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;
MIME-Version: 1.0
References: <20201215164315.3666-1-calvin.johnson@oss.nxp.com>
 <20201215164315.3666-11-calvin.johnson@oss.nxp.com> <X9jrrMJIj2EQBykI@pendragon.ideasonboard.com>
In-Reply-To: <X9jrrMJIj2EQBykI@pendragon.ideasonboard.com>
From:   Andy Shevchenko <andy.shevchenko@gmail.com>
Date:   Tue, 15 Dec 2020 19:49:14 +0200
Message-ID: <CAHp75VdA9Qzvg0TqOes1B1PXF+z8YMM2hK1zOJYMZusmHz=hdQ@mail.gmail.com>
Subject: Re: [net-next PATCH v2 10/14] device property: Introduce fwnode_get_id()
To:     Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
        Rob Herring <robh+dt@kernel.org>
Cc:     Calvin Johnson <calvin.johnson@oss.nxp.com>,
        Grant Likely <grant.likely@arm.com>,
        "Rafael J . Wysocki" <rafael@kernel.org>,
        Jeremy Linton <jeremy.linton@arm.com>,
        Andrew Lunn <andrew@lunn.ch>,
        Florian Fainelli <f.fainelli@gmail.com>,
        Russell King - ARM Linux admin <linux@armlinux.org.uk>,
        Cristi Sovaiala <cristian.sovaiala@nxp.com>,
        Florin Laurentiu Chiculita <florinlaurentiu.chiculita@nxp.com>,
        Ioana Ciornei <ioana.ciornei@nxp.com>,
        Madalin Bucur <madalin.bucur@oss.nxp.com>,
        Heikki Krogerus <heikki.krogerus@linux.intel.com>,
        Marcin Wojtas <mw@semihalf.com>,
        Pieter Jansen Van Vuuren <pieter.jansenvv@bamboosystems.io>,
        Jon <jon@solid-run.com>, "linux.cj" <linux.cj@gmail.com>,
        Laurentiu Tudor <laurentiu.tudor@nxp.com>,
        Diana Madalina Craciun <diana.craciun@nxp.com>,
        ACPI Devel Maling List <linux-acpi@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        linux-arm Mailing List <linux-arm-kernel@lists.infradead.org>,
        netdev <netdev@vger.kernel.org>,
        Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
        Bartosz Golaszewski <bgolaszewski@baylibre.com>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>,
        Randy Dunlap <rdunlap@infradead.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Dec 15, 2020 at 7:00 PM Laurent Pinchart
<laurent.pinchart@ideasonboard.com> wrote:
> On Tue, Dec 15, 2020 at 10:13:11PM +0530, Calvin Johnson wrote:
> > Using fwnode_get_id(), get the reg property value for DT node
> > and get the _ADR object value for ACPI node.

...

> > +/**
> > + * fwnode_get_id - Get the id of a fwnode.
> > + * @fwnode: firmware node
> > + * @id: id of the fwnode
>
> Is the concept of fwnode ID documented clearly somewhere ? I think this
> function should otherwise have more documentation, at least to explain
> what the ID is.

I'm afraid that OF has no clear concept of this either. It's usually
used as a unique ID for the children of some device (like MFD) and can
represent a lot of things. But I agree it should be documented.

Rob, any ideas about this?

> > + * Returns 0 on success or a negative errno.
> > + */
> > +int fwnode_get_id(struct fwnode_handle *fwnode, u32 *id)
> > +{
> > +     unsigned long long adr;
> > +     acpi_status status;
> > +
> > +     if (is_of_node(fwnode)) {
> > +             return of_property_read_u32(to_of_node(fwnode), "reg", id);
> > +     } else if (is_acpi_node(fwnode)) {
> > +             status = acpi_evaluate_integer(ACPI_HANDLE_FWNODE(fwnode),
> > +                                            METHOD_NAME__ADR, NULL, &adr);
> > +             if (ACPI_FAILURE(status))
> > +                     return -ENODATA;
>
> Would it make sense to standardize error codes ? of_property_read_u32()
> can return -EINVAL, -ENODATA or -EOVERFLOW. I don't think the caller of
> this function would be very interested to tell those three cases apart.
> Maybe we should return -EINVAL in all error cases ? Or maybe different
> error codes to mean "the backend doesn't support the concept of IDs",
> and "the device doesn't have an ID" ?

We may actually get mapping to all three if first we check for the
method/name existence followed by value check.
But I don't think we need to bloat this simple one.

> > +             *id = (u32)adr;
> > +             return 0;
> > +     }
> > +     return -EINVAL;
> > +}

-- 
With Best Regards,
Andy Shevchenko