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: (37 commits)
  [XFS] Fix lockdep annotations for xfs_lock_inodes
  [LIB]: export radix_tree_preload()
  [XFS] Fix XFS_IOC_FSBULKSTAT{,_SINGLE} & XFS_IOC_FSINUMBERS in compat mode
  [XFS] Compat ioctl handler for handle operations
  [XFS] Compat ioctl handler for XFS_IOC_FSGEOMETRY_V1.
  [XFS] Clean up function name handling in tracing code
  [XFS] Quota inode has no parent.
  [XFS] Concurrent Multi-File Data Streams
  [XFS] Use uninitialized_var macro to stop warning about rtx
  [XFS] XFS should not be looking at filp reference counts
  [XFS] Use is_power_of_2 instead of open coding checks
  [XFS] Reduce shouting by removing unnecessary macros from dir2 code.
  [XFS] Simplify XFS min/max macros.
  [XFS] Kill off xfs_count_bits
  [XFS] Cancel transactions on xfs_itruncate_start error.
  [XFS] Use do_div() on 64 bit types.
  [XFS] Fix remount,readonly path to flush everything correctly.
  [XFS] Cleanup inode extent size hint extraction
  [XFS] Prevent ENOSPC from aborting transactions that need to succeed
  [XFS] Prevent deadlock when flushing inodes on unmount
  ...

fs/xfs/Makefile-linux-2.6

@@ -64,6 +64,7 @@ xfs-y				+= xfs_alloc.o \
 				   xfs_dir2_sf.o \
 				   xfs_error.o \
 				   xfs_extfree_item.o \
+				   xfs_filestream.o \
 				   xfs_fsops.o \
 				   xfs_ialloc.o \
 				   xfs_ialloc_btree.o \
@@ -77,6 +78,7 @@ xfs-y				+= xfs_alloc.o \
 				   xfs_log.o \
 				   xfs_log_recover.o \
 				   xfs_mount.o \
+				   xfs_mru_cache.o \
 				   xfs_rename.o \
 				   xfs_trans.o \
 				   xfs_trans_ail.o \

fs/xfs/linux-2.6/kmem.h

@@ -100,25 +100,6 @@ kmem_zone_destroy(kmem_zone_t *zone)
 extern void *kmem_zone_alloc(kmem_zone_t *, unsigned int __nocast);
 extern void *kmem_zone_zalloc(kmem_zone_t *, unsigned int __nocast);
 
-/*
- * Low memory cache shrinkers
- */
-
-typedef struct shrinker *kmem_shaker_t;
-typedef int (*kmem_shake_func_t)(int, gfp_t);
-
-static inline kmem_shaker_t
-kmem_shake_register(kmem_shake_func_t sfunc)
-{
-	return set_shrinker(DEFAULT_SEEKS, sfunc);
-}
-
-static inline void
-kmem_shake_deregister(kmem_shaker_t shrinker)
-{
-	remove_shrinker(shrinker);
-}
-
 static inline int
 kmem_shake_allow(gfp_t gfp_mask)
 {

fs/xfs/linux-2.6/xfs_aops.c

@@ -108,14 +108,19 @@ xfs_page_trace(
 
 /*
  * Schedule IO completion handling on a xfsdatad if this was
- * the final hold on this ioend.
+ * the final hold on this ioend. If we are asked to wait,
+ * flush the workqueue.
  */
 STATIC void
 xfs_finish_ioend(
-	xfs_ioend_t		*ioend)
+	xfs_ioend_t	*ioend,
+	int		wait)
 {
-	if (atomic_dec_and_test(&ioend->io_remaining))
+	if (atomic_dec_and_test(&ioend->io_remaining)) {
 		queue_work(xfsdatad_workqueue, &ioend->io_work);
+		if (wait)
+			flush_workqueue(xfsdatad_workqueue);
+	}
 }
 
 /*
@@ -156,6 +161,8 @@ xfs_setfilesize(
 	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);
@@ -334,7 +341,7 @@ xfs_end_bio(
 	bio->bi_end_io = NULL;
 	bio_put(bio);
 
-	xfs_finish_ioend(ioend);
+	xfs_finish_ioend(ioend, 0);
 	return 0;
 }
 
@@ -470,7 +477,7 @@ xfs_submit_ioend(
 		}
 		if (bio)
 			xfs_submit_ioend_bio(ioend, bio);
-		xfs_finish_ioend(ioend);
+		xfs_finish_ioend(ioend, 0);
 	} while ((ioend = next) != NULL);
 }
 
@@ -1003,6 +1010,8 @@ xfs_page_state_convert(
 		if (buffer_unwritten(bh) || buffer_delay(bh) ||
 		    ((buffer_uptodate(bh) || PageUptodate(page)) &&
 		     !buffer_mapped(bh) && (unmapped || startio))) {
+			int new_ioend = 0;
+
 			/*
 			 * Make sure we don't use a read-only iomap
 			 */
@@ -1021,6 +1030,15 @@ xfs_page_state_convert(
 			}
 
 			if (!iomap_valid) {
+				/*
+				 * if we didn't have a valid mapping then we
+				 * need to ensure that we put the new mapping
+				 * in a new ioend structure. This needs to be
+				 * done to ensure that the ioends correctly
+				 * reflect the block mappings at io completion
+				 * for unwritten extent conversion.
+				 */
+				new_ioend = 1;
 				if (type == IOMAP_NEW) {
 					size = xfs_probe_cluster(inode,
 							page, bh, head, 0);
@@ -1040,7 +1058,7 @@ xfs_page_state_convert(
 				if (startio) {
 					xfs_add_to_ioend(inode, bh, offset,
 							type, &ioend,
-							!iomap_valid);
+							new_ioend);
 				} else {
 					set_buffer_dirty(bh);
 					unlock_buffer(bh);
@@ -1416,6 +1434,13 @@ xfs_end_io_direct(
 	 * This is not necessary for synchronous direct I/O, but we do
 	 * it anyway to keep the code uniform and simpler.
 	 *
+	 * Well, if only it were that simple. Because synchronous direct I/O
+	 * requires extent conversion to occur *before* we return to userspace,
+	 * we have to wait for extent conversion to complete. Look at the
+	 * iocb that has been passed to us to determine if this is AIO or
+	 * not. If it is synchronous, tell xfs_finish_ioend() to kick the
+	 * workqueue and wait for it to complete.
+	 *
 	 * The core direct I/O code might be changed to always call the
 	 * completion handler in the future, in which case all this can
 	 * go away.
@@ -1423,9 +1448,9 @@ xfs_end_io_direct(
 	ioend->io_offset = offset;
 	ioend->io_size = size;
 	if (ioend->io_type == IOMAP_READ) {
-		xfs_finish_ioend(ioend);
+		xfs_finish_ioend(ioend, 0);
 	} else if (private && size > 0) {
-		xfs_finish_ioend(ioend);
+		xfs_finish_ioend(ioend, is_sync_kiocb(iocb));
 	} else {
 		/*
 		 * A direct I/O write ioend starts it's life in unwritten
@@ -1434,7 +1459,7 @@ xfs_end_io_direct(
 		 * handler.
 		 */
 		INIT_WORK(&ioend->io_work, xfs_end_bio_written);
-		xfs_finish_ioend(ioend);
+		xfs_finish_ioend(ioend, 0);
 	}
 
 	/*

fs/xfs/linux-2.6/xfs_buf.c

@@ -35,7 +35,7 @@
 #include <linux/freezer.h>
 
 static kmem_zone_t *xfs_buf_zone;
-static kmem_shaker_t xfs_buf_shake;
+static struct shrinker *xfs_buf_shake;
 STATIC int xfsbufd(void *);
 STATIC int xfsbufd_wakeup(int, gfp_t);
 STATIC void xfs_buf_delwri_queue(xfs_buf_t *, int);
@@ -314,7 +314,7 @@ xfs_buf_free(
 
 	ASSERT(list_empty(&bp->b_hash_list));
 
-	if (bp->b_flags & _XBF_PAGE_CACHE) {
+	if (bp->b_flags & (_XBF_PAGE_CACHE|_XBF_PAGES)) {
 		uint		i;
 
 		if ((bp->b_flags & XBF_MAPPED) && (bp->b_page_count > 1))
@@ -323,18 +323,11 @@ xfs_buf_free(
 		for (i = 0; i < bp->b_page_count; i++) {
 			struct page	*page = bp->b_pages[i];
 
-			ASSERT(!PagePrivate(page));
+			if (bp->b_flags & _XBF_PAGE_CACHE)
+				ASSERT(!PagePrivate(page));
 			page_cache_release(page);
 		}
 		_xfs_buf_free_pages(bp);
-	} else if (bp->b_flags & _XBF_KMEM_ALLOC) {
-		 /*
-		  * XXX(hch): bp->b_count_desired might be incorrect (see
-		  * xfs_buf_associate_memory for details), but fortunately
-		  * the Linux version of kmem_free ignores the len argument..
-		  */
-		kmem_free(bp->b_addr, bp->b_count_desired);
-		_xfs_buf_free_pages(bp);
 	}
 
 	xfs_buf_deallocate(bp);
@@ -764,43 +757,44 @@ xfs_buf_get_noaddr(
 	size_t			len,
 	xfs_buftarg_t		*target)
 {
-	size_t			malloc_len = len;
+	unsigned long		page_count = PAGE_ALIGN(len) >> PAGE_SHIFT;
+	int			error, i;
 	xfs_buf_t		*bp;
-	void			*data;
-	int			error;
 
 	bp = xfs_buf_allocate(0);
 	if (unlikely(bp == NULL))
 		goto fail;
 	_xfs_buf_initialize(bp, target, 0, len, 0);
 
- try_again:
-	data = kmem_alloc(malloc_len, KM_SLEEP | KM_MAYFAIL | KM_LARGE);
-	if (unlikely(data == NULL))
+	error = _xfs_buf_get_pages(bp, page_count, 0);
+	if (error)
 		goto fail_free_buf;
 
-	/* check whether alignment matches.. */
-	if ((__psunsigned_t)data !=
-	    ((__psunsigned_t)data & ~target->bt_smask)) {
-		/* .. else double the size and try again */
-		kmem_free(data, malloc_len);
-		malloc_len <<= 1;
-		goto try_again;
+	for (i = 0; i < page_count; i++) {
+		bp->b_pages[i] = alloc_page(GFP_KERNEL);
+		if (!bp->b_pages[i])
+			goto fail_free_mem;
 	}
+	bp->b_flags |= _XBF_PAGES;
 
-	error = xfs_buf_associate_memory(bp, data, len);
-	if (error)
+	error = _xfs_buf_map_pages(bp, XBF_MAPPED);
+	if (unlikely(error)) {
+		printk(KERN_WARNING "%s: failed to map pages\n",
+				__FUNCTION__);
 		goto fail_free_mem;
-	bp->b_flags |= _XBF_KMEM_ALLOC;
+	}
 
 	xfs_buf_unlock(bp);
 
-	XB_TRACE(bp, "no_daddr", data);
+	XB_TRACE(bp, "no_daddr", len);
 	return bp;
+
  fail_free_mem:
-	kmem_free(data, malloc_len);
+	while (--i >= 0)
+		__free_page(bp->b_pages[i]);
+	_xfs_buf_free_pages(bp);
  fail_free_buf:
-	xfs_buf_free(bp);
+	xfs_buf_deallocate(bp);
  fail:
 	return NULL;
 }
@@ -1453,6 +1447,7 @@ xfs_free_buftarg(
 	int			external)
 {
 	xfs_flush_buftarg(btp, 1);
+	xfs_blkdev_issue_flush(btp);
 	if (external)
 		xfs_blkdev_put(btp->bt_bdev);
 	xfs_free_bufhash(btp);
@@ -1837,7 +1832,7 @@ xfs_buf_init(void)
 	if (!xfsdatad_workqueue)
 		goto out_destroy_xfslogd_workqueue;
 
-	xfs_buf_shake = kmem_shake_register(xfsbufd_wakeup);
+	xfs_buf_shake = set_shrinker(DEFAULT_SEEKS, xfsbufd_wakeup);
 	if (!xfs_buf_shake)
 		goto out_destroy_xfsdatad_workqueue;
 
@@ -1859,7 +1854,7 @@ xfs_buf_init(void)
 void
 xfs_buf_terminate(void)
 {
-	kmem_shake_deregister(xfs_buf_shake);
+	remove_shrinker(xfs_buf_shake);
 	destroy_workqueue(xfsdatad_workqueue);
 	destroy_workqueue(xfslogd_workqueue);
 	kmem_zone_destroy(xfs_buf_zone);

fs/xfs/linux-2.6/xfs_buf.h

@@ -63,7 +63,7 @@ typedef enum {
 
 	/* flags used only internally */
 	_XBF_PAGE_CACHE = (1 << 17),/* backed by pagecache		   */
-	_XBF_KMEM_ALLOC = (1 << 18),/* backed by kmem_alloc()		   */
+	_XBF_PAGES = (1 << 18),	    /* backed by refcounted pages	   */
 	_XBF_RUN_QUEUES = (1 << 19),/* run block device task queue	   */
 	_XBF_DELWRI_Q = (1 << 21),   /* buffer on delwri queue		   */
 } xfs_buf_flags_t;

fs/xfs/linux-2.6/xfs_file.c

@@ -183,15 +183,6 @@ xfs_file_open(
 	return -bhv_vop_open(vn_from_inode(inode), NULL);
 }
 
-STATIC int
-xfs_file_close(
-	struct file	*filp,
-	fl_owner_t	id)
-{
-	return -bhv_vop_close(vn_from_inode(filp->f_path.dentry->d_inode), 0,
-				file_count(filp) > 1 ? L_FALSE : L_TRUE, NULL);
-}
-
 STATIC int
 xfs_file_release(
 	struct inode	*inode,
@@ -436,7 +427,6 @@ const struct file_operations xfs_file_operations = {
 #endif
 	.mmap		= xfs_file_mmap,
 	.open		= xfs_file_open,
-	.flush		= xfs_file_close,
 	.release	= xfs_file_release,
 	.fsync		= xfs_file_fsync,
 #ifdef HAVE_FOP_OPEN_EXEC
@@ -458,7 +448,6 @@ const struct file_operations xfs_invis_file_operations = {
 #endif
 	.mmap		= xfs_file_mmap,
 	.open		= xfs_file_open,
-	.flush		= xfs_file_close,
 	.release	= xfs_file_release,
 	.fsync		= xfs_file_fsync,
 };

fs/xfs/linux-2.6/xfs_globals.c

@@ -46,6 +46,7 @@ xfs_param_t xfs_params = {
 	.inherit_nosym	= {	0,		0,		1	},
 	.rotorstep	= {	1,		1,		255	},
 	.inherit_nodfrg	= {	0,		1,		1	},
+	.fstrm_timer	= {	1,		50,		3600*100},
 };
 
 /*

fs/xfs/linux-2.6/xfs_ioctl.c

@@ -1019,7 +1019,7 @@ xfs_ioc_bulkstat(
 
 	if (cmd == XFS_IOC_FSINUMBERS)
 		error = xfs_inumbers(mp, &inlast, &count,
-						bulkreq.ubuffer);
+					bulkreq.ubuffer, xfs_inumbers_fmt);
 	else if (cmd == XFS_IOC_FSBULKSTAT_SINGLE)
 		error = xfs_bulkstat_single(mp, &inlast,
 						bulkreq.ubuffer, &done);

fs/xfs/linux-2.6/xfs_ioctl32.c

@@ -23,10 +23,25 @@
 #include <linux/fs.h>
 #include <asm/uaccess.h>
 #include "xfs.h"
-#include "xfs_types.h"
 #include "xfs_fs.h"
+#include "xfs_bit.h"
+#include "xfs_log.h"
+#include "xfs_inum.h"
+#include "xfs_trans.h"
+#include "xfs_sb.h"
+#include "xfs_ag.h"
+#include "xfs_dir2.h"
+#include "xfs_dmapi.h"
+#include "xfs_mount.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_attr_sf.h"
+#include "xfs_dir2_sf.h"
 #include "xfs_vfs.h"
 #include "xfs_vnode.h"
+#include "xfs_dinode.h"
+#include "xfs_inode.h"
+#include "xfs_itable.h"
+#include "xfs_error.h"
 #include "xfs_dfrag.h"
 
 #define  _NATIVE_IOC(cmd, type) \
@@ -34,6 +49,7 @@
 
 #if defined(CONFIG_IA64) || defined(CONFIG_X86_64)
 #define BROKEN_X86_ALIGNMENT
+#define _PACKED __attribute__((packed))
 /* on ia32 l_start is on a 32-bit boundary */
 typedef struct xfs_flock64_32 {
 	__s16		l_type;
@@ -75,35 +91,276 @@ xfs_ioctl32_flock(
 	return (unsigned long)p;
 }
 
+typedef struct compat_xfs_fsop_geom_v1 {
+	__u32		blocksize;	/* filesystem (data) block size */
+	__u32		rtextsize;	/* realtime extent size		*/
+	__u32		agblocks;	/* fsblocks in an AG		*/
+	__u32		agcount;	/* number of allocation groups	*/
+	__u32		logblocks;	/* fsblocks in the log		*/
+	__u32		sectsize;	/* (data) sector size, bytes	*/
+	__u32		inodesize;	/* inode size in bytes		*/
+	__u32		imaxpct;	/* max allowed inode space(%)	*/
+	__u64		datablocks;	/* fsblocks in data subvolume	*/
+	__u64		rtblocks;	/* fsblocks in realtime subvol	*/
+	__u64		rtextents;	/* rt extents in realtime subvol*/
+	__u64		logstart;	/* starting fsblock of the log	*/
+	unsigned char	uuid[16];	/* unique id of the filesystem	*/
+	__u32		sunit;		/* stripe unit, fsblocks	*/
+	__u32		swidth;		/* stripe width, fsblocks	*/
+	__s32		version;	/* structure version		*/
+	__u32		flags;		/* superblock version flags	*/
+	__u32		logsectsize;	/* log sector size, bytes	*/
+	__u32		rtsectsize;	/* realtime sector size, bytes	*/
+	__u32		dirblocksize;	/* directory block size, bytes	*/
+} __attribute__((packed)) compat_xfs_fsop_geom_v1_t;
+
+#define XFS_IOC_FSGEOMETRY_V1_32  \
+	_IOR ('X', 100, struct compat_xfs_fsop_geom_v1)
+
+STATIC unsigned long xfs_ioctl32_geom_v1(unsigned long arg)
+{
+	compat_xfs_fsop_geom_v1_t __user *p32 = (void __user *)arg;
+	xfs_fsop_geom_v1_t __user *p = compat_alloc_user_space(sizeof(*p));
+
+	if (copy_in_user(p, p32, sizeof(*p32)))
+		return -EFAULT;
+	return (unsigned long)p;
+}
+
+typedef struct compat_xfs_inogrp {
+	__u64		xi_startino;	/* starting inode number	*/
+	__s32		xi_alloccount;	/* # bits set in allocmask	*/
+	__u64		xi_allocmask;	/* mask of allocated inodes	*/
+} __attribute__((packed)) compat_xfs_inogrp_t;
+
+STATIC int xfs_inumbers_fmt_compat(
+	void __user *ubuffer,
+	const xfs_inogrp_t *buffer,
+	long count,
+	long *written)
+{
+	compat_xfs_inogrp_t *p32 = ubuffer;
+	long i;
+
+	for (i = 0; i < count; i++) {
+		if (put_user(buffer[i].xi_startino,   &p32[i].xi_startino) ||
+		    put_user(buffer[i].xi_alloccount, &p32[i].xi_alloccount) ||
+		    put_user(buffer[i].xi_allocmask,  &p32[i].xi_allocmask))
+			return -EFAULT;
+	}
+	*written = count * sizeof(*p32);
+	return 0;
+}
+
 #else
 
-typedef struct xfs_fsop_bulkreq32 {
+#define xfs_inumbers_fmt_compat xfs_inumbers_fmt
+#define _PACKED
+
+#endif
+
+/* XFS_IOC_FSBULKSTAT and friends */
+
+typedef struct compat_xfs_bstime {
+	__s32		tv_sec;		/* seconds		*/
+	__s32		tv_nsec;	/* and nanoseconds	*/
+} compat_xfs_bstime_t;
+
+STATIC int xfs_bstime_store_compat(
+	compat_xfs_bstime_t __user *p32,
+	const xfs_bstime_t *p)
+{
+	__s32 sec32;
+
+	sec32 = p->tv_sec;
+	if (put_user(sec32, &p32->tv_sec) ||
+	    put_user(p->tv_nsec, &p32->tv_nsec))
+		return -EFAULT;
+	return 0;
+}
+
+typedef struct compat_xfs_bstat {
+	__u64		bs_ino;		/* inode number			*/
+	__u16		bs_mode;	/* type and mode		*/
+	__u16		bs_nlink;	/* number of links		*/
+	__u32		bs_uid;		/* user id			*/
+	__u32		bs_gid;		/* group id			*/
+	__u32		bs_rdev;	/* device value			*/
+	__s32		bs_blksize;	/* block size			*/
+	__s64		bs_size;	/* file size			*/
+	compat_xfs_bstime_t bs_atime;	/* access time			*/
+	compat_xfs_bstime_t bs_mtime;	/* modify time			*/
+	compat_xfs_bstime_t bs_ctime;	/* inode change time		*/
+	int64_t		bs_blocks;	/* number of blocks		*/
+	__u32		bs_xflags;	/* extended flags		*/
+	__s32		bs_extsize;	/* extent size			*/
+	__s32		bs_extents;	/* number of extents		*/
+	__u32		bs_gen;		/* generation count		*/
+	__u16		bs_projid;	/* project id			*/
+	unsigned char	bs_pad[14];	/* pad space, unused		*/
+	__u32		bs_dmevmask;	/* DMIG event mask		*/
+	__u16		bs_dmstate;	/* DMIG state info		*/
+	__u16		bs_aextents;	/* attribute number of extents	*/
+} _PACKED compat_xfs_bstat_t;
+
+STATIC int xfs_bulkstat_one_fmt_compat(
+	void			__user *ubuffer,
+	const xfs_bstat_t	*buffer)
+{
+	compat_xfs_bstat_t __user *p32 = ubuffer;
+
+	if (put_user(buffer->bs_ino, &p32->bs_ino) ||
+	    put_user(buffer->bs_mode, &p32->bs_mode) ||
+	    put_user(buffer->bs_nlink, &p32->bs_nlink) ||
+	    put_user(buffer->bs_uid, &p32->bs_uid) ||
+	    put_user(buffer->bs_gid, &p32->bs_gid) ||
+	    put_user(buffer->bs_rdev, &p32->bs_rdev) ||
+	    put_user(buffer->bs_blksize, &p32->bs_blksize) ||
+	    put_user(buffer->bs_size, &p32->bs_size) ||
+	    xfs_bstime_store_compat(&p32->bs_atime, &buffer->bs_atime) ||
+	    xfs_bstime_store_compat(&p32->bs_mtime, &buffer->bs_mtime) ||
+	    xfs_bstime_store_compat(&p32->bs_ctime, &buffer->bs_ctime) ||
+	    put_user(buffer->bs_blocks, &p32->bs_blocks) ||
+	    put_user(buffer->bs_xflags, &p32->bs_xflags) ||
+	    put_user(buffer->bs_extsize, &p32->bs_extsize) ||
+	    put_user(buffer->bs_extents, &p32->bs_extents) ||
+	    put_user(buffer->bs_gen, &p32->bs_gen) ||
+	    put_user(buffer->bs_projid, &p32->bs_projid) ||
+	    put_user(buffer->bs_dmevmask, &p32->bs_dmevmask) ||
+	    put_user(buffer->bs_dmstate, &p32->bs_dmstate) ||
+	    put_user(buffer->bs_aextents, &p32->bs_aextents))
+		return -EFAULT;
+	return sizeof(*p32);
+}
+
+
+
+typedef struct compat_xfs_fsop_bulkreq {
 	compat_uptr_t	lastip;		/* last inode # pointer		*/
 	__s32		icount;		/* count of entries in buffer	*/
 	compat_uptr_t	ubuffer;	/* user buffer for inode desc.	*/
-	__s32		ocount;		/* output count pointer		*/
-} xfs_fsop_bulkreq32_t;
+	compat_uptr_t	ocount;		/* output count pointer		*/
+} compat_xfs_fsop_bulkreq_t;
 
-STATIC unsigned long
-xfs_ioctl32_bulkstat(
-	unsigned long		arg)
+#define XFS_IOC_FSBULKSTAT_32 \
+	_IOWR('X', 101, struct compat_xfs_fsop_bulkreq)
+#define XFS_IOC_FSBULKSTAT_SINGLE_32 \
+	_IOWR('X', 102, struct compat_xfs_fsop_bulkreq)
+#define XFS_IOC_FSINUMBERS_32 \
+	_IOWR('X', 103, struct compat_xfs_fsop_bulkreq)
+
+/* copied from xfs_ioctl.c */
+STATIC int
+xfs_ioc_bulkstat_compat(
+	xfs_mount_t		*mp,
+	unsigned int		cmd,
+	void			__user *arg)
 {
-	xfs_fsop_bulkreq32_t	__user *p32 = (void __user *)arg;
-	xfs_fsop_bulkreq_t	__user *p = compat_alloc_user_space(sizeof(*p));
+	compat_xfs_fsop_bulkreq_t __user *p32 = (void __user *)arg;
 	u32			addr;
+	xfs_fsop_bulkreq_t	bulkreq;
+	int			count;	/* # of records returned */
+	xfs_ino_t		inlast;	/* last inode number */
+	int			done;
+	int			error;
+
+	/* done = 1 if there are more stats to get and if bulkstat */
+	/* should be called again (unused here, but used in dmapi) */
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return -XFS_ERROR(EIO);
+
+	if (get_user(addr, &p32->lastip))
+		return -EFAULT;
+	bulkreq.lastip = compat_ptr(addr);
+	if (get_user(bulkreq.icount, &p32->icount) ||
+	    get_user(addr, &p32->ubuffer))
+		return -EFAULT;
+	bulkreq.ubuffer = compat_ptr(addr);
+	if (get_user(addr, &p32->ocount))
+		return -EFAULT;
+	bulkreq.ocount = compat_ptr(addr);
+
+	if (copy_from_user(&inlast, bulkreq.lastip, sizeof(__s64)))
+		return -XFS_ERROR(EFAULT);
+
+	if ((count = bulkreq.icount) <= 0)
+		return -XFS_ERROR(EINVAL);
+
+	if (cmd == XFS_IOC_FSINUMBERS)
+		error = xfs_inumbers(mp, &inlast, &count,
+				bulkreq.ubuffer, xfs_inumbers_fmt_compat);
+	else {
+		/* declare a var to get a warning in case the type changes */
+		bulkstat_one_fmt_pf formatter = xfs_bulkstat_one_fmt_compat;
+		error = xfs_bulkstat(mp, &inlast, &count,
+			xfs_bulkstat_one, formatter,
+			sizeof(compat_xfs_bstat_t), bulkreq.ubuffer,
+			BULKSTAT_FG_QUICK, &done);
+	}
+	if (error)
+		return -error;
+
+	if (bulkreq.ocount != NULL) {
+		if (copy_to_user(bulkreq.lastip, &inlast,
+						sizeof(xfs_ino_t)))
+			return -XFS_ERROR(EFAULT);
+
+		if (copy_to_user(bulkreq.ocount, &count, sizeof(count)))
+			return -XFS_ERROR(EFAULT);
+	}
+
+	return 0;
+}
+
+
+
+typedef struct compat_xfs_fsop_handlereq {
+	__u32		fd;		/* fd for FD_TO_HANDLE		*/
+	compat_uptr_t	path;		/* user pathname		*/
+	__u32		oflags;		/* open flags			*/
+	compat_uptr_t	ihandle;	/* user supplied handle		*/
+	__u32		ihandlen;	/* user supplied length		*/
+	compat_uptr_t	ohandle;	/* user buffer for handle	*/
+	compat_uptr_t	ohandlen;	/* user buffer length		*/
+} compat_xfs_fsop_handlereq_t;
+
+#define XFS_IOC_PATH_TO_FSHANDLE_32 \
+	_IOWR('X', 104, struct compat_xfs_fsop_handlereq)
+#define XFS_IOC_PATH_TO_HANDLE_32 \
+	_IOWR('X', 105, struct compat_xfs_fsop_handlereq)
+#define XFS_IOC_FD_TO_HANDLE_32 \
+	_IOWR('X', 106, struct compat_xfs_fsop_handlereq)
+#define XFS_IOC_OPEN_BY_HANDLE_32 \
+	_IOWR('X', 107, struct compat_xfs_fsop_handlereq)
+#define XFS_IOC_READLINK_BY_HANDLE_32 \
+	_IOWR('X', 108, struct compat_xfs_fsop_handlereq)
+
+STATIC unsigned long xfs_ioctl32_fshandle(unsigned long arg)
+{
+	compat_xfs_fsop_handlereq_t __user *p32 = (void __user *)arg;
+	xfs_fsop_handlereq_t __user *p = compat_alloc_user_space(sizeof(*p));
+	u32 addr;
 
-	if (get_user(addr, &p32->lastip) ||
-	    put_user(compat_ptr(addr), &p->lastip) ||
-	    copy_in_user(&p->icount, &p32->icount, sizeof(s32)) ||
-	    get_user(addr, &p32->ubuffer) ||
-	    put_user(compat_ptr(addr), &p->ubuffer) ||
-	    get_user(addr, &p32->ocount) ||
-	    put_user(compat_ptr(addr), &p->ocount))
+	if (copy_in_user(&p->fd, &p32->fd, sizeof(__u32)) ||
+	    get_user(addr, &p32->path) ||
+	    put_user(compat_ptr(addr), &p->path) ||
+	    copy_in_user(&p->oflags, &p32->oflags, sizeof(__u32)) ||
+	    get_user(addr, &p32->ihandle) ||
+	    put_user(compat_ptr(addr), &p->ihandle) ||
+	    copy_in_user(&p->ihandlen, &p32->ihandlen, sizeof(__u32)) ||
+	    get_user(addr, &p32->ohandle) ||
+	    put_user(compat_ptr(addr), &p->ohandle) ||
+	    get_user(addr, &p32->ohandlen) ||
+	    put_user(compat_ptr(addr), &p->ohandlen))
 		return -EFAULT;
 
 	return (unsigned long)p;
 }
-#endif
+
 
 STATIC long
 xfs_compat_ioctl(
@@ -118,7 +375,6 @@ xfs_compat_ioctl(
 
 	switch (cmd) {
 	case XFS_IOC_DIOINFO:
-	case XFS_IOC_FSGEOMETRY_V1:
 	case XFS_IOC_FSGEOMETRY:
 	case XFS_IOC_GETVERSION:
 	case XFS_IOC_GETXFLAGS:
@@ -131,12 +387,7 @@ xfs_compat_ioctl(
 	case XFS_IOC_GETBMAPA:
 	case XFS_IOC_GETBMAPX:
 /* not handled
-	case XFS_IOC_FD_TO_HANDLE:
-	case XFS_IOC_PATH_TO_HANDLE:
-	case XFS_IOC_PATH_TO_FSHANDLE:
-	case XFS_IOC_OPEN_BY_HANDLE:
 	case XFS_IOC_FSSETDM_BY_HANDLE:
-	case XFS_IOC_READLINK_BY_HANDLE:
 	case XFS_IOC_ATTRLIST_BY_HANDLE:
 	case XFS_IOC_ATTRMULTI_BY_HANDLE:
 */
@@ -166,6 +417,10 @@ xfs_compat_ioctl(
 		arg = xfs_ioctl32_flock(arg);
 		cmd = _NATIVE_IOC(cmd, struct xfs_flock64);
 		break;
+	case XFS_IOC_FSGEOMETRY_V1_32:
+		arg = xfs_ioctl32_geom_v1(arg);
+		cmd = _NATIVE_IOC(cmd, struct xfs_fsop_geom_v1);
+		break;
 
 #else /* These are handled fine if no alignment issues */
 	case XFS_IOC_ALLOCSP:
@@ -176,18 +431,28 @@ xfs_compat_ioctl(
 	case XFS_IOC_FREESP64:
 	case XFS_IOC_RESVSP64:
 	case XFS_IOC_UNRESVSP64:
+	case XFS_IOC_FSGEOMETRY_V1:
 		break;
 
 	/* xfs_bstat_t still has wrong u32 vs u64 alignment */
 	case XFS_IOC_SWAPEXT:
 		break;
 
-	case XFS_IOC_FSBULKSTAT_SINGLE:
-	case XFS_IOC_FSBULKSTAT:
-	case XFS_IOC_FSINUMBERS:
-		arg = xfs_ioctl32_bulkstat(arg);
-		break;
 #endif
+	case XFS_IOC_FSBULKSTAT_32:
+	case XFS_IOC_FSBULKSTAT_SINGLE_32:
+	case XFS_IOC_FSINUMBERS_32:
+		cmd = _NATIVE_IOC(cmd, struct xfs_fsop_bulkreq);
+		return xfs_ioc_bulkstat_compat(XFS_BHVTOI(VNHEAD(vp))->i_mount,
+				cmd, (void*)arg);
+	case XFS_IOC_FD_TO_HANDLE_32:
+	case XFS_IOC_PATH_TO_HANDLE_32:
+	case XFS_IOC_PATH_TO_FSHANDLE_32:
+	case XFS_IOC_OPEN_BY_HANDLE_32:
+	case XFS_IOC_READLINK_BY_HANDLE_32:
+		arg = xfs_ioctl32_fshandle(arg);
+		cmd = _NATIVE_IOC(cmd, struct xfs_fsop_handlereq);
+		break;
 	default:
 		return -ENOIOCTLCMD;
 	}

fs/xfs/linux-2.6/xfs_linux.h

@@ -123,6 +123,7 @@
 #define xfs_inherit_nosymlinks	xfs_params.inherit_nosym.val
 #define xfs_rotorstep		xfs_params.rotorstep.val
 #define xfs_inherit_nodefrag	xfs_params.inherit_nodfrg.val
+#define xfs_fstrm_centisecs	xfs_params.fstrm_timer.val
 
 #define current_cpu()		(raw_smp_processor_id())
 #define current_pid()		(current->pid)

fs/xfs/linux-2.6/xfs_super.c

@@ -547,7 +547,8 @@ vfs_sync_worker(
 
 	if (!(vfsp->vfs_flag & VFS_RDONLY))
 		error = bhv_vfs_sync(vfsp, SYNC_FSDATA | SYNC_BDFLUSH | \
-					SYNC_ATTR | SYNC_REFCACHE, NULL);
+					SYNC_ATTR | SYNC_REFCACHE | SYNC_SUPER,
+					NULL);
 	vfsp->vfs_sync_seq++;
 	wake_up(&vfsp->vfs_wait_single_sync_task);
 }
@@ -663,7 +664,7 @@ xfs_fs_sync_super(
 		 * occur here so don't bother flushing the buftarg (i.e
 		 * SYNC_QUIESCE) because it'll just get dirty again.
 		 */
-		flags = SYNC_FSDATA | SYNC_DELWRI | SYNC_WAIT | SYNC_IOWAIT;
+		flags = SYNC_DATA_QUIESCE;
 	} else
 		flags = SYNC_FSDATA | (wait ? SYNC_WAIT : 0);
 

fs/xfs/linux-2.6/xfs_sysctl.c

@@ -210,6 +210,17 @@ static ctl_table xfs_table[] = {
 		.extra1		= &xfs_params.inherit_nodfrg.min,
 		.extra2		= &xfs_params.inherit_nodfrg.max
 	},
+	{
+		.ctl_name	= XFS_FILESTREAM_TIMER,
+		.procname	= "filestream_centisecs",
+		.data		= &xfs_params.fstrm_timer.val,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1		= &xfs_params.fstrm_timer.min,
+		.extra2		= &xfs_params.fstrm_timer.max,
+	},
 	/* please keep this the last entry */
 #ifdef CONFIG_PROC_FS
 	{

fs/xfs/linux-2.6/xfs_sysctl.h

@@ -47,6 +47,7 @@ typedef struct xfs_param {
 	xfs_sysctl_val_t inherit_nosym;	/* Inherit the "nosymlinks" flag. */
 	xfs_sysctl_val_t rotorstep;	/* inode32 AG rotoring control knob */
 	xfs_sysctl_val_t inherit_nodfrg;/* Inherit the "nodefrag" inode flag. */
+	xfs_sysctl_val_t fstrm_timer;	/* Filestream dir-AG assoc'n timeout. */
 } xfs_param_t;
 
 /*
@@ -86,6 +87,7 @@ enum {
 	XFS_INHERIT_NOSYM = 19,
 	XFS_ROTORSTEP = 20,
 	XFS_INHERIT_NODFRG = 21,
+	XFS_FILESTREAM_TIMER = 22,
 };
 
 extern xfs_param_t	xfs_params;

fs/xfs/linux-2.6/xfs_vfs.h

@@ -92,6 +92,21 @@ typedef enum {
 #define SYNC_REFCACHE		0x0040  /* prune some of the nfs ref cache */
 #define SYNC_REMOUNT		0x0080  /* remount readonly, no dummy LRs */
 #define SYNC_IOWAIT		0x0100  /* wait for all I/O to complete */
+#define SYNC_SUPER		0x0200  /* flush superblock to disk */
+
+/*
+ * When remounting a filesystem read-only or freezing the filesystem,
+ * we have two phases to execute. This first phase is syncing the data
+ * before we quiesce the fielsystem, and the second is flushing all the
+ * inodes out after we've waited for all the transactions created by
+ * the first phase to complete. The second phase uses SYNC_INODE_QUIESCE
+ * to ensure that the inodes are written to their location on disk
+ * rather than just existing in transactions in the log. This means
+ * after a quiesce there is no log replay required to write the inodes
+ * to disk (this is the main difference between a sync and a quiesce).
+ */
+#define SYNC_DATA_QUIESCE	(SYNC_DELWRI|SYNC_FSDATA|SYNC_WAIT|SYNC_IOWAIT)
+#define SYNC_INODE_QUIESCE	(SYNC_REMOUNT|SYNC_ATTR|SYNC_WAIT)
 
 #define SHUTDOWN_META_IO_ERROR	0x0001	/* write attempt to metadata failed */
 #define SHUTDOWN_LOG_IO_ERROR	0x0002	/* write attempt to the log failed */

fs/xfs/linux-2.6/xfs_vnode.h

@@ -129,10 +129,7 @@ typedef enum bhv_vchange {
 	VCHANGE_FLAGS_IOEXCL_COUNT	= 4
 } bhv_vchange_t;
 
-typedef enum { L_FALSE, L_TRUE } lastclose_t;
-
 typedef int	(*vop_open_t)(bhv_desc_t *, struct cred *);
-typedef int	(*vop_close_t)(bhv_desc_t *, int, lastclose_t, struct cred *);
 typedef ssize_t (*vop_read_t)(bhv_desc_t *, struct kiocb *,
 				const struct iovec *, unsigned int,
 				loff_t *, int, struct cred *);
@@ -200,7 +197,6 @@ typedef int	(*vop_iflush_t)(bhv_desc_t *, int);
 typedef struct bhv_vnodeops {
 	bhv_position_t  vn_position;    /* position within behavior chain */
 	vop_open_t		vop_open;
-	vop_close_t		vop_close;
 	vop_read_t		vop_read;
 	vop_write_t		vop_write;
 	vop_splice_read_t	vop_splice_read;
@@ -245,7 +241,6 @@ typedef struct bhv_vnodeops {
 #define VNHEAD(vp)	((vp)->v_bh.bh_first)
 #define VOP(op, vp)	(*((bhv_vnodeops_t *)VNHEAD(vp)->bd_ops)->op)
 #define bhv_vop_open(vp, cr)		VOP(vop_open, vp)(VNHEAD(vp),cr)
-#define bhv_vop_close(vp, f,last,cr)	VOP(vop_close, vp)(VNHEAD(vp),f,last,cr)
 #define bhv_vop_read(vp,file,iov,segs,offset,ioflags,cr)		\
 		VOP(vop_read, vp)(VNHEAD(vp),file,iov,segs,offset,ioflags,cr)
 #define bhv_vop_write(vp,file,iov,segs,offset,ioflags,cr)		\

fs/xfs/quota/xfs_qm.c

@@ -62,10 +62,9 @@ uint		ndquot;
 
 kmem_zone_t	*qm_dqzone;
 kmem_zone_t	*qm_dqtrxzone;
-static kmem_shaker_t	xfs_qm_shaker;
+static struct shrinker *xfs_qm_shaker;
 
 static cred_t	xfs_zerocr;
-static xfs_inode_t	xfs_zeroino;
 
 STATIC void	xfs_qm_list_init(xfs_dqlist_t *, char *, int);
 STATIC void	xfs_qm_list_destroy(xfs_dqlist_t *);
@@ -150,7 +149,7 @@ xfs_Gqm_init(void)
 	} else
 		xqm->qm_dqzone = qm_dqzone;
 
-	xfs_qm_shaker = kmem_shake_register(xfs_qm_shake);
+	xfs_qm_shaker = set_shrinker(DEFAULT_SEEKS, xfs_qm_shake);
 
 	/*
 	 * The t_dqinfo portion of transactions.
@@ -182,7 +181,7 @@ xfs_qm_destroy(
 
 	ASSERT(xqm != NULL);
 	ASSERT(xqm->qm_nrefs == 0);
-	kmem_shake_deregister(xfs_qm_shaker);
+	remove_shrinker(xfs_qm_shaker);
 	hsize = xqm->qm_dqhashmask + 1;
 	for (i = 0; i < hsize; i++) {
 		xfs_qm_list_destroy(&(xqm->qm_usr_dqhtable[i]));
@@ -1415,7 +1414,7 @@ xfs_qm_qino_alloc(
 		return error;
 	}
 
-	if ((error = xfs_dir_ialloc(&tp, &xfs_zeroino, S_IFREG, 1, 0,
+	if ((error = xfs_dir_ialloc(&tp, NULL, S_IFREG, 1, 0,
 				   &xfs_zerocr, 0, 1, ip, &committed))) {
 		xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES |
 				 XFS_TRANS_ABORT);

fs/xfs/xfs.h

@@ -38,6 +38,7 @@
 #define XFS_RW_TRACE 1
 #define XFS_BUF_TRACE 1
 #define XFS_VNODE_TRACE 1
+#define XFS_FILESTREAMS_TRACE 1
 #endif
 
 #include <linux-2.6/xfs_linux.h>

fs/xfs/xfs_ag.h

@@ -68,6 +68,7 @@ typedef struct xfs_agf {
 	__be32		agf_flcount;	/* count of blocks in freelist */
 	__be32		agf_freeblks;	/* total free blocks */
 	__be32		agf_longest;	/* longest free space */
+	__be32		agf_btreeblks;	/* # of blocks held in AGF btrees */
 } xfs_agf_t;
 
 #define	XFS_AGF_MAGICNUM	0x00000001
@@ -81,7 +82,8 @@ typedef struct xfs_agf {
 #define	XFS_AGF_FLCOUNT		0x00000100
 #define	XFS_AGF_FREEBLKS	0x00000200
 #define	XFS_AGF_LONGEST		0x00000400
-#define	XFS_AGF_NUM_BITS	11
+#define	XFS_AGF_BTREEBLKS	0x00000800
+#define	XFS_AGF_NUM_BITS	12
 #define	XFS_AGF_ALL_BITS	((1 << XFS_AGF_NUM_BITS) - 1)
 
 /* disk block (xfs_daddr_t) in the AG */
@@ -186,12 +188,15 @@ typedef struct xfs_perag
 	__uint32_t	pagf_flcount;	/* count of blocks in freelist */
 	xfs_extlen_t	pagf_freeblks;	/* total free blocks */
 	xfs_extlen_t	pagf_longest;	/* longest free space */
+	__uint32_t	pagf_btreeblks;	/* # of blocks held in AGF btrees */
 	xfs_agino_t	pagi_freecount;	/* number of free inodes */
+	xfs_agino_t	pagi_count;	/* number of allocated inodes */
+	int		pagb_count;	/* pagb slots in use */
 #ifdef __KERNEL__
 	lock_t		pagb_lock;	/* lock for pagb_list */
 #endif
-	int		pagb_count;	/* pagb slots in use */
 	xfs_perag_busy_t *pagb_list;	/* unstable blocks */
+	atomic_t        pagf_fstrms;    /* # of filestreams active in this AG */
 } xfs_perag_t;
 
 #define	XFS_AG_MAXLEVELS(mp)		((mp)->m_ag_maxlevels)

fs/xfs/xfs_alloc.c

@@ -55,17 +55,17 @@ xfs_alloc_search_busy(xfs_trans_t *tp,
 ktrace_t *xfs_alloc_trace_buf;
 
 #define	TRACE_ALLOC(s,a)	\
-	xfs_alloc_trace_alloc(fname, s, a, __LINE__)
+	xfs_alloc_trace_alloc(__FUNCTION__, s, a, __LINE__)
 #define	TRACE_FREE(s,a,b,x,f)	\
-	xfs_alloc_trace_free(fname, s, mp, a, b, x, f, __LINE__)
+	xfs_alloc_trace_free(__FUNCTION__, s, mp, a, b, x, f, __LINE__)
 #define	TRACE_MODAGF(s,a,f)	\
-	xfs_alloc_trace_modagf(fname, s, mp, a, f, __LINE__)
-#define	TRACE_BUSY(fname,s,ag,agb,l,sl,tp)	\
-	xfs_alloc_trace_busy(fname, s, mp, ag, agb, l, sl, tp, XFS_ALLOC_KTRACE_BUSY, __LINE__)
-#define	TRACE_UNBUSY(fname,s,ag,sl,tp)	\
-	xfs_alloc_trace_busy(fname, s, mp, ag, -1, -1, sl, tp, XFS_ALLOC_KTRACE_UNBUSY, __LINE__)
-#define	TRACE_BUSYSEARCH(fname,s,ag,agb,l,sl,tp)	\
-	xfs_alloc_trace_busy(fname, s, mp, ag, agb, l, sl, tp, XFS_ALLOC_KTRACE_BUSYSEARCH, __LINE__)
+	xfs_alloc_trace_modagf(__FUNCTION__, s, mp, a, f, __LINE__)
+#define	TRACE_BUSY(__FUNCTION__,s,ag,agb,l,sl,tp)	\
+	xfs_alloc_trace_busy(__FUNCTION__, s, mp, ag, agb, l, sl, tp, XFS_ALLOC_KTRACE_BUSY, __LINE__)
+#define	TRACE_UNBUSY(__FUNCTION__,s,ag,sl,tp)	\
+	xfs_alloc_trace_busy(__FUNCTION__, s, mp, ag, -1, -1, sl, tp, XFS_ALLOC_KTRACE_UNBUSY, __LINE__)
+#define	TRACE_BUSYSEARCH(__FUNCTION__,s,ag,agb,l,sl,tp)	\
+	xfs_alloc_trace_busy(__FUNCTION__, s, mp, ag, agb, l, sl, tp, XFS_ALLOC_KTRACE_BUSYSEARCH, __LINE__)
 #else
 #define	TRACE_ALLOC(s,a)
 #define	TRACE_FREE(s,a,b,x,f)
@@ -420,7 +420,7 @@ xfs_alloc_read_agfl(
  */
 STATIC void
 xfs_alloc_trace_alloc(
-	char		*name,		/* function tag string */
+	const char	*name,		/* function tag string */
 	char		*str,		/* additional string */
 	xfs_alloc_arg_t	*args,		/* allocation argument structure */
 	int		line)		/* source line number */
@@ -453,7 +453,7 @@ xfs_alloc_trace_alloc(
  */
 STATIC void
 xfs_alloc_trace_free(
-	char		*name,		/* function tag string */
+	const char	*name,		/* function tag string */
 	char		*str,		/* additional string */
 	xfs_mount_t	*mp,		/* file system mount point */
 	xfs_agnumber_t	agno,		/* allocation group number */
@@ -479,7 +479,7 @@ xfs_alloc_trace_free(
  */
 STATIC void
 xfs_alloc_trace_modagf(
-	char		*name,		/* function tag string */
+	const char	*name,		/* function tag string */
 	char		*str,		/* additional string */
 	xfs_mount_t	*mp,		/* file system mount point */
 	xfs_agf_t	*agf,		/* new agf value */
@@ -507,7 +507,7 @@ xfs_alloc_trace_modagf(
 
 STATIC void
 xfs_alloc_trace_busy(
-	char		*name,		/* function tag string */
+	const char	*name,		/* function tag string */
 	char		*str,		/* additional string */
 	xfs_mount_t	*mp,		/* file system mount point */
 	xfs_agnumber_t	agno,		/* allocation group number */
@@ -549,9 +549,6 @@ xfs_alloc_ag_vextent(
 	xfs_alloc_arg_t	*args)	/* argument structure for allocation */
 {
 	int		error=0;
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_ag_vextent";
-#endif
 
 	ASSERT(args->minlen > 0);
 	ASSERT(args->maxlen > 0);
@@ -635,9 +632,6 @@ xfs_alloc_ag_vextent_exact(
 	xfs_agblock_t	fbno;	/* start block of found extent */
 	xfs_agblock_t	fend;	/* end block of found extent */
 	xfs_extlen_t	flen;	/* length of found extent */
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_ag_vextent_exact";
-#endif
 	int		i;	/* success/failure of operation */
 	xfs_agblock_t	maxend;	/* end of maximal extent */
 	xfs_agblock_t	minend;	/* end of minimal extent */
@@ -737,9 +731,6 @@ xfs_alloc_ag_vextent_near(
 	xfs_btree_cur_t	*bno_cur_gt;	/* cursor for bno btree, right side */
 	xfs_btree_cur_t	*bno_cur_lt;	/* cursor for bno btree, left side */
 	xfs_btree_cur_t	*cnt_cur;	/* cursor for count btree */
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_ag_vextent_near";
-#endif
 	xfs_agblock_t	gtbno;		/* start bno of right side entry */
 	xfs_agblock_t	gtbnoa;		/* aligned ... */
 	xfs_extlen_t	gtdiff;		/* difference to right side entry */
@@ -1270,9 +1261,6 @@ xfs_alloc_ag_vextent_size(
 	int		error;		/* error result */
 	xfs_agblock_t	fbno;		/* start of found freespace */
 	xfs_extlen_t	flen;		/* length of found freespace */
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_ag_vextent_size";
-#endif
 	int		i;		/* temp status variable */
 	xfs_agblock_t	rbno;		/* returned block number */
 	xfs_extlen_t	rlen;		/* length of returned extent */
@@ -1427,9 +1415,6 @@ xfs_alloc_ag_vextent_small(
 	int		error;
 	xfs_agblock_t	fbno;
 	xfs_extlen_t	flen;
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_ag_vextent_small";
-#endif
 	int		i;
 
 	if ((error = xfs_alloc_decrement(ccur, 0, &i)))
@@ -1447,7 +1432,8 @@ xfs_alloc_ag_vextent_small(
 	else if (args->minlen == 1 && args->alignment == 1 && !args->isfl &&
 		 (be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_flcount)
 		  > args->minleft)) {
-		if ((error = xfs_alloc_get_freelist(args->tp, args->agbp, &fbno)))
+		error = xfs_alloc_get_freelist(args->tp, args->agbp, &fbno, 0);
+		if (error)
 			goto error0;
 		if (fbno != NULLAGBLOCK) {
 			if (args->userdata) {
@@ -1515,9 +1501,6 @@ xfs_free_ag_extent(
 	xfs_btree_cur_t	*bno_cur;	/* cursor for by-block btree */
 	xfs_btree_cur_t	*cnt_cur;	/* cursor for by-size btree */
 	int		error;		/* error return value */
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_free_ag_extent";
-#endif
 	xfs_agblock_t	gtbno;		/* start of right neighbor block */
 	xfs_extlen_t	gtlen;		/* length of right neighbor block */
 	int		haveleft;	/* have a left neighbor block */
@@ -1923,7 +1906,8 @@ xfs_alloc_fix_freelist(
 	while (be32_to_cpu(agf->agf_flcount) > need) {
 		xfs_buf_t	*bp;
 
-		if ((error = xfs_alloc_get_freelist(tp, agbp, &bno)))
+		error = xfs_alloc_get_freelist(tp, agbp, &bno, 0);
+		if (error)
 			return error;
 		if ((error = xfs_free_ag_extent(tp, agbp, args->agno, bno, 1, 1)))
 			return error;
@@ -1973,8 +1957,9 @@ xfs_alloc_fix_freelist(
 		 * Put each allocated block on the list.
 		 */
 		for (bno = targs.agbno; bno < targs.agbno + targs.len; bno++) {
-			if ((error = xfs_alloc_put_freelist(tp, agbp, agflbp,
-					bno)))
+			error = xfs_alloc_put_freelist(tp, agbp,
+							agflbp, bno, 0);
+			if (error)
 				return error;
 		}
 	}
@@ -1991,16 +1976,15 @@ int				/* error */
 xfs_alloc_get_freelist(
 	xfs_trans_t	*tp,	/* transaction pointer */
 	xfs_buf_t	*agbp,	/* buffer containing the agf structure */
-	xfs_agblock_t	*bnop)	/* block address retrieved from freelist */
+	xfs_agblock_t	*bnop,	/* block address retrieved from freelist */
+	int		btreeblk) /* destination is a AGF btree */
 {
 	xfs_agf_t	*agf;	/* a.g. freespace structure */
 	xfs_agfl_t	*agfl;	/* a.g. freelist structure */
 	xfs_buf_t	*agflbp;/* buffer for a.g. freelist structure */
 	xfs_agblock_t	bno;	/* block number returned */
 	int		error;
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_get_freelist";
-#endif
+	int		logflags;
 	xfs_mount_t	*mp;	/* mount structure */
 	xfs_perag_t	*pag;	/* per allocation group data */
 
@@ -2032,8 +2016,16 @@ xfs_alloc_get_freelist(
 	be32_add(&agf->agf_flcount, -1);
 	xfs_trans_agflist_delta(tp, -1);
 	pag->pagf_flcount--;
-	TRACE_MODAGF(NULL, agf, XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT);
-	xfs_alloc_log_agf(tp, agbp, XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT);
+
+	logflags = XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT;
+	if (btreeblk) {
+		be32_add(&agf->agf_btreeblks, 1);
+		pag->pagf_btreeblks++;
+		logflags |= XFS_AGF_BTREEBLKS;
+	}
+
+	TRACE_MODAGF(NULL, agf, logflags);
+	xfs_alloc_log_agf(tp, agbp, logflags);
 	*bnop = bno;
 
 	/*
@@ -2071,6 +2063,7 @@ xfs_alloc_log_agf(
 		offsetof(xfs_agf_t, agf_flcount),
 		offsetof(xfs_agf_t, agf_freeblks),
 		offsetof(xfs_agf_t, agf_longest),
+		offsetof(xfs_agf_t, agf_btreeblks),
 		sizeof(xfs_agf_t)
 	};
 
@@ -2106,15 +2099,14 @@ xfs_alloc_put_freelist(
 	xfs_trans_t		*tp,	/* transaction pointer */
 	xfs_buf_t		*agbp,	/* buffer for a.g. freelist header */
 	xfs_buf_t		*agflbp,/* buffer for a.g. free block array */
-	xfs_agblock_t		bno)	/* block being freed */
+	xfs_agblock_t		bno,	/* block being freed */
+	int			btreeblk) /* block came from a AGF btree */
 {
 	xfs_agf_t		*agf;	/* a.g. freespace structure */
 	xfs_agfl_t		*agfl;	/* a.g. free block array */
 	__be32			*blockp;/* pointer to array entry */
 	int			error;
-#ifdef XFS_ALLOC_TRACE
-	static char		fname[] = "xfs_alloc_put_freelist";
-#endif
+	int			logflags;
 	xfs_mount_t		*mp;	/* mount structure */
 	xfs_perag_t		*pag;	/* per allocation group data */
 
@@ -2132,11 +2124,22 @@ xfs_alloc_put_freelist(
 	be32_add(&agf->agf_flcount, 1);
 	xfs_trans_agflist_delta(tp, 1);
 	pag->pagf_flcount++;
+
+	logflags = XFS_AGF_FLLAST | XFS_AGF_FLCOUNT;
+	if (btreeblk) {
+		be32_add(&agf->agf_btreeblks, -1);
+		pag->pagf_btreeblks--;
+		logflags |= XFS_AGF_BTREEBLKS;
+	}
+
+	TRACE_MODAGF(NULL, agf, logflags);
+	xfs_alloc_log_agf(tp, agbp, logflags);
+
 	ASSERT(be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp));
 	blockp = &agfl->agfl_bno[be32_to_cpu(agf->agf_fllast)];
 	*blockp = cpu_to_be32(bno);
-	TRACE_MODAGF(NULL, agf, XFS_AGF_FLLAST | XFS_AGF_FLCOUNT);
-	xfs_alloc_log_agf(tp, agbp, XFS_AGF_FLLAST | XFS_AGF_FLCOUNT);
+	TRACE_MODAGF(NULL, agf, logflags);
+	xfs_alloc_log_agf(tp, agbp, logflags);
 	xfs_trans_log_buf(tp, agflbp,
 		(int)((xfs_caddr_t)blockp - (xfs_caddr_t)agfl),
 		(int)((xfs_caddr_t)blockp - (xfs_caddr_t)agfl +
@@ -2196,6 +2199,7 @@ xfs_alloc_read_agf(
 	pag = &mp->m_perag[agno];
 	if (!pag->pagf_init) {
 		pag->pagf_freeblks = be32_to_cpu(agf->agf_freeblks);
+		pag->pagf_btreeblks = be32_to_cpu(agf->agf_btreeblks);
 		pag->pagf_flcount = be32_to_cpu(agf->agf_flcount);
 		pag->pagf_longest = be32_to_cpu(agf->agf_longest);
 		pag->pagf_levels[XFS_BTNUM_BNOi] =
@@ -2235,9 +2239,6 @@ xfs_alloc_vextent(
 	xfs_agblock_t	agsize;	/* allocation group size */
 	int		error;
 	int		flags;	/* XFS_ALLOC_FLAG_... locking flags */
-#ifdef XFS_ALLOC_TRACE
-	static char	fname[] = "xfs_alloc_vextent";
-#endif
 	xfs_extlen_t	minleft;/* minimum left value, temp copy */
 	xfs_mount_t	*mp;	/* mount structure pointer */
 	xfs_agnumber_t	sagno;	/* starting allocation group number */

fs/xfs/xfs_alloc.h

@@ -136,7 +136,8 @@ int				/* error */
 xfs_alloc_get_freelist(
 	struct xfs_trans *tp,	/* transaction pointer */
 	struct xfs_buf	*agbp,	/* buffer containing the agf structure */
-	xfs_agblock_t	*bnop);	/* block address retrieved from freelist */
+	xfs_agblock_t	*bnop,	/* block address retrieved from freelist */
+	int		btreeblk); /* destination is a AGF btree */
 
 /*
  * Log the given fields from the agf structure.
@@ -165,7 +166,8 @@ xfs_alloc_put_freelist(
 	struct xfs_trans *tp,	/* transaction pointer */
 	struct xfs_buf	*agbp,	/* buffer for a.g. freelist header */
 	struct xfs_buf	*agflbp,/* buffer for a.g. free block array */
-	xfs_agblock_t	bno);	/* block being freed */
+	xfs_agblock_t	bno,	/* block being freed */
+	int		btreeblk); /* owner was a AGF btree */
 
 /*
  * Read in the allocation group header (free/alloc section).

fs/xfs/xfs_alloc_btree.c

@@ -226,8 +226,9 @@ xfs_alloc_delrec(
 			/*
 			 * Put this buffer/block on the ag's freelist.
 			 */
-			if ((error = xfs_alloc_put_freelist(cur->bc_tp,
-					cur->bc_private.a.agbp, NULL, bno)))
+			error = xfs_alloc_put_freelist(cur->bc_tp,
+					cur->bc_private.a.agbp, NULL, bno, 1);
+			if (error)
 				return error;
 			/*
 			 * Since blocks move to the free list without the
@@ -549,8 +550,9 @@ xfs_alloc_delrec(
 	/*
 	 * Free the deleting block by putting it on the freelist.
 	 */
-	if ((error = xfs_alloc_put_freelist(cur->bc_tp, cur->bc_private.a.agbp,
-			NULL, rbno)))
+	error = xfs_alloc_put_freelist(cur->bc_tp,
+					 cur->bc_private.a.agbp, NULL, rbno, 1);
+	if (error)
 		return error;
 	/*
 	 * Since blocks move to the free list without the coordination
@@ -1320,8 +1322,9 @@ xfs_alloc_newroot(
 	/*
 	 * Get a buffer from the freelist blocks, for the new root.
 	 */
-	if ((error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_private.a.agbp,
-			&nbno)))
+	error = xfs_alloc_get_freelist(cur->bc_tp,
+					cur->bc_private.a.agbp, &nbno, 1);
+	if (error)
 		return error;
 	/*
 	 * None available, we fail.
@@ -1604,8 +1607,9 @@ xfs_alloc_split(
 	 * Allocate the new block from the freelist.
 	 * If we can't do it, we're toast.  Give up.
 	 */
-	if ((error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_private.a.agbp,
-			&rbno)))
+	error = xfs_alloc_get_freelist(cur->bc_tp,
+					 cur->bc_private.a.agbp, &rbno, 1);
+	if (error)
 		return error;
 	if (rbno == NULLAGBLOCK) {
 		*stat = 0;

fs/xfs/xfs_bit.c

@@ -65,44 +65,6 @@ static const char xfs_highbit[256] = {
 };
 #endif
 
-/*
- * Count of bits set in byte, 0..8.
- */
-static const char xfs_countbit[256] = {
-	0, 1, 1, 2, 1, 2, 2, 3,			/* 00 .. 07 */
-	1, 2, 2, 3, 2, 3, 3, 4,			/* 08 .. 0f */
-	1, 2, 2, 3, 2, 3, 3, 4,			/* 10 .. 17 */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 18 .. 1f */
-	1, 2, 2, 3, 2, 3, 3, 4,			/* 20 .. 27 */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 28 .. 2f */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 30 .. 37 */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* 38 .. 3f */
-	1, 2, 2, 3, 2, 3, 3, 4,			/* 40 .. 47 */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 48 .. 4f */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 50 .. 57 */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* 58 .. 5f */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 60 .. 67 */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* 68 .. 6f */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* 70 .. 77 */
-	4, 5, 5, 6, 5, 6, 6, 7,			/* 78 .. 7f */
-	1, 2, 2, 3, 2, 3, 3, 4,			/* 80 .. 87 */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 88 .. 8f */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* 90 .. 97 */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* 98 .. 9f */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* a0 .. a7 */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* a8 .. af */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* b0 .. b7 */
-	4, 5, 5, 6, 5, 6, 6, 7,			/* b8 .. bf */
-	2, 3, 3, 4, 3, 4, 4, 5,			/* c0 .. c7 */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* c8 .. cf */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* d0 .. d7 */
-	4, 5, 5, 6, 5, 6, 6, 7,			/* d8 .. df */
-	3, 4, 4, 5, 4, 5, 5, 6,			/* e0 .. e7 */
-	4, 5, 5, 6, 5, 6, 6, 7,			/* e8 .. ef */
-	4, 5, 5, 6, 5, 6, 6, 7,			/* f0 .. f7 */
-	5, 6, 6, 7, 6, 7, 7, 8,			/* f8 .. ff */
-};
-
 /*
  * xfs_highbit32: get high bit set out of 32-bit argument, -1 if none set.
  */
@@ -167,56 +129,21 @@ xfs_highbit64(
 
 
 /*
- * Count the number of bits set in the bitmap starting with bit
- * start_bit.  Size is the size of the bitmap in words.
- *
- * Do the counting by mapping a byte value to the number of set
- * bits for that value using the xfs_countbit array, i.e.
- * xfs_countbit[0] == 0, xfs_countbit[1] == 1, xfs_countbit[2] == 1,
- * xfs_countbit[3] == 2, etc.
+ * Return whether bitmap is empty.
+ * Size is number of words in the bitmap, which is padded to word boundary
+ * Returns 1 for empty, 0 for non-empty.
  */
 int
-xfs_count_bits(uint *map, uint size, uint start_bit)
+xfs_bitmap_empty(uint *map, uint size)
 {
-	register int	bits;
-	register unsigned char	*bytep;
-	register unsigned char	*end_map;
-	int		byte_bit;
-
-	bits = 0;
-	end_map = (char*)(map + size);
-	bytep = (char*)(map + (start_bit & ~0x7));
-	byte_bit = start_bit & 0x7;
-
-	/*
-	 * If the caller fell off the end of the map, return 0.
-	 */
-	if (bytep >= end_map) {
-		return (0);
-	}
-
-	/*
-	 * If start_bit is not byte aligned, then process the
-	 * first byte separately.
-	 */
-	if (byte_bit != 0) {
-		/*
-		 * Shift off the bits we don't want to look at,
-		 * before indexing into xfs_countbit.
-		 */
-		bits += xfs_countbit[(*bytep >> byte_bit)];
-		bytep++;
-	}
+	uint i;
+	uint ret = 0;
 
-	/*
-	 * Count the bits in each byte until the end of the bitmap.
-	 */
-	while (bytep < end_map) {
-		bits += xfs_countbit[*bytep];
-		bytep++;
+	for (i = 0; i < size; i++) {
+		ret |= map[i];
 	}
 
-	return (bits);
+	return (ret == 0);
 }
 
 /*

fs/xfs/xfs_bit.h

@@ -55,8 +55,8 @@ extern int xfs_lowbit64(__uint64_t v);
 /* Get high bit set out of 64-bit argument, -1 if none set */
 extern int xfs_highbit64(__uint64_t);
 
-/* Count set bits in map starting with start_bit */
-extern int xfs_count_bits(uint *map, uint size, uint start_bit);
+/* Return whether bitmap is empty (1 == empty) */
+extern int xfs_bitmap_empty(uint *map, uint size);
 
 /* Count continuous one bits in map starting with start_bit */
 extern int xfs_contig_bits(uint *map, uint size, uint start_bit);

fs/xfs/xfs_bmap.h

@@ -144,12 +144,14 @@ extern ktrace_t	*xfs_bmap_trace_buf;
  */
 void
 xfs_bmap_trace_exlist(
-	char			*fname,		/* function name */
+	const char		*fname,		/* function name */
 	struct xfs_inode	*ip,		/* incore inode pointer */
 	xfs_extnum_t		cnt,		/* count of entries in list */
 	int			whichfork);	/* data or attr fork */
+#define	XFS_BMAP_TRACE_EXLIST(ip,c,w)	\
+	xfs_bmap_trace_exlist(__FUNCTION__,ip,c,w)
 #else
-#define	xfs_bmap_trace_exlist(f,ip,c,w)
+#define	XFS_BMAP_TRACE_EXLIST(ip,c,w)
 #endif
 
 /*

fs/xfs/xfs_bmap_btree.c

@@ -76,7 +76,7 @@ static char	EXIT[] = "exit";
  */
 STATIC void
 xfs_bmbt_trace_enter(
-	char		*func,
+	const char	*func,
 	xfs_btree_cur_t	*cur,
 	char		*s,
 	int		type,
@@ -117,7 +117,7 @@ xfs_bmbt_trace_enter(
  */
 STATIC void
 xfs_bmbt_trace_argbi(
-	char		*func,
+	const char	*func,
 	xfs_btree_cur_t	*cur,
 	xfs_buf_t	*b,
 	int		i,
@@ -134,7 +134,7 @@ xfs_bmbt_trace_argbi(
  */
 STATIC void
 xfs_bmbt_trace_argbii(
-	char		*func,
+	const char	*func,
 	xfs_btree_cur_t	*cur,
 	xfs_buf_t	*b,
 	int		i0,
@@ -153,7 +153,7 @@ xfs_bmbt_trace_argbii(
  */
 STATIC void
 xfs_bmbt_trace_argfffi(
-	char			*func,
+	const char		*func,
 	xfs_btree_cur_t		*cur,
 	xfs_dfiloff_t		o,
 	xfs_dfsbno_t		b,
@@ -172,7 +172,7 @@ xfs_bmbt_trace_argfffi(
  */
 STATIC void
 xfs_bmbt_trace_argi(
-	char		*func,
+	const char	*func,
 	xfs_btree_cur_t	*cur,
 	int		i,
 	int		line)
@@ -188,7 +188,7 @@ xfs_bmbt_trace_argi(
  */
 STATIC void
 xfs_bmbt_trace_argifk(
-	char			*func,
+	const char		*func,
 	xfs_btree_cur_t		*cur,
 	int			i,
 	xfs_fsblock_t		f,
@@ -206,7 +206,7 @@ xfs_bmbt_trace_argifk(
  */
 STATIC void
 xfs_bmbt_trace_argifr(
-	char			*func,
+	const char		*func,
 	xfs_btree_cur_t		*cur,
 	int			i,
 	xfs_fsblock_t		f,
@@ -235,7 +235,7 @@ xfs_bmbt_trace_argifr(
  */
 STATIC void
 xfs_bmbt_trace_argik(
-	char			*func,
+	const char		*func,
 	xfs_btree_cur_t		*cur,
 	int			i,
 	xfs_bmbt_key_t		*k,
@@ -255,7 +255,7 @@ xfs_bmbt_trace_argik(
  */
 STATIC void
 xfs_bmbt_trace_cursor(
-	char		*func,
+	const char	*func,
 	xfs_btree_cur_t	*cur,
 	char		*s,
 	int		line)
@@ -274,21 +274,21 @@ xfs_bmbt_trace_cursor(
 }
 
 #define	XFS_BMBT_TRACE_ARGBI(c,b,i)	\
-	xfs_bmbt_trace_argbi(fname, c, b, i, __LINE__)
+	xfs_bmbt_trace_argbi(__FUNCTION__, c, b, i, __LINE__)
 #define	XFS_BMBT_TRACE_ARGBII(c,b,i,j)	\
-	xfs_bmbt_trace_argbii(fname, c, b, i, j, __LINE__)
+	xfs_bmbt_trace_argbii(__FUNCTION__, c, b, i, j, __LINE__)
 #define	XFS_BMBT_TRACE_ARGFFFI(c,o,b,i,j)	\
-	xfs_bmbt_trace_argfffi(fname, c, o, b, i, j, __LINE__)
+	xfs_bmbt_trace_argfffi(__FUNCTION__, c, o, b, i, j, __LINE__)
 #define	XFS_BMBT_TRACE_ARGI(c,i)	\
-	xfs_bmbt_trace_argi(fname, c, i, __LINE__)
+	xfs_bmbt_trace_argi(__FUNCTION__, c, i, __LINE__)
 #define	XFS_BMBT_TRACE_ARGIFK(c,i,f,s)	\
-	xfs_bmbt_trace_argifk(fname, c, i, f, s, __LINE__)
+	xfs_bmbt_trace_argifk(__FUNCTION__, c, i, f, s, __LINE__)
 #define	XFS_BMBT_TRACE_ARGIFR(c,i,f,r)	\
-	xfs_bmbt_trace_argifr(fname, c, i, f, r, __LINE__)
+	xfs_bmbt_trace_argifr(__FUNCTION__, c, i, f, r, __LINE__)
 #define	XFS_BMBT_TRACE_ARGIK(c,i,k)	\
-	xfs_bmbt_trace_argik(fname, c, i, k, __LINE__)
+	xfs_bmbt_trace_argik(__FUNCTION__, c, i, k, __LINE__)
 #define	XFS_BMBT_TRACE_CURSOR(c,s)	\
-	xfs_bmbt_trace_cursor(fname, c, s, __LINE__)
+	xfs_bmbt_trace_cursor(__FUNCTION__, c, s, __LINE__)
 #else
 #define	XFS_BMBT_TRACE_ARGBI(c,b,i)
 #define	XFS_BMBT_TRACE_ARGBII(c,b,i,j)
@@ -318,9 +318,6 @@ xfs_bmbt_delrec(
 	xfs_fsblock_t		bno;		/* fs-relative block number */
 	xfs_buf_t		*bp;		/* buffer for block */
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_delrec";
-#endif
 	int			i;		/* loop counter */
 	int			j;		/* temp state */
 	xfs_bmbt_key_t		key;		/* bmap btree key */
@@ -694,9 +691,6 @@ xfs_bmbt_insrec(
 	xfs_bmbt_block_t	*block;		/* bmap btree block */
 	xfs_buf_t		*bp;		/* buffer for block */
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_insrec";
-#endif
 	int			i;		/* loop index */
 	xfs_bmbt_key_t		key;		/* bmap btree key */
 	xfs_bmbt_key_t		*kp=NULL;	/* pointer to bmap btree key */
@@ -880,9 +874,6 @@ xfs_bmbt_killroot(
 	xfs_bmbt_ptr_t		*cpp;
 #ifdef DEBUG
 	int			error;
-#endif
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_killroot";
 #endif
 	int			i;
 	xfs_bmbt_key_t		*kp;
@@ -973,9 +964,6 @@ xfs_bmbt_log_keys(
 	int		kfirst,
 	int		klast)
 {
-#ifdef XFS_BMBT_TRACE
-	static char	fname[] = "xfs_bmbt_log_keys";
-#endif
 	xfs_trans_t	*tp;
 
 	XFS_BMBT_TRACE_CURSOR(cur, ENTRY);
@@ -1012,9 +1000,6 @@ xfs_bmbt_log_ptrs(
 	int		pfirst,
 	int		plast)
 {
-#ifdef XFS_BMBT_TRACE
-	static char	fname[] = "xfs_bmbt_log_ptrs";
-#endif
 	xfs_trans_t	*tp;
 
 	XFS_BMBT_TRACE_CURSOR(cur, ENTRY);
@@ -1055,9 +1040,6 @@ xfs_bmbt_lookup(
 	xfs_daddr_t		d;
 	xfs_sfiloff_t		diff;
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char	fname[] = "xfs_bmbt_lookup";
-#endif
 	xfs_fsblock_t		fsbno=0;
 	int			high;
 	int			i;
@@ -1195,9 +1177,6 @@ xfs_bmbt_lshift(
 	int			*stat)		/* success/failure */
 {
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_lshift";
-#endif
 #ifdef DEBUG
 	int			i;		/* loop counter */
 #endif
@@ -1331,9 +1310,6 @@ xfs_bmbt_rshift(
 	int			*stat)		/* success/failure */
 {
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_rshift";
-#endif
 	int			i;		/* loop counter */
 	xfs_bmbt_key_t		key;		/* bmap btree key */
 	xfs_buf_t		*lbp;		/* left buffer pointer */
@@ -1492,9 +1468,6 @@ xfs_bmbt_split(
 {
 	xfs_alloc_arg_t		args;		/* block allocation args */
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_split";
-#endif
 	int			i;		/* loop counter */
 	xfs_fsblock_t		lbno;		/* left sibling block number */
 	xfs_buf_t		*lbp;		/* left buffer pointer */
@@ -1640,9 +1613,6 @@ xfs_bmbt_updkey(
 	xfs_buf_t		*bp;
 #ifdef DEBUG
 	int			error;
-#endif
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_updkey";
 #endif
 	xfs_bmbt_key_t		*kp;
 	int			ptr;
@@ -1712,9 +1682,6 @@ xfs_bmbt_decrement(
 	xfs_bmbt_block_t	*block;
 	xfs_buf_t		*bp;
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_decrement";
-#endif
 	xfs_fsblock_t		fsbno;
 	int			lev;
 	xfs_mount_t		*mp;
@@ -1785,9 +1752,6 @@ xfs_bmbt_delete(
 	int		*stat)		/* success/failure */
 {
 	int		error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char	fname[] = "xfs_bmbt_delete";
-#endif
 	int		i;
 	int		level;
 
@@ -2000,9 +1964,6 @@ xfs_bmbt_increment(
 	xfs_bmbt_block_t	*block;
 	xfs_buf_t		*bp;
 	int			error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_increment";
-#endif
 	xfs_fsblock_t		fsbno;
 	int			lev;
 	xfs_mount_t		*mp;
@@ -2080,9 +2041,6 @@ xfs_bmbt_insert(
 	int		*stat)		/* success/failure */
 {
 	int		error;		/* error return value */
-#ifdef XFS_BMBT_TRACE
-	static char	fname[] = "xfs_bmbt_insert";
-#endif
 	int		i;
 	int		level;
 	xfs_fsblock_t	nbno;
@@ -2142,9 +2100,6 @@ xfs_bmbt_log_block(
 	int			fields)
 {
 	int			first;
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_log_block";
-#endif
 	int			last;
 	xfs_trans_t		*tp;
 	static const short	offsets[] = {
@@ -2181,9 +2136,6 @@ xfs_bmbt_log_recs(
 {
 	xfs_bmbt_block_t	*block;
 	int			first;
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_log_recs";
-#endif
 	int			last;
 	xfs_bmbt_rec_t		*rp;
 	xfs_trans_t		*tp;
@@ -2245,9 +2197,6 @@ xfs_bmbt_newroot(
 	xfs_bmbt_key_t		*ckp;		/* child key pointer */
 	xfs_bmbt_ptr_t		*cpp;		/* child ptr pointer */
 	int			error;		/* error return code */
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_newroot";
-#endif
 #ifdef DEBUG
 	int			i;		/* loop counter */
 #endif
@@ -2630,9 +2579,6 @@ xfs_bmbt_update(
 	xfs_bmbt_block_t	*block;
 	xfs_buf_t		*bp;
 	int			error;
-#ifdef XFS_BMBT_TRACE
-	static char		fname[] = "xfs_bmbt_update";
-#endif
 	xfs_bmbt_key_t		key;
 	int			ptr;
 	xfs_bmbt_rec_t		*rp;

fs/xfs/xfs_btree.h

@@ -444,30 +444,14 @@ xfs_btree_setbuf(
 /*
  * Min and max functions for extlen, agblock, fileoff, and filblks types.
  */
-#define	XFS_EXTLEN_MIN(a,b)	\
-	((xfs_extlen_t)(a) < (xfs_extlen_t)(b) ? \
-		(xfs_extlen_t)(a) : (xfs_extlen_t)(b))
-#define	XFS_EXTLEN_MAX(a,b)	\
-	((xfs_extlen_t)(a) > (xfs_extlen_t)(b) ? \
-		(xfs_extlen_t)(a) : (xfs_extlen_t)(b))
-#define	XFS_AGBLOCK_MIN(a,b)	\
-	((xfs_agblock_t)(a) < (xfs_agblock_t)(b) ? \
-		(xfs_agblock_t)(a) : (xfs_agblock_t)(b))
-#define	XFS_AGBLOCK_MAX(a,b)	\
-	((xfs_agblock_t)(a) > (xfs_agblock_t)(b) ? \
-		(xfs_agblock_t)(a) : (xfs_agblock_t)(b))
-#define	XFS_FILEOFF_MIN(a,b)	\
-	((xfs_fileoff_t)(a) < (xfs_fileoff_t)(b) ? \
-		(xfs_fileoff_t)(a) : (xfs_fileoff_t)(b))
-#define	XFS_FILEOFF_MAX(a,b)	\
-	((xfs_fileoff_t)(a) > (xfs_fileoff_t)(b) ? \
-		(xfs_fileoff_t)(a) : (xfs_fileoff_t)(b))
-#define	XFS_FILBLKS_MIN(a,b)	\
-	((xfs_filblks_t)(a) < (xfs_filblks_t)(b) ? \
-		(xfs_filblks_t)(a) : (xfs_filblks_t)(b))
-#define	XFS_FILBLKS_MAX(a,b)	\
-	((xfs_filblks_t)(a) > (xfs_filblks_t)(b) ? \
-		(xfs_filblks_t)(a) : (xfs_filblks_t)(b))
+#define	XFS_EXTLEN_MIN(a,b)	min_t(xfs_extlen_t, (a), (b))
+#define	XFS_EXTLEN_MAX(a,b)	max_t(xfs_extlen_t, (a), (b))
+#define	XFS_AGBLOCK_MIN(a,b)	min_t(xfs_agblock_t, (a), (b))
+#define	XFS_AGBLOCK_MAX(a,b)	max_t(xfs_agblock_t, (a), (b))
+#define	XFS_FILEOFF_MIN(a,b)	min_t(xfs_fileoff_t, (a), (b))
+#define	XFS_FILEOFF_MAX(a,b)	max_t(xfs_fileoff_t, (a), (b))
+#define	XFS_FILBLKS_MIN(a,b)	min_t(xfs_filblks_t, (a), (b))
+#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 && \

fs/xfs/xfs_buf_item.c

@@ -580,8 +580,8 @@ xfs_buf_item_unlock(
 	 * If the buf item isn't tracking any data, free it.
 	 * Otherwise, if XFS_BLI_HOLD is set clear it.
 	 */
-	if (xfs_count_bits(bip->bli_format.blf_data_map,
-			      bip->bli_format.blf_map_size, 0) == 0) {
+	if (xfs_bitmap_empty(bip->bli_format.blf_data_map,
+			     bip->bli_format.blf_map_size)) {
 		xfs_buf_item_relse(bp);
 	} else if (hold) {
 		bip->bli_flags &= ~XFS_BLI_HOLD;

fs/xfs/xfs_clnt.h

@@ -99,5 +99,7 @@ struct xfs_mount_args {
  */
 #define XFSMNT2_COMPAT_IOSIZE	0x00000001	/* don't report large preferred
 						 * I/O size in stat(2) */
+#define XFSMNT2_FILESTREAMS	0x00000002	/* enable the filestreams
+						 * allocator */
 
