This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: rs6000 -mcpu overrides -msoft-float
- From: Alan Modra <amodra at bigpond dot net dot au>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Mon, 8 Mar 2004 15:32:10 +1030
- Subject: Re: rs6000 -mcpu overrides -msoft-float
- References: <20040221125205.GD3427@bubble.modra.org> <200403080249.i282n0T34526@makai.watson.ibm.com>
On Sun, Mar 07, 2004 at 09:48:59PM -0500, David Edelsohn wrote:
> * config/rs6000/rs6000.c (rs6000_override_options): Don't override
> -msoft-float by -mcpu. Consolidate similar code for MASK_MULTIPLE
> and MASK_STRING.
>
> Okay for both. Next patch in this function, I want to see confirmation by
> a VHDL logic verifier. :-)
Thanks. Committed. I also committed the fix you okayed on irc a few
days ago for the offset range check in legitimate_offset_address_p.
For the record, I think my rs6000_override_options fix doesn't go far
enough. Options other than -msoft-float will suffer a similar fate,
eg. -mcpu=620 -mno-powerpc-gfxopt will result in MASK_PPC_GFXOPT being
set.
Perhaps the answer is to use
set_masks &= ~target_flags_explicit;
instead of
set_masks &= ~(target_flags_explicit
& (MASK_MULTIPLE | MASK_STRING | MASK_SOFT_FLOAT));
--
Alan Modra
IBM OzLabs - Linux Technology Centre