remote.c and a re-entrant event-loop

[Andrew Cagney <ac131313 at cygnus dot com>]

Hello,

Ref:

  non-blocking reads/writes and event loops
    http://sources.redhat.com/ml/gdb/2000-06/msg00088.html
  remote.c's "async", "extended-async" back ends
    http://sources.redhat.com/ml/gdb/2001-02/msg00016.html

In http://sources.redhat.com/ml/gdb/2000-06/msg00144.html I sketched out
two possible ways of progressing the event-loop problem vis:

o       invert the targets first
	Would involve re-entering the
	event-loop

o	invert the expression evaluator first
	Would involve wrapping targets

Thinking over this again.  I would argue that right now we have the
first case - a re-entrant event loop - only it is carefully hidden in
select/read and timers in serial.c, /proc, ...

Assuming that is accepted as the current reality, can I propose that the
real event loop be made re-entrant so that the existing code is at least
able to use that instead of the current hacks.  It also gives people
motivated to work on the event loop problem the freedom to evolve things
in which ever direction suits them (from the top or bottom).  (but more
on that later :-)

	Andrew

PS: What has this to do with remote.c?  It is the other reason why
remote.c can't be simplified.  In the end you still end up needing code
for the blocking and async cases.