> -----Original Message----- > From: Andrew Lunn [mailto:and...@lunn.ch] > Sent: Wednesday, January 11, 2017 6:00 PM > To: Kwok, WingMan > Cc: rmk+ker...@arm.linux.org.uk; charles-antoine.cou...@nexvision.fr; > Karicheri, Muralidharan; netdev@vger.kernel.org > Subject: Re: Marvell Phy (1510) issue since v4.7 kernel > > > looping in Charles-Antoine (author of patch > > with commit id 6cfb3bcc) > > > > Charles-Antoine, > > > > Do you have the ID of Marvell PHY 88E1512? > > I suspect that is the wrong question to ask. > > The Marvell driver is being loaded, so it must be using on of the IDs > in the driver. There is no ID in the driver specifically for the > 88E1512. It seems like the 88E1512 uses the 88E1510 ID. > > So i think the correct question should be, how can we tell the 88E1512 > from the 88E1510 if they have the same ID in register 3. > > It appears that for the 88E1512, page 0 are the copper registers and > page 1 is the fibre registers. Maybe the 88E1512 has an ID in page 1 > register 3? Maybe the 88E1510 does not have an ID in page 1 register > 3? > > Andrew
Andrew, Would Charles-Antoine be the better person to submit a patch to fix the original problem then, since he tested the original fiber support patch with 1512? I unfortunately don't have the datasheet for 1512, and it does not seem to be available publicly. Regards, WingMan