Re: RFC: #pragma optimization_level

[Mark Mitchell <mark at codesourcery dot com>]

Richard Guenther wrote:
> On Apr 1, 2005 9:36 PM, Mark Mitchell <mark@codesourcery.com> wrote:
>
> > Diego Novillo wrote:
> >
> > > On Fri, Apr 01, 2005 at 11:24:06AM -0800, Mark Mitchell wrote:
> > >
> > >
> > > > Dale Johannesen wrote:
> > > >
> > > >
> > > >
> > > > > So I guess question 1 is, Mark, do you feel negatively enough about this
> > > > > feature to block its acceptance in mainline?
> > > >
> > > > I'm not sure that I *could* block it, but, no, I don't feel that negatively.
> > >
> > > I don't mind either way.  But I do have a question, what's the
> > > granularity of this #pragma?  Function-level, I hope?
> >
> > That's what I assumed.  Anything finer than that is insane. :-)
>
>
> Well, if we're inventing something that can annotate functions we may as
> well invent it in a way that it can be extended to handle statement
> annotations.  Just like for
>
>    for (;;)
> #pragma inline
>       foo();

The concept I was calling insane was trying to optimize the first half 
of a function with -O2 and the second half with -O0.

Your example is different; you're talking about how to treat the call to 
"foo", not how to optimize the entire function (including the "for"). 
Something like what you're suggesting doesn't seem entirely unreasonable 
to me.

--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com
(916) 791-8304