[Dwarf-discuss] Encoding fixed-point types

Todd Allen todd.allen
Wed Jun 29 11:12:51 GMT 2005

> The lack of rational constants, alas, would seem to limit the generality of 
> the supported 'Small values.  I observe that such a facility would also give
> a way to represent Ada named numbers.  

Well, without them things aren't perfect.  But providing smalls as
double-precision floating-point is pretty good.  And likewise with named
numbers.  And it saves debuggers from having to support abitrary-length
integers.  But I certainly wouldn't object to a new DW_TAG_rational_constant
that provided perfect representations.

> Are the vendor extensions for Ada that you mention publicly available?

Yes and no.  I have permission to distibute Concurrent's, but it isn't
anywhere that's easy to reach at the moment.  I'll send you a copy in a
subsequent email (to keep the mailing list bandwidth low).

Michael Eager is planning to place DDC-I's spec on the Dwarf 3 website.  And
I'll be sending him a copy of Concurrent's spec to go in the same place.

Todd Allen
Concurrent Computer Corporation

More information about the Dwarf-discuss mailing list