You are not logged in.
No I didn't but also I wouldn't know what to look for.
The GPU sits tight if you ask me
MSI Cause: This is a great Idea I will test this with a 60hz Screen. I have one.
Last edited by i34sti1337 (2021-09-09 13:15:26)
Offline
Another Try, Another Freeze; this time It even has a new Error. because of the NVIDIA Command startx -logverbose 6
and now I see the rest of the Error Message.
The Full Log is here.
https://pastebin.pl/view/619da66d
I don't like that, I don't like that at all.
Sep 09 22:58:37 seb kernel: NVRM: GPU at PCI:0000:01:00: GPU-17becba7-9540-63e6-c6a2-6f931a9958e3
Sep 09 22:58:38 seb kernel: NVRM: Xid (PCI:0000:01:00): 79, pid=490, GPU has fallen off the bus.
Sep 09 22:58:38 seb kernel: NVRM: GPU 0000:01:00.0: GPU has fallen off the bus.
Sep 09 22:58:38 seb kernel: NVRM: GPU 0000:01:00.0: GPU serial number is .
Sep 09 22:58:40 seb steam.desktop[2697]: STEAM_RUNTIME_HEAVY: ./steam-runtime-heavy
Sep 09 22:58:40 seb steam.desktop[2699]: ./steamwebhelper: symbol lookup error: /usr/lib/libfreetype.so.6: undefined symbol: hb_ot_tags_from_script_and_language
Sep 09 22:58:49 seb steam.desktop[2282]: ERROR: Can't load a null sound.
Sep 09 22:58:49 seb kernel: nvidia-modeset: ERROR: GPU:0: Failed to query display engine channel state: 0x0000c57e:0:0:0x0000000f
Sep 09 22:58:49 seb kernel: nvidia-modeset: ERROR: GPU:0: Failed to query display engine channel state: 0x0000c57e:0:0:0x0000000f
Sep 09 22:58:49 seb kernel: nvidia-modeset: ERROR: GPU:0: Failed to query display engine channel state: 0x0000c57e:0:0:0x0000000f
Sep 09 22:58:50 seb kernel: nvidia-gpu 0000:01:00.3: can't change power state from D3cold to D0 (config space inaccessible)
Sep 09 22:58:50 seb kernel: nvidia-gpu 0000:01:00.3: nv_msi_ht_cap_quirk_leaf+0x0/0x20 took 14267 usecs
Last edited by i34sti1337 (2021-09-09 21:28:57)
Offline
The "Failed to query display engine channel state" and "can't change power state from D3cold to D0" and "nv_msi_ht_cap_quirk_leaf" stuff happens after and because the GPU fell off the bus (what means it disappeared from the system)
That's not really surprising.
The MSI gaming controller is still there, though?
Online
I did not test it with the different screen yet, I will today
Last edited by i34sti1337 (2021-09-10 10:36:44)
Offline
so I tested it with the smaller screern 1920x1080 Refreshrate: 60Hz and of course it was the MSI Screen
This sucks that Screen is awesome. There is no way that can be fixed. Why MSI Why
No Errors or GPU fallen off the bus Entry and I Installed it with the normal Nvidia Package
https://pastebin.com/jNqVcQ48
Last edited by i34sti1337 (2021-09-10 13:37:03)
Offline
Hold your horses! ;-)
What if you drive the MSI Screen (why does it register as game controller?) at the lower resolution and refresh rate?
Online
I don't know what you mean, but here is the device I was talking about
In BIOS (Apprently it's essential for a monitor these days -.-)
https://ibb.co/ZS2rPYJ
And in Windows 10
https://ibb.co/XY9rLw1
The Monitor is this one:
https://de.msi.com/Monitor/Optix-MAG322CQRV/Overview
There is no way I can stop that from happening.
It's integrated in the Hardware.
Last edited by i34sti1337 (2021-09-10 14:42:19)
Offline
xrandr --output DP-1 --mode 1920x1080 --rate 60 # assuming the MSI is wired to DP-1, check "xrandr -q"
I'm not yet sure that the fake usb drive is the direct cause. Rendering a game at 4k/200Hz is considerably more demanding to the GPU that running it at FullHD/60Hz
Online
Do you think it could work with DP? I use the HDMI Port and Cable.
