Re: [ECOS] How to handle it???

[Gary Thomas <gary at mlbassoc dot com>]

On Tue, 2003-12-02 at 18:25, zheng wujun wrote:
> Thanks RandyLin:
>        Now my redboot can work,and I can use redboot command "load ......-m 
> tftp....."to download app to snds board.
>        Howerer, if not using -r option,the error as displayed "***Abort! 
> Attempt to load ELF data to address: 0x00020000 which is not in RAM".
>        If using  -r option,app looks like to be downloaded correctly,but the 
> display is "Raw file loaded 0x00300000-0x003c6116, assumed entry at 
> 0x00300000",and then if  using redboot command "go 0x00300000" to execute 
> the app,the redboot will display some strange strings as
> " 
> $T0a0f:86000000;0d:80060200;#17$T0a0f:86000000;0d:80060200;#17$T0a0f:8600"on 
> console.

When you startup RedBoot, what does it tell you the available RAM 
regions are?  For example, on this board:
  RAM: 0x00000000-0x00800000, [0x00038de8-0x007ec000] available
The board has 8MB of RAM, starting at 0x00000000, but RedBoot uses all of
the memory up to 0x00038de8.  So, for this board, eCos programs are linked
to start at 0x40000.

Note: eCos programs are typically *not* position independent.  If the program
is linked to load & run at 0x20000, that's the only location in RAM where
you can put it.  Your example above where you overrode this and put the
code at 0x30000 will never work.

Loading a file with the "-r" option tells RedBoot not to interpret the contents
in any way.  For an ELF file, this is deadly, as you'll be trying to execute
the ELF binary file header which is just a set of illegal ARM instructions!

> 
> to above,the config is SECTION_fixed_vectors (ram, 0x20, LMA_EQ_VMA)   and 
> SECTION_rom_vectors (ram, 0x20000, LMA_EQ_VMA) in file mlt_arm_snds_ram.ldi 
> and  SECTION_fixed_vectors (ram, 0x20, LMA_EQ_VMA)  and  SECTION_data (ram, 
> 0x20000, FOLLOWING (.gcc_except_table)) in mlt_arm_snds_rom.ldi
> 
> If using arm-elf-gdb to download app,the steps is as following:
> 1 [root@zheng root]#arm-elf-gdb
> 2(gdb) target remote /dev/ttyS0
> Remote debugging using /dev/ttyS0
> 0x01804b30 in ?? ()
> (gdb) load /tftpboot/hello
> Loading section .rom_vectors, size 0x40 lma 0x20000
> Loading section .text, size 0xfe00 lma 0x20040
> Ignoring packet error, continuing...
> Ignoring packet error, continuing...
> Ignoring packet error, continuing...
> 
> and gdb use the same serial line as the hyperspatial,the baudrate is 38400.
> 
> 
> Why did these error happen?Can some one have good idea to handle it?
> 
> Thanks a lot!
> 
> _________________________________________________________________
> Help STOP SPAM with the new MSN 8 and get 2 months FREE*  
> http://join.msn.com/?page=features/junkmail
-- 
Gary Thomas <gary@mlbassoc.com>
MLB Associates


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