This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch] e500 port
- From: Aldy Hernandez <aldyh at redhat dot com>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Wed, 24 Jul 2002 13:40:22 -0700
- Subject: Re: [patch] e500 port
- References: <aldyh@redhat.com> <200207242011.QAA24420@makai.watson.ibm.com>
On Wed, Jul 24, 2002 at 04:11:40PM -0400, David Edelsohn wrote:
> >>>>> Aldy Hernandez writes:
>
> Aldy> Not a problem. We could do the same thing as we've done for
> Aldy> DEFAULT_ABI. Declare rs6000_fprs to be an enum. #define TARGET_FPRS
> Aldy> to a known value in the individual system .h files. And define
> Aldy> TARGET_FPRS to rs6000_fprs if the macro is not defined.
>
> Aldy> I can do this if you want.
>
> I don't think that TARGET_SPRS needs to be an enum, just 0 (FALSE)
> or rs6000_fprs.
Ok
> Then it seems that rs6000.h should define
>
> + #define TARGET_SPE_ABI 0
> + #define TARGET_SPE 0
> + #define TARGET_ISEL 0
> + #define TARGET_FPRS 0
>
> and only rs6000/eabi.h and rs6000/eabisim.h should define
>
> + #define TARGET_SPE_ABI rs6000_spe_abi
> + #define TARGET_SPE (rs6000_cpu == PROCESSOR_PPC8540)
> + #define TARGET_ISEL rs6000_isel
> + #define TARGET_FPRS rs6000_fprs
Fair enough.
> Is there a reason that you want to define TARGET_8540 and only use
> it to set TARGET_SPE instead of defining TARGET_SPE directly?
Nope. Originally I thought TARGET_8540 could be use for other things
(scheduling, perhaps?). But we can get rid of it.
Aldy