This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: fix gimplification of compound lvalues
- From: Arnaud Charlet <charlet at ACT-Europe dot FR>
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: rth at twiddle dot net, gcc-patches at gcc dot gnu dot org
- Date: Wed, 21 Jul 2004 11:14:52 +0200
- Subject: Re: fix gimplification of compound lvalues
- References: <10407202333.AA08567@vlsi1.ultra.nyu.edu>
> I just did a good Ada bootstrap a bit ago, but that was with a tree about
> five days old and on x86-64. I have to resolve the confict in my tree
> with your change to tree-ssa-operands.c and then I'll try again with the
> current tree and see how far I get.
OK, forget my previous message about failure using stage1 compiler, I
hadn't seen richard henderson check-in/email.
I'm rebuilding right now on x86-linux, I'll report my results/findings,
if any.
> The Ada tools don't build because of a change in the Ada part of the
> front end that involves variable-sized returns. I need to convert a
> function call into a procedure call and that's going to be a little work.
This can actually easily be worked around with a simple change in sem_util.adb:
*** sem_util.adb.old 2004-07-16 22:48:10.000000000 +0200
--- sem_util.adb 2004-07-21 11:13:52.000000000 +0200
***************
*** 5772,5780 ****
-- scope because the back end otherwise tries to allocate a
-- variable length temporary for the particular variant.
! if Opt.GCC_Version = 2
! and then Has_Discriminants (Typ)
! then
return True;
-- For GCC 3, or for a non-discriminated record in GCC 2, we are
--- 5772,5778 ----
-- scope because the back end otherwise tries to allocate a
-- variable length temporary for the particular variant.
! if Has_Discriminants (Typ) then
return True;
-- For GCC 3, or for a non-discriminated record in GCC 2, we are