Often "segmentation fault" crashes since 4.4.17

[AsciiWolf]

Originally reported here, but it does not seem to be a Flatpak specific issue.

It looks like that there is a regression in the new xnec2c version 4.4.17 causing very frequent crashes:

[crit] [crit] [crit] segmentation fault, exiting
segmentation fault, exiting
segmentation fault, exiting
[crit] segmentation fault, exiting
[crit] segmentation fault, exiting
[err] early EOF?, child 0  length 39321  return 0
[err] Failed to read data from forked child
[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11write(): Broken pipe
write(): Broken pipe
read(): No such file or directory
[err] child 4  length 7  return 0
read(): No such file or directory
[err] child 3  length 7  return 0
read(): No such file or directory
[err] child 1  length 7  return 0

I will try to narrow down the problem, but unfortunately my limited free time does not allow me to work on this properly.

However, it is important to fix this since the version 4.4.17 is already in Debian Testing and will be in the upcoming Ubuntu 26.04 LTS.

[KJ7LNW]

Do you have the .nec file that triggers this particular problem?

[AsciiWolf]

I have created a new one with random configuration to test Flatpak builds. Feel free to try it, but I doubt it will be any useful.

[KJ7LNW]

I tried that file on the build on my system and did not hit a problem, so I am not able to reproduce it yet.

[AsciiWolf]

So, I have re-tested it using a new Flatpak build and I was not able to reproduce the original crash anymore either. But it still crashed two times (on a slightly different error) when randomly setting some values like gain or frequency:

$ flatpak run org.xnec2c.Xnec2c//test
[info] Detected 8 CPUs

(xnec2c:2): Gtk-WARNING **: 20:29:35.025: Can't set a parent on widget which has a parent
[warn] ATLAS, Threaded was not detected: Unable to set the mathlib index to 0
[warn] ATLAS, Threaded was not detected: Unable to set the mathlib index to 0

(xnec2c:2): Gtk-WARNING **: 20:29:37.749: Content added to the action area of a dialog using header bars

(xnec2c:2): Gtk-WARNING **: 20:29:37.749: Content added to the action area of a dialog using header bars

(xnec2c:2): Gtk-CRITICAL **: 20:39:24.840: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET (widget)' failed

(xnec2c:2): Gtk-CRITICAL **: 20:39:29.034: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET (widget)' failed
[crit] segmentation fault, exiting
[crit] [crit] segmentation fault, exiting
segmentation fault, exiting
[err] early EOF?, child 1  length 39321  return 0
[err] [crit] Failed to read data from forked child
segmentation fault, exiting
[crit] segmentation fault, exiting
[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11
(xnec2c:2): Gtk-CRITICAL **: 20:39:34.611: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET (widget)' failed

(xnec2c:2): Gtk-CRITICAL **: 20:39:35.774: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET (widget)' failed

(xnec2c:2): Gtk-CRITICAL **: 20:39:36.337: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET (widget)' failed
[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[err] fstep=-1 is out of bounds. calc_data.steps_total=11[crit] segmentation fault, exiting

First time it was directly in the NEC2 Input File Editor window, second time it was in the Radiation Patterns window.

[KJ7LNW]

But it still crashed two times (on a slightly different error) when randomly setting some values like gain or frequency

Can you provide a series of steps that reliably trigger the problem? If you can, can you run a backtrace from within GDB so I can see where it is crashing? If you have specifics and I can reproduce it, then I can do GDB on my side, but at the moment I do not have enough information to reproduce the issue.

[AsciiWolf]

The original problem does not seem to happen anymore and the other (similar) crashes happen randomly.

[KJ7LNW]

The original problem does not seem to happen anymore and the other (similar) crashes happen randomly.

I would love to fix them if you can get the program to run under GDB so that whenever it crashes, you can get me a back trace that would be very useful. Be sure to build it so that configure --disable-optimizations disables optimizations, so you can get clean backtraces.

[AsciiWolf]

It is already on my TODO list. :)

[KJ7LNW]

I just released 4.4.18, so please give it a try. I'm not sure if it will trigger the issues that you were having, but it would be worth testing, and the latest version should get out so that people can get it on FlatHub.

[KJ7LNW]

I'm preparing a big release with lots of changes including OpenGL. If you have a few minutes, can you give this master branch a test build and see if it behaves the way it should?

The high level new features are SY tuning/auto-optimization, OpenGL, Noise plots, and i18n language translations for every country that has an amateur radio program. The latest documentation is on the website. However, I have not yet officially released all of these changes in a new version yet: https://www.xnec2c.org/

[AsciiWolf]

I will try later this week (weekend), but I have been unfortunately very short on time lately. I was not ever able to properly test the 4.4.17 version (the same one that had issues in Flatpak version) that is in Ubuntu 26.04 LTS (and will be there for years to come). :-/

[KJ7LNW]

We're getting close to release. There's a bunch of new features. This will be version five.

Can you see if you still have any build/runtime problems?

[AsciiWolf]

I apologize, but I unfortunately had no time to test it yet. :(