[gnugo-devel] EB25 considered harmful

[Daniel Bump]

I ran the regressions and got two crashes. These are
13x13 test 80 and trevora test 300. Both patterns are
in new pattern EB25.

Backtrace below for 13x13 test 80. Crash is in new pattern EB25
with the anchor at N3 in this position:

   A B C D E F G H J K L M N
13 . . . X . . . . . . . . . 13
12 . O X . X . . . . . . . . 12
11 . . O X X . . . X . . . . 11
10 . . O + . . + . . + X . . 10
 9 . . O X . . . . . . . . . 9
 8 . . . X . . X . . X . . . 8
 7 . . O O X X + . X + . O . 7
 6 . . . . O O O . . . X . . 6
 5 . . . . . . . . O O O O O 5
 4 . . O + . . + . . X X X O 4
 3 . . . . O . . O . . . . X 3
 2 . . . . . . . . . . X . . 2
 1 . . . . . . . . . . . . . 1
   A B C D E F G H J K L M N

The problem is that the pattern doesn't contain a * so when 
we get to line 58 of connections.c we get a crash. Other
B patterns in conn.db have a * location marked.

I had revised this pattern per Gunnar's comment. It
currently reads:

Pattern EB25
# tm New Pattern (3.1.20)

O!.  cutable keima on edge.
.!O
.!.
---

:8,B

Obc
daf
.!.
---

; !xplay_attack_either(a,b,c,a,c) ||
; !xplay_attack_either(b,a,d,b,d)

Yesterday I pointed out another problem with the pattern in
that the second constraint looks wrong---usually if x
plays b O can secure a connection at d instead of a
even if there is a stone one point to the left of d.

I propose we take the pattern out of the CVS for the
time being then revise it.

Dan

#0  0x400bfa01 in __kill () from /lib/i686/libc.so.6
#1  0x400bf7da in raise (sig=6) at ../sysdeps/posix/raise.c:27
#2  0x400c0f82 in abort () at ../sysdeps/generic/abort.c:88
#3  0x0805663d in do_trymove (pos=214, color=1, ignore_ko=0) at board.c:537
#4  0x080561e5 in trymove (pos=214, color=1, message=0x80d5eda "safe_move-A", 
    str=0, komaster=0, kom_pos=0) at board.c:392
#5  0x0808f8e5 in safe_move (move=214, color=1) at reading.c:5410
#6  0x0809e4b1 in cut_connect_callback (m=10, n=12, color=2, 
    pattern=0x80eb104, ll=7, data=0x0) at connections.c:58
#7  0x0806ce16 in do_matchpat (m=10, n=12, callback=0x809ea20 <cut_callback>, 
    color=2, pattern=0x80eb104, callback_data=0x0, goal=0x0) at matchpat.c:580
#8  0x0806cebf in matchpat_loop (callback=0x809ea20 <cut_callback>, color=2, 
    anchor=2, pdb=0x80ed10c, callback_data=0x0, goal=0x0) at matchpat.c:621
#9  0x0806d774 in matchpat (callback=0x809ea20 <cut_callback>, color=6, 
    pdb=0x80ed10c, callback_data=0x0, goal=0x0) at matchpat.c:1071
#10 0x0809eab7 in find_cuts () at connections.c:259
#11 0x08075e91 in make_domains (b_eye=0x84c8c80, w_eye=0x84d32a0, owl_call=0)
    at optics.c:214
#12 0x080600e6 in make_dragons (color=1, stop_before_owl=0) at dragon.c:112
#13 0x08064495 in examine_position (color=1, how_much=99) at genmove.c:135
#14 0x08064916 in do_genmove (move=0xbffff748, color=1, 
    pure_threat_value=0.400000006) at genmove.c:326
#15 0x08064788 in genmove (i=0xbffff774, j=0xbffff778, color=1)
    at genmove.c:260
#16 0x0805205f in load_and_analyze_sgf_file (head=0x8657bc0, 
    gameinfo=0xbffff870, untilstr=0xbffffcc5 "41", benchmark=0, to_move=1)
    at play_solo.c:163
#17 0x0804a386 in main (argc=9, argv=0xbffffb24) at main.c:824
#18 0x400ad507 in __libc_start_main (main=0x80491d0 <main>, argc=9, 
    ubp_av=0xbffffb24, init=0x8048d4c <_init>, fini=0x80c8160 <_fini>, 
    rtld_fini=0x4000dc14 <_dl_fini>, stack_end=0xbffffb1c)
    at ../sysdeps/generic/libc-start.c:129