#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>
{
xfs_perag_t *pag = &mp->m_perag[ag];
int nr_found;
- int first_index = 0;
+ uint32_t 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 */
-
- 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;
- }
do {
struct inode *inode;
- boolean_t inode_refed;
xfs_inode_t *ip = NULL;
+ int lock_flags = XFS_ILOCK_SHARED;
/*
* use a gang lookup to find the next inode in the tree
break;
}
- /* update the index for the next lookup */
- first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
-
/*
- * skip inodes in reclaim. Let xfs_syncsub do that for
- * us so we don't need to worry.
+ * 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.
*/
- if (xfs_iflags_test(ip, (XFS_IRECLAIM|XFS_IRECLAIMABLE))) {
- read_unlock(&pag->pag_ici_lock);
- continue;
- }
-
- /* bad inodes are dealt with elsewhere */
- inode = VFS_I(ip);
- if (is_bad_inode(inode)) {
+ first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
+ if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino)) {
read_unlock(&pag->pag_ici_lock);
- continue;
+ break;
}
/* nothing to sync during shutdown */
}
/*
- * 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.
+ * If we can't get a reference on the inode, it must be
+ * in reclaim. Leave it for the reclaim code to flush.
*/
- 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;
- }
+ inode = VFS_I(ip);
+ if (!igrab(inode)) {
read_unlock(&pag->pag_ici_lock);
+ continue;
+ }
+ read_unlock(&pag->pag_ici_lock);
+
+ /* avoid new or bad inodes */
+ if (is_bad_inode(inode) ||
+ xfs_iflags_test(ip, XFS_INEW)) {
+ IRELE(ip);
+ continue;
}
/*
* 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.
+ * we need to hold the iolock.
*/
- 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 (flags & SYNC_DELWRI) {
+ if (VN_DIRTY(inode)) {
+ if (flags & SYNC_TRYLOCK) {
+ if (xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED))
+ lock_flags |= XFS_IOLOCK_SHARED;
+ } else {
+ xfs_ilock(ip, XFS_IOLOCK_SHARED);
+ lock_flags |= XFS_IOLOCK_SHARED;
+ }
+ if (lock_flags & XFS_IOLOCK_SHARED) {
+ error = xfs_flush_pages(ip, 0, -1,
+ (flags & SYNC_WAIT) ? 0
+ : XFS_B_ASYNC,
+ FI_NONE);
+ }
+ }
+ if (VN_CACHED(inode) && (flags & SYNC_IOWAIT))
+ xfs_ioend_wait(ip);
}
+ xfs_ilock(ip, XFS_ILOCK_SHARED);
if ((flags & SYNC_ATTR) && !xfs_inode_clean(ip)) {
if (flags & SYNC_WAIT) {
xfs_ifunlock(ip);
}
}
-
- if (lock_flags)
- xfs_iunlock(ip, lock_flags);
-
- if (inode_refed) {
- IRELE(ip);
- }
+ xfs_iput(ip, lock_flags);
if (error)
last_error = error;
}
/*
- * First stage of freeze - no more writers will make progress now we are here,
+ * When remounting a filesystem read-only or freezing the filesystem, we have
+ * two phases to execute. This first phase is syncing the data before we
+ * quiesce the filesystem, and the second is flushing all the inodes out after
+ * we've waited for all the transactions created by the first phase to
+ * complete. The second phase ensures that the inodes are written to their
+ * location on disk rather than just existing in transactions in the log. This
+ * means after a quiesce there is no log replay required to write the inodes to
+ * disk (this is the main difference between a sync and a quiesce).
+ */
+/*
+ * First stage of freeze - no writers will make progress now we are here,
* so we flush delwri and delalloc buffers here, then wait for all I/O to
* complete. Data is frozen at that point. Metadata is not frozen,
- * transactions can still occur here so don't bother flushing the buftarg (i.e
- * SYNC_QUIESCE) because it'll just get dirty again.
+ * transactions can still occur here so don't bother flushing the buftarg
+ * because it'll just get dirty again.
*/
int
xfs_quiesce_data(
/* push non-blocking */
xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_BDFLUSH);
- XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
+ xfs_qm_sync(mp, SYNC_BDFLUSH);
xfs_filestream_flush(mp);
/* push and block */
xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_WAIT|SYNC_IOWAIT);
- XFS_QM_DQSYNC(mp, SYNC_WAIT);
+ xfs_qm_sync(mp, SYNC_WAIT);
- /* write superblock and hoover shutdown errors */
+ /* write superblock and hoover up shutdown errors */
error = xfs_sync_fsdata(mp, 0);
- /* flush devices */
- XFS_bflush(mp->m_ddev_targp);
+ /* flush data-only devices */
if (mp->m_rtdev_targp)
XFS_bflush(mp->m_rtdev_targp);
return error;
}
-/*
- * xfs_sync flushes any pending I/O to file system vfsp.
