Hello,
Has anyone else already run into the locking complaint shown below ?
All I had to do to trigger that complaint is to enable VN2VN support:
# modprobe fcoe
# echo eth0 > /sys/module/libfcoe/parameters/create_vn2vn
Bart.
======================================================
[ INFO: possible circular locking dependency detected ]
3.13.0-rc1-debug+ #1 Not tainted
-------------------------------------------------------
bash/16148 is trying to acquire lock:
(fcoe_config_mutex){+.+.+.}, at: [<ffffffffa0229037>] _fcoe_create+0x27/0x780
[fcoe]
but task is already holding lock:
(ft_mutex){+.+.+.}, at: [<ffffffffa025a690>] fcoe_transport_create+0x20/0x210
[libfcoe]
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #5 (ft_mutex){+.+.+.}:
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffff814e4f82>] mutex_lock_nested+0x72/0x430
[<ffffffffa025a690>] fcoe_transport_create+0x20/0x210 [libfcoe]
[<ffffffff81065bbd>] param_attr_store+0x3d/0x70
[<ffffffff81065b08>] module_attr_store+0x18/0x30
[<ffffffff811ad7e9>] sysfs_write_file+0x179/0x1c0
[<ffffffff81141432>] vfs_write+0xb2/0x1f0
[<ffffffff81141b27>] SyS_write+0x47/0xa0
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
-> #4 (param_lock){+.+.+.}:
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffff814e4f82>] mutex_lock_nested+0x72/0x430
[<ffffffff81065baf>] param_attr_store+0x2f/0x70
[<ffffffff81065b08>] module_attr_store+0x18/0x30
[<ffffffff811ad7e9>] sysfs_write_file+0x179/0x1c0
[<ffffffff81141432>] vfs_write+0xb2/0x1f0
[<ffffffff81141b27>] SyS_write+0x47/0xa0
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
-> #3 (&of->mutex){+.+.+.}:
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffff814e4f82>] mutex_lock_nested+0x72/0x430
[<ffffffff811ac6da>] sysfs_bin_mmap+0x4a/0x120
[<ffffffff811189ed>] mmap_region+0x3cd/0x5a0
[<ffffffff81118f17>] do_mmap_pgoff+0x357/0x3e0
[<ffffffff81105395>] vm_mmap_pgoff+0x65/0xa0
[<ffffffff81117583>] SyS_mmap_pgoff+0x1d3/0x270
[<ffffffff81006a3d>] SyS_mmap+0x1d/0x20
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
-> #2 (&mm->mmap_sem){++++++}:
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffff8110de04>] might_fault+0x84/0xb0
[<ffffffff813e3b07>] dev_ethtool+0x47/0x1b00
[<ffffffff813f534b>] dev_ioctl+0x16b/0x660
[<ffffffff813c198d>] sock_ioctl+0x14d/0x280
[<ffffffff81154ac0>] do_vfs_ioctl+0x300/0x520
[<ffffffff81154d20>] SyS_ioctl+0x40/0x80
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
-> #1 (rtnl_mutex){+.+.+.}:
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffff814e4f82>] mutex_lock_nested+0x72/0x430
[<ffffffff813edab2>] rtnl_lock+0x12/0x20
[<ffffffff813d7b49>] register_netdevice_notifier+0x19/0x1d0
[<ffffffffa024c131>] 0xffffffffa024c131
[<ffffffff810002f2>] do_one_initcall+0xd2/0x180
[<ffffffff810ba356>] load_module+0x1dd6/0x24b0
[<ffffffff810baac5>] SyS_init_module+0x95/0xf0
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
-> #0 (fcoe_config_mutex){+.+.+.}:
[<ffffffff8108a41b>] __lock_acquire+0x17ab/0x1b10
