You are not logged in.
Hi all,
when booting 4.15 kernels (I tried 4.15.1 and 4.15.2), I only see a black screen after KMS and the lightDM login screen is never shown. Here are the journalctl entries for a working 4.14 kernel and a failing 4.15
4.14 kernel:
https://ptpb.pw/V6Po
4.15 kernel:
https://ptpb.pw/K1bo
A diff shows a lot of changes between the kernel versions, but I cannot identify an obvious problem for 4.15.
One message that is new when using pacman to upgrade the kernel is this (translation to english is mine):
Warning: Folder permissions are different for /boot/
Filesystem: 700 Package: 755Original:
Warnung: Verzeichnis-Berechtigungen unterscheiden sich für /boot/
Dateisystem: 700 Paket: 755I don't think this should be a major issue though.
Another thing that might be of interest is that the /boot drive has only 16MB free space left with 4.15 kernels. It was 18MB with 4.14 kernels. But here also, I don't think this is a critical issue.
I'm grateful for any suggestions how to fix this!
Last edited by YesMen (2018-02-10 23:08:04)
Offline
I think the chances you're using a UEFI system and your boot partition was not mounted or got corrupted while updating the kernel are high.
Your pacman log would be nice at this point, but first: how did you install Arch? How do you keep it updated (I mean, post the exact command you use to upgrade your packages).
Also, you created this thread. Leave aside that you didn't mind to mark it as [SOLVED], loqs helped you and at this point you should know how to keep your system updated nicely.
Offline
Hi lo1,
I updated the kernel to 4.15. and returned to 4.14 several times to check if anything went wrong on installing but there was no obvious trouble except the warning message posted above.
Edit: Also, there have been numerous kernel updates in the last months which worked fine, so I wouldn't suspect a corrupt UEFI partition to be the problem.
As to how I installed the system: I used a LiveCD as far as I remember. It was a couple of years ago so I cannot recall the steps taken. However I keep the system current with
pacman -Syuwhich works fine almost all of the time. I use this on every system startup so I am confident it is not a general issue with instaling updates.
Regarding my other thread that you pointed to: I must have missed the switch to set the threat to SOLVED. I will look into it.
However the problem I encountered there is not related to the current one, since now, the partition has free space left.
Last edited by YesMen (2018-02-10 19:33:06)
Offline
Details about your graphics stack would be helpful. Try disbaling your DM.
Moving to NC...
Offline
Details about your graphics stack would be helpful. Try disbaling your DM.
I'm not sure how this issue has to do with OP's graphics stack:
One message that is new when using pacman to upgrade the kernel is this (translation to english is mine):
Warning: Folder permissions are different for /boot/ Filesystem: 700 Package: 755
anyway, looking at his graphic setup (and maybe to an Xorg log) may be sane. But, my best guess is that the issue here is a wrong update or installation procedure (and most importantly this may have *everything* to do with OP's previous thread).
Offline
jasonwryan wrote:Details about your graphics stack would be helpful. Try disbaling your DM.
I'm not sure how this issue has to do with OP's graphics stack
The journal shows that it is not kernel related. His DM is not starting (which is mostly what they seem to do: not start).
Offline
Hi jasonwryan,
thanks for chiming in. Could you elaborate on how I can supply info about my graphics stack?
I tried to disable the DM for 4.15 by setting systemd.unit=rescue.target as well as multi-user.target but the screen stays black and no terminal appears. So there are no bootup messages either (the ones that are marked off with a green OK, when the service/job is succesfully started).
Last edited by YesMen (2018-02-10 20:38:30)
Offline
If you boot to multi-user.target and also use the nomodeset option do you reach a command prompt?
Offline
Hi loqs
thanks for your reply and yes, with nomodeset, the login prompt is reached, independet of a systemd.unit setting. Once logged in there, I tried to get lightdm working but it crashes with a seg fault.
I will try to switch lightdm with something else and see whree it goes from there.
Offline
Could you elaborate on how I can supply info about my graphics stack?
Details of your card(s), the driver you are using (including version number), X or Wayland. Xorg.log if X, etc.
Offline
Using nomodeset would expect you to encounter the same issue with other display managers and X in general unless you switch to using the unaccelerated vesa driver.
Adding the nomodeset option was to diagnose if the issue was caused by the kernel using modesetting not a fix. Please post the journal from a boot without nomodeset and without using the -x option.
See https://lists.freedesktop.org/archives/ … 13029.html.
Offline
Ah yes, sorry.
lspci:
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Ellesmere [Radeon RX 470/480/570/580] (rev cf)amdgpu driver:
xf86-video-amdgpu-1.4.0-1-x86_64.pkg.tar.xzUsing X.
Edit: I think this Xorg.0.log is from a failed boot:
https://ptpb.pw/1Pjy
Last edited by YesMen (2018-02-10 22:13:30)
Offline
Offline
That Xorg.log if from a boot with nomodeset. Looking at the boot from the journal
Feb 10 22:59:12 kernel: [drm] Cannot find any crtc or sizesinstead of from 14.4
Feb 10 17:43:59 kernel: [drm] AMDGPU Display Connectors
Feb 10 17:43:59 kernel: [drm] Connector 0:
Feb 10 17:43:59 kernel: [drm] DP-1
Feb 10 17:43:59 kernel: [drm] HPD6
Feb 10 17:43:59 kernel: [drm] DDC: 0x4868 0x4868 0x4869 0x4869 0x486a 0x486a 0x486b 0x486b
Feb 10 17:43:59 kernel: [drm] Encoders:
Feb 10 17:43:59 kernel: [drm] DFP1: INTERNAL_UNIPHY2
Feb 10 17:43:59 kernel: [drm] Connector 1:
Feb 10 17:43:59 kernel: [drm] DP-2
Feb 10 17:43:59 kernel: [drm] HPD4
Feb 10 17:43:59 kernel: [drm] DDC: 0x4870 0x4870 0x4871 0x4871 0x4872 0x4872 0x4873 0x4873
Feb 10 17:43:59 kernel: [drm] Encoders:
Feb 10 17:43:59 kernel: [drm] DFP2: INTERNAL_UNIPHY2
Feb 10 17:43:59 kernel: [drm] Connector 2:
Feb 10 17:43:59 kernel: [drm] DP-3
Feb 10 17:43:59 kernel: [drm] HPD1
Feb 10 17:43:59 kernel: [drm] DDC: 0x486c 0x486c 0x486d 0x486d 0x486e 0x486e 0x486f 0x486f
Feb 10 17:43:59 kernel: [drm] Encoders:
Feb 10 17:43:59 kernel: [drm] DFP3: INTERNAL_UNIPHY1
Feb 10 17:43:59 kernel: [drm] Connector 3:
Feb 10 17:43:59 kernel: [drm] HDMI-A-1
Feb 10 17:43:59 kernel: [drm] HPD5
Feb 10 17:43:59 kernel: [drm] DDC: 0x4874 0x4874 0x4875 0x4875 0x4876 0x4876 0x4877 0x4877
Feb 10 17:43:59 kernel: [drm] Encoders:
Feb 10 17:43:59 kernel: [drm] DFP4: INTERNAL_UNIPHY1
Feb 10 17:43:59 kernel: [drm] Connector 4:
Feb 10 17:43:59 kernel: [drm] DVI-D-1
Feb 10 17:43:59 kernel: [drm] HPD3
Feb 10 17:43:59 kernel: [drm] DDC: 0x487c 0x487c 0x487d 0x487d 0x487e 0x487e 0x487f 0x487f
Feb 10 17:43:59 kernel: [drm] Encoders:
Feb 10 17:43:59 kernel: [drm] DFP5: INTERNAL_UNIPHYOffline
So drm is the problem. But since the driver is the current version (as far as I know), I cannot fix this?
Edit: Aparently someone else has the same problem:
https://bugzilla.kernel.org/show_bug.cgi?id=198745
So it was a kernel issue to begin with ![]()
Thanks everybody for your help!
Last edited by YesMen (2018-02-10 22:49:12)
Offline
If it is an issue with DisplayCore then adding amdgpu.dc=0 to disable it might help.
Offline
Hmm, it should have been unset with 4.15. for Ellesmere chips. But I'll give it a try...
Edit: You've been spot on. It works with the setting =0. Guess you cannot trust phoronix articles then ![]()
Thanks!
Last edited by YesMen (2018-02-10 23:07:38)
Offline