Merge branch 'for-linus' of git://oss.sgi.com:8090/xfs/xfs-2.6

* 'for-linus' of git://oss.sgi.com:8090/xfs/xfs-2.6: (59 commits)
  [XFS] eagerly remove vmap mappings to avoid upsetting Xen
  [XFS] simplify validata_fields
  [XFS] no longer using io_vnode, as was remaining from 23 cherrypick
  [XFS] Remove STATIC which was missing from prior manual merge
  [XFS] Put back the QUEUE_ORDERED_NONE test in the barrier check.
  [XFS] Turn off XBF_ASYNC flag before re-reading superblock.
  [XFS] avoid race in sync_inodes() that can fail to write out all dirty data
  [XFS] This fix prevents bulkstat from spinning in an infinite loop.
  [XFS] simplify xfs_create/mknod/symlink prototype
  [XFS] avoid xfs_getattr in XFS_IOC_FSGETXATTR ioctl
  [XFS] get_bulkall() could return incorrect inode state
  [XFS] Kill unused IOMAP_EOF flag
  [XFS] fix when DMAPI mount option processing happens
  [XFS] ensure file size is logged on synchronous writes
  [XFS] growlock should be a mutex
  [XFS] replace some large xfs_log_priv.h macros by proper functions
  [XFS] kill struct bhv_vfs
  [XFS] move syncing related members from struct bhv_vfs to struct xfs_mount
  [XFS] kill the vfs_flags member in struct bhv_vfs
  [XFS] kill the vfs_fsid and vfs_altfsid members in struct bhv_vfs
  ...

22 files changed, 656 insertions, 1514 deletions

diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 52bd08c0a27..2e34b104107 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -37,6 +37,7 @@
 #include "xfs_error.h"
 #include "xfs_rw.h"
 #include "xfs_iomap.h"
+#include "xfs_vnodeops.h"
 #include <linux/mpage.h>
 #include <linux/pagevec.h>
 #include <linux/writeback.h>
@@ -139,9 +140,11 @@ xfs_destroy_ioend(
 		next = bh->b_private;
 		bh->b_end_io(bh, !ioend->io_error);
 	}
-	if (unlikely(ioend->io_error))
-		vn_ioerror(ioend->io_vnode, ioend->io_error, __FILE__,__LINE__);
-	vn_iowake(ioend->io_vnode);
+	if (unlikely(ioend->io_error)) {
+		vn_ioerror(XFS_I(ioend->io_inode), ioend->io_error,
+				__FILE__,__LINE__);
+	}
+	vn_iowake(XFS_I(ioend->io_inode));
 	mempool_free(ioend, xfs_ioend_pool);
 }
 
@@ -156,14 +159,10 @@ STATIC void
 xfs_setfilesize(
 	xfs_ioend_t		*ioend)
 {
-	xfs_inode_t		*ip;
+	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
 	xfs_fsize_t		isize;
 	xfs_fsize_t		bsize;
 
-	ip = xfs_vtoi(ioend->io_vnode);
-	if (!ip)
-		return;
-
 	ASSERT((ip->i_d.di_mode & S_IFMT) == S_IFREG);
 	ASSERT(ioend->io_type != IOMAP_READ);
 
@@ -181,7 +180,7 @@ xfs_setfilesize(
 		ip->i_d.di_size = isize;
 		ip->i_update_core = 1;
 		ip->i_update_size = 1;
-		mark_inode_dirty_sync(vn_to_inode(ioend->io_vnode));
+		mark_inode_dirty_sync(ioend->io_inode);
 	}
 
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
@@ -227,12 +226,12 @@ xfs_end_bio_unwritten(
 {
 	xfs_ioend_t		*ioend =
 		container_of(work, xfs_ioend_t, io_work);
-	bhv_vnode_t		*vp = ioend->io_vnode;
 	xfs_off_t		offset = ioend->io_offset;
 	size_t			size = ioend->io_size;
 
 	if (likely(!ioend->io_error)) {
-		bhv_vop_bmap(vp, offset, size, BMAPI_UNWRITTEN, NULL, NULL);
+		xfs_bmap(XFS_I(ioend->io_inode), offset, size,
+				BMAPI_UNWRITTEN, NULL, NULL);
 		xfs_setfilesize(ioend);
 	}
 	xfs_destroy_ioend(ioend);
@@ -275,10 +274,10 @@ xfs_alloc_ioend(
 	ioend->io_error = 0;
 	ioend->io_list = NULL;
 	ioend->io_type = type;
-	ioend->io_vnode = vn_from_inode(inode);
+	ioend->io_inode = inode;
 	ioend->io_buffer_head = NULL;
 	ioend->io_buffer_tail = NULL;
-	atomic_inc(&ioend->io_vnode->v_iocount);
+	atomic_inc(&XFS_I(ioend->io_inode)->i_iocount);
 	ioend->io_offset = 0;
 	ioend->io_size = 0;
 
