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


Groups > comp.programming.threads > #2195 > unrolled thread

Another solution to my concurrent FIFO queue...

Started byaminer <aminer@toto.net>
First post2014-04-19 21:13 -0700
Last post2014-04-19 21:13 -0700
Articles 1 — 1 participant

Back to article view | Back to comp.programming.threads


Contents

  Another solution to my concurrent FIFO queue... aminer <aminer@toto.net> - 2014-04-19 21:13 -0700

#2195 — Another solution to my concurrent FIFO queue...

Fromaminer <aminer@toto.net>
Date2014-04-19 21:13 -0700
SubjectAnother solution to my concurrent FIFO queue...
Message-ID<liv707$84e$1@news.albasani.net>
Hello,

As i told you before in my previous post called "Please read again"
i have discovered a serious problem with the Ticketspinlock and
also with the array based locks and also with the queue locks.. so you 
have to be very carefull about this serious problem that look like
lock convoy , i have just tested it and it has showed up on
my computer in less than 2 minutes, but i have found a solution
to my concurrent FIFO queue , i have used this time a Spinlock
wht a backoff inside my SemaCondvar and used my SemaCondvar with my 
concurrent FIFO queue and  my concurrent FIFO queue has given more than 
2x throughput at exactly 7.4 millions of transactions per second with
my 2.4 GHz Quadcore even with my semacondvar and even with blocking wait 
if there is no item in the queue and it's still FIFO fair on the push() 
side.

I will soon upgrade my concurrent FIFO queue and i am sure
that you will be happy.


Thank you,
Amine Moulay Ramdane.



[toc] | [standalone]


Back to top | Article view | comp.programming.threads


csiph-web