Resend message to [email protected] as there was
a Majordomo configuration error on my side.
----- Forwarded message from Sven Utcke <[EMAIL PROTECTED]> -----
Delivered-To: [EMAIL PROTECTED]
Delivered-To: [EMAIL PROTECTED]
Date: Thu, 6 Oct 2005 21:18:03 +0200
From: Sven Utcke <[EMAIL PROTECTED]>
To: Marcus Brinkmann <[EMAIL PROTECTED]>
Cc: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED]
Subject: Re: Bug in Configure for gpgme under Solaris
In-Reply-To: <[EMAIL PROTECTED]>
User-Agent: Mutt/1.4.1i
Hello Marcus,
> > Then that, too, is a pth-bug. Obviously it _ought_ to output the
> > correct values, or the flag would have been named --bogus instead of
> > --all :-)
>
> I am not going to second guess the intentions of the Pth authors. You
> will have to ask them.
>
> > > It is for this reason that I asked you to take this up with the Pth
> > > maintainers.
> >
> > Believe me, I tried :-)
>
> You will need to try harder :)
> > > However, things are probably much easier at least for GPGME, because
> > > we need "socket stuff" in assuan, and thus I think we want to test for
> > > it anyway.
> >
> > the same reasoning as above of course also applies to libassuan-config...
>
> Sure. I can take that into my hands. The only problem is I can't
> test on Solaris. Can you give out shell accounts for testing?
Yes. I will privately mail you the details.
> > Ok, simply calling autoconf seems to do the right thing.
>
> I think you will want to run automake, too. But I am not sure.
It seems to have worked.
> > Ok, so does adding AC_CHECK_LIB(socket, connect) help?
> >
> > Hmm. Adding just somewhere obviously can't help the fact that the
> > pth-sanity check will fail, so we need to add this in there
> > somewhere. Ok, can't add it in there, so I add it just before the
> > pth-sanity check. Does this work...? Yes, configure runs through.
> > What about a make? yepp, make and make check run nicely, and doing
> > the same on a Linux system shows no ill effects there either.
>
> Does it link GPGME against -lsocket?
yes:
kogs2>/software/gpgme-1.0.3/SunOS-5.8/objs% ldd ./gpgme/.libs/libgpgme.so |
grep socket
libsocket.so.1 => /usr/lib/libsocket.so.1
> gpgme-config also needs to print
> -lsocket.
This does not work:
kogs2>/software/gpgme-1.0.3/SunOS-5.8/objs% ./gpgme/gpgme-config --libs
-L/software/gpgme-1.0.3/SunOS-5.8/lib -lgpgme
-L/software/libgpg-error-1.1/SunOS-5.8/lib -lgpg-error
> > Ok, this seems to be a possible workaround.
>
> Well, if Assuan is included, this isn't even a work around, but we
> need to do it anyway.
Not in order to work around a bug in Assuan or libassuan-config, I
hope...
Thanks
Sven
Here is the original mail:
> > configure fails under solaris
> >
> > kogs1>/raid0/sw/gpgme-1.0.3% uname -a
> > SunOS kogs1 5.8 Generic_117350-27 sun4u sparc SUNW,Ultra-250
> >
> > when trying to link against pth. Please change
> >
> > _pth_cflags=`pth-config --cflags`
> > _pth_ldflags=`pth-config --ldflags`
> > _pth_libs=`pth-config --libs`
> >
> > to
> >
> > _pth_cflags=`pth-config --cflags`
> > _pth_ldflags=`pth-config --ldflags`
> > _pth_libs=`pth-config --libs --all`
> >
> > so that it will also link agains -lsocket...
>
> The file pth.m4 comes from the pth distribution.
Oh. Ok. That explains why I had this problem twice.
> If there is something wrong with it, you should contact them and
> resolve it with the pth maintainers. Looking at the output of
> --cflags --all and --libs --all under GNU/Linux, I have serious
> doubts that using --libs --all is the correct approach.
Well, the code in question tests whether it is possible to link
something against the libpth. "pth-config --libs --all" will give you
all libraries needed to sucessfully link a program against libpth, as
opposed to "pth-config --libs", which essentially just tells you what
the name of libpth is...
> Maybe a check for the socket library should be added to GPGME.
I don't think this is the way to go --- after all, there might be
systems out there which contain a libsocket but don't need it for
linking against libpth, while other systems may need other libs. And
it is after all just shear dump luck that the test in question doesn't
also require libnsl...
--
_ __ The Cognitive Systems Group
| |/ /___ __ _ ___ University of Hamburg
| ' </ _ \/ _` (_-< phone: +49 (0)40 42883-2576 Vogt-Koelln-Strasse 30
|_|\_\___/\__, /__/ fax : +49 (0)40 42883-2572 D-22527 Hamburg
|___/ http://kogs-www.informatik.uni-hamburg.de/~utcke/home.html
----- End forwarded message -----
Ralf S. Engelschall
[EMAIL PROTECTED]
www.engelschall.com
______________________________________________________________________
GNU Portable Threads (Pth) http://www.gnu.org/software/pth/
Development Site http://www.ossp.org/pkg/lib/pth/
Distribution Files ftp://ftp.gnu.org/gnu/pth/
Distribution Snapshots ftp://ftp.ossp.org/pkg/lib/pth/
User Support Mailing List [email protected]
Automated List Manager (Majordomo) [EMAIL PROTECTED]
