Facing a malformed packet error using Insight debugger

[Insightuser]

Hi,
I am using insight to debug my target mips2.
If I use the command line debugger provided by montavista, after connecting to my target , the resposne is

(gdb) set solib-absolute-prefix .
(gdb) target remote 192.168.12.162:1000
Remote debugging using 192.168.12.162:1000
0x2aaa8830 in __start () from ./lib/ld.so.1
(gdb) c

and this works fine.


If I use the insight gdb, the following problem is faced.
$ /cygdrive/c/Insight6.8/tuxbuilder-1.1/mipsel-unknown-linux-gnu/cross/usr/bin/
mipsel-linux-gdb.org.exe ./philips/apps/amApp
GNU gdb 6.6
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "--host=i686-cross-cygwin --target=mipsel-linux"...


entry: ~/.gdbinit
exit: ~/.gdbinit
(gdb) set solib-absolute-prefix .
(gdb) target remote 192.168.12.162:1000
Remote debugging using 192.168.12.162:1000
0x2aaa8830 in __start () at ../ports/sysdeps/mips/dl-machine.h:81
81      ../ports/sysdeps/mips/dl-machine.h: No such file or directory.
        in ../ports/sysdeps/mips/dl-machine.h
(gdb) c
Continuing.
[New Thread 240]
warning: Invalid remote reply: nf1
Malformed packet(b) (missing colon): 25:bcc9d62a;1d:f8aa9c7f;thread:f0;
Packet: 'TR242;25:bcc9d62a;1d:f8aa9c7f;thread:f0;'

(gdb)

please let me know How do I solve the malformed packet error.

[Keith Seitz]

Insightuser wrote:

> Hi,
> I am using insight to debug my target mips2.
> If I use the command line debugger provided by montavista, after connecting
> to my target , the resposne is
>
> (gdb) set solib-absolute-prefix .
> (gdb) target remote 192.168.12.162:1000
> Remote debugging using 192.168.12.162:1000
> 0x2aaa8830 in __start () from ./lib/ld.so.1
> (gdb) c
>
> and this works fine.

You left off the version information -- what does "show version" say?


> GNU gdb 6.6
> Copyright (C) 2006 Free Software Foundation, Inc.

Wow, that is one old debugger!! I really suggest you try at least 6.8.
Even better: use a snapshot. Insight and gdb are always kept in a usable
state. I only ever use CVS HEAD for my daily work.

> [New Thread 240]
> warning: Invalid remote reply: nf1
> Malformed packet(b) (missing colon): 25:bcc9d62a;1d:f8aa9c7f;thread:f0;
> Packet: 'TR242;25:bcc9d62a;1d:f8aa9c7f;thread:f0;'

I'm long out of practice with the remote protocol, but this packet does
look malformed. According to GDB's remote.c and the documentation, the
proper reply packet starting with 'T' should be
"TssN1:R1;N2:R2;...;Nn:Rn". ss = Signal number; N1..Nn = register
numbers; R1..Rn = register values. I have no idea what "nf1" could
possibly be.

According to remote_parse_stop_reply in remote.c, the code skips the
first three bytes of the packet, parsers the next two bytes as a
register number and looks for the colon that separates this number from
the value. It doesn't find the colon, and the error above is emitted.

As far as I can tell, gdb/insight are working as designed. [But again,
let me warn you that I am no pro at the remote protocol.] Are you
perhaps using some sort of custom Monta Vista extension to the remote
protocol?

You might have better luck with this question if you email the gdb
mailing list; that is where the experts hang out, many of them from
Monta Vista.

Keith

[Insightuser]

Hi Keith,
Thanks for your reply.
The show version in montavista gdb shows:

(gdb) show version
GNU gdb 6.6.50.20070301 (MontaVista 6.6.50-2.0.5.0703840 2007-08-23)
Copyright (C) 2007 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "--host=i686-pc-cygwin --target=mipsel-montavista-lin
ux-gnu".

And the insight gdb shows :

(gdb) show version
GNU gdb 6.6
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "--host=i686-cross-cygwin --target=mipsel-linux".

My gdbserver doesn't print any version information when it is started.
as you suggested, I would mail this to gdb mailing list as well.

thanks,
Raju

Keith Seitz wrote:

Insightuser wrote:
> Hi,
> I am using insight to debug my target mips2.
> If I use the command line debugger provided by montavista, after connecting
> to my target , the resposne is
>
> (gdb) set solib-absolute-prefix .
> (gdb) target remote 192.168.12.162:1000
> Remote debugging using 192.168.12.162:1000
> 0x2aaa8830 in __start () from ./lib/ld.so.1
> (gdb) c
>
> and this works fine.

You left off the version information -- what does "show version" say?


> GNU gdb 6.6
> Copyright (C) 2006 Free Software Foundation, Inc.

Wow, that is one old debugger!! I really suggest you try at least 6.8.
Even better: use a snapshot. Insight and gdb are always kept in a usable
state. I only ever use CVS HEAD for my daily work.

