[m-rev.] for review: threadsafe version_arrays in Java

[Paul Bone]

On Mon, Apr 19, 2010 at 12:58:44PM +1000, Peter Wang wrote:
> On 19 April 2010 12:52, Paul Bone <pbone at csse.unimelb.edu.au> wrote:
> > On Mon, Apr 19, 2010 at 12:17:15PM +1000, Peter Wang wrote:
> >> >> @@ -691,6 +761,9 @@ ML_va_set(ML_va_ptr VA0, MR_Integer I, MR_Word X,
> >> >> ML_va_ptr *VAptr)
> >> >>          VA1->index      = -1;
> >> >>          VA1->value      = (MR_Word) NULL;
> >> >>          VA1->rest.array = VA0->rest.array;
> >> >> +#ifdef MR_THREAD_SAFE
> >> >> +        VA1->lock       = VA0->lock;
> >> >> +#endif
> ...
> >
> > But do you know that this is a reference to the same lock (pointer to a lock)
> > or just a copy of the lock?
> 
> Same lock.
> 
> struct ML_va {
>     ...
> #ifdef MR_THREAD_SAFE
>     MercuryLock         *lock;  /* NULL or lock                  */
> #endif
> };
> 

Doh!

I forgot you where using a pointer as a 'maybe there is a lock' type.

Cheers.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 489 bytes
Desc: Digital signature
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20100419/23e5b2b9/attachment.sig>