This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: patch ping


On Tue, Mar 31, 2009 at 10:03:02PM +0200, Daniel Franke wrote:
> On Tuesday 31 March 2009 06:23:04 Steve Kargl wrote:
> 
> Steve, thanks for the reviews.
> 
> > On Mon, Mar 30, 2009 at 07:49:13PM +0200, Daniel Franke wrote:
> > > http://gcc.gnu.org/ml/fortran/2009-03/msg00127.html (cshift, eoshift)
> >
> > This one is OK with the following comment.  You have
> >
> > +	  char buffer[80];
> > +	  snprintf (buffer, 80, "arguments '%s' and '%s' for intrinsic %s",
> > +		    gfc_current_intrinsic_arg[1], gfc_current_intrinsic_arg[2],
> > +		    gfc_current_intrinsic);
> > +	  if (gfc_check_conformance (buffer, shift, boundary) == FAILURE)
> >
> > I'm wondering if gfortran should dynamic construct the string
> > instead of using an 80 character static buffer.  A symbol in
> > F2003 can be 63 characters.
> 
> I found this in four other places (minmaxloc, reduction, pack, check_rest). 
> Increasing the respective buffers to a bigger value would do for now?
> 

I'm not sure we want at least five larger buffers.  Perhaps,
the 5 functions can share a static buffer of 150 bytes.

-- 
Steve


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]