Mailing-List: contact stdcxx-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: stdcxx-dev@incubator.apache.org
Received-SPF: pass (herse.apache.org: local policy)
Message-ID: <45F74477.5080506@roguewave.com>
Date: Tue, 13 Mar 2007 17:40:23 -0700
From: Martin Sebor <sebor@roguewave.com>
Organization: Rogue Wave Software
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
 rv:1.8.1.2) Gecko/20070221 SeaMonkey/1.1.1
MIME-Version: 1.0
To: stdcxx-dev@incubator.apache.org
Subject: Re: [PATCH] for STDCXX-262 "Multiple vtable definitions on Darwin
 platforms"
References: <D730FF7CEDDCA64483F9E99D999A158BEF016C@qxvcexch01.ad.quovadx.com>
 <45F70AB5.6030702@roguewave.com>
 <D730FF7CEDDCA64483F9E99D999A158BEF02FA@qxvcexch01.ad.quovadx.com>
In-Reply-To: 
 <D730FF7CEDDCA64483F9E99D999A158BEF02FA@qxvcexch01.ad.quovadx.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Eric Lemings wrote:
> If LDSOFLAGS is only used when linking the library then I don't see
> a problem with that since -dynamiclib is only need when linking the
> library.  I'm not sure what will happen if that flag is given when
> linking executables.

LDSOFLAGS is only used when building a shared library.
I suppose the name could be improved :)

> 
> Now that I think about it, the -dynamiclib flag probably needs to be
> within the build mode conditionals that check see if a shared
> library is being built.  Not sure what will happen if that flag is
> specified when building a static library.

Right, it needs to be used only for shared library builds.

Martin

> 
> I guess you could define yet another platform-specific variable in
> gcc.config that gets added to LDFLAGS only when linking the library
> (if one doesn't already exist of course).
> 
> Brad.
> 
>> -----Original Message-----
>> From: Martin Sebor [mailto:sebor@roguewave.com] 
>> Sent: Tuesday, March 13, 2007 2:34 PM
>> To: stdcxx-dev@incubator.apache.org
>> Subject: Re: [PATCH] for STDCXX-262 "Multiple vtable 
>> definitions on Darwin platforms"
>>
>> Eric Lemings wrote:
>>> "[gcc/Mac OS X 10.4.6 Tiger] Multiple vtable definitions for 
>>> bad_exception and bad_alloc"
>> Thanks Brad! Good to know there is an easy solution for this.
>>
>> As for the patch, my comment is the same as what I just said 
>> to Farid, i.e., that (like stdcxx C++ sources) the makefiles 
>> are intended to be platform-agnostic and free of platform 
>> specific ifs (and ifdefs).
>>
>> Fortunately, unlike in Farid's case, the solution here should 
>> be straightforward: add the -dynamiclib option to LDSOFLAGS 
>> in gcc.config instead of GNUmakefile.lib.
>>
>> If you post a (tested) patch that does that I'll be happy to 
>> commit it on trunk.
>>
>> Thanks
>> Martin
>>
>>> $ svn diff etc/config/GNUmakefile.lib
>>> Index: etc/config/GNUmakefile.lib
>>> ===================================================================
>>> --- etc/config/GNUmakefile.lib  (revision 517771)
>>> +++ etc/config/GNUmakefile.lib  (working copy)
>>> @@ -39,6 +39,10 @@
>>>  ONE_REPOSITORY = 1
>>>  include ../makefile.common
>>>  
>>> +ifeq ($(findstring darwin,$(PLATFORM)),darwin)
>>> +LDSOFLAGS   += -dynamiclib
>>> +endif
>>> +
>>>  CXXFLAGS    += $(PICFLAGS)
>>>  LDFLAGS     += $(LDSOFLAGS) $(MAPFLAGS) $(MAPFILE)
>>>  
>>>
>>> Tested with gcc-4.2-20070307 snapshot on Mac OS X 10.4.8.
>>>  
>>> Eric.
>>