But i try it.
Last edited by i34sti1337 (2021-09-10 15:05:44)
Offline
Same result with 1920x1080 and 60hz. Also it does not crash on Windows
And it's not 4k it's 2k QHD -> also known as 1440p
Last edited by i34sti1337 (2021-09-10 15:26:50)
Offline
Same result with 1920x1080 and 60hz.
Ok, so at least it's the stupid fake USB device.
If you
echo 0 | sudo tee /sys/devices/pci0000:00/0000:00:14.0/usb1/1-8/1-8.1/1-8.1.3/authorized # check the journal for the actual bus address
and then play the game… does it still crash?
DisplayPort supports USB as well, so while it might be more stable, you won't be able to ignore the device this way.
Edit: do you have a dedicated usb connection from the monitor or is this some weird usb-over-cec stuff?
Last edited by seth (2021-09-10 15:46:20)
Online
I missed a second connection cable which was plugged in: from the MSI Screen, which apparently was the MSI Gaming Controller, because it doesn't show the MSI Device anymore when I type in lsusb
Maybe it works when I plugged it off.
I will try it now. This could actually work!
Last edited by i34sti1337 (2021-09-10 19:58:48)
Offline
It's getting more and more fun:
Now the /dev/sde is gone but it still freezes
Sep 10 21:17:20 seb kernel: NVRM: GPU at PCI:0000:01:00: GPU-17becba7-9540-63e6-c6a2-6f931a9958e3
Sep 10 21:17:20 seb kernel: NVRM: Xid (PCI:0000:01:00): 32, pid=2824, Channel ID 00000030 intr0 00040000
Sep 10 21:17:20 seb kernel: NVRM: Xid (PCI:0000:01:00): 32, pid=2824, Channel ID 00000030 intr0 00040000
Sep 10 21:18:11 seb kernel: NVRM: Xid (PCI:0000:01:00): 79, pid=0, GPU has fallen off the bus.
Sep 10 21:18:11 seb kernel: NVRM: GPU 0000:01:00.0: GPU has fallen off the bus.
Sep 10 21:18:11 seb kernel: [133B blob data]
Sep 10 21:18:11 seb kernel: NVRM: A GPU crash dump has been created. If possible, please run
NVRM: nvidia-bug-report.sh as root to collect this data before
NVRM: the NVIDIA kernel module is unloaded.
Sep 10 21:18:12 seb kernel: irq 16: nobody cared (try booting with the "irqpoll" option)
Sep 10 21:18:12 seb kernel: CPU: 4 PID: 0 Comm: swapper/4 Tainted: P OE 5.14.2-arch1-2 #1 848e2e7fc8cf29819b6cce53109477af0de996f1
Sep 10 21:18:12 seb kernel: Hardware name: LENOVO 90JF005HGF/36E9, BIOS O3NKT30A 12/01/2020
Sep 10 21:18:12 seb kernel: Call Trace:
Sep 10 21:18:12 seb kernel: <IRQ>
Sep 10 21:18:12 seb kernel: dump_stack_lvl+0x46/0x5a
Sep 10 21:18:12 seb kernel: __report_bad_irq+0x35/0xaa
Sep 10 21:18:12 seb kernel: note_interrupt.cold+0xb/0x64
Sep 10 21:18:12 seb kernel: handle_irq_event+0xa9/0xb0
Sep 10 21:18:12 seb kernel: handle_fasteoi_irq+0x8a/0x1f0
Sep 10 21:18:12 seb kernel: __common_interrupt+0x41/0xa0
Sep 10 21:18:12 seb kernel: common_interrupt+0x7e/0xa0
Sep 10 21:18:12 seb kernel: </IRQ>
Sep 10 21:18:12 seb kernel: asm_common_interrupt+0x1e/0x40
Sep 10 21:18:12 seb kernel: RIP: 0010:cpuidle_enter_state+0xc7/0x380
Sep 10 21:18:12 seb kernel: Code: 8b 3d e5 71 7e 47 e8 88 7b 8a ff 49 89 c5 0f 1f 44 00 00 31 ff e8 a9 88 8a ff 45 84 ff 0f 85 da 01 00 00 fb 66 0f 1f 44 00 00 <45> 85 f6 0f 88 11 01 00 00 49 63 d6 4c 2b 2c 24 48 8d 04 52 48 8d
Sep 10 21:18:12 seb kernel: RSP: 0018:ffffb1f70014bea8 EFLAGS: 00200246
Sep 10 21:18:12 seb kernel: RAX: ffff89201ed2d700 RBX: 0000000000000008 RCX: 000000000000001f
Sep 10 21:18:12 seb kernel: RDX: 0000000000000000 RSI: 000000002d958513 RDI: 0000000000000000
