Instead of waiting to be allocated, actively search
for another split point to join when finishes its
search. Also modify split conditions.
This patch has been tested with 7 threads SMP and
passed both STC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 2885 W: 519 L: 410 D: 1956
And a reduced-LTC at 25+0.05
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 4401 W: 684 L: 566 D: 3151
Was then retested against regression in 3 thread case
at standard LTC of 60+0.05:
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 40809 W: 5446 L: 5406 D: 29957
bench: 8802105
Thanks to std::bitset we can easily increase
the limit of active threads above 64.
Thanks to Lucas Braesch for pointing at the
correct solution of using std::bitset.
No functional change.
Because we test for available slaves before
entering split(), we almost always allocate a
slave, only in the rare case of a race (less
then 2% of cases) this is not true, but to
special case this occurrence is not worth
the added complexity.
bench: 7451319
A great simplification that shows no regression
and it seems even a bit scalable.
Tested with fixed number of games:
Short TC
ELO: 0.60 +-2.1 (95%) LOS: 71.1%
Total: 39554 W: 7477 L: 7409 D: 24668
Long TC
ELO: 2.97 +-2.0 (95%) LOS: 99.8%
Total: 36424 W: 5894 L: 5583 D: 24947
bench: 8184352
And remove a complex (and broken) formula.
Indeed previous code was broken in case of TC with big
time increments where available_time() was too similar
to total time yielding to many time losses, so for instance:
go wtime 2600 winc 2600
info nodes 4432770 time 2601 <-- time forfeit!
maximum search time = 2530 ms
available_time = 2300 ms
For a reference and further details see:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/dCPAvQDcm2E
Speed tested with bench disabling timer alltogheter vs timer set at
max resolution, showed we have no speed regressions both in single
core and when using all physical cores.
No functional change.
This reverts commit 800410eef1 and instead increases
stack size.
I went through the old emails with Daylen that reported the
crash issue on Mac OS X and was fixed by 0049d3f337.
It was reported default stack size for a thread in Mac OS X is 8
megabytes while the patch that we are reverting allows to reduce
stack size at max of about 217KB, so the reason for the crash was
only marginal in MAX_MOVES value. On those emails Daylen also
hinted how to increase stack size for Mac OS X to 16MB.
So prefer to increase stack size to 16MB instad of re-inventing
the wheel and do our home grown stack as we did with the patch
that we are now reverting (it will remain anyhow in git history
for documentation purposes).
No functional change.
This greately reduces stack usage and is a
prerequisite for next patch.
Verified with 40K games both in single and SMP
case that there are no regressions.
No functional change.
Set threads number always to 1 at startup and let the
user explicitly to chose the number of threads.
Also preserve the useful behavior of automatically set
"Min Split Depth" according to the requested threads,
indeed this parameter is too technical for a casual user,
so, when left to zero, we set it on a sensible value.
No functional change
Introduce ThreadBase struct that is search
agnostic and just handles low level stuff,
and derive all the other specialized classes
form here.
In particular TimerThread does not hinerits
anymore all the search related stuff from Thread.
Also some renaming while there.
Suggested by Steven Edwards
No functional change.
At thread creation start_routine() is called
and from there the virtual function idle_loop()
because we do this inside Thread c'tor, where the
virtual mechanism is disabled, it could happen that
the base class idle_loop() is called instead.
The issue happens with TimerThread and MainThread
where, at launch, start_routine calls
Thread::idle_loop instead of the derived ones.
Normally this bug is hidden because c'tor finishes
before start_routine() is actually called in the
just created execution thread, but on some platforms
and in some cases this is not guaranteed and the
engine hangs.
Reported by Ted Wong on talkchess
No functional change.
Follow Don Dailey definition of cut/all node:
"If the previous node was a cut node, we consider this an ALL node.
The only exception is for PV nodes which are a special case of ALL nodes.
In the PVS framework, the first zero width window searched from a PV
node is by our definition a CUT node and if you have to do a re-search
then it is suddenly promoted to a PV nodes (as per PVS search) and only
then can the cut and all nodes swap positions. In other words, these
internal search failures can force the status of every node in the subtree
to swap if it propagates back to the last PV nodes."
http://talkchess.com/forum/viewtopic.php?topic_view=threads&p=519741&t=47577
With this definition we have an hit rate higher than 90% on:
if (!PvNode && depth > 4 * ONE_PLY)
dbg_hit_on_c(cutNode, (bestValue >= beta));
And an hit rate of just 28% on:
if (!PvNode && depth > 4 * ONE_PLY)
dbg_hit_on_c(!cutNode, (bestValue >= beta));
No functional change.
Search is started after setting a position and
issuing UCI 'go' command. Then if we stop the search
and call 'go' again without setting a new position it
is assumed that the previous setup is preserved, but
this is not the case because what happens is that
SetupStates is reset to NULL, leading to a crash as
soon as RootPos.is_draw() is called because st->previous
is now stale.
UCI protocol is not very clear about requiring that a
position is setup always before launching a search,
so here we easy the life of GUI developers assuming
that the current state is preserved after returning
from a 'stop' command.
Bug reported by Gregor Cramer.
No functional change.
Add MOVE_NONE at the tail, this allows to loop
across MoveList checking for *it != MOVE_NONE,
and because *it is used imediately after compiler
is able to reuse it.
With this small patch perft speed increased of 3%
And it is also a semplification !
No functional change.
This reverts commit 0d68b523a3.
After easy move semplification this machinery is not
needed anymore (because of we don't need to know if a
root move is a recapture)
No functional change.
After previous patch if split point master is
waiting for job and "Use Sleeping Threads" is
false (our condition for official releases) then
it will lock/unlock splitPoint mutex in a super
tight loop badly affecting performance.
Rewrite the code to lock only when we are about
to finish. Note that race condition on slavesMask
is anyhow fixed.
No functional change.
Save the current active position in each Thread
instead of keeping a centralized array in struct
SplitPoint.
This allow to skip a memset() call at each split.
No functional change.
The syntax splitPoints() should force the compiler to
value-initialize the array and because there is no
user defined c'tor it falls back on zero-initialization.
Unfortunatly this is broken in MSVC compilers, because
value initialization for non-POD types is not supported,
so left splitPoints un-initialized and add in split()
initialization of slavesPositions, that is the only
member not already set at split time.
This fixes an assert under MSVC when running with
more than one thread.
Spotted and reported by Jundery.
No functional change.
To return a pointer to the available
thread instead of a bool. This allows
to simplify the core loop in split().
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
This function "returns" two values: bestValue and bestMove
Instead of returning one and passing as pointer the other
be consistent and pass as pointers both.
No functional change.
Handling of History and Gains is almost the same, with
the exception of the update logic, so unify both
classes under a single Stats struct.
No functional change.
When a thread is allocated a bit is set in slavesMask.
This bit corresponds to the thread's index field that,
because it happens to be the position in the threads
array, eventually it is equal to the loop index 'i'.
But instead of relying on this 'coincidence', explicitly
use the 'idx' field so to clarify slavesMask usage.
Backported from c++11 branch.
No functional change.
Fix again TimerThread::idle_loop() to prevent a
theoretical race with 'exit' flag in ~Thread().
Indeed in Thread d'tor we raise 'exit' and then
call notify() that is lock protected, so we
have to check again for 'exit' before going to
sleep in idle_loop().
Also same change in Thread::idle_loop() where we
now check for 'exit' before to go to sleep.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Silly logic bug introduced in dda7de17e7
Timer thread, when msec = 0, instead of going
to sleep, calls check_time() in an endless loop.
Spotted and reported by snino64 due to abnormally
high CPU usage.
No functional change.
Subclass MainThread and TimerThread and declare
idle_loop() virtual. This allow us to cleanly
remove a good bunch of hacks, relying on C++
polymorphism to do the job.
No functional change.
Rename it is_finished and use it only in main
thread to signal search is finished. This allows
us to simplify the complex SMP logic.
Ultra tricky patch: deep test is required under
wide conditions like pondering on and option
"Use Sleeping Threads" set to false.
No functional change.
This reverts commit 869c924410
I misunderstood here. Actually it can happen that
thread is created but still not entered idle_loop
and at the same time start_searching() is called.
Becuase 'do_sleep' is set start_searching() will
set it to false and start the search, but when,
at last, the thread enters idle_loop(), resets
the flag and goes to sleep: not what we want.
Revert the hack waiting for a better solution
in the next patches.
No functional change.
