Prev: VIDEO: ivtvfb, remove unneeded NULL test
Next: [PATCH] nilfs2: remove BKL uses
From: Michael Kerrisk on 4 Jul 2010 00:30
[CC+=linux-api]

On Fri, Jul 2, 2010 at 1:57 AM, David Howells <dhowells(a)redhat.com> wrote:
> [This is, for the moment, to be considered an example. �Do we actually want to
> �export these flags? �Should they be a full member of struct xstat?]


Since I suggested the idea, obviously I'm inclined to think they should ;-).

Cheers,

Michael


> Allow an extra result to be requested that makes available some inode flags,
> along the lines of BSD's st_flags and Ext2/3/4's inode flags. �This is
> requested by setting XSTAT_REQUEST_INODE_FLAGS in the request_mask. �If the
> filesystem supports it for that file, then this will be set in result_mask and
> 16 bytes of information will be appended to the xstat buffer, if sufficient
> buffer space is available.
>
> The extra result is laid out according to the following structure:
>
> � � � �struct xstat_inode_flags {
> � � � � � � � �unsigned long long � � �st_flags;
> � � � � � � � �unsigned long long � � �st_supported_flags;
> � � � �};
>
> where the filesystem indicates the flags it supports for that file and the
> flags that are set on that file. �The structure is of length:
>
> � � � �XSTAT_LENGTH_INODE_FLAGS
>
> The flags come in three sets:
>
> �(1) User settable flags (to be consistent with the BSD st_flags field):
>
> � � � �UF_NODUMP � � � Do not dump this file.
> � � � �UF_IMMUTABLE � �This file is immutable.
> � � � �UF_APPEND � � � This file is append-only.
> � � � �UF_OPAQUE � � � This directory is opaque (unionfs).
> � � � �UF_NOUNLINK � � This file can't be removed or renamed.
> � � � �UF_COMPRESSED � This file is compressed.
> � � � �UF_HIDDEN � � � This file shouldn't be displayed in a GUI.
>
> � � The UF_SETTABLE constant is the union of the above flags.
>
> �(2) Superuser settable flags (to be consistent with the BSD st_flags field):
>
> � � � �SF_ARCHIVED � � This file has been archived.
> � � � �SF_IMMUTABLE � �This file is immutable.
> � � � �SF_APPEND � � � This file is append-only.
> � � � �SF_NOUNLINK � � This file can't be removed or renamed.
> � � � �SF_HIDDEN � � � This file is a snapshot inode.
>
> � � The SF_SETTABLE constant is the union of the above flags.
>
> �(3) Linux-specific flags:
>
> � � � �XSTAT_LF_MAGIC_FILE � � Magic file, such as found in procfs and sysfs.
> � � � �XSTAT_LF_SYNC � � � � � File is written synchronously.
> � � � �XSTAT_LF_NOATIME � � � �Atime is not updated on this file.
> � � � �XSTAT_LF_JOURNALLED_DATA Data modifications to this file are journalled.
> � � � �XSTAT_LF_ENCRYPTED � � �This file is encrypted.
> � � � �XSTAT_LF_SYSTEM � � � � This file is a system file (FAT/NTFS/CIFS).
> � � � �XSTAT_LF_TEMPORARY � � �This file is a temporary file (NTFS/CIFS).
> � � � �XSTAT_LF_OFFLINE � � � �file is currently unavailable (CIFS).
>
>
> The Ext4 filesystem has been modified to map certain Ext4 inode flags to the
> above:
>
> � � � �EXT4 FLAG � � � � � � � MAPPED TO
> � � � �======================= =======================================
> � � � �EXT4_COMPR_FL � � � � � UF_COMPRESSED
> � � � �EXT4_SYNC_FL � � � � � �XSTAT_LF_SYNC
> � � � �EXT4_IMMUTABLE_FL � � � UF_IMMUTABLE and SF_IMMUTABLE
> � � � �EXT4_APPEND_FL � � � � �UF_APPEND and SF_APPEND
> � � � �EXT4_NODUMP_FL � � � � �UF_NODUMP
> � � � �EXT4_NOATIME_FL � � � � XSTAT_LF_NOATIME
> � � � �EXT4_JOURNAL_DATA_FL � �XSTAT_LF_JOURNALLED_DATA
> � � � �EXT4_DIRSYNC_FL � � � � XSTAT_LF_SYNC (directories only)
