you have lots of other third party modules on your system, but the ones in my report were the only ones running at the time of the crash - so applying past experience then says that I know that driver causes issues, so it would be my first port of call. As I said, that is hoping that the issue is purely software related, as they can equally be hardware.

Out of curiosity, why is it so bad that you have to diable your playstation controller? I wouldn't have thought it was world crumbling!

@PC Tailor @Colif
I've completed the DDU reinstallation and let memtest do his thing for a whole night. Results of the latter were... Pretty bad
https://1drv.ms/u/s!AszPI4pVTqe0gUOvTXGPTHqwkZ7a?e=35XOOD

Microsoft OneDrive

1drv.ms

First test gave 1 error, another one during the second, I went to sleep and I woke up to 12 total errors. 12 is a lot for this, right?
Should I carry on with my maintenance routine (I still have to upgrade the BIOS and disable the drivers through autoruns, plus eventually using driver verification), or does the memetest result point at my RAM being the most likely culprit?

 

[PC Tailor]

@PC Tailor @Colif
I've completed the DDU reinstallation and let memtest do his thing for a whole night. Results of the latter were... Pretty bad
https://1drv.ms/u/s!AszPI4pVTqe0gUOvTXGPTHqwkZ7a?e=35XOOD

Microsoft OneDrive

1drv.ms

First test gave 1 error, another one during the second, I went to sleep and I woke up to 12 total errors. 12 is a lot for this, right?
Should I carry on with my maintenance routine (I still have to upgrade the BIOS and disable the drivers through autoruns, plus eventually using driver verification), or does the memetest result point at my RAM being the most likely culprit?

Even a single error on Memtest indicates RAM that needs replacing. Don't bother with the rest until you fix the glaring issue first. Let alone 12+ errors.
RAM from reputable manufacturers all come with a lifetime warranty so should be able to return the pack easily.

 

[Colif]

check motherboard web site and make sure to buy ram that is on the listing - https://www.gigabyte.com/au/Motherboard/GA-Z97-HD3-rev-10/support#support-doc

ram might be cause of all the errors you were having

 

[cloudropis_1]

Even a single error on Memtest indicates RAM that needs replacing. Don't bother with the rest until you fix the glaring issue first. Let alone 12+ errors.
RAM from reputable manufacturers all come with a lifetime warranty so should be able to return the pack easily.

I've ordered a new stick of RAM, should arrive by Thursday. Once I install I'll run memtest and then stress the PC in those scenarios that lead to software crashes or BSODs.
Thanks a lot for the help. Can this thread be left open until then, or should I pick a best answer and open a new one referencing this if I encounter issues?

 

[PC Tailor]

I've ordered a new stick of RAM, should arrive by Thursday. Once I install I'll run memtest and then stress the PC in those scenarios that lead to software crashes or BSODs.
Thanks a lot for the help. Can this thread be left open until then, or should I pick a best answer and open a new one referencing this if I encounter issues?

Leave it open for now, (Colif is mod so can technically answer better than I) - but we want to verify that your problem is fixed before running away
because there is no guarantee that the memory is the sole cause of your issue, but you've identified a big problem, that needed fixing regardless, and it very could be the sole cause. So we'll have to see.

 

[Colif]

just leave it open until you sure its fixed. So wait a while and see if you get any bsod in a week or so (though in some cases, 1 week isn't enough time, I have had questions come back to bite me after a month of no errors).

NO rush to close threads

 

[cloudropis_1]

Sup again guys
My new RAM just arrived, I installed it and everything fine. I'm gonna procede the "stress test" by playing the game in which I could reliably get crashes or BSODs with the SCP drivers on and, if they keep happening, off (I managed to snag a different controller from a friend that will let me turn off the SCP drivers for a bit). I will also run a memtest tonight, just to be sure

 

[cloudropis_1]

Sorry for the long absence @Colif and @PC Tailor, I'm finally reporting back.
The new RAM is great, memtest gives me 0 errors, I tested it plenty on the game where I'd get continuos crashes and BSODs on and now works like a charm.
It didn't completely fix my BSOD problems though. For example, shortly after changing the RAM, I woke up the PC from sleep and noticed my browser wouldn't open and my desktop screensaver was slightly distorted (telltale sign of my PC about to self combust in a matter of seconds, I rebooted manually before it went to BSOD). Just now, I restarted my PC and got an IRQL_NOT_LESS_OR_EQUAL before shutdown. Those are the only ones I've got though, and I never got them again in the moments that mattered (games and whatnot), so I'm overall happy

 

[Colif]

you are supposed to be fixed, go away (joking)

you could give us minidumps and we could see if we figure out why you crashing still.

 

[PC Tailor]

Agreed minidumps may help

the IRQL one is usually a driver, so who knows, maybe we've eradicated one problem to reveal another. Either way, the previous RAM needed replacing, and by the fact that it's gone down is a good indicator at least.

my desktop screensaver was slightly distorted

Hmm... I'd be interested to see the dump files, but maybe if it's having graphical implications it could be an indicator...

 

[Colif]

people still use screensavers? Thats the bigger question for 2019

 

[cloudropis_1]

people still use screensavers? Thats the bigger question for 2019

I actually meant the desktop background picture, ESL strikes again.

https://1drv.ms/u/s!AszPI4pVTqe0gUQWFe1TuDoY9CgO?e=oxPRcW

I do agree it's time to lay this to rest, don't worry, unless you find something surprising in this dump file I'm quite happy to finally close the thread

 

[Colif]

prefer not to close it till it over.

Weirdness after waking from sleep is sure sign of old drivers not liking sleep. Could be display drivers need an update, though might be something else. I can't look at the old dumps that PC Tailor converted before as the site that hosts them deletes them after a month.