Commit e89c0413 authored by Darrick J. Wong's avatar Darrick J. Wong

xfs: implement the GETFSMAP ioctl

Introduce a new ioctl that uses the reverse mapping btree to return
information about the physical layout of the filesystem.
Signed-off-by: default avatarDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: default avatarBrian Foster <bfoster@redhat.com>
parent fb3c3de2
...@@ -79,6 +79,7 @@ xfs-y += xfs_aops.o \ ...@@ -79,6 +79,7 @@ xfs-y += xfs_aops.o \
xfs_extent_busy.o \ xfs_extent_busy.o \
xfs_file.o \ xfs_file.o \
xfs_filestream.o \ xfs_filestream.o \
xfs_fsmap.o \
xfs_fsops.o \ xfs_fsops.o \
xfs_globals.o \ xfs_globals.o \
xfs_icache.o \ xfs_icache.o \
......
...@@ -92,6 +92,18 @@ struct getbmapx { ...@@ -92,6 +92,18 @@ struct getbmapx {
#define BMV_OF_LAST 0x4 /* segment is the last in the file */ #define BMV_OF_LAST 0x4 /* segment is the last in the file */
#define BMV_OF_SHARED 0x8 /* segment shared with another file */ #define BMV_OF_SHARED 0x8 /* segment shared with another file */
/* fmr_owner special values for FS_IOC_GETFSMAP */
#define XFS_FMR_OWN_FREE FMR_OWN_FREE /* free space */
#define XFS_FMR_OWN_UNKNOWN FMR_OWN_UNKNOWN /* unknown owner */
#define XFS_FMR_OWN_FS FMR_OWNER('X', 1) /* static fs metadata */
#define XFS_FMR_OWN_LOG FMR_OWNER('X', 2) /* journalling log */
#define XFS_FMR_OWN_AG FMR_OWNER('X', 3) /* per-AG metadata */
#define XFS_FMR_OWN_INOBT FMR_OWNER('X', 4) /* inode btree blocks */
#define XFS_FMR_OWN_INODES FMR_OWNER('X', 5) /* inodes */
#define XFS_FMR_OWN_REFC FMR_OWNER('X', 6) /* refcount tree */
#define XFS_FMR_OWN_COW FMR_OWNER('X', 7) /* cow staging */
#define XFS_FMR_OWN_DEFECTIVE FMR_OWNER('X', 8) /* bad blocks */
/* /*
* Structure for XFS_IOC_FSSETDM. * Structure for XFS_IOC_FSSETDM.
* For use by backup and restore programs to set the XFS on-disk inode * For use by backup and restore programs to set the XFS on-disk inode
...@@ -502,6 +514,7 @@ typedef struct xfs_swapext ...@@ -502,6 +514,7 @@ typedef struct xfs_swapext
#define XFS_IOC_GETBMAPX _IOWR('X', 56, struct getbmap) #define XFS_IOC_GETBMAPX _IOWR('X', 56, struct getbmap)
#define XFS_IOC_ZERO_RANGE _IOW ('X', 57, struct xfs_flock64) #define XFS_IOC_ZERO_RANGE _IOW ('X', 57, struct xfs_flock64)
#define XFS_IOC_FREE_EOFBLOCKS _IOR ('X', 58, struct xfs_fs_eofblocks) #define XFS_IOC_FREE_EOFBLOCKS _IOR ('X', 58, struct xfs_fs_eofblocks)
/* XFS_IOC_GETFSMAP ------ hoisted 59 */
/* /*
* ioctl commands that replace IRIX syssgi()'s * ioctl commands that replace IRIX syssgi()'s
......
