This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: ia64 performance regression
- To: Richard Henderson <rth at redhat dot com>
- Subject: Re: ia64 performance regression
- From: law at redhat dot com
- Date: Wed, 17 Oct 2001 18:51:02 -0600
- cc: Steve Christiansen <smc at us dot ibm dot com>, gcc at gcc dot gnu dot org
- Reply-To: law at redhat dot com
In message <20011016172801.B5704@redhat.com>you write:
> On Tue, Oct 09, 2001 at 01:53:37PM -0700, Steve Christiansen wrote:
> > What is the purpose of the following code in loop.c (loop_regs_scan)?
> >
> > for (i = 0; i < FIRST_PSEUDO_REGISTER; i++)
> > {
> > regs->array[i].may_not_optimize = 1;
> > regs->array[i].set_in_loop = 1;
> > }
>
> I don't know. Paranoia, probably.
FWIW, this code has been around since 1990 or earlier -- I ran into this on
the PA a decade or so ago.
I was relatively inexperienced with the optimizers themselves, so rather than
hack loop.c, I twiddled the PA backend to not expose the PIC register on
symbolic address computations, loads, and stores.
jeff