[Bug fortran/15548] New: enable-languages broken wrt libgfortran

[Bug fortran/15548] New: enable-languages broken wrt libgfortran

[green at redhat dot com]

When I configure with --enable-languages=c,c++,java, it still tries to build
libgfortran (and fails, since there's no gfortran compiler).

Specifically, the build failure is:
/home/green/build/gcj/gcc/gfortran -B/home/green/build/gcj/gcc/
-B/home/green/install/jhbuild-inst/i686-pc-linux-gnu/bin/
-B/home/green/install/jhbuild-inst/i686-pc-linux-gnu/lib/ -isystem
/home/green/install/jhbuild-inst/i686-pc-linux-gnu/include -isystem
/home/green/install/jhbuild-inst/i686-pc-linux-gnu/sys-include -Wall
-fno-repack-arrays -fno-underscoring -c
/home/green/sources/jhbuild-src/gcc/libgfortran/intrinsics/selected_kind.f90 
-fPIC -DPIC -o .libs/selected_kind.o./libtool: line 1:
/home/green/build/gcj/gcc/gfortran: No such file or directory
make[2]: *** [selected_kind.lo] Error 1


And, I'm configuring like so:
/home/green/sources/jhbuild-src/gcc/configure  
--prefix=/home/green/install/jhbuild-inst
--bindir=/home/green/install/jhbuild-inst/gcj-bin --enable-java-awt=gtk
--enable-gtk-cairo --enable-libgcj-multifile --disable-static --disable-gtk-doc
--with-gcc-version-trigger=/home/green/sources/jhbuild-src/gcc/gcc/version.c
--enable-languages=c,c++,java

-- 
           Summary: enable-languages broken wrt libgfortran
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: green at redhat dot com
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug fortran/15548] enable-languages broken wrt libgfortran

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-05-23 16:23 -------
I could not reproduce this with "--prefix=${HOME}/fsf-clean-nocheck_c --disable-checking --enable-
languages=c,c++".

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug fortran/15548] enable-languages broken wrt libgfortran

[green at redhat dot com]

------- Additional Comments From green at redhat dot com  2004-05-24 14:22 -------
I think this has something to do with how I checked out the code.

I get this problem when just check out gcc-core gcc-java gcc-g++ and
gcc/gcc/fixinc, but not when I add gcc/gcc/fortran.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug other/15548] [3.5 Regression] enable-languages broken wrt libgfortran

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-05-24 14:34 -------
Confirmed, this is the way the modules are.  The problem is that libgfortran should be included in gcc-
fortran and not in gcc-core.  Someone needs to change this.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
          Component|fortran                     |other
     Ever Confirmed|                            |1
           Keywords|                            |build
   Last reconfirmed|0000-00-00 00:00:00         |2004-05-24 14:34:44
               date|                            |
            Summary|enable-languages broken wrt |[3.5 Regression] enable-
                   |libgfortran                 |languages broken wrt
                   |                            |libgfortran
   Target Milestone|---                         |3.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug other/15548] [3.5 Regression] libgfortran in wrong CVS module, breaks build

[tobi at gcc dot gnu dot org]

------- Additional Comments From tobi at gcc dot gnu dot org  2004-06-09 17:48 -------
(In reply to comment #3)
> Confirmed, this is the way the modules are.  The problem is that libgfortran
should be included in gcc-
> fortran and not in gcc-core.  Someone needs to change this.

Who can do this? Changed the summary to attract more attention :-)



-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[3.5 Regression] enable-    |[3.5 Regression] libgfortran
                   |languages broken wrt        |in wrong CVS module, breaks
                   |libgfortran                 |build


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug other/15548] [3.5 Regression] libgfortran in wrong CVS module, breaks build

[wilson at gcc dot gnu dot org]

------- Additional Comments From wilson at gcc dot gnu dot org  2004-08-18 03:52 -------
Proposed patch at
    http://gcc.gnu.org/ml/gcc-patches/2004-08/msg01301.html

The reason we have a problem here is because of an undocumented feature of cvs.
 It uses strncmp (test_dir, ignore_dir, strlen (ignore_dir)), thus putting
!gcc/f in the gcc-core module means that we also lose gcc/fortran even though it
wasn't listed.  Maybe they meant to use strlen()+1?  Maybe I should report this
as a bug?  Anyways, this isn't a serious problem after adding the fortran
support, as we wanted to lose gcc/fortran anyways.

However, I just realized that we are probably also losing gcc/fixinc, and I
forgot to check for that.  You probably wouldn't notice if you were on a linux
machine.  This could be fixed by explicitly adding this dir after the !gcc/f. 
Checking...yes I see it is lost from the gcc-core module.  Oh well.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug other/15548] [3.5 Regression] libgfortran in wrong CVS module, breaks build

[wilson at gcc dot gnu dot org]

------- Additional Comments From wilson at gcc dot gnu dot org  2004-08-21 03:46 -------
The patch is checked in, and the cvs bug has been reported to bug-cvs. 
gcc/fixinc is still broken, I forgot to try fixing that, but I really think that
is a cvs bug, so I will wait for a response to that.  I believe this is fixed
now.  I didn't try a build, but the list of directories checked out with
gcc-core is correct now.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |wilson at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug other/15548] libgfortran in wrong CVS module, breaks build

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-10-11 12:50 -------
gcc/fixinc does not exist on the mainline any more (as it was moved to the toplevel) so this is not a 
mainline problem, only a CVS problem so I am removing the target milestone and the regression 
marker in the summary.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[4.0 Regression] libgfortran|libgfortran in wrong CVS
                   |in wrong CVS module, breaks |module, breaks build
                   |build                       |
   Target Milestone|4.0.0                       |---


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548

[Bug other/15548] libgfortran in wrong CVS module, breaks build

[wilson at gcc dot gnu dot org]

------- Additional Comments From wilson at gcc dot gnu dot org  2004-10-12 01:18 -------
My CVS patch was accepted and installed into the master CVS sources.  We can
patch the local cvs on gcc.gnu.org if necessary, but I don't see any real need
for that, given that gcc/fixinc is now gone.  There are no longer any problems
using modules to check out mainline.  Using modules to check out old gcc
versions may fail, but I don't know of anyone doing that, and fixing that is
problematic anyways as CVS doesn't version the modules file.


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15548