Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:7 as permitted sender) client-ip=2620:137:e000::3:7;
Message-ID: <018982972b43bdac3f6d50bde03287282364e357.camel@redhat.com>
Subject: Re: [PATCH] fjes: Add missing check for vzalloc
From:   Paolo Abeni <pabeni@redhat.com>
To:     Petr Machata <petrm@nvidia.com>, Chen Ni <nichen@iscas.ac.cn>
Cc:     davem@davemloft.net, edumazet@google.com, kuba@kernel.org,
        izumi.taku@jp.fujitsu.com, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Date:   Tue, 03 Oct 2023 13:07:35 +0200
In-Reply-To: <87r0mms4a5.fsf@nvidia.com>
References: <20230925085318.1228225-1-nichen@iscas.ac.cn>
         <87r0mms4a5.fsf@nvidia.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.4 (3.46.4-1.fc37) 
MIME-Version: 1.0
Precedence: bulk

On Mon, 2023-09-25 at 16:40 +0200, Petr Machata wrote:
> Chen Ni <nichen@iscas.ac.cn> writes:
>=20
> > Because of the potential failure of the vzalloc(), the hw->hw_info.trac=
e
> > could be NULL.
> > Therefore, we need to check it and return -ENOMEM in order to transfer
> > the error.
> >=20
> > Fixes: b6ba737d0b29 ("fjes: ethtool -w and -W support for fjes driver")
> > Signed-off-by: Chen Ni <nichen@iscas.ac.cn>
> > ---
> >  drivers/net/fjes/fjes_hw.c | 3 +++
> >  1 file changed, 3 insertions(+)
> >=20
> > diff --git a/drivers/net/fjes/fjes_hw.c b/drivers/net/fjes/fjes_hw.c
> > index 704e949484d0..3a06a3cf021d 100644
> > --- a/drivers/net/fjes/fjes_hw.c
> > +++ b/drivers/net/fjes/fjes_hw.c
> > @@ -330,6 +330,9 @@ int fjes_hw_init(struct fjes_hw *hw)
> >  	ret =3D fjes_hw_setup(hw);
> > =20
> >  	hw->hw_info.trace =3D vzalloc(FJES_DEBUG_BUFFER_SIZE);
> > +	if (!hw->hw_info.trace)
> > +		return -ENOMEM;
> > +
>=20
> I'm not sure, but shouldn't this call fjes_hw_cleanup() to mirror the
> setup() above? Also only if ret=3D0 I suppose.

Yes, that looks needed, or memory will be leaked. Additionally it looks
like the rest of the driver handles correctly the case where
hw_info.trace is NULL, so this fix is likely not needed at all.

Cheers,

Paolo