====================================================== [ INFO: possible circular locking dependency detected ] 4.4.174+ #4 Not tainted ------------------------------------------------------- syz-executor.0/3286 is trying to acquire lock: (sel_mutex){+.+.+.}, at: [] sel_read_policy+0x59/0x110 security/selinux/selinuxfs.c:441 but task is already holding lock: (sb_writers#4){.+.+.+}, at: [] file_start_write include/linux/fs.h:2543 [inline] (sb_writers#4){.+.+.+}, at: [] do_sendfile+0x8a6/0xba0 fs/read_write.c:1228 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: [] lock_acquire+0x15e/0x450 kernel/locking/lockdep.c:3592 [] percpu_down_read include/linux/percpu-rwsem.h:26 [inline] [] __sb_start_write+0x1af/0x310 fs/super.c:1239 [] sb_start_write include/linux/fs.h:1517 [inline] [] ext4_run_li_request fs/ext4/super.c:2685 [inline] [] ext4_lazyinit_thread fs/ext4/super.c:2784 [inline] [] ext4_lazyinit_thread+0x1e4/0x7b0 fs/ext4/super.c:2760 [] kthread+0x273/0x310 kernel/kthread.c:211 [] ret_from_fork+0x55/0x80 arch/x86/entry/entry_64.S:537 -> #3 (&eli->li_list_mtx){+.+...}: [] lock_acquire+0x15e/0x450 kernel/locking/lockdep.c:3592 [] __mutex_lock_common kernel/locking/mutex.c:521 [inline] [] mutex_lock_nested+0xc1/0xb80 kernel/locking/mutex.c:621 [] ext4_register_li_request+0x2fd/0x7d0 fs/ext4/super.c:2972 [] ext4_remount+0x1366/0x1b90 fs/ext4/super.c:4922 [] do_remount_sb2+0x41b/0x7a0 fs/super.c:781 [] do_remount fs/namespace.c:2347 [inline] [] do_mount+0xfdb/0x2a40 fs/namespace.c:2860 [] SYSC_mount fs/namespace.c:3063 [inline] [] SyS_mount+0x130/0x1d0 fs/namespace.c:3041 [] entry_SYSCALL_64_fastpath+0x1e/0x9a -> #2 (&ext4_li_mtx){+.+.+.}: [] lock_acquire+0x15e/0x450 kernel/locking/lockdep.c:3592 [] __mutex_lock_common kernel/locking/mutex.c:521 [inline] [] mutex_lock_nested+0xc1/0xb80 kernel/locking/mutex.c:621 [] ext4_register_li_request+0x89/0x7d0 fs/ext4/super.c:2945 [] ext4_remount+0x1366/0x1b90 fs/ext4/super.c:4922 [] do_remount_sb2+0x41b/0x7a0 fs/super.c:781 [] do_remount fs/namespace.c:2347 [inline] [] do_mount+0xfdb/0x2a40 fs/namespace.c:2860 [] SYSC_mount fs/namespace.c:3063 [inline] [] SyS_mount+0x130/0x1d0 fs/namespace.c:3041 [] entry_SYSCALL_64_fastpath+0x1e/0x9a -> #1 (&type->s_umount_key#34){++++++}: [] lock_acquire+0x15e/0x450 kernel/locking/lockdep.c:3592 [] down_read+0x42/0x60 kernel/locking/rwsem.c:22 [] iterate_supers+0xe1/0x250 fs/super.c:547 [] selinux_complete_init+0x2f/0x31 security/selinux/hooks.c:6154 [] security_load_policy+0x69d/0x9c0 security/selinux/ss/services.c:2060 [] sel_write_load+0x175/0xf90 security/selinux/selinuxfs.c:535 [] __vfs_write+0x116/0x3d0 fs/read_write.c:491 [] vfs_write+0x182/0x4e0 fs/read_write.c:540 [] SYSC_write fs/read_write.c:587 [inline] [] SyS_write+0xdc/0x1c0 fs/read_write.c:579 [] entry_SYSCALL_64_fastpath+0x1e/0x9a -> #0 (sel_mutex){+.+.+.}: [] check_prev_add kernel/locking/lockdep.c:1853 [inline] [] check_prevs_add kernel/locking/lockdep.c:1958 [inline] [] validate_chain kernel/locking/lockdep.c:2144 [inline] [] __lock_acquire+0x37d6/0x4f50 kernel/locking/lockdep.c:3213 [] lock_acquire+0x15e/0x450 kernel/locking/lockdep.c:3592 [] __mutex_lock_common kernel/locking/mutex.c:521 [inline] [] mutex_lock_nested+0xc1/0xb80 kernel/locking/mutex.c:621 [] sel_read_policy+0x59/0x110 security/selinux/selinuxfs.c:441 [] do_loop_readv_writev+0x148/0x1e0 fs/read_write.c:682 [] do_readv_writev+0x573/0x6e0 fs/read_write.c:812 [] vfs_readv+0x7a/0xb0 fs/read_write.c:836 [] kernel_readv fs/splice.c:586 [inline] [] default_file_splice_read+0x3ac/0x8b0 fs/splice.c:662 [] do_splice_to+0xff/0x160 fs/splice.c:1154 [] splice_direct_to_actor+0x249/0x850 fs/splice.c:1226 [] do_splice_direct+0x1a5/0x260 fs/splice.c:1337 [] do_sendfile+0x4ed/0xba0 fs/read_write.c:1229 [] SYSC_sendfile64 fs/read_write.c:1290 [inline] [] SyS_sendfile64+0x137/0x150 fs/read_write.c:1276 [] entry_SYSCALL_64_fastpath+0x1e/0x9a other info that might help us debug this: Chain exists of: sel_mutex --> &eli->li_list_mtx --> sb_writers#4 Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(sb_writers#4); lock(&eli->li_list_mtx); lock(sb_writers#4); lock(sel_mutex); *** DEADLOCK *** 1 lock held by syz-executor.0/3286: #0: (sb_writers#4){.+.+.+}, at: [] file_start_write include/linux/fs.h:2543 [inline] #0: (sb_writers#4){.+.+.+}, at: [] do_sendfile+0x8a6/0xba0 fs/read_write.c:1228 stack backtrace: CPU: 0 PID: 3286 Comm: syz-executor.0 Not tainted 4.4.174+ #4 0000000000000000 046ea0a0f12e5cc1 ffff8800b54bf0a0 ffffffff81aad1a1 ffffffff84057a80 ffff8801d3978000 ffffffff83ab8870 ffffffff83abc380 ffffffff83ac8080 ffff8800b54bf0f0 ffffffff813abcda ffffffff83e1eb80 Call Trace: [] __dump_stack lib/dump_stack.c:15 [inline] [] dump_stack+0xc1/0x120 lib/dump_stack.c:51 [] print_circular_bug.cold+0x2f7/0x44e kernel/locking/lockdep.c:1226 [] check_prev_add kernel/locking/lockdep.c:1853 [inline] [] check_prevs_add kernel/locking/lockdep.c:1958 [inline] [] validate_chain kernel/locking/lockdep.c:2144 [inline] [] __lock_acquire+0x37d6/0x4f50 kernel/locking/lockdep.c:3213 [] lock_acquire+0x15e/0x450 kernel/locking/lockdep.c:3592 [] __mutex_lock_common kernel/locking/mutex.c:521 [inline] [] mutex_lock_nested+0xc1/0xb80 kernel/locking/mutex.c:621 [] sel_read_policy+0x59/0x110 security/selinux/selinuxfs.c:441 [] do_loop_readv_writev+0x148/0x1e0 fs/read_write.c:682 [] do_readv_writev+0x573/0x6e0 fs/read_write.c:812 [] vfs_readv+0x7a/0xb0 fs/read_write.c:836 [] kernel_readv fs/splice.c:586 [inline] [] default_file_splice_read+0x3ac/0x8b0 fs/splice.c:662 [] do_splice_to+0xff/0x160 fs/splice.c:1154 [] splice_direct_to_actor+0x249/0x850 fs/splice.c:1226 [] do_splice_direct+0x1a5/0x260 fs/splice.c:1337 [] do_sendfile+0x4ed/0xba0 fs/read_write.c:1229 [] SYSC_sendfile64 fs/read_write.c:1290 [inline] [] SyS_sendfile64+0x137/0x150 fs/read_write.c:1276 [] entry_SYSCALL_64_fastpath+0x1e/0x9a audit: type=1400 audit(1574852176.774:20): avc: denied { create } for pid=3307 comm="syz-executor.3" scontext=unconfined_u:system_r:insmod_t:s0-s0:c0.c1023 tcontext=unconfined_u:system_r:insmod_t:s0-s0:c0.c1023 tclass=netlink_crypto_socket permissive=1 syz-executor.2 uses obsolete (PF_INET,SOCK_PACKET)