Gentoo Forums :: View topic - [Enlightenment] won't start

[Enlightenment] won't start

View unanswered posts
View posts from last 24 hours
View posts from last 7 days

Gentoo Forums Forum Index

Desktop Environments

View previous topic :: View next topic

Author

Message

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Sat Apr 18, 2020 7:43 pm Post subject: [Enlightenment] won't start

Hi!

Trying to use Enlightenment wm but startx won't lauch it with following error:

Code:

Closing log file.t bus namexinit: connection to X server lost) was not propagated to all the parent classes? Forcibly removing it. This may leak! Refs: 0:0ERR<3892>

Code:

~/.xinitrc
exec dbus-launch --exit-with-session enlightenment_start

Xorg.0.log file, though it doesn't seem to come from Xorg-server.

This is with enlightenment-0.23.1 and efl-1.23.3
For xorg-drivers package we have:

Code:

INPUT_DEVICES="libinput joystick"
GRAPHIC_CARDS="i915 i965"

Any hints on this one?

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Sat Apr 25, 2020 1:40 pm Post subject:

bump...

Perfect Gentleman
Veteran
Veteran

Joined: 18 May 2014
Posts: 1252

Posted: Sat Apr 25, 2020 2:38 pm Post subject:

try

Code:

startx /usr/bin/enlightenment_start

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Sat Apr 25, 2020 6:26 pm Post subject:

Perfect Gentleman wrote:

try

Code:

startx /usr/bin/enlightenment_start

Thanks for joining

Just tried but I get the very same error message.

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Sun Apr 26, 2020 6:36 am Post subject:

Just

Code:

enlightenment_start

should be enough, since it needs to choose between wayland or xorg session.

Is your dbus running? That log snippet isn't telling much. Your dpaste got removed already, too. elogind/systemd running?

Please also tell

Code:

emerge -pv efl enlightenment

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Sun Apr 26, 2020 9:28 am Post subject:

Hi Juippisi,
A simple enlightenment_start gives the same error message.
Dbus and elogind are both running, according to rc-status. I am using openrc.

Here is the Xorg.0.log back.

Code:

# emerge -pv efl enlightenment
These are the packages that would be merged, in order:

Calculating dependencies ... done!
[ebuild R ] dev-libs/efl-1.23.3::gentoo USE="X bmp connman dds debug doc eet elogind fbcon gif gstreamer luajit nls opengl pdf sdl sound ssl svg system-lz4 tiff v4l xpm -drm -examples -fontconfig -fribidi -gles2 -glib -gnutls -harfbuzz -hyphen -ibus -ico -jpeg2k -json -libressl -lua -mono -physics -pmaps -postscript -psd -pulseaudio -raw -scim -static-libs -systemd -tga -tgv -tslib -unwind -vlc -vnc -wayland -webp -xcf -xim -xine -xpresent -zeroconf" 0 KiB
[ebuild R ] x11-wm/enlightenment-0.23.1:0.17/0.23.1::gentoo USE="acpi bluetooth connman doc geolocation nls pam udisks wifi (-packagekit) -systemd -wayland -xwayland" 0 KiB

Total: 2 packages (2 reinstalls), Size of downloads: 0 KiB

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Sun Apr 26, 2020 11:25 am Post subject:

Looks to me like your X is broken. What's even more weird is that your Xorg.0.log has some calls to systemd-logind...?

Does any other WM work? (A WM that requires logind or dbus)

What GPU are you using? It seems to try to load modules for pretty much everything, and fails. Have you installed drivers correctly for your GPU? Then I've never seen this before, but

Code:

[ 107.239] (EE) modeset(G0): glamor initialization failed

All in all your Xorg.0.log looks weird.

EDIT: If the problem only persists with enlightenment, then please capture the whole stder log with something like

Code:

enlightenment_start >> e_error_log.log 2>&1

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Sun Apr 26, 2020 2:59 pm Post subject:

Juippisi wrote:

Looks to me like your X is broken.

Nope it works fine:

Code:

[ 109.936] (II) Server terminated successfully (0). Closing log file.

I set it up with a big hand of NeddySeagoon: https://forums.gentoo.org/viewtopic-t-1110992-highlight-adrien.html

Quote:

Does any other WM work? (A WM that requires logind or dbus)

I am currently using LXDE and it works fine afaik

Juippisi wrote:

What GPU are you using? It seems to try to load modules for pretty much everything, and fails. Have you installed drivers correctly for your GPU?

I am using i915 driver, got a nvidia gpu too but it seems nouveau is not ready for it yet.

Here we go:

Code:

# enlightenment_start >> e_error_log.log 2>&1

output here

Thanks for your assistance!

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Mon Apr 27, 2020 7:25 am Post subject:

Adrien wrote:

Juippisi wrote:

What GPU are you using? It seems to try to load modules for pretty much everything, and fails. Have you installed drivers correctly for your GPU?

I am using i915 driver, got a nvidia gpu too but it seems nouveau is not ready for it yet.

and I see you had opengl enabled in efl. This may be problematic, since i915 doesn't support opengl-3 which if I remember correctly, efl does require. Try enabling gles2 instead and disabling opengl. Outcome will be the same.

https://en.wikipedia.org/wiki/List_of_Intel_graphics_processing_units#Third_generation
https://www.phoronix.com/scan.php?page=news_item&px=Mesa-i915-OpenGL-2-Drop

If the problem still persists (which I doubt) we need a more precise debugging. First install strace, then do

Code:

export E_START=1
strace -o strace.log enlightenment
(if it doesn't start, use startx enlightenment_start)

If the gles2-thing fixes it for you, I believe it may be handled in ebuild properly.

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Mon Apr 27, 2020 6:43 pm Post subject:

Juippisi wrote:

and I see you had opengl enabled in efl. This may be problematic, since i915 doesn't support opengl-3 which if I remember correctly, efl does require. Try enabling gles2 instead and disabling opengl. Outcome will be the same.
https://en.wikipedia.org/wiki/List_of_Intel_graphics_processing_units#Third_generation
https://www.phoronix.com/scan.php?page=news_item&px=Mesa-i915-OpenGL-2-Drop

Thanks for advice and links.
Just tried but still complains about msgbus

Xorg.0.log updated here

Juippisi wrote:

If the problem still persists (which I doubt) we need a more precise debugging.

Here is my strace.log file.
Did it with:

Code:

$ strace -o strace.log enlightenment_start

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Tue Apr 28, 2020 9:00 am Post subject:

I'm sorry to still bounce it back, but that strace doesn't tell much. It just tries to fork a process called "enlightenment" and fails.

Code:

export E_START=1
export DISPLAY=:0 (or whatever your DISPLAY usually is)
strace -o log enlightenment

should have a more meaningful log entry. You may need X env variables that login manager usually imports, but I doubt it. If you have en_US.utf8 locale installed, it'd be better to export that too for LC_ALL to see error messages in english.

Still I find it weird your xorg.0.log has systemd calls.. it may be the reason too. Could you paste your

Code:

rc-status

and

Code:

rc-update

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Tue Apr 28, 2020 6:53 pm Post subject:

Juippisi wrote:

I'm sorry to still bounce it back, but that strace doesn't tell much. It just tries to fork a process called "enlightenment" and fails.

Code:

export E_START=1
export DISPLAY=:0 (or whatever your DISPLAY usually is)
strace -o log enlightenment

No worries, apologies for that, here we go with the proper strace.log

I do not have en_US.utf8 installed but it shouldn't be too complicated for parsing the log:

Code:

"Aucun fichier ou dossier de ce type" <=> No such file or directory
"Argument invalide" <=> Invalid argument
"Le fichier existe" <=> File already exists

Code:

# rc-status
Runlevel: default
sysklogd [ started ]
acpid [ started ]
cronie [ started ]
dbus [ started ]
elogind [ started ]
net.wlan0 [ started ]
netmount [ started ]
numlock [ started ]
pcscd [ started ]
local [ started ]
Dynamic Runlevel: hotplugged
pcscd [ started ]
net.wlan0 [ started ]

and

Code:

Thanks for your patience.

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Wed Apr 29, 2020 9:36 am Post subject:

Adrien wrote:

Thanks for your patience. :)

Thanks for yours! Unfortunately I haven't seen this kind of error happen ever before.

Your elogind is started in "default", but it should be started at "boot". However I don't believe is the issue, since my elogind used to be "default" a long time as well. Still something you may want to check.

Good news and bad news. There is definitely something wrong in your dbus. For some reason, your login doesn't invocate a dbus session for your user. Can I ask a bit about the environment? Is this a virtual machine or something? Do you have known issues with your dbus sessions? Have you made some systemd -> openrc switch (that perhaps isn't fully completed)?

So what I would suggest now, is after you log in, do something like:

Code:

dbus-launch /bin/bash
startx enlightenment_startx

seeing the previous strace, this ... should at least work better. If it still doesn't launch Enlightenment, then within this dbus-launch'd bash shell, do the straces again.

Code:

dbus-launch /bin/bash
export E_START=1
export DISPLAY=:0 (or whatever your DISPLAY usually is)
strace -o log enlightenment

dbus is a hard-dependency of Enlightenment so it needs to be launched and working before Enlightenment can work.

EDIT: Well one thing did come to my mind... if you have enabled rc_parallel (for faster system startup), the misplaced elogind could be the cause... But in theory, as said, I've never seen a problem like this before.

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Wed Apr 29, 2020 11:53 am Post subject:

Juippisi wrote:

Thanks for yours! Unfortunately I haven't seen this kind of error happen ever before.

Makes your help even more appreciated

Juippisi wrote:

Your elogind is started in "default", but it should be started at "boot". However I don't believe is the issue, since my elogind used to be "default" a long time as well. Still something you may want to check.

Thanks, I've just corrected this

Juippisi wrote:

Can I ask a bit about the environment? Is this a virtual machine or something? Do you have known issues with your dbus sessions? Have you made some systemd -> openrc switch (that perhaps isn't fully completed)?

No virtual machine here, just a fresh Dell laptop and fresh install with openrc, did not ever use systemd on this machine and no issues with dbus afaik.

Juippisi wrote:

So what I would suggest now, is after you log in, do something like:

Code:

dbus-launch /bin/bash
startx enlightenment_startx

Still won't launch but the dbus error message seems to be gone.
Xorg.0.log here

And here's the new strace output

Juippisi wrote:

EDIT: Well one thing did come to my mind... if you have enabled rc_parallel (for faster system startup), the misplaced elogind could be the cause... But in theory, as said, I've never seen a problem like this before.

rc_parallel is set to "NO" (commented)
But just an hypothesis: could there be a kernel-related dbus issue? Indeed, I'm back to Gentoo after years and might have forgotten some option in .config.

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Wed Apr 29, 2020 3:39 pm Post subject:

So the new strace shows something,

Code:

connect(45, {sa_family=AF_UNIX, sun_path=@"/tmp/.X11-unix/X0"}, 20) = -1 ECONNREFUSED (Connexion refusée)

This is the problematic part. Now, there could be a 1000 different reasons why this happens...

Your user may not have correct permissions, your xorg-server most likely needs to be emerged with "suid" USE, your /tmp may have wrong permissions... and yeah, it could also be something with misconfigured kernel. It can also be due to you already having one instance of X open somewhere. You wouldn't happen to have some autolanchers? Or do you use a login manager and then Ctrl+Alt+F1 to go into TTY without turning off the login manager? Check "ps aux" in case.

So check that your user is in every correct group.
Check that your tmp and /tmp/.X11-unix/ and /tmp/.X11-unix/X0 have the correct permissions, for me they are

Code:

/
drwxrwxrwt 26 root root 580 Apr 29 18:26 tmp

/tmp
-r--r--r-- 1 root root 11 Apr 29 09:17 .X0-lock
drwxrwxrwt 2 root root 60 Apr 29 09:17 .X11-unix

/tmp/.X0-lock
srwxrwxrwx 1 root root 0 Apr 29 09:17 X0

note the "s" in "srwxrwxrwx". My /etc/fstab:

Code:

tmpfs /tmp tmpfs noatime,nodiratime,size=20G 0 0

(do note that these files probably only appear after X has been started once)

And in your kernel, if you use tmpfs for /tmp you will want xattr patch enabled and in use.

Code:

CONFIG_TMPFS_XATTR=y

Well there's something to start from, but missing / not having access to the X socket seems to cause it. But still, if LXDE works... I'm at awe really. Somehow it really seems to be a disconnection between dbus and X and has nothing to do with E. LXDE does also require dbus to my knowledge. So, there could be a 1000 other reasons, but I hope some of the questions above solve it.

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Wed Apr 29, 2020 6:35 pm Post subject:

So, these were ok already:

Code:

CONFIG_TMPFS_XATTR=y

Code:

[ebuild R ] x11-base/xorg-server-1.20.7:0/1.20.7::gentoo USE="elogind ipv6 libglvnd suid udev xorg -debug -dmx -doc -kdrive -libressl -minimal (-selinux) -static-libs -systemd -unwind -wayland -xcsecurity -xephyr -xnest -xvfb" 6 156 KiB

I added this tmpfs line to /etc/fstab (was not present in mine)

Code:

tmpfs /tmp tmpfs noatime,nodiratime 0 0

And here we get:

Code:

# mount
proc on /proc type proc (rw,nosuid,nodev,noexec,relatime)
udev on /dev type devtmpfs (rw,nosuid,relatime,size=10240k,nr_inodes=2020089,mode=755)
devpts on /dev/pts type devpts (rw,relatime,gid=5,mode=620,ptmxmode=000)
sysfs on /sys type sysfs (rw,nosuid,nodev,noexec,relatime)
/dev/nvme0n1p4 on / type jfs (rw,noatime)
tmpfs on /run type tmpfs (rw,nodev,relatime,size=1616736k,mode=755)
debugfs on /sys/kernel/debug type debugfs (rw,nosuid,nodev,noexec,relatime)
configfs on /sys/kernel/config type configfs (rw,nosuid,nodev,noexec,relatime)
fusectl on /sys/fs/fuse/connections type fusectl (rw,nosuid,nodev,noexec,relatime)
efivarfs on /sys/firmware/efi/efivars type efivarfs (rw,nosuid,nodev,noexec,relatime)
cgroup_root on /sys/fs/cgroup type tmpfs (rw,nosuid,nodev,noexec,relatime,size=10240k,mode=755)
openrc on /sys/fs/cgroup/openrc type cgroup (rw,nosuid,nodev,noexec,relatime,release_agent=/lib/rc/sh/cgroup-release-agent.sh,name=openrc)
none on /sys/fs/cgroup/unified type cgroup2 (rw,nosuid,nodev,noexec,relatime,nsdelegate)
shm on /dev/shm type tmpfs (rw,nosuid,nodev,noexec,relatime)
/dev/nvme0n1p5 on /home/user type jfs (rw,nosuid,nodev,noexec,noatime)
tmpfs on /tmp type tmpfs (rw,noatime,nodiratime)
binfmt_misc on /proc/sys/fs/binfmt_misc type binfmt_misc (rw,nosuid,nodev,noexec,relatime)
tmpfs on /run/user/0 type tmpfs (rw,nosuid,nodev,relatime,size=1616732k,mode=700)
tmpfs on /run/user/1000 type tmpfs (rw,nosuid,nodev,relatime,size=1616732k,mode=700,uid=1000,gid=1000)

Just showing you in case I did something wrong. :roll:

Now, that is the /tmp check after starting X and shutting it down:

Code:

# ls -al /tmp
drwxrwxrwt 5 root root 120 29 avril 20:01 .
drwxr-xr-x 21 root root 4096 24 avril 10:06 ..
drwxrwxrwt 2 root root 40 29 avril 19:59 .ICE-unix
drwx------ 2 atreyu atreyu 60 29 avril 20:01 ssh-RaJ0ajHQb3wo
srwxr-xr-x 1 root root 0 29 avril 19:59 wpa_ctrl_3331-1
drwxrwxrwt 2 root root 40 29 avril 20:01 .X11-unix

As you can see, on my system, /tmp/X0-lock does not exist... is that relevant in any way?

