February 2011 - libdispatch-dev - lists.macosforge.org

libpthread_workqueue ported to Windows
by Mark Heily 22 Feb '11

22 Feb '11

I have ported libpthread_workqueue to Windows using the MS Visual Studio toolchain and the built-in Windows threadpool API. For compatibility with Windows XP, I am using the "legacy thread pool API" as described here: http://msdn.microsoft.com/en-us/library/ms686756%28v=vs.85%29.aspx The library also builds using the MinGW toolchain, but there are still a few bugs in the build system that need to be fixed before it can be considered fully functional. I havn't tried building a combined libdispatch+libpthread_workqueue on Windows, but will give it a try soon. Cheers, - Mark

1 0

Fully functional libdispatch port for Windows and Linux
by Marius Zwicker 20 Feb '11

20 Feb '11

Hi everyone, I'm glad to announce that I finished a port of libdispatch to Linux and Windows during the last days. Except the source interfaces (which differ too much between Windows and Linux) the API should be fully operational. The majority of your original tests is passing - some of them still need to be ported to build and run on all platforms. As I did not manage to understand your code completely, I did a complete rewrite and merely reused your api and general ideas. For achieving "blocks support" on the majority of the used compilers I added lambda support from the upcoming C++0x standard. Additionally I implemented (for those who need and use it) an interface to integrate libdispatch into applications developed using the Qt framework. As my port is a bit more than a patch, I named it libxdispatch and it is currently available by going to http://opensource.mlba-team.de/xdispatch - I'm really curious about your comments. Regards, Marius Zwicker

4 5

[libdispatch-dev] _dispatch_queue_wakeup_global/_dispatch_worker_thread2 use of dgq_pending - question
by Joakim Johansson 14 Feb '11

14 Feb '11

Hi, I am spending a bit of time investigating possible improvements to libpthread_workqueue and reviewed some of the libdispatch code and found one small thing I didn’t quite understand (but it might simply be my lack of understanding the semantics of the gcc atomic ops fully): ------------ queue.c: 1490: void 1491: _dispatch_worker_thread2(void *context) { . . . 1502: qc->dgq_pending = 0; —————— Shouldn’t this be set using atomic primitives to ensure visibility of the update of dgq_pending? (_dispatch_queue_wakeup_global() uses dispatch_atomic_cmpxchg() to set this) I would have expected something like (#ifdef:ed here to not warn when running without pthread workqueues, as it would be called by a path that doesn’t set dgq_pending): —— #if HAVE_PTHREAD_WORKQUEUES if (dispatch_atomic_cmpxchg(&qc->dgq_pending, 1, 0)) // zero out, ensure visibility { _dispatch_debug("got new worker thread for global queue: %p", dq); } else { _dispatch_debug(“unexpectedly got a new worker thread for global queue: %p", dq); } #endif —— Any enlightenment appreciated. Thanks, Joakim

2 2