From mboxrd@z Thu Jan  1 00:00:00 1970
From: Andrew Lunn <andrew@lunn.ch>
Subject: Re: [PATCH 1/4] net: macb: Fix regression breaking non-MDIO
 fixed-link PHYs
Date: Mon, 20 Aug 2018 21:06:30 +0200
Message-ID: <20180820190630.GB12238@lunn.ch>
References: <20180820121238.7779-1-a.fatoum@pengutronix.de>
 <20180820135536.GD6583@lunn.ch>
 <b5445ad9-70bd-12fe-a8e7-1dc3ec73585e@pengutronix.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: "David S. Miller" <davem@davemloft.net>,
        Nicolas Ferre <nicolas.ferre@microchip.com>,
        kernel@pengutronix.de, netdev@vger.kernel.org, mdf@kernel.org,
        Brad Mouring <brad.mouring@ni.com>,
        Florian Fainelli <f.fainelli@gmail.com>, stable@vger.kernel.org
To: Ahmad Fatoum <a.fatoum@pengutronix.de>
Return-path: <stable-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <b5445ad9-70bd-12fe-a8e7-1dc3ec73585e@pengutronix.de>
Sender: stable-owner@vger.kernel.org
List-Id: netdev.vger.kernel.org

On Mon, Aug 20, 2018 at 05:56:53PM +0200, Ahmad Fatoum wrote:
> On 08/20/2018 03:55 PM, Andrew Lunn wrote:
> > Why is of_phy_register_fixed_link(np) failing?
> 
> Apparently, the fixed-link's gpio's FLAG_REQUESTED bit remained set
> causing gpiod_request_commit to return -EBUSY in (v4.18.0):
> 
> [<c0667750>] (gpiod_request_commit) from [<c066890c>] (gpiod_request+0x64/0x88)             
> [<c066890c>] (gpiod_request) from [<c066a338>] (gpiod_get_from_of_node+0x48/0x13c)          
> [<c066a338>] (gpiod_get_from_of_node) from [<c094b92c>] (mdiobus_register_device+0x70/0x124)
> [<c094b92c>] (mdiobus_register_device) from [<c0949d2c>] (phy_device_register+0xc/0xa0)     
> [<c0949d2c>] (phy_device_register) from [<c094e824>] (fixed_phy_register+0xe8/0x1f8)        
> [<c094e824>] (fixed_phy_register) from [<c0b84260>] (of_phy_register_fixed_link+0x150/0x1e4)
> [<c0b84260>] (of_phy_register_fixed_link) from [<c0964908>] (macb_probe+0x548/0xa7c)        
> [<c0964908>] (macb_probe) from [<c086ebec>] (platform_drv_probe+0x48/0x98)                  
> 
> But that's a separate issue, I'll remove this line from the commit message...

I would actually say, this is your real issue here. The warnings are
annoying, but i don't think they are fatal. This -EBUSY is what is
stopping the driver from loading, causing the real regression.

I'm guessing, but i think you will find the driver is loading once,
but hits a EPROBE_DEFFER condition, after getting the gpio. It does
not release the gpio correctly. Sometime later, it gets loaded again,
but the gpio is now in use, so you get the -EBUSY.

So check the error paths, and make sure cleanup is being done correct.
It could also be a phylib core bug...

   Andrew