Inconsistent lock state
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
linaro-landing-team-ti |
Incomplete
|
Undecided
|
Unassigned |
Bug Description
I'm seeing this when trying to download a tarball of a rootfs and write it to sd card in several lava tests. This is on the master image (the one that we hope is stable) not the test image, so it never makes it through deployment. It *seems* that maybe it happens more frequently with larger images such as ubuntu-desktop.
root@master:~# [rc=0]: wget -qO- http://
wget -qO- http://
tgz |tar --numeric-owner -C /mnt/root -xzf -
[ 9251.570739]
[ 9251.570770] =======
[ 9251.584320] [ INFO: inconsistent lock state ]
[ 9251.592620] 3.0.0-1005-
[ 9251.601745] -------
[ 9251.609985] inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage.
[ 9251.620513] kswapd0/428 [HC0[0]
[ 9251.629516] (&sb->s_
[ 9251.645996] {RECLAIM_FS-ON-W} state was registered at:
[ 9251.655059] [<8005c239>] mark_held_
[ 9251.664001] [<8005c693>] lockdep_
[ 9251.673248] [<800adaaf>] kmem_cache_
[ 9251.682128] [<800bd8c5>] d_alloc+0x1d/0x13c
[ 9251.690185] [<800b67b7>] d_alloc_
[ 9251.699096] [<800b7b73>] do_lookup+
[ 9251.707336] [<800b829f>] path_lookupat+
[ 9251.715759] [<800b8625>] do_path_
[ 9251.724121] [<800b88e3>] kern_path+0x1f/0x32
[ 9251.731994] [<800c3b1f>] do_mount+0x55/0x176
[ 9251.739807] [<800c3e73>] sys_mount+0x57/0x80
[ 9251.747619] [<8000c2c1>] ret_fast_
[ 9251.756011] irq event stamp: 291334
[ 9251.762725] hardirqs last enabled at (291334): [<80073651>] __call_
[ 9251.776763] hardirqs last disabled at (291333): [<800735a5>] __call_
[ 9251.790832] softirqs last enabled at (288839): [<8003b2e7>] irq_exit+0x43/0x84
[ 9251.805114] softirqs last disabled at (288830): [<8003b2e7>] irq_exit+0x43/0x84
[ 9251.819641]
[ 9251.819671] other info that might help us debug this:
[ 9251.833282] Possible unsafe locking scenario:
[ 9251.833282]
[ 9251.846191] CPU0
[ 9251.851959] ----
[ 9251.857574] lock(&sb-
[ 9251.865203] <Interrupt>
[ 9251.870880] lock(&sb-
[ 9251.878631]
[ 9251.878631] *** DEADLOCK ***
[ 9251.878631]
[ 9251.893402] 2 locks held by kswapd0/428:
[ 9251.900360] #0: (shrinker_
[ 9251.914398] #1: (iprune_
[ 9251.928527]
[ 9251.928558] stack backtrace:
[ 9251.939361] [<800110ad>] (unwind_
[ 9251.955535] [<8005adc7>] (print_
[ 9251.971710] [<8005ae9d>] (mark_lock_
[ 9251.987335] [<8005b195>] (mark_lock+
[ 9252.002868] [<8005b345>] (mark_irqflags+
[ 9252.018890] [<8005b78b>] (__lock_
[ 9252.035186] [<8005bcf5>] (lock_acquire+
[ 9252.051971] [<803e15bf>] (mutex_
[ 9252.069763] [<80104f39>] (ext4_evict_
[ 9252.086761] [<800bf7a1>] (evict+0x65/0xec) from [<800bf851>] (dispose_
[ 9252.103576] [<800bf851>] (dispose_
[ 9252.121765] [<800bfdf9>] (prune_
[ 9252.141052] [<800bfe39>] (shrink_
[ 9252.160156] [<80095ac7>] (shrink_
[ 9252.178955] [<800970b9>] (balance_
[ 9252.197753] [<8009740f>] (kswapd+
[ 9252.215850] [<8004b823>] (kthread+0x57/0x68) from [<8000ccc5>] (kernel_
If you want to see the full log from one of these, check here: http://
#######
Behaviour described in comment #2 can be reproduced in following images:
http://
http://
http://
http://
Changed in linux-linaro: | |
status: | New → Incomplete |
affects: | linux-linaro → linaro-landing-team-ti |
description: | updated |
description: | updated |
Hm, I wouldn't recommend using the linux-linaro kernel for Panda, as the TILT generally works a lot better.