This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC 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, Fortan] Depricate -M<dir> in favour of -J<dir>


On Wed, 2008-01-30 at 20:23 +0100, Daniel Franke wrote:
> On Wednesday 30 January 2008 20:16:31 Thomas Koenig wrote:
> > On Tue, 2008-01-29 at 21:09 +0100, Tobias Burnus wrote:
> > > Build and regtested on x86-64-linux.
> > > OK for the trunk (4.3.0)?
> >
> > Because we're in regression-only mode, I think it would be
> > better to wait for 4.4.
> 
> That's the point. If -M is deprecated in 4.4 only, we can re-use it for 
> Makefile-dependency-generation in 4.5 at the earliest?!

> To phrase it differently: are we allowed to change the meaning of an option 
> from one release to another, i.e. from 4.3 to 4.4, without deprecating its 
> usage in the earlier meaning first?

We can always document the obsoletion in the manual.  I think bringing
the options into line with gcc should be an OK change.

BTW, just a thought... suppose foo.F90 contains

program main
  use foo
#include "bar.inc"
end program main

what should the output be?  Should the "gfortran -M" output "foo.mod",
"bar.inc", or both?  -M is a preprocessor right now...

	Thomas


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