vfs: add renameat2 syscall

Add new renameat2 syscall, which is the same as renameat with an added flags argument. Pass flags to vfs_rename() and to i_op->rename() as well. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Reviewed-by: J. Bruce Fields <bfields@redhat.com>

vfs: add renameat2 syscall
Add new renameat2 syscall, which is the same as renameat with an added flags argument. Pass flags to vfs_rename() and to i_op->rename() as well. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Reviewed-by: J. Bruce Fields <bfields@redhat.com>
Miklos Szeredi
1 parent bc27027a73
Showing 10 changed files with 58 additions and 15 deletions Side-by-side Diff
Documentation/filesystems/Locking
Documentation/filesystems/vfs.txt
arch/x86/syscalls/syscall_64.tbl
drivers/staging/lustre/lustre/include/linux/lustre_compat25.h
drivers/staging/lustre/lustre/lvfs/lvfs_linux.c
fs/cachefiles/namei.c
fs/ecryptfs/inode.c
fs/namei.c
fs/nfsd/vfs.c
include/linux/fs.h
@@ -47,6 +47,8 @@
 	int (*mknod) (struct inode *,struct dentry *,umode_t,dev_t);
 	int (*rename) (struct inode *, struct dentry *,
 			struct inode *, struct dentry *);
+	int (*rename2) (struct inode *, struct dentry *,
+			struct inode *, struct dentry *, unsigned int);
 	int (*readlink) (struct dentry *, char __user *,int);
 	void * (*follow_link) (struct dentry *, struct nameidata *);
 	void (*put_link) (struct dentry *, struct nameidata *, void *);
@@ -78,6 +80,7 @@
 unlink:		yes (both)
 rmdir:		yes (both)	(see below)
 rename:		yes (all)	(see below)
+rename2:	yes (all)	(see below)
 readlink:	no
 follow_link:	no
 put_link:	no
@@ -96,7 +99,8 @@
  
 	Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
 victim.
-	cross-directory ->rename() has (per-superblock) ->s_vfs_rename_sem.
+	cross-directory ->rename() and rename2() has (per-superblock)
+->s_vfs_rename_sem.
  
 See Documentation/filesystems/directory-locking for more detailed discussion
 of the locking scheme for directory operations.
@@ -347,6 +347,8 @@
 	int (*mknod) (struct inode *,struct dentry *,umode_t,dev_t);
 	int (*rename) (struct inode *, struct dentry *,
 			struct inode *, struct dentry *);
+	int (*rename2) (struct inode *, struct dentry *,
+			struct inode *, struct dentry *, unsigned int);
 	int (*readlink) (struct dentry *, char __user *,int);
         void * (*follow_link) (struct dentry *, struct nameidata *);
         void (*put_link) (struct dentry *, struct nameidata *, void *);
@@ -413,6 +415,20 @@
  
   rename: called by the rename(2) system call to rename the object to
 	have the parent and name given by the second inode and dentry.
+
+  rename2: this has an additional flags argument compared to rename.
+	If no flags are supported by the filesystem then this method
+	need not be implemented.  If some flags are supported then the
+	filesystem must return -EINVAL for any unsupported or unknown
+	flags.  Currently the following flags are implemented:
+	(1) RENAME_NOREPLACE: this flag indicates that if the target
+	of the rename exists the rename should fail with -EEXIST
+	instead of replacing the target.  The VFS already checks for
+	existence, so for local filesystems the RENAME_NOREPLACE
+	implementation is equivalent to plain rename.
+	(2) RENAME_EXCHANGE: exchange source and target.  Both must
+	exist; this is checked by the VFS.  Unlike plain rename,
+	source and target may be of different type.
  
   readlink: called by the readlink(2) system call. Only required if
 	you want to support reading symbolic links
@@ -322,6 +322,7 @@
 313	common	finit_module		sys_finit_module
 314	common	sched_setattr		sys_sched_setattr
 315	common	sched_getattr		sys_sched_getattr
+316	common	renameat2		sys_renameat2
  
 #
 # x32-specific system call numbers start at 512 to avoid cache impact
@@ -105,8 +105,8 @@
 #define ll_vfs_unlink(inode,entry,mnt)	  vfs_unlink(inode,entry)
 #define ll_vfs_mknod(dir,entry,mnt,mode,dev)    vfs_mknod(dir,entry,mode,dev)
 #define ll_security_inode_unlink(dir,entry,mnt) security_inode_unlink(dir,entry)
-#define ll_vfs_rename(old,old_dir,mnt,new,new_dir,mnt1,delegated_inode) \
-		vfs_rename(old,old_dir,new,new_dir,delegated_inode)
+#define ll_vfs_rename(old, old_dir, mnt, new, new_dir, mnt1) \
+		vfs_rename(old, old_dir, new, new_dir, NULL, 0)
  
 #define cfs_bio_io_error(a,b)   bio_io_error((a))
 #define cfs_bio_endio(a,b,c)    bio_endio((a),(c))
@@ -223,7 +223,7 @@
 		GOTO(put_old, err = PTR_ERR(dchild_new));
  
 	err = ll_vfs_rename(dir->d_inode, dchild_old, mnt,
-			    dir->d_inode, dchild_new, mnt, NULL);
+			    dir->d_inode, dchild_new, mnt);
  
 	dput(dchild_new);
 put_old:
@@ -396,7 +396,7 @@
 		cachefiles_io_error(cache, "Rename security error %d", ret);
 	} else {
 		ret = vfs_rename(dir->d_inode, rep,
-				 cache->graveyard->d_inode, grave, NULL);
+				 cache->graveyard->d_inode, grave, NULL, 0);
 		if (ret != 0 && ret != -ENOMEM)
 			cachefiles_io_error(cache,
 					    "Rename failed with error %d", ret);
@@ -641,7 +641,7 @@
 	}
 	rc = vfs_rename(lower_old_dir_dentry->d_inode, lower_old_dentry,
 			lower_new_dir_dentry->d_inode, lower_new_dentry,
-			NULL);
+			NULL, 0);
 	if (rc)
 		goto out_lock;
 	if (target_inode)
@@ -3980,6 +3980,7 @@
  * @new_dir:	parent of destination
  * @new_dentry:	destination
  * @delegated_inode: returns an inode needing a delegation break
+ * @flags:	rename flags
  *
  * The caller must hold multiple mutexes--see lock_rename()).
  *
@@ -4023,7 +4024,7 @@
  */
 int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	       struct inode *new_dir, struct dentry *new_dentry,
-	       struct inode **delegated_inode)
+	       struct inode **delegated_inode, unsigned int flags)
 {
 	int error;
 	bool is_dir = d_is_dir(old_dentry);
@@ -4048,6 +4049,9 @@
 	if (!old_dir->i_op->rename)
 		return -EPERM;
  
+	if (flags && !old_dir->i_op->rename2)
+		return -EINVAL;
+
 	/*
 	 * If we are going to change the parent - check write permissions,
 	 * we'll need to flip '..'.
@@ -4093,7 +4097,13 @@
 				goto out;
 		}
 	}
-	error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
+	if (!flags) {
+		error = old_dir->i_op->rename(old_dir, old_dentry,
+					      new_dir, new_dentry);
+	} else {
+		error = old_dir->i_op->rename2(old_dir, old_dentry,
+					       new_dir, new_dentry, flags);
+	}
 	if (error)
 		goto out;
  
@@ -4118,8 +4128,8 @@
 	return error;
 }
  
-SYSCALL_DEFINE4(renameat, int, olddfd, const char __user *, oldname,
-		int, newdfd, const char __user *, newname)
+SYSCALL_DEFINE5(renameat2, int, olddfd, const char __user *, oldname,
+		int, newdfd, const char __user *, newname, unsigned int, flags)
 {
 	struct dentry *old_dir, *new_dir;
 	struct dentry *old_dentry, *new_dentry;
@@ -4131,6 +4141,10 @@
 	unsigned int lookup_flags = 0;
 	bool should_retry = false;
 	int error;
+
+	if (flags)
+		return -EINVAL;
+
 retry:
 	from = user_path_parent(olddfd, oldname, &oldnd, lookup_flags);
 	if (IS_ERR(from)) {
@@ -4202,8 +4216,8 @@
 	if (error)
 		goto exit5;
 	error = vfs_rename(old_dir->d_inode, old_dentry,
-				   new_dir->d_inode, new_dentry,
-				   &delegated_inode);
+			   new_dir->d_inode, new_dentry,
+			   &delegated_inode, flags);
 exit5:
 	dput(new_dentry);
 exit4:
  