 #endif	/* __XFS_CLNT_H__ */

fs/xfs/xfs_dinode.h

@@ -257,6 +257,7 @@ typedef enum xfs_dinode_fmt
 #define XFS_DIFLAG_EXTSIZE_BIT      11	/* inode extent size allocator hint */
 #define XFS_DIFLAG_EXTSZINHERIT_BIT 12	/* inherit inode extent size */
 #define XFS_DIFLAG_NODEFRAG_BIT     13	/* do not reorganize/defragment */
+#define XFS_DIFLAG_FILESTREAM_BIT   14  /* use filestream allocator */
 #define XFS_DIFLAG_REALTIME      (1 << XFS_DIFLAG_REALTIME_BIT)
 #define XFS_DIFLAG_PREALLOC      (1 << XFS_DIFLAG_PREALLOC_BIT)
 #define XFS_DIFLAG_NEWRTBM       (1 << XFS_DIFLAG_NEWRTBM_BIT)
@@ -271,12 +272,13 @@ typedef enum xfs_dinode_fmt
 #define XFS_DIFLAG_EXTSIZE       (1 << XFS_DIFLAG_EXTSIZE_BIT)
 #define XFS_DIFLAG_EXTSZINHERIT  (1 << XFS_DIFLAG_EXTSZINHERIT_BIT)
 #define XFS_DIFLAG_NODEFRAG      (1 << XFS_DIFLAG_NODEFRAG_BIT)
+#define XFS_DIFLAG_FILESTREAM    (1 << XFS_DIFLAG_FILESTREAM_BIT)
 
 #define XFS_DIFLAG_ANY \
 	(XFS_DIFLAG_REALTIME | XFS_DIFLAG_PREALLOC | XFS_DIFLAG_NEWRTBM | \
 	 XFS_DIFLAG_IMMUTABLE | XFS_DIFLAG_APPEND | XFS_DIFLAG_SYNC | \
 	 XFS_DIFLAG_NOATIME | XFS_DIFLAG_NODUMP | XFS_DIFLAG_RTINHERIT | \
 	 XFS_DIFLAG_PROJINHERIT | XFS_DIFLAG_NOSYMLINKS | XFS_DIFLAG_EXTSIZE | \
-	 XFS_DIFLAG_EXTSZINHERIT | XFS_DIFLAG_NODEFRAG)
+	 XFS_DIFLAG_EXTSZINHERIT | XFS_DIFLAG_NODEFRAG | XFS_DIFLAG_FILESTREAM)
 
 #endif	/* __XFS_DINODE_H__ */

fs/xfs/xfs_dir2.c

@@ -55,9 +55,9 @@ xfs_dir_mount(
 	       XFS_MAX_BLOCKSIZE);
 	mp->m_dirblksize = 1 << (mp->m_sb.sb_blocklog + mp->m_sb.sb_dirblklog);
 	mp->m_dirblkfsbs = 1 << mp->m_sb.sb_dirblklog;
-	mp->m_dirdatablk = XFS_DIR2_DB_TO_DA(mp, XFS_DIR2_DATA_FIRSTDB(mp));
-	mp->m_dirleafblk = XFS_DIR2_DB_TO_DA(mp, XFS_DIR2_LEAF_FIRSTDB(mp));
-	mp->m_dirfreeblk = XFS_DIR2_DB_TO_DA(mp, XFS_DIR2_FREE_FIRSTDB(mp));
+	mp->m_dirdatablk = xfs_dir2_db_to_da(mp, XFS_DIR2_DATA_FIRSTDB(mp));
+	mp->m_dirleafblk = xfs_dir2_db_to_da(mp, XFS_DIR2_LEAF_FIRSTDB(mp));
+	mp->m_dirfreeblk = xfs_dir2_db_to_da(mp, XFS_DIR2_FREE_FIRSTDB(mp));
 	mp->m_attr_node_ents =
 		(mp->m_sb.sb_blocksize - (uint)sizeof(xfs_da_node_hdr_t)) /
 		(uint)sizeof(xfs_da_node_entry_t);
@@ -554,7 +554,7 @@ xfs_dir2_grow_inode(
 	 */
 	if (mapp != &map)
 		kmem_free(mapp, sizeof(*mapp) * count);
-	*dbp = XFS_DIR2_DA_TO_DB(mp, (xfs_dablk_t)bno);
+	*dbp = xfs_dir2_da_to_db(mp, (xfs_dablk_t)bno);
 	/*
 	 * Update file's size if this is the data space and it grew.
 	 */
@@ -706,7 +706,7 @@ xfs_dir2_shrink_inode(
 	dp = args->dp;
 	mp = dp->i_mount;
 	tp = args->trans;
-	da = XFS_DIR2_DB_TO_DA(mp, db);
+	da = xfs_dir2_db_to_da(mp, db);
 	/*
 	 * Unmap the fsblock(s).
 	 */
@@ -742,7 +742,7 @@ xfs_dir2_shrink_inode(
 	/*
 	 * If the block isn't the last one in the directory, we're done.
 	 */
-	if (dp->i_d.di_size > XFS_DIR2_DB_OFF_TO_BYTE(mp, db + 1, 0))
+	if (dp->i_d.di_size > xfs_dir2_db_off_to_byte(mp, db + 1, 0))
 		return 0;
 	bno = da;
 	if ((error = xfs_bmap_last_before(tp, dp, &bno, XFS_DATA_FORK))) {

fs/xfs/xfs_dir2_block.c

@@ -115,13 +115,13 @@ xfs_dir2_block_addname(
 		xfs_da_brelse(tp, bp);
 		return XFS_ERROR(EFSCORRUPTED);
 	}
-	len = XFS_DIR2_DATA_ENTSIZE(args->namelen);
+	len = xfs_dir2_data_entsize(args->namelen);
 	/*
 	 * Set up pointers to parts of the block.
 	 */
 	bf = block->hdr.bestfree;
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * No stale entries?  Need space for entry and new leaf.
 	 */
@@ -396,7 +396,7 @@ xfs_dir2_block_addname(
 	 * Fill in the leaf entry.
 	 */
 	blp[mid].hashval = cpu_to_be32(args->hashval);
-	blp[mid].address = cpu_to_be32(XFS_DIR2_BYTE_TO_DATAPTR(mp,
+	blp[mid].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
 				(char *)dep - (char *)block));
 	xfs_dir2_block_log_leaf(tp, bp, lfloglow, lfloghigh);
 	/*
@@ -411,7 +411,7 @@ xfs_dir2_block_addname(
 	dep->inumber = cpu_to_be64(args->inumber);
 	dep->namelen = args->namelen;
 	memcpy(dep->name, args->name, args->namelen);
-	tagp = XFS_DIR2_DATA_ENTRY_TAG_P(dep);
+	tagp = xfs_dir2_data_entry_tag_p(dep);
 	*tagp = cpu_to_be16((char *)dep - (char *)block);
 	/*
 	 * Clean up the bestfree array and log the header, tail, and entry.
@@ -455,7 +455,7 @@ xfs_dir2_block_getdents(
 	/*
 	 * If the block number in the offset is out of range, we're done.
 	 */
-	if (XFS_DIR2_DATAPTR_TO_DB(mp, uio->uio_offset) > mp->m_dirdatablk) {
+	if (xfs_dir2_dataptr_to_db(mp, uio->uio_offset) > mp->m_dirdatablk) {
 		*eofp = 1;
 		return 0;
 	}
@@ -471,15 +471,15 @@ xfs_dir2_block_getdents(
 	 * Extract the byte offset we start at from the seek pointer.
 	 * We'll skip entries before this.
 	 */
-	wantoff = XFS_DIR2_DATAPTR_TO_OFF(mp, uio->uio_offset);
+	wantoff = xfs_dir2_dataptr_to_off(mp, uio->uio_offset);
 	block = bp->data;
 	xfs_dir2_data_check(dp, bp);
 	/*
 	 * Set up values for the loop.
 	 */
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, block);
 	ptr = (char *)block->u;
-	endptr = (char *)XFS_DIR2_BLOCK_LEAF_P(btp);
+	endptr = (char *)xfs_dir2_block_leaf_p(btp);
 	p.dbp = dbp;
 	p.put = put;
 	p.uio = uio;
@@ -502,7 +502,7 @@ xfs_dir2_block_getdents(
 		/*
 		 * Bump pointer for the next iteration.
 		 */
-		ptr += XFS_DIR2_DATA_ENTSIZE(dep->namelen);
+		ptr += xfs_dir2_data_entsize(dep->namelen);
 		/*
 		 * The entry is before the desired starting point, skip it.
 		 */
@@ -513,7 +513,7 @@ xfs_dir2_block_getdents(
 		 */
 		p.namelen = dep->namelen;
 
-		p.cook = XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+		p.cook = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 						    ptr - (char *)block);
 		p.ino = be64_to_cpu(dep->inumber);
 #if XFS_BIG_INUMS
@@ -531,7 +531,7 @@ xfs_dir2_block_getdents(
 		 */
 		if (!p.done) {
 			uio->uio_offset =
-				XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+				xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 					(char *)dep - (char *)block);
 			xfs_da_brelse(tp, bp);
 			return error;
@@ -545,7 +545,7 @@ xfs_dir2_block_getdents(
 	*eofp = 1;
 
 	uio->uio_offset =
-		XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk + 1, 0);
+		xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk + 1, 0);
 
 	xfs_da_brelse(tp, bp);
 
@@ -569,8 +569,8 @@ xfs_dir2_block_log_leaf(
 
 	mp = tp->t_mountp;
 	block = bp->data;
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	xfs_da_log_buf(tp, bp, (uint)((char *)&blp[first] - (char *)block),
 		(uint)((char *)&blp[last + 1] - (char *)block - 1));
 }
@@ -589,7 +589,7 @@ xfs_dir2_block_log_tail(
 
 	mp = tp->t_mountp;
 	block = bp->data;
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, block);
 	xfs_da_log_buf(tp, bp, (uint)((char *)btp - (char *)block),
 		(uint)((char *)(btp + 1) - (char *)block - 1));
 }
@@ -623,13 +623,13 @@ xfs_dir2_block_lookup(
 	mp = dp->i_mount;
 	block = bp->data;
 	xfs_dir2_data_check(dp, bp);
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Get the offset from the leaf entry, to point to the data.
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)block + XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(blp[ent].address)));
+	      ((char *)block + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
 	/*
 	 * Fill in inode number, release the block.
 	 */
@@ -675,8 +675,8 @@ xfs_dir2_block_lookup_int(
 	ASSERT(bp != NULL);
 	block = bp->data;
 	xfs_dir2_data_check(dp, bp);
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Loop doing a binary search for our hash value.
 	 * Find our entry, ENOENT if it's not there.
@@ -713,7 +713,7 @@ xfs_dir2_block_lookup_int(
 		 * Get pointer to the entry from the leaf.
 		 */
 		dep = (xfs_dir2_data_entry_t *)
-			((char *)block + XFS_DIR2_DATAPTR_TO_OFF(mp, addr));
+			((char *)block + xfs_dir2_dataptr_to_off(mp, addr));
 		/*
 		 * Compare, if it's right give back buffer & entry number.
 		 */
@@ -768,20 +768,20 @@ xfs_dir2_block_removename(
 	tp = args->trans;
 	mp = dp->i_mount;
 	block = bp->data;
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Point to the data entry using the leaf entry.
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)block + XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(blp[ent].address)));
+	      ((char *)block + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
 	/*
 	 * Mark the data entry's space free.
 	 */
 	needlog = needscan = 0;
 	xfs_dir2_data_make_free(tp, bp,
 		(xfs_dir2_data_aoff_t)((char *)dep - (char *)block),
-		XFS_DIR2_DATA_ENTSIZE(dep->namelen), &needlog, &needscan);
+		xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
 	/*
 	 * Fix up the block tail.
 	 */
@@ -843,13 +843,13 @@ xfs_dir2_block_replace(
 	dp = args->dp;
 	mp = dp->i_mount;
 	block = bp->data;
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Point to the data entry we need to change.
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)block + XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(blp[ent].address)));
+	      ((char *)block + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
 	ASSERT(be64_to_cpu(dep->inumber) != args->inumber);
 	/*
 	 * Change the inode number to the new value.
@@ -912,7 +912,7 @@ xfs_dir2_leaf_to_block(
 	mp = dp->i_mount;
 	leaf = lbp->data;
 	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	/*
 	 * If there are data blocks other than the first one, take this
 	 * opportunity to remove trailing empty data blocks that may have
@@ -920,7 +920,7 @@ xfs_dir2_leaf_to_block(
 	 * These will show up in the leaf bests table.
 	 */
 	while (dp->i_d.di_size > mp->m_dirblksize) {
-		bestsp = XFS_DIR2_LEAF_BESTS_P(ltp);
+		bestsp = xfs_dir2_leaf_bests_p(ltp);
 		if (be16_to_cpu(bestsp[be32_to_cpu(ltp->bestcount) - 1]) ==
 		    mp->m_dirblksize - (uint)sizeof(block->hdr)) {
 			if ((error =
@@ -974,14 +974,14 @@ xfs_dir2_leaf_to_block(
 	/*
 	 * Initialize the block tail.
 	 */
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, block);
 	btp->count = cpu_to_be32(be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale));
 	btp->stale = 0;
 	xfs_dir2_block_log_tail(tp, dbp);
 	/*
 	 * Initialize the block leaf area.  We compact out stale entries.
 	 */
-	lep = XFS_DIR2_BLOCK_LEAF_P(btp);
+	lep = xfs_dir2_block_leaf_p(btp);
 	for (from = to = 0; from < be16_to_cpu(leaf->hdr.count); from++) {
 		if (be32_to_cpu(leaf->ents[from].address) == XFS_DIR2_NULL_DATAPTR)
 			continue;
@@ -1067,7 +1067,7 @@ xfs_dir2_sf_to_block(
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
 	/*
 	 * Copy the directory into the stack buffer.
 	 * Then pitch the incore inode data so we can make extents.
@@ -1119,10 +1119,10 @@ xfs_dir2_sf_to_block(
 	/*
 	 * Fill in the tail.
 	 */
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, block);
 	btp->count = cpu_to_be32(sfp->hdr.count + 2);	/* ., .. */
 	btp->stale = 0;
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	blp = xfs_dir2_block_leaf_p(btp);
 	endoffset = (uint)((char *)blp - (char *)block);
 	/*
 	 * Remove the freespace, we'll manage it.
@@ -1138,25 +1138,25 @@ xfs_dir2_sf_to_block(
 	dep->inumber = cpu_to_be64(dp->i_ino);
 	dep->namelen = 1;
 	dep->name[0] = '.';
-	tagp = XFS_DIR2_DATA_ENTRY_TAG_P(dep);
+	tagp = xfs_dir2_data_entry_tag_p(dep);
 	*tagp = cpu_to_be16((char *)dep - (char *)block);
 	xfs_dir2_data_log_entry(tp, bp, dep);
 	blp[0].hashval = cpu_to_be32(xfs_dir_hash_dot);
-	blp[0].address = cpu_to_be32(XFS_DIR2_BYTE_TO_DATAPTR(mp,
+	blp[0].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
 				(char *)dep - (char *)block));
 	/*
 	 * Create entry for ..
 	 */
 	dep = (xfs_dir2_data_entry_t *)
 		((char *)block + XFS_DIR2_DATA_DOTDOT_OFFSET);
-	dep->inumber = cpu_to_be64(XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent));
+	dep->inumber = cpu_to_be64(xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent));
 	dep->namelen = 2;
 	dep->name[0] = dep->name[1] = '.';
-	tagp = XFS_DIR2_DATA_ENTRY_TAG_P(dep);
+	tagp = xfs_dir2_data_entry_tag_p(dep);
 	*tagp = cpu_to_be16((char *)dep - (char *)block);
 	xfs_dir2_data_log_entry(tp, bp, dep);
 	blp[1].hashval = cpu_to_be32(xfs_dir_hash_dotdot);
-	blp[1].address = cpu_to_be32(XFS_DIR2_BYTE_TO_DATAPTR(mp,
+	blp[1].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
 				(char *)dep - (char *)block));
 	offset = XFS_DIR2_DATA_FIRST_OFFSET;
 	/*
@@ -1165,7 +1165,7 @@ xfs_dir2_sf_to_block(
 	if ((i = 0) == sfp->hdr.count)
 		sfep = NULL;
 	else
-		sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+		sfep = xfs_dir2_sf_firstentry(sfp);
 	/*
 	 * Need to preserve the existing offset values in the sf directory.
 	 * Insert holes (unused entries) where necessary.
@@ -1177,7 +1177,7 @@ xfs_dir2_sf_to_block(
 		if (sfep == NULL)
 			newoffset = endoffset;
 		else
-			newoffset = XFS_DIR2_SF_GET_OFFSET(sfep);
+			newoffset = xfs_dir2_sf_get_offset(sfep);
 		/*
 		 * There should be a hole here, make one.
 		 */
@@ -1186,7 +1186,7 @@ xfs_dir2_sf_to_block(
 			      ((char *)block + offset);
 			dup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 			dup->length = cpu_to_be16(newoffset - offset);
-			*XFS_DIR2_DATA_UNUSED_TAG_P(dup) = cpu_to_be16(
+			*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16(
 				((char *)dup - (char *)block));
 			xfs_dir2_data_log_unused(tp, bp, dup);
 			(void)xfs_dir2_data_freeinsert((xfs_dir2_data_t *)block,
@@ -1198,22 +1198,22 @@ xfs_dir2_sf_to_block(
 		 * Copy a real entry.
 		 */
 		dep = (xfs_dir2_data_entry_t *)((char *)block + newoffset);
-		dep->inumber = cpu_to_be64(XFS_DIR2_SF_GET_INUMBER(sfp,
-				XFS_DIR2_SF_INUMBERP(sfep)));
+		dep->inumber = cpu_to_be64(xfs_dir2_sf_get_inumber(sfp,
+				xfs_dir2_sf_inumberp(sfep)));
 		dep->namelen = sfep->namelen;
 		memcpy(dep->name, sfep->name, dep->namelen);
-		tagp = XFS_DIR2_DATA_ENTRY_TAG_P(dep);
+		tagp = xfs_dir2_data_entry_tag_p(dep);
 		*tagp = cpu_to_be16((char *)dep - (char *)block);
 		xfs_dir2_data_log_entry(tp, bp, dep);
 		blp[2 + i].hashval = cpu_to_be32(xfs_da_hashname(
 					(char *)sfep->name, sfep->namelen));
-		blp[2 + i].address = cpu_to_be32(XFS_DIR2_BYTE_TO_DATAPTR(mp,
+		blp[2 + i].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
 						 (char *)dep - (char *)block));
 		offset = (int)((char *)(tagp + 1) - (char *)block);
 		if (++i == sfp->hdr.count)
 			sfep = NULL;
 		else
-			sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep);
+			sfep = xfs_dir2_sf_nextentry(sfp, sfep);
 	}
 	/* Done with the temporary buffer */
 	kmem_free(buf, buf_len);

fs/xfs/xfs_dir2_block.h

@@ -60,7 +60,6 @@ typedef struct xfs_dir2_block {
 /*
  * Pointer to the leaf header embedded in a data block (1-block format)
  */
-#define	XFS_DIR2_BLOCK_TAIL_P(mp,block)	xfs_dir2_block_tail_p(mp,block)
 static inline xfs_dir2_block_tail_t *
 xfs_dir2_block_tail_p(struct xfs_mount *mp, xfs_dir2_block_t *block)
 {
@@ -71,7 +70,6 @@ xfs_dir2_block_tail_p(struct xfs_mount *mp, xfs_dir2_block_t *block)
 /*
  * Pointer to the leaf entries embedded in a data block (1-block format)
  */
-#define	XFS_DIR2_BLOCK_LEAF_P(btp)	xfs_dir2_block_leaf_p(btp)
 static inline struct xfs_dir2_leaf_entry *
 xfs_dir2_block_leaf_p(xfs_dir2_block_tail_t *btp)
 {

fs/xfs/xfs_dir2_data.c

@@ -72,8 +72,8 @@ xfs_dir2_data_check(
 	bf = d->hdr.bestfree;
 	p = (char *)d->u;
 	if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
-		btp = XFS_DIR2_BLOCK_TAIL_P(mp, (xfs_dir2_block_t *)d);
-		lep = XFS_DIR2_BLOCK_LEAF_P(btp);
+		btp = xfs_dir2_block_tail_p(mp, (xfs_dir2_block_t *)d);
+		lep = xfs_dir2_block_leaf_p(btp);
 		endp = (char *)lep;
 	} else
 		endp = (char *)d + mp->m_dirblksize;
@@ -107,7 +107,7 @@ xfs_dir2_data_check(
 		 */
 		if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
 			ASSERT(lastfree == 0);
-			ASSERT(be16_to_cpu(*XFS_DIR2_DATA_UNUSED_TAG_P(dup)) ==
+			ASSERT(be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)) ==
 			       (char *)dup - (char *)d);
 			dfp = xfs_dir2_data_freefind(d, dup);
 			if (dfp) {
@@ -131,12 +131,12 @@ xfs_dir2_data_check(
 		dep = (xfs_dir2_data_entry_t *)p;
 		ASSERT(dep->namelen != 0);
 		ASSERT(xfs_dir_ino_validate(mp, be64_to_cpu(dep->inumber)) == 0);
-		ASSERT(be16_to_cpu(*XFS_DIR2_DATA_ENTRY_TAG_P(dep)) ==
+		ASSERT(be16_to_cpu(*xfs_dir2_data_entry_tag_p(dep)) ==
 		       (char *)dep - (char *)d);
 		count++;
 		lastfree = 0;
 		if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
-			addr = XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+			addr = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 				(xfs_dir2_data_aoff_t)
 				((char *)dep - (char *)d));
 			hash = xfs_da_hashname((char *)dep->name, dep->namelen);
@@ -147,7 +147,7 @@ xfs_dir2_data_check(
 			}
 			ASSERT(i < be32_to_cpu(btp->count));
 		}
-		p += XFS_DIR2_DATA_ENTSIZE(dep->namelen);
+		p += xfs_dir2_data_entsize(dep->namelen);
 	}
 	/*
 	 * Need to have seen all the entries and all the bestfree slots.
@@ -346,8 +346,8 @@ xfs_dir2_data_freescan(
 	 */
 	p = (char *)d->u;
 	if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
-		btp = XFS_DIR2_BLOCK_TAIL_P(mp, (xfs_dir2_block_t *)d);
-		endp = (char *)XFS_DIR2_BLOCK_LEAF_P(btp);
+		btp = xfs_dir2_block_tail_p(mp, (xfs_dir2_block_t *)d);
+		endp = (char *)xfs_dir2_block_leaf_p(btp);
 	} else
 		endp = (char *)d + mp->m_dirblksize;
 	/*
@@ -360,7 +360,7 @@ xfs_dir2_data_freescan(
 		 */
 		if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
 			ASSERT((char *)dup - (char *)d ==
-			       be16_to_cpu(*XFS_DIR2_DATA_UNUSED_TAG_P(dup)));
+			       be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)));
 			xfs_dir2_data_freeinsert(d, dup, loghead);
 			p += be16_to_cpu(dup->length);
 		}
@@ -370,8 +370,8 @@ xfs_dir2_data_freescan(
 		else {
 			dep = (xfs_dir2_data_entry_t *)p;
 			ASSERT((char *)dep - (char *)d ==
-			       be16_to_cpu(*XFS_DIR2_DATA_ENTRY_TAG_P(dep)));
-			p += XFS_DIR2_DATA_ENTSIZE(dep->namelen);
+			       be16_to_cpu(*xfs_dir2_data_entry_tag_p(dep)));
+			p += xfs_dir2_data_entsize(dep->namelen);
 		}
 	}
 }
@@ -402,7 +402,7 @@ xfs_dir2_data_init(
 	/*
 	 * Get the buffer set up for the block.
 	 */
-	error = xfs_da_get_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, blkno), -1, &bp,
+	error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, blkno), -1, &bp,
 		XFS_DATA_FORK);
 	if (error) {
 		return error;
@@ -427,7 +427,7 @@ xfs_dir2_data_init(
 	t=mp->m_dirblksize - (uint)sizeof(d->hdr);
 	d->hdr.bestfree[0].length = cpu_to_be16(t);
 	dup->length = cpu_to_be16(t);
-	*XFS_DIR2_DATA_UNUSED_TAG_P(dup) = cpu_to_be16((char *)dup - (char *)d);
+	*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16((char *)dup - (char *)d);
 	/*
 	 * Log it and return it.
 	 */
@@ -452,7 +452,7 @@ xfs_dir2_data_log_entry(
 	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
 	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
 	xfs_da_log_buf(tp, bp, (uint)((char *)dep - (char *)d),
-		(uint)((char *)(XFS_DIR2_DATA_ENTRY_TAG_P(dep) + 1) -
+		(uint)((char *)(xfs_dir2_data_entry_tag_p(dep) + 1) -
 		       (char *)d - 1));
 }
 
@@ -497,8 +497,8 @@ xfs_dir2_data_log_unused(
 	 * Log the end (tag) of the unused entry.
 	 */
 	xfs_da_log_buf(tp, bp,
-		(uint)((char *)XFS_DIR2_DATA_UNUSED_TAG_P(dup) - (char *)d),
-		(uint)((char *)XFS_DIR2_DATA_UNUSED_TAG_P(dup) - (char *)d +
+		(uint)((char *)xfs_dir2_data_unused_tag_p(dup) - (char *)d),
+		(uint)((char *)xfs_dir2_data_unused_tag_p(dup) - (char *)d +
 		       sizeof(xfs_dir2_data_off_t) - 1));
 }
 
@@ -535,8 +535,8 @@ xfs_dir2_data_make_free(
 		xfs_dir2_block_tail_t	*btp;	/* block tail */
 
 		ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
-		btp = XFS_DIR2_BLOCK_TAIL_P(mp, (xfs_dir2_block_t *)d);
-		endptr = (char *)XFS_DIR2_BLOCK_LEAF_P(btp);
+		btp = xfs_dir2_block_tail_p(mp, (xfs_dir2_block_t *)d);
+		endptr = (char *)xfs_dir2_block_leaf_p(btp);
 	}
 	/*
 	 * If this isn't the start of the block, then back up to
@@ -587,7 +587,7 @@ xfs_dir2_data_make_free(
 		 * Fix up the new big freespace.
 		 */
 		be16_add(&prevdup->length, len + be16_to_cpu(postdup->length));
-		*XFS_DIR2_DATA_UNUSED_TAG_P(prevdup) =
+		*xfs_dir2_data_unused_tag_p(prevdup) =
 			cpu_to_be16((char *)prevdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, prevdup);
 		if (!needscan) {
@@ -621,7 +621,7 @@ xfs_dir2_data_make_free(
 	else if (prevdup) {
 		dfp = xfs_dir2_data_freefind(d, prevdup);
 		be16_add(&prevdup->length, len);
-		*XFS_DIR2_DATA_UNUSED_TAG_P(prevdup) =
+		*xfs_dir2_data_unused_tag_p(prevdup) =
 			cpu_to_be16((char *)prevdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, prevdup);
 		/*
@@ -649,7 +649,7 @@ xfs_dir2_data_make_free(
 		newdup = (xfs_dir2_data_unused_t *)((char *)d + offset);
 		newdup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup->length = cpu_to_be16(len + be16_to_cpu(postdup->length));
-		*XFS_DIR2_DATA_UNUSED_TAG_P(newdup) =
+		*xfs_dir2_data_unused_tag_p(newdup) =
 			cpu_to_be16((char *)newdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		/*
@@ -676,7 +676,7 @@ xfs_dir2_data_make_free(
 		newdup = (xfs_dir2_data_unused_t *)((char *)d + offset);
 		newdup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup->length = cpu_to_be16(len);
-		*XFS_DIR2_DATA_UNUSED_TAG_P(newdup) =
+		*xfs_dir2_data_unused_tag_p(newdup) =
 			cpu_to_be16((char *)newdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		(void)xfs_dir2_data_freeinsert(d, newdup, needlogp);
@@ -712,7 +712,7 @@ xfs_dir2_data_use_free(
 	ASSERT(be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG);
 	ASSERT(offset >= (char *)dup - (char *)d);
 	ASSERT(offset + len <= (char *)dup + be16_to_cpu(dup->length) - (char *)d);
-	ASSERT((char *)dup - (char *)d == be16_to_cpu(*XFS_DIR2_DATA_UNUSED_TAG_P(dup)));
+	ASSERT((char *)dup - (char *)d == be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)));
 	/*
 	 * Look up the entry in the bestfree table.
 	 */
@@ -745,7 +745,7 @@ xfs_dir2_data_use_free(
 		newdup = (xfs_dir2_data_unused_t *)((char *)d + offset + len);
 		newdup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup->length = cpu_to_be16(oldlen - len);
-		*XFS_DIR2_DATA_UNUSED_TAG_P(newdup) =
+		*xfs_dir2_data_unused_tag_p(newdup) =
 			cpu_to_be16((char *)newdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		/*
@@ -772,7 +772,7 @@ xfs_dir2_data_use_free(
 	else if (matchback) {
 		newdup = dup;
 		newdup->length = cpu_to_be16(((char *)d + offset) - (char *)newdup);
-		*XFS_DIR2_DATA_UNUSED_TAG_P(newdup) =
+		*xfs_dir2_data_unused_tag_p(newdup) =
 			cpu_to_be16((char *)newdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		/*
@@ -799,13 +799,13 @@ xfs_dir2_data_use_free(
 	else {
 		newdup = dup;
 		newdup->length = cpu_to_be16(((char *)d + offset) - (char *)newdup);
-		*XFS_DIR2_DATA_UNUSED_TAG_P(newdup) =
+		*xfs_dir2_data_unused_tag_p(newdup) =
 			cpu_to_be16((char *)newdup - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		newdup2 = (xfs_dir2_data_unused_t *)((char *)d + offset + len);
 		newdup2->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup2->length = cpu_to_be16(oldlen - len - be16_to_cpu(newdup->length));
-		*XFS_DIR2_DATA_UNUSED_TAG_P(newdup2) =
+		*xfs_dir2_data_unused_tag_p(newdup2) =
 			cpu_to_be16((char *)newdup2 - (char *)d);
 		xfs_dir2_data_log_unused(tp, bp, newdup2);
 		/*

fs/xfs/xfs_dir2_data.h

@@ -44,7 +44,7 @@ struct xfs_trans;
 #define	XFS_DIR2_DATA_SPACE	0
 #define	XFS_DIR2_DATA_OFFSET	(XFS_DIR2_DATA_SPACE * XFS_DIR2_SPACE_SIZE)
 #define	XFS_DIR2_DATA_FIRSTDB(mp)	\
-	XFS_DIR2_BYTE_TO_DB(mp, XFS_DIR2_DATA_OFFSET)
+	xfs_dir2_byte_to_db(mp, XFS_DIR2_DATA_OFFSET)
 
 /*
  * Offsets of . and .. in data space (always block 0)
@@ -52,9 +52,9 @@ struct xfs_trans;
 #define	XFS_DIR2_DATA_DOT_OFFSET	\
 	((xfs_dir2_data_aoff_t)sizeof(xfs_dir2_data_hdr_t))
 #define	XFS_DIR2_DATA_DOTDOT_OFFSET	\
-	(XFS_DIR2_DATA_DOT_OFFSET + XFS_DIR2_DATA_ENTSIZE(1))
+	(XFS_DIR2_DATA_DOT_OFFSET + xfs_dir2_data_entsize(1))
 #define	XFS_DIR2_DATA_FIRST_OFFSET		\
-	(XFS_DIR2_DATA_DOTDOT_OFFSET + XFS_DIR2_DATA_ENTSIZE(2))
+	(XFS_DIR2_DATA_DOTDOT_OFFSET + xfs_dir2_data_entsize(2))
 
 /*
  * Structures.
@@ -123,7 +123,6 @@ typedef struct xfs_dir2_data {
 /*
  * Size of a data entry.
  */
-#define XFS_DIR2_DATA_ENTSIZE(n)	xfs_dir2_data_entsize(n)
 static inline int xfs_dir2_data_entsize(int n)
 {
 	return (int)roundup(offsetof(xfs_dir2_data_entry_t, name[0]) + (n) + \
@@ -133,19 +132,16 @@ static inline int xfs_dir2_data_entsize(int n)
 /*
  * Pointer to an entry's tag word.
  */
-#define	XFS_DIR2_DATA_ENTRY_TAG_P(dep)	xfs_dir2_data_entry_tag_p(dep)
 static inline __be16 *
 xfs_dir2_data_entry_tag_p(xfs_dir2_data_entry_t *dep)
 {
 	return (__be16 *)((char *)dep +
-		XFS_DIR2_DATA_ENTSIZE(dep->namelen) - sizeof(__be16));
+		xfs_dir2_data_entsize(dep->namelen) - sizeof(__be16));
 }
 
 /*
  * Pointer to a freespace's tag word.
  */
-#define	XFS_DIR2_DATA_UNUSED_TAG_P(dup) \
-	xfs_dir2_data_unused_tag_p(dup)
 static inline __be16 *
 xfs_dir2_data_unused_tag_p(xfs_dir2_data_unused_t *dup)
 {

fs/xfs/xfs_dir2_leaf.c

@@ -92,7 +92,7 @@ xfs_dir2_block_to_leaf(
 	if ((error = xfs_da_grow_inode(args, &blkno))) {
 		return error;
 	}
-	ldb = XFS_DIR2_DA_TO_DB(mp, blkno);
+	ldb = xfs_dir2_da_to_db(mp, blkno);
 	ASSERT(ldb == XFS_DIR2_LEAF_FIRSTDB(mp));
 	/*
 	 * Initialize the leaf block, get a buffer for it.
@@ -104,8 +104,8 @@ xfs_dir2_block_to_leaf(
 	leaf = lbp->data;
 	block = dbp->data;
 	xfs_dir2_data_check(dp, dbp);
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Set the counts in the leaf header.
 	 */
@@ -137,9 +137,9 @@ xfs_dir2_block_to_leaf(
 	/*
 	 * Set up leaf tail and bests table.
 	 */
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	ltp->bestcount = cpu_to_be32(1);
-	bestsp = XFS_DIR2_LEAF_BESTS_P(ltp);
+	bestsp = xfs_dir2_leaf_bests_p(ltp);
 	bestsp[0] =  block->hdr.bestfree[0].length;
 	/*
 	 * Log the data header and leaf bests table.
@@ -209,9 +209,9 @@ xfs_dir2_leaf_addname(
 	 */
 	index = xfs_dir2_leaf_search_hash(args, lbp);
 	leaf = lbp->data;
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
-	bestsp = XFS_DIR2_LEAF_BESTS_P(ltp);
-	length = XFS_DIR2_DATA_ENTSIZE(args->namelen);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
+	bestsp = xfs_dir2_leaf_bests_p(ltp);
+	length = xfs_dir2_data_entsize(args->namelen);
 	/*
 	 * See if there are any entries with the same hash value
 	 * and space in their block for the new entry.
@@ -223,7 +223,7 @@ xfs_dir2_leaf_addname(
 	     index++, lep++) {
 		if (be32_to_cpu(lep->address) == XFS_DIR2_NULL_DATAPTR)
 			continue;
-		i = XFS_DIR2_DATAPTR_TO_DB(mp, be32_to_cpu(lep->address));
+		i = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 		ASSERT(i < be32_to_cpu(ltp->bestcount));
 		ASSERT(be16_to_cpu(bestsp[i]) != NULLDATAOFF);
 		if (be16_to_cpu(bestsp[i]) >= length) {
@@ -378,7 +378,7 @@ xfs_dir2_leaf_addname(
 	 */
 	else {
 		if ((error =
-		    xfs_da_read_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, use_block),
+		    xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, use_block),
 			    -1, &dbp, XFS_DATA_FORK))) {
 			xfs_da_brelse(tp, lbp);
 			return error;
@@ -407,7 +407,7 @@ xfs_dir2_leaf_addname(
 	dep->inumber = cpu_to_be64(args->inumber);
 	dep->namelen = args->namelen;
 	memcpy(dep->name, args->name, dep->namelen);
-	tagp = XFS_DIR2_DATA_ENTRY_TAG_P(dep);
+	tagp = xfs_dir2_data_entry_tag_p(dep);
 	*tagp = cpu_to_be16((char *)dep - (char *)data);
 	/*
 	 * Need to scan fix up the bestfree table.
@@ -529,7 +529,7 @@ xfs_dir2_leaf_addname(
 	 * Fill in the new leaf entry.
 	 */
 	lep->hashval = cpu_to_be32(args->hashval);
-	lep->address = cpu_to_be32(XFS_DIR2_DB_OFF_TO_DATAPTR(mp, use_block,
+	lep->address = cpu_to_be32(xfs_dir2_db_off_to_dataptr(mp, use_block,
 				be16_to_cpu(*tagp)));
 	/*
 	 * Log the leaf fields and give up the buffers.
@@ -567,13 +567,13 @@ xfs_dir2_leaf_check(
 	 * Should factor in the size of the bests table as well.
 	 * We can deduce a value for that from di_size.
 	 */
-	ASSERT(be16_to_cpu(leaf->hdr.count) <= XFS_DIR2_MAX_LEAF_ENTS(mp));
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ASSERT(be16_to_cpu(leaf->hdr.count) <= xfs_dir2_max_leaf_ents(mp));
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	/*
 	 * Leaves and bests don't overlap.
 	 */
 	ASSERT((char *)&leaf->ents[be16_to_cpu(leaf->hdr.count)] <=
-	       (char *)XFS_DIR2_LEAF_BESTS_P(ltp));
+	       (char *)xfs_dir2_leaf_bests_p(ltp));
 	/*
 	 * Check hash value order, count stale entries.
 	 */
@@ -815,12 +815,12 @@ xfs_dir2_leaf_getdents(
 	 * Inside the loop we keep the main offset value as a byte offset
 	 * in the directory file.
 	 */
-	curoff = XFS_DIR2_DATAPTR_TO_BYTE(mp, uio->uio_offset);
+	curoff = xfs_dir2_dataptr_to_byte(mp, uio->uio_offset);
 	/*
 	 * Force this conversion through db so we truncate the offset
 	 * down to get the start of the data block.
 	 */
-	map_off = XFS_DIR2_DB_TO_DA(mp, XFS_DIR2_BYTE_TO_DB(mp, curoff));
+	map_off = xfs_dir2_db_to_da(mp, xfs_dir2_byte_to_db(mp, curoff));
 	/*
 	 * Loop over directory entries until we reach the end offset.
 	 * Get more blocks and readahead as necessary.
@@ -870,7 +870,7 @@ xfs_dir2_leaf_getdents(
 			 */
 			if (1 + ra_want > map_blocks &&
 			    map_off <
-			    XFS_DIR2_BYTE_TO_DA(mp, XFS_DIR2_LEAF_OFFSET)) {
+			    xfs_dir2_byte_to_da(mp, XFS_DIR2_LEAF_OFFSET)) {
 				/*
 				 * Get more bmaps, fill in after the ones
 				 * we already have in the table.
@@ -878,7 +878,7 @@ xfs_dir2_leaf_getdents(
 				nmap = map_size - map_valid;
 				error = xfs_bmapi(tp, dp,
 					map_off,
-					XFS_DIR2_BYTE_TO_DA(mp,
+					xfs_dir2_byte_to_da(mp,
 						XFS_DIR2_LEAF_OFFSET) - map_off,
 					XFS_BMAPI_METADATA, NULL, 0,
 					&map[map_valid], &nmap, NULL, NULL);
@@ -903,7 +903,7 @@ xfs_dir2_leaf_getdents(
 					map[map_valid + nmap - 1].br_blockcount;
 				else
 					map_off =
-						XFS_DIR2_BYTE_TO_DA(mp,
+						xfs_dir2_byte_to_da(mp,
 							XFS_DIR2_LEAF_OFFSET);
 				/*
 				 * Look for holes in the mapping, and
@@ -931,14 +931,14 @@ xfs_dir2_leaf_getdents(
 			 * No valid mappings, so no more data blocks.
 			 */
 			if (!map_valid) {
-				curoff = XFS_DIR2_DA_TO_BYTE(mp, map_off);
+				curoff = xfs_dir2_da_to_byte(mp, map_off);
 				break;
 			}
 			/*
 			 * Read the directory block starting at the first
 			 * mapping.
 			 */
-			curdb = XFS_DIR2_DA_TO_DB(mp, map->br_startoff);
+			curdb = xfs_dir2_da_to_db(mp, map->br_startoff);
 			error = xfs_da_read_buf(tp, dp, map->br_startoff,
 				map->br_blockcount >= mp->m_dirblkfsbs ?
 				    XFS_FSB_TO_DADDR(mp, map->br_startblock) :
@@ -1014,7 +1014,7 @@ xfs_dir2_leaf_getdents(
 			/*
 			 * Having done a read, we need to set a new offset.
 			 */
-			newoff = XFS_DIR2_DB_OFF_TO_BYTE(mp, curdb, 0);
+			newoff = xfs_dir2_db_off_to_byte(mp, curdb, 0);
 			/*
 			 * Start of the current block.
 			 */
@@ -1024,7 +1024,7 @@ xfs_dir2_leaf_getdents(
 			 * Make sure we're in the right block.
 			 */
 			else if (curoff > newoff)
-				ASSERT(XFS_DIR2_BYTE_TO_DB(mp, curoff) ==
+				ASSERT(xfs_dir2_byte_to_db(mp, curoff) ==
 				       curdb);
 			data = bp->data;
 			xfs_dir2_data_check(dp, bp);
@@ -1032,7 +1032,7 @@ xfs_dir2_leaf_getdents(
 			 * Find our position in the block.
 			 */
 			ptr = (char *)&data->u;
-			byteoff = XFS_DIR2_BYTE_TO_OFF(mp, curoff);
+			byteoff = xfs_dir2_byte_to_off(mp, curoff);
 			/*
 			 * Skip past the header.
 			 */
@@ -1054,15 +1054,15 @@ xfs_dir2_leaf_getdents(
 					}
 					dep = (xfs_dir2_data_entry_t *)ptr;
 					length =
-					   XFS_DIR2_DATA_ENTSIZE(dep->namelen);
+					   xfs_dir2_data_entsize(dep->namelen);
 					ptr += length;
 				}
 				/*
 				 * Now set our real offset.
 				 */
 				curoff =
-					XFS_DIR2_DB_OFF_TO_BYTE(mp,
-					    XFS_DIR2_BYTE_TO_DB(mp, curoff),
+					xfs_dir2_db_off_to_byte(mp,
+					    xfs_dir2_byte_to_db(mp, curoff),
 					    (char *)ptr - (char *)data);
 				if (ptr >= (char *)data + mp->m_dirblksize) {
 					continue;
@@ -1091,9 +1091,9 @@ xfs_dir2_leaf_getdents(
 
 		p->namelen = dep->namelen;
 
-		length = XFS_DIR2_DATA_ENTSIZE(p->namelen);
+		length = xfs_dir2_data_entsize(p->namelen);
 
-		p->cook = XFS_DIR2_BYTE_TO_DATAPTR(mp, curoff + length);
+		p->cook = xfs_dir2_byte_to_dataptr(mp, curoff + length);
 
 		p->ino = be64_to_cpu(dep->inumber);
 #if XFS_BIG_INUMS
@@ -1121,10 +1121,10 @@ xfs_dir2_leaf_getdents(
 	 * All done.  Set output offset value to current offset.
 	 */
 	*eofp = eof;
-	if (curoff > XFS_DIR2_DATAPTR_TO_BYTE(mp, XFS_DIR2_MAX_DATAPTR))
+	if (curoff > xfs_dir2_dataptr_to_byte(mp, XFS_DIR2_MAX_DATAPTR))
 		uio->uio_offset = XFS_DIR2_MAX_DATAPTR;
 	else
-		uio->uio_offset = XFS_DIR2_BYTE_TO_DATAPTR(mp, curoff);
+		uio->uio_offset = xfs_dir2_byte_to_dataptr(mp, curoff);
 	kmem_free(map, map_size * sizeof(*map));
 	kmem_free(p, sizeof(*p));
 	if (bp)
@@ -1159,7 +1159,7 @@ xfs_dir2_leaf_init(
 	/*
 	 * Get the buffer for the block.
 	 */
-	error = xfs_da_get_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, bno), -1, &bp,
+	error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, bno), -1, &bp,
 		XFS_DATA_FORK);
 	if (error) {
 		return error;
@@ -1181,7 +1181,7 @@ xfs_dir2_leaf_init(
 	 * the block.
 	 */
 	if (magic == XFS_DIR2_LEAF1_MAGIC) {
-		ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+		ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 		ltp->bestcount = 0;
 		xfs_dir2_leaf_log_tail(tp, bp);
 	}
@@ -1206,9 +1206,9 @@ xfs_dir2_leaf_log_bests(
 
 	leaf = bp->data;
 	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
-	ltp = XFS_DIR2_LEAF_TAIL_P(tp->t_mountp, leaf);
-	firstb = XFS_DIR2_LEAF_BESTS_P(ltp) + first;
-	lastb = XFS_DIR2_LEAF_BESTS_P(ltp) + last;
+	ltp = xfs_dir2_leaf_tail_p(tp->t_mountp, leaf);
+	firstb = xfs_dir2_leaf_bests_p(ltp) + first;
+	lastb = xfs_dir2_leaf_bests_p(ltp) + last;
 	xfs_da_log_buf(tp, bp, (uint)((char *)firstb - (char *)leaf),
 		(uint)((char *)lastb - (char *)leaf + sizeof(*lastb) - 1));
 }
@@ -1268,7 +1268,7 @@ xfs_dir2_leaf_log_tail(
 	mp = tp->t_mountp;
 	leaf = bp->data;
 	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	xfs_da_log_buf(tp, bp, (uint)((char *)ltp - (char *)leaf),
 		(uint)(mp->m_dirblksize - 1));
 }
@@ -1312,7 +1312,7 @@ xfs_dir2_leaf_lookup(
 	 */
 	dep = (xfs_dir2_data_entry_t *)
 	      ((char *)dbp->data +
-	       XFS_DIR2_DATAPTR_TO_OFF(dp->i_mount, be32_to_cpu(lep->address)));
+	       xfs_dir2_dataptr_to_off(dp->i_mount, be32_to_cpu(lep->address)));
 	/*
 	 * Return the found inode number.
 	 */
@@ -1381,7 +1381,7 @@ xfs_dir2_leaf_lookup_int(
 		/*
 		 * Get the new data block number.
 		 */
-		newdb = XFS_DIR2_DATAPTR_TO_DB(mp, be32_to_cpu(lep->address));
+		newdb = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 		/*
 		 * If it's not the same as the old data block number,
 		 * need to pitch the old one and read the new one.
@@ -1391,7 +1391,7 @@ xfs_dir2_leaf_lookup_int(
 				xfs_da_brelse(tp, dbp);
 			if ((error =
 			    xfs_da_read_buf(tp, dp,
-				    XFS_DIR2_DB_TO_DA(mp, newdb), -1, &dbp,
+				    xfs_dir2_db_to_da(mp, newdb), -1, &dbp,
 				    XFS_DATA_FORK))) {
 				xfs_da_brelse(tp, lbp);
 				return error;
@@ -1404,7 +1404,7 @@ xfs_dir2_leaf_lookup_int(
 		 */
 		dep = (xfs_dir2_data_entry_t *)
 		      ((char *)dbp->data +
-		       XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(lep->address)));
+		       xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
 		/*
 		 * If it matches then return it.
 		 */
@@ -1469,20 +1469,20 @@ xfs_dir2_leaf_removename(
 	 * Point to the leaf entry, use that to point to the data entry.
 	 */
 	lep = &leaf->ents[index];
-	db = XFS_DIR2_DATAPTR_TO_DB(mp, be32_to_cpu(lep->address));
+	db = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)data + XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(lep->address)));
+	      ((char *)data + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
 	needscan = needlog = 0;
 	oldbest = be16_to_cpu(data->hdr.bestfree[0].length);
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
-	bestsp = XFS_DIR2_LEAF_BESTS_P(ltp);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
+	bestsp = xfs_dir2_leaf_bests_p(ltp);
 	ASSERT(be16_to_cpu(bestsp[db]) == oldbest);
 	/*
 	 * Mark the former data entry unused.
 	 */
 	xfs_dir2_data_make_free(tp, dbp,
 		(xfs_dir2_data_aoff_t)((char *)dep - (char *)data),
-		XFS_DIR2_DATA_ENTSIZE(dep->namelen), &needlog, &needscan);
+		xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
 	/*
 	 * We just mark the leaf entry stale by putting a null in it.
 	 */
@@ -1602,7 +1602,7 @@ xfs_dir2_leaf_replace(
 	 */
 	dep = (xfs_dir2_data_entry_t *)
 	      ((char *)dbp->data +
-	       XFS_DIR2_DATAPTR_TO_OFF(dp->i_mount, be32_to_cpu(lep->address)));
+	       xfs_dir2_dataptr_to_off(dp->i_mount, be32_to_cpu(lep->address)));
 	ASSERT(args->inumber != be64_to_cpu(dep->inumber));
 	/*
 	 * Put the new inode number in, log it.
@@ -1698,7 +1698,7 @@ xfs_dir2_leaf_trim_data(
 	/*
 	 * Read the offending data block.  We need its buffer.
 	 */
-	if ((error = xfs_da_read_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, db), -1, &dbp,
+	if ((error = xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, db), -1, &dbp,
 			XFS_DATA_FORK))) {
 		return error;
 	}
@@ -1712,7 +1712,7 @@ xfs_dir2_leaf_trim_data(
 	 */
 
 	leaf = lbp->data;
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	ASSERT(be16_to_cpu(data->hdr.bestfree[0].length) ==
 	       mp->m_dirblksize - (uint)sizeof(data->hdr));
 	ASSERT(db == be32_to_cpu(ltp->bestcount) - 1);
@@ -1727,7 +1727,7 @@ xfs_dir2_leaf_trim_data(
 	/*
 	 * Eliminate the last bests entry from the table.
 	 */
-	bestsp = XFS_DIR2_LEAF_BESTS_P(ltp);
+	bestsp = xfs_dir2_leaf_bests_p(ltp);
 	be32_add(&ltp->bestcount, -1);
 	memmove(&bestsp[1], &bestsp[0], be32_to_cpu(ltp->bestcount) * sizeof(*bestsp));
 	xfs_dir2_leaf_log_tail(tp, lbp);
@@ -1838,12 +1838,12 @@ xfs_dir2_node_to_leaf(
 	/*
 	 * Set up the leaf tail from the freespace block.
 	 */
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	ltp->bestcount = free->hdr.nvalid;
 	/*
 	 * Set up the leaf bests table.
 	 */
-	memcpy(XFS_DIR2_LEAF_BESTS_P(ltp), free->bests,
+	memcpy(xfs_dir2_leaf_bests_p(ltp), free->bests,
 		be32_to_cpu(ltp->bestcount) * sizeof(leaf->bests[0]));
 	xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
 	xfs_dir2_leaf_log_tail(tp, lbp);

fs/xfs/xfs_dir2_leaf.h

@@ -32,7 +32,7 @@ struct xfs_trans;
 #define	XFS_DIR2_LEAF_SPACE	1
 #define	XFS_DIR2_LEAF_OFFSET	(XFS_DIR2_LEAF_SPACE * XFS_DIR2_SPACE_SIZE)
 #define	XFS_DIR2_LEAF_FIRSTDB(mp)	\
-	XFS_DIR2_BYTE_TO_DB(mp, XFS_DIR2_LEAF_OFFSET)
+	xfs_dir2_byte_to_db(mp, XFS_DIR2_LEAF_OFFSET)
 
 /*
  * Offset in data space of a data entry.
@@ -82,7 +82,6 @@ typedef struct xfs_dir2_leaf {
  * DB blocks here are logical directory block numbers, not filesystem blocks.
  */
 
-#define	XFS_DIR2_MAX_LEAF_ENTS(mp)	xfs_dir2_max_leaf_ents(mp)
 static inline int xfs_dir2_max_leaf_ents(struct xfs_mount *mp)
 {
 	return (int)(((mp)->m_dirblksize - (uint)sizeof(xfs_dir2_leaf_hdr_t)) /
@@ -92,7 +91,6 @@ static inline int xfs_dir2_max_leaf_ents(struct xfs_mount *mp)
 /*
  * Get address of the bestcount field in the single-leaf block.
  */
-#define	XFS_DIR2_LEAF_TAIL_P(mp,lp)	xfs_dir2_leaf_tail_p(mp, lp)
 static inline xfs_dir2_leaf_tail_t *
 xfs_dir2_leaf_tail_p(struct xfs_mount *mp, xfs_dir2_leaf_t *lp)
 {
@@ -104,7 +102,6 @@ xfs_dir2_leaf_tail_p(struct xfs_mount *mp, xfs_dir2_leaf_t *lp)
 /*
  * Get address of the bests array in the single-leaf block.
  */
-#define	XFS_DIR2_LEAF_BESTS_P(ltp)	xfs_dir2_leaf_bests_p(ltp)
 static inline __be16 *
 xfs_dir2_leaf_bests_p(xfs_dir2_leaf_tail_t *ltp)
 {
@@ -114,7 +111,6 @@ xfs_dir2_leaf_bests_p(xfs_dir2_leaf_tail_t *ltp)
 /*
  * Convert dataptr to byte in file space
  */
-#define	XFS_DIR2_DATAPTR_TO_BYTE(mp,dp)	xfs_dir2_dataptr_to_byte(mp, dp)
 static inline xfs_dir2_off_t
 xfs_dir2_dataptr_to_byte(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
 {
@@ -124,7 +120,6 @@ xfs_dir2_dataptr_to_byte(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
 /*
  * Convert byte in file space to dataptr.  It had better be aligned.
  */
-#define	XFS_DIR2_BYTE_TO_DATAPTR(mp,by)	xfs_dir2_byte_to_dataptr(mp,by)
 static inline xfs_dir2_dataptr_t
 xfs_dir2_byte_to_dataptr(struct xfs_mount *mp, xfs_dir2_off_t by)
 {
@@ -134,7 +129,6 @@ xfs_dir2_byte_to_dataptr(struct xfs_mount *mp, xfs_dir2_off_t by)
 /*
  * Convert byte in space to (DB) block
  */
-#define	XFS_DIR2_BYTE_TO_DB(mp,by)	xfs_dir2_byte_to_db(mp, by)
 static inline xfs_dir2_db_t
 xfs_dir2_byte_to_db(struct xfs_mount *mp, xfs_dir2_off_t by)
 {
@@ -145,17 +139,15 @@ xfs_dir2_byte_to_db(struct xfs_mount *mp, xfs_dir2_off_t by)
 /*
  * Convert dataptr to a block number
  */
-#define	XFS_DIR2_DATAPTR_TO_DB(mp,dp)	xfs_dir2_dataptr_to_db(mp, dp)
 static inline xfs_dir2_db_t
 xfs_dir2_dataptr_to_db(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
 {
-	return XFS_DIR2_BYTE_TO_DB(mp, XFS_DIR2_DATAPTR_TO_BYTE(mp, dp));
+	return xfs_dir2_byte_to_db(mp, xfs_dir2_dataptr_to_byte(mp, dp));
 }
 
 /*
  * Convert byte in space to offset in a block
  */
-#define	XFS_DIR2_BYTE_TO_OFF(mp,by)	xfs_dir2_byte_to_off(mp, by)
 static inline xfs_dir2_data_aoff_t
 xfs_dir2_byte_to_off(struct xfs_mount *mp, xfs_dir2_off_t by)
 {
@@ -166,18 +158,15 @@ xfs_dir2_byte_to_off(struct xfs_mount *mp, xfs_dir2_off_t by)
 /*
  * Convert dataptr to a byte offset in a block
  */
-#define	XFS_DIR2_DATAPTR_TO_OFF(mp,dp)	xfs_dir2_dataptr_to_off(mp, dp)
 static inline xfs_dir2_data_aoff_t
 xfs_dir2_dataptr_to_off(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
 {
-	return XFS_DIR2_BYTE_TO_OFF(mp, XFS_DIR2_DATAPTR_TO_BYTE(mp, dp));
+	return xfs_dir2_byte_to_off(mp, xfs_dir2_dataptr_to_byte(mp, dp));
 }
 
 /*
  * Convert block and offset to byte in space
  */
-#define	XFS_DIR2_DB_OFF_TO_BYTE(mp,db,o)	\
-	xfs_dir2_db_off_to_byte(mp, db, o)
 static inline xfs_dir2_off_t
 xfs_dir2_db_off_to_byte(struct xfs_mount *mp, xfs_dir2_db_t db,
 			xfs_dir2_data_aoff_t o)
@@ -189,7 +178,6 @@ xfs_dir2_db_off_to_byte(struct xfs_mount *mp, xfs_dir2_db_t db,
 /*
  * Convert block (DB) to block (dablk)
  */
-#define	XFS_DIR2_DB_TO_DA(mp,db)	xfs_dir2_db_to_da(mp, db)
 static inline xfs_dablk_t
 xfs_dir2_db_to_da(struct xfs_mount *mp, xfs_dir2_db_t db)
 {
@@ -199,29 +187,25 @@ xfs_dir2_db_to_da(struct xfs_mount *mp, xfs_dir2_db_t db)
 /*
  * Convert byte in space to (DA) block
  */
-#define	XFS_DIR2_BYTE_TO_DA(mp,by)	xfs_dir2_byte_to_da(mp, by)
 static inline xfs_dablk_t
 xfs_dir2_byte_to_da(struct xfs_mount *mp, xfs_dir2_off_t by)
 {
-	return XFS_DIR2_DB_TO_DA(mp, XFS_DIR2_BYTE_TO_DB(mp, by));
+	return xfs_dir2_db_to_da(mp, xfs_dir2_byte_to_db(mp, by));
 }
 
 /*
  * Convert block and offset to dataptr
  */
-#define	XFS_DIR2_DB_OFF_TO_DATAPTR(mp,db,o)	\
-	xfs_dir2_db_off_to_dataptr(mp, db, o)
 static inline xfs_dir2_dataptr_t
 xfs_dir2_db_off_to_dataptr(struct xfs_mount *mp, xfs_dir2_db_t db,
 			   xfs_dir2_data_aoff_t o)
 {
-	return XFS_DIR2_BYTE_TO_DATAPTR(mp, XFS_DIR2_DB_OFF_TO_BYTE(mp, db, o));
+	return xfs_dir2_byte_to_dataptr(mp, xfs_dir2_db_off_to_byte(mp, db, o));
 }
 
 /*
  * Convert block (dablk) to block (DB)
  */
-#define	XFS_DIR2_DA_TO_DB(mp,da)	xfs_dir2_da_to_db(mp, da)
 static inline xfs_dir2_db_t
 xfs_dir2_da_to_db(struct xfs_mount *mp, xfs_dablk_t da)
 {
@@ -231,11 +215,10 @@ xfs_dir2_da_to_db(struct xfs_mount *mp, xfs_dablk_t da)
 /*
  * Convert block (dablk) to byte offset in space
  */
-#define XFS_DIR2_DA_TO_BYTE(mp,da)	xfs_dir2_da_to_byte(mp, da)
 static inline xfs_dir2_off_t
 xfs_dir2_da_to_byte(struct xfs_mount *mp, xfs_dablk_t da)
 {
-	return XFS_DIR2_DB_OFF_TO_BYTE(mp, XFS_DIR2_DA_TO_DB(mp, da), 0);
+	return xfs_dir2_db_off_to_byte(mp, xfs_dir2_da_to_db(mp, da), 0);
 }
 
 /*

fs/xfs/xfs_dir2_node.c

@@ -136,14 +136,14 @@ xfs_dir2_leaf_to_node(
 	/*
 	 * Get the buffer for the new freespace block.
 	 */
-	if ((error = xfs_da_get_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, fdb), -1, &fbp,
+	if ((error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, fdb), -1, &fbp,
 			XFS_DATA_FORK))) {
 		return error;
 	}
 	ASSERT(fbp != NULL);
 	free = fbp->data;
 	leaf = lbp->data;
-	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
+	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	/*
 	 * Initialize the freespace block header.
 	 */
@@ -155,7 +155,7 @@ xfs_dir2_leaf_to_node(
 	 * Copy freespace entries from the leaf block to the new block.
 	 * Count active entries.
 	 */
-	for (i = n = 0, from = XFS_DIR2_LEAF_BESTS_P(ltp), to = free->bests;
+	for (i = n = 0, from = xfs_dir2_leaf_bests_p(ltp), to = free->bests;
 	     i < be32_to_cpu(ltp->bestcount); i++, from++, to++) {
 		if ((off = be16_to_cpu(*from)) != NULLDATAOFF)
 			n++;
@@ -215,7 +215,7 @@ xfs_dir2_leafn_add(
 	 * a compact.
 	 */
 
-	if (be16_to_cpu(leaf->hdr.count) == XFS_DIR2_MAX_LEAF_ENTS(mp)) {
+	if (be16_to_cpu(leaf->hdr.count) == xfs_dir2_max_leaf_ents(mp)) {
 		if (!leaf->hdr.stale)
 			return XFS_ERROR(ENOSPC);
 		compact = be16_to_cpu(leaf->hdr.stale) > 1;
@@ -327,7 +327,7 @@ xfs_dir2_leafn_add(
 	 * Insert the new entry, log everything.
 	 */
 	lep->hashval = cpu_to_be32(args->hashval);
-	lep->address = cpu_to_be32(XFS_DIR2_DB_OFF_TO_DATAPTR(mp,
+	lep->address = cpu_to_be32(xfs_dir2_db_off_to_dataptr(mp,
 				args->blkno, args->index));
 	xfs_dir2_leaf_log_header(tp, bp);
 	xfs_dir2_leaf_log_ents(tp, bp, lfloglow, lfloghigh);
@@ -352,7 +352,7 @@ xfs_dir2_leafn_check(
 	leaf = bp->data;
 	mp = dp->i_mount;
 	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
-	ASSERT(be16_to_cpu(leaf->hdr.count) <= XFS_DIR2_MAX_LEAF_ENTS(mp));
+	ASSERT(be16_to_cpu(leaf->hdr.count) <= xfs_dir2_max_leaf_ents(mp));
 	for (i = stale = 0; i < be16_to_cpu(leaf->hdr.count); i++) {
 		if (i + 1 < be16_to_cpu(leaf->hdr.count)) {
 			ASSERT(be32_to_cpu(leaf->ents[i].hashval) <=
@@ -440,7 +440,7 @@ xfs_dir2_leafn_lookup_int(
 	if (args->addname) {
 		curfdb = curbp ? state->extrablk.blkno : -1;
 		curdb = -1;
-		length = XFS_DIR2_DATA_ENTSIZE(args->namelen);
+		length = xfs_dir2_data_entsize(args->namelen);
 		if ((free = (curbp ? curbp->data : NULL)))
 			ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
 	}
@@ -465,7 +465,7 @@ xfs_dir2_leafn_lookup_int(
 		/*
 		 * Pull the data block number from the entry.
 		 */
-		newdb = XFS_DIR2_DATAPTR_TO_DB(mp, be32_to_cpu(lep->address));
+		newdb = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 		/*
 		 * For addname, we're looking for a place to put the new entry.
 		 * We want to use a data block with an entry of equal
@@ -482,7 +482,7 @@ xfs_dir2_leafn_lookup_int(
 				 * Convert the data block to the free block
 				 * holding its freespace information.
 				 */
-				newfdb = XFS_DIR2_DB_TO_FDB(mp, newdb);
+				newfdb = xfs_dir2_db_to_fdb(mp, newdb);
 				/*
 				 * If it's not the one we have in hand,
 				 * read it in.
@@ -497,7 +497,7 @@ xfs_dir2_leafn_lookup_int(
 					 * Read the free block.
 					 */
 					if ((error = xfs_da_read_buf(tp, dp,
-							XFS_DIR2_DB_TO_DA(mp,
+							xfs_dir2_db_to_da(mp,
 								newfdb),
 							-1, &curbp,
 							XFS_DATA_FORK))) {
@@ -517,7 +517,7 @@ xfs_dir2_leafn_lookup_int(
 				/*
 				 * Get the index for our entry.
 				 */
-				fi = XFS_DIR2_DB_TO_FDINDEX(mp, curdb);
+				fi = xfs_dir2_db_to_fdindex(mp, curdb);
 				/*
 				 * If it has room, return it.
 				 */
@@ -561,7 +561,7 @@ xfs_dir2_leafn_lookup_int(
 				 */
 				if ((error =
 				    xfs_da_read_buf(tp, dp,
-					    XFS_DIR2_DB_TO_DA(mp, newdb), -1,
+					    xfs_dir2_db_to_da(mp, newdb), -1,
 					    &curbp, XFS_DATA_FORK))) {
 					return error;
 				}
@@ -573,7 +573,7 @@ xfs_dir2_leafn_lookup_int(
 			 */
 			dep = (xfs_dir2_data_entry_t *)
 			      ((char *)curbp->data +
-			       XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(lep->address)));
+			       xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
 			/*
 			 * Compare the entry, return it if it matches.
 			 */
@@ -876,9 +876,9 @@ xfs_dir2_leafn_remove(
 	/*
 	 * Extract the data block and offset from the entry.
 	 */
-	db = XFS_DIR2_DATAPTR_TO_DB(mp, be32_to_cpu(lep->address));
+	db = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 	ASSERT(dblk->blkno == db);
-	off = XFS_DIR2_DATAPTR_TO_OFF(mp, be32_to_cpu(lep->address));
+	off = xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address));
 	ASSERT(dblk->index == off);
 	/*
 	 * Kill the leaf entry by marking it stale.
@@ -898,7 +898,7 @@ xfs_dir2_leafn_remove(
 	longest = be16_to_cpu(data->hdr.bestfree[0].length);
 	needlog = needscan = 0;
 	xfs_dir2_data_make_free(tp, dbp, off,
-		XFS_DIR2_DATA_ENTSIZE(dep->namelen), &needlog, &needscan);
+		xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
 	/*
 	 * Rescan the data block freespaces for bestfree.
 	 * Log the data block header if needed.
@@ -924,8 +924,8 @@ xfs_dir2_leafn_remove(
 		 * Convert the data block number to a free block,
 		 * read in the free block.
 		 */
-		fdb = XFS_DIR2_DB_TO_FDB(mp, db);
-		if ((error = xfs_da_read_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, fdb),
+		fdb = xfs_dir2_db_to_fdb(mp, db);
+		if ((error = xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, fdb),
 				-1, &fbp, XFS_DATA_FORK))) {
 			return error;
 		}
@@ -937,7 +937,7 @@ xfs_dir2_leafn_remove(
 		/*
 		 * Calculate which entry we need to fix.
 		 */
-		findex = XFS_DIR2_DB_TO_FDINDEX(mp, db);
+		findex = xfs_dir2_db_to_fdindex(mp, db);
 		longest = be16_to_cpu(data->hdr.bestfree[0].length);
 		/*
 		 * If the data block is now empty we can get rid of it
@@ -1073,7 +1073,7 @@ xfs_dir2_leafn_split(
 	/*
 	 * Initialize the new leaf block.
 	 */
-	error = xfs_dir2_leaf_init(args, XFS_DIR2_DA_TO_DB(mp, blkno),
+	error = xfs_dir2_leaf_init(args, xfs_dir2_da_to_db(mp, blkno),
 		&newblk->bp, XFS_DIR2_LEAFN_MAGIC);
 	if (error) {
 		return error;
@@ -1385,7 +1385,7 @@ xfs_dir2_node_addname_int(
 	dp = args->dp;
 	mp = dp->i_mount;
 	tp = args->trans;
-	length = XFS_DIR2_DATA_ENTSIZE(args->namelen);
+	length = xfs_dir2_data_entsize(args->namelen);
 	/*
 	 * If we came in with a freespace block that means that lookup
 	 * found an entry with our hash value.  This is the freespace
@@ -1438,7 +1438,7 @@ xfs_dir2_node_addname_int(
 
 		if ((error = xfs_bmap_last_offset(tp, dp, &fo, XFS_DATA_FORK)))
 			return error;
-		lastfbno = XFS_DIR2_DA_TO_DB(mp, (xfs_dablk_t)fo);
+		lastfbno = xfs_dir2_da_to_db(mp, (xfs_dablk_t)fo);
 		fbno = ifbno;
 	}
 	/*
@@ -1474,7 +1474,7 @@ xfs_dir2_node_addname_int(
 			 * to avoid it.
 			 */
 			if ((error = xfs_da_read_buf(tp, dp,
-					XFS_DIR2_DB_TO_DA(mp, fbno), -2, &fbp,
+					xfs_dir2_db_to_da(mp, fbno), -2, &fbp,
 					XFS_DATA_FORK))) {
 				return error;
 			}
@@ -1550,9 +1550,9 @@ xfs_dir2_node_addname_int(
 		 * Get the freespace block corresponding to the data block
 		 * that was just allocated.
 		 */
-		fbno = XFS_DIR2_DB_TO_FDB(mp, dbno);
+		fbno = xfs_dir2_db_to_fdb(mp, dbno);
 		if (unlikely(error = xfs_da_read_buf(tp, dp,
-				XFS_DIR2_DB_TO_DA(mp, fbno), -2, &fbp,
+				xfs_dir2_db_to_da(mp, fbno), -2, &fbp,
 				XFS_DATA_FORK))) {
 			xfs_da_buf_done(dbp);
 			return error;
@@ -1567,14 +1567,14 @@ xfs_dir2_node_addname_int(
 				return error;
 			}
 
-			if (unlikely(XFS_DIR2_DB_TO_FDB(mp, dbno) != fbno)) {
+			if (unlikely(xfs_dir2_db_to_fdb(mp, dbno) != fbno)) {
 				cmn_err(CE_ALERT,
 					"xfs_dir2_node_addname_int: dir ino "
 					"%llu needed freesp block %lld for\n"
 					"  data block %lld, got %lld\n"
 					"  ifbno %llu lastfbno %d\n",
 					(unsigned long long)dp->i_ino,
-					(long long)XFS_DIR2_DB_TO_FDB(mp, dbno),
+					(long long)xfs_dir2_db_to_fdb(mp, dbno),
 					(long long)dbno, (long long)fbno,
 					(unsigned long long)ifbno, lastfbno);
 				if (fblk) {
@@ -1598,7 +1598,7 @@ xfs_dir2_node_addname_int(
 			 * Get a buffer for the new block.
 			 */
 			if ((error = xfs_da_get_buf(tp, dp,
-						   XFS_DIR2_DB_TO_DA(mp, fbno),
+						   xfs_dir2_db_to_da(mp, fbno),
 						   -1, &fbp, XFS_DATA_FORK))) {
 				return error;
 			}
@@ -1623,7 +1623,7 @@ xfs_dir2_node_addname_int(
 		/*
 		 * Set the freespace block index from the data block number.
 		 */
-		findex = XFS_DIR2_DB_TO_FDINDEX(mp, dbno);
+		findex = xfs_dir2_db_to_fdindex(mp, dbno);
 		/*
 		 * If it's after the end of the current entries in the
 		 * freespace block, extend that table.
@@ -1669,7 +1669,7 @@ xfs_dir2_node_addname_int(
 		 * Read the data block in.
 		 */
 		if (unlikely(
-		    error = xfs_da_read_buf(tp, dp, XFS_DIR2_DB_TO_DA(mp, dbno),
+		    error = xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, dbno),
 				-1, &dbp, XFS_DATA_FORK))) {
 			if ((fblk == NULL || fblk->bp == NULL) && fbp != NULL)
 				xfs_da_buf_done(fbp);
@@ -1698,7 +1698,7 @@ xfs_dir2_node_addname_int(
 	dep->inumber = cpu_to_be64(args->inumber);
 	dep->namelen = args->namelen;
 	memcpy(dep->name, args->name, dep->namelen);
-	tagp = XFS_DIR2_DATA_ENTRY_TAG_P(dep);
+	tagp = xfs_dir2_data_entry_tag_p(dep);
 	*tagp = cpu_to_be16((char *)dep - (char *)data);
 	xfs_dir2_data_log_entry(tp, dbp, dep);
 	/*
@@ -1904,7 +1904,7 @@ xfs_dir2_node_replace(
 		ASSERT(be32_to_cpu(data->hdr.magic) == XFS_DIR2_DATA_MAGIC);
 		dep = (xfs_dir2_data_entry_t *)
 		      ((char *)data +
-		       XFS_DIR2_DATAPTR_TO_OFF(state->mp, be32_to_cpu(lep->address)));
+		       xfs_dir2_dataptr_to_off(state->mp, be32_to_cpu(lep->address)));
 		ASSERT(inum != be64_to_cpu(dep->inumber));
 		/*
 		 * Fill in the new inode number and log the entry.
@@ -1980,7 +1980,7 @@ xfs_dir2_node_trim_free(
 	 * Blow the block away.
 	 */
 	if ((error =
-	    xfs_dir2_shrink_inode(args, XFS_DIR2_DA_TO_DB(mp, (xfs_dablk_t)fo),
+	    xfs_dir2_shrink_inode(args, xfs_dir2_da_to_db(mp, (xfs_dablk_t)fo),
 		    bp))) {
 		/*
 		 * Can't fail with ENOSPC since that only happens with no

fs/xfs/xfs_dir2_node.h

@@ -36,7 +36,7 @@ struct xfs_trans;
 #define	XFS_DIR2_FREE_SPACE	2
 #define	XFS_DIR2_FREE_OFFSET	(XFS_DIR2_FREE_SPACE * XFS_DIR2_SPACE_SIZE)
 #define	XFS_DIR2_FREE_FIRSTDB(mp)	\
-	XFS_DIR2_BYTE_TO_DB(mp, XFS_DIR2_FREE_OFFSET)
+	xfs_dir2_byte_to_db(mp, XFS_DIR2_FREE_OFFSET)
 
 #define	XFS_DIR2_FREE_MAGIC	0x58443246	/* XD2F */
 
@@ -60,7 +60,6 @@ typedef struct xfs_dir2_free {
 /*
  * Convert data space db to the corresponding free db.
  */
-#define	XFS_DIR2_DB_TO_FDB(mp,db)	xfs_dir2_db_to_fdb(mp, db)
 static inline xfs_dir2_db_t
 xfs_dir2_db_to_fdb(struct xfs_mount *mp, xfs_dir2_db_t db)
 {
@@ -70,7 +69,6 @@ xfs_dir2_db_to_fdb(struct xfs_mount *mp, xfs_dir2_db_t db)
 /*
  * Convert data space db to the corresponding index in a free db.
  */
-#define	XFS_DIR2_DB_TO_FDINDEX(mp,db)	xfs_dir2_db_to_fdindex(mp, db)
 static inline int
 xfs_dir2_db_to_fdindex(struct xfs_mount *mp, xfs_dir2_db_t db)
 {

fs/xfs/xfs_dir2_sf.c

@@ -89,8 +89,8 @@ xfs_dir2_block_sfsize(
 	mp = dp->i_mount;
 
 	count = i8count = namelen = 0;
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
-	blp = XFS_DIR2_BLOCK_LEAF_P(btp);
+	btp = xfs_dir2_block_tail_p(mp, block);
+	blp = xfs_dir2_block_leaf_p(btp);
 
 	/*
 	 * Iterate over the block's data entries by using the leaf pointers.
@@ -102,7 +102,7 @@ xfs_dir2_block_sfsize(
 		 * Calculate the pointer to the entry at hand.
 		 */
 		dep = (xfs_dir2_data_entry_t *)
-		      ((char *)block + XFS_DIR2_DATAPTR_TO_OFF(mp, addr));
+		      ((char *)block + xfs_dir2_dataptr_to_off(mp, addr));
 		/*
 		 * Detect . and .., so we can special-case them.
 		 * . is not included in sf directories.
@@ -124,7 +124,7 @@ xfs_dir2_block_sfsize(
 		/*
 		 * Calculate the new size, see if we should give up yet.
 		 */
-		size = XFS_DIR2_SF_HDR_SIZE(i8count) +		/* header */
+		size = xfs_dir2_sf_hdr_size(i8count) +		/* header */
 		       count +					/* namelen */
 		       count * (uint)sizeof(xfs_dir2_sf_off_t) + /* offset */
 		       namelen +				/* name */
@@ -139,7 +139,7 @@ xfs_dir2_block_sfsize(
 	 */
 	sfhp->count = count;
 	sfhp->i8count = i8count;
-	XFS_DIR2_SF_PUT_INUMBER((xfs_dir2_sf_t *)sfhp, &parent, &sfhp->parent);
+	xfs_dir2_sf_put_inumber((xfs_dir2_sf_t *)sfhp, &parent, &sfhp->parent);
 	return size;
 }
 
@@ -199,15 +199,15 @@ xfs_dir2_block_to_sf(
 	 * Copy the header into the newly allocate local space.
 	 */
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	memcpy(sfp, sfhp, XFS_DIR2_SF_HDR_SIZE(sfhp->i8count));
+	memcpy(sfp, sfhp, xfs_dir2_sf_hdr_size(sfhp->i8count));
 	dp->i_d.di_size = size;
 	/*
 	 * Set up to loop over the block's entries.
 	 */
-	btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, block);
 	ptr = (char *)block->u;
-	endptr = (char *)XFS_DIR2_BLOCK_LEAF_P(btp);
-	sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+	endptr = (char *)xfs_dir2_block_leaf_p(btp);
+	sfep = xfs_dir2_sf_firstentry(sfp);
 	/*
 	 * Loop over the active and unused entries.
 	 * Stop when we reach the leaf/tail portion of the block.
@@ -233,22 +233,22 @@ xfs_dir2_block_to_sf(
 		else if (dep->namelen == 2 &&
 			 dep->name[0] == '.' && dep->name[1] == '.')
 			ASSERT(be64_to_cpu(dep->inumber) ==
-			       XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent));
+			       xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent));
 		/*
 		 * Normal entry, copy it into shortform.
 		 */
 		else {
 			sfep->namelen = dep->namelen;
-			XFS_DIR2_SF_PUT_OFFSET(sfep,
+			xfs_dir2_sf_put_offset(sfep,
 				(xfs_dir2_data_aoff_t)
 				((char *)dep - (char *)block));
 			memcpy(sfep->name, dep->name, dep->namelen);
 			temp = be64_to_cpu(dep->inumber);
-			XFS_DIR2_SF_PUT_INUMBER(sfp, &temp,
-				XFS_DIR2_SF_INUMBERP(sfep));
-			sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep);
+			xfs_dir2_sf_put_inumber(sfp, &temp,
+				xfs_dir2_sf_inumberp(sfep));
+			sfep = xfs_dir2_sf_nextentry(sfp, sfep);
 		}
-		ptr += XFS_DIR2_DATA_ENTSIZE(dep->namelen);
+		ptr += xfs_dir2_data_entsize(dep->namelen);
 	}
 	ASSERT((char *)sfep - (char *)sfp == size);
 	xfs_dir2_sf_check(args);
@@ -294,11 +294,11 @@ xfs_dir2_sf_addname(
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
 	/*
 	 * Compute entry (and change in) size.
 	 */
-	add_entsize = XFS_DIR2_SF_ENTSIZE_BYNAME(sfp, args->namelen);
+	add_entsize = xfs_dir2_sf_entsize_byname(sfp, args->namelen);
 	incr_isize = add_entsize;
 	objchange = 0;
 #if XFS_BIG_INUMS
@@ -392,7 +392,7 @@ xfs_dir2_sf_addname_easy(
 	/*
 	 * Grow the in-inode space.
 	 */
-	xfs_idata_realloc(dp, XFS_DIR2_SF_ENTSIZE_BYNAME(sfp, args->namelen),
+	xfs_idata_realloc(dp, xfs_dir2_sf_entsize_byname(sfp, args->namelen),
 		XFS_DATA_FORK);
 	/*
 	 * Need to set up again due to realloc of the inode data.
@@ -403,10 +403,10 @@ xfs_dir2_sf_addname_easy(
 	 * Fill in the new entry.
 	 */
 	sfep->namelen = args->namelen;
-	XFS_DIR2_SF_PUT_OFFSET(sfep, offset);
+	xfs_dir2_sf_put_offset(sfep, offset);
 	memcpy(sfep->name, args->name, sfep->namelen);
-	XFS_DIR2_SF_PUT_INUMBER(sfp, &args->inumber,
-		XFS_DIR2_SF_INUMBERP(sfep));
+	xfs_dir2_sf_put_inumber(sfp, &args->inumber,
+		xfs_dir2_sf_inumberp(sfep));
 	/*
 	 * Update the header and inode.
 	 */
@@ -463,14 +463,14 @@ xfs_dir2_sf_addname_hard(
 	 * If it's going to end up at the end then oldsfep will point there.
 	 */
 	for (offset = XFS_DIR2_DATA_FIRST_OFFSET,
-	      oldsfep = XFS_DIR2_SF_FIRSTENTRY(oldsfp),
-	      add_datasize = XFS_DIR2_DATA_ENTSIZE(args->namelen),
+	      oldsfep = xfs_dir2_sf_firstentry(oldsfp),
+	      add_datasize = xfs_dir2_data_entsize(args->namelen),
 	      eof = (char *)oldsfep == &buf[old_isize];
 	     !eof;
-	     offset = new_offset + XFS_DIR2_DATA_ENTSIZE(oldsfep->namelen),
-	      oldsfep = XFS_DIR2_SF_NEXTENTRY(oldsfp, oldsfep),
+	     offset = new_offset + xfs_dir2_data_entsize(oldsfep->namelen),
+	      oldsfep = xfs_dir2_sf_nextentry(oldsfp, oldsfep),
 	      eof = (char *)oldsfep == &buf[old_isize]) {
-		new_offset = XFS_DIR2_SF_GET_OFFSET(oldsfep);
+		new_offset = xfs_dir2_sf_get_offset(oldsfep);
 		if (offset + add_datasize <= new_offset)
 			break;
 	}
@@ -495,10 +495,10 @@ xfs_dir2_sf_addname_hard(
 	 * Fill in the new entry, and update the header counts.
 	 */
 	sfep->namelen = args->namelen;
-	XFS_DIR2_SF_PUT_OFFSET(sfep, offset);
+	xfs_dir2_sf_put_offset(sfep, offset);
 	memcpy(sfep->name, args->name, sfep->namelen);
-	XFS_DIR2_SF_PUT_INUMBER(sfp, &args->inumber,
-		XFS_DIR2_SF_INUMBERP(sfep));
+	xfs_dir2_sf_put_inumber(sfp, &args->inumber,
+		xfs_dir2_sf_inumberp(sfep));
 	sfp->hdr.count++;
 #if XFS_BIG_INUMS
 	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && !objchange)
@@ -508,7 +508,7 @@ xfs_dir2_sf_addname_hard(
 	 * If there's more left to copy, do that.
 	 */
 	if (!eof) {
-		sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep);
+		sfep = xfs_dir2_sf_nextentry(sfp, sfep);
 		memcpy(sfep, oldsfep, old_isize - nbytes);
 	}
 	kmem_free(buf, old_isize);
@@ -544,9 +544,9 @@ xfs_dir2_sf_addname_pick(
 	mp = dp->i_mount;
 
