[linux.git] / fs / logfs / file.c

/*
 * fs/logfs/file.c	- prepare_write, commit_write and friends
 *
 * As should be obvious for Linux kernel code, license is GPLv2
 *
 * Copyright (c) 2005-2008 Joern Engel <[email protected]>
 */
#include "logfs.h"
#include <linux/sched.h>
#include <linux/writeback.h>

static int logfs_write_begin(struct file *file, struct address_space *mapping,
		loff_t pos, unsigned len, unsigned flags,
		struct page **pagep, void **fsdata)
{
	struct inode *inode = mapping->host;
	struct page *page;
	pgoff_t index = pos >> PAGE_CACHE_SHIFT;

	page = grab_cache_page_write_begin(mapping, index, flags);
	if (!page)
		return -ENOMEM;
	*pagep = page;

	if ((len == PAGE_CACHE_SIZE) || PageUptodate(page))
		return 0;
	if ((pos & PAGE_CACHE_MASK) >= i_size_read(inode)) {
		unsigned start = pos & (PAGE_CACHE_SIZE - 1);
		unsigned end = start + len;

		/* Reading beyond i_size is simple: memset to zero */
		zero_user_segments(page, 0, start, end, PAGE_CACHE_SIZE);
		return 0;
	}
	return logfs_readpage_nolock(page);
}

static int logfs_write_end(struct file *file, struct address_space *mapping,
		loff_t pos, unsigned len, unsigned copied, struct page *page,
		void *fsdata)
{
	struct inode *inode = mapping->host;
	pgoff_t index = page->index;
	unsigned start = pos & (PAGE_CACHE_SIZE - 1);
	unsigned end = start + copied;
	int ret = 0;

	BUG_ON(PAGE_CACHE_SIZE != inode->i_sb->s_blocksize);
	BUG_ON(page->index > I3_BLOCKS);

	if (copied < len) {
		/*
		 * Short write of a non-initialized paged.  Just tell userspace
		 * to retry the entire page.
		 */
		if (!PageUptodate(page)) {
			copied = 0;
			goto out;
		}
	}
	if (copied == 0)
		goto out; /* FIXME: do we need to update inode? */

	if (i_size_read(inode) < (index << PAGE_CACHE_SHIFT) + end) {
		i_size_write(inode, (index << PAGE_CACHE_SHIFT) + end);
		mark_inode_dirty_sync(inode);
	}

	SetPageUptodate(page);
	if (!PageDirty(page)) {
		if (!get_page_reserve(inode, page))
			__set_page_dirty_nobuffers(page);
		else
			ret = logfs_write_buf(inode, page, WF_LOCK);
	}
out:
	unlock_page(page);
	page_cache_release(page);
	return ret ? ret : copied;
}

int logfs_readpage(struct file *file, struct page *page)
{
	int ret;

	ret = logfs_readpage_nolock(page);
	unlock_page(page);
	return ret;
}

/* Clear the page's dirty flag in the radix tree. */
/* TODO: mucking with PageWriteback is silly.  Add a generic function to clear
 * the dirty bit from the radix tree for filesystems that don't have to wait
 * for page writeback to finish (i.e. any compressing filesystem).
 */
static void clear_radix_tree_dirty(struct page *page)
{
	BUG_ON(PagePrivate(page) || page->private);
	set_page_writeback(page);
	end_page_writeback(page);
}

static int __logfs_writepage(struct page *page)
{
	struct inode *inode = page->mapping->host;
	int err;

	err = logfs_write_buf(inode, page, WF_LOCK);
	if (err)
		set_page_dirty(page);
	else
		clear_radix_tree_dirty(page);
	unlock_page(page);
	return err;
}

static int logfs_writepage(struct page *page, struct writeback_control *wbc)
{
	struct inode *inode = page->mapping->host;
	loff_t i_size = i_size_read(inode);
	pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
	unsigned offset;
	u64 bix;
	level_t level;

	log_file("logfs_writepage(%lx, %lx, %p)\n", inode->i_ino, page->index,
			page);

	logfs_unpack_index(page->index, &bix, &level);

	/* Indirect blocks are never truncated */
	if (level != 0)
		return __logfs_writepage(page);

	/*
	 * TODO: everything below is a near-verbatim copy of nobh_writepage().
	 * The relevant bits should be factored out after logfs is merged.
	 */

	/* Is the page fully inside i_size? */
	if (bix < end_index)
		return __logfs_writepage(page);

	 /* Is the page fully outside i_size? (truncate in progress) */
	offset = i_size & (PAGE_CACHE_SIZE-1);
	if (bix > end_index || offset == 0) {
		unlock_page(page);
		return 0; /* don't care */
	}

	/*
	 * The page straddles i_size.  It must be zeroed out on each and every
	 * writepage invokation because it may be mmapped.  "A file is mapped
	 * in multiples of the page size.  For a file that is not a multiple of
	 * the  page size, the remaining memory is zeroed when mapped, and
	 * writes to that region are not written out to the file."
	 */
	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
	return __logfs_writepage(page);
}

static void logfs_invalidatepage(struct page *page, unsigned long offset)
{
	struct logfs_block *block = logfs_block(page);

	if (block->reserved_bytes) {
		struct super_block *sb = page->mapping->host->i_sb;
		struct logfs_super *super = logfs_super(sb);

		super->s_dirty_pages -= block->reserved_bytes;
		block->ops->free_block(sb, block);
		BUG_ON(bitmap_weight(block->alias_map, LOGFS_BLOCK_FACTOR));
	} else
		move_page_to_btree(page);
	BUG_ON(PagePrivate(page) || page->private);
}

static int logfs_releasepage(struct page *page, gfp_t only_xfs_uses_this)
{
	return 0; /* None of these are easy to release */
}


long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
{
	struct inode *inode = file->f_path.dentry->d_inode;
	struct logfs_inode *li = logfs_inode(inode);
	unsigned int oldflags, flags;
	int err;

	switch (cmd) {
	case FS_IOC_GETFLAGS:
		flags = li->li_flags & LOGFS_FL_USER_VISIBLE;
		return put_user(flags, (int __user *)arg);
	case FS_IOC_SETFLAGS:
		if (IS_RDONLY(inode))
			return -EROFS;

		if (!inode_owner_or_capable(inode))
			return -EACCES;

		err = get_user(flags, (int __user *)arg);
		if (err)
			return err;

		mutex_lock(&inode->i_mutex);
		oldflags = li->li_flags;
		flags &= LOGFS_FL_USER_MODIFIABLE;
		flags |= oldflags & ~LOGFS_FL_USER_MODIFIABLE;
		li->li_flags = flags;
		mutex_unlock(&inode->i_mutex);

		inode->i_ctime = CURRENT_TIME;
		mark_inode_dirty_sync(inode);
		return 0;

	default:
		return -ENOTTY;
	}
}

int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
{
	struct super_block *sb = file->f_mapping->host->i_sb;
	struct inode *inode = file->f_mapping->host;
	int ret;

	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
	if (ret)
		return ret;

	mutex_lock(&inode->i_mutex);
	logfs_write_anchor(sb);
	mutex_unlock(&inode->i_mutex);

	return 0;
}

static int logfs_setattr(struct dentry *dentry, struct iattr *attr)
{
	struct inode *inode = dentry->d_inode;
	int err = 0;

	err = inode_change_ok(inode, attr);
	if (err)
		return err;

	if (attr->ia_valid & ATTR_SIZE) {
		err = logfs_truncate(inode, attr->ia_size);
		if (err)
			return err;
	}

	setattr_copy(inode, attr);
	mark_inode_dirty(inode);
	return 0;
}

const struct inode_operations logfs_reg_iops = {
	.setattr	= logfs_setattr,
};

const struct file_operations logfs_reg_fops = {
	.aio_read	= generic_file_aio_read,
	.aio_write	= generic_file_aio_write,
	.fsync		= logfs_fsync,
	.unlocked_ioctl	= logfs_ioctl,
	.llseek		= generic_file_llseek,
	.mmap		= generic_file_readonly_mmap,
	.open		= generic_file_open,
	.read		= do_sync_read,
	.write		= do_sync_write,
};

const struct address_space_operations logfs_reg_aops = {
	.invalidatepage	= logfs_invalidatepage,
	.readpage	= logfs_readpage,
	.releasepage	= logfs_releasepage,
	.set_page_dirty	= __set_page_dirty_nobuffers,
	.writepage	= logfs_writepage,
	.writepages	= generic_writepages,
	.write_begin	= logfs_write_begin,
	.write_end	= logfs_write_end,
};
Commit	Line	Data
5db53f3e JE	1	/*
	2	* fs/logfs/file.c - prepare_write, commit_write and friends
	3	*
	4	* As should be obvious for Linux kernel code, license is GPLv2
	5	*
	6	* Copyright (c) 2005-2008 Joern Engel <[email protected]>
	7	*/
	8	#include "logfs.h"
	9	#include <linux/sched.h>
	10	#include <linux/writeback.h>
	11
	12	static int logfs_write_begin(struct file file, struct address_space mapping,
	13	loff_t pos, unsigned len, unsigned flags,
	14	struct page pagep, void fsdata)
	15	{
	16	struct inode *inode = mapping->host;
	17	struct page *page;
	18	pgoff_t index = pos >> PAGE_CACHE_SHIFT;
	19
	20	page = grab_cache_page_write_begin(mapping, index, flags);
	21	if (!page)
	22	return -ENOMEM;
	23	*pagep = page;
	24
	25	if ((len == PAGE_CACHE_SIZE) \|\| PageUptodate(page))
	26	return 0;
	27	if ((pos & PAGE_CACHE_MASK) >= i_size_read(inode)) {
	28	unsigned start = pos & (PAGE_CACHE_SIZE - 1);
	29	unsigned end = start + len;
	30
	31	/* Reading beyond i_size is simple: memset to zero */
	32	zero_user_segments(page, 0, start, end, PAGE_CACHE_SIZE);
	33	return 0;
	34	}
	35	return logfs_readpage_nolock(page);
	36	}
	37
	38	static int logfs_write_end(struct file file, struct address_space mapping,
	39	loff_t pos, unsigned len, unsigned copied, struct page *page,
	40	void *fsdata)
	41	{
	42	struct inode *inode = mapping->host;
	43	pgoff_t index = page->index;
	44	unsigned start = pos & (PAGE_CACHE_SIZE - 1);
	45	unsigned end = start + copied;
	46	int ret = 0;
	47
	48	BUG_ON(PAGE_CACHE_SIZE != inode->i_sb->s_blocksize);
	49	BUG_ON(page->index > I3_BLOCKS);
	50
	51	if (copied < len) {
	52	/*
	53	* Short write of a non-initialized paged. Just tell userspace
	54	* to retry the entire page.
	55	*/
	56	if (!PageUptodate(page)) {
	57	copied = 0;
	58	goto out;
	59	}
	60	}
	61	if (copied == 0)
	62	goto out; /* FIXME: do we need to update inode? */
	63
	64	if (i_size_read(inode) < (index << PAGE_CACHE_SHIFT) + end) {
65	i_size_write(inode, (index << PAGE_CACHE_SHIFT) + end);
66	mark_inode_dirty_sync(inode);
67	}
68
69	SetPageUptodate(page);
70	if (!PageDirty(page)) {
71	if (!get_page_reserve(inode, page))
72	__set_page_dirty_nobuffers(page);
73	else
74	ret = logfs_write_buf(inode, page, WF_LOCK);
75	}
76	out:
77	unlock_page(page);
78	page_cache_release(page);
79	return ret ? ret : copied;
80	}
81
82	int logfs_readpage(struct file file, struct page page)
83	{
84	int ret;
85
86	ret = logfs_readpage_nolock(page);
87	unlock_page(page);
88	return ret;
89	}
90
91	/* Clear the page's dirty flag in the radix tree. */
92	/* TODO: mucking with PageWriteback is silly. Add a generic function to clear
93	* the dirty bit from the radix tree for filesystems that don't have to wait
94	* for page writeback to finish (i.e. any compressing filesystem).
95	*/
96	static void clear_radix_tree_dirty(struct page *page)
97	{
98	BUG_ON(PagePrivate(page) \|\| page->private);
99	set_page_writeback(page);
100	end_page_writeback(page);
101	}
102
103	static int __logfs_writepage(struct page *page)
104	{
105	struct inode *inode = page->mapping->host;
106	int err;
107
108	err = logfs_write_buf(inode, page, WF_LOCK);
109	if (err)
110	set_page_dirty(page);
111	else
112	clear_radix_tree_dirty(page);
113	unlock_page(page);
114	return err;
115	}
116
117	static int logfs_writepage(struct page page, struct writeback_control wbc)
118	{
119	struct inode *inode = page->mapping->host;
120	loff_t i_size = i_size_read(inode);
121	pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
122	unsigned offset;
123	u64 bix;
124	level_t level;
125
126	log_file("logfs_writepage(%lx, %lx, %p)\n", inode->i_ino, page->index,
127	page);
128
129	logfs_unpack_index(page->index, &bix, &level);
130
131	/* Indirect blocks are never truncated */
132	if (level != 0)
133	return __logfs_writepage(page);
134
135	/*
136	* TODO: everything below is a near-verbatim copy of nobh_writepage().
137	* The relevant bits should be factored out after logfs is merged.
138	*/
139
140	/* Is the page fully inside i_size? */
141	if (bix < end_index)
142	return __logfs_writepage(page);
143
144	/* Is the page fully outside i_size? (truncate in progress) */
145	offset = i_size & (PAGE_CACHE_SIZE-1);
146	if (bix > end_index \|\| offset == 0) {
147	unlock_page(page);
148	return 0; /* don't care */
149	}
150
151	/*
152	* The page straddles i_size. It must be zeroed out on each and every
153	* writepage invokation because it may be mmapped. "A file is mapped
154	* in multiples of the page size. For a file that is not a multiple of
155	* the page size, the remaining memory is zeroed when mapped, and
156	* writes to that region are not written out to the file."
157	*/
158	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
159	return __logfs_writepage(page);
160	}
161
162	static void logfs_invalidatepage(struct page *page, unsigned long offset)
163	{
05ebad85 JE	164	struct logfs_block *block = logfs_block(page);
	165
	166	if (block->reserved_bytes) {
	167	struct super_block *sb = page->mapping->host->i_sb;
	168	struct logfs_super *super = logfs_super(sb);
	169
	170	super->s_dirty_pages -= block->reserved_bytes;
	171	block->ops->free_block(sb, block);
	172	BUG_ON(bitmap_weight(block->alias_map, LOGFS_BLOCK_FACTOR));
	173	} else
	174	move_page_to_btree(page);
5db53f3e JE	175	BUG_ON(PagePrivate(page) \|\| page->private);
	176	}
	177
	178	static int logfs_releasepage(struct page *page, gfp_t only_xfs_uses_this)
	179	{
	180	return 0; /* None of these are easy to release */
	181	}
	182
	183
02d6d685	184	long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
5db53f3e	185	{
02d6d685	186	struct inode *inode = file->f_path.dentry->d_inode;
5db53f3e JE	187	struct logfs_inode *li = logfs_inode(inode);
	188	unsigned int oldflags, flags;
	189	int err;
	190
	191	switch (cmd) {
	192	case FS_IOC_GETFLAGS:
	193	flags = li->li_flags & LOGFS_FL_USER_VISIBLE;
	194	return put_user(flags, (int __user *)arg);
	195	case FS_IOC_SETFLAGS:
	196	if (IS_RDONLY(inode))
	197	return -EROFS;
	198
2e149670	199	if (!inode_owner_or_capable(inode))
5db53f3e JE	200	return -EACCES;
	201
	202	err = get_user(flags, (int __user *)arg);
	203	if (err)
	204	return err;
	205
	206	mutex_lock(&inode->i_mutex);
	207	oldflags = li->li_flags;
	208	flags &= LOGFS_FL_USER_MODIFIABLE;
	209	flags \|= oldflags & ~LOGFS_FL_USER_MODIFIABLE;
	210	li->li_flags = flags;
	211	mutex_unlock(&inode->i_mutex);
	212
	213	inode->i_ctime = CURRENT_TIME;
	214	mark_inode_dirty_sync(inode);
	215	return 0;
	216
	217	default:
	218	return -ENOTTY;
	219	}
	220	}
	221
02c24a82	222	int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
5db53f3e	223	{
7ea80859	224	struct super_block *sb = file->f_mapping->host->i_sb;
02c24a82 JB	225	struct inode *inode = file->f_mapping->host;
	226	int ret;
	227
	228	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
	229	if (ret)
	230	return ret;
5db53f3e	231
02c24a82	232	mutex_lock(&inode->i_mutex);
c0c79c31	233	logfs_write_anchor(sb);
02c24a82 JB	234	mutex_unlock(&inode->i_mutex);
02c24a82 JB	235
5db53f3e JE	236	return 0;
	237	}
	238
	239	static int logfs_setattr(struct dentry dentry, struct iattr attr)
	240	{
	241	struct inode *inode = dentry->d_inode;
	242	int err = 0;
	243
db78b877 CH	244	err = inode_change_ok(inode, attr);
	245	if (err)
	246	return err;
	247
1025774c	248	if (attr->ia_valid & ATTR_SIZE) {
5db53f3e	249	err = logfs_truncate(inode, attr->ia_size);
1025774c CH	250	if (err)
	251	return err;
	252	}
5db53f3e	253
1025774c CH	254	setattr_copy(inode, attr);
	255	mark_inode_dirty(inode);
	256	return 0;
5db53f3e JE	257	}
	258
	259	const struct inode_operations logfs_reg_iops = {
	260	.setattr = logfs_setattr,
	261	};
	262
	263	const struct file_operations logfs_reg_fops = {
	264	.aio_read = generic_file_aio_read,
	265	.aio_write = generic_file_aio_write,
	266	.fsync = logfs_fsync,
02d6d685	267	.unlocked_ioctl = logfs_ioctl,
5db53f3e JE	268	.llseek = generic_file_llseek,
	269	.mmap = generic_file_readonly_mmap,
	270	.open = generic_file_open,
	271	.read = do_sync_read,
	272	.write = do_sync_write,
	273	};
	274
	275	const struct address_space_operations logfs_reg_aops = {
	276	.invalidatepage = logfs_invalidatepage,
	277	.readpage = logfs_readpage,
	278	.releasepage = logfs_releasepage,
	279	.set_page_dirty = __set_page_dirty_nobuffers,
	280	.writepage = logfs_writepage,
	281	.writepages = generic_writepages,
	282	.write_begin = logfs_write_begin,
	283	.write_end = logfs_write_end,
	284	};