Peter Skinner wrote:
Here is my rc file, and logging _is_ turned off:
mt=2
egtb
hash=192M
hashp=128M
cache=32M
log off
whisper 2
exit
There will still be a game.xxx produced, but no log.xxx file.
Peter
Do you still need to use smpnice=1 in conjunction with ponder=off for ponder off use , and what is the "whisper 2" command ?
.
Since I have never really used the smpnice=1 feature, possibly Robert can answer that for you. I always use ICS with two computers (both dual core) to play matches, or I just match an opponent on FICS/ICC for a few hundred games. We have Robert's cluster that we use for matches, and since that produces results much faster than I can attain them locally, I really only test the opening books that I/we make online.
The whisper 2 command is purely for ICS play where Crafty will "whisper" it's eval to observers. You can also use the kibitz command, but that can become an annoyance to human opponents.
Peter
smpnice = 1 is the default and has been for a while.. The previous problem was using processes and not threads on linux meant that when a search ended. each process (except the original) would exit(0); But we used threads on windows and the exit() in a thread terminates the whole mess, not just that thread (which should simply "return" to exit.
Now we use threads for linux and windows, and I have run tons of tests with max cpus set to 8, and when a search ends, the 7 extra threads terminate just fine and then re-start when the search continues. I have not seen any problems at all, although I have only tested under linux. With ponder=off, it will burn up 8 cpus when thinking, and then after making a move, it uses no CPU cycles until the opponent moves... And it behaves that way by default... For very fast games I would not use that mode myself, but then I don't use ponder=off for playing games anyway if I am using a parallel search.
Werner wrote:But a short test here showed me: with ponder off the smp-engine (2CPU) uses no cpu time after the move now (that´s good).
best
So smpnice=1 is no longer necessary. Very good. I'll try it myself later
It wasn't "smpnice=1" that was needed, it was smpnice=0. Otherwise under windows it would simply exit after the first search. Now you can run it either way and it should not cause any further problems....
glorfindel wrote:It is good news that crafty is getting stronger, and fortunately with 22.2 I can take back moves in analysis mode with 4 threads running without a crash and with smpnice=1.
My problem is now in positions from the endgame TBs:
1. In analysis mode (in XBoard) I can see no output about the best move or the score. I only see depth 0 score +0.00.
2. If I play out a position from the endgame TBs against crafty and I take back a move, crafty crashes immediately like this:
79230 >first : remove
xboard: Error: first chess program (/usr/games/bin/crafty) exited unexpectedly
GameEnds(0, (null), 2)
I managed to crash Crafty 2 times out of 2 in a K+2P v K+P ending, 2 times out of 2 in a K+B+P v K+N ending, but in a K+R+B v K+R ending it refused to crash and worked normally.
I used the Makefile provided, target linux-amd64, only with NUMA disabled.
My advice is to download the intel icc compiler and use that. There is a target "make linux-icc" that should work and will do a profile-guided optimization run that will be significantly faster than gcc. ICC is free for linux from Intel's web site.
Werner wrote:I do not remember which problems you mean with wb2uci adapter.
I mean using Crafty as UCI, with wb2uci adapter (not Winboard)
With the previous Crafty versions (and ShredderClassic GUI), I had this, with my dual-core:
"When I do infinite analysis, I obtain 100% cpu use on the Crafty process (which is normal, since I use both cores). But when I stop the analysis, I still have 50% cpu use on the Crafty process, which of course is not normal. If I then start again the infinite analysis, Crafty hangs, with 0% cpu use."
Does Crafty 22.2 now work properly as MP in infinite analysis?
bob wrote:My advice is to download the intel icc compiler and use that. There is a target "make linux-icc" that should work and will do a profile-guided optimization run that will be significantly faster than gcc. ICC is free for linux from Intel's web site.
Perhaps I will try that, although I would prefer to use gcc.
But I am not sure which one to download. Intel(R) Compiler Suite Professional Edition for Linux or Intel® C++ Compiler Professional Edition for Linux?
Thank you for your offer to have a look at it but it's as simple as that: taking back a move when analyzing causes exception by Crafty but only in endgame positions, so the suspicion it depended on handling the tablebases seems rather probable to me. Is the .rc- line
egtb=C:\xxx
correct?
Thanks again and best regards!
Peter.
Werner wrote:I do not remember which problems you mean with wb2uci adapter.
I mean using Crafty as UCI, with wb2uci adapter (not Winboard)
With the previous Crafty versions (and ShredderClassic GUI), I had this, with my dual-core:
"When I do infinite analysis, I obtain 100% cpu use on the Crafty process (which is normal, since I use both cores). But when I stop the analysis, I still have 50% cpu use on the Crafty process, which of course is not normal. If I then start again the infinite analysis, Crafty hangs, with 0% cpu use."
Does Crafty 22.2 now work properly as MP in infinite analysis?
Yes. that is what smpnice=1 does, it just did not work properly under windows in previous versions, where it should work fine here.