[PATCH] Re: Yet another sparc-sun-solaris2.8 bootstrap failure

[Gerald Pfeifer]

On Tue, 9 Apr 2002, Richard Henderson wrote:
>> We should really make multilibbing work in this case I think.
> *shrug* I don't think this is that important.  How many folks
> are running the broken ultras anyway?

Lots, as far as I know.  And (surprisingly?) many run non-broken Ultras
with a 32-bit kernel for others reasons.

On Tue, 9 Apr 2002, Richard Henderson wrote:
> Actually, I think I'd rather config.guess detected this host as
> sparc32-solaris or something, so that we can know what happened
> when some user builds the compiler and files a bug report.
>
> Or something.  Perhaps "sparc32" is confusing.  Perhaps
>
> 	sparc-solaris		32-bit only
> 	sparcv9-solaris		32-bit default, 64-bit capable
> 	sparc64-solaris		64-bit default
>
> It would change the current behaviour of sparcv9, but if we
> document that, I don't think folk will mind *too* much.

Having some kind of clear distinction between these three cases would
be very helpful, yes.  (How to actually name the targets probably isn't
that important, in the end.)

Gerald
-- 
Gerald "Jerry" pfeifer@dbai.tuwien.ac.at http://www.dbai.tuwien.ac.at/~pfeifer/