Re: [PATCH] DWARFv5. Handle DW_TAG_atomic_type _Atomic type modifier.

[Mark Wielaard <mjw at redhat dot com>]

On Mon, 2014-06-23 at 11:01 -0700, Eric Christopher wrote:
> FWIW a new vendor tag, even for such as this, would be a better
> solution. The numbers you choose for any attributes, if the proposals
> are accepted, would not necessarily be the same ones you chose. That
> confusion between numbers in consumers is worse than an unknown tag
> IMO.

Most certainly agreed. I normally work on other DWARF consumers and I
know all too well that whatever GCC/GDB adopts is what the rest of the
toolchain ends up having to support. The goal of the prototype patches
wasn't to get them integrated/adopted before DWARFv5 is finalized. The
goal was just to create something to see if it is doable (both GCC and
GDB seem to have nice datastructures already setup, so even for someone
like me without any prior knowledge of either program it was easy to
add) and give feedback on the proposal to see if it is worth it to adopt
for the next DWARF spec.

I'll think the answer is yes and I'll keep the patches around. So that
if there is a new DWARF standard version then we have an implementation
for the GNU toolchain as soon as it is final. But there isn't even a
first working draft at this point, just a bunch of proposals which might
or might not be adopted in their current or in some completely different
form. So these patches might not be integrated till next year.

For anything that does get integrated into GCC or GDB before DWARFv5 is
final we should certainly use a GNU vendor extension so that the rest of
the toolchain can easily adopt it. But that won't be possible in the
current form since DWARF type qualifier tags aren't vendor extensible.
And I don't know if anybody even needs it right now. Maybe if it is
useful for the GDB/GCC compiler expression effort we could create a GNU
vendor extension?

Cheers,

Mark