netdev - RE: [EXT] Re: [PATCH net-next] net: fec: Use a proper ID allocation scheme

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <VI1PR0402MB360000C02868DB471237E08AFF110@VI1PR0402MB3600.eurprd04.prod.outlook.com>
Date:   Tue, 18 Feb 2020 14:30:01 +0000
From:   Andy Duan <fugang.duan@....com>
To:     Fabio Estevam <festevam@...il.com>
CC:     David Miller <davem@...emloft.net>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "andrew@...n.ch" <andrew@...n.ch>
Subject: RE: [EXT] Re: [PATCH net-next] net: fec: Use a proper ID allocation
 scheme

From: Fabio Estevam <festevam@...il.com> Sent: Tuesday, February 18, 2020 10:05 PM
> Hi Andy,
> 
> On Tue, Feb 18, 2020 at 10:54 AM Andy Duan <fugang.duan@....com>
> wrote:
> 
> > For imx6sl/imx8mp/imx8mm/imx8mn, soc only has one instance, bind
> > operation is supported and has no problem.
> 
> This is not true.
> 
> As per the commit log, here is the result of unbind/bind on a i.mx6qp, which
> only has a single FEC instance:
I mean if apply the patch, it should work for one instance.

> 
> # echo 2188000.ethernet > /sys/bus/platform/drivers/fec/unbind
> # echo 2188000.ethernet > /sys/bus/platform/drivers/fec/bind
> [   10.756519] pps pps0: new PPS source ptp0
> [   10.792626] libphy: fec_enet_mii_bus: probed
> [   10.799330] fec 2188000.ethernet eth0: registered PHC device 1
> # udhcpc -i eth0
> udhcpc: started, v1.31.1
> [   14.985211] fec 2188000.ethernet eth0: no PHY, assuming direct
> connection to switch
> [   14.993140] libphy: PHY fixed-0:00 not found
> [   14.997643] fec 2188000.ethernet eth0: could not attach to PHY
> 
> After performing unbind/bind operation the network is not functional at all.
> 
> Don't you agree that unbind/bind is currently broken here even for SoCs with
> a single FEC?
> 
> Should we prevent unbind? Or any other suggestion?
Suppose apply the patch, it can work for one instance, but not for two instances.
Currently, I agree to prevent unbind operation.