4.9.21-stable review

@@ -7,6 +7,7 @@
 #include <linux/kthread.h>
 #include <linux/net.h>
 #include <linux/nsproxy.h>
+#include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/socket.h>
 #include <linux/string.h>

@@ -469,11 +470,16 @@ static int ceph_tcp_connect(struct ceph_
 {
 	struct sockaddr_storage *paddr = &con->peer_addr.in_addr;
 	struct socket *sock;
+	unsigned int noio_flag;
 	int ret;
 
 	BUG_ON(con->sock);
+
+	/* sock_create_kern() allocates with GFP_KERNEL */
+	noio_flag = memalloc_noio_save();
 	ret = sock_create_kern(read_pnet(&con->msgr->net), paddr->ss_family,
 			       SOCK_STREAM, IPPROTO_TCP, &sock);
+	memalloc_noio_restore(noio_flag);
 	if (ret)
 		return ret;
 	sock->sk->sk_allocation = GFP_NOFS;

@@ -713,10 +713,9 @@ static void __init xen_reserve_xen_mfnli
 		size = PFN_PHYS(xen_start_info->nr_p2m_frames);
 	}
 
-	if (!xen_is_e820_reserved(start, size)) {
-		memblock_reserve(start, size);
+	memblock_reserve(start, size);
+	if (!xen_is_e820_reserved(start, size))
 		return;
-	}
 
 #ifdef CONFIG_X86_32
 	/*

@@ -727,6 +726,7 @@ static void __init xen_reserve_xen_mfnli
 	BUG();
 #else
 	xen_relocate_p2m();
+	memblock_free(start, size);
 #endif
 }
 

@@ -200,22 +200,27 @@ __xfs_ag_resv_init(
 	struct xfs_mount		*mp = pag->pag_mount;
 	struct xfs_ag_resv		*resv;
 	int				error;
+	xfs_extlen_t			reserved;
 
-	resv = xfs_perag_resv(pag, type);
 	if (used > ask)
 		ask = used;
-	resv->ar_asked = ask;
-	resv->ar_reserved = resv->ar_orig_reserved = ask - used;
-	mp->m_ag_max_usable -= ask;
+	reserved = ask - used;
 
-	trace_xfs_ag_resv_init(pag, type, ask);
-
-	error = xfs_mod_fdblocks(mp, -(int64_t)resv->ar_reserved, true);
-	if (error)
+	error = xfs_mod_fdblocks(mp, -(int64_t)reserved, true);
+	if (error) {
 		trace_xfs_ag_resv_init_error(pag->pag_mount, pag->pag_agno,
 				error, _RET_IP_);
+		return error;
+	}
 
-	return error;
+	mp->m_ag_max_usable -= ask;
+
+	resv = xfs_perag_resv(pag, type);
+	resv->ar_asked = ask;
+	resv->ar_reserved = resv->ar_orig_reserved = reserved;
+
+	trace_xfs_ag_resv_init(pag, type, ask);
+	return 0;
 }
 
 /* Create a per-AG block reservation. */

@@ -39,6 +39,7 @@
 #include "xfs_rmap_btree.h"
 #include "xfs_btree.h"
 #include "xfs_refcount_btree.h"
+#include "xfs_ialloc_btree.h"
 
 /*
  * Per-AG Block Reservations

@@ -210,6 +211,9 @@ __xfs_ag_resv_init(
 	if (error) {
 		trace_xfs_ag_resv_init_error(pag->pag_mount, pag->pag_agno,
 				error, _RET_IP_);
+		xfs_warn(mp,
+"Per-AG reservation for AG %u failed.  Filesystem may run out of space.",
+				pag->pag_agno);
 		return error;
 	}
 

@@ -228,6 +232,8 @@ int
 xfs_ag_resv_init(
 	struct xfs_perag		*pag)
 {
+	struct xfs_mount		*mp = pag->pag_mount;
+	xfs_agnumber_t			agno = pag->pag_agno;
 	xfs_extlen_t			ask;
 	xfs_extlen_t			used;
 	int				error = 0;

@@ -236,23 +242,45 @@ xfs_ag_resv_init(
 	if (pag->pag_meta_resv.ar_asked == 0) {
 		ask = used = 0;
 
-		error = xfs_refcountbt_calc_reserves(pag->pag_mount,
-				pag->pag_agno, &ask, &used);
+		error = xfs_refcountbt_calc_reserves(mp, agno, &ask, &used);
 		if (error)
 			goto out;
 
-		error = __xfs_ag_resv_init(pag, XFS_AG_RESV_METADATA,
-				ask, used);
+		error = xfs_finobt_calc_reserves(mp, agno, &ask, &used);
 		if (error)
 			goto out;
+
+		error = __xfs_ag_resv_init(pag, XFS_AG_RESV_METADATA,
+				ask, used);
+		if (error) {
+			/*
+			 * Because we didn't have per-AG reservations when the
+			 * finobt feature was added we might not be able to
+			 * reserve all needed blocks.  Warn and fall back to the
+			 * old and potentially buggy code in that case, but
+			 * ensure we do have the reservation for the refcountbt.
+			 */
+			ask = used = 0;
+
+			mp->m_inotbt_nores = true;
+
+			error = xfs_refcountbt_calc_reserves(mp, agno, &ask,
+					&used);
+			if (error)
+				goto out;
+
+			error = __xfs_ag_resv_init(pag, XFS_AG_RESV_METADATA,
+					ask, used);
+			if (error)
+				goto out;
+		}
 	}
 
 	/* Create the AGFL metadata reservation */
 	if (pag->pag_agfl_resv.ar_asked == 0) {
 		ask = used = 0;
 
-		error = xfs_rmapbt_calc_reserves(pag->pag_mount, pag->pag_agno,
-				&ask, &used);
+		error = xfs_rmapbt_calc_reserves(mp, agno, &ask, &used);
 		if (error)
 			goto out;
 

@@ -261,9 +289,16 @@ xfs_ag_resv_init(
 			goto out;
 	}
 
+#ifdef DEBUG
+	/* need to read in the AGF for the ASSERT below to work */
+	error = xfs_alloc_pagf_init(pag->pag_mount, NULL, pag->pag_agno, 0);
+	if (error)
+		return error;
+
 	ASSERT(xfs_perag_resv(pag, XFS_AG_RESV_METADATA)->ar_reserved +
 	       xfs_perag_resv(pag, XFS_AG_RESV_AGFL)->ar_reserved <=
 	       pag->pagf_freeblks + pag->pagf_flcount);
+#endif
 out:
 	return error;
 }

120 lines not on this page · raw message

@@ -72,4 +72,7 @@ int xfs_inobt_rec_check_count(struct xfs
 #define xfs_inobt_rec_check_count(mp, rec)	0
 #endif	/* DEBUG */
 
+int xfs_finobt_calc_reserves(struct xfs_mount *mp, xfs_agnumber_t agno,
+		xfs_extlen_t *ask, xfs_extlen_t *used);
+
 #endif	/* __XFS_IALLOC_BTREE_H__ */

@@ -1801,22 +1801,23 @@ xfs_inactive_ifree(
 	int			error;
 
 	/*
-	 * The ifree transaction might need to allocate blocks for record
-	 * insertion to the finobt. We don't want to fail here at ENOSPC, so
-	 * allow ifree to dip into the reserved block pool if necessary.
-	 *
-	 * Freeing large sets of inodes generally means freeing inode chunks,
-	 * directory and file data blocks, so this should be relatively safe.
-	 * Only under severe circumstances should it be possible to free enough
-	 * inodes to exhaust the reserve block pool via finobt expansion while
-	 * at the same time not creating free space in the filesystem.
+	 * We try to use a per-AG reservation for any block needed by the finobt
+	 * tree, but as the finobt feature predates the per-AG reservation
+	 * support a degraded file system might not have enough space for the
+	 * reservation at mount time.  In that case try to dip into the reserved
+	 * pool and pray.
 	 *
 	 * Send a warning if the reservation does happen to fail, as the inode
 	 * now remains allocated and sits on the unlinked list until the fs is
 	 * repaired.
 	 */
-	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ifree,
-			XFS_IFREE_SPACE_RES(mp), 0, XFS_TRANS_RESERVE, &tp);
+	if (unlikely(mp->m_inotbt_nores)) {
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ifree,
+				XFS_IFREE_SPACE_RES(mp), 0, XFS_TRANS_RESERVE,
+				&tp);
+	} else {
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ifree, 0, 0, 0, &tp);
+	}
 	if (error) {
 		if (error == -ENOSPC) {
 			xfs_warn_ratelimited(mp,

@@ -140,6 +140,7 @@ typedef struct xfs_mount {
 	int			m_fixedfsid[2];	/* unchanged for life of FS */
 	uint			m_dmevmask;	/* DMI events for this FS */
 	__uint64_t		m_flags;	/* global mount flags */
+	bool			m_inotbt_nores; /* no per-AG finobt resv. */
 	int			m_ialloc_inos;	/* inodes in inode allocation */
 	int			m_ialloc_blks;	/* blocks in inode allocation */
 	int			m_ialloc_min_blks;/* min blocks in sparse inode

@@ -917,17 +917,18 @@ xfs_can_free_eofblocks(struct xfs_inode
  */
 int
 xfs_free_eofblocks(
-	xfs_mount_t	*mp,
-	xfs_inode_t	*ip,
-	bool		need_iolock)
+	struct xfs_inode	*ip)
 {
-	xfs_trans_t	*tp;
-	int		error;
-	xfs_fileoff_t	end_fsb;
-	xfs_fileoff_t	last_fsb;
-	xfs_filblks_t	map_len;
-	int		nimaps;
-	xfs_bmbt_irec_t	imap;
+	struct xfs_trans	*tp;
+	int			error;
+	xfs_fileoff_t		end_fsb;
+	xfs_fileoff_t		last_fsb;
+	xfs_filblks_t		map_len;
+	int			nimaps;
+	struct xfs_bmbt_irec	imap;
+	struct xfs_mount	*mp = ip->i_mount;
+
+	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
 
 	/*
 	 * Figure out if there are any blocks beyond the end

@@ -944,6 +945,10 @@ xfs_free_eofblocks(
 	error = xfs_bmapi_read(ip, end_fsb, map_len, &imap, &nimaps, 0);
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
+	/*
+	 * If there are blocks after the end of file, truncate the file to its
+	 * current size to free them up.
+	 */
 	if (!error && (nimaps != 0) &&
 	    (imap.br_startblock != HOLESTARTBLOCK ||
 	     ip->i_delayed_blks)) {

@@ -954,22 +959,10 @@ xfs_free_eofblocks(
 		if (error)
 			return error;
 
-		/*
-		 * There are blocks after the end of file.
-		 * Free them up now by truncating the file to
-		 * its current size.
-		 */
-		if (need_iolock) {
-			if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL))
-				return -EAGAIN;
-		}
-
 		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0,
 				&tp);
 		if (error) {
 			ASSERT(XFS_FORCED_SHUTDOWN(mp));
-			if (need_iolock)
-				xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 			return error;
 		}
 

@@ -997,8 +990,6 @@ xfs_free_eofblocks(
 		}
 
 		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-		if (need_iolock)
-			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 	}
 	return error;
 }