@@ -302,12 +301,13 @@ xfs_map_blocks(
 	xfs_iomap_t		*mapp,
 	int			flags)
 {
-	bhv_vnode_t		*vp = vn_from_inode(inode);
+	xfs_inode_t		*ip = XFS_I(inode);
 	int			error, nmaps = 1;
 
-	error = bhv_vop_bmap(vp, offset, count, flags, mapp, &nmaps);
+	error = xfs_bmap(ip, offset, count,
+				flags, mapp, &nmaps);
 	if (!error && (flags & (BMAPI_WRITE|BMAPI_ALLOCATE)))
-		VMODIFY(vp);
+		xfs_iflags_set(ip, XFS_IMODIFIED);
 	return -error;
 }
 
@@ -497,7 +497,7 @@ xfs_cancel_ioend(
 			unlock_buffer(bh);
 		} while ((bh = next_bh) != NULL);
 
-		vn_iowake(ioend->io_vnode);
+		vn_iowake(XFS_I(ioend->io_inode));
 		mempool_free(ioend, xfs_ioend_pool);
 	} while ((ioend = next) != NULL);
 }
@@ -1237,10 +1237,7 @@ xfs_vm_writepages(
 	struct address_space	*mapping,
 	struct writeback_control *wbc)
 {
-	struct bhv_vnode	*vp = vn_from_inode(mapping->host);
-
-	if (VN_TRUNC(vp))
-		VUNTRUNCATE(vp);
+	xfs_iflags_clear(XFS_I(mapping->host), XFS_ITRUNCATED);
 	return generic_writepages(mapping, wbc);
 }
 
@@ -1317,7 +1314,6 @@ __xfs_get_blocks(
 	int			direct,
 	bmapi_flags_t		flags)
 {
-	bhv_vnode_t		*vp = vn_from_inode(inode);
 	xfs_iomap_t		iomap;
 	xfs_off_t		offset;
 	ssize_t			size;
@@ -1327,7 +1323,7 @@ __xfs_get_blocks(
 	offset = (xfs_off_t)iblock << inode->i_blkbits;
 	ASSERT(bh_result->b_size >= (1 << inode->i_blkbits));
 	size = bh_result->b_size;
-	error = bhv_vop_bmap(vp, offset, size,
+	error = xfs_bmap(XFS_I(inode), offset, size,
 			     create ? flags : BMAPI_READ, &iomap, &niomap);
 	if (error)
 		return -error;
@@ -1475,13 +1471,13 @@ xfs_vm_direct_IO(
 {
 	struct file	*file = iocb->ki_filp;
 	struct inode	*inode = file->f_mapping->host;
-	bhv_vnode_t	*vp = vn_from_inode(inode);
 	xfs_iomap_t	iomap;
 	int		maps = 1;
 	int		error;
 	ssize_t		ret;
 
-	error = bhv_vop_bmap(vp, offset, 0, BMAPI_DEVICE, &iomap, &maps);
+	error = xfs_bmap(XFS_I(inode), offset, 0,
+				BMAPI_DEVICE, &iomap, &maps);
 	if (error)
 		return -error;
 
@@ -1527,12 +1523,13 @@ xfs_vm_bmap(
 	sector_t		block)
 {
 	struct inode		*inode = (struct inode *)mapping->host;
-	bhv_vnode_t		*vp = vn_from_inode(inode);
+	struct xfs_inode	*ip = XFS_I(inode);
 
-	vn_trace_entry(vp, __FUNCTION__, (inst_t *)__return_address);
-	bhv_vop_rwlock(vp, VRWLOCK_READ);
-	bhv_vop_flush_pages(vp, (xfs_off_t)0, -1, 0, FI_REMAPF);
-	bhv_vop_rwunlock(vp, VRWLOCK_READ);
+	vn_trace_entry(XFS_I(inode), __FUNCTION__,
+			(inst_t *)__return_address);
+	xfs_rwlock(ip, VRWLOCK_READ);
+	xfs_flush_pages(ip, (xfs_off_t)0, -1, 0, FI_REMAPF);
+	xfs_rwunlock(ip, VRWLOCK_READ);
 	return generic_block_bmap(mapping, block, xfs_get_blocks);
 }
 
diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h
index 2244e516b66..3ba0631a381 100644
--- a/fs/xfs/linux-2.6/xfs_aops.h
+++ b/fs/xfs/linux-2.6/xfs_aops.h
@@ -32,7 +32,7 @@ typedef struct xfs_ioend {
 	unsigned int		io_type;	/* delalloc / unwritten */
 	int			io_error;	/* I/O error code */
 	atomic_t		io_remaining;	/* hold count */
-	struct bhv_vnode	*io_vnode;	/* file being written to */
+	struct inode		*io_inode;	/* file being written to */
 	struct buffer_head	*io_buffer_head;/* buffer linked list head */
 	struct buffer_head	*io_buffer_tail;/* buffer linked list tail */
 	size_t			io_size;	/* size of the extent */
diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
index 39f44ee572e..b9c8589e05c 100644
--- a/fs/xfs/linux-2.6/xfs_buf.c
+++ b/fs/xfs/linux-2.6/xfs_buf.c
@@ -187,6 +187,19 @@ free_address(
 {
 	a_list_t	*aentry;
 
