Nxnode.bin segfaults on logout

[Joseph]

Frequently nxnode.bin will segfault when a user logs out of their Gnome desktop and kills the connection before they get presented the gdm login screen again. This puts NoMachine in a state where trying to make a new connection fails and claims there is no display detected (we do the edid trick to make a local display). systemctl restart nxserver.service gets NoMachine to see the local display again, but we are having to do this for several users every day, though not always the same users.

 

Remote:

• NoMachine-enterprise-desktop.x86_64     7.7.4-1
• Linux host.example.org 3.10.0-1062.18.1.el7.x86_64 #1 SMP Tue Mar 17 23:49:17 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
• CentOS Linux release 7.7.1908 (Core)
• gdm.x86_64                              1:3.28.2-16.el7
• gnome-desktop3.x86_64                   3.28.2-2.el7
• gnome-shell.x86_64                      3.28.3-14.el7_7

 

Local:

• NoMachine-enterprise-client  7.7.4
• Windows 10 Enterprise

 

I won’t be able to attach/email the 10 ~500MB core files it creates.

-rw-r—–. 1 gdm   gdm    503M Oct 27 16:23 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid14361-time1635366230

-rw-r—–. 1 gdm   gdm    501M Oct 27 16:23 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid14161-time1635366206

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:23 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid13893-time1635366180

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:22 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid13640-time1635366155

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:22 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid13341-time1635366130

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:21 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid12759-time1635366105

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:21 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid12562-time1635366080

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:20 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid12372-time1635366055

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:20 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid12147-time1635366030

-rw-r—–. 1 gdm   gdm    493M Oct 27 16:20 core-host.example.org-nxnode.bin-sig11-uid42-gid42-pid11885-time1635366005

[Joseph]

I see log files didn’t attach, but a bunch of this:

Oct 27 16:20:05 host.example.org kernel: nxnode.bin[12076]: segfault at 11f8 ip 00007f367693201e sp 00007f360d5c3a38 error 4

Oct 27 16:20:30 host.example.org kernel: nxnode.bin[12302]: segfault at 11f8 ip 00007f726851901e sp 00007f71ff7fda38 error 4 in libnx.so[7f726843b000+13a000]

Oct 27 16:20:55 host.example.org kernel: nxnode.bin[12490]: segfault at 11f8 ip 00007fb15559a01e sp 00007fb0ecb96a38 error 4 in libnx.so[7fb1554bc000+13a000]

Oct 27 16:21:20 host.example.org kernel: nxnode.bin[12689]: segfault at 11f8 ip 00007f8760ec001e sp 00007f86f7ffea38 error 4 in libnx.so[7f8760de2000+13a000]

Oct 27 16:21:45 host.example.org kernel: nxnode.bin[13270]: segfault at 11f8 ip 00007f36879b201e sp 00007f361f393a38 error 4

Oct 27 16:22:10 host.example.org kernel: nxnode.bin[13569]: segfault at 11f8 ip 00007f2dcb0b201e sp 00007f2d61dc4a38 error 4

Oct 27 16:22:35 host.example.org kernel: nxnode.bin[13823]: segfault at 11f8 ip 00007f1cd113b01e sp 00007f1c63ffea38 error 4 in libnx.so[7f1cd105d000+13a000]

Oct 27 16:23:00 host.example.org kernel: nxnode.bin[14090]: segfault at 11f8 ip 00007f705f76f01e sp 00007f6ff6b92a38 error 4 in libnx.so[7f705f691000+13a000]

Oct 27 16:23:26 host.example.org kernel: nxnode.bin[14290]: segfault at 11f8 ip 00007f6ecb5b101e sp 00007f6e62b92a38 error 4 in libnx.so[7f6ecb4d3000+13a000]

Oct 27 16:23:50 host.example.org kernel: nxnode.bin[14543]: segfault at 11f8 ip 00007f48b1d4e01e sp 00007f4848b96a38 error 4 in libnx.so[7f48b1c70000+13a000]

[Carin]

Hi,

We could not reproduce the problem in our labs. Would it be possible to upgrade your system to 7.9 and see if the problem persists?

[Joseph]

We are evaluating 7.9 and seeing if we can update, in the mean time we’ve made gdm restart nxserver on session end, which seems to be lessening the number of issues people are having. However, it doesn’t make for the cleanest experience, just having NoMachine suddenly close once someone logs out instead of being brought back to the login screen.

Is there any other data I can provide to help look into this, in case we can’t update to 7.9 any time soon?

[Britgirl]

We aren’t able to reproduce this on our systems with your same set-up. Can you send us the full backtrace of the crash?

Instructions are here:

Using GDB to debug a core file on Linux
https://knowledgebase.nomachine.com/AR09L00810

Send it to forum[at]nomachine[dot]com. In the subject of your email please use the topic’s title. Thanks!

[Joseph]

Backtraces have been sent.

Thank you.

[Britgirl]

Joseph, we did not receive any attachments from you. Could you send them again?

[Joseph]

Resent. Also sent an email with a zip in case the mail server doesn’t like txt attachments for some reason.

[Britgirl]

We got them, thank you!

[Britgirl]

The backtraces don’t show enough information to understand why a crash is occurring where it is. Can we send you a debug rpm packages (64bit)? We’re not able to reproduce similar behaviour, so you’ll need to install the debug package and extract the logs. We’ll give you instructions when we send you the link.

[Joseph]

Yeah, perfectly happy to try a debug package.

[Britgirl]

I sent you a link via private email, check your inbox 😉

[Joseph]

I’m not seeing the email. Mind sending it again?

[Britgirl]

Check again 🙂 (I just resent it)

[Joseph]

Got it. Thanks!

[Author]

Posts