@@ -4233,9 +4247,15 @@
 	return error;
 }
  
+SYSCALL_DEFINE4(renameat, int, olddfd, const char __user *, oldname,
+		int, newdfd, const char __user *, newname)
+{
+	return sys_renameat2(olddfd, oldname, newdfd, newname, 0);
+}
+
 SYSCALL_DEFINE2(rename, const char __user *, oldname, const char __user *, newname)
 {
-	return sys_renameat(AT_FDCWD, oldname, AT_FDCWD, newname);
+	return sys_renameat2(AT_FDCWD, oldname, AT_FDCWD, newname, 0);
 }
  
 int vfs_readlink(struct dentry *dentry, char __user *buffer, int buflen, const char *link)
@@ -1694,7 +1694,7 @@
 	if (ffhp->fh_export->ex_path.dentry != tfhp->fh_export->ex_path.dentry)
 		goto out_dput_new;
  
-	host_err = vfs_rename(fdir, odentry, tdir, ndentry, NULL);
+	host_err = vfs_rename(fdir, odentry, tdir, ndentry, NULL, 0);
 	if (!host_err) {
 		host_err = commit_metadata(tfhp);
 		if (!host_err)
@@ -1460,7 +1460,7 @@
 extern int vfs_link(struct dentry *, struct inode *, struct dentry *, struct inode **);
 extern int vfs_rmdir(struct inode *, struct dentry *);
 extern int vfs_unlink(struct inode *, struct dentry *, struct inode **);
-extern int vfs_rename(struct inode *, struct dentry *, struct inode *, struct dentry *, struct inode **);
+extern int vfs_rename(struct inode *, struct dentry *, struct inode *, struct dentry *, struct inode **, unsigned int);
  
