System hangs at randome times (mostly while writing a lot of text)

ku_respawned · 2017-01-14 22:00:12

Hello,

my system starts to hang at random times, most often when I write a long text and try to save it, no matter if in Firefox, Thunderbird, Pidgin or Gedit. Mouse cursor is moving, but there is no responce to clicks and keyboard input. When i push my power button (set to hybrid sleep), it works, but after resuming the system still hangs.

I suspect my SSD, but smartctl shows no new errors, is there any other way to test it? Though, it also happened once when I booted a copy of Arch from a HDD, the SSD was connected, but not currently used. Windows doesn't hang completely, but once most of the software (but not all) that I had running was completely unresponsive.

I suspected a faulty motherboard, but now I have completely new motherboard, CPU and RAM and when I wrote a long e-mail in Thunderbird 10 minutes ago, the system hanged again

ku_respawned · 2017-01-15 13:15:39

Another thing is - Memtest86+ (from December Arch boot disc) hangs in the middle of test 7 (block move) with multi-core support, with both the old and the new hardware. With single-core support I managed 1.2 passes with the old hardware, then I rebooted. I tested it with my RAM (old and new) and with my brother's RAM, which should be OK. My brother himself couldn't run it with multi-core support on his PC (i7-860), it only ran in single-core.

P.S. I downloaded Gsmartcontrol for windows and ran an extended (10 minutes) test on the SSD, no new problems were detected.

Last edited by ku_respawned (2017-01-15 13:31:05)

ku_respawned · 2017-01-15 15:47:39

BIOS shows a little over 5 and 12V, there's no 3.3V value.

Yesterday I ran AIDA64 stress test with memory, CPU, FPU and GPU for 13 hours without anything to notice, today I restarted the test with addinational disk stress, very shortly after that Explorer has crashed and restarted. Of course, I can't know if it has restarted during the nightly test.

ooo · 2017-01-15 16:29:21

Last time I checked, and at least with the version bundled in arch ISOs, the multi-core support/test in Memtest86+ had some bug which caused the test to fail, even though single-core test completed fine.

I've also encountered similar freezes a couple of times during the past month or so. In my case the system resumes from the feeze eventually in a few minutes, and I don't have to hard reboot. Is that the case with your system as well?

Have you tried if you get the freezes with linux-lts kernel?

ku_respawned · 2017-01-15 18:16:56

I'm able to put the system into suspend/hibernation mode through the power button, but it doesn't recover from the freeze after resuming, I haven't waited more than for a couple of minutes, though. I haven't tried other kernels, in the other thread people suggested that it could be caused by a faulty PSU.

As for Memtest86+, do you remember if it was freezing or what kind of failing? In my brother's case it just started in sigle-core mode, even though he pressed F2, in my case it starts in multi-core mode, but freezes in the middle of test 7.

ku_respawned · 2017-01-15 19:07:58

P.S. During those "input freezes" I continue receiving mail in Thunderbird and new instant messages in Pidgin, and I'm still able to move the mouse, but no clicks or keyboard input are taking effect. My keaboard and mouse are connected through a KMV-switch, I tried changing the output channel and disconnecting the USB cable, nothing helps. Also I never had "input freezes" on my Windows notebook (second KVM-port).

In Windows (on this PC) I didn't get a complete "input freeze" yet, but once Thunderbird, Firefox and Total Commander stopped responding completely, but Edge and Control Panel still did respond.

ooo · 2017-01-15 21:58:23

ku_respawned wrote:

in my case it starts in multi-core mode, but freezes in the middle of test 7.

Yes indeed. If you search yourself, you should find discussions about this issue on internet.
I doubt it has anything to do with your system hanging issue, at least if you can complete the single core memtest just fine.

Also I doubt your other issue would be related to system hangs. As others have suggested, it sounds like a separate hardware issue, and nothing to do with your linux system.

