linux/fs/bfs/inode.c

533 lines
12 KiB
C
Raw Normal View History

/*
* fs/bfs/inode.c
* BFS superblock and inode operations.
* Copyright (C) 1999-2006 Tigran Aivazian <tigran@aivazian.fsnet.co.uk>
* From fs/minix, Copyright (C) 1991, 1992 Linus Torvalds.
*
* Made endianness-clean by Andrew Stribblehill <ads@wompom.org>, 2005.
*/
#include <linux/module.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/fs.h>
#include <linux/smp_lock.h>
#include <linux/buffer_head.h>
#include <linux/vfs.h>
#include <linux/writeback.h>
#include <asm/uaccess.h>
#include "bfs.h"
MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>");
MODULE_DESCRIPTION("SCO UnixWare BFS filesystem for Linux");
MODULE_LICENSE("GPL");
#undef DEBUG
#ifdef DEBUG
#define dprintf(x...) printf(x)
#else
#define dprintf(x...)
#endif
static void bfs_write_super(struct super_block *s);
void dump_imap(const char *prefix, struct super_block *s);
struct inode *bfs_iget(struct super_block *sb, unsigned long ino)
{
struct bfs_inode *di;
struct inode *inode;
struct buffer_head *bh;
int block, off;
inode = iget_locked(sb, ino);
if (IS_ERR(inode))
return ERR_PTR(-ENOMEM);
if (!(inode->i_state & I_NEW))
return inode;
if ((ino < BFS_ROOT_INO) || (ino > BFS_SB(inode->i_sb)->si_lasti)) {
printf("Bad inode number %s:%08lx\n", inode->i_sb->s_id, ino);
goto error;
}
block = (ino - BFS_ROOT_INO) / BFS_INODES_PER_BLOCK + 1;
bh = sb_bread(inode->i_sb, block);
if (!bh) {
printf("Unable to read inode %s:%08lx\n", inode->i_sb->s_id,
ino);
goto error;
}
off = (ino - BFS_ROOT_INO) % BFS_INODES_PER_BLOCK;
di = (struct bfs_inode *)bh->b_data + off;
inode->i_mode = 0x0000FFFF & le32_to_cpu(di->i_mode);
if (le32_to_cpu(di->i_vtype) == BFS_VDIR) {
inode->i_mode |= S_IFDIR;
inode->i_op = &bfs_dir_inops;
inode->i_fop = &bfs_dir_operations;
} else if (le32_to_cpu(di->i_vtype) == BFS_VREG) {
inode->i_mode |= S_IFREG;
inode->i_op = &bfs_file_inops;
inode->i_fop = &bfs_file_operations;
inode->i_mapping->a_ops = &bfs_aops;
}
BFS_I(inode)->i_sblock = le32_to_cpu(di->i_sblock);
BFS_I(inode)->i_eblock = le32_to_cpu(di->i_eblock);
BFS_I(inode)->i_dsk_ino = le16_to_cpu(di->i_ino);
inode->i_uid = le32_to_cpu(di->i_uid);
inode->i_gid = le32_to_cpu(di->i_gid);
inode->i_nlink = le32_to_cpu(di->i_nlink);
inode->i_size = BFS_FILESIZE(di);
inode->i_blocks = BFS_FILEBLOCKS(di);
inode->i_atime.tv_sec = le32_to_cpu(di->i_atime);
inode->i_mtime.tv_sec = le32_to_cpu(di->i_mtime);
inode->i_ctime.tv_sec = le32_to_cpu(di->i_ctime);
inode->i_atime.tv_nsec = 0;
inode->i_mtime.tv_nsec = 0;
inode->i_ctime.tv_nsec = 0;
brelse(bh);
unlock_new_inode(inode);
return inode;
error:
iget_failed(inode);
return ERR_PTR(-EIO);
}
static int bfs_write_inode(struct inode *inode, struct writeback_control *wbc)
{
struct bfs_sb_info *info = BFS_SB(inode->i_sb);
unsigned int ino = (u16)inode->i_ino;
unsigned long i_sblock;
struct bfs_inode *di;
struct buffer_head *bh;
int block, off;
int err = 0;
dprintf("ino=%08x\n", ino);
if ((ino < BFS_ROOT_INO) || (ino > BFS_SB(inode->i_sb)->si_lasti)) {
printf("Bad inode number %s:%08x\n", inode->i_sb->s_id, ino);
return -EIO;
}
mutex_lock(&info->bfs_lock);
block = (ino - BFS_ROOT_INO) / BFS_INODES_PER_BLOCK + 1;
bh = sb_bread(inode->i_sb, block);
if (!bh) {
printf("Unable to read inode %s:%08x\n",
inode->i_sb->s_id, ino);
mutex_unlock(&info->bfs_lock);
return -EIO;
}
off = (ino - BFS_ROOT_INO) % BFS_INODES_PER_BLOCK;
di = (struct bfs_inode *)bh->b_data + off;
if (ino == BFS_ROOT_INO)
di->i_vtype = cpu_to_le32(BFS_VDIR);
else
di->i_vtype = cpu_to_le32(BFS_VREG);
di->i_ino = cpu_to_le16(ino);
di->i_mode = cpu_to_le32(inode->i_mode);
di->i_uid = cpu_to_le32(inode->i_uid);
di->i_gid = cpu_to_le32(inode->i_gid);
di->i_nlink = cpu_to_le32(inode->i_nlink);
di->i_atime = cpu_to_le32(inode->i_atime.tv_sec);
di->i_mtime = cpu_to_le32(inode->i_mtime.tv_sec);
di->i_ctime = cpu_to_le32(inode->i_ctime.tv_sec);
i_sblock = BFS_I(inode)->i_sblock;
di->i_sblock = cpu_to_le32(i_sblock);
di->i_eblock = cpu_to_le32(BFS_I(inode)->i_eblock);
di->i_eoffset = cpu_to_le32(i_sblock * BFS_BSIZE + inode->i_size - 1);
mark_buffer_dirty(bh);
if (wbc->sync_mode == WB_SYNC_ALL) {
sync_dirty_buffer(bh);
if (buffer_req(bh) && !buffer_uptodate(bh))
err = -EIO;
}
brelse(bh);
mutex_unlock(&info->bfs_lock);
return err;
}
static void bfs_delete_inode(struct inode *inode)
{
unsigned long ino = inode->i_ino;
struct bfs_inode *di;
struct buffer_head *bh;
int block, off;
struct super_block *s = inode->i_sb;
struct bfs_sb_info *info = BFS_SB(s);
struct bfs_inode_info *bi = BFS_I(inode);
dprintf("ino=%08lx\n", ino);
truncate_inode_pages(&inode->i_data, 0);
if ((ino < BFS_ROOT_INO) || (ino > info->si_lasti)) {
printf("invalid ino=%08lx\n", ino);
return;
}
inode->i_size = 0;
inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME_SEC;
mutex_lock(&info->bfs_lock);
mark_inode_dirty(inode);
block = (ino - BFS_ROOT_INO) / BFS_INODES_PER_BLOCK + 1;
bh = sb_bread(s, block);
if (!bh) {
printf("Unable to read inode %s:%08lx\n",
inode->i_sb->s_id, ino);
mutex_unlock(&info->bfs_lock);
return;
}
off = (ino - BFS_ROOT_INO) % BFS_INODES_PER_BLOCK;
di = (struct bfs_inode *)bh->b_data + off;
memset((void *)di, 0, sizeof(struct bfs_inode));
mark_buffer_dirty(bh);
brelse(bh);
if (bi->i_dsk_ino) {
if (bi->i_sblock)
info->si_freeb += bi->i_eblock + 1 - bi->i_sblock;
info->si_freei++;
clear_bit(ino, info->si_imap);
dump_imap("delete_inode", s);
}
/*
* If this was the last file, make the previous block
* "last block of the last file" even if there is no
* real file there, saves us 1 gap.
*/
if (info->si_lf_eblk == bi->i_eblock) {
info->si_lf_eblk = bi->i_sblock - 1;
mark_buffer_dirty(info->si_sbh);
}
mutex_unlock(&info->bfs_lock);
clear_inode(inode);
}
static int bfs_sync_fs(struct super_block *sb, int wait)
{
struct bfs_sb_info *info = BFS_SB(sb);
mutex_lock(&info->bfs_lock);
mark_buffer_dirty(info->si_sbh);
sb->s_dirt = 0;
mutex_unlock(&info->bfs_lock);
return 0;
}
static void bfs_write_super(struct super_block *sb)
{
if (!(sb->s_flags & MS_RDONLY))
bfs_sync_fs(sb, 1);
else
sb->s_dirt = 0;
}
static void bfs_put_super(struct super_block *s)
{
struct bfs_sb_info *info = BFS_SB(s);
if (!info)
return;
lock_kernel();
if (s->s_dirt)
bfs_write_super(s);
brelse(info->si_sbh);
mutex_destroy(&info->bfs_lock);
kfree(info->si_imap);
kfree(info);
s->s_fs_info = NULL;
unlock_kernel();
}
static int bfs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
struct super_block *s = dentry->d_sb;
struct bfs_sb_info *info = BFS_SB(s);
u64 id = huge_encode_dev(s->s_bdev->bd_dev);
buf->f_type = BFS_MAGIC;
buf->f_bsize = s->s_blocksize;
buf->f_blocks = info->si_blocks;
buf->f_bfree = buf->f_bavail = info->si_freeb;
buf->f_files = info->si_lasti + 1 - BFS_ROOT_INO;
buf->f_ffree = info->si_freei;
buf->f_fsid.val[0] = (u32)id;
buf->f_fsid.val[1] = (u32)(id >> 32);
buf->f_namelen = BFS_NAMELEN;
return 0;
}
static struct kmem_cache *bfs_inode_cachep;
static struct inode *bfs_alloc_inode(struct super_block *sb)
{
struct bfs_inode_info *bi;
bi = kmem_cache_alloc(bfs_inode_cachep, GFP_KERNEL);
if (!bi)
return NULL;
return &bi->vfs_inode;
}
static void bfs_destroy_inode(struct inode *inode)
{
kmem_cache_free(bfs_inode_cachep, BFS_I(inode));
}
static void init_once(void *foo)
{
struct bfs_inode_info *bi = foo;
inode_init_once(&bi->vfs_inode);
}
static int init_inodecache(void)
{
bfs_inode_cachep = kmem_cache_create("bfs_inode_cache",
sizeof(struct bfs_inode_info),
0, (SLAB_RECLAIM_ACCOUNT|
SLAB_MEM_SPREAD),
init_once);
if (bfs_inode_cachep == NULL)
return -ENOMEM;
return 0;
}
static void destroy_inodecache(void)
{
kmem_cache_destroy(bfs_inode_cachep);
}
static const struct super_operations bfs_sops = {
.alloc_inode = bfs_alloc_inode,
.destroy_inode = bfs_destroy_inode,
.write_inode = bfs_write_inode,
.delete_inode = bfs_delete_inode,
.put_super = bfs_put_super,
.write_super = bfs_write_super,
.sync_fs = bfs_sync_fs,
.statfs = bfs_statfs,
};
void dump_imap(const char *prefix, struct super_block *s)
{
#ifdef DEBUG
int i;
char *tmpbuf = (char *)get_zeroed_page(GFP_KERNEL);
if (!tmpbuf)
return;
for (i = BFS_SB(s)->si_lasti; i >= 0; i--) {
if (i > PAGE_SIZE - 100) break;
if (test_bit(i, BFS_SB(s)->si_imap))
strcat(tmpbuf, "1");
else
strcat(tmpbuf, "0");
}
printf("BFS-fs: %s: lasti=%08lx <%s>\n",
prefix, BFS_SB(s)->si_lasti, tmpbuf);
free_page((unsigned long)tmpbuf);
#endif
}
static int bfs_fill_super(struct super_block *s, void *data, int silent)
{
struct buffer_head *bh;
struct bfs_super_block *bfs_sb;
struct inode *inode;
unsigned i, imap_len;
struct bfs_sb_info *info;
int ret = -EINVAL;
unsigned long i_sblock, i_eblock, i_eoff, s_size;
info = kzalloc(sizeof(*info), GFP_KERNEL);
if (!info)
return -ENOMEM;
mutex_init(&info->bfs_lock);
s->s_fs_info = info;
sb_set_blocksize(s, BFS_BSIZE);
info->si_sbh = sb_bread(s, 0);
if (!info->si_sbh)
goto out;
bfs_sb = (struct bfs_super_block *)info->si_sbh->b_data;
if (le32_to_cpu(bfs_sb->s_magic) != BFS_MAGIC) {
if (!silent)
printf("No BFS filesystem on %s (magic=%08x)\n",
s->s_id, le32_to_cpu(bfs_sb->s_magic));
goto out1;
}
if (BFS_UNCLEAN(bfs_sb, s) && !silent)
printf("%s is unclean, continuing\n", s->s_id);
s->s_magic = BFS_MAGIC;
if (le32_to_cpu(bfs_sb->s_start) > le32_to_cpu(bfs_sb->s_end)) {
printf("Superblock is corrupted\n");
goto out1;
}
info->si_lasti = (le32_to_cpu(bfs_sb->s_start) - BFS_BSIZE) /
sizeof(struct bfs_inode)
+ BFS_ROOT_INO - 1;
imap_len = (info->si_lasti / 8) + 1;
info->si_imap = kzalloc(imap_len, GFP_KERNEL);
if (!info->si_imap)
goto out1;
for (i = 0; i < BFS_ROOT_INO; i++)
set_bit(i, info->si_imap);
s->s_op = &bfs_sops;
inode = bfs_iget(s, BFS_ROOT_INO);
if (IS_ERR(inode)) {
ret = PTR_ERR(inode);
goto out2;
}
s->s_root = d_alloc_root(inode);
if (!s->s_root) {
iput(inode);
ret = -ENOMEM;
goto out2;
}
info->si_blocks = (le32_to_cpu(bfs_sb->s_end) + 1) >> BFS_BSIZE_BITS;
info->si_freeb = (le32_to_cpu(bfs_sb->s_end) + 1
- le32_to_cpu(bfs_sb->s_start)) >> BFS_BSIZE_BITS;
info->si_freei = 0;
info->si_lf_eblk = 0;
/* can we read the last block? */
bh = sb_bread(s, info->si_blocks - 1);
if (!bh) {
printf("Last block not available: %lu\n", info->si_blocks - 1);
ret = -EIO;
goto out3;
}
brelse(bh);
bh = NULL;
for (i = BFS_ROOT_INO; i <= info->si_lasti; i++) {
struct bfs_inode *di;
int block = (i - BFS_ROOT_INO) / BFS_INODES_PER_BLOCK + 1;
int off = (i - BFS_ROOT_INO) % BFS_INODES_PER_BLOCK;
unsigned long eblock;
if (!off) {
brelse(bh);
bh = sb_bread(s, block);
}
if (!bh)
continue;
di = (struct bfs_inode *)bh->b_data + off;
/* test if filesystem is not corrupted */
i_eoff = le32_to_cpu(di->i_eoffset);
i_sblock = le32_to_cpu(di->i_sblock);
i_eblock = le32_to_cpu(di->i_eblock);
s_size = le32_to_cpu(bfs_sb->s_end);
if (i_sblock > info->si_blocks ||
i_eblock > info->si_blocks ||
i_sblock > i_eblock ||
i_eoff > s_size ||
i_sblock * BFS_BSIZE > i_eoff) {
printf("Inode 0x%08x corrupted\n", i);
brelse(bh);
ret = -EIO;
goto out3;
}
if (!di->i_ino) {
info->si_freei++;
continue;
}
set_bit(i, info->si_imap);
info->si_freeb -= BFS_FILEBLOCKS(di);
eblock = le32_to_cpu(di->i_eblock);
if (eblock > info->si_lf_eblk)
info->si_lf_eblk = eblock;
}
brelse(bh);
if (!(s->s_flags & MS_RDONLY)) {
mark_buffer_dirty(info->si_sbh);
s->s_dirt = 1;
}
dump_imap("read_super", s);
return 0;
out3:
dput(s->s_root);
s->s_root = NULL;
out2:
kfree(info->si_imap);
out1:
brelse(info->si_sbh);
out:
mutex_destroy(&info->bfs_lock);
kfree(info);
s->s_fs_info = NULL;
return ret;
}
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 11:02:57 +02:00
static int bfs_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *data, struct vfsmount *mnt)
{
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 11:02:57 +02:00
return get_sb_bdev(fs_type, flags, dev_name, data, bfs_fill_super, mnt);
}
static struct file_system_type bfs_fs_type = {
.owner = THIS_MODULE,
.name = "bfs",
.get_sb = bfs_get_sb,
.kill_sb = kill_block_super,
.fs_flags = FS_REQUIRES_DEV,
};
static int __init init_bfs_fs(void)
{
int err = init_inodecache();
if (err)
goto out1;
err = register_filesystem(&bfs_fs_type);
if (err)
goto out;
return 0;
out:
destroy_inodecache();
out1:
return err;
}
static void __exit exit_bfs_fs(void)
{
unregister_filesystem(&bfs_fs_type);
destroy_inodecache();
}
module_init(init_bfs_fs)
module_exit(exit_bfs_fs)