dwarf patches fail bootstrapping coff platforms]

[mark@codesourcery.com]

>>>>> "Jeffrey" == Jeffrey A Law <law@cygnus.com> writes:

    >> >>>>> "Robert" == Robert Lipe <robertlipe@usa.net> writes:
    >> 
    Robert> In init.c:expand_aggr_vbase_init() if I comment out the
    Robert> call to expand_cleanup_for_base(), it assembles once
    Robert> again.  I have no idea if the code actually works, if the
    Robert> debugging records are valid, etc., but hopefully this will
    Robert> be enough to illuminate the little light bulb over your
    Robert> head. :-)
    >>  I'm afraid I'm still in the dark.  That was indeed the only
    >> substantive change in that patch (the rest consisted of moving
    >> code out of one function so that it could be called from
    >> expand_aggr_vbase_init).  We're going to need a COFF expert,
    >> I'm afraid.
    Jeffrey> If it's just a question of the debugging record
    Jeffrey> correctness, then it's unlikely that you need to know
    Jeffrey> anything about COFF -- just dwarf :-)

Sorry, that's what I meant.  I can actually do COFF just fine. :-(

I could probably figure out DWARF, eventually, but it would be much
quicker if Jason or Jim could take a look.

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com