@@ -1415,7 +1406,7 @@ xfs_shift_file_space(
 	 * into the accessible region of the file.
 	 */
 	if (xfs_can_free_eofblocks(ip, true)) {
-		error = xfs_free_eofblocks(mp, ip, false);
+		error = xfs_free_eofblocks(ip);
 		if (error)
 			return error;
 	}

@@ -63,8 +63,7 @@ int	xfs_insert_file_space(struct xfs_ino
 
 /* EOF block manipulation functions */
 bool	xfs_can_free_eofblocks(struct xfs_inode *ip, bool force);
-int	xfs_free_eofblocks(struct xfs_mount *mp, struct xfs_inode *ip,
-			   bool need_iolock);
+int	xfs_free_eofblocks(struct xfs_inode *ip);
 
 int	xfs_swap_extents(struct xfs_inode *ip, struct xfs_inode *tip,
 			 struct xfs_swapext *sx);

@@ -1324,7 +1324,7 @@ xfs_inode_free_eofblocks(
 	int			flags,
 	void			*args)
 {
-	int ret;
+	int ret = 0;
 	struct xfs_eofblocks *eofb = args;
 	bool need_iolock = true;
 	int match;

@@ -1360,19 +1360,25 @@ xfs_inode_free_eofblocks(
 			return 0;
 
 		/*
-		 * A scan owner implies we already hold the iolock. Skip it in
-		 * xfs_free_eofblocks() to avoid deadlock. This also eliminates
-		 * the possibility of EAGAIN being returned.
+		 * A scan owner implies we already hold the iolock. Skip it here
+		 * to avoid deadlock.
 		 */
 		if (eofb->eof_scan_owner == ip->i_ino)
 			need_iolock = false;
 	}
 
-	ret = xfs_free_eofblocks(ip->i_mount, ip, need_iolock);
-
-	/* don't revisit the inode if we're not waiting */
-	if (ret == -EAGAIN && !(flags & SYNC_WAIT))
-		ret = 0;
+	/*
+	 * If the caller is waiting, return -EAGAIN to keep the background
+	 * scanner moving and revisit the inode in a subsequent pass.
+	 */
+	if (need_iolock && !xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) {
+		if (flags & SYNC_WAIT)
+			ret = -EAGAIN;
+		return ret;
+	}
+	ret = xfs_free_eofblocks(ip);
+	if (need_iolock)
+		xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 
 	return ret;
 }

68 lines not on this page · raw message

@@ -675,8 +675,10 @@ xfs_file_buffered_aio_write(
 	struct xfs_inode	*ip = XFS_I(inode);
 	ssize_t			ret;
 	int			enospc = 0;
-	int			iolock = XFS_IOLOCK_EXCL;
+	int			iolock;
 
+write_retry:
+	iolock = XFS_IOLOCK_EXCL;
 	xfs_rw_ilock(ip, iolock);
 
 	ret = xfs_file_aio_write_checks(iocb, from, &iolock);

@@ -686,7 +688,6 @@ xfs_file_buffered_aio_write(
 	/* We can write back this queue in page reclaim */
 	current->backing_dev_info = inode_to_bdi(inode);
 
-write_retry:
 	trace_xfs_file_buffered_write(ip, iov_iter_count(from), iocb->ki_pos);
 	ret = iomap_file_buffered_write(iocb, from, &xfs_iomap_ops);
 	if (likely(ret >= 0))

@@ -702,18 +703,21 @@ write_retry:
 	 * running at the same time.
 	 */
 	if (ret == -EDQUOT && !enospc) {
+		xfs_rw_iunlock(ip, iolock);
 		enospc = xfs_inode_free_quota_eofblocks(ip);
 		if (enospc)
 			goto write_retry;
 		enospc = xfs_inode_free_quota_cowblocks(ip);
 		if (enospc)
 			goto write_retry;
+		iolock = 0;
 	} else if (ret == -ENOSPC && !enospc) {
 		struct xfs_eofblocks eofb = {0};
 
 		enospc = 1;
 		xfs_flush_inodes(ip->i_mount);
-		eofb.eof_scan_owner = ip->i_ino; /* for locking */
+
+		xfs_rw_iunlock(ip, iolock);
 		eofb.eof_flags = XFS_EOF_FLAGS_SYNC;
 		xfs_icache_free_eofblocks(ip->i_mount, &eofb);
 		goto write_retry;

@@ -721,7 +725,8 @@ write_retry:
 
 	current->backing_dev_info = NULL;
 out:
-	xfs_rw_iunlock(ip, iolock);
+	if (iolock)
+		xfs_rw_iunlock(ip, iolock);
 	return ret;
 }
 

100 lines not on this page · raw message

@@ -27,7 +27,6 @@ struct xfs_eofblocks {
 	kgid_t		eof_gid;
 	prid_t		eof_prid;
 	__u64		eof_min_file_size;
-	xfs_ino_t	eof_scan_owner;
 };
 
 #define SYNC_WAIT		0x0001	/* wait for i/o to complete */

@@ -102,7 +101,6 @@ xfs_fs_eofblocks_from_user(
 	dst->eof_flags = src->eof_flags;
 	dst->eof_prid = src->eof_prid;
 	dst->eof_min_file_size = src->eof_min_file_size;
-	dst->eof_scan_owner = NULLFSINO;
 
 	dst->eof_uid = INVALID_UID;
 	if (src->eof_flags & XFS_EOF_FLAGS_UID) {

@@ -959,6 +959,9 @@ xfs_free_eofblocks(
 		if (error)
 			return error;
 
+		/* wait on dio to ensure i_size has settled */
+		inode_dio_wait(VFS_I(ip));
+
 		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0,
 				&tp);
 		if (error) {

@@ -497,15 +497,14 @@ xfs_iread_extents(
 	 * We know that the size is valid (it's checked in iformat_btree)
 	 */
 	ifp->if_bytes = ifp->if_real_bytes = 0;
-	ifp->if_flags |= XFS_IFEXTENTS;
 	xfs_iext_add(ifp, 0, nextents);
 	error = xfs_bmap_read_extents(tp, ip, whichfork);
 	if (error) {
 		xfs_iext_destroy(ifp);
-		ifp->if_flags &= ~XFS_IFEXTENTS;
 		return error;
 	}
 	xfs_validate_extents(ifp, nextents, XFS_EXTFMT_INODE(ip));
+	ifp->if_flags |= XFS_IFEXTENTS;
 	return 0;
 }
 /*

@@ -2633,7 +2633,7 @@ out_free:
 /*
  * Readahead the dir/attr block.
  */
-xfs_daddr_t
+int
 xfs_da_reada_buf(
 	struct xfs_inode	*dp,
 	xfs_dablk_t		bno,

@@ -2664,7 +2664,5 @@ out_free:
 	if (mapp != &map)
 		kmem_free(mapp);
 
-	if (error)
-		return -1;
-	return mappedbno;
+	return error;
 }

@@ -201,7 +201,7 @@ int	xfs_da_read_buf(struct xfs_trans *tr
 			       xfs_dablk_t bno, xfs_daddr_t mappedbno,
 			       struct xfs_buf **bpp, int whichfork,
 			       const struct xfs_buf_ops *ops);
-xfs_daddr_t	xfs_da_reada_buf(struct xfs_inode *dp, xfs_dablk_t bno,
+int	xfs_da_reada_buf(struct xfs_inode *dp, xfs_dablk_t bno,
 				xfs_daddr_t mapped_bno, int whichfork,
 				const struct xfs_buf_ops *ops);
 int	xfs_da_shrink_inode(xfs_da_args_t *args, xfs_dablk_t dead_blkno,

@@ -992,9 +992,9 @@ xfs_dir_open(
 	 */
 	mode = xfs_ilock_data_map_shared(ip);
 	if (ip->i_d.di_nextents > 0)
-		xfs_dir3_data_readahead(ip, 0, -1);
+		error = xfs_dir3_data_readahead(ip, 0, -1);
 	xfs_iunlock(ip, mode);
-	return 0;
+	return error;
 }
 
 STATIC int

@@ -1278,7 +1278,6 @@ xfs_bmap_read_extents(
 	/* REFERENCED */
 	xfs_extnum_t		room;	/* number of entries there's room for */
 
-	bno = NULLFSBLOCK;
 	mp = ip->i_mount;
 	ifp = XFS_IFORK_PTR(ip, whichfork);
 	exntf = (whichfork != XFS_DATA_FORK) ? XFS_EXTFMT_NOSTATE :

@@ -1291,9 +1290,7 @@ xfs_bmap_read_extents(
 	ASSERT(level > 0);
 	pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
 	bno = be64_to_cpu(*pp);
-	ASSERT(bno != NULLFSBLOCK);
-	ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
-	ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
+
 	/*
 	 * Go down the tree until leaf level is reached, following the first
 	 * pointer (leftmost) at each level.

@@ -810,7 +810,8 @@ xfs_btree_read_bufl(
 	xfs_daddr_t		d;		/* real disk block address */
 	int			error;
 
-	ASSERT(fsbno != NULLFSBLOCK);
+	if (!XFS_FSB_SANITY_CHECK(mp, fsbno))
+		return -EFSCORRUPTED;
 	d = XFS_FSB_TO_DADDR(mp, fsbno);
 	error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, d,
 				   mp->m_bsize, lock, &bp, ops);

@@ -491,7 +491,7 @@ static inline int xfs_btree_get_level(st
 #define	XFS_FILBLKS_MAX(a,b)	max_t(xfs_filblks_t, (a), (b))
 
 #define	XFS_FSB_SANITY_CHECK(mp,fsb)	\
-	(XFS_FSB_TO_AGNO(mp, fsb) < mp->m_sb.sb_agcount && \
+	(fsb && XFS_FSB_TO_AGNO(mp, fsb) < mp->m_sb.sb_agcount && \
 		XFS_FSB_TO_AGBNO(mp, fsb) < mp->m_sb.sb_agblocks)
 
 /*

@@ -26,6 +26,7 @@
 #include "xfs_inode.h"
 #include "xfs_trans.h"
 #include "xfs_inode_item.h"
+#include "xfs_btree.h"
 #include "xfs_bmap_btree.h"
 #include "xfs_bmap.h"
 #include "xfs_error.h"

@@ -429,11 +430,13 @@ xfs_iformat_btree(
 	/* REFERENCED */
 	int			nrecs;
 	int			size;
+	int			level;
 
 	ifp = XFS_IFORK_PTR(ip, whichfork);
 	dfp = (xfs_bmdr_block_t *)XFS_DFORK_PTR(dip, whichfork);
 	size = XFS_BMAP_BROOT_SPACE(mp, dfp);
 	nrecs = be16_to_cpu(dfp->bb_numrecs);
+	level = be16_to_cpu(dfp->bb_level);
 
 	/*
 	 * blow out if -- fork has less extents than can fit in

@@ -446,7 +449,8 @@ xfs_iformat_btree(
 					XFS_IFORK_MAXEXT(ip, whichfork) ||
 		     XFS_BMDR_SPACE_CALC(nrecs) >
 					XFS_DFORK_SIZE(dip, mp, whichfork) ||
-		     XFS_IFORK_NEXTENTS(ip, whichfork) > ip->i_d.di_nblocks)) {
+		     XFS_IFORK_NEXTENTS(ip, whichfork) > ip->i_d.di_nblocks) ||
+		     level == 0 || level > XFS_BTREE_MAXLEVELS) {
 		xfs_warn(mp, "corrupt inode %Lu (btree).",
 					(unsigned long long) ip->i_ino);
 		XFS_CORRUPTION_ERROR("xfs_iformat_btree", XFS_ERRLEVEL_LOW,