You should test if the hanging disappears with linux-lts as I suggested.

ku_respawned · 2017-01-15 23:23:25

Thanks!
I tried linux-lts and wrote myself a gibberish mail in Thunderbird (which so far caused the most freezes most quickly) over five screens, constantly clicking on "Save". So far no freezes. But like I said, they appeared most randomly, sometimes once in two days (even though, I wrote a couple of posts in Firefox and a couple of short mails in Thunderbird) and sometimes twice in five minutes. If the freezes will reappear, I will report here.

ku_respawned · 2017-02-01 21:41:35

With new hardware I encountered no issues in 6-7 days, with the old, Firefox crashes after pretty much every suspending and I got 2 more "input freezes" in 6-7 days. I'm gonna exchange the new motherboard tomorrow (2 slots are not working) and gonna test it 7 for more days with new hardware. Then I have to decide, whether I keep it. Not sure whether it could actually have something to do with the PSU.

ku_respawned · 2017-02-04 20:52:56

OK, i exchanged the new motherboard yesterday and got another "input freeze" today, with the new hardware. I dunno, after I switched to -lts, it helped for a week or a little more (no problems whatsoever), now freezes happen on both hardware sets.

ku_respawned · 2017-02-13 21:55:40

OK, I exchanged the PSU and got another "input freeze". Now I'm clueless. A SSD (but then I got one with system started from the HDD)? A HDD? KVM Switch (but no such problems with two notebooks connected to it from time to time)? Or software?

killer1804 · 2017-02-14 05:05:47

may be it keyboard problem ? bad cabel, or usb or ps/2 ?

seth · 2017-02-14 15:09:12

Determine the nature of the freeze.
Does ctrl+alt+f1 still kick you to another VT?
Can you access the system from outside (ping/ssh)?

ewaller · 2017-02-14 15:15:11

I don't know how I had not seen this thread before. Rather than many consecutive posts, please use the edit function to add more information to your posts when they are the last post in the thread. It keeps things cleaner.

Have you installed and configured the Microcode updates for your processor? Doing so it step one and doing so solves a myriad of issues.

ku_respawned · 2017-02-14 16:41:24

Installed and configured? I wrote a line in grub to load them when I installed Arch, is that what you mean?

As for SSH - need to install Putty or something on my Windows notebook (actually, with Win10 it should be possible to install Linux and SSH client, if it's better). I'm sure, SSH would work as I'm still able to get mail/instant messages and watch a video to the end if it was started. The only thing I'm not able to do is not register clicks or keyboard input. I tried changing channels on the KVM Switch (with everything working at the same time on my Windows notebook) and dis-/reconnecting the KVM Switch USB cable and keyboard/mouse' USB cables on the KVM Switch. Still not able to do anything besides moving the mouse

Last edited by ku_respawned (2017-02-14 16:49:27)

ewaller · 2017-02-14 16:51:51

Do you know that it is working? That's all. It is easy to get wrong, and it had not been mentioned in the thread -- or if it was I missed it.

ewaller@turing ~ 1010 %journalctl -b | grep \[mM\]icrocode
Feb 11 21:43:46 turing kernel: microcode: microcode updated early to revision 0x20, date = 2016-03-16
Feb 11 21:43:46 turing kernel: microcode: sig=0x306c3, pf=0x20, revision=0x20
Feb 11 21:43:46 turing kernel: microcode: Microcode Update Driver: v2.01 <tigran@aivazian.fsnet.co.uk>, Peter Oruba
ewaller@turing ~ 1011 %

This output shows that my microcode is being updated to rev 0x20.
Your output may or may not show that the an update is applied, but the driver should identify itself a la that last line. Make sure you see that, and also see if there is an update being applied.

Last edited by ewaller (2017-02-14 16:52:11)

ku_respawned · 2017-02-14 17:27:21

Feb 13 20:00:32 entheogenesis kernel: microcode: CPU0 sig=0x306a9, pf=0x2, revision=0x1c
Feb 13 20:00:32 entheogenesis kernel: microcode: CPU1 sig=0x306a9, pf=0x2, revision=0x1c
Feb 13 20:00:32 entheogenesis kernel: microcode: CPU2 sig=0x306a9, pf=0x2, revision=0x1c
Feb 13 20:00:32 entheogenesis kernel: microcode: CPU3 sig=0x306a9, pf=0x2, revision=0x1c
Feb 13 20:00:32 entheogenesis kernel: microcode: Microcode Update Driver: v2.01 <tigran@aivazian.fsnet.co.uk>, Peter Oruba
Feb 14 14:03:45 entheogenesis kernel: microcode: CPU0 microcode updated early to revision 0x1c, date = 2015-02-26
Feb 14 14:03:45 entheogenesis kernel: microcode: CPU1 microcode updated early to revision 0x1c, date = 2015-02-26
Feb 14 14:03:45 entheogenesis kernel: microcode: CPU2 microcode updated early to revision 0x1c, date = 2015-02-26
Feb 14 14:03:45 entheogenesis kernel: microcode: CPU3 microcode updated early to revision 0x1c, date = 2015-02-26

seth · 2017-02-14 17:34:32

Check whether you can still enter VT1 - sounds like a stale keyboard/mouse grab. If you can either enter VT1 or ssh into the machine, it's possible to figure why the input fails (and hopefully fix that)

ku_respawned · 2017-02-14 17:37:54

seth wrote:

Check whether you can still enter VT1 - sounds like a stale keyboard/mouse grab. If you can either enter VT1 or ssh into the machine, it's possible to figure why the input fails (and hopefully fix that)

I can't. That was the first thing I tried. What can I do when I login to SSH?

seth · 2017-02-14 20:04:59

Depends severely on whether this is an xorg or wayland (default for gnome) session (which would actually be a nice cross test anyway)
=> this is what kind of session? ;-)

ku_respawned · 2017-02-14 22:35:33

Xorg with Fluxbox.

seth · 2017-02-14 23:50:50

install xdotool and run

export DISPLAY=:0
xterm & # see whether the server still responds: likely
xdotool key "XF86LogGrabInfo"

Your xorg log afterwards contains a grab status which will tell you whether and which client holds passive and active grabs (in doubt paste the section - in code tags, if here)

Is this fluxbox from arch or git (1.4 rc)?

ku_respawned · 2017-02-16 13:52:49

Run it from SSH after another input freeze? OK. So far that didn't happen anymore (and I still need to install and set up a SSH server here and a client on my notebook).

It's fluxbox from arch' repos, not git.

seth · 2017-02-16 15:52:46

Run it from SSH after another input freeze?

Yes. The point is to regain "some" input access and then start to inspect the system. If input via ssh still works, that's the way to go.
Also check dmesg if the above is unspecific (look out for usb/xhci_hid error messages)

ku_respawned · 2017-02-19 22:16:02

OK, so, xdotool's output here: https://bpaste.net/show/15623e999cbb (note that I tried it out when you told me about it and it's probably doubled)
Can you really make out what I wrote here the whole time by that? Maybe I should remove it, but I couldn't find anything useful, so I uploaded the whole thing.

Anyway, I was wrong, I was indeed able to switch to VT2 and kill Xorg. After that the screen blacked out. I noticed that the backlight is probably gradually going off, but the monitor stays on. Then I rebooted through Ctrl-Alt-Delete and after some time the backlight gradually went on and Windows booted.

In dmesg log (https://bpaste.net/show/72fd0d762b64) I found following in the first third:

[    7.614006] input: Logitech USB Keyboard as /devices/pci0000:00/0000:00:1d.0/usb4/4-1/4-1.5/4-1.5.1/4-1.5.1:1.0/0003:046D:C31D.0001/input/input6
[    7.672010] hid-generic 0003:046D:C31D.0001: input,hidraw0: USB HID v1.10 Keyboard [Logitech USB Keyboard] on usb-0000:00:1d.0-1.5.1/input0
[    7.678361] input: Logitech USB Keyboard as /devices/pci0000:00/0000:00:1d.0/usb4/4-1/4-1.5/4-1.5.1/4-1.5.1:1.1/0003:046D:C31D.0002/input/input7
[    7.732368] hid-generic 0003:046D:C31D.0002: input,hidraw1: USB HID v1.10 Device [Logitech USB Keyboard] on usb-0000:00:1d.0-1.5.1/input1

and notice how in the end only the mouse is reattached (but buttons don't work). I believe, the first one is a dissapearing of connection and the second is switching to my notebook for SSH:

