Re: [MiNT] MiNT lib 55

[Johan Klockars <rand@cd.chalmers.se>]

> > > This 68020-60 flag is not a bad idea, in itself, as it simplifies
> > > compiling binaries for a larger range of hardware, but it _does_
> > > mean code that is somewhat less optimized than it could be.

Indeed it does, but that should mostly concern those of us with '040's,
since those are generally the 'odd man out'.

> > Sorry, I disagree. With m68020-60 the gcc generate very good code that
> > run efficient on 68020, 68030, 68040 and 68060.

I'd assume that m68020-60 generates code that's relatively good for the
'020/'030 and reasonably so for the '060 (it probably doesn't take
advantage of the quick integer multiply, since that would hurt the older
processors), but somewhat worse for the '040.

Motorola used to have a document that outlined the general optimization
strategies for their m68k processors, but I can't find it now.

> I would agree here. The68k family is in fact splitted into two
> sub-families :) i.e. 68000/10 at one side, and anything else at the
> other one. Considering user code, there are almost no differences between
> 68020 and 68060, so no real need to make specific optimizations.

There are actually a great many differences that are important when
it comes to optimization. Sure, the available user mode instructions
are more or less the same (I believe the '020 is the only one with
CALLM, and some things were removed in the '060), but the timings and
the cache behaviour vary significantly. And, after all, the timings are
usually what's interesting when you're talking about optimization.

-- 
  Chalmers University   | Why are these |  e-mail:   rand@cd.chalmers.se
     of Technology      |  .signatures  |            johan@rand.thn.htu.se
                        | so hard to do |  WWW/ftp:  rand.thn.htu.se
   Gothenburg, Sweden   |     well?     |            (fVDI, MGIFv5, QLem)