Talk:Mutex: Difference between revisions

Content added Content deleted
(Win32 API CRITICAL_SECTION)
(Not familiar enough with critical sections.)
Line 1: Line 1:
I added some C/C++ code for the Win32 API. My reasoning is that--parallelization algorithms aside--how to access synchronization objects in a language is becoming almost as important as that language's conditional and loop processing facilities. I think code should be offered for other languages (and APIs...I don't know how to do it in pthreads or C++0x) as well.--[[User:Short Circuit|Short Circuit]] 04:57, 3 November 2008 (UTC)
I added some C/C++ code for the Win32 API. My reasoning is that--parallelization algorithms aside--how to access synchronization objects in a language is becoming almost as important as that language's conditional and loop processing facilities. I think code should be offered for other languages (and APIs...I don't know how to do it in pthreads or C++0x) as well.--[[User:Short Circuit|Short Circuit]] 04:57, 3 November 2008 (UTC)
: With Win32 API CRITICAL_SECTION objects are often used as local mutexes, non-reentrant, if I correctly remember. They are less heavy weighted than mutex objects, which can serve as global mutexes. --[[User:Dmitry-kazakov|Dmitry-kazakov]] 09:25, 3 November 2008 (UTC)
: With Win32 API CRITICAL_SECTION objects are often used as local mutexes, non-reentrant, if I correctly remember. They are less heavy weighted than mutex objects, which can serve as global mutexes. --[[User:Dmitry-kazakov|Dmitry-kazakov]] 09:25, 3 November 2008 (UTC)
:: I know of CRITICAL_SECTIONs and of some of their tradeoffs, but I haven't used them enough to write code from memory. I also don't have a Win32 programming environment readily available at home, and I'm prohibited from working on Rosetta Code on company hardware. If someone wanted to add setup, lock, unlock and destruction code for CRITICAL_SECTIONs, that would be entirely appropriate, though. --[[User:Short Circuit|Short Circuit]] 03:58, 4 November 2008 (UTC)