Sep 10 21:18:12 seb kernel: RBP: ffff89201ed37f00 R08: 000000ea4fcdb054 R09: 0000000000000008
Sep 10 21:18:12 seb kernel: R10: 0000000000000003 R11: 0000000000000003 R12: ffffffffb9b49200
Sep 10 21:18:12 seb kernel: R13: 000000ea4fcdb054 R14: 0000000000000008 R15: 0000000000000000
Sep 10 21:18:12 seb kernel: ? cpuidle_enter_state+0xb7/0x380
Sep 10 21:18:12 seb kernel: cpuidle_enter+0x29/0x40
Sep 10 21:18:12 seb kernel: do_idle+0x1e1/0x270
Sep 10 21:18:12 seb kernel: cpu_startup_entry+0x19/0x20
Sep 10 21:18:12 seb kernel: secondary_startup_64_no_verify+0xc2/0xcb
Sep 10 21:18:12 seb kernel: handlers:
Sep 10 21:18:12 seb kernel: [<0000000068cd52de>] i801_isr [i2c_i801]
Sep 10 21:18:12 seb kernel: Disabling IRQ #16
lines 5762-5824/5824 (END)
Offline
DId you retry the other monitor to ensure the previous test wasn't a fluke?
Sep 10 21:18:12 seb kernel: RIP: 0010:cpuidle_enter_state+0xc7/0x380
https://wiki.archlinux.org/title/Intel_ … ete_freeze (ignore the baytrail part)
Online
I tried both Monitors again.
The old one didn't crash; I played and went to bed without any freezes.
The MSI did crash even with the MSI Gaming Controller plugged off.
It seems I have to write to NVIDIA. It's an internal problem, we can't fix.
Last edited by i34sti1337 (2021-09-11 11:47:05)
Offline
There used to be issues w/ certain monitors, but those were rather related to DP, not HDMI.
Regardless, do you use the same output (HDMI) and did you try DP?
Online
Okay also the DP Output crashed. Back to the Start, i guess..
Last edited by i34sti1337 (2021-09-11 20:18:02)
Offline
I found the Solution!
Again I looked at the code for the nivida-installation
&
/usr/bin/ldconfig
/usr/bin/depmod -a
seems to fix it.
Offline
Unlikely.
How sure are you about (how long did you test) the other output?
(Since the problem doesn't seem to be reliably reproducible)
Online
I played very long: I finished the game this evening.
I was out with friends before using the computer
I'm not sure what else would make it not crash other than those 2 commands.
Yes, I tried it without the installer and it works.
Last edited by i34sti1337 (2021-09-12 22:17:16)
Offline
Cosmic rays. Room temperature dropped.
I'll point out that due to the unpredictable nature of your issue, you had previously considered random other stuff to be the solution.
ldconfig updates the linker cache and depmod the kernel module dependency/map.
This happens pretty much every time you install something relevant (notably the kernel or kernel modules) and if it didn't, you'd fail to load kernel modules/shared objects and that would not lead to the deescribed symptoms but an error message that the module/library could not be found.
Online
It's a secret cheat code from MSI
I contacted NVIDIA anyhow
https://forums.developer.nvidia.com/t/m … -01/189026
Maybe they will fix it in the next version.
Last edited by i34sti1337 (2021-09-14 14:43:27)
Offline
I used the Linux Long Term Support Kernel and it seems to stop the freezes, why didn't I think of that more early.
EDIT: They updated the LTS Kernel Last Night; It doesn't work anymore.
still it works with the depmod thing.
Last edited by i34sti1337 (2021-09-20 09:54:06)
Offline
Update: The Bug was finally fixed in NVIDIA Version 470.74 no freezes anymore
Offline