> [New Thread 240]
> warning: Invalid remote reply: nf1
> Malformed packet(b) (missing colon): 25:bcc9d62a;1d:f8aa9c7f;thread:f0;
> Packet: 'TR242;25:bcc9d62a;1d:f8aa9c7f;thread:f0;'

I'm long out of practice with the remote protocol, but this packet does
look malformed. According to GDB's remote.c and the documentation, the
proper reply packet starting with 'T' should be
"TssN1:R1;N2:R2;...;Nn:Rn". ss = Signal number; N1..Nn = register
numbers; R1..Rn = register values. I have no idea what "nf1" could
possibly be.

According to remote_parse_stop_reply in remote.c, the code skips the
first three bytes of the packet, parsers the next two bytes as a
register number and looks for the colon that separates this number from
the value. It doesn't find the colon, and the error above is emitted.

As far as I can tell, gdb/insight are working as designed. [But again,
let me warn you that I am no pro at the remote protocol.] Are you
perhaps using some sort of custom Monta Vista extension to the remote
protocol?

You might have better luck with this question if you email the gdb
mailing list; that is where the experts hang out, many of them from
Monta Vista.

Keith

[Insightuser]

Hi Keith,
I have downloaded the insight-weekly-6.8.50-20090427.tar of insight available on ftp://sourceware.org/pub/insight/snapshots/current/
I am trying to build it using cygwin.
Facing the following error.

config.status: creating po/Makefile.in
config.status: executing depfiles commands
config.status: executing libtool commands
config.status: executing default-1 commands
config.status: executing bfd_stdint.h commands
config.status: executing default commands
make[3]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
rc/bfd/po'
make[3]: Entering directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/
src/bfd/po'
make[3]: Nothing to be done for `info'.
make[3]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
rc/bfd/po'
make[3]: Entering directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/
src/bfd'
make[3]: Nothing to be done for `info-am'.
make[3]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
rc/bfd'
make[2]: *** [info-recursive] Error 1
make[2]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
rc/bfd'
make[1]: *** [all-bfd] Error 2
make[1]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
rc'
make: *** [all] Error 2

ing01536@INGBTCPIC6NB322 /cygdrive/o/space/__insight_MontaVistaMips32_Debug/src
$

Can you please let me know how to fix this error.

regards,
Raju

-----Original Message-----
From: Keith Seitz [mailto:keiths@...]
Sent: 2009 Jul 06 9:17 PM
To: Ramaraju G
Cc: insight@...
Subject: Re: Facing a malformed packet error using Insight debugger

Insightuser wrote:

> Hi,
> I am using insight to debug my target mips2.
> If I use the command line debugger provided by montavista, after connecting
> to my target , the resposne is
>
> (gdb) set solib-absolute-prefix .
> (gdb) target remote 192.168.12.162:1000
> Remote debugging using 192.168.12.162:1000
> 0x2aaa8830 in __start () from ./lib/ld.so.1
> (gdb) c
>
> and this works fine.

You left off the version information -- what does "show version" say?


> GNU gdb 6.6
> Copyright (C) 2006 Free Software Foundation, Inc.

Wow, that is one old debugger!! I really suggest you try at least 6.8.
Even better: use a snapshot. Insight and gdb are always kept in a usable
state. I only ever use CVS HEAD for my daily work.

> [New Thread 240]
> warning: Invalid remote reply: nf1
> Malformed packet(b) (missing colon): 25:bcc9d62a;1d:f8aa9c7f;thread:f0;
> Packet: 'TR242;25:bcc9d62a;1d:f8aa9c7f;thread:f0;'

I'm long out of practice with the remote protocol, but this packet does
look malformed. According to GDB's remote.c and the documentation, the
proper reply packet starting with 'T' should be
"TssN1:R1;N2:R2;...;Nn:Rn". ss = Signal number; N1..Nn = register
numbers; R1..Rn = register values. I have no idea what "nf1" could
possibly be.

According to remote_parse_stop_reply in remote.c, the code skips the
first three bytes of the packet, parsers the next two bytes as a
register number and looks for the colon that separates this number from
the value. It doesn't find the colon, and the error above is emitted.

As far as I can tell, gdb/insight are working as designed. [But again,
let me warn you that I am no pro at the remote protocol.] Are you
perhaps using some sort of custom Monta Vista extension to the remote
protocol?

You might have better luck with this question if you email the gdb
mailing list; that is where the experts hang out, many of them from
Monta Vista.

Keith

[Keith Seitz]

On 07/16/2009 11:09 PM, Ramaraju G wrote:

> config.status: creating po/Makefile.in
> config.status: executing depfiles commands
> config.status: executing libtool commands
> config.status: executing default-1 commands
> config.status: executing bfd_stdint.h commands
> config.status: executing default commands
> make[3]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
> rc/bfd/po'
> make[3]: Entering directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/
> src/bfd/po'
> make[3]: Nothing to be done for `info'.
> make[3]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
> rc/bfd/po'
> make[3]: Entering directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/
> src/bfd'
> make[3]: Nothing to be done for `info-am'.
> make[3]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
> rc/bfd'
> make[2]: *** [info-recursive] Error 1
> make[2]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
> rc/bfd'
> make[1]: *** [all-bfd] Error 2
> make[1]: Leaving directory `/cygdrive/o/space/__insight_MontaVistaMips32_Debug/s
> rc'
> make: *** [all] Error 2

You haven't left me enough of the build output to actually see the error...

