Talos Principle hanging , Nvidia Bug Report included ( GTX 1050 )

It happens with 384.98 , 384.111 , 387.34. Vulkan and OpenGL.

Here is the bug report.
nvidia-bug-report.log.gz (97.9 KB)
acpidump.txt (1.12 MB)

Hmmmm… XID 39: Unused
Seems like Nvidia has to update their docs.
Were you running Vulkan or OpenGL when the hang occured? What other settings in TP were you using?

I was on Vulkan.

Setting were this:

Window Mode : Fullscreen

Resolution: 1920x1080

Max 3d rendering MPIX: 2.1 (HD 1080)

Fov :75.00

Vsync: Disabled

Performance :

Cpu speed : Ultra

Gpu speed: Medium

Gpu memory : Ultra

Level caching : Medium

Max fps: Unlimited

Can you please check if this is depending on PRIME sync, i.e. turn off sync using
xrandr --output eDP-1-1 --set “PRIME Synchronization” “0”
and then start TP?

It get past that certain hanging area this time but it stuttered real bad like it will gonna hang.

But i’ve first encountered this issue when Prime Sync wasn’t active. So i think i just got the luck on my side this time.

Also i have this odd issue too , with and without Prime Sync.

So this is always happening in a specific area? Which one?

Now that’s a load of problems.
Did you ever try to use a clean user profile? Maybe also try to delete the shader cache ~/.nv

It happens at where you had numbered puzzle worlds with teleporting points at the main hall.

It either happens one of the worlds there some times , but it happens every time when you get the elevator. Although it got past there this time with Psync off.

Ps: Game bought from Steam. Not a pirated version , because when i search this i found out that elevator thing was a pirated version easter egg.

Although at pirated version , elevator wouldn’t work as described. On my issue , it works and Elohim starts talking but when elevator go halfway through it , it hangs.

It is not bad as Cs Go lockup , because Cs Go locks whole system while Talos only locks itself.

Also i had this lockup issue with several games too , Bioshock Infinite and Everspace. These are happened as system lockups while Talos only hangs itself.

I deleted shader cache as some user from Reddit suggested but it didn’t help.

Also i have a question. New Steam pre cache shader page shows that i have 338 mb pre cached shaders. Is this normal and acceptable?

Also if you want to know anything more ; using this card was painful.

Nvidia prop drivers always caused a NMI watchdog lockup with Gdm and Lightdm.

And this an ongoing issue with most of the display managers.

I did a workaround with start using Mdm instead of Lightdm and using terminal to switch between Intel and Nvidia.

I bought an Nvidia based laptop with thinking that was the ultimate solution for Linux but it became a complete problem mess.

This depends on the type of games you have installed, this should be a resonable size.

I have 17 installed games , that looks normal i guess.

Laptop hanging after switch to intel is a known problem, see:
[url]Invalid Bug ID
The common workaround is to use kernel parameter
acpi_osi=! acpi_osi=“Windows 2009”
You can run
sudo acpidump >acpidump.txt
and attach that, I have a look at it.

Sure , i added it to first message.

Meh, your internal model number is GL553VD, two users showed up here already with that, so I know this will be a tough ride.
You have to use the kernel parameter
acpi_osi=! acpi_osi=“Windows 2009”
for the Nvidia GPU to work correctly due to the aforementioned kernel bug. Unfortunately, then your touchpad won’t work. To work around that, you have to change an acpi table and load that on boot. Luckily one of the previous users wrote a howto after fiddling this out.
[url]https://github.com/Bumblebee-Project/Bumblebee/issues/764#issuecomment-306543064[/url]

Yep , GL553 VD is basically the same laptop while it advertised as a gaming one while FX is much more general use.

I looked at it but it is so much hassle for me. Maybe i can try that when i decided to go on another adventure. For now , the solution i wrote on the link i gave you is works.

But is this related to my hanging problem on Talos , Cs Go etc?

And shouldn’t it fixed by kernel side for a permanent solution?

The reason for this bug is unknown, the kernel developers are clueless so there’s no fix in sight.
Maybe take a shortcut and PM the user medakk for his compiled initrd. I’ve compared your acpi tables with his and it’s the same.
https://devtalk.nvidia.com/member/2192263/
Then you will only have to copy it to /boot and add the line to grub.

Apart from this, it shouldn’t have any impact on games.

Thank you so much. I will reach him for sure.

I hope it works but things like these are big negatives for people who got sick of Windows and then decided on Linux. But that is not the place for that rant.

So if there is a connection between these lock , hang issues and Feral ports and got fixed ; i would be so happy.

Is there anything more for investigating that lockup issue more?

I noticed something when these Cs Go complete system lockup happened , hdd led went crazy. Just like it is using swap partition , but the thing is i don’t have a swap partition.

Though there’s nothing in dmesg about the hdd, maybe try to set up a test user with home on the ssd to rule out problems coming from there.
That aside, not having swap at all can lead to serious problems in terms of memory management, even a 512MB swap file would work around that. So that could be a candidate.

So can i add swap partition to a working setup by live usb?

I have 8 gb system ram , should i set a 8 gb swap partition?