Re: Occasional hangups with forcedeth
On Sun, 2011-01-16 at 11:56 +0100, Klaus Pieper wrote:
> Hi gurus,
>
> using
>
> 00:0a.0 Ethernet controller: nVidia Corporation MCP79 Ethernet (rev b1)
>
> I have occasional (i.e. occuring about every 6 weeks) hangups of this
> device.
> Ping to and from the machine does work, but loss is over 70%.
>
> Any hints or workarounds?
>
> Regards,
> Klaus
>
>
> # dmesg |grep -i forcedet
> [ 1.528115] forcedeth: Reverse Engineered nForce ethernet driver.
> Version 0.64.
> [ 1.529046] forcedeth 0000:00:0a.0: PCI INT A -> Link[LMAC] -> GSI 20
> (level, low) -> IRQ 20
> [ 1.529055] forcedeth 0000:00:0a.0: setting latency timer to 64
> [ 1.594154] forcedeth 0000:00:0a.0: ifname eth0, PHY OUI 0x732 @ 3,
> addr 00:01:2e:27:c5:1c
> [ 1.594164] forcedeth 0000:00:0a.0: highdma csum pwrctl gbit lnktim
> msi desc-v3
>
> eth0 Link encap:Ethernet HWaddr .........
> inet addr:192.168.1.2 Bcast:192.168.1.255 Mask:255.255.255.0
> inet6 addr: fe80::201:2eff:fe27:c51c/64 Scope:Link
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:40 errors:86 dropped:0 overruns:0 frame:86
> TX packets:197 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:3391 (3.3 KiB) TX bytes:24074 (23.5 KiB)
> Interrupt:22
>
> Kernel IP routing table
> Destination Gateway Genmask Flags Metric Ref Use
> Iface
> 192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
> 0.0.0.0 192.168.1.1 0.0.0.0 UG 0 0 0 eth0
>
>
> Settings for eth0:
> Supported ports: [ MII ]
> Supported link modes: 10baseT/Half 10baseT/Full
> 100baseT/Half 100baseT/Full
> 1000baseT/Full
> Supports auto-negotiation: Yes
> Advertised link modes: 10baseT/Half 10baseT/Full
> 100baseT/Half 100baseT/Full
> 1000baseT/Full
> Advertised pause frame use: No
> Advertised auto-negotiation: Yes
> Speed: 1000Mb/s
> Duplex: Full
> Port: MII
> PHYAD: 3
> Transceiver: external
> Auto-negotiation: on
> Supports Wake-on: g
> Wake-on: g
> Link detected: yes
>
>
We have had disastrous results with this driver on kernels 2.6.28 and
2.6.29. In our case, they would simply stop working altogether even
when bonded. We were never able to decisively blame it on the driver
but, it was enough of a problem that we replaced them all with Intel
cards. I can decisively mention that the performance and, especially
the packet handling latency -critical for I/O bound operations like
iSCSI - were significantly worse for forcedeth than for the Intel cards
and drivers - John
Reply to: