]>
git.sesse.net Git - stockfish/log
Marco Costalba [Fri, 31 Dec 2010 17:18:23 +0000 (18:18 +0100)]
Send correct searched nodes statistic
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 31 Dec 2010 13:32:25 +0000 (14:32 +0100)]
Remove dubious castle detector
It was introduced by patch
66d16592 of 22/3/2009 merging
from Glarurung iPhone.
Tord says:
That change is only found in the Glaurung iPhone app, and not
in the latest Glaurung UCI source code. I don't remember why
this was added (and the iPhone app, unlike the UCI engine,
was never version controlled), but it was almost certainly
because it was somehow needed in the communication between
the engine and the iPhone GUI, and that it was never meant to be
included in the UCI engine. My guess is that it has something to
do with castling moves being entered as e1-g1 in the GUI, but
represented as e1-h1 in the chess engine.
Removing it in Stockfish should be completely safe, and won't harm
the iPhone version. Initially the iPhone GUI called functions in the
chess engine for checking for legality of moves, writing the move
list in SAN format, and various other tasks, but this is no longer
the case in the current version.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 31 Dec 2010 12:19:05 +0000 (13:19 +0100)]
Implement "seldepth" UCI info
This is the "selective search depth in plies" and we set
equal to PV line length.
Tested that works under FritzGUI.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 15:27:16 +0000 (16:27 +0100)]
Retire a couple of unused debug functions
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 15:18:22 +0000 (16:18 +0100)]
Leave threads go to sleep when waiting for a ponderhit
No need to heat up CPU in this case ;-)
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 14:45:43 +0000 (15:45 +0100)]
Move printing of best move on think
It seems a more appropiate place (IMHO) and helps to clarify
that idle_loop() should return a move, not a score.
Fix also handling of stalemate positions (we were not
sending any score) and we don't need to wait on "ponderhit",
this is done when returning in think().
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 11:53:02 +0000 (12:53 +0100)]
Unify single and multi PV 'new best move' handling
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 10:55:25 +0000 (11:55 +0100)]
Move print_pv_info() under RootMove
And rename to pv_info_to_uci()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 10:46:53 +0000 (11:46 +0100)]
Better document value_to_uci()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 30 Dec 2010 09:38:02 +0000 (10:38 +0100)]
Standardize root_search() signature
Now pass alpha and beta by copy as in serach()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Wed, 29 Dec 2010 10:00:32 +0000 (11:00 +0100)]
Move extract_pv_from_tt() and insert_pv_in_tt() under RootMove
Functional change only due to additional do/undo move
but absolutly harmless.
Also handle re-insertion in tt of PV lines also for multi PV case.
Marco Costalba [Wed, 29 Dec 2010 07:47:14 +0000 (08:47 +0100)]
Use rml[0].pv[] instead of dedicated pv[] array
We have a small functionality change in case we have a
fail-high so that both rml[].pv and pv[] are updated, but if,
after researching, we have a fail-low then rml score is updated
again but pv[] remains the same and coming back from search we
used a PV line that has failed-low (after having failed-high).
With this patch we always use the 'correct' PV line, i.e. the
line with highest score at the end of the whole search.
Retire also redundant RootMove's 'move' member and directly
use pv[0] instead.
Marco Costalba [Tue, 28 Dec 2010 18:05:24 +0000 (19:05 +0100)]
Last small touches in RootMoveList
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 20 Dec 2010 15:42:07 +0000 (16:42 +0100)]
Retire LMR intermediate research
It does not seem to improve anything.
After 8344 games Mod - Orig:
1362 - 1321 - 5661 ELO +2
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 28 Dec 2010 12:29:53 +0000 (13:29 +0100)]
Use insertion_sort() in RootMoveList
Simplify code and get a bit of extra speed, about +0.5%
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 28 Dec 2010 10:45:57 +0000 (11:45 +0100)]
Redefine MoveStack comparison as the natural one
Define symbol '<' to mean 'minor of', as it should be. Its meaning
was reversed to be used with std::sort() that sorts in ascending order
while we want a descending order.
