System freezes after some uptime

**amakaresu** · 17-Jan-2017, 05:38

Hello!
I have one big problem with my fresh opensuse installation:
Everything is good, but sometimes (and always after 12h of uptime) my system just freezes, without any logical explanation

Behavior after freeze:
Can't do almost anything, hotkeys doesn't work (tryed ctrl+esc, ctrl+alt+t, ctrl+alt+l), except ctrl+alt+f1 (i get to tty1 and can do anything there, without any glitches)
Still, after switching back to x-session workspace remains frozen
When in "frozen" state my mouse cursor movable, but any clicks do nothing.

Environment:
Kernel: 4.4.36-8
OS: OpenSuSe Leap 42.2 x64
Hardware:
ThinkPad T410
intel i5 m420
6gb RAM ddr3
integrated graphics (driver=i915)
chipset: intel 5 Series/3400 Series Chipset
Software: nothing special, fresh install. I have virtualbox 5.1.8 and LAMP server, also multimedia codecs from "pacman" repo.

What i've tried:
When my system frozen i've tried to (in tty1):

Code:

journalctl

the only error which is always happens right before freeze is:

Code:

Jan 17 14:37:15 linux-aqta.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
Jan 17 14:37:15 linux-aqta.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder A FIFO underrun

I don't understand this error, and wasn't able to find anything helpful in google.
also looked in Xord log file

Code:

cat /var/log/Xorg.0.log | grep EE
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[    51.995] (EE) evdev: TPPS/2 IBM TrackPoint: Unable to query fd: Inappropriate ioctl for device
[    52.024] (EE) PreInit returned 2 for "TPPS/2 IBM TrackPoint"

Tried to SIG_TERM processes, killing plasmashell don't help - system still in same state
Commands like "init 3" or "systemctl restart sddm" make my system responsive again

i've even tried to reinstall my system (bad habbit after long time on windows) - didn't help, everything same.

I need your help to debug this error! Sorry for poor english!

**OrsoBruno** · 17-Jan-2017, 06:52

Originally Posted by amakaresu

...
the only error which is always happens right before freeze is:

Code:

Jan 17 14:37:15 linux-aqta.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
Jan 17 14:37:15 linux-aqta.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder A FIFO underrun

I don't understand this error, and wasn't able to find anything helpful in google.

Hi, I remember seeing something similar when on openSUSE 13.2 and a 965GM Intel GPU.
Something related to the i915 driver, but can't recall at the moment. Maybe switching back to the UXA acceleration might help, or some other option of the i915 video driver.
Will write back if I find something useful.

**amakaresu** · 17-Jan-2017, 07:28

Thanks for reply!
Found some suggestions on arch wiki
Added /etc/X11/xorg.conf.d/20-intel.conf
Added

Code:

Section "Device"
   Identifier  "Intel Graphics"
   Driver      "intel"
   Option      "NoAccel" "True"
EndSection

Will report back after 2 days (or earlier, if freeze will occur)

some additional info:
Before switching to opensuse 42.2 i was on opensuse 42.1
there wasn't such error.
some other outputs:

Code:

sudo journalctl | grep i915
root's password:
янв 16 13:48:57 linux-aqta kernel: [drm] Initialized i915 1.6.0 20151010 for 0000:00:02.0 on minor 0
янв 16 13:48:58 linux-aqta kernel: [drm] GMBUS [i915 gmbus dpb] timed out, falling back to bit banging on pin 5
янв 16 13:48:59 linux-aqta kernel: i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
янв 16 17:17:22 linux-aqta kernel: [drm] Initialized i915 1.6.0 20151010 for 0000:00:02.0 on minor 0
янв 16 17:17:23 linux-aqta kernel: [drm] GMBUS [i915 gmbus dpb] timed out, falling back to bit banging on pin 5
янв 16 17:17:24 linux-aqta kernel: i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
янв 16 18:49:58 linux-aqta kernel: [drm] Initialized i915 1.6.0 20151010 for 0000:00:02.0 on minor 0
янв 16 18:49:58 linux-aqta kernel: [drm] GMBUS [i915 gmbus dpb] timed out, falling back to bit banging on pin 5
янв 16 18:49:58 linux-aqta kernel: i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
янв 17 14:37:15 linux-aqta.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
янв 17 14:37:15 linux-aqta.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder A FIFO underrun
янв 17 14:49:51 linux-aqta kernel: [drm] Initialized i915 1.6.0 20151010 for 0000:00:02.0 on minor 0
янв 17 14:49:51 linux-aqta kernel: [drm] GMBUS [i915 gmbus dpb] timed out, falling back to bit banging on pin 5
янв 17 14:49:52 linux-aqta kernel: i915 0000:00:02.0: fb0: inteldrmfb frame buffer devi

