Merge cantab.net:/usr/src/bklinux-2.5

into cantab.net:/usr/src/tng-2.0.12

Documentation/filesystems/ntfs.txt

@@ -247,6 +247,12 @@ ChangeLog
 
 Note, a technical ChangeLog aimed at kernel hackers is in fs/ntfs/ChangeLog.
 
+2.0.12:
+	- Internal cleanups in address space operations made possible by the
+	  changes introduced in the previous release.
+2.0.11:
+	- Internal updates and cleanups introducing the first step towards
+	  fake inode based attribute i/o.
 2.0.10:
 	- Microsoft says that the maximum number of inodes is 2^32 - 1. Update
 	  the driver accordingly to only use 32-bits to store inode numbers on

fs/ntfs/ChangeLog

@@ -6,7 +6,7 @@ ToDo:
 	  user open()s a file with i_size > s_maxbytes? Should read_inode()
 	  truncate the visible i_size? Will the user just get -E2BIG (or
 	  whatever) on open()? Or will (s)he be able to open() but lseek() and
-	  read() will fail when s_maxbytes is reached? -> Investigate this!
+	  read() will fail when s_maxbytes is reached? -> Investigate this.
 	- Implement/allow non-resident index bitmaps in dir.c::ntfs_readdir()
 	  and then also consider initialized_size w.r.t. the bitmaps, etc.
 	- vcn_to_lcn() should somehow return the correct pointer within the
@@ -17,11 +17,52 @@ ToDo:
 	- Consider if ntfs_file_read_compressed_block() shouldn't be coping
 	  with initialized_size < data_size. I don't think it can happen but
 	  it requires more careful consideration.
-	- CLEANUP: Modularise and reuse code in aops.c. At the moment we have
-	  several copies of almost identicall functions and the functions are
-	  quite big. Modularising them a bit, e.g. a-la get_block(), will make
-	  them cleaner and make code reuse easier.
-	- Want to use dummy inodes for address space i/o.
+	- CLEANUP: At the moment we have two copies of almost identical
+	  functions in aops.c, can merge them once fake inode address space
+	  based attribute i/o is further developed.
+	- CLEANUP: Modularising code in aops.c a bit, e.g. a-la get_block(),
+	  will be cleaner and make code reuse easier.
+	- Enable NFS exporting of NTFS.
+	- Use iget5_locked() and friends instead of conventional iget().
+	- Use fake inodes for address space i/o.
+
+2.0.12 - Initial cleanup of address space operations following 2.0.11 changes.
+
+	- Merge fs/ntfs/aops.c::end_buffer_read_mst_async() and
+	  fs/ntfs/aops.c::end_buffer_read_file_async() into one function
+	  fs/ntfs/aops.c::end_buffer_read_attr_async() using NInoMstProtected()
+	  to determine whether to apply mst fixups or not.
+	- Above change allows merging fs/ntfs/aops.c::ntfs_file_read_block()
+	  and fs/ntfs/aops.c::ntfs_mst_readpage() into one function
+	  fs/ntfs/aops.c::ntfs_attr_read_block(). Also, create a tiny wrapper
+	  fs/ntfs/aops.c::ntfs_mst_readpage() to transform the parameters from
+	  the VFS readpage function prototype to the ntfs_attr_read_block()
+	  function prototype.
+
+2.0.11 - Initial preparations for fake inode based attribute i/o.
+
+	- Move definition of ntfs_inode_state_bits to fs/ntfs/inode.h and
+	  do some macro magic (adapted from include/linux/buffer_head.h) to
+	  expand all the helper functions NInoFoo(), NInoSetFoo(), and
+	  NInoClearFoo().
+	- Add new flag to ntfs_inode_state_bits: NI_Sparse.
+	- Add new fields to ntfs_inode structure to allow use of fake inodes
+	  for attribute i/o: type, name, name_len. Also add new state bits:
+	  NI_Attr, which, if set, indicates the inode is a fake inode, and
+	  NI_MstProtected, which, if set, indicates the attribute uses multi
+	  sector transfer protection, i.e. fixups need to be applied after
+	  reads and before/after writes.
+	- Rename fs/ntfs/inode.c::ntfs_{new,clear,destroy}_inode() to
+	  ntfs_{new,clear,destroy}_extent_inode() and update callers.
+	- Use ntfs_clear_extent_inode() in fs/ntfs/inode.c::__ntfs_clear_inode()
+	  instead of ntfs_destroy_extent_inode().
+	- Cleanup memory deallocations in {__,}ntfs_clear_{,big_}inode().
+	- Make all operations on ntfs inode state bits use the NIno* functions.
+	- Set up the new ntfs inode fields and state bits in
+	  fs/ntfs/inode.c::ntfs_read_inode() and add appropriate cleanup of
+	  allocated memory to __ntfs_clear_inode().
+	- Cleanup ntfs_inode structure a bit for better ordering of elements
+	  w.r.t. their size to allow better packing of the structure in memory.
 
 2.0.10 - There can only be 2^32 - 1 inodes on an NTFS volume.
 
@@ -38,7 +79,10 @@ ToDo:
 
 	- Change decompression engine to use a single buffer protected by a
 	  spin lock instead of per-CPU buffers. (Rusty Russell)
-	- Switch to using the new KM_BIO_SRC_IRQ for atomic kmaps. (Andrew
+	- Do not update cb_pos when handling a partial final page during
+	  decompression of a sparse compression block, as the value is later
+	  reset without being read/used. (Rusty Russell)
+	- Switch to using the new KM_BIO_SRC_IRQ for atomic kmap()s. (Andrew
 	  Morton)
 	- Change buffer size in ntfs_readdir()/ntfs_filldir() to use
 	  NLS_MAX_CHARSET_SIZE which makes the buffers almost 1kiB each but

fs/ntfs/Makefile

@@ -5,7 +5,7 @@ obj-$(CONFIG_NTFS_FS) += ntfs.o
 ntfs-objs := aops.o attrib.o compress.o debug.o dir.o file.o inode.o mft.o \
 	     mst.o namei.o super.o sysctl.o time.o unistr.o upcase.o
 
-EXTRA_CFLAGS = -DNTFS_VERSION=\"2.0.10\"
+EXTRA_CFLAGS = -DNTFS_VERSION=\"2.0.12\"
 
 ifeq ($(CONFIG_NTFS_DEBUG),y)
 EXTRA_CFLAGS += -DDEBUG

fs/ntfs/dir.c

@@ -27,7 +27,7 @@
 /**
  * The little endian Unicode string $I30 as a global constant.
  */
-const uchar_t I30[5] = { const_cpu_to_le16('$'), const_cpu_to_le16('I'),
+uchar_t I30[5] = { const_cpu_to_le16('$'), const_cpu_to_le16('I'),
 		const_cpu_to_le16('3'),	const_cpu_to_le16('0'),
 		const_cpu_to_le16(0) };
 

fs/ntfs/dir.h

@@ -38,7 +38,7 @@ typedef struct {
 } __attribute__ ((__packed__)) ntfs_name;
 
 /* The little endian Unicode string $I30 as a global constant. */
-extern const uchar_t I30[5];
+extern uchar_t I30[5];
 
 extern MFT_REF ntfs_lookup_inode_by_name(ntfs_inode *dir_ni,
 		const uchar_t *uname, const int uname_len, ntfs_name **res);

fs/ntfs/inode.c

@@ -49,7 +49,7 @@ void ntfs_destroy_big_inode(struct inode *inode)
 	kmem_cache_free(ntfs_big_inode_cache, NTFS_I(inode));
 }
 
-ntfs_inode *ntfs_alloc_inode(void)
+ntfs_inode *ntfs_alloc_extent_inode(void)
 {
 	ntfs_inode *ni = (ntfs_inode *)kmem_cache_alloc(ntfs_inode_cache,
 			SLAB_NOFS);
@@ -59,7 +59,7 @@ ntfs_inode *ntfs_alloc_inode(void)
 	return ni;
 }
 
-void ntfs_destroy_inode(ntfs_inode *ni)
+void ntfs_destroy_extent_inode(ntfs_inode *ni)
 {
 	ntfs_debug("Entering.");
 	BUG_ON(atomic_read(&ni->mft_count) || !atomic_dec_and_test(&ni->count));
@@ -102,9 +102,9 @@ static void ntfs_init_big_inode(struct inode *vi)
 	return;
 }
 
-ntfs_inode *ntfs_new_inode(struct super_block *sb)
+ntfs_inode *ntfs_new_extent_inode(struct super_block *sb)
 {
-	ntfs_inode *ni = ntfs_alloc_inode();
+	ntfs_inode *ni = ntfs_alloc_extent_inode();
 
 	ntfs_debug("Entering.");
 	if (ni)
@@ -239,7 +239,8 @@ void ntfs_read_inode(struct inode *vi)
 
