sungem on imac G5

[Markus Demleitner]

Hi all,

Sorry for lowering the discussion level, but I believe this is most
appropriate mailing list for my woes.

My problem is that with the 2.6.11.2 kernel, patched with benh's
recent patches (the sungem driver has date 
2005-03-10 23:04:36.000000000 +0000), networking on an iMac G5 works
like a charm when OF hasn't touched the network card.

If I boot from the network (yaboot, getting yaboot, the kernel, and
initrd via tftp) *or* have OF put its console on the network (i.e.,
as long as OF touches the chip in a serious way), however, I get
"GMAC PHY not responding" when the chip is initialized by the kernel,
more exactly:

sungem.c: v0.98 8/24/03 David S. Miller (davem at redhat.com)
eth%d: GMAC PHY not responding !
[ Just an aside: the blank before the bang should go away:-)
I admit to not having tried to investigate why the %d is left in
gp->dev->name, but since it's filled in later I figured that wouldn't
be the problem ]
eth0: Sun GEM (PCI) 10/100/1000BaseT Ethernet 00:0a:XXXXXXXX
eht0: Found no PHY

Experimentally, I raised the delay in
#ifdef CONFIG_PPC_PMAC
			pmac_call_feature(PMAC_FTR_GMAC_PHY_RESET, gp->of_node, 0, 0);
			msleep(20);
#endif
to up to 2000 without really knowing what I was doing, but
(fortunately) that didn't help.  I've checked that that code really
is executed.

Now, I don't even know what a PHY is, really, and thus this cry for
help:  Is there any way to perform a really deep, deep reset on the
sungem device?  Is it just that Apple chose some not-quite-known
PHY in the G5?

Thanks for any pointers,

        Markus