1 files changed, 201 insertions, 171 deletions

diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index 3e61d0cc23f8..13851c0d640b 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -17,105 +17,149 @@
 #include "xfs_fsops.h"
 #include "xfs_trans_space.h"
 #include "xfs_log.h"
+#include "xfs_log_priv.h"
 #include "xfs_ag.h"
 #include "xfs_ag_resv.h"
+#include "xfs_trace.h"
+
+/*
+ * Write new AG headers to disk. Non-transactional, but need to be
+ * written and completed prior to the growfs transaction being logged.
+ * To do this, we use a delayed write buffer list and wait for
+ * submission and IO completion of the list as a whole. This allows the
+ * IO subsystem to merge all the AG headers in a single AG into a single
+ * IO and hide most of the latency of the IO from us.
+ *
+ * This also means that if we get an error whilst building the buffer
+ * list to write, we can cancel the entire list without having written
+ * anything.
+ */
+static int
+xfs_resizefs_init_new_ags(
+	struct xfs_trans	*tp,
+	struct aghdr_init_data	*id,
+	xfs_agnumber_t		oagcount,
+	xfs_agnumber_t		nagcount,
+	xfs_rfsblock_t		delta,
+	struct xfs_perag	*last_pag,
+	bool			*lastag_extended)
+{
+	struct xfs_mount	*mp = tp->t_mountp;
+	xfs_rfsblock_t		nb = mp->m_sb.sb_dblocks + delta;
+	int			error;
+
+	*lastag_extended = false;
+
+	INIT_LIST_HEAD(&id->buffer_list);
+	for (id->agno = nagcount - 1;
+	     id->agno >= oagcount;
+	     id->agno--, delta -= id->agsize) {
+
+		if (id->agno == nagcount - 1)
+			id->agsize = nb - (id->agno *
+					(xfs_rfsblock_t)mp->m_sb.sb_agblocks);
+		else
+			id->agsize = mp->m_sb.sb_agblocks;
+
+		error = xfs_ag_init_headers(mp, id);
+		if (error) {
+			xfs_buf_delwri_cancel(&id->buffer_list);
+			return error;
+		}
+	}
+
+	error = xfs_buf_delwri_submit(&id->buffer_list);
+	if (error)
+		return error;
+
+	if (delta) {
+		*lastag_extended = true;
+		error = xfs_ag_extend_space(last_pag, tp, delta);
+	}
+	return error;
+}
 
 /*
  * growfs operations
  */
 static int
 xfs_growfs_data_private(
-	xfs_mount_t		*mp,		/* mount point for filesystem */
-	xfs_growfs_data_t	*in)		/* growfs data input struct */
+	struct xfs_mount	*mp,		/* mount point for filesystem */
+	struct xfs_growfs_data	*in)		/* growfs data input struct */
 {
-	xfs_buf_t		*bp;
+	struct xfs_buf		*bp;
 	int			error;
 	xfs_agnumber_t		nagcount;
 	xfs_agnumber_t		nagimax = 0;
-	xfs_rfsblock_t		nb, nb_mod;
-	xfs_rfsblock_t		new;
+	xfs_rfsblock_t		nb, nb_div, nb_mod;
+	int64_t			delta;
+	bool			lastag_extended;
 	xfs_agnumber_t		oagcount;
-	xfs_trans_t		*tp;
+	struct xfs_trans	*tp;
 	struct aghdr_init_data	id = {};
+	struct xfs_perag	*last_pag;
 
 	nb = in->newblocks;
-	if (nb < mp->m_sb.sb_dblocks)
-		return -EINVAL;
-	if ((error = xfs_sb_validate_fsb_count(&mp->m_sb, nb)))
+	error = xfs_sb_validate_fsb_count(&mp->m_sb, nb);
+	if (error)
 		return error;
-	error = xfs_buf_read_uncached(mp->m_ddev_targp,
+
+	if (nb > mp->m_sb.sb_dblocks) {
+		error = xfs_buf_read_uncached(mp->m_ddev_targp,
 				XFS_FSB_TO_BB(mp, nb) - XFS_FSS_TO_BB(mp, 1),
 				XFS_FSS_TO_BB(mp, 1), 0, &bp, NULL);
-	if (error)
-		return error;
-	xfs_buf_relse(bp);
+		if (error)
+			return error;
+		xfs_buf_relse(bp);
+	}
 
