Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <240155f20aae47e9f7461e2b7416120ba6238886.1679650087.git.geert+renesas@glider.be>
In-Reply-To: <240155f20aae47e9f7461e2b7416120ba6238886.1679650087.git.geert+renesas@glider.be>
From:   Saravana Kannan <saravanak@google.com>
Date:   Wed, 29 Mar 2023 12:47:31 -0700
Message-ID: <CAGETcx-5a3ZKRiG1MbZaYiJUPVSPH1O18+E-HoTRnAUz0mc07A@mail.gmail.com>
Subject: Re: [PATCH v2] treewide: Fix instantiation of devices in DT overlays
To:     geert+renesas@glider.be
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Shawn Guo <shawnguo@kernel.org>,
        Sascha Hauer <s.hauer@pengutronix.de>,
        Pengutronix Kernel Team <kernel@pengutronix.de>,
        Fabio Estevam <festevam@gmail.com>,
        NXP Linux Team <linux-imx@nxp.com>,
        Wolfram Sang <wsa@kernel.org>,
        Rob Herring <robh+dt@kernel.org>,
        Frank Rowand <frowand.list@gmail.com>,
        Mark Brown <broonie@kernel.org>,
        linux-arm-kernel@lists.infradead.org, linux-i2c@vger.kernel.org,
        devicetree@vger.kernel.org, linux-spi@vger.kernel.org,
        linux-renesas-soc@vger.kernel.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Fri, Mar 24, 2023 at 2:30=E2=80=AFAM Geert Uytterhoeven
<geert+renesas@glider.be> wrote:
>

I saw all the other replies. Let me give more details on the commit
text and inline comments so it's clearer and not a "he said he said"

> When loading a DT overlay that creates a device, the device is not
> instantiated,

IIRC I think the devices are instantiated, but not probed.

> unless the DT overlay is unloaded and reloaded again.
>
> Saravana explains:
>   Basically for all overlays (I hope the function is only used for
>   overlays) we assume all nodes are NOT devices until they actually
>   get added as a device.

This comment was about what the patch is doing. I think a better
commit text would be something like below. Feel free to reword as you
see fit.

After the recent refactor to improve fw_devlink[1], it no longer depends on
"compatible" property to identify which device tree nodes will
become struct device. fw_devlink now picks up dangling consumers (consumers
pointing to descendent device tree nodes of a device that aren't converted
to child devices) when a device is successfully bound to a driver. See
__fw_devlink_pickup_dangling_consumers().

However, during DT overlay, a device's device tree node can have sub-nodes
added/removed without unbinding/rebinding the driver. This difference in
behavior between the normal device instantiation and probing flow vs the DT
overlay flow has a bunch of implications that are pointed out elsewhere[2].
One of them is that the fw_devlink logic to pick up dangling consumers is
never exercised.

This patch solves the fw_devlink issue by marking all DT nodes added by DT
overlay with FWNODE_FLAG_NOT_DEVICE (fwnode that won't become device) and
then clearing the flag when a struct device is actually created for the DT
node. This way, fw_devlink knows not to have consumers waiting on these
newly added DT nodes and to propagate the dependency to an ancestor DT node
that has corresponding struct device.

[1] - https://lore.kernel.org/lkml/20230207014207.1678715-1-saravanak@googl=
e.com/
[2] - https://lore.kernel.org/lkml/CAGETcx_bkuFaLCiPrAWCPQz+w79ccDp6=3D9e88=
1qmK=3Dvx3hBMyg@mail.gmail.com/#t


> Based on a patch by Saravana Kannan, which covered only platform and spi
> devices.

You can keep this in the commit text.

