[Flightgear-devel] Re: core dump with latest cvs

2002-02-07 Thread Alex Romosan
Curtis L. Olson [EMAIL PROTECTED] writes: This is just plain weird. Remind me what opengl drivers you are i found the problem. the check in configure.in for FX is wrong. it should check for fxmesa.h not xmesa.h. with this patch i can run fgfs again: --- configure.in5 Feb 2002

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-07 Thread Curtis L. Olson
Ok, I'll update the configure script. Very glad you were able to track down the problem. Thanks for your persistance. Curt. Alex Romosan writes: Curtis L. Olson [EMAIL PROTECTED] writes: This is just plain weird. Remind me what opengl drivers you are i found the problem. the check

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-06 Thread Erik Hofman
Curtis L. Olson wrote: Erik Hofman [EMAIL PROTECTED] writes: Just to be sure, you didn't remove MetaKit and re-install it in between did you? Or maybe some other libraries that might have a fresh install inbetween? Erik, What are you seeing (that I'm not) that would lead you to believe

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-06 Thread Curtis L. Olson
Alex Romosan writes: Curtis L. Olson [EMAIL PROTECTED] writes: reports isn't a trust worthy indicator. I'm not seeing anything like this on my systems though ... Debian 3.0, gcc version 2.95.4, latest nvidia drivers, glibc 2.2.4-7 ... pretty much the same thing except i update my

[Flightgear-devel] Re: core dump with latest cvs

2002-02-05 Thread Alex Romosan
Curtis L. Olson [EMAIL PROTECTED] writes: Have you tried recompiling simgear and flightgear completely from scratch? i recompiled everything from scratch, same problem. i've trying to debug this with gdb for the past 2 hrs with no success. it looks to me like the problem is with the property

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-05 Thread Curtis L. Olson
Alex, It looks like there definitely is some sort of memory corruption going on with your system. It's strange that this isn't more widely reported if it is a problem with the code. globals-get_props() can't be non-zero at this point in the code. Are you running out of RAM or something?

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-05 Thread Erik Hofman
Alex Romosan wrote: Curtis L. Olson [EMAIL PROTECTED] writes: Have you tried recompiling simgear and flightgear completely from scratch? i recompiled everything from scratch, same problem. i've trying to debug this with gdb for the past 2 hrs with no success. it looks to me like the

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-05 Thread Cameron Moore
* [EMAIL PROTECTED] (Alex Romosan) [2002.02.05 15:15]: btw, i've noticed that my messages take for ever to make it to the list. any ideas? (I'm posting for a different address than the one i used to subscribe to the list). This is because we moderate non-member posts to the -devel list. Curt

[Flightgear-devel] Re: core dump with latest cvs

2002-02-05 Thread Alex Romosan
Curtis L. Olson [EMAIL PROTECTED] writes: What is the value of globals before and after this call? Are you saying when this is done, globals == 0? no. this is what happens. we call globals = new FGGlobals; (main.cxx:1433). if i print globals.props this is what i get: {_name = {static npos =

Re: [Flightgear-devel] Re: core dump with latest cvs

2002-02-05 Thread Curtis L. Olson
This is just plain weird. Remind me what opengl drivers you are running? What version of XFree86? Can you run any other opengl programs on your system? Are you building with a RedHat/Mandrake/et.al broken compiler? Which compiler are you using? This might be one of those things that just

[Flightgear-devel] Re: core dump with latest cvs

2002-02-03 Thread Alex Romosan
Curtis L. Olson [EMAIL PROTECTED] writes: This is weird. Your back trace isn't showing where exactly in the options.cxx:fgSetDefaults() routine you are crashing. But, the value= shows -110.6642444 so this is clearly dying on the first fgSetDouble():