[Bug libfortran/110651] New: libgfortran.spec links twice with libgcc spec

[Bug libfortran/110651] New: libgfortran.spec links twice with libgcc spec

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

            Bug ID: 110651
           Summary: libgfortran.spec links twice with libgcc spec
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libfortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ro at gcc dot gnu.org
                CC: burnus at gcc dot gnu.org, fxcoudert at gcc dot gnu.org,
                    iains at gcc dot gnu.org
  Target Milestone: ---
              Host: x86_64-apple-darwin23.0.0
            Target: x86_64-apple-darwin23.0.0
             Build: x86_64-apple-darwin23.0.0

When bootstrapping current trunk on macOS 14.0 beta 3 with Xcode 15 beta 4,
every single fortran link test FAILs like

FAIL: gfortran.dg/c-interop/allocatable-dummy.f90   -O0  (test for excess
errors)
Excess errors: 
ld: warning: ignoring duplicate library '-lemutls_w'
ld: warning: ignoring duplicate library '-lgcc'

The link line ends in

-lgfortran -lemutls_w -lgcc -lquadmath -lemutls_w -lgcc -lSystem
-no_compact_unwind -idsym -dsym

I could trace this to libgfortran.spec, which has

%rename lib liborig
*lib: %{static-libquadmath:libquadmath.a%s;:-lquadmath}  %(libgcc) %(liborig)

The libgcc spec is included twice, once explicitly here and another time in
%(liborig).  I have no idea why this is done this way: this already was in 
the original patch that introduced libgfortran.spec.in:

commit 1ec601bf9fb0fbc39b3a6cb90450500f857adae8
Author: Francois-Xavier Coudert <fxcoudert@gcc.gnu.org>
Date:   Tue Nov 16 21:23:19 2010 +0000

    re PR fortran/32049 (Support on x86_64 also kind=16)

I see quite a number of possible solutions:

* Get rid of %(libgcc) in libgfortran.spec.in.

* Include it conditionally depending on a configure test.

* Disable ld warnings with -w in the spec, probably using some
@TARGET_LDFLAGS@.

* Disable ld warnings globally in the Darwin driver code.  That may be
  undisable since it would disable possibly benign warnings, too.

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |14.0

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

--- Comment #1 from Iain Sandoe <iains at gcc dot gnu.org> ---
(In reply to Rainer Orth from comment #0)
> When bootstrapping current trunk on macOS 14.0 beta 3 with Xcode 15 beta 4,
> every single fortran link test FAILs like

> * Get rid of %(libgcc) in libgfortran.spec.in.
> 
> * Include it conditionally depending on a configure test.

Hmm .. I thought we already had configure tests to customise the spec for
Darwin?
FX?

> * Disable ld warnings with -w in the spec, probably using some
> @TARGET_LDFLAGS@.
> 
> * Disable ld warnings globally in the Darwin driver code.  That may be
>   undisable since it would disable possibly benign warnings, too.

I think we need to fix the specs not work around with disabled warnings.

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

--- Comment #2 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #1 from Iain Sandoe <iains at gcc dot gnu.org> ---
> (In reply to Rainer Orth from comment #0)
>> When bootstrapping current trunk on macOS 14.0 beta 3 with Xcode 15 beta 4,
>> every single fortran link test FAILs like
>
>> * Get rid of %(libgcc) in libgfortran.spec.in.
>> 
>> * Include it conditionally depending on a configure test.
>
> Hmm .. I thought we already had configure tests to customise the spec for
> Darwin?
> FX?

We do: @LIBM@ is handled that way.

>> * Disable ld warnings with -w in the spec, probably using some
>> @TARGET_LDFLAGS@.
>> 
>> * Disable ld warnings globally in the Darwin driver code.  That may be
>>   undisable since it would disable possibly benign warnings, too.
>
> I think we need to fix the specs not work around with disabled warnings.

Agreed: for me modifying the spec was a quick hack to avoid half the
failures in the testsuite run.

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

--- Comment #3 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #2 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot
> Uni-Bielefeld.DE> ---
>> --- Comment #1 from Iain Sandoe <iains at gcc dot gnu.org> ---
>> (In reply to Rainer Orth from comment #0)
>>> When bootstrapping current trunk on macOS 14.0 beta 3 with Xcode 15 beta 4,
>>> every single fortran link test FAILs like
>>
>>> * Get rid of %(libgcc) in libgfortran.spec.in.
>>> 
>>> * Include it conditionally depending on a configure test.
>>
>> Hmm .. I thought we already had configure tests to customise the spec for
>> Darwin?
>> FX?
>
> We do: @LIBM@ is handled that way.

FWIW, I've now removed the %(libgcc) from libgfortran.spec.in locally:
on all of *-*-solaris2.11, x86_64-pc-unix-gnu, and
x86_64-apple-darwin23.0.0 there were no regressions.

It seems ever more important to understand why it was introduced in the
first place, before even considering adding it conditionally.

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[fxcoudert at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2023-12-19
                URL|                            |https://gcc.gnu.org/piperma
                   |                            |il/gcc-patches/2023-Decembe
                   |                            |r/640943.html
     Ever confirmed|0                           |1

--- Comment #4 from Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> ---
I introduced the original behaviour and I'm not sure I remember why. I've
propose to remove the duplication, because I cannot see a single case where it
could be useful. Patch posted at
https://gcc.gnu.org/pipermail/gcc-patches/2023-December/640943.html

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

--- Comment #5 from Iain Sandoe <iains at gcc dot gnu.org> ---
AFAICT, this was fixed on trunk by r14-6721-gd31c54c7da7661 (which seems to
have a reference to the PR so not sure why it did not show up here).

I think we need this on any open branch which we want to work with macOS14+.

[Bug libfortran/110651] libgfortran.spec links twice with libgcc spec

[fxcoudert at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110651

Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> changed:

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

--- Comment #6 from Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> ---
Didn't have any fallout reported, so let's go.

Backported to 13 as
https://gcc.gnu.org/git?p=gcc.git;a=commit;h=0ea5efea8f24e182b17c7f4f2e3923a2938a2de3

Backported to 12 as
https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=56071adae867e63248816f5879339eec27785383

Backported to 11 as
https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=407100bffa8c66e92fd4f6bfd2a083995b9dab93