Collision-resistant hashing

[Jonathan Schilling]

> From: Mark Mitchell <mark at codesourcery.com>
> 
> >>>>> "Jim" == Jim Dehnert <dehnert at baalbek.engr.sgi.com> writes:
> 
>     Jim> I'm intrigued by Jonathon's idea of separating the link name
>     Jim> from the "real" mangled name.  Suppose we used the
> 
> I thought that was my idea. :-) :-) 

It was!  I just made a comment on one point and Jim must have misread
the mail quotings.  

In fact, I'm a little skeptical about this approach.  It probably
violates the user expectation that the linker symbol table name is
also the mangled name; they won't realize that there's a side table
somewhere that is needed to correlate the two.  And I think "c++filt" 
has to work all the time, not just when code is built with -g.  

On the other hand space/speed savings are certainly desirable.  If
simple mangled name compression can't come close to what the hashing
approach offers, then the latter is worth considering; my point is
just that it should be as transparent and unsurprising as possible.

Jonathan Schilling		SCO, Inc.		jls at sco.com