Gabor, thanks for the feedback.
>> + ar71xx_eth0_data.phy_if_mode = PHY_INTERFACE_MODE_MII;
>> + ar71xx_eth0_data.speed = SPEED_100;
>> + ar71xx_eth0_data.duplex = DUPLEX_FULL;
>
> You don't have to specify the speed and duplex values if you are using a
> phy_mask.
OK, fixed.
>
>> + ar71xx_eth0_data.phy_mask = 0xffffffff; /* I don't really
>> know the correct value here.
>> + * This selects any present
>> phy.
>> + */
>
> You should be able to determine the correct value from dmesg.
>
Here's the relevant lines from dmesg:
ag71xx_mdio: probed
eth0: Atheros AG71xx at 0xb9000000, irq 4
eth0: connected to PHY at ag71xx-mdio:0c [uid=00008201, driver=Generic PHY]
I don't really understand the mask. I know it determines what phy is
selected, but I'm in the dark about the correct value.
I've tried using both b9000000 and 00008201.
both result in a 'no PHY found with phy_mask' message.
>> set_preinit_iface() {
>> ifname=eth0
>> + [ $(ar71xx_board_name) = "tl-wa901nd-v2" ] && mii-tool -R
>> + /* The TP-link WA901ND V2 requires a hardware reset of
>> + * its ethernet port before it will connect.
>> + */
>> +
>
> This looks strange. The phy core should reset the PHY automatically. Please
> try
> to add this line into the machine setup code, somewhere before the
> 'ar71xx_add_device_eth(0)' call:
>
> ar71xx_eth0_data.reset_bit = RESET_MODULE_GE0_MAC | RESET_MODULE_GE0_PHY;
That works perfectly. Thank you! Using mii-tool was really bugging me, as well.
Once the correct phy_mask value is determined, I'll resubmit the patch.
>
> Regards,
> Gabor
>
Cheers,
Jonathan
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
