[2/2] memory: omap-gpmc: expand the description of the debug facility

Most register values for the chip select setup depend on the frequency
of the fck clock.
So add a hint that the values setup by the bootloader might differ from
the right setup for Linux if the bootloader uses a different frequency.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
---
 drivers/memory/Kconfig | 3 +++
 1 file changed, 3 insertions(+)

On 06/10/15 23:07, Uwe Kleine-König wrote:
> Most register values for the chip select setup depend on the frequency
> of the fck clock.
> So add a hint that the values setup by the bootloader might differ from
> the right setup for Linux if the bootloader uses a different frequency.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> ---
>  drivers/memory/Kconfig | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/memory/Kconfig b/drivers/memory/Kconfig
> index c6a644b22af4..1414dd53be57 100644
> --- a/drivers/memory/Kconfig
> +++ b/drivers/memory/Kconfig
> @@ -64,6 +64,9 @@ config OMAP_GPMC_DEBUG
>  	  Enables verbose debugging mostly to decode the bootloader provided
>  	  timings. Enable this during development to configure devices
>  	  connected to the GPMC bus.
> +	  Note that you cannot just tweak your device tree until the registers
> +	  setup by linux match what the bootloader did because that one might
> +	  use a different fck frequency influencing most register settings.

Looks like we can't know for sure the GPMC fclk used at the bootloader
else we could have just printed the GPMC fclk pre and post gpmc settings.

How about this instead?

NOTE: Apart from matching the register setup with the bootloader you also need to
match the GPMC FCLK frequency used by the bootloader else the GPMC timings
won't be identical with the bootloader timings.

Also you might need to build this patch on top of
http://article.gmane.org/gmane.linux.kernel/2054796

>  
>  config MVEBU_DEVBUS
>  	bool "Marvell EBU Device Bus Controller"
> 

cheers,
-roger

Hello Roger,

On Wed, Oct 07, 2015 at 10:45:50AM +0300, Roger Quadros wrote:
> On 06/10/15 23:07, Uwe Kleine-König wrote:
> > Most register values for the chip select setup depend on the frequency
> > of the fck clock.
> > So add a hint that the values setup by the bootloader might differ from
> > the right setup for Linux if the bootloader uses a different frequency.
> > 
> > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> > ---
> >  drivers/memory/Kconfig | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/drivers/memory/Kconfig b/drivers/memory/Kconfig
> > index c6a644b22af4..1414dd53be57 100644
> > --- a/drivers/memory/Kconfig
> > +++ b/drivers/memory/Kconfig
> > @@ -64,6 +64,9 @@ config OMAP_GPMC_DEBUG
> >  	  Enables verbose debugging mostly to decode the bootloader provided
> >  	  timings. Enable this during development to configure devices
> >  	  connected to the GPMC bus.
> > +	  Note that you cannot just tweak your device tree until the registers
> > +	  setup by linux match what the bootloader did because that one might
> > +	  use a different fck frequency influencing most register settings.
> 
> Looks like we can't know for sure the GPMC fclk used at the bootloader
> else we could have just printed the GPMC fclk pre and post gpmc settings.
> 
> How about this instead?
> 
> NOTE: Apart from matching the register setup with the bootloader you also need to
> match the GPMC FCLK frequency used by the bootloader else the GPMC timings
> won't be identical with the bootloader timings.
Yeah, sounds better, thanks.

> Also you might need to build this patch on top of
> http://article.gmane.org/gmane.linux.kernel/2054796
I talked to Tony about this patch yesterday on irc, but I didn't find it
in the archives yet when I sent my mail.

Thanks
Uwe