The new machine (see signature) runs very stable and issues no messages to the journal. However opening a Konsole, logging in to root with su - and issuing some commands for some 10 minutes frequently results in a freeze of the system. Need to press reset button or power off to continue. Even after pressing the reset button it takes many seconds until the machine reboots. No messages are written to the journal.
If this is a brand new installation,
Assuming you installed from DVD and did not enable online sources (default install),
You should always update your machine immediately, and then depending on your preference may or may not update again later(You should at least patch which would apply security fixes but wouldn’t install improvements).
To update your machine so you get all fixes and improvements since the DVD was first released, you should run the following in a root console
zypper update
TSU
Have you tried the SysRq (Sys Request) methods to see if you can get some messages to the journal?
First, check that SysRq is enabled:
cat /proc/sys/kernel/sysrq
If not enabled, this will return a 0.
You would probably prefer it to return a 1. To enable it (as root):
echo 1 > /proc/sys/kernel/sysrq
Next time it freezes, try these keystrokes in the following order:
Alt-SysRq-r (Takes the control of the keyboard away from the X server and out of raw mode.)
After this, you might be able to use the Ctrl-Alt-F(n) keystroke to get to another console and restart X.
If not, continue with:
Alt-SysRq-k (kills X and all programs on the current console. You might now be able to press the Ctrl-Alt-Backspace combo to restart X Server.
Still did not work? Then:
Alt-SysRq-r (release the keyboard)
Alt-SysRq-e (send the terminate signal to all running processes)
Alt-SysRq-i (send the kill signal to alll processes to force them to quit)
Alt-SysRq-s (sync, flushes the data from your cache to disk)
Alt-SysRq-u (remounts all the file systems as read-only)
Alt-SysRq-b (reboots your computer)
or
Alt-SysRq-o shuts off the computer.
Good luck.
I installed Leap 42.1 also on harddisk. Screen freezing occurs too, but now there is a hint in the journal:
Aug 21 07:54:54 linux-cqkj kernel: [drm] stuck on render ring
Aug 21 07:54:54 linux-cqkj kernel: [drm] GPU HANG: ecode 9:0:0x85dffffb, in java [4350], reason: Ring hung, action: reset
Aug 21 07:54:54 linux-cqkj kernel: [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
Aug 21 07:54:54 linux-cqkj kernel: [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
Aug 21 07:54:54 linux-cqkj kernel: [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
Aug 21 07:54:54 linux-cqkj kernel: [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
Aug 21 07:54:54 linux-cqkj kernel: [drm] GPU crash dump saved to /sys/class/drm/card0/error
Aug 21 07:54:54 linux-cqkj kernel: ------------ cut here ]------------
Aug 21 07:54:54 linux-cqkj kernel: WARNING: CPU: 5 PID: 5013 at ../drivers/gpu/drm/i915/intel_display.c:10211 intel_mmio_flip_work_func+0x2b2/0x2e0 [i915]()
Aug 21 07:54:54 linux-cqkj kernel: WARN_ON(__i915_wait_request(mmio_flip->req, crtc->reset_counter, false, NULL, NULL) != 0)
Aug 21 07:54:54 linux-cqkj kernel: Modules linked in:
Aug 21 07:54:54 linux-cqkj kernel: fuse bnep bluetooth rfkill nf_log_ipv6 xt_pkttype nf_log_ipv4 nf_log_common xt_LOG xt_limit af_packet iscsi_ibft iscsi_boot_sysfs ip6t_REJECT nf_reject_ipv6 xt_tcpudp nf_conntrack_ipv6 nf_defrag_ipv6 ip6table_raw ipt_REJECT nf_reject_i
Aug 21 07:54:54 linux-cqkj kernel: serio_raw lrw i2c_hid wmi gf128mul acpi_cpufreq processor tpm_tis tpm glue_helper ablk_helper cryptd pcspkr battery acpi_pad 8250_fintek crc32c_intel i915 xhci_pci i2c_algo_bit xhci_hcd drm_kms_helper usbcore drm usb_common video butto
Aug 21 07:54:54 linux-cqkj kernel: CPU: 5 PID: 5013 Comm: kworker/5:1 Not tainted 4.1.27-27-default #1
Aug 21 07:54:54 linux-cqkj kernel: Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./Z170 Pro4S, BIOS P3.50 06/23/2016
Aug 21 07:54:54 linux-cqkj kernel: Workqueue: events intel_mmio_flip_work_func [i915]
Aug 21 07:54:54 linux-cqkj kernel: 0000000000000286 0000000000000000 ffffffff8165f3d0 0000000000000000
Aug 21 07:54:54 linux-cqkj kernel: ffff88032128bd58 ffffffffa02420a8 ffffffff81068951 ffff880076bdc8b8
Aug 21 07:54:54 linux-cqkj kernel: ffff8804600c8b80 ffff880475d55f00 ffff880076bdc000 0000000000000000
Aug 21 07:54:54 linux-cqkj kernel: Call Trace:
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff810055cc>] dump_trace+0x8c/0x340
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff8100597c>] show_stack_log_lvl+0xfc/0x1a0
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff81006ec1>] show_stack+0x21/0x50
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff8165f3d0>] dump_stack+0x5d/0x79
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff81068951>] warn_slowpath_common+0x81/0xb0
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff810689ca>] warn_slowpath_fmt+0x4a/0x50
Aug 21 07:54:54 linux-cqkj kernel: <ffffffffa01dd932>] intel_mmio_flip_work_func+0x2b2/0x2e0 [i915]
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff81080fc5>] process_one_work+0x145/0x450
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff810813e4>] worker_thread+0x114/0x470
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff81086971>] kthread+0xc1/0xe0
Aug 21 07:54:54 linux-cqkj kernel: <ffffffff81665fe2>] ret_from_fork+0x42/0x70
Aug 21 07:54:54 linux-cqkj kernel: --- end trace 32737fd6891ce8a3 ]---
Aug 21 07:54:54 linux-cqkj kernel: drm/i915: Resetting chip after gpu hang
Aug 21 07:54:56 linux-cqkj kernel: [drm] RC6 on
Aug 21 07:55:25 linux-cqkj kernel: [drm:gen8_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
Many thanks for pointing to Alt-SysRq-r. Suspecting the Samsung 950 PRO could be the culprit was helpful, since when booting from the HDD the traceback showed up pointing to https://bugs.freedesktop.org/show_bug.cgi?id=94101. Modified the Grub command line:
linux-lfh1:~ # grep i915.enable_rc6=0 /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="resume=/dev/nvme0n1p1 splash=silent quiet showopts i915.enable_rc6=0"
linux-lfh1:~ #
No more hangs occur since booting with the above.
Welcome.
No more hangs occur since booting with the above.
… and that is good to know. Great!
Well, there were still some freezes. Needed to disable hardware acceleration: Intel graphics - ArchWiki
The boot now survived until a kernel security patch required rebooting.
Currently, the Leap 42.2 Beta is using the 4.4 (LTS) kernel.
Arch Linux: Intel graphics: Skylake support
Note: Fixes for the GPU/DRM bugs are pending in kernel 4.6. The following steps are unnecessary if you have testing repository enabled, or once 4.6 lands in core.
Upgraded a Haswell machine to Tumbleweed and got some screen flicker. Turned off Panel Self Refresh (PSR) as follows:
hofkirchen:~ # grep enable /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="video=1920x1200 resume=/dev/disk/by-id/ata-Samsung_SSD_840_EVO_250GB_S1DBNSCFA21657V-part1 splash=silent quiet showopts vga=894 i915.enable_psr=0"
hofkirchen:~ #
Flicker now is gone.