From: Christian Brauner <brauner@kernel.org>
To: Josef Bacik <josef@toxicpanda.com>
Cc: linux-fsdevel@vger.kernel.org, linux-btrfs@vger.kernel.org,
kernel-team@fb.com, linux-ext4@vger.kernel.org,
linux-xfs@vger.kernel.org, viro@zeniv.linux.org.uk
Subject: Re: [PATCH 18/50] fs: disallow 0 reference count inodes
Date: Mon, 25 Aug 2025 12:54:01 +0200 [thread overview]
Message-ID: <20250825-person-knapp-e802daccfe5b@brauner> (raw)
In-Reply-To: <6f4fb1baddecbdab4231c6094bbb05a98bbb7365.1755806649.git.josef@toxicpanda.com>
On Thu, Aug 21, 2025 at 04:18:29PM -0400, Josef Bacik wrote:
> Now that we take a full reference for inodes on the LRU, move the logic
> to add the inode to the LRU to before we drop our last reference. This
> allows us to ensure that if the inode has a reference count it can be
> used, and we no longer hold onto inodes that have a 0 reference count.
>
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
> ---
> fs/inode.c | 53 +++++++++++++++++++++++++++++++++--------------------
> 1 file changed, 33 insertions(+), 20 deletions(-)
>
> diff --git a/fs/inode.c b/fs/inode.c
> index de0ec791f9a3..b4145ddbaf8e 100644
> --- a/fs/inode.c
> +++ b/fs/inode.c
> @@ -614,7 +614,7 @@ static void __inode_add_lru(struct inode *inode, bool rotate)
>
> if (inode->i_state & (I_FREEING | I_WILL_FREE))
> return;
> - if (atomic_read(&inode->i_count))
> + if (atomic_read(&inode->i_count) != 1)
> return;
> if (inode->__i_nlink == 0)
> return;
> @@ -1966,28 +1966,11 @@ EXPORT_SYMBOL(generic_delete_inode);
> * in cache if fs is alive, sync and evict if fs is
> * shutting down.
> */
> -static void iput_final(struct inode *inode, bool skip_lru)
> +static void iput_final(struct inode *inode, bool drop)
> {
> - struct super_block *sb = inode->i_sb;
> - const struct super_operations *op = inode->i_sb->s_op;
> unsigned long state;
> - int drop;
>
> WARN_ON(inode->i_state & I_NEW);
> -
> - if (op->drop_inode)
> - drop = op->drop_inode(inode);
> - else
> - drop = generic_drop_inode(inode);
> -
> - if (!drop && !skip_lru &&
> - !(inode->i_state & I_DONTCACHE) &&
> - (sb->s_flags & SB_ACTIVE)) {
> - __inode_add_lru(inode, true);
> - spin_unlock(&inode->i_lock);
> - return;
> - }
> -
> WARN_ON(!list_empty(&inode->i_lru));
>
> state = inode->i_state;
> @@ -2009,8 +1992,29 @@ static void iput_final(struct inode *inode, bool skip_lru)
> evict(inode);
> }
>
> +static bool maybe_add_lru(struct inode *inode, bool skip_lru)
> +{
> + const struct super_operations *op = inode->i_sb->s_op;
> + struct super_block *sb = inode->i_sb;
> + bool drop = false;
> +
> + if (op->drop_inode)
> + drop = op->drop_inode(inode);
> + else
> + drop = generic_drop_inode(inode);
> +
> + if (!drop && !skip_lru &&
> + !(inode->i_state & I_DONTCACHE) &&
> + (sb->s_flags & SB_ACTIVE))
> + __inode_add_lru(inode, true);
> +
> + return drop;
> +}
Can we rewrite this as:
static bool maybe_add_lru(struct inode *inode, bool skip_lru)
{
const struct super_operations *op = inode->i_sb->s_op;
const struct super_block *sb = inode->i_sb;
bool drop = false;
if (op->drop_inode)
drop = op->drop_inode(inode);
else
drop = generic_drop_inode(inode);
if (drop)
return drop;
if (skip_lru)
return drop;
if (inode->i_state & I_DONTCACHE)
return drop;
if (!(sb->s_flags & SB_ACTIVE))
return drop;
__inode_add_lru(inode, true);
return drop;
}
so it's a lot easier to follow. I really dislike munging conditions
together with a bunch of ands and negations mixed in.
And btw for both I_DONTCACHE and !SB_ACTIVE it seems that returning
anything other than false from op->drop_inode() would be a bug probably
a technicality but I find it pretty odd.
Maybe we add a VFS_WARN_ON_ONCE() at least in your local testing to see
whether you see anything that ever hits this case.
> +
> static void __iput(struct inode *inode, bool skip_lru)
> {
> + bool drop;
> +
> if (!inode)
> return;
> BUG_ON(inode->i_state & I_CLEAR);
> @@ -2026,8 +2030,17 @@ static void __iput(struct inode *inode, bool skip_lru)
> }
>
> spin_lock(&inode->i_lock);
> +
> + /*
> + * If we want to keep the inode around on an LRU we will grab a ref to
> + * the inode when we add it to the LRU list, so we can safely drop the
> + * callers reference after this. If we didn't add the inode to the LRU
> + * then the refcount will still be 1 and we can do the final iput.
> + */
> + drop = maybe_add_lru(inode, skip_lru);
> +
> if (atomic_dec_and_test(&inode->i_count))
> - iput_final(inode, skip_lru);
> + iput_final(inode, drop);
> else
> spin_unlock(&inode->i_lock);
>
> --
> 2.49.0
>
next prev parent reply other threads:[~2025-08-25 10:54 UTC|newest]
Thread overview: 85+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-21 20:18 [PATCH 00/50] fs: rework inode reference counting Josef Bacik
2025-08-21 20:18 ` [PATCH 01/50] fs: add an i_obj_count refcount to the inode Josef Bacik
2025-08-21 20:18 ` [PATCH 02/50] fs: make the i_state flags an enum Josef Bacik
2025-08-22 11:08 ` Christian Brauner
2025-08-22 13:31 ` Josef Bacik
2025-08-22 14:36 ` David Sterba
2025-08-22 11:18 ` Sun YangKai
2025-08-22 11:42 ` [PATCH 02/50] " Alan Huang
2025-08-22 12:11 ` Sun YangKai
2025-08-22 14:40 ` [PATCH 02/50] fs: " Josef Bacik
2025-08-21 20:18 ` [PATCH 03/50] fs: hold an i_obj_count reference in wait_sb_inodes Josef Bacik
2025-08-21 20:18 ` [PATCH 04/50] fs: hold an i_obj_count reference for the i_wb_list Josef Bacik
2025-08-22 11:27 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 05/50] fs: hold an i_obj_count reference for the i_io_list Josef Bacik
2025-08-21 20:18 ` [PATCH 06/50] fs: hold an i_obj_count reference in writeback_sb_inodes Josef Bacik
2025-08-22 12:20 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 07/50] fs: hold an i_obj_count reference while on the hashtable Josef Bacik
2025-08-21 20:18 ` [PATCH 08/50] fs: hold an i_obj_count reference while on the LRU list Josef Bacik
2025-08-21 20:18 ` [PATCH 09/50] fs: hold an i_obj_count reference while on the sb inode list Josef Bacik
2025-08-21 20:18 ` [PATCH 10/50] fs: stop accessing ->i_count directly in f2fs and gfs2 Josef Bacik
2025-08-22 12:38 ` (subset) " Christian Brauner
2025-08-21 20:18 ` [PATCH 11/50] fs: hold an i_obj_count when we have an i_count reference Josef Bacik
2025-08-21 20:18 ` [PATCH 12/50] fs: rework iput logic Josef Bacik
2025-08-22 12:54 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 13/50] fs: add an I_LRU flag to the inode Josef Bacik
2025-08-21 20:18 ` [PATCH 14/50] fs: maintain a list of pinned inodes Josef Bacik
2025-08-22 14:55 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 15/50] fs: delete the inode from the LRU list on lookup Josef Bacik
2025-08-22 15:27 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 16/50] fs: change evict_inodes to use iput instead of evict directly Josef Bacik
2025-08-25 9:07 ` Christian Brauner
2025-08-25 19:35 ` Josef Bacik
2025-08-26 9:56 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 17/50] fs: hold a full ref while the inode is on a LRU Josef Bacik
2025-08-25 9:20 ` Christian Brauner
2025-08-25 10:40 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 18/50] fs: disallow 0 reference count inodes Josef Bacik
2025-08-25 10:54 ` Christian Brauner [this message]
2025-08-25 19:26 ` Josef Bacik
2025-08-26 9:28 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 19/50] fs: make evict_inodes add to the dispose list under the i_lock Josef Bacik