+#ifdef CONFIG_XEN
+	/*
+	 * Xen needs to be able to make sure it can get an exclusive
+	 * RO mapping of pages it wants to turn into a pagetable.  If
+	 * a newly allocated page is also still being vmap()ed by xfs,
+	 * it will cause pagetable construction to fail.  This is a
+	 * quick workaround to always eagerly unmap pages so that Xen
+	 * is happy.
+	 */
+	vunmap(addr);
+	return;
+#endif
+
 	aentry = kmalloc(sizeof(a_list_t), GFP_NOWAIT);
 	if (likely(aentry)) {
 		spin_lock(&as_lock);
@@ -997,7 +1010,18 @@ xfs_buf_iodone_work(
 	xfs_buf_t		*bp =
 		container_of(work, xfs_buf_t, b_iodone_work);
 
-	if (bp->b_iodone)
+	/*
+	 * We can get an EOPNOTSUPP to ordered writes.  Here we clear the
+	 * ordered flag and reissue them.  Because we can't tell the higher
+	 * layers directly that they should not issue ordered I/O anymore, they
+	 * need to check if the ordered flag was cleared during I/O completion.
+	 */
+	if ((bp->b_error == EOPNOTSUPP) &&
+	    (bp->b_flags & (XBF_ORDERED|XBF_ASYNC)) == (XBF_ORDERED|XBF_ASYNC)) {
+		XB_TRACE(bp, "ordered_retry", bp->b_iodone);
+		bp->b_flags &= ~XBF_ORDERED;
+		xfs_buf_iorequest(bp);
+	} else if (bp->b_iodone)
 		(*(bp->b_iodone))(bp);
 	else if (bp->b_flags & XBF_ASYNC)
 		xfs_buf_relse(bp);
diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c
index e3a5fedac1b..726449d4fd2 100644
--- a/fs/xfs/linux-2.6/xfs_export.c
+++ b/fs/xfs/linux-2.6/xfs_export.c
@@ -17,12 +17,18 @@
  */
 #include "xfs.h"
 #include "xfs_types.h"
-#include "xfs_dmapi.h"
+#include "xfs_inum.h"
 #include "xfs_log.h"
 #include "xfs_trans.h"
 #include "xfs_sb.h"
+#include "xfs_ag.h"
+#include "xfs_dmapi.h"
 #include "xfs_mount.h"
 #include "xfs_export.h"
+#include "xfs_vnodeops.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_inode.h"
+#include "xfs_vfsops.h"
 
 static struct dentry dotdot = { .d_name.name = "..", .d_name.len = 2, };
 
@@ -96,9 +102,7 @@ xfs_fs_encode_fh(
 	int			len;
 	int			is64 = 0;
 #if XFS_BIG_INUMS
-	bhv_vfs_t		*vfs = vfs_from_sb(inode->i_sb);
-
-	if (!(vfs->vfs_flag & VFS_32BITINODES)) {
+	if (!(XFS_M(inode->i_sb)->m_flags & XFS_MOUNT_SMALL_INUMS)) {
 		/* filesystem may contain 64bit inode numbers */
 		is64 = XFS_FILEID_TYPE_64FLAG;
 	}
@@ -138,10 +142,9 @@ xfs_fs_get_dentry(
 	bhv_vnode_t		*vp;
 	struct inode		*inode;
 	struct dentry		*result;
-	bhv_vfs_t		*vfsp = vfs_from_sb(sb);
 	int			error;
 
-	error = bhv_vfs_vget(vfsp, &vp, (fid_t *)data);
+	error = xfs_vget(XFS_M(sb), &vp, (fid_t *)data);
 	if (error || vp == NULL)
 		return ERR_PTR(-ESTALE) ;
 
@@ -159,12 +162,11 @@ xfs_fs_get_parent(
 	struct dentry		*child)
 {
 	int			error;
-	bhv_vnode_t		*vp, *cvp;
+	bhv_vnode_t		*cvp;
 	struct dentry		*parent;
 
 	cvp = NULL;
-	vp = vn_from_inode(child->d_inode);
-	error = bhv_vop_lookup(vp, &dotdot, &cvp, 0, NULL, NULL);
+	error = xfs_lookup(XFS_I(child->d_inode), &dotdot, &cvp);
 	if (unlikely(error))
 		return ERR_PTR(-error);
 
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index 0d4001eafd1..fb8dd34041e 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -37,6 +37,7 @@
 #include "xfs_error.h"
 #include "xfs_rw.h"
 #include "xfs_ioctl32.h"
+#include "xfs_vnodeops.h"
 
 #include <linux/dcache.h>
 #include <linux/smp_lock.h>
@@ -55,13 +56,12 @@ __xfs_file_read(
 	loff_t			pos)
 {
 	struct file		*file = iocb->ki_filp;
-	bhv_vnode_t		*vp = vn_from_inode(file->f_path.dentry->d_inode);
 