 	/*
 	 * Initialize the ntfs specific part of @vi special casing
-	 * FILE_MFT which we need to do at mount time.
+	 * FILE_MFT which we need to do at mount time. This also sets
+	 * ni->mft_no to vi->i_ino.
 	 */
 	if (vi->i_ino != FILE_MFT)
 		ntfs_init_big_inode(vi);
@@ -358,13 +359,14 @@ void ntfs_read_inode(struct inode *vi)
 		if (vi->i_ino == FILE_MFT)
 			goto skip_attr_list_load;
 		ntfs_debug("Attribute list found in inode 0x%lx.", vi->i_ino);
-		ni->state |= 1 << NI_AttrList;
+		NInoSetAttrList(ni);
 		if (ctx->attr->flags & ATTR_IS_ENCRYPTED ||
-				ctx->attr->flags & ATTR_COMPRESSION_MASK) {
+				ctx->attr->flags & ATTR_COMPRESSION_MASK ||
+				ctx->attr->flags & ATTR_IS_SPARSE) {
 			ntfs_error(vi->i_sb, "Attribute list attribute is "
-					"compressed/encrypted. Not allowed. "
-					"Corrupt inode. You should run "
-					"chkdsk.");
+					"compressed/encrypted/sparse. Not "
+					"allowed. Corrupt inode. You should "
+					"run chkdsk.");
 			goto put_unm_err_out;
 		}
 		/* Now allocate memory for the attribute list. */
@@ -377,7 +379,7 @@ void ntfs_read_inode(struct inode *vi)
 			goto ec_put_unm_err_out;
 		}
 		if (ctx->attr->non_resident) {
-			ni->state |= 1 << NI_AttrListNonResident;
+			NInoSetAttrListNonResident(ni);
 			if (ctx->attr->_ANR(lowest_vcn)) {
 				ntfs_error(vi->i_sb, "Attribute list has non "
 						"zero lowest_vcn. Inode is "
@@ -459,7 +461,7 @@ void ntfs_read_inode(struct inode *vi)
 		 * encrypted.
 		 */
 		if (ctx->attr->flags & ATTR_COMPRESSION_MASK)
-			ni->state |= 1 << NI_Compressed;
+			NInoSetCompressed(ni);
 		if (ctx->attr->flags & ATTR_IS_ENCRYPTED) {
 			if (ctx->attr->flags & ATTR_COMPRESSION_MASK) {
 				ntfs_error(vi->i_sb, "Found encrypted and "
@@ -467,8 +469,10 @@ void ntfs_read_inode(struct inode *vi)
 						"allowed.");
 				goto put_unm_err_out;
 			}
-			ni->state |= 1 << NI_Encrypted;
+			NInoSetEncrypted(ni);
 		}
+		if (ctx->attr->flags & ATTR_IS_SPARSE)
+			NInoSetSparse(ni);
 		ir = (INDEX_ROOT*)((char*)ctx->attr +
 				le16_to_cpu(ctx->attr->_ARA(value_offset)));
 		ir_end = (char*)ir + le32_to_cpu(ctx->attr->_ARA(value_length));
@@ -530,12 +534,19 @@ void ntfs_read_inode(struct inode *vi)
 			ni->_IDM(index_vcn_size) = vol->sector_size;
 			ni->_IDM(index_vcn_size_bits) = vol->sector_size_bits;
 		}
+
+		/* Setup the index allocation attribute, even if not present. */
+		NInoSetMstProtected(ni);
+		ni->type = AT_INDEX_ALLOCATION;
+		ni->name = I30;
+		ni->name_len = 4;
+
 		if (!(ir->index.flags & LARGE_INDEX)) {
 			/* No index allocation. */
 			vi->i_size = ni->initialized_size = 0;
 			goto skip_large_dir_stuff;
 		} /* LARGE_INDEX: Index allocation present. Setup state. */
-		ni->state |= 1 << NI_NonResident;
+		NInoSetIndexAllocPresent(ni);
 		/* Find index allocation attribute. */
 		reinit_attr_search_ctx(ctx);
 		if (!lookup_attr(AT_INDEX_ALLOCATION, I30, 4, CASE_SENSITIVE,
@@ -555,6 +566,11 @@ void ntfs_read_inode(struct inode *vi)
 					"is encrypted.");
 			goto put_unm_err_out;
 		}
+		if (ctx->attr->flags & ATTR_IS_SPARSE) {
+			ntfs_error(vi->i_sb, "$INDEX_ALLOCATION attribute "
+					"is sparse.");
+			goto put_unm_err_out;
+		}
 		if (ctx->attr->flags & ATTR_COMPRESSION_MASK) {
 			ntfs_error(vi->i_sb, "$INDEX_ALLOCATION attribute "
 					"is compressed.");
@@ -581,13 +597,13 @@ void ntfs_read_inode(struct inode *vi)
 			goto put_unm_err_out;
 		}
 		if (ctx->attr->flags & (ATTR_COMPRESSION_MASK |
-				ATTR_IS_ENCRYPTED)) {
+				ATTR_IS_ENCRYPTED | ATTR_IS_SPARSE)) {
 			ntfs_error(vi->i_sb, "$BITMAP attribute is compressed "
-					"and/or encrypted.");
+					"and/or encrypted and/or sparse.");
 			goto put_unm_err_out;
 		}
 		if (ctx->attr->non_resident) {
-			ni->state |= 1 << NI_BmpNonResident;
+			NInoSetBmpNonResident(ni);
 			if (ctx->attr->_ANR(lowest_vcn)) {
 				ntfs_error(vi->i_sb, "First extent of $BITMAP "
 						"attribute has non zero "
@@ -647,6 +663,12 @@ void ntfs_read_inode(struct inode *vi)
 	} else {
 		/* It is a file: find first extent of unnamed data attribute. */
 		reinit_attr_search_ctx(ctx);
+
+		/* Setup the data attribute, even if not present. */
+		ni->type = AT_DATA;
+		ni->name = NULL;
+		ni->name_len = 0;
+
 		if (!lookup_attr(AT_DATA, NULL, 0, 0, 0, NULL, 0, ctx)) {
 			vi->i_size = ni->initialized_size =
 					ni->allocated_size = 0LL;
@@ -675,9 +697,9 @@ void ntfs_read_inode(struct inode *vi)
 		}
 		/* Setup the state. */
 		if (ctx->attr->non_resident) {
-			ni->state |= 1 << NI_NonResident;
+			NInoSetNonResident(ni);
 			if (ctx->attr->flags & ATTR_COMPRESSION_MASK) {
-				ni->state |= 1 << NI_Compressed;
+				NInoSetCompressed(ni);
 				if (vol->cluster_size > 4096) {
 					ntfs_error(vi->i_sb, "Found "
 						"compressed data but "
@@ -707,8 +729,9 @@ void ntfs_read_inode(struct inode *vi)
 					goto ec_put_unm_err_out;
 				}
 				ni->_ICF(compression_block_size) = 1U << (
-					       ctx->attr->_ANR(compression_unit)
-						+ vol->cluster_size_bits);
+						ctx->attr->_ANR(
+						compression_unit) +
+						vol->cluster_size_bits);
 				ni->_ICF(compression_block_size_bits) = ffs(
 					ni->_ICF(compression_block_size)) - 1;
 			}
@@ -718,8 +741,10 @@ void ntfs_read_inode(struct inode *vi)
 							"and compressed data.");
 					goto put_unm_err_out;
 				}
-				ni->state |= 1 << NI_Encrypted;
+				NInoSetEncrypted(ni);
 			}
+			if (ctx->attr->flags & ATTR_IS_SPARSE)
+				NInoSetSparse(ni);
 			if (ctx->attr->_ANR(lowest_vcn)) {
 				ntfs_error(vi->i_sb, "First extent of $DATA "
 						"attribute has non zero "
@@ -861,6 +886,13 @@ void ntfs_read_inode_mount(struct inode *vi)
 		goto err_out;
 	}
 
