This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: pattern problem with register assignment
- From: Ian Lance Taylor <iant at google dot com>
- To: Christian GrÃssler <chris at groessler dot org>
- Cc: gcc at gcc dot gnu dot org
- Date: Wed, 09 Mar 2011 16:39:30 -0800
- Subject: Re: pattern problem with register assignment
- References: <4D65738B.9040501@groessler.org> <4D666D72.8090203@joguin.com> <4D66A32B.60001@groessler.org> <4D77F1FE.6010803@groessler.org>
Christian GrÃssler <chris@groessler.org> writes:
>>>> It seems that gcc wants to create an instruction like "p2=add(p1,p3)".
>>>> How can I tell him not to do that?
>>>> I tried to fiddle with the "p_d_general_operand" predication and use a
>>>> modified one for operand 2, but
>>>> at the time the constraint is called, I only see pseudo registers, and
>>>> don't know in which hard register
>>>> they will appear at the end.
I would guess that the addition instruction is being generated by
reload. Look at the RTL dumps generated by -da and see where the insn
first appears. If it is being created by reload then you need to find
out why. I'm not sure what to suggest beyond that because the answer is
likely to be highly processor specific.
Ian