Re: [ECOS] vectors.s

["Rupesh S" <rupeshs at myw dot ltindia dot com>]

thanx for the info...
As u said, the content of _reset_platform is data (the address of
__reset_platform , which is in vectors.S , and does the hardware inits...).
May be the dissambler would have interpreted the address as a match of the
bra instruction. I was reading that fluke code.
Good leason not to rely on the disassembled code....!!!

I have yet another doubt regarding the same...
if the startup is ROM , then the data there is address of
__reset_platform+0x20000000, why is it done like that...??


----- Original Message -----
From: "Jonathan Larmour" <jifl at eCosCentric dot com>
To: "Rupesh S" <rupeshs at myw dot ltindia dot com>
Cc: "Andrew Lunn" <andrew dot lunn at ascom dot ch>; <ecos-discuss at sources dot redhat dot com>
Sent: Wednesday, April 16, 2003 2:37 PM
Subject: Re: [ECOS] vectors.s


> Rupesh S wrote:
> > I din't feel that the code is linked for Position independency, 'coz the
> > jump and bra instructions in the dissambled code seems to have 4 byte
full
> > address operands.
>
>
> Looking at the memory layout of the SE77x9 HAL, it should be linked to run
> at 0x80000000 anyway, not 0x0.
>
> I suggest you look at and read the comments in
> hal/sh/arch/current/include/hal_var_bank.inc. From that file it seems that
> what is at _reset_platform is data, not code. Although presumably due to
> some fluke of opcodes, it looks like one. Instead it looks like it's
> jumping to the umapped shadow address of the function __reset_platform
> (note the underscores) which is defined in vectors.S.
>
> HTH,
>
> Jifl
>
> > ----- Original Message -----
> > From: "Andrew Lunn" <andrew dot lunn at ascom dot ch>
> > To: "Rupesh S" <rupeshs at myw dot ltindia dot com>
> > Cc: <ecos-discuss at sources dot redhat dot com>
> > Sent: Wednesday, April 16, 2003 1:56 AM
> > Subject: Re: [ECOS] vectors.s
> >
> >
> >
> >>On Mon, Apr 16, 2001 at 02:28:21PM -0700, Rupesh S wrote:
> >>
> >>>I looked in the dissambled code generated using sh-elf-objdump,
> >>
> >>The code will be position independent. It might looked to be linked to
> >>run at 0x0, but since its position independent, it will run OK at the
> >>flash high address 0x80000000.
> >>
> >>      Andrew
> >>
> >>
> >>>----- Original Message -----
> >>>From: "Andrew Lunn" <andrew dot lunn at ascom dot ch>
> >>>To: "Rupesh S" <rupeshs at myw dot ltindia dot com>
> >>>Cc: <ecos-discuss at sources dot redhat dot com>
> >>>Sent: Wednesday, April 16, 2003 1:00 AM
> >>>Subject: Re: [ECOS] vectors.s
> >>>
> >>>
> >>>
> >>>>On Mon, Apr 16, 2001 at 01:35:26PM -0700, Rupesh S wrote:
> >>>>
> >>>>>My architecture is sh3 processor ,7729.
> >>>>>As you said, Flash is mapped at 0x0 & 0x80000000.
> >>>>>But what I don't understand is that, when it jumps to the High
> >
> > memory,
> >
> >>>>>(0x8000000e in my case), shouldn't there  be some or the other pice
> >
> > of
> >
> >>>code
> >>>
> >>>>>attached there,
> >>>>>But I'm not seeing any instruction placed at that address.
> >>>>
> >>>>How are you looking?
> >>>>
> >>>>    Andrew
> >>>
> >
> >
>
>
> --
> eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
> --[ "You can complain because roses have thorns, or you ]--
> --[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine
>


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