#include "xfs_trans_priv.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
-#include "xfs_dir.h"
#include "xfs_dir2.h"
#include "xfs_dmapi.h"
#include "xfs_mount.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
-#include "xfs_dir_sf.h"
#include "xfs_dir2_sf.h"
#include "xfs_attr_sf.h"
#include "xfs_dinode.h"
#include "xfs_utils.h"
#include "xfs_dir2_trace.h"
#include "xfs_quota.h"
-#include "xfs_mac.h"
#include "xfs_acl.h"
xfs_daddr_t bno,
uint imap_flags)
{
+ xfs_imap_t imap;
xfs_buf_t *bp;
int error;
- xfs_imap_t imap;
-#ifdef __KERNEL__
int i;
int ni;
-#endif
if (ip->i_blkno == (xfs_daddr_t)0) {
/*
*/
error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, imap.im_blkno,
(int)imap.im_len, XFS_BUF_LOCK, &bp);
-
if (error) {
#ifdef DEBUG
xfs_fs_cmn_err(CE_ALERT, mp, "xfs_itobp: "
#endif /* DEBUG */
return error;
}
-#ifdef __KERNEL__
+
/*
* Validate the magic number and version of every inode in the buffer
* (if DEBUG kernel) or the first inode in the buffer, otherwise.
+ * No validation is done here in userspace (xfs_repair).
*/
-#ifdef DEBUG
- ni = (imap_flags & XFS_IMAP_BULKSTAT) ? 0 :
- (BBTOB(imap.im_len) >> mp->m_sb.sb_inodelog);
-#else
- ni = (imap_flags & XFS_IMAP_BULKSTAT) ? 0 : 1;
+#if !defined(__KERNEL__)
+ ni = 0;
+#elif defined(DEBUG)
+ ni = BBTOB(imap.im_len) >> mp->m_sb.sb_inodelog;
+#else /* usual case */
+ ni = 1;
#endif
+
for (i = 0; i < ni; i++) {
int di_ok;
xfs_dinode_t *dip;
(i << mp->m_sb.sb_inodelog));
di_ok = INT_GET(dip->di_core.di_magic, ARCH_CONVERT) == XFS_DINODE_MAGIC &&
XFS_DINODE_GOOD_VERSION(INT_GET(dip->di_core.di_version, ARCH_CONVERT));
- if (unlikely(XFS_TEST_ERROR(!di_ok, mp, XFS_ERRTAG_ITOBP_INOTOBP,
- XFS_RANDOM_ITOBP_INOTOBP))) {
+ if (unlikely(XFS_TEST_ERROR(!di_ok, mp,
+ XFS_ERRTAG_ITOBP_INOTOBP,
+ XFS_RANDOM_ITOBP_INOTOBP))) {
+ if (imap_flags & XFS_IMAP_BULKSTAT) {
+ xfs_trans_brelse(tp, bp);
+ return XFS_ERROR(EINVAL);
+ }
#ifdef DEBUG
- prdev("bad inode magic/vsn daddr %lld #%d (magic=%x)",
- mp->m_ddev_targp,
+ cmn_err(CE_ALERT,
+ "Device %s - bad inode magic/vsn "
+ "daddr %lld #%d (magic=%x)",
+ XFS_BUFTARG_NAME(mp->m_ddev_targp),
(unsigned long long)imap.im_blkno, i,
INT_GET(dip->di_core.di_magic, ARCH_CONVERT));
#endif
return XFS_ERROR(EFSCORRUPTED);
}
}
-#endif /* __KERNEL__ */
xfs_inobp_check(mp, bp);
return XFS_ERROR(EFSCORRUPTED);
}
ip->i_d.di_size = 0;
+ ip->i_size = 0;
ip->i_df.if_u2.if_rdev = INT_GET(dip->di_u.di_dev, ARCH_CONVERT);
break;
STATIC uint
_xfs_dic2xflags(
- xfs_dinode_core_t *dic,
__uint16_t di_flags)
{
uint flags = 0;
flags |= XFS_XFLAG_EXTSIZE;
if (di_flags & XFS_DIFLAG_EXTSZINHERIT)
flags |= XFS_XFLAG_EXTSZINHERIT;
+ if (di_flags & XFS_DIFLAG_NODEFRAG)
+ flags |= XFS_XFLAG_NODEFRAG;
}
return flags;
