Brad schrieb:
On Monday 27 July 2009 02:00:33 Dorian B|ttner wrote:
Kevin Lo schrieb:
On Sun, 2009-07-26 at 13:13 +0200, Dorian B|ttner wrote:
It looks like this NIC doesn't agree to 1000 full with the switch, other
cards do. Problem didn't however arise with this patch, it also shows
with booting obsd.
Could you try attached diff? Thanks.
>From DragonflyBSD: don't read/write 1000baseT specific MII registers
on 10/100M only device.
Kevin
AT_PCVT -D_KERNEL -Di386 -c ../../../../dev/pci/if_ale.c
../../../../dev/pci/if_ale.c: In function `ale_miibus_writereg':
../../../../dev/pci/if_ale.c:186: warning: `return' with a value, in
function returning void
*** Error code 1
Stop in /usr/src/sys/arch/i386/compile/EEE (line 92 of
/usr/share/mk/sys.mk).
I made it compile by editing return (0); to return; but it doesn't seem
to have any effect on the 1000baseT issue?
The diff won't make any difference for potential GigE operation. It only
affects handling of the L2E Fast Ethernet chipset and preventing certain
registers from being accessed on the PHY.
You're certain this system has a GigE NIC and not FastE? Whether it is
because the system says it does or you've tested another OS. I just want to be
100% certain.
Aw crab. Yes it took me some serious googling to figure out this eee is
supposed to 10/100 only. I was in belief those L1E chips were GigE
capable througout the bank, and I never took really care of it before
(setting it to 1000 full didn't even yield a negative result, though).
Was just that this test drew my attention to it somehow.
