Received-SPF: pass (google.com: domain of linux-wireless-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: <ZK9mXE00xEHZV4fi@makrotopia.org>
In-Reply-To: <ZK9mXE00xEHZV4fi@makrotopia.org>
From:   Jonas Gorski <jonas.gorski@gmail.com>
Date:   Wed, 26 Jul 2023 18:04:18 +0200
Message-ID: <CAOiHx=m5jRYd70ymvX=-9Xh4ZDApJ04G39j9+U0YB18W1QPv7A@mail.gmail.com>
Subject: Re: [PATCH] wifi: mt76: support per-band MAC addresses from OF child nodes
To:     Daniel Golle <daniel@makrotopia.org>
Cc:     linux-wireless@vger.kernel.org, linux-mediatek@lists.infradead.org,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        Felix Fietkau <nbd@nbd.name>,
        Lorenzo Bianconi <lorenzo@kernel.org>,
        Ryder Lee <ryder.lee@mediatek.com>,
        Shayne Chen <shayne.chen@mediatek.com>,
        Sean Wang <sean.wang@mediatek.com>,
        Kalle Valo <kvalo@kernel.org>,
        Matthias Brugger <matthias.bgg@gmail.com>,
        AngeloGioacchino Del Regno 
        <angelogioacchino.delregno@collabora.com>,
        Christian Marangi <ansuelsmth@gmail.com>,
        John Crispin <john@phrozen.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi,

On Thu, 13 Jul 2023 at 04:53, Daniel Golle <daniel@makrotopia.org> wrote:
>
> With dual-band-dual-congruent front-ends which appear as two independent
> radios it is desirable to assign a per-band MAC address from device-tree,
> eg. using nvmem-cells.
> Support specifying MAC-address related properties in band-specific child
> nodes, e.g.
>         mt7915@0,0 {
>                 reg = <0x0000 0 0 0 0>;
>                 #addr-cells = <1>;
>                 #size-cells = <0>;
>
>                 band@0 {
>                         /* 2.4 GHz */
>                         reg = <0>;
>                         nvmem-cells = <&macaddr 2>;
>                         nvmem-cell-names = "mac-address";
>                 };
>
>                 band@1 {
>                         /* 5 GHz */
>                         reg = <1>;
>                         nvmem-cells = <&macaddr 3>;
>                         nvmem-cell-names = "mac-address";
>                 };
>         };
>
> Signed-off-by: Daniel Golle <daniel@makrotopia.org>
> ---
>  drivers/net/wireless/mediatek/mt76/eeprom.c | 15 ++++++++++++++-
>  1 file changed, 14 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/wireless/mediatek/mt76/eeprom.c b/drivers/net/wireless/mediatek/mt76/eeprom.c
> index dce851d42e083..90ee138843a55 100644
> --- a/drivers/net/wireless/mediatek/mt76/eeprom.c
> +++ b/drivers/net/wireless/mediatek/mt76/eeprom.c
> @@ -106,7 +106,20 @@ void
>  mt76_eeprom_override(struct mt76_phy *phy)
>  {
>         struct mt76_dev *dev = phy->dev;
> -       struct device_node *np = dev->dev->of_node;
> +       struct device_node *child_np, *np = dev->dev->of_node;
> +       u32 reg;
> +       int ret;
> +
> +       for_each_child_of_node(np, child_np) {
> +               ret = of_property_read_u32(child_np, "reg", &reg);
> +               if (ret)
> +                       continue;
> +
> +               if (reg == phy->band_idx) {
> +                       np = child_np;
> +                       break;

When breaking out of the loop here you still hold an additional
reference to child_np, so you need to call of_node_put() for it,
probably after calling of_get_mac_address().

> +               }
> +       }
>
>         of_get_mac_address(np, phy->macaddr);
>

probably just add a of_node_put(child_np); here, of_node_put() seems
to be NULL safe.


Regards,
Jonas