Subj : Re: posix and lock-free algorithms
To   : comp.programming.threads
From : Alexander Terekhov
Date : Thu Aug 11 2005 02:38 pm


John Doug Reynolds wrote:
[...]
> But the "reader" thread does place a memory barrier between its read
> of pInstance and any attempt to access *pInstance.  

Not really. Because apart from the fact that your code is data-race 
UNfree, lock-unlock under POSIX-compatible release consistency models 
doesn't really substitute a bidirectional fence. It would sorta work 
in your case (violation of POSIX data-race free requirement aside for 
a moment), but only if you lock-unlock the *same* mutex on both slow 
and "fast" paths, not different mutexes on the stacks of 
participating threads.

regards,
alexander.

.