bfd_get_relocated_section_contents on hppa and ia64

Camm Maguire camm@maguirefamily.org
Mon Feb 1 22:57:00 GMT 2010


Greetings!  OK will do.  

Could you please advise me on a bit higher level.  GCL used bfd to
relocate loaded code prior to execution from the .data section of the
running process.  We have been maintaining (loosely) patches for this
purpose for some time.  I would greatly love to use an unpatched bfd.

On most systems, this is in fact the way it is -- we use bfd out of
the box without issue.  (x86, amd64, arm, m68k, sparc, ppc, s390, ???)

We have a macosx patch contributed a long time ago which I'd greatly
appreciate merging or removing.

We have mips, mipsel, and alpha modifications which only require one
change to the bfd tree, namely to make the howto struct non-constant,
so the special function can be altered.  We then use
sfaslbfd_{mips,alpha}.c from within GCL.  I'd be most happy to
contribute this code if you would consider using it and maintaining
it. 

On ia64, hppa, the relocation algorithm terminates in the error shown,
which I will try to detail on bugzilla.  I don't believe this is a
programming error, rather this routine was never written for these
cpu's because it is never used by the linker.  This is my
understanding only.

If you like the mips and alpha code, and can point out relocation
documentation for ia64 and hppa, I may be able to assist in generating
analogous files for these cpus.

Take care,

Nick Clifton <nickc@redhat.com> writes:

> Hi Camm,
>
>> Greetings!  Any plans on implementing this?
>
> If by implementing you mean fixing then the answer is yes.  But please
> could you file a bug report here:
>
>   http://sourceware.org/bugzilla/
>
> If you include a testcase to reproduce the problem that will really help.
>
> Cheers
>   Nick
>
>
>
>
>

-- 
Camm Maguire			     		    camm@maguirefamily.org
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah



More information about the Binutils mailing list