@@ -155,6 +155,42 @@ const struct xfs_buf_ops xfs_dir3_free_b
 	.verify_write = xfs_dir3_free_write_verify,
 };
 
+/* Everything ok in the free block header? */
+static bool
+xfs_dir3_free_header_check(
+	struct xfs_inode	*dp,
+	xfs_dablk_t		fbno,
+	struct xfs_buf		*bp)
+{
+	struct xfs_mount	*mp = dp->i_mount;
+	unsigned int		firstdb;
+	int			maxbests;
+
+	maxbests = dp->d_ops->free_max_bests(mp->m_dir_geo);
+	firstdb = (xfs_dir2_da_to_db(mp->m_dir_geo, fbno) -
+		   xfs_dir2_byte_to_db(mp->m_dir_geo, XFS_DIR2_FREE_OFFSET)) *
+			maxbests;
+	if (xfs_sb_version_hascrc(&mp->m_sb)) {
+		struct xfs_dir3_free_hdr *hdr3 = bp->b_addr;
+
+		if (be32_to_cpu(hdr3->firstdb) != firstdb)
+			return false;
+		if (be32_to_cpu(hdr3->nvalid) > maxbests)
+			return false;
+		if (be32_to_cpu(hdr3->nvalid) < be32_to_cpu(hdr3->nused))
+			return false;
+	} else {
+		struct xfs_dir2_free_hdr *hdr = bp->b_addr;
+
+		if (be32_to_cpu(hdr->firstdb) != firstdb)
+			return false;
+		if (be32_to_cpu(hdr->nvalid) > maxbests)
+			return false;
+		if (be32_to_cpu(hdr->nvalid) < be32_to_cpu(hdr->nused))
+			return false;
+	}
+	return true;
+}
 
 static int
 __xfs_dir3_free_read(

@@ -168,11 +204,22 @@ __xfs_dir3_free_read(
 
 	err = xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp,
 				XFS_DATA_FORK, &xfs_dir3_free_buf_ops);
+	if (err || !*bpp)
+		return err;
+
+	/* Check things that we can't do in the verifier. */
+	if (!xfs_dir3_free_header_check(dp, fbno, *bpp)) {
+		xfs_buf_ioerror(*bpp, -EFSCORRUPTED);
+		xfs_verifier_error(*bpp);
+		xfs_trans_brelse(tp, *bpp);
+		return -EFSCORRUPTED;
+	}
 
 	/* try read returns without an error or *bpp if it lands in a hole */
-	if (!err && tp && *bpp)
+	if (tp)
 		xfs_trans_buf_set_type(tp, *bpp, XFS_BLFT_DIR_FREE_BUF);
-	return err;
+
+	return 0;
 }
 
 int

@@ -1952,6 +1952,7 @@ xfs_bmap_add_extent_delay_real(
 		 */
 		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
 		xfs_bmbt_set_startblock(ep, new->br_startblock);
+		xfs_bmbt_set_state(ep, new->br_state);
 		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
 		(*nextents)++;

@@ -2290,6 +2291,7 @@ STATIC int				/* error */
 xfs_bmap_add_extent_unwritten_real(
 	struct xfs_trans	*tp,
 	xfs_inode_t		*ip,	/* incore inode pointer */
+	int			whichfork,
 	xfs_extnum_t		*idx,	/* extent number to update/insert */
 	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
 	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */

@@ -2309,12 +2311,14 @@ xfs_bmap_add_extent_unwritten_real(
 					/* left is 0, right is 1, prev is 2 */
 	int			rval=0;	/* return value (logging flags) */
 	int			state = 0;/* state bits, accessed thru macros */
-	struct xfs_mount	*mp = tp->t_mountp;
+	struct xfs_mount	*mp = ip->i_mount;
 
 	*logflagsp = 0;
 
 	cur = *curp;
-	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	ifp = XFS_IFORK_PTR(ip, whichfork);
+	if (whichfork == XFS_COW_FORK)
+		state |= BMAP_COWFORK;
 
 	ASSERT(*idx >= 0);
 	ASSERT(*idx <= xfs_iext_count(ifp));

@@ -2373,7 +2377,7 @@ xfs_bmap_add_extent_unwritten_real(
 	 * Don't set contiguous if the combined extent would be too large.
 	 * Also check for all-three-contiguous being too large.
 	 */
-	if (*idx < xfs_iext_count(&ip->i_df) - 1) {
+	if (*idx < xfs_iext_count(ifp) - 1) {
 		state |= BMAP_RIGHT_VALID;
 		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT);
 		if (isnullstartblock(RIGHT.br_startblock))

@@ -2413,7 +2417,8 @@ xfs_bmap_add_extent_unwritten_real(
 		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
 
 		xfs_iext_remove(ip, *idx + 1, 2, state);
-		ip->i_d.di_nextents -= 2;
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) - 2);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {

@@ -2456,7 +2461,8 @@ xfs_bmap_add_extent_unwritten_real(
 		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
 
 		xfs_iext_remove(ip, *idx + 1, 1, state);
-		ip->i_d.di_nextents--;
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {

@@ -2491,7 +2497,8 @@ xfs_bmap_add_extent_unwritten_real(
 		xfs_bmbt_set_state(ep, newext);
 		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
 		xfs_iext_remove(ip, *idx + 1, 1, state);
-		ip->i_d.di_nextents--;
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {

@@ -2603,7 +2610,8 @@ xfs_bmap_add_extent_unwritten_real(
 		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
 
 		xfs_iext_insert(ip, *idx, 1, new, state);
-		ip->i_d.di_nextents++;
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {

@@ -2681,7 +2689,8 @@ xfs_bmap_add_extent_unwritten_real(
 		++*idx;
 		xfs_iext_insert(ip, *idx, 1, new, state);
 
-		ip->i_d.di_nextents++;
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {

@@ -2729,7 +2738,8 @@ xfs_bmap_add_extent_unwritten_real(
 		++*idx;
 		xfs_iext_insert(ip, *idx, 2, &r[0], state);
 
-		ip->i_d.di_nextents += 2;
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) + 2);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {

@@ -2783,17 +2793,17 @@ xfs_bmap_add_extent_unwritten_real(
 	}
 
 	/* update reverse mappings */
-	error = xfs_rmap_convert_extent(mp, dfops, ip, XFS_DATA_FORK, new);
+	error = xfs_rmap_convert_extent(mp, dfops, ip, whichfork, new);
 	if (error)
 		goto done;
 
 	/* convert to a btree if necessary */
-	if (xfs_bmap_needs_btree(ip, XFS_DATA_FORK)) {
+	if (xfs_bmap_needs_btree(ip, whichfork)) {
 		int	tmp_logflags;	/* partial log flag return val */
 
 		ASSERT(cur == NULL);
 		error = xfs_bmap_extents_to_btree(tp, ip, first, dfops, &cur,
-				0, &tmp_logflags, XFS_DATA_FORK);
+				0, &tmp_logflags, whichfork);
 		*logflagsp |= tmp_logflags;
 		if (error)
 			goto done;

@@ -2805,7 +2815,7 @@ xfs_bmap_add_extent_unwritten_real(
 		*curp = cur;
 	}
 
-	xfs_bmap_check_leaf_extents(*curp, ip, XFS_DATA_FORK);
+	xfs_bmap_check_leaf_extents(*curp, ip, whichfork);
 done:
 	*logflagsp |= rval;
 	return error;

@@ -4458,10 +4468,16 @@ xfs_bmapi_allocate(
 	bma->got.br_state = XFS_EXT_NORM;
 
 	/*
-	 * A wasdelay extent has been initialized, so shouldn't be flagged
-	 * as unwritten.
+	 * In the data fork, a wasdelay extent has been initialized, so
+	 * shouldn't be flagged as unwritten.
+	 *
+	 * For the cow fork, however, we convert delalloc reservations
+	 * (extents allocated for speculative preallocation) to
+	 * allocated unwritten extents, and only convert the unwritten
+	 * extents to real extents when we're about to write the data.
 	 */
-	if (!bma->wasdel && (bma->flags & XFS_BMAPI_PREALLOC) &&
+	if ((!bma->wasdel || (bma->flags & XFS_BMAPI_COWFORK)) &&
+	    (bma->flags & XFS_BMAPI_PREALLOC) &&
 	    xfs_sb_version_hasextflgbit(&mp->m_sb))
 		bma->got.br_state = XFS_EXT_UNWRITTEN;
 

@@ -4512,8 +4528,6 @@ xfs_bmapi_convert_unwritten(
 			(XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT))
 		return 0;
 
-	ASSERT(whichfork != XFS_COW_FORK);
-
 	/*
 	 * Modify (by adding) the state flag, if writing.
 	 */

@@ -4538,8 +4552,8 @@ xfs_bmapi_convert_unwritten(
 			return error;
 	}
 
-	error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, &bma->idx,
-			&bma->cur, mval, bma->firstblock, bma->dfops,
+	error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, whichfork,
+			&bma->idx, &bma->cur, mval, bma->firstblock, bma->dfops,
 			&tmp_logflags);
 	/*
 	 * Log the inode core unconditionally in the unwritten extent conversion

@@ -4548,8 +4562,12 @@ xfs_bmapi_convert_unwritten(
 	 * in the transaction for the sake of fsync(), even if nothing has
 	 * changed, because fsync() will not force the log for this transaction
 	 * unless it sees the inode pinned.
+	 *
+	 * Note: If we're only converting cow fork extents, there aren't
+	 * any on-disk updates to make, so we don't need to log anything.
 	 */
-	bma->logflags |= tmp_logflags | XFS_ILOG_CORE;
+	if (whichfork != XFS_COW_FORK)
+		bma->logflags |= tmp_logflags | XFS_ILOG_CORE;
 	if (error)
 		return error;
 

@@ -4623,15 +4641,15 @@ xfs_bmapi_write(
 	ASSERT(*nmap >= 1);
 	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
 	ASSERT(!(flags & XFS_BMAPI_IGSTATE));
-	ASSERT(tp != NULL);
+	ASSERT(tp != NULL ||
+	       (flags & (XFS_BMAPI_CONVERT | XFS_BMAPI_COWFORK)) ==
+			(XFS_BMAPI_CONVERT | XFS_BMAPI_COWFORK));
 	ASSERT(len > 0);
 	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 	ASSERT(!(flags & XFS_BMAPI_REMAP) || whichfork == XFS_DATA_FORK);
 	ASSERT(!(flags & XFS_BMAPI_PREALLOC) || !(flags & XFS_BMAPI_REMAP));
 	ASSERT(!(flags & XFS_BMAPI_CONVERT) || !(flags & XFS_BMAPI_REMAP));
-	ASSERT(!(flags & XFS_BMAPI_PREALLOC) || whichfork != XFS_COW_FORK);
-	ASSERT(!(flags & XFS_BMAPI_CONVERT) || whichfork != XFS_COW_FORK);
 
 	/* zeroing is for currently only for data extents, not metadata */
 	ASSERT((flags & (XFS_BMAPI_METADATA | XFS_BMAPI_ZERO)) !=