...@@ -2305,3 +2305,31 @@ xfs_rmap_free_extent( ...@@ -2305,3 +2305,31 @@ xfs_rmap_free_extent(
return __xfs_rmap_add(mp, dfops, XFS_RMAP_FREE, owner, return __xfs_rmap_add(mp, dfops, XFS_RMAP_FREE, owner,
XFS_DATA_FORK, &bmap); XFS_DATA_FORK, &bmap);
} }
/* Compare rmap records. Returns -1 if a < b, 1 if a > b, and 0 if equal. */
int
xfs_rmap_compare(
const struct xfs_rmap_irec *a,
const struct xfs_rmap_irec *b)
{
__u64 oa;
__u64 ob;
oa = xfs_rmap_irec_offset_pack(a);
ob = xfs_rmap_irec_offset_pack(b);
if (a->rm_startblock < b->rm_startblock)
return -1;
else if (a->rm_startblock > b->rm_startblock)
return 1;
else if (a->rm_owner < b->rm_owner)
return -1;
else if (a->rm_owner > b->rm_owner)
return 1;
else if (oa < ob)
return -1;
else if (oa > ob)
return 1;
else
return 0;
}
...@@ -214,5 +214,7 @@ int xfs_rmap_find_left_neighbor(struct xfs_btree_cur *cur, xfs_agblock_t bno, ...@@ -214,5 +214,7 @@ int xfs_rmap_find_left_neighbor(struct xfs_btree_cur *cur, xfs_agblock_t bno,
int xfs_rmap_lookup_le_range(struct xfs_btree_cur *cur, xfs_agblock_t bno, int xfs_rmap_lookup_le_range(struct xfs_btree_cur *cur, xfs_agblock_t bno,
uint64_t owner, uint64_t offset, unsigned int flags, uint64_t owner, uint64_t offset, unsigned int flags,
struct xfs_rmap_irec *irec, int *stat); struct xfs_rmap_irec *irec, int *stat);
int xfs_rmap_compare(const struct xfs_rmap_irec *a,
const struct xfs_rmap_irec *b);
#endif /* __XFS_RMAP_H__ */ #endif /* __XFS_RMAP_H__ */
This diff is collapsed.
/*
* Copyright (C) 2017 Oracle. All Rights Reserved.
*
* Author: Darrick J. Wong <darrick.wong@oracle.com>
*
* 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; either version 2
* of the License, or (at your option) any later version.
*
* 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.
*/
#ifndef __XFS_FSMAP_H__
#define __XFS_FSMAP_H__
struct fsmap;
/* internal fsmap representation */
struct xfs_fsmap {
dev_t fmr_device; /* device id */
uint32_t fmr_flags; /* mapping flags */
uint64_t fmr_physical; /* device offset of segment */
uint64_t fmr_owner; /* owner id */
xfs_fileoff_t fmr_offset; /* file offset of segment */
xfs_filblks_t fmr_length; /* length of segment, blocks */
};
struct xfs_fsmap_head {
uint32_t fmh_iflags; /* control flags */
uint32_t fmh_oflags; /* output flags */
unsigned int fmh_count; /* # of entries in array incl. input */
unsigned int fmh_entries; /* # of entries filled in (output). */
struct xfs_fsmap fmh_keys[2]; /* low and high keys */
};
void xfs_fsmap_from_internal(struct fsmap *dest, struct xfs_fsmap *src);
void xfs_fsmap_to_internal(struct xfs_fsmap *dest, struct fsmap *src);
/* fsmap to userspace formatter - copy to user & advance pointer */
typedef int (*xfs_fsmap_format_t)(struct xfs_fsmap *, void *);
int xfs_getfsmap(struct xfs_mount *mp, struct xfs_fsmap_head *head,
xfs_fsmap_format_t formatter, void *arg);
#endif /* __XFS_FSMAP_H__ */
...@@ -41,6 +41,9 @@ ...@@ -41,6 +41,9 @@
#include "xfs_trans.h" #include "xfs_trans.h"
#include "xfs_pnfs.h" #include "xfs_pnfs.h"
#include "xfs_acl.h" #include "xfs_acl.h"
#include "xfs_btree.h"
#include <linux/fsmap.h>
#include "xfs_fsmap.h"
#include <linux/capability.h> #include <linux/capability.h>
#include <linux/cred.h> #include <linux/cred.h>
...@@ -1609,6 +1612,84 @@ xfs_ioc_getbmapx( ...@@ -1609,6 +1612,84 @@ xfs_ioc_getbmapx(
return 0; return 0;
