Re: [PATCH 2/2] soc: renesas: rz-sysc: Populate readable_reg/writeable_reg in regmap config

[Geert Uytterhoeven]

Hi Claudiu,

On Fri, 14 Nov 2025 at 12:46, Claudiu Beznea <claudiu.beznea@xxxxxxxxx> wrote:
> On 11/13/25 21:15, Geert Uytterhoeven wrote:
> > On Wed, 5 Nov 2025 at 08:05, Claudiu <claudiu.beznea@xxxxxxxxx> wrote:
> >> From: Claudiu Beznea <claudiu.beznea.uj@xxxxxxxxxxxxxx>
> >>
> >> Not all system controller registers are accessible from Linux. Accessing
> >> such registers generates synchronous external abort. Populate the
> >> readable_reg and writeable_reg members of the regmap config to inform the
> >> regmap core which registers can be accessed. The list will need to be
> >> updated whenever new system controller functionality is exported through
> >> regmap.
> >>
> >> Fixes: 2da2740fb9c8 ("soc: renesas: rz-sysc: Add syscon/regmap support")
> >> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@xxxxxxxxxxxxxx>
> >> ---
> >>
> >> Changes in v2:
> >> - added all SYSC registers IP specific, except the SPI
> >>   registers on RZ/V2H and RZ/V2N as these are accessible only from EL3
> >
> > Reviewed-by: Geert Uytterhoeven <geert+renesas@xxxxxxxxx>
> > i.e. will queue in renesas-devel for v6.19.
> >
> >> --- a/drivers/soc/renesas/r9a08g045-sysc.c
> >> +++ b/drivers/soc/renesas/r9a08g045-sysc.c
> >
> >> @@ -18,7 +37,57 @@ static const struct rz_sysc_soc_id_init_data rzg3s_sysc_soc_id_init_data __initc
> >>         .specific_id_mask = GENMASK(27, 0),
> >>  };
> >>
> >> +static bool rzg3s_regmap_readable_reg(struct device *dev, unsigned int reg)
> >> +{
> >> +       switch (reg) {
> >> +       case SYS_XSPI_MAP_STAADD_CS0:
> >> +       case SYS_XSPI_MAP_ENDADD_CS0:
> >> +       case SYS_XSPI_MAP_STAADD_CS1:
> >> +       case SYS_XSPI_MAP_ENDADD_CS1:
> >> +       case SYS_GETH0_CFG:
> >> +       case SYS_GETH1_CFG:
> >> +       case SYS_PCIE_CFG:
> >> +       case SYS_PCIE_MON:
> >> +       case SYS_PCIE_ERR_MON:
> >> +       case SYS_PCIE_PHY:
> >> +       case SYS_I2C0_CFG:
> >> +       case SYS_I2C1_CFG:
> >> +       case SYS_I2C2_CFG:
> >> +       case SYS_I2C3_CFG:
> >> +       case SYS_I3C_CFG:
> >> +       case SYS_USB_PWRRDY:
> >> +       case SYS_PCIE_RST_RSM_B:
> >> +               return true;
> >> +       default:
> >> +               return false;
> >> +       }
> >> +}
> >> +
> >> +static bool rzg3s_regmap_writeable_reg(struct device *dev, unsigned int reg)
> >> +{
> >> +       switch (reg) {
> >> +       case SYS_XSPI_MAP_STAADD_CS0:
> >> +       case SYS_XSPI_MAP_ENDADD_CS0:
> >> +       case SYS_XSPI_MAP_STAADD_CS1:
> >> +       case SYS_XSPI_MAP_ENDADD_CS1:
> >> +       case SYS_PCIE_CFG:
> >> +       case SYS_PCIE_PHY:
> >> +       case SYS_I2C0_CFG:
> >> +       case SYS_I2C1_CFG:
> >> +       case SYS_I2C2_CFG:
> >> +       case SYS_I2C3_CFG:
> >> +       case SYS_I3C_CFG:
> >> +       case SYS_USB_PWRRDY:
> >> +       case SYS_PCIE_RST_RSM_B:
> >> +               return true;
> >> +       default:
> >> +               return false;
> >> +       }
> >> +}
> >
> > As all the writeable regs are a subset of the readable regs, do you
> > think it would be worthwhile to write e.g.
> >
> >     static bool rzg3s_regmap_readable_reg(struct device *dev, unsigned int reg)
> >     {
> >             if (rzg3s_regmap_writeable_reg(dev, reg))
> >                     return true;
> >
> >             switch (reg) {
> >             case SYS_GETH0_CFG:
> >             case SYS_GETH1_CFG:
> >             case SYS_PCIE_MON:
> >             case SYS_PCIE_ERR_MON:
> >                     return true;
> >             default:
> >                     return false;
> >             }
> >     }
>
> Looks ok to me as well. I chose to have it like this as most of the
> readable/writeable function that I remember to have seen in the past were
> in the format I've presented in this patch.
>
> I noticed you already sent the PR. Would you prefer to return with a follow
> up patch and adjust it as you suggested?

Up to you, I haven't looked at the impact on code size 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