Re: [PATCH] mm/damon: validate addr_unit to be power of 2

[Liew Rui Yan <aethernet65535@gmail.com>]

Hi SeongJae,

Thank you for the detailed explaination and for pointing out the
'PAGE_SIZE' limitation. You are right; and I should have accounted for
that.

However, I performend further testing to address your confusion
regarding kdamond's termination. My patch didn't fix anything.

Here is my reproduction:

# Log (Both with and without this patch):

	# cd /sys/module/damon_lru_sort/parameters/
	# echo Y > enabled 
	# echo 3 > addr_unit 
	# ps aux | rg "[k]damond"
	root    71  0.0 0.0 0   0   ?   I   22:26   0:00 [kdamond.0]
	# echo Y > commit_inputs 
	bash: echo: write error: Invalid argument
	# cat kdamond_pid 
	71
	# ps aux | rg "[k]damond"
	# ... kdamond terminated

I am very sorry for the noise caused by my misunderstanding of the
sashiko review.

The real question I am facing now is:
Should kdamond terminate itself when damon_commit_ctx() fails due to
invalid user inputs?

Best regards,
Rui Yan