Re: [bug report] WARNING: block/blk-mq-debugfs.c:620 at debugfs_create_files.isra.0+0x54/0x60, CPU#220: kworker/u1036:7/70421

["Yu Kuai" <yukuai@fnnas.com>]

Hi,

在 2026/2/3 20:28, Yi Zhang 写道:
> Hello
>
> CKI reported the following failure today. Seems it was introduced from
> the recently blk-mq-debugfs update, please help check and let me know
> if you need any testing/info.
>
> Reproducer: blktests nvme/tcp nvme/005
> Commit message: Merge branch 'for-7.0/block' into for-next
>
> dmesg:
> [  951.319384] run blktests nvme/005 at 2026-02-02 17:59:35
> [  951.346116] loop0: detected capacity change from 0 to 2097152
> [  951.356428] nvmet: adding nsid 1 to subsystem blktests-subsystem-1
> [  951.370823] nvmet_tcp: enabling port 0 (127.0.0.1:4420)
> [  951.397438] nvmet: Created nvm controller 1 for subsystem
> blktests-subsystem-1 for NQN
> nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349.
> [  951.412081] nvme nvme0: creating 128 I/O queues.
> [  951.425914] nvme nvme0: mapped 128/0/0 default/read/poll queues.
> [  951.467027] nvme nvme0: new ctrl: NQN "blktests-subsystem-1", addr
> 127.0.0.1:4420, hostnqn:
> nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349
> [  951.784608] nvmet: Created nvm controller 2 for subsystem
> blktests-subsystem-1 for NQN
> nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349.
> [  951.798920] nvme nvme0: creating 128 I/O queues.
> [  951.871600] nvme nvme0: mapped 128/0/0 default/read/poll queues.

Thanks for the test!

Looks like this is because blk_mq_update_nr_hw_queues() is called with queue
already frozen from nvme-tcp, nvme-rdma, nvme-apple:

nvme_start_freeze()
blk_mq_update_nr_hw_queues()
nvme_unfreeze()

And recent commit add a WARN_ON() check for this. This looks fine, but I'm not
quite sure if it's ok to remove the freeze queue before blk_mq_update_nr_hw_queues().

> [  951.880680] ------------[ cut here ]------------
> [  951.885309] WARNING: block/blk-mq-debugfs.c:620 at
> debugfs_create_files.isra.0+0x54/0x60, CPU#220: kworker/u1036:7/70421
> [  951.896190] Modules linked in: nvmet_tcp nvmet nvme_tcp
> nvme_fabrics nvme nvme_core nvme_keyring nvme_auth rtrs_core rdma_cm
> iw_cm ib_cm hkdf nbd loop mlx5_ib ib_uverbs macsec sunrpc ib_core
> mlx5_fwctl fwctl rfkill vfat fat amd_atl intel_rapl_msr
> intel_rapl_common amd64_edac edac_mce_amd jc42 kvm_amd ee1004 joydev
> wmi_bmof mlx5_core kvm irqbypass mlxfw rapl psample tls ptdma ipmi_si
> ipmi_devintf pcspkr ipmi_msghandler acpi_cpufreq i2c_piix4 k10temp
> i2c_smbus fuse nfnetlink zram lz4hc_compress lz4_compress xfs ast
> i2c_algo_bit ghash_clmulni_intel wmi sp5100_tco i2c_dev [last
> unloaded: nvmet]
> [  951.948577] CPU: 220 UID: 0 PID: 70421 Comm: kworker/u1036:7 Not
> tainted 6.19.0-rc7 #1 PREEMPT(lazy)
> [  951.957814] Hardware name: AMD Corporation DAYTONA_X/DAYTONA_X,
> BIOS RYM100GA 01/13/2025
> [  951.965941] Workqueue: nvme-reset-wq nvme_reset_ctrl_work [nvme_tcp]
> [  951.972321] RIP: 0010:debugfs_create_files.isra.0+0x54/0x60
> [  951.977911] Code: 48 89 d9 4d 89 e0 48 89 ea 49 c7 c1 20 20 73 9a
> 48 83 c3 28 e8 bd 57 d9 ff 48 8b 3b 48 85 ff 75 db 5b 5d 41 5c e9 f7
> b6 51 ff <0f> 0b eb be 0f 1f 84 00 00 00 00 00 90 90 90 90 90 90 90 90
> 90 90
> [  951.996673] RSP: 0018:ffffd137f8ee3c90 EFLAGS: 00010202
> [  952.001919] RAX: ffff8e3a49aa6c00 RBX: ffffffff9a731e00 RCX: ffffffff9a731e00
> [  952.009070] RDX: ffff8e3a831ab000 RSI: ffff8e3a49aa6c00 RDI: 0000000000000001
> [  952.016227] RBP: ffff8e3a49aa6c00 R08: 0000000000000064 R09: 0000000000000000
> [  952.023396] R10: ffff8e3a1c6a2540 R11: 0000000000000000 R12: ffff8e3a831ab000
> [  952.030607] R13: ffff8e3a64c03620 R14: ffffd137f8ee3d38 R15: ffff8e3a8c43e400
> [  952.037815] FS:  0000000000000000(0000) GS:ffff8e5601f7b000(0000)
> knlGS:0000000000000000
> [  952.045946] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [  952.051708] CR2: 00007f310853ef58 CR3: 00000037aac2e003 CR4: 0000000000f70ef0
> [  952.058856] PKRU: 55555554
> [  952.061579] Call Trace:
> [  952.064048]  <TASK>
> [  952.066159]  blk_mq_debugfs_register_hctx+0x97/0x150
> [  952.071136]  ? srso_alias_return_thunk+0x5/0xfbef5
> [  952.075945]  blk_mq_debugfs_register_hctxs+0x4e/0x70
> [  952.080924]  __blk_mq_update_nr_hw_queues+0x218/0x390
> [  952.085991]  blk_mq_update_nr_hw_queues+0x39/0x50
> [  952.090714]  nvme_tcp_configure_io_queues.cold+0x1ac/0x35e [nvme_tcp]
> [  952.097175]  nvme_tcp_setup_ctrl+0xab/0x250 [nvme_tcp]
> [  952.102327]  ? nvme_change_ctrl_state+0x135/0x1b0 [nvme_core]
> [  952.108090]  nvme_reset_ctrl_work+0x9a/0xd0 [nvme_tcp]
> [  952.113246]  process_one_work+0x18d/0x340
> [  952.117266]  worker_thread+0x256/0x3a0
> [  952.121030]  ? __pfx_worker_thread+0x10/0x10
> [  952.125320]  kthread+0xfc/0x240
> [  952.128481]  ? __pfx_kthread+0x10/0x10
> [  952.132245]  ? __pfx_kthread+0x10/0x10
> [  952.136009]  ret_from_fork+0x126/0x190
> [  952.139773]  ? __pfx_kthread+0x10/0x10
> [  952.143539]  ret_from_fork_asm+0x1a/0x30
> [  952.147484]  </TASK>
> [  952.149687] ---[ end trace 0000000000000000 ]---
> [  952.167930] nvme nvme0: Removing ctrl: NQN "blktests-subsystem-1"
> [  952.195424] block nvme0n1: no available path - failing I/O
> [  952.200938] block nvme0n1: no available path - failing I/O
> [  952.206483] buffer_io_error: 439 callbacks suppressed
> [  952.206486] Buffer I/O error on dev nvme0n1, logical block 262142,
> async page read
>
>
>
-- 
Thansk,
Kuai