Hi Julien,
On Wed, Sep 22, 2021 at 11:54 AM Julien Massot <julien.massot@xxxxxxx> wrote:
> > In general, I think this looks like a good abstraction, which we can
> > also use for further abstraction of R-Car Gen2 (cfr. [1]).
> Yes I was also thinking about future generation like Gen4, but I don't have the documentation
> at this point.
> From what I understand in [1]: CA7BAR in Gen2 is managed by the SYSC module and not by the RST module
> as for CR7BAR in Gen3. So despite the fact that the procedure is similar, we won't be able to set CA7BAR in
> rcar-rst.c.
On R-Car Gen2, CA7BAR is managed by the RST module.
On R-Mobile APE6, it is managed by the SYSC module.
> > I'm just wondering if we should pass the BAR offset to
> > rcar_rst_set_rproc_boot_addr() explicitly (and rename the function),
> > or have multiple functions for the different BARs.
> Passing BAR offset will imply to spread RST module offsets across different subsystems,
> and the second question will be to be able to do the correct boundary check for the targeted
> processor: CR7BAR is aligned on 4k an it looks like CA7BAR is on 256k. It looks like it's
> manageable thanks to the driver data which already holds the 'mode monitor register' offset per SoC generation.
>
> One missing point will be for future R-Car SoCs: trends on others SoC seems to be to go to have
> several 'realtime', or remote processor. In this case this function will not scale up.
On R-Car V3U, CR52BAR is managed by the APMU module.
There's not much we can do about future processors yet ;-)
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
