This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] Implement x86 SIZE32/SIZE64 relocations
On 01/16/2013 03:42 PM, David Miller wrote:
> From: Roland McGrath <roland@hack.frob.com>
> Date: Wed, 16 Jan 2013 12:33:21 -0800 (PST)
>
>> If it's so well settled, then you can commit the changes to binutils and
>> report here that they are in binutils already when you propose your patch.
>
> I strongly disagree with this ordering requirement.
>
> The whole world is not binutils, there are many tool sets that might
> want to emit these ABI standard defined relocations. The most
> important dependency is the dynamic linker, and it should support
> all cpu ABI defined relocations regardless of whatever binutils can
> emit it or not.
Agreed.
The ABI standard is the arbiter of the relocations.
Failing that, yes, binutils is the arbiter of the relocations.
Once an ABI defines the relocations I have no problem
including them in glibc.
Even if the binutils work exposes a problem with the relocations
there isn't anything that uses them yet so you can just change
their meaning in the dynamic linker anyway?
I see no clear reason to wait for binutils to implement a relocation.
Cheers,
Carlos.