[PATCH] Dwarf location expressions, tracing and printing

[Stan Shebs]

Tom Tromey wrote:
>>>>>> "Stan" == Stan Shebs <stan@codesourcery.com> writes:
>>>>>>             
>
> Stan> Perhaps the most interesting part is that I finally got tired of
> Stan> working blind and filled out description location functions, so
> Stan> now both "info scope" and "info address" have quite a lot of say
> Stan> about variables.  Try things like "info scope
> Stan> handle_inferior_event" to get some truly awesome verbiage. :-)
>
> Thanks, I think this is great.
>
> I think Jan had a similar patch that would enable disassembling the
> DWARF expression.  I'm not sure what happened to that; it seems like
> something that would at least be useful when working on gdb or other
> parts of the toolchain.
>   

What I did is not quite the same thing, in that it restricts itself to 
interpreting location expressions, and tries to do it in a 
user-intelligible way, no opcodes shown.  But yeah, we could use more 
display power; with dwarf things being so data-driven, GDB backtrace 
alone isn't that useful.

Stan