Every now and then (luckily, not too frequent) my notebooks screen just turns
black - no more output.
So far I was not able to pin it down to anything in particular, but could find
the following ouput logged to dmesg when this happens:
[78091.998108] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in
has_drrs (expected 1, found 0)
[78091.998121] ------------[ cut here ]------------
[78091.998166] WARNING: CPU: 2 PID: 1745 at
../drivers/gpu/drm/i915/intel_display.c:12700
intel_modeset_check_state+0x5aa/0x870 [i915]()
[78091.998169] pipe state doesn't match!
[78091.998172] Modules linked in: tun uas usb_storage hid_plantronics
snd_usb_audio snd_usbmidi_lib snd_rawmidi snd_seq_device usbhid hid_generic
hidp drbg ansi_cprng ctr ccm cmac ecb rfcomm fuse af_packet bnep uvcvideo
videobuf2_vmalloc videobuf2_memops videobuf2_core v4l2_common videodev btusb
btrtl btbcm btintel bluetooth xfs libcrc32c arc4 nls_iso8859_1 nls_cp437 vfat
fat tpm_infineon dm_mod x86_pkg_temp_thermal intel_powerclamp coretemp iTCO_wdt
kvm_intel iTCO_vendor_support kvm iwlmvm ppdev hp_wmi sparse_keymap
snd_hda_codec_hdmi crct10dif_pclmul mac80211 snd_hda_codec_idt crc32_pclmul
snd_hda_codec_generic aesni_intel aes_x86_64 snd_hda_intel lrw iwlwifi gf128mul
glue_helper ablk_helper cryptd snd_hda_codec snd_hda_core snd_hwdep rtsx_pci_ms
snd_pcm joydev pcspkr lpc_ich memstick cfg80211 snd_timer
[78091.998281] i2c_i801 hp_accel lis3lv02d parport_pc 8250_fintek parport
rfkill snd input_polldev wmi thermal shpchp battery mei_me mei tpm_tis
soundcore tpm hp_wireless intel_smartconnect processor ac btrfs xor raid6_pq
rtsx_pci_sdmmc mmc_core crc32c_intel rtsx_pci mfd_core serio_raw ehci_pci
ehci_hcd xhci_pci xhci_hcd e1000e usbcore ptp pps_core usb_common fjes i915
video button i2c_algo_bit drm_kms_helper syscopyarea sysfillrect sysimgblt
fb_sys_fops drm sg efivarfs
[78091.998351] CPU: 2 PID: 1745 Comm: Xorg Tainted: G W
4.3.0-2-default #1
[78091.998355] Hardware name: Hewlett-Packard HP EliteBook 820 G1/1991, BIOS
L71 Ver. 01.07 02/09/2014
[78091.998359] ffffffffa01cf898 ffff88009a397b20 ffffffff81363719
ffff88009a397b68
[78091.998365] ffff88009a397b58 ffffffff81068cf2 ffff8801087dd400
ffff8800b8195000
[78091.998371] ffff88007eb7ec00 ffff880139363800 0000000000000000
ffff88009a397bb8
[78091.998377] Call Trace:
[78091.998398] [<ffffffff81008385>] try_stack_unwind+0x175/0x190
[78091.998413] [<ffffffff81007203>] dump_trace+0x93/0x3a0
[78091.998423] [<ffffffff810083eb>] show_trace_log_lvl+0x4b/0x60
[78091.998431] [<ffffffff8100761c>] show_stack_log_lvl+0x10c/0x180
[78091.998438] [<ffffffff81008485>] show_stack+0x25/0x50
[78091.998450] [<ffffffff81363719>] dump_stack+0x4b/0x72
[78091.998463] [<ffffffff81068cf2>] warn_slowpath_common+0x82/0xc0
[78091.998473] [<ffffffff81068d7c>] warn_slowpath_fmt+0x4c/0x50
[78091.998516] [<ffffffffa01655ba>] intel_modeset_check_state+0x5aa/0x870
[i915]
[78091.998564] [<ffffffffa01706a4>] intel_atomic_commit+0x434/0x640 [i915]
[78091.998608] [<ffffffffa0035487>] drm_atomic_commit+0x37/0x60 [drm]
[78091.998629] [<ffffffffa00b690f>] drm_atomic_helper_set_config+0x1bf/0x420
[drm_kms_helper]
[78091.998658] [<ffffffffa0024d02>] drm_mode_set_config_internal+0x62/0x100
[drm]
[78091.998688] [<ffffffffa0029148>] drm_mode_setcrtc+0x3a8/0x4c0 [drm]
[78091.998708] [<ffffffffa001a458>] drm_ioctl+0x138/0x500 [drm]
[78091.998718] [<ffffffff811f5645>] do_vfs_ioctl+0x285/0x460
[78091.998732] [<ffffffff811f5899>] SyS_ioctl+0x79/0x90
[78091.998745] [<ffffffff8166e4b6>] entry_SYSCALL_64_fastpath+0x16/0x75
[78092.003293] DWARF2 unwinder stuck at entry_SYSCALL_64_fastpath+0x16/0x75
[78092.003299] Leftover inexact backtrace:
[78092.003307] ---[ end trace b9851fcbefc0ed9f ]---
My current quickest approach to recover from this is to push the suspend
buttont - let the system go to Suspend-to-ram and then start it up again (so it
costs me about 30 seconds total time).
I attach dmesg output which contains the time when it goes black, suspend,
resume and re-operational states.
(note: I cleared dmesg earlier during the day, so the file is not that big -
but covers the entire period of the issue described)