- *
- * This routine is called by vfs_sync() to make sure that things make it
- * out to disk eventually, on sync() system calls to flush out everything,
- * and when the file system is unmounted. For the vfs_sync() case, all
- * we really need to do is sync out the log to make all of our meta-data
- * updates permanent (except for timestamps). For calls from pflushd(),
- * dirty pages are kept moving by calling pdflush() on the inodes
- * containing them. We also flush the inodes that we can lock without
- * sleeping and the superblock if we can lock it without sleeping from
- * vfs_sync() so that items at the tail of the log are always moving out.
- *
- * Flags:
- * SYNC_BDFLUSH - We're being called from vfs_sync() so we don't want
- * to sleep if we can help it. All we really need
- * to do is ensure that the log is synced at least
- * periodically. We also push the inodes and
- * superblock if we can lock them without sleeping
- * and they are not pinned.
- * SYNC_ATTR - We need to flush the inodes. Now handled by direct calls
- * to xfs_sync_inodes().
- * SYNC_WAIT - All the flushes that take place in this call should
- * be synchronous.
- * SYNC_DELWRI - This tells us to push dirty pages associated with
- * inodes. SYNC_WAIT and SYNC_BDFLUSH are used to
- * determine if they should be flushed sync, async, or
- * delwri.
- * SYNC_FSDATA - This indicates that the caller would like to make
- * sure the superblock is safe on disk. We can ensure
- * this by simply making sure the log gets flushed
- * if SYNC_BDFLUSH is set, and by actually writing it
- * out otherwise.
- * SYNC_IOWAIT - The caller wants us to wait for all data I/O to complete
- * before we return (including direct I/O). Forms the drain
- * side of the write barrier needed to safely quiesce the
- * filesystem.
- *
- */
-int
-xfs_sync(
- xfs_mount_t *mp,
- int flags)
+STATIC void
+xfs_quiesce_fs(
+ struct xfs_mount *mp)
{
- int error;
- int last_error = 0;
- uint log_flags = XFS_LOG_FORCE;
+ int count = 0, pincount;
- ASSERT(!(flags & SYNC_ATTR));
+ xfs_flush_buftarg(mp->m_ddev_targp, 0);
+ xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
/*
- * Get the Quota Manager to flush the dquots.
- *
- * If XFS quota support is not enabled or this filesystem
- * instance does not use quotas XFS_QM_DQSYNC will always
- * return zero.
+ * This loop must run at least twice. The first instance of the loop
+ * will flush most meta data but that will generate more meta data
+ * (typically directory updates). Which then must be flushed and
+ * logged before we can write the unmount record.
*/
- error = XFS_QM_DQSYNC(mp, flags);
- if (error) {
- /*
- * If we got an IO error, we will be shutting down.
- * So, there's nothing more for us to do here.
- */
- ASSERT(error != EIO || XFS_FORCED_SHUTDOWN(mp));
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(error);
- }
-
- if (flags & SYNC_IOWAIT)
- xfs_filestream_flush(mp);
-
- /*
- * Sync out the log. This ensures that the log is periodically
- * flushed even if there is not enough activity to fill it up.
- */
- if (flags & SYNC_WAIT)
- log_flags |= XFS_LOG_SYNC;
+ do {
+ xfs_sync_inodes(mp, SYNC_ATTR|SYNC_WAIT);
+ pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
+ if (!pincount) {
+ delay(50);
+ count++;
+ }
+ } while (count < 2);
+}
- xfs_log_force(mp, (xfs_lsn_t)0, log_flags);
+/*
+ * Second stage of a quiesce. The data is already synced, now we have to take
+ * care of the metadata. New transactions are already blocked, so we need to
+ * wait for any remaining transactions to drain out before proceding.
+ */
+void
+xfs_quiesce_attr(
+ struct xfs_mount *mp)
+{
+ int error = 0;
- if (flags & SYNC_DELWRI) {
- if (flags & SYNC_BDFLUSH)
- xfs_finish_reclaim_all(mp, 1, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
- else
- error = xfs_sync_inodes(mp, flags);
- /*
- * Flushing out dirty data above probably generated more
- * log activity, so if this isn't vfs_sync() then flush
- * the log again.
- */
- xfs_log_force(mp, 0, log_flags);
- }
+ /* wait for all modifications to complete */
+ while (atomic_read(&mp->m_active_trans) > 0)
+ delay(100);
- if (flags & SYNC_FSDATA) {
- error = xfs_sync_fsdata(mp, flags);
- if (error)
- last_error = error;
- }
+ /* flush inodes and push all remaining buffers out to disk */
+ xfs_quiesce_fs(mp);
/*
- * Now check to see if the log needs a "dummy" transaction.
+ * Just warn here till VFS can correctly support
+ * read-only remount without racing.
