| Message ID | 20241025003722.3630252-4-ming.lei@redhat.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | block: model freeze/enter queue as lock for lockdep | expand |
On 25.10.2024 02:37, Ming Lei wrote: > Recently we got several deadlock report[1][2][3] caused by > blk_mq_freeze_queue and blk_enter_queue(). > > Turns out the two are just like acquiring read/write lock, so model them > as read/write lock for supporting lockdep: > > 1) model q->q_usage_counter as two locks(io and queue lock) > > - queue lock covers sync with blk_enter_queue() > > - io lock covers sync with bio_enter_queue() > > 2) make the lockdep class/key as per-queue: > > - different subsystem has very different lock use pattern, shared lock > class causes false positive easily > > - freeze_queue degrades to no lock in case that disk state becomes DEAD > because bio_enter_queue() won't be blocked any more > > - freeze_queue degrades to no lock in case that request queue becomes dying > because blk_enter_queue() won't be blocked any more > > 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock > - it is exclusive lock, so dependency with blk_enter_queue() is covered > > - it is trylock because blk_mq_freeze_queue() are allowed to run > concurrently > > 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() > - nested blk_enter_queue() are allowed > > - dependency with blk_mq_freeze_queue() is covered > > - blk_queue_exit() is often called from other contexts(such as irq), and > it can't be annotated as lock_release(), so simply do it in > blk_enter_queue(), this way still covered cases as many as possible > > With lockdep support, such kind of reports may be reported asap and > needn't wait until the real deadlock is triggered. > > For example, lockdep report can be triggered in the report[3] with this > patch applied. > > [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' > https://bugzilla.kernel.org/show_bug.cgi?id=219166 > > [2] del_gendisk() vs blk_queue_enter() race condition > https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ > > [3] queue_freeze & queue_enter deadlock in scsi > https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u > > Reviewed-by: Christoph Hellwig <hch@lst.de> > Signed-off-by: Ming Lei <ming.lei@redhat.com> This patch landed yesterday in linux-next as commit f1be1788a32e ("block: model freeze & enter queue as lock for supporting lockdep"). In my tests I found that it introduces the following 2 lockdep warnings: 1. On Samsung Exynos 4412-based Odroid U3 board (ARM 32bit), observed when booting it: ====================================================== WARNING: possible circular locking dependency detected 6.12.0-rc4-00037-gf1be1788a32e #9290 Not tainted ------------------------------------------------------ find/1284 is trying to acquire lock: cf3b8534 (&mm->mmap_lock){++++}-{3:3}, at: __might_fault+0x30/0x70 but task is already holding lock: c203a0c8 (&sb->s_type->i_mutex_key#2){++++}-{3:3}, at: iterate_dir+0x30/0x140 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (&sb->s_type->i_mutex_key#2){++++}-{3:3}: down_write+0x44/0xc4 start_creating+0x8c/0x170 debugfs_create_dir+0x1c/0x178 blk_register_queue+0xa0/0x1c0 add_disk_fwnode+0x210/0x434 brd_alloc+0x1cc/0x210 brd_init+0xac/0x104 do_one_initcall+0x64/0x30c kernel_init_freeable+0x1c4/0x228 kernel_init+0x1c/0x12c ret_from_fork+0x14/0x28 -> #3 (&q->debugfs_mutex){+.+.}-{3:3}: __mutex_lock+0x94/0x94c mutex_lock_nested+0x1c/0x24 blk_mq_init_sched+0x140/0x204 elevator_init_mq+0xb8/0x130 add_disk_fwnode+0x3c/0x434 mmc_blk_alloc_req+0x34c/0x464 mmc_blk_probe+0x1f4/0x6f8 really_probe+0xe0/0x3d8 __driver_probe_device+0x9c/0x1e4 driver_probe_device+0x30/0xc0 __device_attach_driver+0xa8/0x120 bus_for_each_drv+0x80/0xcc __device_attach+0xac/0x1fc bus_probe_device+0x8c/0x90 device_add+0x5a4/0x7cc mmc_add_card+0x118/0x2c8 mmc_attach_mmc+0xd8/0x174 mmc_rescan+0x2ec/0x3a8 process_one_work+0x240/0x6d0 worker_thread+0x1a0/0x398 kthread+0x104/0x138 ret_from_fork+0x14/0x28 -> #2 (&q->q_usage_counter(io)#17){++++}-{0:0}: blk_mq_submit_bio+0x8dc/0xb34 __submit_bio+0x78/0x148 submit_bio_noacct_nocheck+0x204/0x32c ext4_mpage_readpages+0x558/0x7b0 read_pages+0x64/0x28c page_cache_ra_unbounded+0x118/0x1bc filemap_get_pages+0x124/0x7ec filemap_read+0x174/0x5b0 __kernel_read+0x128/0x2c0 bprm_execve+0x230/0x7a4 kernel_execve+0xec/0x194 try_to_run_init_process+0xc/0x38 kernel_init+0xdc/0x12c ret_from_fork+0x14/0x28 -> #1 (mapping.invalidate_lock#2){++++}-{3:3}: down_read+0x44/0x224 filemap_fault+0x648/0x10f0 __do_fault+0x38/0xd4 handle_mm_fault+0xaf8/0x14d0 do_page_fault+0xe0/0x5c8 do_DataAbort+0x3c/0xb0 __dabt_svc+0x50/0x80 __clear_user_std+0x34/0x68 elf_load+0x1a8/0x208 load_elf_binary+0x3f4/0x13cc bprm_execve+0x28c/0x7a4 do_execveat_common+0x150/0x198 sys_execve+0x30/0x38 ret_fast_syscall+0x0/0x1c -> #0 (&mm->mmap_lock){++++}-{3:3}: __lock_acquire+0x158c/0x2970 lock_acquire+0x130/0x384 __might_fault+0x50/0x70 filldir64+0x94/0x28c dcache_readdir+0x174/0x260 iterate_dir+0x64/0x140 sys_getdents64+0x60/0x130 ret_fast_syscall+0x0/0x1c other info that might help us debug this: Chain exists of: &mm->mmap_lock --> &q->debugfs_mutex --> &sb->s_type->i_mutex_key#2 Possible unsafe locking scenario: CPU0 CPU1 ---- ---- rlock(&sb->s_type->i_mutex_key#2); lock(&q->debugfs_mutex); lock(&sb->s_type->i_mutex_key#2); rlock(&mm->mmap_lock); *** DEADLOCK *** 2 locks held by find/1284: #0: c3df1e88 (&f->f_pos_lock){+.+.}-{3:3}, at: fdget_pos+0x88/0xd0 #1: c203a0c8 (&sb->s_type->i_mutex_key#2){++++}-{3:3}, at: iterate_dir+0x30/0x140 stack backtrace: CPU: 1 UID: 0 PID: 1284 Comm: find Not tainted 6.12.0-rc4-00037-gf1be1788a32e #9290 Hardware name: Samsung Exynos (Flattened Device Tree) Call trace: unwind_backtrace from show_stack+0x10/0x14 show_stack from dump_stack_lvl+0x68/0x88 dump_stack_lvl from print_circular_bug+0x31c/0x394 print_circular_bug from check_noncircular+0x16c/0x184 check_noncircular from __lock_acquire+0x158c/0x2970 __lock_acquire from lock_acquire+0x130/0x384 lock_acquire from __might_fault+0x50/0x70 __might_fault from filldir64+0x94/0x28c filldir64 from dcache_readdir+0x174/0x260 dcache_readdir from iterate_dir+0x64/0x140 iterate_dir from sys_getdents64+0x60/0x130 sys_getdents64 from ret_fast_syscall+0x0/0x1c Exception stack(0xf22b5fa8 to 0xf22b5ff0) 5fa0: 004b4fa0 004b4f80 00000004 004b4fa0 00008000 00000000 5fc0: 004b4fa0 004b4f80 00000001 000000d9 00000000 004b4af0 00000000 000010ea 5fe0: 004b1eb4 bea05af0 b6da4b08 b6da4a28 --->8--- 2. On QEMU's ARM64 virt machine, observed during system suspend/resume cycle: # time rtcwake -s10 -mmem rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 PM: suspend entry (s2idle) Filesystems sync: 0.004 seconds Freezing user space processes Freezing user space processes completed (elapsed 0.007 seconds) OOM killer disabled. Freezing remaining freezable tasks Freezing remaining freezable tasks completed (elapsed 0.004 seconds) ====================================================== WARNING: possible circular locking dependency detected 6.12.0-rc4+ #9291 Not tainted ------------------------------------------------------ rtcwake/1299 is trying to acquire lock: ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 but task is already holding lock: ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: virtblk_freeze+0x24/0x60 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&q->q_usage_counter(io)#5){++++}-{0:0}: blk_mq_submit_bio+0x7c0/0x9d8 __submit_bio+0x74/0x164 submit_bio_noacct_nocheck+0x2d4/0x3b4 submit_bio_noacct+0x148/0x3fc submit_bio+0x130/0x204 submit_bh_wbc+0x148/0x1bc submit_bh+0x18/0x24 ext4_read_bh_nowait+0x70/0x100 ext4_sb_breadahead_unmovable+0x50/0x80 __ext4_get_inode_loc+0x354/0x654 ext4_get_inode_loc+0x40/0xa8 ext4_reserve_inode_write+0x40/0xf0 __ext4_mark_inode_dirty+0x88/0x300 ext4_ext_tree_init+0x40/0x4c __ext4_new_inode+0x99c/0x1614 ext4_create+0xe4/0x1d4 lookup_open.isra.0+0x47c/0x540 path_openat+0x370/0x9e8 do_filp_open+0x80/0x130 do_sys_openat2+0xb4/0xe8 __arm64_compat_sys_openat+0x5c/0xa4 invoke_syscall+0x48/0x110 el0_svc_common.constprop.0+0x40/0xe8 do_el0_svc_compat+0x20/0x3c el0_svc_compat+0x44/0xe0 el0t_32_sync_handler+0x98/0x148 el0t_32_sync+0x194/0x198 -> #2 (jbd2_handle){++++}-{0:0}: start_this_handle+0x178/0x4e8 jbd2__journal_start+0x110/0x298 __ext4_journal_start_sb+0x9c/0x274 ext4_dirty_inode+0x38/0x88 __mark_inode_dirty+0x90/0x568 generic_update_time+0x50/0x64 touch_atime+0x2c0/0x324 ext4_file_mmap+0x68/0x88 mmap_region+0x448/0xa38 do_mmap+0x3dc/0x540 vm_mmap_pgoff+0xf8/0x1b4 ksys_mmap_pgoff+0x148/0x1f0 __arm64_compat_sys_aarch32_mmap2+0x20/0x2c invoke_syscall+0x48/0x110 el0_svc_common.constprop.0+0x40/0xe8 do_el0_svc_compat+0x20/0x3c el0_svc_compat+0x44/0xe0 el0t_32_sync_handler+0x98/0x148 el0t_32_sync+0x194/0x198 -> #1 (&mm->mmap_lock){++++}-{3:3}: __might_fault+0x5c/0x80 inet_gifconf+0xcc/0x278 dev_ifconf+0xc4/0x1f8 sock_ioctl+0x234/0x384 compat_sock_ioctl+0x180/0x35c __arm64_compat_sys_ioctl+0x14c/0x16c invoke_syscall+0x48/0x110 el0_svc_common.constprop.0+0x40/0xe8 do_el0_svc_compat+0x20/0x3c el0_svc_compat+0x44/0xe0 el0t_32_sync_handler+0x98/0x148 el0t_32_sync+0x194/0x198 -> #0 (rtnl_mutex){+.+.}-{3:3}: __lock_acquire+0x1374/0x2224 lock_acquire+0x200/0x340 __mutex_lock+0x98/0x428 mutex_lock_nested+0x24/0x30 rtnl_lock+0x1c/0x28 virtnet_freeze_down.isra.0+0x20/0x9c virtnet_freeze+0x44/0x60 virtio_device_freeze+0x68/0x94 virtio_mmio_freeze+0x14/0x20 platform_pm_suspend+0x2c/0x6c dpm_run_callback+0xa4/0x218 device_suspend+0x12c/0x52c dpm_suspend+0x10c/0x2e4 dpm_suspend_start+0x70/0x78 suspend_devices_and_enter+0x130/0x798 pm_suspend+0x1ac/0x2e8 state_store+0x8c/0x110 kobj_attr_store+0x18/0x2c sysfs_kf_write+0x4c/0x78 kernfs_fop_write_iter+0x120/0x1b4 vfs_write+0x2b0/0x35c ksys_write+0x68/0xf4 __arm64_sys_write+0x1c/0x28 invoke_syscall+0x48/0x110 el0_svc_common.constprop.0+0x40/0xe8 do_el0_svc_compat+0x20/0x3c el0_svc_compat+0x44/0xe0 el0t_32_sync_handler+0x98/0x148 el0t_32_sync+0x194/0x198 other info that might help us debug this: Chain exists of: rtnl_mutex --> jbd2_handle --> &q->q_usage_counter(io)#5 Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&q->q_usage_counter(io)#5); lock(jbd2_handle); lock(&q->q_usage_counter(io)#5); lock(rtnl_mutex); *** DEADLOCK *** 9 locks held by rtcwake/1299: #0: ffff0000069103f8 (sb_writers#7){.+.+}-{0:0}, at: vfs_write+0x1e4/0x35c #1: ffff000007c0ae88 (&of->mutex#2){+.+.}-{3:3}, at: kernfs_fop_write_iter+0xf0/0x1b4 #2: ffff0000046906e8 (kn->active#24){.+.+}-{0:0}, at: kernfs_fop_write_iter+0xf8/0x1b4 #3: ffff800082fd9908 (system_transition_mutex){+.+.}-{3:3}, at: pm_suspend+0x88/0x2e8 #4: ffff000006137678 (&q->q_usage_counter(io)#6){++++}-{0:0}, at: virtblk_freeze+0x24/0x60 #5: ffff0000061376b0 (&q->q_usage_counter(queue)#2){++++}-{0:0}, at: virtblk_freeze+0x24/0x60 #6: ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: virtblk_freeze+0x24/0x60 #7: ffff000006136da0 (&q->q_usage_counter(queue)){++++}-{0:0}, at: virtblk_freeze+0x24/0x60 #8: ffff0000056208f8 (&dev->mutex){....}-{3:3}, at: device_suspend+0xf8/0x52c stack backtrace: CPU: 0 UID: 0 PID: 1299 Comm: rtcwake Not tainted 6.12.0-rc4+ #9291 Hardware name: linux,dummy-virt (DT) Call trace: dump_backtrace+0x94/0xec show_stack+0x18/0x24 dump_stack_lvl+0x90/0xd0 dump_stack+0x18/0x24 print_circular_bug+0x298/0x37c check_noncircular+0x15c/0x170 __lock_acquire+0x1374/0x2224 lock_acquire+0x200/0x340 __mutex_lock+0x98/0x428 mutex_lock_nested+0x24/0x30 rtnl_lock+0x1c/0x28 virtnet_freeze_down.isra.0+0x20/0x9c virtnet_freeze+0x44/0x60 virtio_device_freeze+0x68/0x94 virtio_mmio_freeze+0x14/0x20 platform_pm_suspend+0x2c/0x6c dpm_run_callback+0xa4/0x218 device_suspend+0x12c/0x52c dpm_suspend+0x10c/0x2e4 dpm_suspend_start+0x70/0x78 suspend_devices_and_enter+0x130/0x798 pm_suspend+0x1ac/0x2e8 state_store+0x8c/0x110 kobj_attr_store+0x18/0x2c sysfs_kf_write+0x4c/0x78 kernfs_fop_write_iter+0x120/0x1b4 vfs_write+0x2b0/0x35c ksys_write+0x68/0xf4 __arm64_sys_write+0x1c/0x28 invoke_syscall+0x48/0x110 el0_svc_common.constprop.0+0x40/0xe8 do_el0_svc_compat+0x20/0x3c el0_svc_compat+0x44/0xe0 el0t_32_sync_handler+0x98/0x148 el0t_32_sync+0x194/0x198 virtio_blk virtio2: 1/0/0 default/read/poll queues virtio_blk virtio3: 1/0/0 default/read/poll queues OOM killer enabled. Restarting tasks ... done. random: crng reseeded on system resumption PM: suspend exit --->8--- Let me know if You need more information about my test systems. > --- > block/blk-core.c | 18 ++++++++++++++++-- > block/blk-mq.c | 26 ++++++++++++++++++++++---- > block/blk.h | 29 ++++++++++++++++++++++++++--- > block/genhd.c | 15 +++++++++++---- > include/linux/blkdev.h | 6 ++++++ > 5 files changed, 81 insertions(+), 13 deletions(-) > ... Best regards