{
xfs_dinode_core_t *dic = &ip->i_d;
- return _xfs_dic2xflags(dic, dic->di_flags) |
- (XFS_CFORK_Q(dic) ? XFS_XFLAG_HASATTR : 0);
+ return _xfs_dic2xflags(dic->di_flags) |
+ (XFS_CFORK_Q(dic) ? XFS_XFLAG_HASATTR : 0);
}
uint
xfs_dic2xflags(
xfs_dinode_core_t *dic)
{
- return _xfs_dic2xflags(dic, INT_GET(dic->di_flags, ARCH_CONVERT)) |
- (XFS_CFORK_Q_DISK(dic) ? XFS_XFLAG_HASATTR : 0);
+ return _xfs_dic2xflags(INT_GET(dic->di_flags, ARCH_CONVERT)) |
+ (XFS_CFORK_Q_DISK(dic) ? XFS_XFLAG_HASATTR : 0);
}
/*
xfs_trans_t *tp,
xfs_ino_t ino,
xfs_inode_t **ipp,
- xfs_daddr_t bno)
+ xfs_daddr_t bno,
+ uint imap_flags)
{
xfs_buf_t *bp;
xfs_dinode_t *dip;
ip = kmem_zone_zalloc(xfs_inode_zone, KM_SLEEP);
ip->i_ino = ino;
ip->i_mount = mp;
+ spin_lock_init(&ip->i_flags_lock);
/*
* Get pointer's to the on-disk inode and the buffer containing it.
* return NULL as well. Set i_blkno to 0 so that xfs_itobp() will
* know that this is a new incore inode.
*/
- error = xfs_itobp(mp, tp, ip, &dip, &bp, bno, 0);
+ error = xfs_itobp(mp, tp, ip, &dip, &bp, bno, imap_flags);
if (error) {
kmem_zone_free(xfs_inode_zone, ip);
return error;
}
ip->i_delayed_blks = 0;
+ ip->i_size = ip->i_d.di_size;
/*
* Mark the buffer containing the inode as something to keep
{
xfs_ino_t ino;
xfs_inode_t *ip;
- vnode_t *vp;
+ bhv_vnode_t *vp;
uint flags;
int error;
* to prevent others from looking at until we're done.
*/
error = xfs_trans_iget(tp->t_mountp, tp, ino,
- IGET_CREATE, XFS_ILOCK_EXCL, &ip);
+ XFS_IGET_CREATE, XFS_ILOCK_EXCL, &ip);
if (error != 0) {
return error;
}
}
ip->i_d.di_size = 0;
+ ip->i_size = 0;
ip->i_d.di_nextents = 0;
ASSERT(ip->i_d.di_nblocks == 0);
xfs_ichgtime(ip, XFS_ICHGTIME_CHG|XFS_ICHGTIME_ACC|XFS_ICHGTIME_MOD);
di_flags |= XFS_DIFLAG_NOSYMLINKS;
if (pip->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
di_flags |= XFS_DIFLAG_PROJINHERIT;
+ if ((pip->i_d.di_flags & XFS_DIFLAG_NODEFRAG) &&
+ xfs_inherit_nodefrag)
+ di_flags |= XFS_DIFLAG_NODEFRAG;
ip->i_d.di_flags |= di_flags;
}
/* FALLTHROUGH */
*/
xfs_trans_log_inode(tp, ip, flags);
- /* now that we have an i_mode we can set Linux inode ops (& unlock) */
- VFS_INIT_VNODE(XFS_MTOVFS(tp->t_mountp), vp, XFS_ITOBHV(ip), 1);
+ /* now that we have an i_mode we can setup inode ops and unlock */
+ bhv_vfs_init_vnode(XFS_MTOVFS(tp->t_mountp), vp, XFS_ITOBHV(ip), 1);
*ipp = ip;
return 0;
} else {
last_block = 0;
}
- size_last_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)ip->i_d.di_size);
+ size_last_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)ip->i_size);
last_block = XFS_FILEOFF_MAX(last_block, size_last_block);
last_byte = XFS_FSB_TO_B(mp, last_block);
* must be called again with all the same restrictions as the initial
* call.
*/
-void
+int
xfs_itruncate_start(
xfs_inode_t *ip,
uint flags,
xfs_fsize_t last_byte;
xfs_off_t toss_start;
xfs_mount_t *mp;
- vnode_t *vp;
+ bhv_vnode_t *vp;
+ int error = 0;
ASSERT(ismrlocked(&ip->i_iolock, MR_UPDATE) != 0);
- ASSERT((new_size == 0) || (new_size <= ip->i_d.di_size));
+ ASSERT((new_size == 0) || (new_size <= ip->i_size));
ASSERT((flags == XFS_ITRUNC_DEFINITE) ||
(flags == XFS_ITRUNC_MAYBE));
vn_iowait(vp); /* wait for the completion of any pending DIOs */
/*
- * Call VOP_TOSS_PAGES() or VOP_FLUSHINVAL_PAGES() to get rid of pages and buffers
+ * Call toss_pages or flushinval_pages to get rid of pages
* overlapping the region being removed. We have to use
- * the less efficient VOP_FLUSHINVAL_PAGES() in the case that the
+ * the less efficient flushinval_pages in the case that the
* caller may not be able to finish the truncate without
* dropping the inode's I/O lock. Make sure
* to catch any pages brought in by buffers overlapping
* so that we don't toss things on the same block as
* new_size but before it.
*
- * Before calling VOP_TOSS_PAGES() or VOP_FLUSHINVAL_PAGES(), make sure to
+ * Before calling toss_page or flushinval_pages, make sure to
* call remapf() over the same region if the file is mapped.
* This frees up mapped file references to the pages in the
- * given range and for the VOP_FLUSHINVAL_PAGES() case it ensures
+ * given range and for the flushinval_pages case it ensures
* that we get the latest mapped changes flushed out.
*/
toss_start = XFS_B_TO_FSB(mp, (xfs_ufsize_t)new_size);
* file size, so there is no way that the data extended
* out there.
*/
- return;
+ return 0;
}
last_byte = xfs_file_last_byte(ip);
xfs_itrunc_trace(XFS_ITRUNC_START, ip, flags, new_size, toss_start,
last_byte);
if (last_byte > toss_start) {
if (flags & XFS_ITRUNC_DEFINITE) {
- VOP_TOSS_PAGES(vp, toss_start, -1, FI_REMAPF_LOCKED);
+ bhv_vop_toss_pages(vp, toss_start, -1, FI_REMAPF_LOCKED);
} else {
- VOP_FLUSHINVAL_PAGES(vp, toss_start, -1, FI_REMAPF_LOCKED);
+ error = bhv_vop_flushinval_pages(vp, toss_start, -1, FI_REMAPF_LOCKED);
}
}
ASSERT(VN_CACHED(vp) == 0);
}
#endif
+ return error;
}
/*
ASSERT(ismrlocked(&ip->i_iolock, MR_UPDATE) != 0);
ASSERT(ismrlocked(&ip->i_lock, MR_UPDATE) != 0);
- ASSERT((new_size == 0) || (new_size <= ip->i_d.di_size));
+ ASSERT((new_size == 0) || (new_size <= ip->i_size));
ASSERT(*tp != NULL);
ASSERT((*tp)->t_flags & XFS_TRANS_PERM_LOG_RES);
ASSERT(ip->i_transp == *tp);
*/
if (fork == XFS_DATA_FORK) {
if (ip->i_d.di_nextents > 0) {
- ip->i_d.di_size = new_size;
- xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
+ /*
+ * If we are not changing the file size then do
+ * not update the on-disk file size - we may be
+ * called from xfs_inactive_free_eofblocks(). If we
+ * update the on-disk file size and then the system
+ * crashes before the contents of the file are
+ * flushed to disk then the files may be full of
+ * holes (ie NULL files bug).
