fs/userfaultd: Fix a recently introduced lockdep complaint
Subsystems (3)
PATCH 1/1 fs/userfaultd: Fix a recently introduced lockdep complaint
Avoid that lockdep reports the following:
=====================================================
WARNING: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected
5.0.0-rc4+ #56 Not tainted
-----------------------------------------------------
syz-executor5/9727 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:00000000a4278d31 (&ctx->fault_pending_wqh){+.+.}, at: spin_lock include/linux/spinlock.h:329 [inline] 00000000a4278d31 (&ctx->fault_pending_wqh){+.+.}, at: userfaultfd_ctx_read fs/userfaultfd.c:1040 [inline] 00000000a4278d31 (&ctx->fault_pending_wqh){+.+.}, at: userfaultfd_read+0x540/0x1940 fs/userfaultfd.c:1198
and this task is already holding:000000000e5b4350 (&ctx->fd_wqh){....}, at: spin_lock_irq include/linux/spinlock.h:354 [inline] 000000000e5b4350 (&ctx->fd_wqh){....}, at: userfaultfd_ctx_read fs/userfaultfd.c:1036 [inline] 000000000e5b4350 (&ctx->fd_wqh){....}, at: userfaultfd_read+0x27a/0x1940 fs/userfaultfd.c:1198 which would create a new lock dependency:
(&ctx->fd_wqh){....} -> (&ctx->fault_pending_wqh){+.+.}
but this new dependency connects a SOFTIRQ-irq-safe lock:
(&(&ctx->ctx_lock)->rlock){..-.}
... which became SOFTIRQ-irq-safe at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841 __raw_spin_lock_irq include/linux/spinlock_api_smp.h:128 [inline] _raw_spin_lock_irq+0x60/0x80 kernel/locking/spinlock.c:160 spin_lock_irq include/linux/spinlock.h:354 [inline] free_ioctx_users+0x2d/0x4a0 fs/aio.c:610 percpu_ref_put_many include/linux/percpu-refcount.h:285 [inline] percpu_ref_put include/linux/percpu-refcount.h:301 [inline] percpu_ref_call_confirm_rcu lib/percpu-refcount.c:123 [inline] percpu_ref_switch_to_atomic_rcu+0x3e7/0x520 lib/percpu-refcount.c:158 __rcu_reclaim kernel/rcu/rcu.h:240 [inline] rcu_do_batch kernel/rcu/tree.c:2452 [inline] invoke_rcu_callbacks kernel/rcu/tree.c:2773 [inline] rcu_process_callbacks+0x928/0x1390 kernel/rcu/tree.c:2754 __do_softirq+0x266/0x95a kernel/softirq.c:292 run_ksoftirqd kernel/softirq.c:654 [inline] run_ksoftirqd+0x8e/0x110 kernel/softirq.c:646 smpboot_thread_fn+0x6ab/0xa10 kernel/smpboot.c:164 kthread+0x357/0x430 kernel/kthread.c:246 ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:352
to a SOFTIRQ-irq-unsafe lock:
(&ctx->fault_pending_wqh){+.+.}
... which became SOFTIRQ-irq-unsafe at:
...
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841 __raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline] _raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144 spin_lock include/linux/spinlock.h:329 [inline] userfaultfd_release+0x497/0x6d0 fs/userfaultfd.c:916 __fput+0x2df/0x8d0 fs/file_table.c:278 ____fput+0x16/0x20 fs/file_table.c:309 task_work_run+0x14a/0x1c0 kernel/task_work.c:113 tracehook_notify_resume include/linux/tracehook.h:188 [inline] exit_to_usermode_loop+0x273/0x2c0 arch/x86/entry/common.c:166 prepare_exit_to_usermode arch/x86/entry/common.c:197 [inline] syscall_return_slowpath arch/x86/entry/common.c:268 [inline] do_syscall_64+0x52d/0x610 arch/x86/entry/common.c:293 entry_SYSCALL_64_after_hwframe+0x49/0xbe
other info that might help us debug this:
Chain exists of: &(&ctx->ctx_lock)->rlock --> &ctx->fd_wqh --> &ctx->fault_pending_wqh
Possible interrupt unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&ctx->fault_pending_wqh);
local_irq_disable();
lock(&(&ctx->ctx_lock)->rlock);
lock(&ctx->fd_wqh);
<Interrupt>
lock(&(&ctx->ctx_lock)->rlock);*** DEADLOCK ***
1 lock held by syz-executor5/9727:
#0: 000000000e5b4350 (&ctx->fd_wqh){....}, at: spin_lock_irq include/linux/spinlock.h:354 [inline] #0: 000000000e5b4350 (&ctx->fd_wqh){....}, at: userfaultfd_ctx_read fs/userfaultfd.c:1036 [inline] #0: 000000000e5b4350 (&ctx->fd_wqh){....}, at: userfaultfd_read+0x27a/0x1940 fs/userfaultfd.c:1198
the dependencies between SOFTIRQ-irq-safe lock and the holding lock:
-> (&(&ctx->ctx_lock)->rlock){..-.} {
IN-SOFTIRQ-W at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock_irq include/linux/spinlock_api_smp.h:128 [inline]
_raw_spin_lock_irq+0x60/0x80 kernel/locking/spinlock.c:160
spin_lock_irq include/linux/spinlock.h:354 [inline]
free_ioctx_users+0x2d/0x4a0 fs/aio.c:610
percpu_ref_put_many include/linux/percpu-refcount.h:285 [inline]
percpu_ref_put include/linux/percpu-refcount.h:301 [inline]
percpu_ref_call_confirm_rcu lib/percpu-refcount.c:123 [inline]
percpu_ref_switch_to_atomic_rcu+0x3e7/0x520 lib/percpu-refcount.c:158
__rcu_reclaim kernel/rcu/rcu.h:240 [inline]
rcu_do_batch kernel/rcu/tree.c:2452 [inline]
invoke_rcu_callbacks kernel/rcu/tree.c:2773 [inline]
rcu_process_callbacks+0x928/0x1390 kernel/rcu/tree.c:2754
__do_softirq+0x266/0x95a kernel/softirq.c:292
run_ksoftirqd kernel/softirq.c:654 [inline]
run_ksoftirqd+0x8e/0x110 kernel/softirq.c:646
smpboot_thread_fn+0x6ab/0xa10 kernel/smpboot.c:164
kthread+0x357/0x430 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:352
INITIAL USE at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock_irq include/linux/spinlock_api_smp.h:128 [inline]
_raw_spin_lock_irq+0x60/0x80 kernel/locking/spinlock.c:160
spin_lock_irq include/linux/spinlock.h:354 [inline]
free_ioctx_users+0x2d/0x4a0 fs/aio.c:610
percpu_ref_put_many include/linux/percpu-refcount.h:285 [inline]
percpu_ref_put include/linux/percpu-refcount.h:301 [inline]
percpu_ref_call_confirm_rcu lib/percpu-refcount.c:123 [inline]
percpu_ref_switch_to_atomic_rcu+0x3e7/0x520 lib/percpu-refcount.c:158
__rcu_reclaim kernel/rcu/rcu.h:240 [inline]
rcu_do_batch kernel/rcu/tree.c:2452 [inline]
invoke_rcu_callbacks kernel/rcu/tree.c:2773 [inline]
rcu_process_callbacks+0x928/0x1390 kernel/rcu/tree.c:2754
__do_softirq+0x266/0x95a kernel/softirq.c:292
run_ksoftirqd kernel/softirq.c:654 [inline]
run_ksoftirqd+0x8e/0x110 kernel/softirq.c:646
smpboot_thread_fn+0x6ab/0xa10 kernel/smpboot.c:164
kthread+0x357/0x430 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:352