*/
- if (!(flags & SYNC_REMOUNT) && xfs_log_need_covered(mp)) {
- error = xfs_commit_dummy_trans(mp, log_flags);
- if (error)
- return error;
- }
-
- return XFS_ERROR(last_error);
+ WARN_ON(atomic_read(&mp->m_active_trans) != 0);
+
+ /* Push the superblock and write an unmount record */
+ error = xfs_log_sbcount(mp, 1);
+ if (error)
+ xfs_fs_cmn_err(CE_WARN, mp,
+ "xfs_attr_quiesce: failed to log sb changes. "
+ "Frozen image may not be consistent.");
+ xfs_log_unmount_write(mp);
+ xfs_unmountfs_writesb(mp);
}
/*
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_inodes(mp, SYNC_DELWRI | SYNC_TRYLOCK);
+ xfs_sync_inodes(mp, SYNC_DELWRI | SYNC_TRYLOCK | SYNC_IOWAIT);
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_finish_reclaim_all(mp, 1, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+ xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
/* dgc: errors ignored here */
- error = XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
+ error = xfs_qm_sync(mp, SYNC_BDFLUSH);
error = xfs_sync_fsdata(mp, SYNC_BDFLUSH);
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);
kthread_stop(mp->m_sync_task);
}
+int
+xfs_reclaim_inode(
+ xfs_inode_t *ip,
+ int locked,
+ int sync_mode)
+{
+ xfs_perag_t *pag = xfs_get_perag(ip->i_mount, ip->i_ino);
+
+ /* The hash lock here protects a thread in xfs_iget_core from
+ * racing with us on linking the inode back with a vnode.
+ * Once we have the XFS_IRECLAIM flag set it will not touch
+ * us.
+ */
+ write_lock(&pag->pag_ici_lock);
+ spin_lock(&ip->i_flags_lock);
+ if (__xfs_iflags_test(ip, XFS_IRECLAIM) ||
+ !__xfs_iflags_test(ip, XFS_IRECLAIMABLE)) {
+ spin_unlock(&ip->i_flags_lock);
+ write_unlock(&pag->pag_ici_lock);
+ if (locked) {
+ xfs_ifunlock(ip);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ }
+ return 1;
+ }
+ __xfs_iflags_set(ip, XFS_IRECLAIM);
+ spin_unlock(&ip->i_flags_lock);
+ write_unlock(&pag->pag_ici_lock);
+ xfs_put_perag(ip->i_mount, pag);
+
+ /*
+ * If the inode is still dirty, then flush it out. If the inode
+ * is not in the AIL, then it will be OK to flush it delwri as
+ * long as xfs_iflush() does not keep any references to the inode.
+ * We leave that decision up to xfs_iflush() since it has the
+ * knowledge of whether it's OK to simply do a delwri flush of
+ * the inode or whether we need to wait until the inode is
+ * pulled from the AIL.
+ * We get the flush lock regardless, though, just to make sure
+ * we don't free it while it is being flushed.
+ */
+ if (!locked) {
+ xfs_ilock(ip, XFS_ILOCK_EXCL);
+ xfs_iflock(ip);
+ }
+
+ /*
+ * In the case of a forced shutdown we rely on xfs_iflush() to
+ * wait for the inode to be unpinned before returning an error.
+ */
+ if (!is_bad_inode(VFS_I(ip)) && xfs_iflush(ip, sync_mode) == 0) {
+ /* synchronize with xfs_iflush_done */
+ xfs_iflock(ip);
+ xfs_ifunlock(ip);
+ }
+
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ xfs_ireclaim(ip);
+ return 0;
+}
+
+/*
+ * We set the inode flag atomically with the radix tree tag.
+ * Once we get tag lookups on the radix tree, this inode flag
+ * can go away.
+ */
+void
+xfs_inode_set_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);
+ radix_tree_tag_set(&pag->pag_ici_root,
+ XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG);
+ __xfs_iflags_set(ip, XFS_IRECLAIMABLE);
+ spin_unlock(&ip->i_flags_lock);
+ read_unlock(&pag->pag_ici_lock);
+ xfs_put_perag(mp, pag);
+}
+
+void
+__xfs_inode_clear_reclaim_tag(
+ xfs_mount_t *mp,
+ xfs_perag_t *pag,
+ xfs_inode_t *ip)
+{
+ radix_tree_tag_clear(&pag->pag_ici_root,
+ 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)
+{
+ xfs_inode_t *ip = NULL;
+ xfs_perag_t *pag = &mp->m_perag[ag];
+ int nr_found;
+ uint32_t 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. 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)) {
+ read_unlock(&pag->pag_ici_lock);
+ break;
+ }
+
+ /* 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;
+ }
+ }
+ 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;
+
+}
+
+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;
+}
+
+
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff