[Dwarf-Discuss] PROPOSAL: Constant expressions in location lists

[Jim Blandy]

On Jan 4, 2008 1:28 PM, Frank Ch. Eigler <fche at redhat.com> wrote:
> I just joined the list to rescussitate an old thread from June last
> year [1].  I work on the systemtap project, and we are avid consumers
> of DWARF data accompanying highly optimized binaries.  I bring it up
> because we are starting to encounter real-world cases where variables
> are optimized away from addressable locations, but can nevertheless be
> readily rematerialized for debugging purposes.  Having a
> DW_OP_computed_value (or extension of DW_OP_implicit_value) of the
> sort Roland proposed would be a good fit to expose this data.

When you say "rematerialized for debugging purposes", do you mean that
the value is a constant known at build time, or do you mean that it
can be computed from the current state of the program, even though the
variable has no location allocated to it at the moment?

The DW_OP_constant_value suggestion would help with the former, but
not the latter.  (What you wrote makes me think you have the latter in
mind.)