Hi David, This is a very hard subject, as validating a memory model is an incredible hard task.
Right now we provide loose semantics close to ecma backed by the architecture you're running. This means we do a very different thing in, say, X86 and Itanium. Having a consistent memory model is a very hard task that requires a huge amount code reviewing and testing. Right now, as you can see, we haven't done all the required work. We still have work to do on the runtime itself (we have some racy code that deal with memory in a hazardous way); and work on the JIT as well. The JIT side is going to suffer a major change right after the 2.0 branch so we are not very inclined in spending much time chaising any possible corner case of the ecma memory model on the current one. Anyway, if you get reproducible bugs please fill bug reports so we can work on improving mono. Cheers, Rodrigo On Tue, Jul 1, 2008 at 1:21 PM, David Wolinsky <[EMAIL PROTECTED]> wrote: > I am so entirely confused about what the memory model for Mono is. If > we are using gmcs can we assume .Net 2.0 Framework or are we safe always > falling back on what ECMA says? This article ( > http://msdn.microsoft.com/en-us/magazine/cc163715.aspx ) makes me want > to code for ECMA and not .Net, but we have a project that was already > coded for the .Net memory model and it is having issues in Mono that are > incredibly difficult to track down! > > > This post ( > http://lists.ximian.com/pipermail/mono-devel-list/2008-May/027885.html ) > doesn't help but confuse me more. Is there somewhere where this is > listed now, if not could you put up the memory models used by Mono on > www.mono-project.org? > > Cheers, > David > _______________________________________________ > Mono-devel-list mailing list > Mono-devel-list@lists.ximian.com > http://lists.ximian.com/mailman/listinfo/mono-devel-list >
_______________________________________________ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
