From: Jan Kara <jack@suse.cz>
To: Paul Moore <paul@paul-moore.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>,
linux-audit@redhat.com, <linux-fsdevel@vger.kernel.org>,
rgb@redhat.com, Amir Goldstein <amir73il@gmail.com>,
Jan Kara <jack@suse.cz>
Subject: [PATCH 05/11] audit: Make hash table insertion safe against concurrent lookups
Date: Tue, 4 Sep 2018 18:06:25 +0200 [thread overview]
Message-ID: <20180904160632.21210-6-jack@suse.cz> (raw)
In-Reply-To: <20180904160632.21210-1-jack@suse.cz>
Currently, the audit tree code does not make sure that when a chunk is
inserted into the hash table, it is fully initialized. So in theory a
user of RCU lookup could see uninitialized structure in the hash table
and crash. Add appropriate barriers between initialization of the
structure and its insertion into hash table.
Signed-off-by: Jan Kara <jack@suse.cz>
---
kernel/audit_tree.c | 32 +++++++++++++++++++++++++++++---
1 file changed, 29 insertions(+), 3 deletions(-)
diff --git a/kernel/audit_tree.c b/kernel/audit_tree.c
index bac5dd90c629..307749d6773c 100644
--- a/kernel/audit_tree.c
+++ b/kernel/audit_tree.c
@@ -186,6 +186,12 @@ static void insert_hash(struct audit_chunk *chunk)
if (!(chunk->mark.flags & FSNOTIFY_MARK_FLAG_ATTACHED))
return;
+ /*
+ * Make sure chunk is fully initialized before making it visible in the
+ * hash. Pairs with a data dependency barrier in READ_ONCE() in
+ * audit_tree_lookup().
+ */
+ smp_wmb();
WARN_ON_ONCE(!chunk->key);
list = chunk_hash(chunk->key);
list_add_rcu(&chunk->hash, list);
@@ -199,7 +205,11 @@ struct audit_chunk *audit_tree_lookup(const struct inode *inode)
struct audit_chunk *p;
list_for_each_entry_rcu(p, list, hash) {
- if (p->key == key) {
+ /*
+ * We use a data dependency barrier in READ_ONCE() to make sure
+ * the chunk we see is fully initialized.
+ */
+ if (READ_ONCE(p->key) == key) {
atomic_long_inc(&p->refs);
return p;
}
@@ -304,9 +314,15 @@ static void untag_chunk(struct node *p)
list_replace_init(&chunk->owners[j].list, &new->owners[i].list);
}
- list_replace_rcu(&chunk->hash, &new->hash);
list_for_each_entry(owner, &new->trees, same_root)
owner->root = new;
+ /*
+ * Make sure chunk is fully initialized before making it visible in the
+ * hash. Pairs with a data dependency barrier in READ_ONCE() in
+ * audit_tree_lookup().
+ */
+ smp_wmb();
+ list_replace_rcu(&chunk->hash, &new->hash);
spin_unlock(&hash_lock);
fsnotify_detach_mark(entry);
mutex_unlock(&entry->group->mark_mutex);
@@ -368,6 +384,10 @@ static int create_chunk(struct inode *inode, struct audit_tree *tree)
list_add(&tree->same_root, &chunk->trees);
}
chunk->key = inode_to_key(inode);
+ /*
+ * Inserting into the hash table has to go last as once we do that RCU
+ * readers can see the chunk.
+ */
insert_hash(chunk);
spin_unlock(&hash_lock);
mutex_unlock(&audit_tree_group->mark_mutex);
@@ -459,7 +479,6 @@ static int tag_chunk(struct inode *inode, struct audit_tree *tree)
p->owner = tree;
get_tree(tree);
list_add(&p->list, &tree->chunks);
- list_replace_rcu(&old->hash, &chunk->hash);
list_for_each_entry(owner, &chunk->trees, same_root)
owner->root = chunk;
old->dead = 1;
@@ -467,6 +486,13 @@ static int tag_chunk(struct inode *inode, struct audit_tree *tree)
tree->root = chunk;
list_add(&tree->same_root, &chunk->trees);
}
+ /*
+ * Make sure chunk is fully initialized before making it visible in the
+ * hash. Pairs with a data dependency barrier in READ_ONCE() in
+ * audit_tree_lookup().
+ */
+ smp_wmb();
+ list_replace_rcu(&old->hash, &chunk->hash);
spin_unlock(&hash_lock);
fsnotify_detach_mark(old_entry);
mutex_unlock(&audit_tree_group->mark_mutex);
--
2.16.4
next prev parent reply other threads:[~2018-09-04 20:32 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-04 16:06 [PATCH 0/11 v3] audit: Fix various races when tagging and untagging mounts Jan Kara
2018-09-04 16:06 ` [PATCH 01/11] audit_tree: Remove mark->lock locking Jan Kara
2018-09-04 16:06 ` [PATCH 02/11] audit: Fix possible spurious -ENOSPC error Jan Kara
2018-09-04 16:06 ` [PATCH 03/11] audit: Fix possible tagging failures Jan Kara
2018-09-04 16:06 ` [PATCH 04/11] audit: Embed key into chunk Jan Kara
2018-09-13 20:06 ` Richard Guy Briggs
2018-09-04 16:06 ` Jan Kara [this message]
2018-09-04 16:06 ` [PATCH 06/11] audit: Factor out chunk replacement code Jan Kara
2018-09-04 16:06 ` [PATCH 07/11] audit: Remove pointless check in insert_hash() Jan Kara
2018-09-04 16:06 ` [PATCH 08/11] audit: Provide helper for dropping mark's chunk reference Jan Kara
2018-09-04 16:06 ` [PATCH 09/11] audit: Allocate fsnotify mark independently of chunk Jan Kara
2018-09-14 14:09 ` Richard Guy Briggs
2018-09-17 16:46 ` Jan Kara
2018-10-03 22:11 ` Paul Moore
2018-10-03 22:08 ` Paul Moore
2018-10-03 22:39 ` Richard Guy Briggs
2018-10-04 6:57 ` Jan Kara
2018-09-04 16:06 ` [PATCH 10/11] audit: Replace chunk attached to mark instead of replacing mark Jan Kara
2018-09-04 16:06 ` [PATCH 11/11] audit: Use 'mark' name for fsnotify_mark variables Jan Kara
2018-09-14 18:29 ` Richard Guy Briggs
2018-09-17 16:44 ` Jan Kara
2018-09-17 18:13 ` Richard Guy Briggs
2018-09-04 16:06 ` [PATCH 12/11 TESTSUITE] audit_testsuite: Add stress test for tree watches Jan Kara
2018-09-14 18:21 ` Richard Guy Briggs
2018-09-17 16:56 ` Jan Kara
2018-10-05 21:06 ` Paul Moore
2018-10-09 7:40 ` Jan Kara
2018-10-10 6:43 ` Paul Moore
2018-10-11 11:39 ` Jan Kara
2018-10-11 23:03 ` Paul Moore
2018-10-15 10:04 ` Jan Kara
2018-10-15 15:39 ` Paul Moore
2018-10-17 10:09 ` Jan Kara
2018-11-14 0:34 ` Paul Moore
2018-11-14 12:16 ` Jan Kara
2018-11-19 15:19 ` Paul Moore
2018-09-14 19:13 ` [PATCH 0/11 v3] audit: Fix various races when tagging and untagging mounts Richard Guy Briggs
2018-09-17 16:57 ` Jan Kara
2018-10-04 1:20 ` Paul Moore
2018-10-04 6:59 ` Jan Kara
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=20180904160632.21210-6-jack@suse.cz \
--to=jack@suse.cz \
--cc=amir73il@gmail.com \
--cc=linux-audit@redhat.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=paul@paul-moore.com \
--cc=rgb@redhat.com \
--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).