other/8888: Linking shared libraries with -pthread fails to link to libpthread

[Sebastian Wilhelmi]

The following reply was made to PR other/8888; it has been noted by GNATS.

From: Sebastian Wilhelmi <wilhelmi@ira.uka.de>
To: bangerth@dealii.org, gcc-bugs@gcc.gnu.org, gcc-prs@gcc.gnu.org,  gcc-gnats@gcc.gnu.org
Cc:  
Subject: Re: other/8888: Linking shared libraries with -pthread fails to
	link to libpthread
Date: 10 Dec 2002 15:31:02 +0100

 Hi,
 
 > Synopsis: Linking shared libraries with -pthread fails to link to libpthread
 > 
 > State-Changed-From-To: open->feedback
 > State-Changed-By: bangerth
 > State-Changed-When: Tue Dec 10 06:13:52 2002
 > State-Changed-Why:
 >     I fail to see the reasoning behind the report: what is wrong
 >     with passing -lpthread to the linker line? The -pthread
 >     changes some things when compiling (as far as I understand,
 >     for example the initialization of static variables), so
 >     has nothing to do with what libraries are linked. You 
 >     would want to specify this separately. Why is this bad?
 > 
 > http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=8888
 
 the man page for gcc-3.2 says:
 
        -pthread
            Adds support for multithreading with the pthreads library.
            This option sets flags for both the preprocessor and linker.
 
 So we consider -pthread in GLib the preferred way to indicate, that a
 program or library is multithreaded. We use it both for linking and
 compiling programs and libraries. I got the impression, that also gcc
 consideres using -pthread superior to using -D_REENTRANT and -lpthread.
 
 Also for non-shared binaries gcc does the right thing now, but not for
 shared objects, why is that?
 
 Bye,
 Sebastian
 
 P.S.: I have no write access to GNATS, so I could only write this
 e-mail, or is there some way to attach that message to the bug report?
 -- 
 Sebastian Wilhelmi
 mailto:wilhelmi@ira.uka.de
 http://goethe.ira.uka.de/~wilhelmi