| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20171121125047.tlhyaemp5vtqkyvc@wfg-t540p.sh.intel.com>
Date: Tue, 21 Nov 2017 20:50:47 +0800
From: Fengguang Wu <fengguang.wu@...el.com>
To: Peter Zijlstra <peterz@...radead.org>
Cc: Ingo Molnar <mingo@...hat.com>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Chris Wilson <chris@...is-wilson.co.uk>,
linux-kernel@...r.kernel.org, lkp@...org
Subject: [test_cycle_work] WARNING: possible circular locking dependency
detected
Hello,
FYI this happens in mainline kernel 4.14.0-06692-g5bbcc0f.
It looks like a new regression around 4.14-rc1.
It occurs in 2 out of 2 boots.
[ 2.026528] spin_lock-torture: Creating lock_torture_stats task
[ 2.027461] spin_lock-torture: lock_torture_writer task started
[ 2.028686] spin_lock-torture: lock_torture_stats task started
[ 2.531239]
[ 2.531439] ======================================================
[ 2.532005] WARNING: possible circular locking dependency detected
[ 2.532570] 4.14.0-06692-g5bbcc0f #1 Not tainted
[ 2.532979] ------------------------------------------------------
[ 2.533515] kworker/u4:1/54 is trying to acquire lock:
[ 2.533967] (ww_class_mutex){+.+.}, at: [<44567940>] test_cycle_work+0xf0/0x460:
test_cycle_work at kernel/locking/test-ww_mutex.c:268
[ 2.534473]
[ 2.534473] but now in release context of a crosslock acquired at the following:
[ 2.534473] ((completion)&cycle->b_signal){+.+.}, at: [<4539ccd5>] wait_for_completion+0x25/0x40:
wait_for_completion at kernel/sched/completion.c:144
[ 2.534473]
[ 2.534473] which lock already depends on the new lock.
[ 2.534473]
[ 2.536776]
[ 2.536776] the existing dependency chain (in reverse order) is:
[ 2.536776]
[ 2.536776] -> #1 ((completion)&cycle->b_signal){+.+.}:
[ 2.536776]
[ 2.536776] -> #0 (ww_class_mutex){+.+.}:
[ 2.536776] test_cycle_work+0xf0/0x460:
test_cycle_work at kernel/locking/test-ww_mutex.c:268
[ 2.536776]
[ 2.536776] other info that might help us debug this:
[ 2.536776]
[ 2.536776] Possible unsafe locking scenario by crosslock:
[ 2.536776]
[ 2.536776] CPU0 CPU1
[ 2.536776] ---- ----
[ 2.536776] lock(ww_class_mutex);
[ 2.536776] lock((completion)&cycle->b_signal);
[ 2.536776] lock(ww_class_mutex);
[ 2.536776] unlock((completion)&cycle->b_signal);
[ 2.536776]
[ 2.536776] *** DEADLOCK ***
[ 2.536776]
[ 2.536776] 5 locks held by kworker/u4:1/54:
[ 2.536776] #0: ((wq_completion)"test-ww_mutex"){+.+.}, at: [<444d5dc4>] process_one_work+0x294/0x1210:
process_one_work at kernel/workqueue.c:2083
[ 2.536776] #1: ((work_completion)(&cycle->work)){+.+.}, at: [<444d5dc4>] process_one_work+0x294/0x1210:
process_one_work at kernel/workqueue.c:2083
[ 2.536776] #2: (ww_class_acquire){+.+.}, at: [<444d5f2d>] process_one_work+0x3fd/0x1210:
__read_once_size at include/linux/compiler.h:178
(inlined by) atomic_read at arch/x86/include/asm/atomic.h:27
(inlined by) static_key_count at include/linux/jump_label.h:191
(inlined by) static_key_false at include/linux/jump_label.h:201
(inlined by) trace_workqueue_execute_end at include/trace/events/workqueue.h:112
(inlined by) process_one_work at kernel/workqueue.c:2117
