[concurrency-interest] question the about the JMM
dcholmes at optusnet.com.au
Fri Dec 7 00:03:53 EST 2007
> The problem with using "cache coherence" as a logical model for the JMM is
> two-fold. First, it is insufficient as it doesn't address reorderings that
> the runtime compiler is allowed to do.
I misspoke. I was confusing hardware cache-coherency protocols (which know
nothing about compiler reorderings) and the logical "cache" model as used in
the original JMM. The latter can of course address compiler re-orderings.
Sorry for the confusion.
More information about the Concurrency-interest