On Tue, Oct 29, 2024 at 12:13:35PM +0100, Marek Szyprowski wrote: > On 25.10.2024 02:37, Ming Lei wrote: > > Recently we got several deadlock report[1][2][3] caused by > > blk_mq_freeze_queue and blk_enter_queue(). > > > > Turns out the two are just like acquiring read/write lock, so model them > > as read/write lock for supporting lockdep: > > > > 1) model q->q_usage_counter as two locks(io and queue lock) > > > > - queue lock covers sync with blk_enter_queue() > > > > - io lock covers sync with bio_enter_queue() > > > > 2) make the lockdep class/key as per-queue: > > > > - different subsystem has very different lock use pattern, shared lock > > class causes false positive easily > > > > - freeze_queue degrades to no lock in case that disk state becomes DEAD > > because bio_enter_queue() won't be blocked any more > > > > - freeze_queue degrades to no lock in case that request queue becomes dying > > because blk_enter_queue() won't be blocked any more > > > > 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock > > - it is exclusive lock, so dependency with blk_enter_queue() is covered > > > > - it is trylock because blk_mq_freeze_queue() are allowed to run > > concurrently > > > > 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() > > - nested blk_enter_queue() are allowed > > > > - dependency with blk_mq_freeze_queue() is covered > > > > - blk_queue_exit() is often called from other contexts(such as irq), and > > it can't be annotated as lock_release(), so simply do it in > > blk_enter_queue(), this way still covered cases as many as possible > > > > With lockdep support, such kind of reports may be reported asap and > > needn't wait until the real deadlock is triggered. > > > > For example, lockdep report can be triggered in the report[3] with this > > patch applied. > > > > [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' > > https://bugzilla.kernel.org/show_bug.cgi?id=219166 > > > > [2] del_gendisk() vs blk_queue_enter() race condition > > https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ > > > > [3] queue_freeze & queue_enter deadlock in scsi > > https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u > > > > Reviewed-by: Christoph Hellwig <hch@lst.de> > > Signed-off-by: Ming Lei <ming.lei@redhat.com> > > This patch landed yesterday in linux-next as commit f1be1788a32e > ("block: model freeze & enter queue as lock for supporting lockdep"). > In my tests I found that it introduces the following 2 lockdep warnings: > > 1. On Samsung Exynos 4412-based Odroid U3 board (ARM 32bit), observed > when booting it: > > ====================================================== > WARNING: possible circular locking dependency detected > 6.12.0-rc4-00037-gf1be1788a32e #9290 Not tainted > ------------------------------------------------------ > find/1284 is trying to acquire lock: > cf3b8534 (&mm->mmap_lock){++++}-{3:3}, at: __might_fault+0x30/0x70 > > but task is already holding lock: > c203a0c8 (&sb->s_type->i_mutex_key#2){++++}-{3:3}, at: > iterate_dir+0x30/0x140 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #4 (&sb->s_type->i_mutex_key#2){++++}-{3:3}: > down_write+0x44/0xc4 > start_creating+0x8c/0x170 > debugfs_create_dir+0x1c/0x178 > blk_register_queue+0xa0/0x1c0 > add_disk_fwnode+0x210/0x434 > brd_alloc+0x1cc/0x210 > brd_init+0xac/0x104 > do_one_initcall+0x64/0x30c > kernel_init_freeable+0x1c4/0x228 > kernel_init+0x1c/0x12c > ret_from_fork+0x14/0x28 > > -> #3 (&q->debugfs_mutex){+.+.}-{3:3}: > __mutex_lock+0x94/0x94c > mutex_lock_nested+0x1c/0x24 > blk_mq_init_sched+0x140/0x204 > elevator_init_mq+0xb8/0x130 > add_disk_fwnode+0x3c/0x434 The above chain can be cut by the following patch because disk state can be thought as DEAD in add_disk(), can you test it? diff --git a/block/elevator.c b/block/elevator.c index 4122026b11f1..efa6ff941a25 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -600,12 +600,14 @@ void elevator_init_mq(struct request_queue *q) * requests, then no need to quiesce queue which may add long boot * latency, especially when lots of disks are involved. */ - blk_mq_freeze_queue(q); + if (__blk_freeze_queue_start(q)) + blk_freeze_acquire_lock(q, true, false); blk_mq_cancel_work_sync(q); err = blk_mq_init_sched(q, e); - blk_mq_unfreeze_queue(q); + if (__blk_mq_unfreeze_queue(q, false)) + blk_unfreeze_release_lock(q, true, false); if (err) { pr_warn("\"%s\" elevator initialization failed, " ... > 2 locks held by find/1284: > #0: c3df1e88 (&f->f_pos_lock){+.+.}-{3:3}, at: fdget_pos+0x88/0xd0 > #1: c203a0c8 (&sb->s_type->i_mutex_key#2){++++}-{3:3}, at: > iterate_dir+0x30/0x140 > > stack backtrace: > CPU: 1 UID: 0 PID: 1284 Comm: find Not tainted > 6.12.0-rc4-00037-gf1be1788a32e #9290 > Hardware name: Samsung Exynos (Flattened Device Tree) > Call trace: > unwind_backtrace from show_stack+0x10/0x14 > show_stack from dump_stack_lvl+0x68/0x88 > dump_stack_lvl from print_circular_bug+0x31c/0x394 > print_circular_bug from check_noncircular+0x16c/0x184 > check_noncircular from __lock_acquire+0x158c/0x2970 > __lock_acquire from lock_acquire+0x130/0x384 > lock_acquire from __might_fault+0x50/0x70 > __might_fault from filldir64+0x94/0x28c > filldir64 from dcache_readdir+0x174/0x260 > dcache_readdir from iterate_dir+0x64/0x140 > iterate_dir from sys_getdents64+0x60/0x130 > sys_getdents64 from ret_fast_syscall+0x0/0x1c > Exception stack(0xf22b5fa8 to 0xf22b5ff0) > 5fa0: 004b4fa0 004b4f80 00000004 004b4fa0 00008000 > 00000000 > 5fc0: 004b4fa0 004b4f80 00000001 000000d9 00000000 004b4af0 00000000 > 000010ea > 5fe0: 004b1eb4 bea05af0 b6da4b08 b6da4a28 > > --->8--- > > > 2. On QEMU's ARM64 virt machine, observed during system suspend/resume > cycle: > > # time rtcwake -s10 -mmem > rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 > PM: suspend entry (s2idle) > Filesystems sync: 0.004 seconds > Freezing user space processes > Freezing user space processes completed (elapsed 0.007 seconds) > OOM killer disabled. > Freezing remaining freezable tasks > Freezing remaining freezable tasks completed (elapsed 0.004 seconds) > > ====================================================== > WARNING: possible circular locking dependency detected > 6.12.0-rc4+ #9291 Not tainted > ------------------------------------------------------ > rtcwake/1299 is trying to acquire lock: > ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 > > but task is already holding lock: > ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: > virtblk_freeze+0x24/0x60 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: This one looks a real thing, at least the added lockdep code works as expected, also the blk_mq_freeze_queue() use in virtio-blk's ->suspend() is questionable. I will take a further look. Thanks, Ming
On 29.10.2024 16:58, Ming Lei wrote: > On Tue, Oct 29, 2024 at 12:13:35PM +0100, Marek Szyprowski wrote: >> On 25.10.2024 02:37, Ming Lei wrote: >>> Recently we got several deadlock report[1][2][3] caused by >>> blk_mq_freeze_queue and blk_enter_queue(). >>> >>> Turns out the two are just like acquiring read/write lock, so model them >>> as read/write lock for supporting lockdep: >>> >>> 1) model q->q_usage_counter as two locks(io and queue lock) >>> >>> - queue lock covers sync with blk_enter_queue() >>> >>> - io lock covers sync with bio_enter_queue() >>> >>> 2) make the lockdep class/key as per-queue: >>> >>> - different subsystem has very different lock use pattern, shared lock >>> class causes false positive easily >>> >>> - freeze_queue degrades to no lock in case that disk state becomes DEAD >>> because bio_enter_queue() won't be blocked any more >>> >>> - freeze_queue degrades to no lock in case that request queue becomes dying >>> because blk_enter_queue() won't be blocked any more >>> >>> 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock >>> - it is exclusive lock, so dependency with blk_enter_queue() is covered >>> >>> - it is trylock because blk_mq_freeze_queue() are allowed to run >>> concurrently >>> >>> 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() >>> - nested blk_enter_queue() are allowed >>> >>> - dependency with blk_mq_freeze_queue() is covered >>> >>> - blk_queue_exit() is often called from other contexts(such as irq), and >>> it can't be annotated as lock_release(), so simply do it in >>> blk_enter_queue(), this way still covered cases as many as possible >>> >>> With lockdep support, such kind of reports may be reported asap and >>> needn't wait until the real deadlock is triggered. >>> >>> For example, lockdep report can be triggered in the report[3] with this >>> patch applied. >>> >>> [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' >>> https://bugzilla.kernel.org/show_bug.cgi?id=219166 >>> >>> [2] del_gendisk() vs blk_queue_enter() race condition >>> https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ >>> >>> [3] queue_freeze & queue_enter deadlock in scsi >>> https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u >>> >>> Reviewed-by: Christoph Hellwig <hch@lst.de> >>> Signed-off-by: Ming Lei <ming.lei@redhat.com> >> This patch landed yesterday in linux-next as commit f1be1788a32e >> ("block: model freeze & enter queue as lock for supporting lockdep"). >> In my tests I found that it introduces the following 2 lockdep warnings: >> >> 1. On Samsung Exynos 4412-based Odroid U3 board (ARM 32bit), observed >> when booting it: >> >> ====================================================== >> WARNING: possible circular locking dependency detected >> 6.12.0-rc4-00037-gf1be1788a32e #9290 Not tainted >> ------------------------------------------------------ >> find/1284 is trying to acquire lock: >> cf3b8534 (&mm->mmap_lock){++++}-{3:3}, at: __might_fault+0x30/0x70 >> >> but task is already holding lock: >> c203a0c8 (&sb->s_type->i_mutex_key#2){++++}-{3:3}, at: >> iterate_dir+0x30/0x140 >> >> which lock already depends on the new lock. >> >> >> the existing dependency chain (in reverse order) is: >> >> -> #4 (&sb->s_type->i_mutex_key#2){++++}-{3:3}: >> down_write+0x44/0xc4 >> start_creating+0x8c/0x170 >> debugfs_create_dir+0x1c/0x178 >> blk_register_queue+0xa0/0x1c0 >> add_disk_fwnode+0x210/0x434 >> brd_alloc+0x1cc/0x210 >> brd_init+0xac/0x104 >> do_one_initcall+0x64/0x30c >> kernel_init_freeable+0x1c4/0x228 >> kernel_init+0x1c/0x12c >> ret_from_fork+0x14/0x28 >> >> -> #3 (&q->debugfs_mutex){+.+.}-{3:3}: >> __mutex_lock+0x94/0x94c >> mutex_lock_nested+0x1c/0x24 >> blk_mq_init_sched+0x140/0x204 >> elevator_init_mq+0xb8/0x130 >> add_disk_fwnode+0x3c/0x434 > The above chain can be cut by the following patch because disk state > can be thought as DEAD in add_disk(), can you test it? Seems to be fixing this issue. Feel free to add: Reported-by: Marek Szyprowski <m.szyprowski@samsung.com> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> > diff --git a/block/elevator.c b/block/elevator.c > index 4122026b11f1..efa6ff941a25 100644 > --- a/block/elevator.c > +++ b/block/elevator.c > @@ -600,12 +600,14 @@ void elevator_init_mq(struct request_queue *q) > * requests, then no need to quiesce queue which may add long boot > * latency, especially when lots of disks are involved. > */ > - blk_mq_freeze_queue(q); > + if (__blk_freeze_queue_start(q)) > + blk_freeze_acquire_lock(q, true, false); > blk_mq_cancel_work_sync(q); > > err = blk_mq_init_sched(q, e); > > - blk_mq_unfreeze_queue(q); > + if (__blk_mq_unfreeze_queue(q, false)) > + blk_unfreeze_release_lock(q, true, false); > > if (err) { > pr_warn("\"%s\" elevator initialization failed, " > > ... > >> 2 locks held by find/1284: >> #0: c3df1e88 (&f->f_pos_lock){+.+.}-{3:3}, at: fdget_pos+0x88/0xd0 >> #1: c203a0c8 (&sb->s_type->i_mutex_key#2){++++}-{3:3}, at: >> iterate_dir+0x30/0x140 >> >> stack backtrace: >> CPU: 1 UID: 0 PID: 1284 Comm: find Not tainted >> 6.12.0-rc4-00037-gf1be1788a32e #9290 >> Hardware name: Samsung Exynos (Flattened Device Tree) >> Call trace: >> unwind_backtrace from show_stack+0x10/0x14 >> show_stack from dump_stack_lvl+0x68/0x88 >> dump_stack_lvl from print_circular_bug+0x31c/0x394 >> print_circular_bug from check_noncircular+0x16c/0x184 >> check_noncircular from __lock_acquire+0x158c/0x2970 >> __lock_acquire from lock_acquire+0x130/0x384 >> lock_acquire from __might_fault+0x50/0x70 >> __might_fault from filldir64+0x94/0x28c >> filldir64 from dcache_readdir+0x174/0x260 >> dcache_readdir from iterate_dir+0x64/0x140 >> iterate_dir from sys_getdents64+0x60/0x130 >> sys_getdents64 from ret_fast_syscall+0x0/0x1c >> Exception stack(0xf22b5fa8 to 0xf22b5ff0) >> 5fa0: 004b4fa0 004b4f80 00000004 004b4fa0 00008000 >> 00000000 >> 5fc0: 004b4fa0 004b4f80 00000001 000000d9 00000000 004b4af0 00000000 >> 000010ea >> 5fe0: 004b1eb4 bea05af0 b6da4b08 b6da4a28 >> >> --->8--- >> >> >> 2. On QEMU's ARM64 virt machine, observed during system suspend/resume >> cycle: >> >> # time rtcwake -s10 -mmem >> rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 >> PM: suspend entry (s2idle) >> Filesystems sync: 0.004 seconds >> Freezing user space processes >> Freezing user space processes completed (elapsed 0.007 seconds) >> OOM killer disabled. >> Freezing remaining freezable tasks >> Freezing remaining freezable tasks completed (elapsed 0.004 seconds) >> >> ====================================================== >> WARNING: possible circular locking dependency detected >> 6.12.0-rc4+ #9291 Not tainted >> ------------------------------------------------------ >> rtcwake/1299 is trying to acquire lock: >> ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 >> >> but task is already holding lock: >> ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: >> virtblk_freeze+0x24/0x60 >> >> which lock already depends on the new lock. >> >> >> the existing dependency chain (in reverse order) is: > This one looks a real thing, at least the added lockdep code works as > expected, also the blk_mq_freeze_queue() use in virtio-blk's ->suspend() > is questionable. I will take a further look. > > > Thanks, > Ming > > Best regards
Hi Ming,
Greetings!
I used Syzkaller and found that there is possible deadlock in __submit_bio in linux-next next-20241029.
After bisection and the first bad commit is:
"
f1be1788a32e block: model freeze & enter queue as lock for supporting lockdep
"
All detailed into can be found at:
https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio
Syzkaller repro code:
https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.c
Syzkaller repro syscall steps:
https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.prog
Syzkaller report:
https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.report
Kconfig(make olddefconfig):
https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/kconfig_origin
Bisect info:
https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/bisect_info.log
bzImage:
https://github.com/laifryiee/syzkaller_logs/raw/refs/heads/main/241029_183511___submit_bio/bzImage_6fb2fa9805c501d9ade047fc511961f3273cdcb5
Issue dmesg:
https://github.com/laifryiee/syzkaller_logs/blob/main/241029_183511___submit_bio/6fb2fa9805c501d9ade047fc511961f3273cdcb5_dmesg.log
"
[ 22.219103] 6.12.0-rc5-next-20241029-6fb2fa9805c5 #1 Not tainted
[ 22.219512] ------------------------------------------------------
[ 22.219827] repro/735 is trying to acquire lock:
[ 22.220066] ffff888010f1a768 (&q->q_usage_counter(io)#25){++++}-{0:0}, at: __submit_bio+0x39f/0x550
[ 22.220568]
[ 22.220568] but task is already holding lock:
[ 22.220884] ffffffff872322a0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x76b/0x21e0
[ 22.221453]
[ 22.221453] which lock already depends on the new lock.
[ 22.221453]
[ 22.221862]
[ 22.221862] the existing dependency chain (in reverse order) is:
[ 22.222247]
[ 22.222247] -> #1 (fs_reclaim){+.+.}-{0:0}:
[ 22.222630] lock_acquire+0x80/0xb0
[ 22.222920] fs_reclaim_acquire+0x116/0x160
[ 22.223244] __kmalloc_cache_node_noprof+0x59/0x470
[ 22.223528] blk_mq_init_tags+0x79/0x1a0
[ 22.223771] blk_mq_alloc_map_and_rqs+0x1f4/0xdd0
[ 22.224127] blk_mq_init_sched+0x33d/0x6d0
[ 22.224376] elevator_init_mq+0x2b2/0x400
[ 22.224619] add_disk_fwnode+0x11c/0x1300
[ 22.224866] device_add_disk+0x31/0x40
[ 22.225097] sd_probe+0xa79/0x1080
[ 22.225324] really_probe+0x27c/0xac0
[ 22.225556] __driver_probe_device+0x1f3/0x460
[ 22.225819] driver_probe_device+0x56/0x1b0
[ 22.226069] __device_attach_driver+0x1e7/0x300
[ 22.226336] bus_for_each_drv+0x159/0x1e0
[ 22.226576] __device_attach_async_helper+0x1e4/0x2a0
[ 22.226871] async_run_entry_fn+0xa3/0x450
[ 22.227127] process_one_work+0x92e/0x1b50
[ 22.227380] worker_thread+0x68d/0xe90
[ 22.227614] kthread+0x35a/0x470
[ 22.227834] ret_from_fork+0x56/0x90
[ 22.228119] ret_from_fork_asm+0x1a/0x30
[ 22.228365]
[ 22.228365] -> #0 (&q->q_usage_counter(io)#25){++++}-{0:0}:
[ 22.228739] __lock_acquire+0x2ff8/0x5d60
[ 22.228982] lock_acquire.part.0+0x142/0x390
[ 22.229237] lock_acquire+0x80/0xb0
[ 22.229452] blk_mq_submit_bio+0x1cbe/0x2590
[ 22.229708] __submit_bio+0x39f/0x550
[ 22.229931] submit_bio_noacct_nocheck+0x647/0xcc0
[ 22.230212] submit_bio_noacct+0x620/0x1e00
[ 22.230463] submit_bio+0xce/0x480
[ 22.230677] __swap_writepage+0x2f1/0xdf0
[ 22.230923] swap_writepage+0x464/0xbc0
[ 22.231157] shmem_writepage+0xdeb/0x1340
[ 22.231406] pageout+0x3bc/0x9b0
[ 22.231617] shrink_folio_list+0x16b9/0x3b60
[ 22.231884] shrink_lruvec+0xd78/0x2790
[ 22.232220] shrink_node+0xb29/0x2870
[ 22.232447] do_try_to_free_pages+0x2e3/0x1790
[ 22.232789] try_to_free_pages+0x2bc/0x750
[ 22.233065] __alloc_pages_slowpath.constprop.0+0x812/0x21e0
[ 22.233528] __alloc_pages_noprof+0x5d4/0x6f0
[ 22.233861] alloc_pages_mpol_noprof+0x30a/0x580
[ 22.234233] alloc_pages_noprof+0xa9/0x180
[ 22.234555] kimage_alloc_pages+0x79/0x240
[ 22.234808] kimage_alloc_control_pages+0x1cb/0xa60
[ 22.235119] do_kexec_load+0x3a6/0x8e0
[ 22.235418] __x64_sys_kexec_load+0x1cc/0x240
[ 22.235759] x64_sys_call+0xf0f/0x20d0
[ 22.236058] do_syscall_64+0x6d/0x140
[ 22.236343] entry_SYSCALL_64_after_hwframe+0x76/0x7e
[ 22.236643]
[ 22.236643] other info that might help us debug this:
[ 22.236643]
[ 22.237042] Possible unsafe locking scenario:
[ 22.237042]
[ 22.237359] CPU0 CPU1
[ 22.237662] ---- ----
[ 22.237926] lock(fs_reclaim);
[ 22.238098] lock(&q->q_usage_counter(io)#25);
[ 22.238535] lock(fs_reclaim);
[ 22.238935] rlock(&q->q_usage_counter(io)#25);
[ 22.239182]
[ 22.239182] *** DEADLOCK ***
[ 22.239182]
[ 22.239521] 1 lock held by repro/735:
[ 22.239778] #0: ffffffff872322a0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x76b/0x21e0
[ 22.240320]
[ 22.240320] stack backtrace:
[ 22.240550] CPU: 1 UID: 0 PID: 735 Comm: repro Not tainted 6.12.0-rc5-next-20241029-6fb2fa9805c5 #1
[ 22.241079] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
[ 22.241780] Call Trace:
[ 22.241928] <TASK>
[ 22.242095] dump_stack_lvl+0xea/0x150
[ 22.242351] dump_stack+0x19/0x20
[ 22.242534] print_circular_bug+0x47f/0x750
[ 22.242761] check_noncircular+0x2f4/0x3e0
[ 22.242992] ? __pfx_check_noncircular+0x10/0x10
[ 22.243302] ? __pfx_lock_release+0x10/0x10
[ 22.243586] ? lockdep_lock+0xd0/0x1d0
[ 22.243865] ? __pfx_lockdep_lock+0x10/0x10
[ 22.244103] __lock_acquire+0x2ff8/0x5d60
[ 22.244382] ? __kernel_text_address+0x16/0x50
[ 22.244633] ? __pfx___lock_acquire+0x10/0x10
[ 22.244871] ? stack_trace_save+0x97/0xd0
[ 22.245102] lock_acquire.part.0+0x142/0x390
[ 22.245394] ? __submit_bio+0x39f/0x550
[ 22.245646] ? __pfx_lock_acquire.part.0+0x10/0x10
"
I hope you find it useful.
Regards,
Yi Lai
---
If you don't need the following environment to reproduce the problem or if you
already have one reproduced environment, please ignore the following information.
How to reproduce:
git clone https://gitlab.com/xupengfe/repro_vm_env.git
cd repro_vm_env
tar -xvf repro_vm_env.tar.gz
cd repro_vm_env; ./start3.sh // it needs qemu-system-x86_64 and I used v7.1.0
// start3.sh will load bzImage_2241ab53cbb5cdb08a6b2d4688feb13971058f65 v6.2-rc5 kernel
// You could change the bzImage_xxx as you want
// Maybe you need to remove line "-drive if=pflash,format=raw,readonly=on,file=./OVMF_CODE.fd \" for different qemu version
You could use below command to log in, there is no password for root.
ssh -p 10023 root@localhost
After login vm(virtual machine) successfully, you could transfer reproduced
binary to the vm by below way, and reproduce the problem in vm:
gcc -pthread -o repro repro.c
scp -P 10023 repro root@localhost:/root/
Get the bzImage for target kernel:
Please use target kconfig and copy it to kernel_src/.config
make olddefconfig
make -jx bzImage //x should equal or less than cpu num your pc has
Fill the bzImage file into above start3.sh to load the target kernel in vm.
Tips:
If you already have qemu-system-x86_64, please ignore below info.
If you want to install qemu v7.1.0 version:
git clone https://github.com/qemu/qemu.git
cd qemu
git checkout -f v7.1.0
mkdir build
cd build
yum install -y ninja-build.x86_64
yum -y install libslirp-devel.x86_64
../configure --target-list=x86_64-softmmu --enable-kvm --enable-vnc --enable-gtk --enable-sdl --enable-usb-redir --enable-slirp
make
make install
On Fri, Oct 25, 2024 at 08:37:20AM +0800, Ming Lei wrote:
> Recently we got several deadlock report[1][2][3] caused by
> blk_mq_freeze_queue and blk_enter_queue().