-	new = nb;	/* use new as a temporary here */
-	nb_mod = do_div(new, mp->m_sb.sb_agblocks);
-	nagcount = new + (nb_mod != 0);
+	nb_div = nb;
+	nb_mod = do_div(nb_div, mp->m_sb.sb_agblocks);
+	nagcount = nb_div + (nb_mod != 0);
 	if (nb_mod && nb_mod < XFS_MIN_AG_BLOCKS) {
 		nagcount--;
 		nb = (xfs_rfsblock_t)nagcount * mp->m_sb.sb_agblocks;
-		if (nb < mp->m_sb.sb_dblocks)
-			return -EINVAL;
 	}
-	new = nb - mp->m_sb.sb_dblocks;
-	oagcount = mp->m_sb.sb_agcount;
+	delta = nb - mp->m_sb.sb_dblocks;
+	/*
+	 * Reject filesystems with a single AG because they are not
+	 * supported, and reject a shrink operation that would cause a
+	 * filesystem to become unsupported.
+	 */
+	if (delta < 0 && nagcount < 2)
+		return -EINVAL;
 
+	oagcount = mp->m_sb.sb_agcount;
 	/* allocate the new per-ag structures */
 	if (nagcount > oagcount) {
-		error = xfs_initialize_perag(mp, nagcount, &nagimax);
+		error = xfs_initialize_perag(mp, nagcount, nb, &nagimax);
 		if (error)
 			return error;
+	} else if (nagcount < oagcount) {
+		/* TODO: shrinking the entire AGs hasn't yet completed */
+		return -EINVAL;
 	}
 
 	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_growdata,
-			XFS_GROWFS_SPACE_RES(mp), 0, XFS_TRANS_RESERVE, &tp);
+			(delta > 0 ? XFS_GROWFS_SPACE_RES(mp) : -delta), 0,
+			XFS_TRANS_RESERVE, &tp);
 	if (error)
 		return error;
 
-	/*
-	 * Write new AG headers to disk. Non-transactional, but need to be
-	 * written and completed prior to the growfs transaction being logged.
-	 * To do this, we use a delayed write buffer list and wait for
-	 * submission and IO completion of the list as a whole. This allows the
-	 * IO subsystem to merge all the AG headers in a single AG into a single
-	 * IO and hide most of the latency of the IO from us.
-	 *
-	 * This also means that if we get an error whilst building the buffer
-	 * list to write, we can cancel the entire list without having written
-	 * anything.
-	 */
-	INIT_LIST_HEAD(&id.buffer_list);
-	for (id.agno = nagcount - 1;
-	     id.agno >= oagcount;
-	     id.agno--, new -= id.agsize) {
-
-		if (id.agno == nagcount - 1)
-			id.agsize = nb -
-				(id.agno * (xfs_rfsblock_t)mp->m_sb.sb_agblocks);
-		else
-			id.agsize = mp->m_sb.sb_agblocks;
+	last_pag = xfs_perag_get(mp, oagcount - 1);
+	if (delta > 0) {
+		error = xfs_resizefs_init_new_ags(tp, &id, oagcount, nagcount,
+				delta, last_pag, &lastag_extended);
+	} else {
+		xfs_warn_mount(mp, XFS_OPSTATE_WARNED_SHRINK,
+	"EXPERIMENTAL online shrink feature in use. Use at your own risk!");
 
-		error = xfs_ag_init_headers(mp, &id);
-		if (error) {
-			xfs_buf_delwri_cancel(&id.buffer_list);
-			goto out_trans_cancel;
-		}
+		error = xfs_ag_shrink_space(last_pag, &tp, -delta);
 	}
-	error = xfs_buf_delwri_submit(&id.buffer_list);
+	xfs_perag_put(last_pag);
 	if (error)
 		goto out_trans_cancel;
 
-	xfs_trans_agblocks_delta(tp, id.nfree);
-
-	/* If there are new blocks in the old last AG, extend it. */
-	if (new) {
-		error = xfs_ag_extend_space(mp, tp, &id, new);
-		if (error)
-			goto out_trans_cancel;
-	}
-
 	/*
 	 * Update changed superblock fields transactionally. These are not
 	 * seen by the rest of the world until the transaction commit applies
@@ -123,11 +167,19 @@ xfs_growfs_data_private(
 	 */
 	if (nagcount > oagcount)
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_AGCOUNT, nagcount - oagcount);
-	if (nb > mp->m_sb.sb_dblocks)
-		xfs_trans_mod_sb(tp, XFS_TRANS_SB_DBLOCKS,
-				 nb - mp->m_sb.sb_dblocks);
+	if (delta)
+		xfs_trans_mod_sb(tp, XFS_TRANS_SB_DBLOCKS, delta);
 	if (id.nfree)
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_FDBLOCKS, id.nfree);
+
+	/*
+	 * Sync sb counters now to reflect the updated values. This is
+	 * particularly important for shrink because the write verifier
+	 * will fail if sb_fdblocks is ever larger than sb_dblocks.
+	 */
+	if (xfs_has_lazysbcount(mp))
+		xfs_log_sb(tp);
+
 	xfs_trans_set_sync(tp);
 	error = xfs_trans_commit(tp);
 	if (error)
@@ -139,28 +191,29 @@ xfs_growfs_data_private(
 	xfs_set_low_space_thresholds(mp);
 	mp->m_alloc_set_aside = xfs_alloc_set_aside(mp);
 
-	/*
-	 * If we expanded the last AG, free the per-AG reservation
-	 * so we can reinitialize it with the new size.
-	 */
-	if (new) {
-		struct xfs_perag	*pag;
-
-		pag = xfs_perag_get(mp, id.agno);
-		error = xfs_ag_resv_free(pag);
-		xfs_perag_put(pag);
-		if (error)
-			return error;
+	if (delta > 0) {
+		/*
+		 * If we expanded the last AG, free the per-AG reservation
+		 * so we can reinitialize it with the new size.
+		 */
+		if (lastag_extended) {
+			struct xfs_perag	*pag;
+
+			pag = xfs_perag_get(mp, id.agno);
+			error = xfs_ag_resv_free(pag);
+			xfs_perag_put(pag);
+			if (error)
+				return error;
+		}
+		/*
+		 * Reserve AG metadata blocks. ENOSPC here does not mean there
+		 * was a growfs failure, just that there still isn't space for
+		 * new user data after the grow has been run.
+		 */
+		error = xfs_fs_reserve_ag_blocks(mp);
+		if (error == -ENOSPC)
+			error = 0;
 	}
-
-	/*
-	 * Reserve AG metadata blocks. ENOSPC here does not mean there was a
-	 * growfs failure, just that there still isn't space for new user data
-	 * after the grow has been run.
-	 */
-	error = xfs_fs_reserve_ag_blocks(mp);
-	if (error == -ENOSPC)
-		error = 0;
 	return error;
 
 out_trans_cancel:
@@ -170,8 +223,8 @@ out_trans_cancel:
 
 static int
 xfs_growfs_log_private(
-	xfs_mount_t		*mp,	/* mount point for filesystem */
-	xfs_growfs_log_t	*in)	/* growfs log input struct */
+	struct xfs_mount	*mp,	/* mount point for filesystem */
+	struct xfs_growfs_log	*in)	/* growfs log input struct */
 {
 	xfs_extlen_t		nb;
 
@@ -268,7 +321,7 @@ out_error:
 int
 xfs_growfs_log(
 	xfs_mount_t		*mp,
-	xfs_growfs_log_t	*in)
+	struct xfs_growfs_log	*in)
 {
 	int error;
 
@@ -293,11 +346,8 @@ xfs_fs_counts(
 	cnt->allocino = percpu_counter_read_positive(&mp->m_icount);
 	cnt->freeino = percpu_counter_read_positive(&mp->m_ifree);
 	cnt->freedata = percpu_counter_read_positive(&mp->m_fdblocks) -
-						mp->m_alloc_set_aside;
-
-	spin_lock(&mp->m_sb_lock);
-	cnt->freertx = mp->m_sb.sb_frextents;
-	spin_unlock(&mp->m_sb_lock);
+						xfs_fdblocks_unavailable(mp);
+	cnt->freertx = percpu_counter_read_positive(&mp->m_frextents);
 }
 
