Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.programming.threads > #2308

Re: And improved algorithm of a concurrent FIFO queue

Cross-posted to 3 groups.

Show all headers | View raw

> "animer":

> my new and improved algorithm has now 50% better throughput on the pop() 
> side than the following Chris Thomasson concurrent FIFO queue that uses 
> the bakery algorithm, and it has the same throughput on the push() side 
> than the following Chriss Thomasson concurrent FIFO Queue.

50% better throughput on the pop side! AFAICT, that is a total
crock of fuc%ing bullshi%! Anyway, what the hell are all of those
nasty Sleep(0)'s littered around your code?

BTW, can you tell me the usage pattern for an eventcount? I
have doubts that you know how to properly use one, let alone
implement one.


FWIW, here my first implementation of an eventcount here back
in 2005

https://groups.google.com/d/topic/comp.programming.threads/qoxirQbbs4A/discussion


Joe Seigh has some excellent feedback dealing with the proper
memory barriers and common usage pattern in the thread...


Now, learn the basics' of this stuff, and leave my name out of
it in the meantime!

Personally, I don't trust ANY of your code because you fail to at
least try to check it against verification tools. 

Back to comp.programming.threads | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

And improved algorithm of a concurrent FIFO queue aminer <aminer@toto.net> - 2014-05-06 20:42 -0700
  Re: And improved algorithm of a concurrent FIFO queue "Chris M. Thomasson" <no@spam.invalid> - 2014-05-06 18:59 -0700
    Re: And improved algorithm of a concurrent FIFO queue Melzzzzz <mel@zzzzz.invalid> - 2014-05-07 05:38 +0200

csiph-web