>
> Turns out the two are just like acquiring read/write lock, so model them
> as read/write lock for supporting lockdep:
>
> 1) model q->q_usage_counter as two locks(io and queue lock)
>
> - queue lock covers sync with blk_enter_queue()
>
> - io lock covers sync with bio_enter_queue()
>
> 2) make the lockdep class/key as per-queue:
>
> - different subsystem has very different lock use pattern, shared lock
> class causes false positive easily
>
> - freeze_queue degrades to no lock in case that disk state becomes DEAD
> because bio_enter_queue() won't be blocked any more
>
> - freeze_queue degrades to no lock in case that request queue becomes dying
> because blk_enter_queue() won't be blocked any more
>
> 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock
> - it is exclusive lock, so dependency with blk_enter_queue() is covered
>
> - it is trylock because blk_mq_freeze_queue() are allowed to run
> concurrently
>
> 4) model blk_enter_queue() & bio_enter_queue() as acquire_read()
> - nested blk_enter_queue() are allowed
>
> - dependency with blk_mq_freeze_queue() is covered
>
> - blk_queue_exit() is often called from other contexts(such as irq), and
> it can't be annotated as lock_release(), so simply do it in
> blk_enter_queue(), this way still covered cases as many as possible
>
> With lockdep support, such kind of reports may be reported asap and
> needn't wait until the real deadlock is triggered.
>
> For example, lockdep report can be triggered in the report[3] with this
> patch applied.
>
> [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler'
> https://bugzilla.kernel.org/show_bug.cgi?id=219166
>
> [2] del_gendisk() vs blk_queue_enter() race condition
> https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/
>
> [3] queue_freeze & queue_enter deadlock in scsi
> https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u
>
> Reviewed-by: Christoph Hellwig <hch@lst.de>
> Signed-off-by: Ming Lei <ming.lei@redhat.com>
> ---
> block/blk-core.c | 18 ++++++++++++++++--
> block/blk-mq.c | 26 ++++++++++++++++++++++----
> block/blk.h | 29 ++++++++++++++++++++++++++---
> block/genhd.c | 15 +++++++++++----
> include/linux/blkdev.h | 6 ++++++
> 5 files changed, 81 insertions(+), 13 deletions(-)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index bc5e8c5eaac9..09d10bb95fda 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -261,6 +261,8 @@ static void blk_free_queue(struct request_queue *q)
> blk_mq_release(q);
>
> ida_free(&blk_queue_ida, q->id);
> + lockdep_unregister_key(&q->io_lock_cls_key);
> + lockdep_unregister_key(&q->q_lock_cls_key);
> call_rcu(&q->rcu_head, blk_free_queue_rcu);
> }
>
> @@ -278,18 +280,20 @@ void blk_put_queue(struct request_queue *q)
> }
> EXPORT_SYMBOL(blk_put_queue);
>
> -void blk_queue_start_drain(struct request_queue *q)
> +bool blk_queue_start_drain(struct request_queue *q)
> {
> /*
> * When queue DYING flag is set, we need to block new req
> * entering queue, so we call blk_freeze_queue_start() to
> * prevent I/O from crossing blk_queue_enter().
> */
> - blk_freeze_queue_start(q);
> + bool freeze = __blk_freeze_queue_start(q);
> if (queue_is_mq(q))
> blk_mq_wake_waiters(q);
> /* Make blk_queue_enter() reexamine the DYING flag. */
> wake_up_all(&q->mq_freeze_wq);
> +
> + return freeze;
> }
>
> /**
> @@ -321,6 +325,8 @@ int blk_queue_enter(struct request_queue *q, blk_mq_req_flags_t flags)
> return -ENODEV;
> }
>
> + rwsem_acquire_read(&q->q_lockdep_map, 0, 0, _RET_IP_);
> + rwsem_release(&q->q_lockdep_map, _RET_IP_);
> return 0;
> }
>
> @@ -352,6 +358,8 @@ int __bio_queue_enter(struct request_queue *q, struct bio *bio)
> goto dead;
> }
>
> + rwsem_acquire_read(&q->io_lockdep_map, 0, 0, _RET_IP_);
> + rwsem_release(&q->io_lockdep_map, _RET_IP_);
> return 0;
> dead:
> bio_io_error(bio);
> @@ -441,6 +449,12 @@ struct request_queue *blk_alloc_queue(struct queue_limits *lim, int node_id)
> PERCPU_REF_INIT_ATOMIC, GFP_KERNEL);
> if (error)
> goto fail_stats;
> + lockdep_register_key(&q->io_lock_cls_key);
> + lockdep_register_key(&q->q_lock_cls_key);
> + lockdep_init_map(&q->io_lockdep_map, "&q->q_usage_counter(io)",
> + &q->io_lock_cls_key, 0);
> + lockdep_init_map(&q->q_lockdep_map, "&q->q_usage_counter(queue)",
> + &q->q_lock_cls_key, 0);
>
> q->nr_requests = BLKDEV_DEFAULT_RQ;
>
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index 96858fb3b9ff..76f277a30c11 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -120,17 +120,29 @@ void blk_mq_in_flight_rw(struct request_queue *q, struct block_device *part,
> inflight[1] = mi.inflight[1];
> }
>
> -void blk_freeze_queue_start(struct request_queue *q)
> +bool __blk_freeze_queue_start(struct request_queue *q)
> {
> + int freeze;
> +
> mutex_lock(&q->mq_freeze_lock);
> if (++q->mq_freeze_depth == 1) {
> percpu_ref_kill(&q->q_usage_counter);
> mutex_unlock(&q->mq_freeze_lock);
> if (queue_is_mq(q))
> blk_mq_run_hw_queues(q, false);
> + freeze = true;
> } else {
> mutex_unlock(&q->mq_freeze_lock);
> + freeze = false;
> }
> +
> + return freeze;
> +}
> +
> +void blk_freeze_queue_start(struct request_queue *q)
> +{
> + if (__blk_freeze_queue_start(q))
> + blk_freeze_acquire_lock(q, false, false);
> }
> EXPORT_SYMBOL_GPL(blk_freeze_queue_start);
>
> @@ -176,8 +188,10 @@ void blk_mq_freeze_queue(struct request_queue *q)
> }
> EXPORT_SYMBOL_GPL(blk_mq_freeze_queue);
>
> -void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic)
> +bool __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic)
> {
> + int unfreeze = false;
> +
> mutex_lock(&q->mq_freeze_lock);
> if (force_atomic)
> q->q_usage_counter.data->force_atomic = true;
> @@ -186,13 +200,17 @@ void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic)
> if (!q->mq_freeze_depth) {
> percpu_ref_resurrect(&q->q_usage_counter);
> wake_up_all(&q->mq_freeze_wq);
> + unfreeze = true;
> }
> mutex_unlock(&q->mq_freeze_lock);
> +
> + return unfreeze;
> }
>
> void blk_mq_unfreeze_queue(struct request_queue *q)
> {
> - __blk_mq_unfreeze_queue(q, false);
> + if (__blk_mq_unfreeze_queue(q, false))
> + blk_unfreeze_release_lock(q, false, false);
> }
> EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue);
>
> @@ -205,7 +223,7 @@ EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue);
> */
> void blk_freeze_queue_start_non_owner(struct request_queue *q)
> {
> - blk_freeze_queue_start(q);
> + __blk_freeze_queue_start(q);
> }
> EXPORT_SYMBOL_GPL(blk_freeze_queue_start_non_owner);
>
> diff --git a/block/blk.h b/block/blk.h
> index c718e4291db0..832e54c5a271 100644
> --- a/block/blk.h
> +++ b/block/blk.h
> @@ -4,6 +4,7 @@
>
> #include <linux/bio-integrity.h>
> #include <linux/blk-crypto.h>
> +#include <linux/lockdep.h>
> #include <linux/memblock.h> /* for max_pfn/max_low_pfn */
> #include <linux/sched/sysctl.h>
> #include <linux/timekeeping.h>
> @@ -35,8 +36,9 @@ struct blk_flush_queue *blk_alloc_flush_queue(int node, int cmd_size,
> void blk_free_flush_queue(struct blk_flush_queue *q);
>
> void blk_freeze_queue(struct request_queue *q);
> -void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
> -void blk_queue_start_drain(struct request_queue *q);
> +bool __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
> +bool blk_queue_start_drain(struct request_queue *q);
> +bool __blk_freeze_queue_start(struct request_queue *q);
> int __bio_queue_enter(struct request_queue *q, struct bio *bio);
> void submit_bio_noacct_nocheck(struct bio *bio);
> void bio_await_chain(struct bio *bio);
> @@ -69,8 +71,11 @@ static inline int bio_queue_enter(struct bio *bio)
> {
> struct request_queue *q = bdev_get_queue(bio->bi_bdev);
>
> - if (blk_try_enter_queue(q, false))
> + if (blk_try_enter_queue(q, false)) {
> + rwsem_acquire_read(&q->io_lockdep_map, 0, 0, _RET_IP_);
> + rwsem_release(&q->io_lockdep_map, _RET_IP_);
> return 0;
> + }
> return __bio_queue_enter(q, bio);
> }
>
> @@ -734,4 +739,22 @@ void blk_integrity_verify(struct bio *bio);
> void blk_integrity_prepare(struct request *rq);
> void blk_integrity_complete(struct request *rq, unsigned int nr_bytes);
>
> +static inline void blk_freeze_acquire_lock(struct request_queue *q, bool
> + disk_dead, bool queue_dying)
> +{
> + if (!disk_dead)
> + rwsem_acquire(&q->io_lockdep_map, 0, 1, _RET_IP_);
> + if (!queue_dying)
> + rwsem_acquire(&q->q_lockdep_map, 0, 1, _RET_IP_);
> +}
> +
> +static inline void blk_unfreeze_release_lock(struct request_queue *q, bool
> + disk_dead, bool queue_dying)
> +{
> + if (!queue_dying)
> + rwsem_release(&q->q_lockdep_map, _RET_IP_);
> + if (!disk_dead)
> + rwsem_release(&q->io_lockdep_map, _RET_IP_);
> +}
> +
> #endif /* BLK_INTERNAL_H */
> diff --git a/block/genhd.c b/block/genhd.c
> index 1c05dd4c6980..6ad3fcde0110 100644
> --- a/block/genhd.c
> +++ b/block/genhd.c
> @@ -581,13 +581,13 @@ static void blk_report_disk_dead(struct gendisk *disk, bool surprise)
> rcu_read_unlock();
> }
>
> -static void __blk_mark_disk_dead(struct gendisk *disk)
> +static bool __blk_mark_disk_dead(struct gendisk *disk)
> {
> /*
> * Fail any new I/O.
> */
> if (test_and_set_bit(GD_DEAD, &disk->state))
> - return;
> + return false;
>
> if (test_bit(GD_OWNS_QUEUE, &disk->state))
> blk_queue_flag_set(QUEUE_FLAG_DYING, disk->queue);
> @@ -600,7 +600,7 @@ static void __blk_mark_disk_dead(struct gendisk *disk)
> /*
> * Prevent new I/O from crossing bio_queue_enter().
> */
> - blk_queue_start_drain(disk->queue);
> + return blk_queue_start_drain(disk->queue);
> }
>
> /**
> @@ -641,6 +641,7 @@ void del_gendisk(struct gendisk *disk)
> struct request_queue *q = disk->queue;
> struct block_device *part;
> unsigned long idx;
> + bool start_drain, queue_dying;
>
> might_sleep();
>
> @@ -668,7 +669,10 @@ void del_gendisk(struct gendisk *disk)
> * Drop all partitions now that the disk is marked dead.
