Re: 6.6 kernel block: blk_mq_freeze_queue_wait in suspend path but userspace task held the queue->q_usage_counter in 'TASK_FROZEN' state

[Bart Van Assche <bvanassche@xxxxxxx>]

On 6/19/24 11:53 PM, Kassey Li wrote:
> hello, linux block team:

Please repost this message on the linux-scsi mailing list. I think this
is a UFSHCD driver issue rather than a block layer issue.

> userspace task A  ['TASK_FROZEN']
>
>      [<ffffffdc0c527f10>] __switch_to+0x1e8
>      [<ffffffdc0c5287ec>] __schedule+0x6cc
>      [<ffffffdc0c528c28>] schedule+0x78
>      [<ffffffdc0c532bbc>] schedule_timeout+0x50
>      [<ffffffdc0c529e48>] do_wait_for_common+0x10c
>      [<ffffffdc0c529238>] wait_for_completion+0x48
>      [<ffffffdc0b4def4c>] __flush_work+0xcc
>      [<ffffffdc0b4dee70>] flush_work+0x14
>      [<ffffffdc0c091aec>] ufshcd_hold+0xc0

Is ufshcd_hold() executing flush_work(&hba->clk_gating.ungate_work)?
If so, why does the ungate work not complete?

Thanks,

Bart.