[m-rev.] for review: version array thread_safe attributes

Julien Fischer juliensf at csse.unimelb.edu.au
Fri Apr 13 21:25:47 AEST 2012


On Fri, 13 Apr 2012, Peter Wang wrote:

> It turns out that the C implementation of version arrays wasn't ever
> thread-unsafe due to the global lock.  The Java (and C#) backend don't
> implement a global lock and that's why we saw thread-safety issues, and
> added locking to version arrays.  As a result, the C implementation is
> extremely thread-safe!  Anyway, per-array locks are better than relying
> on the global lock so it all worked out, eventually.
> --
>
> Branches: main, 11.07
>
> library/version_array.m:
> 	Set `thread_safe' attributes on version array predicates,
> 	with justification.
>
> 	Conform to renaming of `new' to `init' previously.
> 	The Java and C# foreign_procs were not renamed.
>
> 	Reword a bit of documentation.

That looks fine.

Julien.
--------------------------------------------------------------------------
mercury-reviews mailing list
Post messages to:       mercury-reviews at csse.unimelb.edu.au
Administrative Queries: owner-mercury-reviews at csse.unimelb.edu.au
Subscriptions:          mercury-reviews-request at csse.unimelb.edu.au
--------------------------------------------------------------------------



More information about the reviews mailing list