Date: Tue, 11 Oct 2011 10:10:29 -0700 From: YongHyeon PYUN <pyunyh@gmail.com> To: Karim <fodillemlinkarim@gmail.com> Cc: freebsd-net@freebsd.org Subject: Re: if_msk.c link negotiation / packet drops Message-ID: <20111011171029.GA5661@michelle.cdnetworks.com> In-Reply-To: <4E94637A.5090607@gmail.com> References: <4E94637A.5090607@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Oct 11, 2011 at 11:40:42AM -0400, Karim wrote:
> Hi List,
>
> Using a Marvell NIC plugged into a CISCO switch I see the
> auto-negotiation failing and even when forcing the device to full-duplex
> we sometimes see packet drops.
>
> Here is the device description from dmesg:
>
> mskc0: <Marvell Yukon 88E8053 Gigabit Ethernet> port 0xbe00-0xbeff mem
> 0xfdefc000-0xfdefffff irq 16 at device 0.0 on pci1
> msk0: <Marvell Technology Group Ltd. Yukon EC Id 0xb6 Rev 0x02> on mskc0
> msk0: Ethernet address: 00:03:2d:09:94:52
> miibus0: <MII bus> on msk0
> e1000phy0: <Marvell 88E1111 Gigabit PHY> PHY 0 on miibus0
> e1000phy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT,
> 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, auto, auto-flow
> mskc0: [ITHREAD]
>
> The switch its plugged in (Cisco) is configured for 100baseTX full-duplex.
>
> ifconfig reports:
>
> msk0:
> flags=608843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST,SATELLITE,LAN_NET>
> metric 0 mtu 1500
> options=40018<VLAN_MTU,VLAN_HWTAGGING>
The flags and options show that you're using very customized
driver, right?
> ether 00:03:2d:09:94:52
> inet 192.168.122.7 netmask 0xffffff00 broadcast 192.168.122.255
> media: Ethernet autoselect (100baseTX <half-duplex>)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Resolved duplex is half so I guess it would be normal to see
dropped frames which may be triggered by collision.
> status: active
>
> I added a bit of debugging code in e1000phy_status in the hope to
> understand what was going on:
>
> bmsr = PHY_READ(sc, E1000_SR) | PHY_READ(sc, E1000_SR);
> bmcr = PHY_READ(sc, E1000_CR);
> ssr = PHY_READ(sc, E1000_SSR);
>
> printf("%s ssr 0x%x bmsr 0x%x bmcr 0x%x\n", __func__, ssr, bmsr, bmcr);
>
> Which consistently gives:
>
> kernel: e1000phy_status ssr 0x4d00 bmsr 0x796d bmcr 0x1000
>
This register value is consistent with media status of ifconfig
output. The real question is why auto-negotiation does not work
against your switch. It could be a bug in switch side and if you
have to manually set full-duplex with msk(4), please make sure your
switch's duplex configuration also should be manually set to
full-duplex. Otherwise resolved duplex is half.
> Now and then the system will report the device is going inactive and we
> can see packet drops when that happens. This repeat itself once in a
> while without being predictable.
>
Check MAC H/W statistics counter and see any
collisions(dev.msk.0.stats)
> By the way on a 7.4 system this was happening every 10 min. Using a
> driver from FBSD9 back ported to 7.4 we see the issue at a much lower
> frequency (every 30 min) but the issue is still there.
>
> Cheers,
>
> Karim.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20111011171029.GA5661>