Re: rs6000 struct returns

[Michael Meissner <meissner at cygnus dot com>]

On Mon, Mar 19, 2001 at 09:58:24PM -0500, David Edelsohn wrote:
> >>>>> Michael Meissner writes:
> 
> Michael> Yeah, its been a bug for 5 years ago, and as Geoff mentions, fixing it will
> Michael> break linking objects compiled with previous compilers (particularly the dbm_*
> Michael> functions under Linux).  As I recall, structure returns is the one thing that
> Michael> is different on Solaris than System V (of course PowerPC Solaris is long since
> Michael> dead).
> 
> 	The PowerPC Solaris and PowerPC SVR4 ABIs are identical.  Maybe
> this was a late change -- I don't know.  Solaris and SVR4 implementations
> should not have had different behaviors.

Yes I know that, but bugs do happen.  At one time I believe that I had Solaris
returning structures in the registers, and System V doing it the way GCC
currently does it.  I think I did it because I wasn't sure which was the
official ABI and I was attempting to match the behavior of the actual Solaris
compiler instead of an ABI, as I only had drafts, and sometimes I either
misread them or they conflicted.  In the current sources and 2.95.2 have V.4
and Solaris only differing in some minor assembly differences & how MI thunks
are done on 2.95.2.  Gcc 2.7.2.3 doesn't have Solaris support in it, and I
don't have any other early GCC's, such as egcs-1.1.2 on this system.

-- 
Michael Meissner, Red Hat, Inc.  (GCC group)
PMB 198, 174 Littleton Road #3, Westford, Massachusetts 01886, USA
Work:	  meissner@redhat.com		phone: +1 978-486-9304
Non-work: meissner@spectacle-pond.org	fax:   +1 978-692-4482