I'll take this one.

greg



On Tue, 4 Mar 2014, Michael Ferguson wrote:

> Modify threads-pthreads to possibly use __thread for TLS
>
> In an approach taken from qthreads, we modify the threads-pthreads
> implementation to use __thread when it is available by defining
> macros for declaring, initializing, deleting, reading, and writing
> thread local storage. When __thread is not available, we use
> the usual pthread_key_create/delete pthread_getspecific/setspecific calls.
>
> Lastly, I noticed that in pthread_func, the threadEnd function is
> not called if the thread is not canceled (but the thread function
> completes normally). I believe that it should be called in either
> case -- it might free memory for example (and does for FIFO), so this
> patch calls pthread_cleanup_pop(1) -- the 1 meaning to invoke the
> cleanup handler. I don't think that this comes up with FIFO's current
> use of threads-pthreads because it seems that thread_begin never exits;
> nonetheless I think this is a worthwhile bug fix in case the future
> usage pattern changes.
>
>

------------------------------------------------------------------------------
Subversion Kills Productivity. Get off Subversion & Make the Move to Perforce.
With Perforce, you get hassle-free workflows. Merge that actually works. 
Faster operations. Version large binaries.  Built-in WAN optimization and the
freedom to use Git, Perforce or both. Make the move to Perforce.
http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk
_______________________________________________
Chapel-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/chapel-developers

Reply via email to