RE: [ECOS] ecos redboot compile problem.

["Beymer, Mike" <Mike dot Beymer at itron dot com>]

I'm still unable to get the same results as Patrick and Gary. Here's a list of things tried.
  - reinstall ecos and tool chains.
  - build tool chain from scratch.
  - looked at hex dumps of img file and found same byte order at same offset for the 
     12345678 87654321 pattern.
  - looked at the source for this pattern. According to a comment in source the 12345678 and
     87654321 when loaded into the first 2 words of sram is a trick to get the reset to branch 
     to the 3rd location in sram. 
  - 12345678 decodes into a valid arm instruction. "eornes  r5, r4, #125829120"  87654321
      decodes into a valid instruction as well. Maybe the better question to ask is why
      isn't this code being looked at by obj-dump -d as an instruction in Patrick and Gary's' 
      generated code. Is there somewhere that this "trick" is documented outside of ECOS?
     
     Any ideas?    

-Thanks Mike Beymer

-----Original Message-----
From: Gary D. Thomas [mailto:gary dot thomas at mind dot be]
Sent: Friday, April 18, 2003 1:54 AM
To: Beymer, Mike
Cc: eCos Discussion
Subject: RE: [ECOS] ecos redboot compile problem.


On Thu, 2003-04-17 at 16:01, Beymer, Mike wrote:
> I'm building on Linux RH 8.0. I went on vacation and can't remember exactly where the toolchains came from. The version information from arm-elf-gcc returns 
> 
> arm-elf-gcc (GCC) 3.2.1 (eCosCentric)

Then you got these from the eCos 2.0 Beta distribution.

> 
> Do you know which tool chain you are using? Maybe I need to replace mine.

I build my own toolchains, but the ones that you've downloaded
are known to work fine.

What commands did you use to build?  Here's what I did that worked:
  % ecosconfig new innovator redboot
  % ecosconfig import /work2/ecos/packages/hal/arm/arm9/innovator/current/misc/redboot_SRAM.ecm
  % ecosconfig tree
  % make
(my eCos repository is stored at "/work2/ecos/packages")

>  -Thanks
>     -Mike Beymer 
> 
> -----Original Message-----
> From: Gary D. Thomas [mailto:gary dot thomas at mind dot be]
> Sent: Thursday, April 17, 2003 2:18 PM
> To: Beymer, Mike
> Cc: eCos Discussion
> Subject: Re: [ECOS] ecos redboot compile problem.
> 
> 
> On Thu, 2003-04-17 at 14:21, Beymer, Mike wrote:
> > I'm working on a port of redboot to the PSI Minno OMAP5910 PCA. Which is very similar
> > to the PSI Innovator. In trying to verify my build process, I am attempting to rebuild Patrick Doyle's SRAM image for the innovator. ( his works mine doesn't). In looking at an objdump -d
> > of Patrick's object file compared to mine, it appears to have a problem with endian'ism
> > Patrick's looks as follows:
> > 
> > redboot-sram.out:     file format elf32-littlearm
> > 
> > Disassembly of section .rom_vectors:
> > 
> > 20000000 <__rom_vectors_lma>:
> > 20000000:       78 56 34 12 21 43 65 87                             xV4.!Ce.
> > 
> > 20000008 <__exception_handlers>:
> > 
> > 
> > Mine comes out as: 
> > 
> > redboot.img:     file format elf32-littlearm
> > 
> > Disassembly of section .rom_vectors:
> > 
> > 20000000 <__rom_vectors_lma>:
> > 20000000:       12345678        eornes  r5, r4, #125829120      ; 0x7800000
> > 20000004:       87654321        strhib  r4, [r5, -r1, lsr #6]!
> > 
> > 20000008 <__exception_handlers>:
> > 
> > The same numbers but a different order. I've tried playing with the endian switches in the 
> > compiler with no change in the results. I stumped, Patrick sent me what he remembers doing to build the file, but that didn't work. Any ideas??
> >  -Mike Beymer
> 
> Must be something with your tools.  I just built this and I
> got identical results to Patrick's.
> 
> What host are you using?
> Where did you get your toolchain from?
> 
> -- 
> .--------------------------------------------------------.
> |       Mind: Embedded Linux and eCos Development        |
> |--------------------------------------------------------|
> | Gary Thomas              email:  gary dot thomas at mind dot be   |
> | Mind ( http://mind.be )  tel:    +1 (970) 229-1963     |
> | gpg: http://www.chez-thomas.org/gary/gpg_key.asc       |
> '--------------------------------------------------------'
> 
> 
> 
> 
> 
> This message was scanned for viruses.
> 
> 
-- 
.--------------------------------------------------------.
|       Mind: Embedded Linux and eCos Development        |
|--------------------------------------------------------|
| Gary Thomas              email:  gary dot thomas at mind dot be   |
| Mind ( http://mind.be )  tel:    +1 (970) 229-1963     |
| gpg: http://www.chez-thomas.org/gary/gpg_key.asc       |
'--------------------------------------------------------'





This message was scanned for viruses.




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