[RFC][PATCH 0/6] Enhanced file stat system call

classic Classic list List threaded Threaded
51 messages Options
123
Reply | Threaded
Open this post in threaded view
|

[RFC][PATCH 0/6] Enhanced file stat system call

David Howells

Implement a new system call to provide enhanced file stats.  The patches can
be found here:

        http://git.kernel.org/cgit/linux/kernel/git/dhowells/linux-fs.git/log/?h=xstat


===========
DESCRIPTION
===========

The first patch provides this new system call:

        long ret = statx(int dfd,
                         const char *filename,
                         unsigned atflag,
                         unsigned mask,
                         struct statx *buffer);

This is an enhanced file stat function that provides a number of useful
features, in summary:

 (1) More information: creation time, data version number, inode generation
     number and flags.  A subset of these is available through a number of
     filesystems (such as CIFS, NFS, AFS, Ext4 and BTRFS).

 (2) Lightweight stat (AT_NO_ATTR_SYNC): Ask for just those details of
     interest, and allow a network fs to approximate anything not of
     interest, without going to the server.

 (3) Heavyweight stat (AT_FORCE_ATTR_SYNC): Force a network fs to flush
     buffers and go to the server, even if it thinks its cached attributes
     are up to date.

 (4) Allow the filesystem to indicate what it can/cannot provide: A
     filesystem can now say it doesn't support a standard stat feature if
     that isn't available.

 (5) Make the fields a consistent size on all arches, and make them large.

 (6) Can be extended by using more request flags and using up the padding
     space in the statx struct.

Note that no lstat() equivalent is required as that can be implemented
through statx() with atflag == 0.  There is also no fstat() equivalent as
that can be implemented through statx() with filename == NULL and the
relevant fd passed as dfd.


=======
TESTING
=======

A test program is added into samples/statx/ by the first patch.

David
---
David Howells (6):
      statx: Add a system call to make enhanced file info available
      statx: AFS: Return enhanced file attributes
      statx: Ext4: Return enhanced file attributes
      statx: NFS: Return enhanced file attributes
      statx: Make windows attributes available for CIFS, NTFS and FAT to use
      statx: CIFS: Return enhanced attributes


 arch/x86/entry/syscalls/syscall_32.tbl |    1
 arch/x86/entry/syscalls/syscall_64.tbl |    1
 fs/afs/inode.c                         |   24 ++-
 fs/cifs/cifsfs.h                       |    4
 fs/cifs/cifsglob.h                     |    8 +
 fs/cifs/dir.c                          |    2
 fs/cifs/inode.c                        |  117 +++++++++---
 fs/exportfs/expfs.c                    |    4
 fs/ext4/ext4.h                         |    2
 fs/ext4/file.c                         |    2
 fs/ext4/inode.c                        |   30 +++
 fs/ext4/namei.c                        |    2
 fs/ext4/symlink.c                      |    2
 fs/nfs/inode.c                         |   41 ++++
 fs/stat.c                              |  306 +++++++++++++++++++++++++++++---
 include/linux/fs.h                     |    5 -
 include/linux/stat.h                   |   15 +-
 include/linux/syscalls.h               |    4
 include/uapi/linux/fcntl.h             |    2
 include/uapi/linux/stat.h              |  135 ++++++++++++++
 samples/Makefile                       |    2
 samples/statx/Makefile                 |   10 +
 samples/statx/test-statx.c             |  274 +++++++++++++++++++++++++++++
 23 files changed, 910 insertions(+), 83 deletions(-)
 create mode 100644 samples/statx/Makefile
 create mode 100644 samples/statx/test-statx.c

Reply | Threaded
Open this post in threaded view
|

[PATCH 2/6] statx: AFS: Return enhanced file attributes

David Howells
Return enhanced file attributes from the AFS filesystem.  This includes the
following:

 (1) The data version number as st_version, setting STATX_VERSION.

 (2) STATX_INFO_AUTOMOUNT will be set on automount directories by virtue of
     S_AUTOMOUNT being set on the inode.  These are referrals to other
     volumes or other cells.

 (3) STATX_INFO_AUTODIR on a directory that does cell lookup for
     non-existent names and mounts them (typically mounted on /afs with -o
     autocell).  The resulting directories are marked STATX_INFO_FABRICATED
     as they do not actually exist in the mounted AFS directory.

 (4) Files, directories and symlinks accessed over AFS are marked
     STATX_INFO_REMOTE.  Local fake directories are marked
     STATX_INFO_FABRICATED.

 (5) STATX_INFO_NONSYSTEM_OWNERSHIP is set as the UID and GID retrieved
     from an AFS share may not be applicable on the system.

STATX_ATIME, STATX_CTIME and STATX_BLOCKS are cleared as AFS does not
support them.

Example output:

        [root@andromeda ~]# ./samples/statx/test-statx /afs
        statx(/afs) = 0
        results=7ef
          Size: 2048            Blocks: 0          IO Block: 4096    directory
        Device: 00:25           Inode: 1           Links: 2
        Access: (0777/drwxrwxrwx)  Uid:     0   Gid:     0
        Access: 2006-05-07 00:21:15.000000000+0100
        Modify: 2006-05-07 00:21:15.000000000+0100
        Change: 2006-05-07 00:21:15.000000000+0100
        IO-blocksize: blksize=4096

Signed-off-by: David Howells <[hidden email]>
---

 fs/afs/inode.c |   24 +++++++++++++++++++-----
 1 file changed, 19 insertions(+), 5 deletions(-)

diff --git a/fs/afs/inode.c b/fs/afs/inode.c
index 86cc7264c21c..e9958d5e267e 100644
--- a/fs/afs/inode.c
+++ b/fs/afs/inode.c
@@ -72,9 +72,9 @@ static int afs_inode_map_status(struct afs_vnode *vnode, struct key *key)
  inode->i_uid = vnode->status.owner;
  inode->i_gid = GLOBAL_ROOT_GID;
  inode->i_size = vnode->status.size;
- inode->i_ctime.tv_sec = vnode->status.mtime_server;
- inode->i_ctime.tv_nsec = 0;
- inode->i_atime = inode->i_mtime = inode->i_ctime;
+ inode->i_mtime.tv_sec = vnode->status.mtime_server;
+ inode->i_mtime.tv_nsec = 0;
+ inode->i_atime = inode->i_ctime = inode->i_mtime;
  inode->i_blocks = 0;
  inode->i_generation = vnode->fid.unique;
  inode->i_version = vnode->status.data_version;
@@ -375,8 +375,7 @@ error_unlock:
 /*
  * read the attributes of an inode
  */
-int afs_getattr(struct vfsmount *mnt, struct dentry *dentry,
-      struct kstat *stat)
+int afs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 {
  struct inode *inode;
 
@@ -385,6 +384,21 @@ int afs_getattr(struct vfsmount *mnt, struct dentry *dentry,
  _enter("{ ino=%lu v=%u }", inode->i_ino, inode->i_generation);
 
  generic_fillattr(inode, stat);
+
+ stat->result_mask &= ~(STATX_ATIME | STATX_CTIME | STATX_BLOCKS);
+ stat->result_mask |= STATX_VERSION | STATX_GEN;
+ stat->version = inode->i_version;
+ stat->gen = inode->i_generation;
+
+ if (test_bit(AFS_VNODE_AUTOCELL, &AFS_FS_I(inode)->flags))
+ stat->information |= STATX_INFO_AUTODIR;
+
+ if (test_bit(AFS_VNODE_PSEUDODIR, &AFS_FS_I(inode)->flags))
+ stat->information |= STATX_INFO_FABRICATED;
+ else
+ stat->information |= STATX_INFO_REMOTE;
+
+ stat->information |= STATX_INFO_NONSYSTEM_OWNERSHIP;
  return 0;
 }
 

Reply | Threaded
Open this post in threaded view
|

[PATCH 1/6] statx: Add a system call to make enhanced file info available

David Howells
In reply to this post by David Howells
Add a system call to make extended file information available, including
file creation time, inode version and data version where available through
the underlying filesystem.


========
OVERVIEW
========

The idea was initially proposed as a set of xattrs that could be retrieved
with getxattr(), but the general preferance proved to be for a new syscall
with an extended stat structure.

This has a number of uses:

 (1) Better support for the y2038 problem [Arnd Bergmann].

 (2) Creation time: The SMB protocol carries the creation time, which could
     be exported by Samba, which will in turn help CIFS make use of
     FS-Cache as that can be used for coherency data.

     This is also specified in NFSv4 as a recommended attribute and could
     be exported by NFSD [Steve French].

 (3) Lightweight stat: Ask for just those details of interest, and allow a
     netfs (such as NFS) to approximate anything not of interest, possibly
     without going to the server [Trond Myklebust, Ulrich Drepper, Andreas
     Dilger].

 (4) Heavyweight stat: Force a netfs to go to the server, even if it thinks
     its cached attributes are up to date [Trond Myklebust].

 (5) Data version number: Could be used by userspace NFS servers [Aneesh
     Kumar].

     Can also be used to modify fill_post_wcc() in NFSD which retrieves
     i_version directly, but has just called vfs_getattr().  It could get
     it from the kstat struct if it used vfs_xgetattr() instead.

 (6) BSD stat compatibility: Including more fields from the BSD stat such
     as creation time (st_btime) and inode generation number (st_gen)
     [Jeremy Allison, Bernd Schubert].

 (7) Inode generation number: Useful for FUSE and userspace NFS servers
     [Bernd Schubert].  This was asked for but later deemed unnecessary
     with the open-by-handle capability available

 (8) Extra coherency data may be useful in making backups [Andreas Dilger].

 (9) Allow the filesystem to indicate what it can/cannot provide: A
     filesystem can now say it doesn't support a standard stat feature if
     that isn't available, so if, for instance, inode numbers or UIDs don't
     exist or are fabricated locally...

(10) Make the fields a consistent size on all arches and make them large.

(11) Store a 16-byte volume ID in the superblock that can be returned in
     struct xstat [Steve French].

(12) Include granularity fields in the time data to indicate the
     granularity of each of the times (NFSv4 time_delta) [Steve French].

(13) FS_IOC_GETFLAGS value.  These could be translated to BSD's st_flags.
     Note that the Linux IOC flags are a mess and filesystems such as Ext4
     define flags that aren't in linux/fs.h, so translation in the kernel
     may be a necessity (or, possibly, we provide the filesystem type too).

(14) Mask of features available on file (eg: ACLs, seclabel) [Brad Boyer,
     Michael Kerrisk].

(15) Spare space, request flags and information flags are provided for
     future expansion.

Note that not all of the above are implemented here.


===============
NEW SYSTEM CALL
===============

The new system call is:

        int ret = statx(int dfd,
                        const char *filename,
                        unsigned int flags,
                        unsigned int mask,
                        struct statx *buffer);

The dfd, filename and flags parameters indicate the file to query.  There
is no equivalent of lstat() as that can be emulated with statx() by passing
AT_SYMLINK_NOFOLLOW in flags.  There is also no equivalent of fstat() as
that can be emulated by passing a NULL filename to statx() with the fd of
interest in dfd.

AT_FORCE_ATTR_SYNC can be set in flags.  This will require a network
filesystem to synchronise its attributes with the server.

AT_NO_ATTR_SYNC can be set in flags.  This will suppress synchronisation
with the server in a network filesystem.  The resulting values should be
considered approximate.

mask is a bitmask indicating the fields in struct statx that are of
interest to the caller.  The user should set this to STATX_BASIC_STATS to
get the basic set returned by stat().

buffer points to the destination for the data.  This must be 256 bytes in
size.


======================
MAIN ATTRIBUTES RECORD
======================

The following structures are defined in which to return the main attribute
set:

        struct statx {
                __u32 st_mask;
                __u32 st_information;
                __u32 st_blksize;
                __u32 st_nlink;
                __u32 st_gen;
                __u32 st_uid;
                __u32 st_gid;
                __u16 st_mode;
                __u16 __spare0[1];
                __u64 st_ino;
                __u64 st_size;
                __u64 st_blocks;
                __u64 st_version;
                __s64 st_atime;
                __s64 st_btime;
                __s64 st_ctime;
                __s64 st_mtime;
                __s32 st_atime_ns;
                __s32 st_btime_ns;
                __s32 st_ctime_ns;
                __s32 st_mtime_ns;
                __u32 st_rdev_major;
                __u32 st_rdev_minor;
                __u32 st_dev_major;
                __u32 st_dev_minor;
                __u64 __spare1[16];
        };

where st_information is local system information about the file, st_gen is
the inode generation number, st_btime is the file creation time, st_version
is the data version number (i_version), st_mask is a bitmask indicating the
data provided and __spares*[] are where as-yet undefined fields can be
placed.

Time fields are split into separate seconds and nanoseconds fields to make
packing easier and the granularities can be queried with the filesystem
info system call.  Note that times will be negative if before 1970; in such
a case, the nanosecond fields should also be negative if not zero.

The defined bits in request_mask and st_mask are:

        STATX_MODE Want/got st_mode
        STATX_NLINK Want/got st_nlink
        STATX_UID Want/got st_uid
        STATX_GID Want/got st_gid
        STATX_RDEV Want/got st_rdev_*
        STATX_ATIME Want/got st_atime
        STATX_MTIME Want/got st_mtime
        STATX_CTIME Want/got st_ctime
        STATX_INO Want/got st_ino
        STATX_SIZE Want/got st_size
        STATX_BLOCKS Want/got st_blocks
        STATX_BASIC_STATS [The stuff in the normal stat struct]
        STATX_BTIME Want/got st_btime
        STATX_VERSION Want/got st_data_version
        STATX_GEN Want/got st_gen
        STATX_ALL_STATS [All currently available stuff]

The defined bits in the st_information field give local system data on a
file, how it is accessed, where it is and what it does:

        STATX_INFO_ENCRYPTED File is encrypted
        STATX_INFO_TEMPORARY File is temporary
        STATX_INFO_FABRICATED File was made up by filesystem
        STATX_INFO_KERNEL_API File is kernel API (eg: procfs/sysfs)
        STATX_INFO_REMOTE File is remote
        STATX_INFO_AUTOMOUNT Dir is automount trigger
        STATX_INFO_AUTODIR Dir provides unlisted automounts
        STATX_INFO_NONSYSTEM_OWNERSHIP File has non-system ownership details

These are for the use of GUI tools that might want to mark files specially,
depending on what they are.

Fields in struct statx come in a number of classes:

 (0) st_information, st_dev_*, st_blksize.

     These are local data and are always available.

 (1) st_nlinks, st_uid, st_gid, st_[amc]time*, st_ino, st_size, st_blocks.

     These will be returned whether the caller asks for them or not.  The
     corresponding bits in st_mask will be set to indicate whether they
     actually have valid values.

     If the caller didn't ask for them, then they may be approximated.  For
     example, NFS won't waste any time updating them from the server,
     unless as a byproduct of updating something requested.

     If the values don't actually exist for the underlying object (such as
     UID or GID on a DOS file), then the bit won't be set in the st_mask,
     even if the caller asked for the value.  In such a case, the returned
     value will be a fabrication.

 (2) st_mode.

     The part of this that identifies the file type will always be
     available, irrespective of the setting of STATX_MODE.  The access
     flags and sticky bit are as for class (1).

 (3) st_rdev_*.

     As for class (1), but this will be cleared if the file is not a
     blockdev or chardev.  The bit will be cleared if the value is not
     returned.

 (4) File creation time (st_btime*), data version (st_version), inode
     generation number (st_gen).

     These will be returned if available whether the caller asked for them or
     not.  The corresponding bits in st_mask will be set or cleared as
     appropriate to indicate a valid value.

     If the caller didn't ask for them, then they may be approximated.  For
     example, NFS won't waste any time updating them from the server, unless
     as a byproduct of updating something requested.


=======
TESTING
=======

The following test program can be used to test the statx system call:

        samples/statx/test-statx.c

Just compile and run, passing it paths to the files you want to examine.
The file is built automatically if CONFIG_SAMPLES is enabled.

Here's some example output.  Firstly, an NFS directory that crosses to
another FSID.  Note that the FABRICATED and AUTOMOUNT info flags are set.
The former because the directory is invented locally as we don't see the
underlying dir on the server, the latter because transiting this directory
will cause d_automount to be invoked by the VFS.

        [root@andromeda tmp]# ./samples/statx/test-statx -A /warthog/data
        statx(/warthog/data) = 0
        results=4fef
          Size: 4096            Blocks: 8          IO Block: 1048576  directory
        Device: 00:1d           Inode: 2           Links: 110
        Access: (3777/drwxrwxrwx)  Uid: -2
        Gid: 4294967294
        Access: 2012-04-30 09:01:55.283819565+0100
        Modify: 2012-03-28 19:01:19.405465361+0100
        Change: 2012-03-28 19:01:19.405465361+0100
        Data version: ef51734f11e92a18h
        Information: 00000134 (-------- -------- -------a --mr-f--)

Secondly, the result of automounting on that directory.

        [root@andromeda tmp]# ./samples/statx/test-statx /warthog/data
        statx(/warthog/data) = 0
        results=14fef
          Size: 4096            Blocks: 8          IO Block: 1048576  directory
        Device: 00:1e           Inode: 2           Links: 110
        Access: (3777/drwxrwxrwx)  Uid: -2
        Gid: 4294967294
        Access: 2012-04-30 09:01:55.283819565+0100
        Modify: 2012-03-28 19:01:19.405465361+0100
        Change: 2012-03-28 19:01:19.405465361+0100
        Data version: ef51734f11e92a18h
        Information: 00000110 (-------- -------- -------a ---r----)

Signed-off-by: David Howells <[hidden email]>
---

 arch/x86/entry/syscalls/syscall_32.tbl |    1
 arch/x86/entry/syscalls/syscall_64.tbl |    1
 fs/exportfs/expfs.c                    |    4
 fs/stat.c                              |  303 +++++++++++++++++++++++++++++---
 include/linux/fs.h                     |    5 -
 include/linux/stat.h                   |   15 +-
 include/linux/syscalls.h               |    4
 include/uapi/linux/fcntl.h             |    2
 include/uapi/linux/stat.h              |  109 ++++++++++++
 samples/Makefile                       |    2
 samples/statx/Makefile                 |   10 +
 samples/statx/test-statx.c             |  243 ++++++++++++++++++++++++++
 12 files changed, 662 insertions(+), 37 deletions(-)
 create mode 100644 samples/statx/Makefile
 create mode 100644 samples/statx/test-statx.c

diff --git a/arch/x86/entry/syscalls/syscall_32.tbl b/arch/x86/entry/syscalls/syscall_32.tbl
index b30dd8154cc2..b99a6b3a167c 100644
--- a/arch/x86/entry/syscalls/syscall_32.tbl
+++ b/arch/x86/entry/syscalls/syscall_32.tbl
@@ -386,3 +386,4 @@
 377 i386 copy_file_range sys_copy_file_range
 378 i386 preadv2 sys_preadv2
 379 i386 pwritev2 sys_pwritev2
+380 i386 statx sys_statx
diff --git a/arch/x86/entry/syscalls/syscall_64.tbl b/arch/x86/entry/syscalls/syscall_64.tbl
index cac6d17ce5db..6d5ef6c87cdc 100644
--- a/arch/x86/entry/syscalls/syscall_64.tbl
+++ b/arch/x86/entry/syscalls/syscall_64.tbl
@@ -335,6 +335,7 @@
 326 common copy_file_range sys_copy_file_range
 327 64 preadv2 sys_preadv2
 328 64 pwritev2 sys_pwritev2
+329 common statx sys_statx
 
 #
 # x32-specific system call numbers start at 512 to avoid cache impact
diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c
index c46f1a190b8d..cd6d9cbc9300 100644
--- a/fs/exportfs/expfs.c
+++ b/fs/exportfs/expfs.c
@@ -295,7 +295,9 @@ static int get_name(const struct path *path, char *name, struct dentry *child)
  * filesystem supports 64-bit inode numbers.  So we need to
  * actually call ->getattr, not just read i_ino:
  */
- error = vfs_getattr_nosec(&child_path, &stat);
+ stat.query_flags = 0;
+ stat.request_mask = STATX_BASIC_STATS;
+ error = vfs_xgetattr_nosec(&child_path, &stat);
  if (error)
  return error;
  buffer.ino = stat.ino;
diff --git a/fs/stat.c b/fs/stat.c
index bc045c7994e1..c2f8370dab13 100644
--- a/fs/stat.c
+++ b/fs/stat.c
@@ -18,6 +18,15 @@
 #include <asm/uaccess.h>
 #include <asm/unistd.h>
 
+/**
+ * generic_fillattr - Fill in the basic attributes from the inode struct
+ * @inode: Inode to use as the source
+ * @stat: Where to fill in the attributes
+ *
+ * Fill in the basic attributes in the kstat structure from data that's to be
+ * found on the VFS inode structure.  This is the default if no getattr inode
+ * operation is supplied.
+ */
 void generic_fillattr(struct inode *inode, struct kstat *stat)
 {
  stat->dev = inode->i_sb->s_dev;
@@ -27,87 +36,197 @@ void generic_fillattr(struct inode *inode, struct kstat *stat)
  stat->uid = inode->i_uid;
  stat->gid = inode->i_gid;
  stat->rdev = inode->i_rdev;
- stat->size = i_size_read(inode);
- stat->atime = inode->i_atime;
  stat->mtime = inode->i_mtime;
  stat->ctime = inode->i_ctime;
- stat->blksize = (1 << inode->i_blkbits);
+ stat->size = i_size_read(inode);
  stat->blocks = inode->i_blocks;
-}
+ stat->blksize = 1 << inode->i_blkbits;
 
+ stat->result_mask |= STATX_BASIC_STATS & ~STATX_RDEV;
+ if (IS_NOATIME(inode))
+ stat->result_mask &= ~STATX_ATIME;
+ else
+ stat->atime = inode->i_atime;
+
+ if (S_ISREG(stat->mode) && stat->nlink == 0)
+ stat->information |= STATX_INFO_TEMPORARY;
+ if (IS_AUTOMOUNT(inode))
+ stat->information |= STATX_INFO_AUTOMOUNT;
+
+ if (unlikely(S_ISBLK(stat->mode) || S_ISCHR(stat->mode)))
+ stat->result_mask |= STATX_RDEV;
+}
 EXPORT_SYMBOL(generic_fillattr);
 