> */
> mutex_lock(&disk->open_mutex);
> - __blk_mark_disk_dead(disk);
> + start_drain = __blk_mark_disk_dead(disk);
> + queue_dying = blk_queue_dying(q);
> + if (start_drain)
> + blk_freeze_acquire_lock(q, true, queue_dying);
> xa_for_each_start(&disk->part_tbl, idx, part, 1)
> drop_partition(part);
> mutex_unlock(&disk->open_mutex);
> @@ -725,6 +729,9 @@ void del_gendisk(struct gendisk *disk)
> if (queue_is_mq(q))
> blk_mq_exit_queue(q);
> }
> +
> + if (start_drain)
> + blk_unfreeze_release_lock(q, true, queue_dying);
> }
> EXPORT_SYMBOL(del_gendisk);
>
> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> index 50c3b959da28..57f1ee386b57 100644
> --- a/include/linux/blkdev.h
> +++ b/include/linux/blkdev.h
> @@ -25,6 +25,7 @@
> #include <linux/uuid.h>
> #include <linux/xarray.h>
> #include <linux/file.h>
> +#include <linux/lockdep.h>
>
> struct module;
> struct request_queue;
> @@ -471,6 +472,11 @@ struct request_queue {
> struct xarray hctx_table;
>
> struct percpu_ref q_usage_counter;
> + struct lock_class_key io_lock_cls_key;
> + struct lockdep_map io_lockdep_map;
> +
> + struct lock_class_key q_lock_cls_key;
> + struct lockdep_map q_lockdep_map;
>
> struct request *last_merge;
>
> --
> 2.46.0
>
On Wed, Oct 30, 2024 at 02:45:03PM +0800, Lai, Yi wrote: > Hi Ming, > > Greetings! > > I used Syzkaller and found that there is possible deadlock in __submit_bio in linux-next next-20241029. > > After bisection and the first bad commit is: > " > f1be1788a32e block: model freeze & enter queue as lock for supporting lockdep > " > > All detailed into can be found at: > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio > Syzkaller repro code: > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.c > Syzkaller repro syscall steps: > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.prog > Syzkaller report: > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.report > Kconfig(make olddefconfig): > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/kconfig_origin > Bisect info: > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/bisect_info.log > bzImage: > https://github.com/laifryiee/syzkaller_logs/raw/refs/heads/main/241029_183511___submit_bio/bzImage_6fb2fa9805c501d9ade047fc511961f3273cdcb5 > Issue dmesg: > https://github.com/laifryiee/syzkaller_logs/blob/main/241029_183511___submit_bio/6fb2fa9805c501d9ade047fc511961f3273cdcb5_dmesg.log > > " > [ 22.219103] 6.12.0-rc5-next-20241029-6fb2fa9805c5 #1 Not tainted > [ 22.219512] ------------------------------------------------------ > [ 22.219827] repro/735 is trying to acquire lock: > [ 22.220066] ffff888010f1a768 (&q->q_usage_counter(io)#25){++++}-{0:0}, at: __submit_bio+0x39f/0x550 > [ 22.220568] > [ 22.220568] but task is already holding lock: > [ 22.220884] ffffffff872322a0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x76b/0x21e0 > [ 22.221453] > [ 22.221453] which lock already depends on the new lock. > [ 22.221453] > [ 22.221862] > [ 22.221862] the existing dependency chain (in reverse order) is: > [ 22.222247] > [ 22.222247] -> #1 (fs_reclaim){+.+.}-{0:0}: > [ 22.222630] lock_acquire+0x80/0xb0 > [ 22.222920] fs_reclaim_acquire+0x116/0x160 > [ 22.223244] __kmalloc_cache_node_noprof+0x59/0x470 > [ 22.223528] blk_mq_init_tags+0x79/0x1a0 > [ 22.223771] blk_mq_alloc_map_and_rqs+0x1f4/0xdd0 > [ 22.224127] blk_mq_init_sched+0x33d/0x6d0 > [ 22.224376] elevator_init_mq+0x2b2/0x400 It should be addressed by the following patch: https://lore.kernel.org/linux-block/ZyEGLdg744U_xBjp@fedora/ Thanks, Ming
On Wed, Oct 30, 2024 at 03:13:09PM +0800, Ming Lei wrote: > On Wed, Oct 30, 2024 at 02:45:03PM +0800, Lai, Yi wrote: > > Hi Ming, > > > > Greetings! > > > > I used Syzkaller and found that there is possible deadlock in __submit_bio in linux-next next-20241029. > > > > After bisection and the first bad commit is: > > " > > f1be1788a32e block: model freeze & enter queue as lock for supporting lockdep > > " > > > > All detailed into can be found at: > > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio > > Syzkaller repro code: > > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.c > > Syzkaller repro syscall steps: > > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.prog > > Syzkaller report: > > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/repro.report > > Kconfig(make olddefconfig): > > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/kconfig_origin > > Bisect info: > > https://github.com/laifryiee/syzkaller_logs/tree/main/241029_183511___submit_bio/bisect_info.log > > bzImage: > > https://github.com/laifryiee/syzkaller_logs/raw/refs/heads/main/241029_183511___submit_bio/bzImage_6fb2fa9805c501d9ade047fc511961f3273cdcb5 > > Issue dmesg: > > https://github.com/laifryiee/syzkaller_logs/blob/main/241029_183511___submit_bio/6fb2fa9805c501d9ade047fc511961f3273cdcb5_dmesg.log > > > > " > > [ 22.219103] 6.12.0-rc5-next-20241029-6fb2fa9805c5 #1 Not tainted > > [ 22.219512] ------------------------------------------------------ > > [ 22.219827] repro/735 is trying to acquire lock: > > [ 22.220066] ffff888010f1a768 (&q->q_usage_counter(io)#25){++++}-{0:0}, at: __submit_bio+0x39f/0x550 > > [ 22.220568] > > [ 22.220568] but task is already holding lock: > > [ 22.220884] ffffffff872322a0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x76b/0x21e0 > > [ 22.221453] > > [ 22.221453] which lock already depends on the new lock. > > [ 22.221453] > > [ 22.221862] > > [ 22.221862] the existing dependency chain (in reverse order) is: > > [ 22.222247] > > [ 22.222247] -> #1 (fs_reclaim){+.+.}-{0:0}: > > [ 22.222630] lock_acquire+0x80/0xb0 > > [ 22.222920] fs_reclaim_acquire+0x116/0x160 > > [ 22.223244] __kmalloc_cache_node_noprof+0x59/0x470 > > [ 22.223528] blk_mq_init_tags+0x79/0x1a0 > > [ 22.223771] blk_mq_alloc_map_and_rqs+0x1f4/0xdd0 > > [ 22.224127] blk_mq_init_sched+0x33d/0x6d0 > > [ 22.224376] elevator_init_mq+0x2b2/0x400 > > It should be addressed by the following patch: > > https://lore.kernel.org/linux-block/ZyEGLdg744U_xBjp@fedora/ > I have applied proposed fix patch on top of next-20241029. Issue can still be reproduced. It seems the dependency chain is different from Marek's log and mine. > Thanks, > Ming >
On Wed, Oct 30, 2024 at 4:51 PM Lai, Yi <yi1.lai@linux.intel.com> wrote: > > On Wed, Oct 30, 2024 at 03:13:09PM +0800, Ming Lei wrote: > > On Wed, Oct 30, 2024 at 02:45:03PM +0800, Lai, Yi wrote: ... > > > > It should be addressed by the following patch: > > > > https://lore.kernel.org/linux-block/ZyEGLdg744U_xBjp@fedora/ > > > > I have applied proposed fix patch on top of next-20241029. Issue can > still be reproduced. > > It seems the dependency chain is different from Marek's log and mine. Can you post the new log since q->q_usage_counter(io)->fs_reclaim from blk_mq_init_sched is cut down by the patch? Thanks, Ming
On Wed, Oct 30, 2024 at 05:50:15PM +0800, Ming Lei wrote: > On Wed, Oct 30, 2024 at 4:51 PM Lai, Yi <yi1.lai@linux.intel.com> wrote: > > > > On Wed, Oct 30, 2024 at 03:13:09PM +0800, Ming Lei wrote: > > > On Wed, Oct 30, 2024 at 02:45:03PM +0800, Lai, Yi wrote: > ... > > > > > > It should be addressed by the following patch: > > > > > > https://lore.kernel.org/linux-block/ZyEGLdg744U_xBjp@fedora/ > > > > > > > I have applied proposed fix patch on top of next-20241029. Issue can > > still be reproduced. > > > > It seems the dependency chain is different from Marek's log and mine. > > Can you post the new log since q->q_usage_counter(io)->fs_reclaim from > blk_mq_init_sched is cut down by the patch? > New possible deadlock log after patch applied: [ 52.485023] repro: page allocation failure: order:1, mode:0x10cc0(GFP_KERNEL|__GFP_NORETRY), nodemask=(null),cpuset=/,mems0 [ 52.486074] CPU: 1 UID: 0 PID: 635 Comm: repro Not tainted 6.12.0-rc5-next-20241029-kvm-dirty #6 [ 52.486752] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/014 [ 52.487616] Call Trace: [ 52.487820] <TASK> [ 52.488001] dump_stack_lvl+0x121/0x150 [ 52.488345] dump_stack+0x19/0x20 [ 52.488616] warn_alloc+0x218/0x350 [ 52.488913] ? __pfx_warn_alloc+0x10/0x10 [ 52.489263] ? __alloc_pages_direct_compact+0x130/0xa10 [ 52.489699] ? __pfx___alloc_pages_direct_compact+0x10/0x10 [ 52.490151] ? __drain_all_pages+0x27b/0x480 [ 52.490522] __alloc_pages_slowpath.constprop.0+0x14d6/0x21e0 [ 52.491018] ? __pfx___alloc_pages_slowpath.constprop.0+0x10/0x10 [ 52.491519] ? lock_is_held_type+0xef/0x150 [ 52.491875] ? __pfx_get_page_from_freelist+0x10/0x10 [ 52.492291] ? lock_acquire+0x80/0xb0 [ 52.492619] __alloc_pages_noprof+0x5d4/0x6f0 [ 52.492992] ? __pfx___alloc_pages_noprof+0x10/0x10 [ 52.493405] ? __sanitizer_cov_trace_switch+0x58/0xa0 [ 52.493830] ? policy_nodemask+0xf9/0x450 [ 52.494169] alloc_pages_mpol_noprof+0x30a/0x580 [ 52.494561] ? __pfx_alloc_pages_mpol_noprof+0x10/0x10 [ 52.494982] ? sysvec_apic_timer_interrupt+0x6a/0xd0 [ 52.495396] ? asm_sysvec_apic_timer_interrupt+0x1f/0x30 [ 52.495845] alloc_pages_noprof+0xa9/0x180 [ 52.496201] kimage_alloc_pages+0x79/0x240 [ 52.496558] kimage_alloc_control_pages+0x1cb/0xa60 [ 52.496982] ? __pfx_kimage_alloc_control_pages+0x10/0x10 [ 52.497437] ? __sanitizer_cov_trace_const_cmp1+0x1e/0x30 [ 52.497897] do_kexec_load+0x3a6/0x8e0 [ 52.498228] ? __pfx_do_kexec_load+0x10/0x10 [ 52.498593] ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30 [ 52.499035] ? _copy_from_user+0xb6/0xf0 [ 52.499371] __x64_sys_kexec_load+0x1cc/0x240 [ 52.499740] x64_sys_call+0xf0f/0x20d0 [ 52.500055] do_syscall_64+0x6d/0x140 [ 52.500367] entry_SYSCALL_64_after_hwframe+0x76/0x7e [ 52.500778] RIP: 0033:0x7f310423ee5d [ 52.501077] Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c88 [ 52.502494] RSP: 002b:00007fffcecca558 EFLAGS: 00000207 ORIG_RAX: 00000000000000f6 [ 52.503087] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f310423ee5d [ 52.503644] RDX: 0000000020000040 RSI: 0000000000000009 RDI: 0000000000000000 [ 52.504198] RBP: 00007fffcecca560 R08: 00007fffcecc9fd0 R09: 00007fffcecca590 [ 52.504767] R10: 0000000000000000 R11: 0000000000000207 R12: 00007fffcecca6d8 [ 52.505345] R13: 0000000000401c72 R14: 0000000000403e08 R15: 00007f3104469000 [ 52.505949] </TASK> [ 52.506239] Mem-Info: [ 52.506449] active_anon:119 inactive_anon:14010 isolated_anon:0 [ 52.506449] active_file:17895 inactive_file:87 isolated_file:0 [ 52.506449] unevictable:0 dirty:15 writeback:0 [ 52.506449] slab_reclaimable:6957 slab_unreclaimable:20220 [ 52.506449] mapped:11598 shmem:1150 pagetables:766 [ 52.506449] sec_pagetables:0 bounce:0 [ 52.506449] kernel_misc_reclaimable:0 [ 52.506449] free:13776 free_pcp:99 free_cma:0 [ 52.509456] Node 0 active_anon:476kB inactive_anon:56040kB active_file:71580kB inactive_file:348kB unevictable:0kB isolateo [ 52.511881] Node 0 DMA free:440kB boost:0kB min:440kB low:548kB high:656kB reserved_highatomic:0KB active_anon:0kB inactivB [ 52.513883] lowmem_reserve[]: 0 1507 0 0 0 [ 52.514269] Node 0 DMA32 free:54664kB boost:0kB min:44612kB low:55764kB high:66916kB reserved_highatomic:0KB active_anon:4B [ 52.516485] lowmem_reserve[]: 0 0 0 0 0 [ 52.516831] Node 0 DMA: 0*4kB 1*8kB (U) 1*16kB (U) 1*32kB (U) 0*64kB 1*128kB (U) 1*256kB (U) 0*512kB 0*1024kB 0*2048kB 0*4B [ 52.517895] Node 0 DMA32: 2970*4kB (UME) 1123*8kB (UME) 532*16kB (UME) 280*32kB (UM) 126*64kB (UM) 27*128kB (UME) 9*256kB B [ 52.519279] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB [ 52.519387] [ 52.519971] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB [ 52.520138] ====================================================== [ 52.520805] 19113 total pagecache pages [ 52.521702] WARNING: possible circular locking dependency detected [ 52.522016] 0 pages in swap cache [ 52.522022] Free swap = 124984kB [ 52.522027] Total swap = 124996kB [ 52.523720] 6.12.0-rc5-next-20241029-kvm-dirty #6 Not tainted [ 52.523741] ------------------------------------------------------ [ 52.524059] 524158 pages RAM [ 52.525050] kswapd0/56 is trying to acquire lock: [ 52.525452] 0 pages HighMem/MovableOnly [ 52.525461] 129765 pages reserved [ 52.525465] 0 pages cma reserved [ 52.525469] 0 pages hwpoisoned [ 52.527163] ffff8880104374e8 (&q->q_usage_counter(io)#25){++++}-{0:0}, at: __submit_bio+0x39f/0x550 [ 52.532396] [ 52.532396] but task is already holding lock: [ 52.533293] ffffffff872322a0 (fs_reclaim){+.