linux/fs/fuse/dir.c

1795 lines
44 KiB
C
Raw Normal View History

/*
FUSE: Filesystem in Userspace
Copyright (C) 2001-2008 Miklos Szeredi <miklos@szeredi.hu>
This program can be distributed under the terms of the GNU GPL.
See the file COPYING.
*/
#include "fuse_i.h"
#include <linux/pagemap.h>
#include <linux/file.h>
#include <linux/sched.h>
#include <linux/namei.h>
#include <linux/slab.h>
#include <linux/xattr.h>
#include <linux/iversion.h>
#include <linux/posix_acl.h>
static void fuse_advise_use_readdirplus(struct inode *dir)
{
struct fuse_inode *fi = get_fuse_inode(dir);
set_bit(FUSE_I_ADVISE_RDPLUS, &fi->state);
}
#if BITS_PER_LONG >= 64
static inline void __fuse_dentry_settime(struct dentry *entry, u64 time)
{
entry->d_fsdata = (void *) time;
}
static inline u64 fuse_dentry_time(const struct dentry *entry)
{
return (u64)entry->d_fsdata;
}
#else
union fuse_dentry {
u64 time;
struct rcu_head rcu;
};
static inline void __fuse_dentry_settime(struct dentry *dentry, u64 time)
{
((union fuse_dentry *) dentry->d_fsdata)->time = time;
}
static inline u64 fuse_dentry_time(const struct dentry *entry)
{
return ((union fuse_dentry *) entry->d_fsdata)->time;
}
#endif
static void fuse_dentry_settime(struct dentry *dentry, u64 time)
{
struct fuse_conn *fc = get_fuse_conn_super(dentry->d_sb);
bool delete = !time && fc->delete_stale;
/*
* Mess with DCACHE_OP_DELETE because dput() will be faster without it.
* Don't care about races, either way it's just an optimization
*/
if ((!delete && (dentry->d_flags & DCACHE_OP_DELETE)) ||
(delete && !(dentry->d_flags & DCACHE_OP_DELETE))) {
spin_lock(&dentry->d_lock);
if (!delete)
dentry->d_flags &= ~DCACHE_OP_DELETE;
else
dentry->d_flags |= DCACHE_OP_DELETE;
spin_unlock(&dentry->d_lock);
}
__fuse_dentry_settime(dentry, time);
}
/*
* FUSE caches dentries and attributes with separate timeout. The
* time in jiffies until the dentry/attributes are valid is stored in
* dentry->d_fsdata and fuse_inode->i_time respectively.
*/
/*
* Calculate the time in jiffies until a dentry/attributes are valid
*/
static u64 time_to_jiffies(u64 sec, u32 nsec)
{
if (sec || nsec) {
struct timespec64 ts = {
sec,
min_t(u32, nsec, NSEC_PER_SEC - 1)
};
return get_jiffies_64() + timespec64_to_jiffies(&ts);
} else
return 0;
}
/*
* Set dentry and possibly attribute timeouts from the lookup/mk*
* replies
*/
void fuse_change_entry_timeout(struct dentry *entry, struct fuse_entry_out *o)
{
fuse_dentry_settime(entry,
time_to_jiffies(o->entry_valid, o->entry_valid_nsec));
}
static u64 attr_timeout(struct fuse_attr_out *o)
{
return time_to_jiffies(o->attr_valid, o->attr_valid_nsec);
}
u64 entry_attr_timeout(struct fuse_entry_out *o)
{
return time_to_jiffies(o->attr_valid, o->attr_valid_nsec);
}
static void fuse_invalidate_attr_mask(struct inode *inode, u32 mask)
{
set_mask_bits(&get_fuse_inode(inode)->inval_mask, 0, mask);
}
/*
* Mark the attributes as stale, so that at the next call to
* ->getattr() they will be fetched from userspace
*/
void fuse_invalidate_attr(struct inode *inode)
{
fuse_invalidate_attr_mask(inode, STATX_BASIC_STATS);
}
static void fuse_dir_changed(struct inode *dir)
{
fuse_invalidate_attr(dir);
inode_maybe_inc_iversion(dir, false);
}
/**
* Mark the attributes as stale due to an atime change. Avoid the invalidate if
* atime is not used.
*/
void fuse_invalidate_atime(struct inode *inode)
{
if (!IS_RDONLY(inode))
fuse_invalidate_attr_mask(inode, STATX_ATIME);
}
/*
* Just mark the entry as stale, so that a next attempt to look it up
* will result in a new lookup call to userspace
*
* This is called when a dentry is about to become negative and the
* timeout is unknown (unlink, rmdir, rename and in some cases
* lookup)
*/
void fuse_invalidate_entry_cache(struct dentry *entry)
{
fuse_dentry_settime(entry, 0);
}
/*
* Same as fuse_invalidate_entry_cache(), but also try to remove the
* dentry from the hash
*/
static void fuse_invalidate_entry(struct dentry *entry)
{
d_invalidate(entry);
fuse_invalidate_entry_cache(entry);
}
static void fuse_lookup_init(struct fuse_conn *fc, struct fuse_args *args,
u64 nodeid, const struct qstr *name,
struct fuse_entry_out *outarg)
{
memset(outarg, 0, sizeof(struct fuse_entry_out));
args->opcode = FUSE_LOOKUP;
args->nodeid = nodeid;
args->in_numargs = 1;
args->in_args[0].size = name->len + 1;
args->in_args[0].value = name->name;
args->out_numargs = 1;
args->out_args[0].size = sizeof(struct fuse_entry_out);
args->out_args[0].value = outarg;
}
/*
* Check whether the dentry is still valid
*
* If the entry validity timeout has expired and the dentry is
* positive, try to redo the lookup. If the lookup results in a
* different inode, then let the VFS invalidate the dentry and redo
* the lookup once more. If the lookup results in the same inode,
* then refresh the attributes, timeouts and mark the dentry valid.
