[LU-8970] Kernel warning on client mount if pool is defined multiple times

• Type: Bug
• Resolution: Unresolved
• Priority: Minor
• Severity: 3
• Bugzilla ID: 18,266
• Assignee: WC Triage
• Reporter: Ben Evans (Inactive)

Description

Steps to reproduce:

1. unmount the lustre FS completely (clients and all targets)
2. Mount just the MGS
3. lctl pool_new (fsname).(poolname)
4. repeat step 3
5. Remount the FS. Client will generate a kernel warning.

This should also work on a filesystem with an MGS mounted on it's own node with no MDT or client.

Generates the following stack trace:
Dec 23 10:31:31 centclient29 kernel: -----------[ cut here ]-----------
Dec 23 10:31:31 centclient29 kernel: WARNING: at fs/proc/generic.c:599 proc_register+0xb9/0x170() (Not tainted)
Dec 23 10:31:31 centclient29 kernel: Hardware name: VMware Virtual Platform
Dec 23 10:31:31 centclient29 kernel: proc_dir_entry 'pools/ssu1' already registered
Dec 23 10:31:31 centclient29 kernel: Modules linked in: osc(U) mgc(U) lustre(U) lov(U) mdc(U) fid(U) lmv(U) fld(U) ksocklnd(U) ptlrpc(U) obdclass(U) lnet(U) libcfs(U) sha512_generic crc32c_intel nfs lockd fscache auth_rpcgss nfs_acl sunrpc autofs4 ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables ipv6 ppdev vmware_balloon parport_pc parport vmxnet3 i2c_piix4 i2c_core sg shpchp ext4 jbd2 mbcache sd_mod crc_t10dif sr_mod cdrom vmw_pvscsi pata_acpi ata_generic ata_piix dm_mirror dm_region_hash dm_log dm_mod [last unloaded: libcfs]
Dec 23 10:31:31 centclient29 kernel: Pid: 16403, comm: llog_process_th Not tainted 2.6.32-573.12.1.el6_lustre.x86_64 #1
Dec 23 10:31:31 centclient29 kernel: Call Trace:
Dec 23 10:31:31 centclient29 kernel: [<ffffffff81077461>] ? warn_slowpath_common+0x91/0xe0
Dec 23 10:31:31 centclient29 kernel: [<ffffffff81077566>] ? warn_slowpath_fmt+0x46/0x60
Dec 23 10:31:31 centclient29 kernel: [<ffffffff81204b19>] ? proc_register+0xb9/0x170
Dec 23 10:31:31 centclient29 kernel: [<ffffffff81204c45>] ? proc_create_data+0x75/0xd0
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa06c0df9>] ? lprocfs_add_simple+0x49/0xd0 [obdclass]
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa0c169a2>] ? lov_pool_new+0x3d2/0x6f0 [lov]
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa06d78cd>] ? class_process_config+0x1e0d/0x24c0 [obdclass]
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa05aac11>] ? libcfs_debug_msg+0x41/0x50 [libcfs]
Dec 23 10:31:31 centclient29 kernel: [<ffffffff81178a2c>] ? __kmalloc+0x21c/0x230
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa06d980a>] ? class_config_llog_handler+0xbfa/0x1d10 [obdclass]
Dec 23 10:31:31 centclient29 kernel: [<ffffffff8153a65e>] ? mutex_lock+0x1e/0x50
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa069d7aa>] ? llog_process_thread+0x8fa/0x1050 [obdclass]
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa069e405>] ? llog_process_thread_daemonize+0x45/0x70 [obdclass]
Dec 23 10:31:31 centclient29 kernel: [<ffffffffa069e3c0>] ? llog_process_thread_daemonize+0x0/0x70 [obdclass]
Dec 23 10:31:31 centclient29 kernel: [<ffffffff810a0fce>] ? kthread+0x9e/0xc0
Dec 23 10:31:31 centclient29 kernel: [<ffffffff8100c28a>] ? child_rip+0xa/0x20
Dec 23 10:31:31 centclient29 kernel: [<ffffffff810a0f30>] ? kthread+0x0/0xc0
Dec 23 10:31:31 centclient29 kernel: [<ffffffff8100c280>] ? child_rip+0x0/0x20
Dec 23 10:31:31 centclient29 kernel: --[ end trace 1aaa1b32eaf573d4 ]--

Issue Links

• is related to: LU-8066 Move lustre procfs handling to sysfs and debugfs.
• is related to: LU-17308 makes "lctl pool_*" more reliable for big configurations