Bug 1785858

Summary: startx script does not start X
Product: [Fedora] Fedora Reporter: Andrew <travneff>
Component: xorg-x11-xinitAssignee: X/OpenGL Maintenance List <xgl-maint>
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 31CC: airlied, ajax, caillon+fedoraproject, jglisse, john.j5live, negativo17, rhughes, rstrode, travneff, xgl-maint
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2020-01-03 08:17:04 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
kernel log with drm.debug=14
none
xorg log file
none
/var/log/messages for several sessions none

Description Andrew 2019-12-21 16:45:41 UTC
Created attachment 1647038 [details]
kernel log with drm.debug=14

Usecase: boot with /etc/systemd/system/default.target symlinked to /lib/systemd/system/multi-user.target, authenticate in text console and run startx.

It worked recently but now seems like X does start and immediate exit for unknown reason - I see no obvious errors in X log, dmesg and /var/log/messages.
Can start X with launching kdm.service from root and selecting plasma session ("default" or "custom" seem just do the same - X does not start, KDM restarts). First try showed a fresh plasma session (lost all my desktop settings) but subsequent try handled my settings back, currently writing from it.
Root user (tried just for test) and another test user have the same in general.

If run this instead of startx, it does not exit like startx but shows black screen only:
  /usr/libexec/Xorg :0 vt1 -nolisten tcp

If run `systemctl isolate graphical.target`, then `systemctl status graphical.target` shows success but X is not started.

Versions:
  xorg-x11-xinit-1.4.0-5.fc31.x86_64
  xorg-x11-server-Xorg-1.20.6-1.fc31.x86_64
  xorg-x11-drv-intel-2.99.917-43.20180618.fc31.x86_64
  kernel-5.3.16-300.fc31.x86_64
  (tried as well: 5.3.12-300.fc31.x86_64, kernel-5.3.13-300.fc31.x86_64)

HW:
CPU: Intel(R) Core(TM) i5-6440HQ CPU @ 2.60GHz
VGA: 00:02.0 VGA compatible controller [0300]: Intel Corporation HD Graphics 530 [8086:191b] (rev 06)
     DeviceName:  Onboard IGD
     Subsystem: Dell Device [1028:06de]
     Kernel driver in use: i915
     Kernel modules: i915

Comment 1 Andrew 2019-12-21 16:46:37 UTC
Created attachment 1647039 [details]
xorg log file

Comment 2 Andrew 2019-12-21 16:47:52 UTC
Created attachment 1647040 [details]
/var/log/messages for several sessions

Comment 3 Michel Dänzer 2019-12-23 11:53:18 UTC
(In reply to Andrew Travneff from comment #0)
> It worked recently but now seems like X does start and immediate exit for
> unknown reason - I see no obvious errors in X log, dmesg and
> /var/log/messages.

That sounds like there's no problem with the X server, but rather with the client session, causing it to exit immediately. Check the files referenced by the startx(1) manpage, in particular ~/.startxrc and ~/.xinitrc .


> If run this instead of startx, it does not exit like startx but shows black
> screen only:
>   /usr/libexec/Xorg :0 vt1 -nolisten tcp

That's the default behaviour of Xorg waiting for clients to connect.

Comment 4 Andrew 2019-12-25 11:09:52 UTC
Thank you for the info, I found the reason with it maybe.
`man startx` lists these files:

    $(HOME)/.startxrc
    /usr/lib64/sys.startxrc
    $(HOME)/.xinitrc
    /etc/X11/xinit/xinitrc

Only the last exists for me.
Firstly it calls $HOME/.Xclients (absent for me) and then /etc/X11/xinit/Xclients
The latter reads /etc/sysconfig/desktop[1] and shall run startkde - but startkde is absent in plasma-workspace-5.17.4-1.fc31.x86_64 (my installed version) and was present in plasma-workspace-5.16.5-1.fc31.x86_64 (the only location of startkde reported by dnf[2]).
So, this is the reason I guess (did not trace it for sure, please tell if necessary).

  1:

$ cat /etc/sysconfig/desktop
DESKTOP="KDE"
DISPLAYMANAGER="KDE"

  2:

# dnf -Cq provides '*/startkde'
plasma-workspace-5.16.5-1.fc31.i686 : Plasma workspace, applications and applets
Repo        : fedora
Matched from:
Filename    : /usr/bin/startkde

plasma-workspace-5.16.5-1.fc31.x86_64 : Plasma workspace, applications and applets
Repo        : fedora
Matched from:
Filename    : /usr/bin/startkde

Comment 5 Andrew 2019-12-25 21:53:56 UTC
I downloaded plasma-workspace-5.16.5-1.fc31.x86_64 rpm and unpacked startkde to my /user/bin
After this got error in GUI window after startx:
    kstartupconfig5 does not exist or fails. The error code is 127. Check your installation

Then I unpacked all files which are absent under my /usr/bin/ : startkde, kstartupconfig5, kdostartupconfig5, kcheckrunning (total = 4).
For this case startx gives black screen with mouse cursor. At least some processes for X and KDE are shown running by `ps` while this.

Comment 6 Michel Dänzer 2020-01-03 08:17:04 UTC

*** This bug has been marked as a duplicate of bug 1785826 ***