This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: [RFA:] Use refcount in SH ELF backend
- From: Alan Modra <amodra at bigpond dot net dot au>
- To: "Clarke, Stephen" <stephen dot clarke at superh dot com>
- Cc: Daniel Jacobowitz <drow at mvista dot com>, kaz Kojima <kkojima at rr dot iij4u dot or dot jp>, binutils at sources dot redhat dot com
- Date: Fri, 23 Aug 2002 22:09:31 +0930
- Subject: Re: [RFA:] Use refcount in SH ELF backend
- References: <287E4644B5249D449C56FA5409A874AE03EF92@sh-us-ex01.us.w2k.superh.com>
On Thu, Aug 22, 2002 at 02:00:43PM -0700, Clarke, Stephen wrote:
> > From: Daniel Jacobowitz [mailto:drow@mvista.com]
> > Sent: Thursday, August 22, 2002 12:58 PM
>
> > I see that Debian's been using this patch for some time; so has the
> > SH-Linux community. I'd really appreciate it if an SH5-aware
> > individual would step forward and look at this so it can be approved.
>
> I'd really like to see this patch go in too ...
I had a good look over the patch since it is porting the delayed
allocation scheme for dynamic relocs I implemented for x86, and it
looks good to me. Besides a few cosmetic mods, eg. replacing
"elf_hash_table (info)->" with "htab->root.", I made only one change
of any substance, adding
if (h->root.type == bfd_link_hash_warning)
h = (struct elf_link_hash_entry *) h->root.u.i.link;
to the start of readonly_dynrelocs.
That said, it clearly needs the fixes you mention in
http://sources.redhat.com/ml/binutils/2002-07/msg00190.html for
SHmedia, and a careful look at just what happens with those indirect
syms you create. Any objections to me committing the patch as is,
and you fixing the Shmedia breakage later? I've gotten a little
interested in SH since my assembler reloc processing changes proved
difficult to get right for this port. :)
--
Alan Modra
IBM OzLabs - Linux Technology Centre