The following BUG was found when running on a kernel with CONFIG_DEBUG_MUTEXES=y set: [ 502.794510] DEBUG_LOCKS_WARN_ON(lock->magic != lock) [ 502.794519] WARNING: CPU: 16 PID: 2571 at kernel/locking/mutex.c:1085 mutex_trylock+0x10d/0x120 [ 502.794537] Modules linked in: pds_vfio_pci(OE) pds_core xt_CHECKSUM xt_MASQUERADE xt_conntrack ipt_REJECT nft_compat nf_nat_tftp nf_conntrack_tftp bridge stp llc nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nft_chain_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_set nf_tables nfnetlink intel_rapl_msr intel_rapl_common isst_if_common nfit libnvdimm x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel rapl mei_me ses intel_cstate ipmi_ssif sunrpc enclosure mei nvme intel_uncore nvme_core hpwdt hpilo ioatdma acpi_tad acpi_ipmi pcspkr lpc_ich intel_pch_thermal acpi_power_meter ipmi_si xfs libcrc32c sd_mod t10_pi crc64_rocksoft crc64 sg mgag200 drm_kms_helper drm_shmem_helper crc32c_intel serio_raw igb smartpqi dca drm scsi_transport_sas i2c_algo_bit ionic wmi dm_mirror dm_region_hash dm_log dm_mod ipmi_devintf ipmi_msghandler fuse [last unloaded: pds_core] [ 502.794673] CPU: 16 PID: 2571 Comm: bash Tainted: G S OE 6.6.0-rc1-next-20230911 #1 [ 502.794676] Hardware name: HPE ProLiant DL360 Gen10/ProLiant DL360 Gen10, BIOS U32 01/23/2021 [ 502.794678] RIP: 0010:mutex_trylock+0x10d/0x120 [ 502.794681] Code: ff 85 c0 0f 84 40 ff ff ff 8b 3d f2 35 eb 00 85 ff 0f 85 32 ff ff ff 48 c7 c6 47 ff d3 bc 48 c7 c7 80 69 d3 bc e8 33 c7 4a ff <0f> 0b e9 18 ff ff ff b8 01 00 00 00 e9 5c ff ff ff 66 90 90 90 90 [ 502.794684] RSP: 0018:ffffabadc750fda8 EFLAGS: 00010282 [ 502.794686] RAX: 0000000000000000 RBX: ffff930000451000 RCX: 0000000000000027 [ 502.794688] RDX: 0000000000000027 RSI: ffffabadc750fca0 RDI: ffff930f7f920948 [ 502.794690] RBP: ffff930000451708 R08: 0000000000000000 R09: c0000000ffff7fff [ 502.794692] R10: 0000000000000001 R11: ffffabadc750fc40 R12: 0000000000000000 [ 502.794694] R13: fffffffffffffff2 R14: ffffabadc750fea0 R15: ffff92ffc85dc520 [ 502.794695] FS: 00007f3d8e322740(0000) GS:ffff930f7f900000(0000) knlGS:0000000000000000 [ 502.794697] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 502.794698] CR2: 000055cc4f6d9310 CR3: 0000000109894005 CR4: 00000000007706e0 [ 502.794700] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 502.794701] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 502.794703] PKRU: 55555554 [ 502.794704] Call Trace: [ 502.794707] <TASK> [ 502.794710] ? __warn+0x85/0x140 [ 502.794718] ? mutex_trylock+0x10d/0x120 [ 502.794720] ? report_bug+0xfc/0x1e0 [ 502.794727] ? handle_bug+0x3f/0x70 [ 502.794732] ? exc_invalid_op+0x17/0x70 [ 502.794735] ? asm_exc_invalid_op+0x1a/0x20 [ 502.794742] ? mutex_trylock+0x10d/0x120 [ 502.794744] ? mutex_trylock+0x10d/0x120 [ 502.794748] pds_vfio_reset+0x3a/0x60 [pds_vfio_pci] [ 502.794756] pci_reset_function+0x4b/0x70 [ 502.794763] reset_store+0x5b/0xa0 [ 502.794770] kernfs_fop_write_iter+0x137/0x1d0 [ 502.794776] vfs_write+0x2de/0x410 [ 502.794784] ksys_write+0x5d/0xd0 [ 502.794787] do_syscall_64+0x3b/0x90 [ 502.794790] entry_SYSCALL_64_after_hwframe+0x6e/0xd8 [ 502.794794] RIP: 0033:0x7f3d8d51fa28 [ 502.794796] Code: 89 02 48 c7 c0 ff ff ff ff eb b3 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 15 4d 2a 00 8b 00 85 c0 75 17 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 41 54 49 89 d4 55 [ 502.794799] RSP: 002b:00007ffc7b6c76f8 EFLAGS: 00000246 ORIG_RAX: 0000000000000001 [ 502.794801] RAX: ffffffffffffffda RBX: 0000000000000002 RCX: 00007f3d8d51fa28 [ 502.794803] RDX: 0000000000000002 RSI: 000055cc4f742aa0 RDI: 0000000000000001 [ 502.794804] RBP: 000055cc4f742aa0 R08: 000000000000000a R09: 00007f3d8d57fae0 [ 502.794805] R10: 000000000000000a R11: 0000000000000246 R12: 00007f3d8d7c06e0 [ 502.794807] R13: 0000000000000002 R14: 00007f3d8d7bb860 R15: 0000000000000002 [ 502.794809] </TASK> As shown, lock->magic != lock. This is because mutex_init(&pds_vfio->state_mutex) is called in the VFIO open path. So, if a reset is initiated before the VFIO device is opened the mutex will have never been initialized. Fix this by calling mutex_init(&pds_vfio->state_mutex) in the VFIO init path. Signed-off-by: Brett Creeley <brett.creeley@xxxxxxx> Reviewed-by: Shannon Nelson <shannon.nelson@xxxxxxx> --- drivers/vfio/pci/pds/vfio_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/vfio/pci/pds/vfio_dev.c b/drivers/vfio/pci/pds/vfio_dev.c index 147a543a7c39..9db5f2c8f1ea 100644 --- a/drivers/vfio/pci/pds/vfio_dev.c +++ b/drivers/vfio/pci/pds/vfio_dev.c @@ -155,6 +155,7 @@ static int pds_vfio_init_device(struct vfio_device *vdev) pds_vfio->vf_id = vf_id; + mutex_init(&pds_vfio->state_mutex); spin_lock_init(&pds_vfio->reset_lock); vdev->migration_flags = VFIO_MIGRATION_STOP_COPY | VFIO_MIGRATION_P2P; @@ -181,7 +182,6 @@ static int pds_vfio_open_device(struct vfio_device *vdev) if (err) return err; - mutex_init(&pds_vfio->state_mutex); pds_vfio->state = VFIO_DEVICE_STATE_RUNNING; pds_vfio->deferred_reset_state = VFIO_DEVICE_STATE_RUNNING; -- 2.17.1
