Re: nfe(4) strangeness on the 7-CURRENT

From: Andre Oppermann <andre_at_freebsd.org>
Date: Mon, 11 Jun 2007 21:09:18 +0200
Eygene Ryabinkin wrote:
> Good day.
> 
> I had recently upgraded my amd64 box to the 7-CURRENT and started
> using nfe(4) instead of nve(4),  because the latter was almost
> unusable on the moderate traffic flow from the amd64 box to some
> other machine due to the watchdog timeouts.
> 
> But the stock nfe(4) was not good too: it provoked some kernel
> crashes.  So, I resorted to the nfe(4) driver from
> http://www.f.csce.kyushu-u.ac.jp/~shigeaki/software/freebsd-nfe.html
> 
> Things are working, but I have many kernel messages like this:
> -----
> TCP: [third_local_machine]:65385 to [second_local_machine]:443 tcpflags 0x4<RST>; tcp_input
> : Listen socket: Spurious RST, segment rejected
> TCP: [127.0.0.1]:61775 to [127.0.0.1]:40001 tcpflags 0x11<FIN,ACK>; syncache_expand: Segment failed SYNCOOKIE authentication, segment rejected (probably spoofed)
> TCP: [third_local_machine]:63851 to [second_local_machine]:443 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [third_local_machine]:62901 to [second_local_machine]:443 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [third_local_machine]:55359 to [second_local_machine]:443 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [209.132.176.167]:60779 to [first_local_machine]:25 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [69.147.83.53]:54604 to [first_local_machine]:25 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [third_local_machine]:63187 to [second_local_machine]:443 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [209.132.176.167]:38662 to [first_local_machine]:25 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [209.132.176.167]:38662 to [first_local_machine]:25 tcpflags 0x4<RST>; tcp_input: Listen socket: Spurious RST, segment rejected
> TCP: [127.0.0.1]:60500 to [127.0.0.1]:40001 tcpflags 0x11<FIN,ACK>; syncache_expand: Segment failed SYNCOOKIE authentication, segment rejected (probably spoofed)
> -----
> I would count this as the hardware or driver failures, but the
> messages for the 127.0.0.1 shows that this is not related to the
> packet corruption in the NIC driver.  And SSH transfers are working
> fine, so if there would be packet corruption, then it will broke
> SSH MAC layer.
> 
> The motherboard is ASUS M2NPV-VM.  The device itself is detected
> as
> -----
> nfe0: <NVIDIA nForce 430 MCP13 Networking Adapter> port 0xc800-0xc807 mem 0xfe02b000-0xfe02bfff irq 22 at device 20.0 on pci0
> miibus0: <MII bus> on nfe0
> e1000phy0: <Marvell 88E1116 Gigabit PHY> PHY 1 on miibus0
> e1000phy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseTX-FDX, auto
> nfe0: Ethernet address: 00:18:f3:97:38:85
> nfe0: [FILTER]
> -----
> The kernel signature is '7.0-CURRENT FreeBSD 7.0-CURRENT #3: Fri Jun  8
> 06:26:42 MSD 2007'.
> 
> The non-default part of my -CURRENT is the usb4bsd stack instead
> of the stock one.  But I had the same error messages without usb4bsd.
> 
> Should I just relax and ignore the messages, or this is the sign
> of the brokenness?

These messages are unrelated to your hardware.  There you can relax.
We have a bug in the TCP FSM state transitions which I'm currently
tracking down that indirectly causes the log messages.  You don't
have to worry about it for your case.  TCP connections work fine.

-- 
Andre
Received on Mon Jun 11 2007 - 17:09:17 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:12 UTC