J. Bruce Fields wrote: > On Fri, Jan 22, 2010 at 03:03:25PM -0800, Andrew Morton wrote: > >> (cc linux-nfs) >> >> On Thu, 14 Jan 2010 09:53:04 -0800 >> Harry Edmon wrote: >> >> >>> I have a single quad-core Xeon machine running 2.6.32.2 that keeps >>> giving me "WARNING: at lib/kref.c:43 kref_get+0x,23/0x2b()". This >>> machine is mainaly used as a NFS server. I have attached the kern.log >>> and my config file. Any ideas? >>> > > Are you using NFSv4, or only v2/v3? Looking at the trace: > > Jan 11 06:01:51 t kernel: [1311646.832173] ------------[ cut here ]------------ > Jan 11 06:01:53 t kernel: [1311646.860891] WARNING: at lib/kref.c:43 kref_get+0x23/0x2b() > NFSv4 and NFSv3. > OK, so that's the warning that we're trying to get something with > reference count 0. > > Jan 11 06:01:53 t kernel: [1311646.894782] Hardware name: X7DB8 > Jan 11 06:01:53 t kernel: [1311646.915165] Modules linked in: cpufreq_userspace cpufreq_stats cpufreq_powersave cpufreq_conservative nfsd nfs lockd nfs_acl auth_rpcgss sunrpc bridge stp llc ipv6 i5000_edac edac_core i2c_i801 parport_pc parport psmouse container i2c_core ioatdma dca i5k_amb serio_raw pcspkr button processor joydev evdev xfs exportfs sg usbhid sr_mod hid cdrom ata_piix floppy uhci_hcd ehci_hcd ata_generic 3w_9xxx e1000e thermal fan raid1 md_mod sd_mod ahci libata scsi_mod [last unloaded: scsi_wait_scan] > Jan 11 06:01:53 t kernel: [1311647.178038] Pid: 3068, comm: nfsd Tainted: G D W 2.6.32.2 #1 > > What does the very first oops, bug, or warning you get look like? (If > I'm reading the tain information right, this isn't the first one.) > Attached is the kern.log that contains the first warning. -- Dr. Harry Edmon E-MAIL: harry@atmos.washington.edu 206-543-0547 FAX: 206-543-0308 harry@washington.edu, harry@uw.edu Director of IT, College of the Environment and Director of Computing, Dept of Atmospheric Sciences University of Washington, Box 351640, Seattle, WA 98195-1640