[LU-4127] Loading fld module crashes kernel in mainline 3.12-rc.5

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.4.1
• Type: Bug
• Priority: Critical
• Reporter: Roland Fehrenbacher
• Assignee: WC Triage
• Resolution: Fixed

Environment

Plain 3.12-rc5

• Severity: 4

Description

Trigger Bug

• modprobe fld

[ 395.749857] LustreError: 8638:0:(lu_object.c:1314:lu_context_key_register()) ASSERTION( key->lct_init != ((void *)0) ) failed:
[ 395.761503] LustreError: 8638:0:(lu_object.c:1314:lu_context_key_register()) LBUG
[ 395.769982] Kernel panic - not syncing: LBUG
[ 395.774365] CPU: 1 PID: 8638 Comm: modprobe Tainted: G C 3.12.0-rc5-qlustar-gen-1 #1
[ 395.783134] Hardware name: Supermicro X8DTT-H/X8DTT-H, BIOS 080016 03/08/2010
[ 395.790525] ffffffffa0a27cd8 ffff880c186b1c38 ffffffff81549200 0000000000000007
[ 395.798332] ffffffffa051873d ffff880c186b1cb8 ffffffff81546827 ffffffff816746fd
[ 395.806149] ffffffff00000008 ffff880c186b1cc8 ffff880c186b1c68 ffffffffa06af350
[ 395.813963] Call Trace:
[ 395.816516] [<ffffffff81549200>] dump_stack+0x46/0x58
[ 395.821757] [<ffffffff81546827>] panic+0xb6/0x1c6
[ 395.826667] [<ffffffffa04fcb50>] lbug_with_loc+0xb0/0xb0 [libcfs]
[ 395.832963] [<ffffffffa06902b1>] lu_context_key_register+0x131/0x140 [obdclass]
[ 395.840514] [<ffffffffa0a2a000>] ? 0xffffffffa0a29fff
[ 395.845756] [<ffffffffa0a2a000>] ? 0xffffffffa0a29fff
[ 395.850999] [<ffffffffa0a2a041>] fld_mod_init+0x41/0x45 [fld]
[ 395.856936] [<ffffffff81000239>] do_one_initcall+0x49/0x170
[ 395.862709] [<ffffffff8106ad3e>] ? __blocking_notifier_call_chain+0x5e/0x80
[ 395.869863] [<ffffffff810a7f7a>] load_module+0x1bca/0x22f0
[ 395.875543] [<ffffffff810a56b0>] ? __symbol_get+0x70/0x70
[ 395.881132] [<ffffffff810a8738>] SyS_init_module+0x98/0xc0
[ 395.886810] [<ffffffff81556da6>] system_call_fastpath+0x1a/0x1f

Issue Links

• is related to: LU-4416 support for 3.12 linux kernel

Comments

[Roland Fehrenbacher]

Attached used .config

[Roland Fehrenbacher]

The attached patch that partly reverts e62e5d9251bf8efd08318e78bcaca86fde228383 (staging/lustre/fld: prepare FLD module for client server split) fixes the problem. Might be a good idea to get this into 3.12 final, so it will at least be possible to do a mount and some writing. Unfortunately deleting won't be possible due to another critical bug (see related Jira report https://jira.hpdd.intel.com/browse/LU-4157).

[Dmitry Eremin (Inactive)]

Peng Tao fixed this in upstream. See https://git.kernel.org/cgit/linux/kernel/git/gregkh/staging.git/commit/?h=staging-next&id=86bac591def1a5b7060c0834828b1eaabfe7f0a7