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	size = XFS_DIR2_DATA_ENTSIZE(args->namelen);
+	size = xfs_dir2_data_entsize(args->namelen);
 	offset = XFS_DIR2_DATA_FIRST_OFFSET;
-	sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+	sfep = xfs_dir2_sf_firstentry(sfp);
 	holefit = 0;
 	/*
 	 * Loop over sf entries.
@@ -555,10 +555,10 @@ xfs_dir2_sf_addname_pick(
 	 */
 	for (i = 0; i < sfp->hdr.count; i++) {
 		if (!holefit)
-			holefit = offset + size <= XFS_DIR2_SF_GET_OFFSET(sfep);
-		offset = XFS_DIR2_SF_GET_OFFSET(sfep) +
-			 XFS_DIR2_DATA_ENTSIZE(sfep->namelen);
-		sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep);
+			holefit = offset + size <= xfs_dir2_sf_get_offset(sfep);
+		offset = xfs_dir2_sf_get_offset(sfep) +
+			 xfs_dir2_data_entsize(sfep->namelen);
+		sfep = xfs_dir2_sf_nextentry(sfp, sfep);
 	}
 	/*
 	 * Calculate data bytes used excluding the new entry, if this
@@ -617,18 +617,18 @@ xfs_dir2_sf_check(
 
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
 	offset = XFS_DIR2_DATA_FIRST_OFFSET;
-	ino = XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent);
+	ino = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
 	i8count = ino > XFS_DIR2_MAX_SHORT_INUM;
 
-	for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
 	     i < sfp->hdr.count;
-	     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep)) {
-		ASSERT(XFS_DIR2_SF_GET_OFFSET(sfep) >= offset);
-		ino = XFS_DIR2_SF_GET_INUMBER(sfp, XFS_DIR2_SF_INUMBERP(sfep));
+	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
+		ASSERT(xfs_dir2_sf_get_offset(sfep) >= offset);
+		ino = xfs_dir2_sf_get_inumber(sfp, xfs_dir2_sf_inumberp(sfep));
 		i8count += ino > XFS_DIR2_MAX_SHORT_INUM;
 		offset =
-			XFS_DIR2_SF_GET_OFFSET(sfep) +
-			XFS_DIR2_DATA_ENTSIZE(sfep->namelen);
+			xfs_dir2_sf_get_offset(sfep) +
+			xfs_dir2_data_entsize(sfep->namelen);
 	}
 	ASSERT(i8count == sfp->hdr.i8count);
 	ASSERT(XFS_BIG_INUMS || i8count == 0);
@@ -671,7 +671,7 @@ xfs_dir2_sf_create(
 	ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
 	ASSERT(dp->i_df.if_bytes == 0);
 	i8count = pino > XFS_DIR2_MAX_SHORT_INUM;
-	size = XFS_DIR2_SF_HDR_SIZE(i8count);
+	size = xfs_dir2_sf_hdr_size(i8count);
 	/*
 	 * Make a buffer for the data.
 	 */
@@ -684,7 +684,7 @@ xfs_dir2_sf_create(
 	/*
 	 * Now can put in the inode number, since i8count is set.
 	 */
-	XFS_DIR2_SF_PUT_INUMBER(sfp, &pino, &sfp->hdr.parent);
+	xfs_dir2_sf_put_inumber(sfp, &pino, &sfp->hdr.parent);
 	sfp->hdr.count = 0;
 	dp->i_d.di_size = size;
 	xfs_dir2_sf_check(args);
@@ -727,12 +727,12 @@ xfs_dir2_sf_getdents(
 
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
 
-	ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
 
 	/*
 	 * If the block number in the offset is out of range, we're done.
 	 */
-	if (XFS_DIR2_DATAPTR_TO_DB(mp, dir_offset) > mp->m_dirdatablk) {
+	if (xfs_dir2_dataptr_to_db(mp, dir_offset) > mp->m_dirdatablk) {
 		*eofp = 1;
 		return 0;
 	}
@@ -747,9 +747,9 @@ xfs_dir2_sf_getdents(
 	 * Put . entry unless we're starting past it.
 	 */
 	if (dir_offset <=
-		    XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+		    xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 					       XFS_DIR2_DATA_DOT_OFFSET)) {
-		p.cook = XFS_DIR2_DB_OFF_TO_DATAPTR(mp, 0,
+		p.cook = xfs_dir2_db_off_to_dataptr(mp, 0,
 						XFS_DIR2_DATA_DOTDOT_OFFSET);
 		p.ino = dp->i_ino;
 #if XFS_BIG_INUMS
@@ -762,7 +762,7 @@ xfs_dir2_sf_getdents(
 
 		if (!p.done) {
 			uio->uio_offset =
-				XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+				xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 						XFS_DIR2_DATA_DOT_OFFSET);
 			return error;
 		}
@@ -772,11 +772,11 @@ xfs_dir2_sf_getdents(
 	 * Put .. entry unless we're starting past it.
 	 */
 	if (dir_offset <=
-		    XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+		    xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 					       XFS_DIR2_DATA_DOTDOT_OFFSET)) {
-		p.cook = XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+		p.cook = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 						XFS_DIR2_DATA_FIRST_OFFSET);
-		p.ino = XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent);
+		p.ino = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
 #if XFS_BIG_INUMS
 		p.ino += mp->m_inoadd;
 #endif
@@ -787,7 +787,7 @@ xfs_dir2_sf_getdents(
 
 		if (!p.done) {
 			uio->uio_offset =
-				XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
+				xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 					XFS_DIR2_DATA_DOTDOT_OFFSET);
 			return error;
 		}
@@ -796,23 +796,23 @@ xfs_dir2_sf_getdents(
 	/*
 	 * Loop while there are more entries and put'ing works.
 	 */
-	for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
 		     i < sfp->hdr.count;
-			     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep)) {
+			     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 
-		off = XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
-				XFS_DIR2_SF_GET_OFFSET(sfep));
+		off = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
+				xfs_dir2_sf_get_offset(sfep));
 
 		if (dir_offset > off)
 			continue;
 
 		p.namelen = sfep->namelen;
 
-		p.cook = XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk,
-			XFS_DIR2_SF_GET_OFFSET(sfep) +
-			XFS_DIR2_DATA_ENTSIZE(p.namelen));
+		p.cook = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
+			xfs_dir2_sf_get_offset(sfep) +
+			xfs_dir2_data_entsize(p.namelen));
 
-		p.ino = XFS_DIR2_SF_GET_INUMBER(sfp, XFS_DIR2_SF_INUMBERP(sfep));
+		p.ino = xfs_dir2_sf_get_inumber(sfp, xfs_dir2_sf_inumberp(sfep));
 #if XFS_BIG_INUMS
 		p.ino += mp->m_inoadd;
 #endif
@@ -832,7 +832,7 @@ xfs_dir2_sf_getdents(
 	*eofp = 1;
 
 	uio->uio_offset =
-		XFS_DIR2_DB_OFF_TO_DATAPTR(mp, mp->m_dirdatablk + 1, 0);
+		xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk + 1, 0);
 
 	return 0;
 }
@@ -865,7 +865,7 @@ xfs_dir2_sf_lookup(
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
 	/*
 	 * Special case for .
 	 */
@@ -878,21 +878,21 @@ xfs_dir2_sf_lookup(
 	 */
 	if (args->namelen == 2 &&
 	    args->name[0] == '.' && args->name[1] == '.') {
-		args->inumber = XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent);
+		args->inumber = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
 		return XFS_ERROR(EEXIST);
 	}
 	/*
 	 * Loop over all the entries trying to match ours.
 	 */
-	for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
 	     i < sfp->hdr.count;
-	     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep)) {
+	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 		if (sfep->namelen == args->namelen &&
 		    sfep->name[0] == args->name[0] &&
 		    memcmp(args->name, sfep->name, args->namelen) == 0) {
 			args->inumber =
-				XFS_DIR2_SF_GET_INUMBER(sfp,
-					XFS_DIR2_SF_INUMBERP(sfep));
+				xfs_dir2_sf_get_inumber(sfp,
+					xfs_dir2_sf_inumberp(sfep));
 			return XFS_ERROR(EEXIST);
 		}
 	}
@@ -934,19 +934,19 @@ xfs_dir2_sf_removename(
 	ASSERT(dp->i_df.if_bytes == oldsize);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(oldsize >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
+	ASSERT(oldsize >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
 	/*
 	 * Loop over the old directory entries.
 	 * Find the one we're deleting.
 	 */
-	for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
 	     i < sfp->hdr.count;
-	     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep)) {
+	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 		if (sfep->namelen == args->namelen &&
 		    sfep->name[0] == args->name[0] &&
 		    memcmp(sfep->name, args->name, args->namelen) == 0) {
-			ASSERT(XFS_DIR2_SF_GET_INUMBER(sfp,
-					XFS_DIR2_SF_INUMBERP(sfep)) ==
+			ASSERT(xfs_dir2_sf_get_inumber(sfp,
+					xfs_dir2_sf_inumberp(sfep)) ==
 				args->inumber);
 			break;
 		}
@@ -961,7 +961,7 @@ xfs_dir2_sf_removename(
 	 * Calculate sizes.
 	 */
 	byteoff = (int)((char *)sfep - (char *)sfp);
-	entsize = XFS_DIR2_SF_ENTSIZE_BYNAME(sfp, args->namelen);
+	entsize = xfs_dir2_sf_entsize_byname(sfp, args->namelen);
 	newsize = oldsize - entsize;
 	/*
 	 * Copy the part if any after the removed entry, sliding it down.
@@ -1027,7 +1027,7 @@ xfs_dir2_sf_replace(
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
 #if XFS_BIG_INUMS
 	/*
 	 * New inode number is large, and need to convert to 8-byte inodes.
@@ -1067,28 +1067,28 @@ xfs_dir2_sf_replace(
 	if (args->namelen == 2 &&
 	    args->name[0] == '.' && args->name[1] == '.') {
 #if XFS_BIG_INUMS || defined(DEBUG)
-		ino = XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent);
+		ino = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
 		ASSERT(args->inumber != ino);
 #endif
-		XFS_DIR2_SF_PUT_INUMBER(sfp, &args->inumber, &sfp->hdr.parent);
+		xfs_dir2_sf_put_inumber(sfp, &args->inumber, &sfp->hdr.parent);
 	}
 	/*
 	 * Normal entry, look for the name.
 	 */
 	else {
-		for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
+		for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
 		     i < sfp->hdr.count;
-		     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep)) {
+		     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 			if (sfep->namelen == args->namelen &&
 			    sfep->name[0] == args->name[0] &&
 			    memcmp(args->name, sfep->name, args->namelen) == 0) {
 #if XFS_BIG_INUMS || defined(DEBUG)
-				ino = XFS_DIR2_SF_GET_INUMBER(sfp,
-					XFS_DIR2_SF_INUMBERP(sfep));
+				ino = xfs_dir2_sf_get_inumber(sfp,
+					xfs_dir2_sf_inumberp(sfep));
 				ASSERT(args->inumber != ino);
 #endif
-				XFS_DIR2_SF_PUT_INUMBER(sfp, &args->inumber,
-					XFS_DIR2_SF_INUMBERP(sfep));
+				xfs_dir2_sf_put_inumber(sfp, &args->inumber,
+					xfs_dir2_sf_inumberp(sfep));
 				break;
 			}
 		}
@@ -1189,22 +1189,22 @@ xfs_dir2_sf_toino4(
 	 */
 	sfp->hdr.count = oldsfp->hdr.count;
 	sfp->hdr.i8count = 0;
-	ino = XFS_DIR2_SF_GET_INUMBER(oldsfp, &oldsfp->hdr.parent);
-	XFS_DIR2_SF_PUT_INUMBER(sfp, &ino, &sfp->hdr.parent);
+	ino = xfs_dir2_sf_get_inumber(oldsfp, &oldsfp->hdr.parent);
+	xfs_dir2_sf_put_inumber(sfp, &ino, &sfp->hdr.parent);
 	/*
 	 * Copy the entries field by field.
 	 */
-	for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp),
-		    oldsfep = XFS_DIR2_SF_FIRSTENTRY(oldsfp);
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp),
+		    oldsfep = xfs_dir2_sf_firstentry(oldsfp);
 	     i < sfp->hdr.count;
-	     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep),
-		  oldsfep = XFS_DIR2_SF_NEXTENTRY(oldsfp, oldsfep)) {
+	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep),
+		  oldsfep = xfs_dir2_sf_nextentry(oldsfp, oldsfep)) {
 		sfep->namelen = oldsfep->namelen;
 		sfep->offset = oldsfep->offset;
 		memcpy(sfep->name, oldsfep->name, sfep->namelen);
-		ino = XFS_DIR2_SF_GET_INUMBER(oldsfp,
-			XFS_DIR2_SF_INUMBERP(oldsfep));
-		XFS_DIR2_SF_PUT_INUMBER(sfp, &ino, XFS_DIR2_SF_INUMBERP(sfep));
+		ino = xfs_dir2_sf_get_inumber(oldsfp,
+			xfs_dir2_sf_inumberp(oldsfep));
+		xfs_dir2_sf_put_inumber(sfp, &ino, xfs_dir2_sf_inumberp(sfep));
 	}
 	/*
 	 * Clean up the inode.
@@ -1266,22 +1266,22 @@ xfs_dir2_sf_toino8(
 	 */
 	sfp->hdr.count = oldsfp->hdr.count;
 	sfp->hdr.i8count = 1;
-	ino = XFS_DIR2_SF_GET_INUMBER(oldsfp, &oldsfp->hdr.parent);
-	XFS_DIR2_SF_PUT_INUMBER(sfp, &ino, &sfp->hdr.parent);
+	ino = xfs_dir2_sf_get_inumber(oldsfp, &oldsfp->hdr.parent);
+	xfs_dir2_sf_put_inumber(sfp, &ino, &sfp->hdr.parent);
 	/*
 	 * Copy the entries field by field.
 	 */
-	for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp),
-		    oldsfep = XFS_DIR2_SF_FIRSTENTRY(oldsfp);
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp),
+		    oldsfep = xfs_dir2_sf_firstentry(oldsfp);
 	     i < sfp->hdr.count;
-	     i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep),
-		  oldsfep = XFS_DIR2_SF_NEXTENTRY(oldsfp, oldsfep)) {
+	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep),
+		  oldsfep = xfs_dir2_sf_nextentry(oldsfp, oldsfep)) {
 		sfep->namelen = oldsfep->namelen;
 		sfep->offset = oldsfep->offset;
 		memcpy(sfep->name, oldsfep->name, sfep->namelen);
-		ino = XFS_DIR2_SF_GET_INUMBER(oldsfp,
-			XFS_DIR2_SF_INUMBERP(oldsfep));
-		XFS_DIR2_SF_PUT_INUMBER(sfp, &ino, XFS_DIR2_SF_INUMBERP(sfep));
+		ino = xfs_dir2_sf_get_inumber(oldsfp,
+			xfs_dir2_sf_inumberp(oldsfep));
+		xfs_dir2_sf_put_inumber(sfp, &ino, xfs_dir2_sf_inumberp(sfep));
 	}
 	/*
 	 * Clean up the inode.

fs/xfs/xfs_dir2_sf.h

@@ -90,7 +90,6 @@ typedef struct xfs_dir2_sf {
 	xfs_dir2_sf_entry_t	list[1];	/* shortform entries */
 } xfs_dir2_sf_t;
 
-#define	XFS_DIR2_SF_HDR_SIZE(i8count)	xfs_dir2_sf_hdr_size(i8count)
 static inline int xfs_dir2_sf_hdr_size(int i8count)
 {
 	return ((uint)sizeof(xfs_dir2_sf_hdr_t) - \
@@ -98,14 +97,11 @@ static inline int xfs_dir2_sf_hdr_size(int i8count)
 		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t)));
 }
 
-#define	XFS_DIR2_SF_INUMBERP(sfep)	xfs_dir2_sf_inumberp(sfep)
 static inline xfs_dir2_inou_t *xfs_dir2_sf_inumberp(xfs_dir2_sf_entry_t *sfep)
 {
 	return (xfs_dir2_inou_t *)&(sfep)->name[(sfep)->namelen];
 }
 
-#define	XFS_DIR2_SF_GET_INUMBER(sfp, from) \
-	xfs_dir2_sf_get_inumber(sfp, from)
 static inline xfs_intino_t
 xfs_dir2_sf_get_inumber(xfs_dir2_sf_t *sfp, xfs_dir2_inou_t *from)
 {
@@ -114,8 +110,6 @@ xfs_dir2_sf_get_inumber(xfs_dir2_sf_t *sfp, xfs_dir2_inou_t *from)
 		(xfs_intino_t)XFS_GET_DIR_INO8((from)->i8));
 }
 
-#define	XFS_DIR2_SF_PUT_INUMBER(sfp,from,to) \
-	xfs_dir2_sf_put_inumber(sfp,from,to)
 static inline void xfs_dir2_sf_put_inumber(xfs_dir2_sf_t *sfp, xfs_ino_t *from,
 						xfs_dir2_inou_t *to)
 {
@@ -125,24 +119,18 @@ static inline void xfs_dir2_sf_put_inumber(xfs_dir2_sf_t *sfp, xfs_ino_t *from,
 		XFS_PUT_DIR_INO8(*(from), (to)->i8);
 }
 
-#define	XFS_DIR2_SF_GET_OFFSET(sfep)	\
-	xfs_dir2_sf_get_offset(sfep)
 static inline xfs_dir2_data_aoff_t
 xfs_dir2_sf_get_offset(xfs_dir2_sf_entry_t *sfep)
 {
 	return INT_GET_UNALIGNED_16_BE(&(sfep)->offset.i);
 }
 
-#define	XFS_DIR2_SF_PUT_OFFSET(sfep,off) \
-	xfs_dir2_sf_put_offset(sfep,off)
 static inline void
 xfs_dir2_sf_put_offset(xfs_dir2_sf_entry_t *sfep, xfs_dir2_data_aoff_t off)
 {
 	INT_SET_UNALIGNED_16_BE(&(sfep)->offset.i, off);
 }
 
-#define XFS_DIR2_SF_ENTSIZE_BYNAME(sfp,len)	\
-	xfs_dir2_sf_entsize_byname(sfp,len)
 static inline int xfs_dir2_sf_entsize_byname(xfs_dir2_sf_t *sfp, int len)
 {
 	return ((uint)sizeof(xfs_dir2_sf_entry_t) - 1 + (len) - \
@@ -150,8 +138,6 @@ static inline int xfs_dir2_sf_entsize_byname(xfs_dir2_sf_t *sfp, int len)
 		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t)));
 }
 
-#define XFS_DIR2_SF_ENTSIZE_BYENTRY(sfp,sfep)	\
-	xfs_dir2_sf_entsize_byentry(sfp,sfep)
 static inline int
 xfs_dir2_sf_entsize_byentry(xfs_dir2_sf_t *sfp, xfs_dir2_sf_entry_t *sfep)
 {
@@ -160,19 +146,17 @@ xfs_dir2_sf_entsize_byentry(xfs_dir2_sf_t *sfp, xfs_dir2_sf_entry_t *sfep)
 		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t)));
 }
 
-#define XFS_DIR2_SF_FIRSTENTRY(sfp)	xfs_dir2_sf_firstentry(sfp)
 static inline xfs_dir2_sf_entry_t *xfs_dir2_sf_firstentry(xfs_dir2_sf_t *sfp)
 {
 	return ((xfs_dir2_sf_entry_t *) \
-		((char *)(sfp) + XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count)));
+		((char *)(sfp) + xfs_dir2_sf_hdr_size(sfp->hdr.i8count)));
 }
 
-#define XFS_DIR2_SF_NEXTENTRY(sfp,sfep)	xfs_dir2_sf_nextentry(sfp,sfep)
 static inline xfs_dir2_sf_entry_t *
 xfs_dir2_sf_nextentry(xfs_dir2_sf_t *sfp, xfs_dir2_sf_entry_t *sfep)
 {
 	return ((xfs_dir2_sf_entry_t *) \
-		((char *)(sfep) + XFS_DIR2_SF_ENTSIZE_BYENTRY(sfp,sfep)));
+		((char *)(sfep) + xfs_dir2_sf_entsize_byentry(sfp,sfep)));
 }
 
 /*

fs/xfs/xfs_filestream.c

@@ -0,0 +1,771 @@
+/*
+ * Copyright (c) 2006-2007 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+#include "xfs.h"
+#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"
+#include "xfs_ag.h"
+#include "xfs_dmapi.h"
+#include "xfs_log.h"
+#include "xfs_trans.h"
+#include "xfs_sb.h"
+#include "xfs_mount.h"
+#include "xfs_bmap.h"
+#include "xfs_alloc.h"
+#include "xfs_utils.h"
+#include "xfs_mru_cache.h"
+#include "xfs_filestream.h"
+
+#ifdef XFS_FILESTREAMS_TRACE
+
+ktrace_t *xfs_filestreams_trace_buf;
+
+STATIC void
+xfs_filestreams_trace(
+	xfs_mount_t	*mp,	/* mount point */
+	int		type,	/* type of trace */
+	const char	*func,	/* source function */
+	int		line,	/* source line number */
+	__psunsigned_t	arg0,
+	__psunsigned_t	arg1,
+	__psunsigned_t	arg2,
+	__psunsigned_t	arg3,
+	__psunsigned_t	arg4,
+	__psunsigned_t	arg5)
+{
+	ktrace_enter(xfs_filestreams_trace_buf,
+		(void *)(__psint_t)(type | (line << 16)),
+		(void *)func,
+		(void *)(__psunsigned_t)current_pid(),
+		(void *)mp,
+		(void *)(__psunsigned_t)arg0,
+		(void *)(__psunsigned_t)arg1,
+		(void *)(__psunsigned_t)arg2,
+		(void *)(__psunsigned_t)arg3,
+		(void *)(__psunsigned_t)arg4,
+		(void *)(__psunsigned_t)arg5,
+		NULL, NULL, NULL, NULL, NULL, NULL);
+}
+
+#define TRACE0(mp,t)			TRACE6(mp,t,0,0,0,0,0,0)
+#define TRACE1(mp,t,a0)			TRACE6(mp,t,a0,0,0,0,0,0)
+#define TRACE2(mp,t,a0,a1)		TRACE6(mp,t,a0,a1,0,0,0,0)
+#define TRACE3(mp,t,a0,a1,a2)		TRACE6(mp,t,a0,a1,a2,0,0,0)
+#define TRACE4(mp,t,a0,a1,a2,a3)	TRACE6(mp,t,a0,a1,a2,a3,0,0)
+#define TRACE5(mp,t,a0,a1,a2,a3,a4)	TRACE6(mp,t,a0,a1,a2,a3,a4,0)
+#define TRACE6(mp,t,a0,a1,a2,a3,a4,a5) \
+	xfs_filestreams_trace(mp, t, __FUNCTION__, __LINE__, \
+				(__psunsigned_t)a0, (__psunsigned_t)a1, \
+				(__psunsigned_t)a2, (__psunsigned_t)a3, \
+				(__psunsigned_t)a4, (__psunsigned_t)a5)
+
+#define TRACE_AG_SCAN(mp, ag, ag2) \
+		TRACE2(mp, XFS_FSTRM_KTRACE_AGSCAN, ag, ag2);
+#define TRACE_AG_PICK1(mp, max_ag, maxfree) \
+		TRACE2(mp, XFS_FSTRM_KTRACE_AGPICK1, max_ag, maxfree);
+#define TRACE_AG_PICK2(mp, ag, ag2, cnt, free, scan, flag) \
+		TRACE6(mp, XFS_FSTRM_KTRACE_AGPICK2, ag, ag2, \
+			 cnt, free, scan, flag)
+#define TRACE_UPDATE(mp, ip, ag, cnt, ag2, cnt2) \
+		TRACE5(mp, XFS_FSTRM_KTRACE_UPDATE, ip, ag, cnt, ag2, cnt2)
+#define TRACE_FREE(mp, ip, pip, ag, cnt) \
+		TRACE4(mp, XFS_FSTRM_KTRACE_FREE, ip, pip, ag, cnt)
+#define TRACE_LOOKUP(mp, ip, pip, ag, cnt) \
+		TRACE4(mp, XFS_FSTRM_KTRACE_ITEM_LOOKUP, ip, pip, ag, cnt)
+#define TRACE_ASSOCIATE(mp, ip, pip, ag, cnt) \
+		TRACE4(mp, XFS_FSTRM_KTRACE_ASSOCIATE, ip, pip, ag, cnt)
+#define TRACE_MOVEAG(mp, ip, pip, oag, ocnt, nag, ncnt) \
+		TRACE6(mp, XFS_FSTRM_KTRACE_MOVEAG, ip, pip, oag, ocnt, nag, ncnt)
+#define TRACE_ORPHAN(mp, ip, ag) \
+		TRACE2(mp, XFS_FSTRM_KTRACE_ORPHAN, ip, ag);
+
+
+#else
+#define TRACE_AG_SCAN(mp, ag, ag2)
+#define TRACE_AG_PICK1(mp, max_ag, maxfree)
+#define TRACE_AG_PICK2(mp, ag, ag2, cnt, free, scan, flag)
+#define TRACE_UPDATE(mp, ip, ag, cnt, ag2, cnt2)
+#define TRACE_FREE(mp, ip, pip, ag, cnt)
+#define TRACE_LOOKUP(mp, ip, pip, ag, cnt)
+#define TRACE_ASSOCIATE(mp, ip, pip, ag, cnt)
+#define TRACE_MOVEAG(mp, ip, pip, oag, ocnt, nag, ncnt)
+#define TRACE_ORPHAN(mp, ip, ag)
+#endif
+
+static kmem_zone_t *item_zone;
+
+/*
+ * Structure for associating a file or a directory with an allocation group.
+ * The parent directory pointer is only needed for files, but since there will
+ * generally be vastly more files than directories in the cache, using the same
+ * data structure simplifies the code with very little memory overhead.
+ */
+typedef struct fstrm_item
+{
+	xfs_agnumber_t	ag;	/* AG currently in use for the file/directory. */
+	xfs_inode_t	*ip;	/* inode self-pointer. */
+	xfs_inode_t	*pip;	/* Parent directory inode pointer. */
+} fstrm_item_t;
+
+
+/*
+ * Scan the AGs starting at startag looking for an AG that isn't in use and has
+ * at least minlen blocks free.
+ */
+static int
+_xfs_filestream_pick_ag(
+	xfs_mount_t	*mp,
+	xfs_agnumber_t	startag,
+	xfs_agnumber_t	*agp,
+	int		flags,
+	xfs_extlen_t	minlen)
+{
+	int		err, trylock, nscan;
+	xfs_extlen_t	delta, longest, need, free, minfree, maxfree = 0;
+	xfs_agnumber_t	ag, max_ag = NULLAGNUMBER;
+	struct xfs_perag *pag;
+
+	/* 2% of an AG's blocks must be free for it to be chosen. */
+	minfree = mp->m_sb.sb_agblocks / 50;
+
+	ag = startag;
+	*agp = NULLAGNUMBER;
+
+	/* For the first pass, don't sleep trying to init the per-AG. */
+	trylock = XFS_ALLOC_FLAG_TRYLOCK;
+
+	for (nscan = 0; 1; nscan++) {
+
+		TRACE_AG_SCAN(mp, ag, xfs_filestream_peek_ag(mp, ag));
+
+		pag = mp->m_perag + ag;
+
+		if (!pag->pagf_init) {
+			err = xfs_alloc_pagf_init(mp, NULL, ag, trylock);
+			if (err && !trylock)
+				return err;
+		}
+
+		/* Might fail sometimes during the 1st pass with trylock set. */
+		if (!pag->pagf_init)
+			goto next_ag;
+
+		/* Keep track of the AG with the most free blocks. */
+		if (pag->pagf_freeblks > maxfree) {
+			maxfree = pag->pagf_freeblks;
+			max_ag = ag;
+		}
+
+		/*
+		 * The AG reference count does two things: it enforces mutual
+		 * exclusion when examining the suitability of an AG in this
+		 * loop, and it guards against two filestreams being established
+		 * in the same AG as each other.
+		 */
+		if (xfs_filestream_get_ag(mp, ag) > 1) {
+			xfs_filestream_put_ag(mp, ag);
+			goto next_ag;
+		}
+
+		need = XFS_MIN_FREELIST_PAG(pag, mp);
+		delta = need > pag->pagf_flcount ? need - pag->pagf_flcount : 0;
+		longest = (pag->pagf_longest > delta) ?
+		          (pag->pagf_longest - delta) :
+		          (pag->pagf_flcount > 0 || pag->pagf_longest > 0);
+
+		if (((minlen && longest >= minlen) ||
+		     (!minlen && pag->pagf_freeblks >= minfree)) &&
+		    (!pag->pagf_metadata || !(flags & XFS_PICK_USERDATA) ||
+		     (flags & XFS_PICK_LOWSPACE))) {
+
+			/* Break out, retaining the reference on the AG. */
+			free = pag->pagf_freeblks;
+			*agp = ag;
+			break;
+		}
+
+		/* Drop the reference on this AG, it's not usable. */
+		xfs_filestream_put_ag(mp, ag);
+next_ag:
+		/* Move to the next AG, wrapping to AG 0 if necessary. */
+		if (++ag >= mp->m_sb.sb_agcount)
+			ag = 0;
+
+		/* If a full pass of the AGs hasn't been done yet, continue. */
+		if (ag != startag)
+			continue;
+
+		/* Allow sleeping in xfs_alloc_pagf_init() on the 2nd pass. */
+		if (trylock != 0) {
+			trylock = 0;
+			continue;
+		}
+
+		/* Finally, if lowspace wasn't set, set it for the 3rd pass. */
+		if (!(flags & XFS_PICK_LOWSPACE)) {
+			flags |= XFS_PICK_LOWSPACE;
+			continue;
+		}
+
+		/*
+		 * Take the AG with the most free space, regardless of whether
+		 * it's already in use by another filestream.
+		 */
+		if (max_ag != NULLAGNUMBER) {
+			xfs_filestream_get_ag(mp, max_ag);
+			TRACE_AG_PICK1(mp, max_ag, maxfree);
+			free = maxfree;
+			*agp = max_ag;
+			break;
+		}
+
+		/* take AG 0 if none matched */
+		TRACE_AG_PICK1(mp, max_ag, maxfree);
+		*agp = 0;
+		return 0;
+	}
+
+	TRACE_AG_PICK2(mp, startag, *agp, xfs_filestream_peek_ag(mp, *agp),
+			free, nscan, flags);
+
+	return 0;
+}
+
+/*
+ * Set the allocation group number for a file or a directory, updating inode
+ * references and per-AG references as appropriate.  Must be called with the
+ * m_peraglock held in read mode.
+ */
+static int
+_xfs_filestream_update_ag(
+	xfs_inode_t	*ip,
+	xfs_inode_t	*pip,
+	xfs_agnumber_t	ag)
+{
+	int		err = 0;
+	xfs_mount_t	*mp;
+	xfs_mru_cache_t	*cache;
+	fstrm_item_t	*item;
+	xfs_agnumber_t	old_ag;
+	xfs_inode_t	*old_pip;
+
+	/*
+	 * Either ip is a regular file and pip is a directory, or ip is a
+	 * directory and pip is NULL.
+	 */
+	ASSERT(ip && (((ip->i_d.di_mode & S_IFREG) && pip &&
+	               (pip->i_d.di_mode & S_IFDIR)) ||
+	              ((ip->i_d.di_mode & S_IFDIR) && !pip)));
+
+	mp = ip->i_mount;
+	cache = mp->m_filestream;
+
+	item = xfs_mru_cache_lookup(cache, ip->i_ino);
+	if (item) {
+		ASSERT(item->ip == ip);
+		old_ag = item->ag;
+		item->ag = ag;
+		old_pip = item->pip;
+		item->pip = pip;
+		xfs_mru_cache_done(cache);
+
+		/*
+		 * If the AG has changed, drop the old ref and take a new one,
+		 * effectively transferring the reference from old to new AG.
+		 */
+		if (ag != old_ag) {
+			xfs_filestream_put_ag(mp, old_ag);
+			xfs_filestream_get_ag(mp, ag);
+		}
+
+		/*
+		 * If ip is a file and its pip has changed, drop the old ref and
+		 * take a new one.
+		 */
+		if (pip && pip != old_pip) {
+			IRELE(old_pip);
+			IHOLD(pip);
+		}
+
+		TRACE_UPDATE(mp, ip, old_ag, xfs_filestream_peek_ag(mp, old_ag),
+				ag, xfs_filestream_peek_ag(mp, ag));
+		return 0;
+	}
+
+	item = kmem_zone_zalloc(item_zone, KM_MAYFAIL);
+	if (!item)
+		return ENOMEM;
+
+	item->ag = ag;
+	item->ip = ip;
+	item->pip = pip;
+
+	err = xfs_mru_cache_insert(cache, ip->i_ino, item);
+	if (err) {
+		kmem_zone_free(item_zone, item);
+		return err;
+	}
+
+	/* Take a reference on the AG. */
+	xfs_filestream_get_ag(mp, ag);
+
+	/*
+	 * Take a reference on the inode itself regardless of whether it's a
+	 * regular file or a directory.
+	 */
+	IHOLD(ip);
+
+	/*
+	 * In the case of a regular file, take a reference on the parent inode
+	 * as well to ensure it remains in-core.
+	 */
+	if (pip)
+		IHOLD(pip);
+
+	TRACE_UPDATE(mp, ip, ag, xfs_filestream_peek_ag(mp, ag),
+			ag, xfs_filestream_peek_ag(mp, ag));
+
+	return 0;
+}
+
+/* xfs_fstrm_free_func(): callback for freeing cached stream items. */
+void
+xfs_fstrm_free_func(
+	xfs_ino_t	ino,
+	fstrm_item_t	*item)
+{
+	xfs_inode_t	*ip = item->ip;
+	int ref;
+
+	ASSERT(ip->i_ino == ino);
+
+	xfs_iflags_clear(ip, XFS_IFILESTREAM);
+
+	/* Drop the reference taken on the AG when the item was added. */
+	ref = xfs_filestream_put_ag(ip->i_mount, item->ag);
+
+	ASSERT(ref >= 0);
+	TRACE_FREE(ip->i_mount, ip, item->pip, item->ag,
+		xfs_filestream_peek_ag(ip->i_mount, item->ag));
+
+	/*
+	 * _xfs_filestream_update_ag() always takes a reference on the inode
+	 * itself, whether it's a file or a directory.  Release it here.
+	 * This can result in the inode being freed and so we must
+	 * not hold any inode locks when freeing filesstreams objects
+	 * otherwise we can deadlock here.
+	 */
+	IRELE(ip);
+
+	/*
+	 * In the case of a regular file, _xfs_filestream_update_ag() also
+	 * takes a ref on the parent inode to keep it in-core.  Release that
+	 * too.
+	 */
+	if (item->pip)
+		IRELE(item->pip);
+
+	/* Finally, free the memory allocated for the item. */
+	kmem_zone_free(item_zone, item);
+}
+
+/*
+ * xfs_filestream_init() is called at xfs initialisation time to set up the
+ * memory zone that will be used for filestream data structure allocation.
+ */
+int
+xfs_filestream_init(void)
+{
+	item_zone = kmem_zone_init(sizeof(fstrm_item_t), "fstrm_item");
+#ifdef XFS_FILESTREAMS_TRACE
+	xfs_filestreams_trace_buf = ktrace_alloc(XFS_FSTRM_KTRACE_SIZE, KM_SLEEP);
+#endif
+	return item_zone ? 0 : -ENOMEM;
+}
+
+/*
+ * xfs_filestream_uninit() is called at xfs termination time to destroy the
+ * memory zone that was used for filestream data structure allocation.
+ */
+void
+xfs_filestream_uninit(void)
+{
+#ifdef XFS_FILESTREAMS_TRACE
+	ktrace_free(xfs_filestreams_trace_buf);
+#endif
+	kmem_zone_destroy(item_zone);
+}
+
+/*
+ * xfs_filestream_mount() is called when a file system is mounted with the
+ * filestream option.  It is responsible for allocating the data structures
+ * needed to track the new file system's file streams.
+ */
+int
+xfs_filestream_mount(
+	xfs_mount_t	*mp)
+{
+	int		err;
+	unsigned int	lifetime, grp_count;
+
+	/*
+	 * The filestream timer tunable is currently fixed within the range of
+	 * one second to four minutes, with five seconds being the default.  The
+	 * group count is somewhat arbitrary, but it'd be nice to adhere to the
+	 * timer tunable to within about 10 percent.  This requires at least 10
+	 * groups.
+	 */
+	lifetime  = xfs_fstrm_centisecs * 10;
+	grp_count = 10;
+
+	err = xfs_mru_cache_create(&mp->m_filestream, lifetime, grp_count,
+	                     (xfs_mru_cache_free_func_t)xfs_fstrm_free_func);
+
+	return err;
+}
+
+/*
+ * xfs_filestream_unmount() is called when a file system that was mounted with
+ * the filestream option is unmounted.  It drains the data structures created
+ * to track the file system's file streams and frees all the memory that was
+ * allocated.
+ */
+void
+xfs_filestream_unmount(
+	xfs_mount_t	*mp)
+{
+	xfs_mru_cache_destroy(mp->m_filestream);
+}
+
+/*
+ * If the mount point's m_perag array is going to be reallocated, all
+ * outstanding cache entries must be flushed to avoid accessing reference count
+ * addresses that have been freed.  The call to xfs_filestream_flush() must be
+ * made inside the block that holds the m_peraglock in write mode to do the
+ * reallocation.
+ */
+void
+xfs_filestream_flush(
+	xfs_mount_t	*mp)
+{
+	/* point in time flush, so keep the reaper running */
+	xfs_mru_cache_flush(mp->m_filestream, 1);
+}
+
+/*
+ * Return the AG of the filestream the file or directory belongs to, or
+ * NULLAGNUMBER otherwise.
+ */
+xfs_agnumber_t
+xfs_filestream_lookup_ag(
+	xfs_inode_t	*ip)
+{
+	xfs_mru_cache_t	*cache;
+	fstrm_item_t	*item;
+	xfs_agnumber_t	ag;
+	int		ref;
+
+	if (!(ip->i_d.di_mode & (S_IFREG | S_IFDIR))) {
+		ASSERT(0);
+		return NULLAGNUMBER;
+	}
+
+	cache = ip->i_mount->m_filestream;
+	item = xfs_mru_cache_lookup(cache, ip->i_ino);
+	if (!item) {
+		TRACE_LOOKUP(ip->i_mount, ip, NULL, NULLAGNUMBER, 0);
+		return NULLAGNUMBER;
+	}
+
+	ASSERT(ip == item->ip);
+	ag = item->ag;
+	ref = xfs_filestream_peek_ag(ip->i_mount, ag);
+	xfs_mru_cache_done(cache);
+
+	TRACE_LOOKUP(ip->i_mount, ip, item->pip, ag, ref);
+	return ag;
+}
+
+/*
+ * xfs_filestream_associate() should only be called to associate a regular file
+ * with its parent directory.  Calling it with a child directory isn't
+ * appropriate because filestreams don't apply to entire directory hierarchies.
+ * Creating a file in a child directory of an existing filestream directory
+ * starts a new filestream with its own allocation group association.
+ *
+ * Returns < 0 on error, 0 if successful association occurred, > 0 if
+ * we failed to get an association because of locking issues.
+ */
+int
+xfs_filestream_associate(
+	xfs_inode_t	*pip,
+	xfs_inode_t	*ip)
+{
+	xfs_mount_t	*mp;
+	xfs_mru_cache_t	*cache;
+	fstrm_item_t	*item;
+	xfs_agnumber_t	ag, rotorstep, startag;
+	int		err = 0;
+
+	ASSERT(pip->i_d.di_mode & S_IFDIR);
+	ASSERT(ip->i_d.di_mode & S_IFREG);
+	if (!(pip->i_d.di_mode & S_IFDIR) || !(ip->i_d.di_mode & S_IFREG))
+		return -EINVAL;
+
+	mp = pip->i_mount;
+	cache = mp->m_filestream;
+	down_read(&mp->m_peraglock);
+
+	/*
+	 * We have a problem, Houston.
+	 *
+	 * Taking the iolock here violates inode locking order - we already
+	 * hold the ilock. Hence if we block getting this lock we may never
+	 * wake. Unfortunately, that means if we can't get the lock, we're
+	 * screwed in terms of getting a stream association - we can't spin
+	 * waiting for the lock because someone else is waiting on the lock we
+	 * hold and we cannot drop that as we are in a transaction here.
+	 *
+	 * Lucky for us, this inversion is rarely a problem because it's a
+	 * directory inode that we are trying to lock here and that means the
+	 * only place that matters is xfs_sync_inodes() and SYNC_DELWRI is
+	 * used. i.e. freeze, remount-ro, quotasync or unmount.
+	 *
+	 * So, if we can't get the iolock without sleeping then just give up
+	 */
+	if (!xfs_ilock_nowait(pip, XFS_IOLOCK_EXCL)) {
+		up_read(&mp->m_peraglock);
+		return 1;
+	}
+
+	/* If the parent directory is already in the cache, use its AG. */
+	item = xfs_mru_cache_lookup(cache, pip->i_ino);
+	if (item) {
+		ASSERT(item->ip == pip);
+		ag = item->ag;
+		xfs_mru_cache_done(cache);
+
+		TRACE_LOOKUP(mp, pip, pip, ag, xfs_filestream_peek_ag(mp, ag));
+		err = _xfs_filestream_update_ag(ip, pip, ag);
+
+		goto exit;
+	}
+
+	/*
+	 * Set the starting AG using the rotor for inode32, otherwise
+	 * use the directory inode's AG.
+	 */
+	if (mp->m_flags & XFS_MOUNT_32BITINODES) {
+		rotorstep = xfs_rotorstep;
+		startag = (mp->m_agfrotor / rotorstep) % mp->m_sb.sb_agcount;
+		mp->m_agfrotor = (mp->m_agfrotor + 1) %
+		                 (mp->m_sb.sb_agcount * rotorstep);
+	} else
+		startag = XFS_INO_TO_AGNO(mp, pip->i_ino);
+
+	/* Pick a new AG for the parent inode starting at startag. */
+	err = _xfs_filestream_pick_ag(mp, startag, &ag, 0, 0);
+	if (err || ag == NULLAGNUMBER)
+		goto exit_did_pick;
+
+	/* Associate the parent inode with the AG. */
+	err = _xfs_filestream_update_ag(pip, NULL, ag);
+	if (err)
+		goto exit_did_pick;
+
+	/* Associate the file inode with the AG. */
+	err = _xfs_filestream_update_ag(ip, pip, ag);
+	if (err)
+		goto exit_did_pick;
+
+	TRACE_ASSOCIATE(mp, ip, pip, ag, xfs_filestream_peek_ag(mp, ag));
+
+exit_did_pick:
+	/*
+	 * If _xfs_filestream_pick_ag() returned a valid AG, remove the
+	 * reference it took on it, since the file and directory will have taken
+	 * their own now if they were successfully cached.
+	 */
+	if (ag != NULLAGNUMBER)
+		xfs_filestream_put_ag(mp, ag);
+
+exit:
+	xfs_iunlock(pip, XFS_IOLOCK_EXCL);
+	up_read(&mp->m_peraglock);
+	return -err;
+}
+
+/*
+ * Pick a new allocation group for the current file and its file stream.  This
+ * function is called by xfs_bmap_filestreams() with the mount point's per-ag
+ * lock held.
+ */
+int
+xfs_filestream_new_ag(
+	xfs_bmalloca_t	*ap,
+	xfs_agnumber_t	*agp)
+{
+	int		flags, err;
+	xfs_inode_t	*ip, *pip = NULL;
+	xfs_mount_t	*mp;
+	xfs_mru_cache_t	*cache;
+	xfs_extlen_t	minlen;
+	fstrm_item_t	*dir, *file;
+	xfs_agnumber_t	ag = NULLAGNUMBER;
+
+	ip = ap->ip;
+	mp = ip->i_mount;
+	cache = mp->m_filestream;
+	minlen = ap->alen;
+	*agp = NULLAGNUMBER;
+
+	/*
+	 * Look for the file in the cache, removing it if it's found.  Doing
+	 * this allows it to be held across the dir lookup that follows.
+	 */
+	file = xfs_mru_cache_remove(cache, ip->i_ino);
+	if (file) {
+		ASSERT(ip == file->ip);
+
+		/* Save the file's parent inode and old AG number for later. */
+		pip = file->pip;
+		ag = file->ag;
+
+		/* Look for the file's directory in the cache. */
+		dir = xfs_mru_cache_lookup(cache, pip->i_ino);
+		if (dir) {
+			ASSERT(pip == dir->ip);
+
+			/*
+			 * If the directory has already moved on to a new AG,
+			 * use that AG as the new AG for the file. Don't
+			 * forget to twiddle the AG refcounts to match the
+			 * movement.
+			 */
+			if (dir->ag != file->ag) {
+				xfs_filestream_put_ag(mp, file->ag);
+				xfs_filestream_get_ag(mp, dir->ag);
+				*agp = file->ag = dir->ag;
+			}
+
+			xfs_mru_cache_done(cache);
+		}
+
+		/*
+		 * Put the file back in the cache.  If this fails, the free
+		 * function needs to be called to tidy up in the same way as if
+		 * the item had simply expired from the cache.
+		 */
+		err = xfs_mru_cache_insert(cache, ip->i_ino, file);
+		if (err) {
+			xfs_fstrm_free_func(ip->i_ino, file);
+			return err;
+		}
+
+		/*
+		 * If the file's AG was moved to the directory's new AG, there's
+		 * nothing more to be done.
+		 */
+		if (*agp != NULLAGNUMBER) {
+			TRACE_MOVEAG(mp, ip, pip,
+					ag, xfs_filestream_peek_ag(mp, ag),
+					*agp, xfs_filestream_peek_ag(mp, *agp));
+			return 0;
+		}
+	}
+
+	/*
+	 * If the file's parent directory is known, take its iolock in exclusive
+	 * mode to prevent two sibling files from racing each other to migrate
+	 * themselves and their parent to different AGs.
+	 */
+	if (pip)
+		xfs_ilock(pip, XFS_IOLOCK_EXCL);
+
+	/*
+	 * A new AG needs to be found for the file.  If the file's parent
+	 * directory is also known, it will be moved to the new AG as well to
+	 * ensure that files created inside it in future use the new AG.
+	 */
+	ag = (ag == NULLAGNUMBER) ? 0 : (ag + 1) % mp->m_sb.sb_agcount;
+	flags = (ap->userdata ? XFS_PICK_USERDATA : 0) |
+	        (ap->low ? XFS_PICK_LOWSPACE : 0);
+
+	err = _xfs_filestream_pick_ag(mp, ag, agp, flags, minlen);
+	if (err || *agp == NULLAGNUMBER)
+		goto exit;
+
+	/*
+	 * If the file wasn't found in the file cache, then its parent directory
+	 * inode isn't known.  For this to have happened, the file must either
+	 * be pre-existing, or it was created long enough ago that its cache
+	 * entry has expired.  This isn't the sort of usage that the filestreams
+	 * allocator is trying to optimise, so there's no point trying to track
+	 * its new AG somehow in the filestream data structures.
+	 */
+	if (!pip) {
+		TRACE_ORPHAN(mp, ip, *agp);
+		goto exit;
+	}
+
+	/* Associate the parent inode with the AG. */
+	err = _xfs_filestream_update_ag(pip, NULL, *agp);
+	if (err)
+		goto exit;
+
+	/* Associate the file inode with the AG. */
+	err = _xfs_filestream_update_ag(ip, pip, *agp);
+	if (err)
+		goto exit;
+
+	TRACE_MOVEAG(mp, ip, pip, NULLAGNUMBER, 0,
+			*agp, xfs_filestream_peek_ag(mp, *agp));
+
+exit:
+	/*
+	 * If _xfs_filestream_pick_ag() returned a valid AG, remove the
+	 * reference it took on it, since the file and directory will have taken
+	 * their own now if they were successfully cached.
+	 */
+	if (*agp != NULLAGNUMBER)
+		xfs_filestream_put_ag(mp, *agp);
+	else
+		*agp = 0;
+
+	if (pip)
+		xfs_iunlock(pip, XFS_IOLOCK_EXCL);
+
+	return err;
+}
+
+/*
+ * Remove an association between an inode and a filestream object.
+ * Typically this is done on last close of an unlinked file.
+ */
+void
+xfs_filestream_deassociate(
+	xfs_inode_t	*ip)
+{
+	xfs_mru_cache_t	*cache = ip->i_mount->m_filestream;
+
+	xfs_mru_cache_delete(cache, ip->i_ino);
+}

