This is the mail archive of the
mailing list for the binutils project.
Re: [RFC PATCH] bfd/elf32-arm.c: Set st_value to zero for undefined symbols.
- From: Will Newton <will dot newton at linaro dot org>
- To: Will Newton <will dot newton at linaro dot org>, "binutils at sourceware dot org" <binutils at sourceware dot org>
- Cc: Richard Earnshaw <richard dot earnshaw at arm dot com>
- Date: Mon, 27 Jan 2014 12:00:34 +0000
- Subject: Re: [RFC PATCH] bfd/elf32-arm.c: Set st_value to zero for undefined symbols.
- Authentication-results: sourceware.org; auth=none
- References: <1389868125-7349-1-git-send-email-will dot newton at linaro dot org> <CANu=DmhOVH1FEMwsNa8qKrSMet2z+7CTpwzLHHnasYR6jtHy+w at mail dot gmail dot com> <20140124041743 dot GE5390 at bubble dot grove dot modra dot org> <CANu=Dmj5hoWdvu-ded9qi8L_1gWq_c5DFL4usKdb6+25=HY8aA at mail dot gmail dot com>
On 24 January 2014 10:11, Will Newton <firstname.lastname@example.org> wrote:
> On 24 January 2014 04:17, Alan Modra <email@example.com> wrote:
>> On Thu, Jan 23, 2014 at 03:22:21PM +0000, Will Newton wrote:
>>> On 16 January 2014 10:28, Will Newton <firstname.lastname@example.org> wrote:
>>> > I'm not sure what the implications of this patch are, so I thought I would
>>> > post it here in the hope that someone can explain what might break.
>> You're probably being a bit too honest here. :)
>> The patch looks OK to me as far as it goes, but I count myself
>> unqualified to ack it. I don't know enough about ARM to say whether
>> you've set pointer_equality_needed for all the relocs types you
>> should. It needs to be set whenever you might be taking the address
>> of a function in non-PIC code or data.
> Thanks, I wasn't sure of the semantics of pointer_equality_needed.
> I'll investigate and post an updated patch.
After investigating the relocs involved I believe the original patch
is correct from that point of view. Richard, have you got any opinion
on the patch?
Toolchain Working Group, Linaro