1 files changed, 147 insertions, 212 deletions

diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 96c997ed2ec8..322eb2ee6c55 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -39,13 +39,6 @@ STATIC int
 xlog_clear_stale_blocks(
 	struct xlog	*,
 	xfs_lsn_t);
-#if defined(DEBUG)
-STATIC void
-xlog_recover_check_summary(
-	struct xlog *);
-#else
-#define	xlog_recover_check_summary(log)
-#endif
 STATIC int
 xlog_do_recovery_pass(
         struct xlog *, xfs_daddr_t, xfs_daddr_t, int, xfs_daddr_t *);
@@ -129,7 +122,7 @@ xlog_do_io(
 	xfs_daddr_t		blk_no,
 	unsigned int		nbblks,
 	char			*data,
-	unsigned int		op)
+	enum req_op		op)
 {
 	int			error;
 
@@ -1800,6 +1793,8 @@ static const struct xlog_recover_item_ops *xlog_recover_item_ops[] = {
 	&xlog_cud_item_ops,
 	&xlog_bui_item_ops,
 	&xlog_bud_item_ops,
+	&xlog_attri_item_ops,
+	&xlog_attrd_item_ops,
 };
 
 static const struct xlog_recover_item_ops *
@@ -2485,7 +2480,7 @@ xlog_finish_defer_ops(
 		error = xfs_trans_alloc(mp, &resv, dfc->dfc_blkres,
 				dfc->dfc_rtxres, XFS_TRANS_RESERVE, &tp);
 		if (error) {
-			xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
+			xlog_force_shutdown(mp->m_log, SHUTDOWN_LOG_IO_ERROR);
 			return error;
 		}
 
@@ -2519,21 +2514,22 @@ xlog_abort_defer_ops(
 		xfs_defer_ops_capture_free(mp, dfc);
 	}
 }
+
 /*
  * When this is called, all of the log intent items which did not have
- * corresponding log done items should be in the AIL.  What we do now
- * is update the data structures associated with each one.
+ * corresponding log done items should be in the AIL.  What we do now is update
+ * the data structures associated with each one.
  *
- * Since we process the log intent items in normal transactions, they
- * will be removed at some point after the commit.  This prevents us
- * from just walking down the list processing each one.  We'll use a
- * flag in the intent item to skip those that we've already processed
- * and use the AIL iteration mechanism's generation count to try to
- * speed this up at least a bit.
+ * Since we process the log intent items in normal transactions, they will be
+ * removed at some point after the commit.  This prevents us from just walking
+ * down the list processing each one.  We'll use a flag in the intent item to
+ * skip those that we've already processed and use the AIL iteration mechanism's
+ * generation count to try to speed this up at least a bit.
  *
- * When we start, we know that the intents are the only things in the
- * AIL.  As we process them, however, other items are added to the
- * AIL.
+ * When we start, we know that the intents are the only things in the AIL. As we
+ * process them, however, other items are added to the AIL. Hence we know we
+ * have started recovery on all the pending intents when we find an non-intent
+ * item in the AIL.
  */
 STATIC int
 xlog_recover_process_intents(
@@ -2556,17 +2552,10 @@ xlog_recover_process_intents(
 	for (lip = xfs_trans_ail_cursor_first(ailp, &cur, 0);
 	     lip != NULL;
 	     lip = xfs_trans_ail_cursor_next(ailp, &cur)) {
-		/*
-		 * We're done when we see something other than an intent.
-		 * There should be no intents left in the AIL now.
-		 */
-		if (!xlog_item_is_intent(lip)) {
-#ifdef DEBUG
-			for (; lip; lip = xfs_trans_ail_cursor_next(ailp, &cur))
-				ASSERT(!xlog_item_is_intent(lip));
-#endif
+		const struct xfs_item_ops	*ops;
+
+		if (!xlog_item_is_intent(lip))
 			break;
-		}
 
 		/*
 		 * We should never see a redo item with a LSN higher than
@@ -2580,13 +2569,17 @@ xlog_recover_process_intents(
 		 * deferred ops, you /must/ attach them to the capture list in
 		 * the recover routine or else those subsequent intents will be
 		 * replayed in the wrong order!
+		 *
+		 * The recovery function can free the log item, so we must not
+		 * access lip after it returns.
 		 */
 		spin_unlock(&ailp->ail_lock);
-		error = lip->li_ops->iop_recover(lip, &capture_list);
+		ops = lip->li_ops;
+		error = ops->iop_recover(lip, &capture_list);
 		spin_lock(&ailp->ail_lock);
 		if (error) {
 			trace_xlog_intent_recovery_failed(log->l_mp, error,
-					lip->li_ops->iop_recover);
+					ops->iop_recover);
 			break;
 		}
 	}
@@ -2607,8 +2600,9 @@ err:
 }
 
 /*
- * A cancel occurs when the mount has failed and we're bailing out.
- * Release all pending log intent items so they don't pin the AIL.
+ * A cancel occurs when the mount has failed and we're bailing out.  Release all
+ * pending log intent items that we haven't started recovery on so they don't
+ * pin the AIL.
  */
 STATIC void
 xlog_recover_cancel_intents(
@@ -2622,17 +2616,8 @@ xlog_recover_cancel_intents(
 	spin_lock(&ailp->ail_lock);
 	lip = xfs_trans_ail_cursor_first(ailp, &cur, 0);
 	while (lip != NULL) {
-		/*
-		 * We're done when we see something other than an intent.
-		 * There should be no intents left in the AIL now.
-		 */
-		if (!xlog_item_is_intent(lip)) {
-#ifdef DEBUG
-			for (; lip; lip = xfs_trans_ail_cursor_next(ailp, &cur))
-				ASSERT(!xlog_item_is_intent(lip));
-#endif
+		if (!xlog_item_is_intent(lip))
 			break;
-		}
 
 		spin_unlock(&ailp->ail_lock);
 		lip->li_ops->iop_release(lip);
@@ -2650,21 +2635,21 @@ xlog_recover_cancel_intents(
  */
 STATIC void
 xlog_recover_clear_agi_bucket(
-	xfs_mount_t	*mp,
-	xfs_agnumber_t	agno,
-	int		bucket)
+	struct xfs_perag	*pag,
+	int			bucket)
 {
-	xfs_trans_t	*tp;
-	xfs_agi_t	*agi;
-	struct xfs_buf	*agibp;
-	int		offset;
-	int		error;
+	struct xfs_mount	*mp = pag->pag_mount;
+	struct xfs_trans	*tp;
+	struct xfs_agi		*agi;
+	struct xfs_buf		*agibp;
+	int			offset;
+	int			error;
 
 	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_clearagi, 0, 0, 0, &tp);
 	if (error)
 		goto out_error;
 
-	error = xfs_read_agi(mp, tp, agno, &agibp);
+	error = xfs_read_agi(pag, tp, &agibp);
 	if (error)
 		goto out_abort;
 
@@ -2683,60 +2668,62 @@ xlog_recover_clear_agi_bucket(
 out_abort:
 	xfs_trans_cancel(tp);
 out_error:
-	xfs_warn(mp, "%s: failed to clear agi %d. Continuing.", __func__, agno);
+	xfs_warn(mp, "%s: failed to clear agi %d. Continuing.", __func__,
+			pag->pag_agno);
 	return;
 }
 
-STATIC xfs_agino_t
-xlog_recover_process_one_iunlink(
-	struct xfs_mount		*mp,
-	xfs_agnumber_t			agno,
-	xfs_agino_t			agino,
-	int				bucket)
+static int
+xlog_recover_iunlink_bucket(
+	struct xfs_perag	*pag,
+	struct xfs_agi		*agi,
+	int			bucket)
 {
-	struct xfs_buf			*ibp;
-	struct xfs_dinode		*dip;
-	struct xfs_inode		*ip;
-	xfs_ino_t			ino;
-	int				error;
-
-	ino = XFS_AGINO_TO_INO(mp, agno, agino);
-	error = xfs_iget(mp, NULL, ino, 0, 0, &ip);
-	if (error)
-		goto fail;
+	struct xfs_mount	*mp = pag->pag_mount;
+	struct xfs_inode	*prev_ip = NULL;
+	struct xfs_inode	*ip;
+	xfs_agino_t		prev_agino, agino;
+	int			error = 0;
 
-	/*
-	 * Get the on disk inode to find the next inode in the bucket.
-	 */
-	error = xfs_imap_to_bp(mp, NULL, &ip->i_imap, &ibp);
-	if (error)
-		goto fail_iput;
-	dip = xfs_buf_offset(ibp, ip->i_imap.im_boffset);
+	agino = be32_to_cpu(agi->agi_unlinked[bucket]);
+	while (agino != NULLAGINO) {
+		error = xfs_iget(mp, NULL,
+				XFS_AGINO_TO_INO(mp, pag->pag_agno, agino),
+				0, 0, &ip);
+		if (error)
+			break;
 
-	xfs_iflags_clear(ip, XFS_IRECOVERY);
-	ASSERT(VFS_I(ip)->i_nlink == 0);
-	ASSERT(VFS_I(ip)->i_mode != 0);
+		ASSERT(VFS_I(ip)->i_nlink == 0);
+		ASSERT(VFS_I(ip)->i_mode != 0);
+		xfs_iflags_clear(ip, XFS_IRECOVERY);
+		agino = ip->i_next_unlinked;
 
-	/* setup for the next pass */
-	agino = be32_to_cpu(dip->di_next_unlinked);
-	xfs_buf_relse(ibp);
+		if (prev_ip) {
+			ip->i_prev_unlinked = prev_agino;
+			xfs_irele(prev_ip);
 
-	xfs_irele(ip);
-	return agino;
+			/*
+			 * Ensure the inode is removed from the unlinked list
+			 * before we continue so that it won't race with
+			 * building the in-memory list here. This could be
+			 * serialised with the agibp lock, but that just
+			 * serialises via lockstepping and it's much simpler
+			 * just to flush the inodegc queue and wait for it to
+			 * complete.
+			 */
+			xfs_inodegc_flush(mp);
+		}
 
- fail_iput:
-	xfs_irele(ip);
- fail:
-	/*
-	 * We can't read in the inode this bucket points to, or this inode
-	 * is messed up.  Just ditch this bucket of inodes.  We will lose
-	 * some inodes and space, but at least we won't hang.
-	 *
-	 * Call xlog_recover_clear_agi_bucket() to perform a transaction to
-	 * clear the inode pointer in the bucket.
-	 */
-	xlog_recover_clear_agi_bucket(mp, agno, bucket);
-	return NULLAGINO;
+		prev_agino = agino;
+		prev_ip = ip;
+	}
+
+	if (prev_ip) {
+		ip->i_prev_unlinked = prev_agino;
+		xfs_irele(prev_ip);
+	}
+	xfs_inodegc_flush(mp);
+	return error;
 }
 
 /*
@@ -2762,59 +2749,70 @@ xlog_recover_process_one_iunlink(
  * scheduled on this CPU to ensure other scheduled work can run without undue
  * latency.
  */
-STATIC void
-xlog_recover_process_iunlinks(
-	struct xlog	*log)
+static void
+xlog_recover_iunlink_ag(
+	struct xfs_perag	*pag)
 {
-	struct xfs_mount	*mp = log->l_mp;
-	struct xfs_perag	*pag;
-	xfs_agnumber_t		agno;
 	struct xfs_agi		*agi;
 	struct xfs_buf		*agibp;
-	xfs_agino_t		agino;
 	int			bucket;
 	int			error;
 
-	for_each_perag(mp, agno, pag) {
-		error = xfs_read_agi(mp, NULL, pag->pag_agno, &agibp);
+	error = xfs_read_agi(pag, NULL, &agibp);
+	if (error) {
+		/*
+		 * AGI is b0rked. Don't process it.
+		 *
+		 * We should probably mark the filesystem as corrupt after we've
+		 * recovered all the ag's we can....
+		 */
+		return;
+	}
+
+	/*
+	 * Unlock the buffer so that it can be acquired in the normal course of
+	 * the transaction to truncate and free each inode.  Because we are not
+	 * racing with anyone else here for the AGI buffer, we don't even need
+	 * to hold it locked to read the initial unlinked bucket entries out of
+	 * the buffer. We keep buffer reference though, so that it stays pinned
+	 * in memory while we need the buffer.
+	 */
+	agi = agibp->b_addr;
+	xfs_buf_unlock(agibp);
+
+	for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++) {
+		error = xlog_recover_iunlink_bucket(pag, agi, bucket);
 		if (error) {
 			/*
-			 * AGI is b0rked. Don't process it.
-			 *
-			 * We should probably mark the filesystem as corrupt
-			 * after we've recovered all the ag's we can....
+			 * Bucket is unrecoverable, so only a repair scan can
+			 * free the remaining unlinked inodes. Just empty the
+			 * bucket and remaining inodes on it unreferenced and
+			 * unfreeable.
 			 */
-			continue;
-		}
-		/*
-		 * Unlock the buffer so that it can be acquired in the normal
-		 * course of the transaction to truncate and free each inode.
-		 * Because we are not racing with anyone else here for the AGI
-		 * buffer, we don't even need to hold it locked to read the
-		 * initial unlinked bucket entries out of the buffer. We keep
-		 * buffer reference though, so that it stays pinned in memory
-		 * while we need the buffer.
-		 */
-		agi = agibp->b_addr;
-		xfs_buf_unlock(agibp);
-
-		for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++) {
-			agino = be32_to_cpu(agi->agi_unlinked[bucket]);
-			while (agino != NULLAGINO) {
-				agino = xlog_recover_process_one_iunlink(mp,
-						pag->pag_agno, agino, bucket);
-				cond_resched();
-			}
+			xfs_inodegc_flush(pag->pag_mount);
+			xlog_recover_clear_agi_bucket(pag, bucket);
 		}
-		xfs_buf_rele(agibp);
 	}
 
+	xfs_buf_rele(agibp);
+}
+
+static void
+xlog_recover_process_iunlinks(
+	struct xlog	*log)
+{
+	struct xfs_perag	*pag;
+	xfs_agnumber_t		agno;
+
+	for_each_perag(log->l_mp, agno, pag)
+		xlog_recover_iunlink_ag(pag);
+
 	/*
 	 * Flush the pending unlinked inodes to ensure that the inactivations
 	 * are fully completed on disk and the incore inodes can be reclaimed
 	 * before we signal that recovery is complete.
 	 */
-	xfs_inodegc_flush(mp);
+	xfs_inodegc_flush(log->l_mp);
 }
 
 STATIC void
@@ -3244,7 +3242,7 @@ xlog_do_log_recovery(
 	xfs_daddr_t	head_blk,
 	xfs_daddr_t	tail_blk)
 {
-	int		error, i;
+	int		error;
 
 	ASSERT(head_blk != tail_blk);
 
@@ -3252,37 +3250,25 @@ xlog_do_log_recovery(
 	 * First do a pass to find all of the cancelled buf log items.
 	 * Store them in the buf_cancel_table for use in the second pass.
 	 */
-	log->l_buf_cancel_table = kmem_zalloc(XLOG_BC_TABLE_SIZE *
-						 sizeof(struct list_head),
-						 0);
-	for (i = 0; i < XLOG_BC_TABLE_SIZE; i++)
-		INIT_LIST_HEAD(&log->l_buf_cancel_table[i]);
+	error = xlog_alloc_buf_cancel_table(log);
+	if (error)
+		return error;
 
 	error = xlog_do_recovery_pass(log, head_blk, tail_blk,
 				      XLOG_RECOVER_PASS1, NULL);
-	if (error != 0) {
-		kmem_free(log->l_buf_cancel_table);
-		log->l_buf_cancel_table = NULL;
-		return error;
-	}
+	if (error != 0)
+		goto out_cancel;
+
 	/*
 	 * Then do a second pass to actually recover the items in the log.
 	 * When it is complete free the table of buf cancel items.
 	 */
 	error = xlog_do_recovery_pass(log, head_blk, tail_blk,
 				      XLOG_RECOVER_PASS2, NULL);
-#ifdef DEBUG
-	if (!error) {
-		int	i;
-
-		for (i = 0; i < XLOG_BC_TABLE_SIZE; i++)
-			ASSERT(list_empty(&log->l_buf_cancel_table[i]));
-	}
-#endif	/* DEBUG */
-
-	kmem_free(log->l_buf_cancel_table);
-	log->l_buf_cancel_table = NULL;
-
+	if (!error)
+		xlog_check_buf_cancel_table(log);
+out_cancel:
+	xlog_free_buf_cancel_table(log);
 	return error;
 }
 
@@ -3346,15 +3332,14 @@ xlog_do_recover(
 	/* re-initialise in-core superblock and geometry structures */
 	mp->m_features |= xfs_sb_version_to_features(sbp);
 	xfs_reinit_percpu_counters(mp);
-	error = xfs_initialize_perag(mp, sbp->sb_agcount, &mp->m_maxagi);
+	error = xfs_initialize_perag(mp, sbp->sb_agcount, sbp->sb_dblocks,
+			&mp->m_maxagi);
 	if (error) {
 		xfs_warn(mp, "Failed post-recovery per-ag init: %d", error);
 		return error;
 	}
 	mp->m_alloc_set_aside = xfs_alloc_set_aside(mp);
 
-	xlog_recover_check_summary(log);
-
 	/* Normal transactions can now occur */
 	clear_bit(XLOG_ACTIVE_RECOVERY, &log->l_opstate);
 	return 0;
@@ -3470,7 +3455,7 @@ xlog_recover_finish(
 		 */
 		xlog_recover_cancel_intents(log);
 		xfs_alert(log->l_mp, "Failed to recover intents");
-		xfs_force_shutdown(log->l_mp, SHUTDOWN_LOG_IO_ERROR);
+		xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR);
 		return error;
 	}
 
@@ -3497,7 +3482,6 @@ xlog_recover_finish(
 	}
 
 	xlog_recover_process_iunlinks(log);
-	xlog_recover_check_summary(log);
 
 	/*
 	 * Recover any CoW staging blocks that are still referenced by the
@@ -3517,7 +3501,7 @@ xlog_recover_finish(
 		 * end of intents processing can be pushed through the CIL
 		 * and AIL.
 		 */
-		xfs_force_shutdown(log->l_mp, SHUTDOWN_LOG_IO_ERROR);
+		xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR);
 	}
 
 	return 0;
@@ -3531,52 +3515,3 @@ xlog_recover_cancel(
 		xlog_recover_cancel_intents(log);
 }
 
-#if defined(DEBUG)
-/*
- * Read all of the agf and agi counters and check that they
- * are consistent with the superblock counters.
- */
-STATIC void
-xlog_recover_check_summary(
-	struct xlog		*log)
-{
-	struct xfs_mount	*mp = log->l_mp;
-	struct xfs_perag	*pag;
-	struct xfs_buf		*agfbp;
-	struct xfs_buf		*agibp;
-	xfs_agnumber_t		agno;
-	uint64_t		freeblks;
-	uint64_t		itotal;
-	uint64_t		ifree;
-	int			error;
-
-	freeblks = 0LL;
-	itotal = 0LL;
-	ifree = 0LL;
-	for_each_perag(mp, agno, pag) {
-		error = xfs_read_agf(mp, NULL, pag->pag_agno, 0, &agfbp);
-		if (error) {
-			xfs_alert(mp, "%s agf read failed agno %d error %d",
-						__func__, pag->pag_agno, error);
-		} else {
-			struct xfs_agf	*agfp = agfbp->b_addr;
-
-			freeblks += be32_to_cpu(agfp->agf_freeblks) +
-				    be32_to_cpu(agfp->agf_flcount);
-			xfs_buf_relse(agfbp);
-		}
-
-		error = xfs_read_agi(mp, NULL, pag->pag_agno, &agibp);
-		if (error) {
-			xfs_alert(mp, "%s agi read failed agno %d error %d",
-						__func__, pag->pag_agno, error);
-		} else {
-			struct xfs_agi	*agi = agibp->b_addr;
-
-			itotal += be32_to_cpu(agi->agi_count);
-			ifree += be32_to_cpu(agi->agi_freecount);
-			xfs_buf_relse(agibp);
-		}
-	}
-}
-#endif /* DEBUG */