public inbox for fortran@gcc.gnu.org
 help / color / mirror / Atom feed
From: Iain Sandoe <idsandoe@googlemail.com>
To: feenberg@nber.org
Cc: Fortran List <fortran@gcc.gnu.org>
Subject: Re: gfortran static linking under OS X (was: Re: )
Date: Sat, 20 Mar 2021 21:07:05 +0000	[thread overview]
Message-ID: <FB19ECDC-4158-4310-8C4E-828A3EFA0233@googlemail.com> (raw)
In-Reply-To: <CAFOAwsoFaKCiaqKtnDNdrJOvBfazmqQKNC=ZtsEzAc88AgO51A@mail.gmail.com>

Daniel Feenberg <feenberg@gmail.com> wrote:

> On 3/19/21, Iain Sandoe <idsandoe@googlemail.com> wrote:
>
>>> On Fri, 19 Mar 2021, Tobias Burnus wrote:
>
>> you should be able to work around this without changing the compiler or
>> rebuilding it,

>> %rename lib liborig
>> *lib:  -lquadmath -lm %(libgcc) %(liborig)
>>
>> change the last line line to :
>>
>> *lib: %{!static-libgfortran: -lquadmath } %{static-libgfortran:
>> libquadmath.a%s} %(libgcc) %(liborig)

^^ this line is presumably not broken in your file - i.e. this is a mailer
phenomenon?

>> ===
>>
>> and try your link again
>
> Tried that, the error message is:
>
>   ld: library not found for -lSystem

Please could you either file a PR or send the output of

/your complete command line/ with ‘-save-temps -v’ appended

to me off-list.

----

For the record, this was not completely theoretical on my part; I actually  
tried this on a macOS 11.2.3 system with gcc-6.5 (which is the closest I  
have to what you’re using)

…./gcc-6-5/bin/gfortran /source/test/fortran/hello.f -static-libgfortran  
-static-libgcc -o h

$  otool -Lv h
h:
	/usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 1292.60.1)
	time stamp 2 Thu Jan  1 01:00:02 1970

This ^^ is what I expect to see;
libgfortran, libquadmath and libgcc have all been linked statically,  
leaving only
the "/usr/lib/libSystem.B.dylib” which is fine and expected.

> Indeed there is no libsystem.a anywhere on the computer.

macOS does not allow the equivalent of a “completely statically linked”  
user-mode
executable.  Exes are always started by “dyld” (dynamic linker) and they  
always refer
to libSystem.dylib.

So you should never be needing to find a libSystem.a (there isn’t one, as  
you found).
Nor do you need to care about it - since it’s available to all your  
end-users as part
of their system (as are things like the Accelerate framework).

> Any ideas?

I am somewhat surprised that 6.3 behaves any differently from 6.5 - let’s  
investigate a little more.

cheers
Iain


  reply	other threads:[~2021-03-20 21:07 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-19 20:22 Daniel Feenberg
2021-03-19 20:49 ` gfortran static linking under OS X (was: Re: ) Tobias Burnus
2021-03-19 21:06   ` Iain Sandoe
2021-03-19 22:11   ` Daniel Feenberg
2021-03-19 22:39     ` Iain Sandoe
2021-03-20 20:47       ` Daniel Feenberg
2021-03-20 21:07         ` Iain Sandoe [this message]
2021-03-22 22:20           ` Daniel Feenberg

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=FB19ECDC-4158-4310-8C4E-828A3EFA0233@googlemail.com \
    --to=idsandoe@googlemail.com \
    --cc=feenberg@nber.org \
    --cc=fortran@gcc.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).