[PATCH v2] net: mtk_sgmii: implement mtk_pcs_ops

Russell King (Oracle) linux at armlinux.org.uk
Mon Jan 16 08:43:18 PST 2023


On Mon, Jan 16, 2023 at 05:33:53PM +0100, Bjørn Mork wrote:
> "Russell King (Oracle)" <linux at armlinux.org.uk> writes:
> 
> > On Mon, Jan 16, 2023 at 04:21:30PM +0100, Bjørn Mork wrote:
> >> [   54.539438] mtk_soc_eth 15100000.ethernet wan: Link is Down
> >> [   56.619937] mtk_sgmii_select_pcs: id=1
> >> [   56.623690] mtk_pcs_config: interface=4
> >> [   56.627511] offset:0 0x140
> >> [   56.627513] offset:4 0x4d544950
> >> [   56.630215] offset:8 0x20
> >> [   56.633340] forcing AN
> >> [   56.638292] mtk_pcs_config: rgc3=0x0, advertise=0x1 (changed), link_timer=1600000,  sgm_mode=0x103, bmcr=0x1000, use_an=1
> >> [   56.649226] mtk_pcs_link_up: interface=4
> >> [   56.653135] offset:0 0x81140
> >> [   56.653137] offset:4 0x4d544950
> >> [   56.656001] offset:8 0x1
> >> [   56.659137] mtk_soc_eth 15100000.ethernet wan: Link is Up - 1Gbps/Full - flow control rx/tx
> >
> > Thanks - there seems to be something weird with the bmcr value printed
> > above in the mtk_pcs_config line.
> >
> > You have bmcr=0x1000, but the code sets two bits - SGMII_AN_RESTART and
> > SGMII_AN_ENABLE which are bits 9 and 12, so bmcr should be 0x1200, not
> > 0x1000. Any ideas why?
> 
> No, not really
> 
> > Can you also hint at what the bits in the PHY register you quote mean
> > please?
> 
> This could very well be a red herring.  It's the only difference I've
> been able to spot, but I have no idea what it means.
> 
> This is an attempt at reformatting the pdf tables for email.  Hope it's
> readable: 

I found the document for the PHY at:

https://assets.maxlinear.com/web/documents/617792_gpy212b1vc_gpy212c0vc_ds_rev1.3.pdf

It seems as I suspected, the PHY has not completed SGMII AN. Please
can you read register 8 when operating at 1G speeds as well
(VSPEC1_SGMII_CTRL)? Thanks.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!



More information about the Linux-mediatek mailing list