[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20081130130726.GA4365@x200.localdomain>
Date: Sun, 30 Nov 2008 16:07:26 +0300
From: Alexey Dobriyan <adobriyan@...il.com>
To: netdev@...r.kernel.org
Subject: net-next: lockdep complains re percpu counters
This is net-next as of c5419e6f054c877339f754e02c3b1dafd88cd96c
aka "cxgb3: Fix sparse warning and micro-optimize is_pure_response()".
=================================
[ INFO: inconsistent lock state ]
2.6.28-rc6-netns #2
---------------------------------
inconsistent {softirq-on-W} -> {in-softirq-W} usage.
recvfrom01/7632 [HC0[0]:SC1[1]:HE1:SE0] takes:
(&fbc->lock){-+..}, at: [<ffffffff803292d3>] __percpu_counter_add+0x63/0xc0
{softirq-on-W} state was registered at:
[<ffffffff80257709>] __lock_acquire+0x4b9/0x9c0
[<ffffffff802586b6>] lock_acquire+0x56/0x80
[<ffffffff804eb256>] _spin_lock+0x36/0x50
[<ffffffff803292d3>] __percpu_counter_add+0x63/0xc0
[<ffffffff802907b9>] get_empty_filp+0x59/0x110
[<ffffffff80298fa0>] path_lookup_open+0x30/0xc0
[<ffffffff80299a5e>] do_filp_open+0xae/0x8a0
[<ffffffff8028f036>] do_sys_open+0x76/0xd0
[<ffffffff8028f0ab>] sys_open+0x1b/0x20
[<ffffffff8020b6db>] system_call_fastpath+0x16/0x1b
[<ffffffffffffffff>] 0xffffffffffffffff
irq event stamp: 5698
hardirqs last enabled at (5698): [<ffffffff80258de1>] debug_check_no_locks_freed+0xa1/0x150
hardirqs last disabled at (5697): [<ffffffff80258d80>] debug_check_no_locks_freed+0x40/0x150
softirqs last enabled at (5656): [<ffffffff803d8031>] release_sock+0xc1/0xf0
softirqs last disabled at (5657): [<ffffffff8020cb8c>] call_softirq+0x1c/0x30
other info that might help us debug this:
2 locks held by recvfrom01/7632:
#0: (slock-AF_INET/1){-+..}, at: [<ffffffff8044106b>] tcp_v4_rcv+0x50b/0x840
#1: (slock-AF_INET){-+..}, at: [<ffffffff803d75f1>] sk_clone+0xe1/0x340
stack backtrace:
Pid: 7632, comm: recvfrom01 Not tainted 2.6.28-rc6-netns #2
Call Trace:
<IRQ> [<ffffffff80255e3d>] print_usage_bug+0x18d/0x1e0
[<ffffffff80256a67>] mark_lock+0x887/0xb70
[<ffffffff80257692>] __lock_acquire+0x442/0x9c0
[<ffffffff802586b6>] lock_acquire+0x56/0x80
[<ffffffff803292d3>] ? __percpu_counter_add+0x63/0xc0
[<ffffffff804eb256>] _spin_lock+0x36/0x50
[<ffffffff803292d3>] ? __percpu_counter_add+0x63/0xc0
[<ffffffff803292d3>] __percpu_counter_add+0x63/0xc0
[<ffffffff803d7817>] sk_clone+0x307/0x340
[<ffffffff8042ddb1>] inet_csk_clone+0x11/0xa0
[<ffffffff80442c34>] tcp_create_openreq_child+0x24/0x470
[<ffffffff80440773>] tcp_v4_syn_recv_sock+0x53/0x210
[<ffffffff804431a5>] tcp_check_req+0x125/0x450
[<ffffffff802377b4>] ? local_bh_enable+0xa4/0x110
[<ffffffff80440a6d>] tcp_v4_do_rcv+0x13d/0x230
[<ffffffff804412a5>] tcp_v4_rcv+0x745/0x840
[<ffffffff80423304>] ip_local_deliver_finish+0x124/0x2b0
[<ffffffff80423530>] ip_local_deliver+0xa0/0xb0
[<ffffffff8042364c>] ip_rcv_finish+0x10c/0x3c0
[<ffffffff80423ac0>] ip_rcv+0x1c0/0x300
[<ffffffff803e3f79>] netif_receive_skb+0x379/0x400
[<ffffffff803e411b>] ? process_backlog+0x8b/0x100
[<ffffffff803e4127>] process_backlog+0x97/0x100
[<ffffffff803e42da>] net_rx_action+0x14a/0x210
[<ffffffff802379b8>] __do_softirq+0x88/0x150
[<ffffffff803d8031>] ? release_sock+0xc1/0xf0
[<ffffffff8020cb8c>] call_softirq+0x1c/0x30
<EOI> [<ffffffff8020e5b7>] do_softirq+0x77/0xd0
[<ffffffff803d8031>] ? release_sock+0xc1/0xf0
[<ffffffff8023790b>] local_bh_enable_ip+0xeb/0x110
[<ffffffff804eb929>] _spin_unlock_bh+0x39/0x40
[<ffffffff803d7ffe>] ? release_sock+0x8e/0xf0
[<ffffffff803d8031>] release_sock+0xc1/0xf0
[<ffffffff8044f178>] inet_stream_connect+0x198/0x2e0
[<ffffffff80247e70>] ? autoremove_wake_function+0x0/0x40
[<ffffffff80247e70>] ? autoremove_wake_function+0x0/0x40
[<ffffffff803d557d>] sys_connect+0x7d/0xc0
[<ffffffff8020b70c>] ? sysret_check+0x27/0x62
[<ffffffff80256eca>] ? trace_hardirqs_on_caller+0xba/0x130
[<ffffffff804eaf07>] ? trace_hardirqs_on_thunk+0x3a/0x3f
[<ffffffff8020b6db>] system_call_fastpath+0x16/0x1b
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists