weak symbol testcase

[Ulf Carlsson]

On Fri, 5 May 2000, Koundinya.K wrote:

> Hi Ulf,
>   In connection to gas handling weak symbols for mips, as we know we have 
> got this working correctly from the patches we have cooked up and have been 
> using. Though they address the same issue the patches are different. I think 
> if we can work out the patch on tc-mips.c and get it checked in by someone 
> it will be better since this is a common problem for all mips targets.. This 
> is a serious one and affects g++ primarily since for linkonce sections 
> handling of weak symbols correctly is very important.

I was just about to send a patch.  I'm just recompiling glibc to check whether
this patch really is enough for glibc.  I think they've changed some things in
the generic gas code.  It looks like the weak symbol handling in the patch you
sent actually is enough to fix the problems I have.  I could drop all the weak
symbol changes I had except for the one we had in common and it still works.
I haven't investigated linkonce.

Ulf