[PATCH] vfs: *at functions: core

Here is a series of patches which introduce in total 13 new system calls
which take a file descriptor/filename pair instead of a single file
name.  These functions, openat etc, have been discussed on numerous
occasions.  They are needed to implement race-free filesystem traversal,
they are necessary to implement a virtual per-thread current working
directory (think multi-threaded backup software), etc.

We have in glibc today implementations of the interfaces which use the
/proc/self/fd magic.  But this code is rather expensive.  Here are some
results (similar to what Jim Meyering posted before).

The test creates a deep directory hierarchy on a tmpfs filesystem.  Then
rm -fr is used to remove all directories.  Without syscall support I get
this:

real    0m31.921s
user    0m0.688s
sys     0m31.234s

With syscall support the results are much better:

real    0m20.699s
user    0m0.536s
sys     0m20.149s

The interfaces are for obvious reasons currently not much used.  But they'll
be used.  coreutils (and Jeff's posixutils) are already using them.
Furthermore, code like ftw/fts in libc (maybe even glob) will also start using
them.  I expect a patch to make follow soon.  Every program which is walking
the filesystem tree will benefit.
Signed-off-by: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@ftp.linux.org.uk>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

arch/alpha/kernel/osf_sys.c

@@ -960,7 +960,7 @@ osf_utimes(char __user *filename, struct timeval32 __user *tvs)
 			return -EFAULT;
 	}
 
-	return do_utimes(filename, tvs ? ktvs : NULL);
+	return do_utimes(AT_FDCWD, filename, tvs ? ktvs : NULL);
 }
 
 #define MAX_SELECT_SECONDS \

arch/sparc64/kernel/sys_sparc32.c

@@ -821,7 +821,7 @@ asmlinkage long sys32_utimes(char __user *filename,
 			return -EFAULT;
 	}
 
-	return do_utimes(filename, (tvs ? &ktvs[0] : NULL));
+	return do_utimes(AT_FDCWD, filename, (tvs ? &ktvs[0] : NULL));
 }
 
 /* These are here just in case some old sparc32 binary calls it. */
@@ -1003,7 +1003,7 @@ asmlinkage long sys32_adjtimex(struct timex32 __user *utp)
 asmlinkage long sparc32_open(const char __user *filename,
 			     int flags, int mode)
 {
-	return do_sys_open(filename, flags, mode);
+	return do_sys_open(AT_FDCWD, filename, flags, mode);
 }
 
 extern unsigned long do_mremap(unsigned long addr,

fs/compat.c

@@ -68,10 +68,10 @@ asmlinkage long compat_sys_utime(char __user *filename, struct compat_utimbuf __
 		tv[0].tv_usec = 0;
 		tv[1].tv_usec = 0;
 	}
-	return do_utimes(filename, t ? tv : NULL);
+	return do_utimes(AT_FDCWD, filename, t ? tv : NULL);
 }
 
-asmlinkage long compat_sys_utimes(char __user *filename, struct compat_timeval __user *t)
+asmlinkage long compat_sys_futimesat(int dfd, char __user *filename, struct compat_timeval __user *t)
 {
 	struct timeval tv[2];
 
@@ -82,14 +82,19 @@ asmlinkage long compat_sys_utimes(char __user *filename, struct compat_timeval _
 		    get_user(tv[1].tv_usec, &t[1].tv_usec))
 			return -EFAULT; 
 	} 
-	return do_utimes(filename, t ? tv : NULL);
+	return do_utimes(dfd, filename, t ? tv : NULL);
+}
+
+asmlinkage long compat_sys_utimes(char __user *filename, struct compat_timeval __user *t)
+{
+	return compat_sys_futimesat(AT_FDCWD, filename, t);
 }
 
 asmlinkage long compat_sys_newstat(char __user * filename,
 		struct compat_stat __user *statbuf)
 {
 	struct kstat stat;
-	int error = vfs_stat(filename, &stat);
+	int error = vfs_stat_fd(AT_FDCWD, filename, &stat);
 
 	if (!error)
 		error = cp_compat_stat(&stat, statbuf);
@@ -100,10 +105,31 @@ asmlinkage long compat_sys_newlstat(char __user * filename,
 		struct compat_stat __user *statbuf)
 {
 	struct kstat stat;
-	int error = vfs_lstat(filename, &stat);
+	int error = vfs_lstat_fd(AT_FDCWD, filename, &stat);
+
+	if (!error)
+		error = cp_compat_stat(&stat, statbuf);
+	return error;
+}
+
+asmlinkage long compat_sys_newfstatat(int dfd, char __user *filename,
+		struct compat_stat __user *statbuf, int flag)
+{
+	struct kstat stat;
+	int error = -EINVAL;
+
+	if ((flag & ~AT_SYMLINK_NOFOLLOW) != 0)
+		goto out;
+
+	if (flag & AT_SYMLINK_NOFOLLOW)
+		error = vfs_lstat_fd(dfd, filename, &stat);
+	else
+		error = vfs_stat_fd(dfd, filename, &stat);
 
 	if (!error)
 		error = cp_compat_stat(&stat, statbuf);
+
+out:
 	return error;
 }
 
@@ -1290,7 +1316,17 @@ compat_sys_writev(unsigned long fd, const struct compat_iovec __user *vec, unsig
 asmlinkage long
 compat_sys_open(const char __user *filename, int flags, int mode)
 {
-	return do_sys_open(filename, flags, mode);
+	return do_sys_open(AT_FDCWD, filename, flags, mode);
+}
+
+/*
+ * Exactly like fs/open.c:sys_openat(), except that it doesn't set the
+ * O_LARGEFILE flag.
+ */
+asmlinkage long
+compat_sys_openat(int dfd, const char __user *filename, int flags, int mode)
+{
+	return do_sys_open(dfd, filename, flags, mode);
 }
 
 /*

fs/exec.c

@@ -477,7 +477,7 @@ struct file *open_exec(const char *name)
 	int err;
 	struct file *file;
 
-	err = path_lookup_open(name, LOOKUP_FOLLOW, &nd, FMODE_READ);
+	err = path_lookup_open(AT_FDCWD, name, LOOKUP_FOLLOW, &nd, FMODE_READ);
 	file = ERR_PTR(err);
 
 	if (!err) {

fs/open.c

@@ -20,6 +20,7 @@
 #include <linux/security.h>
 #include <linux/mount.h>
 #include <linux/vfs.h>
+#include <linux/fcntl.h>
 #include <asm/uaccess.h>
 #include <linux/fs.h>
 #include <linux/personality.h>
@@ -414,14 +415,14 @@ asmlinkage long sys_utime(char __user * filename, struct utimbuf __user * times)
  * must be owner or have write permission.
  * Else, update from *times, must be owner or super user.
  */
-long do_utimes(char __user * filename, struct timeval * times)
+long do_utimes(int dfd, char __user *filename, struct timeval *times)
 {
 	int error;
 	struct nameidata nd;
 	struct inode * inode;
 	struct iattr newattrs;
 
-	error = user_path_walk(filename, &nd);
+	error = __user_walk_fd(dfd, filename, LOOKUP_FOLLOW, &nd);
 
 	if (error)
 		goto out;
@@ -461,13 +462,18 @@ long do_utimes(char __user * filename, struct timeval * times)
 	return error;
 }
 
-asmlinkage long sys_utimes(char __user * filename, struct timeval __user * utimes)
+asmlinkage long sys_futimesat(int dfd, char __user *filename, struct timeval __user *utimes)
 {
 	struct timeval times[2];
 
 	if (utimes && copy_from_user(&times, utimes, sizeof(times)))
 		return -EFAULT;
-	return do_utimes(filename, utimes ? times : NULL);
+	return do_utimes(dfd, filename, utimes ? times : NULL);
+}
+
+asmlinkage long sys_utimes(char __user *filename, struct timeval __user *utimes)
+{
+	return sys_futimesat(AT_FDCWD, filename, utimes);
 }
 
 
@@ -476,7 +482,7 @@ asmlinkage long sys_utimes(char __user * filename, struct timeval __user * utime
  * We do this by temporarily clearing all FS-related capabilities and
  * switching the fsuid/fsgid around to the real ones.
  */
-asmlinkage long sys_access(const char __user * filename, int mode)
+asmlinkage long sys_faccessat(int dfd, const char __user *filename, int mode)
 {
 	struct nameidata nd;
 	int old_fsuid, old_fsgid;
@@ -506,7 +512,7 @@ asmlinkage long sys_access(const char __user * filename, int mode)
 	else
 		current->cap_effective = current->cap_permitted;
 
-	res = __user_walk(filename, LOOKUP_FOLLOW|LOOKUP_ACCESS, &nd);
+	res = __user_walk_fd(dfd, filename, LOOKUP_FOLLOW|LOOKUP_ACCESS, &nd);
 	if (!res) {
 		res = vfs_permission(&nd, mode);
 		/* SuS v2 requires we report a read only fs too */
@@ -523,6 +529,11 @@ asmlinkage long sys_access(const char __user * filename, int mode)
 	return res;
 }
 
+asmlinkage long sys_access(const char __user *filename, int mode)
+{
+	return sys_faccessat(AT_FDCWD, filename, mode);
+}
+
 asmlinkage long sys_chdir(const char __user * filename)
 {
 	struct nameidata nd;
@@ -635,14 +646,15 @@ asmlinkage long sys_fchmod(unsigned int fd, mode_t mode)
 	return err;
 }
 
-asmlinkage long sys_chmod(const char __user * filename, mode_t mode)
+asmlinkage long sys_fchmodat(int dfd, const char __user *filename,
+			     mode_t mode)
 {
 	struct nameidata nd;
 	struct inode * inode;
 	int error;
 	struct iattr newattrs;
 
-	error = user_path_walk(filename, &nd);
+	error = __user_walk_fd(dfd, filename, LOOKUP_FOLLOW, &nd);
 	if (error)
 		goto out;
 	inode = nd.dentry->d_inode;
@@ -669,6 +681,11 @@ asmlinkage long sys_chmod(const char __user * filename, mode_t mode)
 	return error;
 }
 
+asmlinkage long sys_chmod(const char __user *filename, mode_t mode)
+{
+	return sys_fchmodat(AT_FDCWD, filename, mode);
+}
+
 static int chown_common(struct dentry * dentry, uid_t user, gid_t group)
 {
 	struct inode * inode;
@@ -717,6 +734,26 @@ asmlinkage long sys_chown(const char __user * filename, uid_t user, gid_t group)
 	return error;
 }
 
+asmlinkage long sys_fchownat(int dfd, const char __user *filename, uid_t user,
+			     gid_t group, int flag)
+{
+	struct nameidata nd;
+	int error = -EINVAL;
+	int follow;
+
+	if ((flag & ~AT_SYMLINK_NOFOLLOW) != 0)
+		goto out;
+
+	follow = (flag & AT_SYMLINK_NOFOLLOW) ? 0 : LOOKUP_FOLLOW;
+	error = __user_walk_fd(dfd, filename, follow, &nd);
+	if (!error) {
+		error = chown_common(nd.dentry, user, group);
+		path_release(&nd);
+	}
+out:
+	return error;
+}
+
 asmlinkage long sys_lchown(const char __user * filename, uid_t user, gid_t group)
 {
 	struct nameidata nd;
@@ -820,7 +857,8 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
  * for the internal routines (ie open_namei()/follow_link() etc). 00 is
  * used by symlinks.
  */
-struct file *filp_open(const char * filename, int flags, int mode)
+static struct file *do_filp_open(int dfd, const char *filename, int flags,
+				 int mode)
 {
 	int namei_flags, error;
 	struct nameidata nd;
@@ -829,12 +867,17 @@ struct file *filp_open(const char * filename, int flags, int mode)
 	if ((namei_flags+1) & O_ACCMODE)
 		namei_flags++;
 
-	error = open_namei(filename, namei_flags, mode, &nd);
+	error = open_namei(dfd, filename, namei_flags, mode, &nd);
 	if (!error)
 		return nameidata_to_filp(&nd, flags);
 
 	return ERR_PTR(error);
 }
+
+struct file *filp_open(const char *filename, int flags, int mode)
+{
+	return do_filp_open(AT_FDCWD, filename, flags, mode);
+}
 EXPORT_SYMBOL(filp_open);
 
