#include "xfs_buf_item.h"
#include "xfs_inode_item.h"
#include "xfs_rw.h"
+#include "xfs_quota.h"
#include <linux/kthread.h>
#include <linux/freezer.h>
-/*
- * Sync all the inodes in the given AG according to the
- * direction given by the flags.
- */
-STATIC int
-xfs_sync_inodes_ag(
- xfs_mount_t *mp,
- int ag,
- int flags)
+
+STATIC xfs_inode_t *
+xfs_inode_ag_lookup(
+ struct xfs_mount *mp,
+ struct xfs_perag *pag,
+ uint32_t *first_index,
+ int tag)
{
- xfs_perag_t *pag = &mp->m_perag[ag];
- int nr_found;
- int first_index = 0;
- int error = 0;
- int last_error = 0;
- int fflag = XFS_B_ASYNC;
- int lock_flags = XFS_ILOCK_SHARED;
-
- if (flags & SYNC_DELWRI)
- fflag = XFS_B_DELWRI;
- if (flags & SYNC_WAIT)
- fflag = 0; /* synchronous overrides all */
+ int nr_found;
+ struct xfs_inode *ip;
- if (flags & SYNC_DELWRI) {
- /*
- * We need the I/O lock if we're going to call any of
- * the flush/inval routines.
- */
- lock_flags |= XFS_IOLOCK_SHARED;
+ /*
+ * use a gang lookup to find the next inode in the tree
+ * as the tree is sparse and a gang lookup walks to find
+ * the number of objects requested.
+ */
+ read_lock(&pag->pag_ici_lock);
+ if (tag == XFS_ICI_NO_TAG) {
+ nr_found = radix_tree_gang_lookup(&pag->pag_ici_root,
+ (void **)&ip, *first_index, 1);
+ } else {
+ nr_found = radix_tree_gang_lookup_tag(&pag->pag_ici_root,
+ (void **)&ip, *first_index, 1, tag);
}
+ if (!nr_found)
+ goto unlock;
+ /*
+ * Update the index for the next lookup. Catch overflows
+ * into the next AG range which can occur if we have inodes
+ * in the last block of the AG and we are currently
+ * pointing to the last inode.
+ */
+ *first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
+ if (*first_index < XFS_INO_TO_AGINO(mp, ip->i_ino))
+ goto unlock;
+
+ return ip;
+
+unlock:
+ read_unlock(&pag->pag_ici_lock);
+ return NULL;
+}
+
+STATIC int
+xfs_inode_ag_walk(
+ struct xfs_mount *mp,
+ xfs_agnumber_t ag,
+ int (*execute)(struct xfs_inode *ip,
+ struct xfs_perag *pag, int flags),
+ int flags,
+ int tag)
+{
+ struct xfs_perag *pag = &mp->m_perag[ag];
+ uint32_t first_index;
+ int last_error = 0;
+ int skipped;
+
+restart:
+ skipped = 0;
+ first_index = 0;
do {
- struct inode *inode;
- boolean_t inode_refed;
- xfs_inode_t *ip = NULL;
+ int error = 0;
+ xfs_inode_t *ip;
+ ip = xfs_inode_ag_lookup(mp, pag, &first_index, tag);
+ if (!ip)
+ break;
+
+ error = execute(ip, pag, flags);
+ if (error == EAGAIN) {
+ skipped++;
+ continue;
+ }
+ if (error)
+ last_error = error;
/*
- * use a gang lookup to find the next inode in the tree
- * as the tree is sparse and a gang lookup walks to find
- * the number of objects requested.
+ * bail out if the filesystem is corrupted.
*/
- read_lock(&pag->pag_ici_lock);
- nr_found = radix_tree_gang_lookup(&pag->pag_ici_root,
- (void**)&ip, first_index, 1);
-
- if (!nr_found) {
- read_unlock(&pag->pag_ici_lock);
+ if (error == EFSCORRUPTED)
break;
- }
- /* update the index for the next lookup */
- first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
+ } while (1);
- /*
- * skip inodes in reclaim. Let xfs_syncsub do that for
- * us so we don't need to worry.
