Re: [patch] m32r gas relocation processing

[Jeffrey A Law <law at cygnus dot com>]

  In message <200003100302.TAA26748@cygint.cygnus.com>you write:
  > : How about this instead?
  > : I've been slow in fully testing this [or it would have been installed
  > : already].  Maybe you can help?
  > 
  >   I didn't realize that you were looking at the same problem.  I had
  >   considered reworking the relocation routines in elf32-m32r.c but
  >   scrapped that approach in favor of mimicking the mips port.  It
  >   seems like we do some unecessary processing with this patch.  For
  >   example, do we need to build the hilist if we're generating relocs?
  >   My patch didn't really address that issue either.
  > 
  >   In general, I feel that the processing for REL relocations (when 
  >   we're actually emitting the relocs to a file), is too intertwined
  >   with the final relocation processing. This patch certainly fixes
  >   the test case and I'm fine with using it.
  > 
  >   This is dlindsay's issue though (I was just helping) so I'm going
  >   to let him do the tool chain testing.  Thanks.
  > 
  >   Catherine
  >   
  > 2000-03-09  Doug Evans  <dje@casey.transmeta.com>
  > 
  > 	* elf32-m32r.c (m32r_elf_lo16_reloc): Rewrite.
Doug probably has as much experience as anyone with the m32r port (did you
ever think you would really get away Doug? :-)

If Doug and you are happy with the change, then go ahead and check in
the patch.

Thanks,
jeff