Code:

dmesg | grep i915
[    3.432661] [drm] Initialized i915 1.6.0 20151010 for 0000:00:02.0 on minor 0
[    3.494897] [drm] GMBUS [i915 gmbus dpb] timed out, falling back to bit banging on pin 5
[    4.094246] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device

**OrsoBruno** · 17-Jan-2017, 07:47

Originally Posted by amakaresu

Thanks for reply!
Found some suggestions on arch wiki
Added /etc/X11/xorg.conf.d/20-intel.conf
Added

Code:

Section "Device"
   Identifier  "Intel Graphics"
   Driver      "intel"
   Option      "NoAccel" "True"
EndSection

Will report back after 2 days (or earlier, if freeze will occur)

OK, this is rather extreme but if you don't need acceleration that will solve your problem if I recall correctly.
Using:

Code:

Option "AccelMethod" "UXA"

or:

Code:

Option "AccelMethod" "blt"

might solve your problem while still allowing some form of graphics acceleration.
At the time I had a similar problem an update to the (kernel) i915 driver eventually solved the problem.

Code:

dmesg | grep i915
[    3.432661] [drm] Initialized i915 1.6.0 20151010 for 0000:00:02.0 on minor 0
[    3.494897] [drm] GMBUS [i915 gmbus dpb] timed out, falling back to bit banging on pin 5
[    4.094246] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device

That is normal, nothing to worry about.

**amakaresu** · 18-Jan-2017, 01:42

Originally Posted by OrsoBruno

OK, this is rather extreme but if you don't need acceleration that will solve your problem if I recall correctly.

I want to be sure that problem comes from my v-card driver.
So far so good, i have some minor glitches on workspace (which tells me that my setting working)
Still need more time, will post back in 1 day or so, if problem will not apper, than i will use other rendering option as you suggested

**amakaresu** · 18-Jan-2017, 03:23

Hello again, i took less time than usual to catch a freeze.
So changing "NoAccel" to "True" in i915 driver didnt help

Posting last lines of journalctl and xorg.0.log

Code:

