Re: [Bug-gnubg] GNU Backgammon 0.16 stable

[Jim Segrave]

On Thu 03 May 2007 (14:37 +0200), Christian Anthon wrote:
> Hi all,
> 
> since the new file choosing interface and the multi-thread code seems
> reasonably stable, I think  we ought to "feature freeze" 0.16 and
> start a new branch. Anything that needs to be done before we do that?
> Øystein's batch code could use a bit of a polish I guess.

Linux versions on SMP amd-64 sigfaults when started in GUI mode with an
sgf file on the command line. Below is a gdb trace of this failure mode

==============================

Current directory is /usr/home/jes/gnubg/
GNU gdb 6.6
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "x86_64-pc-linux-gnu"...
Using host libthread_db library "/lib/libthread_db.so.1".
(gdb) r ../bg/sillymonkey-8.sgf
Starting program: /usr/home/jes/gnubg/gnubg ../bg/sillymonkey-8.sgf
[Thread debugging using libthread_db enabled]
[New Thread 47863903738736 (LWP 19138)]
GNU Backgammon 0.16-devel Copyright 1999, 2000, 2001, 2002, 2003, 2004 by Gary 
Wong.
GNU Backgammon is free software, covered by the GNU General Public License
version 2, and you are welcome to change it and/or distribute copies of it
under certain conditions.  Type "show copying" to see the conditions.
There is absolutely no warranty for GNU Backgammon.  Type "show warranty" for
details.

Python supported.
Window system supported.
External players supported.
XML match equity files supported.
Long RNG seeds supported.
3d Boards supported.
External commands supported.
ESD sound system supported.
Multiple threads supported.
SSE supported and available.
[New Thread 1082460496 (LWP 19141)]
[New Thread 1091180880 (LWP 19142)]
[New Thread 1099901264 (LWP 19143)]
[Thread 1082460496 (LWP 19141) exited]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 47863903738736 (LWP 19138)]
0x000000000049a058 in RenderBoard (prd=0x44e7ee0, 
    puch=0x4d8c910 "ð\031\210.\210+", nStride=-324) at render.c:1608
(gdb) bt
#0  0x000000000049a058 in RenderBoard (prd=0x44e7ee0, 
    puch=0x4d8c910 "ð\031\210.\210+", nStride=-324) at render.c:1608
#1  0x000000000049a8e0 in RenderImages (prd=0x44e7ee0, pri=0x45cd7a8)
    at render.c:2800
#2  0x00000000004b5b3c in board_create_pixmaps (board=<value optimized out>, 
    bd=0x45cd4e0) at gtkboard.c:3211
#3  0x00000000004ba7e2 in board_set (board=<value optimized out>, 
    board_text=0x7fff80278fe7 "", resigned=<value optimized out>, cube_use=1)
    at gtkboard.c:2814
#4  0x00000000004baee1 in game_set (board=0x45c5b20, points=0x6abc40, roll=0, 
    name=0x6adaa8 "SillyMonkey", opp_name=0x6ad280 "jim_segrave", match=9, 
    score=9, opp_score=2, die0=0, die1=0, computer_turn=1, nchequers=15)
    at gtkboard.c:3139
#5  0x000000000043399e in ShowBoard () at gnubg.c:2923
#6  0x0000000000434bd4 in UpdateSettings () at gnubg.c:2435
#7  0x0000000000471dbf in CommandLoadMatch (
    sz=0x44effd0 "../bg/sillymonkey-8.sgf") at sgf.c:1581
#8  0x0000000000439084 in main (argc=2, argv=0x7fff8027b578) at gnubg.c:6840
(gdb) 


Match analysis still has a progress bar whose behaviour is beyond
strange. This happens in both gui and cli mode. Displays liek
Analysing match: move: 13/3055 when the correct values is something
like move: 130/305.


-- 
Jim Segrave           address@hidden