*/
static int fuse_dentry_revalidate(struct dentry *entry, unsigned int flags)
{
struct inode *inode;
struct dentry *parent;
struct fuse_conn *fc;
struct fuse_inode *fi;
int ret;
inode = d_inode_rcu(entry);
if (inode && is_bad_inode(inode))
goto invalid;
else if (time_before64(fuse_dentry_time(entry), get_jiffies_64()) ||
(flags & LOOKUP_REVAL)) {
struct fuse_entry_out outarg;
FUSE_ARGS(args);
struct fuse_forget_link *forget;
u64 attr_version;
/* For negative dentries, always do a fresh lookup */
if (!inode)
goto invalid;
ret = -ECHILD;
if (flags & LOOKUP_RCU)
goto out;
fc = get_fuse_conn(inode);
forget = fuse_alloc_forget();
ret = -ENOMEM;
if (!forget)
goto out;
attr_version = fuse_get_attr_version(fc);
parent = dget_parent(entry);
fuse_lookup_init(fc, &args, get_node_id(d_inode(parent)),
&entry->d_name, &outarg);
ret = fuse_simple_request(fc, &args);
dput(parent);
/* Zero nodeid is same as -ENOENT */
if (!ret && !outarg.nodeid)
ret = -ENOENT;
if (!ret) {
fi = get_fuse_inode(inode);
if (outarg.nodeid != get_node_id(inode)) {
fuse_queue_forget(fc, forget, outarg.nodeid, 1);
goto invalid;
}
spin_lock(&fi->lock);
fi->nlookup++;
spin_unlock(&fi->lock);
}
kfree(forget);
if (ret == -ENOMEM)
goto out;
if (ret || fuse_invalid_attr(&outarg.attr) ||
(outarg.attr.mode ^ inode->i_mode) & S_IFMT)
goto invalid;
forget_all_cached_acls(inode);
fuse_change_attributes(inode, &outarg.attr,
entry_attr_timeout(&outarg),
attr_version);
fuse_change_entry_timeout(entry, &outarg);
} else if (inode) {
fi = get_fuse_inode(inode);
if (flags & LOOKUP_RCU) {
if (test_bit(FUSE_I_INIT_RDPLUS, &fi->state))
return -ECHILD;
} else if (test_and_clear_bit(FUSE_I_INIT_RDPLUS, &fi->state)) {
parent = dget_parent(entry);
fuse_advise_use_readdirplus(d_inode(parent));
dput(parent);
}
}
ret = 1;
out:
return ret;
invalid:
ret = 0;
goto out;
}
#if BITS_PER_LONG < 64
static int fuse_dentry_init(struct dentry *dentry)
{
dentry->d_fsdata = kzalloc(sizeof(union fuse_dentry),
GFP_KERNEL_ACCOUNT | __GFP_RECLAIMABLE);
return dentry->d_fsdata ? 0 : -ENOMEM;
}
static void fuse_dentry_release(struct dentry *dentry)
{
union fuse_dentry *fd = dentry->d_fsdata;
kfree_rcu(fd, rcu);
}
#endif
static int fuse_dentry_delete(const struct dentry *dentry)
{
return time_before64(fuse_dentry_time(dentry), get_jiffies_64());
}
const struct dentry_operations fuse_dentry_operations = {
.d_revalidate = fuse_dentry_revalidate,
.d_delete = fuse_dentry_delete,
#if BITS_PER_LONG < 64
.d_init = fuse_dentry_init,
.d_release = fuse_dentry_release,
#endif
};
const struct dentry_operations fuse_root_dentry_operations = {
#if BITS_PER_LONG < 64
.d_init = fuse_dentry_init,
.d_release = fuse_dentry_release,
#endif
};
int fuse_valid_type(int m)
{
return S_ISREG(m) || S_ISDIR(m) || S_ISLNK(m) || S_ISCHR(m) ||
S_ISBLK(m) || S_ISFIFO(m) || S_ISSOCK(m);
}
bool fuse_invalid_attr(struct fuse_attr *attr)
{
return !fuse_valid_type(attr->mode) ||
attr->size > LLONG_MAX;
}
int fuse_lookup_name(struct super_block *sb, u64 nodeid, const struct qstr *name,
struct fuse_entry_out *outarg, struct inode **inode)
{
struct fuse_conn *fc = get_fuse_conn_super(sb);
FUSE_ARGS(args);
struct fuse_forget_link *forget;
u64 attr_version;
int err;
*inode = NULL;
err = -ENAMETOOLONG;
if (name->len > FUSE_NAME_MAX)
goto out;
forget = fuse_alloc_forget();
err = -ENOMEM;
if (!forget)
goto out;
attr_version = fuse_get_attr_version(fc);
fuse_lookup_init(fc, &args, nodeid, name, outarg);
err = fuse_simple_request(fc, &args);
/* Zero nodeid is same as -ENOENT, but with valid timeout */
if (err || !outarg->nodeid)
goto out_put_forget;
err = -EIO;
if (!outarg->nodeid)
goto out_put_forget;
if (fuse_invalid_attr(&outarg->attr))
goto out_put_forget;
*inode = fuse_iget(sb, outarg->nodeid, outarg->generation,
&outarg->attr, entry_attr_timeout(outarg),
attr_version);
err = -ENOMEM;
if (!*inode) {
fuse_queue_forget(fc, forget, outarg->nodeid, 1);
goto out;
}
err = 0;
out_put_forget:
kfree(forget);
out:
return err;
}
static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry,
unsigned int flags)
{
int err;
struct fuse_entry_out outarg;
struct inode *inode;
struct dentry *newent;
bool outarg_valid = true;
bool locked;
locked = fuse_lock_inode(dir);
err = fuse_lookup_name(dir->i_sb, get_node_id(dir), &entry->d_name,
&outarg, &inode);
fuse_unlock_inode(dir, locked);
if (err == -ENOENT) {
outarg_valid = false;
err = 0;
}
if (err)
goto out_err;
err = -EIO;
if (inode && get_node_id(inode) == FUSE_ROOT_ID)
goto out_iput;
newent = d_splice_alias(inode, entry);
err = PTR_ERR(newent);
if (IS_ERR(newent))
goto out_err;
entry = newent ? newent : entry;
if (outarg_valid)
fuse_change_entry_timeout(entry, &outarg);
else
fuse_invalidate_entry_cache(entry);
if (inode)
fuse_advise_use_readdirplus(dir);
return newent;
out_iput:
iput(inode);
out_err:
return ERR_PTR(err);
}
/*
* Atomic create+open operation
*
* If the filesystem doesn't support this, then fall back to separate
* 'mknod' + 'open' requests.
*/
static int fuse_create_open(struct inode *dir, struct dentry *entry,
struct file *file, unsigned flags,
umode_t mode)
{
int err;
struct inode *inode;
struct fuse_conn *fc = get_fuse_conn(dir);
FUSE_ARGS(args);
struct fuse_forget_link *forget;
struct fuse_create_in inarg;
struct fuse_open_out outopen;
struct fuse_entry_out outentry;
struct fuse_inode *fi;
struct fuse_file *ff;
/* Userspace expects S_IFREG in create mode */
BUG_ON((mode & S_IFMT) != S_IFREG);
forget = fuse_alloc_forget();
err = -ENOMEM;
if (!forget)
goto out_err;
err = -ENOMEM;
ff = fuse_file_alloc(fc);
if (!ff)
goto out_put_forget_req;
if (!fc->dont_mask)
mode &= ~current_umask();
flags &= ~O_NOCTTY;
memset(&inarg, 0, sizeof(inarg));
memset(&outentry, 0, sizeof(outentry));
inarg.flags = flags;
inarg.mode = mode;
inarg.umask = current_umask();
args.opcode = FUSE_CREATE;
args.nodeid = get_node_id(dir);
args.in_numargs = 2;
args.in_args[0].size = sizeof(inarg);
args.in_args[0].value = &inarg;
args.in_args[1].size = entry->d_name.len + 1;
args.in_args[1].value = entry->d_name.name;
args.out_numargs = 2;
args.out_args[0].size = sizeof(outentry);
args.out_args[0].value = &outentry;
args.out_args[1].size = sizeof(outopen);
args.out_args[1].value = &outopen;
err = fuse_simple_request(fc, &args);
if (err)
goto out_free_ff;
err = -EIO;
if (!S_ISREG(outentry.attr.mode) || invalid_nodeid(outentry.nodeid) ||
fuse_invalid_attr(&outentry.attr))
goto out_free_ff;
ff->fh = outopen.fh;
ff->nodeid = outentry.nodeid;
ff->open_flags = outopen.open_flags;
inode = fuse_iget(dir->i_sb, outentry.nodeid, outentry.generation,
&outentry.attr, entry_attr_timeout(&outentry), 0);
if (!inode) {
flags &= ~(O_CREAT | O_EXCL | O_TRUNC);
fuse_sync_release(NULL, ff, flags);
fuse_queue_forget(fc, forget, outentry.nodeid, 1);
err = -ENOMEM;
goto out_err;
}
kfree(forget);
d_instantiate(entry, inode);
fuse_change_entry_timeout(entry, &outentry);
fuse_dir_changed(dir);
err = finish_open(file, entry, generic_file_open);
if (err) {
fi = get_fuse_inode(inode);
fuse_sync_release(fi, ff, flags);
} else {
file->private_data = ff;
fuse_finish_open(inode, file);
}
return err;
out_free_ff:
fuse_file_free(ff);
out_put_forget_req:
kfree(forget);
out_err:
return err;
}
static int fuse_mknod(struct inode *, struct dentry *, umode_t, dev_t);
static int fuse_atomic_open(struct inode *dir, struct dentry *entry,
struct file *file, unsigned flags,
umode_t mode)
{
int err;
struct fuse_conn *fc = get_fuse_conn(dir);
struct dentry *res = NULL;
if (d_in_lookup(entry)) {
res = fuse_lookup(dir, entry, 0);
if (IS_ERR(res))
return PTR_ERR(res);
if (res)
entry = res;
}
if (!(flags & O_CREAT) || d_really_is_positive(entry))
goto no_open;
/* Only creates */
file->f_mode |= FMODE_CREATED;
if (fc->no_create)
goto mknod;
err = fuse_create_open(dir, entry, file, flags, mode);
if (err == -ENOSYS) {
fc->no_create = 1;
goto mknod;
}
