Re: NFS panic: ^C on cp

From: Attilio Rao <attilio_at_freebsd.org>
Date: Fri, 1 Jan 2010 19:32:53 +0100
2009/12/30 Lucius Windschuh <lwindschuh_at_googlemail.com>:
> Hi,
> I'm using CURRENT r201138 and attending the 26C3, where we have access
> to a NetApp filer (nfs://81.163.130.102/vol/fast, may work also for
> external IPs).
> It works well. But the following command sequence results with high
> probability, not always, in a panic:
> # mount_nfs -o nfsv3,rsize=32768,wsize=32768,intr 81.163.130.102:/vol/fast /mnt
> # cd /mnt/
> # cp some_large_file /some/where/local/
> (wait a while)
> ^C
> (panic: Assertion lk->lk_exslpfail == 0 failed at
> /usr/src/sys/kern/kern_lock.c:962)
>
> I
> Backtrace:
> Tracing pid 2754 tid 100209 td 0xc7580240
> kdb_enter(c09ec974,c09ec974,c09c2936,eb402af4,1,...) at kdb_enter+0x3a
> panic(c09c2936,c09e9af9,c09e98f2,3c2,c6511030,...) at panic+0x136
> __lockmgr_args(daf26180,100000,0,0,0,...) at __lockmgr_args+0xe50
> bqrelse(daf26120,0,c09f7b16,d3e,eb402c54,...) at bqrelse+0x44a
> bufdone_finish(daf26120,daf26120,daf26120,0,eb402cbc,...) at
> bufdone_finish+0x3ff
> bufdone(daf26120,eb402c80,c8119080,eb402cbc,c06c7795,...) at bufdone+0x6a
> nfs_doio(c8de7440,daf26120,c8119080,0,3c00,...) at nfs_doio+0x67b
> nfssvc_iod(c0c23f80,eb402d38,c09e7dea,343,c7df3000,...) at nfssvc_iod+0x27c
> fork_exit(c0878520,c0c23f80,eb402d38) at fork_exit+0xb8
> fork_trampoline() at fork_trampoline+0x8
> --- trap 0, eip = 0, esp = 0xeb402d70, ebp = 0 ---
>
> With PID  2754 being: [nfsiod 0]


Lucius,
may you be willing to test this patch and re-create the test case please?:
http://www.freebsd.org/~attilio/lockmgr_fix.diff

Thanks,
Attilio


-- 
Peace can only be achieved by understanding - A. Einstein
Received on Fri Jan 01 2010 - 17:33:02 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:59 UTC