 	BUG_ON(iocb->ki_pos != pos);
 	if (unlikely(file->f_flags & O_DIRECT))
 		ioflags |= IO_ISDIRECT;
-	return bhv_vop_read(vp, iocb, iov, nr_segs, &iocb->ki_pos,
-				ioflags, NULL);
+	return xfs_read(XFS_I(file->f_path.dentry->d_inode), iocb, iov,
+				nr_segs, &iocb->ki_pos, ioflags);
 }
 
 STATIC ssize_t
@@ -93,14 +93,12 @@ __xfs_file_write(
 	loff_t			pos)
 {
 	struct file	*file = iocb->ki_filp;
-	struct inode	*inode = file->f_mapping->host;
-	bhv_vnode_t	*vp = vn_from_inode(inode);
 
 	BUG_ON(iocb->ki_pos != pos);
 	if (unlikely(file->f_flags & O_DIRECT))
 		ioflags |= IO_ISDIRECT;
-	return bhv_vop_write(vp, iocb, iov, nr_segs, &iocb->ki_pos,
-				ioflags, NULL);
+	return xfs_write(XFS_I(file->f_mapping->host), iocb, iov, nr_segs,
+				&iocb->ki_pos, ioflags);
 }
 
 STATIC ssize_t
@@ -131,8 +129,8 @@ xfs_file_splice_read(
 	size_t			len,
 	unsigned int		flags)
 {
-	return bhv_vop_splice_read(vn_from_inode(infilp->f_path.dentry->d_inode),
-				   infilp, ppos, pipe, len, flags, 0, NULL);
+	return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode),
+				   infilp, ppos, pipe, len, flags, 0);
 }
 
 STATIC ssize_t
@@ -143,9 +141,8 @@ xfs_file_splice_read_invis(
 	size_t			len,
 	unsigned int		flags)
 {
-	return bhv_vop_splice_read(vn_from_inode(infilp->f_path.dentry->d_inode),
-				   infilp, ppos, pipe, len, flags, IO_INVIS,
-				   NULL);
+	return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode),
+				   infilp, ppos, pipe, len, flags, IO_INVIS);
 }
 
 STATIC ssize_t
@@ -156,8 +153,8 @@ xfs_file_splice_write(
 	size_t			len,
 	unsigned int		flags)
 {
-	return bhv_vop_splice_write(vn_from_inode(outfilp->f_path.dentry->d_inode),
-				    pipe, outfilp, ppos, len, flags, 0, NULL);
+	return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode),
+				    pipe, outfilp, ppos, len, flags, 0);
 }
 
 STATIC ssize_t
@@ -168,9 +165,8 @@ xfs_file_splice_write_invis(
 	size_t			len,
 	unsigned int		flags)
 {
-	return bhv_vop_splice_write(vn_from_inode(outfilp->f_path.dentry->d_inode),
-				    pipe, outfilp, ppos, len, flags, IO_INVIS,
-				    NULL);
+	return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode),
+				    pipe, outfilp, ppos, len, flags, IO_INVIS);
 }
 
 STATIC int
@@ -180,7 +176,7 @@ xfs_file_open(
 {
 	if (!(filp->f_flags & O_LARGEFILE) && i_size_read(inode) > MAX_NON_LFS)
 		return -EFBIG;
-	return -bhv_vop_open(vn_from_inode(inode), NULL);
+	return -xfs_open(XFS_I(inode));
 }
 
 STATIC int
@@ -188,11 +184,7 @@ xfs_file_release(
 	struct inode	*inode,
 	struct file	*filp)
 {
-	bhv_vnode_t	*vp = vn_from_inode(inode);
-
-	if (vp)
-		return -bhv_vop_release(vp);
-	return 0;
+	return -xfs_release(XFS_I(inode));
 }
 
 STATIC int
@@ -201,14 +193,13 @@ xfs_file_fsync(
 	struct dentry	*dentry,
 	int		datasync)
 {
-	bhv_vnode_t	*vp = vn_from_inode(dentry->d_inode);
 	int		flags = FSYNC_WAIT;
 
 	if (datasync)
 		flags |= FSYNC_DATA;
-	if (VN_TRUNC(vp))
-		VUNTRUNCATE(vp);
-	return -bhv_vop_fsync(vp, flags, NULL, (xfs_off_t)0, (xfs_off_t)-1);
+	xfs_iflags_clear(XFS_I(dentry->d_inode), XFS_ITRUNCATED);
+	return -xfs_fsync(XFS_I(dentry->d_inode), flags,
+			(xfs_off_t)0, (xfs_off_t)-1);
 }
 
 #ifdef CONFIG_XFS_DMAPI
@@ -233,74 +224,30 @@ xfs_file_readdir(
 	void		*dirent,
 	filldir_t	filldir)
 {
-	int		error = 0;
-	bhv_vnode_t	*vp = vn_from_inode(filp->f_path.dentry->d_inode);
-	uio_t		uio;
-	iovec_t		iov;
-	int		eof = 0;
-	caddr_t		read_buf;
-	int		namelen, size = 0;
-	size_t		rlen = PAGE_CACHE_SIZE;
-	xfs_off_t	start_offset, curr_offset;
-	xfs_dirent_t	*dbp = NULL;
-
-	/* Try fairly hard to get memory */
-	do {
-		if ((read_buf = kmalloc(rlen, GFP_KERNEL)))
-			break;
-		rlen >>= 1;
-	} while (rlen >= 1024);
-
-	if (read_buf == NULL)
-		return -ENOMEM;
-
-	uio.uio_iov = &iov;
-	uio.uio_segflg = UIO_SYSSPACE;
-	curr_offset = filp->f_pos;
-	if (filp->f_pos != 0x7fffffff)
-		uio.uio_offset = filp->f_pos;
-	else
-		uio.uio_offset = 0xffffffff;
-
-	while (!eof) {
-		uio.uio_resid = iov.iov_len = rlen;
-		iov.iov_base = read_buf;
-		uio.uio_iovcnt = 1;
-
-		start_offset = uio.uio_offset;
-
-		error = bhv_vop_readdir(vp, &uio, NULL, &eof);
-		if ((uio.uio_offset == start_offset) || error) {
-			size = 0;
-			break;
-		}
-
-		size = rlen - uio.uio_resid;
-		dbp = (xfs_dirent_t *)read_buf;
-		while (size > 0) {
-			namelen = strlen(dbp->d_name);
-
-			if (filldir(dirent, dbp->d_name, namelen,
-					(loff_t) curr_offset & 0x7fffffff,
-					(ino_t) dbp->d_ino,
-					DT_UNKNOWN)) {
-				goto done;
-			}
-			size -= dbp->d_reclen;
-			curr_offset = (loff_t)dbp->d_off /* & 0x7fffffff */;
-			dbp = (xfs_dirent_t *)((char *)dbp + dbp->d_reclen);
-		}
-	}
-done:
-	if (!error) {
-		if (size == 0)
-			filp->f_pos = uio.uio_offset & 0x7fffffff;
-		else if (dbp)
-			filp->f_pos = curr_offset;
-	}
+	struct inode	*inode = filp->f_path.dentry->d_inode;
+	xfs_inode_t	*ip = XFS_I(inode);
+	int		error;
+	size_t		bufsize;
+
+	/*
+	 * The Linux API doesn't pass down the total size of the buffer
+	 * we read into down to the filesystem.  With the filldir concept
+	 * it's not needed for correct information, but the XFS dir2 leaf
+	 * code wants an estimate of the buffer size to calculate it's
+	 * readahead window and size the buffers used for mapping to
+	 * physical blocks.
+	 *
+	 * Try to give it an estimate that's good enough, maybe at some
+	 * point we can change the ->readdir prototype to include the
+	 * buffer size.
+	 */
+	bufsize = (size_t)min_t(loff_t, PAGE_SIZE, inode->i_size);
 
-	kfree(read_buf);
-	return -error;
+	error = xfs_readdir(ip, dirent, bufsize,
+				(xfs_off_t *)&filp->f_pos, filldir);
+	if (error)
+		return -error;
+	return 0;
 }
 
 STATIC int
@@ -312,7 +259,7 @@ xfs_file_mmap(
 	vma->vm_flags |= VM_CAN_NONLINEAR;
 
 #ifdef CONFIG_XFS_DMAPI
-	if (vn_from_inode(filp->f_path.dentry->d_inode)->v_vfsp->vfs_flag & VFS_DMI)
+	if (XFS_M(filp->f_path.dentry->d_inode->i_sb)->m_flags & XFS_MOUNT_DMAPI)
 		vma->vm_ops = &xfs_dmapi_file_vm_ops;
 #endif /* CONFIG_XFS_DMAPI */
 
@@ -328,10 +275,9 @@ xfs_file_ioctl(
 {
 	int		error;
 	struct inode	*inode = filp->f_path.dentry->d_inode;
-	bhv_vnode_t	*vp = vn_from_inode(inode);
 
-	error = bhv_vop_ioctl(vp, inode, filp, 0, cmd, (void __user *)p);
-	VMODIFY(vp);
+	error = xfs_ioctl(XFS_I(inode), filp, 0, cmd, (void __user *)p);
+	xfs_iflags_set(XFS_I(inode), XFS_IMODIFIED);
 
 	/* NOTE:  some of the ioctl's return positive #'s as a
 	 *	  byte count indicating success, such as
@@ -350,10 +296,9 @@ xfs_file_ioctl_invis(
 {
 	int		error;
 	struct inode	*inode = filp->f_path.dentry->d_inode;
-	bhv_vnode_t	*vp = vn_from_inode(inode);
 
-	error = bhv_vop_ioctl(vp, inode, filp, IO_INVIS, cmd, (void __user *)p);
-	VMODIFY(vp);
+	error = xfs_ioctl(XFS_I(inode), filp, IO_INVIS, cmd, (void __user *)p);
+	xfs_iflags_set(XFS_I(inode), XFS_IMODIFIED);
 
 	/* NOTE:  some of the ioctl's return positive #'s as a
 	 *	  byte count indicating success, such as
@@ -371,16 +316,14 @@ xfs_vm_mprotect(
 	struct vm_area_struct *vma,
 	unsigned int	newflags)
 {
-	bhv_vnode_t	*vp = vn_from_inode(vma->vm_file->f_path.dentry->d_inode);
+	struct inode	*inode = vma->vm_file->f_path.dentry->d_inode;
+	struct xfs_mount *mp = XFS_M(inode->i_sb);
 	int		error = 0;
 
-	if (vp->v_vfsp->vfs_flag & VFS_DMI) {
+	if (mp->m_flags & XFS_MOUNT_DMAPI) {
 		if ((vma->vm_flags & VM_MAYSHARE) &&
-		    (newflags & VM_WRITE) && !(vma->vm_flags & VM_WRITE)) {
-			xfs_mount_t	*mp = XFS_VFSTOM(vp->v_vfsp);
-
+		    (newflags & VM_WRITE) && !(vma->vm_flags & VM_WRITE))
 			error = XFS_SEND_MMAP(mp, vma, VM_WRITE);
-		    }
 	}
 	return error;
 }
@@ -397,18 +340,17 @@ STATIC int
 xfs_file_open_exec(
 	struct inode	*inode)
 {
-	bhv_vnode_t	*vp = vn_from_inode(inode);
+	struct xfs_mount *mp = XFS_M(inode->i_sb);
 
-	if (unlikely(vp->v_vfsp->vfs_flag & VFS_DMI)) {
-		xfs_mount_t	*mp = XFS_VFSTOM(vp->v_vfsp);
-		xfs_inode_t	*ip = xfs_vtoi(vp);
+	if (unlikely(mp->m_flags & XFS_MOUNT_DMAPI)) {
+		if (DM_EVENT_ENABLED(XFS_I(inode), DM_EVENT_READ)) {
+			bhv_vnode_t *vp = vn_from_inode(inode);
 
-		if (!ip)
-			return -EINVAL;
-		if (DM_EVENT_ENABLED(vp->v_vfsp, ip, DM_EVENT_READ))
-			return -XFS_SEND_DATA(mp, DM_EVENT_READ, vp,
-					       0, 0, 0, NULL);
+			return -XFS_SEND_DATA(mp, DM_EVENT_READ,
+						vp, 0, 0, 0, NULL);
+		}
 	}
+
 	return 0;
 }
 #endif /* HAVE_FOP_OPEN_EXEC */
diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.c b/fs/xfs/linux-2.6/xfs_fs_subr.c
index 2eb87cd082a..ac6d34cc355 100644
--- a/fs/xfs/linux-2.6/xfs_fs_subr.c
+++ b/fs/xfs/linux-2.6/xfs_fs_subr.c
@@ -16,66 +16,78 @@
  * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  */
 #include "xfs.h"
+#include "xfs_vnodeops.h"
+
+/*
+ * The following six includes are needed so that we can include
+ * xfs_inode.h.  What a mess..
+ */
+#include "xfs_bmap_btree.h"
+#include "xfs_inum.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_sf.h"
+#include "xfs_attr_sf.h"
+#include "xfs_dinode.h"
+
+#include "xfs_inode.h"
 
 int  fs_noerr(void) { return 0; }
 int  fs_nosys(void) { return ENOSYS; }
 void fs_noval(void) { return; }
 
 void
-fs_tosspages(
-	bhv_desc_t	*bdp,
+xfs_tosspages(
+	xfs_inode_t	*ip,
 	xfs_off_t	first,
 	xfs_off_t	last,
 	int		fiopt)
 {
-	bhv_vnode_t	*vp = BHV_TO_VNODE(bdp);
-	struct inode	*ip = vn_to_inode(vp);
+	bhv_vnode_t	*vp = XFS_ITOV(ip);
+	struct inode	*inode = vn_to_inode(vp);
 
 	if (VN_CACHED(vp))
-		truncate_inode_pages(ip->i_mapping, first);
+		truncate_inode_pages(inode->i_mapping, first);
 }
 
 int
-fs_flushinval_pages(
-	bhv_desc_t	*bdp,
+xfs_flushinval_pages(
+	xfs_inode_t	*ip,
 	xfs_off_t	first,
 	xfs_off_t	last,
 	int		fiopt)
 {
-	bhv_vnode_t	*vp = BHV_TO_VNODE(bdp);
-	struct inode	*ip = vn_to_inode(vp);
+	bhv_vnode_t	*vp = XFS_ITOV(ip);
+	struct inode	*inode = vn_to_inode(vp);
 	int		ret = 0;
 
