Hi Nico!

Thanks for explaining. :) So what can I do in this case? Which programmer
is the easiest to get the setup working? Rayer_SPI didn't even detect it,
nor with shorter wires, this is why I went to ICH7, at least it can
recognize KB9012, and read some random data from it. I will not connect
both programmers at the same time, but I have a theory Rayer_SPI doesn't
give SPI clock signal continuously, thus KB9012 didn't enter EDI mode.

Nico Huber <[email protected]> ezt írta (időpont: 2019. jan. 20., V, 13:39):

> Hi Nyúlós,
>
> On 20.01.19 12:46, Nyúlós Lónyál wrote:
> > There are some programmer_delay values in ichspi.c, here is one which
> says
> > something about frequency:
> >
> > timeout = 100 * 60; /* 60 ms are 9.6 million cycles at 16 MHz. */
> > while ((REGREAD16(ICH7_REG_SPIS) & SPIS_SCIP) && --timeout) {
> > programmer_delay(10);
> >
> > So I just need to increase programmer_delay, for example to 20 and maybe
> > the timeout to 200 * 60?
>
> no, this has nothing to do with the transmission speed, it's really just
> a polling loop with a timeout. Generally, the ichspi (both hardware and
> driver) expect a perfect connection to the SPI slave... plus the hard-
> ware might make assumptions about the type of commands used. It's not
> meant to work with your setup, so you need a lot of luck.
>
> >
> > I can try another method too as the KB9012 datasheet says "To enable EDI,
> > it is by detecting any SPI command with EDI_CLK frequency between 1MHz
> to 8
> > MHz.",
>
> According to its datasheet the ICH7 runs its SPI clock at 17.9MHz. It's
> not completely hopeless, though.
>
> > so maybe after probing with flashrom in DIP8 socket, and connecting
> > it to rayer_spi LPT it will detect and flash KB9012 correctly with lower
> > SPI frequency.
>
> Sounds like a fragile procedure, make sure that you never have the ICH
> and Rayer connected at the same time. Also, what Mike said about the
> wire length applies to your setup with the Rayer, too.
>
> Nico
>
_______________________________________________
flashrom mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to