nfs_complete_rename() calls d_move() without i_mutex

PATCH Superseded Benjamin Coddington 4 participants 10 messages Jun 15, 2017

Subsystems (2)

NFS client NFS server

lore

Share & tools

lore

Benjamin Coddington

author

2017-06-15 17:31 UTC #0 lore raw

cover nfs_complete_rename() calls d_move() without i_mutex

Once commit 920b4530fb80430ff30ef83efe21ba1fa5623731 "NFS: nfs_rename() handle -ERESTARTSYS dentry left behind" moved the local d_move() into the RPC asyncronous context, d_move() could be called without holding the directories' i_mutex.

Let's revert that commit, and a follow-up fix for it in 1/2, and then fix the original problem once more by forcing a revalidation of the old and new directories if we notice that the rename was interrupted in 2/2.

Benjamin Coddington (2):
Revert "NFS: nfs_rename() handle -ERESTARTSYS dentry left behind"
NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

 fs/nfs/dir.c            | 47 ++++++++++++++++++++---------------------------
 fs/nfs/unlink.c         |  7 +++++++
 include/linux/nfs_xdr.h |  1 +
 3 files changed, 28 insertions(+), 27 deletions(-)

Signature · 1 line

2.9.3

Benjamin Coddington

author

2017-06-15 17:31 UTC #3 lore raw

PATCH 2/2 NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

An interrupted rename will leave the old dentry behind if the rename succeeds. Fix this by forcing a lookup the next time through ->d_revalidate.

A previous attempt at solving this problem took the approach to complete the work of the rename asynchronously, however that approach was wrong since it would allow the d_move() to occur after the directory's i_mutex had been dropped by the original process.

Signed-off-by Benjamin Coddington <[email protected]>

---
 fs/nfs/dir.c            | 2 ++
 fs/nfs/unlink.c         | 7 +++++++
 include/linux/nfs_xdr.h | 1 +
 3 files changed, 10 insertions(+)

fs/nfs/dir.c +2 -0

index 1faf337b316f..bb697e5c3f6c 100644
@@ -2037,6 +2037,8 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	error = rpc_wait_for_completion_task(task);
 	if (error == 0)
 		error = task->tk_status;
+	else
+		((struct nfs_renamedata *)task->tk_calldata)->cancelled = 1;
 	rpc_put_task(task);
 	nfs_mark_for_revalidate(old_inode);
 out:

fs/nfs/unlink.c +7 -0

index 191aa577dd1f..b47158a34879 100644
@@ -288,6 +288,13 @@ static void nfs_async_rename_release(void *calldata)
 	if (d_really_is_positive(data->old_dentry))
 		nfs_mark_for_revalidate(d_inode(data->old_dentry));
 
+	/* The result of the rename is unknown. Play it safe by
+	 * forcing a new lookup */
+	if (data->cancelled) {
+		nfs_force_lookup_revalidate(data->old_dir);
+		nfs_force_lookup_revalidate(data->new_dir);
+	}
+
 	dput(data->old_dentry);
 	dput(data->new_dentry);
 	iput(data->old_dir);

include/linux/nfs_xdr.h +1 -0

index b28c83475ee8..2a8406b4b353 100644
@@ -1533,6 +1533,7 @@ struct nfs_renamedata {
 	struct nfs_fattr	new_fattr;
 	void (*complete)(struct rpc_task *, struct nfs_renamedata *);
 	long timeout;
+	int cancelled;
 };
 
 struct nfs_access_entry;

Signature · 1 line

2.9.3

Benjamin Coddington

author

2017-06-15 17:31 UTC #1 lore raw

PATCH 1/2 Revert "NFS: nfs_rename() handle -ERESTARTSYS dentry left behind"

In reply to Benjamin Coddington

This reverts commit 920b4530fb80430ff30ef83efe21ba1fa5623731 which could call d_move() without holding the directory's i_mutex, and reverts commit d4ea7e3c5c0e341c15b073016dbf3ab6c65f12f3 "NFS: Fix old dentry rehash after move", which was a follow-up fix.