out_dput:
dput(res);
return err;
mknod:
err = fuse_mknod(dir, entry, mode, 0);
if (err)
goto out_dput;
no_open:
return finish_no_open(file, res);
}
/*
* Code shared between mknod, mkdir, symlink and link
*/
static int create_new_entry(struct fuse_conn *fc, struct fuse_args *args,
struct inode *dir, struct dentry *entry,
umode_t mode)
{
struct fuse_entry_out outarg;
struct inode *inode;
struct dentry *d;
int err;
struct fuse_forget_link *forget;
forget = fuse_alloc_forget();
if (!forget)
return -ENOMEM;
memset(&outarg, 0, sizeof(outarg));
args->nodeid = get_node_id(dir);
args->out_numargs = 1;
args->out_args[0].size = sizeof(outarg);
args->out_args[0].value = &outarg;
err = fuse_simple_request(fc, args);
if (err)
goto out_put_forget_req;
err = -EIO;
if (invalid_nodeid(outarg.nodeid) || fuse_invalid_attr(&outarg.attr))
goto out_put_forget_req;
if ((outarg.attr.mode ^ mode) & S_IFMT)
goto out_put_forget_req;
inode = fuse_iget(dir->i_sb, outarg.nodeid, outarg.generation,
&outarg.attr, entry_attr_timeout(&outarg), 0);
if (!inode) {
fuse_queue_forget(fc, forget, outarg.nodeid, 1);
return -ENOMEM;
}
kfree(forget);
d_drop(entry);
d = d_splice_alias(inode, entry);
if (IS_ERR(d))
return PTR_ERR(d);
if (d) {
fuse_change_entry_timeout(d, &outarg);
dput(d);
} else {
fuse_change_entry_timeout(entry, &outarg);
}
fuse_dir_changed(dir);
return 0;
out_put_forget_req:
kfree(forget);
return err;
}
static int fuse_mknod(struct inode *dir, struct dentry *entry, umode_t mode,
dev_t rdev)
{
struct fuse_mknod_in inarg;
struct fuse_conn *fc = get_fuse_conn(dir);
FUSE_ARGS(args);
if (!fc->dont_mask)
mode &= ~current_umask();
memset(&inarg, 0, sizeof(inarg));
inarg.mode = mode;
inarg.rdev = new_encode_dev(rdev);
inarg.umask = current_umask();
args.opcode = FUSE_MKNOD;
args.in_numargs = 2;
args.in_args[0].size = sizeof(inarg);
args.in_args[0].value = &inarg;
args.in_args[1].size = entry->d_name.len + 1;
args.in_args[1].value = entry->d_name.name;
return create_new_entry(fc, &args, dir, entry, mode);
}
static int fuse_create(struct inode *dir, struct dentry *entry, umode_t mode,
bool excl)
{
return fuse_mknod(dir, entry, mode, 0);
}
static int fuse_mkdir(struct inode *dir, struct dentry *entry, umode_t mode)
{
struct fuse_mkdir_in inarg;
struct fuse_conn *fc = get_fuse_conn(dir);
FUSE_ARGS(args);
if (!fc->dont_mask)
mode &= ~current_umask();
memset(&inarg, 0, sizeof(inarg));
inarg.mode = mode;
inarg.umask = current_umask();
args.opcode = FUSE_MKDIR;
args.in_numargs = 2;
args.in_args[0].size = sizeof(inarg);
args.in_args[0].value = &inarg;
args.in_args[1].size = entry->d_name.len + 1;
args.in_args[1].value = entry->d_name.name;
return create_new_entry(fc, &args, dir, entry, S_IFDIR);
}
static int fuse_symlink(struct inode *dir, struct dentry *entry,
const char *link)
{
struct fuse_conn *fc = get_fuse_conn(dir);
unsigned len = strlen(link) + 1;
FUSE_ARGS(args);
args.opcode = FUSE_SYMLINK;
args.in_numargs = 2;
args.in_args[0].size = entry->d_name.len + 1;
args.in_args[0].value = entry->d_name.name;
args.in_args[1].size = len;
args.in_args[1].value = link;
return create_new_entry(fc, &args, dir, entry, S_IFLNK);
}
void fuse_update_ctime(struct inode *inode)
{
if (!IS_NOCMTIME(inode)) {
inode->i_ctime = current_time(inode);
mark_inode_dirty_sync(inode);
}
}
static int fuse_unlink(struct inode *dir, struct dentry *entry)
{
int err;
struct fuse_conn *fc = get_fuse_conn(dir);
FUSE_ARGS(args);
args.opcode = FUSE_UNLINK;
args.nodeid = get_node_id(dir);
args.in_numargs = 1;
args.in_args[0].size = entry->d_name.len + 1;
args.in_args[0].value = entry->d_name.name;
err = fuse_simple_request(fc, &args);
if (!err) {
struct inode *inode = d_inode(entry);
struct fuse_inode *fi = get_fuse_inode(inode);
spin_lock(&fi->lock);
fi->attr_version = atomic64_inc_return(&fc->attr_version);
/*
* If i_nlink == 0 then unlink doesn't make sense, yet this can
* happen if userspace filesystem is careless. It would be
* difficult to enforce correct nlink usage so just ignore this
* condition here
*/
if (inode->i_nlink > 0)
drop_nlink(inode);
spin_unlock(&fi->lock);
fuse_invalidate_attr(inode);
fuse_dir_changed(dir);
fuse_invalidate_entry_cache(entry);
fuse_update_ctime(inode);
} else if (err == -EINTR)
fuse_invalidate_entry(entry);
return err;
}
static int fuse_rmdir(struct inode *dir, struct dentry *entry)
{
int err;
struct fuse_conn *fc = get_fuse_conn(dir);
FUSE_ARGS(args);
args.opcode = FUSE_RMDIR;
args.nodeid = get_node_id(dir);
args.in_numargs = 1;
args.in_args[0].size = entry->d_name.len + 1;
args.in_args[0].value = entry->d_name.name;
err = fuse_simple_request(fc, &args);
if (!err) {
clear_nlink(d_inode(entry));
fuse_dir_changed(dir);
fuse_invalidate_entry_cache(entry);
} else if (err == -EINTR)
fuse_invalidate_entry(entry);
return err;
}
static int fuse_rename_common(struct inode *olddir, struct dentry *oldent,
struct inode *newdir, struct dentry *newent,
unsigned int flags, int opcode, size_t argsize)
{
int err;
struct fuse_rename2_in inarg;
struct fuse_conn *fc = get_fuse_conn(olddir);
FUSE_ARGS(args);
memset(&inarg, 0, argsize);
inarg.newdir = get_node_id(newdir);
inarg.flags = flags;
args.opcode = opcode;
args.nodeid = get_node_id(olddir);
args.in_numargs = 3;
args.in_args[0].size = argsize;
args.in_args[0].value = &inarg;
args.in_args[1].size = oldent->d_name.len + 1;
args.in_args[1].value = oldent->d_name.name;
args.in_args[2].size = newent->d_name.len + 1;
args.in_args[2].value = newent->d_name.name;
err = fuse_simple_request(fc, &args);
if (!err) {
/* ctime changes */
fuse_invalidate_attr(d_inode(oldent));
fuse_update_ctime(d_inode(oldent));
if (flags & RENAME_EXCHANGE) {
fuse_invalidate_attr(d_inode(newent));
fuse_update_ctime(d_inode(newent));
}
fuse_dir_changed(olddir);
if (olddir != newdir)
fuse_dir_changed(newdir);
/* newent will end up negative */
if (!(flags & RENAME_EXCHANGE) && d_really_is_positive(newent)) {
fuse_invalidate_attr(d_inode(newent));
fuse_invalidate_entry_cache(newent);
fuse_update_ctime(d_inode(newent));
}
} else if (err == -EINTR) {
/* If request was interrupted, DEITY only knows if the
rename actually took place. If the invalidation
fails (e.g. some process has CWD under the renamed
directory), then there can be inconsistency between
the dcache and the real filesystem. Tough luck. */
fuse_invalidate_entry(oldent);
if (d_really_is_positive(newent))
fuse_invalidate_entry(newent);
}
return err;
}
static int fuse_rename2(struct inode *olddir, struct dentry *oldent,
struct inode *newdir, struct dentry *newent,
unsigned int flags)
{
struct fuse_conn *fc = get_fuse_conn(olddir);
int err;
if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT))
return -EINVAL;
if (flags) {
if (fc->no_rename2 || fc->minor < 23)
return -EINVAL;
err = fuse_rename_common(olddir, oldent, newdir, newent, flags,
FUSE_RENAME2,
sizeof(struct fuse_rename2_in));
if (err == -ENOSYS) {
fc->no_rename2 = 1;
err = -EINVAL;
}
} else {
err = fuse_rename_common(olddir, oldent, newdir, newent, 0,
FUSE_RENAME,
sizeof(struct fuse_rename_in));
}
return err;
}
static int fuse_link(struct dentry *entry, struct inode *newdir,
struct dentry *newent)
{
int err;
struct fuse_link_in inarg;
struct inode *inode = d_inode(entry);
struct fuse_conn *fc = get_fuse_conn(inode);
FUSE_ARGS(args);
memset(&inarg, 0, sizeof(inarg));
inarg.oldnodeid = get_node_id(inode);
args.opcode = FUSE_LINK;
args.in_numargs = 2;
args.in_args[0].size = sizeof(inarg);
args.in_args[0].value = &inarg;
args.in_args[1].size = newent->d_name.len + 1;
args.in_args[1].value = newent->d_name.name;
err = create_new_entry(fc, &args, newdir, newent, inode->i_mode);
/* Contrary to "normal" filesystems it can happen that link
makes two "logical" inodes point to the same "physical"
inode. We invalidate the attributes of the old one, so it
will reflect changes in the backing inode (link count,
etc.)
*/
if (!err) {
struct fuse_inode *fi = get_fuse_inode(inode);
spin_lock(&fi->lock);
fi->attr_version = atomic64_inc_return(&fc->attr_version);
if (likely(inode->i_nlink < UINT_MAX))
inc_nlink(inode);
spin_unlock(&fi->lock);
fuse_invalidate_attr(inode);
fuse_update_ctime(inode);
} else if (err == -EINTR) {
fuse_invalidate_attr(inode);
}
return err;
}
static void fuse_fillattr(struct inode *inode, struct fuse_attr *attr,
struct kstat *stat)
{
unsigned int blkbits;
struct fuse_conn *fc = get_fuse_conn(inode);
/* see the comment in fuse_change_attributes() */
if (fc->writeback_cache && S_ISREG(inode->i_mode)) {
attr->size = i_size_read(inode);
attr->mtime = inode->i_mtime.tv_sec;
attr->mtimensec = inode->i_mtime.tv_nsec;
attr->ctime = inode->i_ctime.tv_sec;
attr->ctimensec = inode->i_ctime.tv_nsec;
}
stat->dev = inode->i_sb->s_dev;
stat->ino = attr->ino;
stat->mode = (inode->i_mode & S_IFMT) | (attr->mode & 07777);
stat->nlink = attr->nlink;
fuse: Support fuse filesystems outside of init_user_ns In order to support mounts from namespaces other than init_user_ns, fuse must translate uids and gids to/from the userns of the process servicing requests on /dev/fuse. This patch does that, with a couple of restrictions on the namespace: - The userns for the fuse connection is fixed to the namespace from which /dev/fuse is opened. - The namespace must be the same as s_user_ns. These restrictions simplify the implementation by avoiding the need to pass around userns references and by allowing fuse to rely on the checks in setattr_prepare for ownership changes. Either restriction could be relaxed in the future if needed. For cuse the userns used is the opener of /dev/cuse. Semantically the cuse support does not appear safe for unprivileged users. Practically the permissions on /dev/cuse only make it accessible to the global root user. If something slips through the cracks in a user namespace the only users who will be able to use the cuse device are those users mapped into the user namespace. Translation in the posix acl is updated to use the uuser namespace of the filesystem. Avoiding cases which might bypass this translation is handled in a following change. This change is stronlgy based on a similar change from Seth Forshee and Dongsu Park. Cc: Seth Forshee <seth.forshee@canonical.com> Cc: Dongsu Park <dongsu@kinvolk.io> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
2018-02-21 17:18:07 +00:00
stat->uid = make_kuid(fc->user_ns, attr->uid);
stat->gid = make_kgid(fc->user_ns, attr->gid);
stat->rdev = inode->i_rdev;
stat->atime.tv_sec = attr->atime;
stat->atime.tv_nsec = attr->atimensec;
stat->mtime.tv_sec = attr->mtime;
stat->mtime.tv_nsec = attr->mtimensec;
stat->ctime.tv_sec = attr->ctime;
stat->ctime.tv_nsec = attr->ctimensec;
stat->size = attr->size;
stat->blocks = attr->blocks;
if (attr->blksize != 0)
blkbits = ilog2(attr->blksize);
else
blkbits = inode->i_sb->s_blocksize_bits;
stat->blksize = 1 << blkbits;
}
static int fuse_do_getattr(struct inode *inode, struct kstat *stat,
struct file *file)
{
int err;
struct fuse_getattr_in inarg;
struct fuse_attr_out outarg;
struct fuse_conn *fc = get_fuse_conn(inode);
FUSE_ARGS(args);
u64 attr_version;
attr_version = fuse_get_attr_version(fc);
memset(&inarg, 0, sizeof(inarg));
memset(&outarg, 0, sizeof(outarg));
/* Directories have separate file-handle space */
if (file && S_ISREG(inode->i_mode)) {
struct fuse_file *ff = file->private_data;
inarg.getattr_flags |= FUSE_GETATTR_FH;
inarg.fh = ff->fh;
}
args.opcode = FUSE_GETATTR;
args.nodeid = get_node_id(inode);
args.in_numargs = 1;
args.in_args[0].size = sizeof(inarg);
args.in_args[0].value = &inarg;
args.out_numargs = 1;
args.out_args[0].size = sizeof(outarg);
args.out_args[0].value = &outarg;
err = fuse_simple_request(fc, &args);
if (!err) {
if (fuse_invalid_attr(&outarg.attr) ||
(inode->i_mode ^ outarg.attr.mode) & S_IFMT) {
make_bad_inode(inode);
err = -EIO;
} else {
fuse_change_attributes(inode, &outarg.attr,
attr_timeout(&outarg),
attr_version);
if (stat)
fuse_fillattr(inode, &outarg.attr, stat);
}
}
return err;
}
static int fuse_update_get_attr(struct inode *inode, struct file *file,
struct kstat *stat, u32 request_mask,
unsigned int flags)
{
struct fuse_inode *fi = get_fuse_inode(inode);
int err = 0;
bool sync;
if (flags & AT_STATX_FORCE_SYNC)
sync = true;
else if (flags & AT_STATX_DONT_SYNC)
sync = false;
else if (request_mask & READ_ONCE(fi->inval_mask))
sync = true;
else
sync = time_before64(fi->i_time, get_jiffies_64());
if (sync) {
forget_all_cached_acls(inode);
err = fuse_do_getattr(inode, stat, file);
} else if (stat) {
generic_fillattr(inode, stat);
stat->mode = fi->orig_i_mode;
stat->ino = fi->orig_ino;
}
return err;
}
int fuse_update_attributes(struct inode *inode, struct file *file)
{
/* Do *not* need to get atime for internal purposes */
return fuse_update_get_attr(inode, file, NULL,
STATX_BASIC_STATS & ~STATX_ATIME, 0);
}
int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid,
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
u64 child_nodeid, struct qstr *name)
{
int err = -ENOTDIR;
struct inode *parent;
struct dentry *dir;
struct dentry *entry;
parent = ilookup5(sb, parent_nodeid, fuse_inode_eq, &parent_nodeid);
if (!parent)
return -ENOENT;
inode_lock(parent);
if (!S_ISDIR(parent->i_mode))
goto unlock;
err = -ENOENT;
dir = d_find_alias(parent);
if (!dir)
goto unlock;
name->hash = full_name_hash(dir, name->name, name->len);
entry = d_lookup(dir, name);
dput(dir);
if (!entry)
goto unlock;
fuse_dir_changed(parent);
fuse_invalidate_entry(entry);
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
if (child_nodeid != 0 && d_really_is_positive(entry)) {
inode_lock(d_inode(entry));
if (get_node_id(d_inode(entry)) != child_nodeid) {
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
err = -ENOENT;
goto badentry;
}
if (d_mountpoint(entry)) {
err = -EBUSY;
goto badentry;
}
VFS: (Scripted) Convert S_ISLNK/DIR/REG(dentry->d_inode) to d_is_*(dentry) Convert the following where appropriate: (1) S_ISLNK(dentry->d_inode) to d_is_symlink(dentry). (2) S_ISREG(dentry->d_inode) to d_is_reg(dentry). (3) S_ISDIR(dentry->d_inode) to d_is_dir(dentry). This is actually more complicated than it appears as some calls should be converted to d_can_lookup() instead. The difference is whether the directory in question is a real dir with a ->lookup op or whether it's a fake dir with a ->d_automount op. In some circumstances, we can subsume checks for dentry->d_inode not being NULL into this, provided we the code isn't in a filesystem that expects d_inode to be NULL if the dirent really *is* negative (ie. if we're going to use d_inode() rather than d_backing_inode() to get the inode pointer). Note that the dentry type field may be set to something other than DCACHE_MISS_TYPE when d_inode is NULL in the case of unionmount, where the VFS manages the fall-through from a negative dentry to a lower layer. In such a case, the dentry type of the negative union dentry is set to the same as the type of the lower dentry. However, if you know d_inode is not NULL at the call site, then you can use the d_is_xxx() functions even in a filesystem. There is one further complication: a 0,0 chardev dentry may be labelled DCACHE_WHITEOUT_TYPE rather than DCACHE_SPECIAL_TYPE. Strictly, this was intended for special directory entry types that don't have attached inodes. The following perl+coccinelle script was used: use strict; my @callers; open($fd, 'git grep -l \'S_IS[A-Z].*->d_inode\' |') || die "Can't grep for S_ISDIR and co. callers"; @callers = <$fd>; close($fd); unless (@callers) { print "No matches\n"; exit(0); } my @cocci = ( '@@', 'expression E;', '@@', '', '- S_ISLNK(E->d_inode->i_mode)', '+ d_is_symlink(E)', '', '@@', 'expression E;', '@@', '', '- S_ISDIR(E->d_inode->i_mode)', '+ d_is_dir(E)', '', '@@', 'expression E;', '@@', '', '- S_ISREG(E->d_inode->i_mode)', '+ d_is_reg(E)' ); my $coccifile = "tmp.sp.cocci"; open($fd, ">$coccifile") || die $coccifile; print($fd "$_\n") || die $coccifile foreach (@cocci); close($fd); foreach my $file (@callers) { chomp $file; print "Processing ", $file, "\n"; system("spatch", "--sp-file", $coccifile, $file, "--in-place", "--no-show-diff") == 0 || die "spatch failed"; } [AV: overlayfs parts skipped] Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-01-29 12:02:35 +00:00
if (d_is_dir(entry)) {
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
shrink_dcache_parent(entry);
if (!simple_empty(entry)) {
err = -ENOTEMPTY;
goto badentry;
}
d_inode(entry)->i_flags |= S_DEAD;
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
}
dont_mount(entry);
clear_nlink(d_inode(entry));
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
err = 0;
badentry:
inode_unlock(d_inode(entry));
FUSE: Notifying the kernel of deletion. Allows a FUSE file-system to tell the kernel when a file or directory is deleted. If the specified dentry has the specified inode number, the kernel will unhash it. The current 'fuse_notify_inval_entry' does not cause the kernel to clean up directories that are in use properly, and as a result the users of those directories see incorrect semantics from the file-system. The error condition seen when 'fuse_notify_inval_entry' is used to notify of a deleted directory is avoided when 'fuse_notify_delete' is used instead. The following scenario demonstrates the difference: 1. User A chdirs into 'testdir' and starts reading 'testfile'. 2. User B rm -rf 'testdir'. 3. User B creates 'testdir'. 4. User C chdirs into 'testdir'. If you run the above within the same machine on any file-system (including fuse file-systems), there is no problem: user C is able to chdir into the new testdir. The old testdir is removed from the dentry tree, but still open by user A. If operations 2 and 3 are performed via the network such that the fuse file-system uses one of the notify functions to tell the kernel that the nodes are gone, then the following error occurs for user C while user A holds the original directory open: muirj@empacher:~> ls /test/testdir ls: cannot access /test/testdir: No such file or directory The issue here is that the kernel still has a dentry for testdir, and so it is requesting the attributes for the old directory, while the file-system is responding that the directory no longer exists. If on the other hand, if the file-system can notify the kernel that the directory is deleted using the new 'fuse_notify_delete' function, then the above ls will find the new directory as expected. Signed-off-by: John Muir <john@jmuir.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-12-06 20:50:06 +00:00
if (!err)
d_delete(entry);
} else {
err = 0;
}
dput(entry);
unlock:
inode_unlock(parent);
iput(parent);
return err;
}
/*
* Calling into a user-controlled filesystem gives the filesystem
* daemon ptrace-like capabilities over the current process. This
* means, that the filesystem daemon is able to record the exact
* filesystem operations performed, and can also control the behavior
* of the requester process in otherwise impossible ways. For example
* it can delay the operation for arbitrary length of time allowing
* DoS against the requester.
