[<prev] [next>] [day] [month] [year] [list]
Message-ID: <CAHvQdo2yzJC89K74c_CZFjPydDQ5i22w36XPR5tKVv_W8a2vcg@mail.gmail.com>
Date: Mon, 7 Jun 2021 14:13:31 +0200
From: Johannes Pointner <h4nn35.work@...il.com>
To: dmurphy@...com
Cc: netdev@...r.kernel.org, andrew@...n.ch, hkallweit1@...il.com
Subject: net: phy: DP83822: not able to get a link
Hello,
I just updated a i.MX6 device which uses the DP83822 from 5.4 to 5.10
and can't get a link anymore.
ip addr shows:
1: lo: <LOOPBACK> mtu 65536 qdisc noop qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
link/ether 00:60:65:54:32:10 brd ff:ff:ff:ff:ff:ff
It seems to me that the commit 5dc39fd5ef35 ("net: phy: DP83822: Add
ability to advertise Fiber connection") causes this.
If I revert the following part of this commit, it is working again:
@@ -314,13 +451,85 @@ static int dp83822_phy_reset(struct phy_device *phydev)
{
int err;
- err = phy_write(phydev, MII_DP83822_RESET_CTRL, DP83822_HW_RESET);
+ err = phy_write(phydev, MII_DP83822_RESET_CTRL, DP83822_SW_RESET);
Was this change intentional? Because I took a look at the driver
DP83867 which has similar bits to reset the phy and there this wasn't
changed.
Maybe the naming of the defines in case of the DP83822 is a bit misleading?
DP83882:
#define DP83822_HW_RESET BIT(15)
#define DP83822_SW_RESET BIT(14)
datasheet description:
15 Software Reset
14 Digital Restart
DP83867:
#define DP83867_SW_RESET BIT(15)
#define DP83867_SW_RESTART BIT(14)
datasheet description:
15 SW_RESET
14 SW_RESTART
Thanks,
Hannes
Powered by blists - more mailing lists