+	/* Setup the data attribute. It is special as it is mst protected. */
+	NInoSetNonResident(ni);
+	NInoSetMstProtected(ni);
+	ni->type = AT_DATA;
+	ni->name = NULL;
+	ni->name_len = 0;
+
 	/*
 	 * This sets up our little cheat allowing us to reuse the async io
 	 * completion handler for directories.
@@ -930,13 +962,14 @@ void ntfs_read_inode_mount(struct inode *vi)
 		u8 *al_end;
 
 		ntfs_debug("Attribute list attribute found in $MFT.");
-		ni->state |= 1 << NI_AttrList;
+		NInoSetAttrList(ni);
 		if (ctx->attr->flags & ATTR_IS_ENCRYPTED ||
-				ctx->attr->flags & ATTR_COMPRESSION_MASK) {
+				ctx->attr->flags & ATTR_COMPRESSION_MASK ||
+				ctx->attr->flags & ATTR_IS_SPARSE) {
 			ntfs_error(sb, "Attribute list attribute is "
-					"compressed/encrypted. Not allowed. "
-					"$MFT is corrupt. You should run "
-					"chkdsk.");
+					"compressed/encrypted/sparse. Not "
+					"allowed. $MFT is corrupt. You should "
+					"run chkdsk.");
 			goto put_err_out;
 		}
 		/* Now allocate memory for the attribute list. */
@@ -948,7 +981,7 @@ void ntfs_read_inode_mount(struct inode *vi)
 			goto put_err_out;
 		}
 		if (ctx->attr->non_resident) {
-			ni->state |= 1 << NI_AttrListNonResident;
+			NInoSetAttrListNonResident(ni);
 			if (ctx->attr->_ANR(lowest_vcn)) {
 				ntfs_error(sb, "Attribute list has non zero "
 						"lowest_vcn. $MFT is corrupt. "
@@ -1071,11 +1104,13 @@ void ntfs_read_inode_mount(struct inode *vi)
 		}
 		/* $MFT must be uncompressed and unencrypted. */
 		if (attr->flags & ATTR_COMPRESSION_MASK ||
-				attr->flags & ATTR_IS_ENCRYPTED) {
-			ntfs_error(sb, "$MFT must be uncompressed and "
-					"unencrypted but a compressed/"
-					"encrypted extent was found. "
-					"$MFT is corrupt. Run chkdsk.");
+				attr->flags & ATTR_IS_ENCRYPTED ||
+				attr->flags & ATTR_IS_SPARSE) {
+			ntfs_error(sb, "$MFT must be uncompressed, "
+					"non-sparse, and unencrypted but a "
+					"compressed/sparse/encrypted extent "
+					"was found. $MFT is corrupt. Run "
+					"chkdsk.");
 			goto put_err_out;
 		}
 		/*
@@ -1296,29 +1331,42 @@ void __ntfs_clear_inode(ntfs_inode *ni)
 
