gdb remote protocol, winnt, wigglers

[Quality Quorum <qqi at world dot std dot com>]

Hi,

I am working on implmenting standard compliant remote.c and gdbserver.
I found just a few problems with protocol specs so far. 

1. There is no need in 'isThreadAlive' request: remote.c has to maintain 
   local cache of available threads, which is updated by 'threadListQuery'
   if required (I tried a few variants implementing thread list 
   synchronization and it seems this one is the most rational).

   Target methods xxx_is_thread_alive and xxxxx_find_new_thread 
   are simply using this cache without generating any traffic
   on the wire.

   So. 'itThreadAlive' is depreciated.

2. I misread code implementing 'addBWPoint' and 'deleteBWPpoint', 
   so the description was incorrec, it is fixed now.

I put updated document at http://www.std.com/qqi/ftp/protocol.txt .

I put updated rproxy at http://www.std.com/qqi/labslave/rproxy.html .

It is now running on WinNT and sipports Wiggler among other things - 
I tried it with MPC860ADS.  It will be not too complicated to add 
other processors, however, I do have 68360 based board only and no 
other wigglers. I have ICD cable which I am using under linux,
I am planning to provide a target under linux using ICD cable to 68360.

Please, be forgiving: WinNT stuff is barely in alpha state, rest is
barely in beta. 

I have nagging outstanding problem: I do not know what paperwork and 
what changes in the files I have to make in order to turn it over 
to FSF. I signed and submitted the most appropriate paper I found on
gnu website, however, I did not hear from them yet and it seems to me
that it will not be enough.

Can any kind soul help me with this ?

Thanks,

Aleksey




------
Want more information?  See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/
Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com