From: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
We don't want open-by-handle picking half-set-up in-core
struct inode from e.g. mkdir() having failed halfway through.
In other words, we don't want such inodes returned by iget_locked()
on their way to extinction. However, we can't just have them
unhashed - otherwise open-by-handle immediately *after* that would've
ended up creating a new in-core inode over the on-disk one that
is in process of being freed right under us.
Solution: new flag (I_CREATING) set by insert_inode_locked() and
removed by unlock_new_inode() and a new primitive (discard_new_inode())
to be used by such halfway-through-setup failure exits instead of
unlock_new_inode() / iput() combinations. That primitive unlocks new
inode, but leaves I_CREATING in place.
iget_locked() treats finding an I_CREATING inode as failure
(-ESTALE, once we sort out the error propagation).
insert_inode_locked() treats the same as instant -EBUSY.
ilookup() treats those as icache miss.
Signed-off-by: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
---
fs/inode.c | 44 ++++++++++++++++++++++++++++++++++++++++----
include/linux/fs.h | 6 +++++-
2 files changed, 45 insertions(+), 5 deletions(-)
diff --git a/fs/inode.c b/fs/inode.c
index 2c300e981796..04dd7e0d5142 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -804,6 +804,10 @@ static struct inode *find_inode(struct super_block *sb,
__wait_on_freeing_inode(inode);
goto repeat;
}
+ if (unlikely(inode->i_state & I_CREATING)) {
+ spin_unlock(&inode->i_lock);
+ return ERR_PTR(-ESTALE);
+ }
__iget(inode);
spin_unlock(&inode->i_lock);
return inode;
@@ -831,6 +835,10 @@ static struct inode *find_inode_fast(struct super_block *sb,
__wait_on_freeing_inode(inode);
goto repeat;
}
+ if (unlikely(inode->i_state & I_CREATING)) {
+ spin_unlock(&inode->i_lock);
+ return ERR_PTR(-ESTALE);
+ }
__iget(inode);
spin_unlock(&inode->i_lock);
return inode;
@@ -961,13 +969,26 @@ void unlock_new_inode(struct inode *inode)
lockdep_annotate_inode_mutex_key(inode);
spin_lock(&inode->i_lock);
WARN_ON(!(inode->i_state & I_NEW));
- inode->i_state &= ~I_NEW;
+ inode->i_state &= ~I_NEW & ~I_CREATING;
smp_mb();
wake_up_bit(&inode->i_state, __I_NEW);
spin_unlock(&inode->i_lock);
}
EXPORT_SYMBOL(unlock_new_inode);
+void discard_new_inode(struct inode *inode)
+{
+ lockdep_annotate_inode_mutex_key(inode);
+ spin_lock(&inode->i_lock);
+ WARN_ON(!(inode->i_state & I_NEW));
+ inode->i_state &= ~I_NEW;
+ smp_mb();
+ wake_up_bit(&inode->i_state, __I_NEW);
+ spin_unlock(&inode->i_lock);
+ iput(inode);
+}
+EXPORT_SYMBOL(discard_new_inode);
+
/**
* lock_two_nondirectories - take two i_mutexes on non-directory objects
*
@@ -1039,6 +1060,8 @@ struct inode *inode_insert5(struct inode *inode, unsigned long hashval,
* Use the old inode instead of the preallocated one.
*/
spin_unlock(&inode_hash_lock);
+ if (IS_ERR(old))
+ return NULL;
wait_on_inode(old);
if (unlikely(inode_unhashed(old))) {
iput(old);
@@ -1128,6 +1151,8 @@ struct inode *iget_locked(struct super_block *sb, unsigned long ino)
inode = find_inode_fast(sb, head, ino);
spin_unlock(&inode_hash_lock);
if (inode) {
+ if (IS_ERR(inode))
+ return NULL;
wait_on_inode(inode);
if (unlikely(inode_unhashed(inode))) {
iput(inode);
@@ -1165,6 +1190,8 @@ struct inode *iget_locked(struct super_block *sb, unsigned long ino)
*/
spin_unlock(&inode_hash_lock);
destroy_inode(inode);
+ if (IS_ERR(old))
+ return NULL;
inode = old;
wait_on_inode(inode);
if (unlikely(inode_unhashed(inode))) {
@@ -1282,7 +1309,7 @@ struct inode *ilookup5_nowait(struct super_block *sb, unsigned long hashval,
inode = find_inode(sb, head, test, data);
spin_unlock(&inode_hash_lock);
- return inode;
+ return IS_ERR(inode) ? NULL : inode;
}
EXPORT_SYMBOL(ilookup5_nowait);
@@ -1338,6 +1365,8 @@ struct inode *ilookup(struct super_block *sb, unsigned long ino)
spin_unlock(&inode_hash_lock);
if (inode) {
+ if (IS_ERR(inode))
+ return NULL;
wait_on_inode(inode);
if (unlikely(inode_unhashed(inode))) {
iput(inode);
@@ -1421,12 +1450,16 @@ int insert_inode_locked(struct inode *inode)
}
if (likely(!old)) {
spin_lock(&inode->i_lock);
- inode->i_state |= I_NEW;
+ inode->i_state |= I_NEW | I_CREATING;
hlist_add_head(&inode->i_hash, head);
spin_unlock(&inode->i_lock);
spin_unlock(&inode_hash_lock);
return 0;
}
+ if (unlikely(old->i_state & I_CREATING)) {
+ spin_unlock(&old->i_lock);
+ return -EBUSY;
+ }
__iget(old);
spin_unlock(&old->i_lock);
spin_unlock(&inode_hash_lock);
@@ -1443,7 +1476,10 @@ EXPORT_SYMBOL(insert_inode_locked);
int insert_inode_locked4(struct inode *inode, unsigned long hashval,
int (*test)(struct inode *, void *), void *data)
{
- struct inode *old = inode_insert5(inode, hashval, test, NULL, data);
+ struct inode *old;
+
+ inode->i_state |= I_CREATING;
+ old = inode_insert5(inode, hashval, test, NULL, data);
if (old != inode) {
iput(old);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 5c91108846db..a42600565925 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2016,6 +2016,8 @@ static inline void init_sync_kiocb(struct kiocb *kiocb, struct file *filp)
* I_OVL_INUSE Used by overlayfs to get exclusive ownership on upper
* and work dirs among overlayfs mounts.
*
+ * I_CREATING New object's inode in the middle of setting up.
+ *
* Q: What is the difference between I_WILL_FREE and I_FREEING?
*/
#define I_DIRTY_SYNC (1 << 0)
@@ -2036,7 +2038,8 @@ static inline void init_sync_kiocb(struct kiocb *kiocb, struct file *filp)
#define __I_DIRTY_TIME_EXPIRED 12
#define I_DIRTY_TIME_EXPIRED (1 << __I_DIRTY_TIME_EXPIRED)
#define I_WB_SWITCH (1 << 13)
-#define I_OVL_INUSE (1 << 14)
+#define I_OVL_INUSE (1 << 14)
+#define I_CREATING (1 << 15)
#define I_DIRTY_INODE (I_DIRTY_SYNC | I_DIRTY_DATASYNC)
#define I_DIRTY (I_DIRTY_INODE | I_DIRTY_PAGES)
@@ -2919,6 +2922,7 @@ extern void lockdep_annotate_inode_mutex_key(struct inode *inode);
static inline void lockdep_annotate_inode_mutex_key(struct inode *inode) { };
#endif
extern void unlock_new_inode(struct inode *);
+extern void discard_new_inode(struct inode *);
extern unsigned int get_next_ino(void);
extern void evict_inodes(struct super_block *sb);
--
2.11.0
