[LU-12009] osp_fid_diff()) ASSERTION( fid_seq(fid1) == fid_seq(fid2) ) failed

• Status: Open
• Project: Lustre
• Affects Version/s: Lustre 2.13.0
• Type: Bug
• Priority: Minor
• Reporter: Oleg Drokin
• Assignee: WC Triage
• Resolution: Unresolved
• Severity: 3

Description

I have this bizzare assertion hit from time to time in recovery-small test 51 or 50.

It's bizzare because in all cases fid1 equals fid2 in the printout, so it's certainly some weird supernarrow race. Ot the difference is in some part of the fid we are not printing?

[116469.024571] Lustre: lustre-OST0001: deleting orphan objects from 0x0:2043 to 0x0:2081
[116469.027318] Lustre: lustre-OST0000: deleting orphan objects from 0x0:2716 to 0x0:2753
[116469.030348] Lustre: lustre-OST0002: deleting orphan objects from 0x0:2587 to 0x0:2625
[116469.045959] Lustre: lustre-OST0003: deleting orphan objects from 0x0:2620 to 0x0:2657
[116469.097240] LustreError: 8095:0:(osp_internal.h:611:osp_fid_diff()) ASSERTION( fid_seq(fid1) == fid_seq(fid2) ) failed: fid1:[0x100010000:0x821:0x0], fid2:[0x100010000:0x821:0x0]
[116469.106713] LustreError: 8095:0:(osp_internal.h:611:osp_fid_diff()) LBUG
[116469.112423] Pid: 8095, comm: mdt01_002 3.10.0-7.6-debug #1 SMP Wed Nov 7 21:55:08 EST 2018
[116469.117067] Call Trace:
[116469.118219]  [<ffffffffa01ba7dc>] libcfs_call_trace+0x8c/0xc0 [libcfs]
[116469.119636]  [<ffffffffa01ba88c>] lbug_with_loc+0x4c/0xa0 [libcfs]
[116469.121194]  [<ffffffffa11ba58d>] osp_precreate_reserve+0x18d/0x7f0 [osp]
[116469.122568]  [<ffffffffa11af829>] osp_declare_create+0x199/0x5b0 [osp]
[116469.123778]  [<ffffffffa1161662>] lod_sub_declare_create+0xe2/0x210 [lod]
[116469.125188]  [<ffffffffa11598ae>] lod_qos_declare_object_on+0xbe/0x3a0 [lod]
[116469.128786]  [<ffffffffa115b00e>] lod_alloc_rr.constprop.19+0xeee/0x1490 [lod]
[116469.131820]  [<ffffffffa11608dd>] lod_qos_prep_create+0x12dd/0x1880 [lod]
[116469.133066]  [<ffffffffa1161098>] lod_prepare_create+0x218/0x2e0 [lod]
[116469.136813]  [<ffffffffa11506de>] lod_declare_striped_create+0x1ee/0x990 [lod]
[116469.141621]  [<ffffffffa1155434>] lod_declare_create+0x204/0x5a0 [lod]
[116469.142560]  [<ffffffffa10110af>] mdd_declare_create_object_internal+0xdf/0x2f0 [mdd]
[116469.144973]  [<ffffffffa100127c>] mdd_declare_create+0x4c/0xc90 [mdd]
[116469.146237]  [<ffffffffa1005357>] mdd_create+0x827/0x1440 [mdd]
[116469.147556]  [<ffffffffa1093a3c>] mdt_reint_open+0x1aac/0x28b0 [mdt]
[116469.148884]  [<ffffffffa10868d0>] mdt_reint_rec+0x80/0x210 [mdt]
[116469.149845]  [<ffffffffa1063723>] mdt_reint_internal+0x6e3/0xab0 [mdt]
[116469.150755]  [<ffffffffa106fdd2>] mdt_intent_open+0x82/0x350 [mdt]
[116469.151991]  [<ffffffffa106de65>] mdt_intent_policy+0x2f5/0xd10 [mdt]
[116469.153226]  [<ffffffffa075ab13>] ldlm_lock_enqueue+0x353/0x9f0 [ptlrpc]
[116469.155701]  [<ffffffffa0782dc7>] ldlm_handle_enqueue0+0xa37/0x1580 [ptlrpc]
[116469.157774]  [<ffffffffa0807742>] tgt_enqueue+0x62/0x210 [ptlrpc]
[116469.169885]  [<ffffffffa0810525>] tgt_request_handle+0xaf5/0x1590 [ptlrpc]
[116469.171348]  [<ffffffffa07b4466>] ptlrpc_server_handle_request+0x256/0xad0 [ptlrpc]
[116469.175541]  [<ffffffffa07b8359>] ptlrpc_main+0xa99/0x1f60 [ptlrpc]
[116469.176928]  [<ffffffff810b4ed4>] kthread+0xe4/0xf0
[116469.178177]  [<ffffffff817c4c77>] ret_from_fork_nospec_end+0x0/0x39