Comments on the C++ Concurrency TS
Wednesday, 28 May 2014
It's been a while since I wrote any papers for the C++ committee, but I've written two for the committee mailing prior to the upcoming committee in Rapperswil:
- N4032: Comments on continuations and executors
- N4033: synchronized_value<T> for associating a mutex with a value
The first provides comments, and suggestions for improvements on the concurrency TS based on implementing continuations for Just::Thread V2, and executors for an unreleased internal build of Just::Thread.
The second proposes to standardize
the synchronized_value
class template from Just::Thread
Pro, with a couple of modifications.
Let me know if you have any comments.
Posted by Anthony Williams
[/ news /] permanent link
Tags: multithreading, concurrency, C++17
Stumble It! | Submit to Reddit | Submit to DZone
If you liked this post, why not subscribe to the RSS feed or Follow me on Twitter? You can also subscribe to this blog by email using the form on the left.
Design and Content Copyright © 2005-2024 Just Software Solutions Ltd. All rights reserved. | Privacy Policy
6 Comments
Hey dear get youtube subscribers after visiting the given place and you will get more traffic to it as you will be benefited very much.
Great Post !! your blog is very informative information share....
I would prefer synchronized_value<T> created the update_guard for me:
synchronized_value<string> syncString;
auto sref = syncString.get(); // returns update_guard<string>
*sref = "...";
return *sref;
and then you can get rid of the operator overloads on synchronized_value<T> entirely.
Between the advent of move semantics and 'auto', this seems cleaner and simpler to understand.
Also seems unnecessarily restrictive that you can't dictate the type of the mutex that is used.
Hi guys, Its very time taking.. so can anyone suggest me too..??
Thanks for the valuable information