> cool, i've tested your patch and it seems to work. now i will be free
> of that unfriendly netgear driver. :) i tested it on an updated redhat
> 7.2 box. (2.4.9-13smp) it is an asus p2b-d motherboard. (p3 smp,
> 32bitpci).
>
> i did notice some odd dmesg output, however:
>
> eth%d: enabling 64 bit PCI.
> eth%d: enabling optical transceiver
> eth1: ns83820.c v0.13: DP83820 00:40:f4:29:ea:d7 pciaddr=0xe1000000
> irq=12 rev 0x103
> eth1: link now 1000F mbps, full duplex and up.
> eth1: link now 1000F mbps, full duplex and up.
>
> [now keeping in mind i know nothing of linux device drivers...]
>
> this is only a 32bit pci box so why would it enable 64bit pci?
The code reads a 64bit detect flag from the ns chip - so I guess it
must be bogus with some motherboards. Mine is okay. Ben??
> are references to dev->net_dev.name valid before
> register_netdev(&dev->net_dev) in ns83820_init_one()?
Okay, so i'll move the register_netdev call earlier on in the
initialisation and add any necessary unregister call for failures.
> is/why phy_intr() called 2wice?
The card issues multiple interrupts during auto-negotiation. If you
change the dprintk to a printk on the line with the tbisr=, tanar=,
you'll see the details of the phy interrupt. The driver needs a link
status variable so we then only print link status changes when
link status changes. The current problem is purely cosmetic.
> thanks for the patch!
>
> rob.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/