This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
Re: ARM - RAM region at other address than 0
- From: "Gary D. Thomas" <gary dot thomas at mind dot be>
- To: Grant Edwards <grante at visi dot com>
- Cc: "Doyle, Patrick" <WPD at dtccom dot com>,'Backhaus Willy' <w dot backhaus at newage-avkseg dot com>,eCos Discussion <ecos-discuss at sources dot redhat dot com>
- Date: 19 May 2003 14:51:04 -0600
- Subject: Re: [ECOS] ARM - RAM region at other address than 0
- Organization:
- References: <F18ED44D2778844592DFB732956B66516ED169@DTCNT40S4> <20030519153338.A2341@visi.com>
On Mon, 2003-05-19 at 14:33, Grant Edwards wrote:
> On Mon, May 19, 2003 at 04:11:38PM -0400, Doyle, Patrick wrote:
>
> > I glanced at the '5471 data manual and I don't see the
> > "mechanism for swapping ROM and SDRAM at address 0" that I was
> > expecting to see... Hmmm.... Perhaps you should ask TI about
> > this. eCos really likes to see interrupt vectors in RAM.
>
> At one point I had a hacked version of vectors.S that worked
> with ROM at 0. It did the "double vector" thing (not very
> pretty). After tiring of the requirement to maintain my own
> vectors.S, I finally swapped ROM/RAM -- life is easier that
> way.
>
> If the 5471 doesn't have some sort of software-controlled
> chip-select logic, then when you lay out your board you should
> add a couple gates to allow two of the chip selects to be
> swapped.
>
I personally think that using hardware to swap these segments is
the correct solution. However, if anyone wants to experiment
with Grant's approach (from *long* ago), I still have the patches
that he sent me for this :-)
--
Gary D. Thomas <gary.thomas@mind.be>
--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss