Merge tag 'xfs-for-linus-v3.12-rc1-2' of git://oss.sgi.com/xfs/xfs

Pull xfs update Digilent#2 from Ben Myers:
 "Here we have defrag support for v5 superblock, a number of bugfixes
  and a cleanup or two.

   - defrag support for CRC filesystems
   - fix endian worning in xlog_recover_get_buf_lsn
   - fixes for sparse warnings
   - fix for assert in xfs_dir3_leaf_hdr_from_disk
   - fix for log recovery of remote symlinks
   - fix for log recovery of btree root splits
   - fixes formemory allocation failures with ACLs
   - fix for assert in xfs_buf_item_relse
   - fix for assert in xfs_inode_buf_verify
   - fix an assignment in an assert that should be a test in
     xfs_bmbt_change_owner
   - remove dead code in xlog_recover_inode_pass2"

* tag 'xfs-for-linus-v3.12-rc1-2' of git://oss.sgi.com/xfs/xfs:
  xfs: remove dead code from xlog_recover_inode_pass2
  xfs: = vs == typo in ASSERT()
  xfs: don't assert fail on bad inode numbers
  xfs: aborted buf items can be in the AIL.
  xfs: factor all the kmalloc-or-vmalloc fallback allocations
  xfs: fix memory allocation failures with ACLs
  xfs: ensure we copy buffer type in da btree root splits
  xfs: set remote symlink buffer type for recovery
  xfs: recovery of swap extents operations for CRC filesystems
  xfs: swap extents operations for CRC filesystems
  xfs: check magic numbers in dir3 leaf verifier first
  xfs: fix some minor sparse warnings
  xfs: fix endian warning in xlog_recover_get_buf_lsn()

Author: torvalds

fs/xfs/kmem.c

@@ -27,16 +27,14 @@
 
 /*
  * Greedy allocation.  May fail and may return vmalloced memory.
- *
- * Must be freed using kmem_free_large.
  */
 void *
 kmem_zalloc_greedy(size_t *size, size_t minsize, size_t maxsize)
 {
 	void		*ptr;
 	size_t		kmsize = maxsize;
 
-	while (!(ptr = kmem_zalloc_large(kmsize))) {
+	while (!(ptr = vzalloc(kmsize))) {
 		if ((kmsize >>= 1) <= minsize)
 			kmsize = minsize;
 	}
@@ -75,6 +73,17 @@ kmem_zalloc(size_t size, xfs_km_flags_t flags)
 	return ptr;
 }
 
+void *
+kmem_zalloc_large(size_t size, xfs_km_flags_t flags)
+{
+	void	*ptr;
+
+	ptr = kmem_zalloc(size, flags | KM_MAYFAIL);
+	if (ptr)
+		return ptr;
+	return vzalloc(size);
+}
+
 void
 kmem_free(const void *ptr)
 {

fs/xfs/kmem.h

@@ -57,17 +57,10 @@ kmem_flags_convert(xfs_km_flags_t flags)
 
 extern void *kmem_alloc(size_t, xfs_km_flags_t);
 extern void *kmem_zalloc(size_t, xfs_km_flags_t);
+extern void *kmem_zalloc_large(size_t size, xfs_km_flags_t);
 extern void *kmem_realloc(const void *, size_t, size_t, xfs_km_flags_t);
 extern void  kmem_free(const void *);
 
-static inline void *kmem_zalloc_large(size_t size)
-{
-	return vzalloc(size);
-}
-static inline void kmem_free_large(void *ptr)
-{
-	vfree(ptr);
-}
 
 extern void *kmem_zalloc_greedy(size_t *, size_t, size_t);
 

fs/xfs/xfs_acl.c

@@ -152,7 +152,7 @@ xfs_get_acl(struct inode *inode, int type)
 	 * go out to the disk.
 	 */
 	len = XFS_ACL_MAX_SIZE(ip->i_mount);
-	xfs_acl = kzalloc(len, GFP_KERNEL);
+	xfs_acl = kmem_zalloc_large(len, KM_SLEEP);
 	if (!xfs_acl)
 		return ERR_PTR(-ENOMEM);
 
@@ -175,10 +175,10 @@ xfs_get_acl(struct inode *inode, int type)
 	if (IS_ERR(acl))
 		goto out;
 
- out_update_cache:
+out_update_cache:
 	set_cached_acl(inode, type, acl);
- out:
-	kfree(xfs_acl);
+out:
+	kmem_free(xfs_acl);
 	return acl;
 }
 
@@ -209,7 +209,7 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 		struct xfs_acl *xfs_acl;
 		int len = XFS_ACL_MAX_SIZE(ip->i_mount);
 
-		xfs_acl = kzalloc(len, GFP_KERNEL);
+		xfs_acl = kmem_zalloc_large(len, KM_SLEEP);
 		if (!xfs_acl)
 			return -ENOMEM;
 
@@ -222,7 +222,7 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 		error = -xfs_attr_set(ip, ea_name, (unsigned char *)xfs_acl,
 				len, ATTR_ROOT);
 
-		kfree(xfs_acl);
+		kmem_free(xfs_acl);
 	} else {
 		/*
 		 * A NULL ACL argument means we want to remove the ACL.

fs/xfs/xfs_bmap.c

@@ -4450,7 +4450,7 @@ xfs_bmapi_write(
 {
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_ifork	*ifp;
-	struct xfs_bmalloca	bma = { 0 };	/* args for xfs_bmap_alloc */
+	struct xfs_bmalloca	bma = { NULL };	/* args for xfs_bmap_alloc */
 	xfs_fileoff_t		end;		/* end of mapped file region */
 	int			eof;		/* after the end of extents */
 	int			error;		/* error return */

fs/xfs/xfs_bmap_btree.c

@@ -925,3 +925,47 @@ xfs_bmdr_maxrecs(
 		return blocklen / sizeof(xfs_bmdr_rec_t);
 	return blocklen / (sizeof(xfs_bmdr_key_t) + sizeof(xfs_bmdr_ptr_t));
 }
+
+/*
+ * Change the owner of a btree format fork fo the inode passed in. Change it to
+ * the owner of that is passed in so that we can change owners before or after
+ * we switch forks between inodes. The operation that the caller is doing will
+ * determine whether is needs to change owner before or after the switch.
+ *
+ * For demand paged transactional modification, the fork switch should be done
+ * after reading in all the blocks, modifying them and pinning them in the
+ * transaction. For modification when the buffers are already pinned in memory,
+ * the fork switch can be done before changing the owner as we won't need to
+ * validate the owner until the btree buffers are unpinned and writes can occur
+ * again.
+ *
+ * For recovery based ownership change, there is no transactional context and
+ * so a buffer list must be supplied so that we can record the buffers that we
+ * modified for the caller to issue IO on.
+ */
+int
+xfs_bmbt_change_owner(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	int			whichfork,
+	xfs_ino_t		new_owner,
+	struct list_head	*buffer_list)
+{
+	struct xfs_btree_cur	*cur;
+	int			error;
+
+	ASSERT(tp || buffer_list);
+	ASSERT(!(tp && buffer_list));
+	if (whichfork == XFS_DATA_FORK)
+		ASSERT(ip->i_d.di_format == XFS_DINODE_FMT_BTREE);
+	else
+		ASSERT(ip->i_d.di_aformat == XFS_DINODE_FMT_BTREE);
+
+	cur = xfs_bmbt_init_cursor(ip->i_mount, tp, ip, whichfork);
+	if (!cur)
+		return ENOMEM;
+
+	error = xfs_btree_change_owner(cur, new_owner, buffer_list);
+	xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+	return error;
+}

fs/xfs/xfs_bmap_btree.h

@@ -236,6 +236,10 @@ extern int xfs_bmbt_get_maxrecs(struct xfs_btree_cur *, int level);
 extern int xfs_bmdr_maxrecs(struct xfs_mount *, int blocklen, int leaf);
 extern int xfs_bmbt_maxrecs(struct xfs_mount *, int blocklen, int leaf);
 
+extern int xfs_bmbt_change_owner(struct xfs_trans *tp, struct xfs_inode *ip,
+				 int whichfork, xfs_ino_t new_owner,
+				 struct list_head *buffer_list);
+
 extern struct xfs_btree_cur *xfs_bmbt_init_cursor(struct xfs_mount *,
 		struct xfs_trans *, struct xfs_inode *, int);
 

fs/xfs/xfs_bmap_util.c

@@ -612,13 +612,9 @@ xfs_getbmap(
 
 	if (bmv->bmv_count > ULONG_MAX / sizeof(struct getbmapx))
 		return XFS_ERROR(ENOMEM);
-	out = kmem_zalloc(bmv->bmv_count * sizeof(struct getbmapx), KM_MAYFAIL);
-	if (!out) {
-		out = kmem_zalloc_large(bmv->bmv_count *
-					sizeof(struct getbmapx));
-		if (!out)
-			return XFS_ERROR(ENOMEM);
-	}
+	out = kmem_zalloc_large(bmv->bmv_count * sizeof(struct getbmapx), 0);
+	if (!out)
+		return XFS_ERROR(ENOMEM);
 
 	xfs_ilock(ip, XFS_IOLOCK_SHARED);
 	if (whichfork == XFS_DATA_FORK && !(iflags & BMV_IF_DELALLOC)) {
@@ -754,10 +750,7 @@ xfs_getbmap(
 			break;
 	}
 
