This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [ColdFire 9/63] Add -mcpu, -march, -mtune, and related options
On Wed, 2007-01-10 at 10:54 +0000, Richard Sandiford wrote:
> An obvious implication is that -m68020 should be orthogonal to the
> FPU selection, but the other architecture options shouldn't be.
> However, I don't that's necessarily the reason why the code is
> the way it is, because...
I have no idea why it was done that way. It's never been terribly
rational, though the irrationality is easier to see with the removal
of the FPA support a few years ago.
>
> The m68k-* configurations all default to hard-float 68020, and if
> you pass -m68020 to the m68k-* gccs, you still get hard-float code.
> The only configuration for which the special treatment of -m68020 makes
> a difference _in isolation_ is m68010-*-netbsdelf, which continues to
> generate soft-float code when passed -m68020. However, I doubt that
> the netbsd effect was a deliberate decision. It seems more logical
> for "m68010-*-netbsdelf-gcc -m68020" to generate the same code as
> m68k-*-netbsdelf-gcc, which is the equivalent 68020 configuration,
> and which defaults to hard-float code.
Agreed (it probably wasn't a deliberate decision). Particularly if they
used my ancient 68010 HPBSD support as a template for netbsdelf
(netbsdelf is a descendent of the HPBSD code on numerous levels).
Jeff
ps. I'm just going to blindly trust that you've got all the ColdFire
variants right -- you're *far* more familiar with them than I am.