This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PING] Interprocedural Constant Propagation for mainline
- From: Razya Ladelsky <RAZYA at il dot ibm dot com>
- To: Diego Novillo <dnovillo at redhat dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Mon, 11 Jul 2005 15:16:16 +0300
- Subject: Re: [PING] Interprocedural Constant Propagation for mainline
Diego Novillo <dnovillo@redhat.com> wrote on 11/07/2005 15:02:44:
> On Thu, Jul 07, 2005 at 11:35:25AM +0300, Razya Ladelsky wrote:
>
> > Therefore, I prevent adding assignment statements if there's an asm
call.
> >
> You were probably papering over the real bug. You'll need to dig
> up the test case and see what's going on. There is no logical
> reason why we can't do IPCP on functions with asm instructions.
>
> Remember if this was preventing bootstraps or was it just this
> program that was failing?
>
>
Diego Novillo <dnovillo@redhat.com> wrote on 11/07/2005 15:02:44:
> On Thu, Jul 07, 2005 at 11:35:25AM +0300, Razya Ladelsky wrote:
>
> > Therefore, I prevent adding assignment statements if there's an asm
call.
> >
> You were probably papering over the real bug. You'll need to dig
> up the test case and see what's going on. There is no logical
> reason why we can't do IPCP on functions with asm instructions.
>
> Remember if this was preventing bootstraps or was it just this
> program that was failing?
>
>
This was a regression test.
I think I have the testcase, testsuite/gcc.dg/i386-local.c, but
I can't reproduce the problem :)
So I guess it was a problem appearing only in one of the old development
versions of gcc.
I'll remove the restriction regarding asm from IPCP code.
Thanks,
Razya
> Diego.