Subj : Re: Memory Barriers, Compiler Optimizations, etc.
To   : comp.programming.threads
From : Alexander Terekhov
Date : Mon Feb 07 2005 10:28 pm


"Neill Clift [MSFT]" wrote:
[...]
> InterlockedCompareExchange to do
> one time initialization and if the call fails using the return
> as the initialized value.
> I would consider it a bug if we didn't honour this and push
> for it to be fixed in current platforms.

Your logic nicely illustrates superiority of "the original" 
IBM style CAS. ;-)

regards,
alexander.

.