[m-dev.] for review: direct reuse

Tyson Dowd trd at cs.mu.OZ.AU
Mon Sep 25 16:14:15 AEDT 2000


On 22-Sep-2000, Peter Schachte <schachte at cs.mu.OZ.AU> wrote:
> On Fri, Sep 22, 2000 at 12:20:10PM +1100, Zoltan Somogyi wrote:
> > Yes. However, in the far future (say three years from now ;-), it may be
> > possible for the compiler to analyze the program and some profile data,
> > find out where non-uniform cell sizes are limiting reuse, and automatically
> > modify its internal type definitions to *make* the cell sizes uniform
> > (on at least some of the function symbols in the type) if doing so would
> > gain more than it loses.
> 
> That sounds like a good idea, but it requires global analysis and compilation.
> I thought that was philosophically verboten?

IMHO the philosophy behind Mercury is "no global analysis and
compilation required for fundmental efficiency".  I don't see any
problem with it for incremental improvements in efficiency.

Profiling data like that mentioned above is really just an extra
annotation on the type, it just so happens that the annotation was
generated as a result of a global process (e.g. profiling that data).
It could have been produced by local analysis or even guesswork.

-- 
       Tyson Dowd           # 
                            #  Surreal humour isn't everyone's cup of fur.
     trd at cs.mu.oz.au        # 
http://www.cs.mu.oz.au/~trd #
--------------------------------------------------------------------------
mercury-developers mailing list
Post messages to:       mercury-developers at cs.mu.oz.au
Administrative Queries: owner-mercury-developers at cs.mu.oz.au
Subscriptions:          mercury-developers-request at cs.mu.oz.au
--------------------------------------------------------------------------



More information about the developers mailing list