a target creates an inferior

[Andrew Cagney]

Ref: interface VS target (or something else wrong with the target vector)
http://sources.redhat.com/ml/gdb/2003-10/msg00218.html

I think the theory is right but the names are wrong.  Since we're always 
refering to the inferior we might as well refer to the currently running 
instance (well ignoring core files) as the "inferior".

This gives rise to:

- the exec target's run command creates a new "struct inferior" (or is 
that struct inferior_info).

- read the inferior's memory (inferior->memory_read_partial(...))
At present it is read the target's memory.

- the inferior changed
At present we say the target changed

- the inferior has threads
...

does anyone see problems with this arangement.

Andrew

PS: Initially there will likely be the arangement:
	current_target->inferior == &current_inferior
	current_inferior->target == &current_target
that way we can incrementally migrate inferior methods across to the 
inferior object from the target.