This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] S/390: Fix two issues with the IFUNC optimized mem* routines
On 8/30/2012 9:10 AM, H.J. Lu wrote:
> On Thu, Aug 30, 2012 at 12:07 AM, Andreas Krebbel
> <krebbel@linux.vnet.ibm.com> wrote:
>> On 29/08/12 18:24, H.J. Lu wrote:
>>> IFUNC symbol is accessed via GOT. An IFUNC function may have
>>> relocatons, but it shouldn't use GOT to get the resolved symbol.
>>
>> The question is whether we perhaps want/could get rid of this limitation by e.g. doing the
>> relocation processing in two passes. A first walk over the relocations could ignore the IFUNC relocs
>> and a second walk could only invoke the resolver functions.
>>
>
> I think it is unnecessary and overkill.
>
Agreed.
For hppa we had similar problems when processing function pointer
relocations. I also *thought* that doing relocations
in two passes would be a good idea, but at the end of the
day Ulrich convinced me that the right fix is always in the
static linker.
Cheers,
Carlos.
--
Carlos O'Donell
Mentor Graphics / CodeSourcery
carlos_odonell@mentor.com
carlos@codesourcery.com
+1 (613) 963 1026