 /**
@@ -1016,7 +1059,7 @@ void fastcall fd_install(unsigned int fd, struct file * file)
 
 EXPORT_SYMBOL(fd_install);
 
-long do_sys_open(const char __user *filename, int flags, int mode)
+long do_sys_open(int dfd, const char __user *filename, int flags, int mode)
 {
 	char *tmp = getname(filename);
 	int fd = PTR_ERR(tmp);
@@ -1024,7 +1067,7 @@ long do_sys_open(const char __user *filename, int flags, int mode)
 	if (!IS_ERR(tmp)) {
 		fd = get_unused_fd();
 		if (fd >= 0) {
-			struct file *f = filp_open(tmp, flags, mode);
+			struct file *f = do_filp_open(dfd, tmp, flags, mode);
 			if (IS_ERR(f)) {
 				put_unused_fd(fd);
 				fd = PTR_ERR(f);
@@ -1043,10 +1086,20 @@ asmlinkage long sys_open(const char __user *filename, int flags, int mode)
 	if (force_o_largefile())
 		flags |= O_LARGEFILE;
 
-	return do_sys_open(filename, flags, mode);
+	return do_sys_open(AT_FDCWD, filename, flags, mode);
 }
 EXPORT_SYMBOL_GPL(sys_open);
 
+asmlinkage long sys_openat(int dfd, const char __user *filename, int flags,
+			   int mode)
+{
+	if (force_o_largefile())
+		flags |= O_LARGEFILE;
+
+	return do_sys_open(dfd, filename, flags, mode);
+}
+EXPORT_SYMBOL_GPL(sys_openat);
+
 #ifndef __alpha__
 
 /*

fs/stat.c

@@ -63,12 +63,12 @@ int vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 
 EXPORT_SYMBOL(vfs_getattr);
 
-int vfs_stat(char __user *name, struct kstat *stat)
+int vfs_stat_fd(int dfd, char __user *name, struct kstat *stat)
 {
 	struct nameidata nd;
 	int error;
 
-	error = user_path_walk(name, &nd);
+	error = __user_walk_fd(dfd, name, LOOKUP_FOLLOW, &nd);
 	if (!error) {
 		error = vfs_getattr(nd.mnt, nd.dentry, stat);
 		path_release(&nd);
@@ -76,14 +76,19 @@ int vfs_stat(char __user *name, struct kstat *stat)
 	return error;
 }
 
+int vfs_stat(char __user *name, struct kstat *stat)
+{
+	return vfs_stat_fd(AT_FDCWD, name, stat);
+}
+
 EXPORT_SYMBOL(vfs_stat);
 
-int vfs_lstat(char __user *name, struct kstat *stat)
+int vfs_lstat_fd(int dfd, char __user *name, struct kstat *stat)
 {
 	struct nameidata nd;
 	int error;
 
-	error = user_path_walk_link(name, &nd);
+	error = __user_walk_fd(dfd, name, 0, &nd);
 	if (!error) {
 		error = vfs_getattr(nd.mnt, nd.dentry, stat);
 		path_release(&nd);
@@ -91,6 +96,11 @@ int vfs_lstat(char __user *name, struct kstat *stat)
 	return error;
 }
 
+int vfs_lstat(char __user *name, struct kstat *stat)
+{
+	return vfs_lstat_fd(AT_FDCWD, name, stat);
+}
+
 EXPORT_SYMBOL(vfs_lstat);
 
 int vfs_fstat(unsigned int fd, struct kstat *stat)
@@ -151,7 +161,7 @@ static int cp_old_stat(struct kstat *stat, struct __old_kernel_stat __user * sta
 asmlinkage long sys_stat(char __user * filename, struct __old_kernel_stat __user * statbuf)
 {
 	struct kstat stat;
-	int error = vfs_stat(filename, &stat);
+	int error = vfs_stat_fd(AT_FDCWD, filename, &stat);
 
 	if (!error)
 		error = cp_old_stat(&stat, statbuf);
@@ -161,7 +171,7 @@ asmlinkage long sys_stat(char __user * filename, struct __old_kernel_stat __user
 asmlinkage long sys_lstat(char __user * filename, struct __old_kernel_stat __user * statbuf)
 {
 	struct kstat stat;
-	int error = vfs_lstat(filename, &stat);
+	int error = vfs_lstat_fd(AT_FDCWD, filename, &stat);
 
 	if (!error)
 		error = cp_old_stat(&stat, statbuf);
@@ -229,27 +239,50 @@ static int cp_new_stat(struct kstat *stat, struct stat __user *statbuf)
 	return copy_to_user(statbuf,&tmp,sizeof(tmp)) ? -EFAULT : 0;
 }
 
-asmlinkage long sys_newstat(char __user * filename, struct stat __user * statbuf)
+asmlinkage long sys_newstat(char __user *filename, struct stat __user *statbuf)
 {
 	struct kstat stat;
-	int error = vfs_stat(filename, &stat);
+	int error = vfs_stat_fd(AT_FDCWD, filename, &stat);
 
 	if (!error)
 		error = cp_new_stat(&stat, statbuf);
 
 	return error;
 }
-asmlinkage long sys_newlstat(char __user * filename, struct stat __user * statbuf)
+
+asmlinkage long sys_newlstat(char __user *filename, struct stat __user *statbuf)
 {
 	struct kstat stat;
-	int error = vfs_lstat(filename, &stat);
+	int error = vfs_lstat_fd(AT_FDCWD, filename, &stat);
 
 	if (!error)
 		error = cp_new_stat(&stat, statbuf);
 
 	return error;
 }
-asmlinkage long sys_newfstat(unsigned int fd, struct stat __user * statbuf)
+
+asmlinkage long sys_newfstatat(int dfd, char __user *filename,
+				struct stat __user *statbuf, int flag)
+{
+	struct kstat stat;
+	int error = -EINVAL;
+
+	if ((flag & ~AT_SYMLINK_NOFOLLOW) != 0)
+		goto out;
+
+	if (flag & AT_SYMLINK_NOFOLLOW)
+		error = vfs_lstat_fd(dfd, filename, &stat);
+	else
+		error = vfs_stat_fd(dfd, filename, &stat);
+
+	if (!error)
+		error = cp_new_stat(&stat, statbuf);
+
+out:
+	return error;
+}
+
+asmlinkage long sys_newfstat(unsigned int fd, struct stat __user *statbuf)
 {
 	struct kstat stat;
 	int error = vfs_fstat(fd, &stat);
@@ -260,7 +293,8 @@ asmlinkage long sys_newfstat(unsigned int fd, struct stat __user * statbuf)
 	return error;
 }
 
-asmlinkage long sys_readlink(const char __user * path, char __user * buf, int bufsiz)
+asmlinkage long sys_readlinkat(int dfd, const char __user *path,
+				char __user *buf, int bufsiz)
 {
 	struct nameidata nd;
 	int error;
@@ -268,7 +302,7 @@ asmlinkage long sys_readlink(const char __user * path, char __user * buf, int bu
 	if (bufsiz <= 0)
 		return -EINVAL;
 
-	error = user_path_walk_link(path, &nd);
+	error = __user_walk_fd(dfd, path, 0, &nd);
 	if (!error) {
 		struct inode * inode = nd.dentry->d_inode;
 
@@ -285,6 +319,12 @@ asmlinkage long sys_readlink(const char __user * path, char __user * buf, int bu
 	return error;
 }
 
+asmlinkage long sys_readlink(const char __user *path, char __user *buf,
+				int bufsiz)
+{
+	return sys_readlinkat(AT_FDCWD, path, buf, bufsiz);
+}
+
 
 /* ---------- LFS-64 ----------- */
 #ifdef __ARCH_WANT_STAT64

include/linux/fcntl.h

@@ -23,6 +23,13 @@
 #define DN_ATTRIB	0x00000020	/* File changed attibutes */
 #define DN_MULTISHOT	0x80000000	/* Don't remove notifier */
 
+#define AT_FDCWD		-100    /* Special value used to indicate
+                                           openat should use the current
+                                           working directory. */
+#define AT_SYMLINK_NOFOLLOW	0x100   /* Do not follow symbolic links.  */
+#define AT_REMOVEDIR		0x200   /* Remove directory instead of
+                                           unlinking file.  */
+
 #ifdef __KERNEL__
 
