Re: [RFC][PATCH 06/10] fs: Rework i_count

[Al Viro]

On Fri, Feb 24, 2017 at 04:43:35PM +0100, Peter Zijlstra wrote:

>  {
> -	return atomic_read(&inode->i_count);
> +	int i_count = atomic_read(&inode->i_count);
> +
> +	/*
> +	 * In order to preserve the 'old' usage-count semantics, remove the
> +	 * reference that the hash-table has.

What does it have to do with hashtable, when you are bumping it for _all_
inodes, hashed or not hashed?