[Cbe-oss-dev] [ INFO: possible recursive locking detected ] and [ INFO: possible circular locking dependency detected ]

[Stefan Wald]

And two more...
Is this an app problem?  But what causes the problem?
Thanks,
Stefan,

The first:

=============================================
[ INFO: possible recursive locking detected ]
2.6.29-rc6-01506-g624b1ab #4
---------------------------------------------
powerpc-linux-c/1993 is trying to acquire lock:
 (&ctx->state_mutex){--..}, at: [<c00000000003d844>] .destroy_spu_context
+0x80/0x13c

but task is already holding lock:
 (&ctx->state_mutex){--..}, at: [<c00000000003d6c8>] .spu_forget+0x38/0x90

other info that might help us debug this:
1 lock held by powerpc-linux-c/1993:
 #0:  (&ctx->state_mutex){--..}, at: [<c00000000003d6c8>] .spu_forget
+0x38/0x90

stack backtrace:
Call Trace:
[c00000000678a980] [c0000000000104d0] .show_stack+0xb4/0x1e8 (unreliable)
[c00000000678aa50] [c0000000003e0a08] .dump_stack+0x28/0x3c
[c00000000678aad0] [c000000000096f58] .__lock_acquire+0xf9c/0x1844
[c00000000678abd0] [c0000000000978ac] .lock_acquire+0xac/0xf8
[c00000000678aca0] [c0000000003db870] .mutex_lock_nested+0x184/0x430
[c00000000678ada0] [c00000000003d844] .destroy_spu_context+0x80/0x13c
[c00000000678ae30] [c000000000247824] .kref_put+0x80/0xa0
[c00000000678aeb0] [c00000000003d758] .put_spu_context+0x38/0x58
[c00000000678af40] [c000000000035894] .spufs_delete_inode+0x3c/0x7c
[c00000000678afd0] [c000000000126974] .generic_delete_inode+0x140/0x20c
[c00000000678b060] [c0000000001253bc] .iput+0xa8/0xc8
[c00000000678b0f0] [c0000000001212c4] .dentry_iput+0xfc/0x13c
[c00000000678b190] [c00000000012148c] .d_kill+0x54/0x9c
[c00000000678b220] [c000000000123fc0] .dput+0x1d4/0x204
[c00000000678b2c0] [c00000000010c650] .__fput+0x234/0x284
[c00000000678b370] [c00000000010c6f0] .fput+0x50/0x68
[c00000000678b400] [c0000000000e895c] .remove_vma+0x8c/0xf4
[c00000000678b490] [c0000000000e8b64] .exit_mmap+0x1a0/0x1dc
[c00000000678b540] [c00000000005f174] .mmput+0x80/0x134
[c00000000678b5e0] [c00000000003d6f4] .spu_forget+0x64/0x90
[c00000000678b670] [c000000000035fd0] .spufs_dir_close+0x9c/0xdc
[c00000000678b720] [c00000000010c588] .__fput+0x16c/0x284
[c00000000678b7d0] [c00000000010c6f0] .fput+0x50/0x68
[c00000000678b860] [c000000000107fa0] .filp_close+0xcc/0xf8
[c00000000678b900] [c000000000064540] .put_files_struct+0xd0/0x168
[c00000000678b9c0] [c00000000006463c] .exit_files+0x64/0x84
[c00000000678ba60] [c000000000066620] .do_exit+0x224/0x7a8
[c00000000678bb60] [c000000000066c48] .do_group_exit+0xa4/0xd8
[c00000000678bc00] [c000000000076484] .get_signal_to_deliver+0x3a0/0x400)
[c00000000678bce0] [c000000000012e7c] .do_signal+0x7c/0x30c
[c00000000678be30] [c000000000007b08] do_work+0x24/0x28

The second:

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.29-rc6-01506-g624b1ab #4
-------------------------------------------------------
powerpc-linux-2/2114 is trying to acquire lock:
 (&mm->mmap_sem){----}, at: [<c00000000003a6f0>] .spufs_ps_fault
+0x230/0x328

but task is already holding lock:
 (&ctx->state_mutex){--..}, at: [<c00000000003a684>] .spufs_ps_fault
+0x1c4/0x328

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&ctx->state_mutex){--..}:
       [<c0000000000973e4>] .__lock_acquire+0x1428/0x1844
       [<c0000000000978ac>] .lock_acquire+0xac/0xf8
       [<c0000000003db38c>] .mutex_lock_interruptible_nested+0x190/0x4f0
       [<c00000000003cf88>] .spufs_mem_mmap_fault+0xc0/0x174
       [<c0000000000e1218>] .__do_fault+0xbc/0x5e0
       [<c0000000000e3ddc>] .handle_mm_fault+0x418/0x860
       [<c00000000002ab34>] .do_page_fault+0x3b8/0x5a0
       [<c000000000005230>] handle_page_fault+0x20/0x5c

-> #0 (&mm->mmap_sem){----}:
       [<c000000000097094>] .__lock_acquire+0x10d8/0x1844
       [<c0000000000978ac>] .lock_acquire+0xac/0xf8
       [<c0000000003dc48c>] .down_read+0x64/0xb4
       [<c00000000003a6f0>] .spufs_ps_fault+0x230/0x328
       [<c00000000003a828>] .spufs_psmap_mmap_fault+0x40/0x5c
       [<c0000000000e1218>] .__do_fault+0xbc/0x5e0
       [<c0000000000e3ddc>] .handle_mm_fault+0x418/0x860
       [<c00000000002ab34>] .do_page_fault+0x3b8/0x5a0
       [<c000000000005230>] handle_page_fault+0x20/0x5c

other info that might help us debug this:

1 lock held by powerpc-linux-2/2114:
 #0:  (&ctx->state_mutex){--..}, at: [<c00000000003a684>] .spufs_ps_fault
+0x1c4/0x328

stack backtrace:
Call Trace:
[c000000006c834e0] [c0000000000104d0] .show_stack+0xb4/0x1e8 (unreliable)
[c000000006c835b0] [c0000000003e0a08] .dump_stack+0x28/0x3c
[c000000006c83630] [c000000000095b68] .print_circular_bug_tail+0xe0/0x108
[c000000006c83700] [c000000000097094] .__lock_acquire+0x10d8/0x1844
[c000000006c83800] [c0000000000978ac] .lock_acquire+0xac/0xf8
[c000000006c838d0] [c0000000003dc48c] .down_read+0x64/0xb4
[c000000006c83960] [c00000000003a6f0] .spufs_ps_fault+0x230/0x328
[c000000006c83a50] [c00000000003a828] .spufs_psmap_mmap_fault+0x40/0x5c
[c000000006c83ae0] [c0000000000e1218] .__do_fault+0xbc/0x5e0
[c000000006c83c00] [c0000000000e3ddc] .handle_mm_fault+0x418/0x860
[c000000006c83cf0] [c00000000002ab34] .do_page_fault+0x3b8/0x5a0
[c000000006c83e30] [c000000000005230] handle_page_fault+0x20/0x5c