@@ -5653,8 +5671,8 @@ __xfs_bunmapi(
 			}
 			del.br_state = XFS_EXT_UNWRITTEN;
 			error = xfs_bmap_add_extent_unwritten_real(tp, ip,
-					&lastx, &cur, &del, firstblock, dfops,
-					&logflags);
+					whichfork, &lastx, &cur, &del,
+					firstblock, dfops, &logflags);
 			if (error)
 				goto error0;
 			goto nodelete;

@@ -5711,8 +5729,9 @@ __xfs_bunmapi(
 				prev.br_state = XFS_EXT_UNWRITTEN;
 				lastx--;
 				error = xfs_bmap_add_extent_unwritten_real(tp,
-						ip, &lastx, &cur, &prev,
-						firstblock, dfops, &logflags);
+						ip, whichfork, &lastx, &cur,
+						&prev, firstblock, dfops,
+						&logflags);
 				if (error)
 					goto error0;
 				goto nodelete;

@@ -5720,8 +5739,9 @@ __xfs_bunmapi(
 				ASSERT(del.br_state == XFS_EXT_NORM);
 				del.br_state = XFS_EXT_UNWRITTEN;
 				error = xfs_bmap_add_extent_unwritten_real(tp,
-						ip, &lastx, &cur, &del,
-						firstblock, dfops, &logflags);
+						ip, whichfork, &lastx, &cur,
+						&del, firstblock, dfops,
+						&logflags);
 				if (error)
 					goto error0;
 				goto nodelete;

@@ -486,6 +486,12 @@ xfs_submit_ioend(
 	struct xfs_ioend	*ioend,
 	int			status)
 {
+	/* Convert CoW extents to regular */
+	if (!status && ioend->io_type == XFS_IO_COW) {
+		status = xfs_reflink_convert_cow(XFS_I(ioend->io_inode),
+				ioend->io_offset, ioend->io_size);
+	}
+
 	/* Reserve log space if we might write beyond the on-disk inode size. */
 	if (!status &&
 	    ioend->io_type != XFS_IO_UNWRITTEN &&

@@ -685,7 +685,7 @@ xfs_iomap_write_allocate(
 	int		nres;
 
 	if (whichfork == XFS_COW_FORK)
-		flags |= XFS_BMAPI_COWFORK;
+		flags |= XFS_BMAPI_COWFORK | XFS_BMAPI_PREALLOC;
 
 	/*
 	 * Make sure that the dquots are there.

175 lines not on this page · raw message

@@ -30,6 +30,8 @@ extern int xfs_reflink_reserve_cow(struc
 		struct xfs_bmbt_irec *imap, bool *shared);
 extern int xfs_reflink_allocate_cow_range(struct xfs_inode *ip,
 		xfs_off_t offset, xfs_off_t count);
+extern int xfs_reflink_convert_cow(struct xfs_inode *ip, xfs_off_t offset,
+		xfs_off_t count);
 extern bool xfs_reflink_find_cow_mapping(struct xfs_inode *ip, xfs_off_t offset,
 		struct xfs_bmbt_irec *imap, bool *need_alloc);
 extern int xfs_reflink_trim_irec_to_next_cow(struct xfs_inode *ip,

@@ -3183,6 +3183,7 @@ DECLARE_EVENT_CLASS(xfs_inode_irec_class
 		__field(xfs_fileoff_t, lblk)
 		__field(xfs_extlen_t, len)
 		__field(xfs_fsblock_t, pblk)
+		__field(int, state)
 	),
 	TP_fast_assign(
 		__entry->dev = VFS_I(ip)->i_sb->s_dev;

@@ -3190,13 +3191,15 @@ DECLARE_EVENT_CLASS(xfs_inode_irec_class
 		__entry->lblk = irec->br_startoff;
 		__entry->len = irec->br_blockcount;
 		__entry->pblk = irec->br_startblock;
+		__entry->state = irec->br_state;
 	),
-	TP_printk("dev %d:%d ino 0x%llx lblk 0x%llx len 0x%x pblk %llu",
+	TP_printk("dev %d:%d ino 0x%llx lblk 0x%llx len 0x%x pblk %llu st %d",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  __entry->ino,
 		  __entry->lblk,
 		  __entry->len,
-		  __entry->pblk)
+		  __entry->pblk,
+		  __entry->state)
 );
 #define DEFINE_INODE_IREC_EVENT(name) \
 DEFINE_EVENT(xfs_inode_irec_class, name, \

@@ -3345,6 +3348,7 @@ DEFINE_INODE_IREC_EVENT(xfs_reflink_trim
 DEFINE_INODE_IREC_EVENT(xfs_reflink_cow_alloc);
 DEFINE_INODE_IREC_EVENT(xfs_reflink_cow_found);
 DEFINE_INODE_IREC_EVENT(xfs_reflink_cow_enospc);
+DEFINE_INODE_IREC_EVENT(xfs_reflink_convert_cow);
 
 DEFINE_RW_EVENT(xfs_reflink_reserve_cow);
 DEFINE_RW_EVENT(xfs_reflink_allocate_cow_range);

@@ -1162,6 +1162,7 @@ xfs_buf_iodone_callbacks(
 	 */
 	bp->b_last_error = 0;
 	bp->b_retries = 0;
+	bp->b_first_retry_time = 0;
 
 	xfs_buf_do_callbacks(bp);
 	bp->b_fspriv = NULL;

@@ -985,13 +985,14 @@ STATIC int
 xfs_reflink_update_dest(
 	struct xfs_inode	*dest,
 	xfs_off_t		newlen,
-	xfs_extlen_t		cowextsize)
+	xfs_extlen_t		cowextsize,
+	bool			is_dedupe)
 {
 	struct xfs_mount	*mp = dest->i_mount;
 	struct xfs_trans	*tp;
 	int			error;
 
-	if (newlen <= i_size_read(VFS_I(dest)) && cowextsize == 0)
+	if (is_dedupe && newlen <= i_size_read(VFS_I(dest)) && cowextsize == 0)
 		return 0;
 
 	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);

@@ -1012,6 +1013,10 @@ xfs_reflink_update_dest(
 		dest->i_d.di_flags2 |= XFS_DIFLAG2_COWEXTSIZE;
 	}
 
+	if (!is_dedupe) {
+		xfs_trans_ichgtime(tp, dest,
+				   XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
+	}
 	xfs_trans_log_inode(tp, dest, XFS_ILOG_CORE);
 
 	error = xfs_trans_commit(tp);

@@ -1528,7 +1533,8 @@ xfs_reflink_remap_range(
 	    !(dest->i_d.di_flags2 & XFS_DIFLAG2_COWEXTSIZE))
 		cowextsize = src->i_d.di_cowextsize;
 
-	ret = xfs_reflink_update_dest(dest, pos_out + len, cowextsize);
+	ret = xfs_reflink_update_dest(dest, pos_out + len, cowextsize,
+			is_dedupe);
 
 out_unlock:
 	xfs_iunlock(src, XFS_MMAPLOCK_EXCL);

@@ -1263,44 +1263,6 @@ xfs_map_trim_size(
 	bh_result->b_size = mapping_size;
 }
 
-/* Bounce unaligned directio writes to the page cache. */
-static int
-xfs_bounce_unaligned_dio_write(
-	struct xfs_inode	*ip,
-	xfs_fileoff_t		offset_fsb,
-	struct xfs_bmbt_irec	*imap)
-{
-	struct xfs_bmbt_irec	irec;
-	xfs_fileoff_t		delta;
-	bool			shared;
-	bool			x;
-	int			error;
-
-	irec = *imap;
-	if (offset_fsb > irec.br_startoff) {
-		delta = offset_fsb - irec.br_startoff;
-		irec.br_blockcount -= delta;
-		irec.br_startblock += delta;
-		irec.br_startoff = offset_fsb;
-	}
-	error = xfs_reflink_trim_around_shared(ip, &irec, &shared, &x);
-	if (error)
-		return error;
-
-	/*
-	 * We're here because we're trying to do a directio write to a
-	 * region that isn't aligned to a filesystem block.  If any part
-	 * of the extent is shared, fall back to buffered mode to handle
-	 * the RMW.  This is done by returning -EREMCHG ("remote addr
-	 * changed"), which is caught further up the call stack.
-	 */
-	if (shared) {
-		trace_xfs_reflink_bounce_dio_write(ip, imap);
-		return -EREMCHG;
-	}
-	return 0;
-}
-
 STATIC int
 __xfs_get_blocks(
 	struct inode		*inode,

@@ -1438,13 +1400,6 @@ __xfs_get_blocks(
 	if (imap.br_startblock != HOLESTARTBLOCK &&
 	    imap.br_startblock != DELAYSTARTBLOCK &&
 	    (create || !ISUNWRITTEN(&imap))) {
-		if (create && direct && !is_cow) {
-			error = xfs_bounce_unaligned_dio_write(ip, offset_fsb,
-					&imap);
-			if (error)
-				return error;
-		}
-
 		xfs_map_buffer(inode, bh_result, &imap, offset);
 		if (ISUNWRITTEN(&imap))
 			set_buffer_unwritten(bh_result);

@@ -554,6 +554,15 @@ xfs_file_dio_aio_write(
 	if ((iocb->ki_pos & mp->m_blockmask) ||
 	    ((iocb->ki_pos + count) & mp->m_blockmask)) {
 		unaligned_io = 1;
+
+		/*
+		 * We can't properly handle unaligned direct I/O to reflink
+		 * files yet, as we can't unshare a partial block.
+		 */
+		if (xfs_is_reflink_inode(ip)) {
+			trace_xfs_reflink_bounce_dio_write(ip, iocb->ki_pos, count);
+			return -EREMCHG;
+		}
 		iolock = XFS_IOLOCK_EXCL;
 	} else {
 		iolock = XFS_IOLOCK_SHARED;

@@ -3353,7 +3353,7 @@ DEFINE_INODE_IREC_EVENT(xfs_reflink_conv
 DEFINE_RW_EVENT(xfs_reflink_reserve_cow);
 DEFINE_RW_EVENT(xfs_reflink_allocate_cow_range);
 
-DEFINE_INODE_IREC_EVENT(xfs_reflink_bounce_dio_write);
+DEFINE_SIMPLE_IO_EVENT(xfs_reflink_bounce_dio_write);
 DEFINE_IOMAP_EVENT(xfs_reflink_find_cow_mapping);
 DEFINE_INODE_IREC_EVENT(xfs_reflink_trim_irec);
 

@@ -45,18 +45,7 @@ xfs_extent_busy_insert(
 	struct rb_node		**rbp;
 	struct rb_node		*parent = NULL;
 
-	new = kmem_zalloc(sizeof(struct xfs_extent_busy), KM_MAYFAIL);
-	if (!new) {
-		/*
-		 * No Memory!  Since it is now not possible to track the free
-		 * block, make this a synchronous transaction to insure that
-		 * the block is not reused before this transaction commits.
-		 */
-		trace_xfs_extent_busy_enomem(tp->t_mountp, agno, bno, len);
-		xfs_trans_set_sync(tp);
-		return;
-	}
-
+	new = kmem_zalloc(sizeof(struct xfs_extent_busy), KM_SLEEP);
 	new->agno = agno;
 	new->bno = bno;
 	new->length = len;

