Hey folks!

I'm in a bit of a bind with an experiment, and I was hoping to be pointed in a direction as to how to fix this. Here's the situation: LOP v1.44 and smaugfuss v1.9.2 both hang and crash, but only on one machine, which happens to be a Pi Model 3 running Ubuntu MATE (my other server, running Ubuntu Server, runs these just fine). Both MUDs compiled pretty clean and didn't seem to have any glaring issues.

The crash occurs in both codebases during character creation. I can set name, password, class and race- then, when I attempt to select the color support (RIP/ANSI/No Color), any option I choose hangs the MUD.

I have tried multiple classes/races/names/etc., just to make sure I'm not making some kind of funky combo that's throwing the MUD for a loop (unlikely anyway, but still worthy of ruling out).

I have attempted to trace this back with GDB, and here is the output that I encounter:

Core was generated by `â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’â–’'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 0x00037096 in ?? ()
(gdb) bt
#0 0x00037096 in ?? ()
#1 0x76cfc2b0 in ?? ()
Backtrace stopped: previous frame identical to this frame (corrupt stack?)

Then, when I try to inspect the frame:

(gdb) frame 1
#1 0x76cfc2b0 in ?? ()

From my limited experience with this, what I understand of this output is that even GDB can't find the location of the issue. Either that or I lack the know-how to trace it back with the info given.

Can anyone point me in the right direction, or has anyone encountered this before? If other info is needed, please let me know.

Thank you so much!
-Zavor

can you check and see if it crashes based on each of the choices you can make, see if it only crashes on RIP or ANSI or no color might help find it a bit faster. if its on all three then might be just other stuff that happens after that part.

Hey Remcon, thanks for your input.
-BOTH games crash no matter what option I choose: RIP, ANSI, or None.

One thing I'm noticing in the logfiles for smaugfuss:
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/vampire.gui.vault
Fri May 26 22:18:08 2017 :: druid.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/druid.gui.vault
Fri May 26 22:18:08 2017 :: warrior.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/warrior.gui.vault
Fri May 26 22:18:08 2017 :: augurer.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/augurer.gui.vault
Fri May 26 22:18:08 2017 :: thief.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/thief.gui.vault
Fri May 26 22:18:08 2017 :: cleric.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/cleric.gui.vault
Fri May 26 22:18:08 2017 :: mage.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/mage.gui.vault
Fri May 26 22:18:08 2017 :: ranger.gui
Fri May 26 22:18:08 2017 :: Cannot open vault - ../clans/ranger.gui.vault

I'm also noticing that LOP in its log throws this:
Fri May 26 10:21:09PM 2017 :: [*****] BUG: load_pers_table: Can't open system/personal.dat

These may be related, as well; it's still odd that the exact same codebases work just fine on the normal server. I even downloaded and compiled them separately to make sure there was an even comparison.

Thanks!
-Zavor

those messages are actually normal so no worries there.

ok ill see if i can't figure out what possible issue it might be hitting that is crashing, hard to do without a similar server for checking lol but maybe ill get lucky and find something.

what if you put in another thing that isn't an option to see if it gives the actual message about invalid selection. if that goes ok then its just when you get past that point.

It does give me the invalid selection error. I can even put in an invalid selection multiple times. Doesn't hang until I make a valid selection.

in the comm.c nanny function case con_get_want_ansi add some output strings that directly output to the log file to see at which point it stops doing anything. it will help point out which area the issue may be found.