}
... key at: [<ffffffff8a5760a0>] __key.51972+0x0/0x40
... acquired at:
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
aio_poll fs/aio.c:1772 [inline]
__io_submit_one fs/aio.c:1875 [inline]
io_submit_one+0xedf/0x1cf0 fs/aio.c:1908
__do_sys_io_submit fs/aio.c:1953 [inline]
__se_sys_io_submit fs/aio.c:1923 [inline]
__x64_sys_io_submit+0x1bd/0x580 fs/aio.c:1923
do_syscall_64+0x103/0x610 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x49/0xbe-> (&ctx->fd_wqh){....} {
INITIAL USE at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
_raw_spin_lock_irqsave+0x95/0xcd kernel/locking/spinlock.c:152
__wake_up_common_lock+0xc7/0x190 kernel/sched/wait.c:120
__wake_up+0xe/0x10 kernel/sched/wait.c:145
userfaultfd_release+0x4f5/0x6d0 fs/userfaultfd.c:924
__fput+0x2df/0x8d0 fs/file_table.c:278
____fput+0x16/0x20 fs/file_table.c:309
task_work_run+0x14a/0x1c0 kernel/task_work.c:113
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_usermode_loop+0x273/0x2c0 arch/x86/entry/common.c:166
prepare_exit_to_usermode arch/x86/entry/common.c:197 [inline]
syscall_return_slowpath arch/x86/entry/common.c:268 [inline]
do_syscall_64+0x52d/0x610 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe
}
... key at: [<ffffffff8a575e20>] __key.44854+0x0/0x40
... acquired at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
userfaultfd_ctx_read fs/userfaultfd.c:1040 [inline]
userfaultfd_read+0x540/0x1940 fs/userfaultfd.c:1198
__vfs_read+0x116/0x8c0 fs/read_write.c:416
vfs_read+0x194/0x3e0 fs/read_write.c:452
ksys_read+0xea/0x1f0 fs/read_write.c:578
__do_sys_read fs/read_write.c:588 [inline]
__se_sys_read fs/read_write.c:586 [inline]
__x64_sys_read+0x73/0xb0 fs/read_write.c:586
do_syscall_64+0x103/0x610 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x49/0xbethe dependencies between the lock to be acquired
and SOFTIRQ-irq-unsafe lock:
-> (&ctx->fault_pending_wqh){+.+.} {
HARDIRQ-ON-W at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
userfaultfd_release+0x497/0x6d0 fs/userfaultfd.c:916
__fput+0x2df/0x8d0 fs/file_table.c:278
____fput+0x16/0x20 fs/file_table.c:309
task_work_run+0x14a/0x1c0 kernel/task_work.c:113
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_usermode_loop+0x273/0x2c0 arch/x86/entry/common.c:166
prepare_exit_to_usermode arch/x86/entry/common.c:197 [inline]
syscall_return_slowpath arch/x86/entry/common.c:268 [inline]
do_syscall_64+0x52d/0x610 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe
SOFTIRQ-ON-W at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
userfaultfd_release+0x497/0x6d0 fs/userfaultfd.c:916
__fput+0x2df/0x8d0 fs/file_table.c:278
____fput+0x16/0x20 fs/file_table.c:309
task_work_run+0x14a/0x1c0 kernel/task_work.c:113
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_usermode_loop+0x273/0x2c0 arch/x86/entry/common.c:166
prepare_exit_to_usermode arch/x86/entry/common.c:197 [inline]
syscall_return_slowpath arch/x86/entry/common.c:268 [inline]
do_syscall_64+0x52d/0x610 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe
INITIAL USE at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
userfaultfd_release+0x497/0x6d0 fs/userfaultfd.c:916
__fput+0x2df/0x8d0 fs/file_table.c:278
____fput+0x16/0x20 fs/file_table.c:309
task_work_run+0x14a/0x1c0 kernel/task_work.c:113
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_usermode_loop+0x273/0x2c0 arch/x86/entry/common.c:166
prepare_exit_to_usermode arch/x86/entry/common.c:197 [inline]
syscall_return_slowpath arch/x86/entry/common.c:268 [inline]
do_syscall_64+0x52d/0x610 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe
}
... key at: [<ffffffff8a575ee0>] __key.44851+0x0/0x40
... acquired at:
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
userfaultfd_ctx_read fs/userfaultfd.c:1040 [inline]
userfaultfd_read+0x540/0x1940 fs/userfaultfd.c:1198
__vfs_read+0x116/0x8c0 fs/read_write.c:416
vfs_read+0x194/0x3e0 fs/read_write.c:452
ksys_read+0xea/0x1f0 fs/read_write.c:578
__do_sys_read fs/read_write.c:588 [inline]
__se_sys_read fs/read_write.c:586 [inline]
__x64_sys_read+0x73/0xb0 fs/read_write.c:586
do_syscall_64+0x103/0x610 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x49/0xbe
stack backtrace:
CPU: 1 PID: 9727 Comm: syz-executor5 Not tainted 5.0.0-rc4+ #56
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x172/0x1f0 lib/dump_stack.c:113
print_bad_irq_dependency kernel/locking/lockdep.c:1573 [inline]
check_usage.cold+0x60f/0x940 kernel/locking/lockdep.c:1605
check_irq_usage kernel/locking/lockdep.c:1661 [inline]
check_prev_add_irq kernel/locking/lockdep_states.h:8 [inline]
check_prev_add kernel/locking/lockdep.c:1871 [inline]
check_prevs_add kernel/locking/lockdep.c:1979 [inline]
validate_chain kernel/locking/lockdep.c:2350 [inline]
__lock_acquire+0x1f47/0x4700 kernel/locking/lockdep.c:3338
lock_acquire+0x16f/0x3f0 kernel/locking/lockdep.c:3841
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2f/0x40 kernel/locking/spinlock.c:144
spin_lock include/linux/spinlock.h:329 [inline]
userfaultfd_ctx_read fs/userfaultfd.c:1040 [inline]
userfaultfd_read+0x540/0x1940 fs/userfaultfd.c:1198
__vfs_read+0x116/0x8c0 fs/read_write.c:416
vfs_read+0x194/0x3e0 fs/read_write.c:452
ksys_read+0xea/0x1f0 fs/read_write.c:578
__do_sys_read fs/read_write.c:588 [inline]
__se_sys_read fs/read_write.c:586 [inline]
__x64_sys_read+0x73/0xb0 fs/read_write.c:586
do_syscall_64+0x103/0x610 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x49/0xbe--- fs/userfaultfd.c | 32 ++++++++++++++++---------------- 1 file changed, 16 insertions(+), 16 deletions(-)
fs/userfaultfd.c +16 -16
@@ -458,7 +458,7 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason) blocking_state = return_to_userland ?TASK_INTERRUPTIBLE :TASK_KILLABLE ; -spin_lock (&ctx->fault_pending_wqh.lock); +spin_lock_irq (&ctx->fault_pending_wqh.lock);/* * After the __add_wait_queue the uwq is visible to userland * through poll/read().
@@ -470,7 +470,7 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason)* __add_wait_queue. */ set_current_state (blocking_state); -spin_unlock (&ctx->fault_pending_wqh.lock); +spin_unlock_irq (&ctx->fault_pending_wqh.lock);if (!is_vm_hugetlb_page (vmf->vma)) must_wait = userfaultfd_must_wait(ctx, vmf->address, vmf->flags,
@@ -552,13 +552,13 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason)* kernel stack can be released after the list_del_init. */ if (!list_empty_careful (&uwq.wq.entry)) { -spin_lock (&ctx->fault_pending_wqh.lock); +spin_lock_irq (&ctx->fault_pending_wqh.lock);/* * No need of list_del_init(), the uwq on the stack * will be freed shortly anyway. */ list_del (&uwq.wq.entry); -spin_unlock (&ctx->fault_pending_wqh.lock); +spin_unlock_irq (&ctx->fault_pending_wqh.lock); }/*
@@ -583,7 +583,7 @@ static void userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx,init_waitqueue_entry (&ewq->wq, current); release_new_ctx =NULL ; -spin_lock (&ctx->event_wqh.lock); +spin_lock_irq (&ctx->event_wqh.lock);/* * After the __add_wait_queue the uwq is visible to userland * through poll/read().