[226154.323171] usb 4-1.5.2: USB disconnect, device number 7
[226197.696856] usb 4-1.5.2: new full-speed USB device number 8 using ehci-pci
[226197.855515] input: Logitech Gaming Mouse G600 as /devices/pci0000:00/0000:00:1d.0/usb4/4-1/4-1.5/4-1.5.2/4-1.5.2:1.0/0003:046D:C24A.0007/input/input36
[226197.855674] hid-generic 0003:046D:C24A.0007: input,hidraw2: USB HID v1.11 Mouse [Logitech Gaming Mouse G600] on usb-0000:00:1d.0-1.5.2/input0
[226197.879040] input: Logitech Gaming Mouse G600 as /devices/pci0000:00/0000:00:1d.0/usb4/4-1/4-1.5/4-1.5.2/4-1.5.2:1.1/0003:046D:C24A.0008/input/input37
[226197.937196] hid-generic 0003:046D:C24A.0008: input,hiddev0,hidraw3: USB HID v1.11 Keyboard [Logitech Gaming Mouse G600] on usb-0000:00:1d.0-1.5.2/input1
[226198.230022] usb 4-1.5.2: USB disconnect, device number 8

Any ideas what could cause it? I really don't want to connect the keyboard directly to the PC, because it's way more comfotarble than the notebook's keyboard and like I said, it worked on my notebook at the same time as it didn't on my PC.

P.S. Oh, yeah, it shows me the mouse keyboard (I've got the 12 extra buttons), but I can't use it when it freezes as well, at least the last time I tried.
And yes, switching to other VTs and using them works as well, so the normal keyboard actually should be reconnected as well, even though it doesn't appear in the logs. Or is the keyboard just "named" "after the mouse", because they share one port?

Arch Linux

#1 2017-01-14 22:00:12

System hangs at randome times (mostly while writing a lot of text)

#2 2017-01-15 13:15:39

Re: System hangs at randome times (mostly while writing a lot of text)

#3 2017-01-15 15:47:39

Re: System hangs at randome times (mostly while writing a lot of text)

#4 2017-01-15 16:29:21

Re: System hangs at randome times (mostly while writing a lot of text)

#5 2017-01-15 18:16:56

Re: System hangs at randome times (mostly while writing a lot of text)

#6 2017-01-15 19:07:58

Re: System hangs at randome times (mostly while writing a lot of text)

#7 2017-01-15 21:58:23

Re: System hangs at randome times (mostly while writing a lot of text)

#8 2017-01-15 23:23:25

Re: System hangs at randome times (mostly while writing a lot of text)

#9 2017-02-01 21:41:35

Re: System hangs at randome times (mostly while writing a lot of text)

#10 2017-02-04 20:52:56

Re: System hangs at randome times (mostly while writing a lot of text)

#11 2017-02-13 21:55:40

Re: System hangs at randome times (mostly while writing a lot of text)

#12 2017-02-14 05:05:47

Re: System hangs at randome times (mostly while writing a lot of text)

#13 2017-02-14 15:09:12

Re: System hangs at randome times (mostly while writing a lot of text)

#14 2017-02-14 15:15:11

Re: System hangs at randome times (mostly while writing a lot of text)

#15 2017-02-14 16:41:24

Re: System hangs at randome times (mostly while writing a lot of text)

#16 2017-02-14 16:51:51

Re: System hangs at randome times (mostly while writing a lot of text)

#17 2017-02-14 17:27:21

Re: System hangs at randome times (mostly while writing a lot of text)

#18 2017-02-14 17:34:32

Re: System hangs at randome times (mostly while writing a lot of text)

#19 2017-02-14 17:37:54

Re: System hangs at randome times (mostly while writing a lot of text)

#20 2017-02-14 20:04:59

Re: System hangs at randome times (mostly while writing a lot of text)

#21 2017-02-14 22:35:33

Re: System hangs at randome times (mostly while writing a lot of text)

#22 2017-02-14 23:50:50

Re: System hangs at randome times (mostly while writing a lot of text)

#23 2017-02-16 13:52:49

Re: System hangs at randome times (mostly while writing a lot of text)

#24 2017-02-16 15:52:46

Re: System hangs at randome times (mostly while writing a lot of text)

#25 2017-02-19 22:16:02

Re: System hangs at randome times (mostly while writing a lot of text)

Board footer