 /*
@@ -376,46 +426,36 @@ xfs_reserve_blocks(
 	 * If the request is larger than the current reservation, reserve the
 	 * blocks before we update the reserve counters. Sample m_fdblocks and
 	 * perform a partial reservation if the request exceeds free space.
+	 *
+	 * The code below estimates how many blocks it can request from
+	 * fdblocks to stash in the reserve pool.  This is a classic TOCTOU
+	 * race since fdblocks updates are not always coordinated via
+	 * m_sb_lock.  Set the reserve size even if there's not enough free
+	 * space to fill it because mod_fdblocks will refill an undersized
+	 * reserve when it can.
 	 */
-	error = -ENOSPC;
-	do {
-		free = percpu_counter_sum(&mp->m_fdblocks) -
-						mp->m_alloc_set_aside;
-		if (free <= 0)
-			break;
-
-		delta = request - mp->m_resblks;
-		lcounter = free - delta;
-		if (lcounter < 0)
-			/* We can't satisfy the request, just get what we can */
-			fdblks_delta = free;
-		else
-			fdblks_delta = delta;
-
+	free = percpu_counter_sum(&mp->m_fdblocks) -
+						xfs_fdblocks_unavailable(mp);
+	delta = request - mp->m_resblks;
+	mp->m_resblks = request;
+	if (delta > 0 && free > 0) {
 		/*
 		 * We'll either succeed in getting space from the free block
-		 * count or we'll get an ENOSPC. If we get a ENOSPC, it means
-		 * things changed while we were calculating fdblks_delta and so
-		 * we should try again to see if there is anything left to
-		 * reserve.
+		 * count or we'll get an ENOSPC.  Don't set the reserved flag
+		 * here - we don't want to reserve the extra reserve blocks
+		 * from the reserve.
 		 *
-		 * Don't set the reserved flag here - we don't want to reserve
-		 * the extra reserve blocks from the reserve.....
+		 * The desired reserve size can change after we drop the lock.
+		 * Use mod_fdblocks to put the space into the reserve or into
+		 * fdblocks as appropriate.
 		 */
+		fdblks_delta = min(free, delta);
 		spin_unlock(&mp->m_sb_lock);
 		error = xfs_mod_fdblocks(mp, -fdblks_delta, 0);
+		if (!error)
+			xfs_mod_fdblocks(mp, fdblks_delta, 0);
 		spin_lock(&mp->m_sb_lock);
-	} while (error == -ENOSPC);
-
-	/*
-	 * Update the reserve counters if blocks have been successfully
-	 * allocated.
-	 */
-	if (!error && fdblks_delta) {
-		mp->m_resblks += fdblks_delta;
-		mp->m_resblks_avail += fdblks_delta;
 	}
-
 out:
 	if (outval) {
 		outval->resblks = mp->m_resblks;
@@ -433,13 +473,10 @@ xfs_fs_goingdown(
 {
 	switch (inflags) {
 	case XFS_FSOP_GOING_FLAGS_DEFAULT: {
-		struct super_block *sb = freeze_bdev(mp->m_super->s_bdev);
-
-		if (sb && !IS_ERR(sb)) {
+		if (!freeze_bdev(mp->m_super->s_bdev)) {
 			xfs_force_shutdown(mp, SHUTDOWN_FORCE_UMOUNT);
-			thaw_bdev(sb->s_bdev, sb);
+			thaw_bdev(mp->m_super->s_bdev);
 		}
-
 		break;
 	}
 	case XFS_FSOP_GOING_FLAGS_LOGFLUSH:
@@ -461,59 +498,56 @@ xfs_fs_goingdown(
  * consistent. We don't do an unmount here; just shutdown the shop, make sure
  * that absolutely nothing persistent happens to this filesystem after this
  * point.
+ *
+ * The shutdown state change is atomic, resulting in the first and only the
+ * first shutdown call processing the shutdown. This means we only shutdown the
+ * log once as it requires, and we don't spam the logs when multiple concurrent
+ * shutdowns race to set the shutdown flags.
  */
 void
 xfs_do_force_shutdown(
 	struct xfs_mount *mp,
-	int		flags,
+	uint32_t	flags,
 	char		*fname,
 	int		lnnum)
 {
-	bool		logerror = flags & SHUTDOWN_LOG_IO_ERROR;
+	int		tag;
+	const char	*why;
 
-	/*
-	 * No need to duplicate efforts.
-	 */
-	if (XFS_FORCED_SHUTDOWN(mp) && !logerror)
-		return;
 
-	/*
-	 * This flags XFS_MOUNT_FS_SHUTDOWN, makes sure that we don't
-	 * queue up anybody new on the log reservations, and wakes up
-	 * everybody who's sleeping on log reservations to tell them
-	 * the bad news.
-	 */
-	if (xfs_log_force_umount(mp, logerror))
-		return;
-
-	if (flags & SHUTDOWN_FORCE_UMOUNT) {
-		xfs_alert(mp,
-"User initiated shutdown received. Shutting down filesystem");
+	if (test_and_set_bit(XFS_OPSTATE_SHUTDOWN, &mp->m_opstate)) {
+		xlog_shutdown_wait(mp->m_log);
 		return;
 	}
-
-	xfs_notice(mp,
-"%s(0x%x) called from line %d of file %s. Return address = "PTR_FMT,
-		__func__, flags, lnnum, fname, __return_address);
-
-	if (flags & SHUTDOWN_CORRUPT_INCORE) {
-		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_CORRUPT,
-"Corruption of in-memory data detected.  Shutting down filesystem");
-		if (XFS_ERRLEVEL_HIGH <= xfs_error_level)
-			xfs_stack_trace();
-	} else if (logerror) {
-		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_LOGERROR,
-			"Log I/O Error Detected. Shutting down filesystem");
-	} else if (flags & SHUTDOWN_DEVICE_REQ) {
-		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_IOERROR,
-			"All device paths lost. Shutting down filesystem");
-	} else if (!(flags & SHUTDOWN_REMOTE_REQ)) {
-		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_IOERROR,
-			"I/O Error Detected. Shutting down filesystem");
+	if (mp->m_sb_bp)
+		mp->m_sb_bp->b_flags |= XBF_DONE;
+
+	if (flags & SHUTDOWN_FORCE_UMOUNT)
+		xfs_alert(mp, "User initiated shutdown received.");
+
+	if (xlog_force_shutdown(mp->m_log, flags)) {
+		tag = XFS_PTAG_SHUTDOWN_LOGERROR;
+		why = "Log I/O Error";
+	} else if (flags & SHUTDOWN_CORRUPT_INCORE) {
+		tag = XFS_PTAG_SHUTDOWN_CORRUPT;
+		why = "Corruption of in-memory data";
+	} else if (flags & SHUTDOWN_CORRUPT_ONDISK) {
+		tag = XFS_PTAG_SHUTDOWN_CORRUPT;
+		why = "Corruption of on-disk metadata";
+	} else {
+		tag = XFS_PTAG_SHUTDOWN_IOERROR;
+		why = "Metadata I/O Error";
 	}
 
+	trace_xfs_force_shutdown(mp, tag, flags, fname, lnnum);
+
+	xfs_alert_tag(mp, tag,
+"%s (0x%x) detected at %pS (%s:%d).  Shutting down filesystem.",
+			why, flags, __return_address, fname, lnnum);
 	xfs_alert(mp,
 		"Please unmount the filesystem and rectify the problem(s)");
+	if (xfs_error_level >= XFS_ERRLEVEL_HIGH)
+		xfs_stack_trace();
 }
 
 /*
@@ -529,10 +563,8 @@ xfs_fs_reserve_ag_blocks(
 	int			err2;
 
 	mp->m_finobt_nores = false;
-	for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
-		pag = xfs_perag_get(mp, agno);
+	for_each_perag(mp, agno, pag) {
 		err2 = xfs_ag_resv_init(pag, NULL);
-		xfs_perag_put(pag);
 		if (err2 && !error)
 			error = err2;
 	}
@@ -558,10 +590,8 @@ xfs_fs_unreserve_ag_blocks(
 	int			error = 0;
 	int			err2;
 
-	for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
-		pag = xfs_perag_get(mp, agno);
+	for_each_perag(mp, agno, pag) {
 		err2 = xfs_ag_resv_free(pag);
-		xfs_perag_put(pag);
 		if (err2 && !error)
 			error = err2;
 	}