This is the mail archive of the
mailing list for the binutils project.
Re: Another MIPS multigot patch
Daniel Jacobowitz <firstname.lastname@example.org> writes:
> On Sun, Feb 15, 2004 at 10:01:25PM +0000, Richard Sandiford wrote:
>> I think the problem is simply that check_relocs() was only checking
>> whether the symbol was defined, not whether the symbol was defined
>> by a regular or shared object.
> I will do this and let you know if the build works.
> Do you expect the proposed assertion to succeed now?
Yeah, I just wasn't sure whether it was worth keeping in the final version.
If you think it is, I suggest we move it outside the switch statement.
It isn't really specific to GOT_PAGE relocations.
>> I've attached a shell archive (to be run from the top level of the
>> build directory). Before the patch, a.x needlessly had a global GOT
>> entry for gs. After the patch, both a.x and b.x only have a global
>> GOT entry for "us".
> Interested in adding this to the testsuite? :)