Bug 109386

https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_196/fi-ivb-3770/igt@gem_userptr_blits@map-fixed-invalidate-busy.html

<6> [123.903334] Console: switching to colour dummy device 80x25
<6> [123.903381] [IGT] gem_userptr_blits: executing
<6> [123.958532] [IGT] gem_userptr_blits: starting subtest map-fixed-invalidate-busy
<6> [124.287429] userptr_mn_invalidate_range_start+0x0/0x270 [i915] callback failed with -4 in blockable context.
<4> [124.287710] 
<4> [124.287713] ======================================================
<4> [124.287715] WARNING: possible circular locking dependency detected
<4> [124.287718] 5.0.0-rc2-g19fd97ec0d64-drmtip_196+ #1 Tainted: G     U           
<4> [124.287721] ------------------------------------------------------
<4> [124.287723] gem_userptr_bli/2289 is trying to acquire lock:
<4> [124.287726] 000000001e4606f1 (fs_reclaim){+.+.}, at: fs_reclaim_acquire.part.25+0x0/0x30
<4> [124.287732] 
but task is already holding lock:
<4> [124.287734] 00000000092837a8 (&dev->struct_mutex/1){+.+.}, at: userptr_mn_invalidate_range_start+0x173/0x270 [i915]
<4> [124.287767] 
which lock already depends on the new lock.

<4> [124.287769] 
the existing dependency chain (in reverse order) is:
<4> [124.287772] 
-> #1 (&dev->struct_mutex/1){+.+.}:
<4> [124.287798]        i915_address_space_init+0x53/0x130 [i915]
<4> [124.287824]        i915_ggtt_init_hw+0x50/0x140 [i915]
<4> [124.287846]        i915_driver_load+0xe2e/0x1590 [i915]
<4> [124.287868]        i915_pci_probe+0x29/0xa0 [i915]
<4> [124.287873]        pci_device_probe+0xa1/0x130
<4> [124.287877]        really_probe+0xf3/0x3e0
<4> [124.287880]        driver_probe_device+0x10a/0x120
<4> [124.287883]        __driver_attach+0xdb/0x100
<4> [124.287885]        bus_for_each_dev+0x74/0xc0
<4> [124.287888]        bus_add_driver+0x15f/0x250
<4> [124.287891]        driver_register+0x56/0xe0
<4> [124.287894]        do_one_initcall+0x58/0x2e0
<4> [124.287897]        do_init_module+0x56/0x1ea
<4> [124.287900]        load_module+0x2718/0x29f0
<4> [124.287902]        __se_sys_finit_module+0xd3/0xf0
<4> [124.287905]        do_syscall_64+0x55/0x190
<4> [124.287908]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4> [124.287911] 
-> #0 (fs_reclaim){+.+.}:
<4> [124.287915]        fs_reclaim_acquire.part.25+0x24/0x30
<4> [124.287918]        kmem_cache_alloc+0x27/0x280
<4> [124.287922]        vm_area_alloc+0x15/0x50
<4> [124.287925]        mmap_region+0x306/0x640
<4> [124.287928]        do_mmap+0x3d1/0x570
<4> [124.287931]        vm_mmap_pgoff+0x98/0xf0
<4> [124.287934]        ksys_mmap_pgoff+0x191/0x1f0
<4> [124.287936]        do_syscall_64+0x55/0x190
<4> [124.287939]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4> [124.287941] 
other info that might help us debug this:

<4> [124.287944]  Possible unsafe locking scenario:

<4> [124.287947]        CPU0                    CPU1
<4> [124.287948]        ----                    ----
<4> [124.287950]   lock(&dev->struct_mutex/1);
<4> [124.287953]                                lock(fs_reclaim);
<4> [124.287955]                                lock(&dev->struct_mutex/1);
<4> [124.287958]   lock(fs_reclaim);
<4> [124.287960] 
 *** DEADLOCK ***

<4> [124.287963] 2 locks held by gem_userptr_bli/2289:
<4> [124.287965]  #0: 00000000182dc149 (&mm->mmap_sem){++++}, at: vm_mmap_pgoff+0x6a/0xf0
<4> [124.287969]  #1: 00000000092837a8 (&dev->struct_mutex/1){+.+.}, at: userptr_mn_invalidate_range_start+0x173/0x270 [i915]
<4> [124.287998] 
stack backtrace:
<4> [124.288001] CPU: 1 PID: 2289 Comm: gem_userptr_bli Tainted: G     U            5.0.0-rc2-g19fd97ec0d64-drmtip_196+ #1
<4> [124.288005] Hardware name: Hewlett-Packard HP Pro 3500 Series/2ABF, BIOS 8.11 10/24/2012
<4> [124.288008] Call Trace:
<4> [124.288012]  dump_stack+0x67/0x9b
<4> [124.288016]  print_circular_bug.isra.16+0x1c8/0x2b0
<4> [124.288020]  __lock_acquire+0x183a/0x1b00
<4> [124.288023]  ? mark_held_locks+0x50/0x80
<4> [124.288027]  ? _raw_spin_unlock_irqrestore+0x39/0x60
<4> [124.288030]  ? lock_acquire+0xa6/0x1c0
<4> [124.288033]  lock_acquire+0xa6/0x1c0
<4> [124.288035]  ? free_unref_page_commit.isra.20+0xd0/0xd0
<4> [124.288039]  ? vm_area_alloc+0x15/0x50
<4> [124.288041]  fs_reclaim_acquire.part.25+0x24/0x30
<4> [124.288044]  ? free_unref_page_commit.isra.20+0xd0/0xd0
<4> [124.288047]  kmem_cache_alloc+0x27/0x280
<4> [124.288050]  vm_area_alloc+0x15/0x50
<4> [124.288053]  mmap_region+0x306/0x640
<4> [124.288056]  do_mmap+0x3d1/0x570
<4> [124.288059]  vm_mmap_pgoff+0x98/0xf0
<4> [124.288062]  ksys_mmap_pgoff+0x191/0x1f0
<4> [124.288066]  ? lockdep_hardirqs_off+0x79/0xd0
<4> [124.288069]  do_syscall_64+0x55/0x190
<4> [124.288072]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4> [124.288074] RIP: 0033:0x7f808fdc1a13
<4> [124.288077] Code: 89 f5 41 54 41 89 d4 55 48 89 fd 53 4c 89 cb 74 56 49 89 d9 45 89 f8 45 89 f2 44 89 e2 4c 89 ee 48 89 ef b8 09 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 7d 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 2e 0f
<4> [124.288083] RSP: 002b:00007fff5203c808 EFLAGS: 00000206 ORIG_RAX: 0000000000000009
<4> [124.288086] RAX: ffffffffffffffda RBX: 0000000100000000 RCX: 00007f808fdc1a13
<4> [124.288089] RDX: 0000000000000003 RSI: 0000000000002000 RDI: 00007f8090f1b000
<4> [124.288092] RBP: 00007f8090f1b000 R08: 0000000000000005 R09: 0000000100000000
<4> [124.288095] R10: 0000000000000011 R11: 0000000000000206 R12: 0000000000000003
<4> [124.288097] R13: 0000000000002000 R14: 0000000000000011 R15: 0000000000000005

This might be related to https://bugs.freedesktop.org/show_bug.cgi?id=109385