[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <3c6abdaf-b4e4-440f-9d75-745741287007@lunn.ch>
Date: Thu, 17 Apr 2025 20:11:00 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Thangaraj.S@...rochip.com
Cc: netdev@...r.kernel.org, fiona.klute@....de, andrew+netdev@...n.ch,
linux-usb@...r.kernel.org, davem@...emloft.net,
Rengarajan.S@...rochip.com, kernel-list@...pberrypi.com,
linux-kernel@...r.kernel.org, kuba@...nel.org, pabeni@...hat.com,
edumazet@...gle.com, UNGLinuxDriver@...rochip.com
Subject: Re: [PATCH net v2] net: phy: microchip: force IRQ polling mode for
lan88xx
On Thu, Apr 17, 2025 at 02:17:20PM +0000, Thangaraj.S@...rochip.com wrote:
> Hi Flona,
> We haven't observed this issue with our LAN7801 + LAN88xx setup during
> testing in our setups. Issue reported specifically in RPI and NVidea
> with which we have not tested recently. Additionally, community
> discussions suggest that this issue may have been introduced due to
> driver updates in the LAN78xx. There's no hardware errata indicating
> any problems with the interrupts in LAN88xx.
>
> If the issue lies in the interrupt handling within the LAN78xx, it's
> likely that similar problems could arise with other PHYs that support
> interrupt handling. This need to be debugged and addressed from LAN78xx
> driver rather than removing interrupt support in LAN88xx.
Hi Thangaraj
It is an easily available platform. Maybe you can get one and try to
debug it?
This is a self contained simple fix. From a pragmatic standpoint, it
solves an issue which is bothering people. If you do find and fix an
issue in the LAN78xx this change can be easily reverted. So i'm
learning towards merging it.
Andrew
Powered by blists - more mailing lists