 /**
- * vfs_getattr_nosec - getattr without security checks
+ * vfs_xgetattr_nosec - getattr without security checks
  * @path: file to get attributes from
  * @stat: structure to return attributes in
  *
  * Get attributes without calling security_inode_getattr.
  *
- * Currently the only caller other than vfs_getattr is internal to the
- * filehandle lookup code, which uses only the inode number and returns
- * no attributes to any user.  Any other code probably wants
- * vfs_getattr.
+ * Currently the only caller other than vfs_xgetattr is internal to the
+ * filehandle lookup code, which uses only the inode number and returns no
+ * attributes to any user.  Any other code probably wants vfs_xgetattr.
+ *
+ * The caller must set stat->request_mask to indicate what they want and
+ * stat->query_flags to indicate whether the server should be queried.
  */
-int vfs_getattr_nosec(struct path *path, struct kstat *stat)
+int vfs_xgetattr_nosec(struct path *path, struct kstat *stat)
 {
  struct inode *inode = d_backing_inode(path->dentry);
 
+ stat->query_flags &= ~KSTAT_QUERY_FLAGS;
+ if ((stat->query_flags & AT_FORCE_ATTR_SYNC) &&
+    (stat->query_flags & AT_NO_ATTR_SYNC))
+ return -EINVAL;
+
+ stat->result_mask = 0;
+ stat->information = 0;
  if (inode->i_op->getattr)
  return inode->i_op->getattr(path->mnt, path->dentry, stat);
 
  generic_fillattr(inode, stat);
  return 0;
 }
+EXPORT_SYMBOL(vfs_xgetattr_nosec);
 
-EXPORT_SYMBOL(vfs_getattr_nosec);
-
-int vfs_getattr(struct path *path, struct kstat *stat)
+/*
+ * vfs_xgetattr - Get the enhanced basic attributes of a file
+ * @path: The file of interest
+ * @stat: Where to return the statistics
+ *
+ * Ask the filesystem for a file's attributes.  The caller must have preset
+ * stat->request_mask and stat->query_flags to indicate what they want.
+ *
+ * If the file is remote, the filesystem can be forced to update the attributes
+ * from the backing store by passing AT_FORCE_ATTR_SYNC in query_flags or can
+ * suppress the update by passing AT_NO_ATTR_SYNC.
+ *
+ * Bits must have been set in stat->request_mask to indicate which attributes
+ * the caller wants retrieving.  Any such attribute not requested may be
+ * returned anyway, but the value may be approximate, and, if remote, may not
+ * have been synchronised with the server.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_xgetattr(struct path *path, struct kstat *stat)
 {
  int retval;
 
  retval = security_inode_getattr(path);
  if (retval)
  return retval;
- return vfs_getattr_nosec(path, stat);
+ return vfs_xgetattr_nosec(path, stat);
 }
+EXPORT_SYMBOL(vfs_xgetattr);
 
+/**
+ * vfs_getattr - Get the basic attributes of a file
+ * @path: The file of interest
+ * @stat: Where to return the statistics
+ *
+ * Ask the filesystem for a file's attributes.  If remote, the filesystem isn't
+ * forced to update its files from the backing store.  Only the basic set of
+ * attributes will be retrieved; anyone wanting more must use vfs_xgetattr(),
+ * as must anyone who wants to force attributes to be sync'd with the server.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_getattr(struct path *path, struct kstat *stat)
+{
+ stat->query_flags = 0;
+ stat->request_mask = STATX_BASIC_STATS;
+ return vfs_xgetattr(path, stat);
+}
 EXPORT_SYMBOL(vfs_getattr);
 
-int vfs_fstat(unsigned int fd, struct kstat *stat)
+/**
+ * vfs_fstatx - Get the enhanced basic attributes by file descriptor
+ * @fd: The file descriptor referring to the file of interest
+ * @stat: The result structure to fill in.
+ *
+ * This function is a wrapper around vfs_xgetattr().  The main difference is
+ * that it uses a file descriptor to determine the file location.
+ *
+ * The caller must have preset stat->query_flags and stat->request_mask as for
+ * vfs_xgetattr().
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_fstatx(unsigned int fd, struct kstat *stat)
 {
  struct fd f = fdget_raw(fd);
  int error = -EBADF;
 
  if (f.file) {
- error = vfs_getattr(&f.file->f_path, stat);
+ error = vfs_xgetattr(&f.file->f_path, stat);
  fdput(f);
  }
  return error;
 }
+EXPORT_SYMBOL(vfs_fstatx);
+
+/**
+ * vfs_fstat - Get basic attributes by file descriptor
+ * @fd: The file descriptor referring to the file of interest
+ * @stat: The result structure to fill in.
+ *
+ * This function is a wrapper around vfs_getattr().  The main difference is
+ * that it uses a file descriptor to determine the file location.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_fstat(unsigned int fd, struct kstat *stat)
+{
+ stat->query_flags = 0;
+ stat->request_mask = STATX_BASIC_STATS;
+ return vfs_fstatx(fd, stat);
+}
 EXPORT_SYMBOL(vfs_fstat);
 
-int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat,
- int flag)
+/**
+ * vfs_statx - Get basic and extra attributes by filename
+ * @dfd: A file descriptor representing the base dir for a relative filename
+ * @filename: The name of the file of interest
+ * @flags: Flags to control the query
+ * @stat: The result structure to fill in.
+ *
+ * This function is a wrapper around vfs_xgetattr().  The main difference is
+ * that it uses a filename and base directory to determine the file location.
+ * Additionally, the addition of AT_SYMLINK_NOFOLLOW to flags will prevent a
+ * symlink at the given name from being referenced.
+ *
+ * The caller must have preset stat->request_mask as for vfs_xgetattr().  The
+ * flags are also used to load up stat->query_flags.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_statx(int dfd, const char __user *filename, int flags,
+      struct kstat *stat)
 {
  struct path path;
  int error = -EINVAL;
- unsigned int lookup_flags = 0;
+ unsigned int lookup_flags = LOOKUP_FOLLOW | LOOKUP_AUTOMOUNT;
 
- if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_NO_AUTOMOUNT |
-      AT_EMPTY_PATH)) != 0)
- goto out;
+ if ((flags & ~(AT_SYMLINK_NOFOLLOW | AT_NO_AUTOMOUNT |
+       AT_EMPTY_PATH | KSTAT_QUERY_FLAGS)) != 0)
+ return -EINVAL;
 
- if (!(flag & AT_SYMLINK_NOFOLLOW))
- lookup_flags |= LOOKUP_FOLLOW;
- if (flag & AT_EMPTY_PATH)
+ if (flags & AT_SYMLINK_NOFOLLOW)
+ lookup_flags &= ~LOOKUP_FOLLOW;
+ if (flags & AT_NO_AUTOMOUNT)
+ lookup_flags &= ~LOOKUP_AUTOMOUNT;
+ if (flags & AT_EMPTY_PATH)
  lookup_flags |= LOOKUP_EMPTY;
+ stat->query_flags = flags;
+
 retry:
  error = user_path_at(dfd, filename, lookup_flags, &path);
  if (error)
  goto out;
 
- error = vfs_getattr(&path, stat);
+ error = vfs_xgetattr(&path, stat);
  path_put(&path);
  if (retry_estale(error, lookup_flags)) {
  lookup_flags |= LOOKUP_REVAL;
@@ -116,17 +235,65 @@ retry:
 out:
  return error;
 }
+EXPORT_SYMBOL(vfs_statx);
+
+/**
+ * vfs_fstatat - Get basic attributes by filename
+ * @dfd: A file descriptor representing the base dir for a relative filename
+ * @filename: The name of the file of interest
+ * @flags: Flags to control the query
+ * @stat: The result structure to fill in.
+ *
+ * This function is a wrapper around vfs_statx().  The difference is that it
+ * preselects basic stats only.  The flags are used to load up
+ * stat->query_flags in addition to indicating symlink handling during path
+ * resolution.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat,
+ int flags)
+{
+ stat->request_mask = STATX_BASIC_STATS;
+ return vfs_statx(dfd, filename, flags, stat);
+}
 EXPORT_SYMBOL(vfs_fstatat);
 
-int vfs_stat(const char __user *name, struct kstat *stat)
+/**
+ * vfs_stat - Get basic attributes by filename
+ * @filename: The name of the file of interest
+ * @stat: The result structure to fill in.
+ *
+ * This function is a wrapper around vfs_statx().  The difference is that it
+ * preselects basic stats only, terminal symlinks are followed regardless and a
+ * remote filesystem can't be forced to query the server.  If such is desired,
+ * vfs_statx() should be used instead.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
+int vfs_stat(const char __user *filename, struct kstat *stat)
 {
- return vfs_fstatat(AT_FDCWD, name, stat, 0);
+ stat->request_mask = STATX_BASIC_STATS;
+ return vfs_statx(AT_FDCWD, filename, 0, stat);
 }
 EXPORT_SYMBOL(vfs_stat);
 
+/**
+ * vfs_lstat - Get basic attrs by filename, without following terminal symlink
+ * @filename: The name of the file of interest
+ * @stat: The result structure to fill in.
+ *
+ * This function is a wrapper around vfs_statx().  The difference is that it
+ * preselects basic stats only, terminal symlinks are note followed regardless
+ * and a remote filesystem can't be forced to query the server.  If such is
+ * desired, vfs_statx() should be used instead.
+ *
+ * 0 will be returned on success, and a -ve error code if unsuccessful.
+ */
 int vfs_lstat(const char __user *name, struct kstat *stat)
 {
- return vfs_fstatat(AT_FDCWD, name, stat, AT_SYMLINK_NOFOLLOW);
+ stat->request_mask = STATX_BASIC_STATS;
+ return vfs_statx(AT_FDCWD, name, AT_SYMLINK_NOFOLLOW, stat);
 }
 EXPORT_SYMBOL(vfs_lstat);
 
@@ -141,7 +308,7 @@ static int cp_old_stat(struct kstat *stat, struct __old_kernel_stat __user * sta
 {
  static int warncount = 5;
  struct __old_kernel_stat tmp;
-
+
  if (warncount > 0) {
  warncount--;
  printk(KERN_WARNING "VFS: Warning: %s using old stat() call. Recompile your binary.\n",
@@ -166,7 +333,7 @@ static int cp_old_stat(struct kstat *stat, struct __old_kernel_stat __user * sta
 #if BITS_PER_LONG == 32
  if (stat->size > MAX_NON_LFS)
  return -EOVERFLOW;
-#endif
+#endif
  tmp.st_size = stat->size;
  tmp.st_atime = stat->atime.tv_sec;
  tmp.st_mtime = stat->mtime.tv_sec;
@@ -443,6 +610,80 @@ SYSCALL_DEFINE4(fstatat64, int, dfd, const char __user *, filename,
 }
 #endif /* __ARCH_WANT_STAT64 || __ARCH_WANT_COMPAT_STAT64 */
 
+/*
+ * Set the statx results.
+ */
+static long statx_set_result(struct kstat *stat, struct statx __user *buffer)
+{
+ uid_t uid = from_kuid_munged(current_user_ns(), stat->uid);
+ gid_t gid = from_kgid_munged(current_user_ns(), stat->gid);
+
+#define __put_timestamp(kts, uts) ( \
+ __put_user(kts.tv_sec, uts##_s ) || \
+ __put_user(kts.tv_nsec, uts##_ns ))
+
+ if (__put_user(stat->result_mask, &buffer->st_mask ) ||
+    __put_user(stat->mode, &buffer->st_mode ) ||
+    __clear_user(&buffer->__spare0, sizeof(buffer->__spare0))  ||
+    __put_user(stat->nlink, &buffer->st_nlink ) ||
+    __put_user(uid, &buffer->st_uid ) ||
+    __put_user(gid, &buffer->st_gid ) ||
+    __put_user(stat->information, &buffer->st_information ) ||
+    __put_user(stat->blksize, &buffer->st_blksize ) ||
+    __put_user(MAJOR(stat->rdev), &buffer->st_rdev_major ) ||
+    __put_user(MINOR(stat->rdev), &buffer->st_rdev_minor ) ||
+    __put_user(MAJOR(stat->dev), &buffer->st_dev_major ) ||
+    __put_user(MINOR(stat->dev), &buffer->st_dev_minor ) ||
+    __put_timestamp(stat->atime, &buffer->st_atime ) ||
+    __put_timestamp(stat->btime, &buffer->st_btime ) ||
+    __put_timestamp(stat->ctime, &buffer->st_ctime ) ||
+    __put_timestamp(stat->mtime, &buffer->st_mtime ) ||
+    __put_user(stat->ino, &buffer->st_ino ) ||
+    __put_user(stat->size, &buffer->st_size ) ||
+    __put_user(stat->blocks, &buffer->st_blocks ) ||
+    __put_user(stat->version, &buffer->st_version ) ||
+    __put_user(stat->gen, &buffer->st_gen ) ||
+    __clear_user(&buffer->__spare1, sizeof(buffer->__spare1)))
+ return -EFAULT;
+
+ return 0;
+}
+
+/**
+ * sys_statx - System call to get enhanced stats
+ * @dfd: Base directory to pathwalk from *or* fd to stat.
+ * @filename: File to stat *or* NULL.
+ * @flags: AT_* flags to control pathwalk.
+ * @mask: Parts of statx struct actually required.
+ * @buffer: Result buffer.
+ *
+ * Note that if filename is NULL, then it does the equivalent of fstat() using
+ * dfd to indicate the file of interest.
+ */
+SYSCALL_DEFINE5(statx,
+ int, dfd, const char __user *, filename, unsigned, flags,
+ unsigned int, mask,
+ struct statx __user *, buffer)
+{
+ struct kstat stat;
+ int error;
+
+ if (!access_ok(VERIFY_WRITE, buffer, sizeof(*buffer)))
+ return -EFAULT;
+
+ memset(&stat, 0, sizeof(stat));
+ stat.query_flags = flags;
+ stat.request_mask = mask & STATX_ALL_STATS;
+
+ if (filename)
+ error = vfs_statx(dfd, filename, flags, &stat);
+ else
+ error = vfs_fstatx(dfd, &stat);
+ if (error)
+ return error;
+ return statx_set_result(&stat, buffer);
+}
+
 /* Caller is here responsible for sufficient locking (ie. inode->i_lock) */
 void __inode_add_bytes(struct inode *inode, loff_t bytes)
 {
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 70e61b58baaf..8b2f6df924e9 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2827,8 +2827,9 @@ extern const struct inode_operations page_symlink_inode_operations;
 extern void kfree_link(void *);
 extern int generic_readlink(struct dentry *, char __user *, int);
 extern void generic_fillattr(struct inode *, struct kstat *);
-int vfs_getattr_nosec(struct path *path, struct kstat *stat);
+extern int vfs_xgetattr_nosec(struct path *path, struct kstat *stat);
 extern int vfs_getattr(struct path *, struct kstat *);
+extern int vfs_xgetattr(struct path *, struct kstat *);
 void __inode_add_bytes(struct inode *inode, loff_t bytes);
 void inode_add_bytes(struct inode *inode, loff_t bytes);
 void __inode_sub_bytes(struct inode *inode, loff_t bytes);
@@ -2845,6 +2846,8 @@ extern int vfs_stat(const char __user *, struct kstat *);
 extern int vfs_lstat(const char __user *, struct kstat *);
 extern int vfs_fstat(unsigned int, struct kstat *);
 extern int vfs_fstatat(int , const char __user *, struct kstat *, int);
+extern int vfs_xstat(int, const char __user *, int, struct kstat *);
+extern int vfs_xfstat(unsigned int, struct kstat *);
 
 extern int __generic_block_fiemap(struct inode *inode,
   struct fiemap_extent_info *fieinfo,
diff --git a/include/linux/stat.h b/include/linux/stat.h
index 075cb0c7eb2a..4f1902b0cb94 100644
--- a/include/linux/stat.h
+++ b/include/linux/stat.h
@@ -19,6 +19,13 @@
 #include <linux/uidgid.h>
 
 struct kstat {
+ u32 query_flags; /* Operational flags */
+#define KSTAT_QUERY_FLAGS (AT_FORCE_ATTR_SYNC | AT_NO_ATTR_SYNC)
+ u32 request_mask; /* What fields the user asked for */
+ u32 result_mask; /* What fields the user got */
+ u32 information;
+ u32 win_attrs; /* Windows file attributes */
+ u32 gen;
  u64 ino;
  dev_t dev;
  umode_t mode;
@@ -27,11 +34,13 @@ struct kstat {
  kgid_t gid;
  dev_t rdev;
  loff_t size;
- struct timespec  atime;
+ struct timespec atime;
  struct timespec mtime;
  struct timespec ctime;
- unsigned long blksize;
- unsigned long long blocks;
+ struct timespec btime; /* File creation time */
+ uint32_t blksize; /* Preferred I/O size */
+ u64 blocks;
+ u64 version; /* Data version */
 };
 
 #endif
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index d795472c54d8..f6bfbf74e44d 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -48,6 +48,7 @@ struct stat;
 struct stat64;
 struct statfs;
 struct statfs64;
+struct statx;
 struct __sysctl_args;
 struct sysinfo;
 struct timespec;
@@ -898,4 +899,7 @@ asmlinkage long sys_copy_file_range(int fd_in, loff_t __user *off_in,
 
 asmlinkage long sys_mlock2(unsigned long start, size_t len, int flags);
 
+asmlinkage long sys_statx(int dfd, const char __user *path, unsigned flags,
+  unsigned mask, struct statx __user *buffer);
+
 #endif
diff --git a/include/uapi/linux/fcntl.h b/include/uapi/linux/fcntl.h
index beed138bd359..5c8143b04ff7 100644
--- a/include/uapi/linux/fcntl.h
+++ b/include/uapi/linux/fcntl.h
@@ -62,6 +62,8 @@
 #define AT_SYMLINK_FOLLOW 0x400   /* Follow symbolic links.  */
 #define AT_NO_AUTOMOUNT 0x800 /* Suppress terminal automount traversal */
 #define AT_EMPTY_PATH 0x1000 /* Allow empty relative pathname */
+#define AT_FORCE_ATTR_SYNC 0x2000 /* Force the attributes to be sync'd with the server */
+#define AT_NO_ATTR_SYNC 0x4000 /* Don't sync attributes with the server */
 
 
 #endif /* _UAPI_LINUX_FCNTL_H */
diff --git a/include/uapi/linux/stat.h b/include/uapi/linux/stat.h
index 7fec7e36d921..55ce6607dab6 100644
--- a/include/uapi/linux/stat.h
+++ b/include/uapi/linux/stat.h
@@ -1,6 +1,7 @@
 #ifndef _UAPI_LINUX_STAT_H
 #define _UAPI_LINUX_STAT_H
 
+#include <linux/types.h>
 
 #if defined(__KERNEL__) || !defined(__GLIBC__) || (__GLIBC__ < 2)
 
@@ -41,5 +42,113 @@
 
 #endif
 
+/*
+ * Structures for the extended file attribute retrieval system call
+ * (statx()).
+ *
+ * The caller passes a mask of what they're specifically interested in as a
+ * parameter to statx().  What statx() actually got will be indicated in
+ * st_mask upon return.
+ *
+ * For each bit in the mask argument:
+ *
+ * - if the datum is not available at all, the field and the bit will both be
+ *   cleared;
+ *
+ * - otherwise, if explicitly requested:
+ *
+ *   - the datum will be synchronised to the server if AT_FORCE_ATTR_SYNC is
+ *     set or if the datum is considered out of date, and
+ *
+ *   - the field will be filled in and the bit will be set;
+ *
+ * - otherwise, if not requested, but available in approximate form without any
+ *   effort, it will be filled in anyway, and the bit will be set upon return
+ *   (it might not be up to date, however, and no attempt will be made to
+ *   synchronise the internal state first);
+ *
+ * - otherwise the field and the bit will be cleared before returning.
+ *
+ * Items in STATX_BASIC_STATS may be marked unavailable on return, but they
+ * will have values installed for compatibility purposes so that stat() and
+ * co. can be emulated in userspace.
+ */
+struct statx {
+ /* 0x00 */
+ __u32 st_mask; /* What results were written [uncond] */
+ __u32 st_information; /* Information about the file [uncond] */
+ __u32 st_blksize; /* Preferred general I/O size [uncond] */
+ __u32 st_nlink; /* Number of hard links */
+ /* 0x10 */
+ __u32 st_gen; /* Inode generation number */
+ __u32 st_uid; /* User ID of owner */
+ __u32 st_gid; /* Group ID of owner */
+ __u16 st_mode; /* File mode */
+ __u16 __spare0[1];
+ /* 0x20 */
+ __u64 st_ino; /* Inode number */
+ __u64 st_size; /* File size */
+ __u64 st_blocks; /* Number of 512-byte blocks allocated */
+ __u64 st_version; /* Data version number */
+ /* 0x40 */
+ __s64 st_atime_s; /* Last access time */
+ __s64 st_btime_s; /* File creation time */
+ __s64 st_ctime_s; /* Last attribute change time */
+ __s64 st_mtime_s; /* Last data modification time */
+ /* 0x60 */
+ __s32 st_atime_ns; /* Last access time (ns part) */
+ __s32 st_btime_ns; /* File creation time (ns part) */
+ __s32 st_ctime_ns; /* Last attribute change time (ns part) */
+ __s32 st_mtime_ns; /* Last data modification time (ns part) */
+ /* 0x70 */
+ __u32 st_rdev_major; /* Device ID of special file */
+ __u32 st_rdev_minor;
+ __u32 st_dev_major; /* ID of device containing file [uncond] */
+ __u32 st_dev_minor;
+ /* 0x80 */
+ __u64 __spare1[16]; /* Spare space for future expansion */
+ /* 0x100 */
+};
+
+/*
+ * Flags to be st_mask
+ *
+ * Query request/result mask for statx() and struct statx::st_mask.
+ *
+ * These bits should be set in the mask argument of statx() to request
+ * particular items when calling statx().
+ */
+#define STATX_MODE 0x00000001U /* Want/got st_mode */
+#define STATX_NLINK 0x00000002U /* Want/got st_nlink */
+#define STATX_UID 0x00000004U /* Want/got st_uid */
+#define STATX_GID 0x00000008U /* Want/got st_gid */
+#define STATX_RDEV 0x00000010U /* Want/got st_rdev */
+#define STATX_ATIME 0x00000020U /* Want/got st_atime */
+#define STATX_MTIME 0x00000040U /* Want/got st_mtime */
+#define STATX_CTIME 0x00000080U /* Want/got st_ctime */
+#define STATX_INO 0x00000100U /* Want/got st_ino */
+#define STATX_SIZE 0x00000200U /* Want/got st_size */
+#define STATX_BLOCKS 0x00000400U /* Want/got st_blocks */
+#define STATX_BASIC_STATS 0x000007ffU /* The stuff in the normal stat struct */
+#define STATX_BTIME 0x00000800U /* Want/got st_btime */
+#define STATX_VERSION 0x00001000U /* Want/got st_version */
+#define STATX_GEN 0x00002000U /* Want/got st_gen */
+#define STATX_ALL_STATS 0x00003fffU /* All supported stats */
+
+/*
+ * Flags to be found in st_information
+ *
+ * These give information about the features or the state of a file that might
+ * be of use to ordinary userspace programs such as GUIs or ls rather than
+ * specialised tools.
+ */
+#define STATX_INFO_ENCRYPTED 0x00000001U /* File is encrypted */
+#define STATX_INFO_TEMPORARY 0x00000002U /* File is temporary */
+#define STATX_INFO_FABRICATED 0x00000004U /* File was made up by filesystem */
+#define STATX_INFO_KERNEL_API 0x00000008U /* File is kernel API (eg: procfs/sysfs) */
+#define STATX_INFO_REMOTE 0x00000010U /* File is remote */
+#define STATX_INFO_AUTOMOUNT 0x00000020U /* Dir is automount trigger */
+#define STATX_INFO_AUTODIR 0x00000040U /* Dir provides unlisted automounts */
+#define STATX_INFO_NONSYSTEM_OWNERSHIP 0x00000080U /* File has non-system ownership details */
 
