Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 12 Oct 2018 15:55:50 +0100
From:   Stefan Hajnoczi <stefanha@gmail.com>
To:     piaojun <piaojun@huawei.com>
Cc:     pbonzini@redhat.com, linux-kernel@vger.kernel.org,
        rusty@rustcorp.com.au, kvm@vger.kernel.org, penberg@kernel.org,
        mst@redhat.com, michaelc@cs.wisc.edu
Subject: Re: [PATCH] scsi/virio_scsi.c: do not call virtscsi_remove_vqs() in
 virtscsi_init() to avoid crash bug
Message-ID: <20181012145550.GK24267@stefanha-x1.localdomain>
References: <5B7E1AFA.8030506@huawei.com>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha1;
        protocol="application/pgp-signature"; boundary="sU4rRG038CsJurvk"
Content-Disposition: inline
In-Reply-To: <5B7E1AFA.8030506@huawei.com>
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


--sU4rRG038CsJurvk
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Aug 23, 2018 at 10:24:58AM +0800, piaojun wrote:
> If some error happened before find_vqs, error branch will goto
> virtscsi_remove_vqs to free vqs. Actually the vqs have not been allocated
> successfully, so this will cause wild-pointer-free problem. So
> virtscsi_remove_vqs could be deleted as no error will happen after
> find_vqs.
>=20
> Signed-off-by: Jun Piao <piaojun@huawei.com>
> ---
>  drivers/scsi/virtio_scsi.c | 2 --
>  1 file changed, 2 deletions(-)
>=20
> diff --git a/drivers/scsi/virtio_scsi.c b/drivers/scsi/virtio_scsi.c
> index 1c72db9..da0fd74 100644
> --- a/drivers/scsi/virtio_scsi.c
> +++ b/drivers/scsi/virtio_scsi.c
> @@ -833,8 +833,6 @@ static int virtscsi_init(struct virtio_device *vdev,
>  	kfree(names);
>  	kfree(callbacks);
>  	kfree(vqs);
> -	if (err)
> -		virtscsi_remove_vqs(vdev);

Can you provide more details about the problem?

drivers/virtio/virtio_pci_common.c:vp_del_vqs() looks fine to me, it
iterates over vdev->vqs.  The vdev->vqs list has been initialized in
drivers/virtio/virtio.c:register_virtio_device() and it's empty.

Stefan

--sU4rRG038CsJurvk
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iQEcBAEBAgAGBQJbwLX2AAoJEJykq7OBq3PIHPUH/R8uWZ/PQztf5OpFMjs/unVf
v+1840hOhxyCkb7otc+976KbDbDpc/RizlufgZyLoBsE+3mEjRFwg7A3wh9nYUwj
cXEO2Hu4uKO59N9/PLPRoDpLLwVE//24vtx7IfFjKjkcNrcnY+PNhJgD97FccPKT
bkU6IPRpUq63nQl8ZS0UcW9Ho6BCgaNePTIyGysLgviG32Ofkr1lEco4y3S9OsF7
AvqFJM1nQk+FxBaH7QmV1b7PLT7Rpq3DmvpKo5ay+HgLaY11cq+msP9p0YlpKQLI
A07Fo833XET9vu/IR18pwTSKHf1aZ/mfFHDKCyS6sLRX2mQ+mHWpubNkZko4FVc=
=6xR2
-----END PGP SIGNATURE-----

--sU4rRG038CsJurvk--