fs/xfs/xfs_filestream.h

@@ -0,0 +1,136 @@
+/*
+ * Copyright (c) 2006-2007 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+#ifndef __XFS_FILESTREAM_H__
+#define __XFS_FILESTREAM_H__
+
+#ifdef __KERNEL__
+
+struct xfs_mount;
+struct xfs_inode;
+struct xfs_perag;
+struct xfs_bmalloca;
+
+#ifdef XFS_FILESTREAMS_TRACE
+#define XFS_FSTRM_KTRACE_INFO		1
+#define XFS_FSTRM_KTRACE_AGSCAN		2
+#define XFS_FSTRM_KTRACE_AGPICK1	3
+#define XFS_FSTRM_KTRACE_AGPICK2	4
+#define XFS_FSTRM_KTRACE_UPDATE		5
+#define XFS_FSTRM_KTRACE_FREE		6
+#define	XFS_FSTRM_KTRACE_ITEM_LOOKUP	7
+#define	XFS_FSTRM_KTRACE_ASSOCIATE	8
+#define	XFS_FSTRM_KTRACE_MOVEAG		9
+#define	XFS_FSTRM_KTRACE_ORPHAN		10
+
+#define XFS_FSTRM_KTRACE_SIZE	16384
+extern ktrace_t *xfs_filestreams_trace_buf;
+
+#endif
+
+/*
+ * Allocation group filestream associations are tracked with per-ag atomic
+ * counters.  These counters allow _xfs_filestream_pick_ag() to tell whether a
+ * particular AG already has active filestreams associated with it. The mount
+ * point's m_peraglock is used to protect these counters from per-ag array
+ * re-allocation during a growfs operation.  When xfs_growfs_data_private() is
+ * about to reallocate the array, it calls xfs_filestream_flush() with the
+ * m_peraglock held in write mode.
+ *
+ * Since xfs_mru_cache_flush() guarantees that all the free functions for all
+ * the cache elements have finished executing before it returns, it's safe for
+ * the free functions to use the atomic counters without m_peraglock protection.
+ * This allows the implementation of xfs_fstrm_free_func() to be agnostic about
+ * whether it was called with the m_peraglock held in read mode, write mode or
+ * not held at all.  The race condition this addresses is the following:
+ *
+ *  - The work queue scheduler fires and pulls a filestream directory cache
+ *    element off the LRU end of the cache for deletion, then gets pre-empted.
+ *  - A growfs operation grabs the m_peraglock in write mode, flushes all the
+ *    remaining items from the cache and reallocates the mount point's per-ag
+ *    array, resetting all the counters to zero.
+ *  - The work queue thread resumes and calls the free function for the element
+ *    it started cleaning up earlier.  In the process it decrements the
+ *    filestreams counter for an AG that now has no references.
+ *
+ * With a shrinkfs feature, the above scenario could panic the system.
+ *
+ * All other uses of the following macros should be protected by either the
+ * m_peraglock held in read mode, or the cache's internal locking exposed by the
+ * interval between a call to xfs_mru_cache_lookup() and a call to
+ * xfs_mru_cache_done().  In addition, the m_peraglock must be held in read mode
+ * when new elements are added to the cache.
+ *
+ * Combined, these locking rules ensure that no associations will ever exist in
+ * the cache that reference per-ag array elements that have since been
+ * reallocated.
+ */
+STATIC_INLINE int
+xfs_filestream_peek_ag(
+	xfs_mount_t	*mp,
+	xfs_agnumber_t	agno)
+{
+	return atomic_read(&mp->m_perag[agno].pagf_fstrms);
+}
+
+STATIC_INLINE int
+xfs_filestream_get_ag(
+	xfs_mount_t	*mp,
+	xfs_agnumber_t	agno)
+{
+	return atomic_inc_return(&mp->m_perag[agno].pagf_fstrms);
+}
+
+STATIC_INLINE int
+xfs_filestream_put_ag(
+	xfs_mount_t	*mp,
+	xfs_agnumber_t	agno)
+{
+	return atomic_dec_return(&mp->m_perag[agno].pagf_fstrms);
+}
+
+/* allocation selection flags */
+typedef enum xfs_fstrm_alloc {
+	XFS_PICK_USERDATA = 1,
+	XFS_PICK_LOWSPACE = 2,
+} xfs_fstrm_alloc_t;
+
+/* prototypes for filestream.c */
+int xfs_filestream_init(void);
+void xfs_filestream_uninit(void);
+int xfs_filestream_mount(struct xfs_mount *mp);
+void xfs_filestream_unmount(struct xfs_mount *mp);
+void xfs_filestream_flush(struct xfs_mount *mp);
+xfs_agnumber_t xfs_filestream_lookup_ag(struct xfs_inode *ip);
+int xfs_filestream_associate(struct xfs_inode *dip, struct xfs_inode *ip);
+void xfs_filestream_deassociate(struct xfs_inode *ip);
+int xfs_filestream_new_ag(struct xfs_bmalloca *ap, xfs_agnumber_t *agp);
+
+
+/* filestreams for the inode? */
+STATIC_INLINE int
+xfs_inode_is_filestream(
+	struct xfs_inode	*ip)
+{
+	return (ip->i_mount->m_flags & XFS_MOUNT_FILESTREAMS) ||
+		xfs_iflags_test(ip, XFS_IFILESTREAM) ||
+		(ip->i_d.di_flags & XFS_DIFLAG_FILESTREAM);
+}
+
+#endif /* __KERNEL__ */
+
+#endif /* __XFS_FILESTREAM_H__ */

fs/xfs/xfs_fs.h

@@ -66,6 +66,7 @@ struct fsxattr {
 #define XFS_XFLAG_EXTSIZE	0x00000800	/* extent size allocator hint */
 #define XFS_XFLAG_EXTSZINHERIT	0x00001000	/* inherit inode extent size */
 #define XFS_XFLAG_NODEFRAG	0x00002000  	/* do not defragment */
+#define XFS_XFLAG_FILESTREAM	0x00004000	/* use filestream allocator */
 #define XFS_XFLAG_HASATTR	0x80000000	/* no DIFLAG for this	*/
 
 /*
@@ -238,6 +239,7 @@ typedef struct xfs_fsop_resblks {
 #define XFS_FSOP_GEOM_FLAGS_LOGV2	0x0100	/* log format version 2	*/
 #define XFS_FSOP_GEOM_FLAGS_SECTOR	0x0200	/* sector sizes >1BB	*/
 #define XFS_FSOP_GEOM_FLAGS_ATTR2	0x0400	/* inline attributes rework */
+#define XFS_FSOP_GEOM_FLAGS_LAZYSB	0x4000	/* lazy superblock counters */
 
 
 /*

fs/xfs/xfs_fsops.c

@@ -44,6 +44,7 @@
 #include "xfs_trans_space.h"
 #include "xfs_rtalloc.h"
 #include "xfs_rw.h"
+#include "xfs_filestream.h"
 
 /*
  * File system operations
@@ -94,6 +95,8 @@ xfs_fs_geometry(
 				XFS_FSOP_GEOM_FLAGS_DIRV2 : 0) |
 			(XFS_SB_VERSION_HASSECTOR(&mp->m_sb) ?
 				XFS_FSOP_GEOM_FLAGS_SECTOR : 0) |
+			(xfs_sb_version_haslazysbcount(&mp->m_sb) ?
+				XFS_FSOP_GEOM_FLAGS_LAZYSB : 0) |
 			(XFS_SB_VERSION_HASATTR2(&mp->m_sb) ?
 				XFS_FSOP_GEOM_FLAGS_ATTR2 : 0);
 		geo->logsectsize = XFS_SB_VERSION_HASSECTOR(&mp->m_sb) ?
@@ -140,6 +143,8 @@ xfs_growfs_data_private(
 	pct = in->imaxpct;
 	if (nb < mp->m_sb.sb_dblocks || pct < 0 || pct > 100)
 		return XFS_ERROR(EINVAL);
+	if ((error = xfs_sb_validate_fsb_count(&mp->m_sb, nb)))
+		return error;
 	dpct = pct - mp->m_sb.sb_imax_pct;
 	error = xfs_read_buf(mp, mp->m_ddev_targp,
 			XFS_FSB_TO_BB(mp, nb) - XFS_FSS_TO_BB(mp, 1),
@@ -161,6 +166,7 @@ xfs_growfs_data_private(
 	new = nb - mp->m_sb.sb_dblocks;
 	oagcount = mp->m_sb.sb_agcount;
 	if (nagcount > oagcount) {
+		xfs_filestream_flush(mp);
 		down_write(&mp->m_peraglock);
 		mp->m_perag = kmem_realloc(mp->m_perag,
 			sizeof(xfs_perag_t) * nagcount,
@@ -173,6 +179,7 @@ xfs_growfs_data_private(
 		up_write(&mp->m_peraglock);
 	}
 	tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFS);
+	tp->t_flags |= XFS_TRANS_RESERVE;
 	if ((error = xfs_trans_reserve(tp, XFS_GROWFS_SPACE_RES(mp),
 			XFS_GROWDATA_LOG_RES(mp), 0, 0, 0))) {
 		xfs_trans_cancel(tp, 0);
@@ -328,6 +335,7 @@ xfs_growfs_data_private(
 		be32_add(&agf->agf_length, new);
 		ASSERT(be32_to_cpu(agf->agf_length) ==
 		       be32_to_cpu(agi->agi_length));
+		xfs_alloc_log_agf(tp, bp, XFS_AGF_LENGTH);
 		/*
 		 * Free the new space.
 		 */
@@ -494,8 +502,9 @@ xfs_reserve_blocks(
 	unsigned long		s;
 
 	/* If inval is null, report current values and return */
-
 	if (inval == (__uint64_t *)NULL) {
+		if (!outval)
+			return EINVAL;
 		outval->resblks = mp->m_resblks;
 		outval->resblks_avail = mp->m_resblks_avail;
 		return 0;
@@ -558,8 +567,10 @@ retry:
 		}
 	}
 out:
-	outval->resblks = mp->m_resblks;
-	outval->resblks_avail = mp->m_resblks_avail;
+	if (outval) {
+		outval->resblks = mp->m_resblks;
+		outval->resblks_avail = mp->m_resblks_avail;
+	}
 	XFS_SB_UNLOCK(mp, s);
 
 	if (fdblks_delta) {

fs/xfs/xfs_ialloc.c

@@ -123,6 +123,7 @@ xfs_ialloc_ag_alloc(
 	int		blks_per_cluster;  /* fs blocks per inode cluster */
 	xfs_btree_cur_t	*cur;		/* inode btree cursor */
 	xfs_daddr_t	d;		/* disk addr of buffer */
+	xfs_agnumber_t	agno;
 	int		error;
 	xfs_buf_t	*fbuf;		/* new free inodes' buffer */
 	xfs_dinode_t	*free;		/* new free inode structure */
@@ -302,15 +303,15 @@ xfs_ialloc_ag_alloc(
 	}
 	be32_add(&agi->agi_count, newlen);
 	be32_add(&agi->agi_freecount, newlen);
+	agno = be32_to_cpu(agi->agi_seqno);
 	down_read(&args.mp->m_peraglock);
-	args.mp->m_perag[be32_to_cpu(agi->agi_seqno)].pagi_freecount += newlen;
+	args.mp->m_perag[agno].pagi_freecount += newlen;
 	up_read(&args.mp->m_peraglock);
 	agi->agi_newino = cpu_to_be32(newino);
 	/*
 	 * Insert records describing the new inode chunk into the btree.
 	 */
-	cur = xfs_btree_init_cursor(args.mp, tp, agbp,
-			be32_to_cpu(agi->agi_seqno),
+	cur = xfs_btree_init_cursor(args.mp, tp, agbp, agno,
 			XFS_BTNUM_INO, (xfs_inode_t *)0, 0);
 	for (thisino = newino;
 	     thisino < newino + newlen;
@@ -1387,6 +1388,7 @@ xfs_ialloc_read_agi(
 	pag = &mp->m_perag[agno];
 	if (!pag->pagi_init) {
 		pag->pagi_freecount = be32_to_cpu(agi->agi_freecount);
+		pag->pagi_count = be32_to_cpu(agi->agi_count);
 		pag->pagi_init = 1;
 	} else {
 		/*
@@ -1410,3 +1412,23 @@ xfs_ialloc_read_agi(
 	*bpp = bp;
 	return 0;
 }
+
+/*
+ * Read in the agi to initialise the per-ag data in the mount structure
+ */
+int
+xfs_ialloc_pagi_init(
+	xfs_mount_t	*mp,		/* file system mount structure */
+	xfs_trans_t	*tp,		/* transaction pointer */
+	xfs_agnumber_t	agno)		/* allocation group number */
+{
+	xfs_buf_t	*bp = NULL;
+	int		error;
+
+	error = xfs_ialloc_read_agi(mp, tp, agno, &bp);
+	if (error)
+		return error;
+	if (bp)
+		xfs_trans_brelse(tp, bp);
+	return 0;
+}

fs/xfs/xfs_ialloc.h

@@ -149,6 +149,16 @@ xfs_ialloc_read_agi(
 	xfs_agnumber_t	agno,		/* allocation group number */
 	struct xfs_buf	**bpp);		/* allocation group hdr buf */
 
+/*
+ * Read in the allocation group header to initialise the per-ag data
+ * in the mount structure
+ */
+int
+xfs_ialloc_pagi_init(
+	struct xfs_mount *mp,		/* file system mount structure */
+	struct xfs_trans *tp,		/* transaction pointer */
+        xfs_agnumber_t  agno);		/* allocation group number */
+
 #endif	/* __KERNEL__ */
 
 #endif	/* __XFS_IALLOC_H__ */

fs/xfs/xfs_inode.c

@@ -48,7 +48,9 @@
 #include "xfs_dir2_trace.h"
 #include "xfs_quota.h"
 #include "xfs_acl.h"
+#include "xfs_filestream.h"
 
+#include <linux/log2.h>
 
 kmem_zone_t *xfs_ifork_zone;
 kmem_zone_t *xfs_inode_zone;
@@ -643,8 +645,7 @@ xfs_iformat_extents(
 			ep->l1 = INT_GET(get_unaligned((__uint64_t*)&dp->l1),
 								ARCH_CONVERT);
 		}
-		xfs_bmap_trace_exlist("xfs_iformat_extents", ip, nex,
-			whichfork);
+		XFS_BMAP_TRACE_EXLIST(ip, nex, whichfork);
 		if (whichfork != XFS_DATA_FORK ||
 			XFS_EXTFMT_INODE(ip) == XFS_EXTFMT_NOSTATE)
 				if (unlikely(xfs_check_nostate_extents(
@@ -817,6 +818,8 @@ _xfs_dic2xflags(
 			flags |= XFS_XFLAG_EXTSZINHERIT;
 		if (di_flags & XFS_DIFLAG_NODEFRAG)
 			flags |= XFS_XFLAG_NODEFRAG;
+		if (di_flags & XFS_DIFLAG_FILESTREAM)
+			flags |= XFS_XFLAG_FILESTREAM;
 	}
 
 	return flags;
@@ -1074,6 +1077,11 @@ xfs_iread_extents(
  * also returns the [locked] bp pointing to the head of the freelist
  * as ialloc_context.  The caller should hold this buffer across
  * the commit and pass it back into this routine on the second call.
+ *
+ * If we are allocating quota inodes, we do not have a parent inode
+ * to attach to or associate with (i.e. pip == NULL) because they
+ * are not linked into the directory structure - they are attached
+ * directly to the superblock - and so have no parent.
  */
 int
 xfs_ialloc(
@@ -1099,7 +1107,7 @@ xfs_ialloc(
 	 * Call the space management code to pick
 	 * the on-disk inode to be allocated.
 	 */
-	error = xfs_dialloc(tp, pip->i_ino, mode, okalloc,
+	error = xfs_dialloc(tp, pip ? pip->i_ino : 0, mode, okalloc,
 			    ialloc_context, call_again, &ino);
 	if (error != 0) {
 		return error;
@@ -1150,10 +1158,10 @@ xfs_ialloc(
 	/*
 	 * Project ids won't be stored on disk if we are using a version 1 inode.
 	 */
-	if ( (prid != 0) && (ip->i_d.di_version == XFS_DINODE_VERSION_1))
+	if ((prid != 0) && (ip->i_d.di_version == XFS_DINODE_VERSION_1))
 		xfs_bump_ino_vers2(tp, ip);
 
-	if (XFS_INHERIT_GID(pip, vp->v_vfsp)) {
+	if (pip && XFS_INHERIT_GID(pip, vp->v_vfsp)) {
 		ip->i_d.di_gid = pip->i_d.di_gid;
 		if ((pip->i_d.di_mode & S_ISGID) && (mode & S_IFMT) == S_IFDIR) {
 			ip->i_d.di_mode |= S_ISGID;
@@ -1195,8 +1203,16 @@ xfs_ialloc(
 		flags |= XFS_ILOG_DEV;
 		break;
 	case S_IFREG:
+		if (pip && xfs_inode_is_filestream(pip)) {
+			error = xfs_filestream_associate(pip, ip);
+			if (error < 0)
+				return -error;
+			if (!error)
+				xfs_iflags_set(ip, XFS_IFILESTREAM);
+		}
+		/* fall through */
 	case S_IFDIR:
-		if (unlikely(pip->i_d.di_flags & XFS_DIFLAG_ANY)) {
+		if (pip && (pip->i_d.di_flags & XFS_DIFLAG_ANY)) {
 			uint	di_flags = 0;
 
 			if ((mode & S_IFMT) == S_IFDIR) {
@@ -1233,6 +1249,8 @@ xfs_ialloc(
 			if ((pip->i_d.di_flags & XFS_DIFLAG_NODEFRAG) &&
 			    xfs_inherit_nodefrag)
 				di_flags |= XFS_DIFLAG_NODEFRAG;
+			if (pip->i_d.di_flags & XFS_DIFLAG_FILESTREAM)
+				di_flags |= XFS_DIFLAG_FILESTREAM;
 			ip->i_d.di_flags |= di_flags;
 		}
 		/* FALLTHROUGH */
@@ -2875,9 +2893,6 @@ xfs_iextents_copy(
 	int			copied;
 	xfs_bmbt_rec_t		*dest_ep;
 	xfs_bmbt_rec_t		*ep;
-#ifdef XFS_BMAP_TRACE
-	static char		fname[] = "xfs_iextents_copy";
-#endif
 	int			i;
 	xfs_ifork_t		*ifp;
 	int			nrecs;
@@ -2888,7 +2903,7 @@ xfs_iextents_copy(
 	ASSERT(ifp->if_bytes > 0);
 
 	nrecs = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-	xfs_bmap_trace_exlist(fname, ip, nrecs, whichfork);
+	XFS_BMAP_TRACE_EXLIST(ip, nrecs, whichfork);
 	ASSERT(nrecs > 0);
 
 	/*
@@ -4184,7 +4199,7 @@ xfs_iext_realloc_direct(
 			ifp->if_bytes = new_size;
 			return;
 		}
-		if ((new_size & (new_size - 1)) != 0) {
+		if (!is_power_of_2(new_size)){
 			rnew_size = xfs_iroundup(new_size);
 		}
 		if (rnew_size != ifp->if_real_bytes) {
@@ -4207,7 +4222,7 @@ xfs_iext_realloc_direct(
 	 */
 	else {
 		new_size += ifp->if_bytes;
-		if ((new_size & (new_size - 1)) != 0) {
+		if (!is_power_of_2(new_size)) {
 			rnew_size = xfs_iroundup(new_size);
 		}
 		xfs_iext_inline_to_direct(ifp, rnew_size);

fs/xfs/xfs_inode.h

@@ -379,6 +379,7 @@ xfs_iflags_test(xfs_inode_t *ip, unsigned short flags)
 #define XFS_ISTALE	0x0010	/* inode has been staled */
 #define XFS_IRECLAIMABLE 0x0020 /* inode can be reclaimed */
 #define XFS_INEW	0x0040
+#define XFS_IFILESTREAM	0x0080	/* inode is in a filestream directory */
 
 /*
  * Flags for inode locking.
@@ -414,19 +415,22 @@ xfs_iflags_test(xfs_inode_t *ip, unsigned short flags)
  * gets a lockdep subclass of 1 and the second lock will have a lockdep
  * subclass of 0.
  *
- * XFS_I[O]LOCK_INUMORDER - for locking several inodes at the some time
+ * XFS_LOCK_INUMORDER - for locking several inodes at the some time
  * with xfs_lock_inodes().  This flag is used as the starting subclass
  * and each subsequent lock acquired will increment the subclass by one.
  * So the first lock acquired will have a lockdep subclass of 2, the
- * second lock will have a lockdep subclass of 3, and so on.
+ * second lock will have a lockdep subclass of 3, and so on. It is
+ * the responsibility of the class builder to shift this to the correct
+ * portion of the lock_mode lockdep mask.
  */
+#define XFS_LOCK_PARENT		1
+#define XFS_LOCK_INUMORDER	2
+
 #define XFS_IOLOCK_SHIFT	16
-#define	XFS_IOLOCK_PARENT	(1 << XFS_IOLOCK_SHIFT)
-#define	XFS_IOLOCK_INUMORDER	(2 << XFS_IOLOCK_SHIFT)
+#define	XFS_IOLOCK_PARENT	(XFS_LOCK_PARENT << XFS_IOLOCK_SHIFT)
 
 #define XFS_ILOCK_SHIFT		24
-#define	XFS_ILOCK_PARENT	(1 << XFS_ILOCK_SHIFT)
-#define	XFS_ILOCK_INUMORDER	(2 << XFS_ILOCK_SHIFT)
+#define	XFS_ILOCK_PARENT	(XFS_LOCK_PARENT << XFS_ILOCK_SHIFT)
 
 #define XFS_IOLOCK_DEP_MASK	0x00ff0000
 #define XFS_ILOCK_DEP_MASK	0xff000000

fs/xfs/xfs_iomap.c

@@ -451,19 +451,14 @@ xfs_iomap_write_direct(
 		return XFS_ERROR(error);
 
 	rt = XFS_IS_REALTIME_INODE(ip);
-	if (unlikely(rt)) {
-		if (!(extsz = ip->i_d.di_extsize))
-			extsz = mp->m_sb.sb_rextsize;
-	} else {
-		extsz = ip->i_d.di_extsize;
-	}
+	extsz = xfs_get_extsz_hint(ip);
 
 	isize = ip->i_size;
 	if (io->io_new_size > isize)
 		isize = io->io_new_size;
 
-  	offset_fsb = XFS_B_TO_FSBT(mp, offset);
-  	last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count)));
+	offset_fsb = XFS_B_TO_FSBT(mp, offset);
+	last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count)));
 	if ((offset + count) > isize) {
 		error = xfs_iomap_eof_align_last_fsb(mp, io, isize, extsz,
 							&last_fsb);
@@ -489,13 +484,13 @@ xfs_iomap_write_direct(
 	if (unlikely(rt)) {
 		resrtextents = qblocks = resaligned;
 		resrtextents /= mp->m_sb.sb_rextsize;
-  		resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0);
-  		quota_flag = XFS_QMOPT_RES_RTBLKS;
-  	} else {
-  		resrtextents = 0;
+		resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0);
+		quota_flag = XFS_QMOPT_RES_RTBLKS;
+	} else {
+		resrtextents = 0;
 		resblks = qblocks = XFS_DIOSTRAT_SPACE_RES(mp, resaligned);
-  		quota_flag = XFS_QMOPT_RES_REGBLKS;
-  	}
+		quota_flag = XFS_QMOPT_RES_REGBLKS;
+	}
 
 	/*
 	 * Allocate and setup the transaction
@@ -666,13 +661,7 @@ xfs_iomap_write_delay(
 	if (error)
 		return XFS_ERROR(error);
 
-	if (XFS_IS_REALTIME_INODE(ip)) {
-		if (!(extsz = ip->i_d.di_extsize))
-			extsz = mp->m_sb.sb_rextsize;
-	} else {
-		extsz = ip->i_d.di_extsize;
-	}
-
+	extsz = xfs_get_extsz_hint(ip);
 	offset_fsb = XFS_B_TO_FSBT(mp, offset);
 
 retry:
@@ -788,18 +777,12 @@ xfs_iomap_write_allocate(
 		nimaps = 0;
 		while (nimaps == 0) {
 			tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
+			tp->t_flags |= XFS_TRANS_RESERVE;
 			nres = XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK);
 			error = xfs_trans_reserve(tp, nres,
 					XFS_WRITE_LOG_RES(mp),
 					0, XFS_TRANS_PERM_LOG_RES,
 					XFS_WRITE_LOG_COUNT);
-			if (error == ENOSPC) {
-				error = xfs_trans_reserve(tp, 0,
-						XFS_WRITE_LOG_RES(mp),
-						0,
-						XFS_TRANS_PERM_LOG_RES,
-						XFS_WRITE_LOG_COUNT);
-			}
 			if (error) {
 				xfs_trans_cancel(tp, 0);
 				return XFS_ERROR(error);
@@ -917,8 +900,8 @@ xfs_iomap_write_unwritten(
 		 * from unwritten to real. Do allocations in a loop until
 		 * we have covered the range passed in.
 		 */
-
 		tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
+		tp->t_flags |= XFS_TRANS_RESERVE;
 		error = xfs_trans_reserve(tp, resblks,
 				XFS_WRITE_LOG_RES(mp), 0,
 				XFS_TRANS_PERM_LOG_RES,

fs/xfs/xfs_itable.c

@@ -202,6 +202,16 @@ xfs_bulkstat_one_dinode(
 	return 0;
 }
 
+STATIC int
+xfs_bulkstat_one_fmt(
+	void			__user *ubuffer,
+	const xfs_bstat_t	*buffer)
+{
+	if (copy_to_user(ubuffer, buffer, sizeof(*buffer)))
+		return -EFAULT;
+	return sizeof(*buffer);
+}
+
 /*
  * Return stat information for one inode.
  * Return 0 if ok, else errno.
@@ -221,6 +231,7 @@ xfs_bulkstat_one(
 	xfs_bstat_t	*buf;		/* return buffer */
 	int		error = 0;	/* error value */
 	xfs_dinode_t	*dip;		/* dinode inode pointer */
+	bulkstat_one_fmt_pf formatter = private_data ? : xfs_bulkstat_one_fmt;
 
 	dip = (xfs_dinode_t *)dibuff;
 	*stat = BULKSTAT_RV_NOTHING;
@@ -243,14 +254,15 @@ xfs_bulkstat_one(
 		xfs_bulkstat_one_dinode(mp, ino, dip, buf);
 	}
 
-	if (copy_to_user(buffer, buf, sizeof(*buf)))  {
+	error = formatter(buffer, buf);
+	if (error < 0)  {
 		error = EFAULT;
 		goto out_free;
 	}
 
 	*stat = BULKSTAT_RV_DIDONE;
 	if (ubused)
-		*ubused = sizeof(*buf);
+		*ubused = error;
 
  out_free:
 	kmem_free(buf, sizeof(*buf));
@@ -748,6 +760,19 @@ xfs_bulkstat_single(
 	return 0;
 }
 
+int
+xfs_inumbers_fmt(
+	void			__user *ubuffer, /* buffer to write to */
+	const xfs_inogrp_t	*buffer,	/* buffer to read from */
+	long			count,		/* # of elements to read */
+	long			*written)	/* # of bytes written */
+{
+	if (copy_to_user(ubuffer, buffer, count * sizeof(*buffer)))
+		return -EFAULT;
+	*written = count * sizeof(*buffer);
+	return 0;
+}
+
 /*
  * Return inode number table for the filesystem.
  */
@@ -756,7 +781,8 @@ xfs_inumbers(
 	xfs_mount_t	*mp,		/* mount point for filesystem */
 	xfs_ino_t	*lastino,	/* last inode returned */
 	int		*count,		/* size of buffer/count returned */
-	xfs_inogrp_t	__user *ubuffer)/* buffer with inode descriptions */
+	void		__user *ubuffer,/* buffer with inode descriptions */
+	inumbers_fmt_pf	formatter)
 {
 	xfs_buf_t	*agbp;
 	xfs_agino_t	agino;
@@ -835,12 +861,12 @@ xfs_inumbers(
 		bufidx++;
 		left--;
 		if (bufidx == bcount) {
-			if (copy_to_user(ubuffer, buffer,
-					bufidx * sizeof(*buffer))) {
+			long written;
+			if (formatter(ubuffer, buffer, bufidx, &written)) {
 				error = XFS_ERROR(EFAULT);
 				break;
 			}
-			ubuffer += bufidx;
+			ubuffer += written;
 			*count += bufidx;
 			bufidx = 0;
 		}
@@ -862,8 +888,8 @@ xfs_inumbers(
 	}
 	if (!error) {
 		if (bufidx) {
-			if (copy_to_user(ubuffer, buffer,
-					bufidx * sizeof(*buffer)))
+			long written;
+			if (formatter(ubuffer, buffer, bufidx, &written))
 				error = XFS_ERROR(EFAULT);
 			else
 				*count += bufidx;

fs/xfs/xfs_itable.h

@@ -69,6 +69,10 @@ xfs_bulkstat_single(
 	char			__user *buffer,
 	int			*done);
 
+typedef int (*bulkstat_one_fmt_pf)(  /* used size in bytes or negative error */
+	void			__user *ubuffer, /* buffer to write to */
+	const xfs_bstat_t	*buffer);        /* buffer to read from */
+
 int
 xfs_bulkstat_one(
 	xfs_mount_t		*mp,
@@ -86,11 +90,25 @@ xfs_internal_inum(
 	xfs_mount_t		*mp,
 	xfs_ino_t		ino);
 
+typedef int (*inumbers_fmt_pf)(
+	void			__user *ubuffer, /* buffer to write to */
+	const xfs_inogrp_t	*buffer,	/* buffer to read from */
+	long			count,		/* # of elements to read */
+	long			*written);	/* # of bytes written */
+
+int
+xfs_inumbers_fmt(
+	void			__user *ubuffer, /* buffer to write to */
+	const xfs_inogrp_t	*buffer,	/* buffer to read from */
+	long			count,		/* # of elements to read */
+	long			*written);	/* # of bytes written */
+
 int					/* error status */
 xfs_inumbers(
 	xfs_mount_t		*mp,	/* mount point for filesystem */
 	xfs_ino_t		*last,	/* last inode returned */
 	int			*count,	/* size of buffer/count returned */
-	xfs_inogrp_t		__user *buffer);/* buffer with inode info */
+	void			__user *buffer, /* buffer with inode info */
+	inumbers_fmt_pf		formatter);
 
 #endif	/* __XFS_ITABLE_H__ */

fs/xfs/xfs_log.c

@@ -817,10 +817,8 @@ xfs_log_need_covered(xfs_mount_t *mp)
 	SPLDECL(s);
 	int		needed = 0, gen;
 	xlog_t		*log = mp->m_log;
-	bhv_vfs_t	*vfsp = XFS_MTOVFS(mp);
 
-	if (vfs_test_for_freeze(vfsp) || XFS_FORCED_SHUTDOWN(mp) ||
-	    (vfsp->vfs_flag & VFS_RDONLY))
+	if (!xfs_fs_writable(mp))
 		return 0;
 
 	s = LOG_LOCK(log);
@@ -967,14 +965,16 @@ xlog_iodone(xfs_buf_t *bp)
 	} else if (iclog->ic_state & XLOG_STATE_IOERROR) {
 		aborted = XFS_LI_ABORTED;
 	}
+
+	/* log I/O is always issued ASYNC */
+	ASSERT(XFS_BUF_ISASYNC(bp));
 	xlog_state_done_syncing(iclog, aborted);
-	if (!(XFS_BUF_ISASYNC(bp))) {
-		/*
-		 * Corresponding psema() will be done in bwrite().  If we don't
-		 * vsema() here, panic.
-		 */
-		XFS_BUF_V_IODONESEMA(bp);
-	}
+	/*
+	 * do not reference the buffer (bp) here as we could race
+	 * with it being freed after writing the unmount record to the
+	 * log.
+	 */
+
 }	/* xlog_iodone */
 
