igt@gem_exec_parallel@* - dmesg-warn - WARNING: possible circular locking dependency detected, acquire lock at: i915_request_create, ut task is already holding lock at: gen6_ppgtt_pin

https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7492/shard-hsw1/igt@gem_exec_parallel@rcs0-fds.html

<4> [1285.329963] ======================================================
<4> [1285.329965] WARNING: possible circular locking dependency detected
<4> [1285.329967] 5.4.0-rc8-CI-CI_DRM_7492+ #1 Tainted: G     U           
<4> [1285.329969] ------------------------------------------------------
<4> [1285.329971] gem_exec_parall/5950 is trying to acquire lock:
<4> [1285.329973] ffff8884080f6270 (&kernel#2){+.+.}, at: i915_request_create+0x16/0x1c0 [i915]
<4> [1285.330040] 
but task is already holding lock:
<4> [1285.330042] ffff88838884d550 (&ppgtt->pin_mutex){+.+.}, at: gen6_ppgtt_pin+0x4d/0x110 [i915]
<4> [1285.330078] 
which lock already depends on the new lock.

<4> [1285.330082] 
the existing dependency chain (in reverse order) is:
<4> [1285.330086] 
-> #1 (&ppgtt->pin_mutex){+.+.}:
<4> [1285.330093]        __mutex_lock+0x9a/0x9d0
<4> [1285.330128]        gen6_ppgtt_pin+0x4d/0x110 [i915]
<4> [1285.330156]        emit_ppgtt_update+0x1dc/0x370 [i915]
<4> [1285.330196]        context_barrier_task+0x176/0x310 [i915]
<4> [1285.330228]        ctx_setparam+0x400/0xb10 [i915]
<4> [1285.330259]        i915_gem_context_setparam_ioctl+0xc8/0x160 [i915]
<4> [1285.330263]        drm_ioctl_kernel+0xa7/0xf0
<4> [1285.330265]        drm_ioctl+0x2e1/0x390
<4> [1285.330269]        do_vfs_ioctl+0xa0/0x6f0
<4> [1285.330271]        ksys_ioctl+0x35/0x60
<4> [1285.330274]        __x64_sys_ioctl+0x11/0x20
<4> [1285.330278]        do_syscall_64+0x4f/0x210
<4> [1285.330281]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4> [1285.330284] 
-> #0 (&kernel#2){+.+.}:
<4> [1285.330289]        __lock_acquire+0x1328/0x15d0
<4> [1285.330292]        lock_acquire+0xa7/0x1c0
<4> [1285.330295]        __mutex_lock+0x9a/0x9d0
<4> [1285.330338]        i915_request_create+0x16/0x1c0 [i915]
<4> [1285.330373]        intel_engine_flush_barriers+0x4c/0x100 [i915]
<4> [1285.330417]        i915_ggtt_pin+0x7d/0x130 [i915]
<4> [1285.330458]        gen6_ppgtt_pin+0x9c/0x110 [i915]
<4> [1285.330494]        ring_context_pin+0x2e/0xc0 [i915]
<4> [1285.330529]        __intel_context_do_pin+0x6b/0x190 [i915]
<4> [1285.330571]        i915_gem_do_execbuffer+0x180c/0x26b0 [i915]
<4> [1285.330612]        i915_gem_execbuffer2_ioctl+0x11b/0x460 [i915]
<4> [1285.330616]        drm_ioctl_kernel+0xa7/0xf0
<4> [1285.330619]        drm_ioctl+0x2e1/0x390
<4> [1285.330622]        do_vfs_ioctl+0xa0/0x6f0
<4> [1285.330625]        ksys_ioctl+0x35/0x60
<4> [1285.330628]        __x64_sys_ioctl+0x11/0x20
<4> [1285.330631]        do_syscall_64+0x4f/0x210
<4> [1285.330634]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4> [1285.330636] 
other info that might help us debug this:

<4> [1285.330640]  Possible unsafe locking scenario:

<4> [1285.330643]        CPU0                    CPU1
<4> [1285.330645]        ----                    ----
<4> [1285.330648]   lock(&ppgtt->pin_mutex);
<4> [1285.330650]                                lock(&kernel#2);
<4> [1285.330653]                                lock(&ppgtt->pin_mutex);
<4> [1285.330656]   lock(&kernel#2);
<4> [1285.330659] 
 *** DEADLOCK ***