[Bug libfortran/47938] New: libgfortran symbol version node bumped unnecessarily

[Bug libfortran/47938] New: libgfortran symbol version node bumped unnecessarily

[jb at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47938

           Summary: libgfortran symbol version node bumped unnecessarily
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libfortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: jb@gcc.gnu.org


As pointed out by Janus Weil at

http://gcc.gnu.org/ml/fortran/2011-02/msg00298.html

in 4.5 we have symbol version 1.2. The new symbols for 4.6 should thus be in
version node 1.3. However, for some reason 4.6 has added both 1.3 and 1.4,
which is unnecessary. We should remove the 1.4 node and merge the symbols into
the 1.3 node.

OTOH this change would mean that people would need to recompile code compiled
with the pre-release 4.6.

[Bug libfortran/47938] libgfortran symbol version node bumped unnecessarily

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47938

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-01 10:10:12 UTC ---
I'd say this is too late to change it now, the benefit isn't that big and the
pain that would be caused by it is large.

[Bug libfortran/47938] libgfortran symbol version node bumped unnecessarily

[jb at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47938

Janne Blomqvist <jb at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |WONTFIX

--- Comment #2 from Janne Blomqvist <jb at gcc dot gnu.org> 2011-03-01 11:50:33 UTC ---
Yes, I'm leaning that way as well. FWIW, from looking at commit logs and old
mails, it seems 1.4 was added due to some confusion with symbol versioning and
ABI compatibility. 

In any case, closing as wontfix. If anyone comes up with a convincing argument
why this should be fixed, please reopen (and, be quick about it as 4.6 is going
to be released quite soon I think, and after that we certainly can't remove a
symbol node in a released branch).