[m-rev.] [reuse] diff: user-annotated aliasing

[Peter Ross]

On Thu, Mar 22, 2001 at 11:06:23AM +0100, Nancy Mazur wrote:
> Hi Peter,
> 
> Okay... this is the full diff... I will commit it now. 
> I've annotated array.m to test the changes, and it seems
> to work smoothlessly. What I haven't checked yet is the
> purpose of converting the pragma-annotations back to strings. 
> So for the alias-annotatations, it's still incomplete. But
> I don't know to what extent it has to be really complete
> (is a bit more tricky than the other ones, as there are 
> variables involved). 
> 

> Oh yes, next question, how can I change the library, and 
> have all the stuff bootchecked with an annotated string-module ?
> 
You need to install a version of the compiler which understands these
annotations on pragma foreign_code, and then use that to build a new
copy of the library where array.m has the annotations.  Also you might
want to consider modifying the main branch of the compiler so that it
parses and then ignores the aliasing annotations.  Then wait for the next
rotd to roll around and install that and hence avoid bootstrapping
problems for putting the annotations in the library.

Also you could just copy the modify array.m module into the icfp
programming entry directory.

> Nancy
> 
> PS: in attach my annotated module marray.m and the 
> resulting marray.trans_opt.
> 
Just a quick-look but shouldn't marray__from_list have a possible alias
between the elements of the list and elements of the array, because
currently it is bottom.

Pete
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------