Re: [Fortran-Experiments]: patch for error checking on binding labels and derived types, etc.

[Tobias Schlüter <tobias dot schlueter at physik dot uni-muenchen dot de>]

Steve Kargl wrote:
> On Tue, Feb 27, 2007 at 01:53:34AM +0100, Tobias Schl?ter wrote:
> > Christopher D. Rickett wrote:
> > > ChangeLog entry:
> > >
> > > 2007-02-26  Christopher D. Rickett  <crickett@lanl.gov>
> > >    * gcc/testsuite/gfortran.dg/binding_label_tests_10_main.f03: New
> > >    test case to make sure the compiler catches binding label
> > >    collisions across separate files when they're use associated into
> > >    one namespace.
> > There's no need to spell out the whole path to the testcase.  The 
> > ChangeLog entries for testcases goes into gcc/testsuite/ChangeLog, only 
> > a relative path is needed.  "New test" suffices as a description, if you 
> > want to be verbose a comment inside the testcase is much more useful.
>
> While this is generally true, on the fortran-experiments branch
> we've been keeping all ChangeLog entries in ChangeLog.iso_c_binding.
> We'll cut a patch against trunk that includes everything and 
> at that time proper ChangeLog entries will be generated.  This 
> megapatch will be submitted for the usually review on both fortran@
> and gcc-patches@.

That makes a lot of sense for a branch :-)  In order to save some typing 
during the generation of the patches and during the final submission, it 
might still make sense to format the ChangeLogs to something like

testsuite/
         * gfortran.dg/...
         * gfortran.dg/...
fortran/
         * decl.c: ...
         * st.c: ...

On future branches it might make more sense to have a separate 
testsuite/ChangeLog.branch_name.

Cheers,
- Tobi