[LU-15808] ptlrpc lockdep warning "do not call blocking ops when !TASK_RUNNING"

• Type: Bug
• Resolution: Unresolved
• Priority: Minor
• Affects Version/s: Lustre 2.15.0
• Severity: 3
• Assignee: WC Triage
• Reporter: Andreas Dilger

Description

Saw a lockdep warning in Gerrit Janitor client console logs when running runtests:

http://testing.linuxhacker.ru:3333/lustre-reports/22787/results.html
http://testing.linuxhacker.ru:3333/lustre-reports/22787/testresults/runtests-rhel8-zfs-rhel8.5_x86_64-rhel8.5_x86_64/
http://testing.linuxhacker.ru:3333/lustre-reports/22787/testresults/runtests-rhel8-zfs-rhel8.5_x86_64-rhel8.5_x86_64/oleg240-client-console.txt

[  272.868594] do not call blocking ops when !TASK_RUNNING; state=1 set at [<000000007e590e9d>] prepare_to_wait_event+0xc9/0x2a0
[  272.870611] WARNING: CPU: 1 PID: 4916 at kernel/sched/core.c:6733 __might_sleep+0xa3/0xc0
[  272.877791] CPU: 1 PID: 4916 Comm: mount.lustre Kdump: loaded Tainted: G           O     --------- -  - 4.18.0rh8.5-debug #2
[  272.879551] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
[  272.880475] RIP: 0010:__might_sleep+0xa3/0xc0
272.894276] Call Trace:
[  272.894639]  slab_pre_alloc_hook.constprop.45+0x12d/0x1f0
[  272.895479]  __kmalloc+0x8b/0x3d0
[  272.897379]  null_alloc_repbuf+0x172/0x340 [ptlrpc]
[  272.898261]  sptlrpc_cli_alloc_repbuf+0x1fe/0x290 [ptlrpc]
[  272.899160]  ptl_send_rpc+0x131b/0x1a10 [ptlrpc]
[  272.902503]  ptlrpc_check_set+0x279f/0x2f80 [ptlrpc]
[  272.903480]  ptlrpc_set_wait+0x6ef/0xae0 [ptlrpc]
[  272.905053]  ptlrpc_queue_wait+0xb4/0x2e0 [ptlrpc]
[  272.905939]  ldlm_cli_enqueue+0x633/0xda0 [ptlrpc]
[  272.908533]  mgc_enqueue.isra.10.constprop.18+0x2ab/0x470 [mgc]
[  272.911159]  mgc_process_log+0x123/0xb60 [mgc]
[  272.913066]  do_config_log_add+0x5dd/0x720 [mgc]
[  272.913717]  config_log_find_or_add+0xbc/0x100 [mgc]
[  272.915555]  config_log_add+0x535/0xc90 [mgc]
[  272.916238]  mgc_process_config+0xb74/0x1490 [mgc]
[  272.916997]  lustre_process_log+0x7d4/0xda0 [obdclass]
[  272.920469]  ll_fill_super+0x906/0x14a0 [lustre]
[  272.922096]  lustre_fill_super+0xe8/0x560 [lustre]
[  272.923625]  mount_nodev+0x58/0x100
[  272.924201]  lustre_mount+0x1c/0x30 [lustre]
[  272.924893]  legacy_get_tree+0x38/0x90
[  272.925506]  vfs_get_tree+0x30/0x140
[  272.926118]  do_mount+0xcad/0x10b0
[  272.926658]  ksys_mount+0x11d/0x150
[  272.927271]  __x64_sys_mount+0x29/0x40
[  272.927871]  do_syscall_64+0xd4/0x5a0
[  272.928493]  entry_SYSCALL_64_after_hwframe+0x6a/0xdf

[  272.943745] irq event stamp: 44536
[  272.944345] hardirqs last  enabled at (44535): [<ffffffff81250ce4>] ktime_get_real_ts64+0x224/0x2e0
[  272.946562] hardirqs last disabled at (44536): [<ffffffff820e4df2>] __schedule+0x8b2/0xe50
[  272.949759] softirqs last  enabled at (40638): [<ffffffff81d4a7ae>] sk_common_release+0x8e/0x190
[  272.952854] softirqs last disabled at (40636): [<ffffffff81d4a77a>] sk_common_release+0x5a/0x190

Issue Links

• is related to: LU-16807 Resolve newer debug kernel warnings