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);
@@ -1416,16 +1434,23 @@ 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.
 	 */
 	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);