Skip to content
Snippets Groups Projects
New look: Off

TGL ICL: igt@gem_exec_schedule@u-fairslice@.* - dmesg-warn - WARNING: possible circular locking dependency detected, is trying to acquire lock at: rcu_barrier but task is already holding lock at: unmap_vmas

    • Closed Issue created by Tejasree Illipilli @tejasree

    https://intel-gfx-ci.01.org/tree/drm-tip/IGT_5900/shard-tglb7/igt@gem_exec_schedule@u-fairslice@rcs0.html

    <4> [65.250878] ======================================================
<4> [65.250882] WARNING: possible circular locking dependency detected
<4> [65.250887] 5.10.0-CI-CI_DRM_9483+ #1 Tainted: G     U           
<4> [65.250892] ------------------------------------------------------
<4> [65.250896] gem_exec_schedu/1046 is trying to acquire lock:
<4> [65.250902] ffffffff82731ef0 (rcu_state.barrier_mutex){+.+.}-{3:3}, at: rcu_barrier+0x27/0x210
<4> [65.250923] 
but task is already holding lock:
<4> [65.250929] ffffffff827539c0 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}, at: unmap_vmas+0x0/0x130
<4> [65.250944] 
which lock already depends on the new lock.
<4> [65.250950] 
the existing dependency chain (in reverse order) is:
<4> [65.250955] 
-> #3 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}:
<4> [65.250985]        dma_resv_lockdep+0x1dc/0x28c
<4> [65.250991]        do_one_initcall+0x56/0x2e0
<4> [65.250999]        kernel_init_freeable+0x169/0x1b2
<4> [65.251007]        kernel_init+0x5/0xff
<4> [65.251011]        ret_from_fork+0x1f/0x30
<4> [65.251014] 
-> #2 (fs_reclaim){+.+.}-{0:0}:
<4> [65.251026]        fs_reclaim_acquire.part.129+0x24/0x30
<4> [65.251033]        kmem_cache_alloc_trace+0x2f/0x850
<4> [65.251039]        intel_cpuc_prepare+0x3b/0x1b0
<4> [65.251045]        cpuhp_invoke_callback+0x9e/0x890
<4> [65.251050]        _cpu_up+0xa4/0x130
<4> [65.251053]        cpu_up+0x82/0x90
<4> [65.251057]        bringup_nonboot_cpus+0x4a/0x60
<4> [65.251063]        smp_init+0x21/0x5c
<4> [65.251067]        kernel_init_freeable+0x85/0x1b2
<4> [65.251072]        kernel_init+0x5/0xff
<4> [65.251076]        ret_from_fork+0x1f/0x30
<4> [65.251079] 
-> #1 (cpu_hotplug_lock){++++}-{0:0}:
<4> [65.251087]        cpus_read_lock+0x34/0xc0
<4> [65.251091]        rcu_barrier+0xac/0x210
<4> [65.251095]        kernel_init+0x22/0xff
<4> [65.251098]        ret_from_fork+0x1f/0x30
<4> [65.251101] 
-> #0 (rcu_state.barrier_mutex){+.+.}-{3:3}:
<4> [65.251112]        __lock_acquire+0x1523/0x25a0
<4> [65.251117]        lock_acquire+0xdc/0x3c0
<4> [65.251122]        __mutex_lock+0xa6/0x980
<4> [65.251125]        rcu_barrier+0x27/0x210
<4> [65.251212]        i915_gem_object_unbind+0x2f1/0x3e0 [i915]
<4> [65.251294]        userptr_mn_invalidate_range_start+0x82/0x180 [i915]
<4> [65.251300]        __mmu_notifier_invalidate_range_start+0xbd/0x1f0
<4> [65.251305]        unmap_vmas+0x95/0x130
<4> [65.251310]        unmap_region+0xa3/0x100
<4> [65.251314]        __do_munmap+0x29d/0x4d0
<4> [65.251318]        __vm_munmap+0x66/0xc0
<4> [65.251322]        __x64_sys_munmap+0x12/0x20
<4> [65.251328]        do_syscall_64+0x33/0x80
<4> [65.251333]        entry_SYSCALL_64_after_hwframe+0x44/0xa9
<4> [65.251338] 
other info that might help us debug this:
<4> [65.251343] Chain exists of:
  rcu_state.barrier_mutex --> fs_reclaim --> mmu_notifier_invalidate_range_start
<4> [65.251355]  Possible unsafe locking scenario:
<4> [65.251359]        CPU0                    CPU1
<4> [65.251363]        ----                    ----
<4> [65.251367]   lock(mmu_notifier_invalidate_range_start);
<4> [65.251372]                                lock(fs_reclaim);
<4> [65.251377]                                lock(mmu_notifier_invalidate_range_start);
<4> [65.251384]   lock(rcu_state.barrier_mutex);
<4> [65.251388] 
 *** DEADLOCK ***
    To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information

    Child items ...

    • Activity

    Please register or sign in to reply