mirror of
https://mirrors.bfsu.edu.cn/git/linux.git
synced 2024-11-29 15:14:18 +08:00
cac2f8b8d8
The current way of setting and getting posix acls through the generic xattr interface is error prone and type unsafe. The vfs needs to interpret and fixup posix acls before storing or reporting it to userspace. Various hacks exist to make this work. The code is hard to understand and difficult to maintain in it's current form. Instead of making this work by hacking posix acls through xattr handlers we are building a dedicated posix acl api around the get and set inode operations. This removes a lot of hackiness and makes the codepaths easier to maintain. A lot of background can be found in [1]. The current inode operation for getting posix acls takes an inode argument but various filesystems (e.g., 9p, cifs, overlayfs) need access to the dentry. In contrast to the ->set_acl() inode operation we cannot simply extend ->get_acl() to take a dentry argument. The ->get_acl() inode operation is called from: acl_permission_check() -> check_acl() -> get_acl() which is part of generic_permission() which in turn is part of inode_permission(). Both generic_permission() and inode_permission() are called in the ->permission() handler of various filesystems (e.g., overlayfs). So simply passing a dentry argument to ->get_acl() would amount to also having to pass a dentry argument to ->permission(). We should avoid this unnecessary change. So instead of extending the existing inode operation rename it from ->get_acl() to ->get_inode_acl() and add a ->get_acl() method later that passes a dentry argument and which filesystems that need access to the dentry can implement instead of ->get_inode_acl(). Filesystems like cifs which allow setting and getting posix acls but not using them for permission checking during lookup can simply not implement ->get_inode_acl(). This is intended to be a non-functional change. Link: https://lore.kernel.org/all/20220801145520.1532837-1-brauner@kernel.org [1] Suggested-by/Inspired-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
412 lines
9.8 KiB
C
412 lines
9.8 KiB
C
// SPDX-License-Identifier: GPL-2.0
|
|
#include <linux/capability.h>
|
|
#include <linux/fs.h>
|
|
#include <linux/posix_acl.h>
|
|
#include "reiserfs.h"
|
|
#include <linux/errno.h>
|
|
#include <linux/pagemap.h>
|
|
#include <linux/xattr.h>
|
|
#include <linux/slab.h>
|
|
#include <linux/posix_acl_xattr.h>
|
|
#include "xattr.h"
|
|
#include "acl.h"
|
|
#include <linux/uaccess.h>
|
|
|
|
static int __reiserfs_set_acl(struct reiserfs_transaction_handle *th,
|
|
struct inode *inode, int type,
|
|
struct posix_acl *acl);
|
|
|
|
|
|
int
|
|
reiserfs_set_acl(struct user_namespace *mnt_userns, struct dentry *dentry,
|
|
struct posix_acl *acl, int type)
|
|
{
|
|
int error, error2;
|
|
struct reiserfs_transaction_handle th;
|
|
size_t jcreate_blocks;
|
|
int size = acl ? posix_acl_xattr_size(acl->a_count) : 0;
|
|
int update_mode = 0;
|
|
struct inode *inode = d_inode(dentry);
|
|
umode_t mode = inode->i_mode;
|
|
|
|
/*
|
|
* Pessimism: We can't assume that anything from the xattr root up
|
|
* has been created.
|
|
*/
|
|
|
|
jcreate_blocks = reiserfs_xattr_jcreate_nblocks(inode) +
|
|
reiserfs_xattr_nblocks(inode, size) * 2;
|
|
|
|
reiserfs_write_lock(inode->i_sb);
|
|
error = journal_begin(&th, inode->i_sb, jcreate_blocks);
|
|
reiserfs_write_unlock(inode->i_sb);
|
|
if (error == 0) {
|
|
if (type == ACL_TYPE_ACCESS && acl) {
|
|
error = posix_acl_update_mode(&init_user_ns, inode,
|
|
&mode, &acl);
|
|
if (error)
|
|
goto unlock;
|
|
update_mode = 1;
|
|
}
|
|
error = __reiserfs_set_acl(&th, inode, type, acl);
|
|
if (!error && update_mode)
|
|
inode->i_mode = mode;
|
|
unlock:
|
|
reiserfs_write_lock(inode->i_sb);
|
|
error2 = journal_end(&th);
|
|
reiserfs_write_unlock(inode->i_sb);
|
|
if (error2)
|
|
error = error2;
|
|
}
|
|
|
|
return error;
|
|
}
|
|
|
|
/*
|
|
* Convert from filesystem to in-memory representation.
