Linuxcnc is shutting down while loading ngc file

Hello,
I wanted to change the syntax of writing ngc file Eg. for variables we use #<Cnst_var> but i want to change it to $[Const_var]. I made appropriate changes to interp_read.cc and interp_array.cc. But, when i try to load ngc file in gmoccapy, its shutting down and gives an error on terminal...

5774 Segmentation fault $EMCDISPLAY -ini "$INIFILE" $EMCDISPLAYARGS $EXTRA_ARGS

Thanks in advance

Regards
Kaushik

I wanted to change the syntax of writing ngc file Eg. for variables we use #<Cnst_var> but i want to change it to $[Const_var].

Afraid if you want to change the whole way named parameters are designated and used, you are pretty much on your own.
$[name] actually means something in bash, ie access the contents/value of variable, no idea if that is relevant, but seems a good reason not to use it.

If you have existing code which uses named params in this way, you could simply write a filter script to change it to a #<name> syntax

I made appropriate changes to interp_read.cc and interp_array.cc.

Without seeing what you did and checking there are no other files affected, cannot comment

5774 Segmentation fault $EMCDISPLAY -ini "$INIFILE" $EMCDISPLAYARGS $EXTRA_ARGS

This is quite a generic fault and not very useful of itself. dmesg should show which library or process threw the segfault which may assist more

regards

dmesg is giving me


[[16932.404524] hm2/hm2_5i24.0: registered
[16932.404525] hm2_5i24.0: initialized AnyIO board at 0000:04:00.0
[16932.433145] hm2/hm2_5i24.0: The hm2 pet_watchdog function is no longer needed.
[16932.433149] hm2/hm2_5i24.0: It will be removed before LinuxCNC 2.7.
[16932.433151] hm2/hm2_5i24.0: The hm2 write function now pets the watchdog.
[17040.789625] gmoccapy[13840]: segfault at 68 ip b75a689c sp bfee68c4 error 4 in libc-2.13.so[b7563000+15f000]
[17041.451940] hm2_5i24.0: dropping AnyIO board at 0000:04:00.0
[17041.451944] hm2/hm2_5i24.0: unregistered
[17041.451980] hm2_pci: driver unloaded
[17041.453153] hm2: unloading
[17043.535319] RTAI[math]: unloaded.
[17043.536517] SCHED releases registered named ALIEN PEDV$D
[17043.541118] RTAI[malloc]: unloaded.
[17043.637112] RTAI[sched]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[17043.638593] I-pipe: head domain RTAI unregistered.
[17043.638628] RTAI[hal]: unmounted.


Regards
Kaushik

Hi

Unfortunately libc is so widely used, the error is far from illustrative.

The best way to get a segfault is usually to pass a NULL pointer or iterate beyond the end of an array etc. but it could be anything.

You seem to be running this on a real machine, the fault arising during the initialisation of the hm2 drivers, which may or may not be coincidental.

To make it simple, what happens when you run the axis_mm sim or similar on the amended code?

If it still segfaults, zip the entire config plus the amended source files, that is the only way to test anything

regards