Merge branch 'master' of git+ssh://git.melbourne.sgi.com/git/xfs

fs/xfs/linux-2.6/xfs_buf.c

@@ -1085,7 +1085,7 @@ xfs_bawrite(
 	bp->b_flags &= ~(XBF_READ | XBF_DELWRI | XBF_READ_AHEAD);
 	bp->b_flags |= (XBF_WRITE | XBF_ASYNC | _XBF_RUN_QUEUES);
 
-	bp->b_fspriv3 = mp;
+	bp->b_mount = mp;
 	bp->b_strat = xfs_bdstrat_cb;
 	return xfs_bdstrat_cb(bp);
 }
@@ -1098,7 +1098,7 @@ xfs_bdwrite(
 	XB_TRACE(bp, "bdwrite", 0);
 
 	bp->b_strat = xfs_bdstrat_cb;
-	bp->b_fspriv3 = mp;
+	bp->b_mount = mp;
 
 	bp->b_flags &= ~XBF_READ;
 	bp->b_flags |= (XBF_DELWRI | XBF_ASYNC);

fs/xfs/linux-2.6/xfs_buf.h

@@ -168,7 +168,7 @@ typedef struct xfs_buf {
 	struct completion	b_iowait;	/* queue for I/O waiters */
 	void			*b_fspriv;
 	void			*b_fspriv2;
-	void			*b_fspriv3;
+	struct xfs_mount	*b_mount;
 	unsigned short		b_error;	/* error code on I/O */
 	unsigned int		b_page_count;	/* size of page array */
 	unsigned int		b_offset;	/* page offset in first page */
@@ -335,8 +335,6 @@ extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *);
 #define XFS_BUF_SET_FSPRIVATE(bp, val)		((bp)->b_fspriv = (void*)(val))
 #define XFS_BUF_FSPRIVATE2(bp, type)		((type)(bp)->b_fspriv2)
 #define XFS_BUF_SET_FSPRIVATE2(bp, val)		((bp)->b_fspriv2 = (void*)(val))
-#define XFS_BUF_FSPRIVATE3(bp, type)		((type)(bp)->b_fspriv3)
-#define XFS_BUF_SET_FSPRIVATE3(bp, val)		((bp)->b_fspriv3 = (void*)(val))
 #define XFS_BUF_SET_START(bp)			do { } while (0)
 #define XFS_BUF_SET_BRELSE_FUNC(bp, func)	((bp)->b_relse = (func))
 

fs/xfs/linux-2.6/xfs_file.c

@@ -45,80 +45,44 @@
 
 static struct vm_operations_struct xfs_file_vm_ops;
 
-STATIC_INLINE ssize_t
-__xfs_file_read(
+STATIC ssize_t
+xfs_file_aio_read(
 	struct kiocb		*iocb,
 	const struct iovec	*iov,
 	unsigned long		nr_segs,
-	int			ioflags,
 	loff_t			pos)
 {
 	struct file		*file = iocb->ki_filp;
+	int			ioflags = IO_ISAIO;
 
 	BUG_ON(iocb->ki_pos != pos);
 	if (unlikely(file->f_flags & O_DIRECT))
 		ioflags |= IO_ISDIRECT;
+	if (file->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
 	return xfs_read(XFS_I(file->f_path.dentry->d_inode), iocb, iov,
 				nr_segs, &iocb->ki_pos, ioflags);
 }
 
 STATIC ssize_t
-xfs_file_aio_read(
-	struct kiocb		*iocb,
-	const struct iovec	*iov,
-	unsigned long		nr_segs,
-	loff_t			pos)
-{
-	return __xfs_file_read(iocb, iov, nr_segs, IO_ISAIO, pos);
-}
-
-STATIC ssize_t
-xfs_file_aio_read_invis(
-	struct kiocb		*iocb,
-	const struct iovec	*iov,
-	unsigned long		nr_segs,
-	loff_t			pos)
-{
-	return __xfs_file_read(iocb, iov, nr_segs, IO_ISAIO|IO_INVIS, pos);
-}
-
-STATIC_INLINE ssize_t
-__xfs_file_write(
+xfs_file_aio_write(
 	struct kiocb		*iocb,
 	const struct iovec	*iov,
 	unsigned long		nr_segs,
-	int			ioflags,
 	loff_t			pos)
 {
 	struct file		*file = iocb->ki_filp;
+	int			ioflags = IO_ISAIO;
 