>
> With this patch applied, the test program given in the patch that introduced
> the xstat() syscalls now does this:
>
> � � � �[root(a)andromeda ~]# chattr +ia /var/cache/fscache/cull_atimes
> � � � �[root(a)andromeda ~]# lsattr /var/cache/fscache/cull_atimes
> � � � �----ia-------e- /var/cache/fscache/cull_atimes
> � � � �[root(a)andromeda ~]# /tmp/xstat /var/cache/fscache/cull_atimes
> � � � �xstat(/var/cache/fscache/cull_atimes) = 168
> � � � �results=5fef
> � � � � �Size: 78088 � � � � � Blocks: 168 � � � �IO Block: 4096 � �regular file
> � � � �Device: 08:06 � � � � � Inode: 13 � � � � �Links: 1
> � � � �Access: (0600/-rw-------) �Uid: 0
> � � � �Gid: 0
> � � � �Access: 2010-06-29 18:17:41.092290108+0100
> � � � �Modify: 2010-06-25 17:25:53.320261493+0100
> � � � �Change: 2010-07-02 00:46:51.278803967+0100
> � � � �Create: 2010-06-25 15:17:39.711172889+0100
> � � � �Inode version: f585ab73h
> � � � �0098: 0000000000060006 0000000e00060027
>
> The extra results are hex dumped at the end in 64-bit chunks. �As can be seen
> above, st_flags=0x0000000000060006 and st_supported_flags=0000000e00060027.
> That's showing that the file now has [SU]F_IMMUTABLE and [SU]F_APPEND enabled.
>
> Note also that XSTAT_REQUEST_INODE_FLAGS (0x4000) is present in the result_mask
> value (0x5fef) returned to userspace, and the amount of data returned by
> xstat() has increased from 152 to 168 as appropriate for 16 bytes of extra
> data.
>
> Signed-off-by: David Howells <dhowells(a)redhat.com>
> ---
>
> �fs/ext4/ext4.h � � � | � �2 ++
> �fs/ext4/file.c � � � | � �1 +
> �fs/ext4/inode.c � � �| � 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
> �fs/ext4/namei.c � � �| � �2 ++
> �fs/ext4/symlink.c � �| � �2 ++
> �include/linux/stat.h | � 47 ++++++++++++++++++++++++++++++++++++++++++++++-
> �6 files changed, 103 insertions(+), 1 deletions(-)
>
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 96823f3..26b8dd6 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -1573,6 +1573,8 @@ extern int �ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
> � � � � � � � � � � � � � � � �struct kstat *stat);
> �extern int �ext4_file_getattr(struct vfsmount *mnt, struct dentry *dentry,
> � � � � � � � � � � � � � � � �struct kstat *stat);
> +extern int �ext4_getattr_extra(struct vfsmount *, struct dentry *,
> + � � � � � � � � � � � � � � �struct xstat_extra_result *);
> �extern void ext4_delete_inode(struct inode *);
> �extern int �ext4_sync_inode(handle_t *, struct inode *);
> �extern void ext4_dirty_inode(struct inode *);
> diff --git a/fs/ext4/file.c b/fs/ext4/file.c
> index 18c29ab..657ffa0 100644
> --- a/fs/ext4/file.c
> +++ b/fs/ext4/file.c
> @@ -151,6 +151,7 @@ const struct inode_operations ext4_file_inode_operations = {
> � � � �.truncate � � � = ext4_truncate,
> � � � �.setattr � � � �= ext4_setattr,
> � � � �.getattr � � � �= ext4_file_getattr,
> + � � � .getattr_extra �= ext4_getattr_extra,
> �#ifdef CONFIG_EXT4_FS_XATTR
> � � � �.setxattr � � � = generic_setxattr,
> � � � �.getxattr � � � = generic_getxattr,
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index f9a730a..efa17d6 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -5595,6 +5595,56 @@ int ext4_file_getattr(struct vfsmount *mnt, struct dentry *dentry,
> � � � �return 0;
> �}
>
> +int ext4_getattr_inode_flags(struct inode *inode,
> + � � � � � � � � � � � � � �struct xstat_extra_result *extra)
> +{
> + � � � struct ext4_inode_info *ei = EXT4_I(inode);
> + � � � struct xstat_inode_flags xif = { 0, 0 };