 	if (VN_CACHED(vp)) {
-		if (VN_TRUNC(vp))
-			VUNTRUNCATE(vp);
-		ret = filemap_write_and_wait(ip->i_mapping);
+		xfs_iflags_clear(ip, XFS_ITRUNCATED);
+		ret = filemap_write_and_wait(inode->i_mapping);
 		if (!ret)
-			truncate_inode_pages(ip->i_mapping, first);
+			truncate_inode_pages(inode->i_mapping, first);
 	}
 	return ret;
 }
 
 int
-fs_flush_pages(
-	bhv_desc_t	*bdp,
+xfs_flush_pages(
+	xfs_inode_t	*ip,
 	xfs_off_t	first,
 	xfs_off_t	last,
 	uint64_t	flags,
 	int		fiopt)
 {
-	bhv_vnode_t	*vp = BHV_TO_VNODE(bdp);
-	struct inode	*ip = vn_to_inode(vp);
+	bhv_vnode_t	*vp = XFS_ITOV(ip);
+	struct inode	*inode = vn_to_inode(vp);
 	int		ret = 0;
 	int		ret2;
 
 	if (VN_DIRTY(vp)) {
-		if (VN_TRUNC(vp))
-			VUNTRUNCATE(vp);
-		ret = filemap_fdatawrite(ip->i_mapping);
+		xfs_iflags_clear(ip, XFS_ITRUNCATED);
+		ret = filemap_fdatawrite(inode->i_mapping);
 		if (flags & XFS_B_ASYNC)
 			return ret;
-		ret2 = filemap_fdatawait(ip->i_mapping);
+		ret2 = filemap_fdatawait(inode->i_mapping);
 		if (!ret)
 			ret = ret2;
 	}
diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.h b/fs/xfs/linux-2.6/xfs_fs_subr.h
index c1b53118a30..82bb19b2599 100644
--- a/fs/xfs/linux-2.6/xfs_fs_subr.h
+++ b/fs/xfs/linux-2.6/xfs_fs_subr.h
@@ -18,12 +18,8 @@
 #ifndef	__XFS_FS_SUBR_H__
 #define __XFS_FS_SUBR_H__
 
-struct cred;
 extern int  fs_noerr(void);
 extern int  fs_nosys(void);
 extern void fs_noval(void);
-extern void fs_tosspages(bhv_desc_t *, xfs_off_t, xfs_off_t, int);
-extern int  fs_flushinval_pages(bhv_desc_t *, xfs_off_t, xfs_off_t, int);
-extern int  fs_flush_pages(bhv_desc_t *, xfs_off_t, xfs_off_t, uint64_t, int);
 
 #endif	/* __XFS_FS_SUBR_H__ */
diff --git a/fs/xfs/linux-2.6/xfs_globals.c b/fs/xfs/linux-2.6/xfs_globals.c
index 81565dea9af..9febf9dc999 100644
--- a/fs/xfs/linux-2.6/xfs_globals.c
+++ b/fs/xfs/linux-2.6/xfs_globals.c
@@ -20,11 +20,6 @@
 #include "xfs_sysctl.h"
 
 /*
- * System memory size - used to scale certain data structures in XFS.
- */
-unsigned long xfs_physmem;
-
-/*
  * Tunable XFS parameters.  xfs_params is required even when CONFIG_SYSCTL=n,
  * other XFS code uses these values.  Times are measured in centisecs (i.e.
  * 100ths of a second).
diff --git a/fs/xfs/linux-2.6/xfs_globals.h b/fs/xfs/linux-2.6/xfs_globals.h
index e1a22bfcf86..2770b0085ee 100644
--- a/fs/xfs/linux-2.6/xfs_globals.h
+++ b/fs/xfs/linux-2.6/xfs_globals.h
@@ -19,7 +19,6 @@
 #define __XFS_GLOBALS_H__
 
 extern uint64_t	xfs_panic_mask;		/* set to cause more panics */
-extern unsigned long xfs_physmem;
 extern struct cred *sys_cred;
 
 #endif	/* __XFS_GLOBALS_H__ */
diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c
index 5917808abbd..ffec630e7db 100644
--- a/fs/xfs/linux-2.6/xfs_ioctl.c
+++ b/fs/xfs/linux-2.6/xfs_ioctl.c
@@ -47,6 +47,7 @@
 #include "xfs_utils.h"
 #include "xfs_dfrag.h"
 #include "xfs_fsops.h"
+#include "xfs_vnodeops.h"
 
 #include <linux/capability.h>
 #include <linux/dcache.h>
@@ -137,7 +138,8 @@ xfs_find_handle(
 	vp = vn_from_inode(inode);
 
 	/* now we can grab the fsid */
-	memcpy(&handle.ha_fsid, vp->v_vfsp->vfs_altfsid, sizeof(xfs_fsid_t));
+	memcpy(&handle.ha_fsid, XFS_I(inode)->i_mount->m_fixedfsid,
+			sizeof(xfs_fsid_t));
 	hsize = sizeof(xfs_fsid_t);
 
 	if (cmd != XFS_IOC_PATH_TO_FSHANDLE) {
@@ -349,19 +351,44 @@ xfs_open_by_handle(
 	return new_fd;
 }
 
+/*
+ * This is a copy from fs/namei.c:vfs_readlink(), except for removing it's
+ * unused first argument.
+ */
+STATIC int
+do_readlink(
+	char __user		*buffer,
+	int			buflen,
+	const char		*link)
+{
+        int len;
+
+	len = PTR_ERR(link);
+	if (IS_ERR(link))
+		goto out;
+
+	len = strlen(link);
+	if (len > (unsigned) buflen)
+		len = buflen;
+	if (copy_to_user(buffer, link, len))
+		len = -EFAULT;
+ out:
+	return len;
+}
+
+
 STATIC int
 xfs_readlink_by_handle(
 	xfs_mount_t		*mp,
 	void			__user *arg,
 	struct inode		*parinode)
 {
-	int			error;
-	struct iovec		aiov;
-	struct uio		auio;
 	struct inode		*inode;
 	xfs_fsop_handlereq_t	hreq;
 	bhv_vnode_t		*vp;
 	__u32			olen;
+	void			*link;
+	int			error;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -XFS_ERROR(EPERM);
@@ -374,29 +401,31 @@ xfs_readlink_by_handle(
 
 	/* Restrict this handle operation to symlinks only. */
 	if (!S_ISLNK(inode->i_mode)) {
-		VN_RELE(vp);
-		return -XFS_ERROR(EINVAL);
+		error = -XFS_ERROR(EINVAL);
+		goto out_iput;
 	}
 
 	if (copy_from_user(&olen, hreq.ohandlen, sizeof(__u32))) {
-		VN_RELE(vp);
-		return -XFS_ERROR(EFAULT);
+		error = -XFS_ERROR(EFAULT);
+		goto out_iput;
 	}
-	aiov.iov_len	= olen;
-	aiov.iov_base	= hreq.ohandle;
 
-	auio.uio_iov	= (struct kvec *)&aiov;
-	auio.uio_iovcnt	= 1;
-	auio.uio_offset	= 0;
-	auio.uio_segflg	= UIO_USERSPACE;
-	auio.uio_resid	= olen;
+	link = kmalloc(MAXPATHLEN+1, GFP_KERNEL);
+	if (!link)
+		goto out_iput;
 
-	error = bhv_vop_readlink(vp, &auio, IO_INVIS, NULL);
-	VN_RELE(vp);
+	error = -xfs_readlink(XFS_I(inode), link);
 	if (error)
-		return -error;
+		goto out_kfree;
+	error = do_readlink(hreq.ohandle, olen, link);
+	if (error)
+		goto out_kfree;
 
-	return (olen - auio.uio_resid);
+ out_kfree:
+	kfree(link);
+ out_iput:
+	iput(inode);
+	return error;
 }
 
 STATIC int
@@ -409,7 +438,6 @@ xfs_fssetdm_by_handle(
 	struct fsdmidata	fsd;
 	xfs_fsop_setdm_handlereq_t dmhreq;
 	struct inode		*inode;
-	bhv_desc_t		*bdp;
 	bhv_vnode_t		*vp;
 
 	if (!capable(CAP_MKNOD))
@@ -431,8 +459,8 @@ xfs_fssetdm_by_handle(
 		return -XFS_ERROR(EFAULT);
 	}
 
-	bdp = bhv_base_unlocked(VN_BHV_HEAD(vp));
-	error = xfs_set_dmattrs(bdp, fsd.fsd_dmevmask, fsd.fsd_dmstate, NULL);
+	error = xfs_set_dmattrs(xfs_vtoi(vp),
+			fsd.fsd_dmevmask, fsd.fsd_dmstate);
 
 	VN_RELE(vp);
 	if (error)
@@ -470,8 +498,8 @@ xfs_attrlist_by_handle(
 		goto out_vn_rele;
 
 	cursor = (attrlist_cursor_kern_t *)&al_hreq.pos;
-	error = bhv_vop_attr_list(vp, kbuf, al_hreq.buflen, al_hreq.flags,
-					cursor, NULL);
+	error = xfs_attr_list(XFS_I(inode), kbuf, al_hreq.buflen,
+					al_hreq.flags, cursor);
 	if (error)
 		goto out_kfree;
 
@@ -488,7 +516,7 @@ xfs_attrlist_by_handle(
 
 STATIC int
 xfs_attrmulti_attr_get(
-	bhv_vnode_t		*vp,
+	struct inode		*inode,
 	char			*name,
 	char			__user *ubuf,
 	__uint32_t		*len,
@@ -503,7 +531,7 @@ xfs_attrmulti_attr_get(
 	if (!kbuf)
 		return ENOMEM;
 
-	error = bhv_vop_attr_get(vp, name, kbuf, len, flags, NULL);
+	error = xfs_attr_get(XFS_I(inode), name, kbuf, len, flags, NULL);
 	if (error)
 		goto out_kfree;
 
@@ -517,7 +545,7 @@ xfs_attrmulti_attr_get(<