linux/fs/affs/super.c
Linus Torvalds 437589a74b Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace
Pull user namespace changes from Eric Biederman:
 "This is a mostly modest set of changes to enable basic user namespace
  support.  This allows the code to code to compile with user namespaces
  enabled and removes the assumption there is only the initial user
  namespace.  Everything is converted except for the most complex of the
  filesystems: autofs4, 9p, afs, ceph, cifs, coda, fuse, gfs2, ncpfs,
  nfs, ocfs2 and xfs as those patches need a bit more review.

  The strategy is to push kuid_t and kgid_t values are far down into
  subsystems and filesystems as reasonable.  Leaving the make_kuid and
  from_kuid operations to happen at the edge of userspace, as the values
  come off the disk, and as the values come in from the network.
  Letting compile type incompatible compile errors (present when user
  namespaces are enabled) guide me to find the issues.

  The most tricky areas have been the places where we had an implicit
  union of uid and gid values and were storing them in an unsigned int.
  Those places were converted into explicit unions.  I made certain to
  handle those places with simple trivial patches.

  Out of that work I discovered we have generic interfaces for storing
  quota by projid.  I had never heard of the project identifiers before.
  Adding full user namespace support for project identifiers accounts
  for most of the code size growth in my git tree.

  Ultimately there will be work to relax privlige checks from
  "capable(FOO)" to "ns_capable(user_ns, FOO)" where it is safe allowing
  root in a user names to do those things that today we only forbid to
  non-root users because it will confuse suid root applications.

  While I was pushing kuid_t and kgid_t changes deep into the audit code
  I made a few other cleanups.  I capitalized on the fact we process
  netlink messages in the context of the message sender.  I removed
  usage of NETLINK_CRED, and started directly using current->tty.

  Some of these patches have also made it into maintainer trees, with no
  problems from identical code from different trees showing up in
  linux-next.

  After reading through all of this code I feel like I might be able to
  win a game of kernel trivial pursuit."

Fix up some fairly trivial conflicts in netfilter uid/git logging code.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace: (107 commits)
  userns: Convert the ufs filesystem to use kuid/kgid where appropriate
  userns: Convert the udf filesystem to use kuid/kgid where appropriate
  userns: Convert ubifs to use kuid/kgid
  userns: Convert squashfs to use kuid/kgid where appropriate
  userns: Convert reiserfs to use kuid and kgid where appropriate
  userns: Convert jfs to use kuid/kgid where appropriate
  userns: Convert jffs2 to use kuid and kgid where appropriate
  userns: Convert hpfs to use kuid and kgid where appropriate
  userns: Convert btrfs to use kuid/kgid where appropriate
  userns: Convert bfs to use kuid/kgid where appropriate
  userns: Convert affs to use kuid/kgid wherwe appropriate
  userns: On alpha modify linux_to_osf_stat to use convert from kuids and kgids
  userns: On ia64 deal with current_uid and current_gid being kuid and kgid
  userns: On ppc convert current_uid from a kuid before printing.
  userns: Convert s390 getting uid and gid system calls to use kuid and kgid
  userns: Convert s390 hypfs to use kuid and kgid where appropriate
  userns: Convert binder ipc to use kuids
  userns: Teach security_path_chown to take kuids and kgids
  userns: Add user namespace support to IMA
  userns: Convert EVM to deal with kuids and kgids in it's hmac computation
  ...
2012-10-02 11:11:09 -07:00

647 lines
15 KiB
C

/*
* linux/fs/affs/inode.c
*
* (c) 1996 Hans-Joachim Widmaier - Rewritten
*
* (C) 1993 Ray Burr - Modified for Amiga FFS filesystem.
*
* (C) 1992 Eric Youngdale Modified for ISO 9660 filesystem.
*
* (C) 1991 Linus Torvalds - minix filesystem
*/
#include <linux/module.h>
#include <linux/init.h>
#include <linux/statfs.h>
#include <linux/parser.h>
#include <linux/magic.h>
#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/writeback.h>
#include "affs.h"
extern struct timezone sys_tz;
static int affs_statfs(struct dentry *dentry, struct kstatfs *buf);
static int affs_remount (struct super_block *sb, int *flags, char *data);
static void
affs_commit_super(struct super_block *sb, int wait)
{
struct affs_sb_info *sbi = AFFS_SB(sb);
struct buffer_head *bh = sbi->s_root_bh;
struct affs_root_tail *tail = AFFS_ROOT_TAIL(sb, bh);
lock_buffer(bh);
secs_to_datestamp(get_seconds(), &tail->disk_change);
affs_fix_checksum(sb, bh);
unlock_buffer(bh);
mark_buffer_dirty(bh);
if (wait)
sync_dirty_buffer(bh);
}
static void
affs_put_super(struct super_block *sb)
{
struct affs_sb_info *sbi = AFFS_SB(sb);
pr_debug("AFFS: put_super()\n");
cancel_delayed_work_sync(&sbi->sb_work);
kfree(sbi->s_prefix);
affs_free_bitmap(sb);
affs_brelse(sbi->s_root_bh);
kfree(sbi);
sb->s_fs_info = NULL;
}
static int
affs_sync_fs(struct super_block *sb, int wait)
{
affs_commit_super(sb, wait);
return 0;
}
static void flush_superblock(struct work_struct *work)
{
struct affs_sb_info *sbi;
struct super_block *sb;
sbi = container_of(work, struct affs_sb_info, sb_work.work);
sb = sbi->sb;
spin_lock(&sbi->work_lock);
sbi->work_queued = 0;
spin_unlock(&sbi->work_lock);
affs_commit_super(sb, 1);
}
void affs_mark_sb_dirty(struct super_block *sb)
{
struct affs_sb_info *sbi = AFFS_SB(sb);
unsigned long delay;
if (sb->s_flags & MS_RDONLY)
return;
spin_lock(&sbi->work_lock);
if (!sbi->work_queued) {
delay = msecs_to_jiffies(dirty_writeback_interval * 10);
queue_delayed_work(system_long_wq, &sbi->sb_work, delay);
sbi->work_queued = 1;
}
spin_unlock(&sbi->work_lock);
}
static struct kmem_cache * affs_inode_cachep;
static struct inode *affs_alloc_inode(struct super_block *sb)
{
struct affs_inode_info *i;
i = kmem_cache_alloc(affs_inode_cachep, GFP_KERNEL);
if (!i)
return NULL;
i->vfs_inode.i_version = 1;
i->i_lc = NULL;
i->i_ext_bh = NULL;
i->i_pa_cnt = 0;
return &i->vfs_inode;
}
static void affs_i_callback(struct rcu_head *head)
{
struct inode *inode = container_of(head, struct inode, i_rcu);
kmem_cache_free(affs_inode_cachep, AFFS_I(inode));
}
static void affs_destroy_inode(struct inode *inode)
{
call_rcu(&inode->i_rcu, affs_i_callback);
}
static void init_once(void *foo)
{
struct affs_inode_info *ei = (struct affs_inode_info *) foo;
sema_init(&ei->i_link_lock, 1);
sema_init(&ei->i_ext_lock, 1);
inode_init_once(&ei->vfs_inode);
}
static int init_inodecache(void)
{
affs_inode_cachep = kmem_cache_create("affs_inode_cache",
sizeof(struct affs_inode_info),
0, (SLAB_RECLAIM_ACCOUNT|
SLAB_MEM_SPREAD),
init_once);
if (affs_inode_cachep == NULL)
return -ENOMEM;
return 0;
}
static void destroy_inodecache(void)
{
kmem_cache_destroy(affs_inode_cachep);
}
static const struct super_operations affs_sops = {
.alloc_inode = affs_alloc_inode,
.destroy_inode = affs_destroy_inode,
.write_inode = affs_write_inode,
.evict_inode = affs_evict_inode,
.put_super = affs_put_super,
.sync_fs = affs_sync_fs,
.statfs = affs_statfs,
.remount_fs = affs_remount,
.show_options = generic_show_options,
};
enum {
Opt_bs, Opt_mode, Opt_mufs, Opt_prefix, Opt_protect,
Opt_reserved, Opt_root, Opt_setgid, Opt_setuid,
Opt_verbose, Opt_volume, Opt_ignore, Opt_err,
};
static const match_table_t tokens = {
{Opt_bs, "bs=%u"},
{Opt_mode, "mode=%o"},
{Opt_mufs, "mufs"},
{Opt_prefix, "prefix=%s"},
{Opt_protect, "protect"},
{Opt_reserved, "reserved=%u"},
{Opt_root, "root=%u"},
{Opt_setgid, "setgid=%u"},
{Opt_setuid, "setuid=%u"},
{Opt_verbose, "verbose"},
{Opt_volume, "volume=%s"},
{Opt_ignore, "grpquota"},
{Opt_ignore, "noquota"},
{Opt_ignore, "quota"},
{Opt_ignore, "usrquota"},
{Opt_err, NULL},
};
static int
parse_options(char *options, kuid_t *uid, kgid_t *gid, int *mode, int *reserved, s32 *root,
int *blocksize, char **prefix, char *volume, unsigned long *mount_opts)
{
char *p;
substring_t args[MAX_OPT_ARGS];
/* Fill in defaults */
*uid = current_uid();
*gid = current_gid();
*reserved = 2;
*root = -1;
*blocksize = -1;
volume[0] = ':';
volume[1] = 0;
*mount_opts = 0;
if (!options)
return 1;
while ((p = strsep(&options, ",")) != NULL) {
int token, n, option;
if (!*p)
continue;
token = match_token(p, tokens, args);
switch (token) {
case Opt_bs:
if (match_int(&args[0], &n))
return 0;
if (n != 512 && n != 1024 && n != 2048
&& n != 4096) {
printk ("AFFS: Invalid blocksize (512, 1024, 2048, 4096 allowed)\n");
return 0;
}
*blocksize = n;
break;
case Opt_mode:
if (match_octal(&args[0], &option))
return 0;
*mode = option & 0777;
*mount_opts |= SF_SETMODE;
break;
case Opt_mufs:
*mount_opts |= SF_MUFS;
break;
case Opt_prefix:
*prefix = match_strdup(&args[0]);
if (!*prefix)
return 0;
*mount_opts |= SF_PREFIX;
break;
case Opt_protect:
*mount_opts |= SF_IMMUTABLE;
break;
case Opt_reserved:
if (match_int(&args[0], reserved))
return 0;
break;
case Opt_root:
if (match_int(&args[0], root))
return 0;
break;
case Opt_setgid:
if (match_int(&args[0], &option))
return 0;
*gid = make_kgid(current_user_ns(), option);
if (!gid_valid(*gid))
return 0;
*mount_opts |= SF_SETGID;
break;
case Opt_setuid:
if (match_int(&args[0], &option))
return 0;
*uid = make_kuid(current_user_ns(), option);
if (!uid_valid(*uid))
return 0;
*mount_opts |= SF_SETUID;
break;
case Opt_verbose:
*mount_opts |= SF_VERBOSE;
break;
case Opt_volume: {
char *vol = match_strdup(&args[0]);
if (!vol)
return 0;
strlcpy(volume, vol, 32);
kfree(vol);
break;
}
case Opt_ignore:
/* Silently ignore the quota options */
break;
default:
printk("AFFS: Unrecognized mount option \"%s\" "
"or missing value\n", p);
return 0;
}
}
return 1;
}
/* This function definitely needs to be split up. Some fine day I'll
* hopefully have the guts to do so. Until then: sorry for the mess.
*/
static int affs_fill_super(struct super_block *sb, void *data, int silent)
{
struct affs_sb_info *sbi;
struct buffer_head *root_bh = NULL;
struct buffer_head *boot_bh;
struct inode *root_inode = NULL;
s32 root_block;
int size, blocksize;
u32 chksum;
int num_bm;
int i, j;
s32 key;
kuid_t uid;
kgid_t gid;
int reserved;
unsigned long mount_flags;
int tmp_flags; /* fix remount prototype... */
u8 sig[4];
int ret = -EINVAL;
save_mount_options(sb, data);
pr_debug("AFFS: read_super(%s)\n",data ? (const char *)data : "no options");
sb->s_magic = AFFS_SUPER_MAGIC;
sb->s_op = &affs_sops;
sb->s_flags |= MS_NODIRATIME;
sbi = kzalloc(sizeof(struct affs_sb_info), GFP_KERNEL);
if (!sbi)
return -ENOMEM;
sb->s_fs_info = sbi;
sbi->sb = sb;
mutex_init(&sbi->s_bmlock);
spin_lock_init(&sbi->symlink_lock);
spin_lock_init(&sbi->work_lock);
INIT_DELAYED_WORK(&sbi->sb_work, flush_superblock);
if (!parse_options(data,&uid,&gid,&i,&reserved,&root_block,
&blocksize,&sbi->s_prefix,
sbi->s_volume, &mount_flags)) {
printk(KERN_ERR "AFFS: Error parsing options\n");
kfree(sbi->s_prefix);
kfree(sbi);
return -EINVAL;
}
/* N.B. after this point s_prefix must be released */
sbi->s_flags = mount_flags;
sbi->s_mode = i;
sbi->s_uid = uid;
sbi->s_gid = gid;
sbi->s_reserved= reserved;
/* Get the size of the device in 512-byte blocks.
* If we later see that the partition uses bigger
* blocks, we will have to change it.
*/
size = sb->s_bdev->bd_inode->i_size >> 9;
pr_debug("AFFS: initial blocksize=%d, #blocks=%d\n", 512, size);
affs_set_blocksize(sb, PAGE_SIZE);
/* Try to find root block. Its location depends on the block size. */
i = 512;
j = 4096;
if (blocksize > 0) {
i = j = blocksize;
size = size / (blocksize / 512);
}
for (blocksize = i, key = 0; blocksize <= j; blocksize <<= 1, size >>= 1) {
sbi->s_root_block = root_block;
if (root_block < 0)
sbi->s_root_block = (reserved + size - 1) / 2;
pr_debug("AFFS: setting blocksize to %d\n", blocksize);
affs_set_blocksize(sb, blocksize);
sbi->s_partition_size = size;
/* The root block location that was calculated above is not
* correct if the partition size is an odd number of 512-
* byte blocks, which will be rounded down to a number of
* 1024-byte blocks, and if there were an even number of
* reserved blocks. Ideally, all partition checkers should
* report the real number of blocks of the real blocksize,
* but since this just cannot be done, we have to try to
* find the root block anyways. In the above case, it is one
* block behind the calculated one. So we check this one, too.
*/
for (num_bm = 0; num_bm < 2; num_bm++) {
pr_debug("AFFS: Dev %s, trying root=%u, bs=%d, "
"size=%d, reserved=%d\n",
sb->s_id,
sbi->s_root_block + num_bm,
blocksize, size, reserved);
root_bh = affs_bread(sb, sbi->s_root_block + num_bm);
if (!root_bh)
continue;
if (!affs_checksum_block(sb, root_bh) &&
be32_to_cpu(AFFS_ROOT_HEAD(root_bh)->ptype) == T_SHORT &&
be32_to_cpu(AFFS_ROOT_TAIL(sb, root_bh)->stype) == ST_ROOT) {
sbi->s_hashsize = blocksize / 4 - 56;
sbi->s_root_block += num_bm;
key = 1;
goto got_root;
}
affs_brelse(root_bh);
root_bh = NULL;
}
}
if (!silent)
printk(KERN_ERR "AFFS: No valid root block on device %s\n",
sb->s_id);
goto out_error;
/* N.B. after this point bh must be released */
got_root:
root_block = sbi->s_root_block;
/* Find out which kind of FS we have */
boot_bh = sb_bread(sb, 0);
if (!boot_bh) {
printk(KERN_ERR "AFFS: Cannot read boot block\n");
goto out_error;
}
memcpy(sig, boot_bh->b_data, 4);
brelse(boot_bh);
chksum = be32_to_cpu(*(__be32 *)sig);
/* Dircache filesystems are compatible with non-dircache ones
* when reading. As long as they aren't supported, writing is
* not recommended.
*/
if ((chksum == FS_DCFFS || chksum == MUFS_DCFFS || chksum == FS_DCOFS
|| chksum == MUFS_DCOFS) && !(sb->s_flags & MS_RDONLY)) {
printk(KERN_NOTICE "AFFS: Dircache FS - mounting %s read only\n",
sb->s_id);
sb->s_flags |= MS_RDONLY;
}
switch (chksum) {
case MUFS_FS:
case MUFS_INTLFFS:
case MUFS_DCFFS:
sbi->s_flags |= SF_MUFS;
/* fall thru */
case FS_INTLFFS:
case FS_DCFFS:
sbi->s_flags |= SF_INTL;
break;
case MUFS_FFS:
sbi->s_flags |= SF_MUFS;
break;
case FS_FFS:
break;
case MUFS_OFS:
sbi->s_flags |= SF_MUFS;
/* fall thru */
case FS_OFS:
sbi->s_flags |= SF_OFS;
sb->s_flags |= MS_NOEXEC;
break;
case MUFS_DCOFS:
case MUFS_INTLOFS:
sbi->s_flags |= SF_MUFS;
case FS_DCOFS:
case FS_INTLOFS:
sbi->s_flags |= SF_INTL | SF_OFS;
sb->s_flags |= MS_NOEXEC;
break;
default:
printk(KERN_ERR "AFFS: Unknown filesystem on device %s: %08X\n",
sb->s_id, chksum);
goto out_error;
}
if (mount_flags & SF_VERBOSE) {
u8 len = AFFS_ROOT_TAIL(sb, root_bh)->disk_name[0];
printk(KERN_NOTICE "AFFS: Mounting volume \"%.*s\": Type=%.3s\\%c, Blocksize=%d\n",
len > 31 ? 31 : len,
AFFS_ROOT_TAIL(sb, root_bh)->disk_name + 1,
sig, sig[3] + '0', blocksize);
}
sb->s_flags |= MS_NODEV | MS_NOSUID;
sbi->s_data_blksize = sb->s_blocksize;
if (sbi->s_flags & SF_OFS)
sbi->s_data_blksize -= 24;
/* Keep super block in cache */
sbi->s_root_bh = root_bh;
/* N.B. after this point s_root_bh must be released */
tmp_flags = sb->s_flags;
if (affs_init_bitmap(sb, &tmp_flags))
goto out_error;
sb->s_flags = tmp_flags;
/* set up enough so that it can read an inode */
root_inode = affs_iget(sb, root_block);
if (IS_ERR(root_inode)) {
ret = PTR_ERR(root_inode);
goto out_error;
}
if (AFFS_SB(sb)->s_flags & SF_INTL)
sb->s_d_op = &affs_intl_dentry_operations;
else
sb->s_d_op = &affs_dentry_operations;
sb->s_root = d_make_root(root_inode);
if (!sb->s_root) {
printk(KERN_ERR "AFFS: Get root inode failed\n");
goto out_error;
}
pr_debug("AFFS: s_flags=%lX\n",sb->s_flags);
return 0;
/*
* Begin the cascaded cleanup ...
*/
out_error:
kfree(sbi->s_bitmap);
affs_brelse(root_bh);
kfree(sbi->s_prefix);
kfree(sbi);
sb->s_fs_info = NULL;
return ret;
}
static int
affs_remount(struct super_block *sb, int *flags, char *data)
{
struct affs_sb_info *sbi = AFFS_SB(sb);
int blocksize;
kuid_t uid;
kgid_t gid;
int mode;
int reserved;
int root_block;
unsigned long mount_flags;
int res = 0;
char *new_opts = kstrdup(data, GFP_KERNEL);
char volume[32];
char *prefix = NULL;
pr_debug("AFFS: remount(flags=0x%x,opts=\"%s\")\n",*flags,data);
*flags |= MS_NODIRATIME;
memcpy(volume, sbi->s_volume, 32);
if (!parse_options(data, &uid, &gid, &mode, &reserved, &root_block,
&blocksize, &prefix, volume,
&mount_flags)) {
kfree(prefix);
kfree(new_opts);
return -EINVAL;
}
flush_delayed_work(&sbi->sb_work);
replace_mount_options(sb, new_opts);
sbi->s_flags = mount_flags;
sbi->s_mode = mode;
sbi->s_uid = uid;
sbi->s_gid = gid;
/* protect against readers */
spin_lock(&sbi->symlink_lock);
if (prefix) {
kfree(sbi->s_prefix);
sbi->s_prefix = prefix;
}
memcpy(sbi->s_volume, volume, 32);
spin_unlock(&sbi->symlink_lock);
if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY))
return 0;
if (*flags & MS_RDONLY)
affs_free_bitmap(sb);
else
res = affs_init_bitmap(sb, flags);
return res;
}
static int
affs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
struct super_block *sb = dentry->d_sb;
int free;
u64 id = huge_encode_dev(sb->s_bdev->bd_dev);
pr_debug("AFFS: statfs() partsize=%d, reserved=%d\n",AFFS_SB(sb)->s_partition_size,
AFFS_SB(sb)->s_reserved);
free = affs_count_free_blocks(sb);
buf->f_type = AFFS_SUPER_MAGIC;
buf->f_bsize = sb->s_blocksize;
buf->f_blocks = AFFS_SB(sb)->s_partition_size - AFFS_SB(sb)->s_reserved;
buf->f_bfree = free;
buf->f_bavail = free;
buf->f_fsid.val[0] = (u32)id;
buf->f_fsid.val[1] = (u32)(id >> 32);
buf->f_namelen = 30;
return 0;
}
static struct dentry *affs_mount(struct file_system_type *fs_type,
int flags, const char *dev_name, void *data)
{
return mount_bdev(fs_type, flags, dev_name, data, affs_fill_super);
}
static struct file_system_type affs_fs_type = {
.owner = THIS_MODULE,
.name = "affs",
.mount = affs_mount,
.kill_sb = kill_block_super,
.fs_flags = FS_REQUIRES_DEV,
};
static int __init init_affs_fs(void)
{
int err = init_inodecache();
if (err)
goto out1;
err = register_filesystem(&affs_fs_type);
if (err)
goto out;
return 0;
out:
destroy_inodecache();
out1:
return err;
}
static void __exit exit_affs_fs(void)
{
unregister_filesystem(&affs_fs_type);
destroy_inodecache();
}
MODULE_DESCRIPTION("Amiga filesystem support for Linux");
MODULE_LICENSE("GPL");
module_init(init_affs_fs)
module_exit(exit_affs_fs)