This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Fix, was: Re: [Fwd: CVS trunk bootstrap failure on alphaev6-unknown-linux-gnu]


Alexandre Oliva wrote:

> On Jul  8, 2001, Toon Moene <toon@moene.indiv.nluug.nl> wrote:
> 
> > OK to install ?
> 
> I think it would still be a good idea to pass the mode as an argument,
> just in case the two other arguments are CONST_INTs (remember that
> CONST_INTs don't have a mode).  In fact, your proposed patch would
> probably crash or just fail to do right thing should the first
> argument be a CONST_INT.
> 
> However, since all uses of these instructions take a register as an
> argument (and only now, after having written the paragraph above, do I
> see it can only make sense to call have_add2_insn having a REG in the
> first argument, I think it may indeed be ok.  Then perhaps it would be
> enough to check that GET_MODE (x) != VOIDmode, and abort () otherwise.

Good idea - perhaps someone can ACK it with that change assumed (I just
noticed that the MAINTAINERS file lists no-one as the maintainer of
reload - perhaps adding to Mark Mitchell's observation that reload is
not Free Software, because it cannot be modified ;-)

-- 
Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290
Saturnushof 14, 3738 XG  Maartensdijk, The Netherlands
Maintainer, GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.html
Join GNU Fortran 95: http://g95.sourceforge.net/ (under construction)


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]