Re: Change stubs to handle '$' after packet start

[jtc@redback.com (J.T. Conklin)]

>>>>>> "Stan" == Stan Shebs <shebs@cygnus.com> writes:
Stan> OK, I applied the retry patch, it went right in, no problems.
Stan> So what's our next stub hack going to be?? :-)

I'd thought I'd go after some low hanging fruit first.  For example,
support for run length encoding can only be found in sh-stub.c.  And
I'd like to remove the support for 'b' (baudrate) from the sparc stub,
which has been blindly copied to the sparclite and the sparclet stubs.
It's been ifdef'd out for years, with the comment 'Disabled until we
can unscrew this properly'.  It's been 7 years and there are no signs
of this being unscrewed.  In fact strong arguments can (and have been)
made that commands that change transport layer attributes are in
general a bad idea.

Next, I thought I'd take another look at remote.texi, to see if there
has been anything that has been omitted.

After that, I'm not quite sure.  I think new status values for CPU
exceptions instead of encoding them as signals, plus all the assorted
changes within GDB that implies is high on the list.

        --jtc

-- 
J.T. Conklin
RedBack Networks