aboutsummaryrefslogtreecommitdiff
path: root/fs/xfs/linux-2.6/xfs_sync.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/xfs/linux-2.6/xfs_sync.c')
-rw-r--r--fs/xfs/linux-2.6/xfs_sync.c388
1 files changed, 262 insertions, 126 deletions
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index 1f5e4bb5e970..ef7f0218bccb 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -90,14 +90,14 @@ xfs_inode_ag_lookup(
STATIC int
xfs_inode_ag_walk(
struct xfs_mount *mp,
- xfs_agnumber_t ag,
+ struct xfs_perag *pag,
int (*execute)(struct xfs_inode *ip,
struct xfs_perag *pag, int flags),
int flags,
int tag,
- int exclusive)
+ int exclusive,
+ int *nr_to_scan)
{
- struct xfs_perag *pag = &mp->m_perag[ag];
uint32_t first_index;
int last_error = 0;
int skipped;
@@ -135,14 +135,12 @@ restart:
if (error == EFSCORRUPTED)
break;
- } while (1);
+ } while ((*nr_to_scan)--);
if (skipped) {
delay(1);
goto restart;
}
-
- xfs_put_perag(mp, pag);
return last_error;
}
@@ -153,23 +151,32 @@ xfs_inode_ag_iterator(
struct xfs_perag *pag, int flags),
int flags,
int tag,
- int exclusive)
+ int exclusive,
+ int *nr_to_scan)
{
int error = 0;
int last_error = 0;
xfs_agnumber_t ag;
+ int nr;
+ nr = nr_to_scan ? *nr_to_scan : INT_MAX;
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,
- exclusive);
+ struct xfs_perag *pag;
+
+ pag = xfs_perag_get(mp, ag);
+ error = xfs_inode_ag_walk(mp, pag, execute, flags, tag,
+ exclusive, &nr);
+ xfs_perag_put(pag);
if (error) {
last_error = error;
if (error == EFSCORRUPTED)
break;
}
+ if (nr <= 0)
+ break;
}
+ if (nr_to_scan)
+ *nr_to_scan = nr;
return XFS_ERROR(last_error);
}
@@ -231,7 +238,7 @@ xfs_sync_inode_data(
}
error = xfs_flush_pages(ip, 0, -1, (flags & SYNC_WAIT) ?
- 0 : XFS_B_ASYNC, FI_NONE);
+ 0 : XBF_ASYNC, FI_NONE);
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
out_wait:
@@ -267,8 +274,7 @@ xfs_sync_inode_attr(
goto out_unlock;
}
- error = xfs_iflush(ip, (flags & SYNC_WAIT) ?
- XFS_IFLUSH_SYNC : XFS_IFLUSH_DELWRI);
+ error = xfs_iflush(ip, flags);
out_unlock:
xfs_iunlock(ip, XFS_ILOCK_SHARED);
@@ -289,14 +295,11 @@ xfs_sync_data(
ASSERT((flags & ~(SYNC_TRYLOCK|SYNC_WAIT)) == 0);
error = xfs_inode_ag_iterator(mp, xfs_sync_inode_data, flags,
- XFS_ICI_NO_TAG, 0);
+ XFS_ICI_NO_TAG, 0, NULL);
if (error)
return XFS_ERROR(error);
- xfs_log_force(mp, 0,
- (flags & SYNC_WAIT) ?
- XFS_LOG_FORCE | XFS_LOG_SYNC :
- XFS_LOG_FORCE);
+ xfs_log_force(mp, (flags & SYNC_WAIT) ? XFS_LOG_SYNC : 0);
return 0;
}
@@ -311,7 +314,7 @@ xfs_sync_attr(
ASSERT((flags & ~SYNC_WAIT) == 0);
return xfs_inode_ag_iterator(mp, xfs_sync_inode_attr, flags,
- XFS_ICI_NO_TAG, 0);
+ XFS_ICI_NO_TAG, 0, NULL);
}
STATIC int
@@ -322,10 +325,6 @@ xfs_commit_dummy_trans(
struct xfs_inode *ip = mp->m_rootip;
struct xfs_trans *tp;
int error;
- int log_flags = XFS_LOG_FORCE;
-
- if (flags & SYNC_WAIT)
- log_flags |= XFS_LOG_SYNC;
/*
* Put a dummy transaction in the log to tell recovery
@@ -347,74 +346,29 @@ xfs_commit_dummy_trans(
xfs_iunlock(ip, XFS_ILOCK_EXCL);
/* the log force ensures this transaction is pushed to disk */
- xfs_log_force(mp, 0, log_flags);
+ xfs_log_force(mp, (flags & SYNC_WAIT) ? XFS_LOG_SYNC : 0);
return error;
}
-int
+STATIC int
xfs_sync_fsdata(
- struct xfs_mount *mp,
- int flags)
+ struct xfs_mount *mp)
{
struct xfs_buf *bp;
- struct xfs_buf_log_item *bip;
- int error = 0;
-
- /*
- * If this is xfssyncd() then only sync the superblock if we can
- * lock it without sleeping and it is not pinned.
- */
- if (flags & SYNC_TRYLOCK) {
- ASSERT(!(flags & SYNC_WAIT));
-
- bp = xfs_getsb(mp, XFS_BUF_TRYLOCK);
- if (!bp)
- goto out;
-
- bip = XFS_BUF_FSPRIVATE(bp, struct xfs_buf_log_item *);
- if (!bip || !xfs_buf_item_dirty(bip) || XFS_BUF_ISPINNED(bp))
- goto out_brelse;
- } else {
- bp = xfs_getsb(mp, 0);
-
- /*
- * If the buffer is pinned then push on the log so we won't
- * get stuck waiting in the write for someone, maybe
- * ourselves, to flush the log.
- *
- * Even though we just pushed the log above, we did not have
- * the superblock buffer locked at that point so it can
- * become pinned in between there and here.
- */
- if (XFS_BUF_ISPINNED(bp))
- xfs_log_force(mp, 0, XFS_LOG_FORCE);
- }
-
-
- if (flags & SYNC_WAIT)
- XFS_BUF_UNASYNC(bp);
- else
- XFS_BUF_ASYNC(bp);
-
- error = xfs_bwrite(mp, bp);
- if (error)
- return error;
/*
- * If this is a data integrity sync make sure all pending buffers
- * are flushed out for the log coverage check below.
+ * If the buffer is pinned then push on the log so we won't get stuck
+ * waiting in the write for someone, maybe ourselves, to flush the log.
+ *
+ * Even though we just pushed the log above, we did not have the
+ * superblock buffer locked at that point so it can become pinned in
+ * between there and here.
*/
- if (flags & SYNC_WAIT)
- xfs_flush_buftarg(mp->m_ddev_targp, 1);
-
- if (xfs_log_need_covered(mp))
- error = xfs_commit_dummy_trans(mp, flags);
- return error;
+ bp = xfs_getsb(mp, 0);
+ if (XFS_BUF_ISPINNED(bp))
+ xfs_log_force(mp, 0);
- out_brelse:
- xfs_buf_relse(bp);
- out:
- return error;
+ return xfs_bwrite(mp, bp);
}
/*
@@ -438,7 +392,7 @@ int
xfs_quiesce_data(
struct xfs_mount *mp)
{
- int error;
+ int error, error2 = 0;
/* push non-blocking */
xfs_sync_data(mp, 0);
@@ -448,17 +402,21 @@ xfs_quiesce_data(
xfs_sync_data(mp, SYNC_WAIT);
xfs_qm_sync(mp, SYNC_WAIT);
- /* drop inode references pinned by filestreams */
- xfs_filestream_flush(mp);
-
/* write superblock and hoover up shutdown errors */
- error = xfs_sync_fsdata(mp, SYNC_WAIT);
+ error = xfs_sync_fsdata(mp);
+
+ /* make sure all delwri buffers are written out */
+ xfs_flush_buftarg(mp->m_ddev_targp, 1);
+
+ /* mark the log as covered if needed */
+ if (xfs_log_need_covered(mp))
+ error2 = xfs_commit_dummy_trans(mp, SYNC_WAIT);
/* flush data-only devices */
if (mp->m_rtdev_targp)
XFS_bflush(mp->m_rtdev_targp);
- return error;
+ return error ? error : error2;
}
STATIC void
@@ -467,16 +425,18 @@ xfs_quiesce_fs(
{
int count = 0, pincount;
+ xfs_reclaim_inodes(mp, 0);
xfs_flush_buftarg(mp->m_ddev_targp, 0);
- xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
/*
* 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.
+ * logged before we can write the unmount record. We also so sync
+ * reclaim of inodes to catch any that the above delwri flush skipped.
*/
do {
+ xfs_reclaim_inodes(mp, SYNC_WAIT);
xfs_sync_attr(mp, SYNC_WAIT);
pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
if (!pincount) {
@@ -575,13 +535,13 @@ xfs_flush_inodes(
igrab(inode);
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);
+ xfs_log_force(ip->i_mount, XFS_LOG_SYNC);
}
/*
- * Every sync period we need to unpin all items, reclaim inodes, sync
- * quota and write out the superblock. We might need to cover the log
- * to indicate it is idle.
+ * Every sync period we need to unpin all items, reclaim inodes and sync
+ * disk quotas. We might need to cover the log to indicate that the
+ * filesystem is idle.
*/
STATIC void
xfs_sync_worker(
@@ -591,11 +551,12 @@ xfs_sync_worker(
int error;
if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
- xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
- xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+ xfs_log_force(mp, 0);
+ xfs_reclaim_inodes(mp, 0);
/* dgc: errors ignored here */
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, 0);
}
mp->m_sync_seq++;
wake_up(&mp->m_wait_single_sync_task);
@@ -613,7 +574,8 @@ xfssyncd(
set_freezable();
timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10);
for (;;) {
- timeleft = schedule_timeout_interruptible(timeleft);
+ if (list_empty(&mp->m_sync_list))
+ timeleft = schedule_timeout_interruptible(timeleft);
/* swsusp */
try_to_freeze();
if (kthread_should_stop() && list_empty(&mp->m_sync_list))
@@ -633,8 +595,7 @@ xfssyncd(
list_add_tail(&mp->m_sync_work.w_list,
&mp->m_sync_list);
}
- list_for_each_entry_safe(work, n, &mp->m_sync_list, w_list)
- list_move(&work->w_list, &tmp);
+ list_splice_init(&mp->m_sync_list, &tmp);
spin_unlock(&mp->m_sync_lock);
list_for_each_entry_safe(work, n, &tmp, w_list) {
@@ -658,7 +619,7 @@ xfs_syncd_init(
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");
+ mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd/%s", mp->m_fsname);
if (IS_ERR(mp->m_sync_task))
return -PTR_ERR(mp->m_sync_task);
return 0;
@@ -679,6 +640,7 @@ __xfs_inode_set_reclaim_tag(
radix_tree_tag_set(&pag->pag_ici_root,
XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino),
XFS_ICI_RECLAIM_TAG);
+ pag->pag_ici_reclaimable++;
}
/*
@@ -690,16 +652,17 @@ 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);
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_perag *pag;
- read_lock(&pag->pag_ici_lock);
+ pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
+ write_lock(&pag->pag_ici_lock);
spin_lock(&ip->i_flags_lock);
__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_put_perag(mp, pag);
+ write_unlock(&pag->pag_ici_lock);
+ xfs_perag_put(pag);
}
void
@@ -710,14 +673,67 @@ __xfs_inode_clear_reclaim_tag(
{
radix_tree_tag_clear(&pag->pag_ici_root,
XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG);
+ pag->pag_ici_reclaimable--;
}
+/*
+ * Inodes in different states need to be treated differently, and the return
+ * value of xfs_iflush is not sufficient to get this right. The following table
+ * lists the inode states and the reclaim actions necessary for non-blocking
+ * reclaim:
+ *
+ *
+ * inode state iflush ret required action
+ * --------------- ---------- ---------------
+ * bad - reclaim
+ * shutdown EIO unpin and reclaim
+ * clean, unpinned 0 reclaim
+ * stale, unpinned 0 reclaim
+ * clean, pinned(*) 0 requeue
+ * stale, pinned EAGAIN requeue
+ * dirty, delwri ok 0 requeue
+ * dirty, delwri blocked EAGAIN requeue
+ * dirty, sync flush 0 reclaim
+ *
+ * (*) dgc: I don't think the clean, pinned state is possible but it gets
+ * handled anyway given the order of checks implemented.
