Re: [PATCH] pnfsd-file: cause oops when stop the nfs server

[Christoph Hellwig <hch@infradead.org>]

On Wed, Sep 25, 2013 at 10:58:55AM -0400, Benny Halevy wrote:
> Chenditang, I see that this patch was never subimtted upstream.
> Is there anything specific to pnfs that causes this bug?

It's not.  I've just reproduced it without pnfs, by shutting down
the server after running xfstests over a NFSv4.1 mount.

> The fix itself is completely generic
> and if so can be back-ported to stable since v3.8
> 
> Thanks,
> 
> Benny
> 
> On 2013-04-24 00:46, chenditang wrote:
> > mount nfs dir in the client, and then restart the NFS service in MDS,
> > that will cause oops for client_mutex_owner is NULL in the
> > destroy_client() function.
> > 
> > kernel BUG at fs/nfsd/nfs4state.c:1130!
> > invalid opcode: 0000 [#1] SMP
> > Modules linked in: nfsd(OF) lockd exportfs nfs_acl auth_rpcgss autofs4
> > dlm sctp libcrc32c configfs sunrpc be2iscsi iscsi_boot_sysfs bnx2i cnic
> > uio cxgb4i cxgb4 cxgb3i libcxgbi cxgb3 mdio ib_iser rdma_cm ib_cm iw_cm
> > ib_sa ib_mad ib_core ib_addr ipv6 iscsi_tcp libiscsi_tcp libiscsi
> > scsi_transport_iscsi dm_mirror dm_region_hash dm_log dm_mod ppdev
> > parport_pc parport microcode pcspkr i2c_piix4 i2c_core e1000 sg ext4(F)
> > mbcache(F) jbd2(F) sr_mod(F) cdrom(F) sd_mod(F) crc_t10dif(F)
> > pata_acpi(F) ata_generic(F) ata_piix(F) ahci(F) libahci(F) [last
> > unloaded: speedstep_lib]
> > CPU 0
> > Pid: 2893, comm: nfsd Tainted: GF O 3.8.0-rc4_fl+ #2 innotek GmbH
> > VirtualBox/VirtualBox
> > RIP: 0010:[<ffffffffa045a4cf>] [<ffffffffa045a4cf>]
> > destroy_client+0x2ff/0x330 [nfsd]
> > RSP: 0018:ffff8800256b9d38 EFLAGS: 00010203
> > RAX: 0000000000000010 RBX: ffff880025630000 RCX: 0000000000001440
> > RDX: 00000000000025fc RSI: 0000000000000082 RDI: 0000000000000246
> > RBP: ffff8800256b9d88 R08: ffffffff81cdfba0 R09: 0000000000007768
> > R10: 00000000000001b1 R11: 00000000000001b1 R12: ffff880037988400
> > R13: ffff88003d58ea80 R14: ffffffff81ab2080 R15: 0000000000000000
> > FS: 0000000000000000(0000) GS:ffff88003fc00000(0000) knlGS:0000000000000000
> > CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> > CR2: 00000000006d39d8 CR3: 00000000255be000 CR4: 00000000000006f0
> > DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> > DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> > Process nfsd (pid: 2893, threadinfo ffff8800256b8000, task ffff88003d58ea80)
> > Stack:
> > ffff8800256b9d78 ffff88003d51bdc0 ffff8800256b9d58 ffff880037988400
> > ffffffff81ab2080 ffff880037988400 0000000000000020 ffffffffa04709b0
> > ffffffff81ab2080 0000000000000000 ffff8800256b9dc8 ffffffffa045a555
> > Call Trace:
> > [<ffffffffa045a555>] nfs4_state_destroy_net+0x55/0x120 [nfsd]
> > [<ffffffffa045a718>] nfs4_state_shutdown_net+0xf8/0x140 [nfsd]
> > [<ffffffffa04349b0>] ? nfsd_pool_stats_release+0x50/0x50 [nfsd]
> > [<ffffffffa0434835>] nfsd_shutdown_net+0x35/0x60 [nfsd]
> > [<ffffffffa04348ad>] nfsd_last_thread+0x4d/0x80 [nfsd]
> > [<ffffffffa037f3e5>] svc_shutdown_net+0x35/0x40 [sunrpc]
> > [<ffffffffa0434935>] nfsd_destroy+0x55/0x80 [nfsd]
> > [<ffffffffa0434ab4>] nfsd+0x104/0x130 [nfsd]
> > [<ffffffffa04349b0>] ? nfsd_pool_stats_release+0x50/0x50 [nfsd]
> > [<ffffffff8107951e>] kthread+0xce/0xe0
> > [<ffffffff81079450>] ? kthread_freezable_should_stop+0x70/0x70
> > [<ffffffff8155c56c>] ret_from_fork+0x7c/0xb0
> > [<ffffffff81079450>] ? kthread_freezable_should_stop+0x70/0x70
> > Code: 38 d1 e0 48 89 df e8 81 38 d1 e0 e9 2f ff ff ff 0f 1f 40 00 48 c7
> > c7 72 8a 46 a0 31 c0 e8 86 6b 0f e1 e9 70 fd ff ff 0f 0b eb fe <0f> 0b
> > eb fe 0f 0b 66 66 2e 0f 1f 84 00 00 00 00 00 eb f3 0f 0b
> > RIP [<ffffffffa045a4cf>] destroy_client+0x2ff/0x330 [nfsd]
> > RSP <ffff8800256b9d38>
> > ---[ end trace c2d9f251eabc7c2d ]---
> > 
> > Signed-off-by: chendt.fnst <chendt.fnst@cn.fujitsu.com>
> > Reviewed-by: fanchaoting <fanchaoting@cn.fujitsu.com>
> > ---
> > fs/nfsd/nfs4state.c | 3 +++
> > 1 file changed, 3 insertions(+)
> > 
> > diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> > index 13bc266..40b2348 100644
> > --- a/fs/nfsd/nfs4state.c
> > +++ b/fs/nfsd/nfs4state.c
> > @@ -4936,6 +4936,8 @@ nfs4_state_destroy_net(struct net *net)
> > struct nfsd_net *nn = net_generic(net, nfsd_net_id);
> > struct rb_node *node, *tmp;
> > 
> > + nfs4_lock_state();
> > +
> > for (i = 0; i < CLIENT_HASH_SIZE; i++) {
> > while (!list_empty(&nn->conf_id_hashtbl[i])) {
> > clp = list_entry(nn->conf_id_hashtbl[i].next, struct nfs4_client,
> > cl_idhash);
> > @@ -4952,6 +4954,7 @@ nfs4_state_destroy_net(struct net *net)
> > destroy_client(clp);
> > }
> > 
> > + nfs4_unlock_state();
> > kfree(nn->sessionid_hashtbl);
> > kfree(nn->lockowner_ino_hashtbl);
> > kfree(nn->ownerstr_hashtbl);
> > 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
---end quoted text---