kern/106243: [nve] double fault panic in if_nve.c on high loads

[Jason Bacon]

The following reply was made to PR kern/106243; it has been noted by GNATS.

From: Jason Bacon <jbacon at mcw.edu>
To: Kris Kennaway <kris at FreeBSD.org>
Cc: freebsd-gnats-submit at FreeBSD.org, Remko Lodder <remko at elvandar.org>
Subject: Re: kern/106243: [nve] double fault panic in if_nve.c on high loads
Date: Fri, 22 Feb 2008 13:17:05 -0600

 One other thing, that you may already be aware of: The problems seem to 
 have started with FreeBSD 6.2.  I have a couple of PC-BSD 1.3 (FreeBSD 
 6.1) workstations, and a FreeBSD 6.1-AMD64 RAID backup server that use 
 nve with no problems.  The RAID server backs up about 40 workstations 
 every night using rsync, mostly over NFS.
 
 FreeBSD shire bacon ~ 214: uptime
  1:14PM  up 406 days,  2:10, 1 user, load averages: 0.00, 0.01, 0.00
 
 FreeBSD shire bacon ~ 215: vmstat -i
 interrupt                          total       rate
 irq1: atkbd0                        2290          0
 irq6: fdc0                             9          0
 irq12: psm0                        30046          0
 irq14: ata0                     74728804          2
 irq15: ata1                           47          0
 irq16: fwohci0+               2203258885         62
 irq22: nve0 ehci0            36495718160       1040
 cpu0: timer                  70173027567       2000
 Total                       108946765808       3105
 
 FreeBSD shire bacon ~ 216: ifconfig
 fwe0: 
 flags=108943<UP,BROADCAST,RUNNING,PROMISC,SIMPLEX,MULTICAST,NEEDSGIANT> 
 mtu 1500
         options=8<VLAN_MTU>
         inet6 fe80::11:d8ff:feba:18f7%fwe0 prefixlen 64 scopeid 0x1
         ether 02:11:d8:ba:18:f7
         ch 1 dma 0
 nve0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
         inet6 fe80::217:31ff:fece:5854%nve0 prefixlen 64 scopeid 0x2
         inet 141.106.220.35 netmask 0xffffff00 broadcast 141.106.220.255
         ether 00:17:31:ce:58:54
         media: Ethernet autoselect (100baseTX <full-duplex>)
         status: active
 plip0: flags=108810<POINTOPOINT,SIMPLEX,MULTICAST,NEEDSGIANT> mtu 1500
 lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 16384
         inet6 ::1 prefixlen 128
         inet6 fe80::1%lo0 prefixlen 64 scopeid 0x4
         inet 127.0.0.1 netmask 0xff000000
 
 FreeBSD shire bacon ~ 217: df
 Filesystem  1K-blocks        Used      Avail Capacity  Mounted on
 /dev/ad0s1a     507630      90390     376630    19%    /
 devfs                1          1          0   100%    /dev
 /dev/ad0s1d    2010670     289638    1560180    16%    /var
 /dev/ad0s1e     507630       3664     463356     1%    /tmp
 /dev/ad0s1f   70653456   11305760   53695420    17%    /usr
 /dev/da0p1  7093584556 5110803188 1415294604    78%    /backup
 
     -J
 
 Kris Kennaway wrote:
 > Jason Bacon wrote:
 >>
 >> Just to be thorough, I'm going to do a minimal 7.0 install on a spare 
 >> disk, pop it into one of of the problem machines, and run an rsync 
 >> over NFS with the nfe driver.
 >>
 >> I'll get back to you ASAP.
 >
 > Thanks.  We should probably add a warning to the nve manpage and/or 
 > GENERIC that it has known bugs.
 >
 > Kris
 >
 >>
 >>    -J
 >>
 >> Kris Kennaway wrote:
 >>> Jason Bacon wrote:
 >>>> The following reply was made to PR kern/106243; it has been noted 
 >>>> by GNATS.
 >>>>
 >>>> From: Jason Bacon <jbacon at mcw.edu>
 >>>> To: bug-followup at freebsd.org, littlesavage at orionet.ru
 >>>> Cc:  Subject: Re: kern/106243: [nve] double fault panic in if_nve.c 
 >>>> on high loads
 >>>> Date: Thu, 21 Feb 2008 13:55:40 -0600
 >>>>
 >>>>  I may be seeing the same problem on 6.3-RELEASE.  I can 
 >>>> consistently cause a  reboot by running an rsync backup over nfs.  
 >>>> Popped in a Netgear card (dc  driver) and all is well.
 >>>
 >>> And does nfe work for you?
 >>>
 >>> Kris
 >>
 >>
 >