From: Eric Biggers <ebiggers@kernel.org>
To: fsverity@lists.linux.dev
Cc: linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
Christoph Hellwig <hch@lst.de>,
Eric Biggers <ebiggers@kernel.org>,
stable@vger.kernel.org
Subject: [PATCH] fsverity: RCU-delay the freeing of struct fsverity_info
Date: Thu, 1 Oct 2026 10:13:49 -0700 [thread overview]
Message-ID: <20261001171349.81454-1-ebiggers@kernel.org> (raw)
__fsverity_get_info() calls rhashtable_lookup_fast(), which just uses
rcu_read_lock() and doesn't directly synchronize with
fsverity_remove_info().
For the same inode this isn't a problem: its fsverity_info is removed
only at inode eviction time or upon failure to enable verity, when the
inode no longer needs its fsverity_info and it will no longer be
accessed via that inode.
However, this is broken and can cause a use-after-free for concurrent
__fsverity_get_info() for *different* inodes. Those rely on following
fsverity_info::rhash_head in the rhashtable under rcu_read_lock() only.
They also use fsverity_info::inode to do the key comparison.
Fix this by RCU-delaying the freeing of 'struct fsverity_info' after
it's been removed from the rhashtable.
Fixes: f77f281b6118 ("fsverity: use a hashtable to find the fsverity_info")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
---
fs/verity/fsverity_private.h | 15 ++++++++++++++-
fs/verity/open.c | 13 ++++++++++++-
2 files changed, 26 insertions(+), 2 deletions(-)
diff --git a/fs/verity/fsverity_private.h b/fs/verity/fsverity_private.h
index 881d46f25e08..6f85f63861aa 100644
--- a/fs/verity/fsverity_private.h
+++ b/fs/verity/fsverity_private.h
@@ -11,6 +11,7 @@
#define pr_fmt(fmt) "fs-verity: " fmt
#include <linux/fsverity.h>
+#include <linux/rcupdate.h>
#include <linux/rhashtable.h>
/*
@@ -76,7 +77,19 @@ struct merkle_tree_params {
struct fsverity_info {
struct rhash_head rhash_head;
struct merkle_tree_params tree_params;
- u8 root_hash[FS_VERITY_MAX_DIGEST_SIZE];
+ union {
+ u8 root_hash[FS_VERITY_MAX_DIGEST_SIZE];
+
+ /*
+ * Used for final freeing of the struct, in which case root_hash
+ * is no longer needed. Specifically, the only fields needed
+ * besides the rcu_head are those needed by fsverity_free_info()
+ * itself, and those that can be accessed by fsverity_get_info()
+ * (rhash_head and inode) during the RCU-protected rhashtable
+ * chain search trying to find another inode's fsverity_info.
+ */
+ struct rcu_head rcu_head;
+ };
u8 file_digest[FS_VERITY_MAX_DIGEST_SIZE];
struct inode *inode;
unsigned long *hash_block_verified;
diff --git a/fs/verity/open.c b/fs/verity/open.c
index d0c56a7faa3b..52c8cd014921 100644
--- a/fs/verity/open.c
+++ b/fs/verity/open.c
@@ -399,11 +399,22 @@ void fsverity_free_info(struct fsverity_info *vi)
kmem_cache_free(fsverity_info_cachep, vi);
}
+static void fsverity_free_info_rcu(struct rcu_head *head)
+{
+ fsverity_free_info(container_of(head, struct fsverity_info, rcu_head));
+}
+
void fsverity_remove_info(struct fsverity_info *vi)
{
rhashtable_remove_fast(&fsverity_info_hash, &vi->rhash_head,
fsverity_info_hash_params);
- fsverity_free_info(vi);
+ /*
+ * The freeing must be RCU-delayed because __fsverity_get_info() for a
+ * *different* inode's fsverity_info in the same rhashtable hash chain
+ * can still be accessing *this* fsverity_info's rhash_head and inode
+ * fields as part of the RCU-protected hash chain search.
+ */
+ call_rcu(&vi->rcu_head, fsverity_free_info_rcu);
}
void fsverity_cleanup_inode(struct inode *inode)
base-commit: 551c722f40809618230001baccf219193e22fc5a
--
2.56.0
next reply other threads:[~2026-10-01 17:14 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-01 17:13 Eric Biggers [this message]
2026-10-01 21:23 ` [PATCH] fsverity: RCU-delay the freeing of struct fsverity_info Sandeep Dhavale
2026-10-02 21:45 ` Eric Biggers
2026-10-05 8:43 ` Christoph Hellwig
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=20261001171349.81454-1-ebiggers@kernel.org \
--to=ebiggers@kernel.org \
--cc=fsverity@lists.linux.dev \
--cc=hch@lst.de \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.