 #endif /* _UAPI_LINUX_STAT_H */
diff --git a/samples/Makefile b/samples/Makefile
index 48001d7e23f0..d2ebb4e48d19 100644
--- a/samples/Makefile
+++ b/samples/Makefile
@@ -2,4 +2,4 @@
 
 obj-$(CONFIG_SAMPLES) += kobject/ kprobes/ trace_events/ livepatch/ \
    hw_breakpoint/ kfifo/ kdb/ hidraw/ rpmsg/ seccomp/ \
-   configfs/
+   configfs/ statx/
diff --git a/samples/statx/Makefile b/samples/statx/Makefile
new file mode 100644
index 000000000000..6765dabc4c8d
--- /dev/null
+++ b/samples/statx/Makefile
@@ -0,0 +1,10 @@
+# kbuild trick to avoid linker error. Can be omitted if a module is built.
+obj- := dummy.o
+
+# List of programs to build
+hostprogs-y := test-statx
+
+# Tell kbuild to always build the programs
+always := $(hostprogs-y)
+
+HOSTCFLAGS_test-statx.o += -I$(objtree)/usr/include
diff --git a/samples/statx/test-statx.c b/samples/statx/test-statx.c
new file mode 100644
index 000000000000..38ef23c12e7d
--- /dev/null
+++ b/samples/statx/test-statx.c
@@ -0,0 +1,243 @@
+/* Test the statx() system call
+ *
+ * Copyright (C) 2015 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells ([hidden email])
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public Licence
+ * as published by the Free Software Foundation; either version
+ * 2 of the Licence, or (at your option) any later version.
+ */
+
+#define _GNU_SOURCE
+#define _ATFILE_SOURCE
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <ctype.h>
+#include <errno.h>
+#include <time.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <linux/stat.h>
+#include <linux/fcntl.h>
+#include <sys/stat.h>
+
+#define AT_FORCE_ATTR_SYNC 0x2000
+#define AT_NO_ATTR_SYNC 0x4000
+
+static __attribute__((unused))
+ssize_t statx(int dfd, const char *filename, unsigned flags,
+      unsigned int mask, struct statx *buffer)
+{
+ return syscall(__NR_statx, dfd, filename, flags, mask, buffer);
+}
+
+static void print_time(const char *field, __s64 tv_sec, __s32 tv_nsec)
+{
+ struct tm tm;
+ time_t tim;
+ char buffer[100];
+ int len;
+
+ tim = tv_sec;
+ if (!localtime_r(&tim, &tm)) {
+ perror("localtime_r");
+ exit(1);
+ }
+ len = strftime(buffer, 100, "%F %T", &tm);
+ if (len == 0) {
+ perror("strftime");
+ exit(1);
+ }
+ printf("%s", field);
+ fwrite(buffer, 1, len, stdout);
+ printf(".%09u", tv_nsec);
+ len = strftime(buffer, 100, "%z", &tm);
+ if (len == 0) {
+ perror("strftime2");
+ exit(1);
+ }
+ fwrite(buffer, 1, len, stdout);
+ printf("\n");
+}
+
+static void dump_statx(struct statx *stx)
+{
+ char buffer[256], ft = '?';
+
+ printf("results=%x\n", stx->st_mask);
+
+ printf(" ");
+ if (stx->st_mask & STATX_SIZE)
+ printf(" Size: %-15llu", (unsigned long long)stx->st_size);
+ if (stx->st_mask & STATX_BLOCKS)
+ printf(" Blocks: %-10llu", (unsigned long long)stx->st_blocks);
+ printf(" IO Block: %-6llu ", (unsigned long long)stx->st_blksize);
+ if (stx->st_mask & STATX_MODE) {
+ switch (stx->st_mode & S_IFMT) {
+ case S_IFIFO: printf(" FIFO\n"); ft = 'p'; break;
+ case S_IFCHR: printf(" character special file\n"); ft = 'c'; break;
+ case S_IFDIR: printf(" directory\n"); ft = 'd'; break;
+ case S_IFBLK: printf(" block special file\n"); ft = 'b'; break;
+ case S_IFREG: printf(" regular file\n"); ft = '-'; break;
+ case S_IFLNK: printf(" symbolic link\n"); ft = 'l'; break;
+ case S_IFSOCK: printf(" socket\n"); ft = 's'; break;
+ default:
+ printf("unknown type (%o)\n", stx->st_mode & S_IFMT);
+ break;
+ }
+ }
+
+ sprintf(buffer, "%02x:%02x", stx->st_dev_major, stx->st_dev_minor);
+ printf("Device: %-15s", buffer);
+ if (stx->st_mask & STATX_INO)
+ printf(" Inode: %-11llu", (unsigned long long) stx->st_ino);
+ if (stx->st_mask & STATX_SIZE)
+ printf(" Links: %-5u", stx->st_nlink);
+ if (stx->st_mask & STATX_RDEV)
+ printf(" Device type: %u,%u", stx->st_rdev_major, stx->st_rdev_minor);
+ printf("\n");
+
+ if (stx->st_mask & STATX_MODE)
+ printf("Access: (%04o/%c%c%c%c%c%c%c%c%c%c)  ",
+       stx->st_mode & 07777,
+       ft,
+       stx->st_mode & S_IRUSR ? 'r' : '-',
+       stx->st_mode & S_IWUSR ? 'w' : '-',
+       stx->st_mode & S_IXUSR ? 'x' : '-',
+       stx->st_mode & S_IRGRP ? 'r' : '-',
+       stx->st_mode & S_IWGRP ? 'w' : '-',
+       stx->st_mode & S_IXGRP ? 'x' : '-',
+       stx->st_mode & S_IROTH ? 'r' : '-',
+       stx->st_mode & S_IWOTH ? 'w' : '-',
+       stx->st_mode & S_IXOTH ? 'x' : '-');
+ if (stx->st_mask & STATX_UID)
+ printf("Uid: %5d   ", stx->st_uid);
+ if (stx->st_mask & STATX_GID)
+ printf("Gid: %5d\n", stx->st_gid);
+
+ if (stx->st_mask & STATX_ATIME)
+ print_time("Access: ", stx->st_atime_s, stx->st_atime_ns);
+ if (stx->st_mask & STATX_MTIME)
+ print_time("Modify: ", stx->st_mtime_s, stx->st_mtime_ns);
+ if (stx->st_mask & STATX_CTIME)
+ print_time("Change: ", stx->st_ctime_s, stx->st_ctime_ns);
+ if (stx->st_mask & STATX_BTIME)
+ print_time(" Birth: ", stx->st_btime_s, stx->st_btime_ns);
+
+ if (stx->st_mask & STATX_VERSION)
+ printf("Data version: %llxh\n",
+       (unsigned long long)stx->st_version);
+
+ if (stx->st_mask & STATX_GEN)
+ printf("Inode gen   : %xh\n", stx->st_gen);
+
+ if (stx->st_information) {
+ unsigned char bits;
+ int loop, byte;
+
+ static char info_representation[32 + 1] =
+ /* STATX_INFO_ flags: */
+ "????????" /* 31-24 0x00000000-ff000000 */
+ "????????" /* 23-16 0x00000000-00ff0000 */
+ "????????" /* 15- 8 0x00000000-0000ff00 */
+ "ndmrkfte" /*  7- 0 0x00000000-000000ff */
+ ;
+
+ printf("Information: %08x (", stx->st_information);
+ for (byte = 32 - 8; byte >= 0; byte -= 8) {
+ bits = stx->st_information >> byte;
+ for (loop = 7; loop >= 0; loop--) {
+ int bit = byte + loop;
+
+ if (bits & 0x80)
+ putchar(info_representation[31 - bit]);
+ else
+ putchar('-');
+ bits <<= 1;
+ }
+ if (byte)
+ putchar(' ');
+ }
+ printf(")\n");
+ }
+
+ printf("IO-blocksize: blksize=%u\n", stx->st_blksize);
+}
+
+static void dump_hex(unsigned long long *data, int from, int to)
+{
+ unsigned offset, print_offset = 1, col = 0;
+
+ from /= 8;
+ to = (to + 7) / 8;
+
+ for (offset = from; offset < to; offset++) {
+ if (print_offset) {
+ printf("%04x: ", offset * 8);
+ print_offset = 0;
+ }
+ printf("%016llx", data[offset]);
+ col++;
+ if ((col & 3) == 0) {
+ printf("\n");
+ print_offset = 1;
+ } else {
+ printf(" ");
+ }
+ }
+
+ if (!print_offset)
+ printf("\n");
+}
+
+int main(int argc, char **argv)
+{
+ struct statx stx;
+ int ret, raw = 0, atflag = AT_SYMLINK_NOFOLLOW;
+
+ unsigned int mask = STATX_ALL_STATS;
+
+ for (argv++; *argv; argv++) {
+ if (strcmp(*argv, "-F") == 0) {
+ atflag |= AT_FORCE_ATTR_SYNC;
+ continue;
+ }
+ if (strcmp(*argv, "-N") == 0) {
+ atflag |= AT_NO_ATTR_SYNC;
+ continue;
+ }
+ if (strcmp(*argv, "-L") == 0) {
+ atflag &= ~AT_SYMLINK_NOFOLLOW;
+ continue;
+ }
+ if (strcmp(*argv, "-O") == 0) {
+ mask &= ~STATX_BASIC_STATS;
+ continue;
+ }
+ if (strcmp(*argv, "-A") == 0) {
+ atflag |= AT_NO_AUTOMOUNT;
+ continue;
+ }
+ if (strcmp(*argv, "-R") == 0) {
+ raw = 1;
+ continue;
+ }
+
+ memset(&stx, 0xbf, sizeof(stx));
+ ret = statx(AT_FDCWD, *argv, atflag, mask, &stx);
+ printf("statx(%s) = %d\n", *argv, ret);
+ if (ret < 0) {
+ perror(*argv);
+ exit(1);
+ }
+
+ if (raw)
+ dump_hex((unsigned long long *)&stx, 0, sizeof(stx));
+
+ dump_statx(&stx);
+ }
+ return 0;
+}

Reply | Threaded
Open this post in threaded view
|

[PATCH 3/6] statx: Ext4: Return enhanced file attributes

David Howells
In reply to this post by David Howells
Return enhanced file attributes from the Ext4 filesystem.  This includes
the following:

 (1) The inode creation time (i_crtime) as i_btime, setting STATX_BTIME.

 (2) The inode i_version as st_version if a file with I_VERSION set or a
     directory, setting STATX_VERSION.

 (3) FS_xxx_FL flags are returned as for ioctl(FS_IOC_GETFLAGS), setting
     STATX_IOC_FLAGS.

This requires that all ext4 inodes have a getattr call, not just some of
them, so to this end, split the ext4_getattr() function and only call part
of it where appropriate.

Example output:

        [root@andromeda ~]# ./samples/statx/test-statx /usr
        statx(/usr) = 0
        results=37ef
          Size: 4096            Blocks: 16         IO Block: 4096    directory
        Device: 08:02           Inode: 1572865     Links: 14
        Access: (0755/drwxr-xr-x)  Uid:     0   Gid:     0
        Access: 2015-11-03 16:12:30.000000000+0000
        Modify: 2013-10-18 15:29:18.000000000+0100
        Change: 2013-10-18 15:29:18.000000000+0100
        Data version: 2fh
        Inode flags: 00000000 (-------- -------- -------- --------)
        IO-blocksize: blksize=4096

Signed-off-by: David Howells <[hidden email]>
---

 fs/ext4/ext4.h    |    2 ++
 fs/ext4/file.c    |    2 +-
 fs/ext4/inode.c   |   30 +++++++++++++++++++++++++++---
 fs/ext4/namei.c   |    2 ++
 fs/ext4/symlink.c |    2 ++
 5 files changed, 34 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 349afebe21ee..2f25eaa63f39 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2549,6 +2549,8 @@ extern int  ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
  struct kstat *stat);
 extern void ext4_evict_inode(struct inode *);
 extern void ext4_clear_inode(struct inode *);
+extern int  ext4_file_getattr(struct vfsmount *mnt, struct dentry *dentry,
+      struct kstat *stat);
 extern int  ext4_sync_inode(handle_t *, struct inode *);
 extern void ext4_dirty_inode(struct inode *, int);
 extern int ext4_change_inode_journal_flag(struct inode *, int);
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index fa2208bae2e1..45c7b8644d0e 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -715,7 +715,7 @@ const struct file_operations ext4_file_operations = {
 
 const struct inode_operations ext4_file_inode_operations = {
  .setattr = ext4_setattr,
- .getattr = ext4_getattr,
+ .getattr = ext4_file_getattr,
  .setxattr = generic_setxattr,
  .getxattr = generic_getxattr,
  .listxattr = ext4_listxattr,
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 981a1fc30eaa..309b6cff8afc 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5100,11 +5100,35 @@ err_out:
 int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
  struct kstat *stat)
 {
- struct inode *inode;
- unsigned long long delalloc_blocks;
+ struct inode *inode = d_inode(dentry);
+ struct ext4_inode *raw_inode;
+ struct ext4_inode_info *ei = EXT4_I(inode);
+
+ stat->result_mask |= STATX_GEN;
+ stat->gen = inode->i_generation;
+
+ if (EXT4_FITS_IN_INODE(raw_inode, ei, i_crtime)) {
+ stat->result_mask |= STATX_BTIME;
+ stat->btime.tv_sec = ei->i_crtime.tv_sec;
+ stat->btime.tv_nsec = ei->i_crtime.tv_nsec;
+ }
+
+ if (S_ISDIR(inode->i_mode) || IS_I_VERSION(inode)) {
+ stat->result_mask |= STATX_VERSION;
+ stat->version = inode->i_version;
+ }
 
- inode = d_inode(dentry);
  generic_fillattr(inode, stat);
+ return 0;
+}
+
+int ext4_file_getattr(struct vfsmount *mnt, struct dentry *dentry,
+      struct kstat *stat)
+{
+ struct inode *inode = dentry->d_inode;
+ u64 delalloc_blocks;
+
+ ext4_getattr(mnt, dentry, stat);
 
  /*
  * If there is inline data in the inode, the inode will normally not
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 48e4b8907826..9b5fdc971e22 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -3882,6 +3882,7 @@ const struct inode_operations ext4_dir_inode_operations = {
  .tmpfile = ext4_tmpfile,
  .rename2 = ext4_rename2,
  .setattr = ext4_setattr,
+ .getattr = ext4_getattr,
  .setxattr = generic_setxattr,
  .getxattr = generic_getxattr,
  .listxattr = ext4_listxattr,
@@ -3893,6 +3894,7 @@ const struct inode_operations ext4_dir_inode_operations = {
 
 const struct inode_operations ext4_special_inode_operations = {
  .setattr = ext4_setattr,
+ .getattr = ext4_getattr,
  .setxattr = generic_setxattr,
  .getxattr = generic_getxattr,
  .listxattr = ext4_listxattr,
diff --git a/fs/ext4/symlink.c b/fs/ext4/symlink.c
index 75ed5c2f0c16..54015f3d7516 100644
--- a/fs/ext4/symlink.c
+++ b/fs/ext4/symlink.c
@@ -105,6 +105,7 @@ const struct inode_operations ext4_symlink_inode_operations = {
  .readlink = generic_readlink,
  .get_link = page_get_link,
  .setattr = ext4_setattr,
+ .getattr = ext4_getattr,
  .setxattr = generic_setxattr,
  .getxattr = generic_getxattr,
  .listxattr = ext4_listxattr,
@@ -115,6 +116,7 @@ const struct inode_operations ext4_fast_symlink_inode_operations = {
  .readlink = generic_readlink,
  .get_link = simple_get_link,
  .setattr = ext4_setattr,
+ .getattr = ext4_getattr,
  .setxattr = generic_setxattr,
  .getxattr = generic_getxattr,
  .listxattr = ext4_listxattr,

Reply | Threaded
Open this post in threaded view
|

[PATCH 4/6] statx: NFS: Return enhanced file attributes

David Howells
In reply to this post by David Howells
Return enhanced file atrributes from the NFS filesystem.  This includes the
following:

 (1) The change attribute as st_version if NFSv4.

 (2) STATX_INFO_AUTOMOUNT and STATX_INFO_FABRICATED are set on referral or
     submount directories that are automounted upon.  NFS shows one
     directory with a different FSID, but the local filesystem has two: the
     mountpoint directory and the root of the filesystem mounted upon it.

 (3) STATX_INFO_REMOTE is set on files acquired over NFS.

 (4) STATX_IOC_FLAGS is set and if the atime is unavailable on a file,
     st_ioc_flags will have FL_NOATIME_FL set in it.

Furthermore, what nfs_getattr() does can be controlled as follows:

 (1) If AT_NO_ATTR_SYNC is indicated then this will suppress the flushing
     of outstanding writes and the rereading of the inode's attributes with
     the server as detailed below.

 (2) Otherwise:

     (a) If AT_FORCE_ATTR_SYNC is indicated, or mtime, ctime or
      data_version (NFSv4 only) are requested then the outstanding
      writes will be written to the server first.

     (b) The inode's attributes will be reread from the server:

      (i) if AT_FORCE_ATTR_SYNC is indicated;

        (ii) if atime is requested (and atime updating is not suppressed by
          a mount flag); or

       (iii) if the cached attributes have expired;

If the inode isn't synchronised, then the cached attributes will be used -
even if expired - without reference to the server.

Example output:

        [root@andromeda ~]# ./samples/statx/test-statx /warthog/
        statx(/warthog/) = 0
        results=37ef
          Size: 4096            Blocks: 8          IO Block: 1048576  directory
        Device: 00:26           Inode: 2           Links: 122
        Access: (3777/drwxrwxrwx)  Uid:     0   Gid:  4041
        Access: 2015-10-30 16:15:41.730925545+0000
        Modify: 2015-10-07 10:33:19.896108112+0100
        Change: 2015-10-07 10:33:19.896108112+0100
        Data version: 5614e6df35698650h
        Inode flags: 00000000 (-------- -------- -------- --------)
        Information: 00000010 (-------- -------- -------- ---r----)
        IO-blocksize: blksize=1048576

Note that the NFS4 protocol potentially provides a creation time that could
be passed through this interface and system, hidden and archive values that
could be passed as IOC flags.  There is also a backup time that could be
added.

Signed-off-by: David Howells <[hidden email]>
---

 fs/nfs/inode.c |   41 ++++++++++++++++++++++++++++++++++-------
 1 file changed, 34 insertions(+), 7 deletions(-)

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 738c84a42eb0..8637236bca0c 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -655,12 +655,23 @@ static bool nfs_need_revalidate_inode(struct inode *inode)
 int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 {
  struct inode *inode = d_inode(dentry);
- int need_atime = NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATIME;
+ bool force_sync = stat->query_flags & AT_FORCE_ATTR_SYNC;
+ bool suppress_sync = stat->query_flags & AT_NO_ATTR_SYNC;
+ bool need_atime = NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATIME;
  int err = 0;
 
  trace_nfs_getattr_enter(inode);
- /* Flush out writes to the server in order to update c/mtime.  */
- if (S_ISREG(inode->i_mode)) {
+
+ if (NFS_SERVER(inode)->nfs_client->rpc_ops->version < 4)
+ stat->request_mask &= ~STATX_VERSION;
+
+ /* Flush out writes to the server in order to update c/mtime or data
+ * version if the user wants them.
+ */
+ if (S_ISREG(inode->i_mode) && !suppress_sync &&
+    (force_sync || (stat->request_mask &
+    (STATX_MTIME | STATX_CTIME | STATX_VERSION)))
+    ) {
  inode_lock(inode);
  err = nfs_sync_inode(inode);
  inode_unlock(inode);
@@ -677,11 +688,13 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
  *  - NFS never sets MS_NOATIME or MS_NODIRATIME so there is
  *    no point in checking those.
  */
- if ((mnt->mnt_flags & MNT_NOATIME) ||
-    ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)))
- need_atime = 0;
+ if (!(stat->request_mask & STATX_ATIME) ||
+    (mnt->mnt_flags & MNT_NOATIME) ||
+    ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)))
+ need_atime = false;
 
- if (need_atime || nfs_need_revalidate_inode(inode)) {
+ if (!suppress_sync &&
+    (force_sync || need_atime || nfs_need_revalidate_inode(inode))) {
  struct nfs_server *server = NFS_SERVER(inode);
 
  if (server->caps & NFS_CAP_READDIRPLUS)
@@ -694,6 +707,20 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
  if (S_ISDIR(inode->i_mode))
  stat->blksize = NFS_SERVER(inode)->dtsize;
  }
+
+ generic_fillattr(inode, stat);
+ stat->ino = nfs_compat_user_ino64(NFS_FILEID(inode));
+
+ if (stat->request_mask & STATX_VERSION) {
+ stat->version = inode->i_version;
+ stat->result_mask |= STATX_VERSION;
+ }
+
+ if (IS_AUTOMOUNT(inode))
+ stat->information |= STATX_INFO_FABRICATED;
+
+ stat->information |= STATX_INFO_REMOTE;
+
 out:
  trace_nfs_getattr_exit(inode, err);
  return err;

Reply | Threaded
Open this post in threaded view
|

[PATCH 6/6] statx: CIFS: Return enhanced attributes

David Howells
In reply to this post by David Howells
Return enhanced attributes from the CIFS filesystem.  This includes the
following:

 (1) Return the file creation time as btime.  We assume that the creation
     time won't change over the life of the inode.

 (2) Set STATX_INFO_AUTOMOUNT on referral/submount directories.

 (3) Unset STATX_INO if we made up the inode number and didn't get it from
     the server.

 (4) Unset STATX_[UG]ID if we are either returning values passed to mount
     and/or the server doesn't return them.

 (5) Set STATX_WIN_ATTRS and return windows attributes in st_win_attrs.

 (6) Set certain STATX_INFO_xxx to reflect a number of Windows file
     attributes:

        ATTR_TEMPORARY -> STATX_INFO_TEMPORARY;
        ATTR_ENCRYPTED -> STATX_INFO_ENCRYPTED;

 (7) Set STATX_INFO_REMOTE on all files fetched by CIFS.

 (8) Set STATX_INFO_NONSYSTEM_OWNERSHIP on all files as they all have
     Windows ownership details too.

Furthermore, what cifs_getattr() does can be controlled as follows:

 (1) If AT_NO_ATTR_SYNC is indicated then this will suppress the flushing
     of outstanding writes and the rereading of the inode's attributes with
     the server as detailed below.

 (2) Otherwise:

     (a) If AT_FORCE_ATTR_SYNC is indicated, or mtime, ctime or size are
      requested then the outstanding writes will be written to the
      server first.

     (b) The inode's attributes will be reread from the server:

      (i) if AT_FORCE_ATTR_SYNC is indicated;

        (ii) if the cached attributes have expired;

       (iii) extra attributes are requested that aren't normally stored.

If the inode isn't synchronised, then the cached attributes will be used -
even if expired - without reference to the server.  Some attributes may be
unavailable that would otherwise be provided.

Note that cifs_revalidate_dentry() will issue an extra operation to get the
FILE_ALL_INFO in addition to the FILE_UNIX_BASIC_INFO if it needs to
collect creation time and attributes on behalf of cifs_getattr().

[NOTE: THIS PATCH IS UNTESTED!]

Signed-off-by: David Howells <[hidden email]>
---

 fs/cifs/cifsfs.h   |    4 +-
 fs/cifs/cifsglob.h |    8 ++++
 fs/cifs/dir.c      |    2 -
 fs/cifs/inode.c    |  117 ++++++++++++++++++++++++++++++++++++++++------------
 4 files changed, 101 insertions(+), 30 deletions(-)

diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
index 83aac8ba50b0..304f4567d207 100644
--- a/fs/cifs/cifsfs.h
+++ b/fs/cifs/cifsfs.h
@@ -67,9 +67,9 @@ extern int cifs_rmdir(struct inode *, struct dentry *);
 extern int cifs_rename2(struct inode *, struct dentry *, struct inode *,
  struct dentry *, unsigned int);
 extern int cifs_revalidate_file_attr(struct file *filp);
-extern int cifs_revalidate_dentry_attr(struct dentry *);
+extern int cifs_revalidate_dentry_attr(struct dentry *, bool, bool);
 extern int cifs_revalidate_file(struct file *filp);
-extern int cifs_revalidate_dentry(struct dentry *);
+extern int cifs_revalidate_dentry(struct dentry *, bool, bool);
 extern int cifs_invalidate_mapping(struct inode *inode);
 extern int cifs_revalidate_mapping(struct inode *inode);
 extern int cifs_zap_mapping(struct inode *inode);
diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h
index f2cc0b3d1af7..3225b16f7522 100644
--- a/fs/cifs/cifsglob.h
+++ b/fs/cifs/cifsglob.h
@@ -1166,7 +1166,11 @@ struct cifsInodeInfo {
  unsigned long flags;
  spinlock_t writers_lock;
  unsigned int writers; /* Number of writers on this inode */
+ bool btime_valid:1; /* stored creation time is valid */
+ bool uid_faked:1; /* true if i_uid is faked */
+ bool gid_faked:1; /* true if i_gid is faked */
  unsigned long time; /* jiffies of last update of inode */
+ struct timespec btime; /* creation time */
  u64  server_eof; /* current file size on server -- protected by i_lock */
  u64  uniqueid; /* server inode number */
  u64  createtime; /* creation time on server */
@@ -1379,6 +1383,9 @@ struct dfs_info3_param {
 #define CIFS_FATTR_NEED_REVAL 0x4
 #define CIFS_FATTR_INO_COLLISION 0x8
 #define CIFS_FATTR_UNKNOWN_NLINK 0x10
+#define CIFS_FATTR_WINATTRS_VALID 0x20 /* T if cf_btime and cf_cifsattrs valid */
+#define CIFS_FATTR_UID_FAKED 0x40 /* T if cf_uid is faked */
+#define CIFS_FATTR_GID_FAKED 0x80 /* T if cf_gid is faked */
 
 struct cifs_fattr {
  u32 cf_flags;
@@ -1396,6 +1403,7 @@ struct cifs_fattr {
  struct timespec cf_atime;
  struct timespec cf_mtime;
  struct timespec cf_ctime;
+ struct timespec cf_btime;
 };
 
 static inline void free_dfs_info_param(struct dfs_info3_param *param)
diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index c3eb998a99bd..4984f04b0677 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -792,7 +792,7 @@ cifs_d_revalidate(struct dentry *direntry, unsigned int flags)
  return -ECHILD;
 
  if (d_really_is_positive(direntry)) {
- if (cifs_revalidate_dentry(direntry))
+ if (cifs_revalidate_dentry(direntry, false, false))
  return 0;
  else {
  /*
diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c
index 5f9ad5c42180..dc1814b70c81 100644
--- a/fs/cifs/inode.c
+++ b/fs/cifs/inode.c
@@ -166,13 +166,21 @@ cifs_fattr_to_inode(struct inode *inode, struct cifs_fattr *fattr)
  cifs_nlink_fattr_to_inode(inode, fattr);
  inode->i_uid = fattr->cf_uid;
  inode->i_gid = fattr->cf_gid;
+ if (fattr->cf_flags & CIFS_FATTR_UID_FAKED)
+ cifs_i->uid_faked = true;
+ if (fattr->cf_flags & CIFS_FATTR_GID_FAKED)
+ cifs_i->gid_faked = true;
 
  /* if dynperm is set, don't clobber existing mode */
  if (inode->i_state & I_NEW ||
     !(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM))
  inode->i_mode = fattr->cf_mode;
 
- cifs_i->cifsAttrs = fattr->cf_cifsattrs;
+ if (fattr->cf_flags & CIFS_FATTR_WINATTRS_VALID) {
+ cifs_i->cifsAttrs = fattr->cf_cifsattrs;
+ cifs_i->btime = fattr->cf_btime;
+ cifs_i->btime_valid = true;
+ }
 
  if (fattr->cf_flags & CIFS_FATTR_NEED_REVAL)
  cifs_i->time = 0;
@@ -284,18 +292,22 @@ cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, FILE_UNIX_BASIC_INFO *info,
  u64 id = le64_to_cpu(info->Uid);
  if (id < ((uid_t)-1)) {
  kuid_t uid = make_kuid(&init_user_ns, id);
- if (uid_valid(uid))
+ if (uid_valid(uid)) {
  fattr->cf_uid = uid;
+ fattr->cf_flags |= CIFS_FATTR_UID_FAKED;
+ }
  }
  }
-
+
  fattr->cf_gid = cifs_sb->mnt_gid;
  if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID)) {
  u64 id = le64_to_cpu(info->Gid);
  if (id < ((gid_t)-1)) {
  kgid_t gid = make_kgid(&init_user_ns, id);
- if (gid_valid(gid))
+ if (gid_valid(gid)) {
  fattr->cf_gid = gid;
+ fattr->cf_flags |= CIFS_FATTR_GID_FAKED;
+ }
  }
  }
 
@@ -324,7 +336,8 @@ cifs_create_dfs_fattr(struct cifs_fattr *fattr, struct super_block *sb)
  fattr->cf_ctime = CURRENT_TIME;
  fattr->cf_mtime = CURRENT_TIME;
  fattr->cf_nlink = 2;
- fattr->cf_flags |= CIFS_FATTR_DFS_REFERRAL;
+ fattr->cf_flags |= CIFS_FATTR_DFS_REFERRAL |
+ CIFS_FATTR_UID_FAKED | CIFS_FATTR_GID_FAKED;
 }
 
 static int
@@ -590,6 +603,7 @@ cifs_all_info_to_fattr(struct cifs_fattr *fattr, FILE_ALL_INFO *info,
  struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb);
 
  memset(fattr, 0, sizeof(*fattr));
+ fattr->cf_flags = CIFS_FATTR_WINATTRS_VALID;
  fattr->cf_cifsattrs = le32_to_cpu(info->Attributes);
  if (info->DeletePending)
  fattr->cf_flags |= CIFS_FATTR_DELETE_PENDING;
@@ -601,6 +615,7 @@ cifs_all_info_to_fattr(struct cifs_fattr *fattr, FILE_ALL_INFO *info,
 
  fattr->cf_ctime = cifs_NTtimeToUnix(info->ChangeTime);
  fattr->cf_mtime = cifs_NTtimeToUnix(info->LastWriteTime);
+ fattr->cf_btime = cifs_NTtimeToUnix(info->CreationTime);
 
  if (adjust_tz) {
  fattr->cf_ctime.tv_sec += tcon->ses->server->timeAdj;
@@ -1907,7 +1922,8 @@ int cifs_revalidate_file_attr(struct file *filp)
  return rc;
 }
 
-int cifs_revalidate_dentry_attr(struct dentry *dentry)
+int cifs_revalidate_dentry_attr(struct dentry *dentry,
+ bool want_extra_bits, bool force)
 {
  unsigned int xid;
  int rc = 0;
@@ -1918,7 +1934,7 @@ int cifs_revalidate_dentry_attr(struct dentry *dentry)
  if (inode == NULL)
  return -ENOENT;
 
- if (!cifs_inode_needs_reval(inode))
+ if (!force && !cifs_inode_needs_reval(inode))
  return rc;
 
  xid = get_xid();
@@ -1935,9 +1951,12 @@ int cifs_revalidate_dentry_attr(struct dentry *dentry)
  full_path, inode, inode->i_count.counter,
  dentry, dentry->d_time, jiffies);
 
- if (cifs_sb_master_tcon(CIFS_SB(sb))->unix_ext)
+ if (cifs_sb_master_tcon(CIFS_SB(sb))->unix_ext) {
  rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid);
- else
+ if (rc != 0)
+ goto out;
+ }
+ if (!cifs_sb_master_tcon(CIFS_SB(sb))->unix_ext || want_extra_bits)
  rc = cifs_get_inode_info(&inode, full_path, NULL, sb,
  xid, NULL);
 
@@ -1960,12 +1979,13 @@ int cifs_revalidate_file(struct file *filp)
 }
 
 /* revalidate a dentry's inode attributes */
-int cifs_revalidate_dentry(struct dentry *dentry)
+int cifs_revalidate_dentry(struct dentry *dentry,
+   bool want_extra_bits, bool force)
 {
  int rc;
  struct inode *inode = d_inode(dentry);
 
- rc = cifs_revalidate_dentry_attr(dentry);
+ rc = cifs_revalidate_dentry_attr(dentry, want_extra_bits, force);
  if (rc)
  return rc;
 
@@ -1978,28 +1998,62 @@ int cifs_getattr(struct vfsmount *mnt, struct dentry *dentry,
  struct cifs_sb_info *cifs_sb = CIFS_SB(dentry->d_sb);
  struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb);
  struct inode *inode = d_inode(dentry);
+ struct cifsInodeInfo *cifs_i = CIFS_I(inode);
+ bool force = stat->query_flags & AT_FORCE_ATTR_SYNC;
+ bool want_extra_bits = false;
+ u32 info;
+ u32 attrs;
  int rc;
 
- /*
- * We need to be sure that all dirty pages are written and the server
- * has actual ctime, mtime and file length.
- */
- if (!CIFS_CACHE_READ(CIFS_I(inode)) && inode->i_mapping &&
-    inode->i_mapping->nrpages != 0) {
- rc = filemap_fdatawait(inode->i_mapping);
- if (rc) {
- mapping_set_error(inode->i_mapping, rc);
- return rc;
+ if (cifs_i->uid_faked)
+ stat->request_mask &= ~STATX_UID;
+ if (cifs_i->gid_faked)
+ stat->request_mask &= ~STATX_GID;
+
+ if ((stat->request_mask & STATX_BTIME && !cifs_i->btime_valid) ||
+    stat->request_mask & STATX_WIN_ATTRS)
+ want_extra_bits = force = true;
+
+ if (!(stat->query_flags & AT_NO_ATTR_SYNC)) {
+ /* Unless we're explicitly told not to sync, we need to be sure
+ * that all dirty pages are written and the server has actual
+ * ctime, mtime and file length.
+ */
+ bool flush = force;
+
+ if (stat->request_mask &
+    (STATX_CTIME | STATX_MTIME | STATX_SIZE))
+ flush = true;
+
+ if (flush &&
+    !CIFS_CACHE_READ(CIFS_I(inode)) && inode->i_mapping &&
+    inode->i_mapping->nrpages != 0) {
+ rc = filemap_fdatawait(inode->i_mapping);
+ if (rc) {
+ mapping_set_error(inode->i_mapping, rc);
+ return rc;
+ }
  }
- }
 
- rc = cifs_revalidate_dentry_attr(dentry);
- if (rc)
- return rc;
+ rc = cifs_revalidate_dentry(dentry, want_extra_bits, force);
+ if (rc)
+ return rc;
+ }
 
  generic_fillattr(inode, stat);
  stat->blksize = CIFS_MAX_MSGSIZE;
- stat->ino = CIFS_I(inode)->uniqueid;
+
+ info = STATX_INFO_REMOTE | STATX_INFO_NONSYSTEM_OWNERSHIP;
+
+ if (cifs_i->btime_valid) {
+ stat->btime = cifs_i->btime;
+ stat->result_mask |= STATX_BTIME;
+ }
+
+ /* We don't promise an inode number if we made one up */
+ stat->ino = cifs_i->uniqueid;
+ if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM))
+ stat->result_mask &= ~STATX_INO;
 
  /*
  * If on a multiuser mount without unix extensions or cifsacl being
@@ -2013,8 +2067,17 @@ int cifs_getattr(struct vfsmount *mnt, struct dentry *dentry,
  stat->uid = current_fsuid();
  if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID))
  stat->gid = current_fsgid();
+ stat->result_mask &= ~(STATX_UID | STATX_GID);
  }
- return rc;
+
+ attrs = cifs_i->cifsAttrs;
+ stat->win_attrs = attrs;
+ if (attrs & ATTR_TEMPORARY) info |= STATX_INFO_TEMPORARY;
+ if (attrs & ATTR_ENCRYPTED) info |= STATX_INFO_ENCRYPTED;
+ stat->information |= info;
+ stat->result_mask |= STATX_WIN_ATTRS;
+
+ return 0;
 }
 
 static int cifs_truncate_page(struct address_space *mapping, loff_t from)

Reply | Threaded
Open this post in threaded view
|

[PATCH 5/6] statx: Make windows attributes available for CIFS, NTFS and FAT to use

David Howells
In reply to this post by David Howells
Make windows attributes available for CIFS, NTFS and FAT to use in the
statx struct.  The attribute flags map directly by value to those in the
CIFS PDU flags.  Some of these bits can also be used by JFS, UFS and HPFS.

The statx struct acquires:

        __u32 st_win_attrs;

The value in this is present if STATX_WIN_ATTRS is set.

The defined flags in this are:

        STATX_WIN_ATTR_READONLY
        STATX_WIN_ATTR_HIDDEN
        STATX_WIN_ATTR_SYSTEM
        STATX_WIN_ATTR_DIRECTORY
        STATX_WIN_ATTR_ARCHIVE
        STATX_WIN_ATTR_NORMAL
        STATX_WIN_ATTR_TEMPORARY
        STATX_WIN_ATTR_SPARSE_FILE
        STATX_WIN_ATTR_REPARSE_POINT
        STATX_WIN_ATTR_COMPRESSED
        STATX_WIN_ATTR_OFFLINE
        STATX_WIN_ATTR_NOT_CONTENT_INDEXED
        STATX_WIN_ATTR_ENCRYPTED
        STATX_WIN_ATTR_INTEGRITY_STREAM
        STATX_WIN_ATTR_NO_SCRUB_DATA

Signed-off-by: David Howells <[hidden email]>
---

 fs/stat.c                  |    5 ++++-
 include/uapi/linux/stat.h  |   30 ++++++++++++++++++++++++++++--
 samples/statx/test-statx.c |   31 +++++++++++++++++++++++++++++++
 3 files changed, 63 insertions(+), 3 deletions(-)

diff --git a/fs/stat.c b/fs/stat.c
index c2f8370dab13..1552bff154e6 100644
--- a/fs/stat.c
+++ b/fs/stat.c
@@ -83,6 +83,7 @@ int vfs_xgetattr_nosec(struct path *path, struct kstat *stat)
 
  stat->result_mask = 0;
  stat->information = 0;
+ stat->win_attrs = 0;
  if (inode->i_op->getattr)
  return inode->i_op->getattr(path->mnt, path->dentry, stat);
 
@@ -643,7 +644,9 @@ static long statx_set_result(struct kstat *stat, struct statx __user *buffer)
     __put_user(stat->blocks, &buffer->st_blocks ) ||
     __put_user(stat->version, &buffer->st_version ) ||
     __put_user(stat->gen, &buffer->st_gen ) ||
-    __clear_user(&buffer->__spare1, sizeof(buffer->__spare1)))
+    __put_user(stat->win_attrs, &buffer->st_win_attrs ) ||
+    __clear_user(&buffer->__spare1,
+ sizeof(buffer->__spare1) + sizeof(buffer->__spare2)))
  return -EFAULT;
 
  return 0;
diff --git a/include/uapi/linux/stat.h b/include/uapi/linux/stat.h
index 55ce6607dab6..64729b0785bf 100644
--- a/include/uapi/linux/stat.h
+++ b/include/uapi/linux/stat.h
@@ -106,7 +106,10 @@ struct statx {
  __u32 st_dev_major; /* ID of device containing file [uncond] */
  __u32 st_dev_minor;
  /* 0x80 */
- __u64 __spare1[16]; /* Spare space for future expansion */
+ __u32 st_win_attrs; /* Windows file attributes */
+ __u32 __spare1[3];
+ /* 0x90 */
+ __u64 __spare2[14]; /* Spare space for future expansion */
  /* 0x100 */
 };
 
@@ -133,7 +136,8 @@ struct statx {
 #define STATX_BTIME 0x00000800U /* Want/got st_btime */
 #define STATX_VERSION 0x00001000U /* Want/got st_version */
 #define STATX_GEN 0x00002000U /* Want/got st_gen */
-#define STATX_ALL_STATS 0x00003fffU /* All supported stats */
+#define STATX_WIN_ATTRS 0x00004000U /* Want/got st_win_attrs */
+#define STATX_ALL_STATS 0x00007fffU /* All supported stats */
 
 /*
  * Flags to be found in st_information
@@ -151,4 +155,26 @@ struct statx {
 #define STATX_INFO_AUTODIR 0x00000040U /* Dir provides unlisted automounts */
 #define STATX_INFO_NONSYSTEM_OWNERSHIP 0x00000080U /* File has non-system ownership details */
 
+/*
+ * Flags to be found in st_win_attrs.
+ *
+ * These give information about the state of a file on a Windows filesystem
+ * (such as. FAT, NTFS, CIFS).  These values are borrowed from the CIFS fs.
+ */
+#define STATX_WIN_ATTR_READONLY 0x00000001
+#define STATX_WIN_ATTR_HIDDEN 0x00000002
+#define STATX_WIN_ATTR_SYSTEM 0x00000004
+#define STATX_WIN_ATTR_DIRECTORY 0x00000010
+#define STATX_WIN_ATTR_ARCHIVE 0x00000020
+#define STATX_WIN_ATTR_NORMAL 0x00000080
+#define STATX_WIN_ATTR_TEMPORARY 0x00000100
+#define STATX_WIN_ATTR_SPARSE_FILE 0x00000200
+#define STATX_WIN_ATTR_REPARSE_POINT 0x00000400
+#define STATX_WIN_ATTR_COMPRESSED 0x00000800
+#define STATX_WIN_ATTR_OFFLINE 0x00001000
+#define STATX_WIN_ATTR_NOT_CONTENT_INDEXED 0x00002000
+#define STATX_WIN_ATTR_ENCRYPTED 0x00004000
+#define STATX_WIN_ATTR_INTEGRITY_STREAM 0x00008000
+#define STATX_WIN_ATTR_NO_SCRUB_DATA 0x00020000
+
 #endif /* _UAPI_LINUX_STAT_H */
diff --git a/samples/statx/test-statx.c b/samples/statx/test-statx.c
index 38ef23c12e7d..86eac7d16c32 100644
--- a/samples/statx/test-statx.c
+++ b/samples/statx/test-statx.c
@@ -134,6 +134,37 @@ static void dump_statx(struct statx *stx)
  if (stx->st_mask & STATX_GEN)
  printf("Inode gen   : %xh\n", stx->st_gen);
 
+ if (stx->st_mask & STATX_WIN_ATTRS) {
+ unsigned char bits;
+ int loop, byte;
+
+ static char wattr_representation[32 + 1] =
+ /* STATX_WIN_ATTR_ flags: */
+ "????????" /* 31-24 0x00000000-ff000000 */
+ "??????N?" /* 23-16 0x00000000-00ff0000 */
+ "ieNocrst" /* 15- 8 0x00000000-0000ff00 */
+ "n?Ad?SHR" /*  7- 0 0x00000000-000000ff */
+ ;
+
+ printf("Win attrs  : %08llx (",
+       (unsigned long long)stx->st_win_attrs);
+ for (byte = 32 - 8; byte >= 0; byte -= 8) {
+ bits = stx->st_win_attrs >> byte;
+ for (loop = 7; loop >= 0; loop--) {
+ int bit = byte + loop;
+
+ if (bits & 0x80)
+ putchar(wattr_representation[31 - bit]);
+ else
+ putchar('-');
+ bits <<= 1;
+ }
+ if (byte)
+ putchar(' ');
+ }
+ printf(")\n");
+ }
+
  if (stx->st_information) {
  unsigned char bits;
  int loop, byte;

Reply | Threaded
Open this post in threaded view
|

Re: [RFC][PATCH 0/6] Enhanced file stat system call

Jeff Layton-3
In reply to this post by David Howells
On Fri, 2016-04-29 at 13:57 +0100, David Howells wrote:

> Implement a new system call to provide enhanced file stats.  The patches can
> be found here:
>
> http://git.kernel.org/cgit/linux/kernel/git/dhowells/linux-fs.git/log/?h=xstat
>
>
> ===========
> DESCRIPTION
> ===========
>
> The first patch provides this new system call:
>
> long ret = statx(int dfd,
>  const char *filename,
>  unsigned atflag,
>  unsigned mask,
>  struct statx *buffer);
>
> This is an enhanced file stat function that provides a number of useful
> features, in summary:
>
>  (1) More information: creation time, data version number, inode generation
>      number and flags.  A subset of these is available through a number of
>      filesystems (such as CIFS, NFS, AFS, Ext4 and BTRFS).
>
>  (2) Lightweight stat (AT_NO_ATTR_SYNC): Ask for just those details of
>      interest, and allow a network fs to approximate anything not of
>      interest, without going to the server.
>
>  (3) Heavyweight stat (AT_FORCE_ATTR_SYNC): Force a network fs to flush
>      buffers and go to the server, even if it thinks its cached attributes
>      are up to date.
>
>  (4) Allow the filesystem to indicate what it can/cannot provide: A
>      filesystem can now say it doesn't support a standard stat feature if
>      that isn't available.
>
>  (5) Make the fields a consistent size on all arches, and make them large.
>
>  (6) Can be extended by using more request flags and using up the padding
>      space in the statx struct.
>
> Note that no lstat() equivalent is required as that can be implemented
> through statx() with atflag == 0.  There is also no fstat() equivalent as
> that can be implemented through statx() with filename == NULL and the
> relevant fd passed as dfd.
>
>
> =======
> TESTING
> =======
>
> A test program is added into samples/statx/ by the first patch.
>
> David
> ---
> David Howells (6):
>       statx: Add a system call to make enhanced file info available
>       statx: AFS: Return enhanced file attributes
>       statx: Ext4: Return enhanced file attributes
>       statx: NFS: Return enhanced file attributes
>       statx: Make windows attributes available for CIFS, NTFS and FAT to use
>       statx: CIFS: Return enhanced attributes
>
>
>  arch/x86/entry/syscalls/syscall_32.tbl |    1 
>  arch/x86/entry/syscalls/syscall_64.tbl |    1 
>  fs/afs/inode.c                         |   24 ++-
>  fs/cifs/cifsfs.h                       |    4 
>  fs/cifs/cifsglob.h                     |    8 +
>  fs/cifs/dir.c                          |    2 
>  fs/cifs/inode.c                        |  117 +++++++++---
>  fs/exportfs/expfs.c                    |    4 
>  fs/ext4/ext4.h                         |    2 
>  fs/ext4/file.c                         |    2 
>  fs/ext4/inode.c                        |   30 +++
>  fs/ext4/namei.c                        |    2 
>  fs/ext4/symlink.c                      |    2 
>  fs/nfs/inode.c                         |   41 ++++
>  fs/stat.c                              |  306 +++++++++++++++++++++++++++++---
>  include/linux/fs.h                     |    5 -
>  include/linux/stat.h                   |   15 +-
>  include/linux/syscalls.h               |    4 
>  include/uapi/linux/fcntl.h             |    2 
>  include/uapi/linux/stat.h              |  135 ++++++++++++++
>  samples/Makefile                       |    2 
>  samples/statx/Makefile                 |   10 +
>  samples/statx/test-statx.c             |  274 +++++++++++++++++++++++++++++
>  23 files changed, 910 insertions(+), 83 deletions(-)
>  create mode 100644 samples/statx/Makefile
>  create mode 100644 samples/statx/test-statx.c
>


I looked over the patchset and it looks fairly straightforward to me.

Are there things that I'd have probably done differently? Yes, but
nothing really sticks out to me as show-stopping and this has the
potential to be helpful in all sorts of ways.

I think we really ought to resist excessive bikeshedding this time
around. With the way the interface is designed, we should be able to
expand on it later if we want to add new attributes after the interface
itself is merged.

Reviewed-by: Jeff Layton <[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 1/6] statx: Add a system call to make enhanced file info available

Andreas Dilger-7
In reply to this post by David Howells
On Apr 29, 2016, at 6:57 AM, David Howells <[hidden email]> wrote:
>
> Add a system call to make extended file information available, including
> file creation time, inode version and data version where available through
> the underlying filesystem.

Hi David,
thanks for resubmitting the patch series.  No requests to add features here,
just a couple of comments on the patches regarding the implementation...

> ========
> OVERVIEW
> ========
>
> The idea was initially proposed as a set of xattrs that could be retrieved
> with getxattr(), but the general preferance proved to be for a new syscall
> with an extended stat structure.
>
> This has a number of uses:
>
> (1) Better support for the y2038 problem [Arnd Bergmann].
>
> (2) Creation time: The SMB protocol carries the creation time, which could
>     be exported by Samba, which will in turn help CIFS make use of
>     FS-Cache as that can be used for coherency data.
>
>     This is also specified in NFSv4 as a recommended attribute and could
>     be exported by NFSD [Steve French].
>
> (3) Lightweight stat: Ask for just those details of interest, and allow a
>     netfs (such as NFS) to approximate anything not of interest, possibly
>     without going to the server [Trond Myklebust, Ulrich Drepper, Andreas
>     Dilger].
>
> (4) Heavyweight stat: Force a netfs to go to the server, even if it thinks
>     its cached attributes are up to date [Trond Myklebust].
>
> (5) Data version number: Could be used by userspace NFS servers [Aneesh
>     Kumar].
>
>     Can also be used to modify fill_post_wcc() in NFSD which retrieves
>     i_version directly, but has just called vfs_getattr().  It could get
>     it from the kstat struct if it used vfs_xgetattr() instead.
>
> (6) BSD stat compatibility: Including more fields from the BSD stat such
>     as creation time (st_btime) and inode generation number (st_gen)
>     [Jeremy Allison, Bernd Schubert].
>
> (7) Inode generation number: Useful for FUSE and userspace NFS servers
>     [Bernd Schubert].  This was asked for but later deemed unnecessary
>     with the open-by-handle capability available
>
> (8) Extra coherency data may be useful in making backups [Andreas Dilger].
>
> (9) Allow the filesystem to indicate what it can/cannot provide: A
>     filesystem can now say it doesn't support a standard stat feature if
>     that isn't available, so if, for instance, inode numbers or UIDs don't
>     exist or are fabricated locally...
>
> (10) Make the fields a consistent size on all arches and make them large.
>
> (11) Store a 16-byte volume ID in the superblock that can be returned in
>     struct xstat [Steve French].
>
> (12) Include granularity fields in the time data to indicate the
>     granularity of each of the times (NFSv4 time_delta) [Steve French].
>
> (13) FS_IOC_GETFLAGS value.  These could be translated to BSD's st_flags.
>     Note that the Linux IOC flags are a mess and filesystems such as Ext4
>     define flags that aren't in linux/fs.h, so translation in the kernel
>     may be a necessity (or, possibly, we provide the filesystem type too).
>
> (14) Mask of features available on file (eg: ACLs, seclabel) [Brad Boyer,
>     Michael Kerrisk].
>
> (15) Spare space, request flags and information flags are provided for
>     future expansion.
>
> Note that not all of the above are implemented here.
>
>
> ===============
> NEW SYSTEM CALL
> ===============
>
> The new system call is:
>
> int ret = statx(int dfd,
> const char *filename,
> unsigned int flags,
> unsigned int mask,
> struct statx *buffer);
>
> The dfd, filename and flags parameters indicate the file to query.  There
> is no equivalent of lstat() as that can be emulated with statx() by passing
> AT_SYMLINK_NOFOLLOW in flags.  There is also no equivalent of fstat() as
> that can be emulated by passing a NULL filename to statx() with the fd of
> interest in dfd.
>
> AT_FORCE_ATTR_SYNC can be set in flags.  This will require a network
> filesystem to synchronise its attributes with the server.
>
> AT_NO_ATTR_SYNC can be set in flags.  This will suppress synchronisation
> with the server in a network filesystem.  The resulting values should be
> considered approximate.
>
> mask is a bitmask indicating the fields in struct statx that are of
> interest to the caller.  The user should set this to STATX_BASIC_STATS to
> get the basic set returned by stat().
>
> buffer points to the destination for the data.  This must be 256 bytes in
> size.
>
>
> ======================
> MAIN ATTRIBUTES RECORD
> ======================
>
> The following structures are defined in which to return the main attribute
> set:
>
> struct statx {
> __u32 st_mask;
> __u32 st_information;
> __u32 st_blksize;
> __u32 st_nlink;
> __u32 st_gen;
> __u32 st_uid;
> __u32 st_gid;
> __u16 st_mode;
> __u16 __spare0[1];
> __u64 st_ino;
> __u64 st_size;
> __u64 st_blocks;
> __u64 st_version;
> __s64 st_atime;
> __s64 st_btime;
> __s64 st_ctime;
> __s64 st_mtime;
> __s32 st_atime_ns;
> __s32 st_btime_ns;
> __s32 st_ctime_ns;
> __s32 st_mtime_ns;
> __u32 st_rdev_major;
> __u32 st_rdev_minor;
> __u32 st_dev_major;
> __u32 st_dev_minor;
> __u64 __spare1[16];
> };
>
> where st_information is local system information about the file, st_gen is
> the inode generation number, st_btime is the file creation time, st_version
> is the data version number (i_version), st_mask is a bitmask indicating the
> data provided and __spares*[] are where as-yet undefined fields can be
> placed.
>
> Time fields are split into separate seconds and nanoseconds fields to make
> packing easier and the granularities can be queried with the filesystem
> info system call.  Note that times will be negative if before 1970; in such
> a case, the nanosecond fields should also be negative if not zero.
>
> The defined bits in request_mask and st_mask are:
>
> STATX_MODE Want/got st_mode
> STATX_NLINK Want/got st_nlink
> STATX_UID Want/got st_uid
> STATX_GID Want/got st_gid
> STATX_RDEV Want/got st_rdev_*
> STATX_ATIME Want/got st_atime
> STATX_MTIME Want/got st_mtime
> STATX_CTIME Want/got st_ctime
> STATX_INO Want/got st_ino
> STATX_SIZE Want/got st_size
> STATX_BLOCKS Want/got st_blocks
> STATX_BASIC_STATS [The stuff in the normal stat struct]
> STATX_BTIME Want/got st_btime
> STATX_VERSION Want/got st_data_version
> STATX_GEN Want/got st_gen
> STATX_ALL_STATS [All currently available stuff]
>
> The defined bits in the st_information field give local system data on a
> file, how it is accessed, where it is and what it does:
>
> STATX_INFO_ENCRYPTED File is encrypted
This flag overlaps with FS_ENCRYPT_FL that is encoded in the FS_IOC_GETFLAGS
attributes.  Are the FS_* flags expected to be translated into STATX_INFO_*
flags by each filesystem, or will they be partly duplicated in a separate
"st_attrs" field added in the future?

Cheers, Andreas

> STATX_INFO_TEMPORARY File is temporary
> STATX_INFO_FABRICATED File was made up by filesystem
> STATX_INFO_KERNEL_API File is kernel API (eg: procfs/sysfs)
> STATX_INFO_REMOTE File is remote
> STATX_INFO_AUTOMOUNT Dir is automount trigger
> STATX_INFO_AUTODIR Dir provides unlisted automounts
> STATX_INFO_NONSYSTEM_OWNERSHIP File has non-system ownership details
>
> These are for the use of GUI tools that might want to mark files specially,
> depending on what they are.
>
> Fields in struct statx come in a number of classes:
>
> (0) st_information, st_dev_*, st_blksize.
>
>     These are local data and are always available.
>
> (1) st_nlinks, st_uid, st_gid, st_[amc]time*, st_ino, st_size, st_blocks.
>
>     These will be returned whether the caller asks for them or not.  The
>     corresponding bits in st_mask will be set to indicate whether they
>     actually have valid values.
>
>     If the caller didn't ask for them, then they may be approximated.  For
>     example, NFS won't waste any time updating them from the server,
>     unless as a byproduct of updating something requested.
>
>     If the values don't actually exist for the underlying object (such as
>     UID or GID on a DOS file), then the bit won't be set in the st_mask,
>     even if the caller asked for the value.  In such a case, the returned
>     value will be a fabrication.
>
> (2) st_mode.
>
>     The part of this that identifies the file type will always be
>     available, irrespective of the setting of STATX_MODE.  The access
>     flags and sticky bit are as for class (1).
>
> (3) st_rdev_*.
>
>     As for class (1), but this will be cleared if the file is not a
>     blockdev or chardev.  The bit will be cleared if the value is not
>     returned.
>
> (4) File creation time (st_btime*), data version (st_version), inode
>     generation number (st_gen).
>
>     These will be returned if available whether the caller asked for them or
>     not.  The corresponding bits in st_mask will be set or cleared as
>     appropriate to indicate a valid value.
>
>     If the caller didn't ask for them, then they may be approximated.  For
>     example, NFS won't waste any time updating them from the server, unless
>     as a byproduct of updating something requested.
>
>
> =======
> TESTING
> =======
>
> The following test program can be used to test the statx system call:
>
> samples/statx/test-statx.c
>
> Just compile and run, passing it paths to the files you want to examine.
> The file is built automatically if CONFIG_SAMPLES is enabled.
>
> Here's some example output.  Firstly, an NFS directory that crosses to
> another FSID.  Note that the FABRICATED and AUTOMOUNT info flags are set.
> The former because the directory is invented locally as we don't see the
> underlying dir on the server, the latter because transiting this directory
> will cause d_automount to be invoked by the VFS.
>
> [root@andromeda tmp]# ./samples/statx/test-statx -A /warthog/data
> statx(/warthog/data) = 0
> results=4fef
>  Size: 4096            Blocks: 8          IO Block: 1048576  directory
> Device: 00:1d           Inode: 2           Links: 110
> Access: (3777/drwxrwxrwx)  Uid: -2
> Gid: 4294967294
> Access: 2012-04-30 09:01:55.283819565+0100
> Modify: 2012-03-28 19:01:19.405465361+0100
> Change: 2012-03-28 19:01:19.405465361+0100
> Data version: ef51734f11e92a18h
> Information: 00000134 (-------- -------- -------a --mr-f--)
>
> Secondly, the result of automounting on that directory.
>
> [root@andromeda tmp]# ./samples/statx/test-statx /warthog/data
> statx(/warthog/data) = 0
> results=14fef
>  Size: 4096            Blocks: 8          IO Block: 1048576  directory
> Device: 00:1e           Inode: 2           Links: 110
> Access: (3777/drwxrwxrwx)  Uid: -2
> Gid: 4294967294
> Access: 2012-04-30 09:01:55.283819565+0100
> Modify: 2012-03-28 19:01:19.405465361+0100
> Change: 2012-03-28 19:01:19.405465361+0100
> Data version: ef51734f11e92a18h
> Information: 00000110 (-------- -------- -------a ---r----)
>
> Signed-off-by: David Howells <[hidden email]>
> ---
>
> arch/x86/entry/syscalls/syscall_32.tbl |    1
> arch/x86/entry/syscalls/syscall_64.tbl |    1
> fs/exportfs/expfs.c                    |    4
> fs/stat.c                              |  303 +++++++++++++++++++++++++++++---
> include/linux/fs.h                     |    5 -
> include/linux/stat.h                   |   15 +-
> include/linux/syscalls.h               |    4
> include/uapi/linux/fcntl.h             |    2
> include/uapi/linux/stat.h              |  109 ++++++++++++
> samples/Makefile                       |    2
> samples/statx/Makefile                 |   10 +
> samples/statx/test-statx.c             |  243 ++++++++++++++++++++++++++
> 12 files changed, 662 insertions(+), 37 deletions(-)
> create mode 100644 samples/statx/Makefile
> create mode 100644 samples/statx/test-statx.c
>
> diff --git a/arch/x86/entry/syscalls/syscall_32.tbl b/arch/x86/entry/syscalls/syscall_32.tbl
> index b30dd8154cc2..b99a6b3a167c 100644
> --- a/arch/x86/entry/syscalls/syscall_32.tbl
> +++ b/arch/x86/entry/syscalls/syscall_32.tbl
> @@ -386,3 +386,4 @@
> 377 i386 copy_file_range sys_copy_file_range
> 378 i386 preadv2 sys_preadv2
> 379 i386 pwritev2 sys_pwritev2
> +380 i386 statx sys_statx
> diff --git a/arch/x86/entry/syscalls/syscall_64.tbl b/arch/x86/entry/syscalls/syscall_64.tbl
> index cac6d17ce5db..6d5ef6c87cdc 100644
> --- a/arch/x86/entry/syscalls/syscall_64.tbl
> +++ b/arch/x86/entry/syscalls/syscall_64.tbl
> @@ -335,6 +335,7 @@
> 326 common copy_file_range sys_copy_file_range
> 327 64 preadv2 sys_preadv2
> 328 64 pwritev2 sys_pwritev2
> +329 common statx sys_statx
>
> #
> # x32-specific system call numbers start at 512 to avoid cache impact
> diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c
> index c46f1a190b8d..cd6d9cbc9300 100644
> --- a/fs/exportfs/expfs.c
> +++ b/fs/exportfs/expfs.c
> @@ -295,7 +295,9 @@ static int get_name(const struct path *path, char *name, struct dentry *child)
> * filesystem supports 64-bit inode numbers.  So we need to
> * actually call ->getattr, not just read i_ino:
> */
> - error = vfs_getattr_nosec(&child_path, &stat);
> + stat.query_flags = 0;
> + stat.request_mask = STATX_BASIC_STATS;
> + error = vfs_xgetattr_nosec(&child_path, &stat);
> if (error)
> return error;
> buffer.ino = stat.ino;
> diff --git a/fs/stat.c b/fs/stat.c
> index bc045c7994e1..c2f8370dab13 100644
> --- a/fs/stat.c
> +++ b/fs/stat.c
> @@ -18,6 +18,15 @@
> #include <asm/uaccess.h>
> #include <asm/unistd.h>
>
> +/**
> + * generic_fillattr - Fill in the basic attributes from the inode struct
> + * @inode: Inode to use as the source
> + * @stat: Where to fill in the attributes
> + *
> + * Fill in the basic attributes in the kstat structure from data that's to be
> + * found on the VFS inode structure.  This is the default if no getattr inode
> + * operation is supplied.
> + */
> void generic_fillattr(struct inode *inode, struct kstat *stat)
> {
> stat->dev = inode->i_sb->s_dev;
> @@ -27,87 +36,197 @@ void generic_fillattr(struct inode *inode, struct kstat *stat)
> stat->uid = inode->i_uid;
> stat->gid = inode->i_gid;
> stat->rdev = inode->i_rdev;
> - stat->size = i_size_read(inode);
> - stat->atime = inode->i_atime;
> stat->mtime = inode->i_mtime;
> stat->ctime = inode->i_ctime;
> - stat->blksize = (1 << inode->i_blkbits);
> + stat->size = i_size_read(inode);
> stat->blocks = inode->i_blocks;
> -}
> + stat->blksize = 1 << inode->i_blkbits;
>
> + stat->result_mask |= STATX_BASIC_STATS & ~STATX_RDEV;
> + if (IS_NOATIME(inode))
> + stat->result_mask &= ~STATX_ATIME;
> + else
> + stat->atime = inode->i_atime;
> +
> + if (S_ISREG(stat->mode) && stat->nlink == 0)
> + stat->information |= STATX_INFO_TEMPORARY;
> + if (IS_AUTOMOUNT(inode))
> + stat->information |= STATX_INFO_AUTOMOUNT;
> +
> + if (unlikely(S_ISBLK(stat->mode) || S_ISCHR(stat->mode)))
> + stat->result_mask |= STATX_RDEV;
> +}
> EXPORT_SYMBOL(generic_fillattr);
>
> /**
> - * vfs_getattr_nosec - getattr without security checks
> + * vfs_xgetattr_nosec - getattr without security checks
>  * @path: file to get attributes from
>  * @stat: structure to return attributes in
>  *
>  * Get attributes without calling security_inode_getattr.
>  *
> - * Currently the only caller other than vfs_getattr is internal to the
> - * filehandle lookup code, which uses only the inode number and returns
> - * no attributes to any user.  Any other code probably wants
> - * vfs_getattr.
> + * Currently the only caller other than vfs_xgetattr is internal to the
> + * filehandle lookup code, which uses only the inode number and returns no
> + * attributes to any user.  Any other code probably wants vfs_xgetattr.
> + *
> + * The caller must set stat->request_mask to indicate what they want and
> + * stat->query_flags to indicate whether the server should be queried.
>  */
> -int vfs_getattr_nosec(struct path *path, struct kstat *stat)
> +int vfs_xgetattr_nosec(struct path *path, struct kstat *stat)
> {
> struct inode *inode = d_backing_inode(path->dentry);
>
> + stat->query_flags &= ~KSTAT_QUERY_FLAGS;
> + if ((stat->query_flags & AT_FORCE_ATTR_SYNC) &&
> +    (stat->query_flags & AT_NO_ATTR_SYNC))
> + return -EINVAL;
> +
> + stat->result_mask = 0;
> + stat->information = 0;
> if (inode->i_op->getattr)
> return inode->i_op->getattr(path->mnt, path->dentry, stat);
>
> generic_fillattr(inode, stat);
> return 0;
> }
> +EXPORT_SYMBOL(vfs_xgetattr_nosec);
>
> -EXPORT_SYMBOL(vfs_getattr_nosec);
> -
> -int vfs_getattr(struct path *path, struct kstat *stat)
> +/*
> + * vfs_xgetattr - Get the enhanced basic attributes of a file
> + * @path: The file of interest
> + * @stat: Where to return the statistics
> + *
> + * Ask the filesystem for a file's attributes.  The caller must have preset
> + * stat->request_mask and stat->query_flags to indicate what they want.
> + *
> + * If the file is remote, the filesystem can be forced to update the attributes
> + * from the backing store by passing AT_FORCE_ATTR_SYNC in query_flags or can
> + * suppress the update by passing AT_NO_ATTR_SYNC.
> + *
> + * Bits must have been set in stat->request_mask to indicate which attributes
> + * the caller wants retrieving.  Any such attribute not requested may be
> + * returned anyway, but the value may be approximate, and, if remote, may not
> + * have been synchronised with the server.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_xgetattr(struct path *path, struct kstat *stat)
> {
> int retval;
>
> retval = security_inode_getattr(path);
> if (retval)
> return retval;
> - return vfs_getattr_nosec(path, stat);
> + return vfs_xgetattr_nosec(path, stat);
> }
> +EXPORT_SYMBOL(vfs_xgetattr);
>
> +/**
> + * vfs_getattr - Get the basic attributes of a file
> + * @path: The file of interest
> + * @stat: Where to return the statistics
> + *
> + * Ask the filesystem for a file's attributes.  If remote, the filesystem isn't
> + * forced to update its files from the backing store.  Only the basic set of
> + * attributes will be retrieved; anyone wanting more must use vfs_xgetattr(),
> + * as must anyone who wants to force attributes to be sync'd with the server.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_getattr(struct path *path, struct kstat *stat)
> +{
> + stat->query_flags = 0;
> + stat->request_mask = STATX_BASIC_STATS;
> + return vfs_xgetattr(path, stat);
> +}
> EXPORT_SYMBOL(vfs_getattr);
>
> -int vfs_fstat(unsigned int fd, struct kstat *stat)
> +/**
> + * vfs_fstatx - Get the enhanced basic attributes by file descriptor
> + * @fd: The file descriptor referring to the file of interest
> + * @stat: The result structure to fill in.
> + *
> + * This function is a wrapper around vfs_xgetattr().  The main difference is
> + * that it uses a file descriptor to determine the file location.
> + *
> + * The caller must have preset stat->query_flags and stat->request_mask as for
> + * vfs_xgetattr().
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_fstatx(unsigned int fd, struct kstat *stat)
> {
> struct fd f = fdget_raw(fd);
> int error = -EBADF;
>
> if (f.file) {
> - error = vfs_getattr(&f.file->f_path, stat);
> + error = vfs_xgetattr(&f.file->f_path, stat);
> fdput(f);
> }
> return error;
> }
> +EXPORT_SYMBOL(vfs_fstatx);
> +
> +/**
> + * vfs_fstat - Get basic attributes by file descriptor
> + * @fd: The file descriptor referring to the file of interest
> + * @stat: The result structure to fill in.
> + *
> + * This function is a wrapper around vfs_getattr().  The main difference is
> + * that it uses a file descriptor to determine the file location.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_fstat(unsigned int fd, struct kstat *stat)
> +{
> + stat->query_flags = 0;
> + stat->request_mask = STATX_BASIC_STATS;
> + return vfs_fstatx(fd, stat);
> +}
> EXPORT_SYMBOL(vfs_fstat);
>
> -int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat,
> - int flag)
> +/**
> + * vfs_statx - Get basic and extra attributes by filename
> + * @dfd: A file descriptor representing the base dir for a relative filename
> + * @filename: The name of the file of interest
> + * @flags: Flags to control the query
> + * @stat: The result structure to fill in.
> + *
> + * This function is a wrapper around vfs_xgetattr().  The main difference is
> + * that it uses a filename and base directory to determine the file location.
> + * Additionally, the addition of AT_SYMLINK_NOFOLLOW to flags will prevent a
> + * symlink at the given name from being referenced.
> + *
> + * The caller must have preset stat->request_mask as for vfs_xgetattr().  The
> + * flags are also used to load up stat->query_flags.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_statx(int dfd, const char __user *filename, int flags,
> +      struct kstat *stat)
> {
> struct path path;
> int error = -EINVAL;
> - unsigned int lookup_flags = 0;
> + unsigned int lookup_flags = LOOKUP_FOLLOW | LOOKUP_AUTOMOUNT;
>
> - if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_NO_AUTOMOUNT |
> -      AT_EMPTY_PATH)) != 0)
> - goto out;
> + if ((flags & ~(AT_SYMLINK_NOFOLLOW | AT_NO_AUTOMOUNT |
> +       AT_EMPTY_PATH | KSTAT_QUERY_FLAGS)) != 0)
> + return -EINVAL;
>
> - if (!(flag & AT_SYMLINK_NOFOLLOW))
> - lookup_flags |= LOOKUP_FOLLOW;
> - if (flag & AT_EMPTY_PATH)
> + if (flags & AT_SYMLINK_NOFOLLOW)
> + lookup_flags &= ~LOOKUP_FOLLOW;
> + if (flags & AT_NO_AUTOMOUNT)
> + lookup_flags &= ~LOOKUP_AUTOMOUNT;
> + if (flags & AT_EMPTY_PATH)
> lookup_flags |= LOOKUP_EMPTY;
> + stat->query_flags = flags;
> +
> retry:
> error = user_path_at(dfd, filename, lookup_flags, &path);
> if (error)
> goto out;
>
> - error = vfs_getattr(&path, stat);
> + error = vfs_xgetattr(&path, stat);
> path_put(&path);
> if (retry_estale(error, lookup_flags)) {
> lookup_flags |= LOOKUP_REVAL;
> @@ -116,17 +235,65 @@ retry:
> out:
> return error;
> }
> +EXPORT_SYMBOL(vfs_statx);
> +
> +/**
> + * vfs_fstatat - Get basic attributes by filename
> + * @dfd: A file descriptor representing the base dir for a relative filename
> + * @filename: The name of the file of interest
> + * @flags: Flags to control the query
> + * @stat: The result structure to fill in.
> + *
> + * This function is a wrapper around vfs_statx().  The difference is that it
> + * preselects basic stats only.  The flags are used to load up
> + * stat->query_flags in addition to indicating symlink handling during path
> + * resolution.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat,
> + int flags)
> +{
> + stat->request_mask = STATX_BASIC_STATS;
> + return vfs_statx(dfd, filename, flags, stat);
> +}
> EXPORT_SYMBOL(vfs_fstatat);
>
> -int vfs_stat(const char __user *name, struct kstat *stat)
> +/**
> + * vfs_stat - Get basic attributes by filename
> + * @filename: The name of the file of interest
> + * @stat: The result structure to fill in.
> + *
> + * This function is a wrapper around vfs_statx().  The difference is that it
> + * preselects basic stats only, terminal symlinks are followed regardless and a
> + * remote filesystem can't be forced to query the server.  If such is desired,
> + * vfs_statx() should be used instead.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> +int vfs_stat(const char __user *filename, struct kstat *stat)
> {
> - return vfs_fstatat(AT_FDCWD, name, stat, 0);
> + stat->request_mask = STATX_BASIC_STATS;
> + return vfs_statx(AT_FDCWD, filename, 0, stat);
> }
> EXPORT_SYMBOL(vfs_stat);
>
> +/**
> + * vfs_lstat - Get basic attrs by filename, without following terminal symlink
> + * @filename: The name of the file of interest
> + * @stat: The result structure to fill in.
> + *
> + * This function is a wrapper around vfs_statx().  The difference is that it
> + * preselects basic stats only, terminal symlinks are note followed regardless
> + * and a remote filesystem can't be forced to query the server.  If such is
> + * desired, vfs_statx() should be used instead.
> + *
> + * 0 will be returned on success, and a -ve error code if unsuccessful.
> + */
> int vfs_lstat(const char __user *name, struct kstat *stat)
> {
> - return vfs_fstatat(AT_FDCWD, name, stat, AT_SYMLINK_NOFOLLOW);
> + stat->request_mask = STATX_BASIC_STATS;
> + return vfs_statx(AT_FDCWD, name, AT_SYMLINK_NOFOLLOW, stat);
> }
> EXPORT_SYMBOL(vfs_lstat);
>
> @@ -141,7 +308,7 @@ static int cp_old_stat(struct kstat *stat, struct __old_kernel_stat __user * sta
> {
> static int warncount = 5;
> struct __old_kernel_stat tmp;
> -
> +
> if (warncount > 0) {
> warncount--;
> printk(KERN_WARNING "VFS: Warning: %s using old stat() call. Recompile your binary.\n",
> @@ -166,7 +333,7 @@ static int cp_old_stat(struct kstat *stat, struct __old_kernel_stat __user * sta
> #if BITS_PER_LONG == 32
> if (stat->size > MAX_NON_LFS)
> return -EOVERFLOW;
> -#endif
> +#endif
> tmp.st_size = stat->size;
> tmp.st_atime = stat->atime.tv_sec;
> tmp.st_mtime = stat->mtime.tv_sec;
> @@ -443,6 +610,80 @@ SYSCALL_DEFINE4(fstatat64, int, dfd, const char __user *, filename,
> }
> #endif /* __ARCH_WANT_STAT64 || __ARCH_WANT_COMPAT_STAT64 */
>
> +/*
> + * Set the statx results.
> + */
> +static long statx_set_result(struct kstat *stat, struct statx __user *buffer)
> +{
> + uid_t uid = from_kuid_munged(current_user_ns(), stat->uid);
> + gid_t gid = from_kgid_munged(current_user_ns(), stat->gid);
> +
> +#define __put_timestamp(kts, uts) ( \
> + __put_user(kts.tv_sec, uts##_s ) || \
> + __put_user(kts.tv_nsec, uts##_ns ))
> +
> + if (__put_user(stat->result_mask, &buffer->st_mask ) ||
> +    __put_user(stat->mode, &buffer->st_mode ) ||
> +    __clear_user(&buffer->__spare0, sizeof(buffer->__spare0))  ||
> +    __put_user(stat->nlink, &buffer->st_nlink ) ||
> +    __put_user(uid, &buffer->st_uid ) ||
> +    __put_user(gid, &buffer->st_gid ) ||
> +    __put_user(stat->information, &buffer->st_information ) ||
> +    __put_user(stat->blksize, &buffer->st_blksize ) ||
> +    __put_user(MAJOR(stat->rdev), &buffer->st_rdev_major ) ||
> +    __put_user(MINOR(stat->rdev), &buffer->st_rdev_minor ) ||
> +    __put_user(MAJOR(stat->dev), &buffer->st_dev_major ) ||
> +    __put_user(MINOR(stat->dev), &buffer->st_dev_minor ) ||
> +    __put_timestamp(stat->atime, &buffer->st_atime ) ||
> +    __put_timestamp(stat->btime, &buffer->st_btime ) ||
> +    __put_timestamp(stat->ctime, &buffer->st_ctime ) ||
> +    __put_timestamp(stat->mtime, &buffer->st_mtime ) ||
> +    __put_user(stat->ino, &buffer->st_ino ) ||
> +    __put_user(stat->size, &buffer->st_size ) ||
> +    __put_user(stat->blocks, &buffer->st_blocks ) ||
> +    __put_user(stat->version, &buffer->st_version ) ||
> +    __put_user(stat->gen, &buffer->st_gen ) ||
> +    __clear_user(&buffer->__spare1, sizeof(buffer->__spare1)))
> + return -EFAULT;
> +
> + return 0;
> +}
> +
> +/**
> + * sys_statx - System call to get enhanced stats
> + * @dfd: Base directory to pathwalk from *or* fd to stat.
> + * @filename: File to stat *or* NULL.
> + * @flags: AT_* flags to control pathwalk.
> + * @mask: Parts of statx struct actually required.
> + * @buffer: Result buffer.
> + *
> + * Note that if filename is NULL, then it does the equivalent of fstat() using
> + * dfd to indicate the file of interest.
> + */
> +SYSCALL_DEFINE5(statx,
> + int, dfd, const char __user *, filename, unsigned, flags,
> + unsigned int, mask,
> + struct statx __user *, buffer)
> +{
> + struct kstat stat;
> + int error;
> +
> + if (!access_ok(VERIFY_WRITE, buffer, sizeof(*buffer)))
> + return -EFAULT;
> +
> + memset(&stat, 0, sizeof(stat));
> + stat.query_flags = flags;
> + stat.request_mask = mask & STATX_ALL_STATS;
> +
> + if (filename)
> + error = vfs_statx(dfd, filename, flags, &stat);
> + else
> + error = vfs_fstatx(dfd, &stat);
> + if (error)
> + return error;
> + return statx_set_result(&stat, buffer);
> +}
> +
> /* Caller is here responsible for sufficient locking (ie. inode->i_lock) */
> void __inode_add_bytes(struct inode *inode, loff_t bytes)
> {
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 70e61b58baaf..8b2f6df924e9 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -2827,8 +2827,9 @@ extern const struct inode_operations page_symlink_inode_operations;
> extern void kfree_link(void *);
> extern int generic_readlink(struct dentry *, char __user *, int);
> extern void generic_fillattr(struct inode *, struct kstat *);
> -int vfs_getattr_nosec(struct path *path, struct kstat *stat);
> +extern int vfs_xgetattr_nosec(struct path *path, struct kstat *stat);
> extern int vfs_getattr(struct path *, struct kstat *);
> +extern int vfs_xgetattr(struct path *, struct kstat *);
> void __inode_add_bytes(struct inode *inode, loff_t bytes);
> void inode_add_bytes(struct inode *inode, loff_t bytes);
> void __inode_sub_bytes(struct inode *inode, loff_t bytes);
> @@ -2845,6 +2846,8 @@ extern int vfs_stat(const char __user *, struct kstat *);
> extern int vfs_lstat(const char __user *, struct kstat *);
> extern int vfs_fstat(unsigned int, struct kstat *);
> extern int vfs_fstatat(int , const char __user *, struct kstat *, int);
> +extern int vfs_xstat(int, const char __user *, int, struct kstat *);
> +extern int vfs_xfstat(unsigned int, struct kstat *);
>
> extern int __generic_block_fiemap(struct inode *inode,
>  struct fiemap_extent_info *fieinfo,
> diff --git a/include/linux/stat.h b/include/linux/stat.h
> index 075cb0c7eb2a..4f1902b0cb94 100644
> --- a/include/linux/stat.h
> +++ b/include/linux/stat.h
> @@ -19,6 +19,13 @@
> #include <linux/uidgid.h>
>
> struct kstat {
> + u32 query_flags; /* Operational flags */
> +#define KSTAT_QUERY_FLAGS (AT_FORCE_ATTR_SYNC | AT_NO_ATTR_SYNC)
> + u32 request_mask; /* What fields the user asked for */
> + u32 result_mask; /* What fields the user got */
> + u32 information;
> + u32 win_attrs; /* Windows file attributes */
> + u32 gen;
> u64 ino;
> dev_t dev;
> umode_t mode;
> @@ -27,11 +34,13 @@ struct kstat {
> kgid_t gid;
> dev_t rdev;
> loff_t size;
> - struct timespec  atime;
> + struct timespec atime;
> struct timespec mtime;
> struct timespec ctime;
> - unsigned long blksize;
> - unsigned long long blocks;
> + struct timespec btime; /* File creation time */
> + uint32_t blksize; /* Preferred I/O size */
> + u64 blocks;
> + u64 version; /* Data version */
> };
>
> #endif
> diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
> index d795472c54d8..f6bfbf74e44d 100644
> --- a/include/linux/syscalls.h
> +++ b/include/linux/syscalls.h
> @@ -48,6 +48,7 @@ struct stat;
> struct stat64;
> struct statfs;
> struct statfs64;
> +struct statx;
> struct __sysctl_args;
> struct sysinfo;
> struct timespec;
> @@ -898,4 +899,7 @@ asmlinkage long sys_copy_file_range(int fd_in, loff_t __user *off_in,
>
> asmlinkage long sys_mlock2(unsigned long start, size_t len, int flags);
>
> +asmlinkage long sys_statx(int dfd, const char __user *path, unsigned flags,
> +  unsigned mask, struct statx __user *buffer);
> +
> #endif
> diff --git a/include/uapi/linux/fcntl.h b/include/uapi/linux/fcntl.h
> index beed138bd359..5c8143b04ff7 100644
> --- a/include/uapi/linux/fcntl.h
> +++ b/include/uapi/linux/fcntl.h
> @@ -62,6 +62,8 @@
> #define AT_SYMLINK_FOLLOW 0x400   /* Follow symbolic links.  */
> #define AT_NO_AUTOMOUNT 0x800 /* Suppress terminal automount traversal */
> #define AT_EMPTY_PATH 0x1000 /* Allow empty relative pathname */
> +#define AT_FORCE_ATTR_SYNC 0x2000 /* Force the attributes to be sync'd with the server */
> +#define AT_NO_ATTR_SYNC 0x4000 /* Don't sync attributes with the server */
>
>
> #endif /* _UAPI_LINUX_FCNTL_H */
> diff --git a/include/uapi/linux/stat.h b/include/uapi/linux/stat.h
> index 7fec7e36d921..55ce6607dab6 100644
> --- a/include/uapi/linux/stat.h
> +++ b/include/uapi/linux/stat.h
> @@ -1,6 +1,7 @@
> #ifndef _UAPI_LINUX_STAT_H
> #define _UAPI_LINUX_STAT_H
>
> +#include <linux/types.h>
>
> #if defined(__KERNEL__) || !defined(__GLIBC__) || (__GLIBC__ < 2)
>
> @@ -41,5 +42,113 @@
>
> #endif
>
> +/*
> + * Structures for the extended file attribute retrieval system call
> + * (statx()).
> + *
> + * The caller passes a mask of what they're specifically interested in as a
> + * parameter to statx().  What statx() actually got will be indicated in
> + * st_mask upon return.
> + *
> + * For each bit in the mask argument:
> + *
> + * - if the datum is not available at all, the field and the bit will both be
> + *   cleared;
> + *
> + * - otherwise, if explicitly requested:
> + *
> + *   - the datum will be synchronised to the server if AT_FORCE_ATTR_SYNC is
> + *     set or if the datum is considered out of date, and
> + *
> + *   - the field will be filled in and the bit will be set;
> + *
> + * - otherwise, if not requested, but available in approximate form without any
> + *   effort, it will be filled in anyway, and the bit will be set upon return
> + *   (it might not be up to date, however, and no attempt will be made to
> + *   synchronise the internal state first);
> + *
> + * - otherwise the field and the bit will be cleared before returning.
> + *
> + * Items in STATX_BASIC_STATS may be marked unavailable on return, but they
> + * will have values installed for compatibility purposes so that stat() and
> + * co. can be emulated in userspace.
> + */
> +struct statx {
> + /* 0x00 */
> + __u32 st_mask; /* What results were written [uncond] */
> + __u32 st_information; /* Information about the file [uncond] */
> + __u32 st_blksize; /* Preferred general I/O size [uncond] */
> + __u32 st_nlink; /* Number of hard links */
> + /* 0x10 */
> + __u32 st_gen; /* Inode generation number */
> + __u32 st_uid; /* User ID of owner */
> + __u32 st_gid; /* Group ID of owner */
> + __u16 st_mode; /* File mode */
> + __u16 __spare0[1];
> + /* 0x20 */
> + __u64 st_ino; /* Inode number */
> + __u64 st_size; /* File size */
> + __u64 st_blocks; /* Number of 512-byte blocks allocated */
> + __u64 st_version; /* Data version number */
> + /* 0x40 */
> + __s64 st_atime_s; /* Last access time */
> + __s64 st_btime_s; /* File creation time */
> + __s64 st_ctime_s; /* Last attribute change time */
> + __s64 st_mtime_s; /* Last data modification time */
> + /* 0x60 */
> + __s32 st_atime_ns; /* Last access time (ns part) */
> + __s32 st_btime_ns; /* File creation time (ns part) */
> + __s32 st_ctime_ns; /* Last attribute change time (ns part) */
> + __s32 st_mtime_ns; /* Last data modification time (ns part) */
> + /* 0x70 */
> + __u32 st_rdev_major; /* Device ID of special file */
> + __u32 st_rdev_minor;
> + __u32 st_dev_major; /* ID of device containing file [uncond] */
> + __u32 st_dev_minor;
> + /* 0x80 */
> + __u64 __spare1[16]; /* Spare space for future expansion */
> + /* 0x100 */
> +};
> +
> +/*
> + * Flags to be st_mask
> + *
> + * Query request/result mask for statx() and struct statx::st_mask.
> + *
> + * These bits should be set in the mask argument of statx() to request
> + * particular items when calling statx().
> + */
> +#define STATX_MODE 0x00000001U /* Want/got st_mode */
> +#define STATX_NLINK 0x00000002U /* Want/got st_nlink */
> +#define STATX_UID 0x00000004U /* Want/got st_uid */
> +#define STATX_GID 0x00000008U /* Want/got st_gid */
> +#define STATX_RDEV 0x00000010U /* Want/got st_rdev */
> +#define STATX_ATIME 0x00000020U /* Want/got st_atime */
> +#define STATX_MTIME 0x00000040U /* Want/got st_mtime */
> +#define STATX_CTIME 0x00000080U /* Want/got st_ctime */
> +#define STATX_INO 0x00000100U /* Want/got st_ino */
> +#define STATX_SIZE 0x00000200U /* Want/got st_size */
> +#define STATX_BLOCKS 0x00000400U /* Want/got st_blocks */
> +#define STATX_BASIC_STATS 0x000007ffU /* The stuff in the normal stat struct */
> +#define STATX_BTIME 0x00000800U /* Want/got st_btime */
> +#define STATX_VERSION 0x00001000U /* Want/got st_version */
> +#define STATX_GEN 0x00002000U /* Want/got st_gen */
> +#define STATX_ALL_STATS 0x00003fffU /* All supported stats */
> +
> +/*
> + * Flags to be found in st_information
> + *
> + * These give information about the features or the state of a file that might
> + * be of use to ordinary userspace programs such as GUIs or ls rather than
> + * specialised tools.
> + */
> +#define STATX_INFO_ENCRYPTED 0x00000001U /* File is encrypted */
> +#define STATX_INFO_TEMPORARY 0x00000002U /* File is temporary */
> +#define STATX_INFO_FABRICATED 0x00000004U /* File was made up by filesystem */
> +#define STATX_INFO_KERNEL_API 0x00000008U /* File is kernel API (eg: procfs/sysfs) */
> +#define STATX_INFO_REMOTE 0x00000010U /* File is remote */
> +#define STATX_INFO_AUTOMOUNT 0x00000020U /* Dir is automount trigger */
> +#define STATX_INFO_AUTODIR 0x00000040U /* Dir provides unlisted automounts */
> +#define STATX_INFO_NONSYSTEM_OWNERSHIP 0x00000080U /* File has non-system ownership details */
>
> #endif /* _UAPI_LINUX_STAT_H */
> diff --git a/samples/Makefile b/samples/Makefile
> index 48001d7e23f0..d2ebb4e48d19 100644
> --- a/samples/Makefile
> +++ b/samples/Makefile
> @@ -2,4 +2,4 @@
>
> obj-$(CONFIG_SAMPLES) += kobject/ kprobes/ trace_events/ livepatch/ \
>   hw_breakpoint/ kfifo/ kdb/ hidraw/ rpmsg/ seccomp/ \
> -   configfs/
> +   configfs/ statx/
> diff --git a/samples/statx/Makefile b/samples/statx/Makefile
> new file mode 100644
> index 000000000000..6765dabc4c8d
> --- /dev/null
> +++ b/samples/statx/Makefile
> @@ -0,0 +1,10 @@
> +# kbuild trick to avoid linker error. Can be omitted if a module is built.
> +obj- := dummy.o
> +
> +# List of programs to build
> +hostprogs-y := test-statx
> +
> +# Tell kbuild to always build the programs
> +always := $(hostprogs-y)
> +
> +HOSTCFLAGS_test-statx.o += -I$(objtree)/usr/include
> diff --git a/samples/statx/test-statx.c b/samples/statx/test-statx.c
> new file mode 100644
> index 000000000000..38ef23c12e7d
> --- /dev/null
> +++ b/samples/statx/test-statx.c
> @@ -0,0 +1,243 @@
> +/* Test the statx() system call
> + *
> + * Copyright (C) 2015 Red Hat, Inc. All Rights Reserved.
> + * Written by David Howells ([hidden email])
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public Licence
> + * as published by the Free Software Foundation; either version
> + * 2 of the Licence, or (at your option) any later version.
> + */
> +
> +#define _GNU_SOURCE
> +#define _ATFILE_SOURCE
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <string.h>
> +#include <unistd.h>
> +#include <ctype.h>
> +#include <errno.h>
> +#include <time.h>
> +#include <sys/syscall.h>
> +#include <sys/types.h>
> +#include <linux/stat.h>
> +#include <linux/fcntl.h>
> +#include <sys/stat.h>
> +
> +#define AT_FORCE_ATTR_SYNC 0x2000
> +#define AT_NO_ATTR_SYNC 0x4000
> +
> +static __attribute__((unused))
> +ssize_t statx(int dfd, const char *filename, unsigned flags,
> +      unsigned int mask, struct statx *buffer)
> +{
> + return syscall(__NR_statx, dfd, filename, flags, mask, buffer);
> +}
> +
> +static void print_time(const char *field, __s64 tv_sec, __s32 tv_nsec)
> +{
> + struct tm tm;
> + time_t tim;
> + char buffer[100];
> + int len;
> +
> + tim = tv_sec;
> + if (!localtime_r(&tim, &tm)) {
> + perror("localtime_r");
> + exit(1);
> + }
> + len = strftime(buffer, 100, "%F %T", &tm);
> + if (len == 0) {
> + perror("strftime");
> + exit(1);
> + }
> + printf("%s", field);
> + fwrite(buffer, 1, len, stdout);
> + printf(".%09u", tv_nsec);
> + len = strftime(buffer, 100, "%z", &tm);
> + if (len == 0) {
> + perror("strftime2");
> + exit(1);
> + }
> + fwrite(buffer, 1, len, stdout);
> + printf("\n");
> +}
> +
> +static void dump_statx(struct statx *stx)
> +{
> + char buffer[256], ft = '?';
> +
> + printf("results=%x\n", stx->st_mask);
> +
> + printf(" ");
> + if (stx->st_mask & STATX_SIZE)
> + printf(" Size: %-15llu", (unsigned long long)stx->st_size);
> + if (stx->st_mask & STATX_BLOCKS)
> + printf(" Blocks: %-10llu", (unsigned long long)stx->st_blocks);
> + printf(" IO Block: %-6llu ", (unsigned long long)stx->st_blksize);
> + if (stx->st_mask & STATX_MODE) {
> + switch (stx->st_mode & S_IFMT) {
> + case S_IFIFO: printf(" FIFO\n"); ft = 'p'; break;
> + case S_IFCHR: printf(" character special file\n"); ft = 'c'; break;
> + case S_IFDIR: printf(" directory\n"); ft = 'd'; break;
> + case S_IFBLK: printf(" block special file\n"); ft = 'b'; break;
> + case S_IFREG: printf(" regular file\n"); ft = '-'; break;
> + case S_IFLNK: printf(" symbolic link\n"); ft = 'l'; break;
> + case S_IFSOCK: printf(" socket\n"); ft = 's'; break;
> + default:
> + printf("unknown type (%o)\n", stx->st_mode & S_IFMT);
> + break;
> + }
> + }
> +
> + sprintf(buffer, "%02x:%02x", stx->st_dev_major, stx->st_dev_minor);
> + printf("Device: %-15s", buffer);
> + if (stx->st_mask & STATX_INO)
> + printf(" Inode: %-11llu", (unsigned long long) stx->st_ino);
> + if (stx->st_mask & STATX_SIZE)
> + printf(" Links: %-5u", stx->st_nlink);
> + if (stx->st_mask & STATX_RDEV)
> + printf(" Device type: %u,%u", stx->st_rdev_major, stx->st_rdev_minor);
> + printf("\n");
> +
> + if (stx->st_mask & STATX_MODE)
> + printf("Access: (%04o/%c%c%c%c%c%c%c%c%c%c)  ",
> +       stx->st_mode & 07777,
> +       ft,
> +       stx->st_mode & S_IRUSR ? 'r' : '-',
> +       stx->st_mode & S_IWUSR ? 'w' : '-',
> +       stx->st_mode & S_IXUSR ? 'x' : '-',
> +       stx->st_mode & S_IRGRP ? 'r' : '-',
> +       stx->st_mode & S_IWGRP ? 'w' : '-',
> +       stx->st_mode & S_IXGRP ? 'x' : '-',
> +       stx->st_mode & S_IROTH ? 'r' : '-',
> +       stx->st_mode & S_IWOTH ? 'w' : '-',
> +       stx->st_mode & S_IXOTH ? 'x' : '-');
> + if (stx->st_mask & STATX_UID)
> + printf("Uid: %5d   ", stx->st_uid);
> + if (stx->st_mask & STATX_GID)
> + printf("Gid: %5d\n", stx->st_gid);
> +
> + if (stx->st_mask & STATX_ATIME)
> + print_time("Access: ", stx->st_atime_s, stx->st_atime_ns);
> + if (stx->st_mask & STATX_MTIME)
> + print_time("Modify: ", stx->st_mtime_s, stx->st_mtime_ns);
> + if (stx->st_mask & STATX_CTIME)
> + print_time("Change: ", stx->st_ctime_s, stx->st_ctime_ns);
> + if (stx->st_mask & STATX_BTIME)
> + print_time(" Birth: ", stx->st_btime_s, stx->st_btime_ns);
> +
> + if (stx->st_mask & STATX_VERSION)
> + printf("Data version: %llxh\n",
> +       (unsigned long long)stx->st_version);
> +
> + if (stx->st_mask & STATX_GEN)
> + printf("Inode gen   : %xh\n", stx->st_gen);
> +
> + if (stx->st_information) {
> + unsigned char bits;
> + int loop, byte;
> +
> + static char info_representation[32 + 1] =
> + /* STATX_INFO_ flags: */
> + "????????" /* 31-24 0x00000000-ff000000 */
> + "????????" /* 23-16 0x00000000-00ff0000 */
> + "????????" /* 15- 8 0x00000000-0000ff00 */
> + "ndmrkfte" /*  7- 0 0x00000000-000000ff */
> + ;
> +
> + printf("Information: %08x (", stx->st_information);
> + for (byte = 32 - 8; byte >= 0; byte -= 8) {
> + bits = stx->st_information >> byte;
> + for (loop = 7; loop >= 0; loop--) {
> + int bit = byte + loop;
> +
> + if (bits & 0x80)
> + putchar(info_representation[31 - bit]);
> + else
> + putchar('-');
> + bits <<= 1;
> + }
> + if (byte)
> + putchar(' ');
> + }
> + printf(")\n");
> + }
> +
> + printf("IO-blocksize: blksize=%u\n", stx->st_blksize);
> +}
> +
> +static void dump_hex(unsigned long long *data, int from, int to)
> +{
> + unsigned offset, print_offset = 1, col = 0;
> +
> + from /= 8;
> + to = (to + 7) / 8;
> +
> + for (offset = from; offset < to; offset++) {
> + if (print_offset) {
> + printf("%04x: ", offset * 8);
> + print_offset = 0;
> + }
> + printf("%016llx", data[offset]);
> + col++;
> + if ((col & 3) == 0) {
> + printf("\n");
> + print_offset = 1;
> + } else {
> + printf(" ");
> + }
> + }
> +
> + if (!print_offset)
> + printf("\n");
> +}
> +
> +int main(int argc, char **argv)
> +{
> + struct statx stx;
> + int ret, raw = 0, atflag = AT_SYMLINK_NOFOLLOW;
> +
> + unsigned int mask = STATX_ALL_STATS;
> +
> + for (argv++; *argv; argv++) {
> + if (strcmp(*argv, "-F") == 0) {
> + atflag |= AT_FORCE_ATTR_SYNC;
> + continue;
> + }
> + if (strcmp(*argv, "-N") == 0) {
> + atflag |= AT_NO_ATTR_SYNC;
> + continue;
> + }
> + if (strcmp(*argv, "-L") == 0) {
> + atflag &= ~AT_SYMLINK_NOFOLLOW;
> + continue;
> + }
> + if (strcmp(*argv, "-O") == 0) {
> + mask &= ~STATX_BASIC_STATS;
> + continue;
> + }
> + if (strcmp(*argv, "-A") == 0) {
> + atflag |= AT_NO_AUTOMOUNT;
> + continue;
> + }
> + if (strcmp(*argv, "-R") == 0) {
> + raw = 1;
> + continue;
> + }
> +
> + memset(&stx, 0xbf, sizeof(stx));
> + ret = statx(AT_FDCWD, *argv, atflag, mask, &stx);
> + printf("statx(%s) = %d\n", *argv, ret);
> + if (ret < 0) {
> + perror(*argv);
> + exit(1);
> + }
> +
> + if (raw)
> + dump_hex((unsigned long long *)&stx, 0, sizeof(stx));
> +
> + dump_statx(&stx);
> + }
> + return 0;
> +}
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to [hidden email]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Cheers, Andreas