> Fixes: 4a032827daa89350 ("of: property: Simplify of_link_to_phandle()")
> Link: https://lore.kernel.org/r/CAGETcx_+rhHvaC_HJXGrr5_WAd2+k5f=3DrWYnkC=
Z6z5bGX-wj4w@mail.gmail.com
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
> Acked-by: Mark Brown <broonie@kernel.org>
> ---
> v2:
>   - Add Acked-by,
>   - Drop RFC.
> ---
>  drivers/bus/imx-weim.c    | 1 +
>  drivers/i2c/i2c-core-of.c | 1 +
>  drivers/of/dynamic.c      | 1 +
>  drivers/of/platform.c     | 1 +
>  drivers/spi/spi.c         | 1 +
>  5 files changed, 5 insertions(+)
>
> diff --git a/drivers/bus/imx-weim.c b/drivers/bus/imx-weim.c
> index 36d42484142aede2..898e23a4231400fa 100644
> --- a/drivers/bus/imx-weim.c
> +++ b/drivers/bus/imx-weim.c
> @@ -329,6 +329,7 @@ static int of_weim_notify(struct notifier_block *nb, =
unsigned long action,
>                                  "Failed to setup timing for '%pOF'\n", r=
d->dn);
>
>                 if (!of_node_check_flag(rd->dn, OF_POPULATED)) {

It's important this flag clearing is done before the device is added.
So, can you please add a comment before all these clear flag lines
that's something like:

/* Clear the flag before adding the device so that fw_devlink doesn't
skip adding consumers to this device. */

Thanks,
Saravana

> +                       rd->dn->fwnode.flags &=3D ~FWNODE_FLAG_NOT_DEVICE=
;
>                         if (!of_platform_device_create(rd->dn, NULL, &pde=
v->dev)) {
>                                 dev_err(&pdev->dev,
>                                         "Failed to create child device '%=
pOF'\n",
> diff --git a/drivers/i2c/i2c-core-of.c b/drivers/i2c/i2c-core-of.c
> index aa93467784c29c89..303f9003562eed3d 100644
> --- a/drivers/i2c/i2c-core-of.c
> +++ b/drivers/i2c/i2c-core-of.c
> @@ -178,6 +178,7 @@ static int of_i2c_notify(struct notifier_block *nb, u=
nsigned long action,
>                         return NOTIFY_OK;
>                 }
>
> +               rd->dn->fwnode.flags &=3D ~FWNODE_FLAG_NOT_DEVICE;
>                 client =3D of_i2c_register_device(adap, rd->dn);
>                 if (IS_ERR(client)) {
>                         dev_err(&adap->dev, "failed to create client for =
'%pOF'\n",
> diff --git a/drivers/of/dynamic.c b/drivers/of/dynamic.c
> index 07d93753b12f5f4d..e311d406b1705306 100644
> --- a/drivers/of/dynamic.c
> +++ b/drivers/of/dynamic.c
> @@ -226,6 +226,7 @@ static void __of_attach_node(struct device_node *np)
>         np->sibling =3D np->parent->child;
>         np->parent->child =3D np;
>         of_node_clear_flag(np, OF_DETACHED);
> +       np->fwnode.flags |=3D FWNODE_FLAG_NOT_DEVICE;
>  }
>
>  /**
> diff --git a/drivers/of/platform.c b/drivers/of/platform.c
> index b2bd2e783445dd78..17c92cbfb62ee3ef 100644
> --- a/drivers/of/platform.c
> +++ b/drivers/of/platform.c
> @@ -737,6 +737,7 @@ static int of_platform_notify(struct notifier_block *=
nb,
>                 if (of_node_check_flag(rd->dn, OF_POPULATED))
>                         return NOTIFY_OK;
>
> +               rd->dn->fwnode.flags &=3D ~FWNODE_FLAG_NOT_DEVICE;
>                 /* pdev_parent may be NULL when no bus platform device */
>                 pdev_parent =3D of_find_device_by_node(rd->dn->parent);
>                 pdev =3D of_platform_device_create(rd->dn, NULL,
> diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
> index 8e8af148b1dc371e..66ac67580d2a473b 100644
> --- a/drivers/spi/spi.c
> +++ b/drivers/spi/spi.c
> @@ -4527,6 +4527,7 @@ static int of_spi_notify(struct notifier_block *nb,=
 unsigned long action,
>                         return NOTIFY_OK;
>                 }
>
> +               rd->dn->fwnode.flags &=3D ~FWNODE_FLAG_NOT_DEVICE;
>                 spi =3D of_register_spi_device(ctlr, rd->dn);
>                 put_device(&ctlr->dev);
>
> --
> 2.34.1
>