+.}-{0:0}, at: balance_pgdat+0xb0f/0x1500 [ 52.534508] [ 52.534508] which lock already depends on the new lock. [ 52.534508] [ 52.535723] [ 52.535723] the existing dependency chain (in reverse order) is: [ 52.536818] [ 52.536818] -> #2 (fs_reclaim){+.+.}-{0:0}: [ 52.537705] lock_acquire+0x80/0xb0 [ 52.538337] fs_reclaim_acquire+0x116/0x160 [ 52.539076] blk_mq_alloc_and_init_hctx+0x4df/0x1200 [ 52.539906] blk_mq_realloc_hw_ctxs+0x4cf/0x610 [ 52.540676] blk_mq_init_allocated_queue+0x3da/0x11b0 [ 52.541547] blk_mq_alloc_queue+0x22c/0x300 [ 52.542279] __blk_mq_alloc_disk+0x34/0x100 [ 52.543011] loop_add+0x4c9/0xbd0 [ 52.543622] loop_init+0x133/0x1a0 [ 52.544248] do_one_initcall+0x114/0x5d0 [ 52.544954] kernel_init_freeable+0xab0/0xeb0 [ 52.545732] kernel_init+0x28/0x2f0 [ 52.546366] ret_from_fork+0x56/0x90 [ 52.547009] ret_from_fork_asm+0x1a/0x30 [ 52.547698] [ 52.547698] -> #1 (&q->sysfs_lock){+.+.}-{4:4}: [ 52.548625] lock_acquire+0x80/0xb0 [ 52.549276] __mutex_lock+0x17c/0x1540 [ 52.549958] mutex_lock_nested+0x1f/0x30 [ 52.550664] queue_attr_store+0xea/0x180 [ 52.551360] sysfs_kf_write+0x11f/0x180 [ 52.552036] kernfs_fop_write_iter+0x40e/0x630 [ 52.552808] vfs_write+0xc59/0x1140 [ 52.553446] ksys_write+0x14f/0x290 [ 52.554068] __x64_sys_write+0x7b/0xc0 [ 52.554728] x64_sys_call+0x1685/0x20d0 [ 52.555397] do_syscall_64+0x6d/0x140 [ 52.556029] entry_SYSCALL_64_after_hwframe+0x76/0x7e [ 52.556865] [ 52.556865] -> #0 (&q->q_usage_counter(io)#25){++++}-{0:0}: [ 52.557963] __lock_acquire+0x2ff8/0x5d60 [ 52.558667] lock_acquire.part.0+0x142/0x390 [ 52.559427] lock_acquire+0x80/0xb0 [ 52.560057] blk_mq_submit_bio+0x1cbe/0x2590 [ 52.560801] __submit_bio+0x39f/0x550 [ 52.561473] submit_bio_noacct_nocheck+0x647/0xcc0 [ 52.562285] submit_bio_noacct+0x620/0x1e00 [ 52.563017] submit_bio+0xce/0x480 [ 52.563637] __swap_writepage+0x2f1/0xdf0 [ 52.564349] swap_writepage+0x464/0xbc0 [ 52.565022] shmem_writepage+0xdeb/0x1340 [ 52.565745] pageout+0x3bc/0x9b0 [ 52.566353] shrink_folio_list+0x16b9/0x3b60 [ 52.567104] shrink_lruvec+0xd78/0x2790 [ 52.567794] shrink_node+0xb29/0x2870 [ 52.568454] balance_pgdat+0x9c2/0x1500 [ 52.569142] kswapd+0x765/0xe00 [ 52.569741] kthread+0x35a/0x470 [ 52.570340] ret_from_fork+0x56/0x90 [ 52.570993] ret_from_fork_asm+0x1a/0x30 [ 52.571696] [ 52.571696] other info that might help us debug this: [ 52.571696] [ 52.572904] Chain exists of: [ 52.572904] &q->q_usage_counter(io)#25 --> &q->sysfs_lock --> fs_reclaim [ 52.572904] [ 52.574631] Possible unsafe locking scenario: [ 52.574631] [ 52.575547] CPU0 CPU1 [ 52.576246] ---- ---- [ 52.576942] lock(fs_reclaim); [ 52.577467] lock(&q->sysfs_lock); [ 52.578382] lock(fs_reclaim); [ 52.579250] rlock(&q->q_usage_counter(io)#25); [ 52.579974] [ 52.579974] *** DEADLOCK *** [ 52.579974] [ 52.580866] 1 lock held by kswapd0/56: [ 52.581459] #0: ffffffff872322a0 (fs_reclaim){+.+.}-{0:0}, at: balance_pgdat+0xb0f/0x1500 [ 52.582731] [ 52.582731] stack backtrace: [ 52.583404] CPU: 0 UID: 0 PID: 56 Comm: kswapd0 Not tainted 6.12.0-rc5-next-20241029-kvm-dirty #6 [ 52.584735] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/014 [ 52.586439] Call Trace: [ 52.586836] <TASK> [ 52.587190] dump_stack_lvl+0xea/0x150 [ 52.587753] dump_stack+0x19/0x20 [ 52.588253] print_circular_bug+0x47f/0x750 [ 52.588872] check_noncircular+0x2f4/0x3e0 [ 52.589492] ? __pfx_check_noncircular+0x10/0x10 [ 52.590180] ? lockdep_lock+0xd0/0x1d0 [ 52.590741] ? __pfx_lockdep_lock+0x10/0x10 [ 52.590790] kexec: Could not allocate control_code_buffer [ 52.591341] __lock_acquire+0x2ff8/0x5d60 [ 52.592365] ? __pfx___lock_acquire+0x10/0x10 [ 52.593087] ? __pfx_mark_lock.part.0+0x10/0x10 [ 52.593753] ? __kasan_check_read+0x15/0x20 [ 52.594366] lock_acquire.part.0+0x142/0x390 [ 52.594989] ? __submit_bio+0x39f/0x550 [ 52.595554] ? __pfx_lock_acquire.part.0+0x10/0x10 [ 52.596246] ? debug_smp_processor_id+0x20/0x30 [ 52.596900] ? rcu_is_watching+0x19/0xc0 [ 52.597484] ? trace_lock_acquire+0x139/0x1b0 [ 52.598118] lock_acquire+0x80/0xb0 [ 52.598633] ? __submit_bio+0x39f/0x550 [ 52.599191] blk_mq_submit_bio+0x1cbe/0x2590 [ 52.599805] ? __submit_bio+0x39f/0x550 [ 52.600361] ? __kasan_check_read+0x15/0x20 [ 52.600966] ? __pfx_blk_mq_submit_bio+0x10/0x10 [ 52.601632] ? __pfx_mark_lock.part.0+0x10/0x10 [ 52.602285] ? __this_cpu_preempt_check+0x21/0x30 [ 52.602968] ? __this_cpu_preempt_check+0x21/0x30 [ 52.603646] ? lock_release+0x441/0x870 [ 52.604207] __submit_bio+0x39f/0x550 [ 52.604742] ? __pfx___submit_bio+0x10/0x10 [ 52.605364] ? __this_cpu_preempt_check+0x21/0x30 [ 52.606045] ? seqcount_lockdep_reader_access.constprop.0+0xb4/0xd0 [ 52.606940] ? __sanitizer_cov_trace_const_cmp1+0x1e/0x30 [ 52.607707] ? kvm_clock_get_cycles+0x43/0x70 [ 52.608345] submit_bio_noacct_nocheck+0x647/0xcc0 [ 52.609045] ? __pfx_submit_bio_noacct_nocheck+0x10/0x10 [ 52.609820] ? __sanitizer_cov_trace_switch+0x58/0xa0 [ 52.610552] submit_bio_noacct+0x620/0x1e00 [ 52.611167] submit_bio+0xce/0x480 [ 52.611677] __swap_writepage+0x2f1/0xdf0 [ 52.612267] swap_writepage+0x464/0xbc0 [ 52.612837] shmem_writepage+0xdeb/0x1340 [ 52.613441] ? __pfx_shmem_writepage+0x10/0x10 [ 52.614090] ? __kasan_check_write+0x18/0x20 [ 52.614716] ? folio_clear_dirty_for_io+0xc1/0x600 [ 52.615403] pageout+0x3bc/0x9b0 [ 52.615894] ? __pfx_pageout+0x10/0x10 [ 52.616471] ? __pfx_folio_referenced_one+0x10/0x10 [ 52.617169] ? __pfx_folio_lock_anon_vma_read+0x10/0x10 [ 52.617918] ? __pfx_invalid_folio_referenced_vma+0x10/0x10 [ 52.618713] shrink_folio_list+0x16b9/0x3b60 [ 52.619346] ? __pfx_shrink_folio_list+0x10/0x10 [ 52.620021] ? __this_cpu_preempt_check+0x21/0x30 [ 52.620713] ? mark_lock.part.0+0xf3/0x17b0 [ 52.621339] ? isolate_lru_folios+0xcb1/0x1250 [ 52.621991] ? __pfx_mark_lock.part.0+0x10/0x10 [ 52.622655] ? __this_cpu_preempt_check+0x21/0x30 [ 52.623335] ? lock_release+0x441/0x870 [ 52.623900] ? __this_cpu_preempt_check+0x21/0x30 [ 52.624573] ? _raw_spin_unlock_irq+0x2c/0x60 [ 52.625204] ? lockdep_hardirqs_on+0x89/0x110 [ 52.625848] shrink_lruvec+0xd78/0x2790 [ 52.626422] ? __pfx_shrink_lruvec+0x10/0x10 [ 52.627040] ? __this_cpu_preempt_check+0x21/0x30 [ 52.627729] ? __this_cpu_preempt_check+0x21/0x30 [ 52.628423] ? trace_lock_acquire+0x139/0x1b0 [ 52.629061] ? trace_lock_acquire+0x139/0x1b0 [ 52.629752] shrink_node+0xb29/0x2870 [ 52.630305] ? __pfx_shrink_node+0x10/0x10 [ 52.630899] ? pgdat_balanced+0x1d4/0x230 [ 52.631490] balance_pgdat+0x9c2/0x1500 [ 52.632055] ? __pfx_balance_pgdat+0x10/0x10 [ 52.632669] ? __this_cpu_preempt_check+0x21/0x30 [ 52.633380] kswapd+0x765/0xe00 [ 52.633861] ? __pfx_kswapd+0x10/0x10 [ 52.634393] ? local_clock_noinstr+0xb0/0xd0 [ 52.635015] ? __pfx_autoremove_wake_function+0x10/0x10 [ 52.635759] ? __sanitizer_cov_trace_const_cmp1+0x1e/0x30 [ 52.636525] ? __kthread_parkme+0x15d/0x230 [ 52.637134] kthread+0x35a/0x470 [ 52.637616] ? __pfx_kswapd+0x10/0x10 [ 52.638146] ? __pfx_kthread+0x10/0x10 [ 52.638693] ret_from_fork+0x56/0x90 [ 52.639227] ? __pfx_kthread+0x10/0x10 [ 52.639778] ret_from_fork_asm+0x1a/0x30 [ 52.640391] </TASK> > Thanks, > Ming >
On Wed, Oct 30, 2024 at 06:39:13PM +0800, Lai, Yi wrote: > On Wed, Oct 30, 2024 at 05:50:15PM +0800, Ming Lei wrote: > > On Wed, Oct 30, 2024 at 4:51 PM Lai, Yi <yi1.lai@linux.intel.com> wrote: > > > > > > On Wed, Oct 30, 2024 at 03:13:09PM +0800, Ming Lei wrote: > > > > On Wed, Oct 30, 2024 at 02:45:03PM +0800, Lai, Yi wrote: > > ... > > > > > > > > It should be addressed by the following patch: > > > > > > > > https://lore.kernel.org/linux-block/ZyEGLdg744U_xBjp@fedora/ > > > > > > > > > > I have applied proposed fix patch on top of next-20241029. Issue can > > > still be reproduced. > > > > > > It seems the dependency chain is different from Marek's log and mine. > > > > Can you post the new log since q->q_usage_counter(io)->fs_reclaim from > > blk_mq_init_sched is cut down by the patch? > > > > New possible deadlock log after patch applied: This one looks like one real deadlock, any memory allocation with q->sysfs_lock held has such risk. There is another similar report related with queue sysfs store operation: https://lore.kernel.org/linux-scsi/ZxG38G9BuFdBpBHZ@fedora/ Thanks, Ming
On 29.10.2024 16:58, Ming Lei wrote: > On Tue, Oct 29, 2024 at 12:13:35PM +0100, Marek Szyprowski wrote: >> On 25.10.2024 02:37, Ming Lei wrote: >>> Recently we got several deadlock report[1][2][3] caused by >>> blk_mq_freeze_queue and blk_enter_queue(). >>> >>> Turns out the two are just like acquiring read/write lock, so model them >>> as read/write lock for supporting lockdep: >>> >>> 1) model q->q_usage_counter as two locks(io and queue lock) >>> >>> - queue lock covers sync with blk_enter_queue() >>> >>> - io lock covers sync with bio_enter_queue() >>> >>> 2) make the lockdep class/key as per-queue: >>> >>> - different subsystem has very different lock use pattern, shared lock >>> class causes false positive easily >>> >>> - freeze_queue degrades to no lock in case that disk state becomes DEAD >>> because bio_enter_queue() won't be blocked any more >>> >>> - freeze_queue degrades to no lock in case that request queue becomes dying >>> because blk_enter_queue() won't be blocked any more >>> >>> 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock >>> - it is exclusive lock, so dependency with blk_enter_queue() is covered >>> >>> - it is trylock because blk_mq_freeze_queue() are allowed to run >>> concurrently >>> >>> 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() >>> - nested blk_enter_queue() are allowed >>> >>> - dependency with blk_mq_freeze_queue() is covered >>> >>> - blk_queue_exit() is often called from other contexts(such as irq), and >>> it can't be annotated as lock_release(), so simply do it in >>> blk_enter_queue(), this way still covered cases as many as possible >>> >>> With lockdep support, such kind of reports may be reported asap and >>> needn't wait until the real deadlock is triggered. >>> >>> For example, lockdep report can be triggered in the report[3] with this >>> patch applied. >>> >>> [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' >>> https://bugzilla.kernel.org/show_bug.cgi?id=219166 >>> >>> [2] del_gendisk() vs blk_queue_enter() race condition >>> https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ >>> >>> [3] queue_freeze & queue_enter deadlock in scsi >>> https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u >>> >>> Reviewed-by: Christoph Hellwig <hch@lst.de> >>> Signed-off-by: Ming Lei <ming.lei@redhat.com> >> This patch landed yesterday in linux-next as commit f1be1788a32e >> ("block: model freeze & enter queue as lock for supporting lockdep"). >> In my tests I found that it introduces the following 2 lockdep warnings: >> >> > ... >> >> >> 2. On QEMU's ARM64 virt machine, observed during system suspend/resume >> cycle: >> >> # time rtcwake -s10 -mmem >> rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 >> PM: suspend entry (s2idle) >> Filesystems sync: 0.004 seconds >> Freezing user space processes >> Freezing user space processes completed (elapsed 0.007 seconds) >> OOM killer disabled. >> Freezing remaining freezable tasks >> Freezing remaining freezable tasks completed (elapsed 0.004 seconds) >> >> ====================================================== >> WARNING: possible circular locking dependency detected >> 6.12.0-rc4+ #9291 Not tainted >> ------------------------------------------------------ >> rtcwake/1299 is trying to acquire lock: >> ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 >> >> but task is already holding lock: >> ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: >> virtblk_freeze+0x24/0x60 >> >> which lock already depends on the new lock. >> >> >> the existing dependency chain (in reverse order) is: > This one looks a real thing, at least the added lockdep code works as > expected, also the blk_mq_freeze_queue() use in virtio-blk's ->suspend() > is questionable. I will take a further look. Did you find a way to fix this one? I still observe such warnings in my tests, even though your lockdep fixes are already merged to -next: https://lore.kernel.org/all/20241031133723.303835-1-ming.lei@redhat.com/ Best regards
