Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <zajec5@gmail.com>
To:     Rob Herring <robh+dt@kernel.org>,
        Srinivas Kandagatla <srinivas.kandagatla@linaro.org>,
        Michael Walle <michael@walle.cc>
Cc:     linux-mtd@lists.infradead.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
        netdev@vger.kernel.org, Miquel Raynal <miquel.raynal@bootlin.com>,
        Richard Weinberger <richard@nod.at>,
        Vignesh Raghavendra <vigneshr@ti.com>,
        Shawn Guo <shawnguo@kernel.org>, Li Yang <leoyang.li@nxp.com>,
        Frank Rowand <frowand.list@gmail.com>,
        "David S . Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>,
        Ansuel Smith <ansuelsmth@gmail.com>,
        Andrew Lunn <andrew@lunn.ch>,
        Florian Fainelli <f.fainelli@gmail.com>,
        Hauke Mehrtens <hauke@hauke-m.de>,
        =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <rafal@milecki.pl>
Subject: [PATCH 0/2] dt-bindings: nvmem: support describing cells
Date:   Tue, 25 Jan 2022 19:01:12 +0100
Message-Id: <20220125180114.12286-1-zajec5@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Rafał Miłecki <rafal@milecki.pl>

Michael has recently posted a cleaned up patchset for NVMEM
transformations support:
[PATCH 0/8] nvmem: add ethernet address offset support
https://lore.kernel.org/lkml/20211228142549.1275412-1-michael@walle.cc/T/
https://patchwork.ozlabs.org/project/linux-mtd/list/?series=278644&state=*

I find it very important & fully support it. In home routers we very
often deal with just one MAC address that:
1. Is a base for calculating multiple Ethernet addresses
2. Can be stored in binary as well as ASCII format

I'd like to suggest just a slightly different solution though. I think
that using something like:

otp-1 {
        compatible = "kontron,sl28-vpd", "user-otp";
        #address-cells = <1>;
        #size-cells = <1>;

        base_mac_address: base-mac-address@17 {
                #nvmem-cell-cells = <1>;
                reg = <17 6>;
        };
};

isn't clear enough and requires too much conditional code in Linux /
whatever implementation. DT doesn't make it clear which NVMEM cells
are used for what and how should be handled. That has to be hardcoded in
a Linux / whatever driver.

My idea is to add "compatible" & additional flags to NVMEM cells.
Example:

otp-1 {
        compatible = "user-otp";
        #address-cells = <1>;
        #size-cells = <1>;

        base_mac_address: base-mac-address@17 {
                compatible = "mac-address";
                reg = <17 6>;
                #nvmem-cell-cells = <1>;
        };
};

(for more examples see PATCH 2/2 and its mac-address.yaml .

Rafał Miłecki (2):
  dt-bindings: nvmem: extract NVMEM cell to separated file
  dt-bindings: nvmem: cells: add MAC address cell

 .../devicetree/bindings/nvmem/cells/cell.yaml | 35 +++++++
 .../bindings/nvmem/cells/mac-address.yaml     | 94 +++++++++++++++++++
 .../devicetree/bindings/nvmem/nvmem.yaml      | 25 +----
 3 files changed, 131 insertions(+), 23 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/nvmem/cells/cell.yaml
 create mode 100644 Documentation/devicetree/bindings/nvmem/cells/mac-address.yaml

-- 
2.31.1