lkml.org 
[lkml]   [2004]   [Feb]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: Split kgdb into "lite" and "normal" parts
Date
On Wednesday 25 Feb 2004 5:08 am, Tom Rini wrote:
> On Wed, Feb 25, 2004 at 12:27:03AM +0100, Pavel Machek wrote:
> > Hi!
> >
> > > > > > Tested (core-lite.patch + i386-lite.patch + 8250.patch)
> > > > > > combination. Looks good.
> > > > > >
> > > > > > Let's first check this in and then do more cleanups.
> > > > > > Tom, does it sound ok?
> > > > >
> > > > > This sounds fine to me. Pavel, I'm guessing you did this with
> > > > > quilt, could you provide some pointers on how to replicate this in
> > > > > the future?
> > > >
> > > > Unfortunately, I done it by hand :-(. But if -lite parts are not
> > > > merged, soon, I'll be forced to start using quilt. Doing stuff by
> > > > hand is quite painfull...
> > >
> > > There's still a whole bunch of bogons in the -lite patch still, so I
> > > don't think it should be merged yet.
> >
> > Well, it seems to contains a *lot* less bogons than what currently is
> > in -mm series.
> >
> > What big problems do you see? It does not yet use weak symbols, but I
> > do not think that's a serious problem. What else?
>
> The first two big ones are:
> - Doesn't like gdb 6.0 (You cannot assume the first packet is Hc...)

Can you tell me more about this?

> - Wierdities with kgdb_killed_or_detached / kgdb_might_be_resumed
> (both can die).

Yes. These have to be thought over again. I don't think a perfect solution
exists for all problems related to gdb kill/die followed by a reattach. We
should attempt a proper design describing different scenarios.

> - All of the function pointer games (of which the weak symbols, but not
> all of them) are a part of.
> - Issues w/ handling 'D' and 'k' packets cleaner (and I think there was
> a correctness fix in there, too, but it was a while ago).

Is this wrt kgdb_killed.., kgdb_might..., remove breakpoints?

> - Don't ACK packets sitting on the line

More info please.

-Amit

> - kgdb_schedule/process_breakpoint, required for kgdboe, harmless to use
> on serial.
>
> There's still a lot of stuff I checked into linux-2.6-kgdb that's
> non-trivially important
> (http://ppc.bkbits.net:8080/linux-2.6-kgdb/ChangeSet@-4w?nav=index.html)

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 14:01    [from the cache]
©2003-2011 Jasper Spaans