Re: [PATCH] Bug in flag fconserve-stack

[Hariharan Sandanagobalane <hariharans at picochip dot com>]

Committed.

Hari

Richard Guenther wrote:
> On Thu, Jan 29, 2009 at 7:09 PM, Hariharan <hariharans@picochip.com> wrote:
> > Hi Jan,
> > I guess this probably is a copy-paste error in the conserve-stack option. As
> > it stands now, the flag works in exactly opposite ways. Setting
> > fno-conserve-stack seems to be the only way to conserve stack!
> >
> > I am attaching a patch with fix for this. Let me know if it is good to go.
>
> Ok if you make it
>
> > +Common Var(flag_conserve_stack) Init(0) Optimization
>
> and adjust the changelog comment to what you are doing, for example
>
> >        * common.opt (flag_conserve_stack): Initialize to zero.
>
> Thanks,
> Richard.
>
> > Cheers
> > Hari
> >
> > ChangeLog:
> >
> > 2009-01-29  Hariharan Sandanagobalane <hariharan@picochip.com>
> >
> >        * common.opt (flag_conserve_stack): This flag was doing exact
> >        opposite of what it should have. Fixed now.
> >
> >
> > Index: gcc/common.opt
> > ===================================================================
> > --- gcc/common.opt      (revision 143749)
> > +++ gcc/common.opt      (working copy)
> > @@ -386,7 +386,7 @@
> >  Do not put uninitialized globals in the common section
> >
> >  fconserve-stack
> > -Common Var(flag_conserve_stack,0) Optimization
> > +Common Var(flag_conserve_stack) Optimization
> >  Do not perform optimizations increasing noticeably stack usage
> >
> >  fcprop-registers