 /*
@@ -1199,11 +1199,18 @@ xlog_alloc_log(xfs_mount_t	*mp,
 		*iclogp = (xlog_in_core_t *)
 			  kmem_zalloc(sizeof(xlog_in_core_t), KM_SLEEP);
 		iclog = *iclogp;
-		iclog->hic_data = (xlog_in_core_2_t *)
-			  kmem_zalloc(iclogsize, KM_SLEEP | KM_LARGE);
-
 		iclog->ic_prev = prev_iclog;
 		prev_iclog = iclog;
+
+		bp = xfs_buf_get_noaddr(log->l_iclog_size, mp->m_logdev_targp);
+		if (!XFS_BUF_CPSEMA(bp))
+			ASSERT(0);
+		XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone);
+		XFS_BUF_SET_BDSTRAT_FUNC(bp, xlog_bdstrat_cb);
+		XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
+		iclog->ic_bp = bp;
+		iclog->hic_data = bp->b_addr;
+
 		log->l_iclog_bak[i] = (xfs_caddr_t)&(iclog->ic_header);
 
 		head = &iclog->ic_header;
@@ -1216,11 +1223,6 @@ xlog_alloc_log(xfs_mount_t	*mp,
 		INT_SET(head->h_fmt, ARCH_CONVERT, XLOG_FMT);
 		memcpy(&head->h_fs_uuid, &mp->m_sb.sb_uuid, sizeof(uuid_t));
 
-		bp = xfs_buf_get_empty(log->l_iclog_size, mp->m_logdev_targp);
-		XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone);
-		XFS_BUF_SET_BDSTRAT_FUNC(bp, xlog_bdstrat_cb);
-		XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
-		iclog->ic_bp = bp;
 
 		iclog->ic_size = XFS_BUF_SIZE(bp) - log->l_iclog_hsize;
 		iclog->ic_state = XLOG_STATE_ACTIVE;
@@ -1432,7 +1434,7 @@ xlog_sync(xlog_t		*log,
 	} else {
 		iclog->ic_bwritecnt = 1;
 	}
-	XFS_BUF_SET_PTR(bp, (xfs_caddr_t) &(iclog->ic_header), count);
+	XFS_BUF_SET_COUNT(bp, count);
 	XFS_BUF_SET_FSPRIVATE(bp, iclog);	/* save for later */
 	XFS_BUF_ZEROFLAGS(bp);
 	XFS_BUF_BUSY(bp);
@@ -1528,7 +1530,6 @@ xlog_dealloc_log(xlog_t *log)
 		}
 #endif
 		next_iclog = iclog->ic_next;
-		kmem_free(iclog->hic_data, log->l_iclog_size);
 		kmem_free(iclog, sizeof(xlog_in_core_t));
 		iclog = next_iclog;
 	}

fs/xfs/xfs_log_recover.c

@@ -927,6 +927,14 @@ xlog_find_tail(
 			ASSIGN_ANY_LSN_HOST(log->l_last_sync_lsn, log->l_curr_cycle,
 					after_umount_blk);
 			*tail_blk = after_umount_blk;
+
+			/*
+			 * Note that the unmount was clean. If the unmount
+			 * was not clean, we need to know this to rebuild the
+			 * superblock counters from the perag headers if we
+			 * have a filesystem using non-persistent counters.
+			 */
+			log->l_mp->m_flags |= XFS_MOUNT_WAS_CLEAN;
 		}
 	}
 

fs/xfs/xfs_mount.c

@@ -202,6 +202,27 @@ xfs_mount_free(
 	kmem_free(mp, sizeof(xfs_mount_t));
 }
 
+/*
+ * Check size of device based on the (data/realtime) block count.
+ * Note: this check is used by the growfs code as well as mount.
+ */
+int
+xfs_sb_validate_fsb_count(
+	xfs_sb_t	*sbp,
+	__uint64_t	nblocks)
+{
+	ASSERT(PAGE_SHIFT >= sbp->sb_blocklog);
+	ASSERT(sbp->sb_blocklog >= BBSHIFT);
+
+#if XFS_BIG_BLKNOS     /* Limited by ULONG_MAX of page cache index */
+	if (nblocks >> (PAGE_CACHE_SHIFT - sbp->sb_blocklog) > ULONG_MAX)
+		return E2BIG;
+#else                  /* Limited by UINT_MAX of sectors */
+	if (nblocks << (sbp->sb_blocklog - BBSHIFT) > UINT_MAX)
+		return E2BIG;
+#endif
+	return 0;
+}
 
 /*
  * Check the validity of the SB found.
@@ -284,18 +305,8 @@ xfs_mount_validate_sb(
 		return XFS_ERROR(EFSCORRUPTED);
 	}
 
-	ASSERT(PAGE_SHIFT >= sbp->sb_blocklog);
-	ASSERT(sbp->sb_blocklog >= BBSHIFT);
-
-#if XFS_BIG_BLKNOS     /* Limited by ULONG_MAX of page cache index */
-	if (unlikely(
-	    (sbp->sb_dblocks >> (PAGE_SHIFT - sbp->sb_blocklog)) > ULONG_MAX ||
-	    (sbp->sb_rblocks >> (PAGE_SHIFT - sbp->sb_blocklog)) > ULONG_MAX)) {
-#else                  /* Limited by UINT_MAX of sectors */
-	if (unlikely(
-	    (sbp->sb_dblocks << (sbp->sb_blocklog - BBSHIFT)) > UINT_MAX ||
-	    (sbp->sb_rblocks << (sbp->sb_blocklog - BBSHIFT)) > UINT_MAX)) {
-#endif
+	if (xfs_sb_validate_fsb_count(sbp, sbp->sb_dblocks) ||
+	    xfs_sb_validate_fsb_count(sbp, sbp->sb_rblocks)) {
 		xfs_fs_mount_cmn_err(flags,
 			"file system too large to be mounted on this system.");
 		return XFS_ERROR(E2BIG);
@@ -632,6 +643,64 @@ xfs_mount_common(xfs_mount_t *mp, xfs_sb_t *sbp)
 					sbp->sb_inopblock);
 	mp->m_ialloc_blks = mp->m_ialloc_inos >> sbp->sb_inopblog;
 }
+
+/*
+ * xfs_initialize_perag_data
+ *
+ * Read in each per-ag structure so we can count up the number of
+ * allocated inodes, free inodes and used filesystem blocks as this
+ * information is no longer persistent in the superblock. Once we have
+ * this information, write it into the in-core superblock structure.
+ */
+STATIC int
+xfs_initialize_perag_data(xfs_mount_t *mp, xfs_agnumber_t agcount)
+{
+	xfs_agnumber_t	index;
+	xfs_perag_t	*pag;
+	xfs_sb_t	*sbp = &mp->m_sb;
+	uint64_t	ifree = 0;
+	uint64_t	ialloc = 0;
+	uint64_t	bfree = 0;
+	uint64_t	bfreelst = 0;
+	uint64_t	btree = 0;
+	int		error;
+	int		s;
+
+	for (index = 0; index < agcount; index++) {
+		/*
+		 * read the agf, then the agi. This gets us
+		 * all the inforamtion we need and populates the
+		 * per-ag structures for us.
+		 */
+		error = xfs_alloc_pagf_init(mp, NULL, index, 0);
+		if (error)
+			return error;
+
+		error = xfs_ialloc_pagi_init(mp, NULL, index);
+		if (error)
+			return error;
+		pag = &mp->m_perag[index];
+		ifree += pag->pagi_freecount;
+		ialloc += pag->pagi_count;
+		bfree += pag->pagf_freeblks;
+		bfreelst += pag->pagf_flcount;
+		btree += pag->pagf_btreeblks;
+	}
+	/*
+	 * Overwrite incore superblock counters with just-read data
+	 */
+	s = XFS_SB_LOCK(mp);
+	sbp->sb_ifree = ifree;
+	sbp->sb_icount = ialloc;
+	sbp->sb_fdblocks = bfree + bfreelst + btree;
+	XFS_SB_UNLOCK(mp, s);
+
+	/* Fixup the per-cpu counters as well. */
+	xfs_icsb_reinit_counters(mp);
+
+	return 0;
+}
+
 /*
  * xfs_mountfs
  *
@@ -656,7 +725,7 @@ xfs_mountfs(
 	bhv_vnode_t	*rvp = NULL;
 	int		readio_log, writeio_log;
 	xfs_daddr_t	d;
-	__uint64_t	ret64;
+	__uint64_t	resblks;
 	__int64_t	update_flags;
 	uint		quotamount, quotaflags;
 	int		agno;
@@ -773,6 +842,7 @@ xfs_mountfs(
 	 */
 	if ((mfsi_flags & XFS_MFSI_SECOND) == 0 &&
 	    (mp->m_flags & XFS_MOUNT_NOUUID) == 0) {
+		__uint64_t	ret64;
 		if (xfs_uuid_mount(mp)) {
 			error = XFS_ERROR(EINVAL);
 			goto error1;
@@ -975,6 +1045,34 @@ xfs_mountfs(
 		goto error2;
 	}
 
+	/*
+	 * Now the log is mounted, we know if it was an unclean shutdown or
+	 * not. If it was, with the first phase of recovery has completed, we
+	 * have consistent AG blocks on disk. We have not recovered EFIs yet,
+	 * but they are recovered transactionally in the second recovery phase
+	 * later.
+	 *
+	 * Hence we can safely re-initialise incore superblock counters from
+	 * the per-ag data. These may not be correct if the filesystem was not
+	 * cleanly unmounted, so we need to wait for recovery to finish before
+	 * doing this.
+	 *
+	 * If the filesystem was cleanly unmounted, then we can trust the
+	 * values in the superblock to be correct and we don't need to do
+	 * anything here.
+	 *
+	 * If we are currently making the filesystem, the initialisation will
+	 * fail as the perag data is in an undefined state.
+	 */
+
+	if (xfs_sb_version_haslazysbcount(&mp->m_sb) &&
+	    !XFS_LAST_UNMOUNT_WAS_CLEAN(mp) &&
+	     !mp->m_sb.sb_inprogress) {
+		error = xfs_initialize_perag_data(mp, sbp->sb_agcount);
+		if (error) {
+			goto error2;
+		}
+	}
 	/*
 	 * Get and sanity-check the root inode.
 	 * Save the pointer to it in the mount structure.
@@ -1044,6 +1142,23 @@ xfs_mountfs(
 	if ((error = XFS_QM_MOUNT(mp, quotamount, quotaflags, mfsi_flags)))
 		goto error4;
 
+	/*
+	 * Now we are mounted, reserve a small amount of unused space for
+	 * privileged transactions. This is needed so that transaction
+	 * space required for critical operations can dip into this pool
+	 * when at ENOSPC. This is needed for operations like create with
+	 * attr, unwritten extent conversion at ENOSPC, etc. Data allocations
+	 * are not allowed to use this reserved space.
+	 *
+	 * We default to 5% or 1024 fsbs of space reserved, whichever is smaller.
+	 * This may drive us straight to ENOSPC on mount, but that implies
+	 * we were already there on the last unmount.
+	 */
+	resblks = mp->m_sb.sb_dblocks;
+	do_div(resblks, 20);
+	resblks = min_t(__uint64_t, resblks, 1024);
+	xfs_reserve_blocks(mp, &resblks, NULL);
+
 	return 0;
 
  error4:
@@ -1083,7 +1198,19 @@ xfs_unmountfs(xfs_mount_t *mp, struct cred *cr)
 #if defined(DEBUG) || defined(INDUCE_IO_ERROR)
 	int64_t		fsid;
 #endif
+	__uint64_t	resblks;
 
+	/*
+	 * We can potentially deadlock here if we have an inode cluster
+	 * that has been freed has it's buffer still pinned in memory because
+	 * the transaction is still sitting in a iclog. The stale inodes
+	 * on that buffer will have their flush locks held until the
+	 * transaction hits the disk and the callbacks run. the inode
+	 * flush takes the flush lock unconditionally and with nothing to
+	 * push out the iclog we will never get that unlocked. hence we
+	 * need to force the log first.
+	 */
+	xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC);
 	xfs_iflush_all(mp);
 
 	XFS_QM_DQPURGEALL(mp, XFS_QMOPT_QUOTALL | XFS_QMOPT_UMOUNTING);
@@ -1100,10 +1227,26 @@ xfs_unmountfs(xfs_mount_t *mp, struct cred *cr)
 		xfs_binval(mp->m_rtdev_targp);
 	}
 
-	xfs_unmountfs_writesb(mp);
+	/*
+	 * Unreserve any blocks we have so that when we unmount we don't account
+	 * the reserved free space as used. This is really only necessary for
+	 * lazy superblock counting because it trusts the incore superblock
+	 * counters to be aboslutely correct on clean unmount.
+	 *
+	 * We don't bother correcting this elsewhere for lazy superblock
+	 * counting because on mount of an unclean filesystem we reconstruct the
+	 * correct counter value and this is irrelevant.
+	 *
+	 * For non-lazy counter filesystems, this doesn't matter at all because
+	 * we only every apply deltas to the superblock and hence the incore
+	 * value does not matter....
+	 */
+	resblks = 0;
+	xfs_reserve_blocks(mp, &resblks, NULL);
 
+	xfs_log_sbcount(mp, 1);
+	xfs_unmountfs_writesb(mp);
 	xfs_unmountfs_wait(mp); 		/* wait for async bufs */
-
 	xfs_log_unmount(mp);			/* Done! No more fs ops. */
 
 	xfs_freesb(mp);
@@ -1149,6 +1292,62 @@ xfs_unmountfs_wait(xfs_mount_t *mp)
 	xfs_wait_buftarg(mp->m_ddev_targp);
 }
 
+int
+xfs_fs_writable(xfs_mount_t *mp)
+{
+	bhv_vfs_t	*vfsp = XFS_MTOVFS(mp);
+
+	return !(vfs_test_for_freeze(vfsp) || XFS_FORCED_SHUTDOWN(mp) ||
+		(vfsp->vfs_flag & VFS_RDONLY));
+}
+
+/*
+ * xfs_log_sbcount
+ *
+ * Called either periodically to keep the on disk superblock values
+ * roughly up to date or from unmount to make sure the values are
+ * correct on a clean unmount.
+ *
+ * Note this code can be called during the process of freezing, so
+ * we may need to use the transaction allocator which does not not
+ * block when the transaction subsystem is in its frozen state.
+ */
+int
+xfs_log_sbcount(
+	xfs_mount_t	*mp,
+	uint		sync)
+{
+	xfs_trans_t	*tp;
+	int		error;
+
+	if (!xfs_fs_writable(mp))
+		return 0;
+
+	xfs_icsb_sync_counters(mp);
+
+	/*
+	 * we don't need to do this if we are updating the superblock
+	 * counters on every modification.
+	 */
+	if (!xfs_sb_version_haslazysbcount(&mp->m_sb))
+		return 0;
+
+	tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT);
+	error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0,
+					XFS_DEFAULT_LOG_COUNT);
+	if (error) {
+		xfs_trans_cancel(tp, 0);
+		return error;
+	}
+
+	xfs_mod_sb(tp, XFS_SB_IFREE | XFS_SB_ICOUNT | XFS_SB_FDBLOCKS);
+	if (sync)
+		xfs_trans_set_sync(tp);
+	xfs_trans_commit(tp, 0);
+
+	return 0;
+}
+
 int
 xfs_unmountfs_writesb(xfs_mount_t *mp)
 {
@@ -1160,16 +1359,15 @@ xfs_unmountfs_writesb(xfs_mount_t *mp)
 	 * skip superblock write if fs is read-only, or
 	 * if we are doing a forced umount.
 	 */
-	sbp = xfs_getsb(mp, 0);
 	if (!(XFS_MTOVFS(mp)->vfs_flag & VFS_RDONLY ||
 		XFS_FORCED_SHUTDOWN(mp))) {
 
-		xfs_icsb_sync_counters(mp);
+		sbp = xfs_getsb(mp, 0);
+ 		sb = XFS_BUF_TO_SBP(sbp);
 
 		/*
 		 * mark shared-readonly if desired
 		 */
-		sb = XFS_BUF_TO_SBP(sbp);
 		if (mp->m_mk_sharedro) {
 			if (!(sb->sb_flags & XFS_SBF_READONLY))
 				sb->sb_flags |= XFS_SBF_READONLY;
@@ -1178,6 +1376,7 @@ xfs_unmountfs_writesb(xfs_mount_t *mp)
 			xfs_fs_cmn_err(CE_NOTE, mp,
 				"Unmounting, marking shared read-only");
 		}
+
 		XFS_BUF_UNDONE(sbp);
 		XFS_BUF_UNREAD(sbp);
 		XFS_BUF_UNDELAYWRITE(sbp);
@@ -1192,8 +1391,8 @@ xfs_unmountfs_writesb(xfs_mount_t *mp)
 					  mp, sbp, XFS_BUF_ADDR(sbp));
 		if (error && mp->m_mk_sharedro)
 			xfs_fs_cmn_err(CE_ALERT, mp, "Superblock write error detected while unmounting.  Filesystem may not be marked shared readonly");
+		xfs_buf_relse(sbp);
 	}
-	xfs_buf_relse(sbp);
 	return error;
 }
 

fs/xfs/xfs_mount.h

@@ -66,6 +66,7 @@ struct xfs_bmbt_irec;
 struct xfs_bmap_free;
 struct xfs_extdelta;
 struct xfs_swapext;
+struct xfs_mru_cache;
 
 extern struct bhv_vfsops xfs_vfsops;
 extern struct bhv_vnodeops xfs_vnodeops;
@@ -424,17 +425,18 @@ typedef struct xfs_mount {
 	struct notifier_block	m_icsb_notifier; /* hotplug cpu notifier */
 	struct mutex		m_icsb_mutex;	/* balancer sync lock */
 #endif
+	struct xfs_mru_cache	*m_filestream;  /* per-mount filestream data */
 } xfs_mount_t;
 
 /*
  * Flags for m_flags.
  */
-#define	XFS_MOUNT_WSYNC		(1ULL << 0)	/* for nfs - all metadata ops
+#define XFS_MOUNT_WSYNC		(1ULL << 0)	/* for nfs - all metadata ops
 						   must be synchronous except
 						   for space allocations */
-#define	XFS_MOUNT_INO64		(1ULL << 1)
+#define XFS_MOUNT_INO64		(1ULL << 1)
 			     /* (1ULL << 2)	-- currently unused */
-			     /* (1ULL << 3)	-- currently unused */
+#define XFS_MOUNT_WAS_CLEAN	(1ULL << 3)
 #define XFS_MOUNT_FS_SHUTDOWN	(1ULL << 4)	/* atomic stop of all filesystem
 						   operations, typically for
 						   disk errors in metadata */
@@ -463,6 +465,8 @@ typedef struct xfs_mount {
 						 * I/O size in stat() */
 #define XFS_MOUNT_NO_PERCPU_SB	(1ULL << 23)	/* don't use per-cpu superblock
 						   counters */
+#define XFS_MOUNT_FILESTREAMS	(1ULL << 24)	/* enable the filestreams
+						   allocator */
 
 
 /*
@@ -511,6 +515,8 @@ xfs_preferred_iosize(xfs_mount_t *mp)
 
 #define XFS_MAXIOFFSET(mp)	((mp)->m_maxioffset)
 
+#define XFS_LAST_UNMOUNT_WAS_CLEAN(mp)	\
+				((mp)->m_flags & XFS_MOUNT_WAS_CLEAN)
 #define XFS_FORCED_SHUTDOWN(mp)	((mp)->m_flags & XFS_MOUNT_FS_SHUTDOWN)
 #define xfs_force_shutdown(m,f)	\
 	bhv_vfs_force_shutdown((XFS_MTOVFS(m)), f, __FILE__, __LINE__)
@@ -602,6 +608,7 @@ typedef struct xfs_mod_sb {
 
 extern xfs_mount_t *xfs_mount_init(void);
 extern void	xfs_mod_sb(xfs_trans_t *, __int64_t);
+extern int	xfs_log_sbcount(xfs_mount_t *, uint);
 extern void	xfs_mount_free(xfs_mount_t *mp, int remove_bhv);
 extern int	xfs_mountfs(struct bhv_vfs *, xfs_mount_t *mp, int);
 extern void	xfs_mountfs_check_barriers(xfs_mount_t *mp);
@@ -618,12 +625,14 @@ extern int	xfs_mod_incore_sb_batch(xfs_mount_t *, xfs_mod_sb_t *,
 extern struct xfs_buf *xfs_getsb(xfs_mount_t *, int);
 extern int	xfs_readsb(xfs_mount_t *, int);
 extern void	xfs_freesb(xfs_mount_t *);
+extern int	xfs_fs_writable(xfs_mount_t *);
 extern void	xfs_do_force_shutdown(bhv_desc_t *, int, char *, int);
 extern int	xfs_syncsub(xfs_mount_t *, int, int *);
 extern int	xfs_sync_inodes(xfs_mount_t *, int, int *);
 extern xfs_agnumber_t	xfs_initialize_perag(struct bhv_vfs *, xfs_mount_t *,
 						xfs_agnumber_t);
 extern void	xfs_xlatesb(void *, struct xfs_sb *, int, __int64_t);
+extern int	xfs_sb_validate_fsb_count(struct xfs_sb *, __uint64_t);
 
 extern struct xfs_dmops xfs_dmcore_stub;
 extern struct xfs_qmops xfs_qmcore_stub;

fs/xfs/xfs_mru_cache.c

@@ -0,0 +1,608 @@
+/*
+ * Copyright (c) 2006-2007 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+#include "xfs.h"
+#include "xfs_mru_cache.h"
+
+/*
+ * The MRU Cache data structure consists of a data store, an array of lists and
+ * a lock to protect its internal state.  At initialisation time, the client
+ * supplies an element lifetime in milliseconds and a group count, as well as a
+ * function pointer to call when deleting elements.  A data structure for
+ * queueing up work in the form of timed callbacks is also included.
+ *
+ * The group count controls how many lists are created, and thereby how finely
+ * the elements are grouped in time.  When reaping occurs, all the elements in
+ * all the lists whose time has expired are deleted.
+ *
+ * To give an example of how this works in practice, consider a client that
+ * initialises an MRU Cache with a lifetime of ten seconds and a group count of
+ * five.  Five internal lists will be created, each representing a two second
+ * period in time.  When the first element is added, time zero for the data
+ * structure is initialised to the current time.
+ *
+ * All the elements added in the first two seconds are appended to the first
+ * list.  Elements added in the third second go into the second list, and so on.
+ * If an element is accessed at any point, it is removed from its list and
+ * inserted at the head of the current most-recently-used list.
+ *
+ * The reaper function will have nothing to do until at least twelve seconds
+ * have elapsed since the first element was added.  The reason for this is that
+ * if it were called at t=11s, there could be elements in the first list that
+ * have only been inactive for nine seconds, so it still does nothing.  If it is
+ * called anywhere between t=12 and t=14 seconds, it will delete all the
+ * elements that remain in the first list.  It's therefore possible for elements
+ * to remain in the data store even after they've been inactive for up to
+ * (t + t/g) seconds, where t is the inactive element lifetime and g is the
+ * number of groups.
+ *
+ * The above example assumes that the reaper function gets called at least once
+ * every (t/g) seconds.  If it is called less frequently, unused elements will
+ * accumulate in the reap list until the reaper function is eventually called.
+ * The current implementation uses work queue callbacks to carefully time the
+ * reaper function calls, so this should happen rarely, if at all.
+ *
+ * From a design perspective, the primary reason for the choice of a list array
+ * representing discrete time intervals is that it's only practical to reap
+ * expired elements in groups of some appreciable size.  This automatically
+ * introduces a granularity to element lifetimes, so there's no point storing an
+ * individual timeout with each element that specifies a more precise reap time.
+ * The bonus is a saving of sizeof(long) bytes of memory per element stored.
+ *
+ * The elements could have been stored in just one list, but an array of
+ * counters or pointers would need to be maintained to allow them to be divided
+ * up into discrete time groups.  More critically, the process of touching or
+ * removing an element would involve walking large portions of the entire list,
+ * which would have a detrimental effect on performance.  The additional memory
+ * requirement for the array of list heads is minimal.
+ *
+ * When an element is touched or deleted, it needs to be removed from its
+ * current list.  Doubly linked lists are used to make the list maintenance
+ * portion of these operations O(1).  Since reaper timing can be imprecise,
+ * inserts and lookups can occur when there are no free lists available.  When
+ * this happens, all the elements on the LRU list need to be migrated to the end
+ * of the reap list.  To keep the list maintenance portion of these operations
+ * O(1) also, list tails need to be accessible without walking the entire list.
+ * This is the reason why doubly linked list heads are used.
+ */
+
+/*
+ * An MRU Cache is a dynamic data structure that stores its elements in a way
+ * that allows efficient lookups, but also groups them into discrete time
+ * intervals based on insertion time.  This allows elements to be efficiently
+ * and automatically reaped after a fixed period of inactivity.
+ *
+ * When a client data pointer is stored in the MRU Cache it needs to be added to
+ * both the data store and to one of the lists.  It must also be possible to
+ * access each of these entries via the other, i.e. to:
+ *
+ *    a) Walk a list, removing the corresponding data store entry for each item.
+ *    b) Look up a data store entry, then access its list entry directly.
+ *
+ * To achieve both of these goals, each entry must contain both a list entry and
+ * a key, in addition to the user's data pointer.  Note that it's not a good
+ * idea to have the client embed one of these structures at the top of their own
+ * data structure, because inserting the same item more than once would most
+ * likely result in a loop in one of the lists.  That's a sure-fire recipe for
+ * an infinite loop in the code.
+ */
+typedef struct xfs_mru_cache_elem
+{
+	struct list_head list_node;
+	unsigned long	key;
+	void		*value;
+} xfs_mru_cache_elem_t;
+
+static kmem_zone_t		*xfs_mru_elem_zone;
+static struct workqueue_struct	*xfs_mru_reap_wq;
+
+/*
+ * When inserting, destroying or reaping, it's first necessary to update the
+ * lists relative to a particular time.  In the case of destroying, that time
+ * will be well in the future to ensure that all items are moved to the reap
+ * list.  In all other cases though, the time will be the current time.
+ *
+ * This function enters a loop, moving the contents of the LRU list to the reap
+ * list again and again until either a) the lists are all empty, or b) time zero
+ * has been advanced sufficiently to be within the immediate element lifetime.
+ *
+ * Case a) above is detected by counting how many groups are migrated and
+ * stopping when they've all been moved.  Case b) is detected by monitoring the
+ * time_zero field, which is updated as each group is migrated.
+ *
+ * The return value is the earliest time that more migration could be needed, or
+ * zero if there's no need to schedule more work because the lists are empty.
+ */
+STATIC unsigned long
+_xfs_mru_cache_migrate(
+	xfs_mru_cache_t	*mru,
+	unsigned long	now)
+{
+	unsigned int	grp;
+	unsigned int	migrated = 0;
+	struct list_head *lru_list;
+
+	/* Nothing to do if the data store is empty. */
+	if (!mru->time_zero)
+		return 0;
+
+	/* While time zero is older than the time spanned by all the lists. */
+	while (mru->time_zero <= now - mru->grp_count * mru->grp_time) {
+
+		/*
+		 * If the LRU list isn't empty, migrate its elements to the tail
+		 * of the reap list.
+		 */
+		lru_list = mru->lists + mru->lru_grp;
+		if (!list_empty(lru_list))
+			list_splice_init(lru_list, mru->reap_list.prev);
+
+		/*
+		 * Advance the LRU group number, freeing the old LRU list to
+		 * become the new MRU list; advance time zero accordingly.
+		 */
+		mru->lru_grp = (mru->lru_grp + 1) % mru->grp_count;
+		mru->time_zero += mru->grp_time;
+
+		/*
+		 * If reaping is so far behind that all the elements on all the
+		 * lists have been migrated to the reap list, it's now empty.
+		 */
+		if (++migrated == mru->grp_count) {
+			mru->lru_grp = 0;
+			mru->time_zero = 0;
+			return 0;
+		}
+	}
+
+	/* Find the first non-empty list from the LRU end. */
+	for (grp = 0; grp < mru->grp_count; grp++) {
+
+		/* Check the grp'th list from the LRU end. */
+		lru_list = mru->lists + ((mru->lru_grp + grp) % mru->grp_count);
+		if (!list_empty(lru_list))
+			return mru->time_zero +
+			       (mru->grp_count + grp) * mru->grp_time;
+	}
+
+	/* All the lists must be empty. */
+	mru->lru_grp = 0;
+	mru->time_zero = 0;
+	return 0;
+}
+
+/*
+ * When inserting or doing a lookup, an element needs to be inserted into the
+ * MRU list.  The lists must be migrated first to ensure that they're
+ * up-to-date, otherwise the new element could be given a shorter lifetime in
+ * the cache than it should.
+ */
+STATIC void
+_xfs_mru_cache_list_insert(
+	xfs_mru_cache_t		*mru,
+	xfs_mru_cache_elem_t	*elem)
+{
+	unsigned int	grp = 0;
+	unsigned long	now = jiffies;
+
+	/*
+	 * If the data store is empty, initialise time zero, leave grp set to
+	 * zero and start the work queue timer if necessary.  Otherwise, set grp
+	 * to the number of group times that have elapsed since time zero.
+	 */
+	if (!_xfs_mru_cache_migrate(mru, now)) {
+		mru->time_zero = now;
+		if (!mru->next_reap)
+			mru->next_reap = mru->grp_count * mru->grp_time;
+	} else {
+		grp = (now - mru->time_zero) / mru->grp_time;
+		grp = (mru->lru_grp + grp) % mru->grp_count;
+	}
+
+	/* Insert the element at the tail of the corresponding list. */
+	list_add_tail(&elem->list_node, mru->lists + grp);
+}
+
+/*
+ * When destroying or reaping, all the elements that were migrated to the reap
+ * list need to be deleted.  For each element this involves removing it from the
+ * data store, removing it from the reap list, calling the client's free
+ * function and deleting the element from the element zone.
+ */
+STATIC void
+_xfs_mru_cache_clear_reap_list(
+	xfs_mru_cache_t		*mru)
+{
+	xfs_mru_cache_elem_t	*elem, *next;
+	struct list_head	tmp;
+
+	INIT_LIST_HEAD(&tmp);
+	list_for_each_entry_safe(elem, next, &mru->reap_list, list_node) {
+
+		/* Remove the element from the data store. */
+		radix_tree_delete(&mru->store, elem->key);
+
+		/*
+		 * remove to temp list so it can be freed without
+		 * needing to hold the lock
+		 */
+		list_move(&elem->list_node, &tmp);
+	}
+	mutex_spinunlock(&mru->lock, 0);
+
+	list_for_each_entry_safe(elem, next, &tmp, list_node) {
+
+		/* Remove the element from the reap list. */
+		list_del_init(&elem->list_node);
+
+		/* Call the client's free function with the key and value pointer. */
+		mru->free_func(elem->key, elem->value);
+
+		/* Free the element structure. */
+		kmem_zone_free(xfs_mru_elem_zone, elem);
+	}
+
+	mutex_spinlock(&mru->lock);
+}
+
+/*
+ * We fire the reap timer every group expiry interval so
+ * we always have a reaper ready to run. This makes shutdown
+ * and flushing of the reaper easy to do. Hence we need to
+ * keep when the next reap must occur so we can determine
+ * at each interval whether there is anything we need to do.
+ */
+STATIC void
+_xfs_mru_cache_reap(
+	struct work_struct	*work)
+{
+	xfs_mru_cache_t		*mru = container_of(work, xfs_mru_cache_t, work.work);
+	unsigned long		now;
+
+	ASSERT(mru && mru->lists);
+	if (!mru || !mru->lists)
+		return;
+
+	mutex_spinlock(&mru->lock);
+	now = jiffies;
+	if (mru->reap_all ||
+	    (mru->next_reap && time_after(now, mru->next_reap))) {
+		if (mru->reap_all)
+			now += mru->grp_count * mru->grp_time * 2;
+		mru->next_reap = _xfs_mru_cache_migrate(mru, now);
+		_xfs_mru_cache_clear_reap_list(mru);
+	}
+
+	/*
+	 * the process that triggered the reap_all is responsible
+	 * for restating the periodic reap if it is required.
+	 */
+	if (!mru->reap_all)
+		queue_delayed_work(xfs_mru_reap_wq, &mru->work, mru->grp_time);
+	mru->reap_all = 0;
+	mutex_spinunlock(&mru->lock, 0);
+}
+
+int
+xfs_mru_cache_init(void)
+{
+	xfs_mru_elem_zone = kmem_zone_init(sizeof(xfs_mru_cache_elem_t),
+	                                 "xfs_mru_cache_elem");
+	if (!xfs_mru_elem_zone)
+		return ENOMEM;
+
+	xfs_mru_reap_wq = create_singlethread_workqueue("xfs_mru_cache");
+	if (!xfs_mru_reap_wq) {
+		kmem_zone_destroy(xfs_mru_elem_zone);
+		return ENOMEM;
+	}
+
+	return 0;
+}
+
+void
+xfs_mru_cache_uninit(void)
+{
+	destroy_workqueue(xfs_mru_reap_wq);
+	kmem_zone_destroy(xfs_mru_elem_zone);
+}
+
+/*
+ * To initialise a struct xfs_mru_cache pointer, call xfs_mru_cache_create()
+ * with the address of the pointer, a lifetime value in milliseconds, a group
+ * count and a free function to use when deleting elements.  This function
+ * returns 0 if the initialisation was successful.
+ */
+int
+xfs_mru_cache_create(
+	xfs_mru_cache_t		**mrup,
+	unsigned int		lifetime_ms,
+	unsigned int		grp_count,
+	xfs_mru_cache_free_func_t free_func)
+{
+	xfs_mru_cache_t	*mru = NULL;
+	int		err = 0, grp;
+	unsigned int	grp_time;
+
+	if (mrup)
+		*mrup = NULL;
+
+	if (!mrup || !grp_count || !lifetime_ms || !free_func)
+		return EINVAL;
+
+	if (!(grp_time = msecs_to_jiffies(lifetime_ms) / grp_count))
+		return EINVAL;
+
+	if (!(mru = kmem_zalloc(sizeof(*mru), KM_SLEEP)))
+		return ENOMEM;
+
+	/* An extra list is needed to avoid reaping up to a grp_time early. */
+	mru->grp_count = grp_count + 1;
+	mru->lists = kmem_alloc(mru->grp_count * sizeof(*mru->lists), KM_SLEEP);
+
+	if (!mru->lists) {
+		err = ENOMEM;
+		goto exit;
+	}
+
+	for (grp = 0; grp < mru->grp_count; grp++)
+		INIT_LIST_HEAD(mru->lists + grp);
+
+	/*
+	 * We use GFP_KERNEL radix tree preload and do inserts under a
+	 * spinlock so GFP_ATOMIC is appropriate for the radix tree itself.
+	 */
+	INIT_RADIX_TREE(&mru->store, GFP_ATOMIC);
+	INIT_LIST_HEAD(&mru->reap_list);
+	spinlock_init(&mru->lock, "xfs_mru_cache");
+	INIT_DELAYED_WORK(&mru->work, _xfs_mru_cache_reap);
+
+	mru->grp_time  = grp_time;
+	mru->free_func = free_func;
+
+	/* start up the reaper event */
+	mru->next_reap = 0;
+	mru->reap_all = 0;
+	queue_delayed_work(xfs_mru_reap_wq, &mru->work, mru->grp_time);
+
+	*mrup = mru;
+
+exit:
+	if (err && mru && mru->lists)
+		kmem_free(mru->lists, mru->grp_count * sizeof(*mru->lists));
+	if (err && mru)
+		kmem_free(mru, sizeof(*mru));
+
+	return err;
+}
+
+/*
+ * Call xfs_mru_cache_flush() to flush out all cached entries, calling their
+ * free functions as they're deleted.  When this function returns, the caller is
+ * guaranteed that all the free functions for all the elements have finished
+ * executing.
+ *
+ * While we are flushing, we stop the periodic reaper event from triggering.
+ * Normally, we want to restart this periodic event, but if we are shutting
+ * down the cache we do not want it restarted. hence the restart parameter
+ * where 0 = do not restart reaper and 1 = restart reaper.
+ */
+void
+xfs_mru_cache_flush(
+	xfs_mru_cache_t		*mru,
+	int			restart)
+{
+	if (!mru || !mru->lists)
+		return;
+
+	cancel_rearming_delayed_workqueue(xfs_mru_reap_wq, &mru->work);
+
+	mutex_spinlock(&mru->lock);
+	mru->reap_all = 1;
+	mutex_spinunlock(&mru->lock, 0);
+
+	queue_work(xfs_mru_reap_wq, &mru->work.work);
+	flush_workqueue(xfs_mru_reap_wq);
+
+	mutex_spinlock(&mru->lock);
+	WARN_ON_ONCE(mru->reap_all != 0);
+	mru->reap_all = 0;
+	if (restart)
+		queue_delayed_work(xfs_mru_reap_wq, &mru->work, mru->grp_time);
+	mutex_spinunlock(&mru->lock, 0);
+}
+
+void
+xfs_mru_cache_destroy(
+	xfs_mru_cache_t		*mru)
+{
+	if (!mru || !mru->lists)
+		return;
+
+	/* we don't want the reaper to restart here */
+	xfs_mru_cache_flush(mru, 0);
+
+	kmem_free(mru->lists, mru->grp_count * sizeof(*mru->lists));
+	kmem_free(mru, sizeof(*mru));
+}
+
+/*
+ * To insert an element, call xfs_mru_cache_insert() with the data store, the
+ * element's key and the client data pointer.  This function returns 0 on
+ * success or ENOMEM if memory for the data element couldn't be allocated.
+ */
+int
+xfs_mru_cache_insert(
+	xfs_mru_cache_t	*mru,
+	unsigned long	key,
+	void		*value)
+{
+	xfs_mru_cache_elem_t *elem;
+
+	ASSERT(mru && mru->lists);
+	if (!mru || !mru->lists)
+		return EINVAL;
+
+	elem = kmem_zone_zalloc(xfs_mru_elem_zone, KM_SLEEP);
+	if (!elem)
+		return ENOMEM;
+
+	if (radix_tree_preload(GFP_KERNEL)) {
+		kmem_zone_free(xfs_mru_elem_zone, elem);
+		return ENOMEM;
+	}
+
+	INIT_LIST_HEAD(&elem->list_node);
+	elem->key = key;
+	elem->value = value;
+
+	mutex_spinlock(&mru->lock);
+
+	radix_tree_insert(&mru->store, key, elem);
+	radix_tree_preload_end();
+	_xfs_mru_cache_list_insert(mru, elem);
+
+	mutex_spinunlock(&mru->lock, 0);
+
+	return 0;
+}
+
+/*
+ * To remove an element without calling the free function, call
+ * xfs_mru_cache_remove() with the data store and the element's key.  On success
+ * the client data pointer for the removed element is returned, otherwise this
+ * function will return a NULL pointer.
+ */
+void *
+xfs_mru_cache_remove(
+	xfs_mru_cache_t	*mru,
+	unsigned long	key)
+{
+	xfs_mru_cache_elem_t *elem;
+	void		*value = NULL;
+
+	ASSERT(mru && mru->lists);
+	if (!mru || !mru->lists)
+		return NULL;
+
+	mutex_spinlock(&mru->lock);
+	elem = radix_tree_delete(&mru->store, key);
+	if (elem) {
+		value = elem->value;
+		list_del(&elem->list_node);
+	}
+
+	mutex_spinunlock(&mru->lock, 0);
+
+	if (elem)
+		kmem_zone_free(xfs_mru_elem_zone, elem);
+
+	return value;
+}
+
+/*
+ * To remove and element and call the free function, call xfs_mru_cache_delete()
+ * with the data store and the element's key.
+ */
+void
+xfs_mru_cache_delete(
+	xfs_mru_cache_t	*mru,
+	unsigned long	key)
+{
+	void		*value = xfs_mru_cache_remove(mru, key);
+
+	if (value)
+		mru->free_func(key, value);
+}
+
+/*
+ * To look up an element using its key, call xfs_mru_cache_lookup() with the
+ * data store and the element's key.  If found, the element will be moved to the
+ * head of the MRU list to indicate that it's been touched.
+ *
+ * The internal data structures are protected by a spinlock that is STILL HELD
+ * when this function returns.  Call xfs_mru_cache_done() to release it.  Note
+ * that it is not safe to call any function that might sleep in the interim.
+ *
+ * The implementation could have used reference counting to avoid this
+ * restriction, but since most clients simply want to get, set or test a member
+ * of the returned data structure, the extra per-element memory isn't warranted.
+ *
+ * If the element isn't found, this function returns NULL and the spinlock is
+ * released.  xfs_mru_cache_done() should NOT be called when this occurs.
+ */
+void *
+xfs_mru_cache_lookup(
+	xfs_mru_cache_t	*mru,
+	unsigned long	key)
+{
+	xfs_mru_cache_elem_t *elem;
+
+	ASSERT(mru && mru->lists);
+	if (!mru || !mru->lists)
+		return NULL;
+
+	mutex_spinlock(&mru->lock);
+	elem = radix_tree_lookup(&mru->store, key);
+	if (elem) {
+		list_del(&elem->list_node);
+		_xfs_mru_cache_list_insert(mru, elem);
+	}
+	else
+		mutex_spinunlock(&mru->lock, 0);
+
+	return elem ? elem->value : NULL;
+}
+
+/*
+ * To look up an element using its key, but leave its location in the internal
+ * lists alone, call xfs_mru_cache_peek().  If the element isn't found, this
+ * function returns NULL.
+ *
+ * See the comments above the declaration of the xfs_mru_cache_lookup() function
+ * for important locking information pertaining to this call.
+ */
+void *
+xfs_mru_cache_peek(
+	xfs_mru_cache_t	*mru,
+	unsigned long	key)
+{
+	xfs_mru_cache_elem_t *elem;
+
+	ASSERT(mru && mru->lists);
+	if (!mru || !mru->lists)
+		return NULL;
+
+	mutex_spinlock(&mru->lock);
+	elem = radix_tree_lookup(&mru->store, key);
+	if (!elem)
+		mutex_spinunlock(&mru->lock, 0);
+
+	return elem ? elem->value : NULL;
+}
+
+/*
+ * To release the internal data structure spinlock after having performed an
+ * xfs_mru_cache_lookup() or an xfs_mru_cache_peek(), call xfs_mru_cache_done()
+ * with the data store pointer.
+ */
+void
+xfs_mru_cache_done(
+	xfs_mru_cache_t	*mru)
+{
+	mutex_spinunlock(&mru->lock, 0);
+}

fs/xfs/xfs_mru_cache.h

@@ -0,0 +1,57 @@
+/*
+ * Copyright (c) 2006-2007 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+#ifndef __XFS_MRU_CACHE_H__
+#define __XFS_MRU_CACHE_H__
+
+
+/* Function pointer type for callback to free a client's data pointer. */
+typedef void (*xfs_mru_cache_free_func_t)(unsigned long, void*);
+
+typedef struct xfs_mru_cache
+{
+	struct radix_tree_root	store;     /* Core storage data structure.  */
+	struct list_head	*lists;    /* Array of lists, one per grp.  */
+	struct list_head	reap_list; /* Elements overdue for reaping. */
+	spinlock_t		lock;      /* Lock to protect this struct.  */
+	unsigned int		grp_count; /* Number of discrete groups.    */
+	unsigned int		grp_time;  /* Time period spanned by grps.  */
+	unsigned int		lru_grp;   /* Group containing time zero.   */
+	unsigned long		time_zero; /* Time first element was added. */
+	unsigned long		next_reap; /* Time that the reaper should
+					      next do something. */
+	unsigned int		reap_all;  /* if set, reap all lists */
+	xfs_mru_cache_free_func_t free_func; /* Function pointer for freeing. */
+	struct delayed_work	work;      /* Workqueue data for reaping.   */
+} xfs_mru_cache_t;
+
+int xfs_mru_cache_init(void);
+void xfs_mru_cache_uninit(void);
+int xfs_mru_cache_create(struct xfs_mru_cache **mrup, unsigned int lifetime_ms,
+			     unsigned int grp_count,
+			     xfs_mru_cache_free_func_t free_func);
+void xfs_mru_cache_flush(xfs_mru_cache_t *mru, int restart);
+void xfs_mru_cache_destroy(struct xfs_mru_cache *mru);
+int xfs_mru_cache_insert(struct xfs_mru_cache *mru, unsigned long key,
+				void *value);
+void * xfs_mru_cache_remove(struct xfs_mru_cache *mru, unsigned long key);
+void xfs_mru_cache_delete(struct xfs_mru_cache *mru, unsigned long key);
+void *xfs_mru_cache_lookup(struct xfs_mru_cache *mru, unsigned long key);
+void *xfs_mru_cache_peek(struct xfs_mru_cache *mru, unsigned long key);
+void xfs_mru_cache_done(struct xfs_mru_cache *mru);
+
+#endif /* __XFS_MRU_CACHE_H__ */

fs/xfs/xfs_rtalloc.c

@@ -1882,11 +1882,13 @@ xfs_growfs_rt(
 	    (nrblocks = in->newblocks) <= sbp->sb_rblocks ||
 	    (sbp->sb_rblocks && (in->extsize != sbp->sb_rextsize)))
 		return XFS_ERROR(EINVAL);
+	if ((error = xfs_sb_validate_fsb_count(sbp, nrblocks)))
+		return error;
 	/*
 	 * Read in the last block of the device, make sure it exists.
 	 */
 	error = xfs_read_buf(mp, mp->m_rtdev_targp,
-			XFS_FSB_TO_BB(mp, in->newblocks - 1),
+			XFS_FSB_TO_BB(mp, nrblocks - 1),
 			XFS_FSB_TO_BB(mp, 1), 0, &bp);
 	if (error)
 		return error;

fs/xfs/xfs_rw.h

@@ -71,6 +71,34 @@ xfs_fsb_to_db_io(struct xfs_iocore *io, xfs_fsblock_t fsb)
 		 XFS_FSB_TO_DADDR((io)->io_mount, (fsb)));
 }
 
