1 files changed, 139 insertions, 26 deletions

diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 67dec11e34c7..0ac717aad380 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -25,7 +25,7 @@
 #include "xfs_dquot.h"
 #include "xfs_icache.h"
 
-kmem_zone_t	*xfs_trans_zone;
+struct kmem_cache	*xfs_trans_cache;
 
 #if defined(CONFIG_TRACEPOINTS)
 static void
@@ -76,7 +76,7 @@ xfs_trans_free(
 	if (!(tp->t_flags & XFS_TRANS_NO_WRITECOUNT))
 		sb_end_intwrite(tp->t_mountp->m_super);
 	xfs_trans_free_dqinfo(tp);
-	kmem_cache_free(xfs_trans_zone, tp);
+	kmem_cache_free(xfs_trans_cache, tp);
 }
 
 /*
@@ -95,7 +95,7 @@ xfs_trans_dup(
 
 	trace_xfs_trans_dup(tp, _RET_IP_);
 
-	ntp = kmem_cache_zalloc(xfs_trans_zone, GFP_KERNEL | __GFP_NOFAIL);
+	ntp = kmem_cache_zalloc(xfs_trans_cache, GFP_KERNEL | __GFP_NOFAIL);
 
 	/*
 	 * Initialize the new transaction structure.
@@ -263,7 +263,7 @@ xfs_trans_alloc(
 	 * by doing GFP_KERNEL allocations inside sb_start_intwrite().
 	 */
 retry:
-	tp = kmem_cache_zalloc(xfs_trans_zone, GFP_KERNEL | __GFP_NOFAIL);
+	tp = kmem_cache_zalloc(xfs_trans_cache, GFP_KERNEL | __GFP_NOFAIL);
 	if (!(flags & XFS_TRANS_NO_WRITECOUNT))
 		sb_start_intwrite(mp->m_super);
 	xfs_trans_set_context(tp);
@@ -477,7 +477,7 @@ STATIC void
 xfs_trans_apply_sb_deltas(
 	xfs_trans_t	*tp)
 {
-	xfs_dsb_t	*sbp;
+	struct xfs_dsb	*sbp;
 	struct xfs_buf	*bp;
 	int		whole = 0;
 
@@ -541,14 +541,14 @@ xfs_trans_apply_sb_deltas(
 		/*
 		 * Log the whole thing, the fields are noncontiguous.
 		 */
-		xfs_trans_log_buf(tp, bp, 0, sizeof(xfs_dsb_t) - 1);
+		xfs_trans_log_buf(tp, bp, 0, sizeof(struct xfs_dsb) - 1);
 	else
 		/*
 		 * Since all the modifiable fields are contiguous, we
 		 * can get away with this.
 		 */
-		xfs_trans_log_buf(tp, bp, offsetof(xfs_dsb_t, sb_icount),
-				  offsetof(xfs_dsb_t, sb_frextents) +
+		xfs_trans_log_buf(tp, bp, offsetof(struct xfs_dsb, sb_icount),
+				  offsetof(struct xfs_dsb, sb_frextents) +
 				  sizeof(sbp->sb_frextents) - 1);
 }
 
@@ -648,7 +648,7 @@ xfs_trans_add_item(
 	struct xfs_trans	*tp,
 	struct xfs_log_item	*lip)
 {
-	ASSERT(lip->li_mountp == tp->t_mountp);
+	ASSERT(lip->li_log == tp->t_mountp->m_log);
 	ASSERT(lip->li_ailp == tp->t_mountp->m_ail);
 	ASSERT(list_empty(&lip->li_trans));
 	ASSERT(!test_bit(XFS_LI_DIRTY, &lip->li_flags));
@@ -775,7 +775,7 @@ xfs_trans_committed_bulk(
 		 * object into the AIL as we are in a shutdown situation.
 		 */
 		if (aborted) {
-			ASSERT(xfs_is_shutdown(ailp->ail_mount));
+			ASSERT(xlog_is_shutdown(ailp->ail_log));
 			if (lip->li_ops->iop_unpin)
 				lip->li_ops->iop_unpin(lip, 1);
 			continue;
@@ -836,6 +836,7 @@ __xfs_trans_commit(
 	bool			regrant)
 {
 	struct xfs_mount	*mp = tp->t_mountp;
+	struct xlog		*log = mp->m_log;
 	xfs_csn_t		commit_seq = 0;
 	int			error = 0;
 	int			sync = tp->t_flags & XFS_TRANS_SYNC;
@@ -864,7 +865,13 @@ __xfs_trans_commit(
 	if (!(tp->t_flags & XFS_TRANS_DIRTY))
 		goto out_unreserve;
 
-	if (xfs_is_shutdown(mp)) {
+	/*
+	 * We must check against log shutdown here because we cannot abort log
+	 * items and leave them dirty, inconsistent and unpinned in memory while
+	 * the log is active. This leaves them open to being written back to
+	 * disk, and that will lead to on-disk corruption.
+	 */
+	if (xlog_is_shutdown(log)) {
 		error = -EIO;
 		goto out_unreserve;
 	}
@@ -878,7 +885,7 @@ __xfs_trans_commit(
 		xfs_trans_apply_sb_deltas(tp);
 	xfs_trans_apply_dquot_deltas(tp);
 
-	xlog_cil_commit(mp->m_log, tp, &commit_seq, regrant);
+	xlog_cil_commit(log, tp, &commit_seq, regrant);
 
 	xfs_trans_free(tp);
 
@@ -905,10 +912,10 @@ out_unreserve:
 	 */
 	xfs_trans_unreserve_and_mod_dquots(tp);
 	if (tp->t_ticket) {
-		if (regrant && !xlog_is_shutdown(mp->m_log))
-			xfs_log_ticket_regrant(mp->m_log, tp->t_ticket);
+		if (regrant && !xlog_is_shutdown(log))
+			xfs_log_ticket_regrant(log, tp->t_ticket);
 		else
-			xfs_log_ticket_ungrant(mp->m_log, tp->t_ticket);
+			xfs_log_ticket_ungrant(log, tp->t_ticket);
 		tp->t_ticket = NULL;
 	}
 	xfs_trans_free_items(tp, !!error);
@@ -926,36 +933,56 @@ xfs_trans_commit(
 }
 
 /*
- * Unlock all of the transaction's items and free the transaction.
- * The transaction must not have modified any of its items, because
- * there is no way to restore them to their previous state.
+ * Unlock all of the transaction's items and free the transaction.  If the
+ * transaction is dirty, we must shut down the filesystem because there is no
+ * way to restore them to their previous state.
  *
- * If the transaction has made a log reservation, make sure to release
- * it as well.
+ * If the transaction has made a log reservation, make sure to release it as
+ * well.
+ *
+ * This is a high level function (equivalent to xfs_trans_commit()) and so can
+ * be called after the transaction has effectively been aborted due to the mount
+ * being shut down. However, if the mount has not been shut down and the
+ * transaction is dirty we will shut the mount down and, in doing so, that
+ * guarantees that the log is shut down, too. Hence we don't need to be as
+ * careful with shutdown state and dirty items here as we need to be in
+ * xfs_trans_commit().
  */
 void
 xfs_trans_cancel(
 	struct xfs_trans	*tp)
 {
 	struct xfs_mount	*mp = tp->t_mountp;
+	struct xlog		*log = mp->m_log;
 	bool			dirty = (tp->t_flags & XFS_TRANS_DIRTY);
 
 	trace_xfs_trans_cancel(tp, _RET_IP_);
 
-	if (tp->t_flags & XFS_TRANS_PERM_LOG_RES)
+	/*
+	 * It's never valid to cancel a transaction with deferred ops attached,
+	 * because the transaction is effectively dirty.  Complain about this
+	 * loudly before freeing the in-memory defer items.
+	 */
+	if (!list_empty(&tp->t_dfops)) {
+		ASSERT(xfs_is_shutdown(mp) || list_empty(&tp->t_dfops));
+		ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
+		dirty = true;
 		xfs_defer_cancel(tp);
+	}
 
 	/*
-	 * See if the caller is relying on us to shut down the
-	 * filesystem.  This happens in paths where we detect
-	 * corruption and decide to give up.
+	 * See if the caller is relying on us to shut down the filesystem. We
+	 * only want an error report if there isn't already a shutdown in
+	 * progress, so we only need to check against the mount shutdown state
+	 * here.
 	 */
 	if (dirty && !xfs_is_shutdown(mp)) {
 		XFS_ERROR_REPORT("xfs_trans_cancel", XFS_ERRLEVEL_LOW, mp);
 		xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
 	}
 #ifdef DEBUG
-	if (!dirty && !xfs_is_shutdown(mp)) {
+	/* Log items need to be consistent until the log is shut down. */
+	if (!dirty && !xlog_is_shutdown(log)) {
 		struct xfs_log_item *lip;
 
 		list_for_each_entry(lip, &tp->t_items, li_trans)
@@ -966,7 +993,7 @@ xfs_trans_cancel(
 	xfs_trans_unreserve_and_mod_dquots(tp);
 
 	if (tp->t_ticket) {
-		xfs_log_ticket_ungrant(mp->m_log, tp->t_ticket);
+		xfs_log_ticket_ungrant(log, tp->t_ticket);
 		tp->t_ticket = NULL;
 	}
 
@@ -1201,3 +1228,89 @@ out_cancel:
 	xfs_trans_cancel(tp);
 	return error;
 }
+
+/*
+ * Allocate an transaction, lock and join the directory and child inodes to it,
+ * and reserve quota for a directory update.  If there isn't sufficient space,
+ * @dblocks will be set to zero for a reservationless directory update and
+ * @nospace_error will be set to a negative errno describing the space
+ * constraint we hit.
+ *
+ * The caller must ensure that the on-disk dquots attached to this inode have
+ * already been allocated and initialized.  The ILOCKs will be dropped when the
+ * transaction is committed or cancelled.
+ */
+int
+xfs_trans_alloc_dir(
+	struct xfs_inode	*dp,
+	struct xfs_trans_res	*resv,
+	struct xfs_inode	*ip,
+	unsigned int		*dblocks,
+	struct xfs_trans	**tpp,
+	int			*nospace_error)
+{
+	struct xfs_trans	*tp;
+	struct xfs_mount	*mp = ip->i_mount;
+	unsigned int		resblks;
+	bool			retried = false;
+	int			error;
+
+retry:
+	*nospace_error = 0;
+	resblks = *dblocks;
+	error = xfs_trans_alloc(mp, resv, resblks, 0, 0, &tp);
+	if (error == -ENOSPC) {
+		*nospace_error = error;
+		resblks = 0;
+		error = xfs_trans_alloc(mp, resv, resblks, 0, 0, &tp);
+	}
+	if (error)
+		return error;
+
+	xfs_lock_two_inodes(dp, XFS_ILOCK_EXCL, ip, XFS_ILOCK_EXCL);
+
+	xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+
+	error = xfs_qm_dqattach_locked(dp, false);
+	if (error) {
+		/* Caller should have allocated the dquots! */
+		ASSERT(error != -ENOENT);
+		goto out_cancel;
+	}
+
+	error = xfs_qm_dqattach_locked(ip, false);
+	if (error) {
+		/* Caller should have allocated the dquots! */
+		ASSERT(error != -ENOENT);
+		goto out_cancel;
+	}
+
+	if (resblks == 0)
+		goto done;
+
+	error = xfs_trans_reserve_quota_nblks(tp, dp, resblks, 0, false);
+	if (error == -EDQUOT || error == -ENOSPC) {
+		if (!retried) {
+			xfs_trans_cancel(tp);
+			xfs_blockgc_free_quota(dp, 0);
+			retried = true;
+			goto retry;
+		}
+
+		*nospace_error = error;
+		resblks = 0;
+		error = 0;
+	}
+	if (error)
+		goto out_cancel;
+
+done:
+	*tpp = tp;
+	*dblocks = resblks;
+	return 0;
+
+out_cancel:
+	xfs_trans_cancel(tp);
+	return error;
+}