} }
struct getfsmap_info {
struct xfs_mount *mp;
struct fsmap __user *data;
__u32 last_flags;
};
STATIC int
xfs_getfsmap_format(struct xfs_fsmap *xfm, void *priv)
{
struct getfsmap_info *info = priv;
struct fsmap fm;
trace_xfs_getfsmap_mapping(info->mp, xfm);
info->last_flags = xfm->fmr_flags;
xfs_fsmap_from_internal(&fm, xfm);
if (copy_to_user(info->data, &fm, sizeof(struct fsmap)))
return -EFAULT;
info->data++;
return 0;
}
STATIC int
xfs_ioc_getfsmap(
struct xfs_inode *ip,
void __user *arg)
{
struct getfsmap_info info = {0};
struct xfs_fsmap_head xhead = {0};
struct fsmap_head head;
bool aborted = false;
int error;
if (copy_from_user(&head, arg, sizeof(struct fsmap_head)))
return -EFAULT;
if (memchr_inv(head.fmh_reserved, 0, sizeof(head.fmh_reserved)) ||
memchr_inv(head.fmh_keys[0].fmr_reserved, 0,
sizeof(head.fmh_keys[0].fmr_reserved)) ||
memchr_inv(head.fmh_keys[1].fmr_reserved, 0,
sizeof(head.fmh_keys[1].fmr_reserved)))
return -EINVAL;
xhead.fmh_iflags = head.fmh_iflags;
xhead.fmh_count = head.fmh_count;
xfs_fsmap_to_internal(&xhead.fmh_keys[0], &head.fmh_keys[0]);
xfs_fsmap_to_internal(&xhead.fmh_keys[1], &head.fmh_keys[1]);
trace_xfs_getfsmap_low_key(ip->i_mount, &xhead.fmh_keys[0]);
trace_xfs_getfsmap_high_key(ip->i_mount, &xhead.fmh_keys[1]);
info.mp = ip->i_mount;
info.data = ((__force struct fsmap_head *)arg)->fmh_recs;
error = xfs_getfsmap(ip->i_mount, &xhead, xfs_getfsmap_format, &info);
if (error == XFS_BTREE_QUERY_RANGE_ABORT) {
error = 0;
aborted = true;
} else if (error)
return error;
/* If we didn't abort, set the "last" flag in the last fmx */
if (!aborted && xhead.fmh_entries) {
info.data--;
info.last_flags |= FMR_OF_LAST;
if (copy_to_user(&info.data->fmr_flags, &info.last_flags,
sizeof(info.last_flags)))
return -EFAULT;
}
/* copy back header */
head.fmh_entries = xhead.fmh_entries;
head.fmh_oflags = xhead.fmh_oflags;
if (copy_to_user(arg, &head, sizeof(struct fsmap_head)))
return -EFAULT;
return 0;
}
int int
xfs_ioc_swapext( xfs_ioc_swapext(
xfs_swapext_t *sxp) xfs_swapext_t *sxp)
...@@ -1789,6 +1870,9 @@ xfs_file_ioctl( ...@@ -1789,6 +1870,9 @@ xfs_file_ioctl(
case XFS_IOC_GETBMAPX: case XFS_IOC_GETBMAPX:
return xfs_ioc_getbmapx(ip, arg); return xfs_ioc_getbmapx(ip, arg);
case FS_IOC_GETFSMAP:
return xfs_ioc_getfsmap(ip, arg);
case XFS_IOC_FD_TO_HANDLE: case XFS_IOC_FD_TO_HANDLE:
case XFS_IOC_PATH_TO_HANDLE: case XFS_IOC_PATH_TO_HANDLE:
case XFS_IOC_PATH_TO_FSHANDLE: { case XFS_IOC_PATH_TO_FSHANDLE: {
......
...@@ -20,6 +20,7 @@ ...@@ -20,6 +20,7 @@
#include <linux/mount.h> #include <linux/mount.h>
#include <linux/slab.h> #include <linux/slab.h>
#include <linux/uaccess.h> #include <linux/uaccess.h>
#include <linux/fsmap.h>
#include "xfs.h" #include "xfs.h"
#include "xfs_fs.h" #include "xfs_fs.h"
#include "xfs_format.h" #include "xfs_format.h"
...@@ -554,6 +555,7 @@ xfs_file_compat_ioctl( ...@@ -554,6 +555,7 @@ xfs_file_compat_ioctl(
case XFS_IOC_GOINGDOWN: case XFS_IOC_GOINGDOWN:
case XFS_IOC_ERROR_INJECTION: case XFS_IOC_ERROR_INJECTION:
case XFS_IOC_ERROR_CLEARALL: case XFS_IOC_ERROR_CLEARALL:
case FS_IOC_GETFSMAP:
return xfs_file_ioctl(filp, cmd, p); return xfs_file_ioctl(filp, cmd, p);
#ifndef BROKEN_X86_ALIGNMENT #ifndef BROKEN_X86_ALIGNMENT
/* These are handled fine if no alignment issues */ /* These are handled fine if no alignment issues */
......