On Tue, Nov 12, 2024 at 09:36:40AM +0100, Marek Szyprowski wrote: > On 29.10.2024 16:58, Ming Lei wrote: > > On Tue, Oct 29, 2024 at 12:13:35PM +0100, Marek Szyprowski wrote: > >> On 25.10.2024 02:37, Ming Lei wrote: > >>> Recently we got several deadlock report[1][2][3] caused by > >>> blk_mq_freeze_queue and blk_enter_queue(). > >>> > >>> Turns out the two are just like acquiring read/write lock, so model them > >>> as read/write lock for supporting lockdep: > >>> > >>> 1) model q->q_usage_counter as two locks(io and queue lock) > >>> > >>> - queue lock covers sync with blk_enter_queue() > >>> > >>> - io lock covers sync with bio_enter_queue() > >>> > >>> 2) make the lockdep class/key as per-queue: > >>> > >>> - different subsystem has very different lock use pattern, shared lock > >>> class causes false positive easily > >>> > >>> - freeze_queue degrades to no lock in case that disk state becomes DEAD > >>> because bio_enter_queue() won't be blocked any more > >>> > >>> - freeze_queue degrades to no lock in case that request queue becomes dying > >>> because blk_enter_queue() won't be blocked any more > >>> > >>> 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock > >>> - it is exclusive lock, so dependency with blk_enter_queue() is covered > >>> > >>> - it is trylock because blk_mq_freeze_queue() are allowed to run > >>> concurrently > >>> > >>> 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() > >>> - nested blk_enter_queue() are allowed > >>> > >>> - dependency with blk_mq_freeze_queue() is covered > >>> > >>> - blk_queue_exit() is often called from other contexts(such as irq), and > >>> it can't be annotated as lock_release(), so simply do it in > >>> blk_enter_queue(), this way still covered cases as many as possible > >>> > >>> With lockdep support, such kind of reports may be reported asap and > >>> needn't wait until the real deadlock is triggered. > >>> > >>> For example, lockdep report can be triggered in the report[3] with this > >>> patch applied. > >>> > >>> [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' > >>> https://bugzilla.kernel.org/show_bug.cgi?id=219166 > >>> > >>> [2] del_gendisk() vs blk_queue_enter() race condition > >>> https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ > >>> > >>> [3] queue_freeze & queue_enter deadlock in scsi > >>> https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u > >>> > >>> Reviewed-by: Christoph Hellwig <hch@lst.de> > >>> Signed-off-by: Ming Lei <ming.lei@redhat.com> > >> This patch landed yesterday in linux-next as commit f1be1788a32e > >> ("block: model freeze & enter queue as lock for supporting lockdep"). > >> In my tests I found that it introduces the following 2 lockdep warnings: > >> > >> > ... > >> > >> > >> 2. On QEMU's ARM64 virt machine, observed during system suspend/resume > >> cycle: > >> > >> # time rtcwake -s10 -mmem > >> rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 > >> PM: suspend entry (s2idle) > >> Filesystems sync: 0.004 seconds > >> Freezing user space processes > >> Freezing user space processes completed (elapsed 0.007 seconds) > >> OOM killer disabled. > >> Freezing remaining freezable tasks > >> Freezing remaining freezable tasks completed (elapsed 0.004 seconds) > >> > >> ====================================================== > >> WARNING: possible circular locking dependency detected > >> 6.12.0-rc4+ #9291 Not tainted > >> ------------------------------------------------------ > >> rtcwake/1299 is trying to acquire lock: > >> ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 > >> > >> but task is already holding lock: > >> ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: > >> virtblk_freeze+0x24/0x60 > >> > >> which lock already depends on the new lock. > >> > >> > >> the existing dependency chain (in reverse order) is: > > This one looks a real thing, at least the added lockdep code works as > > expected, also the blk_mq_freeze_queue() use in virtio-blk's ->suspend() > > is questionable. I will take a further look. > > Did you find a way to fix this one? I still observe such warnings in my > tests, even though your lockdep fixes are already merged to -next: > https://lore.kernel.org/all/20241031133723.303835-1-ming.lei@redhat.com/ The lockdep fixes in ->next is just for making the added lockdep work correctly, and virtio-blk is another story. It might be fine to annotate it with blk_mq_freeze_queue_no_owner(), but it looks very fragile to call freeze queue in ->suspend(), and the lock is just kept as being grabbed in the whole suspend code path. Can you try the following patch? diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 194417abc105..21488740eb15 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -1594,6 +1594,7 @@ static int virtblk_freeze(struct virtio_device *vdev) /* Ensure no requests in virtqueues before deleting vqs. */ blk_mq_freeze_queue(vblk->disk->queue); + blk_mq_unfreeze_queue(vblk->disk->queue); /* Ensure we don't receive any more interrupts */ virtio_reset_device(vdev); @@ -1617,8 +1618,6 @@ static int virtblk_restore(struct virtio_device *vdev) return ret; virtio_device_ready(vdev); - - blk_mq_unfreeze_queue(vblk->disk->queue); return 0; } #endif Thanks, Ming
On 12.11.2024 11:15, Ming Lei wrote: > On Tue, Nov 12, 2024 at 09:36:40AM +0100, Marek Szyprowski wrote: >> On 29.10.2024 16:58, Ming Lei wrote: >>> On Tue, Oct 29, 2024 at 12:13:35PM +0100, Marek Szyprowski wrote: >>>> On 25.10.2024 02:37, Ming Lei wrote: >>>>> Recently we got several deadlock report[1][2][3] caused by >>>>> blk_mq_freeze_queue and blk_enter_queue(). >>>>> >>>>> Turns out the two are just like acquiring read/write lock, so model them >>>>> as read/write lock for supporting lockdep: >>>>> >>>>> 1) model q->q_usage_counter as two locks(io and queue lock) >>>>> >>>>> - queue lock covers sync with blk_enter_queue() >>>>> >>>>> - io lock covers sync with bio_enter_queue() >>>>> >>>>> 2) make the lockdep class/key as per-queue: >>>>> >>>>> - different subsystem has very different lock use pattern, shared lock >>>>> class causes false positive easily >>>>> >>>>> - freeze_queue degrades to no lock in case that disk state becomes DEAD >>>>> because bio_enter_queue() won't be blocked any more >>>>> >>>>> - freeze_queue degrades to no lock in case that request queue becomes dying >>>>> because blk_enter_queue() won't be blocked any more >>>>> >>>>> 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock >>>>> - it is exclusive lock, so dependency with blk_enter_queue() is covered >>>>> >>>>> - it is trylock because blk_mq_freeze_queue() are allowed to run >>>>> concurrently >>>>> >>>>> 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() >>>>> - nested blk_enter_queue() are allowed >>>>> >>>>> - dependency with blk_mq_freeze_queue() is covered >>>>> >>>>> - blk_queue_exit() is often called from other contexts(such as irq), and >>>>> it can't be annotated as lock_release(), so simply do it in >>>>> blk_enter_queue(), this way still covered cases as many as possible >>>>> >>>>> With lockdep support, such kind of reports may be reported asap and >>>>> needn't wait until the real deadlock is triggered. >>>>> >>>>> For example, lockdep report can be triggered in the report[3] with this >>>>> patch applied. >>>>> >>>>> [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' >>>>> https://bugzilla.kernel.org/show_bug.cgi?id=219166 >>>>> >>>>> [2] del_gendisk() vs blk_queue_enter() race condition >>>>> https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ >>>>> >>>>> [3] queue_freeze & queue_enter deadlock in scsi >>>>> https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u >>>>> >>>>> Reviewed-by: Christoph Hellwig <hch@lst.de> >>>>> Signed-off-by: Ming Lei <ming.lei@redhat.com> >>>> This patch landed yesterday in linux-next as commit f1be1788a32e >>>> ("block: model freeze & enter queue as lock for supporting lockdep"). >>>> In my tests I found that it introduces the following 2 lockdep warnings: >>>> >>>>> ... >>>> >>>> 2. On QEMU's ARM64 virt machine, observed during system suspend/resume >>>> cycle: >>>> >>>> # time rtcwake -s10 -mmem >>>> rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 >>>> PM: suspend entry (s2idle) >>>> Filesystems sync: 0.004 seconds >>>> Freezing user space processes >>>> Freezing user space processes completed (elapsed 0.007 seconds) >>>> OOM killer disabled. >>>> Freezing remaining freezable tasks >>>> Freezing remaining freezable tasks completed (elapsed 0.004 seconds) >>>> >>>> ====================================================== >>>> WARNING: possible circular locking dependency detected >>>> 6.12.0-rc4+ #9291 Not tainted >>>> ------------------------------------------------------ >>>> rtcwake/1299 is trying to acquire lock: >>>> ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 >>>> >>>> but task is already holding lock: >>>> ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: >>>> virtblk_freeze+0x24/0x60 >>>> >>>> which lock already depends on the new lock. >>>> >>>> >>>> the existing dependency chain (in reverse order) is: >>> This one looks a real thing, at least the added lockdep code works as >>> expected, also the blk_mq_freeze_queue() use in virtio-blk's ->suspend() >>> is questionable. I will take a further look. >> Did you find a way to fix this one? I still observe such warnings in my >> tests, even though your lockdep fixes are already merged to -next: >> https://lore.kernel.org/all/20241031133723.303835-1-ming.lei@redhat.com/ > The lockdep fixes in ->next is just for making the added lockdep work > correctly, and virtio-blk is another story. > > It might be fine to annotate it with blk_mq_freeze_queue_no_owner(), > but it looks very fragile to call freeze queue in ->suspend(), and the lock > is just kept as being grabbed in the whole suspend code path. > > Can you try the following patch? Yes, this hides this lockdep warning, but imho it looks like a workaround, not a final fix. Reported-by: Marek Szyprowski <m.szyprowski@samsung.com> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> > diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c > index 194417abc105..21488740eb15 100644 > --- a/drivers/block/virtio_blk.c > +++ b/drivers/block/virtio_blk.c > @@ -1594,6 +1594,7 @@ static int virtblk_freeze(struct virtio_device *vdev) > > /* Ensure no requests in virtqueues before deleting vqs. */ > blk_mq_freeze_queue(vblk->disk->queue); > + blk_mq_unfreeze_queue(vblk->disk->queue); > > /* Ensure we don't receive any more interrupts */ > virtio_reset_device(vdev); > @@ -1617,8 +1618,6 @@ static int virtblk_restore(struct virtio_device *vdev) > return ret; > > virtio_device_ready(vdev); > - > - blk_mq_unfreeze_queue(vblk->disk->queue); > return 0; > } > #endif > > > > Thanks, > Ming > > Best regards
