Subj : Re: What does Memory Barriers mean ?? To : comp.programming.threads From : Alexander Terekhov Date : Fri Jan 07 2005 11:07 am David Hopwood wrote: > > SenderX wrote: > >>>>Volatiles are braindead. > > > > The C/C++ std shoud rip its dead brain out, and replace it with a new one > > that can actually comprehend threads and memory visibility... > > > > ;) > > No ;) required. . FWIW, here's what I wrote in November to a 'couple' among "Other participants in this effort" from section 6. ---- I think that POSIX MT memory model should be tackled first. C++ would "simply extend it" with things like atomic<> and "isolated<>". http://www.opengroup.org/sophocles/show_mail.tpl?source=L&listname=austin-group-l&id=7675 http://www.opengroup.org/sophocles/show_mail.tpl?source=L&listname=austin-group-l&id=7682 http://www.opengroup.org/sophocles/show_mail.tpl?source=L&listname=austin-group-l&id=7688 I just hate volatiles. Given that they already have some defined "single threaded" semantics (e.g. longjmp() and automatic storage) in C/POSIX, I'd leave them alone and wouldn't go Java route trying to atomic<>-ize them and impose implicit (and needlessly heavy) msync semantics on top of all that mess. ---- regards, alexander. .