Re: RawData ?

[Andrew Haley <aph at redhat dot com>]

Michael Koch writes:
 > -----BEGIN PGP SIGNED MESSAGE-----
 > Hash: SHA1
 > 
 > Am Dienstag, 8. April 2003 21:08 schrieb Andrew Haley:

 > >
 > > I'm wondering if it might be a good idea to have a type which is
 > > the equivalent of "integer attribute ((mode(pointer)))" or
 > > somesuch. However, it wouldn't need to be an arithmetic type.
 > >
 > > We could then replace all instances of RawData with that.  We
 > > wouldn't be able to use it for an IdentityHashmap key, which would
 > > be a shame.
 > >
 > >  > If Classpath does continue with this, we could consider
 > >  > changing.  A patch to do this would also have to change the
 > >  > compiler (that's easy in this case though).
 > >
 > > Seems reasonable.
 > 
 > It is possible to use long instead of RawData.

The problem is that the reader will find it hard to understand the
code.  When you see long, do you think "pointer"?  

 > But then we need conversion methods I think (void* -> jlong, jlong
 > -> void*). With RawData we can just use a <reinterpret_cast> which
 > has more ore less no cost at runtime.

Well, the cost of long isn't high.  There's no cost at all for jlong->void*.

 > As I understand it using long is in general more safe in regard to 
 > security holes, right ?

Well, I don't know.  If we expose some code that uses a pointer, and
we can fabricate a pointer to it, then we have a security hole.  So
the only solution, as Per said, is to make very sure that methods
which take a pointer in whatever form are private or protected.
("friend" would be nice here...)

 > We will have to do a lot of work to make sure we are on the safe
 > side when we use RawData overall.

Yes.  It would be nice to get rid of RawData, but it is a good
solution to some problems.

Andrew.