@@ -613,15 +613,15 @@ static void userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx,break ; } -spin_unlock (&ctx->event_wqh.lock); +spin_unlock_irq (&ctx->event_wqh.lock);wake_up_poll (&ctx->fd_wqh,EPOLLIN );schedule (); -spin_lock (&ctx->event_wqh.lock); +spin_lock_irq (&ctx->event_wqh.lock); }__set_current_state (TASK_RUNNING ); -spin_unlock (&ctx->event_wqh.lock); +spin_unlock_irq (&ctx->event_wqh.lock);if (release_new_ctx) {struct vm_area_struct *vma;
@@ -913,10 +913,10 @@ static int userfaultfd_release(struct inode *inode, struct file *file)* the last page faults that may have been already waiting on * the fault_*wqh. */ -spin_lock (&ctx->fault_pending_wqh.lock); +spin_lock_irq (&ctx->fault_pending_wqh.lock);__wake_up_locked_key (&ctx->fault_pending_wqh,TASK_NORMAL , &range);__wake_up (&ctx->fault_wqh,TASK_NORMAL ,1 , &range); -spin_unlock (&ctx->fault_pending_wqh.lock); +spin_unlock_irq (&ctx->fault_pending_wqh.lock);/* Flush pending events that may still wait on event_wqh */ wake_up_all (&ctx->event_wqh);
@@ -1129,7 +1129,7 @@ static ssize_t userfaultfd_ctx_read(struct userfaultfd_ctx *ctx, int no_wait,if (!ret && msg->event ==UFFD_EVENT_FORK ) { ret =resolve_userfault_fork (ctx, fork_nctx, msg); -spin_lock (&ctx->event_wqh.lock); +spin_lock_irq (&ctx->event_wqh.lock);if (!list_empty (&fork_event)) {/* * The fork thread didn't abort, so we can
@@ -1175,7 +1175,7 @@ static ssize_t userfaultfd_ctx_read(struct userfaultfd_ctx *ctx, int no_wait,if (ret)userfaultfd_ctx_put (fork_nctx); } -spin_unlock (&ctx->event_wqh.lock); +spin_unlock_irq (&ctx->event_wqh.lock); }return ret;
@@ -1214,14 +1214,14 @@ static ssize_t userfaultfd_read(struct file *file, char __user *buf,static void __wake_userfault (struct userfaultfd_ctx *ctx,struct userfaultfd_wake_range *range) { -spin_lock (&ctx->fault_pending_wqh.lock); +spin_lock_irq (&ctx->fault_pending_wqh.lock);/* wake all in the range and autoremove */ if (waitqueue_active (&ctx->fault_pending_wqh))__wake_up_locked_key (&ctx->fault_pending_wqh,TASK_NORMAL , range);if (waitqueue_active (&ctx->fault_wqh))__wake_up (&ctx->fault_wqh,TASK_NORMAL ,1 , range); -spin_unlock (&ctx->fault_pending_wqh.lock); +spin_unlock_irq (&ctx->fault_pending_wqh.lock); } static __always_inline void wake_userfault(struct userfaultfd_ctx *ctx,
@@ -1872,7 +1872,7 @@ static void userfaultfd_show_fdinfo(struct seq_file *m, struct file *f)wait_queue_entry_t *wq;unsigned long pending =0 , total =0 ; -spin_lock (&ctx->fault_pending_wqh.lock); +spin_lock_irq (&ctx->fault_pending_wqh.lock);list_for_each_entry (wq, &ctx->fault_pending_wqh.head, entry) { pending++; total++;
@@ -1880,7 +1880,7 @@ static void userfaultfd_show_fdinfo(struct seq_file *m, struct file *f)list_for_each_entry (wq, &ctx->fault_wqh.head, entry) { total++; } -spin_unlock (&ctx->fault_pending_wqh.lock); +spin_unlock_irq (&ctx->fault_pending_wqh.lock);/* * If more protocols will be added, there will be all shown
Signature · 1 line
2.20.1.791.gb4d0f1c61a-goog
Quoted 1 line · Bart Van Assche
Avoid that lockdep reports the following: [ ... ]
Hi Christoph,
If patch "aio: Fix locking in aio_poll()" (https://marc.info/?l=linux-fsdevel&m=154967401921791) would be accepted, would that allow to revert commit ae62c16e105a ("userfaultfd: disable irqs when taking the waitqueue lock") instead of applying this patch?
Thanks,
Bart.