@@ -2907,7 +2907,8 @@ xfs_bmap_add_extent_hole_delay(
 		oldlen = startblockval(left.br_startblock) +
 			startblockval(new->br_startblock) +
 			startblockval(right.br_startblock);
-		newlen = xfs_bmap_worst_indlen(ip, temp);
+		newlen = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+					 oldlen);
 		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
 			nullstartblock((int)newlen));
 		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);

@@ -2928,7 +2929,8 @@ xfs_bmap_add_extent_hole_delay(
 		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
 		oldlen = startblockval(left.br_startblock) +
 			startblockval(new->br_startblock);
-		newlen = xfs_bmap_worst_indlen(ip, temp);
+		newlen = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+					 oldlen);
 		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
 			nullstartblock((int)newlen));
 		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);

@@ -2944,7 +2946,8 @@ xfs_bmap_add_extent_hole_delay(
 		temp = new->br_blockcount + right.br_blockcount;
 		oldlen = startblockval(new->br_startblock) +
 			startblockval(right.br_startblock);
-		newlen = xfs_bmap_worst_indlen(ip, temp);
+		newlen = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+					 oldlen);
 		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
 			new->br_startoff,
 			nullstartblock((int)newlen), temp, right.br_state);

@@ -4899,34 +4899,59 @@ xfs_bmap_split_indlen(
 	xfs_filblks_t			len2 = *indlen2;
 	xfs_filblks_t			nres = len1 + len2; /* new total res. */
 	xfs_filblks_t			stolen = 0;
+	xfs_filblks_t			resfactor;
 
 	/*
 	 * Steal as many blocks as we can to try and satisfy the worst case
 	 * indlen for both new extents.
 	 */
-	while (nres > ores && avail) {
-		nres--;
-		avail--;
-		stolen++;
-	}
+	if (ores < nres && avail)
+		stolen = XFS_FILBLKS_MIN(nres - ores, avail);
+	ores += stolen;
+
+	 /* nothing else to do if we've satisfied the new reservation */
+	if (ores >= nres)
+		return stolen;
+
+	/*
+	 * We can't meet the total required reservation for the two extents.
+	 * Calculate the percent of the overall shortage between both extents
+	 * and apply this percentage to each of the requested indlen values.
+	 * This distributes the shortage fairly and reduces the chances that one
+	 * of the two extents is left with nothing when extents are repeatedly
+	 * split.
+	 */
+	resfactor = (ores * 100);
+	do_div(resfactor, nres);
+	len1 *= resfactor;
+	do_div(len1, 100);
+	len2 *= resfactor;
+	do_div(len2, 100);
+	ASSERT(len1 + len2 <= ores);
+	ASSERT(len1 < *indlen1 && len2 < *indlen2);
 
 	/*
-	 * The only blocks available are those reserved for the original
-	 * extent and what we can steal from the extent being removed.
-	 * If this still isn't enough to satisfy the combined
-	 * requirements for the two new extents, skim blocks off of each
-	 * of the new reservations until they match what is available.
+	 * Hand out the remainder to each extent. If one of the two reservations
+	 * is zero, we want to make sure that one gets a block first. The loop
+	 * below starts with len1, so hand len2 a block right off the bat if it
+	 * is zero.
 	 */
-	while (nres > ores) {
-		if (len1) {
-			len1--;
-			nres--;
+	ores -= (len1 + len2);
+	ASSERT((*indlen1 - len1) + (*indlen2 - len2) >= ores);
+	if (ores && !len2 && *indlen2) {
+		len2++;
+		ores--;
+	}
+	while (ores) {
+		if (len1 < *indlen1) {
+			len1++;
+			ores--;
 		}
-		if (nres == ores)
+		if (!ores)
 			break;
-		if (len2) {
-			len2--;
-			nres--;
+		if (len2 < *indlen2) {
+			len2++;
+			ores--;
 		}
 	}
 

@@ -363,7 +363,7 @@ xfs_reflink_convert_cow(
 	xfs_fileoff_t		end_fsb = XFS_B_TO_FSB(mp, offset + count);
 	xfs_extnum_t		idx;
 	bool			found;
-	int			error;
+	int			error = 0;
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 

@@ -1387,10 +1387,16 @@ xfs_shift_file_space(
 	xfs_fileoff_t		stop_fsb;
 	xfs_fileoff_t		next_fsb;
 	xfs_fileoff_t		shift_fsb;
+	uint			resblks;
 
 	ASSERT(direction == SHIFT_LEFT || direction == SHIFT_RIGHT);
 
 	if (direction == SHIFT_LEFT) {
+		/*
+		 * Reserve blocks to cover potential extent merges after left
+		 * shift operations.
+		 */
+		resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0);
 		next_fsb = XFS_B_TO_FSB(mp, offset + len);
 		stop_fsb = XFS_B_TO_FSB(mp, VFS_I(ip)->i_size);
 	} else {

@@ -1398,6 +1404,7 @@ xfs_shift_file_space(
 		 * If right shift, delegate the work of initialization of
 		 * next_fsb to xfs_bmap_shift_extent as it has ilock held.
 		 */
+		resblks = 0;
 		next_fsb = NULLFSBLOCK;
 		stop_fsb = XFS_B_TO_FSB(mp, offset);
 	}

@@ -1439,21 +1446,14 @@ xfs_shift_file_space(
 	}
 
 	while (!error && !done) {
-		/*
-		 * We would need to reserve permanent block for transaction.
-		 * This will come into picture when after shifting extent into
-		 * hole we found that adjacent extents can be merged which
-		 * may lead to freeing of a block during record update.
-		 */
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write,
-				XFS_DIOSTRAT_SPACE_RES(mp, 0), 0, 0, &tp);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, 0, 0,
+					&tp);
 		if (error)
 			break;
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
 		error = xfs_trans_reserve_quota(tp, mp, ip->i_udquot,
-				ip->i_gdquot, ip->i_pdquot,
-				XFS_DIOSTRAT_SPACE_RES(mp, 0), 0,
+				ip->i_gdquot, ip->i_pdquot, resblks, 0,
 				XFS_QMOPT_RES_REGBLKS);
 		if (error)
 			goto out_trans_cancel;

@@ -51,8 +51,7 @@ xfs_ialloc_cluster_alignment(
 	struct xfs_mount	*mp)
 {
 	if (xfs_sb_version_hasalign(&mp->m_sb) &&
-	    mp->m_sb.sb_inoalignmt >=
-			XFS_B_TO_FSBT(mp, mp->m_inode_cluster_size))
+	    mp->m_sb.sb_inoalignmt >= xfs_icluster_size_fsb(mp))
 		return mp->m_sb.sb_inoalignmt;
 	return 1;
 }

@@ -804,9 +804,7 @@ try_another_ag:
 	 */
 	ASSERT(args.fsbno != NULLFSBLOCK);
 	ASSERT(*firstblock == NULLFSBLOCK ||
-	       args.agno == XFS_FSB_TO_AGNO(mp, *firstblock) ||
-	       (dfops->dop_low &&
-		args.agno > XFS_FSB_TO_AGNO(mp, *firstblock)));
+	       args.agno >= XFS_FSB_TO_AGNO(mp, *firstblock));
 	*firstblock = cur->bc_private.b.firstblock = args.fsbno;
 	cur->bc_private.b.allocated++;
 	ip->i_d.di_nblocks++;

@@ -3923,17 +3921,13 @@ xfs_bmap_btalloc(
 		 * the first block that was allocated.
 		 */
 		ASSERT(*ap->firstblock == NULLFSBLOCK ||
-		       XFS_FSB_TO_AGNO(mp, *ap->firstblock) ==
-		       XFS_FSB_TO_AGNO(mp, args.fsbno) ||
-		       (ap->dfops->dop_low &&
-			XFS_FSB_TO_AGNO(mp, *ap->firstblock) <
-			XFS_FSB_TO_AGNO(mp, args.fsbno)));
+		       XFS_FSB_TO_AGNO(mp, *ap->firstblock) <=
+		       XFS_FSB_TO_AGNO(mp, args.fsbno));
 
 		ap->blkno = args.fsbno;
 		if (*ap->firstblock == NULLFSBLOCK)
 			*ap->firstblock = args.fsbno;
-		ASSERT(nullfb || fb_agno == args.agno ||
-		       (ap->dfops->dop_low && fb_agno < args.agno));
+		ASSERT(nullfb || fb_agno <= args.agno);
 		ap->length = args.len;
 		if (!(ap->flags & XFS_BMAPI_COWFORK))
 			ap->ip->i_d.di_nblocks += args.len;

@@ -4858,13 +4852,9 @@ error0:
 	if (bma.cur) {
 		if (!error) {
 			ASSERT(*firstblock == NULLFSBLOCK ||
-			       XFS_FSB_TO_AGNO(mp, *firstblock) ==
+			       XFS_FSB_TO_AGNO(mp, *firstblock) <=
 			       XFS_FSB_TO_AGNO(mp,
-				       bma.cur->bc_private.b.firstblock) ||
-			       (dfops->dop_low &&
-				XFS_FSB_TO_AGNO(mp, *firstblock) <
-				XFS_FSB_TO_AGNO(mp,
-					bma.cur->bc_private.b.firstblock)));
+				       bma.cur->bc_private.b.firstblock));
 			*firstblock = bma.cur->bc_private.b.firstblock;
 		}
 		xfs_btree_del_cursor(bma.cur,

@@ -298,7 +298,7 @@ xfs_end_io(
 			goto done;
 		if (ioend->io_bio->bi_error) {
 			error = xfs_reflink_cancel_cow_range(ip,
-					ioend->io_offset, ioend->io_size);
+					ioend->io_offset, ioend->io_size, true);
 			goto done;
 		}
 		error = xfs_reflink_end_cow(ip, ioend->io_offset,

@@ -1610,7 +1610,7 @@ xfs_inode_free_cowblocks(
 	xfs_ilock(ip, XFS_IOLOCK_EXCL);
 	xfs_ilock(ip, XFS_MMAPLOCK_EXCL);
 
-	ret = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF);
+	ret = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, false);
 
 	xfs_iunlock(ip, XFS_MMAPLOCK_EXCL);
 	xfs_iunlock(ip, XFS_IOLOCK_EXCL);

@@ -1624,7 +1624,7 @@ xfs_itruncate_extents(
 
 	/* Remove all pending CoW reservations. */
 	error = xfs_reflink_cancel_cow_blocks(ip, &tp, first_unmap_block,
-			last_block);
+			last_block, true);
 	if (error)
 		goto out;
 

@@ -571,14 +571,18 @@ xfs_reflink_trim_irec_to_next_cow(
 }
 
 /*
- * Cancel all pending CoW reservations for some block range of an inode.
+ * Cancel CoW reservations for some block range of an inode.
+ *
+ * If cancel_real is true this function cancels all COW fork extents for the
+ * inode; if cancel_real is false, real extents are not cleared.
  */
 int
 xfs_reflink_cancel_cow_blocks(
 	struct xfs_inode		*ip,
 	struct xfs_trans		**tpp,
 	xfs_fileoff_t			offset_fsb,
-	xfs_fileoff_t			end_fsb)
+	xfs_fileoff_t			end_fsb,
+	bool				cancel_real)
 {
 	struct xfs_ifork		*ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK);
 	struct xfs_bmbt_irec		got, prev, del;

