From mboxrd@z Thu Jan 1 00:00:00 1970 From: Shawn Guo Subject: Re: [PATCH][NET] several cleanups and bugfixes for fec.c: fix the .remove code Date: Wed, 7 Dec 2011 21:43:36 +0800 Message-ID: <20111207134336.GS5550@S2100-06.ap.freescale.net> References: <04c11563fee748562d7b53d5e5b3c5cf37fef200.1323163127.git.LW@KARO-electronics.de> <20111207132828.GQ5550@S2100-06.ap.freescale.net> <20191.27018.478290.890636@ipc1.ka-ro> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: , Shawn Guo , , To: Lothar =?iso-8859-1?Q?Wa=DFmann?= Return-path: Content-Disposition: inline In-Reply-To: <20191.27018.478290.890636@ipc1.ka-ro> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Wed, Dec 07, 2011 at 02:26:34PM +0100, Lothar Wa=DFmann wrote: > Hi, >=20 > Shawn Guo writes: > > On Tue, Dec 06, 2011 at 11:27:15AM +0100, Lothar Wa=DFmann wrote: > > > The .remove code is broken in several ways. > > > - mdiobus_unregister() is called twice for the same object in ca= se of dual FEC > > > - phy_disconnect() is being called when the PHY is already disco= nnected > > > - the requested IRQ(s) are not freed > > > - fec_stop() is being called with the inteface already stopped > > >=20 > > > All of those lead to kernel crashes if the remove function is ac= tually used. > > >=20 > > > Signed-off-by: Lothar Wa=DFmann > > > --- > > > drivers/net/ethernet/freescale/fec.c | 30 ++++++++++++++++++++= +--------- > > > 1 files changed, 21 insertions(+), 9 deletions(-) > > >=20 > > > diff --git a/drivers/net/ethernet/freescale/fec.c b/drivers/net/e= thernet/freescale/fec.c > > > index ab0afb5..70ec7ec 100644 > > > --- a/drivers/net/ethernet/freescale/fec.c > > > +++ b/drivers/net/ethernet/freescale/fec.c > > > @@ -235,6 +235,7 @@ struct fec_enet_private { > > > =20 > > > /* Phylib and MDIO interface */ > > > struct mii_bus *mii_bus; > > > + int mii_cnt; > > > struct phy_device *phy_dev; > > > int mii_timeout; > > > uint phy_speed; > > > @@ -1040,8 +1041,12 @@ static int fec_enet_mii_init(struct platfo= rm_device *pdev) > > > */ > > > if ((id_entry->driver_data & FEC_QUIRK_ENET_MAC) && pdev->id > = 0) { > > > /* fec1 uses fec0 mii_bus */ > > > - fep->mii_bus =3D fec0_mii_bus; > > > - return 0; > > > + if (fep->mii_cnt && fec0_mii_bus) { > >=20 > > This seems broken. The second fec has its own fep and fep->mii_cnt= is > > always 0 here. > >=20 > I already noticed that myself. >=20 With a quick fix here, the last two patches:=20 Tested-by: Shawn Guo --=20 Regards, Shawn