On Tue, Nov 12, 2024 at 12:32:29PM +0100, Marek Szyprowski wrote: > On 12.11.2024 11:15, Ming Lei wrote: > > On Tue, Nov 12, 2024 at 09:36:40AM +0100, Marek Szyprowski wrote: > >> On 29.10.2024 16:58, Ming Lei wrote: > >>> On Tue, Oct 29, 2024 at 12:13:35PM +0100, Marek Szyprowski wrote: > >>>> On 25.10.2024 02:37, Ming Lei wrote: > >>>>> Recently we got several deadlock report[1][2][3] caused by > >>>>> blk_mq_freeze_queue and blk_enter_queue(). > >>>>> > >>>>> Turns out the two are just like acquiring read/write lock, so model them > >>>>> as read/write lock for supporting lockdep: > >>>>> > >>>>> 1) model q->q_usage_counter as two locks(io and queue lock) > >>>>> > >>>>> - queue lock covers sync with blk_enter_queue() > >>>>> > >>>>> - io lock covers sync with bio_enter_queue() > >>>>> > >>>>> 2) make the lockdep class/key as per-queue: > >>>>> > >>>>> - different subsystem has very different lock use pattern, shared lock > >>>>> class causes false positive easily > >>>>> > >>>>> - freeze_queue degrades to no lock in case that disk state becomes DEAD > >>>>> because bio_enter_queue() won't be blocked any more > >>>>> > >>>>> - freeze_queue degrades to no lock in case that request queue becomes dying > >>>>> because blk_enter_queue() won't be blocked any more > >>>>> > >>>>> 3) model blk_mq_freeze_queue() as acquire_exclusive & try_lock > >>>>> - it is exclusive lock, so dependency with blk_enter_queue() is covered > >>>>> > >>>>> - it is trylock because blk_mq_freeze_queue() are allowed to run > >>>>> concurrently > >>>>> > >>>>> 4) model blk_enter_queue() & bio_enter_queue() as acquire_read() > >>>>> - nested blk_enter_queue() are allowed > >>>>> > >>>>> - dependency with blk_mq_freeze_queue() is covered > >>>>> > >>>>> - blk_queue_exit() is often called from other contexts(such as irq), and > >>>>> it can't be annotated as lock_release(), so simply do it in > >>>>> blk_enter_queue(), this way still covered cases as many as possible > >>>>> > >>>>> With lockdep support, such kind of reports may be reported asap and > >>>>> needn't wait until the real deadlock is triggered. > >>>>> > >>>>> For example, lockdep report can be triggered in the report[3] with this > >>>>> patch applied. > >>>>> > >>>>> [1] occasional block layer hang when setting 'echo noop > /sys/block/sda/queue/scheduler' > >>>>> https://bugzilla.kernel.org/show_bug.cgi?id=219166 > >>>>> > >>>>> [2] del_gendisk() vs blk_queue_enter() race condition > >>>>> https://lore.kernel.org/linux-block/20241003085610.GK11458@google.com/ > >>>>> > >>>>> [3] queue_freeze & queue_enter deadlock in scsi > >>>>> https://lore.kernel.org/linux-block/ZxG38G9BuFdBpBHZ@fedora/T/#u > >>>>> > >>>>> Reviewed-by: Christoph Hellwig <hch@lst.de> > >>>>> Signed-off-by: Ming Lei <ming.lei@redhat.com> > >>>> This patch landed yesterday in linux-next as commit f1be1788a32e > >>>> ("block: model freeze & enter queue as lock for supporting lockdep"). > >>>> In my tests I found that it introduces the following 2 lockdep warnings: > >>>> > >>>>> ... > >>>> > >>>> 2. On QEMU's ARM64 virt machine, observed during system suspend/resume > >>>> cycle: > >>>> > >>>> # time rtcwake -s10 -mmem > >>>> rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Oct 29 11:54:30 2024 > >>>> PM: suspend entry (s2idle) > >>>> Filesystems sync: 0.004 seconds > >>>> Freezing user space processes > >>>> Freezing user space processes completed (elapsed 0.007 seconds) > >>>> OOM killer disabled. > >>>> Freezing remaining freezable tasks > >>>> Freezing remaining freezable tasks completed (elapsed 0.004 seconds) > >>>> > >>>> ====================================================== > >>>> WARNING: possible circular locking dependency detected > >>>> 6.12.0-rc4+ #9291 Not tainted > >>>> ------------------------------------------------------ > >>>> rtcwake/1299 is trying to acquire lock: > >>>> ffff80008358a7f8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_lock+0x1c/0x28 > >>>> > >>>> but task is already holding lock: > >>>> ffff000006136d68 (&q->q_usage_counter(io)#5){++++}-{0:0}, at: > >>>> virtblk_freeze+0x24/0x60 > >>>> > >>>> which lock already depends on the new lock. > >>>> > >>>> > >>>> the existing dependency chain (in reverse order) is: > >>> This one looks a real thing, at least the added lockdep code works as > >>> expected, also the blk_mq_freeze_queue() use in virtio-blk's ->suspend() > >>> is questionable. I will take a further look. > >> Did you find a way to fix this one? I still observe such warnings in my > >> tests, even though your lockdep fixes are already merged to -next: > >> https://lore.kernel.org/all/20241031133723.303835-1-ming.lei@redhat.com/ > > The lockdep fixes in ->next is just for making the added lockdep work > > correctly, and virtio-blk is another story. > > > > It might be fine to annotate it with blk_mq_freeze_queue_no_owner(), > > but it looks very fragile to call freeze queue in ->suspend(), and the lock > > is just kept as being grabbed in the whole suspend code path. > > > > Can you try the following patch? > > Yes, this hides this lockdep warning, but imho it looks like a > workaround, not a final fix. > > Reported-by: Marek Szyprowski <m.szyprowski@samsung.com> > Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> Thanks for the test! It is actually not workaround, because what virtblk_freeze() needs is to drain all in-flight IOs. One thing missed is to mark the queue as quiesced, and I will post one formal patch with queue quiesce covered. Thanks, Ming
diff --git a/block/blk-core.c b/block/blk-core.c index bc5e8c5eaac9..09d10bb95fda 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -261,6 +261,8 @@ static void blk_free_queue(struct request_queue *q) blk_mq_release(q); ida_free(&blk_queue_ida, q->id); + lockdep_unregister_key(&q->io_lock_cls_key); + lockdep_unregister_key(&q->q_lock_cls_key); call_rcu(&q->rcu_head, blk_free_queue_rcu); } @@ -278,18 +280,20 @@ void blk_put_queue(struct request_queue *q) } EXPORT_SYMBOL(blk_put_queue); -void blk_queue_start_drain(struct request_queue *q) +bool blk_queue_start_drain(struct request_queue *q) { /* * When queue DYING flag is set, we need to block new req * entering queue, so we call blk_freeze_queue_start() to * prevent I/O from crossing blk_queue_enter(). */ - blk_freeze_queue_start(q); + bool freeze = __blk_freeze_queue_start(q); if (queue_is_mq(q)) blk_mq_wake_waiters(q); /* Make blk_queue_enter() reexamine the DYING flag. */ wake_up_all(&q->mq_freeze_wq); + + return freeze; } /** @@ -321,6 +325,8 @@ int blk_queue_enter(struct request_queue *q, blk_mq_req_flags_t flags) return -ENODEV; } + rwsem_acquire_read(&q->q_lockdep_map, 0, 0, _RET_IP_); + rwsem_release(&q->q_lockdep_map, _RET_IP_); return 0; } @@ -352,6 +358,8 @@ int __bio_queue_enter(struct request_queue *q, struct bio *bio) goto dead; } + rwsem_acquire_read(&q->io_lockdep_map, 0, 0, _RET_IP_); + rwsem_release(&q->io_lockdep_map, _RET_IP_); return 0; dead: bio_io_error(bio); @@ -441,6 +449,12 @@ struct request_queue *blk_alloc_queue(struct queue_limits *lim, int node_id) PERCPU_REF_INIT_ATOMIC, GFP_KERNEL); if (error) goto fail_stats; + lockdep_register_key(&q->io_lock_cls_key); + lockdep_register_key(&q->q_lock_cls_key); + lockdep_init_map(&q->io_lockdep_map, "&q->q_usage_counter(io)", + &q->io_lock_cls_key, 0); + lockdep_init_map(&q->q_lockdep_map, "&q->q_usage_counter(queue)", + &q->q_lock_cls_key, 0); q->nr_requests = BLKDEV_DEFAULT_RQ; diff --git a/block/blk-mq.c b/block/blk-mq.c index 96858fb3b9ff..76f277a30c11 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -120,17 +120,29 @@ void blk_mq_in_flight_rw(struct request_queue *q, struct block_device *part, inflight[1] = mi.inflight[1]; } -void blk_freeze_queue_start(struct request_queue *q) +bool __blk_freeze_queue_start(struct request_queue *q) { + int freeze; + mutex_lock(&q->mq_freeze_lock); if (++q->mq_freeze_depth == 1) { percpu_ref_kill(&q->q_usage_counter); mutex_unlock(&q->mq_freeze_lock); if (queue_is_mq(q)) blk_mq_run_hw_queues(q, false); + freeze = true; } else { mutex_unlock(&q->mq_freeze_lock); + freeze = false; } + + return freeze; +} + +void blk_freeze_queue_start(struct request_queue *q) +{ + if (__blk_freeze_queue_start(q)) + blk_freeze_acquire_lock(q, false, false); } EXPORT_SYMBOL_GPL(blk_freeze_queue_start); @@ -176,8 +188,10 @@ void blk_mq_freeze_queue(struct request_queue *q) } EXPORT_SYMBOL_GPL(blk_mq_freeze_queue); -void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic) +bool __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic) { + int unfreeze = false; + mutex_lock(&q->mq_freeze_lock); if (force_atomic) q->q_usage_counter.data->force_atomic = true; @@ -186,13 +200,17 @@ void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic) if (!q->mq_freeze_depth) { percpu_ref_resurrect(&q->q_usage_counter); wake_up_all(&q->mq_freeze_wq); + unfreeze = true; } mutex_unlock(&q->mq_freeze_lock); + + return unfreeze; } void blk_mq_unfreeze_queue(struct request_queue *q) { - __blk_mq_unfreeze_queue(q, false); + if (__blk_mq_unfreeze_queue(q, false)) + blk_unfreeze_release_lock(q, false, false); } EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); @@ -205,7 +223,7 @@ EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); */ void blk_freeze_queue_start_non_owner(struct request_queue *q) { - blk_freeze_queue_start(q); + __blk_freeze_queue_start(q); } EXPORT_SYMBOL_GPL(blk_freeze_queue_start_non_owner); diff --git a/block/blk.h b/block/blk.h index c718e4291db0..832e54c5a271 100644 --- a/block/blk.h +++ b/block/blk.h @@ -4,6 +4,7 @@ #include <linux/bio-integrity.h> #include <linux/blk-crypto.h> +#include <linux/lockdep.h> #include <linux/memblock.h> /* for max_pfn/max_low_pfn */ #include <linux/sched/sysctl.h> #include <linux/timekeeping.h> @@ -35,8 +36,9 @@ struct blk_flush_queue *blk_alloc_flush_queue(int node, int cmd_size, void blk_free_flush_queue(struct blk_flush_queue *q); void blk_freeze_queue(struct request_queue *q); -void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic); -void blk_queue_start_drain(struct request_queue *q); +bool __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic); +bool blk_queue_start_drain(struct request_queue *q); +bool __blk_freeze_queue_start(struct request_queue *q); int __bio_queue_enter(struct request_queue *q, struct bio *bio); void submit_bio_noacct_nocheck(struct bio *bio); void bio_await_chain(struct bio *bio); @@ -69,8 +71,11 @@ static inline int bio_queue_enter(struct bio *bio) { struct request_queue *q = bdev_get_queue(bio->bi_bdev); - if (blk_try_enter_queue(q, false)) + if (blk_try_enter_queue(q, false)) { + rwsem_acquire_read(&q->io_lockdep_map, 0, 0, _RET_IP_); + rwsem_release(&q->io_lockdep_map, _RET_IP_); return 0; + } return __bio_queue_enter(q, bio); } @@ -734,4 +739,22 @@ void blk_integrity_verify(struct bio *bio); void blk_integrity_prepare(struct request *rq); void blk_integrity_complete(struct request *rq, unsigned int nr_bytes); +static inline void blk_freeze_acquire_lock(struct request_queue *q, bool + disk_dead, bool queue_dying) +{ + if (!disk_dead) + rwsem_acquire(&q->io_lockdep_map, 0, 1, _RET_IP_); + if (!queue_dying) + rwsem_acquire(&q->q_lockdep_map, 0, 1, _RET_IP_); +} + +static inline void blk_unfreeze_release_lock(struct request_queue *q, bool + disk_dead, bool queue_dying) +{ + if (!queue_dying) + rwsem_release(&q->q_lockdep_map, _RET_IP_); + if (!disk_dead) + rwsem_release(&q->io_lockdep_map, _RET_IP_); +} + #endif /* BLK_INTERNAL_H */ diff --git a/block/genhd.c b/block/genhd.c index 1c05dd4c6980..6ad3fcde0110 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -581,13 +581,13 @@ static void blk_report_disk_dead(struct gendisk *disk, bool surprise) rcu_read_unlock(); } -static void __blk_mark_disk_dead(struct gendisk *disk) +static bool __blk_mark_disk_dead(struct gendisk *disk) { /* * Fail any new I/O. */ if (test_and_set_bit(GD_DEAD, &disk->state)) - return; + return false; if (test_bit(GD_OWNS_QUEUE, &disk->state)) blk_queue_flag_set(QUEUE_FLAG_DYING, disk->queue); @@ -600,7 +600,7 @@ static void __blk_mark_disk_dead(struct gendisk *disk) /* * Prevent new I/O from crossing bio_queue_enter(). */ - blk_queue_start_drain(disk->queue); + return blk_queue_start_drain(disk->queue); } /** @@ -641,6 +641,7 @@ void del_gendisk(struct gendisk *disk) struct request_queue *q = disk->queue; struct block_device *part; unsigned long idx; + bool start_drain, queue_dying; might_sleep(); @@ -668,7 +669,10 @@ void del_gendisk(struct gendisk *disk) * Drop all partitions now that the disk is marked dead. */ mutex_lock(&disk->open_mutex); - __blk_mark_disk_dead(disk); + start_drain = __blk_mark_disk_dead(disk); + queue_dying = blk_queue_dying(q); + if (start_drain) + blk_freeze_acquire_lock(q, true, queue_dying); xa_for_each_start(&disk->part_tbl, idx, part, 1) drop_partition(part); mutex_unlock(&disk->open_mutex); @@ -725,6 +729,9 @@ void del_gendisk(struct gendisk *disk) if (queue_is_mq(q)) blk_mq_exit_queue(q); } + + if (start_drain) + blk_unfreeze_release_lock(q, true, queue_dying); } EXPORT_SYMBOL(del_gendisk); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 50c3b959da28..57f1ee386b57 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -25,6 +25,7 @@ #include <linux/uuid.h> #include <linux/xarray.h> #include <linux/file.h> +#include <linux/lockdep.h> struct module; struct request_queue; @@ -471,6 +472,11 @@ struct request_queue { struct xarray hctx_table; struct percpu_ref q_usage_counter; + struct lock_class_key io_lock_cls_key; + struct lockdep_map io_lockdep_map; + + struct lock_class_key q_lock_cls_key; + struct lockdep_map q_lockdep_map; struct request *last_merge;