This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
ICE-on-valid-code, rejects-valid and wrong-code
- From: Paul Thomas <paulthomas2 at wanadoo dot fr>
- To: Fortran List <fortran at gcc dot gnu dot org>
- Date: Sun, 28 Jan 2007 21:16:03 +0100
- Subject: ICE-on-valid-code, rejects-valid and wrong-code
Dear All,
I note that we still have 29 such bugs of which:
14 are assigned;
7 are 4.2 and 4.1 only.
Whilst looking onward to F2003 compatibility, I might remind you that
our ambition, at present, is an F95 compiler. How about having a go at
some of the 15 unassigned bugs in this category and to clearing those
which are 4.2 and 4.1 only?
Steve was challenged on comp.lang.fortran, after his end-of-year
progress report about F95 features that are missing from gfortran.
Whilst the correspondence was rather irritating in its tone, under the
circumstances, there was serious intent there too. I know of two
missing features, maybe we can build up a complete list?
(i) Array TRANSFER does not work in initialization expressions(PR18769); and
(ii) Array pointers to components of derived type arrays do not work. ie.
type :: a
real :: r
integer :: i
end type a
type(a), target :: dt(10)
integer, pointer :: ip(:)
ip => dt%i
does not work correctly.
I have just submitted PR30625 for the latter, which also adds one to the
list above!.
Best regards
Paul