> +
> +#define _(FL, ST) � � � � � � � � � �\
> + � � � xif.st_supported_flags |= ST; \
> + � � � if (ei->i_flags & FL) � � � � \
> + � � � � � � � xif.st_flags |= ST;
> +
> + � � � _(EXT4_COMPR_FL, � � � �UF_COMPRESSED);
> + � � � _(EXT4_SYNC_FL, � � � � XSTAT_LF_SYNC);
> + � � � _(EXT4_IMMUTABLE_FL, � �UF_IMMUTABLE | SF_IMMUTABLE);
> + � � � _(EXT4_APPEND_FL, � � � UF_APPEND | SF_APPEND);
> + � � � _(EXT4_NODUMP_FL, � � � UF_NODUMP);
> + � � � _(EXT4_NOATIME_FL, � � �XSTAT_LF_NOATIME);
> + � � � _(EXT4_JOURNAL_DATA_FL, XSTAT_LF_JOURNALLED_DATA);
> +
> + � � � if (S_ISDIR(ei->vfs_inode.i_mode))
> + � � � � � � � _(EXT4_DIRSYNC_FL, � � �XSTAT_LF_SYNC);
> +
> + � � � return extra->pass_result(extra, ilog2(XSTAT_REQUEST_INODE_FLAGS),
> + � � � � � � � � � � � � � � � � &xif, sizeof(xif));
> +}
> +
> +int ext4_getattr_extra(struct vfsmount *mnt, struct dentry *dentry,
> + � � � � � � � � � � �struct xstat_extra_result *extra)
> +{
> + � � � struct inode *inode = dentry->d_inode;
> + � � � u64 request_mask = extra->request_mask;
> + � � � int request, ret;
> +
> + � � � do {
> + � � � � � � � request = __ffs64(request_mask);
> + � � � � � � � request_mask &= ~(1ULL << request);
> +
> + � � � � � � � switch (request) {
> + � � � � � � � case ilog2(XSTAT_REQUEST_INODE_FLAGS):
> + � � � � � � � � � � � ret = ext4_getattr_inode_flags(inode, extra);
> + � � � � � � � � � � � break;
> + � � � � � � � default:
> + � � � � � � � � � � � ret = 0;
> + � � � � � � � � � � � break;
> + � � � � � � � }
> +
> + � � � } while (ret == 0 && request_mask);
> + � � � return ret;
> +}
> +
> �static int ext4_indirect_trans_blocks(struct inode *inode, int nrblocks,
> � � � � � � � � � � � � � � � � � � �int chunk)
> �{
> diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
> index 0f776c7..3c37b3f 100644
> --- a/fs/ext4/namei.c
> +++ b/fs/ext4/namei.c
> @@ -2543,6 +2543,7 @@ const struct inode_operations ext4_dir_inode_operations = {
> � � � �.rename � � � � = ext4_rename,
> � � � �.setattr � � � �= ext4_setattr,
> � � � �.getattr � � � �= ext4_getattr,
> + � � � .getattr_extra �= ext4_getattr_extra,
> �#ifdef CONFIG_EXT4_FS_XATTR
> � � � �.setxattr � � � = generic_setxattr,
> � � � �.getxattr � � � = generic_getxattr,
> @@ -2556,6 +2557,7 @@ const struct inode_operations ext4_dir_inode_operations = {
> �const struct inode_operations ext4_special_inode_operations = {
> � � � �.setattr � � � �= ext4_setattr,
> � � � �.getattr � � � �= ext4_getattr,
> + � � � .getattr_extra �= ext4_getattr_extra,
> �#ifdef CONFIG_EXT4_FS_XATTR
> � � � �.setxattr � � � = generic_setxattr,
> � � � �.getxattr � � � = generic_getxattr,
> diff --git a/fs/ext4/symlink.c b/fs/ext4/symlink.c
> index d8fe7fb..8c206b2 100644
> --- a/fs/ext4/symlink.c
> +++ b/fs/ext4/symlink.c
> @@ -36,6 +36,7 @@ const struct inode_operations ext4_symlink_inode_operations = {
> � � � �.put_link � � � = page_put_link,
> � � � �.setattr � � � �= ext4_setattr,
> � � � �.getattr � � � �= ext4_getattr,
> + � � � .getattr_extra �= ext4_getattr_extra,
> �#ifdef CONFIG_EXT4_FS_XATTR
> � � � �.setxattr � � � = generic_setxattr,
> � � � �.getxattr � � � = generic_getxattr,
> @@ -49,6 +50,7 @@ const struct inode_operations ext4_fast_symlink_inode_operations = {
> � � � �.follow_link � �= ext4_follow_link,
> � � � �.setattr � � � �= ext4_setattr,
> � � � �.getattr � � � �= ext4_getattr,
> + � � � .getattr_extra �= ext4_getattr_extra,
> �#ifdef CONFIG_EXT4_FS_XATTR
> � � � �.setxattr � � � = generic_setxattr,