+ *
+ * As can be seen from the table, the return value of xfs_iflush() is not
+ * sufficient to correctly decide the reclaim action here. The checks in
+ * xfs_iflush() might look like duplicates, but they are not.
+ *
+ * Also, because we get the flush lock first, we know that any inode that has
+ * been flushed delwri has had the flush completed by the time we check that
+ * the inode is clean. The clean inode check needs to be done before flushing
+ * the inode delwri otherwise we would loop forever requeuing clean inodes as
+ * we cannot tell apart a successful delwri flush and a clean inode from the
+ * return value of xfs_iflush().
+ *
+ * Note that because the inode is flushed delayed write by background
+ * writeback, the flush lock may already be held here and waiting on it can
+ * result in very long latencies. Hence for sync reclaims, where we wait on the
+ * flush lock, the caller should push out delayed write inodes first before
+ * trying to reclaim them to minimise the amount of time spent waiting. For
+ * background relaim, we just requeue the inode for the next pass.
+ *
+ * Hence the order of actions after gaining the locks should be:
+ * bad => reclaim
+ * shutdown => unpin and reclaim
+ * pinned, delwri => requeue
+ * pinned, sync => unpin
+ * stale => reclaim
+ * clean => reclaim
+ * dirty, delwri => flush and requeue
+ * dirty, sync => flush, wait and reclaim
+ */
STATIC int
xfs_reclaim_inode(
struct xfs_inode *ip,
struct xfs_perag *pag,
int sync_mode)
{
+ int error = 0;
+
/*
* The radix tree lock here protects a thread in xfs_iget from racing
* with us starting reclaim on the inode. Once we have the
@@ -735,33 +751,70 @@ xfs_reclaim_inode(
spin_unlock(&ip->i_flags_lock);
write_unlock(&pag->pag_ici_lock);
- /*
- * 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.
- */
xfs_ilock(ip, XFS_ILOCK_EXCL);
- xfs_iflock(ip);
+ if (!xfs_iflock_nowait(ip)) {
+ if (!(sync_mode & SYNC_WAIT))
+ goto out;
+ xfs_iflock(ip);
+ }
+
+ if (is_bad_inode(VFS_I(ip)))
+ goto reclaim;
+ if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+ xfs_iunpin_wait(ip);
+ goto reclaim;
+ }
+ if (xfs_ipincount(ip)) {
+ if (!(sync_mode & SYNC_WAIT)) {
+ xfs_ifunlock(ip);
+ goto out;
+ }
+ xfs_iunpin_wait(ip);
+ }
+ if (xfs_iflags_test(ip, XFS_ISTALE))
+ goto reclaim;
+ if (xfs_inode_clean(ip))
+ goto reclaim;
+
+ /* Now we have an inode that needs flushing */
+ error = xfs_iflush(ip, sync_mode);
+ if (sync_mode & SYNC_WAIT) {
+ xfs_iflock(ip);
+ goto reclaim;
+ }
/*
- * In the case of a forced shutdown we rely on xfs_iflush() to
- * wait for the inode to be unpinned before returning an error.
+ * When we have to flush an inode but don't have SYNC_WAIT set, we
+ * flush the inode out using a delwri buffer and wait for the next
+ * call into reclaim to find it in a clean state instead of waiting for
+ * it now. We also don't return errors here - if the error is transient
+ * then the next reclaim pass will flush the inode, and if the error
+ * is permanent then the next sync reclaim will reclaim the inode and
+ * pass on the 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);
+ if (error && error != EAGAIN && !XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+ xfs_fs_cmn_err(CE_WARN, ip->i_mount,
+ "inode 0x%llx background reclaim flush failed with %d",
+ (long long)ip->i_ino, error);
}
+out:
+ xfs_iflags_clear(ip, XFS_IRECLAIM);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ /*
+ * We could return EAGAIN here to make reclaim rescan the inode tree in
+ * a short while. However, this just burns CPU time scanning the tree
+ * waiting for IO to complete and xfssyncd never goes back to the idle
+ * state. Instead, return 0 to let the next scheduled background reclaim
+ * attempt to reclaim the inode again.
+ */
+ return 0;
+reclaim:
+ xfs_ifunlock(ip);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_ireclaim(ip);
- return 0;
+ return error;
+
}
int
@@ -770,5 +823,88 @@ xfs_reclaim_inodes(
int mode)
{
return xfs_inode_ag_iterator(mp, xfs_reclaim_inode, mode,
- XFS_ICI_RECLAIM_TAG, 1);
+ XFS_ICI_RECLAIM_TAG, 1, NULL);
+}
+
+/*
+ * Shrinker infrastructure.
+ *
+ * This is all far more complex than it needs to be. It adds a global list of
+ * mounts because the shrinkers can only call a global context. We need to make
+ * the shrinkers pass a context to avoid the need for global state.
+ */
+static LIST_HEAD(xfs_mount_list);
+static struct rw_semaphore xfs_mount_list_lock;
+
+static int
+xfs_reclaim_inode_shrink(
+ int nr_to_scan,
+ gfp_t gfp_mask)
+{
+ struct xfs_mount *mp;
+ struct xfs_perag *pag;
+ xfs_agnumber_t ag;
+ int reclaimable = 0;
+
+ if (nr_to_scan) {
+ if (!(gfp_mask & __GFP_FS))
+ return -1;
+
+ down_read(&xfs_mount_list_lock);
+ list_for_each_entry(mp, &xfs_mount_list, m_mplist) {
+ xfs_inode_ag_iterator(mp, xfs_reclaim_inode, 0,
+ XFS_ICI_RECLAIM_TAG, 1, &nr_to_scan);
+ if (nr_to_scan <= 0)
+ break;
+ }
+ up_read(&xfs_mount_list_lock);
+ }
+
+ down_read(&xfs_mount_list_lock);
+ list_for_each_entry(mp, &xfs_mount_list, m_mplist) {
+ for (ag = 0; ag < mp->m_sb.sb_agcount; ag++) {
+ pag = xfs_perag_get(mp, ag);
+ reclaimable += pag->pag_ici_reclaimable;
+ xfs_perag_put(pag);
+ }
+ }
+ up_read(&xfs_mount_list_lock);
+ return reclaimable;
+}
+
+static struct shrinker xfs_inode_shrinker = {
+ .shrink = xfs_reclaim_inode_shrink,
+ .seeks = DEFAULT_SEEKS,
+};
+
+void __init
+xfs_inode_shrinker_init(void)
+{
+ init_rwsem(&xfs_mount_list_lock);
+ register_shrinker(&xfs_inode_shrinker);
+}
+
+void
+xfs_inode_shrinker_destroy(void)
+{
+ ASSERT(list_empty(&xfs_mount_list));
+ unregister_shrinker(&xfs_inode_shrinker);
+}
+
+void
+xfs_inode_shrinker_register(
+ struct xfs_mount *mp)
+{
+ down_write(&xfs_mount_list_lock);
+ list_add_tail(&mp->m_mplist, &xfs_mount_list);
+ up_write(&xfs_mount_list_lock);
+}
+
+void
+xfs_inode_shrinker_unregister(
+ struct xfs_mount *mp)
+{
+ down_write(&xfs_mount_list_lock);
+ list_del(&mp->m_mplist);
+ up_write(&xfs_mount_list_lock);
}
Powered by cgit, Gruvboxed by Blaster4385.