BSOD of hell

[Bartimaeus]

Yeah, not sure. Symptoms are all over the place, so I can't rightly guess as to what it might be or whether the manufacturer will be able to fix it.

[Hulk'O'Saurus]

3 hours ago, Bartimaeus said:

Yeah, not sure. Symptoms are all over the place, so I can't rightly guess as to what it might be or whether the manufacturer will be able to fix it.

Can't blame you :D. 

Thanks for trying anyways, man. Appreciate it.

I am literally piling a report for them, atm. I called them today and I spoke with a different person - I have a 3 months extended warranty only for the changed part, and I told him most of what's going on. He was very, very puzzled. However, they are also looking into software issues, and I will get a discount for transportation. I wish I could reset this thing to factory setting and try that, but alas, nope.  

EDIT: just finished a fresh Windows install form USB made on the other laptop. It went to BSOD withing 5 minutes of running. 

Edited December 3, 2019 by Hulk'O'Saurus

[pmp10]

On 12/2/2019 at 6:27 PM, Hulk'O'Saurus said:

Started it up under Ubuntu from flashdrive and it has ran for hours, plugged in and without issue. 

From the remaining components it's still possible to check if it's the hard drive. 
If you have one of these windows-compatible USBs try running win10 directly from that.
Preferably without HDD connected so it can't try to write swap files.

[Hulk'O'Saurus]

2 hours ago, pmp10 said:

From the remaining components it's still possible to check if it's the hard drive. 
If you have one of these windows-compatible USBs try running win10 directly from that.
Preferably without HDD connected so it can't try to write swap files.

The trouble of removing both the SSD and HDD on this laptop notwithstanding, I will look into it tomorrow

[pmp10]

51 minutes ago, Hulk'O'Saurus said:

The trouble of removing both the SSD and HDD on this laptop notwithstanding, I will look into it tomorrow

Disabling them in BIOS will do just fine.

[Hulk'O'Saurus]

3 hours ago, pmp10 said:

Disabling them in BIOS will do just fine.

Thanks for the tip!

[Hulk'O'Saurus]

22 hours ago, pmp10 said:

From the remaining components it's still possible to check if it's the hard drive. 
If you have one of these windows-compatible USBs try running win10 directly from that.
Preferably without HDD connected so it can't try to write swap files.

Small update. 

This advise was next on the list of things to do. I did get one about reinstalling the ''latest DirecX redistributable'' before that.

Being bad with PC specifics I went online and looked what the ''latest redistributable'' might be but couldn't find anything I could go by. I started the new Windows installation(the one I installed from USB on top of the old one) under Diagnostic Startup and uninstalled the Video Adapter driver via Device Manager. I installed DirectX 11 after that and put it under Normal Startup. DxDiag still shows DirectX 12 is installed, but Windows has ran for a few hours now without issue. I've disabled updates for now.  

Edited December 4, 2019 by Hulk'O'Saurus

[Bartimaeus]

If it continues to run fine, then at that point, it might be wise to look at the SMART data for your hard drive (although I would've expected a chkdsk to have been performed by Windows at some point automatically if the drive were dying). Most likely, though, it sounds like a Windows update or driver issue.

[Hulk'O'Saurus]

1 hour ago, Bartimaeus said:

If it continues to run fine, then at that point, it might be wise to look at the SMART data for your hard drive (although I would've expected a chkdsk to have been performed by Windows at some point automatically if the drive were dying). Most likely, though, it sounds like a Windows update or driver issue.

Yeah, I will do that tomorrow

[Hulk'O'Saurus]

18 hours ago, Bartimaeus said:

If it continues to run fine, then at that point, it might be wise to look at the SMART data for your hard drive (although I would've expected a chkdsk to have been performed by Windows at some point automatically if the drive were dying). Most likely, though, it sounds like a Windows update or driver issue.

I used Command Prompt to check the SMART and it didn't return any errors. Is there another application that does this better?

I've also been installing drivers today. 

Windows itself is stable but I encountered a strange error. 

Spoiler

I was installing basic drivers - some from the manufacturers' websites, some from MSI's. In between every install I'd also restart to see if it will work. 

MSI's Dragon Centre is a decent app that can control cpu clocks and fan speed, as well as provide info on temps and a few other things. It has some additional features in the forms of apps, one of which can control blue light levels. I like that one a lot. But after I installed it - MSI True Colour, as it's called, the aforementioned error started appearing. Of course I uninstalled the app and rested again, this time with no error. The error appeared again when I got to sound drivers - Realtek High Definition Audio and Nahamic. The error appeared between installations of both, after which I uninstalled both drivers, and also Nvidia's Audio driver, as well. I installed GeForce Experience and updated gpu drivers, but that didn't fix the issue. The error is still there although the code it gives slightly changes.  

EDIT: I started uninstalling Nvidia display drivers under Normal Startup. Whenever the latest version would uninstall, there'd be an older Nvidia Display Driver installed after that. This thing repeated basically until I went into Diagnostic startup and uninstalled everything Nvidia related. Without reverting to Normal Startup Windows gives an error upon loading connected with Windows update functions being disabled. I am guessing that it wants to automatically install some form of Nvidia display driver. The error given on restart is gone. If I am reading the issue correctly - for I very well may not, Nvidia drivers are having some form of a problem with some of the laptop's drivers and firmware. 

Edited December 5, 2019 by Hulk'O'Saurus

[Bartimaeus]

GSmartControl is what I'd suggest to read the SMART data of drives.

In general, it usually works best to install drivers specific to your motherboard (or in this case, laptop model i.e. from MSI). The GPU drivers are the exception to that rule, since it's usually better to use the latest from Nvidia/AMD, but maybe try MSI's latest driver if they have one for the time eing.

[Hulk'O'Saurus]

On 12/5/2019 at 5:28 PM, Bartimaeus said:

GSmartControl is what I'd suggest to read the SMART data of drives.

In general, it usually works best to install drivers specific to your motherboard (or in this case, laptop model i.e. from MSI). The GPU drivers are the exception to that rule, since it's usually better to use the latest from Nvidia/AMD, but maybe try MSI's latest driver if they have one for the time eing.

I tried GSmartControl today. The HDD - non-system drive, doesn't return any errors. The SSD - system OS drive cannot be read. It shows that it's SMART is enabled and supported but GSmartControl keeps ''unknown'' for it's basic health check. 

I also formatted OS drive and installed another, hopefully cleaner, version of Windows. That one went BSOD within 5 minutes of work. I went into Diagnostic and uninstalled both video adapter and 3d controller drivers. Now it worked. This is when I finally noticed something. The system is SLI and I think all of this time most of these BSODs have been happening when Windows tries to recognise and configure something about the SLI. The new installation of Windows, as it always does, detects first one of the GPUs, then the second one and you can notice that through Device Manager. That installation of Windows encountered some form of a problem when it trying to setup the second GPU. Windows also installs newer version of gpu drivers than the ones on MSI website(at least as far the laptop model is concerned) and that is non-negotiable with Windows. Even if you uninstall them, it will just repeat the setup automatically. Instead I updated GPU drivers to latest with Nvidia Experience and then installed a bunch of other drivers and MSI firmware on top of that. Restarting in between each new install. It's working without any issue so far. 

EDIT: Ok, it actually has frozen when I went to do groceries. 

EDIT 2: It crashed/froze a few times more including on startup. It also refused to load Ubuntu from flash twice.

Edited December 6, 2019 by Hulk'O'Saurus

[Zoraptor]

You can turn off automatic graphics driver updates in windows, it's just obfuscated. If I don't windows happily installs some ancient Catalyst driver over the top of my shiny new Adrenaline drivers which completely breaks pretty much everything since it doesn't bother uninstalling the Adrenaline drivers first. I doubt that's your problem though and not just because your card is nVidia instead but I'll dig out how to do it later since I don't have time now.

[Hulk'O'Saurus]

2 hours ago, Zoraptor said:

You can turn off automatic graphics driver updates in windows, it's just obfuscated. If I don't windows happily installs some ancient Catalyst driver over the top of my shiny new Adrenaline drivers which completely breaks pretty much everything since it doesn't bother uninstalling the Adrenaline drivers first. I doubt that's your problem though and not just because your card is nVidia instead but I'll dig out how to do it later since I don't have time now.

Thanks, man. Appreciated. 

I think that if everything hardware-wise was fine, a brand new Windows install should have fixed these gremlins. As it is I think I am just going to compile all of this into a failure report and ship it over for another inspection. 

Don't bother. 

Edited December 6, 2019 by Hulk'O'Saurus

[Gorgon]

BIOS update and The GFX card's firmware.

Windows might think the card is something other than it is. Fiddling with the cards firmware could brick it, so very last resort.  

[Hulk'O'Saurus]

On 12/8/2019 at 10:19 PM, Gorgon said:

BIOS update and The GFX card's firmware.

Windows might think the card is something other than it is. Fiddling with the cards firmware could brick it, so very last resort.  

Thanks for tuning in.

BIOS update was one of the first things I did way back when. Didn't work. 

How would the gpu's firmware change things, though? I don't think there's anything on MSI's side that could be considered gpu firmware, though. The only thing I can relate to is the Dragon Centre, but that one only shows temps and clocks when it comes to the gpus. 

In any case, I am making another request to them literally today. 

[Gorgon]

What card is it ?

I mean the Vbios, It's the GFX Bios, since we are leaning towards driver incompatibility or maybe slightly wrong detection.  Overclockers use custom ones for tweaks they can't do anywhere else, which might mean there is one available. 

It could absolutely destroy the card though. 

 

 

[Hulk'O'Saurus]

15 minutes ago, Gorgon said:

What card is it ?

I mean the Vbios, It's the GFX Bios, since we are leaning towards driver incompatibility or maybe slightly wrong detection.  Overclockers use custom ones for tweaks they can't do anywhere else, which might mean there is one available. 

It could absolutely destroy the card though. 

I see. It's an GTX 1080 SLI.

I see what you mean there, but in order to make anything happen concerning this I need to read on the subject further. Everything I've ever done to resolve the issue has been due to people's comments, and I have researched quite a number of things during last week. As I literally filed the fault description now - basically making another service request to MSI - there might not be enough time to further delve into this. I could include it as a suggestion to MSI, though. I believe this time they will look harder into it. 

Other than that, the only overclocked component is the cpu - does that change things?

[Hulk'O'Saurus]

I apologise for resurrecting an old thread, but I've got an update on the issue, if anyone's interested. 

First, MSI didn't charge me for service whatsoever. They restored the factory settings on the machine(something I wasn't able to do) and it is running a lot more stable now. They've also performed a few tests of their own but couldn't find any problems whatsoever. They even sent me screenshots.

I've been monitoring the thing for over two weeks now, and while it's running much more stable, it has still produced some bsods(after some considerable up-time). Their frequency has diminished.

I have disabled Windows updates and haven't updated any drivers or anything.

Is anyone still interested in giving this a go :( ?   

[Bartimaeus]

We've run you through all the troubleshooting steps, unfortunately, so there's nothing to really "give a go" to.  Thing is, if it were my machine, I'd have a bunch of different computers to swap out parts with. I'd swap the GPU and PSU at the very least of it and another machine and see if the issues continue or switch to the other machine or neither. If you can't do that, though, I can't really think of anything else to recommend to try.

[Hulk'O'Saurus]

9 hours ago, Bartimaeus said:

We've run you through all the troubleshooting steps, unfortunately, so there's nothing to really "give a go" to.  Thing is, if it were my machine, I'd have a bunch of different computers to swap out parts with. I'd swap the GPU and PSU at the very least of it and another machine and see if the issues continue or switch to the other machine or neither. If you can't do that, though, I can't really think of anything else to recommend to try.

For which I am thankful, by the way, because as I know next to nothing about pc stuff that whole thing has been a learning experience, if thoroughly unpleasant.