[patch] atavia

[Jonathan A. Kollasch]

Add VIA VT6421A LPC programmer driver.

Signed-off-by: Jonathan Kollasch

[Paul Menzel]

Dear Jonathan,

Am Sonntag, den 26.06.2011, 23:33 +0000 schrieb Jonathan A. Kollasch:
> VIA VT6421A LPC programmer

What is atavia? Is that an external programmer? If yes, please post a
link with more information and add that too.

Thanks,

Paul

[Stefan Tauner]

On Mon, 27 Jun 2011 02:04:26 +0200
Paul Menzel  wrote:
> Dear Jonathan,
>
> Am Sonntag, den 26.06.2011, 23:33 +0000 schrieb Jonathan A. Kollasch:
> > VIA VT6421A LPC programmer
>
> What is atavia? Is that an external programmer? If yes, please post a
> link with more information and add that too.

ata via.. named after atahtp.c i guess.

it is a sata pci card see
http://www.via.com.tw/en/products/peripherals/serial-ata_raid/vt6421a/

[Jonathan A. Kollasch]

On Sun, Jun 26, 2011 at 11:33:33PM +0000, Jonathan A. Kollasch wrote:
> Add VIA VT6421A LPC programmer driver.

Okay, so it still doesn't work.  The 6421 is being clever, and coping
with that is really something I don't care to do, at least right now.

Jonathan Kollasch

[Tadas Slotkus]

> Okay, so it still doesn't work.  The 6421 is being clever, and coping
> with that is really something I don't care to do, at least right now.
>
> Jonathan Kollasch

Hi, thanks for the patch. Could you tell us more about the problems you
encountered? I may have some time after summer to play with that
programmer, so some info would be appreciated :)
> +static bool atavia_ready(void)
> +{
> int try;
> uint8_t access;
> bool status;
> for (status = false, try = 0; try < 300; try++) {
> access = pci_read_byte(pcidev_dev, BROM_ACCESS);
> if ((access & BROM_TRIGGER) == 0) {
> 	status = true;
> 	break;
> } else {
> 	programmer_delay(1);
> 	continue;
> }
> }
> msg_pdbg("%s() %s after %d tries\n", __FUNCTION__,
>  status ? "suceeded" : "failed", try);
> return status;
>
> +}

Is 'continue' required? msg_pdbg should be printed only if failed or
should be spew? (generates 3*Chip_size_in_bytes number of lines in
writing at least)

What about bool type usage in this project? We are using integers
here :)

Thanks,
Tadas

[Jonathan A. Kollasch]

On Wed, Jul 27, 2011 at 02:03:10AM +0300, Tadas Slotkus wrote:
> > Okay, so it still doesn't work.  The 6421 is being clever, and coping
> > with that is really something I don't care to do, at least right now.
> >
> > Jonathan Kollasch
>
> Hi, thanks for the patch. Could you tell us more about the problems you
> encountered? I may have some time after summer to play with that
> programmer, so some info would be appreciated :)

Doesn't work with some chips.  Seems 1MiB chips seem to work almost
universally, but not smaller ones.
> Is 'continue' required?

No.
> msg_pdbg should be printed only if failed or
> should be spew? (generates 3*Chip_size_in_bytes number of lines in
> writing at least)

Dunno, it hasn't bothered me yet.
> What about bool type usage in this project? We are using integers
> here :)

Uh, I could care less, I sort of assumed we were C99.

Jonathan Kollasch