On 7/30/25 8:37 AM, Bart Van Assche wrote:
On 7/30/25 5:55 AM, Peter Wang (王信友) wrote:
However, in theory, this issue should still be solvable
without using a lock.
Another idea is to only start ufshcd_devfreq_init
when shost->async_scan = 0.
Does the lockdep complaint mentioned in this email thread occur on
multiple platforms or only on MediaTek platforms? I don't see any
lockdep complaints with Martin's SCSI for-next branch on my
development platform. If this warning only occurs on MediaTek
platforms, why to modify the UFSHCI core driver to eliminate this
lockdep complaint?
Answering my own question: clock scaling is not enabled on my
development platform and that's why I'm not seeing this lockdep
complaint.
Bart.