+/*
+ * Flags for xfs_free_eofblocks
+ */
+#define XFS_FREE_EOF_LOCK	(1<<0)
+#define XFS_FREE_EOF_NOLOCK	(1<<1)
+
+
+/*
+ * helper function to extract extent size hint from inode
+ */
+STATIC_INLINE xfs_extlen_t
+xfs_get_extsz_hint(
+	xfs_inode_t	*ip)
+{
+	xfs_extlen_t	extsz;
+
+	if (unlikely(ip->i_d.di_flags & XFS_DIFLAG_REALTIME)) {
+		extsz = (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE)
+				? ip->i_d.di_extsize
+				: ip->i_mount->m_sb.sb_rextsize;
+		ASSERT(extsz);
+	} else {
+		extsz = (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE)
+				? ip->i_d.di_extsize : 0;
+	}
+	return extsz;
+}
+
 /*
  * Prototypes for functions in xfs_rw.c.
  */
@@ -91,10 +119,12 @@ extern void xfs_ioerror_alert(char *func, struct xfs_mount *mp,
 extern int xfs_rwlock(bhv_desc_t *bdp, bhv_vrwlock_t write_lock);
 extern void xfs_rwunlock(bhv_desc_t *bdp, bhv_vrwlock_t write_lock);
 extern int xfs_setattr(bhv_desc_t *, bhv_vattr_t *vap, int flags,
-		       cred_t *credp);
+			cred_t *credp);
 extern int xfs_change_file_space(bhv_desc_t *bdp, int cmd, xfs_flock64_t *bf,
-				 xfs_off_t offset, cred_t *credp, int flags);
+			xfs_off_t offset, cred_t *credp, int flags);
 extern int xfs_set_dmattrs(bhv_desc_t *bdp, u_int evmask, u_int16_t state,
-			   cred_t *credp);
+			cred_t *credp);
+extern int xfs_free_eofblocks(struct xfs_mount *mp, struct xfs_inode *ip,
+			int flags);
 
 #endif /* __XFS_RW_H__ */

fs/xfs/xfs_sb.h

@@ -74,12 +74,13 @@ struct xfs_mount;
  */
 #define XFS_SB_VERSION2_REALFBITS	0x00ffffff	/* Mask: features */
 #define XFS_SB_VERSION2_RESERVED1BIT	0x00000001
-#define XFS_SB_VERSION2_RESERVED2BIT	0x00000002
+#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_OKREALFBITS	\
-	(XFS_SB_VERSION2_ATTR2BIT)
+	(XFS_SB_VERSION2_LAZYSBCOUNTBIT	| \
+	 XFS_SB_VERSION2_ATTR2BIT)
 #define	XFS_SB_VERSION2_OKSASHFBITS	\
 	(0)
 #define XFS_SB_VERSION2_OKREALBITS	\
@@ -181,6 +182,9 @@ typedef enum {
 #define XFS_SB_SHARED_VN	XFS_SB_MVAL(SHARED_VN)
 #define XFS_SB_UNIT		XFS_SB_MVAL(UNIT)
 #define XFS_SB_WIDTH		XFS_SB_MVAL(WIDTH)
+#define XFS_SB_ICOUNT		XFS_SB_MVAL(ICOUNT)
+#define XFS_SB_IFREE		XFS_SB_MVAL(IFREE)
+#define XFS_SB_FDBLOCKS		XFS_SB_MVAL(FDBLOCKS)
 #define XFS_SB_FEATURES2	XFS_SB_MVAL(FEATURES2)
 #define	XFS_SB_NUM_BITS		((int)XFS_SBS_FIELDCOUNT)
 #define	XFS_SB_ALL_BITS		((1LL << XFS_SB_NUM_BITS) - 1)
@@ -188,7 +192,7 @@ typedef enum {
 	(XFS_SB_UUID | XFS_SB_ROOTINO | XFS_SB_RBMINO | XFS_SB_RSUMINO | \
 	 XFS_SB_VERSIONNUM | XFS_SB_UQUOTINO | XFS_SB_GQUOTINO | \
 	 XFS_SB_QFLAGS | XFS_SB_SHARED_VN | XFS_SB_UNIT | XFS_SB_WIDTH | \
-	 XFS_SB_FEATURES2)
+	 XFS_SB_ICOUNT | XFS_SB_IFREE | XFS_SB_FDBLOCKS | XFS_SB_FEATURES2)
 
 
 /*
@@ -414,6 +418,12 @@ static inline int xfs_sb_version_hasmorebits(xfs_sb_t *sbp)
  *	 ((sbp)->sb_features2 & XFS_SB_VERSION2_FUNBIT)
  */
 
+static inline int xfs_sb_version_haslazysbcount(xfs_sb_t *sbp)
+{
+	return (XFS_SB_VERSION_HASMOREBITS(sbp) &&	\
+		((sbp)->sb_features2 & XFS_SB_VERSION2_LAZYSBCOUNTBIT));
+}
+
 #define XFS_SB_VERSION_HASATTR2(sbp)	xfs_sb_version_hasattr2(sbp)
 static inline int xfs_sb_version_hasattr2(xfs_sb_t *sbp)
 {

fs/xfs/xfs_trans.c

@@ -427,6 +427,14 @@ undo_blocks:
  *
  * Mark the transaction structure to indicate that the superblock
  * needs to be updated before committing.
+ *
+ * Because we may not be keeping track of allocated/free inodes and
+ * used filesystem blocks in the superblock, we do not mark the
+ * superblock dirty in this transaction if we modify these fields.
+ * We still need to update the transaction deltas so that they get
+ * applied to the incore superblock, but we don't want them to
+ * cause the superblock to get locked and logged if these are the
+ * only fields in the superblock that the transaction modifies.
  */
 void
 xfs_trans_mod_sb(
@@ -434,13 +442,19 @@ xfs_trans_mod_sb(
 	uint		field,
 	int64_t		delta)
 {
+	uint32_t	flags = (XFS_TRANS_DIRTY|XFS_TRANS_SB_DIRTY);
+	xfs_mount_t	*mp = tp->t_mountp;
 
 	switch (field) {
 	case XFS_TRANS_SB_ICOUNT:
 		tp->t_icount_delta += delta;
+		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
+			flags &= ~XFS_TRANS_SB_DIRTY;
 		break;
 	case XFS_TRANS_SB_IFREE:
 		tp->t_ifree_delta += delta;
+		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
+			flags &= ~XFS_TRANS_SB_DIRTY;
 		break;
 	case XFS_TRANS_SB_FDBLOCKS:
 		/*
@@ -453,6 +467,8 @@ xfs_trans_mod_sb(
 			ASSERT(tp->t_blk_res_used <= tp->t_blk_res);
 		}
 		tp->t_fdblocks_delta += delta;
+		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
+			flags &= ~XFS_TRANS_SB_DIRTY;
 		break;
 	case XFS_TRANS_SB_RES_FDBLOCKS:
 		/*
@@ -462,6 +478,8 @@ xfs_trans_mod_sb(
 		 */
 		ASSERT(delta < 0);
 		tp->t_res_fdblocks_delta += delta;
+		if (xfs_sb_version_haslazysbcount(&mp->m_sb))
+			flags &= ~XFS_TRANS_SB_DIRTY;
 		break;
 	case XFS_TRANS_SB_FREXTENTS:
 		/*
@@ -515,7 +533,7 @@ xfs_trans_mod_sb(
 		return;
 	}
 
-	tp->t_flags |= (XFS_TRANS_SB_DIRTY | XFS_TRANS_DIRTY);
+	tp->t_flags |= flags;
 }
 
 /*
@@ -544,18 +562,23 @@ xfs_trans_apply_sb_deltas(
 	       (tp->t_ag_freeblks_delta + tp->t_ag_flist_delta +
 		tp->t_ag_btree_delta));
 
-	if (tp->t_icount_delta != 0) {
-		INT_MOD(sbp->sb_icount, ARCH_CONVERT, tp->t_icount_delta);
-	}
-	if (tp->t_ifree_delta != 0) {
-		INT_MOD(sbp->sb_ifree, ARCH_CONVERT, tp->t_ifree_delta);
-	}
+	/*
+	 * Only update the superblock counters if we are logging them
+	 */
+	if (!xfs_sb_version_haslazysbcount(&(tp->t_mountp->m_sb))) {
+		if (tp->t_icount_delta != 0) {
+			INT_MOD(sbp->sb_icount, ARCH_CONVERT, tp->t_icount_delta);
+		}
+		if (tp->t_ifree_delta != 0) {
+			INT_MOD(sbp->sb_ifree, ARCH_CONVERT, tp->t_ifree_delta);
+		}
 
-	if (tp->t_fdblocks_delta != 0) {
-		INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_fdblocks_delta);
-	}
-	if (tp->t_res_fdblocks_delta != 0) {
-		INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_res_fdblocks_delta);
+		if (tp->t_fdblocks_delta != 0) {
+			INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_fdblocks_delta);
+		}
+		if (tp->t_res_fdblocks_delta != 0) {
+			INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_res_fdblocks_delta);
+		}
 	}
 
 	if (tp->t_frextents_delta != 0) {
@@ -615,11 +638,23 @@ xfs_trans_apply_sb_deltas(
 }
 
 /*
- * xfs_trans_unreserve_and_mod_sb() is called to release unused
- * reservations and apply superblock counter changes to the in-core
- * superblock.
+ * xfs_trans_unreserve_and_mod_sb() is called to release unused reservations
+ * and apply superblock counter changes to the in-core superblock.  The
+ * t_res_fdblocks_delta and t_res_frextents_delta fields are explicitly NOT
+ * applied to the in-core superblock.  The idea is that that has already been
+ * done.
  *
  * This is done efficiently with a single call to xfs_mod_incore_sb_batch().
+ * However, we have to ensure that we only modify each superblock field only
+ * once because the application of the delta values may not be atomic. That can
+ * lead to ENOSPC races occurring if we have two separate modifcations of the
+ * free space counter to put back the entire reservation and then take away
+ * what we used.
+ *
+ * If we are not logging superblock counters, then the inode allocated/free and
+ * used block counts are not updated in the on disk superblock. In this case,
+ * XFS_TRANS_SB_DIRTY will not be set when the transaction is updated but we
+ * still need to update the incore superblock with the changes.
  */
 STATIC void
 xfs_trans_unreserve_and_mod_sb(
@@ -627,40 +662,49 @@ xfs_trans_unreserve_and_mod_sb(
 {
 	xfs_mod_sb_t	msb[14];	/* If you add cases, add entries */
 	xfs_mod_sb_t	*msbp;
+	xfs_mount_t	*mp = tp->t_mountp;
 	/* REFERENCED */
 	int		error;
 	int		rsvd;
+	int64_t		blkdelta = 0;
+	int64_t		rtxdelta = 0;
 
 	msbp = msb;
 	rsvd = (tp->t_flags & XFS_TRANS_RESERVE) != 0;
 
-	/*
-	 * Release any reserved blocks.  Any that were allocated
-	 * will be taken back again by fdblocks_delta below.
-	 */
-	if (tp->t_blk_res > 0) {
+	/* calculate free blocks delta */
+	if (tp->t_blk_res > 0)
+		blkdelta = tp->t_blk_res;
+
+	if ((tp->t_fdblocks_delta != 0) &&
+	    (xfs_sb_version_haslazysbcount(&mp->m_sb) ||
+	     (tp->t_flags & XFS_TRANS_SB_DIRTY)))
+	        blkdelta += tp->t_fdblocks_delta;
+
+	if (blkdelta != 0) {
 		msbp->msb_field = XFS_SBS_FDBLOCKS;
-		msbp->msb_delta = tp->t_blk_res;
+		msbp->msb_delta = blkdelta;
 		msbp++;
 	}
 
-	/*
-	 * Release any reserved real time extents .  Any that were
-	 * allocated will be taken back again by frextents_delta below.
-	 */
-	if (tp->t_rtx_res > 0) {
+	/* calculate free realtime extents delta */
+	if (tp->t_rtx_res > 0)
+		rtxdelta = tp->t_rtx_res;
+
+	if ((tp->t_frextents_delta != 0) &&
+	    (tp->t_flags & XFS_TRANS_SB_DIRTY))
+		rtxdelta += tp->t_frextents_delta;
+
+	if (rtxdelta != 0) {
 		msbp->msb_field = XFS_SBS_FREXTENTS;
-		msbp->msb_delta = tp->t_rtx_res;
+		msbp->msb_delta = rtxdelta;
 		msbp++;
 	}
 
-	/*
-	 * Apply any superblock modifications to the in-core version.
-	 * The t_res_fdblocks_delta and t_res_frextents_delta fields are
-	 * explicitly NOT applied to the in-core superblock.
-	 * The idea is that that has already been done.
-	 */
-	if (tp->t_flags & XFS_TRANS_SB_DIRTY) {
+	/* apply remaining deltas */
+
+	if (xfs_sb_version_haslazysbcount(&mp->m_sb) ||
+	     (tp->t_flags & XFS_TRANS_SB_DIRTY)) {
 		if (tp->t_icount_delta != 0) {
 			msbp->msb_field = XFS_SBS_ICOUNT;
 			msbp->msb_delta = tp->t_icount_delta;
@@ -671,16 +715,9 @@ xfs_trans_unreserve_and_mod_sb(
 			msbp->msb_delta = tp->t_ifree_delta;
 			msbp++;
 		}
-		if (tp->t_fdblocks_delta != 0) {
-			msbp->msb_field = XFS_SBS_FDBLOCKS;
-			msbp->msb_delta = tp->t_fdblocks_delta;
-			msbp++;
-		}
-		if (tp->t_frextents_delta != 0) {
-			msbp->msb_field = XFS_SBS_FREXTENTS;
-			msbp->msb_delta = tp->t_frextents_delta;
-			msbp++;
-		}
+	}
+
+	if (tp->t_flags & XFS_TRANS_SB_DIRTY) {
 		if (tp->t_dblocks_delta != 0) {
 			msbp->msb_field = XFS_SBS_DBLOCKS;
 			msbp->msb_delta = tp->t_dblocks_delta;

fs/xfs/xfs_trans.h

@@ -94,7 +94,8 @@ typedef struct xfs_trans_header {
 #define	XFS_TRANS_GROWFSRT_ZERO		38
 #define	XFS_TRANS_GROWFSRT_FREE		39
 #define	XFS_TRANS_SWAPEXT		40
-#define	XFS_TRANS_TYPE_MAX		40
+#define	XFS_TRANS_SB_COUNT		41
+#define	XFS_TRANS_TYPE_MAX		41
 /* new transaction types need to be reflected in xfs_logprint(8) */
 
 

fs/xfs/xfs_vfsops.c

@@ -51,6 +51,8 @@
 #include "xfs_acl.h"
 #include "xfs_attr.h"
 #include "xfs_clnt.h"
+#include "xfs_mru_cache.h"
+#include "xfs_filestream.h"
 #include "xfs_fsops.h"
 
 STATIC int	xfs_sync(bhv_desc_t *, int, cred_t *);
@@ -81,6 +83,8 @@ xfs_init(void)
 	xfs_dabuf_zone = kmem_zone_init(sizeof(xfs_dabuf_t), "xfs_dabuf");
 	xfs_ifork_zone = kmem_zone_init(sizeof(xfs_ifork_t), "xfs_ifork");
 	xfs_acl_zone_init(xfs_acl_zone, "xfs_acl");
+	xfs_mru_cache_init();
+	xfs_filestream_init();
 
 	/*
 	 * The size of the zone allocated buf log item is the maximum
@@ -164,6 +168,8 @@ xfs_cleanup(void)
 	xfs_cleanup_procfs();
 	xfs_sysctl_unregister();
 	xfs_refcache_destroy();
+	xfs_filestream_uninit();
+	xfs_mru_cache_uninit();
 	xfs_acl_zone_destroy(xfs_acl_zone);
 
 #ifdef XFS_DIR2_TRACE
@@ -320,6 +326,9 @@ xfs_start_flags(
 	else
 		mp->m_flags &= ~XFS_MOUNT_BARRIER;
 
+	if (ap->flags2 & XFSMNT2_FILESTREAMS)
+		mp->m_flags |= XFS_MOUNT_FILESTREAMS;
+
 	return 0;
 }
 
@@ -518,6 +527,9 @@ xfs_mount(
 	if (mp->m_flags & XFS_MOUNT_BARRIER)
 		xfs_mountfs_check_barriers(mp);
 
+	if ((error = xfs_filestream_mount(mp)))
+		goto error2;
+
 	error = XFS_IOINIT(vfsp, args, flags);
 	if (error)
 		goto error2;
@@ -575,6 +587,13 @@ xfs_unmount(
 	 */
 	xfs_refcache_purge_mp(mp);
 
+	/*
+	 * Blow away any referenced inode in the filestreams cache.
+	 * This can and will cause log traffic as inodes go inactive
+	 * here.
+	 */
+	xfs_filestream_unmount(mp);
+
 	XFS_bflush(mp->m_ddev_targp);
 	error = xfs_unmount_flush(mp, 0);
 	if (error)
@@ -640,7 +659,7 @@ xfs_quiesce_fs(
 	 * we can write the unmount record.
 	 */
 	do {
-		xfs_syncsub(mp, SYNC_REMOUNT|SYNC_ATTR|SYNC_WAIT, NULL);
+		xfs_syncsub(mp, SYNC_INODE_QUIESCE, NULL);
 		pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
 		if (!pincount) {
 			delay(50);
@@ -651,6 +670,30 @@ xfs_quiesce_fs(
 	return 0;
 }
 
+/*
+ * Second stage of a quiesce. The data is already synced, now we have to take
+ * care of the metadata. New transactions are already blocked, so we need to
+ * wait for any remaining transactions to drain out before proceding.
+ */
+STATIC void
+xfs_attr_quiesce(
+	xfs_mount_t	*mp)
+{
+	/* wait for all modifications to complete */
+	while (atomic_read(&mp->m_active_trans) > 0)
+		delay(100);
+
+	/* flush inodes and push all remaining buffers out to disk */
+	xfs_quiesce_fs(mp);
+
+	ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0);
+
+	/* Push the superblock and write an unmount record */
+	xfs_log_sbcount(mp, 1);
+	xfs_log_unmount_write(mp);
+	xfs_unmountfs_writesb(mp);
+}
+
 STATIC int
 xfs_mntupdate(
 	bhv_desc_t			*bdp,
@@ -670,10 +713,9 @@ xfs_mntupdate(
 			mp->m_flags &= ~XFS_MOUNT_BARRIER;
 		}
 	} else if (!(vfsp->vfs_flag & VFS_RDONLY)) {	/* rw -> ro */
-		bhv_vfs_sync(vfsp, SYNC_FSDATA|SYNC_BDFLUSH|SYNC_ATTR, NULL);
-		xfs_quiesce_fs(mp);
-		xfs_log_unmount_write(mp);
-		xfs_unmountfs_writesb(mp);
+		xfs_filestream_flush(mp);
+		bhv_vfs_sync(vfsp, SYNC_DATA_QUIESCE, NULL);
+		xfs_attr_quiesce(mp);
 		vfsp->vfs_flag |= VFS_RDONLY;
 	}
 	return 0;
@@ -887,6 +929,9 @@ xfs_sync(
 {
 	xfs_mount_t	*mp = XFS_BHVTOM(bdp);
 
+	if (flags & SYNC_IOWAIT)
+		xfs_filestream_flush(mp);
+
 	return xfs_syncsub(mp, flags, NULL);
 }
 
@@ -1128,58 +1173,41 @@ xfs_sync_inodes(
 		 * in the inode list.
 		 */
 
-		if ((flags & SYNC_CLOSE)  && (vp != NULL)) {
-			/*
-			 * This is the shutdown case.  We just need to
-			 * flush and invalidate all the pages associated
-			 * with the inode.  Drop the inode lock since
-			 * we can't hold it across calls to the buffer
-			 * cache.
-			 *
-			 * We don't set the VREMAPPING bit in the vnode
-			 * here, because we don't hold the vnode lock
-			 * exclusively.  It doesn't really matter, though,
-			 * because we only come here when we're shutting
-			 * down anyway.
-			 */
-			xfs_iunlock(ip, XFS_ILOCK_SHARED);
-
-			if (XFS_FORCED_SHUTDOWN(mp)) {
-				bhv_vop_toss_pages(vp, 0, -1, FI_REMAPF);
-			} else {
-				error = bhv_vop_flushinval_pages(vp, 0, -1, FI_REMAPF);
+		/*
+		 * If we have to flush data or wait for I/O completion
+		 * we need to drop the ilock that we currently hold.
+		 * If we need to drop the lock, insert a marker if we
+		 * have not already done so.
+		 */
+		if ((flags & (SYNC_CLOSE|SYNC_IOWAIT)) ||
+		    ((flags & SYNC_DELWRI) && VN_DIRTY(vp))) {
+			if (mount_locked) {
+				IPOINTER_INSERT(ip, mp);
 			}
+			xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
-			xfs_ilock(ip, XFS_ILOCK_SHARED);
-
-		} else if ((flags & SYNC_DELWRI) && (vp != NULL)) {
-			if (VN_DIRTY(vp)) {
-				/* We need to have dropped the lock here,
-				 * so insert a marker if we have not already
-				 * done so.
-				 */
-				if (mount_locked) {
-					IPOINTER_INSERT(ip, mp);
-				}
-
-				/*
-				 * Drop the inode lock since we can't hold it
-				 * across calls to the buffer cache.
-				 */
-				xfs_iunlock(ip, XFS_ILOCK_SHARED);
+			if (flags & SYNC_CLOSE) {
+				/* Shutdown case. Flush and invalidate. */
+				if (XFS_FORCED_SHUTDOWN(mp))
+					bhv_vop_toss_pages(vp, 0, -1, FI_REMAPF);
+				else
+					error = bhv_vop_flushinval_pages(vp, 0,
+								-1, FI_REMAPF);
+			} else if ((flags & SYNC_DELWRI) && VN_DIRTY(vp)) {
 				error = bhv_vop_flush_pages(vp, (xfs_off_t)0,
 							-1, fflag, FI_NONE);
-				xfs_ilock(ip, XFS_ILOCK_SHARED);
 			}
 
+			/*
+			 * When freezing, we need to wait ensure all I/O (including direct
+			 * I/O) is complete to ensure no further data modification can take
+			 * place after this point
+			 */
+			if (flags & SYNC_IOWAIT)
+				vn_iowait(vp);
+
+			xfs_ilock(ip, XFS_ILOCK_SHARED);
 		}
-		/*
-		 * When freezing, we need to wait ensure all I/O (including direct
-		 * I/O) is complete to ensure no further data modification can take
-		 * place after this point
-		 */
-		if (flags & SYNC_IOWAIT)
-			vn_iowait(vp);
 
 		if (flags & SYNC_BDFLUSH) {
 			if ((flags & SYNC_ATTR) &&
@@ -1513,6 +1541,15 @@ xfs_syncsub(
 			xfs_refcache_purge_some(mp);
 	}
 
+	/*
+	 * If asked, update the disk superblock with incore counter values if we
+	 * are using non-persistent counters so that they don't get too far out
+	 * of sync if we crash or get a forced shutdown. We don't want to force
+	 * this to disk, just get a transaction into the iclogs....
+	 */
+	if (flags & SYNC_SUPER)
+		xfs_log_sbcount(mp, 0);
+
 	/*
 	 * Now check to see if the log needs a "dummy" transaction.
 	 */
@@ -1645,6 +1682,7 @@ xfs_vget(
 					 * in stat(). */
 #define MNTOPT_ATTR2	"attr2"		/* do use attr2 attribute format */
 #define MNTOPT_NOATTR2	"noattr2"	/* do not use attr2 attribute format */
+#define MNTOPT_FILESTREAM  "filestreams" /* use filestreams allocator */
 
 STATIC unsigned long
 suffix_strtoul(char *s, char **endp, unsigned int base)
@@ -1831,6 +1869,8 @@ xfs_parseargs(
 			args->flags |= XFSMNT_ATTR2;
 		} else if (!strcmp(this_char, MNTOPT_NOATTR2)) {
 			args->flags &= ~XFSMNT_ATTR2;
+		} else if (!strcmp(this_char, MNTOPT_FILESTREAM)) {
+			args->flags2 |= XFSMNT2_FILESTREAMS;
 		} else if (!strcmp(this_char, "osyncisdsync")) {
 			/* no-op, this is now the default */
 			cmn_err(CE_WARN,
@@ -1959,9 +1999,9 @@ xfs_showargs(
 }
 
 /*
- * Second stage of a freeze. The data is already frozen, now we have to take
- * care of the metadata. New transactions are already blocked, so we need to
- * wait for any remaining transactions to drain out before proceding.
+ * Second stage of a freeze. The data is already frozen so we only
+ * need to take care of themetadata. Once that's done write a dummy
+ * record to dirty the log in case of a crash while frozen.
  */
 STATIC void
 xfs_freeze(
@@ -1969,18 +2009,7 @@ xfs_freeze(
 {
 	xfs_mount_t	*mp = XFS_BHVTOM(bdp);
 
-	/* wait for all modifications to complete */
-	while (atomic_read(&mp->m_active_trans) > 0)
-		delay(100);
-
-	/* flush inodes and push all remaining buffers out to disk */
-	xfs_quiesce_fs(mp);
-
-	ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0);
-
-	/* Push the superblock and write an unmount record */
-	xfs_log_unmount_write(mp);
-	xfs_unmountfs_writesb(mp);
+	xfs_attr_quiesce(mp);
 	xfs_fs_log_dummy(mp);
 }
 

fs/xfs/xfs_vnodeops.c

@@ -51,6 +51,7 @@
 #include "xfs_refcache.h"
 #include "xfs_trans_space.h"
 #include "xfs_log_priv.h"
+#include "xfs_filestream.h"
 
 STATIC int
 xfs_open(
@@ -77,36 +78,6 @@ xfs_open(
 	return 0;
 }
 
-STATIC int
-xfs_close(
-	bhv_desc_t	*bdp,
-	int		flags,
-	lastclose_t	lastclose,
-	cred_t		*credp)
-{
-	bhv_vnode_t	*vp = BHV_TO_VNODE(bdp);
-	xfs_inode_t	*ip = XFS_BHVTOI(bdp);
-
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
-		return XFS_ERROR(EIO);
-
-	if (lastclose != L_TRUE || !VN_ISREG(vp))
-		return 0;
-
-	/*
-	 * If we previously truncated this file and removed old data in
-	 * the process, we want to initiate "early" writeout on the last
-	 * close.  This is an attempt to combat the notorious NULL files
-	 * problem which is particularly noticable from a truncate down,
-	 * buffered (re-)write (delalloc), followed by a crash.  What we
-	 * are effectively doing here is significantly reducing the time
-	 * window where we'd otherwise be exposed to that problem.
-	 */
-	if (VUNTRUNCATE(vp) && VN_DIRTY(vp) && ip->i_delayed_blks > 0)
-		return bhv_vop_flush_pages(vp, 0, -1, XFS_B_ASYNC, FI_NONE);
-	return 0;
-}
-
 /*
  * xfs_getattr
  */
@@ -183,9 +154,8 @@ xfs_getattr(
 			 * realtime extent size or the realtime volume's
 			 * extent size.
 			 */
-			vap->va_blocksize = ip->i_d.di_extsize ?
-				(ip->i_d.di_extsize << mp->m_sb.sb_blocklog) :
-				(mp->m_sb.sb_rextsize << mp->m_sb.sb_blocklog);
+			vap->va_blocksize =
+				xfs_get_extsz_hint(ip) << mp->m_sb.sb_blocklog;
 		}
 		break;
 	}
@@ -814,6 +784,8 @@ xfs_setattr(
 				di_flags |= XFS_DIFLAG_PROJINHERIT;
 			if (vap->va_xflags & XFS_XFLAG_NODEFRAG)
 				di_flags |= XFS_DIFLAG_NODEFRAG;
+			if (vap->va_xflags & XFS_XFLAG_FILESTREAM)
+				di_flags |= XFS_DIFLAG_FILESTREAM;
 			if ((ip->i_d.di_mode & S_IFMT) == S_IFDIR) {
 				if (vap->va_xflags & XFS_XFLAG_RTINHERIT)
 					di_flags |= XFS_DIFLAG_RTINHERIT;
@@ -1201,13 +1173,15 @@ xfs_fsync(
 }
 
 /*
- * This is called by xfs_inactive to free any blocks beyond eof,
- * when the link count isn't zero.
+ * This is called by xfs_inactive to free any blocks beyond eof
+ * when the link count isn't zero and by xfs_dm_punch_hole() when
+ * punching a hole to EOF.
  */
-STATIC int
-xfs_inactive_free_eofblocks(
+int
+xfs_free_eofblocks(
 	xfs_mount_t	*mp,
-	xfs_inode_t	*ip)
+	xfs_inode_t	*ip,
+	int		flags)
 {
 	xfs_trans_t	*tp;
 	int		error;
@@ -1216,6 +1190,7 @@ xfs_inactive_free_eofblocks(
 	xfs_filblks_t	map_len;
 	int		nimaps;
 	xfs_bmbt_irec_t	imap;
+	int		use_iolock = (flags & XFS_FREE_EOF_LOCK);
 
 	/*
 	 * Figure out if there are any blocks beyond the end
@@ -1256,11 +1231,14 @@ xfs_inactive_free_eofblocks(
 		 * cache and we can't
 		 * do that within a transaction.
 		 */
-		xfs_ilock(ip, XFS_IOLOCK_EXCL);
+		if (use_iolock)
+			xfs_ilock(ip, XFS_IOLOCK_EXCL);
 		error = xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE,
 				    ip->i_size);
 		if (error) {
-			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
+			xfs_trans_cancel(tp, 0);
+			if (use_iolock)
+				xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 			return error;
 		}
 
@@ -1297,7 +1275,8 @@ xfs_inactive_free_eofblocks(
 			error = xfs_trans_commit(tp,
 						XFS_TRANS_RELEASE_LOG_RES);
 		}
-		xfs_iunlock(ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL);
+		xfs_iunlock(ip, (use_iolock ? (XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL)
+					    : XFS_ILOCK_EXCL));
 	}
 	return error;
 }
@@ -1560,6 +1539,31 @@ xfs_release(
 	if (vp->v_vfsp->vfs_flag & VFS_RDONLY)
 		return 0;
 
+	if (!XFS_FORCED_SHUTDOWN(mp)) {
+		/*
+		 * If we are using filestreams, and we have an unlinked
+		 * file that we are processing the last close on, then nothing
+		 * will be able to reopen and write to this file. Purge this
+		 * inode from the filestreams cache so that it doesn't delay
+		 * teardown of the inode.
+		 */
+		if ((ip->i_d.di_nlink == 0) && xfs_inode_is_filestream(ip))
+			xfs_filestream_deassociate(ip);
+
+		/*
+		 * If we previously truncated this file and removed old data
+		 * in the process, we want to initiate "early" writeout on
+		 * the last close.  This is an attempt to combat the notorious
+		 * NULL files problem which is particularly noticable from a
+		 * truncate down, buffered (re-)write (delalloc), followed by
+		 * a crash.  What we are effectively doing here is
+		 * significantly reducing the time window where we'd otherwise
+		 * be exposed to that problem.
+		 */
+		if (VUNTRUNCATE(vp) && VN_DIRTY(vp) && ip->i_delayed_blks > 0)
+			bhv_vop_flush_pages(vp, 0, -1, XFS_B_ASYNC, FI_NONE);
+	}
+
 #ifdef HAVE_REFCACHE
 	/* If we are in the NFS reference cache then don't do this now */
 	if (ip->i_refcache)
@@ -1573,7 +1577,8 @@ xfs_release(
 		     (ip->i_df.if_flags & XFS_IFEXTENTS))  &&
 		    (!(ip->i_d.di_flags &
 				(XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)))) {
-			if ((error = xfs_inactive_free_eofblocks(mp, ip)))
+			error = xfs_free_eofblocks(mp, ip, XFS_FREE_EOF_LOCK);
+			if (error)
 				return error;
 			/* Update linux inode block count after free above */
 			vn_to_inode(vp)->i_blocks = XFS_FSB_TO_BB(mp,
@@ -1654,7 +1659,8 @@ xfs_inactive(
 		     (!(ip->i_d.di_flags &
 				(XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)) ||
 		      (ip->i_delayed_blks != 0)))) {
-			if ((error = xfs_inactive_free_eofblocks(mp, ip)))
+			error = xfs_free_eofblocks(mp, ip, XFS_FREE_EOF_LOCK);
+			if (error)
 				return VN_INACTIVE_CACHE;
 			/* Update linux inode block count after free above */
 			vn_to_inode(vp)->i_blocks = XFS_FSB_TO_BB(mp,
@@ -1680,6 +1686,7 @@ xfs_inactive(
 
 		error = xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE, 0);
 		if (error) {
+			xfs_trans_cancel(tp, 0);
 			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 			return VN_INACTIVE_CACHE;
 		}
@@ -2217,9 +2224,9 @@ static inline int
 xfs_lock_inumorder(int lock_mode, int subclass)
 {
 	if (lock_mode & (XFS_IOLOCK_SHARED|XFS_IOLOCK_EXCL))
-		lock_mode |= (subclass + XFS_IOLOCK_INUMORDER) << XFS_IOLOCK_SHIFT;
+		lock_mode |= (subclass + XFS_LOCK_INUMORDER) << XFS_IOLOCK_SHIFT;
 	if (lock_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL))
-		lock_mode |= (subclass + XFS_ILOCK_INUMORDER) << XFS_ILOCK_SHIFT;
+		lock_mode |= (subclass + XFS_LOCK_INUMORDER) << XFS_ILOCK_SHIFT;
 
 	return lock_mode;
 }
@@ -2546,6 +2553,15 @@ xfs_remove(
 	 */
 	xfs_refcache_purge_ip(ip);
 
+	/*
+	 * If we are using filestreams, kill the stream association.
+	 * If the file is still open it may get a new one but that
+	 * will get killed on last close in xfs_close() so we don't
+	 * have to worry about that.
+	 */
+	if (link_zero && xfs_inode_is_filestream(ip))
+		xfs_filestream_deassociate(ip);
+
 	vn_trace_exit(XFS_ITOV(ip), __FUNCTION__, (inst_t *)__return_address);
 
 	/*
@@ -4047,22 +4063,16 @@ xfs_alloc_file_space(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return XFS_ERROR(EIO);
 
-	rt = XFS_IS_REALTIME_INODE(ip);
-	if (unlikely(rt)) {
-		if (!(extsz = ip->i_d.di_extsize))
-			extsz = mp->m_sb.sb_rextsize;
-	} else {
-		extsz = ip->i_d.di_extsize;
-	}
-
 	if ((error = XFS_QM_DQATTACH(mp, ip, 0)))
 		return error;
 
 	if (len <= 0)
 		return XFS_ERROR(EINVAL);
 
+	rt = XFS_IS_REALTIME_INODE(ip);
+	extsz = xfs_get_extsz_hint(ip);
+
 	count = len;
-	error = 0;
 	imapp = &imaps[0];
 	nimaps = 1;
 	bmapi_flag = XFS_BMAPI_WRITE | (alloc_type ? XFS_BMAPI_PREALLOC : 0);
@@ -4678,7 +4688,6 @@ xfs_change_file_space(
 bhv_vnodeops_t xfs_vnodeops = {
 	BHV_IDENTITY_INIT(VN_BHV_XFS,VNODE_POSITION_XFS),
 	.vop_open		= xfs_open,
-	.vop_close		= xfs_close,
 	.vop_read		= xfs_read,
 #ifdef HAVE_SPLICE
 	.vop_splice_read	= xfs_splice_read,

lib/radix-tree.c

@@ -151,6 +151,7 @@ int radix_tree_preload(gfp_t gfp_mask)
 out:
 	return ret;
 }
+EXPORT_SYMBOL(radix_tree_preload);
 
 static inline void tag_set(struct radix_tree_node *node, unsigned int tag,
 		int offset)