Arch Linux

Description:

I was trying to do mkfs.xfs to a partition in a disk inside an ASM2362 USB-NVME enclosure, but the command never returned.

It happens with linux 5.12.12.arch1-1 but not linux-lts 5.10.45-1.

Command output:

meta-data=/dev/sda2 isize=512 agcount=4, agsize=17176661 blks
= sectsz=512 attr=2, projid32bit=1
= crc=1 finobt=1, sparse=1, rmapbt=0
= reflink=1 bigtime=0 inobtcount=0
data = bsize=4096 blocks=68706641, imaxpct=25
= sunit=0 swidth=0 blks
naming =version 2 bsize=4096 ascii-ci=0, ftype=1
log =internal log bsize=4096 blocks=33548, version=2
= sectsz=512 sunit=0 blks, lazy-count=1
realtime =none extsz=4096 blocks=0, rtextents=0

/proc/pid/stack:

[<0>] submit_bio_wait+0x74/0xb0
[<0>] blkdev_issue_zeroout+0x16e/0x250
[<0>] blkdev_fallocate+0x12a/0x190
[<0>] vfs_fallocate+0x14a/0x320
[<0>] __x64_sys_fallocate+0x3c/0x70
[<0>] do_syscall_64+0x33/0x40
[<0>] entry_SYSCALL_64_after_hwframe+0x44/0xae

dmesg:

[ 1597.193239] INFO: task kworker/u32:3:187 blocked for more than 860 seconds.
[ 1597.193252] Not tainted 5.12.12-arch1-1 #1
[ 1597.193257] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1597.193260] task:kworker/u32:3 state:D stack: 0 pid: 187 ppid: 2 flags:0x00004000
[ 1597.193270] Workqueue: scsi_tmf_2 scmd_eh_abort_handler
[ 1597.193281] Call Trace:
[ 1597.193289] __schedule+0x2ff/0x8b0
[ 1597.193300] schedule+0x5b/0xc0
[ 1597.193306] usb_kill_urb+0xa9/0xe0
[ 1597.193371] ? wait_woken+0x80/0x80
[ 1597.193379] uas_eh_abort_handler+0xf2/0x170 [uas]
[ 1597.193387] scmd_eh_abort_handler+0x7e/0x230
[ 1597.193393] process_one_work+0x214/0x3e0
[ 1597.193397] worker_thread+0x4d/0x3d0
[ 1597.193401] ? process_one_work+0x3e0/0x3e0
[ 1597.193404] kthread+0x133/0x150
[ 1597.193409] ? kthread_associate_blkcg+0xc0/0xc0
[ 1597.193414] ret_from_fork+0x1f/0x30

Steps to reproduce:

1. Plug in the USB-NVME drive enclosure
2. Run mkfs.xfs against a partition on it