+ */
+ if (ip->i_size != new_size) {
+ ip->i_d.di_size = new_size;
+ ip->i_size = new_size;
+ xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
+ }
}
} else if (sync) {
ASSERT(!(mp->m_flags & XFS_MOUNT_WSYNC));
* Duplicate the transaction that has the permanent
* reservation and commit the old transaction.
*/
- error = xfs_bmap_finish(tp, &free_list, first_block,
- &committed);
+ error = xfs_bmap_finish(tp, &free_list, &committed);
ntp = *tp;
if (error) {
/*
xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
}
ntp = xfs_trans_dup(ntp);
- (void) xfs_trans_commit(*tp, 0, NULL);
+ (void) xfs_trans_commit(*tp, 0);
*tp = ntp;
error = xfs_trans_reserve(ntp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES,
*/
if (fork == XFS_DATA_FORK) {
xfs_isize_check(mp, ip, new_size);
- ip->i_d.di_size = new_size;
+ /*
+ * If we are not changing the file size then do
+ * not update the on-disk file size - we may be
+ * called from xfs_inactive_free_eofblocks(). If we
+ * update the on-disk file size and then the system
+ * crashes before the contents of the file are
+ * flushed to disk then the files may be full of
+ * holes (ie NULL files bug).
+ */
+ if (ip->i_size != new_size) {
+ ip->i_d.di_size = new_size;
+ ip->i_size = new_size;
+ }
}
xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
ASSERT((new_size != 0) ||
ASSERT(ismrlocked(&(ip->i_lock), MR_UPDATE) != 0);
ASSERT(ismrlocked(&(ip->i_iolock), MR_UPDATE) != 0);
- ASSERT(new_size > ip->i_d.di_size);
+ ASSERT(new_size > ip->i_size);
/*
* Zero any pages that may have been created by
* and any blocks between the old and new file sizes.
*/
error = xfs_zero_eof(XFS_ITOV(ip), &ip->i_iocore, new_size,
- ip->i_d.di_size, new_size);
+ ip->i_size);
return error;
}
ASSERT(ismrlocked(&(ip->i_lock), MR_UPDATE) != 0);
ASSERT(ismrlocked(&(ip->i_iolock), MR_UPDATE) != 0);
ASSERT(ip->i_transp == tp);
- ASSERT(new_size > ip->i_d.di_size);
+ ASSERT(new_size > ip->i_size);
/*
* Update the file size. Update the inode change timestamp
* if change_flag set.
