[Dwarf-Discuss] DWP mixed (DWARFv4/pre-standard + DWARFv5) content

[David Blaikie]

(please add anyone who has a vested interest in Split DWARF in general and
dwp in particular)

tl;dr:
How should DWARFv4 and DWARFv5 coexist in a DWP file:
1) Not at all (invalid/unsupported)
2) Single index table where the section indexes are subjective (look at the
version of the referenced CU/TU to decide what a column number means (eg: 8
is DW_SECT_MACRO in v4, DW_SECT_RNGLISTS in v5)
3) Emit multiple indexes (cu_index and tu_index each would contain two
indexes) - a v4 (well, index version 2) and a v5 (index version 5) index.

Longer discussion:
So I'm working on some (perhaps admittedly a bit hackish at this stage)
support for DWARFv5 in binutils' dwp tool.

So far I've adde support for parsing the DWARFv5 header and for correctly
handling str_offsets in v5 where a header is present. (& appropriately only
skipping the header for a v5 contribution (which means having to look at
the compile unit before parsing the corresponding str_offsets contribution))

But looking at supporting rnglists and loclists it gets a bit fuzzier when
it comes to the cu_index - debug_loc and debug_loclists share the same
DW_SECT numbering (5) in the pre-standard format (
https://gcc.gnu.org/wiki/DebugFissionDWP ) and in DWARFv5 (
http://dwarfstd.org/doc/DWARF5.pdf - Table 7.1), so that's sort of fungible.

rnglists is the trickier one - it doesn't exist in the pre-standard form
doesn't have this field (it has DW_SECT_MACINFO at 7 and DW_SECT_MACRO at
8), the standard version does (DW_SECT_MACRO at 7, DW_SECT_RNGLISTS at 8).

So... that leaves me with the question at the top: Unsupported,
contextually interpreted (read the CU/TU to find the DWARF version, use
that to decide which section name a column index refers to), two indexes
(pre-standard and v5)?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dwarfstd.org/pipermail/dwarf-discuss-dwarfstd.org/attachments/20200225/dfac556d/attachment.html>