| Message ID | 2c4d00dd40124c2ddc0b139cbce7531b108f9052.1686238550.git.chunkeey@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2,1/3] ARM: dts: BCM5301X: MR26: MR32: remove bogus nand-ecc-algo property | expand |
From: Florian Fainelli <f.fainelli@gmail.com> On Thu, 8 Jun 2023 17:36:27 +0200, Christian Lamparter <chunkeey@gmail.com> wrote: > | bcm53015-meraki-mr26.dtb: nand-controller@18028000: > | nand@0:nand-ecc-algo:0: 'hw' is not one of ['hamming', 'bch', 'rs'] > | From schema: Documentation/[...]/nand-controller.yaml > | bcm53016-meraki-mr32.dtb: nand-controller@18028000: > | nand@0:nand-ecc-algo:0: 'hw' is not one of ['hamming', 'bch', 'rs'] > | From schema: Documentation/[...]/nand-controller.yaml > > original ECC values for these old Merakis are sadly not > provided by the vendor. It looks like Meraki just stuck > with what Broadcom's SDK was doing... which left this > up to the proprietary nand driver. > > Note: The invalid setting was and is handled by brcmnand. It > falls back to "bch" in brcmnand_setup_dev() when ecc.algo is > set to NAND_ECC_ALGO_UNKNOWN (since "hw" is not in the list > above). > > A correct nand-ecc-algo = "bch"; is already specified in the > included "bcm5301x-nand-cs0-bch8.dtsi". So this line can be > dropped. > > Reported-by: Rafał Miłecki <zajec5@gmail.com> (per Mail) > Fixes: 935327a73553 ("ARM: dts: BCM5301X: Add DT for Meraki MR26") > Fixes: ec88a9c344d9 ("ARM: BCM5301X: Add DT for Meraki MR32") > Signed-off-by: Christian Lamparter <chunkeey@gmail.com> > --- Applied to https://github.com/Broadcom/stblinux/commits/devicetree/next, thanks! -- Florian
diff --git a/arch/arm/boot/dts/bcm53015-meraki-mr26.dts b/arch/arm/boot/dts/bcm53015-meraki-mr26.dts index 071f2cb97251..83d1b6e0b0d5 100644 --- a/arch/arm/boot/dts/bcm53015-meraki-mr26.dts +++ b/arch/arm/boot/dts/bcm53015-meraki-mr26.dts @@ -72,8 +72,6 @@ &gmac3 { }; &nandcs { - nand-ecc-algo = "hw"; - partitions { compatible = "fixed-partitions"; #address-cells = <0x1>; diff --git a/arch/arm/boot/dts/bcm53016-meraki-mr32.dts b/arch/arm/boot/dts/bcm53016-meraki-mr32.dts index 46c2c93b01d8..16c650595cb9 100644 --- a/arch/arm/boot/dts/bcm53016-meraki-mr32.dts +++ b/arch/arm/boot/dts/bcm53016-meraki-mr32.dts @@ -125,8 +125,6 @@ &pwm { }; &nandcs { - nand-ecc-algo = "hw"; - partitions { /* * The partition autodetection does not work for this device.
| bcm53015-meraki-mr26.dtb: nand-controller@18028000: | nand@0:nand-ecc-algo:0: 'hw' is not one of ['hamming', 'bch', 'rs'] | From schema: Documentation/[...]/nand-controller.yaml | bcm53016-meraki-mr32.dtb: nand-controller@18028000: | nand@0:nand-ecc-algo:0: 'hw' is not one of ['hamming', 'bch', 'rs'] | From schema: Documentation/[...]/nand-controller.yaml original ECC values for these old Merakis are sadly not provided by the vendor. It looks like Meraki just stuck with what Broadcom's SDK was doing... which left this up to the proprietary nand driver. Note: The invalid setting was and is handled by brcmnand. It falls back to "bch" in brcmnand_setup_dev() when ecc.algo is set to NAND_ECC_ALGO_UNKNOWN (since "hw" is not in the list above). A correct nand-ecc-algo = "bch"; is already specified in the included "bcm5301x-nand-cs0-bch8.dtsi". So this line can be dropped. Reported-by: Rafał Miłecki <zajec5@gmail.com> (per Mail) Fixes: 935327a73553 ("ARM: dts: BCM5301X: Add DT for Meraki MR26") Fixes: ec88a9c344d9 ("ARM: BCM5301X: Add DT for Meraki MR32") Signed-off-by: Christian Lamparter <chunkeey@gmail.com> --- v1 -> v2: - add dts and BCM5301X tag (Rafael) - Rebased on top of stblinux devicetree/next - investigated device NAND OOB for clues what is used. => also checked out brcmnand.c --- arch/arm/boot/dts/bcm53015-meraki-mr26.dts | 2 -- arch/arm/boot/dts/bcm53016-meraki-mr32.dts | 2 -- 2 files changed, 4 deletions(-)