[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20201125170714.GK1551@shell.armlinux.org.uk>
Date: Wed, 25 Nov 2020 17:07:14 +0000
From: Russell King - ARM Linux admin <linux@...linux.org.uk>
To: Yonglong Liu <liuyonglong@...wei.com>
Cc: Antonio Borneo <antonio.borneo@...com>,
Andrew Lunn <andrew@...n.ch>,
Heiner Kallweit <hkallweit1@...il.com>,
"David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>, netdev@...r.kernel.org,
Willy Liu <willy.liu@...ltek.com>,
linux-kernel@...r.kernel.org, Salil Mehta <salil.mehta@...wei.com>,
linuxarm@...wei.com, linux-stm32@...md-mailman.stormreply.com
Subject: Re: [PATCH v3 net-next] net: phy: realtek: read actual speed on
rtl8211f to detect downshift
On Thu, Nov 26, 2020 at 12:57:37AM +0800, Yonglong Liu wrote:
> Hi, Antonio:
>
> Could you help to provide a downshift warning message when this happen?
>
> It's a little strange that the adv and the lpa support 1000M, but
> finally the link speed is 100M.
That is an identifying feature of downshift.
Downshift can happen at either end of the link, and since we must not
change the "Advertised link modes" since this is what userspace
configured, if a downshift occurs at the local end, then you will get
the ethtool output you provide, where the speed does not agree with
the reported advertisements.
You should already be getting a warning in the kernel log when this
happens; phy_check_downshift() which is part of the phylib core code
will check this every time the link comes up. You should already
have a message "Downshift occurred ..." in your kernel log. Please
check.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
Powered by blists - more mailing lists