...@@ -47,6 +47,7 @@ ...@@ -47,6 +47,7 @@
#include "xfs_inode_item.h" #include "xfs_inode_item.h"
#include "xfs_bmap_btree.h" #include "xfs_bmap_btree.h"
#include "xfs_filestream.h" #include "xfs_filestream.h"
#include "xfs_fsmap.h"
/* /*
* We include this last to have the helpers above available for the trace * We include this last to have the helpers above available for the trace
......
...@@ -40,6 +40,8 @@ struct xfs_inode_log_format; ...@@ -40,6 +40,8 @@ struct xfs_inode_log_format;
struct xfs_bmbt_irec; struct xfs_bmbt_irec;
struct xfs_btree_cur; struct xfs_btree_cur;
struct xfs_refcount_irec; struct xfs_refcount_irec;
struct xfs_fsmap;
struct xfs_rmap_irec;
DECLARE_EVENT_CLASS(xfs_attr_list_class, DECLARE_EVENT_CLASS(xfs_attr_list_class,
TP_PROTO(struct xfs_attr_list_context *ctx), TP_PROTO(struct xfs_attr_list_context *ctx),
...@@ -3267,6 +3269,88 @@ DEFINE_INODE_IREC_EVENT(xfs_swap_extent_rmap_remap); ...@@ -3267,6 +3269,88 @@ DEFINE_INODE_IREC_EVENT(xfs_swap_extent_rmap_remap);
DEFINE_INODE_IREC_EVENT(xfs_swap_extent_rmap_remap_piece); DEFINE_INODE_IREC_EVENT(xfs_swap_extent_rmap_remap_piece);
DEFINE_INODE_ERROR_EVENT(xfs_swap_extent_rmap_error); DEFINE_INODE_ERROR_EVENT(xfs_swap_extent_rmap_error);
/* fsmap traces */
DECLARE_EVENT_CLASS(xfs_fsmap_class,
TP_PROTO(struct xfs_mount *mp, u32 keydev, xfs_agnumber_t agno,
struct xfs_rmap_irec *rmap),
TP_ARGS(mp, keydev, agno, rmap),
TP_STRUCT__entry(
__field(dev_t, dev)
__field(dev_t, keydev)
__field(xfs_agnumber_t, agno)
__field(xfs_fsblock_t, bno)
__field(xfs_filblks_t, len)
__field(__uint64_t, owner)
__field(__uint64_t, offset)
__field(unsigned int, flags)
),
TP_fast_assign(
__entry->dev = mp->m_super->s_dev;
__entry->keydev = new_decode_dev(keydev);
__entry->agno = agno;
__entry->bno = rmap->rm_startblock;
__entry->len = rmap->rm_blockcount;
__entry->owner = rmap->rm_owner;
__entry->offset = rmap->rm_offset;
__entry->flags = rmap->rm_flags;
),
TP_printk("dev %d:%d keydev %d:%d agno %u bno %llu len %llu owner %lld offset %llu flags 0x%x\n",
MAJOR(__entry->dev), MINOR(__entry->dev),
MAJOR(__entry->keydev), MINOR(__entry->keydev),
__entry->agno,
__entry->bno,
__entry->len,
__entry->owner,
__entry->offset,
__entry->flags)
)
#define DEFINE_FSMAP_EVENT(name) \
DEFINE_EVENT(xfs_fsmap_class, name, \
TP_PROTO(struct xfs_mount *mp, u32 keydev, xfs_agnumber_t agno, \
struct xfs_rmap_irec *rmap), \
TP_ARGS(mp, keydev, agno, rmap))
DEFINE_FSMAP_EVENT(xfs_fsmap_low_key);
DEFINE_FSMAP_EVENT(xfs_fsmap_high_key);
DEFINE_FSMAP_EVENT(xfs_fsmap_mapping);
DECLARE_EVENT_CLASS(xfs_getfsmap_class,
TP_PROTO(struct xfs_mount *mp, struct xfs_fsmap *fsmap),
TP_ARGS(mp, fsmap),
TP_STRUCT__entry(
__field(dev_t, dev)
__field(dev_t, keydev)
__field(xfs_daddr_t, block)
__field(xfs_daddr_t, len)
__field(__uint64_t, owner)
__field(__uint64_t, offset)
__field(__uint64_t, flags)
),
