both running 1.3.68
I strated getting them after one of my NFS servers died:
Feb 23 23:53:43 garfield kernel: NFS server rocker not responding, still trying.
here I killed a running 'df'
Feb 23 23:56:46 garfield kernel: nfs_statfs: statfs error = 512
Feb 23 23:58:08 garfield kernel: double lock on device queue!
Feb 23 23:58:08 garfield kernel: Socket destroy delayed
Feb 23 23:58:18 garfield kernel: Socket destroy delayed
Feb 23 23:58:28 garfield kernel: Socket destroy delayed
Feb 23 23:58:58 garfield last message repeated 3 times
Feb 23 23:59:58 garfield last message repeated 6 times
Feb 24 00:00:48 garfield last message repeated 5 times
.... lot of them...
when the NFS server come back to life:
Feb 24 00:52:51 garfield kernel: NFS server rocker OK.
Feb 24 00:52:51 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:52:58 garfield kernel: Socket destroy delayed
Feb 24 00:53:01 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:53:06 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:53:08 garfield kernel: Socket destroy delayed
Feb 24 00:53:11 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:53:16 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:53:18 garfield kernel: Socket destroy delayed
Feb 24 00:53:21 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:53:26 garfield kernel: nfs_statfs: statfs error = 13
Feb 24 00:53:28 garfield kernel: Socket destroy delayed
....
These was so annoying that I rebooted the machine. After some time this
happened again. An interesting thing is that on the other machine I
stopped getting the "Socket destroy delayed" message after the NFS server
come back. (and not getting statfs=13)
with older kernels I did not get these messages altought the NFS server
often dies.
---------------------- another problem
interesting effect: About an hour ago every elf program stopped
working....(I started migrating to elf but there are a still many a.out
programs on my system)
I brought an a.out strace from another machine and in showed me that
the programs die right after they opened /dev/zero.
After some time I got this:
Feb 24 21:47:20 garfield kernel: Internal error: bad swap-device
Feb 24 21:47:20 garfield kernel: Trying to free nonexistent swap-page
Feb 24 21:47:20 garfield last message repeated 117 times
only only one time.
then my roommate started playing linuxsdoom..... after some time we hear
a beep and doom stopped working... I was able to switch to another console
and kill doom. After that ELF programs WORKED again.
I don't understand it.
What was this??
gyula
-- Gyula Matics mgyula@sch.bme.hu Linux!
486DX4-100, 16M ram, 16M swap,IDE, ne2000 and smc ultra with ip masq. and loop device compiled in.