weak for Darwin

[Aaron W. LaFramboise]

Aaron W. LaFramboise wrote:

> Geoffrey Keating wrote:
> 
>>Index: libstdc++-v3/ChangeLog
>>2004-10-25  Geoffrey Keating  <geoffk@apple.com>
>>
>>	* libsupc++/new_op.cc (new): Make weak.
>>	* libsupc++/new_opnt.cc (new): Make weak.
>>	* libsupc++/new_opv.cc (new): Make weak.
>>	* libsupc++/new_opvnt.cc (new): Make weak.
>>	* libsupc++/delete_op.cc (delete): Make weak.
>>	* libsupc++/delete_opnt.cc (delete): Make weak.
>>	* libsupc++/delete_opv.cc (delete): Make weak.
>>	* libsupc++/delete_opvnt.cc (delete): Make weak.
> 
> 
> Hi Geoffrey,
> 
> Why is it necessary to make these symbols weak?  It doesn't seem like it
> should be necessary, as prior definitions of these symbols will override
> them anyway, and there are no other references to these symbols within
> these translation units.

Well, sorry, I admit I didn't read your entire email before hitting the
send button. :-(

Do you have any ideas on how to fix the problems with non-ELF weak
symbols while preserving this load-time optimization for Darwin?

Aaron W. LaFramboise