-- Logs begin at Mon 2017-01-16 13:48:57 MSK, end at Wed 2017-01-18 13:12:16 MSK. --
Jan 18 13:00:01 ivan-ntbk.suse CRON[14281]: pam_unix(crond:session): session closed for user root
Jan 18 13:00:01 ivan-ntbk.suse systemd[1]: Stopping User Manager for UID 0...
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Reached target Shutdown.
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Starting Exit the Session...
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Stopped target Default.
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Stopped target Basic System.
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Stopped target Timers.
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Stopped target Sockets.
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Stopped target Paths.
Jan 18 13:00:01 ivan-ntbk.suse systemd[14282]: Received SIGRTMIN+24 from PID 14322 (kill).
Jan 18 13:00:01 ivan-ntbk.suse systemd[14286]: pam_unix(systemd-user:session): session closed for user root
Jan 18 13:00:01 ivan-ntbk.suse systemd[1]: Stopped User Manager for UID 0.
Jan 18 13:00:01 ivan-ntbk.suse systemd[1]: Removed slice User Slice of root.
Jan 18 13:00:25 ivan-ntbk.suse smartd[1214]: Device: /dev/sda [SAT], SMART Usage Attribute: 194 Temperature_Celsius changed from 141 to 137
Jan 18 13:02:35 ivan-ntbk.suse winbindd[1665]: [2017/01/18 13:02:35.471530,  0] ../source3/libsmb/cliconnect.c:1895(cli_session_setup_spnego_send)
Jan 18 13:02:35 ivan-ntbk.suse winbindd[1665]:   Kinit for IVAN-NTBK$@ISIDA.RU to access cifs/dc01.isida.ru@ISIDA.RU failed: Preauthentication failed
Jan 18 13:02:35 ivan-ntbk.suse winbindd[1665]: [2017/01/18 13:02:35.553937,  0] ../source3/libsmb/cliconnect.c:1895(cli_session_setup_spnego_send)
Jan 18 13:02:35 ivan-ntbk.suse winbindd[1665]:   Kinit for IVAN-NTBK$@ISIDA.RU to access cifs/dc01.isida.ru@ISIDA.RU failed: Preauthentication failed
Jan 18 13:02:35 ivan-ntbk.suse winbindd[1665]: [2017/01/18 13:02:35.641476,  0] ../source3/libsmb/cliconnect.c:1895(cli_session_setup_spnego_send)
Jan 18 13:02:35 ivan-ntbk.suse winbindd[1665]:   Kinit for IVAN-NTBK$@ISIDA.RU to access cifs/dc01.isida.ru@ISIDA.RU failed: Preauthentication failed
Jan 18 13:05:02 ivan-ntbk.suse kernel: SFW2-INext-DROP-DEFLT IN=wlan0 OUT= MAC=33:33:00:00:00:fb:48:3b:38:71:6f:de:86:dd SRC=fe80:0000:0000:0000:1c8c:31f9:3cd3:fe78 DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=89 TC=0 HOPLIMIT=255 FLOWLBL=915684 PROTO=UDP SPT=5353 DPT=5353 LEN=49 
Jan 18 13:07:36 ivan-ntbk.suse winbindd[1665]: [2017/01/18 13:07:36.910305,  0] ../source3/libsmb/cliconnect.c:1895(cli_session_setup_spnego_send)
Jan 18 13:07:36 ivan-ntbk.suse winbindd[1665]:   Kinit for IVAN-NTBK$@ISIDA.RU to access cifs/dc01.isida.ru@ISIDA.RU failed: Preauthentication failed
Jan 18 13:07:37 ivan-ntbk.suse winbindd[1665]: [2017/01/18 13:07:37.034178,  0] ../source3/libsmb/cliconnect.c:1895(cli_session_setup_spnego_send)
Jan 18 13:07:37 ivan-ntbk.suse winbindd[1665]:   Kinit for IVAN-NTBK$@ISIDA.RU to access cifs/dc01.isida.ru@ISIDA.RU failed: Preauthentication failed
Jan 18 13:07:37 ivan-ntbk.suse winbindd[1665]: [2017/01/18 13:07:37.160633,  0] ../source3/libsmb/cliconnect.c:1895(cli_session_setup_spnego_send)
Jan 18 13:07:37 ivan-ntbk.suse winbindd[1665]:   Kinit for IVAN-NTBK$@ISIDA.RU to access cifs/dc01.isida.ru@ISIDA.RU failed: Preauthentication failed
Jan 18 13:08:45 ivan-ntbk.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
Jan 18 13:08:45 ivan-ntbk.suse kernel: [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder A FIFO underrun
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    address 192.168.10.181
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    plen 24 (255.255.255.0)
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    gateway 192.168.10.1
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    server identifier 192.168.10.7
Jan 18 13:08:47 ivan-ntbk.suse systemd[1]: Starting Network Manager Script Dispatcher Service...
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    lease time 3600
Jan 18 13:08:47 ivan-ntbk.suse systemd[1]: Started Network Manager Script Dispatcher Service.
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    nameserver '192.168.10.6'
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    domain name 'isida.ru'
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>    domain search 'isida.ru.'
Jan 18 13:08:47 ivan-ntbk.suse NetworkManager[1353]: <info>  (wlan0): DHCPv4 state changed bound -> bound
Jan 18 13:08:47 ivan-ntbk.suse dbus[1231]: [system] Activating via systemd: service name='org.freedesktop.nm_dispatcher' unit='dbus-org.freedesktop.nm-dispatcher.service'
Jan 18 13:08:47 ivan-ntbk.suse dbus[1231]: [system] Successfully activated service 'org.freedesktop.nm_dispatcher'
Jan 18 13:08:47 ivan-ntbk.suse nm-dispatcher[14417]: Dispatching action 'dhcp4-change' for wlan0
Jan 18 13:08:54 ivan-ntbk.suse login[1702]: pam_unix(login:session): session opened for user amakaresu by LOGIN(uid=0)
Jan 18 13:08:54 ivan-ntbk.suse systemd-logind[1676]: New session 22 of user amakaresu.
Jan 18 13:08:54 ivan-ntbk.suse systemd[1]: Started Session 22 of user amakaresu.
Jan 18 13:08:54 ivan-ntbk.suse login[1702]: LOGIN ON tty1 BY amakaresu
Jan 18 13:12:16 ivan-ntbk.suse sudo[14505]: amakaresu : TTY=tty1 ; PWD=/home/amakaresu ; USER=root ; COMMAND=/usr/bin/journalctl -n 50
Jan 18 13:12:16 ivan-ntbk.suse sudo[14505]: pam_unix(sudo:session): session opened for user root by amakaresu(uid=0)
Jan 18 13:12:16 ivan-ntbk.suse sudo[14505]: pam_systemd(sudo:session): Cannot create session: Already running in a session

