Dwarf 2 Minutes, March 6, 2001

[Michael Eager <eager at eagercon dot com>]

Issue discussion:

010219.1 -- Duplicate Dwarf elimination

	This proposal describes a scheme which can be used to
	identify and eliminate duplicate Dwarf data.   It does
	not require any changes to the Dwarf specification, and,
	although described in terms of ELF, is not bound to any
	specific object format.  There were several suggestions
	about wording changes, particularly to introduce the 
	concept of "section groups" (also called COMDAT sections)
	earlier in the presentation.  Another suggestion was that
	the name DW_TAG_subunit might be changed to something more
	descriptive.

	A concern was raised about the description referring to and
	appearing to depend upon ELF.   The Dwarf specification 
	does not currently mention ELF, and there is no generally
	available specification for the format.  Opinion was 
	offered that editorial changes to describe the scheme in
	more general, non-ELF dependent, terms would be desirable.

	Proposal accepted.

010222.1 -- Multisection location and range lists

	Variations on this proposal have been discussed at several
	Dwarf committee meetings.  This proposal provides for an
	enhancement to location lists and range list to describe
	multiple, discontiguous address ranges.  It also changes
	the requirements on where to specify DW_AT_low_pc and its
	interpretation.

	Proposal accepted.

Late proposals
	
	We have received a number of proposals which address various
	Ada language issues.  Our previous goal had been to complete
	the Dwarf specification and submit it for public review in
	April.  Further, the committee contains few people who have
	an adequate understanding of Ada who would be able to review
	and discuss such proposals.  

	The committee appreciates the contributions in this area and
	regrets that they arrive as the modifications to the standard
	are being finalized.

	The committee decided to consider proposals which would make
	Dwarf more "Ada friendly", but not to consider full support
	for Ada.  This would mean proposals which address areas where
	the current standard either prevents or significantly impedes
	providing support for Ada in Dwarf would be considered, as well
	as proposals which make minor corrections to changes introduced
	in version 2.1.  Other proposals will not be accepted; they 
	may be addressed by vendor extensions and possibly presented
	for inclusion in a future Dwarf revision.

IEEE-ISTO
	
	Work has been continuing on drafting a procedures document 
	which would formalize the Dwarf Standards Committee as part 
	of IEEE-ISTO.  Suggestions on notifications for public review
	included publishing a letter (or press release) in SigPLAN
	Notices.
	 

-- 
Michael Eager	 Eager Consulting     eager@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077