> � � � �.getxattr � � � = generic_getxattr,
> diff --git a/include/linux/stat.h b/include/linux/stat.h
> index 9e27f88..4c87878 100644
> --- a/include/linux/stat.h
> +++ b/include/linux/stat.h
> @@ -107,7 +107,8 @@ struct xstat_parameters {
> �#define XSTAT_REQUEST_GEN � � � � � � �0x00001000ULL � /* want/got st_gen */
> �#define XSTAT_REQUEST_DATA_VERSION � � 0x00002000ULL � /* want/got st_data_version */
> �#define XSTAT_REQUEST__EXTENDED_STATS �0x00003fffULL � /* the stuff in the xstat struct */
> -#define XSTAT_REQUEST__ALL_STATS � � � 0x00003fffULL � /* the defined set of requestables */
> +#define XSTAT_REQUEST_INODE_FLAGS � � �0x00004000ULL � /* want/got xstat_inode_flags */
> +#define XSTAT_REQUEST__ALL_STATS � � � 0x00007fffULL � /* the defined set of requestables */
> �#define XSTAT_REQUEST__EXTRA_STATS � � (XSTAT_REQUEST__ALL_STATS & ~XSTAT_REQUEST__EXTENDED_STATS)
> �};
>
> @@ -140,6 +141,50 @@ struct xstat {
> � � � �unsigned long long � � �st_extra_results[0]; /* extra requested results */
> �};
>
> +/*
> + * Extra result field for inode flags (XSTAT_REQUEST_INODE_FLAGS)
> + */
> +struct xstat_inode_flags {
> + � � � /* Flags set on the file
> + � � � �* - the LSW matches the BSD st_flags
> + � � � �* - the MSW are Linux-specific
> + � � � �*/
> + � � � unsigned long long � � �st_flags;
> + � � � /* st_flags that users can set */
> +#define UF_SETTABLE � �0x0000ffff
> +#define UF_NODUMP � � �0x00000001 � � �/* do not dump */
> +#define UF_IMMUTABLE � 0x00000002 � � �/* immutable */
> +#define UF_APPEND � � �0x00000004 � � �/* append-only */
> +#define UF_OPAQUE � � �0x00000008 � � �/* directory is opaque (unionfs) */
> +#define UF_NOUNLINK � �0x00000010 � � �/* can't be removed or renamed */
> +#define UF_COMPRESSED �0x00000020 � � �/* file is compressed */
> +#define UF_HIDDEN � � �0x00008000 � � �/* file shouldn't be displayed in a GUI */
> +
> + � � � /* st_flags that only root can set */
> +#define SF_SETTABLE � �0xffff0000
> +#define SF_ARCHIVED � �0x00010000 � � �/* archived */
> +#define SF_IMMUTABLE � 0x00020000 � � �/* immutable */
> +#define SF_APPEND � � �0x00040000 � � �/* append-only */
> +#define SF_NOUNLINK � �0x00100000 � � �/* can't be removed or renamed */
> +#define SF_SNAPSHOT � �0x00200000 � � �/* snapshot inode */
> +
> + � � � /* Linux-specific st_flags */
> +#define XSTAT_LF_MAGIC_FILE � �(1ULL << 32) � �/* magic file, such as /proc/? and /sys/? */
> +#define XSTAT_LF_SYNC � � � � �(1ULL << 33) � �/* file is written synchronously */
> +#define XSTAT_LF_NOATIME � � � (1ULL << 34) � �/* atime is not updated on file */
> +#define XSTAT_LF_JOURNALLED_DATA (1ULL << 35) �/* data modifications to file are journalled */
> +#define XSTAT_LF_ENCRYPTED � � (1ULL << 36) � �/* file is encrypted */
> +#define XSTAT_LF_SYSTEM � � � � � � � �(1ULL << 37) � �/* system file */
> +#define XSTAT_LF_TEMPORARY � � (1ULL << 38) � �/* temporary file */
> +#define XSTAT_LF_OFFLINE � � � (1ULL << 39) � �/* file is currently unavailable */
> +
> + � � � /* Which st_flags are actually supported by this filesystem for this
> + � � � �* file */
> + � � � unsigned long long � � �st_supported_flags;
> +};
> +#define XSTAT_LENGTH_INODE_FLAGS (sizeof(struct xstat_inode_flags))
> +
> +
> �#ifdef __KERNEL__
> �#define S_IRWXUGO � � �(S_IRWXU|S_IRWXG|S_IRWXO)
> �#define S_IALLUGO � � �(S_ISUID|S_ISGID|S_ISVTX|S_IRWXUGO)
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to majordomo(a)vger.kernel.org
> More majordomo info at �http://vger.kernel.org/majordomo-info.html
>



--
Michael Kerrisk Linux man-pages maintainer;
http://www.kernel.org/doc/man-pages/
Author of "The Linux Programming Interface", http://blog.man7.org/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Michael Kerrisk on 4 Jul 2010 00:40
[CC+=linux-api]

On Fri, Jul 2, 2010 at 7:45 PM, Andreas Dilger <adilger(a)dilger.ca> wrote:
> On 2010-07-01, at 17:57, David Howells wrote:
>> [This is, for the moment, to be considered an example. �Do we actually want to
>> export these flags? �Should they be a full member of struct xstat?]
>
> I would say this should be a full-fledged member of struct xstat.
> I think they are fairly standard (available on many filesystems
> today), and requiring an ioctl to access them is unpleasant.
>
>> (1) User settable flags (to be consistent with the BSD st_flags field):
>>
>> � � � UF_NODUMP � � � Do not dump this file.
>> � � � UF_IMMUTABLE � �This file is immutable.
>> � � � UF_APPEND � � � This file is append-only.
>> � � � UF_OPAQUE � � � This directory is opaque (unionfs).
>> � � � UF_NOUNLINK � � This file can't be removed or renamed.
>> � � � UF_COMPRESSED � This file is compressed.
>> � � � UF_HIDDEN � � � This file shouldn't be displayed in a GUI.
>>
>> � �The UF_SETTABLE constant is the union of the above flags.
>>
>> (2) Superuser settable flags (to be consistent with the BSD st_flags field):
>>
>> � � � SF_ARCHIVED � � This file has been archived.
>> � � � SF_IMMUTABLE � �This file is immutable.
>> � � � SF_APPEND � � � This file is append-only.
>> � � � SF_NOUNLINK � � This file can't be removed or renamed.
>> � � � SF_HIDDEN � � � This file is a snapshot inode.
>>
>> � �The SF_SETTABLE constant is the union of the above flags.
>>
>> (3) Linux-specific flags:
>>
>> � � � XSTAT_LF_MAGIC_FILE � � Magic file, such as found in procfs and sysfs.
>> � � � XSTAT_LF_SYNC � � � � � File is written synchronously.
>> � � � XSTAT_LF_NOATIME � � � �Atime is not updated on this file.
>> � � � XSTAT_LF_JOURNALLED_DATA Data modifications to this file are journalled.
>> � � � XSTAT_LF_ENCRYPTED � � �This file is encrypted.
>> � � � XSTAT_LF_SYSTEM � � � � This file is a system file (FAT/NTFS/CIFS).
>> � � � XSTAT_LF_TEMPORARY � � �This file is a temporary file (NTFS/CIFS).
>> � � � XSTAT_LF_OFFLINE � � � �file is currently unavailable (CIFS).
>
> Yuck on the names. �Why not stick with the "UF_" and "SF_" prefixes?
> Since we don't need to keep _binary_ compatibility with these flag values
> (only name portability) we can use the same flag values as the
> FS_*_FL definitions in fs.h. �That is what all of the existing filesystems
> already use (ext2/3/4, ocfs, btrfs, reiserfs, xfs, jfs).

Agree on the naming. Andreas expresses what I intended when I proposed the idea.

Cheers,

Michael


--
Michael Kerrisk Linux man-pages maintainer;
http://www.kernel.org/doc/man-pages/
Author of "The Linux Programming Interface", http://blog.man7.org/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: VIDEO: ivtvfb, remove unneeded NULL test
Next: [PATCH] nilfs2: remove BKL uses