 /*
  * VFS dentry helper functions.
@@ -1571,6 +1571,8 @@
 	int (*mknod) (struct inode *,struct dentry *,umode_t,dev_t);
 	int (*rename) (struct inode *, struct dentry *,
 			struct inode *, struct dentry *);
+	int (*rename2) (struct inode *, struct dentry *,
+			struct inode *, struct dentry *, unsigned int);
 	int (*setattr) (struct dentry *, struct iattr *);
 	int (*getattr) (struct vfsmount *mnt, struct dentry *, struct kstat *);
 	int (*setxattr) (struct dentry *, const char *,const void *,size_t,int);
...	...	@@ -47,6 +47,8 @@
47	47	int (mknod) (struct inode ,struct dentry *,umode_t,dev_t);
48	48	int (rename) (struct inode , struct dentry *,
49	49	struct inode , struct dentry );
	50	+ int (rename2) (struct inode , struct dentry *,
	51	+ struct inode , struct dentry , unsigned int);
50	52	int (readlink) (struct dentry , char __user *,int);
51	53	void * (follow_link) (struct dentry , struct nameidata *);
52	54	void (put_link) (struct dentry , struct nameidata , void );
...	...	@@ -78,6 +80,7 @@
78	80	unlink: yes (both)
79	81	rmdir: yes (both) (see below)
80	82	rename: yes (all) (see below)
	83	+rename2: yes (all) (see below)
81	84	readlink: no
82	85	follow_link: no
83	86	put_link: no
...	...	@@ -96,7 +99,8 @@
96	99
97	100	Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
98	101	victim.
99		- cross-directory ->rename() has (per-superblock) ->s_vfs_rename_sem.
	102	+ cross-directory ->rename() and rename2() has (per-superblock)
	103	+->s_vfs_rename_sem.
100	104
101	105	See Documentation/filesystems/directory-locking for more detailed discussion
102	106	of the locking scheme for directory operations.
...	...	@@ -347,6 +347,8 @@
347	347	int (mknod) (struct inode ,struct dentry *,umode_t,dev_t);
348	348	int (rename) (struct inode , struct dentry *,
349	349	struct inode , struct dentry );
	350	+ int (rename2) (struct inode , struct dentry *,
	351	+ struct inode , struct dentry , unsigned int);
350	352	int (readlink) (struct dentry , char __user *,int);
351	353	void * (follow_link) (struct dentry , struct nameidata *);
352	354	void (put_link) (struct dentry , struct nameidata , void );
...	...	@@ -413,6 +415,20 @@
413	415
414	416	rename: called by the rename(2) system call to rename the object to
415	417	have the parent and name given by the second inode and dentry.
	418	+
	419	+ rename2: this has an additional flags argument compared to rename.
	420	+ If no flags are supported by the filesystem then this method
	421	+ need not be implemented. If some flags are supported then the
	422	+ filesystem must return -EINVAL for any unsupported or unknown
	423	+ flags. Currently the following flags are implemented:
	424	+ (1) RENAME_NOREPLACE: this flag indicates that if the target
	425	+ of the rename exists the rename should fail with -EEXIST
	426	+ instead of replacing the target. The VFS already checks for
	427	+ existence, so for local filesystems the RENAME_NOREPLACE
	428	+ implementation is equivalent to plain rename.
	429	+ (2) RENAME_EXCHANGE: exchange source and target. Both must
	430	+ exist; this is checked by the VFS. Unlike plain rename,
	431	+ source and target may be of different type.
416	432
417	433	readlink: called by the readlink(2) system call. Only required if
418	434	you want to support reading symbolic links
...	...	@@ -322,6 +322,7 @@
322	322	313 common finit_module sys_finit_module
323	323	314 common sched_setattr sys_sched_setattr
324	324	315 common sched_getattr sys_sched_getattr
	325	+316 common renameat2 sys_renameat2
325	326
326	327	#
327	328	# x32-specific system call numbers start at 512 to avoid cache impact
...	...	@@ -105,8 +105,8 @@
105	105	#define ll_vfs_unlink(inode,entry,mnt) vfs_unlink(inode,entry)
106	106	#define ll_vfs_mknod(dir,entry,mnt,mode,dev) vfs_mknod(dir,entry,mode,dev)
107	107	#define ll_security_inode_unlink(dir,entry,mnt) security_inode_unlink(dir,entry)
108		-#define ll_vfs_rename(old,old_dir,mnt,new,new_dir,mnt1,delegated_inode) \
109		- vfs_rename(old,old_dir,new,new_dir,delegated_inode)
	108	+#define ll_vfs_rename(old, old_dir, mnt, new, new_dir, mnt1) \
	109	+ vfs_rename(old, old_dir, new, new_dir, NULL, 0)
110	110
111	111	#define cfs_bio_io_error(a,b) bio_io_error((a))
112	112	#define cfs_bio_endio(a,b,c) bio_endio((a),(c))
...	...	@@ -223,7 +223,7 @@
223	223	GOTO(put_old, err = PTR_ERR(dchild_new));
224	224
225	225	err = ll_vfs_rename(dir->d_inode, dchild_old, mnt,
226		- dir->d_inode, dchild_new, mnt, NULL);
	226	+ dir->d_inode, dchild_new, mnt);
227	227
228	228	dput(dchild_new);
229	229	put_old:
...	...	@@ -396,7 +396,7 @@
396	396	cachefiles_io_error(cache, "Rename security error %d", ret);
397	397	} else {
398	398	ret = vfs_rename(dir->d_inode, rep,
399		- cache->graveyard->d_inode, grave, NULL);
	399	+ cache->graveyard->d_inode, grave, NULL, 0);
400	400	if (ret != 0 && ret != -ENOMEM)
401	401	cachefiles_io_error(cache,
402	402	"Rename failed with error %d", ret);
...	...	@@ -641,7 +641,7 @@
641	641	}
642	642	rc = vfs_rename(lower_old_dir_dentry->d_inode, lower_old_dentry,
643	643	lower_new_dir_dentry->d_inode, lower_new_dentry,
644		- NULL);
	644	+ NULL, 0);
645	645	if (rc)
646	646	goto out_lock;
647	647	if (target_inode)
...	...	@@ -3980,6 +3980,7 @@
3980	3980	* @new_dir: parent of destination
3981	3981	* @new_dentry: destination
3982	3982	* @delegated_inode: returns an inode needing a delegation break
	3983	+ * @flags: rename flags
3983	3984	*
3984	3985	* The caller must hold multiple mutexes--see lock_rename()).
3985	3986	*
...	...	@@ -4023,7 +4024,7 @@
4023	4024	*/
4024	4025	int vfs_rename(struct inode old_dir, struct dentry old_dentry,
4025	4026	struct inode new_dir, struct dentry new_dentry,
4026		- struct inode **delegated_inode)
	4027	+ struct inode **delegated_inode, unsigned int flags)
4027	4028	{
4028	4029	int error;
4029	4030	bool is_dir = d_is_dir(old_dentry);
...	...	@@ -4048,6 +4049,9 @@
4048	4049	if (!old_dir->i_op->rename)
4049	4050	return -EPERM;
4050	4051
	4052	+ if (flags && !old_dir->i_op->rename2)
	4053	+ return -EINVAL;
	4054	+
4051	4055	/*
4052	4056	* If we are going to change the parent - check write permissions,
4053	4057	* we'll need to flip '..'.
...	...	@@ -4093,7 +4097,13 @@
4093	4097	goto out;
4094	4098	}
4095	4099	}
4096		- error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
	4100	+ if (!flags) {
	4101	+ error = old_dir->i_op->rename(old_dir, old_dentry,
	4102	+ new_dir, new_dentry);
	4103	+ } else {
	4104	+ error = old_dir->i_op->rename2(old_dir, old_dentry,
	4105	+ new_dir, new_dentry, flags);
	4106	+ }
4097	4107	if (error)
4098	4108	goto out;
4099	4109
...	...	@@ -4118,8 +4128,8 @@
4118	4128	return error;
4119	4129	}
4120	4130
4121		-SYSCALL_DEFINE4(renameat, int, olddfd, const char __user *, oldname,
4122		- int, newdfd, const char __user *, newname)
	4131	+SYSCALL_DEFINE5(renameat2, int, olddfd, const char __user *, oldname,
	4132	+ int, newdfd, const char __user *, newname, unsigned int, flags)
4123	4133	{
4124	4134	struct dentry old_dir, new_dir;
4125	4135	struct dentry old_dentry, new_dentry;
...	...	@@ -4131,6 +4141,10 @@
4131	4141	unsigned int lookup_flags = 0;
4132	4142	bool should_retry = false;
4133	4143	int error;
	4144	+
	4145	+ if (flags)
	4146	+ return -EINVAL;
	4147	+
4134	4148	retry:
4135	4149	from = user_path_parent(olddfd, oldname, &oldnd, lookup_flags);
4136	4150	if (IS_ERR(from)) {
...	...	@@ -4202,8 +4216,8 @@
4202	4216	if (error)
4203	4217	goto exit5;
4204	4218	error = vfs_rename(old_dir->d_inode, old_dentry,
4205		- new_dir->d_inode, new_dentry,
4206		- &delegated_inode);
	4219	+ new_dir->d_inode, new_dentry,
	4220	+ &delegated_inode, flags);
4207	4221	exit5:
4208	4222	dput(new_dentry);
4209	4223	exit4:
4210	4224
...	...	@@ -4233,9 +4247,15 @@
4233	4247	return error;
4234	4248	}
4235	4249
	4250	+SYSCALL_DEFINE4(renameat, int, olddfd, const char __user *, oldname,
	4251	+ int, newdfd, const char __user *, newname)
	4252	+{
	4253	+ return sys_renameat2(olddfd, oldname, newdfd, newname, 0);
	4254	+}
	4255	+
4236	4256	SYSCALL_DEFINE2(rename, const char __user , oldname, const char __user , newname)
4237	4257	{
4238		- return sys_renameat(AT_FDCWD, oldname, AT_FDCWD, newname);
	4258	+ return sys_renameat2(AT_FDCWD, oldname, AT_FDCWD, newname, 0);
4239	4259	}
4240	4260
4241	4261	int vfs_readlink(struct dentry dentry, char __user buffer, int buflen, const char *link)
...	...	@@ -1694,7 +1694,7 @@
1694	1694	if (ffhp->fh_export->ex_path.dentry != tfhp->fh_export->ex_path.dentry)
1695	1695	goto out_dput_new;
1696	1696
1697		- host_err = vfs_rename(fdir, odentry, tdir, ndentry, NULL);
	1697	+ host_err = vfs_rename(fdir, odentry, tdir, ndentry, NULL, 0);
1698	1698	if (!host_err) {
1699	1699	host_err = commit_metadata(tfhp);
1700	1700	if (!host_err)
...	...	@@ -1460,7 +1460,7 @@
1460	1460	extern int vfs_link(struct dentry , struct inode , struct dentry , struct inode *);
1461	1461	extern int vfs_rmdir(struct inode , struct dentry );
1462	1462	extern int vfs_unlink(struct inode , struct dentry , struct inode **);
1463		-extern int vfs_rename(struct inode , struct dentry , struct inode , struct dentry , struct inode **);
	1463	+extern int vfs_rename(struct inode , struct dentry , struct inode , struct dentry , struct inode **, unsigned int);
1464	1464
1465	1465	/*
1466	1466	* VFS dentry helper functions.
...	...	@@ -1571,6 +1571,8 @@
1571	1571	int (mknod) (struct inode ,struct dentry *,umode_t,dev_t);
1572	1572	int (rename) (struct inode , struct dentry *,
1573	1573	struct inode , struct dentry );
	1574	+ int (rename2) (struct inode , struct dentry *,
	1575	+ struct inode , struct dentry , unsigned int);
1574	1576	int (setattr) (struct dentry , struct iattr *);
1575	1577	int (getattr) (struct vfsmount mnt, struct dentry , struct kstat );
1576	1578	int (setxattr) (struct dentry , const char ,const void ,size_t,int);