Re: [PATCH v2] tg3: clean up PHYLIB resources on probe failure

[Myeonghun Pak]

Hello Andrew,

Sorry for the misunderstanding. I will restore the guard only around
tg3_phy_fini() in v3.

Best regards,
Myeonghun Pak

2026년 9월 15일 (화) 오전 11:37, Andrew Lunn <andrew@xxxxxxx>님이 작성:
>
> On Mon, Sep 14, 2026 at 06:22:46PM -0400, Myeonghun Pak wrote:
> > tg3_get_invariants() can register an MDIO bus and connect a PHY for
> > USE_PHYLIB devices. If tg3_init_one() later fails, its common error path
> > releases the mappings and netdev without undoing those PHYLIB resources.
> >
> > Disconnect the PHY and unregister the MDIO bus before the remaining
> > teardown. The existing IS_CONNECTED and MDIOBUS_INITED flags make both
> > helpers safe when initialization only completed partially.
> >
> > This issue was identified during our ongoing static-analysis research while
> > reviewing kernel code.
> >
> > Fixes: 158d7abdae85 ("tg3: Add mdio bus registration")
> > Assisted-by: OpenAI:GPT-5.6
> > Co-developed-by: Ijae Kim <ae878000@xxxxxxxxx>
> > Signed-off-by: Ijae Kim <ae878000@xxxxxxxxx>
> > Signed-off-by: Myeonghun Pak <mhun512@xxxxxxxxx>
> > ---
> > Changes in v2:
> > - Remove the redundant USE_PHYLIB guard; both cleanup helpers already
> >   check their initialization flags, as discussed with Andrew Lunn.
>
> That is not exactly what i said. I pointed out that tg3_mdio_init() is
> not guarded by USE_PHYLIB. tg3_phy_init() is guarded by
> USE_PHYLIB. Generally, you keep code symmetrical, the conditions you
> apply to init should be also applied to fini. However, the existing
> code is not great, tg3_remove_one() is not symmetrical. However, there
> is no reason to make it worse.
>
>
>     Andrew
>
> ---
> pw-bot: cr
>