Keep getting message on console regarding LD_PRELOAD failure

[Ilgaz]

Hello,

I keep seeing this message on my openSUSE Tumbleweed console after recent updates:

ERROR: ld.so: object ‘/usr/NX/lib/libnxegl.so’ from LD_PRELOAD cannot be preloaded (cannot open shared object file): ignored.

This is what I see on Terminal (in case file is corrupt etc)

ilgaz@think:~> ls -l /usr/NX/lib/libnxegl.so
-rwxr-xr-x 1 root root 54504 Jul 12 11:18 /usr/NX/lib/libnxegl.so
ilgaz@think:~> file /usr/NX/lib/libnxegl.so
/usr/NX/lib/libnxegl.so: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, stripped
ilgaz@think:~> sha256sum /usr/NX/lib/libnxegl.so
f50c6c63608fb847fdbb66de0a74d24139f4a4122d2d84e8751f33cbee8be910  /usr/NX/lib/libnxegl.so

[Britgirl]

Hi, by “recent updates” do you mean the NoMachine update to 8.12 or 8.13? or are you referring to Tumbleweed updates?

[fajny.masz.tornister]

My version is: nomachine-8.14.2-1.x86_64 on RHEL 8.10 and i keep getting it on every application installed with snap.

ERROR: ld.so: object ‘/usr/NX/lib/libnxegl.so’ from LD_PRELOAD cannot be preloaded (cannot open shared object file): ignored.

ERROR: ld.so: object ‘/usr/NX/lib/libnxegl.so’ from LD_PRELOAD cannot be preloaded (cannot open shared object file): ignored.

ERROR: ld.so: object ‘/usr/NX/lib/libnxegl.so’ from LD_PRELOAD cannot be preloaded (cannot open shared object file): ignored.

 

[Britgirl]

Fajny, is it possible that you have installed a 64-bit package on a 32-bit system? This could explain why the library cannot be loaded.

[amprince01]

I’m receiving this error since upgrading to 8.14.2 on OpenSUSE Tumbleweed. I did the upgrade through the application when it prompted me to.

ERROR: ld.so: object ‘/usr/NX/lib/libnxegl.so’ from LD_PRELOAD cannot be preloaded (failed to map segment from shared object): ignored.
ERROR: ld.so: object ‘/usr/NX/lib/libnxegl.so’ from LD_PRELOAD cannot be preloaded (failed to map segment from shared object): ignored.

Free Version 8.14.2 – upgraded from 8.13 – the error comes up when I run commands in bash.

 

[asp100m]

Apparmor probably blocked the shared object.
You need to allow it by Yast -> AppArmor -> Scan Audit Logs.
In my case, it was lessopen.sh called by less.

[hawk]

Hello Ilgaz, fajny.masz.tornister and amprince01
May i ask you for more information?
– please share example snap app that is causing the problem?
– are you connecting to a headless environment?
– which graphical shell you are using?
Logs from /usr/NX/var/log/ directory and output from command env | grep -i nx may be very useful. Send those information to forum[at]nomachine[dot]com, please.

[hawk]

One more thing. Some useful information about things blocked by apparmor can be stored in journalctl. Please run sudo journalctl -g apparmor > apparmor.log command, collect apparmor.log and send it to forum[at]nomachine[dot]com along with other informations.

[Jumpingmanjim]

I get this error too on Ubuntu. Please fix it.

[hawk]

Hello Jumpingmanjim,
May i ask you to share some information i asked above?
– please share example snap app that is causing the problem.
– are you connecting to a headless environment?
– which graphical shell you are using?

Also i will be grateful if you send us nx_journal.log from command sudo journalctl > nx_journal.log. Send those information to forum[at]nomachine[dot]com, please.

[amprince01]

• I get this error when I do something like run sudo zypper ref or just about anything from the command line.
• I am connecting to a laptop that is open, the screen is up – it actually doesn’t matter, I get the error with the screen off as well.
• I’m using KDE6 Plasma v 6.1.5 running on X11 with OpenSUSE Tumbleweed-Slowroll 20241002.
• One thing that I find interesting and might be useful to you is that I’m running NoMachine on 2 other systems. All of them the same same OS and all using Plasma (one is Wayland and the other is X11). I do NOT have the problem on either of them. The difference is that the system having trouble is one that has had NoMachine upgraded a couple times.
• I suspect that you’re looking at a regression of some sort.
• I’m sending the journalctl and nx logs as requested and the command
• alisia@alisia-p14s:~> env | grep -i nx
• LD_PRELOAD=/usr/NX/lib/libnxegl.so

[amprince01]

Is there any update on this issue? It’s really annoying when this happens since I spend a lot of time in the command line on this computer.

[hawk]

We are still investigating this issue.

Are you using this machine as a headless environment, running it in NoMachine virtual
desktop session or some other remote method or you are connecting to a physical session?

If you are not connecting to the physical display with Wayland enabled you can try disabling EGL
capture by command sudo /etc/NX/nxserver --eglcapture no and getting rid of LD_PRELOAD env by command unset LD_PRELOAD

Thank you for your patience.

[fajny.masz.tornister]

LD_PRELOAD mechanism is simply prohibited on some laptops//computers by policy ( corpo machines or other junk hadrware with corpo policy ) so this is the case when on regular ( policy-protected ) user ( not root/Admin ) the LD_PRELOAD is … forbidden which triggers  som “kill” + “reporting” switches and get’s blocked.

Not the best idea / approach, with the LD_PRELOAD ( also got to witness the LD_PRELOAD bocking idea on AIX.. bt this is just a remark not a NoMachine observable ).

I stopped having the env LD_PRELOAD being set ( i use on laptop standard X ) so … problem gone.

[Ilgaz]

I bet Terminal guys knows a trick however, it becomes completely unbearable with ‘distrobox guest distribution shell’ as the message is being repeated tens of times.

[Author]

Posts