TP_fast_assign(
__entry->dev = mp->m_super->s_dev;
__entry->keydev = new_decode_dev(fsmap->fmr_device);
__entry->block = fsmap->fmr_physical;
__entry->len = fsmap->fmr_length;
__entry->owner = fsmap->fmr_owner;
__entry->offset = fsmap->fmr_offset;
__entry->flags = fsmap->fmr_flags;
),
TP_printk("dev %d:%d keydev %d:%d block %llu len %llu owner %lld offset %llu flags 0x%llx\n",
MAJOR(__entry->dev), MINOR(__entry->dev),
MAJOR(__entry->keydev), MINOR(__entry->keydev),
__entry->block,
__entry->len,
__entry->owner,
__entry->offset,
__entry->flags)
)
#define DEFINE_GETFSMAP_EVENT(name) \
DEFINE_EVENT(xfs_getfsmap_class, name, \
TP_PROTO(struct xfs_mount *mp, struct xfs_fsmap *fsmap), \
TP_ARGS(mp, fsmap))
DEFINE_GETFSMAP_EVENT(xfs_getfsmap_low_key);
DEFINE_GETFSMAP_EVENT(xfs_getfsmap_high_key);
DEFINE_GETFSMAP_EVENT(xfs_getfsmap_mapping);
#endif /* _TRACE_XFS_H */ #endif /* _TRACE_XFS_H */
#undef TRACE_INCLUDE_PATH #undef TRACE_INCLUDE_PATH
......
...@@ -262,6 +262,28 @@ xfs_trans_alloc( ...@@ -262,6 +262,28 @@ xfs_trans_alloc(
return 0; return 0;
} }
/*
* Create an empty transaction with no reservation. This is a defensive
* mechanism for routines that query metadata without actually modifying
* them -- if the metadata being queried is somehow cross-linked (think a
* btree block pointer that points higher in the tree), we risk deadlock.
* However, blocks grabbed as part of a transaction can be re-grabbed.
* The verifiers will notice the corrupt block and the operation will fail
* back to userspace without deadlocking.
*
* Note the zero-length reservation; this transaction MUST be cancelled
* without any dirty data.
*/
int
xfs_trans_alloc_empty(
struct xfs_mount *mp,
struct xfs_trans **tpp)
{
struct xfs_trans_res resv = {0};
return xfs_trans_alloc(mp, &resv, 0, 0, XFS_TRANS_NO_WRITECOUNT, tpp);
}
/* /*
* Record the indicated change to the given field for application * Record the indicated change to the given field for application
* to the file system's superblock when the transaction commits. * to the file system's superblock when the transaction commits.
......
...@@ -158,6 +158,8 @@ typedef struct xfs_trans { ...@@ -158,6 +158,8 @@ typedef struct xfs_trans {
int xfs_trans_alloc(struct xfs_mount *mp, struct xfs_trans_res *resp, int xfs_trans_alloc(struct xfs_mount *mp, struct xfs_trans_res *resp,
uint blocks, uint rtextents, uint flags, uint blocks, uint rtextents, uint flags,
struct xfs_trans **tpp); struct xfs_trans **tpp);
int xfs_trans_alloc_empty(struct xfs_mount *mp,
struct xfs_trans **tpp);
void xfs_trans_mod_sb(xfs_trans_t *, uint, int64_t); void xfs_trans_mod_sb(xfs_trans_t *, uint, int64_t);
struct xfs_buf *xfs_trans_get_buf_map(struct xfs_trans *tp, struct xfs_buf *xfs_trans_get_buf_map(struct xfs_trans *tp,
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment