261
Subj : Re: posix and lock-free algorithms
To   : comp.programming.threads
From : Alexander Terekhov
Date : Sat Aug 13 2005 01:30 pm


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

Note that Boehm's atomic<> (still) has "none" that I've sorta 
deprecated (new naked stuff labels naked_competing, 
naked_noncompeting and naked_exclusive replaced msync::none), and 
provides neither cc/dd nor hoist/sink (finer control over memory 
ordering stuff). 

regards,
alexander.

.

0