Juippisi wrote:

It can also be due to you already having one instance of X open somewhere. You wouldn't happen to have some autolanchers? Or do you use a login manager and then Ctrl+Alt+F1 to go into TTY without turning off the login manager? Check "ps aux" in case.

I don't think so, normally I properly closed X every time and ps aux looks clean.

Juippisi wrote:

So check that your user is in every correct group.

Here's my /etc/group , feel free to tell me if I did anything insane/stupid here (I might have...) ^^

Code:

root::0:root
bin::1:root,bin,daemon
daemon::2:root,bin,daemon
sys::3:root,bin,adm
adm::4:root,adm,daemon
tty::5:myuser
disk::6:root,adm,myuser
lp::7:lp
mem::8:
kmem::9:
wheel::10:root,myuser
floppy::11:root
news::13:news
uucp::14:uucp
console::17:
audio::18:myuser
cdrom::19:myuser
dialout::20:
tape::26:root
video::27:root,myuser
cdrw::80:myuser
usb::85:pcscd,myuser
input::97:myuser
users::100:myuser
portage::250:portage,myuser
utmp:x:406:
nogroup::65533:
nobody::65534:
man:x:15:myuser
sshd:x:22:myuser
kvm:x:78:
render:x:28:
messagebus:x:101:
lpadmin:x:106:
polkitd:x:999:
cron:x:16:
crontab:x:998:
locate:x:997:myuser
plugdev:x:272:myuser
dhcp:x:996:
geoclue:x:995:
openct:x:46:pcscd
pcscd:x:47:myuser
myuser:!:1000:
wireshark:x:994:myuser

Now if you don't see anything in these, let's just forget it. We'll be dead before we'd had investigated the 1000 reasons :lol:

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Thu Apr 30, 2020 5:46 am Post subject:

Nothing out of the ordinary there.

Quote:

As you can see, on my system, /tmp/X0-lock does not exist... is that relevant in any way?

When you don't have an active X session, it doesn't exist indeed.

Few things before I'm willing to give up:
I just noticed you have 'fbcon' USE flag enabled in efl, it may try to launch a framebuffer session which then is blocked by missing kernel options. Try disabling 'fbcon' and preferably 'debug' as well.

Can you launch X as root? First from basic log in, then with 'dbus-launch' tricks?

If you could test trying to log in through an up-to-date login manager with your user, like lightdm, gdm, sddm (not slim as its elogind support is reported to be 50-50).

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Thu Apr 30, 2020 11:07 am Post subject:

Hey! Re-emerged efl with -fbcon -debug > no changes X server with LXDE starts just fine as root but enlightenment won't with the same error messages, with dbus-launch it doesn't work either... Just tried lightdm wich works just fine with LXDE but when it comes to enlightenment, just falls back to login screen We're cooked

Juippisi
Developer

Joined: 30 Sep 2005
Posts: 727
Location: /home

Posted: Thu Apr 30, 2020 1:06 pm Post subject:

I'm out of ideas :( Sorry.

swanson
Tux's lil' helper

Joined: 04 Jun 2004
Posts: 147
Location: Edinburgh, Scotland

Posted: Sat May 02, 2020 5:27 pm Post subject:

Have you tried resetting the configuration settings? I've known Enlightenment to occasionally be a bit flaky with it's own settings, though not specifically for your errors.

Code:

mv ~/.e ~/.e-backup

_________________
Alan.

Adrien
Advocate

Joined: 13 Jul 2004
Posts: 2308
Location: Bretagne

Posted: Sun May 03, 2020 4:36 pm Post subject:

swanson wrote:

Have you tried resetting the configuration settings? I've known Enlightenment to occasionally be a bit flaky with it's own settings, though not specifically for your errors.

Code:

mv ~/.e ~/.e-backup

thank you for tip, didn't work, though...

Display posts from previous:

	Gentoo Forums Forum Index Desktop Environments	All times are GMT
Page 1 of 1

Jump to:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

Copyright 2001-2024 Gentoo Foundation, Inc. Designed by Kyle Manna © 2003; Style derived from original subSilver theme. | Hosting by Gossamer Threads Inc. © | Powered by phpBB 2.0.23-gentoo-p11 © 2001, 2002 phpBB Group
Privacy Policy