nfs_common: fix localio to cope with racing nfs_local_probe()

Commit Message

Mike Snitzer Oct. 23, 2024, 8:34 p.m. UTC

Fix the possibility of racing nfs_local_probe() resulting in:
  list_add double add: new=ffff8b99707f9f58, prev=ffff8b99707f9f58, next=ffffffffc0f30000.
  ------------[ cut here ]------------
  kernel BUG at lib/list_debug.c:35!

Add nfs_uuid_init() to properly initialize all nfs_uuid_t members
(particularly its list_head).

Switch to returning bool from nfs_uuid_begin(), returns false if
nfs_uuid_t is already in-use (its list_head is on a list). Update
nfs_local_probe() to return early if the nfs_client's cl_uuid
(nfs_uuid_t) is in-use.

Also, switch nfs_uuid_begin() from using list_add_tail_rcu() to
list_add_tail() -- rculist was used in an earlier version of the
localio code that had a lockless nfs_uuid_lookup interface.

Signed-off-by: Mike Snitzer <[e-mail]>
---
 fs/nfs/client.c            |  3 +--
 fs/nfs/localio.c           |  3 ++-
 fs/nfs_common/nfslocalio.c | 25 +++++++++++++++++++------
 include/linux/nfslocalio.h |  3 ++-
 4 files changed, 24 insertions(+), 10 deletions(-)

Patch

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 114282398716..03ecc7765615 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -181,8 +181,7 @@  struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)
 #if IS_ENABLED(CONFIG_NFS_LOCALIO)
 	seqlock_init(&clp->cl_boot_lock);
 	ktime_get_real_ts64(&clp->cl_nfssvc_boot);
-	clp->cl_uuid.net = NULL;
-	clp->cl_uuid.dom = NULL;
+	nfs_uuid_init(&clp->cl_uuid);
 	spin_lock_init(&clp->cl_localio_lock);
 #endif /* CONFIG_NFS_LOCALIO */
 
diff --git a/fs/nfs/localio.c b/fs/nfs/localio.c
index d0aa680ec816..8f0ce82a677e 100644
--- a/fs/nfs/localio.c
+++ b/fs/nfs/localio.c
@@ -205,7 +205,8 @@  void nfs_local_probe(struct nfs_client *clp)
 		nfs_local_disable(clp);
 	}
 
-	nfs_uuid_begin(&clp->cl_uuid);
+	if (!nfs_uuid_begin(&clp->cl_uuid))
+		return;
 	if (nfs_server_uuid_is_local(clp))
 		nfs_local_enable(clp);
 	nfs_uuid_end(&clp->cl_uuid);
diff --git a/fs/nfs_common/nfslocalio.c b/fs/nfs_common/nfslocalio.c
index 5c8ce5066c16..09404d142d1a 100644
--- a/fs/nfs_common/nfslocalio.c
+++ b/fs/nfs_common/nfslocalio.c
@@ -5,7 +5,7 @@ 
  */
 
 #include <linux/module.h>
-#include <linux/rculist.h>
+#include <linux/list.h>
 #include <linux/nfslocalio.h>
 #include <net/netns/generic.h>
 
@@ -20,15 +20,27 @@  static DEFINE_SPINLOCK(nfs_uuid_lock);
  */
 static LIST_HEAD(nfs_uuids);
 
-void nfs_uuid_begin(nfs_uuid_t *nfs_uuid)
+void nfs_uuid_init(nfs_uuid_t *nfs_uuid)
 {
 	nfs_uuid->net = NULL;
 	nfs_uuid->dom = NULL;
+	INIT_LIST_HEAD(&nfs_uuid->list);
+}
+EXPORT_SYMBOL_GPL(nfs_uuid_init);
+
+bool nfs_uuid_begin(nfs_uuid_t *nfs_uuid)
+{
+	spin_lock(&nfs_uuid_lock);
+	/* Is this nfs_uuid already in use? */
+	if (!list_empty(&nfs_uuid->list)) {
+		spin_unlock(&nfs_uuid_lock);
+		return false;
+	}
 	uuid_gen(&nfs_uuid->uuid);
-
-	spin_lock(&nfs_uuid_lock);
-	list_add_tail_rcu(&nfs_uuid->list, &nfs_uuids);
+	list_add_tail(&nfs_uuid->list, &nfs_uuids);
 	spin_unlock(&nfs_uuid_lock);
+
+	return true;
 }
 EXPORT_SYMBOL_GPL(nfs_uuid_begin);
 
@@ -36,7 +48,8 @@  void nfs_uuid_end(nfs_uuid_t *nfs_uuid)
 {
 	if (nfs_uuid->net == NULL) {
 		spin_lock(&nfs_uuid_lock);
-		list_del_init(&nfs_uuid->list);
+		if (nfs_uuid->net == NULL)
+			list_del_init(&nfs_uuid->list);
 		spin_unlock(&nfs_uuid_lock);
 	}
 }
diff --git a/include/linux/nfslocalio.h b/include/linux/nfslocalio.h
index b0dd9b1eef4f..3982fea79919 100644
--- a/include/linux/nfslocalio.h
+++ b/include/linux/nfslocalio.h
@@ -32,7 +32,8 @@  typedef struct {
 	struct auth_domain *dom; /* auth_domain for localio */
 } nfs_uuid_t;
 
-void nfs_uuid_begin(nfs_uuid_t *);
+void nfs_uuid_init(nfs_uuid_t *);
+bool nfs_uuid_begin(nfs_uuid_t *);
 void nfs_uuid_end(nfs_uuid_t *);
 void nfs_uuid_is_local(const uuid_t *, struct list_head *,
 		       struct net *, struct auth_domain *, struct module *);