Re: Q: access to subprogram

[Ronald Cole <ronald@ridgecrest.ca.us>]

dewar@cs.nyu.edu (Robert Dewar) writes:
> Ronald Cole said (commenting on my claim that GCC did not have multiple
> backends)
> 
> Not true.  A single instance of GCC doesn't support multiple machine
> targets (other than the -m options for some processor families).  GCC
> has to be recompiled to support different machine descriptions (that's
> what the -b and -V options of gcc are for).  I consider this as having
> multiple backends.
> 
> That's not what I meant, yes, of course there are multiple backends at
> the object level, but not at the source level.

Again not true.  Yes, the source contains all the parts for each
backend, but all the parts are not used to build the backend.
Configure() will determine the correct source files and create links
to properly build one target backend.  The multiple backends at the
source level are kept under the config subdirectory and grouped by
architecture.  I count thirty distinct backends:

1750a/1750a.md      fx80/fx80.md        m88k/m88k.md        sh/sh.md
a29k/a29k.md        gmicro/gmicro.md    mips/mips.md        sparc/sparc.md
alpha/alpha.md      h8300/h8300.md      ns32k/ns32k.md      spur/spur.md
arm/arm.md          i370/i370.md        pa/pa.md            tahoe/tahoe.md
clipper/clipper.md  i386/i386.md        pdp11/pdp11.md      vax/vax.md
convex/convex.md    i860/i860.md        pyr/pyr.md          we32k/we32k.md
dsp16xx/dsp16xx.md  i960/i960.md        romp/romp.md
elxsi/elxsi.md      m68k/m68k.md        rs6000/rs6000.md

> Most Ada technologies have a single front end for all targets (though
> Ronald Cole would presumably call these multiple front ends too, because
> there are certanly separate object versions),

You presume incorrectly.  I consider the frontend to be defined by the
language it accepts, not by the host architecture or the number of
object files in an implementation.  Likewise, I consider the backend
to be defined by the object code it produces.

> But for technologies like GCC, where there is only one source program for
> both front end and backend (though in both cases multiple executables),
> this distinction is not significant.

But it is.  While the union of the sources for each target
architecture is the GCC source tree, the difference of the sources for
each target architecture is not the null set.  

-- 
Ronald Cole                                     E-mail: ronald@ridgecrest.ca.us
President, CEO                                          zippy@ecst.csuchico.edu
Forte International                                Fax: (619) 384-2346
     My PGP fingerprint: E9 A8 E3 68 61 88 EF 43  56 2B CE 3E E9 8F 3F 2B