[RFC] varobj deletion after the binary has changed

[Nick Roberts]

 > You might be true. I mentioned that because it's easy to check and there
 > seems to be no matching patch description in your spec file. Anyway
 > fixing that seems simple enough that we don't need to dig through 80
 > RedHat patches to come up with a patch. 

It's not simple for me as I'm not familiar with that part (symbol tables)
of the code yet.

 > AFAIK Denis's working on a version re-evealuating (actually re-parsing)
 > varobjs with no attached blocks and putting the others in some error
 > state that'll return in_scope="false" at the next update. Does that seem
 > reasonable?

In FC5 GDB there doesn't seem to be a problem with varobjs with no attached
blocks i.e global variables, and those which do have attached blocks i.e locals
do currently return in_scope="false", on GNU/Linux at least (maybe it has
something to do with randomisation though).  Like Dennis, I think there should
be some provision for deleting the latter on restarting.  With watchpoints this
happens automatically.

-- 
Nick                                           http://www.inet.net.nz/~nickrob