@@ -605,7 +609,7 @@ xfs_reflink_cancel_cow_blocks(
 					&idx, &got, &del);
 			if (error)
 				break;
-		} else {
+		} else if (del.br_state == XFS_EXT_UNWRITTEN || cancel_real) {
 			xfs_trans_ijoin(*tpp, ip, 0);
 			xfs_defer_init(&dfops, &firstfsb);
 

@@ -648,13 +652,17 @@ xfs_reflink_cancel_cow_blocks(
 }
 
 /*
- * Cancel all pending CoW reservations for some byte range of an inode.
+ * Cancel CoW reservations for some byte range of an inode.
+ *
+ * If cancel_real is true this function cancels all COW fork extents for the
+ * inode; if cancel_real is false, real extents are not cleared.
  */
 int
 xfs_reflink_cancel_cow_range(
 	struct xfs_inode	*ip,
 	xfs_off_t		offset,
-	xfs_off_t		count)
+	xfs_off_t		count,
+	bool			cancel_real)
 {
 	struct xfs_trans	*tp;
 	xfs_fileoff_t		offset_fsb;

@@ -680,7 +688,8 @@ xfs_reflink_cancel_cow_range(
 	xfs_trans_ijoin(tp, ip, 0);
 
 	/* Scrape out the old CoW reservations */
-	error = xfs_reflink_cancel_cow_blocks(ip, &tp, offset_fsb, end_fsb);
+	error = xfs_reflink_cancel_cow_blocks(ip, &tp, offset_fsb, end_fsb,
+			cancel_real);
 	if (error)
 		goto out_cancel;
 

@@ -1686,7 +1695,7 @@ next:
 	 * We didn't find any shared blocks so turn off the reflink flag.
 	 * First, get rid of any leftover CoW mappings.
 	 */
-	error = xfs_reflink_cancel_cow_blocks(ip, tpp, 0, NULLFILEOFF);
+	error = xfs_reflink_cancel_cow_blocks(ip, tpp, 0, NULLFILEOFF, true);
 	if (error)
 		return error;
 

@@ -39,9 +39,9 @@ extern int xfs_reflink_trim_irec_to_next
 
 extern int xfs_reflink_cancel_cow_blocks(struct xfs_inode *ip,
 		struct xfs_trans **tpp, xfs_fileoff_t offset_fsb,
-		xfs_fileoff_t end_fsb);
+		xfs_fileoff_t end_fsb, bool cancel_real);
 extern int xfs_reflink_cancel_cow_range(struct xfs_inode *ip, xfs_off_t offset,
-		xfs_off_t count);
+		xfs_off_t count, bool cancel_real);
 extern int xfs_reflink_end_cow(struct xfs_inode *ip, xfs_off_t offset,
 		xfs_off_t count);
 extern int xfs_reflink_recover_cow(struct xfs_mount *mp);

@@ -948,7 +948,7 @@ xfs_fs_destroy_inode(
 	XFS_STATS_INC(ip->i_mount, vn_remove);
 
 	if (xfs_is_reflink_inode(ip)) {
-		error = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF);
+		error = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, true);
 		if (error && !XFS_FORCED_SHUTDOWN(ip->i_mount))
 			xfs_warn(ip->i_mount,
 "Error %d while evicting CoW blocks for inode %llu.",

@@ -279,54 +279,49 @@ xfs_end_io(
 	struct xfs_ioend	*ioend =
 		container_of(work, struct xfs_ioend, io_work);
 	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
+	xfs_off_t		offset = ioend->io_offset;
+	size_t			size = ioend->io_size;
 	int			error = ioend->io_bio->bi_error;
 
 	/*
-	 * Set an error if the mount has shut down and proceed with end I/O
-	 * processing so it can perform whatever cleanups are necessary.
+	 * Just clean up the in-memory strutures if the fs has been shut down.
 	 */
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
 		error = -EIO;
+		goto done;
+	}
 
 	/*
-	 * For a CoW extent, we need to move the mapping from the CoW fork
-	 * to the data fork.  If instead an error happened, just dump the
-	 * new blocks.
+	 * Clean up any COW blocks on an I/O error.
 	 */
-	if (ioend->io_type == XFS_IO_COW) {
-		if (error)
-			goto done;
-		if (ioend->io_bio->bi_error) {
-			error = xfs_reflink_cancel_cow_range(ip,
-					ioend->io_offset, ioend->io_size, true);
-			goto done;
+	if (unlikely(error)) {
+		switch (ioend->io_type) {
+		case XFS_IO_COW:
+			xfs_reflink_cancel_cow_range(ip, offset, size, true);
+			break;
 		}
-		error = xfs_reflink_end_cow(ip, ioend->io_offset,
-				ioend->io_size);
-		if (error)
-			goto done;
+
+		goto done;
 	}
 
 	/*
-	 * For unwritten extents we need to issue transactions to convert a
-	 * range to normal written extens after the data I/O has finished.
-	 * Detecting and handling completion IO errors is done individually
-	 * for each case as different cleanup operations need to be performed
-	 * on error.
+	 * Success:  commit the COW or unwritten blocks if needed.
 	 */
-	if (ioend->io_type == XFS_IO_UNWRITTEN) {
-		if (error)
-			goto done;
-		error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
-						  ioend->io_size);
-	} else if (ioend->io_append_trans) {
-		error = xfs_setfilesize_ioend(ioend, error);
-	} else {
-		ASSERT(!xfs_ioend_is_append(ioend) ||
-		       ioend->io_type == XFS_IO_COW);
+	switch (ioend->io_type) {
+	case XFS_IO_COW:
+		error = xfs_reflink_end_cow(ip, offset, size);
+		break;
+	case XFS_IO_UNWRITTEN:
+		error = xfs_iomap_write_unwritten(ip, offset, size);
+		break;
+	default:
+		ASSERT(!xfs_ioend_is_append(ioend) || ioend->io_append_trans);
+		break;
 	}
 
 done:
+	if (ioend->io_append_trans)
+		error = xfs_setfilesize_ioend(ioend, error);
 	xfs_destroy_ioend(ioend, error);
 }
 

@@ -502,8 +502,7 @@ STATIC void
 xfs_set_inoalignment(xfs_mount_t *mp)
 {
 	if (xfs_sb_version_hasalign(&mp->m_sb) &&
-	    mp->m_sb.sb_inoalignmt >=
-	    XFS_B_TO_FSBT(mp, mp->m_inode_cluster_size))
+		mp->m_sb.sb_inoalignmt >= xfs_icluster_size_fsb(mp))
 		mp->m_inoalign_mask = mp->m_sb.sb_inoalignmt - 1;
 	else
 		mp->m_inoalign_mask = 0;

@@ -4253,6 +4253,19 @@ xfs_bmapi_read(
 	return 0;
 }
 
+/*
+ * Add a delayed allocation extent to an inode. Blocks are reserved from the
+ * global pool and the extent inserted into the inode in-core extent tree.
+ *
+ * On entry, got refers to the first extent beyond the offset of the extent to
+ * allocate or eof is specified if no such extent exists. On return, got refers
+ * to the extent record that was inserted to the inode fork.
+ *
+ * Note that the allocated extent may have been merged with contiguous extents
+ * during insertion into the inode fork. Thus, got does not reflect the current
+ * state of the inode fork on return. If necessary, the caller can use lastx to
+ * look up the updated record in the inode fork.
+ */
 int
 xfs_bmapi_reserve_delalloc(
 	struct xfs_inode	*ip,

@@ -4339,13 +4352,8 @@ xfs_bmapi_reserve_delalloc(
 	got->br_startblock = nullstartblock(indlen);
 	got->br_blockcount = alen;
 	got->br_state = XFS_EXT_NORM;
-	xfs_bmap_add_extent_hole_delay(ip, whichfork, lastx, got);
 
-	/*
-	 * Update our extent pointer, given that xfs_bmap_add_extent_hole_delay
-	 * might have merged it into one of the neighbouring ones.
-	 */
-	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *lastx), got);
+	xfs_bmap_add_extent_hole_delay(ip, whichfork, lastx, got);
 
 	/*
 	 * Tag the inode if blocks were preallocated. Note that COW fork

@@ -4357,10 +4365,6 @@ xfs_bmapi_reserve_delalloc(
 	if (whichfork == XFS_COW_FORK && (prealloc || aoff < off || alen > len))
 		xfs_inode_set_cowblocks_tag(ip);
 
-	ASSERT(got->br_startoff <= aoff);
-	ASSERT(got->br_startoff + got->br_blockcount >= aoff + alen);
-	ASSERT(isnullstartblock(got->br_startblock));
-	ASSERT(got->br_state == XFS_EXT_NORM);
 	return 0;
 
 out_unreserve_blocks:

@@ -637,6 +637,11 @@ retry:
 		goto out_unlock;
 	}
 
+	/*
+	 * Flag newly allocated delalloc blocks with IOMAP_F_NEW so we punch
+	 * them out if the write happens to fail.
+	 */
+	iomap->flags = IOMAP_F_NEW;
 	trace_xfs_iomap_alloc(ip, offset, count, 0, &got);
 done:
 	if (isnullstartblock(got.br_startblock))

@@ -1061,7 +1066,8 @@ xfs_file_iomap_end_delalloc(
 	struct xfs_inode	*ip,
 	loff_t			offset,
 	loff_t			length,
-	ssize_t			written)
+	ssize_t			written,
+	struct iomap		*iomap)
 {
 	struct xfs_mount	*mp = ip->i_mount;
 	xfs_fileoff_t		start_fsb;

@@ -1080,14 +1086,14 @@ xfs_file_iomap_end_delalloc(
 	end_fsb = XFS_B_TO_FSB(mp, offset + length);
 
 	/*
-	 * Trim back delalloc blocks if we didn't manage to write the whole
-	 * range reserved.
+	 * Trim delalloc blocks if they were allocated by this write and we
+	 * didn't manage to write the whole range.
 	 *
 	 * We don't need to care about racing delalloc as we hold i_mutex
 	 * across the reserve/allocate/unreserve calls. If there are delalloc
 	 * blocks in the range, they are ours.
 	 */
-	if (start_fsb < end_fsb) {
+	if ((iomap->flags & IOMAP_F_NEW) && start_fsb < end_fsb) {
 		truncate_pagecache_range(VFS_I(ip), XFS_FSB_TO_B(mp, start_fsb),
 					 XFS_FSB_TO_B(mp, end_fsb) - 1);
 

@@ -1117,7 +1123,7 @@ xfs_file_iomap_end(
 {
 	if ((flags & IOMAP_WRITE) && iomap->type == IOMAP_DELALLOC)
 		return xfs_file_iomap_end_delalloc(XFS_I(inode), offset,
-				length, written);
+				length, written, iomap);
 	return 0;
 }
 

@@ -769,8 +769,8 @@ xfs_bmap_extents_to_btree(
 		args.type = XFS_ALLOCTYPE_START_BNO;
 		args.fsbno = XFS_INO_TO_FSB(mp, ip->i_ino);
 	} else if (dfops->dop_low) {
-try_another_ag:
 		args.type = XFS_ALLOCTYPE_START_BNO;
+try_another_ag:
 		args.fsbno = *firstblock;
 	} else {
 		args.type = XFS_ALLOCTYPE_NEAR_BNO;

@@ -796,13 +796,17 @@ try_another_ag:
 	if (xfs_sb_version_hasreflink(&cur->bc_mp->m_sb) &&
 	    args.fsbno == NULLFSBLOCK &&
 	    args.type == XFS_ALLOCTYPE_NEAR_BNO) {
-		dfops->dop_low = true;
+		args.type = XFS_ALLOCTYPE_FIRST_AG;
 		goto try_another_ag;
 	}
+	if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
+		xfs_iroot_realloc(ip, -1, whichfork);
+		xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
+		return -ENOSPC;
+	}
 	/*
 	 * Allocation can't fail, the space was reserved.
 	 */
-	ASSERT(args.fsbno != NULLFSBLOCK);
 	ASSERT(*firstblock == NULLFSBLOCK ||
 	       args.agno >= XFS_FSB_TO_AGNO(mp, *firstblock));
 	*firstblock = cur->bc_private.b.firstblock = args.fsbno;

@@ -453,8 +453,8 @@ xfs_bmbt_alloc_block(
 
 	if (args.fsbno == NULLFSBLOCK) {
 		args.fsbno = be64_to_cpu(start->l);
-try_another_ag:
 		args.type = XFS_ALLOCTYPE_START_BNO;
+try_another_ag:
 		/*
 		 * Make sure there is sufficient room left in the AG to
 		 * complete a full tree split for an extent insert.  If

@@ -494,8 +494,8 @@ try_another_ag:
 	if (xfs_sb_version_hasreflink(&cur->bc_mp->m_sb) &&
 	    args.fsbno == NULLFSBLOCK &&
 	    args.type == XFS_ALLOCTYPE_NEAR_BNO) {
-		cur->bc_private.b.dfops->dop_low = true;
 		args.fsbno = cur->bc_private.b.firstblock;
+		args.type = XFS_ALLOCTYPE_FIRST_AG;
 		goto try_another_ag;
 	}
 

@@ -512,7 +512,7 @@ try_another_ag:
 			goto error0;
 		cur->bc_private.b.dfops->dop_low = true;
 	}
-	if (args.fsbno == NULLFSBLOCK) {
+	if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
 		XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT);
 		*stat = 0;
 		return 0;

@@ -998,6 +998,8 @@ sg_ioctl(struct file *filp, unsigned int
 		result = get_user(val, ip);
 		if (result)
 			return result;
+		if (val > SG_MAX_CDB_SIZE)
+			return -ENOMEM;
 		sfp->next_cmd_len = (val > 0) ? val : 0;
 		return 0;
 	case SG_GET_VERSION_NUM:

@@ -221,7 +221,7 @@ static unsigned int sas_ata_qc_issue(str
 		task->num_scatter = qc->n_elem;
 	} else {
 		for_each_sg(qc->sg, sg, qc->n_elem, si)
-			xfer += sg->length;
+			xfer += sg_dma_len(sg);
 
 		task->total_xfer_len = xfer;
 		task->num_scatter = si;

@@ -870,7 +870,7 @@ static void alua_rtpg_queue(struct alua_
 	unsigned long flags;
 	struct workqueue_struct *alua_wq = kaluad_wq;
 
-	if (!pg)
+	if (!pg || scsi_device_get(sdev))
 		return;
 
 	spin_lock_irqsave(&pg->lock, flags);

@@ -884,14 +884,12 @@ static void alua_rtpg_queue(struct alua_
 		pg->flags |= ALUA_PG_RUN_RTPG;
 		kref_get(&pg->kref);
 		pg->rtpg_sdev = sdev;
-		scsi_device_get(sdev);
 		start_queue = 1;
 	} else if (!(pg->flags & ALUA_PG_RUN_RTPG) && force) {
 		pg->flags |= ALUA_PG_RUN_RTPG;
 		/* Do not queue if the worker is already running */
 		if (!(pg->flags & ALUA_PG_RUNNING)) {
 			kref_get(&pg->kref);
-			sdev = NULL;
 			start_queue = 1;
 		}
 	}

@@ -900,13 +898,15 @@ static void alua_rtpg_queue(struct alua_
 		alua_wq = kaluad_sync_wq;
 	spin_unlock_irqrestore(&pg->lock, flags);
 
-	if (start_queue &&
-	    !queue_delayed_work(alua_wq, &pg->rtpg_work,
-				msecs_to_jiffies(ALUA_RTPG_DELAY_MSECS))) {
-		if (sdev)
-			scsi_device_put(sdev);
-		kref_put(&pg->kref, release_port_group);
+	if (start_queue) {
+		if (queue_delayed_work(alua_wq, &pg->rtpg_work,
+				msecs_to_jiffies(ALUA_RTPG_DELAY_MSECS)))
+			sdev = NULL;
+		else
+			kref_put(&pg->kref, release_port_group);
 	}
+	if (sdev)
+		scsi_device_put(sdev);
 }
 
 /*

@@ -113,7 +113,7 @@ struct alua_queue_data {
 #define ALUA_POLICY_SWITCH_ALL		1
 
 static void alua_rtpg_work(struct work_struct *work);
-static void alua_rtpg_queue(struct alua_port_group *pg,
+static bool alua_rtpg_queue(struct alua_port_group *pg,
 			    struct scsi_device *sdev,
 			    struct alua_queue_data *qdata, bool force);
 static void alua_check(struct scsi_device *sdev, bool force);

@@ -862,7 +862,13 @@ static void alua_rtpg_work(struct work_s
 	kref_put(&pg->kref, release_port_group);
 }
 
-static void alua_rtpg_queue(struct alua_port_group *pg,
+/**
+ * alua_rtpg_queue() - cause RTPG to be submitted asynchronously
+ *
+ * Returns true if and only if alua_rtpg_work() will be called asynchronously.
+ * That function is responsible for calling @qdata->fn().
+ */
+static bool alua_rtpg_queue(struct alua_port_group *pg,
 			    struct scsi_device *sdev,
 			    struct alua_queue_data *qdata, bool force)
 {

@@ -871,7 +877,7 @@ static void alua_rtpg_queue(struct alua_
 	struct workqueue_struct *alua_wq = kaluad_wq;
 
 	if (!pg || scsi_device_get(sdev))
-		return;
+		return false;
 
 	spin_lock_irqsave(&pg->lock, flags);
 	if (qdata) {

@@ -907,6 +913,8 @@ static void alua_rtpg_queue(struct alua_
 	}
 	if (sdev)
 		scsi_device_put(sdev);
+
+	return true;
 }
 
 /*

@@ -1007,11 +1015,13 @@ static int alua_activate(struct scsi_dev
 		mutex_unlock(&h->init_mutex);
 		goto out;
 	}
-	fn = NULL;
 	rcu_read_unlock();
 	mutex_unlock(&h->init_mutex);
 
-	alua_rtpg_queue(pg, sdev, qdata, true);
+	if (alua_rtpg_queue(pg, sdev, qdata, true))
+		fn = NULL;
+	else
+		err = SCSI_DH_DEV_OFFLINED;
 	kref_put(&pg->kref, release_port_group);
 out:
 	if (fn)

@@ -265,6 +265,10 @@ int snd_seq_fifo_resize(struct snd_seq_f
 	/* NOTE: overflow flag is not cleared */
 	spin_unlock_irqrestore(&f->lock, flags);
 
+	/* close the old pool and wait until all users are gone */
+	snd_seq_pool_mark_closing(oldpool);
+	snd_use_lock_sync(&f->use_lock);
+
 	/* release cells in old pool */
 	for (cell = oldhead; cell; cell = next) {
 		next = cell->next;

@@ -4846,6 +4846,7 @@ enum {
 	ALC292_FIXUP_DISABLE_AAMIX,
 	ALC293_FIXUP_DISABLE_AAMIX_MULTIJACK,
 	ALC298_FIXUP_DELL1_MIC_NO_PRESENCE,
+	ALC298_FIXUP_DELL_AIO_MIC_NO_PRESENCE,
 	ALC275_FIXUP_DELL_XPS,
 	ALC256_FIXUP_DELL_XPS_13_HEADPHONE_NOISE,
 	ALC293_FIXUP_LENOVO_SPK_NOISE,

@@ -5446,6 +5447,15 @@ static const struct hda_fixup alc269_fix
 		.chained = true,
 		.chain_id = ALC269_FIXUP_HEADSET_MODE
 	},
+	[ALC298_FIXUP_DELL_AIO_MIC_NO_PRESENCE] = {
+		.type = HDA_FIXUP_PINS,
+		.v.pins = (const struct hda_pintbl[]) {
+			{ 0x18, 0x01a1913c }, /* use as headset mic, without its own jack detect */
+			{ }
+		},
+		.chained = true,
+		.chain_id = ALC269_FIXUP_HEADSET_MODE
+	},
 	[ALC275_FIXUP_DELL_XPS] = {
 		.type = HDA_FIXUP_VERBS,
 		.v.verbs = (const struct hda_verb[]) {

@@ -5518,7 +5528,7 @@ static const struct hda_fixup alc269_fix
 		.type = HDA_FIXUP_FUNC,
 		.v.func = alc298_fixup_speaker_volume,
 		.chained = true,
-		.chain_id = ALC298_FIXUP_DELL1_MIC_NO_PRESENCE,
+		.chain_id = ALC298_FIXUP_DELL_AIO_MIC_NO_PRESENCE,
 	},
 	[ALC256_FIXUP_DELL_INSPIRON_7559_SUBWOOFER] = {
 		.type = HDA_FIXUP_PINS,

@@ -349,7 +349,7 @@ static int atmel_classd_codec_dai_digita
 }
 
 #define CLASSD_ACLK_RATE_11M2896_MPY_8 (112896 * 100 * 8)
-#define CLASSD_ACLK_RATE_12M288_MPY_8  (12228 * 1000 * 8)
+#define CLASSD_ACLK_RATE_12M288_MPY_8  (12288 * 1000 * 8)
 
 static struct {
 	int rate;

@@ -448,7 +448,7 @@ static int skl_tplg_set_module_init_data
 			if (bc->set_params != SKL_PARAM_INIT)
 				continue;
 
-			mconfig->formats_config.caps = (u32 *)&bc->params;
+			mconfig->formats_config.caps = (u32 *)bc->params;
 			mconfig->formats_config.caps_size = bc->size;
 
 			break;

@@ -2017,6 +2017,14 @@ static int wacom_parse_and_register(stru
 
 	wacom_update_name(wacom, wireless ? " (WL)" : "");
 
+	/* pen only Bamboo neither support touch nor pad */
+	if ((features->type == BAMBOO_PEN) &&
+	    ((features->device_type & WACOM_DEVICETYPE_TOUCH) ||
+	    (features->device_type & WACOM_DEVICETYPE_PAD))) {
+		error = -ENODEV;
+		goto fail;
+	}
+
 	error = wacom_add_shared_data(hdev);
 	if (error)
 		goto fail;

@@ -2063,14 +2071,6 @@ static int wacom_parse_and_register(stru
 		error = -ENODEV;
 		goto fail_quirks;
 	}
-
-	/* pen only Bamboo neither support touch nor pad */
-	if ((features->type == BAMBOO_PEN) &&
-	    ((features->device_type & WACOM_DEVICETYPE_TOUCH) ||
-	    (features->device_type & WACOM_DEVICETYPE_PAD))) {
-		error = -ENODEV;
-		goto fail_quirks;
-	}
 
 	if (features->device_type & WACOM_DEVICETYPE_WL_MONITOR)
 		error = hid_hw_open(hdev);

@@ -1823,6 +1823,9 @@ static void sdhci_enable_sdio_irq(struct
 	struct sdhci_host *host = mmc_priv(mmc);
 	unsigned long flags;
 
+	if (enable)
+		pm_runtime_get_noresume(host->mmc->parent);
+
 	spin_lock_irqsave(&host->lock, flags);
 	if (enable)
 		host->flags |= SDHCI_SDIO_IRQ_ENABLED;

@@ -1831,6 +1834,9 @@ static void sdhci_enable_sdio_irq(struct
 
 	sdhci_enable_sdio_irq_nolock(host, enable);
 	spin_unlock_irqrestore(&host->lock, flags);
+
+	if (!enable)
+		pm_runtime_put_noidle(host->mmc->parent);
 }
 
 static int sdhci_start_signal_voltage_switch(struct mmc_host *mmc,

@@ -29,6 +29,8 @@
 
 #include "sdhci-pltfm.h"
 
+#define SDMMC_MC1R	0x204
+#define		SDMMC_MC1R_DDR		BIT(3)
 #define SDMMC_CACR	0x230
 #define		SDMMC_CACR_CAPWREN	BIT(0)
 #define		SDMMC_CACR_KEY		(0x46 << 8)

@@ -103,11 +105,18 @@ static void sdhci_at91_set_power(struct
 	sdhci_set_power_noreg(host, mode, vdd);
 }
 
+void sdhci_at91_set_uhs_signaling(struct sdhci_host *host, unsigned int timing)
+{
+	if (timing == MMC_TIMING_MMC_DDR52)
+		sdhci_writeb(host, SDMMC_MC1R_DDR, SDMMC_MC1R);
+	sdhci_set_uhs_signaling(host, timing);
+}
+
 static const struct sdhci_ops sdhci_at91_sama5d2_ops = {
 	.set_clock		= sdhci_at91_set_clock,
 	.set_bus_width		= sdhci_set_bus_width,
 	.reset			= sdhci_reset,
-	.set_uhs_signaling	= sdhci_set_uhs_signaling,
+	.set_uhs_signaling	= sdhci_at91_set_uhs_signaling,
 	.set_power		= sdhci_at91_set_power,
 };
 

@@ -2532,17 +2532,14 @@ static void nfs41_check_delegation_state
 	}
 
 	nfs4_stateid_copy(&stateid, &delegation->stateid);
-	if (test_bit(NFS_DELEGATION_REVOKED, &delegation->flags)) {
+	if (test_bit(NFS_DELEGATION_REVOKED, &delegation->flags) ||
+		!test_and_clear_bit(NFS_DELEGATION_TEST_EXPIRED,
+			&delegation->flags)) {
 		rcu_read_unlock();
 		nfs_finish_clear_delegation_stateid(state, &stateid);
 		return;
 	}
 
-	if (!test_and_clear_bit(NFS_DELEGATION_TEST_EXPIRED, &delegation->flags)) {
-		rcu_read_unlock();
-		return;
-	}
-
 	cred = get_rpccred(delegation->cred);
 	rcu_read_unlock();
 	status = nfs41_test_and_free_expired_stateid(server, &stateid, cred);

@@ -790,6 +790,7 @@ nfserrno (int errno)
 		{ nfserr_serverfault, -ESERVERFAULT },
 		{ nfserr_serverfault, -ENFILE },
 		{ nfserr_io, -EUCLEAN },
+		{ nfserr_perm, -ENOKEY },
 	};
 	int	i;
 

@@ -68,6 +68,15 @@ struct exception_table_entry {
 	".previous\n"
 
 /*
+ * ASM_EXCEPTIONTABLE_ENTRY_EFAULT() creates a special exception table entry
+ * (with lowest bit set) for which the fault handler in fixup_exception() will
+ * load -EFAULT into %r8 for a read or write fault, and zeroes the target
+ * register in case of a read fault in get_user().
+ */
+#define ASM_EXCEPTIONTABLE_ENTRY_EFAULT( fault_addr, except_addr )\
+	ASM_EXCEPTIONTABLE_ENTRY( fault_addr, except_addr + 1)
+
+/*
  * The page fault handler stores, in a per-cpu area, the following information
  * if a fixup routine is available.
  */

@@ -94,7 +103,7 @@ struct exception_data {
 #define __get_user(x, ptr)                               \
 ({                                                       \
 	register long __gu_err __asm__ ("r8") = 0;       \
-	register long __gu_val __asm__ ("r9") = 0;       \
+	register long __gu_val;				 \
 							 \
 	load_sr2();					 \
 	switch (sizeof(*(ptr))) {			 \

@@ -110,22 +119,23 @@ struct exception_data {
 })
 
 #define __get_user_asm(ldx, ptr)                        \
-	__asm__("\n1:\t" ldx "\t0(%%sr2,%2),%0\n\t"	\
-		ASM_EXCEPTIONTABLE_ENTRY(1b, fixup_get_user_skip_1)\
+	__asm__("1: " ldx " 0(%%sr2,%2),%0\n"		\
+		"9:\n"					\
+		ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b)	\
 		: "=r"(__gu_val), "=r"(__gu_err)        \
-		: "r"(ptr), "1"(__gu_err)		\
-		: "r1");
+		: "r"(ptr), "1"(__gu_err));
 
 #if !defined(CONFIG_64BIT)
 
 #define __get_user_asm64(ptr) 				\
-	__asm__("\n1:\tldw 0(%%sr2,%2),%0"		\
-		"\n2:\tldw 4(%%sr2,%2),%R0\n\t"		\
-		ASM_EXCEPTIONTABLE_ENTRY(1b, fixup_get_user_skip_2)\
-		ASM_EXCEPTIONTABLE_ENTRY(2b, fixup_get_user_skip_1)\
+	__asm__("   copy %%r0,%R0\n"			\
+		"1: ldw 0(%%sr2,%2),%0\n"		\
+		"2: ldw 4(%%sr2,%2),%R0\n"		\
+		"9:\n"					\
+		ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b)	\
+		ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 9b)	\
 		: "=r"(__gu_val), "=r"(__gu_err)	\
-		: "r"(ptr), "1"(__gu_err)		\
-		: "r1");
+		: "r"(ptr), "1"(__gu_err));
 
 #endif /* !defined(CONFIG_64BIT) */
 

@@ -151,32 +161,31 @@ struct exception_data {
  * The "__put_user/kernel_asm()" macros tell gcc they read from memory
  * instead of writing. This is because they do not write to any memory
  * gcc knows about, so there are no aliasing issues. These macros must
- * also be aware that "fixup_put_user_skip_[12]" are executed in the
- * context of the fault, and any registers used there must be listed
- * as clobbers. In this case only "r1" is used by the current routines.
- * r8/r9 are already listed as err/val.
+ * also be aware that fixups are executed in the context of the fault,
+ * and any registers used there must be listed as clobbers.
+ * r8 is already listed as err.
  */
 
 #define __put_user_asm(stx, x, ptr)                         \
 	__asm__ __volatile__ (                              \
-		"\n1:\t" stx "\t%2,0(%%sr2,%1)\n\t"	    \
-		ASM_EXCEPTIONTABLE_ENTRY(1b, fixup_put_user_skip_1)\
+		"1: " stx " %2,0(%%sr2,%1)\n"		    \
+		"9:\n"					    \
+		ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b)	    \
 		: "=r"(__pu_err)                            \
-		: "r"(ptr), "r"(x), "0"(__pu_err)	    \
-		: "r1")
+		: "r"(ptr), "r"(x), "0"(__pu_err))
 
 
 #if !defined(CONFIG_64BIT)
 
 #define __put_user_asm64(__val, ptr) do {	    	    \
 	__asm__ __volatile__ (				    \
-		"\n1:\tstw %2,0(%%sr2,%1)"		    \
-		"\n2:\tstw %R2,4(%%sr2,%1)\n\t"		    \
-		ASM_EXCEPTIONTABLE_ENTRY(1b, fixup_put_user_skip_2)\
-		ASM_EXCEPTIONTABLE_ENTRY(2b, fixup_put_user_skip_1)\
+		"1: stw %2,0(%%sr2,%1)\n"		    \
+		"2: stw %R2,4(%%sr2,%1)\n"		    \
+		"9:\n"					    \
+		ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b)	    \
+		ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 9b)	    \
 		: "=r"(__pu_err)                            \
-		: "r"(ptr), "r"(__val), "0"(__pu_err) \
-		: "r1");				    \
+		: "r"(ptr), "r"(__val), "0"(__pu_err));	    \
 } while (0)
 
 #endif /* !defined(CONFIG_64BIT) */

@@ -47,16 +47,6 @@ EXPORT_SYMBOL(__cmpxchg_u64);
 EXPORT_SYMBOL(lclear_user);
 EXPORT_SYMBOL(lstrnlen_user);
 
-/* Global fixups - defined as int to avoid creation of function pointers */
-extern int fixup_get_user_skip_1;
-extern int fixup_get_user_skip_2;
-extern int fixup_put_user_skip_1;
-extern int fixup_put_user_skip_2;
-EXPORT_SYMBOL(fixup_get_user_skip_1);
-EXPORT_SYMBOL(fixup_get_user_skip_2);
-EXPORT_SYMBOL(fixup_put_user_skip_1);
-EXPORT_SYMBOL(fixup_put_user_skip_2);
-
 #ifndef CONFIG_64BIT
 /* Needed so insmod can set dp value */
 extern int $global$;

@@ -149,6 +149,23 @@ int fixup_exception(struct pt_regs *regs
 		d->fault_space = regs->isr;
 		d->fault_addr = regs->ior;
 
+		/*
+		 * Fix up get_user() and put_user().
+		 * ASM_EXCEPTIONTABLE_ENTRY_EFAULT() sets the least-significant
+		 * bit in the relative address of the fixup routine to indicate
+		 * that %r8 should be loaded with -EFAULT to report a userspace
+		 * access error.
+		 */
+		if (fix->fixup & 1) {
+			regs->gr[8] = -EFAULT;
+
+			/* zero target register for get_user() */
+			if (parisc_acctyp(0, regs->iir) == VM_READ) {
+				int treg = regs->iir & 0x1f;
+				regs->gr[treg] = 0;
+			}
+		}
+
 		regs->iaoq[0] = (unsigned long)&fix->fixup + fix->fixup;
 		regs->iaoq[0] &= ~3;
 		/*

@@ -140,6 +140,8 @@ void machine_power_off(void)
 	printk(KERN_EMERG "System shut down completed.\n"
 	       "Please power this system off now.");
 
+	/* prevent soft lockup/stalled CPU messages for endless loop. */
+	rcu_sysrq_start();
 	for (;;);
 }
 

489 lines not on this page · raw message

@@ -25,9 +25,11 @@
 ACPI_MODULE_NAME("platform");
 
 static const struct acpi_device_id forbidden_id_list[] = {
-	{"PNP0000", 0},	/* PIC */
-	{"PNP0100", 0},	/* Timer */
-	{"PNP0200", 0},	/* AT DMA Controller */
+	{"PNP0000",  0},	/* PIC */
+	{"PNP0100",  0},	/* Timer */
+	{"PNP0200",  0},	/* AT DMA Controller */
+	{"ACPI0009", 0},	/* IOxAPIC */
+	{"ACPI000A", 0},	/* IOAPIC */
 	{"", 0},
 };