 		// FIXME: Handle dirty case for each extent inode!
 		for (i = 0; i < ni->nr_extents; i++)
-			ntfs_destroy_inode(ni->_INE(extent_ntfs_inos)[i]);
+			ntfs_clear_extent_inode(ni->_INE(extent_ntfs_inos)[i]);
 		kfree(ni->_INE(extent_ntfs_inos));
 	}
 	/* Free all alocated memory. */
 	down_write(&ni->run_list.lock);
-	ntfs_free(ni->run_list.rl);
-	ni->run_list.rl = NULL;
+	if (ni->run_list.rl) {
+		ntfs_free(ni->run_list.rl);
+		ni->run_list.rl = NULL;
+	}
 	up_write(&ni->run_list.lock);
 
-	ntfs_free(ni->attr_list);
+	if (ni->attr_list) {
+		ntfs_free(ni->attr_list);
+		ni->attr_list = NULL;
+	}
 
 	down_write(&ni->attr_list_rl.lock);
-	ntfs_free(ni->attr_list_rl.rl);
-	ni->attr_list_rl.rl = NULL;
+	if (ni->attr_list_rl.rl) {
+		ntfs_free(ni->attr_list_rl.rl);
+		ni->attr_list_rl.rl = NULL;
+	}
 	up_write(&ni->attr_list_rl.lock);
+
+	if (ni->name_len && ni->name != I30) {
+		/* Catch bugs... */
+		BUG_ON(!ni->name);
+		kfree(ni->name);
+	}
 }
 
-void ntfs_clear_inode(ntfs_inode *ni)
+void ntfs_clear_extent_inode(ntfs_inode *ni)
 {
 	__ntfs_clear_inode(ni);
 
 	/* Bye, bye... */
-	ntfs_destroy_inode(ni);
+	ntfs_destroy_extent_inode(ni);
 }
 
 /**
@@ -1339,7 +1387,8 @@ void ntfs_clear_big_inode(struct inode *vi)
 
 	if (S_ISDIR(vi->i_mode)) {
 		down_write(&ni->_IDM(bmp_rl).lock);
-		ntfs_free(ni->_IDM(bmp_rl).rl);
+		if (ni->_IDM(bmp_rl).rl)
+			ntfs_free(ni->_IDM(bmp_rl).rl);
 		up_write(&ni->_IDM(bmp_rl).lock);
 	}
 	return;

fs/ntfs/inode.h

@@ -3,7 +3,7 @@
  *	     the Linux-NTFS project.
  *
  * Copyright (c) 2001,2002 Anton Altaparmakov.
- * Copyright (C) 2002 Richard Russon.
+ * Copyright (c) 2002 Richard Russon.
  *
  * This program/include file is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License as published
@@ -26,6 +26,7 @@
 
 #include <linux/seq_file.h>
 
+#include "layout.h"
 #include "volume.h"
 
 typedef struct _ntfs_inode ntfs_inode;
@@ -38,21 +39,39 @@ struct _ntfs_inode {
 	s64 initialized_size;	/* Copy from $DATA/$INDEX_ALLOCATION. */
 	s64 allocated_size;	/* Copy from $DATA/$INDEX_ALLOCATION. */
 	unsigned long state;	/* NTFS specific flags describing this inode.
-				   See fs/ntfs/ntfs.h:ntfs_inode_state_bits. */
+				   See ntfs_inode_state_bits below. */
 	unsigned long mft_no;	/* Number of the mft record / inode. */
 	u16 seq_no;		/* Sequence number of the mft record. */
 	atomic_t count;		/* Inode reference count for book keeping. */
 	ntfs_volume *vol;	/* Pointer to the ntfs volume of this inode. */
+	/*
+	 * If NInoAttr() is true, the below fields describe the attribute which
+	 * this fake inode belongs to. The actual inode of this attribute is
+	 * pointed to by base_ntfs_ino and nr_extents is always set to -1 (see
+	 * below). For real inodes, we also set the type (AT_DATA for files and
+	 * AT_INDEX_ALLOCATION for directories), with the name = NULL and
+	 * name_len = 0 for files and name = I30 (global constant) and
+	 * name_len = 4 for directories.
+	 */
+	ATTR_TYPES type;	/* Attribute type of this fake inode. */
+	uchar_t *name;		/* Attribute name of this fake inode. */
+	u32 name_len;		/* Attribute name length of this fake inode. */
 	run_list run_list;	/* If state has the NI_NonResident bit set,
 				   the run list of the unnamed data attribute
 				   (if a file) or of the index allocation
-				   attribute (directory). If run_list.rl is
-				   NULL, the run list has not been read in or
-				   has been unmapped. If NI_NonResident is
-				   clear, the unnamed data attribute is
-				   resident (file) or there is no $I30 index
-				   allocation attribute (directory). In that
-				   case run_list.rl is always NULL.*/
+				   attribute (directory) or of the attribute
+				   described by the fake inode (if NInoAttr()).
+				   If run_list.rl is NULL, the run list has not
+				   been read in yet or has been unmapped. If
+				   NI_NonResident is clear, the attribute is
+				   resident (file and fake inode) or there is
+				   no $I30 index allocation attribute
+				   (small directory). In the latter case
+				   run_list.rl is always NULL.*/
+	/*
+	 * The following fields are only valid for real inodes and extent
+	 * inodes.
+	 */
 	struct rw_semaphore mrec_lock;	/* Lock for serializing access to the
 				   mft record belonging to this inode. */
 	atomic_t mft_count;	/* Mapping reference count for book keeping. */
@@ -74,17 +93,18 @@ struct _ntfs_inode {
 	union {
 		struct { /* It is a directory or $MFT. */
 			u32 index_block_size;	/* Size of an index block. */
-			u8 index_block_size_bits; /* Log2 of the above. */
 			u32 index_vcn_size;	/* Size of a vcn in this
 						   directory index. */
-			u8 index_vcn_size_bits;	/* Log2 of the above. */
 			s64 bmp_size;		/* Size of the $I30 bitmap. */
 			s64 bmp_initialized_size; /* Copy from $I30 bitmap. */
 			s64 bmp_allocated_size;	/* Copy from $I30 bitmap. */
 			run_list bmp_rl;	/* Run list for the $I30 bitmap
 						   if it is non-resident. */
+			u8 index_block_size_bits; /* Log2 of the above. */
+			u8 index_vcn_size_bits;	/* Log2 of the above. */
 		} SN(idm);
-		struct { /* It is a compressed file. */
+		struct { /* It is a compressed file or fake inode. */
+			s64 compressed_size;		/* Copy from $DATA. */
 			u32 compression_block_size;     /* Size of a compression
 						           block (cb). */
 			u8 compression_block_size_bits; /* Log2 of the size of
@@ -92,13 +112,13 @@ struct _ntfs_inode {
 			u8 compression_block_clusters;  /* Number of clusters
 							   per compression
 							   block. */
-			s64 compressed_size;		/* Copy from $DATA. */
 		} SN(icf);
 	} SN(idc);
 	struct semaphore extent_lock;	/* Lock for accessing/modifying the
 					   below . */
 	s32 nr_extents;	/* For a base mft record, the number of attached extent
-			   inodes (0 if none), for extent records this is -1. */
+			   inodes (0 if none), for extent records and for fake
+			   inodes describing an attribute this is -1. */
 	union {		/* This union is only used if nr_extents != 0. */
 		ntfs_inode **extent_ntfs_inos;	/* For nr_extents > 0, array of
 						   the ntfs inodes of the extent
@@ -107,7 +127,9 @@ struct _ntfs_inode {
 						   been loaded. */
 		ntfs_inode *base_ntfs_ino;	/* For nr_extents == -1, the
 						   ntfs inode of the base mft
-						   record. */
+						   record. For fake inodes, the
+						   real (base) inode to which
+						   the attribute belongs. */
 	} SN(ine);
 };
 
@@ -115,6 +137,79 @@ struct _ntfs_inode {
 #define _ICF(X)  SC(idc.icf,X)
 #define _INE(X)  SC(ine,X)
 
+/*
+ * Defined bits for the state field in the ntfs_inode structure.
+ * (f) = files only, (d) = directories only, (a) = attributes/fake inodes only
+ */
+typedef enum {
+	NI_Dirty,		/* 1: Mft record needs to be written to disk. */
+	NI_AttrList,		/* 1: Mft record contains an attribute list. */
+	NI_AttrListNonResident,	/* 1: Attribute list is non-resident. Implies
+				      NI_AttrList is set. */
+
+	NI_Attr,		/* 1: Fake inode for attribute i/o.
+				   0: Real inode or extent inode. */
+
+	NI_MstProtected,	/* 1: Attribute is protected by MST fixups.
+				   0: Attribute is not protected by fixups. */
+	NI_NonResident,		/* 1: Unnamed data attr is non-resident (f).
+				   1: Attribute is non-resident (a). */
+	NI_IndexAllocPresent = NI_NonResident,	/* 1: $I30 index alloc attr is
+						   present (d). */
+	NI_Compressed,		/* 1: Unnamed data attr is compressed (f).
+				   1: Create compressed files by default (d).
+				   1: Attribute is compressed (a). */
+	NI_Encrypted,		/* 1: Unnamed data attr is encrypted (f).
+				   1: Create encrypted files by default (d).
+				   1: Attribute is encrypted (a). */
+	NI_Sparse,		/* 1: Unnamed data attr is sparse (f).
+				   1: Create sparse files by default (d).
+				   1: Attribute is sparse (a). */
+	NI_BmpNonResident,	/* 1: $I30 bitmap attr is non resident (d). */
+} ntfs_inode_state_bits;
+
+/*
+ * NOTE: We should be adding dirty mft records to a list somewhere and they
+ * should be independent of the (ntfs/vfs) inode structure so that an inode can
+ * be removed but the record can be left dirty for syncing later.
+ */
+
+/*
+ * Macro tricks to expand the NInoFoo(), NInoSetFoo(), and NInoClearFoo()
+ * functions.
+ */
+#define NINO_FNS(flag)					\
+static inline int NIno##flag(ntfs_inode *ni)		\
+{							\
+	return test_bit(NI_##flag, &(ni)->state);	\
+}							\
+static inline void NInoSet##flag(ntfs_inode *ni)	\
+{							\
+	set_bit(NI_##flag, &(ni)->state);		\
+}							\
+static inline void NInoClear##flag(ntfs_inode *ni)	\
+{							\
+	clear_bit(NI_##flag, &(ni)->state);		\
+}
+
+/* Emit the ntfs inode bitops functions. */
+NINO_FNS(Dirty)
+NINO_FNS(AttrList)
+NINO_FNS(AttrListNonResident)
+NINO_FNS(Attr)
+NINO_FNS(MstProtected)
+NINO_FNS(NonResident)
+NINO_FNS(IndexAllocPresent)
+NINO_FNS(Compressed)
+NINO_FNS(Encrypted)
+NINO_FNS(Sparse)
+NINO_FNS(BmpNonResident)
+
+/*
+ * The full structure containing a ntfs_inode and a vfs struct inode. Used for
+ * all real and fake inodes but not for extent inodes which lack the vfs struct
+ * inode.
+ */
 typedef struct {
 	ntfs_inode ntfs_inode;
 	struct inode vfs_inode;		/* The vfs inode structure. */
@@ -140,8 +235,8 @@ extern struct inode *ntfs_alloc_big_inode(struct super_block *sb);
 extern void ntfs_destroy_big_inode(struct inode *inode);
 extern void ntfs_clear_big_inode(struct inode *vi);
 
-extern ntfs_inode *ntfs_new_inode(struct super_block *sb);
-extern void ntfs_clear_inode(ntfs_inode *ni);
+extern ntfs_inode *ntfs_new_extent_inode(struct super_block *sb);
+extern void ntfs_clear_extent_inode(ntfs_inode *ni);
 
 extern void ntfs_read_inode(struct inode *vi);
 extern void ntfs_read_inode_mount(struct inode *vi);

fs/ntfs/mft.c

@@ -102,7 +102,7 @@ extern int ntfs_mst_readpage(struct file *, struct page *);
  * ntfs_mft_aops - address space operations for access to $MFT
  *
  * Address space operations for access to $MFT. This allows us to simply use
- * read_cache_page() in map_mft_record().
+ * ntfs_map_page() in map_mft_record_page().
  */
 struct address_space_operations ntfs_mft_aops = {
 	writepage:	NULL,			/* Write dirty page to disk. */
@@ -334,9 +334,9 @@ void unmap_mft_record(const int rw, ntfs_inode *ni)
 