Keith

[Keith Seitz]

On 07/17/2009 12:40 PM, Keith Seitz wrote:
> You haven't left me enough of the build output to actually see the error...

Cincidentally enough, I just ran into this on my own, newly installed
Fedora 11 system. I think this is probably your problem:

WARNING: `makeinfo' is missing on your system.  You should only need it if
          you modified a `.texi' or `.texinfo' file, or any other file
          indirectly affecting the aspect of the manual.  The spurious
          call might also be the consequence of using a buggy `make' (AIX,
          DU, IRIX).  You might want to install the `Texinfo' package or
          the `GNU make' package.  Grab either from any GNU archive site.
make[2]: *** [bfd.info] Error 1

As the warning suggests, you must have the texinfo package installed.
This is required to build a snapshot.

If this isn't the problem you have afterall, then you must send the
actual error message.

Keith

[Insightuser]

Yes, Keith, After installing the texinfo and GNU make, my build is through.
thanks

-----Original Message-----
From: Keith Seitz [mailto:keiths@...]
Sent: 2009 Jul 18 2:27 AM
To: Ramaraju G
Cc: insight@...
Subject: Re: problem in building Insight debugger

On 07/17/2009 12:40 PM, Keith Seitz wrote:
> You haven't left me enough of the build output to actually see the error...

Cincidentally enough, I just ran into this on my own, newly installed
Fedora 11 system. I think this is probably your problem:

WARNING: `makeinfo' is missing on your system.  You should only need it if
          you modified a `.texi' or `.texinfo' file, or any other file
          indirectly affecting the aspect of the manual.  The spurious
          call might also be the consequence of using a buggy `make' (AIX,
          DU, IRIX).  You might want to install the `Texinfo' package or
          the `GNU make' package.  Grab either from any GNU archive site.
make[2]: *** [bfd.info] Error 1

As the warning suggests, you must have the texinfo package installed.
This is required to build a snapshot.

If this isn't the problem you have afterall, then you must send the
actual error message.

Keith

[Insightuser]

Hi Keith,
I have built the latest version of gdb. Facing the following error while debugging with montavista gdbserver.
Any idea on how to fix this ?

O:\space\__insight_MontaVistaMips32_Debug\insightdebugger\bin>mips2_fp_le-gdb.ex
e /cygdrive/y/philips/apps/amApp
GNU gdb (GDB) 6.8.50.20090427-cvs
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=i686-pc-cygwin --target=mips2_fp_le".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
(gdb) cd y:
Working directory /cygdrive/y.
(gdb) set solib-absolute-prefix .
(gdb) target remote 192.168.12.162:1000
Remote debugging using 192.168.12.162:1000
0x2aaa8830 in __start () from ./lib/ld.so.1
(gdb) c
Continuing.
Malformed packet(b) (missing colon): ;25:bcc9d62a;1d:f82aa57f;thread:eb;
Packet: 'TR237;25:bcc9d62a;1d:f82aa57f;thread:eb;'

regards,
raju

-----Original Message-----
From: Keith Seitz [mailto:keiths@...]
Sent: 2009 Jul 18 2:27 AM
To: Ramaraju G
Cc: insight@...
Subject: Re: problem in building Insight debugger

On 07/17/2009 12:40 PM, Keith Seitz wrote:
> You haven't left me enough of the build output to actually see the error...

Cincidentally enough, I just ran into this on my own, newly installed
Fedora 11 system. I think this is probably your problem:

WARNING: `makeinfo' is missing on your system.  You should only need it if
          you modified a `.texi' or `.texinfo' file, or any other file
          indirectly affecting the aspect of the manual.  The spurious
          call might also be the consequence of using a buggy `make' (AIX,
          DU, IRIX).  You might want to install the `Texinfo' package or
          the `GNU make' package.  Grab either from any GNU archive site.
make[2]: *** [bfd.info] Error 1

As the warning suggests, you must have the texinfo package installed.
This is required to build a snapshot.

If this isn't the problem you have afterall, then you must send the
actual error message.

Keith

[Keith Seitz]

On 07/20/2009 12:15 AM, Ramaraju G wrote:
> (gdb) c
> Continuing.
> Malformed packet(b) (missing colon): ;25:bcc9d62a;1d:f82aa57f;thread:eb;
> Packet: 'TR237;25:bcc9d62a;1d:f82aa57f;thread:eb;'

Oddly, enough, this came up just recently...

http://sourceware.org/ml/gdb/2009-07/msg00037.html

Keith