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffff814e4f82>] mutex_lock_nested+0x72/0x430
[<ffffffffa0229037>] _fcoe_create+0x27/0x780 [fcoe]
[<ffffffffa022979e>] fcoe_create+0xe/0x10 [fcoe]
[<ffffffffa025a7b2>] fcoe_transport_create+0x142/0x210 [libfcoe]
[<ffffffff81065bbd>] param_attr_store+0x3d/0x70
[<ffffffff81065b08>] module_attr_store+0x18/0x30
[<ffffffff811ad7e9>] sysfs_write_file+0x179/0x1c0
[<ffffffff81141432>] vfs_write+0xb2/0x1f0
[<ffffffff81141b27>] SyS_write+0x47/0xa0
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
other info that might help us debug this:
Chain exists of:
fcoe_config_mutex --> param_lock --> ft_mutex
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(ft_mutex);
lock(param_lock);
lock(ft_mutex);
lock(fcoe_config_mutex);
*** DEADLOCK ***
5 locks held by bash/16148:
#0: (sb_writers#4){.+.+.+}, at: [<ffffffff811414f3>] vfs_write+0x173/0x1f0
#1: (&of->mutex){+.+.+.}, at: [<ffffffff811ad762>] sysfs_write_file+0xf2/0x1c0
#2: (s_active){++++.+}, at: [<ffffffff811ad76a>] sysfs_write_file+0xfa/0x1c0
#3: (param_lock){+.+.+.}, at: [<ffffffff81065baf>] param_attr_store+0x2f/0x70
#4: (ft_mutex){+.+.+.}, at: [<ffffffffa025a690>]
fcoe_transport_create+0x20/0x210 [libfcoe]
stack backtrace:
CPU: 1 PID: 16148 Comm: bash Not tainted 3.13.0-rc1-debug+ #1
Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
ffffffff8219df80 ffff88000e19fc20 ffffffff814e10ed ffffffff8218c880
ffff88000e19fc60 ffffffff814de549 ffff88000e19fcb0 ffff880009fe2ff8
ffff880009fe28a0 0000000000000005 0000000000000005 ffff880009fe2ff8
Call Trace:
[<ffffffff814e10ed>] dump_stack+0x4e/0x7a
[<ffffffff814de549>] print_circular_bug+0x200/0x20f
[<ffffffff8108a41b>] __lock_acquire+0x17ab/0x1b10
[<ffffffff810781f8>] ? sched_clock_cpu+0xa8/0x100
[<ffffffffa0229037>] ? _fcoe_create+0x27/0x780 [fcoe]
[<ffffffff8108af42>] lock_acquire+0x72/0xa0
[<ffffffffa0229037>] ? _fcoe_create+0x27/0x780 [fcoe]
[<ffffffff814e4f82>] mutex_lock_nested+0x72/0x430
[<ffffffffa0229037>] ? _fcoe_create+0x27/0x780 [fcoe]
[<ffffffffa0229037>] ? _fcoe_create+0x27/0x780 [fcoe]
[<ffffffff810886db>] ? mark_held_locks+0xbb/0x140
[<ffffffff814e7415>] ? __mutex_unlock_slowpath+0xe5/0x190
[<ffffffffa0229037>] _fcoe_create+0x27/0x780 [fcoe]
[<ffffffff8108892d>] ? trace_hardirqs_on+0xd/0x10
[<ffffffffa022979e>] fcoe_create+0xe/0x10 [fcoe]
[<ffffffffa025a7b2>] fcoe_transport_create+0x142/0x210 [libfcoe]
[<ffffffff81065bbd>] param_attr_store+0x3d/0x70
[<ffffffff81065b08>] module_attr_store+0x18/0x30
[<ffffffff811ad7e9>] sysfs_write_file+0x179/0x1c0
[<ffffffff81141432>] vfs_write+0xb2/0x1f0
[<ffffffff81160d77>] ? fget_light+0x387/0x4f0
[<ffffffff81141b27>] SyS_write+0x47/0xa0
[<ffffffff814ef8b9>] system_call_fastpath+0x16/0x1b
scsi2 : FCoE Driver
fcoe: No FDMI support.
host2: libfc: Link up on port (000000)
host2: Assigned Port ID 00d3c4
_______________________________________________
fcoe-devel mailing list
[email protected]
http://lists.open-fcoe.org/mailman/listinfo/fcoe-devel