 #ifndef force_o_largefile

include/linux/fs.h

@@ -1340,7 +1340,8 @@ static inline int break_lease(struct inode *inode, unsigned int mode)
 
 extern int do_truncate(struct dentry *, loff_t start, unsigned int time_attrs,
 		       struct file *filp);
-extern long do_sys_open(const char __user *filename, int flags, int mode);
+extern long do_sys_open(int fdf, const char __user *filename, int flags,
+			int mode);
 extern struct file *filp_open(const char *, int, int);
 extern struct file * dentry_open(struct dentry *, struct vfsmount *, int);
 extern int filp_close(struct file *, fl_owner_t id);
@@ -1479,7 +1480,7 @@ static inline void allow_write_access(struct file *file)
 }
 extern int do_pipe(int *);
 
-extern int open_namei(const char *, int, int, struct nameidata *);
+extern int open_namei(int dfd, const char *, int, int, struct nameidata *);
 extern int may_open(struct nameidata *, int, int);
 
 extern int kernel_read(struct file *, unsigned long, char *, unsigned long);
@@ -1677,6 +1678,8 @@ extern int vfs_readdir(struct file *, filldir_t, void *);
 
 extern int vfs_stat(char __user *, struct kstat *);
 extern int vfs_lstat(char __user *, struct kstat *);
+extern int vfs_stat_fd(int dfd, char __user *, struct kstat *);
+extern int vfs_lstat_fd(int dfd, char __user *, struct kstat *);
 extern int vfs_fstat(unsigned int, struct kstat *);
 
 extern int vfs_ioctl(struct file *, unsigned int, unsigned int, unsigned long);

include/linux/namei.h

@@ -56,10 +56,11 @@ enum {LAST_NORM, LAST_ROOT, LAST_DOT, LAST_DOTDOT, LAST_BIND};
 #define LOOKUP_ACCESS		(0x0400)
 
 extern int FASTCALL(__user_walk(const char __user *, unsigned, struct nameidata *));
+extern int FASTCALL(__user_walk_fd(int dfd, const char __user *, unsigned, struct nameidata *));
 #define user_path_walk(name,nd) \
-	__user_walk(name, LOOKUP_FOLLOW, nd)
+	__user_walk_fd(AT_FDCWD, name, LOOKUP_FOLLOW, nd)
 #define user_path_walk_link(name,nd) \
-	__user_walk(name, 0, nd)
+	__user_walk_fd(AT_FDCWD, name, 0, nd)
 extern int FASTCALL(path_lookup(const char *, unsigned, struct nameidata *));
 extern int FASTCALL(path_walk(const char *, struct nameidata *));
 extern int FASTCALL(link_path_walk(const char *, struct nameidata *));
@@ -67,7 +68,7 @@ extern void path_release(struct nameidata *);
 extern void path_release_on_umount(struct nameidata *);
 
 extern int __user_path_lookup_open(const char __user *, unsigned lookup_flags, struct nameidata *nd, int open_flags);
-extern int path_lookup_open(const char *, unsigned lookup_flags, struct nameidata *, int open_flags);
+extern int path_lookup_open(int dfd, const char *name, unsigned lookup_flags, struct nameidata *, int open_flags);
 extern struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry,
 		int (*open)(struct inode *, struct file *));
 extern struct file *nameidata_to_filp(struct nameidata *nd, int flags);

include/linux/time.h

@@ -74,7 +74,7 @@ extern void do_gettimeofday(struct timeval *tv);
 extern int do_settimeofday(struct timespec *tv);
 extern int do_sys_settimeofday(struct timespec *tv, struct timezone *tz);
 #define do_posix_clock_monotonic_gettime(ts) ktime_get_ts(ts)
-extern long do_utimes(char __user *filename, struct timeval *times);
+extern long do_utimes(int dfd, char __user *filename, struct timeval *times);
 struct itimerval;
 extern int do_setitimer(int which, struct itimerval *value,
 			struct itimerval *ovalue);