 	BUG_ON(iocb->ki_pos != pos);
 	if (unlikely(file->f_flags & O_DIRECT))
 		ioflags |= IO_ISDIRECT;
+	if (file->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
 	return xfs_write(XFS_I(file->f_mapping->host), iocb, iov, nr_segs,
 				&iocb->ki_pos, ioflags);
 }
 
-STATIC ssize_t
-xfs_file_aio_write(
-	struct kiocb		*iocb,
-	const struct iovec	*iov,
-	unsigned long		nr_segs,
-	loff_t			pos)
-{
-	return __xfs_file_write(iocb, iov, nr_segs, IO_ISAIO, pos);
-}
-
-STATIC ssize_t
-xfs_file_aio_write_invis(
-	struct kiocb		*iocb,
-	const struct iovec	*iov,
-	unsigned long		nr_segs,
-	loff_t			pos)
-{
-	return __xfs_file_write(iocb, iov, nr_segs, IO_ISAIO|IO_INVIS, pos);
-}
-
 STATIC ssize_t
 xfs_file_splice_read(
 	struct file		*infilp,
@@ -127,20 +91,13 @@ xfs_file_splice_read(
 	size_t			len,
 	unsigned int		flags)
 {
-	return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode),
-				   infilp, ppos, pipe, len, flags, 0);
-}
+	int			ioflags = 0;
+
+	if (infilp->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
 
-STATIC ssize_t
-xfs_file_splice_read_invis(
-	struct file		*infilp,
-	loff_t			*ppos,
-	struct pipe_inode_info	*pipe,
-	size_t			len,
-	unsigned int		flags)
-{
 	return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode),
-				   infilp, ppos, pipe, len, flags, IO_INVIS);
+				   infilp, ppos, pipe, len, flags, ioflags);
 }
 
 STATIC ssize_t
@@ -151,20 +108,13 @@ xfs_file_splice_write(
 	size_t			len,
 	unsigned int		flags)
 {
-	return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode),
-				    pipe, outfilp, ppos, len, flags, 0);
-}
+	int			ioflags = 0;
+
+	if (outfilp->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
 
-STATIC ssize_t
-xfs_file_splice_write_invis(
-	struct pipe_inode_info	*pipe,
-	struct file		*outfilp,
-	loff_t			*ppos,
-	size_t			len,
-	unsigned int		flags)
-{
 	return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode),
-				    pipe, outfilp, ppos, len, flags, IO_INVIS);
+				    pipe, outfilp, ppos, len, flags, ioflags);
 }
 
 STATIC int
@@ -275,42 +225,6 @@ xfs_file_mmap(
 	return 0;
 }
 
-STATIC long
-xfs_file_ioctl(
-	struct file	*filp,
-	unsigned int	cmd,
-	unsigned long	p)
-{
-	struct inode	*inode = filp->f_path.dentry->d_inode;
-
-
-	/* NOTE:  some of the ioctl's return positive #'s as a
-	 *	  byte count indicating success, such as
-	 *	  readlink_by_handle.  So we don't "sign flip"
-	 *	  like most other routines.  This means true
-	 *	  errors need to be returned as a negative value.
-	 */
-	return xfs_ioctl(XFS_I(inode), filp, 0, cmd, (void __user *)p);
-}
-
-STATIC long
-xfs_file_ioctl_invis(
-	struct file	*filp,
-	unsigned int	cmd,
-	unsigned long	p)
-{
-	struct inode	*inode = filp->f_path.dentry->d_inode;
-
-
-	/* NOTE:  some of the ioctl's return positive #'s as a
-	 *	  byte count indicating success, such as
-	 *	  readlink_by_handle.  So we don't "sign flip"
-	 *	  like most other routines.  This means true
-	 *	  errors need to be returned as a negative value.
-	 */
-	return xfs_ioctl(XFS_I(inode), filp, IO_INVIS, cmd, (void __user *)p);
-}
-
 /*
  * mmap()d file has taken write protection fault and is being made
  * writable. We can set the page state up correctly for a writable
@@ -346,25 +260,6 @@ const struct file_operations xfs_file_operations = {
 #endif
 };
 
-const struct file_operations xfs_invis_file_operations = {
-	.llseek		= generic_file_llseek,
-	.read		= do_sync_read,
-	.write		= do_sync_write,
-	.aio_read	= xfs_file_aio_read_invis,
-	.aio_write	= xfs_file_aio_write_invis,
-	.splice_read	= xfs_file_splice_read_invis,
-	.splice_write	= xfs_file_splice_write_invis,
-	.unlocked_ioctl	= xfs_file_ioctl_invis,
-#ifdef CONFIG_COMPAT
-	.compat_ioctl	= xfs_file_compat_invis_ioctl,
-#endif
-	.mmap		= xfs_file_mmap,
-	.open		= xfs_file_open,
-	.release	= xfs_file_release,
-	.fsync		= xfs_file_fsync,
-};
-
-
 const struct file_operations xfs_dir_file_operations = {
 	.open		= xfs_dir_open,
 	.read		= generic_read_dir,

fs/xfs/linux-2.6/xfs_ioctl.c

@@ -319,10 +319,11 @@ xfs_open_by_handle(
 		put_unused_fd(new_fd);
 		return -XFS_ERROR(-PTR_ERR(filp));
 	}
+
 	if (inode->i_mode & S_IFREG) {
 		/* invisible operation should not change atime */
 		filp->f_flags |= O_NOATIME;
-		filp->f_op = &xfs_invis_file_operations;
+		filp->f_mode |= FMODE_NOCMTIME;
 	}
 
 	fd_install(new_fd, filp);
@@ -1328,21 +1329,31 @@ xfs_ioc_getbmapx(
 	return 0;
 }
 
-int
-xfs_ioctl(
-	xfs_inode_t		*ip,
+/*
+ * Note: some of the ioctl's return positive numbers as a
+ * byte count indicating success, such as readlink_by_handle.
+ * So we don't "sign flip" like most other routines.  This means
+ * true errors need to be returned as a negative value.
+ */
+long
+xfs_file_ioctl(
 	struct file		*filp,
-	int			ioflags,
 	unsigned int		cmd,
-	void			__user *arg)
+	unsigned long		p)
 {
 	struct inode		*inode = filp->f_path.dentry->d_inode;
-	xfs_mount_t		*mp = ip->i_mount;
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+	void			__user *arg = (void __user *)p;
+	int			ioflags = 0;
 	int			error;
 
-	xfs_itrace_entry(XFS_I(inode));
-	switch (cmd) {
+	if (filp->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
 
+	xfs_itrace_entry(ip);
+
+	switch (cmd) {
 	case XFS_IOC_ALLOCSP:
 	case XFS_IOC_FREESP:
 	case XFS_IOC_RESVSP:

fs/xfs/linux-2.6/xfs_ioctl.h

@@ -68,13 +68,13 @@ xfs_attrmulti_attr_remove(
 	__uint32_t		flags);
 
 extern long
-xfs_file_compat_ioctl(
-	struct file		*file,
+xfs_file_ioctl(
+	struct file		*filp,
 	unsigned int		cmd,
-	unsigned long		arg);
+	unsigned long		p);
 
 extern long
-xfs_file_compat_invis_ioctl(
+xfs_file_compat_ioctl(
 	struct file		*file,
 	unsigned int		cmd,
 	unsigned long		arg);

fs/xfs/linux-2.6/xfs_ioctl32.c

@@ -599,19 +599,24 @@ xfs_compat_fssetdm_by_handle(
 	return error;
 }
 
-STATIC long
-xfs_compat_ioctl(
-	xfs_inode_t	*ip,
+long
+xfs_file_compat_ioctl(
 	struct file		*filp,
-	int		ioflags,
 	unsigned		cmd,
-	void		__user *arg)
+	unsigned long		p)
 {
 	struct inode		*inode = filp->f_path.dentry->d_inode;
-	xfs_mount_t	*mp = ip->i_mount;
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+	void			__user *arg = (void __user *)p;
+	int			ioflags = 0;
 	int			error;
 
