Subj : Re: posix and lock-free algorithms
To   : comp.programming.threads
From : David Hopwood
Date : Sat Aug 13 2005 05:54 pm

Alexander Terekhov wrote:
> David Hopwood wrote:
> [...]
> 
>>Why do you assume a release consistency model?
> 
> Well, I actually assume a bit more relaxed one. But see
> 
> http://tinyurl.com/7n27b

enum ordering_constraint {none, acquire, release, ordered};
	// Informally:
	// none ==> This operation is unordered, and may become
	//          visible to other threads in an order that is
	//          constrained only by ordering constraints on other
	//          operations.
	// release ==> All prior memory operations (including ordinary
	//          assignments) become visible to a an acquire
	//          operation on the same object that sees the resulting
	//          value.
	// acquire ==> See above.
	// ordered ==> Both acquire and release ordering properties.

"All prior memory operations (including ordinary assignments)". That's
all read and write accesses performed by the C or C++ abstract machine,
right? Regardless of optimization?

-- 
David Hopwood <david.nospam.hopwood@blueyonder.co.uk>

.