yade-dev team mailing list archive

[Janek Kozicki <janek_listy@xxxxx>]

Václav Šmilauer said:     (by the date of Thu, 29 Apr 2010 21:18:01 +0200)

> That is quite possible. I also had best results with 5 cores on
> Opterons, but Xeons were getting faster up to 8 cores (about 5.5x).
> Depends on memory bus and other stuff I don't understand that much.
> 
> The jumping speed could be also explained bu the GUI thread blocked by
> computation, so it updates only very rarely and shown nonsense values.
> 
> Note that there are still some global locks (such as when creating
> interaction) and non-parallel parts like the collider.
> 
> See https://www.yade-dem.org/wiki/Triaxial_Test_Parallel and
> https://www.yade-dem.org/wiki/Performance_Tuning, in most cases 3-4
> cores give the best performance.

I see, how about using this one, to avoid global locking:

http://www.chaoticmind.net/~hcb/projects/boost.atomic/

It seems however that currently only fundamental types are supported.
I asked the author about that. If you remove 'Something' from
attached test file, it will work nicely.

-- 
Janek Kozicki                               http://janek.kozicki.pl/  |

#include<iostream>
#include<boost/thread.hpp>
#include<boost/bind.hpp>
#include<boost/atomic.hpp>
#include<vector>

class Something
{
	private:
		int val;
	public:
		Something(){};
		Something(int i):val(i){};
		void set_val(int i)	{val=i;};
		int get_val()		{return val;};
};

class Numbers
{
	public:
		boost::atomic<int> sum,t1,t2;
		Numbers():sum(0),t1(0),t2(0){};
};

class Thread1
{
	public:
		void run(Numbers& c)
		{
			int i(100000);
			while(i-->0)
				++c.sum,c.t1++;
		}
};

class Thread2
{
	public:
		void run(Numbers& c)
		{
			int i(100000);
			while(i-->0)
				c.sum++,c.t2++;
		}
};

class Display
{
	public:
		void run(Numbers& n)
		{
			int i(103);
			while(i-->0)
			{
				int sum(n.sum);
				int t1(n.t1);
				int t2(n.t2);
				int SUM(t1+t2);
				std::cout << n.sum << " " << n.t1 << " " << n.t2 << " .. " << SUM << " " << (int)(SUM==sum) << "\n";
			}
		}
};

int main(int argc, char** argv)
{
	Numbers num;
	std::vector<boost::atomic<Something> > v;
	v.resize(100);
	Thread1 t1;
	Thread2 t2;
	Display disp;

	boost::thread_group run;
	run.create_thread(boost::bind(&Thread1::run,boost::ref(t1),  boost::ref(num)));
	run.create_thread(boost::bind(&Thread2::run,boost::ref(t2),  boost::ref(num)));
	run.create_thread(boost::bind(&Display::run,boost::ref(disp),boost::ref(num)));
	run.join_all();
}

Attachment: makefile
Description: Binary data