Code:

[    49.066] (**) Option "Device" "/dev/input/event6"
[    49.066] (**) Option "_source" "server/udev"
[    49.067] (II) input device 'TPPS/2 IBM TrackPoint', /dev/input/event6 is tagged by udev as: Mouse Pointingstick
[    49.067] (II) input device 'TPPS/2 IBM TrackPoint', /dev/input/event6 is a pointer caps
[    49.108] (**) Option "config_info" "udev:/sys/devices/platform/i8042/serio1/serio2/input/input3/event6"
[    49.108] (II) XINPUT: Adding extended input device "TPPS/2 IBM TrackPoint" (type: MOUSE, id 12)
[    49.108] (**) Option "AccelerationScheme" "none"
[    49.108] (**) TPPS/2 IBM TrackPoint: (accel) selected scheme none/0
[    49.108] (**) TPPS/2 IBM TrackPoint: (accel) acceleration factor: 2.000
[    49.108] (**) TPPS/2 IBM TrackPoint: (accel) acceleration threshold: 4
[    49.109] (II) input device 'TPPS/2 IBM TrackPoint', /dev/input/event6 is tagged by udev as: Mouse Pointingstick
[    49.109] (II) input device 'TPPS/2 IBM TrackPoint', /dev/input/event6 is a pointer caps
[    49.109] (II) config/udev: Adding input device TPPS/2 IBM TrackPoint (/dev/input/mouse1)
[    49.109] (**) TPPS/2 IBM TrackPoint: Applying InputClass "Trackpoint Wheel Emulation"
[    49.109] (II) LoadModule: "evdev"
[    49.109] (II) Loading /usr/lib64/xorg/modules/input/evdev_drv.so
[    49.125] (II) Module evdev: vendor="X.Org Foundation"
[    49.125]     compiled for 1.18.3, module version = 2.10.3
[    49.125]     Module class: X.Org XInput Driver
[    49.125]     ABI class: X.Org XInput driver, version 22.1
[    49.125] (II) Using input driver 'evdev' for 'TPPS/2 IBM TrackPoint'
[    49.125] (**) TPPS/2 IBM TrackPoint: always reports core events
[    49.125] (**) evdev: TPPS/2 IBM TrackPoint: Device: "/dev/input/mouse1"
[    49.125] (EE) evdev: TPPS/2 IBM TrackPoint: Unable to query fd: Inappropriate ioctl for device
[    49.176] (EE) PreInit returned 2 for "TPPS/2 IBM TrackPoint"
[    49.176] (II) UnloadModule: "evdev"
[    49.176] (II) config/udev: Adding input device PC Speaker (/dev/input/event7)
[    49.176] (II) No input driver specified, ignoring this device.
[    49.176] (II) This device may have been added with another device file.
[    49.181] (II) config/udev: Adding input device ThinkPad Extra Buttons (/dev/input/event8)
[    49.181] (**) ThinkPad Extra Buttons: Applying InputClass "evdev keyboard catchall"
[    49.181] (**) ThinkPad Extra Buttons: Applying InputClass "system-keyboard"
[    49.181] (**) ThinkPad Extra Buttons: Applying InputClass "evdev keyboard catchall"
[    49.181] (**) ThinkPad Extra Buttons: Applying InputClass "libinput keyboard catchall"
[    49.181] (II) Using input driver 'libinput' for 'ThinkPad Extra Buttons'
[    49.181] (**) ThinkPad Extra Buttons: always reports core events
[    49.181] (**) Option "Device" "/dev/input/event8"
[    49.181] (**) Option "_source" "server/udev"
[    49.182] (II) input device 'ThinkPad Extra Buttons', /dev/input/event8 is tagged by udev as: Keyboard
[    49.182] (II) input device 'ThinkPad Extra Buttons', /dev/input/event8 is a keyboard
[    49.192] (**) Option "config_info" "udev:/sys/devices/platform/thinkpad_acpi/input/input9/event8"
[    49.192] (II) XINPUT: Adding extended input device "ThinkPad Extra Buttons" (type: KEYBOARD, id 13)
[    49.192] (**) Option "xkb_layout" "us"
[    49.193] (II) input device 'ThinkPad Extra Buttons', /dev/input/event8 is tagged by udev as: Keyboard
[    49.193] (II) input device 'ThinkPad Extra Buttons', /dev/input/event8 is a keyboard
[    50.072] (II) intel(0): EDID vendor "LEN", prod id 16438
[    50.072] (II) intel(0): Printing DDC gathered Modelines:
[    50.072] (II) intel(0): Modeline "1440x900"x0.0  114.06  1440 1488 1520 2044  900 903 909 930 -hsync -vsync (55.8 kHz eP)
[    50.072] (II) intel(0): Modeline "1440x900"x0.0  114.06  1440 1488 1520 2204  900 903 909 1035 -hsync -vsync (51.8 kHz e)
[ 17435.913] (II) AIGLX: Suspending AIGLX clients for VT switch

