netdev - Re: [RFC PATCH 1/1] phylink: Set speed to SPEED_UNKNOWN when there is no PHY connected

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: <20190828171431.GR13294@shell.armlinux.org.uk>
Date:   Wed, 28 Aug 2019 18:14:31 +0100
From:   Russell King - ARM Linux admin <linux@...linux.org.uk>
To:     Vladimir Oltean <olteanv@...il.com>
Cc:     andrew@...n.ch, f.fainelli@...il.com, asolokha@...kras.ru,
        netdev@...r.kernel.org
Subject: Re: [RFC PATCH 1/1] phylink: Set speed to SPEED_UNKNOWN when there
 is no PHY connected

On Wed, Aug 28, 2019 at 05:58:02PM +0300, Vladimir Oltean wrote:
> phylink_ethtool_ksettings_get can be called while the interface may not
> even be up, which should not be a problem. But there are drivers (e.g.
> gianfar) which connect to the PHY in .ndo_open and disconnect in
> .ndo_close. While odd, to my knowledge this is again not illegal and
> there may be more that do the same. But PHYLINK for example has this
> check in phylink_ethtool_ksettings_get:
> 
> 	if (pl->phydev) {
> 		phy_ethtool_ksettings_get(pl->phydev, kset);
> 	} else {
> 		kset->base.port = pl->link_port;
> 	}
> 
> So it will not populate kset->base.speed if there is no PHY connected.
> The speed will be 0, by way of a previous memset. Not SPEED_UNKNOWN.
> It is arguable whether that is legal or not. include/uapi/linux/ethtool.h
> says:
> 
> 	All values 0 to INT_MAX are legal.
> 
> By that measure it may be. But it sure would make users of the
> __ethtool_get_link_ksettings API need make more complicated checks
> (against -1, against 0, 1, etc). So far the kernel community has been ok
> with just checking for SPEED_UNKNOWN.
> 
> Take net/sched/sch_taprio.c for example. The check in
> taprio_set_picos_per_byte is currently not robust enough and will
> trigger this division by zero, due to PHYLINK not setting SPEED_UNKNOWN:
> 
> 	if (!__ethtool_get_link_ksettings(dev, &ecmd) &&
> 	    ecmd.base.speed != SPEED_UNKNOWN)
> 		picos_per_byte = div64_s64(NSEC_PER_SEC * 1000LL * 8,
> 					   ecmd.base.speed * 1000 * 1000);

The ethtool API says:

 * If it is enabled then they are read-only; if the link
 * is up they represent the negotiated link mode; if the link is down,
 * the speed is 0, %SPEED_UNKNOWN or the highest enabled speed and
 * @duplex is %DUPLEX_UNKNOWN or the best enabled duplex mode.

So, it seems that taprio is not following the API... I'd suggest either
fixing taprio, or getting agreement to change the ethtool API.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up