GCC not (optimally) exploiting restrict pointers?

Dave Korn dave.korn@artimi.com
Wed Dec 15 18:01:00 GMT 2004


> -----Original Message-----
> From: Joe Buck 
> Sent: 15 December 2004 17:55

> On Wed, Dec 15, 2004 at 05:10:14PM -0000, Dave Korn wrote:
> > > From: gcc-owner On Behalf Of Richard Guenther
> > > So, the value in *a has to be accessed through a exclusively and
> > > cannot be accessed through b, regardless of any type qualifiers
> > > present on b or not.
> > 
> > .... unless of course b was derived from a in the function 
> that calls foo(),
> > perhaps?
> 
> Nope.  The whole point of restrict is that you're making a promise to the
> compiler that an optimization is safe.  If you lie, you lose.
> 

  It's pretty hard to tell from the wording of the standard exactly what that
promise is that is being asked of you.  I thought that the wording allowed for
that case, but the post people have quoted make the terminology a lot clearer.


    cheers, 
      DaveK
-- 
Can't think of a witty .sigline today....



More information about the Gcc mailing list