But now that we use our own sorting code we don't need this
trick anymore.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 28 Dec 2010 09:16:53 +0000 (10:16 +0100)]
Fix broken last patch series
When a reference breaks things !
Here we take a reference (that is a pointer) to an
entry in a vector that changes below us --> BOOM !
References are essential but should be considered with
care in C++ because could lead to nasty surprises.
Restored functionality.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 27 Dec 2010 12:29:52 +0000 (13:29 +0100)]
Speedup moves root list sorting
Instead of a default member by member copy use set_pv()
to copy the useful part of pv[] array and skip the remaining.
This greatly speeds up sorting of root move list !
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 27 Dec 2010 11:20:26 +0000 (12:20 +0100)]
Use a std::vector to store root moves
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 27 Dec 2010 09:52:04 +0000 (10:52 +0100)]
Rename RootMoveList members removing 'move'
It is redundant being a move list ;-)
Also better document the two scores used by root list.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 26 Dec 2010 23:13:10 +0000 (00:13 +0100)]
Retire DirectionTable[]
With this patch even word 'direction' is disappeared !
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 26 Dec 2010 15:27:58 +0000 (16:27 +0100)]
Retire direction.cpp
Move the code to bitboard.cpp
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 26 Dec 2010 11:41:30 +0000 (12:41 +0100)]
Retire enum Direction
Use SquareDelta instead
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 26 Dec 2010 10:58:48 +0000 (11:58 +0100)]
Retire SignedDirectionTable[] and RayBB[]
Function ray_bb() was used just in one endgame where can
be used squares_in_front_of() instead.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 26 Dec 2010 09:54:45 +0000 (10:54 +0100)]
Retire direction.h
Move all to square.h
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 26 Dec 2010 09:21:48 +0000 (10:21 +0100)]
Simplify enum SquareDelta definition
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 25 Dec 2010 17:03:43 +0000 (18:03 +0100)]
Unify MovePicker c'tor call
And use the standard one.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 25 Dec 2010 10:22:55 +0000 (11:22 +0100)]
Use generate_moves() in san.cpp
Instead of MovePicker and cleanup san.cpp
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 25 Dec 2010 07:59:47 +0000 (08:59 +0100)]
Ignore two braindamaged remarks under icc
Remark 1418: external function definition with no prior declaration
and
Remark 1419: external declaration in primary source file
Can be safely ignored because are pure idiocy.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 24 Dec 2010 23:11:53 +0000 (00:11 +0100)]
Renamed thread_should_stop() in cutoff_at_splitpoint()
It is more clear what happened.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 24 Dec 2010 16:07:00 +0000 (17:07 +0100)]
Simplify a condition in update of best move
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 24 Dec 2010 14:25:11 +0000 (15:25 +0100)]
Triviality in struct PieceLetters
And little touches in search() too.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 18 Dec 2010 09:27:24 +0000 (10:27 +0100)]
Better clarify how we use TT depth in qsearch
Namely we use only two types of depth in TT:
DEPTH_QS_CHECKS or DEPTH_QS_NO_CHECKS.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Wed, 15 Dec 2010 08:14:01 +0000 (09:14 +0100)]
Second cleanup wave on check_is_useless()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Joona Kiiski [Tue, 14 Dec 2010 13:49:06 +0000 (15:49 +0200)]
Let check_is_useless() follow SF coding style
No functional change
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Tord Romstad [Tue, 14 Dec 2010 11:07:37 +0000 (12:07 +0100)]
Fixed a bug in move_from_uci(): En passant captures were not handled
correctly.
Joona Kiiski [Sun, 12 Dec 2010 09:54:50 +0000 (11:54 +0200)]
Selective checks at qsearch
After 5821 games
Mod- Orig: 1014 - 869 - 3938 ELO +8 (+- 3.6) LOS 97%
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 13 Dec 2010 10:17:06 +0000 (11:17 +0100)]
Retire uci_main_loop()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 13 Dec 2010 08:48:12 +0000 (09:48 +0100)]
Small cleanup in uci.cpp
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 9 Dec 2010 11:10:02 +0000 (12:10 +0100)]
Retire square_from_string()
And rename move_from/to_string() in a more specific
move_from/to_uci() that is a simple coordinate notation.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 12 Dec 2010 11:46:10 +0000 (12:46 +0100)]
Move the last multi-threads globals to ThreadsManager
Also rename ThreadsManager memeber data to be lower case.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 12 Dec 2010 10:06:36 +0000 (11:06 +0100)]
Further reduce sleep lock contention
Use one sleep lock per thread insted of a single one shared.
Also renamed WaitLock in SleepLock.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 22:45:13 +0000 (23:45 +0100)]
Allow threads to sleep when available
By mean of an an UCI option it is possible let the available
threads to sleep, this should help with Hyper Threading although
is not the best solution when number of threads equals number
of available cores.
Option is disabled by default.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 4 Dec 2010 08:18:18 +0000 (09:18 +0100)]
Increase MAX_THREADS to 16
No speed regression and no functional change.
After 7826 games Mod- Orig:
1188 - 1230 - 5408 ELO -1 (+- 3.1)
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 5 Dec 2010 16:54:35 +0000 (17:54 +0100)]
Increase PV LMR to SF 1.8 levels
Non-PV LMR is left unchanged.
After 8819 games
Mod- Orig: 1442 - 1343 - 6034 ELO +3 (+- 2.9) LOS 86%
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 6 Dec 2010 09:26:01 +0000 (10:26 +0100)]
Various cleanups in Position's ancillary functions
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 4 Dec 2010 20:19:22 +0000 (21:19 +0100)]
Fix a crash due to a broken Book::open()
Bug introduced in
9dcc2aad98b9703
We can be asked to open a non-exsistent file,
in this case we should gracefully handle the
case and silently return instead of exiting.
Bug discovered and bisected down by Joona.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Joona Kiiski [Sat, 13 Nov 2010 16:42:12 +0000 (18:42 +0200)]
New try for unstoppable pawn evaluation
This time we try very hard to avoid false positives.
The obvious downside is that we also miss many true
winning positions.
After 10544 games on RC
Mod- Orig: 1744 - 1646 - 7154 ELO +3 (+- 2.7) LOS 83%
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 30 Oct 2010 14:28:33 +0000 (15:28 +0100)]
Allow razoring after a null move
After 8322 games on Russian Cluster
Mod- Orig: 1341 - 1281 - 5700 ELO +2 (+- 3) LOS 75%
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Joona Kiiski [Tue, 30 Nov 2010 12:46:41 +0000 (14:46 +0200)]
Prune all negative see moves at low depths
After 2036 games Mod- Orig:
381 - 278 - 1377 ELO +17 (+- 6.2) LOS 99%
One of the biggest increases ever !
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 28 Nov 2010 16:05:49 +0000 (17:05 +0100)]
Fix bestmove output in multi PV case
When MultiPV > 1, always take bestmove from the RootMoveList
(and don't bother with a ponder move). Without that the bestmove
is most probably incorrect.
Patch from Peter Petrov.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 22 Nov 2010 17:40:36 +0000 (18:40 +0100)]
Fix 'generation' type to uint8_t
When we store this value in TT we cut this to 9 bits,
so we need a smaller variable otherwise comparisons
like:
replace->generation() == generation
Are always false if generation is bigger then the maximum
TT storable value.
This fixes a very nasty and difficult to spot bug (2 weeks for
regression hunting).
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 19 Nov 2010 06:59:51 +0000 (07:59 +0100)]
Revert sleeping threads
Revert
141caf1d5b64 +
c59efc53c9
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 23 Nov 2010 06:00:30 +0000 (07:00 +0100)]
Revert "Allow split point master to sleep (take 2)"
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 23 Nov 2010 08:09:15 +0000 (09:09 +0100)]
We don't need a stringstream in buildKey()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 22 Nov 2010 13:55:56 +0000 (14:55 +0100)]
Retire Application class
It is a redundant boiler plate, just call initialization and
resource release directly from main()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 22 Nov 2010 10:50:58 +0000 (11:50 +0100)]
Retire piece.cpp
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 21 Nov 2010 22:28:17 +0000 (23:28 +0100)]
UCI options names should not be case sensitive
Correctly handle uci option names in a case insensitive way.
Alos fix some indentation while there.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 21 Nov 2010 13:23:16 +0000 (14:23 +0100)]
Retire "New Game" UCI option
Was introduced by
403db5a6e9f
on 1/12/2009 to correctly handle "loose on time"
LSN filtering functionality, but is now unused.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 21 Nov 2010 12:06:53 +0000 (13:06 +0100)]
Remove hardcoded 16 from benchmark default positions size
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 21 Nov 2010 11:43:16 +0000 (12:43 +0100)]
Micro-optimize pl_move_is_legal()
This L1/L2 optimization has an incredible +4.7% speedup
in perft test where this function is the most time consumer.
Verified a speed up also in normal bench, although smaller.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 20 Nov 2010 13:23:28 +0000 (14:23 +0100)]
Store "true" and "false" in bool options
UCI protocol uses "true" and "false" for check and button types,
so store that values instead of "1" and "0", this simplifies a
bit the code.
Also a bit strictier option's type checking in debug mode.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 20 Nov 2010 10:48:04 +0000 (11:48 +0100)]
Debug counters don't need to be global
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 19 Nov 2010 16:03:44 +0000 (17:03 +0100)]
Small touches to engine_name()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 16 Nov 2010 18:28:09 +0000 (19:28 +0100)]
Use occupied_squares() in book_key()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 16 Nov 2010 08:42:12 +0000 (09:42 +0100)]
Rearrange pawn moves generation
This patch greatly cleanups generation of pawn moves but
we change the order in which moves are generated so there is
a change in functionality, but not in perft.
The only real functionality change is that now when type == CHECK
we generate knight underpromotion captures only if give check and
not always as before.
Perft is 2% faster and fully verified.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Tue, 16 Nov 2010 07:51:07 +0000 (08:51 +0100)]
Introduce generate_promotions()
A bit ugly to guarantee no functional change.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 15 Nov 2010 15:27:05 +0000 (16:27 +0100)]
Simplify generate_pawn_captures()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 15 Nov 2010 12:49:13 +0000 (13:49 +0100)]
Cleanup Bioskey()
And rename in dataAvailable()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 8 Nov 2010 19:49:40 +0000 (20:49 +0100)]
Let bench to have full defaults arguments
Now stockfish bench' defaults to
stockfish bench 128 1 12 default depth
that is the most used line (at least by me)
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 8 Nov 2010 14:30:12 +0000 (15:30 +0100)]
Various cleanup in book.cpp
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 8 Nov 2010 10:08:01 +0000 (11:08 +0100)]
Let benchmark to default to depth 12
And also simplify a lot the code.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 8 Nov 2010 08:44:28 +0000 (09:44 +0100)]
Move PieceValue[] and SlidingArray[] where they belong
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 22:45:13 +0000 (23:45 +0100)]
Allow split point master to sleep (take 2)
Let to sleep even split point master, it will be waken
up by its slaves when they return from the search.
This time let it be enabled by an UCI option, so
people is free to test it on their Hyper Thread box.
Option is disabled by default.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 12:41:21 +0000 (13:41 +0100)]
Use namespace in position.cpp
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 12:22:13 +0000 (13:22 +0100)]
Retire uci.h and benchmark.h
Moved the single prototipes where are needed.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 12:12:48 +0000 (13:12 +0100)]
Object OpeningBook doen't need to be global
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 11:33:17 +0000 (12:33 +0100)]
Make rkiss seed deterministic
Search at fixed depth with one thread must be
reproducible so remove randomess from time().
Also better license description.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 11:11:01 +0000 (12:11 +0100)]
Let rkiss.h to follow SF coding style
Fix also Makefile after mersenne.cpp has been removed
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 7 Nov 2010 10:13:17 +0000 (11:13 +0100)]
Use Heinz's RKiss instead of marsenne
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 5 Nov 2010 12:59:48 +0000 (13:59 +0100)]
Fix an off by one bug in print_uci_options()
Last option was not printed.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Fri, 5 Nov 2010 12:53:13 +0000 (13:53 +0100)]
Retire bitbase.h
Moved the only prototipe where is needed.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 4 Nov 2010 07:03:27 +0000 (08:03 +0100)]
Rewrite options handling in an object oriented fashion
Big rewrite and about 100 lines removed.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Wed, 3 Nov 2010 08:32:14 +0000 (09:32 +0100)]
Retire push_button() and button_was_pressed()
Directly access the underlying bool option.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 1 Nov 2010 12:16:32 +0000 (13:16 +0100)]
Prefer int to uint8_t when possible
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 1 Nov 2010 11:17:54 +0000 (12:17 +0100)]
Add "mingw" compiler to Makefile
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Mon, 1 Nov 2010 10:44:46 +0000 (11:44 +0100)]
Fix MinGW warnings
I finally got SF to compile under MinGW (after adding pthread libraries)
and here are the fixed warnings.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 31 Oct 2010 10:37:43 +0000 (11:37 +0100)]
Allocate RootPosition on the stack
And pass it as an argument.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 31 Oct 2010 09:55:28 +0000 (10:55 +0100)]
Update nodes after a do_move()
And also store the node counter in Position and not in Thread.
This will allow to properly count nodes also in sub trees with
SMP active.
This requires a surprisingly high number of changes
in a lot of places to make it work properly.
No functional change but node count changed for obvious reasons.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 30 Oct 2010 17:44:34 +0000 (18:44 +0100)]
Fix some icc's "statement is unreachable" warnings
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 30 Oct 2010 17:18:48 +0000 (18:18 +0100)]
Move moveCount update near the SpNode case
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 30 Oct 2010 14:22:10 +0000 (15:22 +0100)]
Introduce and use TranspositionTable::refresh()
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 30 Oct 2010 13:22:57 +0000 (14:22 +0100)]
Use margins[] array in evaluate
It will be used by future patches.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 30 Oct 2010 11:31:42 +0000 (12:31 +0100)]
More readable search/qsearch dispatch
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 24 Oct 2010 08:39:33 +0000 (10:39 +0200)]
Fix last leak detected by Valgrind
This was subtle and google was my friend.
The leak was in _dl_allocate_tls called by pthread_create() and
is due to the fact that threads are created in joinable state so that
once terminated are not freed. To make the thread to release
its resources upon termination we should set them in detached state.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 24 Oct 2010 08:15:31 +0000 (10:15 +0200)]
Fix a memcpy() warning under Valgrind
Fix warning: "Source and destination overlap in memcpy"
This happens when we call multiple time do_move() with the
same state, for instance when we don't need to undo the move.
This is what valgrind docs say:
You don't want the two blocks to overlap because one of them could
get partially overwritten by the copying.
You might think that Memcheck is being overly pedantic reporting this
in the case where 'dst' is less than 'src'. For example, the obvious way
to implement memcpy() is by copying from the first byte to the last.
However, the optimisation guides of some architectures recommend copying
from the last byte down to the first. Also, some implementations of
memcpy() zero 'dst' before copying, because zeroing the destination's
cache line(s) can improve performance.
In addition, for many of these functions, the POSIX standards have wording
along the lines "If copying takes place between objects that overlap,
the behavior is undefined." Hence overlapping copies violate the standard.
The moral of the story is: if you want to write truly portable code, don't
make any assumptions about the language implementation.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sun, 24 Oct 2010 08:51:49 +0000 (09:51 +0100)]
Revert previous patch
It seems we have a speed regression under Linux, anyhow
commit and revert to leave some documentation in case we
want to try again in the future.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Thu, 21 Oct 2010 06:46:38 +0000 (07:46 +0100)]
Allow split point master to sleep
Let to sleep even split point master, it will be waken up
by its slaves when they return from the search.
With this patch we get maximum HT speedup
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 23 Oct 2010 06:52:17 +0000 (07:52 +0100)]
Re-add "Pass evalMargin through SearchStack as eval"
It has more sense to treat the two evaluation metrics
in the same way.
As a side effect now we use the correct eval margin when
pruning in a SplitPoint node.
No functional change in single thread.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Marco Costalba [Sat, 23 Oct 2010 06:38:48 +0000 (07:38 +0100)]
Finally retire sp_search()
Fix the movcount updating bug and let search() to completely
subsititute sp_search().
No functional change even with fakes split.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>