*/
ip->i_d.di_size = new_size;
+ ip->i_size = new_size;
if (change_flag)
xfs_ichgtime(ip, XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
xfs_agino_t agino;
xfs_agino_t next_agino;
xfs_buf_t *last_ibp;
- xfs_dinode_t *last_dip;
+ xfs_dinode_t *last_dip = NULL;
short bucket_index;
- int offset, last_offset;
+ int offset, last_offset = 0;
int error;
int agi_ok;
return 0;
}
-static __inline__ int xfs_inode_clean(xfs_inode_t *ip)
+STATIC_INLINE int xfs_inode_clean(xfs_inode_t *ip)
{
return (((ip->i_itemp == NULL) ||
!(ip->i_itemp->ili_format.ilf_fields & XFS_ILOG_ALL)) &&
/* Inode not in memory or we found it already,
* nothing to do
*/
- if (!ip || (ip->i_flags & XFS_ISTALE)) {
+ if (!ip || xfs_iflags_test(ip, XFS_ISTALE)) {
read_unlock(&ih->ih_lock);
continue;
}
if (ip == free_ip) {
if (xfs_iflock_nowait(ip)) {
- ip->i_flags |= XFS_ISTALE;
-
+ xfs_iflags_set(ip, XFS_ISTALE);
if (xfs_inode_clean(ip)) {
xfs_ifunlock(ip);
} else {
if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
if (xfs_iflock_nowait(ip)) {
- ip->i_flags |= XFS_ISTALE;
+ xfs_iflags_set(ip, XFS_ISTALE);
if (xfs_inode_clean(ip)) {
xfs_ifunlock(ip);
AIL_LOCK(mp,s);
iip->ili_flush_lsn = iip->ili_item.li_lsn;
AIL_UNLOCK(mp, s);
- iip->ili_inode->i_flags |= XFS_ISTALE;
+ xfs_iflags_set(iip->ili_inode, XFS_ISTALE);
pre_flushed++;
}
lip = lip->li_bio_list;
ASSERT(ip->i_d.di_nlink == 0);
ASSERT(ip->i_d.di_nextents == 0);
ASSERT(ip->i_d.di_anextents == 0);
- ASSERT((ip->i_d.di_size == 0) ||
+ ASSERT((ip->i_d.di_size == 0 && ip->i_size == 0) ||
((ip->i_d.di_mode & S_IFMT) != S_IFREG));
ASSERT(ip->i_d.di_nblocks == 0);
ktrace_free(ip->i_dir_trace);
#endif
if (ip->i_itemp) {
- /* XXXdpd should be able to assert this but shutdown
- * is leaving the AIL behind. */
- ASSERT(((ip->i_itemp->ili_item.li_flags & XFS_LI_IN_AIL) == 0) ||
- XFS_FORCED_SHUTDOWN(ip->i_mount));
+ /*
+ * Only if we are shutting down the fs will we see an
+ * inode still in the AIL. If it is there, we should remove
+ * it to prevent a use-after-free from occurring.
+ */
+ xfs_mount_t *mp = ip->i_mount;
+ xfs_log_item_t *lip = &ip->i_itemp->ili_item;
+ int s;
+
+ ASSERT(((lip->li_flags & XFS_LI_IN_AIL) == 0) ||
+ XFS_FORCED_SHUTDOWN(ip->i_mount));
+ if (lip->li_flags & XFS_LI_IN_AIL) {
+ AIL_LOCK(mp, s);
+ if (lip->li_flags & XFS_LI_IN_AIL)
+ xfs_trans_delete_ail(mp, lip, s);
+ else
+ AIL_UNLOCK(mp, s);
+ }
xfs_inode_item_destroy(ip);
}
kmem_zone_free(xfs_inode_zone, ip);
{
ASSERT(atomic_read(&ip->i_pincount) > 0);
- if (atomic_dec_and_test(&ip->i_pincount)) {
+ if (atomic_dec_and_lock(&ip->i_pincount, &ip->i_flags_lock)) {
+
/*
- * If the inode is currently being reclaimed, the
- * linux inode _and_ the xfs vnode may have been
- * freed so we cannot reference either of them safely.
- * Hence we should not try to do anything to them
- * if the xfs inode is currently in the reclaim
- * path.
+ * If the inode is currently being reclaimed, the link between
+ * the bhv_vnode and the xfs_inode will be broken after the
+ * XFS_IRECLAIM* flag is set. Hence, if these flags are not
+ * set, then we can move forward and mark the linux inode dirty
+ * knowing that it is still valid as it won't freed until after
+ * the bhv_vnode<->xfs_inode link is broken in xfs_reclaim. The
+ * i_flags_lock is used to synchronise the setting of the
+ * XFS_IRECLAIM* flags and the breaking of the link, and so we
+ * can execute atomically w.r.t to reclaim by holding this lock
+ * here.
