[ECOS] Problems with building Redboot for the iq80321

John Clark j1clark@ucsd.edu
Mon Apr 14 14:06:00 GMT 2003


Am Montag, 14.04.03, um 02:12 Uhr (US/Pacific) schrieb Nick Garnett:

> John Clark <j1clark@ucsd.edu> writes:
>
>> When I attempt to use the ecosconfig built from the recent cvs'd host
>> sources,

Here's my setup script that has worked for the previous versions of
ecosconfig with older Redboot sources. (This works with the appropriate
mods for the iq80310 or 321 and Redboot sources vintage early 2002).

#!/bin/bash

export TOPDIR=`pwd`
export ECOS_REPOSITORY=${TOPDIR}/packages
mkdir ${TOPDIR}/build
cd ${TOPDIR}/build
/usr/local/bin/ecosconfig new iq80321 redboot
/usr/local/bin/ecosconfig import 
${ECOS_REPOSITORY}/hal/arm/xscale/iq80321/current/misc/redboot_ROM.ecm
/usr/local/bin/ecosconfig tree

(I noted that in your e-mail you have a line:

ecosconfig import

This just gives an help listing for me.


>> I get an error on:
>>
>> ecosconfig tree.
>
> It would help if you told us what the error was.

The error is pretty 'unhinful', but here's the gdb back trace when I
run ecosconfig manually with the 'tree' option in the 'build' directory 
created above.

Here's the output from the script above:

Setup_iq80321
U CYGHWR_REDBOOT_ARM_LINUX_EXEC_ADDRESS_DEFAULT, new inferred value 
0xA0008000
U CYGSEM_HAL_USE_ROM_MONITOR, new inferred value 0
./Setup_iq80321: line 9:   488 Segmentation fault      (core dumped) 
/usr/local/bin/ecosconfig tree


Here's the gdb back trace from the last command 'ecosconfig tree", with 
ECOS_REPOSITORY setup
and in the 'build' directory:

[root@linux-pc ecos]# source Setup_Env
[root@linux-pc ecos]# cd build
[root@linux-pc build]# gdb /usr/local/bin/ecosconfig
GNU gdb 5.0
Copyright 2000 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 "i386-redhat-linux"...
(gdb) run tree
Starting program: /usr/local/bin/ecosconfig tree

Program received signal SIGSEGV, Segmentation fault.
0x80c54d1 in basic_string<char, string_char_traits<char>, 
__default_alloc_template<true, 0> >::Rep::grab ()
(gdb) bt
#0  0x80c54d1 in basic_string<char, string_char_traits<char>, 
__default_alloc_template<true, 0> >::Rep::grab ()
#1  0x80c5144 in basic_string<char, string_char_traits<char>, 
__default_alloc_template<true, 0> >::basic_string ()
#2  0x80cac80 in CdlBuildInfo_Loadable::CdlBuildInfo_Loadable ()
#3  0x80ca66b in void construct<CdlBuildInfo_Loadable, 
CdlBuildInfo_Loadable> ()
#4  0x80e31aa in vector<CdlBuildInfo_Loadable, 
allocator<CdlBuildInfo_Loadable> >::_M_insert_aux ()
#5  0x80e26a2 in vector<CdlBuildInfo_Loadable, 
allocator<CdlBuildInfo_Loadable> >::push_back ()
#6  0x80930bc in CdlBuildLoadableBody::update_build_info ()
#7  0x809754a in CdlToplevelBody::get_build_info ()
#8  0x80565a6 in generate_build_tree ()
#9  0x805052a in cdl_exec::cmd_tree ()
#10 0x804d855 in main ()
#11 0x40132b65 in __libc_start_main (main=0x804caec <main>, argc=2, 
ubp_av=0xbffffa14, init=0x804c44c <_init>, fini=0x80f5d1c <_fini>,
     rtld_fini=0x4000df24 <_dl_fini>, stack_end=0xbffffa0c) at 
../sysdeps/generic/libc-start.c:111
(gdb)



The version of ecosconfig was cvs'd from the source.redhat.com archive 
this weekend.

The --version option gives:

[root@linux-pc build]# /usr/local/bin/ecosconfig --version
ecosconfig 2.net (Apr 14 2003 08:28:24)
Copyright (c) 2002 Red Hat, Inc.

Also my older ecosconfig does not work on the current cvs soruces, it 
bombs early on
with some sort of syntax error in parsing  the cdls at some point.

Perhaps if there is a known 'working' snapshot of the source I could 
see if there is just
some stupid error in the cvs'd images.

Thank You
John Clark


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss



More information about the Ecos-discuss mailing list