Re: [Bug-gnubg] Compile & crash on Solaris

[Holger]

Jim Segrave wrote:
> On Sat 04 Jan 2003 (00:58 +0100), address@hidden wrote:
> > I wrote:
> > > And sadly, while playing I got a core dump. I'll attach the terminal
> > > and gdb output.
> > 
> > Got another one. I suppose they are related or the same. If, then a
> > second case might make it clearer.
> 
> OK - I have to admit I'm out of my depth here. 

I thought you geniuses knew always everything! :-)

> But anyway:
> 
> (gdb) where
> #0  0xfe69c534 in _libc_kill ()
> #1  0xfe635884 in abort ()
> #2  0x26300 in MaxTurns (id=-1) at eval.c:1861
> #3  0x25dfc in SanityCheck (anBoard=0xffbe7080, arOutput=0xffbe7060)
>     at eval.c:1676
> #4  0x27190 in EvaluatePositionFull (anBoard=0xffbe7080,
> arOutput=0xffbe7060,
>     pci=0xffbe7760, pec=0xffbe6cd4, nPlies=0, pc=CLASS_RACE) at
> eval.c:2377
> 
> MaxTurns should never get a position id of -1. I can't follow the
> match, as either both players are named Holger or I'm not seeing
> gnubg's moves. I'm wondering why gnubg has called MaxTurns() with a
> patently invalid position ID. I'd love to know what the position was
> when gnubg dumped core.

GNUbg obviously only outputs all commands, not the moves. And I didn't save
the match, unfortunately. Well, better for me because of my play ;-).
Also, I just let the analysis run (2ply large I think). When I looked the
next time I had a core file. :-/ So I don't know when it happened.

Next week I'll try a few more matches. So far it has crashed in both of
them. One while playing, the other during analysis.

Btw, I've solved my pnmtopng problem, too. So the last build ran without
problems.

Regards,

        Holger

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
NEU: Mit GMX ins Internet. Rund um die Uhr für 1 ct/ Min. surfen!