- */
- if (xfs_iflags_test(ip, (XFS_IRECLAIM|XFS_IRECLAIMABLE))) {
- read_unlock(&pag->pag_ici_lock);
- continue;
- }
+ if (skipped) {
+ delay(1);
+ goto restart;
+ }
+
+ xfs_put_perag(mp, pag);
+ return last_error;
+}
+
+int
+xfs_inode_ag_iterator(
+ struct xfs_mount *mp,
+ int (*execute)(struct xfs_inode *ip,
+ struct xfs_perag *pag, int flags),
+ int flags,
+ int tag)
+{
+ int error = 0;
+ int last_error = 0;
+ xfs_agnumber_t ag;
- /* bad inodes are dealt with elsewhere */
- inode = VFS_I(ip);
- if (is_bad_inode(inode)) {
- read_unlock(&pag->pag_ici_lock);
+ for (ag = 0; ag < mp->m_sb.sb_agcount; ag++) {
+ if (!mp->m_perag[ag].pag_ici_init)
continue;
+ error = xfs_inode_ag_walk(mp, ag, execute, flags, tag);
+ if (error) {
+ last_error = error;
+ if (error == EFSCORRUPTED)
+ break;
}
+ }
+ return XFS_ERROR(last_error);
+}
- /* nothing to sync during shutdown */
- if (XFS_FORCED_SHUTDOWN(mp)) {
- read_unlock(&pag->pag_ici_lock);
- return 0;
- }
+/* must be called with pag_ici_lock held and releases it */
+int
+xfs_sync_inode_valid(
+ struct xfs_inode *ip,
+ struct xfs_perag *pag)
+{
+ struct inode *inode = VFS_I(ip);
- /*
- * If we can't get a reference on the VFS_I, the inode must be
- * in reclaim. If we can get the inode lock without blocking,
- * it is safe to flush the inode because we hold the tree lock
- * and xfs_iextract will block right now. Hence if we lock the
- * inode while holding the tree lock, xfs_ireclaim() is
- * guaranteed to block on the inode lock we now hold and hence
- * it is safe to reference the inode until we drop the inode
- * locks completely.
- */
- inode_refed = B_FALSE;
- if (igrab(inode)) {
- read_unlock(&pag->pag_ici_lock);
- xfs_ilock(ip, lock_flags);
- inode_refed = B_TRUE;
- } else {
- if (!xfs_ilock_nowait(ip, lock_flags)) {
- /* leave it to reclaim */
- read_unlock(&pag->pag_ici_lock);
- continue;
- }
- read_unlock(&pag->pag_ici_lock);
- }
+ /* nothing to sync during shutdown */
+ if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+ read_unlock(&pag->pag_ici_lock);
+ return EFSCORRUPTED;
+ }
- /*
- * If we have to flush data or wait for I/O completion
- * we need to drop the ilock that we currently hold.
- * If we need to drop the lock, insert a marker if we
- * have not already done so.
- */
- if ((flags & SYNC_DELWRI) && VN_DIRTY(inode)) {
- xfs_iunlock(ip, XFS_ILOCK_SHARED);
- error = xfs_flush_pages(ip, 0, -1, fflag, FI_NONE);
- if (flags & SYNC_IOWAIT)
- vn_iowait(ip);
- xfs_ilock(ip, XFS_ILOCK_SHARED);
- }
+ /*
+ * If we can't get a reference on the inode, it must be in reclaim.
+ * Leave it for the reclaim code to flush. Also avoid inodes that
+ * haven't been fully initialised.
+ */
+ if (!igrab(inode)) {
+ read_unlock(&pag->pag_ici_lock);
+ return ENOENT;
+ }
+ read_unlock(&pag->pag_ici_lock);
- if ((flags & SYNC_ATTR) && !xfs_inode_clean(ip)) {
- if (flags & SYNC_WAIT) {
- xfs_iflock(ip);
- if (!xfs_inode_clean(ip))
- error = xfs_iflush(ip, XFS_IFLUSH_SYNC);
- else
- xfs_ifunlock(ip);
- } else if (xfs_iflock_nowait(ip)) {
- if (!xfs_inode_clean(ip))
- error = xfs_iflush(ip, XFS_IFLUSH_DELWRI);
- else
- xfs_ifunlock(ip);
- }
- }
+ if (is_bad_inode(inode) || xfs_iflags_test(ip, XFS_INEW)) {
+ IRELE(ip);
+ return ENOENT;
+ }
- if (lock_flags)
- xfs_iunlock(ip, lock_flags);
+ return 0;
+}
- if (inode_refed) {
- IRELE(ip);
- }
+STATIC int
+xfs_sync_inode_data(
+ struct xfs_inode *ip,
+ struct xfs_perag *pag,
+ int flags)
+{
+ struct inode *inode = VFS_I(ip);
+ struct address_space *mapping = inode->i_mapping;
+ int error = 0;
- if (error)
- last_error = error;
- /*
- * bail out if the filesystem is corrupted.