*
- * However, we still need to issue the unpin wakeup
- * call as the inode reclaim may be blocked waiting for
- * the inode to become unpinned.
+ * However, we still need to issue the unpin wakeup call as the
+ * inode reclaim may be blocked waiting for the inode to become
+ * unpinned.
*/
- if (!(ip->i_flags & (XFS_IRECLAIM|XFS_IRECLAIMABLE))) {
- vnode_t *vp = XFS_ITOV_NULL(ip);
- /* make sync come back and flush this inode */
- if (vp) {
- struct inode *inode = vn_to_inode(vp);
+ if (!__xfs_iflags_test(ip, XFS_IRECLAIM|XFS_IRECLAIMABLE)) {
+ bhv_vnode_t *vp = XFS_ITOV_NULL(ip);
+ struct inode *inode = NULL;
- if (!(inode->i_state & I_NEW))
- mark_inode_dirty_sync(inode);
- }
+ BUG_ON(vp == NULL);
+ inode = vn_to_inode(vp);
+ BUG_ON(inode->i_state & I_CLEAR);
+
+ /* make sync come back and flush this inode */
+ if (!(inode->i_state & (I_NEW|I_FREEING)))
+ mark_inode_dirty_sync(inode);
}
+ spin_unlock(&ip->i_flags_lock);
wake_up(&ip->i_ipin_wait);
}
}
ASSERT(ifp->if_bytes <= XFS_IFORK_SIZE(ip, whichfork));
memcpy(cp, ifp->if_u1.if_data, ifp->if_bytes);
}
- if (whichfork == XFS_DATA_FORK) {
- if (unlikely(XFS_DIR_SHORTFORM_VALIDATE_ONDISK(mp, dip))) {
- XFS_ERROR_REPORT("xfs_iflush_fork",
- XFS_ERRLEVEL_LOW, mp);
- return XFS_ERROR(EFSCORRUPTED);
- }
- }
break;
case XFS_DINODE_FMT_EXTENTS:
XFS_STATS_INC(xs_iflush_count);
ASSERT(ismrlocked(&ip->i_lock, MR_UPDATE|MR_ACCESS));
- ASSERT(valusema(&ip->i_flock) <= 0);
+ ASSERT(issemalocked(&(ip->i_flock)));
ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE ||
ip->i_d.di_nextents > ip->i_df.if_ext_max);
corrupt_out:
xfs_buf_relse(bp);
- xfs_force_shutdown(mp, XFS_CORRUPT_INCORE);
+ xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
xfs_iflush_abort(ip);
/*
* Unlocks the flush lock
xfs_buf_relse(bp);
}
- xfs_force_shutdown(mp, XFS_CORRUPT_INCORE);
+ xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
if(!bufwasdelwri) {
/*
SPLDECL(s);
ASSERT(ismrlocked(&ip->i_lock, MR_UPDATE|MR_ACCESS));
- ASSERT(valusema(&ip->i_flock) <= 0);
+ ASSERT(issemalocked(&(ip->i_flock)));
ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE ||
ip->i_d.di_nextents > ip->i_df.if_ext_max);
xfs_mount_t *mp)
{
xfs_inode_t *ip;
- vnode_t *vp;
+ bhv_vnode_t *vp;
again:
XFS_MOUNT_ILOCK(mp);
*/
memcpy(ifp->if_u2.if_inline_ext, ifp->if_u1.if_extents,
nextents * sizeof(xfs_bmbt_rec_t));
- kmem_free(ifp->if_u1.if_extents, KM_SLEEP);
+ kmem_free(ifp->if_u1.if_extents, ifp->if_real_bytes);
ifp->if_u1.if_extents = ifp->if_u2.if_inline_ext;
ifp->if_real_bytes = 0;
}
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff