Merge http://linux-ntfs.bkbits.net/ntfs-tng-2.5

into home.transmeta.com:/home/torvalds/v2.5/linux

Documentation/filesystems/ntfs.txt

@@ -247,6 +247,16 @@ ChangeLog
 
 Note, a technical ChangeLog aimed at kernel hackers is in fs/ntfs/ChangeLog.
 
+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
+	  32-bit architectures. This improves the speed of the driver a little.
+2.0.9:
+	- Change decompression engine to use a single buffer. This should not
+	  affect performance except perhaps on the most heavy i/o on SMP
+	  systems when accessing multiple compressed files from multiple
+	  devices simultaneously.
+	- Minor updates and cleanups.
 2.0.8:
 	- Remove now obsolete show_inodes and posix mount option(s).
 	- Restore show_sys_files mount option.

fs/ntfs/ChangeLog

@@ -23,6 +23,28 @@ ToDo:
 	  them cleaner and make code reuse easier.
 	- Want to use dummy inodes for address space i/o.
 
+2.0.10 - There can only be 2^32 - 1 inodes on an NTFS volume.
+
+	- Add check at mount time to verify that the number of inodes on the
+	  volume does not exceed 2^32 - 1, which is the maximum allowed for
+	  NTFS according to Microsoft.
+	- Change mft_no member of ntfs_inode structure to be unsigned long.
+	  Update all users. This makes ntfs_inode->mft_no just a copy of struct
+	  inode->i_ino. But we can't just always use struct inode->i_ino and
+	  remove mft_no because extent inodes do not have an attached struct
+	  inode.
+
+2.0.9 - Decompression engine now uses a single buffer and other cleanups.
+
+	- 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
+	  Morton)
+	- Change buffer size in ntfs_readdir()/ntfs_filldir() to use
+	  NLS_MAX_CHARSET_SIZE which makes the buffers almost 1kiB each but
+	  it also makes everything safer so it is a good thing.
+	- Miscellaneous minor cleanups to comments.
+
 2.0.8 - Major updates for handling of case sensitivity and dcache aliasing.
 
 	Big thanks go to Al Viro and other inhabitants of #kernel for investing

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=\"2.0.8\"
+EXTRA_CFLAGS = -DNTFS_VERSION=\"2.0.10\"
 
 ifeq ($(CONFIG_NTFS_DEBUG),y)
 EXTRA_CFLAGS += -DDEBUG

fs/ntfs/attrib.c

@@ -1243,8 +1243,7 @@ static BOOL find_external_attr(const ATTR_TYPES type, const uchar_t *name,
 
 	ni = ctx->ntfs_ino;
 	base_ni = ctx->base_ntfs_ino;
-	ntfs_debug("Entering for inode 0x%Lx, type 0x%x.",
-			(unsigned long long)ni->mft_no, type);
+	ntfs_debug("Entering for inode 0x%lx, type 0x%x.", ni->mft_no, type);
 	if (!base_ni) {
 		/* First call happens with the base mft record. */
 		base_ni = ctx->base_ntfs_ino = ctx->ntfs_ino;

fs/ntfs/compress.c

@@ -50,11 +50,13 @@ typedef enum {
 } ntfs_compression_constants;
 
 /**
- * ntfs_compression_buffer - one buffer for the decompression engine.
+ * ntfs_compression_buffer - one buffer for the decompression engine
  */
 static u8 *ntfs_compression_buffer = NULL;
 
-/* This spinlock which protects it */
+/**
+ * ntfs_cb_lock - spinlock which protects ntfs_compression_buffer
+ */
 static spinlock_t ntfs_cb_lock = SPIN_LOCK_UNLOCKED;
 
 /**
@@ -66,8 +68,6 @@ static spinlock_t ntfs_cb_lock = SPIN_LOCK_UNLOCKED;
  */
 int allocate_compression_buffers(void)
 {
-	int i, j;
-
 	BUG_ON(ntfs_compression_buffer);
 
 	ntfs_compression_buffer = vmalloc(NTFS_MAX_CB_SIZE);
@@ -83,8 +83,6 @@ int allocate_compression_buffers(void)
  */
 void free_compression_buffers(void)
 {
-	int i;
-
 	BUG_ON(!ntfs_compression_buffer);
 	vfree(ntfs_compression_buffer);
 	ntfs_compression_buffer = NULL;
@@ -592,7 +590,8 @@ int ntfs_file_read_compressed_block(struct page *page)
 
 	/*
 	 * Get the compression buffer. We must not sleep any more
-	 * until we are finished with it.  */
+	 * until we are finished with it.
+	 */
 	spin_lock(&ntfs_cb_lock);
 	cb = ntfs_compression_buffer;
 
@@ -668,7 +667,10 @@ int ntfs_file_read_compressed_block(struct page *page)
 			if (page)
 				memset(page_address(page) + cur_ofs, 0,
 						cb_max_ofs - cur_ofs);
-			cb_pos += cb_max_ofs - cur_ofs;
+			/*
+			 * No need to update cb_pos at this stage:
+			 *	cb_pos += cb_max_ofs - cur_ofs;
+			 */
 			cur_ofs = cb_max_ofs;
 		}
 	} else if (vcn == start_vcn) {
@@ -742,12 +744,13 @@ int ntfs_file_read_compressed_block(struct page *page)
 				cb_pos,	cb_size - (cb_pos - cb));
 		/*
 		 * We can sleep from now on, lock already dropped by
-		 * ntfs_decompress.  */
+		 * ntfs_decompress().
+		 */
 		if (err) {
 			ntfs_error(vol->sb, "ntfs_decompress() failed in inode "
-					"0x%Lx with error code %i. Skipping "
+					"0x%lx with error code %i. Skipping "
 					"this compression block.\n",
-					(unsigned long long)ni->mft_no, -err);
+					ni->mft_no, -err);
 			/* Release the unfinished pages. */
 			for (; prev_cur_page < cur_page; prev_cur_page++) {
 				page = pages[prev_cur_page];

fs/ntfs/dir.h

@@ -40,8 +40,8 @@ typedef struct {
 /* The little endian Unicode string $I30 as a global constant. */
 extern const uchar_t I30[5];
 
-extern u64 ntfs_lookup_inode_by_name(ntfs_inode *dir_ni, const uchar_t *uname,
-		const int uname_len, ntfs_name **res);
+extern MFT_REF ntfs_lookup_inode_by_name(ntfs_inode *dir_ni,
+		const uchar_t *uname, const int uname_len, ntfs_name **res);
 
 #endif /* _LINUX_NTFS_FS_DIR_H */
 

fs/ntfs/inode.c

@@ -357,8 +357,7 @@ void ntfs_read_inode(struct inode *vi)
 	if (lookup_attr(AT_ATTRIBUTE_LIST, NULL, 0, 0, 0, NULL, 0, ctx)) {
 		if (vi->i_ino == FILE_MFT)
 			goto skip_attr_list_load;
-		ntfs_debug("Attribute list found in inode %li (0x%lx).",
-				vi->i_ino, vi->i_ino);
+		ntfs_debug("Attribute list found in inode 0x%lx.", vi->i_ino);
 		ni->state |= 1 << NI_AttrList;
 		if (ctx->attr->flags & ATTR_IS_ENCRYPTED ||
 				ctx->attr->flags & ATTR_COMPRESSION_MASK) {
@@ -805,8 +804,8 @@ void ntfs_read_inode(struct inode *vi)
 ec_unm_err_out:
 	unmap_mft_record(READ, ni);
 err_out:
-	ntfs_error(vi->i_sb, "Failed with error code %i. Marking inode "
-			"%li (0x%lx) as bad.", -err, vi->i_ino, vi->i_ino);
+	ntfs_error(vi->i_sb, "Failed with error code %i. Marking inode 0x%lx "
+			"as bad.", -err, vi->i_ino);
 	make_bad_inode(vi);
 	return;
 }
@@ -857,7 +856,7 @@ void ntfs_read_inode_mount(struct inode *vi)
 	ntfs_init_big_inode(vi);
 	ni = NTFS_I(vi);
 	if (vi->i_ino != FILE_MFT) {
-		ntfs_error(sb, "Called for inode %ld but only inode %d "
+		ntfs_error(sb, "Called for inode 0x%lx but only inode %d "
 				"allowed.", vi->i_ino, FILE_MFT);
 		goto err_out;
 	}
@@ -1034,7 +1033,7 @@ void ntfs_read_inode_mount(struct inode *vi)
 			if (al_entry->lowest_vcn)
 				goto em_put_err_out;
 			/* First entry has to be in the base mft record. */
-			if (MREF_LE(al_entry->mft_reference) != ni->mft_no) {
+			if (MREF_LE(al_entry->mft_reference) != vi->i_ino) {
 				/* MFT references do not match, logic fails. */
 				ntfs_error(sb, "BUG: The first $DATA extent "
 						"of $MFT is not in the base "
@@ -1096,6 +1095,8 @@ void ntfs_read_inode_mount(struct inode *vi)
 
 		/* Are we in the first extent? */
 		if (!next_vcn) {
+			u64 ll;
+
 			if (attr->_ANR(lowest_vcn)) {
 				ntfs_error(sb, "First extent of $DATA "
 						"attribute has non zero "
@@ -1113,8 +1114,16 @@ void ntfs_read_inode_mount(struct inode *vi)
 			ni->allocated_size = sle64_to_cpu(
 					attr->_ANR(allocated_size));
 			/* Set the number of mft records. */
-			vol->_VMM(nr_mft_records) = vi->i_size >>
-					vol->mft_record_size_bits;
+			ll = vi->i_size >> vol->mft_record_size_bits;
+			/*
+			 * Verify the number of mft records does not exceed
+			 * 2^32 - 1.
+			 */
+			if (ll >= (1ULL << 32)) {
+				ntfs_error(sb, "$MFT is too big! Aborting.");
+				goto put_err_out;
+			}
+			vol->_VMM(nr_mft_records) = ll;
 			/*
 			 * We have got the first extent of the run_list for
 			 * $MFT which means it is now relatively safe to call
@@ -1255,8 +1264,7 @@ void ntfs_dirty_inode(struct inode *vi)
  */
 int ntfs_commit_inode(ntfs_inode *ni)
 {
-	ntfs_debug("Entering for inode 0x%Lx.",
-			(unsigned long long)ni->mft_no);
+	ntfs_debug("Entering for inode 0x%lx.", ni->mft_no);
 	NInoClearDirty(ni);
 	return 0;
 }
@@ -1265,8 +1273,7 @@ void __ntfs_clear_inode(ntfs_inode *ni)
 {
 	int err;
 
-	ntfs_debug("Entering for inode 0x%Lx.",
-			(unsigned long long)ni->mft_no);
+	ntfs_debug("Entering for inode 0x%lx.", ni->mft_no);
 	if (NInoDirty(ni)) {
 		err = ntfs_commit_inode(ni);
 		if (err) {

fs/ntfs/inode.h

@@ -39,7 +39,7 @@ struct _ntfs_inode {
 	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. */
-	u64 mft_no;		/* Mft record number (inode number). */
+	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. */

fs/ntfs/layout.h

@@ -278,9 +278,9 @@ typedef enum {
 
 typedef u64 MFT_REF;
 
-#define MREF(x)		((u64)((x) & MFT_REF_MASK_CPU))
+#define MREF(x)		((unsigned long)((x) & MFT_REF_MASK_CPU))
 #define MSEQNO(x)	((u16)(((x) >> 48) & 0xffff))
-#define MREF_LE(x)	((u64)(le64_to_cpu(x) & MFT_REF_MASK_CPU))
+#define MREF_LE(x)	((unsigned long)(le64_to_cpu(x) & MFT_REF_MASK_CPU))
 #define MSEQNO_LE(x)	((u16)((le64_to_cpu(x) >> 48) & 0xffff))
 
 #define IS_ERR_MREF(x)	(((x) & 0x0000800000000000ULL) ? 1 : 0)
@@ -948,8 +948,7 @@ typedef struct {
 					   data attribute. NOTE: Is a multiple
 					   of the cluster size. */
 /* 30*/	s64 data_size;			/* Byte size of actual data in data
-					   attribute. NOTE: Only present when
-					   lowest_vcn is 0. */
+					   attribute. */
 /* 38*/	FILE_ATTR_FLAGS file_attributes;	/* Flags describing the file. */
 /* 3c*/	union {
 	/* 3c*/	struct {

fs/ntfs/mft.c

@@ -240,8 +240,7 @@ static inline void unmap_mft_record_page(ntfs_inode *ni)
  *
  * The mft record is now ours and we return a pointer to it. You need to check
  * the returned pointer with IS_ERR() and if that is true, PTR_ERR() will return
- * the error code. The following error codes are defined:
- * 	TODO: Fill in the possible error codes.
+ * the error code.
  *
  * NOTE: Caller is responsible for setting the mft record dirty before calling
  * unmap_mft_record(). This is obviously only necessary if the caller really
@@ -254,8 +253,7 @@ MFT_RECORD *map_mft_record(const int rw, ntfs_inode *ni)
 {
 	MFT_RECORD *m;
 
-	ntfs_debug("Entering for i_ino 0x%Lx, mapping for %s.",
-			(unsigned long long)ni->mft_no,
+	ntfs_debug("Entering for mft_no 0x%lx, mapping for %s.", ni->mft_no,
 			rw == READ ? "READ" : "WRITE");
 
 	/* Make sure the ntfs inode doesn't go away. */
@@ -318,8 +316,7 @@ void unmap_mft_record(const int rw, ntfs_inode *ni)
 
 	BUG_ON(!atomic_read(&ni->mft_count) || !page);
 
-	ntfs_debug("Entering for mft_no 0x%Lx, unmapping from %s.",
-			(unsigned long long)ni->mft_no,
+	ntfs_debug("Entering for mft_no 0x%lx, unmapping from %s.", ni->mft_no,
 			rw == READ ? "READ" : "WRITE");
 
 	/* Only release the actual page mapping if this is the last one. */
@@ -369,13 +366,12 @@ MFT_RECORD *map_extent_mft_record(ntfs_inode *base_ni, MFT_REF mref,
 	ntfs_inode *ni = NULL;
 	ntfs_inode **extent_nis = NULL;
 	int i;
-	u64 mft_no = MREF_LE(mref);
+	unsigned long mft_no = MREF_LE(mref);
 	u16 seq_no = MSEQNO_LE(mref);
 	BOOL destroy_ni = FALSE;
 
-	ntfs_debug("Mapping extent mft record 0x%Lx (base mft record 0x%Lx).",
-			(unsigned long long)mft_no,
-			(unsigned long long)base_ni->mft_no);
+	ntfs_debug("Mapping extent mft record 0x%lx (base mft record 0x%lx).",
+			mft_no, base_ni->mft_no);
 	/* Make sure the base ntfs inode doesn't go away. */
 	atomic_inc(&base_ni->count);
 	/*

fs/ntfs/namei.c

@@ -93,7 +93,7 @@ static struct dentry *ntfs_lookup(struct inode *dir_ino, struct dentry *dent)
 	struct inode *dent_inode;
 	uchar_t *uname;
 	ntfs_name *name = NULL;
-	u64 mref;
+	MFT_REF mref;
 	unsigned long dent_ino;
 	int uname_len;
 
@@ -110,7 +110,7 @@ static struct dentry *ntfs_lookup(struct inode *dir_ino, struct dentry *dent)
 			&name);
 	kmem_cache_free(ntfs_name_cache, uname);
 	if (!IS_ERR_MREF(mref)) {
-		dent_ino = (unsigned long)MREF(mref);
+		dent_ino = MREF(mref);
 		ntfs_debug("Found inode 0x%lx. Calling iget.", dent_ino);
 		dent_inode = iget(vol->sb, dent_ino);
 		if (dent_inode) {
@@ -130,17 +130,15 @@ static struct dentry *ntfs_lookup(struct inode *dir_ino, struct dentry *dent)
 				goto handle_name;
 			}
 			ntfs_error(vol->sb, "Found stale reference to inode "
-					"0x%Lx (reference sequence number = "
+					"0x%lx (reference sequence number = "
 					"0x%x, inode sequence number = 0x%x, "
 					"returning -EACCES. Run chkdsk.",
-					(unsigned long long)MREF(mref),
-					MSEQNO(mref),
+					dent_ino, MSEQNO(mref),
 					NTFS_I(dent_inode)->seq_no);
 			iput(dent_inode);
 		} else
-			ntfs_error(vol->sb, "iget(0x%Lx) failed, returning "
-					"-EACCES.",
-					(unsigned long long)MREF(mref));
+			ntfs_error(vol->sb, "iget(0x%lx) failed, returning "
+					"-EACCES.", dent_ino);
 		if (name)
 			kfree(name);
 		return ERR_PTR(-EACCES);
@@ -235,8 +233,9 @@ static struct dentry *ntfs_lookup(struct inode *dir_ino, struct dentry *dent)
 	}
 	nls_name.hash = full_name_hash(nls_name.name, nls_name.len);
 
-	// FIXME: Do we need dcache_lock or dparent_lock here or is the
-	// fact that i_sem is held on the parent inode sufficient? (AIA)
+	/*
+	 * Note: No need for dparent_lock as i_sem is held on the parent inode.
+	 */
 
 	/* Does a dentry matching the nls_name exist already? */
 	real_dent = d_lookup(dent->d_parent, &nls_name);

fs/ntfs/super.c

@@ -1236,14 +1236,13 @@ s64 get_nr_free_clusters(ntfs_volume *vol)
  * Errors are ignored and we just return the number of free inodes we have
  * found. This means we return an underestimate on error.
  */
-s64 get_nr_free_mft_records(ntfs_volume *vol)
+unsigned long get_nr_free_mft_records(ntfs_volume *vol)
 {
 	struct address_space *mapping;
 	filler_t *readpage;
 	struct page *page;
-	unsigned long index, max_index;
+	unsigned long index, max_index, nr_free = 0;
 	unsigned int max_size, i;
-	s64 nr_free = 0LL;
 	u32 *b;
 
 	ntfs_debug("Entering.");
@@ -1285,7 +1284,7 @@ s64 get_nr_free_mft_records(ntfs_volume *vol)
 		b = (u32*)kmap(page);
 		/* For each 4 bytes, add up the number of zero bits. */
 	  	for (i = 0; i < max_size; i++)
-			nr_free += (s64)(32 - hweight32(b[i]));
+			nr_free += 32 - hweight32(b[i]);
 		kunmap(page);
 		page_cache_release(page);
 	}
@@ -1300,7 +1299,7 @@ s64 get_nr_free_mft_records(ntfs_volume *vol)
 		if (max_size) {
 			/* Compensate for out of bounds zero bits. */
 			if ((i = vol->_VMM(nr_mft_records) & 31))
-				nr_free -= (s64)(32 - i);
+				nr_free -= 32 - i;
 			ntfs_debug("Handling partial page, max_size = 0x%x",
 					max_size);
 			goto handle_partial_page;
@@ -1358,10 +1357,7 @@ int ntfs_statfs(struct super_block *sb, struct statfs *sfs)
 	/* Total file nodes in file system (at this moment in time). */
 	sfs->f_files  = vol->mft_ino->i_size >> vol->mft_record_size_bits;
 	/* Free file nodes in fs (based on current total count). */
-	size	      = get_nr_free_mft_records(vol);
-	if (size < 0LL)
-		size = 0LL;
-	sfs->f_ffree = size;
+	sfs->f_ffree = get_nr_free_mft_records(vol);
 	/*
 	 * File system id. This is extremely *nix flavour dependent and even
 	 * within Linux itself all fs do their own thing. I interpret this to

fs/ntfs/volume.h

@@ -105,8 +105,8 @@ typedef struct {
 	struct rw_semaphore mftbmp_lock; /* Lock for serializing accesses to the
 					    mft record bitmap ($MFT/$BITMAP). */
 	union {
-		s64 nr_mft_records;	/* Number of records in the mft. */
-		s64 nr_mft_bits;	/* Number of bits in mft bitmap. */
+		unsigned long nr_mft_records; /* Number of mft records. */
+		unsigned long nr_mft_bits; /* Number of bits in mft bitmap. */
 	} SN(vmm);
 	struct address_space mftbmp_mapping; /* Page cache for $MFT/$BITMAP. */
 	run_list mftbmp_rl;		/* Run list for $MFT/$BITMAP. */