[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gnugo-devel] bug
From: |
bump |
Subject: |
Re: [gnugo-devel] bug |
Date: |
Tue, 12 Oct 2004 19:09:10 -0700 |
> > No, unless there is some uninitialized memory access. I am now testing
> > for this with valgrind.
>
> Can anyone make sense out of the following?
> ==15012== Conditional jump or move depends on uninitialised value(s)
> ==15012== at 0x8057F92: gtp_decode_move (gtp.c:394)
> ==15012== by 0x804FD35: gtp_play (play_gtp.c:670)
> ==15012== by 0x8058624: gtp_main_loop (gtp.c:133)
> ==15012== by 0x804E708: play_gtp (play_gtp.c:358)
> ==15012==
> ==15012== Conditional jump or move depends on uninitialised value(s)
> ==15012== at 0x8057F98: gtp_decode_move (gtp.c:394)
> ==15012== by 0x804FD35: gtp_play (play_gtp.c:670)
> ==15012== by 0x8058624: gtp_main_loop (gtp.c:133)
> ==15012== by 0x804E708: play_gtp (play_gtp.c:358)
>
> It occurs at the GTP command "play white PASS" and does not make sense
> to me.
>
> This was the only error message by valgrind for Tristan's GTP stream.
I wasn't able to reproduce this.
I tried running this with valgrind (under gcc 3.3.3) and I got
no errors. I did get vastly different moves generated (so there
were many illegal moves reported).
I also got a reference to some memory that was not freed but still
reachable. Is this harmless?
valgrind --skin=memcheck -v --leak-check=yes gnugo --mode gtp <gtpstream
...
==4370==
==4370== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)
==4370== malloc/free: in use at exit: 11410420 bytes in 86 blocks.
==4370== malloc/free: 33990 allocs, 33904 frees, 98908407 bytes allocated.
==4370==
==4370== searching for pointers to 86 not-freed blocks.
==4370== checked 16934860 bytes.
==4370==
==4370== LEAK SUMMARY:
==4370== definitely lost: 0 bytes in 0 blocks.
==4370== possibly lost: 0 bytes in 0 blocks.
==4370== still reachable: 11410420 bytes in 86 blocks.
==4370== suppressed: 0 bytes in 0 blocks.
==4370== Reachable blocks (those to which a pointer was found) are not shown.
==4370== To see them, rerun with: --show-reachable=yes
--4370-- TT/TC: 0 tc sectors discarded.
--4370-- 25778 chainings, 0 unchainings.
--4370-- translate: new 21597 (407308 -> 5324185; ratio 130:10)
--4370-- discard 0 (0 -> 0; ratio 0:10).
--4370-- dispatch: 9143450000 jumps (bb entries), of which 558426180 (6%)
were unchained.
--4370-- 182871/552380 major/minor sched events. 297280 tt_fast
misses.
--4370-- reg-alloc: 3711 t-req-spill, 997259+30487 orig+spill uis, 111303
total-reg-r.
--4370-- sanity: 182872 cheap, 7315 expensive checks.
--4370-- ccalls: 108524 C calls, 50% saves+restores avoided (323312 bytes)
--4370-- 143154 args, avg 0.79 setup instrs each (58244 bytes)
--4370-- 0% clear the stack (325572 bytes)
--4370-- 34398 retvals, 29% of reg-reg movs avoided (19320 bytes)
Dan
- Re: [gnugo-devel] bug, (continued)
- Re: [gnugo-devel] bug, bump, 2004/10/08
- Re: [gnugo-devel] bug, Tristan Cazenave, 2004/10/08
- Re: [gnugo-devel] bug, bump, 2004/10/08
- Re: [gnugo-devel] bug, Tristan Cazenave, 2004/10/12
- Re: [gnugo-devel] bug, bump, 2004/10/12
- Re: [gnugo-devel] bug, bump, 2004/10/12
- Re: [gnugo-devel] bug, Tristan Cazenave, 2004/10/12
- Re: [gnugo-devel] bug, Arend Bayer, 2004/10/12
- Re: [gnugo-devel] bug, Arend Bayer, 2004/10/12
- Re: [gnugo-devel] bug, Trevor Morris, 2004/10/12
- Re: [gnugo-devel] bug,
bump <=
- Re: [gnugo-devel] bug, bump, 2004/10/12
- Re: [gnugo-devel] bug, Gunnar Farnebäck, 2004/10/13
- Re: [gnugo-devel] bug, Arend Bayer, 2004/10/09
- Re: [gnugo-devel] gnugo 3.6-pre2 (seed 1914539883), Paul Pogonyshev, 2004/10/07