From mboxrd@z Thu Jan 1 00:00:00 1970 From: Uwe =?iso-8859-1?Q?Kleine-K=F6nig?= Subject: Re: Crashes in -next due to 'phy: add support for a reset-gpio specification' Date: Sun, 22 May 2016 20:21:12 +0200 Message-ID: <20160522182112.GC23704@pengutronix.de> References: <573BF177.6090507@roeck-us.net> <20160522101051.GA23704@pengutronix.de> <5741D323.8010509@roeck-us.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from metis.ext.4.pengutronix.de ([92.198.50.35]:48914 "EHLO metis.ext.4.pengutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752403AbcEVSVT (ORCPT ); Sun, 22 May 2016 14:21:19 -0400 Content-Disposition: inline In-Reply-To: <5741D323.8010509@roeck-us.net> Sender: linux-next-owner@vger.kernel.org List-ID: To: Guenter Roeck Cc: "linux-next@vger.kernel.org" , Network Development , "linux-kernel@vger.kernel.org" , "David S. Miller" , Linus Walleij , linux-gpio@vger.kernel.org Hello Guenter, [extending Cc: a bit] On Sun, May 22, 2016 at 08:41:23AM -0700, Guenter Roeck wrote: > On 05/22/2016 03:10 AM, Uwe Kleine-K=F6nig wrote: > >On Tue, May 17, 2016 at 09:37:11PM -0700, Guenter Roeck wrote: > >>[ 9.366256] libphy: ethoc-mdio: probed > >>[ 9.367389] (null): could not attach to PHY > >>[ 9.368555] (null): failed to probe MDIO bus > >>[ 9.371540] Unable to handle kernel paging request at virtual ad= dress 0000001c > >>[ 9.371540] pc =3D d0320926, ra =3D 903209d1 > >>[ 9.375358] Oops: sig: 11 [#1] > >>[ 9.376081] PREEMPT > >>[ 9.377080] CPU: 0 PID: 1 Comm: swapper Not tainted 4.6.0-next-2= 0160517 #1 > >>[ 9.378397] task: d7c2c000 ti: d7c30000 task.ti: d7c30000 > >>[ 9.379394] a00: 903209d1 d7c31bd0 d7fb5810 00000001 00000000 00= 000000 d7f45c00 d7c31bd0 > >>[ 9.382298] a08: 00000000 00000000 00000000 00000000 00060100 d0= 4b0c10 d7f45dfc d7c31bb0 > >>[ 9.385732] pc: d0320926, ps: 00060110, depc: 00000018, excvaddr= : 0000001c > >>[ 9.387061] lbeg: d0322e35, lend: d0322e57 lcount: 00000000, sar= : 00000011 > >>[ 9.388173] > >>Stack: d7c31be0 00060700 d7f45c00 d7c31bd0 9021d509 d7c31c30 d7f45c= 00 00000000 > >> d0485dcc d0485dcc d7fb5810 d7c2c000 00000000 d7c31c30 d7f45= c00 d025befc > >> d0485dcc d7c30000 d7f45c34 d7c31bf0 9021c985 d7c31c50 d7f45= c00 d7f45c34 > >>[ 9.396652] Call Trace: > >>[ 9.397469] [] __device_release_driver+0x7d/0x98 > >>[ 9.398869] [] device_release_driver+0x15/0x20 > >>[ 9.400247] [] bus_remove_device+0xc1/0xd4 > >>[ 9.401569] [] device_del+0x109/0x15c > >>[ 9.402794] [] phy_mdio_device_remove+0xd/0x18 > >>[ 9.404124] [] mdiobus_unregister+0x40/0x5c > >>[ 9.405444] [] ethoc_probe+0x534/0x5b8 > >>[ 9.406742] [] platform_drv_probe+0x28/0x48 > >>[ 9.408122] [] driver_probe_device+0x101/0x234 > >>[ 9.409499] [] __driver_attach+0x7d/0x98 > >>[ 9.410809] [] bus_for_each_dev+0x30/0x5c > >>[ 9.412104] [] driver_attach+0x14/0x18 > >>[ 9.413385] [] bus_add_driver+0xc9/0x198 > >>[ 9.414686] [] driver_register+0x70/0xa0 > >>[ 9.416001] [] __platform_driver_register+0x24/0x28 > >>[ 9.417463] [] ethoc_driver_init+0x10/0x14 > >>[ 9.418824] [] do_one_initcall+0x80/0x1ac > >>[ 9.420083] [] kernel_init_freeable+0x131/0x198 > >>[ 9.421504] [] kernel_init+0xc/0xb0 > >>[ 9.422693] [] ret_from_kernel_thread+0x8/0xc > > > >Guenter, can you please test if the following patch fixes your setup= : > > > >diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_devi= ce.c > >index 307f72a0f2e2..efa85fb31574 100644 > >--- a/drivers/net/phy/phy_device.c > >+++ b/drivers/net/phy/phy_device.c > >@@ -1573,14 +1573,14 @@ static int phy_probe(struct device *dev) > > int err =3D 0; > > struct gpio_descs *reset_gpios; > > > >- phydev->drv =3D phydrv; > >- > > /* take phy out of reset */ > > reset_gpios =3D devm_gpiod_get_array_optional(dev, "reset", > > GPIOD_OUT_LOW); > > if (IS_ERR(reset_gpios)) > > return PTR_ERR(reset_gpios); > > > >+ phydev->drv =3D phydrv; > >+ > > /* Disable the interrupt if the PHY doesn't support it > > * but the interrupt is still a valid one > > */ > > > >This doesn't make your ethernet work, but at least the driver should > >fail in a clean way. > > > I tried, but it still fails with exactly the same error, meaning > it still crashes with the same traceback. This is strange. If probe fails the device should stay unbound and it shouldn't enter the if block in __device_release_driver at all. > >Together with enabling GPIOLIB this should put ethernet in a working > >state again. > > >=20 > I am not exactly in favor of forcing GPIOLIB to be enabled for every > system with Ethernet support, just because a few of them may require > a gpio based phy reset. The same is true, really, for every other > driver using _optional gpiolib functions. >=20 > Personally, I think that the _optional functions in gpiolib _should_ > return no error if gpiolib is not configured. After all, those > gpio pins _are_ supposed to be optional. gpiolib should be enabled > for affected configurations, ie on systems with gpio support which > do need the optional gpio pins. Forcing gpiolib to be enabled even > in systems with no gpio support seems to be a bit heavy-handed. > It just bloats the kernel on such systems with no added benefit. That's wrong. The usage of gpio_get_optional and friends means that the gpio is optional for the *driver*, that is there are devices that make use of said gpio and others don't. For the devices where the gpio is specified its usage is not optional. So it must not be ignored e.g. by GPIOLIB=3Dn configurations. IMHO the best would be to unconditionally enable that part of GPIOLIB such that the _optional variants do the following with GPIOLIB=3Dn: if a GPIO is specified: return -ENOSYS else: return NULL Best regards Uwe --=20 Pengutronix e.K. | Uwe Kleine-K=F6nig = | Industrial Linux Solutions | http://www.pengutronix.de/= |