signature.asc (850 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 3/6] statx: Ext4: Return enhanced file attributes

Andreas Dilger-7
In reply to this post by David Howells
On Apr 29, 2016, at 6:57 AM, David Howells <[hidden email]> wrote:

>
> Return enhanced file attributes from the Ext4 filesystem.  This includes
> the following:
>
> (1) The inode creation time (i_crtime) as i_btime, setting STATX_BTIME.
>
> (2) The inode i_version as st_version if a file with I_VERSION set or a
>     directory, setting STATX_VERSION.
>
> (3) FS_xxx_FL flags are returned as for ioctl(FS_IOC_GETFLAGS), setting
>     STATX_IOC_FLAGS.
I don't see where this is implemented in this patch, and it should be
removed from the commit message to avoid confusion.

> This requires that all ext4 inodes have a getattr call, not just some of
> them, so to this end, split the ext4_getattr() function and only call part
> of it where appropriate.
>
> Example output:
>
> [root@andromeda ~]# ./samples/statx/test-statx /usr
> statx(/usr) = 0
> results=37ef
>  Size: 4096            Blocks: 16         IO Block: 4096    directory
> Device: 08:02           Inode: 1572865     Links: 14
> Access: (0755/drwxr-xr-x)  Uid:     0   Gid:     0
> Access: 2015-11-03 16:12:30.000000000+0000
> Modify: 2013-10-18 15:29:18.000000000+0100
> Change: 2013-10-18 15:29:18.000000000+0100
> Data version: 2fh
> Inode flags: 00000000 (-------- -------- -------- --------)
> IO-blocksize: blksize=4096
One minor strangeness in this commit message is that the above usage example
doesn't include nanoseconds or Birth time for ext4, when this should be
supported for all ext4 filesystems?  Was this example run on an older ext3
filesystem (or some filesystem upgraded from ext3)?  Any ext4-formatted fs
should have the larger inode size to handle these extra fields.

Also a bit strange that "IO Block: 4096" and "IO-blocksize: blksize=4096"
is listed twice, but that seems to be an issue with the test-statx tool
and not necessarily the data being returned.

Cheers, Andreas

>
> Signed-off-by: David Howells <[hidden email]>
> ---
>
> fs/ext4/ext4.h    |    2 ++
> fs/ext4/file.c    |    2 +-
> fs/ext4/inode.c   |   30 +++++++++++++++++++++++++++---
> fs/ext4/namei.c   |    2 ++
> fs/ext4/symlink.c |    2 ++
> 5 files changed, 34 insertions(+), 4 deletions(-)
>
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 349afebe21ee..2f25eaa63f39 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -2549,6 +2549,8 @@ extern int  ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
> struct kstat *stat);
> extern void ext4_evict_inode(struct inode *);
> extern void ext4_clear_inode(struct inode *);
> +extern int  ext4_file_getattr(struct vfsmount *mnt, struct dentry *dentry,
> +      struct kstat *stat);
> extern int  ext4_sync_inode(handle_t *, struct inode *);
> extern void ext4_dirty_inode(struct inode *, int);
> extern int ext4_change_inode_journal_flag(struct inode *, int);
> diff --git a/fs/ext4/file.c b/fs/ext4/file.c
> index fa2208bae2e1..45c7b8644d0e 100644
> --- a/fs/ext4/file.c
> +++ b/fs/ext4/file.c
> @@ -715,7 +715,7 @@ const struct file_operations ext4_file_operations = {
>
> const struct inode_operations ext4_file_inode_operations = {
> .setattr = ext4_setattr,
> - .getattr = ext4_getattr,
> + .getattr = ext4_file_getattr,
> .setxattr = generic_setxattr,
> .getxattr = generic_getxattr,
> .listxattr = ext4_listxattr,
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 981a1fc30eaa..309b6cff8afc 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -5100,11 +5100,35 @@ err_out:
> int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
> struct kstat *stat)
> {
> - struct inode *inode;
> - unsigned long long delalloc_blocks;
> + struct inode *inode = d_inode(dentry);
> + struct ext4_inode *raw_inode;
> + struct ext4_inode_info *ei = EXT4_I(inode);
> +
> + stat->result_mask |= STATX_GEN;
> + stat->gen = inode->i_generation;
> +
> + if (EXT4_FITS_IN_INODE(raw_inode, ei, i_crtime)) {
> + stat->result_mask |= STATX_BTIME;
> + stat->btime.tv_sec = ei->i_crtime.tv_sec;
> + stat->btime.tv_nsec = ei->i_crtime.tv_nsec;
> + }
> +
> + if (S_ISDIR(inode->i_mode) || IS_I_VERSION(inode)) {
> + stat->result_mask |= STATX_VERSION;
> + stat->version = inode->i_version;
> + }
>
> - inode = d_inode(dentry);
> generic_fillattr(inode, stat);
> + return 0;
> +}
> +
> +int ext4_file_getattr(struct vfsmount *mnt, struct dentry *dentry,
> +      struct kstat *stat)
> +{
> + struct inode *inode = dentry->d_inode;
> + u64 delalloc_blocks;
> +
> + ext4_getattr(mnt, dentry, stat);
>
> /*
> * If there is inline data in the inode, the inode will normally not
> diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
> index 48e4b8907826..9b5fdc971e22 100644
> --- a/fs/ext4/namei.c
> +++ b/fs/ext4/namei.c
> @@ -3882,6 +3882,7 @@ const struct inode_operations ext4_dir_inode_operations = {
> .tmpfile = ext4_tmpfile,
> .rename2 = ext4_rename2,
> .setattr = ext4_setattr,
> + .getattr = ext4_getattr,
> .setxattr = generic_setxattr,
> .getxattr = generic_getxattr,
> .listxattr = ext4_listxattr,
> @@ -3893,6 +3894,7 @@ const struct inode_operations ext4_dir_inode_operations = {
>
> const struct inode_operations ext4_special_inode_operations = {
> .setattr = ext4_setattr,
> + .getattr = ext4_getattr,
> .setxattr = generic_setxattr,
> .getxattr = generic_getxattr,
> .listxattr = ext4_listxattr,
> diff --git a/fs/ext4/symlink.c b/fs/ext4/symlink.c
> index 75ed5c2f0c16..54015f3d7516 100644
> --- a/fs/ext4/symlink.c
> +++ b/fs/ext4/symlink.c
> @@ -105,6 +105,7 @@ const struct inode_operations ext4_symlink_inode_operations = {
> .readlink = generic_readlink,
> .get_link = page_get_link,
> .setattr = ext4_setattr,
> + .getattr = ext4_getattr,
> .setxattr = generic_setxattr,
> .getxattr = generic_getxattr,
> .listxattr = ext4_listxattr,
> @@ -115,6 +116,7 @@ const struct inode_operations ext4_fast_symlink_inode_operations = {
> .readlink = generic_readlink,
> .get_link = simple_get_link,
> .setattr = ext4_setattr,
> + .getattr = ext4_getattr,
> .setxattr = generic_setxattr,
> .getxattr = generic_getxattr,
> .listxattr = ext4_listxattr,
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to [hidden email]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Cheers, Andreas






signature.asc (850 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 4/6] statx: NFS: Return enhanced file attributes

Andreas Dilger-7
In reply to this post by David Howells
On Apr 29, 2016, at 6:58 AM, David Howells <[hidden email]> wrote:

>
> Return enhanced file atrributes from the NFS filesystem.  This includes the
> following:
>
> (1) The change attribute as st_version if NFSv4.
>
> (2) STATX_INFO_AUTOMOUNT and STATX_INFO_FABRICATED are set on referral or
>     submount directories that are automounted upon.  NFS shows one
>     directory with a different FSID, but the local filesystem has two: the
>     mountpoint directory and the root of the filesystem mounted upon it.
>
> (3) STATX_INFO_REMOTE is set on files acquired over NFS.
>
> (4) STATX_IOC_FLAGS is set and if the atime is unavailable on a file,
>     st_ioc_flags will have FL_NOATIME_FL set in it.
This is not implemented in this patch, and should removed from the commit
message.  The generic_fillattr() change in [1/6] will clear the STATX_ATIME
flag if IS_NOATIME() is true, so maybe that is enough for what you intended?

Cheers, Andreas

> Furthermore, what nfs_getattr() does can be controlled as follows:
>
> (1) If AT_NO_ATTR_SYNC is indicated then this will suppress the flushing
>     of outstanding writes and the rereading of the inode's attributes with
>     the server as detailed below.
>
> (2) Otherwise:
>
>     (a) If AT_FORCE_ATTR_SYNC is indicated, or mtime, ctime or
>     data_version (NFSv4 only) are requested then the outstanding
>     writes will be written to the server first.
>
>     (b) The inode's attributes will be reread from the server:
>
>     (i) if AT_FORCE_ATTR_SYNC is indicated;
>
>        (ii) if atime is requested (and atime updating is not suppressed by
>         a mount flag); or
>
>       (iii) if the cached attributes have expired;
>
> If the inode isn't synchronised, then the cached attributes will be used -
> even if expired - without reference to the server.
>
> Example output:
>
> [root@andromeda ~]# ./samples/statx/test-statx /warthog/
> statx(/warthog/) = 0
> results=37ef
>  Size: 4096            Blocks: 8          IO Block: 1048576  directory
> Device: 00:26           Inode: 2           Links: 122
> Access: (3777/drwxrwxrwx)  Uid:     0   Gid:  4041
> Access: 2015-10-30 16:15:41.730925545+0000
> Modify: 2015-10-07 10:33:19.896108112+0100
> Change: 2015-10-07 10:33:19.896108112+0100
> Data version: 5614e6df35698650h
> Inode flags: 00000000 (-------- -------- -------- --------)
> Information: 00000010 (-------- -------- -------- ---r----)
> IO-blocksize: blksize=1048576
>
> Note that the NFS4 protocol potentially provides a creation time that could
> be passed through this interface and system, hidden and archive values that
> could be passed as IOC flags.  There is also a backup time that could be
> added.
>
> Signed-off-by: David Howells <[hidden email]>
> ---
>
> fs/nfs/inode.c |   41 ++++++++++++++++++++++++++++++++++-------
> 1 file changed, 34 insertions(+), 7 deletions(-)
>
> diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
> index 738c84a42eb0..8637236bca0c 100644
> --- a/fs/nfs/inode.c
> +++ b/fs/nfs/inode.c
> @@ -655,12 +655,23 @@ static bool nfs_need_revalidate_inode(struct inode *inode)
> int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
> {
> struct inode *inode = d_inode(dentry);
> - int need_atime = NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATIME;
> + bool force_sync = stat->query_flags & AT_FORCE_ATTR_SYNC;
> + bool suppress_sync = stat->query_flags & AT_NO_ATTR_SYNC;
> + bool need_atime = NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATIME;
> int err = 0;
>
> trace_nfs_getattr_enter(inode);
> - /* Flush out writes to the server in order to update c/mtime.  */
> - if (S_ISREG(inode->i_mode)) {
> +
> + if (NFS_SERVER(inode)->nfs_client->rpc_ops->version < 4)
> + stat->request_mask &= ~STATX_VERSION;
> +
> + /* Flush out writes to the server in order to update c/mtime or data
> + * version if the user wants them.
> + */
> + if (S_ISREG(inode->i_mode) && !suppress_sync &&
> +    (force_sync || (stat->request_mask &
> +    (STATX_MTIME | STATX_CTIME | STATX_VERSION)))
> +    ) {
> inode_lock(inode);
> err = nfs_sync_inode(inode);
> inode_unlock(inode);
> @@ -677,11 +688,13 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
> *  - NFS never sets MS_NOATIME or MS_NODIRATIME so there is
> *    no point in checking those.
> */
> - if ((mnt->mnt_flags & MNT_NOATIME) ||
> -    ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)))
> - need_atime = 0;
> + if (!(stat->request_mask & STATX_ATIME) ||
> +    (mnt->mnt_flags & MNT_NOATIME) ||
> +    ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)))
> + need_atime = false;
>
> - if (need_atime || nfs_need_revalidate_inode(inode)) {
> + if (!suppress_sync &&
> +    (force_sync || need_atime || nfs_need_revalidate_inode(inode))) {
> struct nfs_server *server = NFS_SERVER(inode);
>
> if (server->caps & NFS_CAP_READDIRPLUS)
> @@ -694,6 +707,20 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
> if (S_ISDIR(inode->i_mode))
> stat->blksize = NFS_SERVER(inode)->dtsize;
> }
> +
> + generic_fillattr(inode, stat);
> + stat->ino = nfs_compat_user_ino64(NFS_FILEID(inode));
> +
> + if (stat->request_mask & STATX_VERSION) {
> + stat->version = inode->i_version;
> + stat->result_mask |= STATX_VERSION;
> + }
> +
> + if (IS_AUTOMOUNT(inode))
> + stat->information |= STATX_INFO_FABRICATED;
> +
> + stat->information |= STATX_INFO_REMOTE;
> +
> out:
> trace_nfs_getattr_exit(inode, err);
> return err;
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to [hidden email]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Cheers, Andreas






signature.asc (850 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 5/6] statx: Make windows attributes available for CIFS, NTFS and FAT to use

Andreas Dilger-7
In reply to this post by David Howells
On Apr 29, 2016, at 6:58 AM, David Howells <[hidden email]> wrote:
>
> Make windows attributes available for CIFS, NTFS and FAT to use in the
> statx struct.  The attribute flags map directly by value to those in the
> CIFS PDU flags.  Some of these bits can also be used by JFS, UFS and HPFS.
>
> The statx struct acquires:
>
> __u32 st_win_attrs;

It seems some of these flags are duplicated with the st_information field,
and some are duplicate with FS_IOC_GETFLAGS values, and returning the same
information in multiple ways is confusing.