Maybe there are some other debugging methods? Other logs?

**OrsoBruno** · 18-Jan-2017, 07:33

Originally Posted by amakaresu

Hello again, i took less time than usual to catch a freeze.
So changing "NoAccel" to "True" in i915 driver didnt help

You are unlucky, it seems. This bugreport apparently describes the same issue (please note comment #5, also a T410...) https://bugs.freedesktop.org/show_bug.cgi?id=93310
The only thing I can say at this stage is: try booting with the "nomodeset" option: that way you completely bypass the i915 driver and use a generic driver.
To do that:
- at the GRUB boot screen press E
- look for a line beginning with "linux" (or "linuxefi", but I guess your laptop is a legacy model)
- add "nomodeset" (without quotes) at the end of that line, then press F10 to boot.
If that works, you can make it permanent by launching Yast Bootloader and adding nomodeset to the "Optional Kernel Command Line Parameter" line in the "Kernel Parameters" tab.
That should allow you to work until the issue is sorted out.
Or hope that somebody with more experience with Ironlakes helps you beyond this point...

**amakaresu** · 26-Jan-2017, 03:29

Originally Posted by OrsoBruno

You are unlucky, it seems. This bugreport apparently describes the same issue (please note comment #5, also a T410...) https://bugs.freedesktop.org/show_bug.cgi?id=93310
The only thing I can say at this stage is: try booting with the "nomodeset" option: that way you completely bypass the i915 driver and use a generic driver.
To do that:
- at the GRUB boot screen press E
- look for a line beginning with "linux" (or "linuxefi", but I guess your laptop is a legacy model)
- add "nomodeset" (without quotes) at the end of that line, then press F10 to boot.
If that works, you can make it permanent by launching Yast Bootloader and adding nomodeset to the "Optional Kernel Command Line Parameter" line in the "Kernel Parameters" tab.
That should allow you to work until the issue is sorted out.
Or hope that somebody with more experience with Ironlakes helps you beyond this point...

I've tried to work with "nomodeset" on, but there is too many glitches so i was forced to switch it off again
Main problem was that after "sleep" state my machine wasnt able to resume.

Problem still persists

**karlmistelberger** · 29-Jan-2017, 22:38

Originally Posted by amakaresu

I've tried to work with "nomodeset" on, but there is too many glitches so i was forced to switch it off again
Main problem was that after "sleep" state my machine wasnt able to resume.

Problem still persists

Experienced frequent bad freezes with Leap 42.1 related to i915. Upgraded to Tumbleweed. No more freezes since then.

**scapeuce** · 07-Aug-2017, 02:12

Two weeks ago I had decided to delete Opensuse Leap for ever, from my laptop. Too much work to do and the annoying freeze came 2 or 3 times a day. I was sos sad because I love this OS, so I chose to give it a last chance and started googling again for a recommendation no tried yet. I found, in another Opensuse thread a the same suggestion that amakaresu shows below, but with the other option, so my /etc/X11/xorg.conf.d/20-intel.conf file looks as:

Section "Device"
Identifier "Intel Graphics"
Driver "intel"
Option "AccelMethod" "uxa"
EndSection

My gosh !!! Since I applied this (2 weeks ago) I haven't experienced the problem anymore (I'am working more than 12 hours a day) ....
I couldn't find the original thread again to thank directly to whom posted it so I thank here to and the OpenSuse community for this high valuable help. My laptop is a Lenovo E550.
Thanks,
Andres