 	/*
 	 * If pure ntfs_inode, i.e. no vfs inode attached, we leave it to
-	 * ntfs_clear_inode() in the extent inode case, and to the caller in
-	 * the non-extent, yet pure ntfs inode case, to do the actual tear
-	 * down of all structures and freeing of all allocated memory.
+	 * ntfs_clear_extent_inode() in the extent inode case, and to the
+	 * caller in the non-extent, yet pure ntfs inode case, to do the actual
+	 * tear down of all structures and freeing of all allocated memory.
 	 */
 	return;
 }
@@ -417,7 +417,7 @@ MFT_RECORD *map_extent_mft_record(ntfs_inode *base_ni, MFT_REF mref,
 		return m;
 	}
 	/* Record wasn't there. Get a new ntfs inode and initialize it. */
-	ni = ntfs_new_inode(base_ni->vol->sb);
+	ni = ntfs_new_extent_inode(base_ni->vol->sb);
 	if (!ni) {
 		up(&base_ni->extent_lock);
 		atomic_dec(&base_ni->count);
@@ -433,7 +433,7 @@ MFT_RECORD *map_extent_mft_record(ntfs_inode *base_ni, MFT_REF mref,
 	if (IS_ERR(m)) {
 		up(&base_ni->extent_lock);
 		atomic_dec(&base_ni->count);
-		ntfs_clear_inode(ni);
+		ntfs_clear_extent_inode(ni);
 		goto map_err_out;
 	}
 	/* Verify the sequence number. */
@@ -479,7 +479,7 @@ MFT_RECORD *map_extent_mft_record(ntfs_inode *base_ni, MFT_REF mref,
 	 * release it or we will leak memory.
 	 */
 	if (destroy_ni)
-		ntfs_clear_inode(ni);
+		ntfs_clear_extent_inode(ni);
 	return m;
 }
 

fs/ntfs/ntfs.h

@@ -53,41 +53,6 @@ typedef enum {
 	NTFS_MAX_NAME_LEN	= 255,
 } NTFS_CONSTANTS;
 
-/*
- * Defined bits for the state field in the ntfs_inode structure.
- * (f) = files only, (d) = directories only
- */
-typedef enum {
-	NI_Dirty,		/* 1: Mft record needs to be written to disk. */
-	NI_AttrList,		/* 1: Mft record contains an attribute list. */
-	NI_AttrListNonResident,	/* 1: Attribute list is non-resident. Implies
-				      NI_AttrList is set. */
-	NI_NonResident,		/* 1: Unnamed data attr is non-resident (f).
-				   1: $I30 index alloc attr is present (d). */
-	NI_Compressed,		/* 1: Unnamed data attr is compressed (f).
-				   1: Create compressed files by default (d). */
-	NI_Encrypted,		/* 1: Unnamed data attr is encrypted (f).
-				   1: Create encrypted files by default (d). */
-	NI_BmpNonResident,	/* 1: $I30 bitmap attr is non resident (d). */
-} ntfs_inode_state_bits;
-
-/*
- * NOTE: We should be adding dirty mft records to a list somewhere and they
- * should be independent of the (ntfs/vfs) inode structure so that an inode can
- * be removed but the record can be left dirty for syncing later.
- */
-
-#define NInoDirty(n_ino)	  test_bit(NI_Dirty, &(n_ino)->state)
-#define NInoSetDirty(n_ino)	  set_bit(NI_Dirty, &(n_ino)->state)
-#define NInoClearDirty(n_ino)	  clear_bit(NI_Dirty, &(n_ino)->state)
-
-#define NInoAttrList(n_ino)	  test_bit(NI_AttrList, &(n_ino)->state)
-#define NInoNonResident(n_ino)	  test_bit(NI_NonResident, &(n_ino)->state)
-#define NInoIndexAllocPresent(n_ino) test_bit(NI_NonResident, &(n_ino)->state)
-#define NInoCompressed(n_ino)	  test_bit(NI_Compressed, &(n_ino)->state)
-#define NInoEncrypted(n_ino)	  test_bit(NI_Encrypted, &(n_ino)->state)
-#define NInoBmpNonResident(n_ino) test_bit(NI_BmpNonResident, &(n_ino)->state)
-
 /* Global variables. */
 
 /* Slab caches (from super.c). */

fs/ntfs/super.c

@@ -1709,10 +1709,11 @@ static int ntfs_fill_super(struct super_block *sb, void *opt, const int silent)
 	}
 #undef OGIN
 	/*
-	 * This is needed to get ntfs_clear_inode() called for each inode we
-	 * have ever called iget()/iput() on, otherwise we A) leak resources
-	 * and B) a subsequent mount fails automatically due to iget() never
-	 * calling down into our ntfs_read_inode{_mount}() methods again...
+	 * This is needed to get ntfs_clear_extent_inode() called for each
+	 * inode we have ever called iget()/iput() on, otherwise we A) leak
+	 * resources and B) a subsequent mount fails automatically due to
+	 * iget() never calling down into our ntfs_read_inode{_mount}() methods
+	 * again...
 	 */
 	if (invalidate_inodes(sb)) {
 		ntfs_error(sb, "Busy inodes left. This is most likely a NTFS "