Sunday, November 27, 2016

Digest for comp.lang.c++@googlegroups.com - 19 updates in 10 topics

David CARLIER <devnexen@gmail.com>: Nov 27 12:29PM -0800

http://devnexen.blogspot.com/2016/11/better-stronger-faster-there-is-zapcc.html
"Alf P. Steinbach" <alf.p.steinbach+usenet@gmail.com>: Nov 27 10:36PM +0100

On 27.11.2016 21:29, David CARLIER wrote:
> http://devnexen.blogspot.com/2016/11/better-stronger-faster-there-is-zapcc.html
 
That looks like spam.
 
Even your last name looks like spam.
 
How about a few words about what zapcc is and what the test showed, and
why you think it's relevant to discuss here or relevant for us to know
about?
 
 
 
Cheers,
 
- Alf (and no, I didn't follow the link)
scott@slp53.sl.home (Scott Lurndal): Nov 27 06:06PM


>Don't forget if any process writes to the shared memory, all access to
>the shared memory must be protected by mutex semaphores. That slows
>things down a bit - but it's still a lot faster than any other IPC.
 
There are many ways to ensure correctness
of shared data (between threads or between processes) without using
traditional mutexes or semaphores. Atomic accesses, for example,
can easily be used to maintain shared counters (e.g. gcc __sync_fetch_and_add,
__sync_and_and_fetch, et alia) while compare-and-swap can be
used to maintain linked lists without locks. Spinlocks,
read-copy-update and lock-free algorithms are all common in such
uses.
 
As a note, if using mmap(2) with MAP_PUBLIC or shmat(2) between processes with
disparate address spaces, you must use special flavors of the pthread
mutex with the PTHREAD_PROCESS_SHARED (pthread_mutexattr_setpshared)
attribute or use traditional unix semop(2) functions. Regular pthread mutexes
are normally limited to within a process.
Christian Steins <cs01@quantentunnel.de>: Nov 27 03:37PM +0100

Am 26.11.2016 um 17:33 schrieb Christian Steins:
 
>> neoGFX is coming soon!
 
> Cool, thanks, will check it out.
> http://neogfx.org/
 
The TestApp (test.exe) on that site is not working (blank window).
 
Chris
"Öö Tiib" <ootiib@hot.ee>: Nov 27 07:02AM -0800

On Sunday, 27 November 2016 16:37:18 UTC+2, Christian Steins wrote:
 
> > Cool, thanks, will check it out.
> > http://neogfx.org/
 
> The TestApp (test.exe) on that site is not working (blank window).
 
Most likely it is because MS has screwed up OpenGL support
of Windows and so people have to install and configure that
support manually.
Ramine <ramine@1.1>: Nov 26 10:17PM -0500

Hello,
 
Sorry i have posted about programming in general..
 
I think i must not post those posts here , comp.programming is well
suited for that.
 
So, sorry about that.
 
 
Thank you,
Amine Moulay Ramdane.
bleachbot <bleachbot@httrack.com>: Nov 27 03:17AM +0100

bleachbot <bleachbot@httrack.com>: Nov 27 03:29AM +0100

bleachbot <bleachbot@httrack.com>: Nov 27 03:55AM +0100

bleachbot <bleachbot@httrack.com>: Nov 27 04:39AM +0100

bleachbot <bleachbot@httrack.com>: Nov 27 05:17AM +0100

Popping mad <rainbow@colition.gov>: Nov 27 03:06AM

What exactly is this
 
T& operator=(T&& other) // move assignment
{
assert(this != &other); // self-assignment check not required
delete[] mArray; // delete this storage
mArray = std::exchange(other.mArray, nullptr); // leave moved-from in
valid state
return *this;
}
 
specifically T&& other
 
http://en.cppreference.com/w/cpp/language/operators
Ian Collins <ian-news@hotmail.com>: Nov 27 04:43PM +1300

On 11/27/16 04:06 PM, Popping mad wrote:
> }
 
> specifically T&& other
 
> http://en.cppreference.com/w/cpp/language/operators
 
http://en.cppreference.com/w/cpp/language/reference
 
and
 
http://en.cppreference.com/w/cpp/language/move_assignment
 
--
Ian
Ramine <ramine@1.1>: Nov 26 09:40PM -0500

Hello,
 
Again about programming...
 
If you have read my previous post, that we are always learning in
programming how to scale to a much bigger and complex problem.. and by
doing it, we are structuring ,like in mathematical logic, our thinking
and reasonning, also by using technics of programming like how to
calculate the time complexity and space complexity etc.
 
Take for example reusability, intelligence and scalability are inherent
also to reusability in programming, if you reuse smartly you will
be able to scale fast and to become smarter, this is one way
to prove my assertion above.
 
 
Thank you,
Amine Moulay Ramdane.
Ramine <ramine@1.1>: Nov 26 08:56PM -0500

Hello....
 
What is programming ?
 
I think the very important goal of programming is to make
us scale.
 
Take for example mathematical logic, you will learn for example
in mathematical logic that:
 
p -> q
 
is equivalent to:
 
not(q) -> not(p)
 
because it is infered mathematically from:
 
p -> q
 
is equivalent to:
 
not(p) or q
 
 
So what is exactly we are doing with this mathematical logic?
 
It permits us to scale, that means it permit us to make
our reasonning more complex by formalizing it with
mathematical logic.
 
Programming is the same:
 
We are always learning in programming how to scale to a much
bigger and complex problem.. and by doing it, we are structuring
,like in mathematical logic, our thinking and reasonning, also by using
technics of programming like how to calculate the time complexity
and space complexity..
 
 
Thank you,
Amine Moulay Ramdane.
Ramine <ramine@1.1>: Nov 26 08:29PM -0500

Hello....
 
What have i learned in Parallel programming during all my years of
learning ?
 
There is an important factor in programming that is called
scalability, and that's where i have specialized in parallel programming,
for example you have the criterion that is called maintainability, and
this factor called maintainability is in relation with scalability,
because if you want to make your code more maintainable you have to
learn the guidelines and the programming patterns to know how to scale
more on the the criterion of the cost of the time that you need to
maintain your program, and this is what you are learning on this C++
group also with templates etc., reusability is another criterion that is
in relation with scalability, because what you are trying to do is make
your classes and functions and programs reusable so that you can scale
on productivity for example.. portability is also another criterion that
is in relation with scalability, because when your code is portable you
will be able to run your code on more operating systems and hardwares
and that's scalability also.. on parallel programming that's the same,
you want your code and program to be scalable.
 
So here is what i have learned more, and how i have made my parallel
programming more scalable:
 
Take for example my efficient Threadpool engine with priorities that
scales well, here it is:
 
https://sites.google.com/site/aminer68/an-efficient-threadpool-engine-with-priorities-that-scales-well
 
So if you take a look at it, here is that i have made to make it scale
well:
 
1- minimize at best the cache-lines transfers, by using multiple queues
or multiple stacks, and by using lock-striping and by minimizing the
the cache-line transfers inside the queues or stacks by using a
better algorithm.
 
2- It uses work-stealing to be more efficient.
 
3- It can use processor groups on windows, so that it can use
more than 64 logical processors , so that it scales well.
 
4- It's NUMA-aware and NUMA efficient.
 
 
That's how i have learned more to make my programs more scalable.
 
 
 
Thank you,
Amine Moulay Ramdane.
Mr Flibble <flibbleREMOVETHISBIT@i42.co.uk>: Nov 27 02:34AM

On 27/11/2016 01:29, Ramine wrote:
> more than 64 logical processors , so that it scales well.
 
> 4- It's NUMA-aware and NUMA efficient.
 
> That's how i have learned more to make my programs more scalable.
 
You posted the same spam post twice so you could correct a single
spelling mistake of "were" instead of "where"? Are you fucking mental?
Mate, fuck off.
 
/Flibble
Ian Collins <ian-news@hotmail.com>: Nov 27 03:41PM +1300

On 11/27/16 03:34 PM, Mr Flibble wrote:
 
> You posted the same spam post twice so you could correct a single
> spelling mistake of "were" instead of "where"? Are you fucking mental?
> Mate, fuck off.
 
Abusing spam bots now?
 
--
Ian
Ramine <ramine@1.1>: Nov 26 08:18PM -0500

Hello....
 
What have i learned in Parallel programming during all my years of
learning ?
 
There is an important factor in programming that is called
scalability, and that's were i have specialized in parallel programming,
for example you have the criterion that is called maintainability, and
this factor called maintainability is in relation with scalability,
because if you want to make your code more maintainable you have to
learn the guidelines and the programming patterns to know how to scale
more on the the criterion of the cost of the time that you need to
maintain your program, and this is what you are learning on this C++
group also with templates etc., reusability is another criterion that is
in relation with scalability, because what you are trying to do is make
your classes and functions and programs reusable so that you can scale
on productivity for example.. portability is also another criterion that
is in relation with scalability, because when your code is portable you
will be able to run your code on more operating systems and hardwares
and that's scalability also.. on parallel programming that's the same,
you want your code and program to be scalable.
 
So here is what i have learned more, and how i have made my parallel
programming more scalable:
 
Take for example my efficient Threadpool engine with priorities that
scales well, here it is:
 
https://sites.google.com/site/aminer68/an-efficient-threadpool-engine-with-priorities-that-scales-well
 
So if you take a look at it, here is that i have made to make it scale
well:
 
1- minimize at best the cache-lines transfers, by using multiple queues
or multiple stacks, and by using lock-striping and by minimizing the
the cache-line transfers inside the queues or stacks by using a
better algorithm.
 
2- It uses work-stealing to be more efficient.
 
3- It can use processor groups on windows, so that it can use
more than 64 logical processors , so that it scales well.
 
4- It's NUMA-aware and NUMA efficient.
 
 
That's how i have learned more to make my programs more scalable.
 
 
 
Thank you,
Amine Moulay Ramdane.
You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page.
To unsubscribe from this group and stop receiving emails from it send an email to comp.lang.c+++unsubscribe@googlegroups.com.
Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)