X260 freezes with Kernel 5.1 and newer.

Hi people,
I have the unfortunate problem that my Lenovo X260 freezes with Kernel 5.1 and 5.2.
It has the 6200U Processor with HT enabled and 1.44 BIOS from 08.05.2019. The RAM was tested with Memtest and seems ok. I don't have a dedicated GPU.

The freezes occur in irregular intervals and last from a few seconds up to a total freeze requiring a hard reset. The problem does not occur with Kernel 4.19. I have no idea about the Kernel or how to diagnose the problem.

Any help in that matter is greatly appreciated!

In that case, it is probably best to stick with 4.19 for a while. It's the newest of the long-term-support kernels, so it'll continue receiving bugfixes for quite a while still. :thinking:

1 Like

This is indeed what I will have to do for some time. It however does not address the problem: It does not tell me if it is a hardware problem, some software I installed or indeed the kernel, nor does it make sure the problem will eventually be fixed.

Well, if you don't have this problem with 4.19, then I would surmise that it's not a hardware problem, but perhaps a problem in the 5.x kernel generation. :thinking:

I can also confirm the exact same behaviour on my Razer Blade Stealth (2017). Here is my inxi -Fxzc0 output:

System:    Host: blade Kernel: 4.19.69-1-MANJARO x86_64 bits: 64 compiler: gcc v: 9.1.0 Console: tty 0 Distro: Manjaro Linux 
Machine:   Type: Laptop System: Razer product: Blade Stealth v: 3.06 serial: <filter> 
           Mobo: Razer model: Blade Stealth serial: N/A UEFI: Razer v: 1.51 date: 05/21/2019 
Battery:   ID-1: BAT0 charge: 22.4 Wh condition: 51.2/53.8 Wh (95%) model: Razer Razer Blade Stealth status: Discharging 
CPU:       Topology: Dual Core model: Intel Core i7-7500U bits: 64 type: MT MCP arch: Kaby Lake rev: 9 L2 cache: 4096 KiB 
           flags: avx avx2 lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx bogomips: 23240 
           Speed: 700 MHz min/max: 400/3500 MHz Core speeds (MHz): 1: 700 2: 700 3: 700 4: 700 
Graphics:  Device-1: Intel HD Graphics 620 vendor: Razer USA driver: i915 v: kernel bus ID: 00:02.0 
           Display: server: X.org 1.20.5 driver: intel unloaded: modesetting resolution: <xdpyinfo missing> 
           OpenGL: renderer: Mesa DRI Intel HD Graphics 620 (Kaby Lake GT2) v: 4.5 Mesa 19.1.5 direct render: Yes 
Audio:     Device-1: Intel Sunrise Point-LP HD Audio vendor: Razer USA driver: snd_hda_intel v: kernel bus ID: 00:1f.3 
           Sound Server: ALSA v: k4.19.69-1-MANJARO 
Network:   Device-1: Qualcomm Atheros QCA6174 802.11ac Wireless Network Adapter vendor: Bigfoot Networks driver: ath10k_pci 
           v: kernel port: f040 bus ID: 01:00.0 
           IF: wlp1s0 state: up mac: <filter> 
           Device-2: Qualcomm Atheros type: USB driver: btusb bus ID: 1-6:2 
Drives:    Local Storage: total: 238.47 GiB used: 9.62 GiB (4.0%) 
           ID-1: /dev/nvme0n1 vendor: Samsung model: MZVLV256HCHP-00000 size: 238.47 GiB 
Partition: ID-1: / size: 54.54 GiB used: 9.59 GiB (17.6%) fs: ext4 dev: /dev/nvme0n1p6 
Sensors:   System Temperatures: cpu: 38.5 C mobo: 29.8 C 
           Fan Speeds (RPM): N/A 
Info:      Processes: 223 Uptime: 25m Memory: 15.56 GiB used: 1.58 GiB (10.1%) Init: systemd Compilers: gcc: 9.1.0 Shell: bash 
           v: 5.0.9 inxi: 3.0.36

It is impossible to even go past GDM and login using Kernel 5.1.xx or 5.2.xx. Any help would be appreciated thanks...

1 Like

You can try to install 5.3rc.

1 Like

Hi! Thanks for the advice. I'll definitely try it today... However, I am wondering why this behavior in current Linux kernels (like 5.1 & 5.2)... Don't you think it should be looked into?

Thanks

It could be anything. The switch from 4.19 to 5.0 wasn't exceptional, but from 5.0 to 5.1 there have been some changes. It would need to find a good release announcement.

Manjaro policy is always "try another kernel" and "stick with the one which works". Because it is much work also on the end user side to find why a kernel doesn't work on his or her machine. dmesg and journalctl are the users friends there.

1 Like

I see... How can I help... I want to...
But the problem is, you can't even get to any tty using 5.0/5.1. How may I help? Please be elaborate...

I don't know exactly, I would experiment.

Select the "fallback" option for the kernel in Grub for example.
Add 3 or 2 to kernel parameters.
Try to boot with a live USB with an ISO with linux50 or linux51.
Try a different distro.

Install wgetpaste.
Each after time you tried to boot without success, boot with linux419 and run
sudo journalctl -b -1 | wgetpaste
it will paste the output online, then share the link to the pasted output.

Edit: If wgetpaste doesn't work, then run
sudo journalctl -b -1 > journal.txt
Then paste the contents of the journal.txt file via a pastebin service of your choice.

1 Like

First and foremost, check the output of journalctl for any errors and warnings.
I have an X270 which works perfectly well with 5.2.
5.1 is EOL and shouldn't be used.

See eugen-b's comment above for journalctl output.
Could be a flaky OEM SSD or something like that.

1 Like

Doesn't work... with 2 or 3 as a kernel parameter, it does not even make it to the GDM Login Screen.

Never got successful.

Here it is:
https://hastebin.com/cilaropeye.bash

Same behaviour with normal or fallback kernel...

It is supposed to boot to TTY terminal, not to gdm with 2 or 3 as boot parameters. There you should enter you username and password and log in to terminal session.

I searched the word "error".
line 1877:

Sep 16 21:21:25 blade kernel: [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun

I would search for thiis error. Maybe someone else got it and that can give further hints.

lines 2033-2036:

Sep 16 21:21:32 blade /usr/lib/gdm-x-session[581]: (EE) intel(0): Failed to set backlight intel_backlight for output eDP1, disabling
Sep 16 21:21:32 blade /usr/lib/gdm-x-session[581]: (WW) xf86CloseConsole: KDSETMODE failed: Input/output error
Sep 16 21:21:32 blade /usr/lib/gdm-x-session[581]: (WW) xf86CloseConsole: VT_GETMODE failed: Input/output error
Sep 16 21:21:32 blade /usr/lib/gdm-x-session[581]: (WW) xf86CloseConsole: VT_ACTIVATE failed: Input/output error

This is exactly what I meant to say. I too was expecting a tty login and not a GUI... There was no login screen with either 2 or 3 as a kernel parameter... So... But, one thing was for sure, that my system did not freeze... i could tell this by toggling caps-lock on/off. Previously, I couldn't even toggle caps-lock.

Yeah I searched... it is related to i915. It has one SOLVED solution in arch forums, but that does not work for me. It mentions to put i915.enable_rc6=0 as a kernel parameter and use the latest LTS Kernel which unfortunately does not work for me...

Could you please explain as I couldn't get you.

I just posted lines 2033-2036, because they have errors, but I don't know.

It seems like a graphics issue with very "deep roots" that even the TTY is too much graphics for the kernel.
Try adding i915.enable_guc=2 to kernel parameters like explained here: https://wiki.archlinux.org/index.php/Intel_graphics#Enable_GuC_/_HuC_firmware_loading

1 Like

with or without passing 2 or 3 as a parameter? or just i915.enable_guc=2?

First without 3, then with 3.

1 Like

just being sure... do you think I should disable TLP and TLP-sleep. do you think they could be the culprits?

I think you should try any ideas you have, but document them somehow. For example here in the topic.

You might have a long test program to test everything in the Arch Wiki article.

1 Like

successfully booted without 3 using i915.enable_guc=2 with the current Manjaro kernel 5.2.11-1-MANJARO. Here is the current journalctl -b:
https://hastebin.com/evimosereq.txt

Please check...

1 Like

Forum kindly sponsored by Bytemark