Just mounted a JFFS2 partition (held in NOR flash on a custom AM335x
CPU board), and I always get the following:-
[ 3.864244]
[ 3.865851] ======================================================
[ 3.872359] [ INFO: possible circular locking dependency detected ]
[ 3.878968] 3.10.0-rc4-00172-gf31c62e-dirty #249 Not tainted
[ 3.884926] -------------------------------------------------------
[ 3.891526] rcS/507 is trying to acquire lock:
[ 3.896206] (&mm->mmap_sem){++++++}, at: [<c00b42c8>] might_fault+0x3c/0x94
[ 3.903684]
[ 3.903684] but task is already holding lock:
[ 3.909826] (&f->sem){+.+.+.}, at: [<c016f208>] jffs2_readdir+0x40/0x1b0
[ 3.917021]
[ 3.917021] which lock already depends on the new lock.
[ 3.917021]
[ 3.925637]
[ 3.925637] the existing dependency chain (in reverse order) is:
[ 3.933516]
-> #1 (&f->sem){+.+.+.}:
[ 3.937489] [<c0079ad0>] lock_acquire+0x68/0x7c
[ 3.942918] [<c041f290>] mutex_lock_nested+0x40/0x31c
[ 3.948893] [<c0170b00>] jffs2_readpage+0x20/0x4c
[ 3.954500] [<c009faf4>] __do_page_cache_readahead+0x28c/0x2b4
[ 3.961305] [<c00a0028>] ra_submit+0x28/0x30
[ 3.966450] [<c009686c>] filemap_fault+0x330/0x3e4
[ 3.972143] [<c00b0308>] __do_fault+0x68/0x4a4
[ 3.977481] [<c00b3214>] handle_pte_fault+0x70/0x6b4
[ 3.983361] [<c00b38f0>] handle_mm_fault+0x98/0xcc
[ 3.989055] [<c001d614>] do_page_fault+0x210/0x398
[ 3.994750] [<c00084a4>] do_DataAbort+0x38/0x98
[ 4.000170] [<c001315c>] __dabt_svc+0x3c/0x60
[ 4.005417] [<c01df448>] __clear_user_std+0x1c/0x64
[ 4.011209]
-> #0 (&mm->mmap_sem){++++++}:
[ 4.015729] [<c0078f40>] __lock_acquire+0x1990/0x1d7c
[ 4.021699] [<c0079ad0>] lock_acquire+0x68/0x7c
[ 4.027118] [<c00b42f0>] might_fault+0x64/0x94
[ 4.032444] [<c00d9bd0>] filldir+0x6c/0x174
[ 4.037507] [<c016f28c>] jffs2_readdir+0xc4/0x1b0
[ 4.043110] [<c00da020>] vfs_readdir+0x94/0xb8
[ 4.048438] [<c00da118>] SyS_getdents+0x64/0xd4
[ 4.053859] [<c00135c0>] ret_fast_syscall+0x0/0x48
[ 4.059556]
[ 4.059556] other info that might help us debug this:
[ 4.059556]
[ 4.067989] Possible unsafe locking scenario:
[ 4.067989]
[ 4.074222] CPU0 CPU1
[ 4.078989] ---- ----
[ 4.083756] lock(&f->sem);
[ 4.086799] lock(&mm->mmap_sem);
[ 4.093042] lock(&f->sem);
[ 4.098735] lock(&mm->mmap_sem);
[ 4.102326]
[ 4.102326] *** DEADLOCK ***
[ 4.102326]
[ 4.108567] 2 locks held by rcS/507:
[ 4.112328] #0: (&type->i_mutex_dir_key){+.+.+.}, at: [<c00d9fe0>]
vfs_readdir+0x54/0xb8
[ 4.121070] #1: (&f->sem){+.+.+.}, at: [<c016f208>]
jffs2_readdir+0x40/0x1b0
[ 4.128713]
[ 4.128713] stack backtrace:
[ 4.133312] CPU: 0 PID: 507 Comm: rcS Not tainted
3.10.0-rc4-00172-gf31c62e-dirty #249
[ 4.141676] [<c0018e70>] (unwind_backtrace+0x0/0xf4) from [<c0016d88>]
(show_stack+0x10/0x14)
[ 4.150669] [<c0016d88>] (show_stack+0x10/0x14) from [<c041bc88>]
(print_circular_bug+0x2e8/0x2f4)
[ 4.160118] [<c041bc88>] (print_circular_bug+0x2e8/0x2f4) from [<c0078f40>]
(__lock_acquire+0x1990/0x1d7c)
[ 4.170297] [<c0078f40>] (__lock_acquire+0x1990/0x1d7c) from [<c0079ad0>]
(lock_acquire+0x68/0x7c)
[ 4.179742] [<c0079ad0>] (lock_acquire+0x68/0x7c) from [<c00b42f0>]
(might_fault+0x64/0x94)
[ 4.188549] [<c00b42f0>] (might_fault+0x64/0x94) from [<c00d9bd0>]
(filldir+0x6c/0x174)
[ 4.196992] [<c00d9bd0>] (filldir+0x6c/0x174) from [<c016f28c>]
(jffs2_readdir+0xc4/0x1b0)
[ 4.205708] [<c016f28c>] (jffs2_readdir+0xc4/0x1b0) from [<c00da020>]
(vfs_readdir+0x94/0xb8)
[ 4.214698] [<c00da020>] (vfs_readdir+0x94/0xb8) from [<c00da118>]
(SyS_getdents+0x64/0xd4)
[ 4.223507] [<c00da118>] (SyS_getdents+0x64/0xd4) from [<c00135c0>]
(ret_fast_syscall+0x0/0x48)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
