This is the mail archive of the
mailing list for the eCos project.
Re: pb installing RedBoot on AT91EB40
> Hi Gary,
> thank you for the answer,
>> On Sun, 2005-12-11 at 01:52 +0100, Olivier Kaloudoff wrote:
>>> Hello all,
>>> following the Redboot user guide, I'm trying the
>>> rdi mode with gdb, using a KeySpan USB to Serial converter;
>>> (gdb) imac:~ kalou$ ls -l /dev/*Key*
>>> crw-rw-rw- 1 root wheel 9, 5 Dec 11 01:42 /dev/cu.KeySerial1
>>> crw-rw-rw- 1 root wheel 9, 4 Dec 11 01:34 /dev/tty.KeySerial1
>>> imac:~ kalou$ /usr/local/bin/arm-elf-gdb
>>> GNU gdb 6.0
>>> Copyright 2003 Free Software Foundation, Inc.
>>> GDB is free software, covered by the GNU General Public License, and
>>> welcome to change it and/or distribute copies of it under certain
>>> Type "show copying" to see the conditions.
>>> There is absolutely no warranty for GDB. Type "show warranty" for
>>> This GDB was configured as "--host=powerpc-apple-darwin6.8
>>> (gdb) tar rdi s=/dev/tty.KeySerial1
>>> Could not open device "s=/dev/tty.KeySerial1"
>>> (gdb) tar rdi s=/dev/cu.KeySerial1
>>> Could not open device "s=/dev/cu.KeySerial1"
>>> Am I doing something obviously wrong ?
>>> I checked that the Serial is ok, using zTerm, I get the Angel debug
>>> monitor fine;
>> GDB often (for reasons known only to the past) is picky about what the
>> name of the serial device is. Try making an alias (hard or soft link)
>> for the device, e.g. /dev/ttyS0
>> What system (OS) is this that uses such arcane names?
> I'm using OSX 10.4;
> imac-g5-de-olivier-kaloudoff:/dev root# uname -a
> Darwin imac-g5-de-olivier-kaloudoff.local 8.3.0 Darwin Kernel Version
> 8.3.0: Mon Oct 3 20:04:04 PDT 2005; root:xnu-792.6.22.obj~2/RELEASE_PPC
> Power Macintosh powerpc
> Unfortunatelly, neither symbolic nor hard link did the trick with gdb.
this is bug #1833 inside gdb
to get rid of it, I had to hack a function inside gdb like this;
imac-g5-de-olivier-kaloudoff:~/src/gdb-6.4 kalou$ vi gdb/rdi-share/unixcomm.c
extern const char *Unix_MatchValidSerialDevice(const char *name)
(replaced all the contents of the function with the lines below).
and now, rdi seems to be fine:
(gdb) tar rdi /dev/tty.KeySerial1
Angel Debug Monitor (serial) 1.04 (Advanced RISC Machines SDT 2.11a) for
Angel Debug Monitor rebuilt on Sep 14 1998 at 14:52:22
Serial Rate: 9600
Connected to ARM RDI target.
Note that I skizzed all parsing ... that's a really bad bug fix, but I'm
ready to use my board now using only open source stuff ;-)
tar rdi s=/dev/tty.KeySerial1
tar rdi /dev/tty.KeySerial1
with this dirty hack.
Hope I'll have the time to dig into this bug more throgoughly and post a
real bugfix to gdb.
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss