Document the new XFS_IOC_GETFSMAPX that returns the physical layout
of a (disk-based) filesystem.
(Yes, the leading 'X' needs to fall off...)
Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
---
man2/ioctl_getfsmapx.2 | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 253 insertions(+)
create mode 100644 man2/ioctl_getfsmapx.2
diff --git a/man2/ioctl_getfsmapx.2 b/man2/ioctl_getfsmapx.2
new file mode 100644
index 0000000..b79a8e5
--- /dev/null
+++ b/man2/ioctl_getfsmapx.2
@@ -0,0 +1,253 @@
+.\" Copyright (C) 2016 Oracle. All rights reserved.
+.\"
+.\" %%%LICENSE_START(VERBATIM)
+.\" This program is free software; you can redistribute it and/or
+.\" modify it under the terms of the GNU General Public License as
+.\" published by the Free Software Foundation.
+.\"
+.\" This program is distributed in the hope that it would be useful,
+.\" but WITHOUT ANY WARRANTY; without even the implied warranty of
+.\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+.\" GNU General Public License for more details.
+.\"
+.\" You should have received a copy of the GNU General Public License
+.\" along with this program; if not, write the Free Software Foundation,
+.\" Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+.\" %%%LICENSE_END
+.TH IOCTL-XFS_IOC_GETFSMAPX 2 2016-05-05 "Linux" "Linux Programmer's Manual"
+.SH NAME
+ioctl_getfsmapx \- retrieve the physical layout of the filesystem
+.SH SYNOPSIS
+.br
+.B #include <sys/ioctl.h>
+.br
+.B #include <linux/fs.h>
+.sp
+.BI "int ioctl(int " fd ", XFS_IOC_GETFSMAPX, struct getfsmapx * " arg );
+.SH DESCRIPTION
+This
+.BR ioctl (2)
+retrieves physical extent mappings for a filesystem. This information can
+be used to discover which files are mapped to a physical block, examine
+free space, or find known bad blocks, among other things.
+
+The sole argument to this ioctl should be an array of the following
+structure:
+.in +4n
+.nf
+
+struct getfsmapx {
+ __s64 fmv_device; /* device id */
+ __s64 fmv_block; /* starting block */
+ __s64 fmv_owner; /* owner id */
+ __s64 fmv_offset; /* file offset of segment */
+ __s64 fmv_length; /* length of segment, blocks */
+ __s32 fmv_oflags; /* mapping flags */
+ __s32 fmv_iflags; /* control flags (1st structure) */
+ __s32 fmv_count; /* # of entries in array incl. input */
+ __s32 fmv_entries; /* # of entries filled in (output). */
+ __s64 fmv_unused1; /* future use, must be zero */
+};
+
+.fi
+.in
+The array must contain at least two elements. The first two array
+elements specify the lowest and highest reverse-mapping keys, respectively,
+for which userspace would like physical mapping information. A reverse
+mapping key consists of the tuple (device, block, owner, offset). The
+owner and offset fields are part of the key because some filesystems
+support sharing physical blocks between multiple files and therefore may
+return multiple mappings for a given physical block.
+
+.SS Fields of struct getfsmapx
+.PP
+The
+.I fmv_device
+field contains a 64-bit cookie to uniquely identify the underlying storage
+device if the filesystem supports multiple devices. If not, the field
+should be
+.BR FMV_DEV_DEFAULT "."
+
+.PP
+The
+.I fmv_block
+field contains the 512-byte sector address of the extent.
+
+.PP
+The
+.I fmv_owner
+field contains the owner of the extent. This is generally an inode
+number, though if
+.B FMV_OF_SPECIAL_OWNER
+is set in the
+.I fmv_oflags
+field, then the owner value is one of the following special values:
+.TP
+.B FMV_OWN_FREE
+Free space.
+.TP
+.B FMV_OWN_UNKNOWN
+This extent has an unknown owner.
+.TP
+.B FMV_OWN_FS
+Static filesystem metadata.
+.TP
+.B FMV_OWN_LOG
+The filesystem journal.
+.TP
+.B FMV_OWN_AG
+Allocation group metadata.
+.TP
+.B FMV_OWN_INOBT
+The inode index, if one is provided.
+.TP
+.B FMV_OWN_INODES
+Inodes.
+.TP
+.B FMV_OWN_REFC
+Reference counting indexes.
+.TP
+.B FMV_OWN_COW
+This extent is being used to stage a copy-on-write.
+.TP
+.B FMV_OWN_DEFECTIVE:
+This extent has been marked defective either by the filesystem or the
+underlying device.
+
+.PP
+The
+.I fmv_offset
+field contains the logical address of the reverse mapping record, in units
+of 512-byte blocks. This field has no meaning if the
+.BR FMV_OF_SPECIAL_OWNER " or " FMV_OF_EXTENT_MAP
+flags are set in
+.IR fmv_oflags "."
+
+.PP
+The
+.I fmv_length
+field contains the length of the extent, in units of 512-byte blocks.
+This field must be zero in the second array element.
+
+.PP
+The
+.I fmv_oflags
+field is a bitmask of extent state flags. The bits are:
+.TP
+.B FMV_OF_PREALLOC
+The extent is allocated but not yet written.
+.TP
+.B FMV_OF_ATTR_FORK
+This extent contains extended attribute data.
+.TP
+.B FMV_OF_EXTENT_MAP
+This extent contains extent map information for the owner.
+.TP
+.B FMV_OF_SHARED
+Parts of this extent may be shared.
+.TP
+.B FMV_OF_SPECIAL_OWNER
+The
+.I fmv_owner
+field contains a special value instead of an inode number.
+.TP
+.B FMV_OF_LAST
+This is the last record in the filesystem.
+
+.PP
+The
+.I fmv_iflags
+field is a bitmask passed to the kernel to alter the output. There are no
+flags defined, so this value must be zero in the first two array elements.
+
+.PP
+The
+.I fmv_count
+field contains the number of elements in the array being passed to the
+kernel. This count must include the two control elements at the start of
+the array. The value must be specified in the first array element; in the
+second element this field must be zero.
+
+If this value is 2,
+.I fmv_entries
+will be set to the number of records that would have been returned had
+the array been large enough; no extent information will be returned.
+
+.PP
+The
+.I fmv_entries
+field contains the number of elements in the array that contain useful
+information if the ioctl returns a non-error value. This value includes
+the two control elements at the start of the array and is only set in the
+first array element; in the second, this field must be zero.
+
+.PP
+The
+.I fmv_unused1
+field must be zero in the first two array elements.
+
+.SS Array Elements
+.PP
+The key fields (fmv_device, fmv_block, fmv_owner, fmv_offset) of the first
+element of the array specify the lowest extent record in the keyspace that
+the caller wants returned. For example, if the key is set to
+(FMV_DEV_DEFAULT, 36, 0, 0), the filesystem will only return records for
+extents starting at or above sector 36 on disk. For convenience, the
+.I fmv_length
+field will be added to the
+.IR fmv_block " and " fmv_offset
+fields as appropriate so that the (fmv_device, fmv_block, fmv_owner,
+fmv_offset, fmv_length) fields in the last array element can be copied
+into the first element to seed the next ioctl call.
+
+The key fields of the second element of the array specify the highest
+extent record in the keyspace that the caller wants returned. Returning
+to our example above, if that example key were instead passed in via the
+second array element, the filesystem will not return records for extents
+going past sector 36 on disk. For convenience, the four key fields can be
+set to ~0 (all ones) to signify "end of filesystem".
+
+If
+.I fmv_count
+in the first element of the array is 2, then
+.I fmv_entries
+in the first element of the array will be set to the number of extent
+records found in the filesystem. Otherwise,
+.I fmv_entries
+will be set to the number of extents actually returned, and the subsequent
+array elements will be filled out with extent information. In these
+subsequent array elements, the fields
+.IR fmv_iflags ", " fmv_count ", " fmv_entries ", and " fmv_unused1
+will be set to zero by the filesystem.
+
+.SH RETURN VALUE
+On error, \-1 is returned, and
+.I errno
+is set to indicate the error.
+.PP
+.SH ERRORS
+Error codes can be one of, but are not limited to, the following:
+.TP
+.B EINVAL
+The array is not long enough, or a non-zero value was passed in one of the
+fields that must be zero.
+.TP
+.B EFAULT
+The pointer passed in was not mapped to a valid memory address.
+.TP
+.B EBADF
+.IR fd
+is not open for reading.
+.TP
+.B EPERM
+This query is not allowed.
+.TP
+.B EOPNOTSUPP
+The filesystem does not support this command.
+
+.SH CONFORMING TO
+This API is Linux-specific. Not all filesystems support it.
+.fi
+.in
+.SH SEE ALSO
+.BR ioctl (2)
_______________________________________________
xfs mailing list
xfs@xxxxxxxxxxx
http://oss.sgi.com/mailman/listinfo/xfs
