MI: Another -var-update bug?

[Vladimir Prus]

Nick Roberts wrote:

[moving this to -devel from -patches]
 
> Variable objects appear to test scope on a frame basis (in c_value_of
> root). If we create a variable object for j in the inner block of the
> program below
> then doing -var-update on line 10 doesn't report it as being out of scope.
>  But
> struct varobj_root has a member struct block *valid_block.  Presumably the
> addresses in this structure can be used to test if the variable is still
> in scope or not

The 'out of scope' and nested scopes are just completely broken. I've tried 
to make use of them some time ago, and could not make them work in any
sensible way.

That's why KDevelop computes addresses of each locals on each step, and if
address changes, manually recreates a variable object. This should be in
gdb, but it's a bit scary change ;-)

- Volodya