Linus, please pull from the for_linus branch at:

 
git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/linux-2.6-kgdb.git
for_linus

These are the fixes since the initial kgdb merge.

Thanks,
Jason.

---
Harvey Harrison (2):
      kgdb: trivial sparse fixes in kgdb test-suite
      kgdb: fix signedness mixmatches, add statics, add declaration to
header

Jan Engelhardt (1):
      kgdb: kconfig fix xconfig/menuconfig element

Jason Wessel (1):
      kgdb: 1000 loops for the single step test in kgdbts

grzegorz.chwesewicz@chilan.com (1):
      kgdb: minor documentation fixes

 Documentation/DocBook/kgdb.tmpl |    8 +++--
 drivers/misc/kgdbts.c           |   75
++++++++++++++++++++++----------------
 include/linux/kgdb.h            |    4 ++-
 kernel/kgdb.c                   |    8 ++--
 lib/Kconfig.kgdb                |   16 +++++----
 5 files changed, 64 insertions(+), 47 deletions(-)

--

So when will we see kgdboe or some other way of getting gdb to connect to the 
target over ethernet upstream?

Jesse
--

i think the proper way would be to extend the kgdb-over-console (kgdboc) 
concept and make it easy to utilize the existing netconsole code via 
KGDB. This brings nice synergy: netconsole is utilized by other debug 
infrastructure as well so anything that supports netconsole would 
automatically support KGDB as well. I dont really like the old kgdboe 
approach from a design POV as it created a parallel infrastructure for 
special-purpose kgdb IO.

likewise, the firewire debugging code could/should be extended to create 
a bidirectional console as well - that has other benefits as well. (we'd 
get printks over firewire - instead of the special-purpose firescope 
utility)

and the early console drivers could be reused to do early debugging via 
KGDB as well.

	Ingo
--

When I looked at the sources right after it was merged this seemed like the 
preferred approach, but I couldn't find any evidence of anyone working on the 
port...  Are you aware of any?

Jesse
--