[PATCH 0/5] collected fallout of porting an ATSAMA5D27 based board
Alexander Dahl
ada at thorsis.com
Mon Oct 21 12:51:35 CEST 2024
Hello Benedikt,
Am Mon, Oct 21, 2024 at 12:20:17PM +0200 schrieb Benedikt Spranger:
> The SPL sits in the first Page in NAND flash. It does it's magic and
> then load the U-Boot payload from an UBI volume (Or directly load a
> kernel). Then it "starts" the payload.
>
> > Did you try reading/writing raw NAND from there?
> I try to read from SPL. That's all what ubispl wants. No need for write
> support here.
>
> > I'm not familiar with SPL, does it consider dts?
> It consider dts.
>
> > I could share my dts parts if that helps you.
> That would be great. I definitly give it a try.
> So here is a quick extract (EBI is also used otherwise) of what I did.
Using EBI for accessing an FPGA here on one of three boards.
So I'll share that dts excerpt from my v2024.10 branch.
> Hopefully I did not forget anything...
>
> --- a/arch/arm/dts/at91-sama5d27_ABC.dts
> +++ b/arch/arm/dts/at91-sama5d27_ABC.dts
> @@ -63,7 +63,62 @@
> bias-disable;
> bootph-all;
> };
> +
> + pinctrl_nand0_default: nand0_default {
> + pinmux =
> + <PIN_PA0__D0>,
> + <PIN_PA1__D1>,
> + <PIN_PA2__D2>,
> + <PIN_PA3__D3>,
> + <PIN_PA4__D4>,
> + <PIN_PA5__D5>,
> + <PIN_PA6__D6>,
> + <PIN_PA7__D7>,
> + <PIN_PA8__NWE_NANDWE>,
> + <PIN_PA9__NCS3>,
> + <PIN_PA10__A21_NANDALE>,
> + <PIN_PA11__A22_NANDCLE>,
> + <PIN_PA12__NRD_NANDOE>,
> + <PIN_PA21__NANDRDY>;
> + bias-disable;
> + };
> };
> };
> };
> };
> +
> +&ebi {
> + pinctrl-names = "default";
> + pinctrl-0 = <&pinctrl_nand0_default>;
> + status = "okay";
> +
> + nand_controller: nand-controller {
> + status = "okay";
> +
> + nand at 3 {
> + reg = <0x3 0x0 0x800000>;
> + //atmel,rb = <0>;
> + nand-bus-width = <8>;
> + nand-ecc-mode = "hw";
> + nand-ecc-strength = <4>;
> + nand-ecc-step-size = <512>;
> + nand-on-flash-bbt;
> +
> + partitions {
> + compatible = "fixed-partitions";
> + #address-cells = <1>;
> + #size-cells = <1>;
> +
> + spl at 0 {
> + label = "spl";
> + reg = <0x0 0x40000>;
> + };
> +
> + ubi at 40000 {
> + label = "UBI";
> + reg = <0x40000 0x0>;
> + };
> + };
> + };
> + };
> +};
I have no partitions defined, but that should not hurt. Split pinctrl
settings here, but that's basically because I have more child nodes on
the EBI node in Linux which I don't use in U-Boot here. Should not
make such a difference.
Do you have external pull-ups for the A21_NANDALE, A22_NANDCLE,
NANDRDY, and NCS3 lines? Maybe cross check that with the board code
you probably have for the old non-DM NAND driver.
Now my example:
&ebi {
pinctrl-names = "default";
pinctrl-0 = <&pinctrl_ebi_default>;
status = "okay";
nand_controller: nand-controller {
pinctrl-names = "default";
pinctrl-0 = <&pinctrl_nand_default>;
status = "okay";
nand at 3 {
reg = <0x3 0x0 0x800000>;
atmel,rb = <0>;
nand-bus-width = <8>;
nand-ecc-mode = "hw";
nand-on-flash-bbt;
nand-ecc-strength = <8>;
nand-ecc-step-size = <512>;
label = "atmel_nand";
};
};
};
&pioA {
pinctrl_ebi_default: ebi_default {
rd_we_data {
pinmux = <PIN_PA22__D0>,
<PIN_PA23__D1>,
<PIN_PA24__D2>,
<PIN_PA25__D3>,
<PIN_PA26__D4>,
<PIN_PA27__D5>,
<PIN_PA28__D6>,
<PIN_PA29__D7>,
<PIN_PA30__NWE_NANDWE>,
<PIN_PB2__NRD_NANDOE>;
bias-disable;
atmel,drive-strength = <ATMEL_PIO_DRVSTR_ME>;
};
};
pinctrl_nand_default: nand_default {
ale_cle_rdy_cs {
pinmux = <PIN_PB0__A21_NANDALE>,
<PIN_PB1__A22_NANDCLE>,
<PIN_PC8__NANDRDY>,
<PIN_PA31__NCS3>;
bias-pull-up;
};
};
};
Notice mine uses IO Set 1 here, but your's should work with IO Set 2 as
well.
> > (What I do here: SoC loads at91bootstrap from raw NAND at offset 0,
> > at91bootstrap loads U-Boot from raw NAND at some offset like 0x20000,
> > U-Boot (proper) loads everything else from UBI.)
> From my experience that can cause significant headache starting after
> 5-6 years in use. That's what I try to avoid :)
I'm aware of that. The data retention time for the NAND flash used
here is 10 years according to datasheet. No big deal on Linux when
running ubihealthd for everything stored in UBI, but everything
outside of that is about to fail some day. :-/
Greets
Alex
More information about the U-Boot
mailing list