Yet Another Linux Crash Bug

-_-

Unfortunately, none of the backtraces given to me has been useful. My Linux computer is out of commission, so until somebody gives me a usable backtrace from gdb, I can’t really solve this issue.

Everytime I do a backtrace it ends at “??” so the debugger itself doesn’t know, which is weird. Might it be some issue with a lib being outdated perhaps? It wasn’t that long ago that I did a full system upgrade.

You should try to compile it with full debug options with no optimizations. That may assist in getting a full stack trace. Also, you could try moving up a frame (using the “up” command) and investigate the variables. Ultimately, it sucks.

How do I do that with gcc? The premake makefiles are messy. :stuck_out_tongue:

Use premake4 to generate the makefile. Go into the projects directory and open up Makefile. Go to the “debug” build section: ifeq ($(config),debug)

Under the “CFLAGS += $(CPPFLAGS) $(ARCH) -g” option, add -rdynamic.
That may help.

Also, it should build a debug build by default.

ya, it’s compiling debug-builds alright. I’ll add that to the makefile. :slight_smile:

I committed some bug fixes today. Can you see if this is still an issue?

When the server options are updated or the server starts, " is added to every empty line (bizarre in itself). But if enough of these stack up, it actually prevents the server from starting.

___Merged doublepost__________________

A single disconnection still crashed the server. Latest build

I don’t have any of those issues. Are you getting your gserver source code from the SVN trunk, or one of the release tags?

Cadavre hosts it, as far as I know he always gets the latest builds.