STPC and Macronix Flash.

STPC and Macronix Flash.

[Fabrice Gautier]

Hi,

I have this cutom STPC board with this odd Macronix Flash (MX29F1610)
i'm trying to deal with. So i've tried to write an mtd driver for this.
For now the driver should just try to read the Device ID.

First problem is: In order to enable the flash i have to write in an
STPC register (a RAM controller register) which open a memory hole.

Then the flash should be mapped to 0x2a00000 so i'm doing an ioremap and
using the return value to access the flash. Just like it is done in
physmap.c and oher.

But I can't even get anything. The sample code from Macronix I'm using
gave good result an VxWorks and eCos. I personnally tested the code for
eCos so i'm quite sure i'm doing the right thing from the flash point of
view, but i'm thinking that maybe i get something wrong with the virtual
memory mapping. (There's no virtual memory stuff in eCos so it's very
simple)

And anyway is it really safe to manipulate the RAM controller in the mtd
driver? Didn't that would mess up with linux virtual memory managment?

Is there anyone doing things like that with an STPC ?

Thanks

-- 
Fabrice Gautier <gautier@email.enstfr>



To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: STPC and Macronix Flash.

[David Woodhouse]

gautier@email.enst.fr said:
>  Then the flash should be mapped to 0x2a00000 so i'm doing an ioremap
> and using the return value to access the flash. Just like it is done
> in physmap.c and oher.

Do what I say, not what I do :)

Try ioremap_nocache and __raw_{read,write}[bwl].

--
dwmw2




To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: STPC and Macronix Flash.

[Fabrice Gautier]

On Mon, 23 Apr 2001 10:20:38 +0100
David Woodhouse <dwmw2@infradead.org> wrote:
> 
> gautier@email.enst.fr said:
> >  Then the flash should be mapped to 0x2a00000 so i'm doing an ioremap
> > and using the return value to access the flash. Just like it is done
> > in physmap.c and oher.
> 
> Do what I say, not what I do :)
> 
> Try ioremap_nocache and __raw_{read,write}[bwl].

No success either. :-(

Thanks,

-- 
Fabrice Gautier <gautier@email.enstfr>



To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org