Fixes 920b4530fb80 ("NFS: nfs_rename() handle -ERESTARTSYS dentry left behind")

Cc [email protected] # v4.10+

Signed-off-by Benjamin Coddington <[email protected]>

---
 fs/nfs/dir.c | 45 ++++++++++++++++++---------------------------
 1 file changed, 18 insertions(+), 27 deletions(-)

fs/nfs/dir.c +18 ~~-27~~

index 32ccd7754f8a..1faf337b316f 100644
@@ -1946,29 +1946,6 @@ nfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
 }
 EXPORT_SYMBOL_GPL(nfs_link);
 
-static void
-nfs_complete_rename(struct rpc_task *task, struct nfs_renamedata *data)
-{
-	struct dentry *old_dentry = data->old_dentry;
-	struct dentry *new_dentry = data->new_dentry;
-	struct inode *old_inode = d_inode(old_dentry);
-	struct inode *new_inode = d_inode(new_dentry);
-
-	nfs_mark_for_revalidate(old_inode);
-
-	switch (task->tk_status) {
-	case 0:
-		if (new_inode != NULL)
-			nfs_drop_nlink(new_inode);
-		d_move(old_dentry, new_dentry);
-		nfs_set_verifier(new_dentry,
-					nfs_save_change_attribute(data->new_dir));
-		break;
-	case -ENOENT:
-		nfs_dentry_handle_enoent(old_dentry);
-	}
-}
-
 /*
  * RENAME
  * FIXME: Some nfsds, like the Linux user space nfsd, may generate a

@@ -1999,7 +1976,7 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 {
 	struct inode *old_inode = d_inode(old_dentry);
 	struct inode *new_inode = d_inode(new_dentry);
-	struct dentry *dentry = NULL;
+	struct dentry *dentry = NULL, *rehash = NULL;
 	struct rpc_task *task;
 	int error = -EBUSY;

@@ -2022,8 +1999,10 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		 * To prevent any new references to the target during the
 		 * rename, we unhash the dentry in advance.
 		 */
-		if (!d_unhashed(new_dentry))
+		if (!d_unhashed(new_dentry)) {
 			d_drop(new_dentry);
+			rehash = new_dentry;
+		}
 
 		if (d_count(new_dentry) > 2) {
 			int err;

@@ -2040,6 +2019,7 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 				goto out;
 
 			new_dentry = dentry;
+			rehash = NULL;
 			new_inode = NULL;
 		}
 	}

@@ -2048,8 +2028,7 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (new_inode != NULL)
 		NFS_PROTO(new_inode)->return_delegation(new_inode);
 
-	task = nfs_async_rename(old_dir, new_dir, old_dentry, new_dentry,
-					nfs_complete_rename);
+	task = nfs_async_rename(old_dir, new_dir, old_dentry, new_dentry, NULL);
 	if (IS_ERR(task)) {
 		error = PTR_ERR(task);
 		goto out;

@@ -2059,9 +2038,21 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (error == 0)
 		error = task->tk_status;
 	rpc_put_task(task);
+	nfs_mark_for_revalidate(old_inode);
 out:
+	if (rehash)
+		d_rehash(rehash);
 	trace_nfs_rename_exit(old_dir, old_dentry,
 			new_dir, new_dentry, error);
+	if (!error) {
+		if (new_inode != NULL)
+			nfs_drop_nlink(new_inode);
+		d_move(old_dentry, new_dentry);
+		nfs_set_verifier(new_dentry,
+					nfs_save_change_attribute(new_dir));
+	} else if (error == -ENOENT)
+		nfs_dentry_handle_enoent(old_dentry);
+
 	/* new dentry created? */
 	if (dentry)
 		dput(dentry);

Signature · 1 line

2.9.3

Jeff Layton

Reviewed-by on v1 1/2

2017-06-15 18:16 UTC #2 lore raw

Re: [PATCH 1/2] Revert "NFS: nfs_rename() handle -ERESTARTSYS dentry left behind"

Quoted 97 lines · Benjamin Coddington

Fixes 920b4530fb80 ("NFS: nfs_rename() handle -ERESTARTSYS dentry left behind")

Cc [email protected] # v4.10+

Signed-off-by Benjamin Coddington <[email protected]>

---
 fs/nfs/dir.c | 45 ++++++++++++++++++---------------------------
 1 file changed, 18 insertions(+), 27 deletions(-)

fs/nfs/dir.c +11 ~~-27~~

index 32ccd7754f8a..1faf337b316f 100644
@@ -1946,29 +1946,6 @@ nfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
 }
 EXPORT_SYMBOL_GPL(nfs_link);
 
-static void
-nfs_complete_rename(struct rpc_task *task, struct nfs_renamedata *data)
-{
-	struct dentry *old_dentry = data->old_dentry;
-	struct dentry *new_dentry = data->new_dentry;
-	struct inode *old_inode = d_inode(old_dentry);
-	struct inode *new_inode = d_inode(new_dentry);
-
-	nfs_mark_for_revalidate(old_inode);
-
-	switch (task->tk_status) {
-	case 0:
-		if (new_inode != NULL)
-			nfs_drop_nlink(new_inode);
-		d_move(old_dentry, new_dentry);
-		nfs_set_verifier(new_dentry,
-					nfs_save_change_attribute(data->new_dir));
-		break;
-	case -ENOENT:
-		nfs_dentry_handle_enoent(old_dentry);
-	}
-}
-
 /*
  * RENAME
  * FIXME: Some nfsds, like the Linux user space nfsd, may generate a

@@ -1999,7 +1976,7 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 {
 	struct inode *old_inode = d_inode(old_dentry);
 	struct inode *new_inode = d_inode(new_dentry);
-	struct dentry *dentry = NULL;
+	struct dentry *dentry = NULL, *rehash = NULL;
 	struct rpc_task *task;
 	int error = -EBUSY;

@@ -2022,8 +1999,10 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		 * To prevent any new references to the target during the
 		 * rename, we unhash the dentry in advance.
 		 */
-		if (!d_unhashed(new_dentry))
+		if (!d_unhashed(new_dentry)) {
 			d_drop(new_dentry);
+			rehash = new_dentry;
+		}
 
 		if (d_count(new_dentry) > 2) {
 			int err;

@@ -2040,6 +2019,7 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 				goto out;
 
 			new_dentry = dentry;
+			rehash = NULL;
 			new_inode = NULL;
 		}
 	}

@@ -2048,8 +2028,7 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (new_inode != NULL)
 		NFS_PROTO(new_inode)->return_delegation(new_inode);
 
-	task = nfs_async_rename(old_dir, new_dir, old_dentry, new_dentry,
-					nfs_complete_rename);
+	task = nfs_async_rename(old_dir, new_dir, old_dentry, new_dentry, NULL);
 	if (IS_ERR(task)) {
 		error = PTR_ERR(task);
 		goto out;

@@ -2059,9 +2038,21 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (error == 0)
 		error = task->tk_status;
 	rpc_put_task(task);
+	nfs_mark_for_revalidate(old_inode);
 out:
+	if (rehash)
+		d_rehash(rehash);
 	trace_nfs_rename_exit(old_dir, old_dentry,
 			new_dir, new_dentry, error);
+	if (!error) {
+		if (new_inode != NULL)

fs/nfs/dir.c int nfs_rename(struct inode *old_dir, struct dentry *old_dentry, · line 2058

Quoted 10 lines

fs/nfs/dir.c +7 -0

+			nfs_drop_nlink(new_inode);
+		d_move(old_dentry, new_dentry);
+		nfs_set_verifier(new_dentry,
+					nfs_save_change_attribute(new_dir));
+	} else if (error == -ENOENT)
+		nfs_dentry_handle_enoent(old_dentry);
+
 	/* new dentry created? */
 	if (dentry)
 		dput(dentry);

Reviewed-by Jeff Layton <[email protected]>

Jeff Layton

2017-06-15 18:18 UTC #4 lore raw

Re: [PATCH 2/2] NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

In reply to Benjamin Coddington

Quoted 21 lines · Benjamin Coddington

An interrupted rename will leave the old dentry behind if the rename succeeds. Fix this by forcing a lookup the next time through ->d_revalidate.
A previous attempt at solving this problem took the approach to complete the work of the rename asynchronously, however that approach was wrong since it would allow the d_move() to occur after the directory's i_mutex had been dropped by the original process.
Signed-off-by Benjamin Coddington <[email protected]>
---
 fs/nfs/dir.c            | 2 ++
 fs/nfs/unlink.c         | 7 +++++++
 include/linux/nfs_xdr.h | 1 +
 3 files changed, 10 insertions(+)
fs/nfs/dir.c +0 -0
index 1faf337b316f..bb697e5c3f6c 100644

fs/nfs/dir.c int nfs_rename(struct inode *old_dir, struct dentry *old_dentry, · line 2041

Quoted 5 lines

fs/nfs/dir.c +2 -0

@@ -2037,6 +2037,8 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	error = rpc_wait_for_completion_task(task);
 	if (error == 0)
 		error = task->tk_status;
+	else
+		((struct nfs_renamedata *)task->tk_calldata)->cancelled = 1;

This looks a bit racy. You could end up setting cancelled just after the reply comes in and the completion callback checks it. I think that you probably either want to do this with an atomic operation or under a lock of some sort.

You could probably do it with an xchg() operation?

Quoted 26 lines

fs/nfs/dir.c +0 -0

 	rpc_put_task(task);
 	nfs_mark_for_revalidate(old_inode);
 out:

fs/nfs/unlink.c +7 -0

index 191aa577dd1f..b47158a34879 100644
@@ -288,6 +288,13 @@ static void nfs_async_rename_release(void *calldata)
 	if (d_really_is_positive(data->old_dentry))
 		nfs_mark_for_revalidate(d_inode(data->old_dentry));
 
+	/* The result of the rename is unknown. Play it safe by
+	 * forcing a new lookup */
+	if (data->cancelled) {
+		nfs_force_lookup_revalidate(data->old_dir);
+		nfs_force_lookup_revalidate(data->new_dir);
+	}
+
 	dput(data->old_dentry);
 	dput(data->new_dentry);
 	iput(data->old_dir);

include/linux/nfs_xdr.h +0 -0

index b28c83475ee8..2a8406b4b353 100644

include/linux/nfs_xdr.h struct nfs_renamedata { · line 1536

Quoted 4 lines

include/linux/nfs_xdr.h +1 -0

@@ -1533,6 +1533,7 @@ struct nfs_renamedata {
 	struct nfs_fattr	new_fattr;
 	void (*complete)(struct rpc_task *, struct nfs_renamedata *);
 	long timeout;
+	int cancelled;

No need for a whole int for a flag and these do get allocated. Make it a bool?

Quoted 3 lines

include/linux/nfs_xdr.h +0 -0
 };
 
 struct nfs_access_entry;

Signature · 1 line

Jeff Layton <[email protected]>

Jeff Layton

2017-06-15 19:06 UTC #5 lore raw

Re: [PATCH 2/2] NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

Quoted 33 lines · Jeff Layton

Quoted 21 lines · Benjamin Coddington
An interrupted rename will leave the old dentry behind if the rename succeeds. Fix this by forcing a lookup the next time through ->d_revalidate.
A previous attempt at solving this problem took the approach to complete the work of the rename asynchronously, however that approach was wrong since it would allow the d_move() to occur after the directory's i_mutex had been dropped by the original process.
Signed-off-by Benjamin Coddington <[email protected]>
---
 fs/nfs/dir.c            | 2 ++
 fs/nfs/unlink.c         | 7 +++++++
 include/linux/nfs_xdr.h | 1 +
 3 files changed, 10 insertions(+)
fs/nfs/dir.c +0 -0
index 1faf337b316f..bb697e5c3f6c 100644
fs/nfs/dir.c int nfs_rename(struct inode *old_dir, struct dentry *old_dentry, · line 2041
Quoted 5 lines
fs/nfs/dir.c +2 -0
@@ -2037,6 +2037,8 @@ int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	error = rpc_wait_for_completion_task(task);
 	if (error == 0)
 		error = task->tk_status;
+	else
+		((struct nfs_renamedata *)task->tk_calldata)->cancelled = 1;
This looks a bit racy. You could end up setting cancelled just after the reply comes in and the completion callback checks it. I think that you probably either want to do this with an atomic operation or under a lock of some sort.

Quoted 2 lines

You could probably do it with an xchg() operation?

As Ben pointed out on IRC, that flag is checked in rpc_release, so as long as we ensure that it's set while we hold a task reference we should be fine here without any locking.

That said, do we need a barrier here? We do need to ensure that cancelled is set before the rpc_put_task occurs.

Quoted 39 lines

Quoted 26 lines

fs/nfs/dir.c +0 -0

 	rpc_put_task(task);
 	nfs_mark_for_revalidate(old_inode);
 out:

fs/nfs/unlink.c +7 -0

index 191aa577dd1f..b47158a34879 100644
@@ -288,6 +288,13 @@ static void nfs_async_rename_release(void *calldata)
 	if (d_really_is_positive(data->old_dentry))
 		nfs_mark_for_revalidate(d_inode(data->old_dentry));
 
+	/* The result of the rename is unknown. Play it safe by
+	 * forcing a new lookup */
+	if (data->cancelled) {
+		nfs_force_lookup_revalidate(data->old_dir);
+		nfs_force_lookup_revalidate(data->new_dir);
+	}
+
 	dput(data->old_dentry);
 	dput(data->new_dentry);
 	iput(data->old_dir);

include/linux/nfs_xdr.h +0 -0

index b28c83475ee8..2a8406b4b353 100644

include/linux/nfs_xdr.h struct nfs_renamedata { · line 1536

Quoted 4 lines

include/linux/nfs_xdr.h +1 -0

@@ -1533,6 +1533,7 @@ struct nfs_renamedata {
 	struct nfs_fattr	new_fattr;
 	void (*complete)(struct rpc_task *, struct nfs_renamedata *);
 	long timeout;
+	int cancelled;

No need for a whole int for a flag and these do get allocated. Make it a bool?

Quoted 3 lines

include/linux/nfs_xdr.h +0 -0
 };
 
 struct nfs_access_entry;

Signature · 1 line

Jeff Layton <[email protected]>

Benjamin Coddington

author

2017-06-15 20:19 UTC #6 lore raw

Re: [PATCH 2/2] NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

Quoted 38 lines · Jeff Layton

Quoted 34 lines · Jeff Layton
Quoted 21 lines · Benjamin Coddington
An interrupted rename will leave the old dentry behind if the rename succeeds. Fix this by forcing a lookup the next time through ->d_revalidate.
A previous attempt at solving this problem took the approach to complete the work of the rename asynchronously, however that approach was wrong since it would allow the d_move() to occur after the directory's i_mutex had been dropped by the original process.
Signed-off-by Benjamin Coddington <[email protected]>
---
 fs/nfs/dir.c            | 2 ++
 fs/nfs/unlink.c         | 7 +++++++
 include/linux/nfs_xdr.h | 1 +
 3 files changed, 10 insertions(+)
fs/nfs/dir.c +0 -0
index 1faf337b316f..bb697e5c3f6c 100644
fs/nfs/dir.c int nfs_rename(struct inode *old_dir, struct · line 2042
Quoted 6 lines
fs/nfs/dir.c +2 -0
@@ -2037,6 +2037,8 @@ int nfs_rename(struct inode *old_dir, struct 
 dentry *old_dentry,
 	error = rpc_wait_for_completion_task(task);
 	if (error == 0)
 		error = task->tk_status;
+	else
+		((struct nfs_renamedata *)task->tk_calldata)->cancelled = 1;
This looks a bit racy. You could end up setting cancelled just after the reply comes in and the completion callback checks it. I think that you probably either want to do this with an atomic operation or under a lock of some sort.
Quoted 2 lines
You could probably do it with an xchg() operation?

Quoted 6 lines

As Ben pointed out on IRC, that flag is checked in rpc_release, so as long as we ensure that it's set while we hold a task reference we should be fine here without any locking.
That said, do we need a barrier here? We do need to ensure that cancelled is set before the rpc_put_task occurs.

Yes, I think we probably do.

Quoted 12 lines

fs/nfs/dir.c +0 -0

 	rpc_put_task(task);
 	nfs_mark_for_revalidate(old_inode);

out:

fs/nfs/unlink.c +0 -0

index 191aa577dd1f..b47158a34879 100644
@@ -288,6 +288,13 @@ static void nfs_async_rename_release(void 
 *calldata)
 	if (d_really_is_positive(data->old_dentry))
 		nfs_mark_for_revalidate(d_inode(data->old_dentry));

Quoted 5 lines

Quoted 5 lines
fs/nfs/unlink.c static void nfs_async_rename_release(void *calldata) · line 295
Quoted 5 lines
fs/nfs/unlink.c +5 -0
+	/* The result of the rename is unknown. Play it safe by
+	 * forcing a new lookup */
+	if (data->cancelled) {
+		nfs_force_lookup_revalidate(data->old_dir);
+		nfs_force_lookup_revalidate(data->new_dir);

Jeff's pointed out on IRC that we should hold the i_lock to call nfs_force_lookup_revalidate(), so I'll add that.

Quoted 9 lines

Quoted 9 lines
Quoted 9 lines
fs/nfs/unlink.c +2 -0
+	}
+
 	dput(data->old_dentry);
 	dput(data->new_dentry);
fs/nfs/unlink.c +0 -0
 	iput(data->old_dir);
include/linux/nfs_xdr.h +0 -0
index b28c83475ee8..2a8406b4b353 100644

Quoted 8 lines

Quoted 8 lines
include/linux/nfs_xdr.h struct nfs_renamedata { · line 1536
Quoted 5 lines
include/linux/nfs_xdr.h +1 -0
@@ -1533,6 +1533,7 @@ struct nfs_renamedata {
 	struct nfs_fattr	new_fattr;
 	void (*complete)(struct rpc_task *, struct nfs_renamedata *);
 	long timeout;
+	int cancelled;
No need for a whole int for a flag and these do get allocated. Make it a bool?

unsigned int : 1

which seems to be often used -- see nfs4_opendata. The cancelled flag could be changed there as well I suppose.

Ben

Jeff Layton

2017-06-15 20:34 UTC #7 lore raw

Re: [PATCH 2/2] NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

Quoted 51 lines · Benjamin Coddington

Quoted 49 lines · Jeff Layton
Quoted 40 lines · Jeff Layton
Quoted 24 lines · Benjamin Coddington
An interrupted rename will leave the old dentry behind if the rename succeeds. Fix this by forcing a lookup the next time through ->d_revalidate.
A previous attempt at solving this problem took the approach to complete
the work of the rename asynchronously, however that approach was wrong
since it would allow the d_move() to occur after the directory's i_mutex
had been dropped by the original process.
Signed-off-by Benjamin Coddington <[email protected]>
---
 fs/nfs/dir.c            | 2 ++
 fs/nfs/unlink.c         | 7 +++++++
 include/linux/nfs_xdr.h | 1 +
 3 files changed, 10 insertions(+)
fs/nfs/dir.c +0 -0
index 1faf337b316f..bb697e5c3f6c 100644
fs/nfs/dir.c int nfs_rename(struct inode *old_dir, struct · line 2042
Quoted 6 lines
fs/nfs/dir.c +2 -0
@@ -2037,6 +2037,8 @@ int nfs_rename(struct inode *old_dir, struct 
 dentry *old_dentry,
 	error = rpc_wait_for_completion_task(task);
 	if (error == 0)
 		error = task->tk_status;
+	else
+		((struct nfs_renamedata *)task->tk_calldata)->cancelled = 1;
This looks a bit racy. You could end up setting cancelled just after the
reply comes in and the completion callback checks it. I think that you
probably either want to do this with an atomic operation or under a lock
of some sort.
Quoted 2 lines
You could probably do it with an xchg() operation?
As Ben pointed out on IRC, that flag is checked in rpc_release, so as long as we ensure that it's set while we hold a task reference we should
be fine here without any locking.

Quoted 5 lines

Quoted 2 lines
That said, do we need a barrier here? We do need to ensure that cancelled is set before the rpc_put_task occurs.
Yes, I think we probably do.

Yeah, I think a smp_wmb() there, paired with the implied barrier in the atomic_dec_and_test in rpc_put_task?

Quoted 42 lines

Quoted 13 lines
Quoted 13 lines
Quoted 13 lines
fs/nfs/dir.c +0 -0
 	rpc_put_task(task);
 	nfs_mark_for_revalidate(old_inode);
out:
fs/nfs/unlink.c +0 -0
index 191aa577dd1f..b47158a34879 100644
@@ -288,6 +288,13 @@ static void nfs_async_rename_release(void 
 *calldata)
 	if (d_really_is_positive(data->old_dentry))
 		nfs_mark_for_revalidate(d_inode(data->old_dentry));
 
Quoted 5 lines
Quoted 5 lines
fs/nfs/unlink.c static void nfs_async_rename_release(void *calldata) · line 295
Quoted 5 lines
fs/nfs/unlink.c +5 -0
+	/* The result of the rename is unknown. Play it safe by
+	 * forcing a new lookup */
+	if (data->cancelled) {
+		nfs_force_lookup_revalidate(data->old_dir);
+		nfs_force_lookup_revalidate(data->new_dir);
Jeff's pointed out on IRC that we should hold the i_lock to call nfs_force_lookup_revalidate(), so I'll add that.
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
fs/nfs/unlink.c +2 -0
+	}
+
 	dput(data->old_dentry);
 	dput(data->new_dentry);
fs/nfs/unlink.c +0 -0
 	iput(data->old_dir);
include/linux/nfs_xdr.h +0 -0
index b28c83475ee8..2a8406b4b353 100644
@@ -1533,6 +1533,7 @@ struct nfs_renamedata {
Quoted 8 lines
Quoted 8 lines
include/linux/nfs_xdr.h struct nfs_renamedata { · line 1536
Quoted 4 lines
include/linux/nfs_xdr.h +1 -0
 	struct nfs_fattr	new_fattr;
 	void (*complete)(struct rpc_task *, struct nfs_renamedata *);
 	long timeout;
+	int cancelled;
No need for a whole int for a flag and these do get allocated. Make it a
bool?

Quoted 7 lines

or
unsigned int : 1
which seems to be often used -- see nfs4_opendata. The cancelled flag could
be changed there as well I suppose.

I'd prefer a bool, but it's really up to Trond and Anna, I suppose.

Signature · 1 line

Jeff Layton <[email protected]>

Benjamin Coddington

author

2017-06-15 20:57 UTC #8 lore raw

Re: [PATCH 2/2] NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

Quoted 2 lines · Jeff Layton

Yeah, I think a smp_wmb() there, paired with the implied barrier in the atomic_dec_and_test in rpc_put_task?

Yes, that should do it.

Quoted 6 lines

Quoted 6 lines
Quoted 3 lines
Quoted 3 lines
No need for a whole int for a flag and these do get allocated. Make it a
bool?
or

Quoted 7 lines

Quoted 5 lines
unsigned int : 1
which seems to be often used -- see nfs4_opendata. The cancelled flag could
be changed there as well I suppose.
I'd prefer a bool, but it's really up to Trond and Anna, I suppose.

If Anna or Trond will tell us how they'd like it, I can follow up with a patch to make them all consistent.

Ben

Anna Schumaker

2017-06-15 21:06 UTC #9 lore raw

Re: [PATCH 2/2] NFS: nfs_rename() - revalidate directories on -ERESTARTSYS

Quoted 15 lines · Benjamin Coddington

Quoted 2 lines · Jeff Layton
Yeah, I think a smp_wmb() there, paired with the implied barrier in the atomic_dec_and_test in rpc_put_task?
Yes, that should do it.
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
No need for a whole int for a flag and these do get allocated. Make it a
bool?
or
unsigned int : 1

Quoted 8 lines

Quoted 5 lines
Quoted 3 lines
which seems to be often used -- see nfs4_opendata. The cancelled flag could
be changed there as well I suppose.
I'd prefer a bool, but it's really up to Trond and Anna, I suppose.
If Anna or Trond will tell us how they'd like it, I can follow up with a patch to make them all consistent.

My preference is for a bool

Anna

Quoted 7 lines

Ben
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html