This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Irix6 long doubles implemented wrong? (27_io/ostream_inserter_arith)
On Wed, Jan 29, 2003 at 11:15:02AM -0500, Kaveh R. Ghazi wrote:
> > From: Richard Henderson <rth@redhat.com>
> >
> > On Tue, Jan 28, 2003 at 08:32:19PM -0500, Kaveh R. Ghazi wrote:
> >
> > floathitf2
> > floatunssitf2
> > floatunsditf2
> > negtf2
> > fix_trunctfhi2
>
> Sorry I wasn't clear. I can conceptually come up with the names, what
> I don't see are entries for them in libfuncs.h so I assume I can't
> just start using them.
>
> Is it safe to simply add the appropriate entries to enum libfunc_index
> and magically everything will work?
No, you'd have to modify code as well. See how the _libfuncs
are used in optabs.c. Probably no sense in adding the HImode bits.
The unsigned conversions on the other hand... Well, a little
rearrangement of the code is needed. GCC does not itself provide
unsigned->float conversion routines (which is silly) and so it
synthesizes with libcall + extra code.
r~