AFS: implement file locking

Implement file locking for AFS.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/afs/Makefile

@@ -8,6 +8,7 @@ kafs-objs := \
 	cmservice.o \
 	dir.o \
 	file.o \
+	flock.o \
 	fsclient.o \
 	inode.o \
 	main.o \

fs/afs/afs.h

@@ -37,6 +37,13 @@ typedef enum {
 	AFS_FTYPE_SYMLINK	= 3,
 } afs_file_type_t;
 
+typedef enum {
+	AFS_LOCK_READ		= 0,	/* read lock request */
+	AFS_LOCK_WRITE		= 1,	/* write lock request */
+} afs_lock_type_t;
+
+#define AFS_LOCKWAIT		(5 * 60) /* time until a lock times out (seconds) */
+
 /*
  * AFS file identifier
  */
@@ -120,6 +127,7 @@ struct afs_file_status {
 	struct afs_fid		parent;		/* parent dir ID for non-dirs only */
 	time_t			mtime_client;	/* last time client changed data */
 	time_t			mtime_server;	/* last time server changed data */
+	s32			lock_count;	/* file lock count (0=UNLK -1=WRLCK +ve=#RDLCK */
 };
 
 /*

fs/afs/afs_fs.h

@@ -31,6 +31,9 @@ enum AFS_FS_Operations {
 	FSGETVOLUMEINFO		= 148,	/* AFS Get information about a volume */
 	FSGETVOLUMESTATUS	= 149,	/* AFS Get volume status information */
 	FSGETROOTVOLUME		= 151,	/* AFS Get root volume name */
+	FSSETLOCK		= 156,	/* AFS Request a file lock */
+	FSEXTENDLOCK		= 157,	/* AFS Extend a file lock */
+	FSRELEASELOCK		= 158,	/* AFS Release a file lock */
 	FSLOOKUP		= 161,	/* AFS lookup file in directory */
 	FSFETCHDATA64		= 65537, /* AFS Fetch file data */
 	FSSTOREDATA64		= 65538, /* AFS Store file data */

fs/afs/callback.c

@@ -125,6 +125,9 @@ static void afs_break_callback(struct afs_server *server,
 		spin_unlock(&server->cb_lock);
 
 		queue_work(afs_callback_update_worker, &vnode->cb_broken_work);
+		if (list_empty(&vnode->granted_locks) &&
+		    !list_empty(&vnode->pending_locks))
+			afs_lock_may_be_available(vnode);
 		spin_unlock(&vnode->lock);
 	}
 }

fs/afs/dir.c

@@ -44,6 +44,7 @@ const struct file_operations afs_dir_file_operations = {
 	.open		= afs_dir_open,
 	.release	= afs_release,
 	.readdir	= afs_readdir,
+	.lock		= afs_lock,
 };
 
 const struct inode_operations afs_dir_inode_operations = {

fs/afs/file.c

@@ -34,6 +34,8 @@ const struct file_operations afs_file_operations = {
 	.mmap		= generic_file_readonly_mmap,
 	.splice_read	= generic_file_splice_read,
 	.fsync		= afs_fsync,
+	.lock		= afs_lock,
+	.flock		= afs_flock,
 };
 
 const struct inode_operations afs_file_inode_operations = {

fs/afs/fsclient.c

@@ -67,7 +67,7 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp,
 	EXTRACT(status->group);
 	bp++; /* sync counter */
 	data_version |= (u64) ntohl(*bp++) << 32;
-	bp++; /* lock count */
+	EXTRACT(status->lock_count);
 	size |= (u64) ntohl(*bp++) << 32;
 	bp++; /* spare 4 */
 	*_bp = bp;
@@ -1748,3 +1748,156 @@ int afs_fs_get_volume_status(struct afs_server *server,
 
 	return afs_make_call(&server->addr, call, GFP_NOFS, wait_mode);
 }
+
+/*
+ * deliver reply data to an FS.SetLock, FS.ExtendLock or FS.ReleaseLock
+ */
+static int afs_deliver_fs_xxxx_lock(struct afs_call *call,
+				    struct sk_buff *skb, bool last)
+{
+	const __be32 *bp;
+
+	_enter("{%u},{%u},%d", call->unmarshall, skb->len, last);
+
+	afs_transfer_reply(call, skb);
+	if (!last)
+		return 0;
+
+	if (call->reply_size != call->reply_max)
+		return -EBADMSG;
+
+	/* unmarshall the reply once we've received all of it */
+	bp = call->buffer;
+	/* xdr_decode_AFSVolSync(&bp, call->replyX); */
+
+	_leave(" = 0 [done]");
+	return 0;
+}
+
+/*
+ * FS.SetLock operation type
+ */
+static const struct afs_call_type afs_RXFSSetLock = {
+	.name		= "FS.SetLock",
+	.deliver	= afs_deliver_fs_xxxx_lock,
+	.abort_to_error	= afs_abort_to_error,
+	.destructor	= afs_flat_call_destructor,
+};
+
+/*
+ * FS.ExtendLock operation type
+ */
+static const struct afs_call_type afs_RXFSExtendLock = {
+	.name		= "FS.ExtendLock",
+	.deliver	= afs_deliver_fs_xxxx_lock,
+	.abort_to_error	= afs_abort_to_error,
+	.destructor	= afs_flat_call_destructor,
+};
+
+/*
+ * FS.ReleaseLock operation type
+ */
+static const struct afs_call_type afs_RXFSReleaseLock = {
+	.name		= "FS.ReleaseLock",
+	.deliver	= afs_deliver_fs_xxxx_lock,
+	.abort_to_error	= afs_abort_to_error,
+	.destructor	= afs_flat_call_destructor,
+};
+
+/*
+ * get a lock on a file
+ */
+int afs_fs_set_lock(struct afs_server *server,
+		    struct key *key,
+		    struct afs_vnode *vnode,
+		    afs_lock_type_t type,
+		    const struct afs_wait_mode *wait_mode)
+{
+	struct afs_call *call;
+	__be32 *bp;
+
+	_enter("");
+
+	call = afs_alloc_flat_call(&afs_RXFSSetLock, 5 * 4, 6 * 4);
+	if (!call)
+		return -ENOMEM;
+
+	call->key = key;
+	call->reply = vnode;
+	call->service_id = FS_SERVICE;
+	call->port = htons(AFS_FS_PORT);
+
+	/* marshall the parameters */
+	bp = call->request;
+	*bp++ = htonl(FSSETLOCK);
+	*bp++ = htonl(vnode->fid.vid);
+	*bp++ = htonl(vnode->fid.vnode);
+	*bp++ = htonl(vnode->fid.unique);
+	*bp++ = htonl(type);
+
+	return afs_make_call(&server->addr, call, GFP_NOFS, wait_mode);
+}
+
+/*
+ * extend a lock on a file
+ */
+int afs_fs_extend_lock(struct afs_server *server,
+		       struct key *key,
+		       struct afs_vnode *vnode,
+		       const struct afs_wait_mode *wait_mode)
+{
+	struct afs_call *call;
+	__be32 *bp;
+
+	_enter("");
+
+	call = afs_alloc_flat_call(&afs_RXFSExtendLock, 4 * 4, 6 * 4);
+	if (!call)
+		return -ENOMEM;
+
+	call->key = key;
+	call->reply = vnode;
+	call->service_id = FS_SERVICE;
+	call->port = htons(AFS_FS_PORT);
+
+	/* marshall the parameters */
+	bp = call->request;
+	*bp++ = htonl(FSEXTENDLOCK);
+	*bp++ = htonl(vnode->fid.vid);
+	*bp++ = htonl(vnode->fid.vnode);
+	*bp++ = htonl(vnode->fid.unique);
+
+	return afs_make_call(&server->addr, call, GFP_NOFS, wait_mode);
+}
+
+/*
+ * release a lock on a file
+ */
+int afs_fs_release_lock(struct afs_server *server,
+			struct key *key,
+			struct afs_vnode *vnode,
+			const struct afs_wait_mode *wait_mode)
+{
+	struct afs_call *call;
+	__be32 *bp;
+
+	_enter("");
+
+	call = afs_alloc_flat_call(&afs_RXFSReleaseLock, 4 * 4, 6 * 4);
+	if (!call)
+		return -ENOMEM;
+
+	call->key = key;
+	call->reply = vnode;
+	call->service_id = FS_SERVICE;
+	call->port = htons(AFS_FS_PORT);
+
+	/* marshall the parameters */
+	bp = call->request;
+	*bp++ = htonl(FSRELEASELOCK);
+	*bp++ = htonl(vnode->fid.vid);
+	*bp++ = htonl(vnode->fid.vnode);
+	*bp++ = htonl(vnode->fid.unique);
+
+	return afs_make_call(&server->addr, call, GFP_NOFS, wait_mode);
+}

fs/afs/internal.h

@@ -351,10 +351,18 @@ struct afs_vnode {
 #define AFS_VNODE_ZAP_DATA	3		/* set if vnode's data should be invalidated */
 #define AFS_VNODE_DELETED	4		/* set if vnode deleted on server */
 #define AFS_VNODE_MOUNTPOINT	5		/* set if vnode is a mountpoint symlink */
+#define AFS_VNODE_LOCKING	6		/* set if waiting for lock on vnode */
+#define AFS_VNODE_READLOCKED	7		/* set if vnode is read-locked on the server */
+#define AFS_VNODE_WRITELOCKED	8		/* set if vnode is write-locked on the server */
+#define AFS_VNODE_UNLOCKING	9		/* set if vnode is being unlocked on the server */
 
 	long			acl_order;	/* ACL check count (callback break count) */
 
 	struct list_head	writebacks;	/* alterations in pagecache that need writing */
+	struct list_head	pending_locks;	/* locks waiting to be granted */
+	struct list_head	granted_locks;	/* locks granted on this file */
+	struct delayed_work	lock_work;	/* work to be done in locking */
+	struct key		*unlock_key;	/* key to be used in unlocking */
 
 	/* outstanding callback notification on this file */
 	struct rb_node		server_rb;	/* link in server->fs_vnodes */
@@ -473,6 +481,15 @@ extern const struct file_operations afs_file_operations;
 extern int afs_open(struct inode *, struct file *);
 extern int afs_release(struct inode *, struct file *);
 
+/*
+ * flock.c
+ */
+extern void __exit afs_kill_lock_manager(void);
+extern void afs_lock_work(struct work_struct *);
+extern void afs_lock_may_be_available(struct afs_vnode *);
+extern int afs_lock(struct file *, int, struct file_lock *);
+extern int afs_flock(struct file *, int, struct file_lock *);
+
 /*
  * fsclient.c
  */
@@ -513,6 +530,15 @@ extern int afs_fs_get_volume_status(struct afs_server *, struct key *,
 				    struct afs_vnode *,
 				    struct afs_volume_status *,
 				    const struct afs_wait_mode *);
+extern int afs_fs_set_lock(struct afs_server *, struct key *,
+			   struct afs_vnode *, afs_lock_type_t,
+			   const struct afs_wait_mode *);
+extern int afs_fs_extend_lock(struct afs_server *, struct key *,
+			      struct afs_vnode *,
+			      const struct afs_wait_mode *);
+extern int afs_fs_release_lock(struct afs_server *, struct key *,
+			       struct afs_vnode *,
+			       const struct afs_wait_mode *);
 
 /*
  * inode.c
@@ -681,6 +707,10 @@ extern int afs_vnode_store_data(struct afs_writeback *, pgoff_t, pgoff_t,
 extern int afs_vnode_setattr(struct afs_vnode *, struct key *, struct iattr *);
 extern int afs_vnode_get_volume_status(struct afs_vnode *, struct key *,
 				       struct afs_volume_status *);
+extern int afs_vnode_set_lock(struct afs_vnode *, struct key *,
+			      afs_lock_type_t);
+extern int afs_vnode_extend_lock(struct afs_vnode *, struct key *);
+extern int afs_vnode_release_lock(struct afs_vnode *, struct key *);
 
 /*
  * volume.c

fs/afs/main.c

@@ -168,6 +168,7 @@ static void __exit afs_exit(void)
 	printk(KERN_INFO "kAFS: Red Hat AFS client v0.1 unregistering.\n");
 
 	afs_fs_exit();
+	afs_kill_lock_manager();
 	afs_close_socket();
 	afs_purge_servers();
 	afs_callback_update_kill();

fs/afs/misc.c

@@ -35,6 +35,7 @@ int afs_abort_to_error(u32 abort_code)
 	case VOVERQUOTA:	return -EDQUOT;
 	case VBUSY:		return -EBUSY;
 	case VMOVED:		return -ENXIO;
+	case 0x2f6df0a:		return -EWOULDBLOCK;
 	case 0x2f6df0c:		return -EACCES;
 	case 0x2f6df0f:		return -EBUSY;
 	case 0x2f6df10:		return -EEXIST;

fs/afs/super.c

@@ -460,6 +460,9 @@ static void afs_i_init_once(void *_vnode, struct kmem_cache *cachep,
 	spin_lock_init(&vnode->writeback_lock);
 	spin_lock_init(&vnode->lock);
 	INIT_LIST_HEAD(&vnode->writebacks);
+	INIT_LIST_HEAD(&vnode->pending_locks);
+	INIT_LIST_HEAD(&vnode->granted_locks);
+	INIT_DELAYED_WORK(&vnode->lock_work, afs_lock_work);
 	INIT_WORK(&vnode->cb_broken_work, afs_broken_callback_work);
 }
 

fs/afs/vnode.c

@@ -887,11 +887,6 @@ int afs_vnode_get_volume_status(struct afs_vnode *vnode, struct key *key,
 	       vnode->fid.unique,
 	       key_serial(key));
 
-	/* this op will fetch the status */
-	spin_lock(&vnode->lock);
-	vnode->update_cnt++;
-	spin_unlock(&vnode->lock);
-
 	do {
 		/* pick a server to query */
 		server = afs_volume_pick_fileserver(vnode);
@@ -905,20 +900,127 @@ int afs_vnode_get_volume_status(struct afs_vnode *vnode, struct key *key,
 	} while (!afs_volume_release_fileserver(vnode, server, ret));
 
 	/* adjust the flags */
-	if (ret == 0) {
-		afs_vnode_finalise_status_update(vnode, server);
+	if (ret == 0)
+		afs_put_server(server);
+
+	_leave(" = %d", ret);
+	return ret;
+
+no_server:
+	return PTR_ERR(server);
+}
+
+/*
+ * get a lock on a file
+ */
+int afs_vnode_set_lock(struct afs_vnode *vnode, struct key *key,
+		       afs_lock_type_t type)
+{
+	struct afs_server *server;
+	int ret;
+
+	_enter("%s{%x:%u.%u},%x,%u",
+	       vnode->volume->vlocation->vldb.name,
+	       vnode->fid.vid,
+	       vnode->fid.vnode,
+	       vnode->fid.unique,
+	       key_serial(key), type);
+
+	do {
+		/* pick a server to query */
+		server = afs_volume_pick_fileserver(vnode);
+		if (IS_ERR(server))
+			goto no_server;
+
+		_debug("USING SERVER: %08x\n", ntohl(server->addr.s_addr));
+
+		ret = afs_fs_set_lock(server, key, vnode, type, &afs_sync_call);
+
+	} while (!afs_volume_release_fileserver(vnode, server, ret));
+
+	/* adjust the flags */
+	if (ret == 0)
+		afs_put_server(server);
+
+	_leave(" = %d", ret);
+	return ret;
+
+no_server:
+	return PTR_ERR(server);
+}
+
+/*
+ * extend a lock on a file
+ */
+int afs_vnode_extend_lock(struct afs_vnode *vnode, struct key *key)
+{
+	struct afs_server *server;
+	int ret;
+
+	_enter("%s{%x:%u.%u},%x",
+	       vnode->volume->vlocation->vldb.name,
+	       vnode->fid.vid,
+	       vnode->fid.vnode,
+	       vnode->fid.unique,
+	       key_serial(key));
+
+	do {
+		/* pick a server to query */
+		server = afs_volume_pick_fileserver(vnode);
+		if (IS_ERR(server))
+			goto no_server;
+
+		_debug("USING SERVER: %08x\n", ntohl(server->addr.s_addr));
+
+		ret = afs_fs_extend_lock(server, key, vnode, &afs_sync_call);
+
+	} while (!afs_volume_release_fileserver(vnode, server, ret));
+
+	/* adjust the flags */
+	if (ret == 0)
+		afs_put_server(server);
+
+	_leave(" = %d", ret);
+	return ret;
+
+no_server:
+	return PTR_ERR(server);
+}
+
+/*
+ * release a lock on a file
+ */
+int afs_vnode_release_lock(struct afs_vnode *vnode, struct key *key)
+{
+	struct afs_server *server;
+	int ret;
+
+	_enter("%s{%x:%u.%u},%x",
+	       vnode->volume->vlocation->vldb.name,
+	       vnode->fid.vid,
+	       vnode->fid.vnode,
+	       vnode->fid.unique,
+	       key_serial(key));
+
+	do {
+		/* pick a server to query */
+		server = afs_volume_pick_fileserver(vnode);
+		if (IS_ERR(server))
+			goto no_server;
+
+		_debug("USING SERVER: %08x\n", ntohl(server->addr.s_addr));
+
+		ret = afs_fs_release_lock(server, key, vnode, &afs_sync_call);
+
+	} while (!afs_volume_release_fileserver(vnode, server, ret));
+
+	/* adjust the flags */
+	if (ret == 0)
 		afs_put_server(server);
-	} else {
-		afs_vnode_status_update_failed(vnode, ret);
-	}
 
 	_leave(" = %d", ret);
 	return ret;
 
 no_server:
-	spin_lock(&vnode->lock);
-	vnode->update_cnt--;
-	ASSERTCMP(vnode->update_cnt, >=, 0);
-	spin_unlock(&vnode->lock);
 	return PTR_ERR(server);
 }

include/linux/fs.h

@@ -820,6 +820,10 @@ struct file_lock {
 	union {
 		struct nfs_lock_info	nfs_fl;
 		struct nfs4_lock_info	nfs4_fl;
+		struct {
+			struct list_head link;	/* link in AFS vnode's pending_locks list */
+			int state;		/* state of grant or error if -ve */
+		} afs;
 	} fl_u;
 };