alloc_final() and alloc()

alloc_final() and alloc()

[Vincent BARAT]

Hi !

I'm new to the list and I am working on a project whose aim is
to permit to write CORBA client and servers in Objective ML
on top of a generated stub (ML and C++) that works under VisiBroker
for the moment but should work with any ORB with very few modifications
(because the generated stub is on top of the C++ mapping of CORBA).
This project is very near to its end (3 or 4 days now !). What I could
say for the moment is that it is far more easy to write clients and
servers
in ML than with C++ !!!

So, I've discovered a problem with alloc(): it seems to return
always the same address if it is used several times in the same function
that contains also calls to alloc_final().

Is it a known bug ? If not, am I missing something ?

Thanks for your answer.

P.S. : for the moment, instead of calling alloc(), I call alloc_shr()
which work well concurrently with alloc_final().

Re: alloc_final() and alloc()

[doligez]

>From: Vincent BARAT <vincent.barat@alcatel.fr>

>So, I've discovered a problem with alloc(): it seems to return
>always the same address if it is used several times in the same function
>that contains also calls to alloc_final().
>
>Is it a known bug ? If not, am I missing something ?

It is possible if you're giving strange arguments to alloc_final,
in which case, it is not a bug.

If you could send the relevant piece of your code to
<caml-light@inria.fr> (which is the correct address for bug reports),
we should be able to tell what's going on.

-- Damien