2025-08-21 20:18 ` [PATCH 20/50] fs: convert i_count to refcount_t Josef Bacik
2025-08-22 12:10 ` Amir Goldstein
2025-08-22 13:56 ` kernel test robot
2025-08-25 11:03 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 21/50] fs: use refcount_inc_not_zero in igrab Josef Bacik
2025-08-25 11:21 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 22/50] fs: use inode_tryget in find_inode* Josef Bacik
2025-08-25 11:26 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 23/50] fs: update find_inode_*rcu to check the i_count count Josef Bacik
2025-08-25 11:27 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 24/50] fs: use igrab in insert_inode_locked Josef Bacik
2025-08-21 20:18 ` [PATCH 25/50] fs: remove I_WILL_FREE|I_FREEING check from __inode_add_lru Josef Bacik
2025-08-21 20:18 ` [PATCH 26/50] fs: remove I_WILL_FREE|I_FREEING check in inode_pin_lru_isolating Josef Bacik
2025-08-21 20:18 ` [PATCH 27/50] fs: use inode_tryget in evict_inodes Josef Bacik
2025-08-25 11:43 ` Christian Brauner
2025-08-25 18:22 ` Josef Bacik
2025-08-21 20:18 ` [PATCH 28/50] fs: change evict_dentries_for_decrypted_inodes to use refcount Josef Bacik
2025-08-21 20:18 ` [PATCH 29/50] block: use igrab in sync_bdevs Josef Bacik
2025-08-21 20:18 ` [PATCH 30/50] bcachefs: use the refcount instead of I_WILL_FREE|I_FREEING Josef Bacik
2025-08-21 20:18 ` [PATCH 31/50] btrfs: don't check I_WILL_FREE|I_FREEING Josef Bacik
2025-08-21 20:18 ` [PATCH 32/50] fs: use igrab in drop_pagecache_sb Josef Bacik
2025-08-21 20:18 ` [PATCH 33/50] fs: stop checking I_FREEING in d_find_alias_rcu Josef Bacik
2025-08-21 20:18 ` [PATCH 34/50] ext4: stop checking I_WILL_FREE|IFREEING in ext4_check_map_extents_env Josef Bacik
2025-08-21 20:18 ` [PATCH 35/50] fs: remove I_WILL_FREE|I_FREEING from fs-writeback.c Josef Bacik
2025-08-25 11:46 ` Christian Brauner
2025-08-21 20:18 ` [PATCH 36/50] gfs2: remove I_WILL_FREE|I_FREEING usage Josef Bacik
2025-08-21 20:18 ` [PATCH 37/50] fs: remove I_WILL_FREE|I_FREEING check from dquot.c Josef Bacik
2025-08-21 20:18 ` [PATCH 38/50] notify: remove I_WILL_FREE|I_FREEING checks in fsnotify_unmount_inodes Josef Bacik
2025-08-21 20:18 ` [PATCH 39/50] xfs: remove I_FREEING check Josef Bacik
2025-08-21 20:18 ` [PATCH 40/50] landlock: remove I_FREEING|I_WILL_FREE check Josef Bacik
2025-08-21 20:18 ` [PATCH 41/50] fs: change inode_is_dirtytime_only to use refcount Josef Bacik
2025-08-21 20:18 ` [PATCH 42/50] btrfs: remove references to I_FREEING Josef Bacik
2025-08-21 20:18 ` [PATCH 43/50] ext4: remove reference to I_FREEING in inode.c Josef Bacik
2025-08-21 20:18 ` [PATCH 44/50] ext4: remove reference to I_FREEING in orphan.c Josef Bacik
2025-08-21 20:18 ` [PATCH 45/50] pnfs: use i_count refcount to determine if the inode is going away Josef Bacik
2025-08-21 20:18 ` [PATCH 46/50] fs: remove some spurious I_FREEING references in inode.c Josef Bacik
2025-08-21 20:18 ` [PATCH 47/50] xfs: remove reference to I_FREEING|I_WILL_FREE Josef Bacik
2025-08-21 20:18 ` [PATCH 48/50] ocfs2: do not set I_WILL_FREE Josef Bacik
2025-08-21 20:19 ` [PATCH 49/50] fs: remove I_FREEING|I_WILL_FREE Josef Bacik
2025-08-25 11:53 ` Christian Brauner
2025-08-21 20:19 ` [PATCH 50/50] fs: add documentation explaining the reference count rules for inodes Josef Bacik
2025-08-25 11:56 ` Christian Brauner
2025-08-22 10:51 ` [PATCH 00/50] fs: rework inode reference counting Christian Brauner
2025-08-22 13:30 ` Josef Bacik
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250825-person-knapp-e802daccfe5b@brauner \
--to=brauner@kernel.org \
--cc=josef@toxicpanda.com \
--cc=kernel-team@fb.com \
--cc=linux-btrfs@vger.kernel.org \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=viro@zeniv.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).