Re: [PATCH 08/11] vfs: inode cache conversion to hash-bl

[Al Viro]

On Wed, Dec 06, 2023 at 05:05:37PM +1100, Dave Chinner wrote:

> +	/*
> +	 * There are some callers that come through here without synchronisation
> +	 * and potentially with multiple references to the inode. Hence we have
> +	 * to handle the case that we might race with a remove and insert to a
> +	 * different list. Coda, in particular, seems to have a userspace API
> +	 * that can directly trigger "unhash/rehash to different list" behaviour
> +	 * without any serialisation at all.
> +	 *
> +	 * Hence we have to handle the situation where the inode->i_hash_head
> +	 * might point to a different list than what we expect, indicating that
> +	 * we raced with another unhash and potentially a new insertion. This
> +	 * means we have to retest the head once we have everything locked up
> +	 * and loop again if it doesn't match.
> +	 */

coda_replace_fid() is an old headache, but it's thankfully unique - nobody else
does that kind of shit (just rechecked).

Note that coda_replace_fid() is not going to have the sucker racing with
removal from another source, and I'm 100% sure that they really want
some serialization for handling those requests.

remove_inode_hash() is misused there - "in the middle of hash key change"
is not the same state as "unhashed".

Any races between insert and unhash are bugs, not something to support.