[ 1628.959890]
[ 1628.959916] ======================================================
[ 1628.961900] WARNING: possible circular locking dependency detected
[ 1628.968026] 5.9.1+ #6 Not tainted
[ 1628.971301] ------------------------------------------------------
[ 1628.977425] kswapd0/874 is trying to acquire lock:
[ 1628.982169] ffff0000489d1c08 (&obj->mappings.lock){+.+.}-{3:3}, at: panfrost_gem_teardown_mappings+0x2c/0x68 [panfrost]
[ 1628.992864]
[ 1628.992864] but task is already holding lock:
[ 1628.998643] ffff0000489d1a90 (&shmem->pages_lock){+.+.}-{3:3}, at: panfrost_gem_shrinker_scan+0xc8/0x188 [panfrost]
[ 1629.008996]
[ 1629.008996] which lock already depends on the new lock.
[ 1629.008996]
[ 1629.017098]
[ 1629.017098] the existing dependency chain (in reverse order) is:
[ 1629.024517]
[ 1629.024517] -> #1 (&shmem->pages_lock){+.+.}-{3:3}:
[ 1629.030820]        __mutex_lock+0x88/0x908
[ 1629.034868]        mutex_lock_nested+0x38/0x68
[ 1629.039269]        drm_gem_shmem_madvise+0x2c/0x60
[ 1629.044022]        panfrost_ioctl_madvise+0x100/0x1e8 [panfrost]
[ 1629.049965]        drm_ioctl_kernel+0xc4/0x118
[ 1629.054361]        drm_ioctl+0x23c/0x4a0
[ 1629.058244]        __arm64_sys_ioctl+0xa8/0xf0
[ 1629.062644]        el0_svc_common.constprop.0+0x74/0x190
[ 1629.067903]        do_el0_svc+0x24/0x90
[ 1629.071697]        el0_sync_handler+0x9c/0x1b8
[ 1629.076095]        el0_sync+0x158/0x180
[ 1629.079889]
[ 1629.079889] -> #0 (&obj->mappings.lock){+.+.}-{3:3}:
[ 1629.086276]        __lock_acquire+0x13d4/0x21c8
[ 1629.090759]        lock_acquire+0xec/0x450
[ 1629.094813]        __mutex_lock+0x88/0x908
[ 1629.098866]        mutex_lock_nested+0x38/0x68
[ 1629.103279]        panfrost_gem_teardown_mappings+0x2c/0x68 [panfrost]
[ 1629.109739]        panfrost_gem_shrinker_scan+0xd4/0x188 [panfrost]
[ 1629.115946]        do_shrink_slab+0x1a4/0x570
[ 1629.120256]        shrink_slab+0xb4/0x2b0
[ 1629.124224]        shrink_node+0x3b8/0x6b8
[ 1629.128276]        balance_pgdat+0x310/0x5d0
[ 1629.132502]        kswapd+0x228/0x6c0
[ 1629.136127]        kthread+0x148/0x168
[ 1629.139843]        ret_from_fork+0x10/0x34
[ 1629.143890]
[ 1629.143890] other info that might help us debug this:
[ 1629.143890]
[ 1629.151821]  Possible unsafe locking scenario:
[ 1629.151821]
[ 1629.157686]        CPU0                    CPU1
[ 1629.162171]        ----                    ----
[ 1629.166655]   lock(&shmem->pages_lock);
[ 1629.170451]                                lock(&obj->mappings.lock);
[ 1629.176833]                                lock(&shmem->pages_lock);
[ 1629.183129]   lock(&obj->mappings.lock);
[ 1629.187011]
[ 1629.187011]  *** DEADLOCK ***
[ 1629.187011]
[ 1629.192877] 4 locks held by kswapd0/874:
[ 1629.196757]  #0: ffff800011da2e10 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x0/0x48
[ 1629.205210]  #1: ffff800011d96aa8 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab+0x64/0x2b0
[ 1629.213404]  #2: ffff0000af210408 (&pfdev->shrinker_lock){+.+.}-{3:3}, at: panfrost_gem_shrinker_scan+0x38/0x188 [panfrost]
[ 1629.224444]  #3: ffff0000489d1a90 (&shmem->pages_lock){+.+.}-{3:3}, at: panfrost_gem_shrinker_scan+0xc8/0x188 [panfrost]
[ 1629.235225]
[ 1629.235225] stack backtrace:
[ 1629.239541] CPU: 0 PID: 874 Comm: kswapd0 Not tainted 5.9.1+ #6
[ 1629.245403] Hardware name: Hardkernel ODROID-C4 (DT)
[ 1629.250320] Call trace:
[ 1629.252738]  dump_backtrace+0x0/0x1d8
[ 1629.256357]  show_stack+0x18/0x28
[ 1629.259637]  dump_stack+0xe8/0x15c
[ 1629.263002]  print_circular_bug.isra.0+0x23c/0x280
[ 1629.267743]  check_noncircular+0x164/0x1e0
[ 1629.271797]  __lock_acquire+0x13d4/0x21c8
[ 1629.275765]  lock_acquire+0xec/0x450
[ 1629.279302]  __mutex_lock+0x88/0x908
[ 1629.282837]  mutex_lock_nested+0x38/0x68
[ 1629.286727]  panfrost_gem_teardown_mappings+0x2c/0x68 [panfrost]
[ 1629.292675]  panfrost_gem_shrinker_scan+0xd4/0x188 [panfrost]
[ 1629.298364]  do_shrink_slab+0x1a4/0x570
[ 1629.302157]  shrink_slab+0xb4/0x2b0
[ 1629.305607]  shrink_node+0x3b8/0x6b8
[ 1629.309144]  balance_pgdat+0x310/0x5d0
[ 1629.312851]  kswapd+0x228/0x6c0
[ 1629.315959]  kthread+0x148/0x168
[ 1629.319149]  ret_from_fork+0x10/0x34