[ 2.536776] #3: (ww_class_mutex){+.+.}, at: [<44567940>] test_cycle_work+0xf0/0x460:
test_cycle_work at kernel/locking/test-ww_mutex.c:268
[ 2.536776] #4: (&x->wait#7){....}, at: [<4453c625>] complete+0x25/0xa0:
complete at kernel/sched/completion.c:35
[ 2.536776]
[ 2.536776] stack backtrace:
[ 2.536776] CPU: 1 PID: 54 Comm: kworker/u4:1 Not tainted 4.14.0-06692-g5bbcc0f #1
[ 2.536776] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
[ 2.536776] Workqueue: test-ww_mutex test_cycle_work
[ 2.550112] Call Trace:
[ 2.550112] ? dump_stack+0x153/0x1fa:
__dump_stack at lib/dump_stack.c:17
(inlined by) dump_stack at lib/dump_stack.c:53
[ 2.550112] ? print_circular_bug+0x1d3/0x1f0:
print_circular_bug at kernel/locking/lockdep.c:1272
[ 2.550112] ? check_prev_add+0xca/0xc90:
check_prev_add at kernel/locking/lockdep.c:1915
[ 2.550112] ? __lock_acquire+0x697/0x1370:
__lock_acquire at kernel/locking/lockdep.c:3503
[ 2.550112] ? commit_xhlock+0x330/0x5d0:
commit_xhlock at kernel/locking/lockdep.c:5026
[ 2.550112] ? __print_lock_name+0x130/0x130:
copy_trace at kernel/locking/lockdep.c:4986
[ 2.550112] ? lock_commit_crosslock+0x1c4/0x3c0:
commit_xhlocks at kernel/locking/lockdep.c:5070
(inlined by) lock_commit_crosslock at kernel/locking/lockdep.c:5109
[ 2.550112] ? complete+0x3d/0xa0:
complete at kernel/sched/completion.c:42
[ 2.550112] ? test_cycle_work+0x106/0x460:
test_cycle_work at kernel/locking/test-ww_mutex.c:269
[ 2.550112] ? process_one_work+0x3fd/0x1210:
__read_once_size at include/linux/compiler.h:178
(inlined by) atomic_read at arch/x86/include/asm/atomic.h:27
(inlined by) static_key_count at include/linux/jump_label.h:191
(inlined by) static_key_false at include/linux/jump_label.h:201
(inlined by) trace_workqueue_execute_end at include/trace/events/workqueue.h:112
(inlined by) process_one_work at kernel/workqueue.c:2117
[ 2.550112] ? worker_thread+0x3e3/0xae0:
worker_thread at kernel/workqueue.c:2246
[ 2.550112] ? kthread+0x1ac/0x220:
kthread at kernel/kthread.c:238
[ 2.550112] ? rescuer_thread+0x640/0x640:
worker_thread at kernel/workqueue.c:2189
[ 2.550112] ? __kthread_bind+0x50/0x50:
kthread at kernel/kthread.c:198
[ 2.550112] ? ret_from_fork+0x19/0x30:
ret_from_fork at arch/x86/entry/entry_32.S:299
[ 8.624965] torture_init_begin: Refusing rcu init: spin_lock running.
[ 8.626961] torture_init_begin: One torture test at a time!
[ 9.419528] Initialise system trusted keyrings
[ 9.420475] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[ 9.421452] zbud: loaded
Attached the full dmesg, kconfig and reproduce scripts.
Thanks,
Fengguang
View attachment "dmesg-vm-lkp-hsw01-yocto-i386-16:20171116073239:i386-randconfig-c0-10271823:4.14.0-06692-g5bbcc0f:1" of type "text/plain" (58558 bytes)
View attachment ".config" of type "text/plain" (101310 bytes)
View attachment "job-script" of type "text/plain" (3769 bytes)
View attachment "reproduce-vm-lkp-hsw01-yocto-i386-16:20171116073239:i386-randconfig-c0-10271823:4.14.0-06692-g5bbcc0f:1" of type "text/plain" (1996 bytes)
Powered by blists - more mailing lists