Re: [PATCH, doc RFA] Remove support for "target m32rsdi" and "target mips/pmon/ddb/rockhopper/lsi"

["Maciej W. Rozycki" <macro@imgtec.com>]

On Tue, 22 Mar 2016, Pedro Alves wrote:

> >> Ref: https://sourceware.org/ml/gdb/2016-03/msg00004.html
> >>
> >> This removes support for:
> >>
> >>   | target            | source                |
> >>   |-------------------+-----------------------|
> >>   | target m32rsdi    | gdb/remote-m32r-sdi.c |
> >>   | target mips       | gdb/remote-mips.c     |
> >>   | target pmon       | gdb/remote-mips.c     |
> >>   | target ddb        | gdb/remote-mips.c     |
> >>   | target rockhopper | gdb/remote-mips.c     |
> >>   | target lsi        | gdb/remote-mips.c     |
> >>
> >> That is:
> >>
> >>   - Remote M32R debugging over SDI.
> >>
> >>   - Debugging boards using the MIPS remote debugging protocol
> >>     over a serial line, PMON, and a few variants.
> >>
> >> These are the last non-"target remote" remote targets in the tree, if
> >> you don't count "target sim".
> > 
> > I am OK on this, but I'd like to hear what does Maciej (cc'ed) think.

 Thanks for the heads-up!

> FWIW, I didn't think this would affect MIPS in any way, given that
> I couldn't find a single bug report for any of these targets in
> bugzilla.  I don't mean open bugs, I mean, any, ever, other than
> build failure issues.  I don't recall anyone mentioning anything
> related to these targets on the gdb or gdb-patches lists, ever,
> either.  It seems like google searches for these targets only hit
> the gdb documentation, and my suggestion to remove them on
> the gdb@ list [1].  :-)

 Indeed, all this is old enough I didn't see any actual use myself either.  
All my activity with remote-mips.c was to adjust the source to keep it
compiling or to match changes made elsewhere.  This was from ~2005 on, 
which is when I got to it first.

> Also, I thought of MIPS's earlier focus on MDI, and knowing that
> even that one was pushed out of tree years ago, my thinking was
> that the risk of this affecting anyone in practice
> approximates zero.

 For the record, the MDI target hasn't actually ever made it in the first 
place.

 Anyway, having read your request to remove remote-mips.c I was fairly 
sure it could go, but to be double sure I asked around too, to see if 
there could be any legacy users.  All replies were negative, so I'm fine 
to let it rest in peace.  I can't speak of remote-m32r-sdi.c.

  Maciej