-	xfs_itrace_entry(XFS_I(inode));
+	if (filp->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
+
+	xfs_itrace_entry(ip);
+
 	switch (cmd) {
 	/* No size or alignment issues on any arch */
 	case XFS_IOC_DIOINFO:
@@ -632,7 +637,7 @@ xfs_compat_ioctl(
 	case XFS_IOC_GOINGDOWN:
 	case XFS_IOC_ERROR_INJECTION:
 	case XFS_IOC_ERROR_CLEARALL:
-		return xfs_ioctl(ip, filp, ioflags, cmd, arg);
+		return xfs_file_ioctl(filp, cmd, p);
 #ifndef BROKEN_X86_ALIGNMENT
 	/* These are handled fine if no alignment issues */
 	case XFS_IOC_ALLOCSP:
@@ -646,7 +651,7 @@ xfs_compat_ioctl(
 	case XFS_IOC_FSGEOMETRY_V1:
 	case XFS_IOC_FSGROWFSDATA:
 	case XFS_IOC_FSGROWFSRT:
-		return xfs_ioctl(ip, filp, ioflags, cmd, arg);
+		return xfs_file_ioctl(filp, cmd, p);
 #else
 	case XFS_IOC_ALLOCSP_32:
 	case XFS_IOC_FREESP_32:
@@ -687,7 +692,7 @@ xfs_compat_ioctl(
 	case XFS_IOC_SETXFLAGS_32:
 	case XFS_IOC_GETVERSION_32:
 		cmd = _NATIVE_IOC(cmd, long);
-		return xfs_ioctl(ip, filp, ioflags, cmd, arg);
+		return xfs_file_ioctl(filp, cmd, p);
 	case XFS_IOC_SWAPEXT: {
 		struct xfs_swapext	  sxp;
 		struct compat_xfs_swapext __user *sxu = arg;
@@ -738,26 +743,3 @@ xfs_compat_ioctl(
 		return -XFS_ERROR(ENOIOCTLCMD);
 	}
 }
-
-long
-xfs_file_compat_ioctl(
-	struct file		*filp,
-	unsigned int		cmd,
-	unsigned long		p)
-{
-	struct inode	*inode = filp->f_path.dentry->d_inode;
-
-	return xfs_compat_ioctl(XFS_I(inode), filp, 0, cmd, (void __user *)p);
-}
-
-long
-xfs_file_compat_invis_ioctl(
-	struct file		*filp,
-	unsigned int		cmd,
-	unsigned long		p)
-{
-	struct inode	*inode = filp->f_path.dentry->d_inode;
-
-	return xfs_compat_ioctl(XFS_I(inode), filp, IO_INVIS, cmd,
-				(void __user *)p);
-}

fs/xfs/linux-2.6/xfs_iops.h

@@ -22,7 +22,6 @@ struct xfs_inode;
 
 extern const struct file_operations xfs_file_operations;
 extern const struct file_operations xfs_dir_file_operations;
-extern const struct file_operations xfs_invis_file_operations;
 
 extern ssize_t xfs_vn_listxattr(struct dentry *, char *data, size_t size);
 

fs/xfs/linux-2.6/xfs_lrw.c

@@ -847,13 +847,7 @@ xfs_write(
 int
 xfs_bdstrat_cb(struct xfs_buf *bp)
 {
-	xfs_mount_t	*mp;
-
-	mp = XFS_BUF_FSPRIVATE3(bp, xfs_mount_t *);
-	if (!XFS_FORCED_SHUTDOWN(mp)) {
-		xfs_buf_iorequest(bp);
-		return 0;
-	} else {
+	if (XFS_FORCED_SHUTDOWN(bp->b_mount)) {
 		xfs_buftrace("XFS__BDSTRAT IOERROR", bp);
 		/*
 		 * Metadata write that didn't get logged but
@@ -866,6 +860,9 @@ xfs_bdstrat_cb(struct xfs_buf *bp)
 		else
 			return (xfs_bioerror(bp));
 	}
+
+	xfs_buf_iorequest(bp);
+	return 0;
 }
 
 /*

fs/xfs/xfs_buf_item.c

@@ -707,8 +707,8 @@ xfs_buf_item_init(
 	 * the first.  If we do already have one, there is
 	 * nothing to do here so return.
 	 */
-	if (XFS_BUF_FSPRIVATE3(bp, xfs_mount_t *) != mp)
-		XFS_BUF_SET_FSPRIVATE3(bp, mp);
+	if (bp->b_mount != mp)
+		bp->b_mount = mp;
 	XFS_BUF_SET_BDSTRAT_FUNC(bp, xfs_bdstrat_cb);
 	if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) {
 		lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);

fs/xfs/xfs_inode.c

@@ -3232,8 +3232,6 @@ xfs_iflush_int(
 
 
 #ifdef XFS_ILOCK_TRACE
-ktrace_t	*xfs_ilock_trace_buf;
-
 void
 xfs_ilock_trace(xfs_inode_t *ip, int lock, unsigned int lockflags, inst_t *ra)
 {

fs/xfs/xfs_inode.h

@@ -223,7 +223,6 @@ struct xfs_dquot;
 
 #if defined(XFS_ILOCK_TRACE)
 #define XFS_ILOCK_KTRACE_SIZE	32
-extern ktrace_t *xfs_ilock_trace_buf;
 extern void xfs_ilock_trace(struct xfs_inode *, int, unsigned int, inst_t *);
 #else
 #define	xfs_ilock_trace(i,n,f,ra)
@@ -482,12 +481,6 @@ static inline void xfs_ifunlock(xfs_inode_t *ip)
 	(((pip)->i_mount->m_flags & XFS_MOUNT_GRPID) || \
 	 ((pip)->i_d.di_mode & S_ISGID))
 
-/*
- * Flags for xfs_iget()
- */
-#define XFS_IGET_CREATE		0x1
-#define XFS_IGET_BULKSTAT	0x2
-
 /*
  * xfs_iget.c prototypes.
  */
@@ -509,8 +502,6 @@ void		xfs_ireclaim(xfs_inode_t *);
 /*
  * xfs_inode.c prototypes.
  */
-int		xfs_iread(struct xfs_mount *, struct xfs_trans *,
-			  struct xfs_inode *, xfs_daddr_t, uint);
 int		xfs_ialloc(struct xfs_trans *, xfs_inode_t *, mode_t,
 			   xfs_nlink_t, xfs_dev_t, struct cred *, xfs_prid_t,
 			   int, struct xfs_buf **, boolean_t *, xfs_inode_t **);
@@ -583,12 +574,20 @@ do { \
 
 #endif /* __KERNEL__ */
 
+/*
+ * Flags for xfs_iget()
+ */
+#define XFS_IGET_CREATE		0x1
+#define XFS_IGET_BULKSTAT	0x2
+
 int		xfs_inotobp(struct xfs_mount *, struct xfs_trans *,
 			    xfs_ino_t, struct xfs_dinode **,
 			    struct xfs_buf **, int *, uint);
 int		xfs_itobp(struct xfs_mount *, struct xfs_trans *,
 			  struct xfs_inode *, struct xfs_dinode **,
 			  struct xfs_buf **, uint);
+int		xfs_iread(struct xfs_mount *, struct xfs_trans *,
+			  struct xfs_inode *, xfs_daddr_t, uint);
 void		xfs_dinode_from_disk(struct xfs_icdinode *,
 				     struct xfs_dinode *);
 void		xfs_dinode_to_disk(struct xfs_dinode *,

fs/xfs/xfs_log_recover.c

@@ -267,21 +267,16 @@ STATIC void
 xlog_recover_iodone(
 	struct xfs_buf	*bp)
 {
-	xfs_mount_t	*mp;
-
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *));
-
 	if (XFS_BUF_GETERROR(bp)) {
 		/*
 		 * We're not going to bother about retrying
 		 * this during recovery. One strike!
 		 */
-		mp = XFS_BUF_FSPRIVATE(bp, xfs_mount_t *);
 		xfs_ioerror_alert("xlog_recover_iodone",
-				  mp, bp, XFS_BUF_ADDR(bp));
-		xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR);
+				  bp->b_mount, bp, XFS_BUF_ADDR(bp));
+		xfs_force_shutdown(bp->b_mount, SHUTDOWN_META_IO_ERROR);
 	}
-	XFS_BUF_SET_FSPRIVATE(bp, NULL);
+	bp->b_mount = NULL;
 	XFS_BUF_CLR_IODONE_FUNC(bp);
 	xfs_biodone(bp);
 }
@@ -2225,9 +2220,8 @@ xlog_recover_do_buffer_trans(
 		XFS_BUF_STALE(bp);
 		error = xfs_bwrite(mp, bp);
 	} else {
-		ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL ||
-		       XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp);
-		XFS_BUF_SET_FSPRIVATE(bp, mp);
+		ASSERT(bp->b_mount == NULL || bp->b_mount == mp);
+		bp->b_mount = mp;
 		XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
 		xfs_bdwrite(mp, bp);
 	}
@@ -2490,9 +2484,8 @@ xlog_recover_do_inode_trans(
 
 write_inode_buffer:
 	if (ITEM_TYPE(item) == XFS_LI_INODE) {
-		ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL ||
-		       XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp);
-		XFS_BUF_SET_FSPRIVATE(bp, mp);
+		ASSERT(bp->b_mount == NULL || bp->b_mount == mp);
+		bp->b_mount = mp;
 		XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
 		xfs_bdwrite(mp, bp);
 	} else {
@@ -2623,9 +2616,8 @@ xlog_recover_do_dquot_trans(
 	memcpy(ddq, recddq, item->ri_buf[1].i_len);
 
 	ASSERT(dq_f->qlf_size == 2);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL ||
-	       XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp);
-	XFS_BUF_SET_FSPRIVATE(bp, mp);
+	ASSERT(bp->b_mount == NULL || bp->b_mount == mp);
+	bp->b_mount = mp;
 	XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
 	xfs_bdwrite(mp, bp);
 

fs/xfs/xfs_mount.h

@@ -18,8 +18,6 @@
 #ifndef __XFS_MOUNT_H__
 #define	__XFS_MOUNT_H__
 
-#include "xfs_sync.h"
-
 typedef struct xfs_trans_reservations {
 	uint	tr_write;	/* extent alloc trans */
 	uint	tr_itruncate;	/* truncate trans */
@@ -53,6 +51,8 @@ typedef struct xfs_trans_reservations {
 
 #else /* __KERNEL__ */
 
+#include "xfs_sync.h"
+
 struct cred;
 struct log;
 struct xfs_mount_args;

fs/xfs/xfs_rename.c

@@ -41,31 +41,6 @@
 #include "xfs_vnodeops.h"
 
 
-/*
- * Given an array of up to 4 inode pointers, unlock the pointed to inodes.
- * If there are fewer than 4 entries in the array, the empty entries will
- * be at the end and will have NULL pointers in them.
- */
-STATIC void
-xfs_rename_unlock4(
-	xfs_inode_t	**i_tab,
-	uint		lock_mode)
-{
-	int	i;
-
-	xfs_iunlock(i_tab[0], lock_mode);
-	for (i = 1; i < 4; i++) {
-		if (i_tab[i] == NULL)
-			break;
-
-		/*
-		 * Watch out for duplicate entries in the table.
-		 */
-		if (i_tab[i] != i_tab[i-1])
-			xfs_iunlock(i_tab[i], lock_mode);
-	}
-}
-
 /*
  * Enter all inodes for a rename transaction into a sorted array.
  */
@@ -204,19 +179,6 @@ xfs_rename(
 	 */
 	xfs_lock_inodes(inodes, num_inodes, XFS_ILOCK_EXCL);
 
-	/*
-	 * If we are using project inheritance, we only allow renames
-	 * into our tree when the project IDs are the same; else the
-	 * tree quota mechanism would be circumvented.
-	 */
-	if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) &&
-		     (target_dp->i_d.di_projid != src_ip->i_d.di_projid))) {
-		error = XFS_ERROR(EXDEV);
-		xfs_rename_unlock4(inodes, XFS_ILOCK_EXCL);
-		xfs_trans_cancel(tp, cancel_flags);
-		goto std_return;
-	}
-
 	/*
 	 * Join all the inodes to the transaction. From this point on,
 	 * we can rely on either trans_commit or trans_cancel to unlock
@@ -241,6 +203,17 @@ xfs_rename(
 		xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL);
 	}
 
+	/*
+	 * If we are using project inheritance, we only allow renames
+	 * into our tree when the project IDs are the same; else the
+	 * tree quota mechanism would be circumvented.
+	 */
+	if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) &&
+		     (target_dp->i_d.di_projid != src_ip->i_d.di_projid))) {
+		error = XFS_ERROR(EXDEV);
+		goto error_return;
+	}
+
 	/*
 	 * Set up the target.
 	 */

fs/xfs/xfs_rw.c

@@ -406,7 +406,7 @@ xfs_bwrite(
 	 * XXXsup how does this work for quotas.
 	 */
 	XFS_BUF_SET_BDSTRAT_FUNC(bp, xfs_bdstrat_cb);
-	XFS_BUF_SET_FSPRIVATE3(bp, mp);
+	bp->b_mount = mp;
 	XFS_BUF_WRITE(bp);
 
 	if ((error = XFS_bwrite(bp))) {

fs/xfs/xfs_sb.h

@@ -79,6 +79,7 @@ struct xfs_mount;
 #define XFS_SB_VERSION2_LAZYSBCOUNTBIT	0x00000002	/* Superblk counters */
 #define XFS_SB_VERSION2_RESERVED4BIT	0x00000004
 #define XFS_SB_VERSION2_ATTR2BIT	0x00000008	/* Inline attr rework */
+#define XFS_SB_VERSION2_PARENTBIT	0x00000010	/* parent pointers */
 
 #define	XFS_SB_VERSION2_OKREALFBITS	\
 	(XFS_SB_VERSION2_LAZYSBCOUNTBIT	| \

fs/xfs/xfs_vnodeops.c

@@ -70,7 +70,6 @@ xfs_setattr(
 	gid_t			gid=0, igid=0;
 	int			timeflags = 0;
 	struct xfs_dquot	*udqp, *gdqp, *olddquot1, *olddquot2;
-	int			file_owner;
 	int			need_iolock = 1;
 
 	xfs_itrace_entry(ip);
@@ -81,6 +80,10 @@ xfs_setattr(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return XFS_ERROR(EIO);
 
+	code = -inode_change_ok(inode, iattr);
+	if (code)
+		return code;
+
 	olddquot1 = olddquot2 = NULL;
 	udqp = gdqp = NULL;
 
@@ -158,56 +161,6 @@ xfs_setattr(
 
 	xfs_ilock(ip, lock_flags);
 
-	/* boolean: are we the file owner? */
-	file_owner = (current_fsuid() == ip->i_d.di_uid);
-
-	/*
-	 * Change various properties of a file.
-	 * Only the owner or users with CAP_FOWNER
-	 * capability may do these things.
-	 */
-	if (mask & (ATTR_MODE|ATTR_UID|ATTR_GID)) {
-		/*
-		 * CAP_FOWNER overrides the following restrictions:
-		 *
-		 * The user ID of the calling process must be equal
-		 * to the file owner ID, except in cases where the
-		 * CAP_FSETID capability is applicable.
-		 */
-		if (!file_owner && !capable(CAP_FOWNER)) {
-			code = XFS_ERROR(EPERM);
-			goto error_return;
-		}
-
-		/*
-		 * CAP_FSETID overrides the following restrictions:
-		 *
-		 * The effective user ID of the calling process shall match
-		 * the file owner when setting the set-user-ID and
-		 * set-group-ID bits on that file.
-		 *
-		 * The effective group ID or one of the supplementary group
-		 * IDs of the calling process shall match the group owner of
-		 * the file when setting the set-group-ID bit on that file
-		 */
-		if (mask & ATTR_MODE) {
-			mode_t m = 0;
-
-			if ((iattr->ia_mode & S_ISUID) && !file_owner)
-				m |= S_ISUID;
-			if ((iattr->ia_mode & S_ISGID) &&
-			    !in_group_p((gid_t)ip->i_d.di_gid))
-				m |= S_ISGID;
-#if 0
-			/* Linux allows this, Irix doesn't. */
-			if ((iattr->ia_mode & S_ISVTX) && !S_ISDIR(ip->i_d.di_mode))
-				m |= S_ISVTX;
-#endif
-			if (m && !capable(CAP_FSETID))
-				iattr->ia_mode &= ~m;
-		}
-	}
-
 	/*
 	 * Change file ownership.  Must be the owner or privileged.
 	 */
@@ -223,22 +176,6 @@ xfs_setattr(
 		gid = (mask & ATTR_GID) ? iattr->ia_gid : igid;
 		uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid;
 
-		/*
-		 * CAP_CHOWN overrides the following restrictions:
-		 *
-		 * If _POSIX_CHOWN_RESTRICTED is defined, this capability
-		 * shall override the restriction that a process cannot
-		 * change the user ID of a file it owns and the restriction
-		 * that the group ID supplied to the chown() function
-		 * shall be equal to either the group ID or one of the
-		 * supplementary group IDs of the calling process.
-		 */
-		if ((iuid != uid ||
-		     (igid != gid && !in_group_p((gid_t)gid))) &&
-		    !capable(CAP_CHOWN)) {
-			code = XFS_ERROR(EPERM);
-			goto error_return;
-		}
 		/*
 		 * Do a quota reservation only if uid/gid is actually
 		 * going to change.
@@ -276,25 +213,13 @@ xfs_setattr(
 			code = XFS_ERROR(EINVAL);
 			goto error_return;
 		}
-		/*
-		 * Make sure that the dquots are attached to the inode.
-		 */
-		if ((code = XFS_QM_DQATTACH(mp, ip, XFS_QMOPT_ILOCKED)))
-			goto error_return;
-	}
 
 		/*
-	 * Change file access or modified times.
+		 * Make sure that the dquots are attached to the inode.
 		 */
-	if (mask & (ATTR_ATIME|ATTR_MTIME)) {
-		if (!file_owner) {
-			if ((mask & (ATTR_MTIME_SET|ATTR_ATIME_SET)) &&
-			    !capable(CAP_FOWNER)) {
-				code = XFS_ERROR(EPERM);
+		code = XFS_QM_DQATTACH(mp, ip, XFS_QMOPT_ILOCKED);
+		if (code)
 			goto error_return;
-			}
-		}
-	}
 
 		/*
 		 * Now we can make the changes.  Before we join the inode
@@ -304,8 +229,6 @@ xfs_setattr(
 		 * to the transaction, because the inode cannot be unlocked
 		 * once it is a part of the transaction.
 		 */
-	if (mask & ATTR_SIZE) {
-		code = 0;
 		if (iattr->ia_size > ip->i_size) {
 			/*
 			 * Do the first part of growing a file: zero any data
@@ -360,17 +283,10 @@ xfs_setattr(
 		}
 		commit_flags = XFS_TRANS_RELEASE_LOG_RES;
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
-	}
 
-	if (tp) {
 		xfs_trans_ijoin(tp, ip, lock_flags);
 		xfs_trans_ihold(tp, ip);
-	}
 
-	/*
-	 * Truncate file.  Must have write permission and not be a directory.
-	 */
-	if (mask & ATTR_SIZE) {
 		/*
 		 * Only change the c/mtime if we are changing the size
 		 * or we are explicitly asked to change it. This handles
@@ -410,20 +326,9 @@ xfs_setattr(
 			 */
 			xfs_iflags_set(ip, XFS_ITRUNCATED);
 		}
-	}
-
-	/*
-	 * Change file access modes.
-	 */
-	if (mask & ATTR_MODE) {
-		ip->i_d.di_mode &= S_IFMT;
-		ip->i_d.di_mode |= iattr->ia_mode & ~S_IFMT;
-
-		inode->i_mode &= S_IFMT;
-		inode->i_mode |= iattr->ia_mode & ~S_IFMT;
-
-		xfs_trans_log_inode (tp, ip, XFS_ILOG_CORE);
-		timeflags |= XFS_ICHGTIME_CHG;
+	} else if (tp) {
+		xfs_trans_ijoin(tp, ip, lock_flags);
+		xfs_trans_ihold(tp, ip);
 	}
 
 	/*
@@ -471,6 +376,24 @@ xfs_setattr(
 		timeflags |= XFS_ICHGTIME_CHG;
 	}
 
+	/*
+	 * Change file access modes.
+	 */
+	if (mask & ATTR_MODE) {
+		umode_t mode = iattr->ia_mode;
+
+		if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
+			mode &= ~S_ISGID;
+
+		ip->i_d.di_mode &= S_IFMT;
+		ip->i_d.di_mode |= mode & ~S_IFMT;
+
+		inode->i_mode &= S_IFMT;
+		inode->i_mode |= mode & ~S_IFMT;
+
+		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+		timeflags |= XFS_ICHGTIME_CHG;
+	}
 
 	/*
 	 * Change file access or modified times.

fs/xfs/xfs_vnodeops.h

@@ -53,8 +53,6 @@ int xfs_attr_set(struct xfs_inode *dp, const char *name, char *value,
 int xfs_attr_remove(struct xfs_inode *dp, const char *name, int flags);
 int xfs_attr_list(struct xfs_inode *dp, char *buffer, int bufsize,
 		int flags, struct attrlist_cursor_kern *cursor);
-int xfs_ioctl(struct xfs_inode *ip, struct file *filp,
-		int ioflags, unsigned int cmd, void __user *arg);
 ssize_t xfs_read(struct xfs_inode *ip, struct kiocb *iocb,
 		const struct iovec *iovp, unsigned int segs,
 		loff_t *offset, int ioflags);

include/linux/fs.h

@@ -81,6 +81,14 @@ extern int dir_notify_enable;
 #define FMODE_WRITE_IOCTL	((__force fmode_t)128)
 #define FMODE_NDELAY_NOW	((__force fmode_t)256)
 
+/*
+ * Don't update ctime and mtime.
+ *
+ * Currently a special hack for the XFS open_by_handle ioctl, but we'll
+ * hopefully graduate it to a proper O_CMTIME flag supported by open(2) soon.
+ */
+#define FMODE_NOCMTIME		((__force fmode_t)2048)
+
 #define RW_MASK		1
 #define RWA_MASK	2
 #define READ 0