If these flags are part of the CIFS protocol, and are directly usable by
Samba then I can understand we wouldn't want to change them once in the
kernel and then convert them back in userspace, but I'm a bit reluctant
to have flags only for CIFS/NTFS/FAT that might also be useful for other filesystems.  Would we want to be able to get translated st_win_attrs
flags in ext4 attrs when it is being exported by Samba?

> The value in this is present if STATX_WIN_ATTRS is set.
>
> The defined flags in this are:
>
> STATX_WIN_ATTR_READONLY
> STATX_WIN_ATTR_HIDDEN
> STATX_WIN_ATTR_SYSTEM
> STATX_WIN_ATTR_DIRECTORY

How does this differ from (st_mode & S_IFMT) == S_IFDIR)?

> STATX_WIN_ATTR_ARCHIVE
> STATX_WIN_ATTR_NORMAL

How does this differ from (st_mode & S_IFMT) == S_IFREG)?

> STATX_WIN_ATTR_TEMPORARY

How does this differ from STATX_INFO_TEMPORARY?

> STATX_WIN_ATTR_SPARSE_FILE
> STATX_WIN_ATTR_REPARSE_POINT
> STATX_WIN_ATTR_COMPRESSED
> STATX_WIN_ATTR_OFFLINE
> STATX_WIN_ATTR_NOT_CONTENT_INDEXED
> STATX_WIN_ATTR_ENCRYPTED

How does this differ from STATX_INFO_ENCRYPTED?

Cheers, Andreas

> STATX_WIN_ATTR_INTEGRITY_STREAM
> STATX_WIN_ATTR_NO_SCRUB_DATA
>
> Signed-off-by: David Howells <[hidden email]>
> ---
>
> fs/stat.c                  |    5 ++++-
> include/uapi/linux/stat.h  |   30 ++++++++++++++++++++++++++++--
> samples/statx/test-statx.c |   31 +++++++++++++++++++++++++++++++
> 3 files changed, 63 insertions(+), 3 deletions(-)
>
> diff --git a/fs/stat.c b/fs/stat.c
> index c2f8370dab13..1552bff154e6 100644
> --- a/fs/stat.c
> +++ b/fs/stat.c
> @@ -83,6 +83,7 @@ int vfs_xgetattr_nosec(struct path *path, struct kstat *stat)
>
> stat->result_mask = 0;
> stat->information = 0;
> + stat->win_attrs = 0;
> if (inode->i_op->getattr)
> return inode->i_op->getattr(path->mnt, path->dentry, stat);
>
> @@ -643,7 +644,9 @@ static long statx_set_result(struct kstat *stat, struct statx __user *buffer)
>    __put_user(stat->blocks, &buffer->st_blocks ) ||
>    __put_user(stat->version, &buffer->st_version ) ||
>    __put_user(stat->gen, &buffer->st_gen ) ||
> -    __clear_user(&buffer->__spare1, sizeof(buffer->__spare1)))
> +    __put_user(stat->win_attrs, &buffer->st_win_attrs ) ||
> +    __clear_user(&buffer->__spare1,
> + sizeof(buffer->__spare1) + sizeof(buffer->__spare2)))
> return -EFAULT;
>
> return 0;
> diff --git a/include/uapi/linux/stat.h b/include/uapi/linux/stat.h
> index 55ce6607dab6..64729b0785bf 100644
> --- a/include/uapi/linux/stat.h
> +++ b/include/uapi/linux/stat.h
> @@ -106,7 +106,10 @@ struct statx {
> __u32 st_dev_major; /* ID of device containing file [uncond] */
> __u32 st_dev_minor;
> /* 0x80 */
> - __u64 __spare1[16]; /* Spare space for future expansion */
> + __u32 st_win_attrs; /* Windows file attributes */
> + __u32 __spare1[3];
> + /* 0x90 */
> + __u64 __spare2[14]; /* Spare space for future expansion */
> /* 0x100 */
> };
>
> @@ -133,7 +136,8 @@ struct statx {
> #define STATX_BTIME 0x00000800U /* Want/got st_btime */
> #define STATX_VERSION 0x00001000U /* Want/got st_version */
> #define STATX_GEN 0x00002000U /* Want/got st_gen */
> -#define STATX_ALL_STATS 0x00003fffU /* All supported stats */
> +#define STATX_WIN_ATTRS 0x00004000U /* Want/got st_win_attrs */
> +#define STATX_ALL_STATS 0x00007fffU /* All supported stats */
>
> /*
>  * Flags to be found in st_information
> @@ -151,4 +155,26 @@ struct statx {
> #define STATX_INFO_AUTODIR 0x00000040U /* Dir provides unlisted automounts */
> #define STATX_INFO_NONSYSTEM_OWNERSHIP 0x00000080U /* File has non-system ownership details */
>
> +/*
> + * Flags to be found in st_win_attrs.
> + *
> + * These give information about the state of a file on a Windows filesystem
> + * (such as. FAT, NTFS, CIFS).  These values are borrowed from the CIFS fs.
> + */
> +#define STATX_WIN_ATTR_READONLY 0x00000001
> +#define STATX_WIN_ATTR_HIDDEN 0x00000002
> +#define STATX_WIN_ATTR_SYSTEM 0x00000004
> +#define STATX_WIN_ATTR_DIRECTORY 0x00000010
> +#define STATX_WIN_ATTR_ARCHIVE 0x00000020
> +#define STATX_WIN_ATTR_NORMAL 0x00000080
> +#define STATX_WIN_ATTR_TEMPORARY 0x00000100
> +#define STATX_WIN_ATTR_SPARSE_FILE 0x00000200
> +#define STATX_WIN_ATTR_REPARSE_POINT 0x00000400
> +#define STATX_WIN_ATTR_COMPRESSED 0x00000800
> +#define STATX_WIN_ATTR_OFFLINE 0x00001000
> +#define STATX_WIN_ATTR_NOT_CONTENT_INDEXED 0x00002000
> +#define STATX_WIN_ATTR_ENCRYPTED 0x00004000
> +#define STATX_WIN_ATTR_INTEGRITY_STREAM 0x00008000
> +#define STATX_WIN_ATTR_NO_SCRUB_DATA 0x00020000
> +
> #endif /* _UAPI_LINUX_STAT_H */
> diff --git a/samples/statx/test-statx.c b/samples/statx/test-statx.c
> index 38ef23c12e7d..86eac7d16c32 100644
> --- a/samples/statx/test-statx.c
> +++ b/samples/statx/test-statx.c
> @@ -134,6 +134,37 @@ static void dump_statx(struct statx *stx)
> if (stx->st_mask & STATX_GEN)
> printf("Inode gen   : %xh\n", stx->st_gen);
>
> + if (stx->st_mask & STATX_WIN_ATTRS) {
> + unsigned char bits;
> + int loop, byte;
> +
> + static char wattr_representation[32 + 1] =
> + /* STATX_WIN_ATTR_ flags: */
> + "????????" /* 31-24 0x00000000-ff000000 */
> + "??????N?" /* 23-16 0x00000000-00ff0000 */
> + "ieNocrst" /* 15- 8 0x00000000-0000ff00 */
> + "n?Ad?SHR" /*  7- 0 0x00000000-000000ff */
> + ;
> +
> + printf("Win attrs  : %08llx (",
> +       (unsigned long long)stx->st_win_attrs);
> + for (byte = 32 - 8; byte >= 0; byte -= 8) {
> + bits = stx->st_win_attrs >> byte;
> + for (loop = 7; loop >= 0; loop--) {
> + int bit = byte + loop;
> +
> + if (bits & 0x80)
> + putchar(wattr_representation[31 - bit]);
> + else
> + putchar('-');
> + bits <<= 1;
> + }
> + if (byte)
> + putchar(' ');
> + }
> + printf(")\n");
> + }
> +
> if (stx->st_information) {
> unsigned char bits;
> int loop, byte;
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to [hidden email]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Cheers, Andreas






signature.asc (850 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 1/6] statx: Add a system call to make enhanced file info available

David Howells
In reply to this post by Andreas Dilger-7
Andreas Dilger <[hidden email]> wrote:

> > STATX_INFO_ENCRYPTED File is encrypted
>
> This flag overlaps with FS_ENCRYPT_FL that is encoded in the FS_IOC_GETFLAGS
> attributes.  Are the FS_* flags expected to be translated into STATX_INFO_*
> flags by each filesystem, or will they be partly duplicated in a separate
> "st_attrs" field added in the future?

I think that most of the FS_IOC_GETFLAGS flags are sufficiently specialised
that they aren't something the ordinary user would necessarily find to be of
interest, so I'm not sure that mapping all of them to STATX_INFO_* flags is
necessary.

That said, I think STATX_INFO_ENCRYPTED *is* usefully deployed here to tell
the user that the file or directory is encrypted and that the user will have
to unlock or provide a key to access it.

I'm also thinking that a STATX_INFO_NEED_AUTHENTICATION flag may be needed to
indicate that the user must authenticate in some way (probably only applicable
to network files) to be able to access the file.

David
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 5/6] statx: Make windows attributes available for CIFS, NTFS and FAT to use

David Howells
In reply to this post by Andreas Dilger-7
Andreas Dilger <[hidden email]> wrote:

> > __u32 st_win_attrs;
>
> It seems some of these flags are duplicated with the st_information field,
> and some are duplicate with FS_IOC_GETFLAGS values, and returning the same
> information in multiple ways is confusing.
>
> If these flags are part of the CIFS protocol,

They're part of the Windows API.  CIFS/SMB and NTFS have them at the same
values.

> and are directly usable by Samba then I can understand we wouldn't want to
> change them once in the kernel and then convert them back in userspace, but
> I'm a bit reluctant to have flags only for CIFS/NTFS/FAT that might also be
> useful for other filesystems.

Where said "other filesystems" have some support for Windows or OS/2 and have
bits defined that map directly in semantics, if not in value, to these bits.

JFS, for example, has IREADONLY, IHIDDEN, ISYSTEM, IDIRECTORY, IARCHIVE and
ISPARSE.

> Would we want to be able to get translated
> st_win_attrs flags in ext4 attrs when it is being exported by Samba?

Actually, I was thinking of proposing an ext4 patch that stored windows flags
in a separate word in the inode on disk.  There isn't sufficient bit space to
map all the windows flags to FS_IOC_GETFLAGS.  However, as you say, there is a
certain amount of overlap...

David
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 3/6] statx: Ext4: Return enhanced file attributes

David Howells
In reply to this post by Andreas Dilger-7
Andreas Dilger <[hidden email]> wrote:

> > (3) FS_xxx_FL flags are returned as for ioctl(FS_IOC_GETFLAGS), setting
> >     STATX_IOC_FLAGS.
>
> I don't see where this is implemented in this patch, and it should be
> removed from the commit message to avoid confusion.

Oops - I forgot to remove it from the other patches.

David
Reply | Threaded
Open this post in threaded view
|

Re: [RFC][PATCH 0/6] Enhanced file stat system call

Arnd Bergmann
In reply to this post by David Howells
On Friday 29 April 2016 13:57:36 David Howells wrote:
>                          struct statx *buffer);
>
> This is an enhanced file stat function that provides a number of useful
> features, in summary:
>
>  (1) More information: creation time, data version number, inode generation
>      number and flags.  A subset of these is available through a number of
>      filesystems (such as CIFS, NFS, AFS, Ext4 and BTRFS).
>

I have a question about birthtime/creationtime: As we are gaining a way
to read this, should we also provide a way to update it using a new variant
of the utimensat syscall in order to have 'cp -a' create an identical copy,
or is the idea that this is defined as the time that is particular copy
of the inode was created?

I've discussed this with Deepa in the past, as she is driving the
convertion of the inode timestamps to timespec64 now, and we will
need a new version of utimensat for her work as well. I can see good
reasons either way (allowing updates of btime or disallowing them).

This should not hold up the statx syscall from getting merged as
soon as we can, but I'd like to know what everyone feels about that
question now that they are looking at the interface already.

        Arnd
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 1/6] statx: Add a system call to make enhanced file info available

Dave Chinner
In reply to this post by David Howells
On Fri, Apr 29, 2016 at 01:57:43PM +0100, David Howells wrote:
>  (4) File creation time (st_btime*), data version (st_version), inode
>      generation number (st_gen).
>
>      These will be returned if available whether the caller asked for them or
>      not.  The corresponding bits in st_mask will be set or cleared as
>      appropriate to indicate a valid value.

IMO, exposing the inode generation number to anyone is a potential
security problem because they are used in file handles.

Most file handles provided by filesystems are simply an encoding of
the inode number + generation number, plus maybe the ino+gen of the
parent dir if the NFS server is configured to do this. This makes it
trivial for an attacker to guess what the likely generation numbers
are going to be for inode numbers surrounding any given inode, hence
greatly reducing the search space for guessing valid file handles.

We've known this to be a problem for a long time - file handles are
not cryptographically secure, so exposing information like this by
default make guessing handles successfully almost trivial for many
filesystems.

In the latest XFS filesystem format, we randomise the generation
value during every inode allocation to make it hard to guess the
handle of adjacent inodes from an existing ino+gen pair, or even
from life time to life time of the same inode. We don't use a secure
random number generator (prandom_u32()) so it's still possible to
guess with enough trial and observation. However, it makes it
several orders of magnitude harder to guess and requires knowledge
of inode allocation order to guess correctly once the random number
sequence has been deduced and so makes brute force the only real
option for guessing a valid handle for an inode.

However, this is definitely a problem for the older format where
each cluster of inodes was initialised with the same seed at cluster
allocation time and the generation number was simply incremented for
each life time. Most filesystems use a similar method for seeding
and incrementing generation numbers, so once the generation numbers
are exposed it makes handles trivial to calculate successfully.

>      If the caller didn't ask for them, then they may be approximated.  For
>      example, NFS won't waste any time updating them from the server, unless
>      as a byproduct of updating something requested.

I would suggest that exposing them from the NFS server is something
we most definitely don't want to do because they are the only thing
that keeps remote users from guessing filehandles with ease....

Cheers,

Dave.
--
Dave Chinner
[hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 1/6] statx: Add a system call to make enhanced file info available

NeilBrown-2
In reply to this post by David Howells
On Fri, Apr 29 2016, David Howells wrote:

> Add a system call to make extended file information available, including
> file creation time, inode version and data version where available through
> the underlying filesystem.
>
>
> ========
> OVERVIEW
> ========

I think all this documentation is invaluable - thanks.
I would really like to see much of it in
    Documentation/filesystems/something.txt
rather than just in the commit log.

>
> The defined bits in the st_information field give local system data on a
> file, how it is accessed, where it is and what it does:

These bits form a channel for communication between the filesystem
developer and the application writer.  As such we should be sure that
channel actually communicates meaning...


>
> STATX_INFO_ENCRYPTED File is encrypted
> STATX_INFO_TEMPORARY File is temporary

What is "temporary"?  Is it a statement about quality of storage
technology (will be destroyed by reboot) or intention of creator
(created with O_TMPFILE) or something else?


> STATX_INFO_FABRICATED File was made up by filesystem
> STATX_INFO_KERNEL_API File is kernel API (eg: procfs/sysfs)

What is the difference between these two?  Both are synthesized by the
kernel.
Maybe the "KERNEL_API" is declared never to change its meaning, while the
fabricated one doesn't make a "stable API" promise?

What is the difference between fabricating a file from a bunch of blocks
spread over a storage device, and fabricating a file from a single field
in the super-block?


> STATX_INFO_REMOTE File is remote

How far is "remote"?  Does Infiniband count?  Fibre channel?  iSCSI?
Is a file on a loop-back mounted NFS filesystem more remote than a
fibre-channel connection to the next town?

Or is this relative?  Within a filesystem there are "remote" files and
"non-remote" files and the distinction is filesystem-dependant??

> STATX_INFO_AUTOMOUNT Dir is automount trigger
> STATX_INFO_AUTODIR Dir provides unlisted automounts

I think this last one means that there are names in the directory which
may not appear in "readdir" but will respond to "stat".  I would prefer
the description to match the behavior without necessarily implying that
those names will be automounts.  e.g
        STATX_INFO_INCOMPLETE_READDIR   getdents may not report all
                                        names that respond to stat

> STATX_INFO_NONSYSTEM_OWNERSHIP File has non-system ownership details

This probably is a well defined meaning that I just don't have the
context to understand.  For me, more words would help here.


I don't object to any of these flag.  I just want to be sure that I
understand them.
I am generally in favour this functionality going in promptly.

Thanks,
NeilBrown

signature.asc (834 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 1/6] statx: Add a system call to make enhanced file info available

NeilBrown-2
In reply to this post by Dave Chinner
On Thu, May 05 2016, Dave Chinner wrote:

> On Fri, Apr 29, 2016 at 01:57:43PM +0100, David Howells wrote:
>>  (4) File creation time (st_btime*), data version (st_version), inode
>>      generation number (st_gen).
>>
>>      These will be returned if available whether the caller asked for them or
>>      not.  The corresponding bits in st_mask will be set or cleared as
>>      appropriate to indicate a valid value.
>
> IMO, exposing the inode generation number to anyone is a potential
> security problem because they are used in file handles.
"security through obscurity".  We have Kerberos working really nicely
for NFS these days.  Do we still care?

What if the generation number were only made available to "root"?  Would
that allay your concerns?
Would that still be useful?
We already have name_to_handle_at().  Exposing the generation number
could/should follow the same rules at that.  Or maybe the exposure of
each field should be guided by the filesystem, depending on (for
example) whether it is used to provide uniqueness to the filehandle.

>
>>      If the caller didn't ask for them, then they may be approximated.  For
>>      example, NFS won't waste any time updating them from the server, unless
>>      as a byproduct of updating something requested.
>
> I would suggest that exposing them from the NFS server is something
> we most definitely don't want to do because they are the only thing
> that keeps remote users from guessing filehandles with ease....

Given that the NFS protocol does not define a "generation number"
attribute, I think there is no risk for them being exposed from the NFS
server ... except implicitly within the filehandle of course.

NeilBrown

signature.asc (834 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [PATCH 1/6] statx: Add a system call to make enhanced file info available

Jeff Layton-3
On Thu, 2016-05-05 at 10:09 +1000, NeilBrown wrote:

> On Thu, May 05 2016, Dave Chinner wrote:
>
> >
> > On Fri, Apr 29, 2016 at 01:57:43PM +0100, David Howells wrote:
> > >
> > >  (4) File creation time (st_btime*), data version (st_version), inode
> > >      generation number (st_gen).
> > >
> > >      These will be returned if available whether the caller asked for them or
> > >      not.  The corresponding bits in st_mask will be set or cleared as
> > >      appropriate to indicate a valid value.
> > IMO, exposing the inode generation number to anyone is a potential
> > security problem because they are used in file handles.
> "security through obscurity".  We have Kerberos working really nicely
> for NFS these days.  Do we still care?
>
> What if the generation number were only made available to "root"?  Would
> that allay your concerns?
> Would that still be useful?
> We already have name_to_handle_at().  Exposing the generation number
> could/should follow the same rules at that.  Or maybe the exposure of
> each field should be guided by the filesystem, depending on (for
> example) whether it is used to provide uniqueness to the filehandle.
>
> >
> >
> > >
> > >      If the caller didn't ask for them, then they may be approximated.  For
> > >      example, NFS won't waste any time updating them from the server, unless
> > >      as a byproduct of updating something requested.
> > I would suggest that exposing them from the NFS server is something
> > we most definitely don't want to do because they are the only thing
> > that keeps remote users from guessing filehandles with ease....
> Given that the NFS protocol does not define a "generation number"
> attribute, I think there is no risk for them being exposed from the NFS
> server ... except implicitly within the filehandle of course.
>
> NeilBrown



I don't see a real attack vector here either, but OTOH is there a
potential user of this at the moment? An earlier chunk of the patch
description says:

(7) Inode generation number: Useful for FUSE and userspace NFS servers
     [Bernd Schubert].  This was asked for but later deemed unnecessary
     with the open-by-handle capability available

...the last bit seems to indicate that we don't really need this
anyway, as most userland servers now work with filehandles from the
kernel.

Maybe leave it out for now? It can always be added later.

--
Jeff Layton <[hidden email]>

123