-	if (is_vmalloc_addr(out))
-		kmem_free_large(out);
-	else
-		kmem_free(out);
+	kmem_free(out);
 	return error;
 }
 
@@ -1789,14 +1782,6 @@ xfs_swap_extents(
 	int		taforkblks = 0;
 	__uint64_t	tmp;
 
-	/*
-	 * We have no way of updating owner information in the BMBT blocks for
-	 * each inode on CRC enabled filesystems, so to avoid corrupting the
-	 * this metadata we simply don't allow extent swaps to occur.
-	 */
-	if (xfs_sb_version_hascrc(&mp->m_sb))
-		return XFS_ERROR(EINVAL);
-
 	tempifp = kmem_alloc(sizeof(xfs_ifork_t), KM_MAYFAIL);
 	if (!tempifp) {
 		error = XFS_ERROR(ENOMEM);
@@ -1920,6 +1905,42 @@ xfs_swap_extents(
 			goto out_trans_cancel;
 	}
 
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
+	xfs_trans_ijoin(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
+
+	/*
+	 * Before we've swapped the forks, lets set the owners of the forks
+	 * appropriately. We have to do this as we are demand paging the btree
+	 * buffers, and so the validation done on read will expect the owner
+	 * field to be correctly set. Once we change the owners, we can swap the
+	 * inode forks.
+	 *
+	 * Note the trickiness in setting the log flags - we set the owner log
+	 * flag on the opposite inode (i.e. the inode we are setting the new
+	 * owner to be) because once we swap the forks and log that, log
+	 * recovery is going to see the fork as owned by the swapped inode,
+	 * not the pre-swapped inodes.
+	 */
+	src_log_flags = XFS_ILOG_CORE;
+	target_log_flags = XFS_ILOG_CORE;
+	if (ip->i_d.di_version == 3 &&
+	    ip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
+		target_log_flags |= XFS_ILOG_DOWNER;
+		error = xfs_bmbt_change_owner(tp, ip, XFS_DATA_FORK,
+					      tip->i_ino, NULL);
+		if (error)
+			goto out_trans_cancel;
+	}
+
+	if (tip->i_d.di_version == 3 &&
+	    tip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
+		src_log_flags |= XFS_ILOG_DOWNER;
+		error = xfs_bmbt_change_owner(tp, tip, XFS_DATA_FORK,
+					      ip->i_ino, NULL);
+		if (error)
+			goto out_trans_cancel;
+	}
+
 	/*
 	 * Swap the data forks of the inodes
 	 */
@@ -1957,7 +1978,6 @@ xfs_swap_extents(
 	tip->i_delayed_blks = ip->i_delayed_blks;
 	ip->i_delayed_blks = 0;
 
-	src_log_flags = XFS_ILOG_CORE;
 	switch (ip->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 		/* If the extents fit in the inode, fix the
@@ -1971,11 +1991,12 @@ xfs_swap_extents(
 		src_log_flags |= XFS_ILOG_DEXT;
 		break;
 	case XFS_DINODE_FMT_BTREE:
+		ASSERT(ip->i_d.di_version < 3 ||
+		       (src_log_flags & XFS_ILOG_DOWNER));
 		src_log_flags |= XFS_ILOG_DBROOT;
 		break;
 	}
 
-	target_log_flags = XFS_ILOG_CORE;
 	switch (tip->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 		/* If the extents fit in the inode, fix the
@@ -1990,13 +2011,11 @@ xfs_swap_extents(
 		break;
 	case XFS_DINODE_FMT_BTREE:
 		target_log_flags |= XFS_ILOG_DBROOT;
+		ASSERT(tip->i_d.di_version < 3 ||
+		       (target_log_flags & XFS_ILOG_DOWNER));
 		break;
 	}
 
-
-	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
-	xfs_trans_ijoin(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
-
 	xfs_trans_log_inode(tp, ip,  src_log_flags);
 	xfs_trans_log_inode(tp, tip, target_log_flags);