- */
- if (error == EFSCORRUPTED)
- return XFS_ERROR(error);
+ error = xfs_sync_inode_valid(ip, pag);
+ if (error)
+ return error;
- } while (nr_found);
+ if (!mapping_tagged(mapping, PAGECACHE_TAG_DIRTY))
+ goto out_wait;
- return last_error;
+ if (!xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED)) {
+ if (flags & SYNC_TRYLOCK)
+ goto out_wait;
+ xfs_ilock(ip, XFS_IOLOCK_SHARED);
+ }
+
+ error = xfs_flush_pages(ip, 0, -1, (flags & SYNC_WAIT) ?
+ 0 : XFS_B_ASYNC, FI_NONE);
+ xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+
+ out_wait:
+ if (flags & SYNC_WAIT)
+ xfs_ioend_wait(ip);
+ IRELE(ip);
+ return error;
}
-int
-xfs_sync_inodes(
- xfs_mount_t *mp,
- int flags)
+STATIC int
+xfs_sync_inode_attr(
+ struct xfs_inode *ip,
+ struct xfs_perag *pag,
+ int flags)
{
- int error;
- int last_error;
- int i;
- int lflags = XFS_LOG_FORCE;
+ int error = 0;
- if (mp->m_flags & XFS_MOUNT_RDONLY)
- return 0;
- error = 0;
- last_error = 0;
+ error = xfs_sync_inode_valid(ip, pag);
+ if (error)
+ return error;
- if (flags & SYNC_WAIT)
- lflags |= XFS_LOG_SYNC;
+ xfs_ilock(ip, XFS_ILOCK_SHARED);
+ if (xfs_inode_clean(ip))
+ goto out_unlock;
+ if (!xfs_iflock_nowait(ip)) {
+ if (!(flags & SYNC_WAIT))
+ goto out_unlock;
+ xfs_iflock(ip);
+ }
- for (i = 0; i < mp->m_sb.sb_agcount; i++) {
- if (!mp->m_perag[i].pag_ici_init)
- continue;
- error = xfs_sync_inodes_ag(mp, i, flags);
- if (error)
- last_error = error;
- if (error == EFSCORRUPTED)
- break;
+ if (xfs_inode_clean(ip)) {
+ xfs_ifunlock(ip);
+ goto out_unlock;
}
- if (flags & SYNC_DELWRI)
- xfs_log_force(mp, 0, lflags);
- return XFS_ERROR(last_error);
+ error = xfs_iflush(ip, (flags & SYNC_WAIT) ?
+ XFS_IFLUSH_SYNC : XFS_IFLUSH_DELWRI);
+
+ out_unlock:
+ xfs_iunlock(ip, XFS_ILOCK_SHARED);
+ IRELE(ip);
+ return error;
+}
+
+/*
+ * Write out pagecache data for the whole filesystem.
+ */
+int
+xfs_sync_data(
+ struct xfs_mount *mp,
+ int flags)
+{
+ int error;
+
+ ASSERT((flags & ~(SYNC_TRYLOCK|SYNC_WAIT)) == 0);
+
+ error = xfs_inode_ag_iterator(mp, xfs_sync_inode_data, flags,
+ XFS_ICI_NO_TAG);
+ if (error)
+ return XFS_ERROR(error);
+
+ xfs_log_force(mp, 0,
+ (flags & SYNC_WAIT) ?
+ XFS_LOG_FORCE | XFS_LOG_SYNC :
+ XFS_LOG_FORCE);
+ return 0;
+}
+
+/*
+ * Write out inode metadata (attributes) for the whole filesystem.
+ */
+int
+xfs_sync_attr(
+ struct xfs_mount *mp,
+ int flags)
+{
+ ASSERT((flags & ~SYNC_WAIT) == 0);
+
+ return xfs_inode_ag_iterator(mp, xfs_sync_inode_attr, flags,
+ XFS_ICI_NO_TAG);
}
STATIC int
* If this is xfssyncd() then only sync the superblock if we can
* lock it without sleeping and it is not pinned.
*/
- if (flags & SYNC_BDFLUSH) {
+ if (flags & SYNC_TRYLOCK) {
ASSERT(!(flags & SYNC_WAIT));
bp = xfs_getsb(mp, XFS_BUF_TRYLOCK);
int error;
/* push non-blocking */
- xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_BDFLUSH);
- XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
+ xfs_sync_data(mp, 0);
+ xfs_qm_sync(mp, SYNC_TRYLOCK);
xfs_filestream_flush(mp);
/* push and block */
- xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_WAIT|SYNC_IOWAIT);
- XFS_QM_DQSYNC(mp, SYNC_WAIT);
+ xfs_sync_data(mp, SYNC_WAIT);
+ xfs_qm_sync(mp, SYNC_WAIT);
/* write superblock and hoover up shutdown errors */
error = xfs_sync_fsdata(mp, 0);
int count = 0, pincount;
xfs_flush_buftarg(mp->m_ddev_targp, 0);
- xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+ xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
/*
* This loop must run at least twice. The first instance of the loop
* logged before we can write the unmount record.
*/
do {
- xfs_sync_inodes(mp, SYNC_ATTR|SYNC_WAIT);
+ xfs_sync_attr(mp, SYNC_WAIT);
pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
if (!pincount) {
delay(50);
/* flush inodes and push all remaining buffers out to disk */
xfs_quiesce_fs(mp);
- ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0);
+ /*
+ * Just warn here till VFS can correctly support
+ * read-only remount without racing.
+ */
+ WARN_ON(atomic_read(&mp->m_active_trans) != 0);
/* Push the superblock and write an unmount record */
error = xfs_log_sbcount(mp, 1);
xfs_syncd_queue_work(
struct xfs_mount *mp,
void *data,
- void (*syncer)(struct xfs_mount *, void *))
+ void (*syncer)(struct xfs_mount *, void *),
+ struct completion *completion)
{
- struct bhv_vfs_sync_work *work;
+ struct xfs_sync_work *work;
- work = kmem_alloc(sizeof(struct bhv_vfs_sync_work), KM_SLEEP);
+ work = kmem_alloc(sizeof(struct xfs_sync_work), KM_SLEEP);
INIT_LIST_HEAD(&work->w_list);
work->w_syncer = syncer;
work->w_data = data;
work->w_mount = mp;
+ work->w_completion = completion;
spin_lock(&mp->m_sync_lock);
list_add_tail(&work->w_list, &mp->m_sync_list);
spin_unlock(&mp->m_sync_lock);
* heads, looking about for more room...
*/
STATIC void
-xfs_flush_inode_work(
+xfs_flush_inodes_work(
struct xfs_mount *mp,
void *arg)
{
struct inode *inode = arg;
- filemap_flush(inode->i_mapping);
+ xfs_sync_data(mp, SYNC_TRYLOCK);
+ xfs_sync_data(mp, SYNC_TRYLOCK | SYNC_WAIT);
iput(inode);
}
void
-xfs_flush_inode(
+xfs_flush_inodes(
xfs_inode_t *ip)
{
struct inode *inode = VFS_I(ip);
+ DECLARE_COMPLETION_ONSTACK(completion);
igrab(inode);
- xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inode_work);
- delay(msecs_to_jiffies(500));
-}
-
-/*
- * This is the "bigger hammer" version of xfs_flush_inode_work...
- * (IOW, "If at first you don't succeed, use a Bigger Hammer").
- */
-STATIC void
-xfs_flush_device_work(
- struct xfs_mount *mp,
- void *arg)
-{
- struct inode *inode = arg;
- sync_blockdev(mp->m_super->s_bdev);
- iput(inode);
-}
-
-void
-xfs_flush_device(
- xfs_inode_t *ip)
-{
- struct inode *inode = VFS_I(ip);
-
- igrab(inode);
- xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_device_work);
- delay(msecs_to_jiffies(500));
+ xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inodes_work, &completion);
+ wait_for_completion(&completion);
xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC);
}
if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
- xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+ xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
/* dgc: errors ignored here */
- error = XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
- error = xfs_sync_fsdata(mp, SYNC_BDFLUSH);
+ error = xfs_qm_sync(mp, SYNC_TRYLOCK);
+ error = xfs_sync_fsdata(mp, SYNC_TRYLOCK);
if (xfs_log_need_covered(mp))
error = xfs_commit_dummy_trans(mp, XFS_LOG_FORCE);
}
{
struct xfs_mount *mp = arg;
long timeleft;
- bhv_vfs_sync_work_t *work, *n;
+ xfs_sync_work_t *work, *n;
LIST_HEAD (tmp);
set_freezable();
list_del(&work->w_list);
if (work == &mp->m_sync_work)
continue;
+ if (work->w_completion)
+ complete(work->w_completion);
kmem_free(work);
}
}
{
mp->m_sync_work.w_syncer = xfs_sync_worker;
mp->m_sync_work.w_mount = mp;
+ mp->m_sync_work.w_completion = NULL;
mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd");
if (IS_ERR(mp->m_sync_task))
return -PTR_ERR(mp->m_sync_task);
xfs_ifunlock(ip);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
}
- return 1;
+ return -EAGAIN;
}
__xfs_iflags_set(ip, XFS_IRECLAIM);
spin_unlock(&ip->i_flags_lock);
return 0;
}
+void
+__xfs_inode_set_reclaim_tag(
+ struct xfs_perag *pag,
+ struct xfs_inode *ip)
+{
+ radix_tree_tag_set(&pag->pag_ici_root,
+ XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino),
+ XFS_ICI_RECLAIM_TAG);
+}
+
/*
* We set the inode flag atomically with the radix tree tag.
* Once we get tag lookups on the radix tree, this inode flag
read_lock(&pag->pag_ici_lock);
spin_lock(&ip->i_flags_lock);
- radix_tree_tag_set(&pag->pag_ici_root,
- XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG);
+ __xfs_inode_set_reclaim_tag(pag, ip);
__xfs_iflags_set(ip, XFS_IRECLAIMABLE);
spin_unlock(&ip->i_flags_lock);
read_unlock(&pag->pag_ici_lock);
XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG);
}
-void
-xfs_inode_clear_reclaim_tag(
- xfs_inode_t *ip)
-{
- xfs_mount_t *mp = ip->i_mount;
- xfs_perag_t *pag = xfs_get_perag(mp, ip->i_ino);
-
- read_lock(&pag->pag_ici_lock);
- spin_lock(&ip->i_flags_lock);
- __xfs_inode_clear_reclaim_tag(mp, pag, ip);
- spin_unlock(&ip->i_flags_lock);
- read_unlock(&pag->pag_ici_lock);
- xfs_put_perag(mp, pag);
-}
-
-
-STATIC void
-xfs_reclaim_inodes_ag(
- xfs_mount_t *mp,
- int ag,
- int noblock,
- int mode)
+STATIC int
+xfs_reclaim_inode_now(
+ struct xfs_inode *ip,
+ struct xfs_perag *pag,
+ int flags)
{
- xfs_inode_t *ip = NULL;
- xfs_perag_t *pag = &mp->m_perag[ag];
- int nr_found;
- int first_index;
- int skipped;
-
-restart:
- first_index = 0;
- skipped = 0;
- do {
- /*
- * use a gang lookup to find the next inode in the tree
- * as the tree is sparse and a gang lookup walks to find
- * the number of objects requested.
- */
- read_lock(&pag->pag_ici_lock);
- nr_found = radix_tree_gang_lookup_tag(&pag->pag_ici_root,
- (void**)&ip, first_index, 1,
- XFS_ICI_RECLAIM_TAG);
-
- if (!nr_found) {
- read_unlock(&pag->pag_ici_lock);
- break;
- }
-
- /* update the index for the next lookup */
- first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
-
- ASSERT(xfs_iflags_test(ip, (XFS_IRECLAIMABLE|XFS_IRECLAIM)));
-
- /* ignore if already under reclaim */
- if (xfs_iflags_test(ip, XFS_IRECLAIM)) {
- read_unlock(&pag->pag_ici_lock);
- continue;
- }
-
- if (noblock) {
- if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
- read_unlock(&pag->pag_ici_lock);
- continue;
- }
- if (xfs_ipincount(ip) ||
- !xfs_iflock_nowait(ip)) {
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
- read_unlock(&pag->pag_ici_lock);
- continue;
- }
- }
+ /* ignore if already under reclaim */
+ if (xfs_iflags_test(ip, XFS_IRECLAIM)) {
read_unlock(&pag->pag_ici_lock);
-
- /*
- * hmmm - this is an inode already in reclaim. Do
- * we even bother catching it here?
- */
- if (xfs_reclaim_inode(ip, noblock, mode))
- skipped++;
- } while (nr_found);
-
- if (skipped) {
- delay(1);
- goto restart;
+ return 0;
}
- return;
+ read_unlock(&pag->pag_ici_lock);
+ return xfs_reclaim_inode(ip, 0, flags);
}
int
xfs_reclaim_inodes(
xfs_mount_t *mp,
- int noblock,
int mode)
{
- int i;
-
- for (i = 0; i < mp->m_sb.sb_agcount; i++) {
- if (!mp->m_perag[i].pag_ici_init)
- continue;
- xfs_reclaim_inodes_ag(mp, i, noblock, mode);
- }
- return 0;
+ return xfs_inode_ag_iterator(mp, xfs_reclaim_inode_now, mode,
+ XFS_ICI_RECLAIM_TAG);
}
-
-
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff