Re: port-amd64/50091

[Vicente Chaves de Melo <vicente.melo.br%ieee.org@localhost>]

The following reply was made to PR port-amd64/50091; it has been noted by GNATS.

From: Vicente Chaves de Melo <vicente.melo.br%ieee.org@localhost>
To: gnats-bugs%NetBSD.org@localhost, port-amd64-maintainer%netbsd.org@localhost,
 gnats-admin%netbsd.org@localhost, netbsd-bugs%netbsd.org@localhost, vchaves%ymail.com@localhost
Cc: 
Subject: Re: port-amd64/50091
Date: Tue, 28 Jul 2015 11:25:24 -0300

 Hi Patrick,
 below is the difference between the GENERIC and conf KGDB I'm using.
 
 dm4# diff GENERIC KGDB
 101c101
 < options       DDB             # in-kernel debugger
 ---
  > #options      DDB             # in-kernel debugger
 104,107c104,107
 < options       DDB_HISTORY_SIZE=512    # enable history editing in DDB
 < #options      KGDB            # remote debugger
 < #options KGDB_DEVNAME="\"com\"",KGDB_DEVADDR=0x3f8,KGDB_DEVRATE=9600
 < #makeoptions  DEBUG="-g"      # compile full symbol table
 ---
  > #options      DDB_HISTORY_SIZE=512    # enable history editing in DDB
  > options       KGDB            # remote debugger
  > options KGDB_DEVNAME="\"com\"",KGDB_DEVADDR=0x3f8,KGDB_DEVRATE=9600
  > makeoptions   DEBUG="-g"      # compile full symbol table
 
 Note that the following options remain commented out.
 
 # Diagnostic/debugging support options
 #options        DIAGNOSTIC      # inexpensive kernel consistency checks
                                  # XXX to be commented out on release branch
 #options        DEBUG           # expensive debugging checks/support
 #options        LOCKDEBUG       # expensive locking checks/support
 #options        KMEMSTATS       # kernel memory statistics (vmstat -m)
 
 Best regards
 Vicente
 
 
 On 28/07/2015 09:55, Patrick Welche wrote:
 > The following reply was made to PR port-amd64/50091; it has been noted by GNATS.
 >
 > From: Patrick Welche <prlw1%cam.ac.uk@localhost>
 > To: gnats-bugs%netbsd.org@localhost
 > Cc:
 > Subject: Re: port-amd64/50091
 > Date: Tue, 28 Jul 2015 13:50:26 +0100
 >
 >   Sadly kgdb is still broken for me even after this change. My gdb
 >   protocol trace is slightly different in that I have a "m" (read
 >   memory) rather than "p" (read register) after the "g" (read
 >   registers). This is today's current, so odd to see a difference:
 >   
 >   ...
 >   Sending packet: $qC#b4...Ack
 >   Packet received:
 >   Sending packet: $qAttached#8f...Ack
 >   Packet received:
 >   Packet qAttached (query-attached) is NOT supported
 >   Sending packet: $g#67...Ack
 >   Packet received: 0f000000000000002000000000000000000000000000000001000000000000000060f280ffffffff0000000000000000c09eee80ffffffffc09eee80ffffffff0060f280ffffffffd0070000000000000060f280ffffffff7300000000000000000000000000000000000000000000000000000000000000000000000000000035c31e80ffffffff860200000000000008000000000000001000000000000000
 >   Sending packet: $m80f26000,1#90...Ack
 >   Timed out.
 >   
 >   The time out is because the target panics with a
 >   
 >   panic: lockdebug_lookup: uninitialized lock (lock=0xffffffff80d01448, from=fff807bc16a)
 >   
 >   after having just done a sys/arch/amd64/amd64/kgdb_machdep.c,
 >   kgdb_acc(va=0x80f26000,len=1)
 >   
 >   i.e., the panic happens in
 >   
 >   pte = kvtopte(va)      (as VM_MN_KERNEL_ADDRESS=0)
 >   
 >   (as per http://mail-index.netbsd.org/tech-kern/2015/07/09/msg019142.html)
 >   
 >