|
|
*/
|
|
static struct posix_acl *reiserfs_posix_acl_from_disk(const void *value, size_t size)
|
|
{
|
|
const char *end = (char *)value + size;
|
|
int n, count;
|
|
struct posix_acl *acl;
|
|
|
|
if (!value)
|
|
return NULL;
|
|
if (size < sizeof(reiserfs_acl_header))
|
|
return ERR_PTR(-EINVAL);
|
|
if (((reiserfs_acl_header *) value)->a_version !=
|
|
cpu_to_le32(REISERFS_ACL_VERSION))
|
|
return ERR_PTR(-EINVAL);
|
|
value = (char *)value + sizeof(reiserfs_acl_header);
|
|
count = reiserfs_acl_count(size);
|
|
if (count < 0)
|
|
return ERR_PTR(-EINVAL);
|
|
if (count == 0)
|
|
return NULL;
|
|
acl = posix_acl_alloc(count, GFP_NOFS);
|
|
if (!acl)
|
|
return ERR_PTR(-ENOMEM);
|
|
for (n = 0; n < count; n++) {
|
|
reiserfs_acl_entry *entry = (reiserfs_acl_entry *) value;
|
|
if ((char *)value + sizeof(reiserfs_acl_entry_short) > end)
|
|
goto fail;
|
|
acl->a_entries[n].e_tag = le16_to_cpu(entry->e_tag);
|
|
acl->a_entries[n].e_perm = le16_to_cpu(entry->e_perm);
|
|
switch (acl->a_entries[n].e_tag) {
|
|
case ACL_USER_OBJ:
|
|
case ACL_GROUP_OBJ:
|
|
case ACL_MASK:
|
|
case ACL_OTHER:
|
|
value = (char *)value +
|
|
sizeof(reiserfs_acl_entry_short);
|
|
break;
|
|
|
|
case ACL_USER:
|
|
value = (char *)value + sizeof(reiserfs_acl_entry);
|
|
if ((char *)value > end)
|
|
goto fail;
|
|
acl->a_entries[n].e_uid =
|
|
make_kuid(&init_user_ns,
|
|
le32_to_cpu(entry->e_id));
|
|
break;
|
|
case ACL_GROUP:
|
|
value = (char *)value + sizeof(reiserfs_acl_entry);
|
|
if ((char *)value > end)
|
|
goto fail;
|
|
acl->a_entries[n].e_gid =
|
|
make_kgid(&init_user_ns,
|
|
le32_to_cpu(entry->e_id));
|
|
break;
|
|
|
|
default:
|
|
goto fail;
|
|
}
|
|
}
|
|
if (value != end)
|
|
goto fail;
|
|
return acl;
|
|
|
|
fail:
|
|
posix_acl_release(acl);
|
|
return ERR_PTR(-EINVAL);
|
|
}
|
|
|
|
/*
|
|
* Convert from in-memory to filesystem representation.
|
|
*/
|
|
static void *reiserfs_posix_acl_to_disk(const struct posix_acl *acl, size_t * size)
|
|
{
|
|
reiserfs_acl_header *ext_acl;
|
|
char *e;
|
|
int n;
|
|
|
|
*size = reiserfs_acl_size(acl->a_count);
|
|
ext_acl = kmalloc(sizeof(reiserfs_acl_header) +
|
|
acl->a_count *
|
|
sizeof(reiserfs_acl_entry),
|
|
GFP_NOFS);
|
|
if (!ext_acl)
|
|
return ERR_PTR(-ENOMEM);
|
|
ext_acl->a_version = cpu_to_le32(REISERFS_ACL_VERSION);
|
|
e = (char *)ext_acl + sizeof(reiserfs_acl_header);
|
|
for (n = 0; n < acl->a_count; n++) {
|
|
const struct posix_acl_entry *acl_e = &acl->a_entries[n];
|
|
reiserfs_acl_entry *entry = (reiserfs_acl_entry *) e;
|
|
entry->e_tag = cpu_to_le16(acl->a_entries[n].e_tag);
|
|
entry->e_perm = cpu_to_le16(acl->a_entries[n].e_perm);
|
|
switch (acl->a_entries[n].e_tag) {
|
|
case ACL_USER:
|
|
entry->e_id = cpu_to_le32(
|
|
from_kuid(&init_user_ns, acl_e->e_uid));
|
|
e += sizeof(reiserfs_acl_entry);
|
|
break;
|
|
case ACL_GROUP:
|
|
entry->e_id = cpu_to_le32(
|
|
from_kgid(&init_user_ns, acl_e->e_gid));
|
|
e += sizeof(reiserfs_acl_entry);
|
|
break;
|
|
|
|
case ACL_USER_OBJ:
|
|
case ACL_GROUP_OBJ:
|
|
case ACL_MASK:
|
|
case ACL_OTHER:
|
|
e += sizeof(reiserfs_acl_entry_short);
|
|
break;
|
|
|
|
default:
|
|
goto fail;
|
|
}
|
|
}
|
|
return (char *)ext_acl;
|
|
|
|
fail:
|
|
kfree(ext_acl);
|
|
return ERR_PTR(-EINVAL);
|
|
}
|
|
|
|
/*
|
|
* Inode operation get_posix_acl().
|
|
*
|
|
* inode->i_mutex: down
|
|
* BKL held [before 2.5.x]
|
|
*/
|
|
struct posix_acl *reiserfs_get_acl(struct inode *inode, int type, bool rcu)
|
|
{
|
|
char *name, *value;
|
|
struct posix_acl *acl;
|
|
int size;
|
|
int retval;
|
|
|
|
if (rcu)
|
|
return ERR_PTR(-ECHILD);
|
|
|
|
switch (type) {
|
|
case ACL_TYPE_ACCESS:
|
|
name = XATTR_NAME_POSIX_ACL_ACCESS;
|
|
break;
|
|
case ACL_TYPE_DEFAULT:
|
|
name = XATTR_NAME_POSIX_ACL_DEFAULT;
|
|
break;
|
|
default:
|
|
BUG();
|
|
}
|
|
|
|
size = reiserfs_xattr_get(inode, name, NULL, 0);
|
|
if (size < 0) {
|
|
if (size == -ENODATA || size == -ENOSYS)
|
|
return NULL;
|
|
return ERR_PTR(size);
|
|
}
|
|
|
|
value = kmalloc(size, GFP_NOFS);
|
|
if (!value)
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
retval = reiserfs_xattr_get(inode, name, value, size);
|
|
if (retval == -ENODATA || retval == -ENOSYS) {
|
|
/*
|
|
* This shouldn't actually happen as it should have
|
|
* been caught above.. but just in case
|
|
*/
|
|
acl = NULL;
|
|
} else if (retval < 0) {
|
|
acl = ERR_PTR(retval);
|
|
} else {
|
|
acl = reiserfs_posix_acl_from_disk(value, retval);
|
|
}
|
|
|
|
kfree(value);
|
|
return acl;
|
|
}
|
|
|
|
/*
|
|
* Inode operation set_posix_acl().
|
|
*
|
|
* inode->i_mutex: down
|
|
* BKL held [before 2.5.x]
|
|
*/
|
|
static int
|
|
__reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode,
|
|
int type, struct posix_acl *acl)
|
|
{
|
|
char *name;
|
|
void *value = NULL;
|
|
size_t size = 0;
|
|
int error;
|
|
|
|
switch (type) {
|
|
case ACL_TYPE_ACCESS:
|
|
name = XATTR_NAME_POSIX_ACL_ACCESS;
|
|
break;
|
|
case ACL_TYPE_DEFAULT:
|
|
name = XATTR_NAME_POSIX_ACL_DEFAULT;
|
|
if (!S_ISDIR(inode->i_mode))
|
|
return acl ? -EACCES : 0;
|
|
break;
|
|
default:
|
|
return -EINVAL;
|
|
}
|
|
|
|
if (acl) {
|
|
value = reiserfs_posix_acl_to_disk(acl, &size);
|
|
if (IS_ERR(value))
|
|
return (int)PTR_ERR(value);
|
|
}
|
|
|
|
error = reiserfs_xattr_set_handle(th, inode, name, value, size, 0);
|
|
|
|
/*
|
|
* Ensure that the inode gets dirtied if we're only using
|
|
* the mode bits and an old ACL didn't exist. We don't need
|
|
* to check if the inode is hashed here since we won't get
|
|
* called by reiserfs_inherit_default_acl().
|
|
*/
|
|
if (error == -ENODATA) {
|
|
error = 0;
|
|
if (type == ACL_TYPE_ACCESS) {
|
|
inode->i_ctime = current_time(inode);
|
|
mark_inode_dirty(inode);
|
|
}
|
|
}
|
|
|
|
kfree(value);
|
|
|
|
if (!error)
|
|
set_cached_acl(inode, type, acl);
|
|
|
|
return error;
|
|
}
|
|
|
|
/*
|
|
* dir->i_mutex: locked,
|
|
* inode is new and not released into the wild yet
|
|
*/
|
|
int
|
|
reiserfs_inherit_default_acl(struct reiserfs_transaction_handle *th,
|
|
struct inode *dir, struct dentry *dentry,
|
|
struct inode *inode)
|
|
{
|
|
struct posix_acl *default_acl, *acl;
|
|
int err = 0;
|
|
|
|
/* ACLs only get applied to files and directories */
|
|
if (S_ISLNK(inode->i_mode))
|
|
return 0;
|
|
|
|
/*
|
|
* ACLs can only be used on "new" objects, so if it's an old object
|
|
* there is nothing to inherit from
|
|
*/
|
|
if (get_inode_sd_version(dir) == STAT_DATA_V1)
|
|
goto apply_umask;
|
|
|
|
/*
|
|
* Don't apply ACLs to objects in the .reiserfs_priv tree.. This
|
|
* would be useless since permissions are ignored, and a pain because
|
|
* it introduces locking cycles
|
|
*/
|
|
if (IS_PRIVATE(inode))
|
|
goto apply_umask;
|
|
|
|
err = posix_acl_create(dir, &inode->i_mode, &default_acl, &acl);
|
|
if (err)
|
|
return err;
|
|
|
|
if (default_acl) {
|
|
err = __reiserfs_set_acl(th, inode, ACL_TYPE_DEFAULT,
|
|
default_acl);
|
|
posix_acl_release(default_acl);
|
|
}
|
|
if (acl) {
|
|
if (!err)
|
|
err = __reiserfs_set_acl(th, inode, ACL_TYPE_ACCESS,
|
|
acl);
|
|
posix_acl_release(acl);
|
|
}
|
|
|
|
return err;
|
|
|
|
apply_umask:
|
|
/* no ACL, apply umask */
|
|
inode->i_mode &= ~current_umask();
|
|
return err;
|
|
}
|
|
|
|
/* This is used to cache the default acl before a new object is created.
|
|
* The biggest reason for this is to get an idea of how many blocks will
|
|
* actually be required for the create operation if we must inherit an ACL.
|
|
* An ACL write can add up to 3 object creations and an additional file write
|
|
* so we'd prefer not to reserve that many blocks in the journal if we can.
|
|
* It also has the advantage of not loading the ACL with a transaction open,
|
|
* this may seem silly, but if the owner of the directory is doing the
|
|
* creation, the ACL may not be loaded since the permissions wouldn't require
|
|
* it.
|
|
* We return the number of blocks required for the transaction.
|
|
*/
|
|
int reiserfs_cache_default_acl(struct inode *inode)
|
|
{
|
|
struct posix_acl *acl;
|
|
int nblocks = 0;
|
|
|
|
if (IS_PRIVATE(inode))
|
|
return 0;
|
|
|
|
acl = get_inode_acl(inode, ACL_TYPE_DEFAULT);
|
|
|
|
if (acl && !IS_ERR(acl)) {
|
|
int size = reiserfs_acl_size(acl->a_count);
|
|
|
|
/* Other xattrs can be created during inode creation. We don't
|
|
* want to claim too many blocks, so we check to see if we
|
|
* need to create the tree to the xattrs, and then we
|
|
* just want two files. */
|
|
nblocks = reiserfs_xattr_jcreate_nblocks(inode);
|
|
nblocks += JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb);
|
|
|
|
REISERFS_I(inode)->i_flags |= i_has_xattr_dir;
|
|
|
|
/* We need to account for writes + bitmaps for two files */
|
|
nblocks += reiserfs_xattr_nblocks(inode, size) * 4;
|
|
posix_acl_release(acl);
|
|
}
|
|
|
|
return nblocks;
|
|
}
|
|
|
|
/*
|
|
* Called under i_mutex
|
|
*/
|
|
int reiserfs_acl_chmod(struct dentry *dentry)
|
|
{
|
|
struct inode *inode = d_inode(dentry);
|
|
|
|
if (IS_PRIVATE(inode))
|
|
return 0;
|
|
if (get_inode_sd_version(inode) == STAT_DATA_V1 ||
|
|
!reiserfs_posixacl(inode->i_sb))
|
|
return 0;
|
|
|
|
return posix_acl_chmod(&init_user_ns, dentry, inode->i_mode);
|
|
}
|