*
* For this reason only those processes can call into the filesystem,
* for which the owner of the mount has ptrace privilege. This
* excludes processes started by other users, suid or sgid processes.
*/
int fuse_allow_current_process(struct fuse_conn *fc)
{
const struct cred *cred;
if (fc->allow_other)
return current_in_userns(fc->user_ns);
cred = current_cred();
userns: Support fuse interacting with multiple user namespaces Use kuid_t and kgid_t in struct fuse_conn and struct fuse_mount_data. The connection between between a fuse filesystem and a fuse daemon is established when a fuse filesystem is mounted and provided with a file descriptor the fuse daemon created by opening /dev/fuse. For now restrict the communication of uids and gids between the fuse filesystem and the fuse daemon to the initial user namespace. Enforce this by verifying the file descriptor passed to the mount of fuse was opened in the initial user namespace. Ensuring the mount happens in the initial user namespace is not necessary as mounts from non-initial user namespaces are not yet allowed. In fuse_req_init_context convert the currrent fsuid and fsgid into the initial user namespace for the request that will be sent to the fuse daemon. In fuse_fill_attr convert the uid and gid passed from the fuse daemon from the initial user namespace into kuids and kgids. In iattr_to_fattr called from fuse_setattr convert kuids and kgids into the uids and gids in the initial user namespace before passing them to the fuse filesystem. In fuse_change_attributes_common called from fuse_dentry_revalidate, fuse_permission, fuse_geattr, and fuse_setattr, and fuse_iget convert the uid and gid from the fuse daemon into a kuid and a kgid to store on the fuse inode. By default fuse mounts are restricted to task whose uid, suid, and euid matches the fuse user_id and whose gid, sgid, and egid matches the fuse group id. Convert the user_id and group_id mount options into kuids and kgids at mount time, and use uid_eq and gid_eq to compare the in fuse_allow_task. Cc: Miklos Szeredi <miklos@szeredi.hu> Acked-by: Serge Hallyn <serge.hallyn@canonical.com> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
2012-02-08 00:26:03 +00:00
if (uid_eq(cred->euid, fc->user_id) &&
uid_eq(cred->suid, fc->user_id) &&
uid_eq(cred->uid, fc->user_id) &&
gid_eq(cred->egid, fc->group_id) &&
gid_eq(cred->sgid, fc->group_id) &&
gid_eq(cred->gid, fc->group_id))
return 1;
return 0;
}
static int fuse_access(struct inode *inode, int mask)
{
struct fuse_conn *fc = get_fuse_conn(inode);
FUSE_ARGS(args);
struct fuse_access_in inarg;
int err;
BUG_ON(mask & MAY_NOT_BLOCK);
if (fc->no_access)
return 0;
memset(&inarg, 0, sizeof(inarg));
inarg.mask = mask & (MAY_READ | MAY_WRITE | MAY_EXEC);
args.opcode = FUSE_ACCESS;
args.nodeid = get_node_id(inode);
args.in_numargs = 1;
args.in_args[0].size = sizeof(inarg);
args.in_args[0].value = &inarg;
err = fuse_simple_request(fc, &args);
if (err == -ENOSYS) {
fc->no_access = 1;
err = 0;
}
return err;
}
static int fuse_perm_getattr(struct inode *inode, int mask)
{
if (mask & MAY_NOT_BLOCK)
return -ECHILD;
forget_all_cached_acls(inode);
return fuse_do_getattr(inode, NULL, NULL);
}
/*
* Check permission. The two basic access models of FUSE are:
*
* 1) Local access checking ('default_permissions' mount option) based
* on file mode. This is the plain old disk filesystem permission
* modell.
*
* 2) "Remote" access checking, where server is responsible for
* checking permission in each inode operation. An exception to this
* is if ->permission() was invoked from sys_access() in which case an
* access request is sent. Execute permission is still checked
* locally based on file mode.
*/
static int fuse_permission(struct inode *inode, int mask)
{
struct fuse_conn *fc = get_fuse_conn(inode);
bool refreshed = false;
int err = 0;
if (!fuse_allow_current_process(fc))
return -EACCES;
/*
* If attributes are needed, refresh them before proceeding
*/
if (fc->default_permissions ||
((mask & MAY_EXEC) && S_ISREG(inode->i_mode))) {
struct fuse_inode *fi = get_fuse_inode(inode);
u32 perm_mask = STATX_MODE | STATX_UID | STATX_GID;
if (perm_mask & READ_ONCE(fi->inval_mask) ||
time_before64(fi->i_time, get_jiffies_64())) {
refreshed = true;
err = fuse_perm_getattr(inode, mask);
if (err)
return err;
}
}
if (fc->default_permissions) {
err = generic_permission(inode, mask);
/* If permission is denied, try to refresh file
attributes. This is also needed, because the root
node will at first have no permissions */
if (err == -EACCES && !refreshed) {
err = fuse_perm_getattr(inode, mask);
if (!err)
err = generic_permission(inode, mask);
}
/* Note: the opposite of the above test does not
exist. So if permissions are revoked this won't be
noticed immediately, only after the attribute
timeout has expired */
} else if (mask & (MAY_ACCESS | MAY_CHDIR)) {
err = fuse_access(inode, mask);
} else if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode)) {
if (!(inode->i_mode & S_IXUGO)) {
if (refreshed)
return -EACCES;
err = fuse_perm_getattr(inode, mask);
if (!err && !(inode->i_mode & S_IXUGO))
return -EACCES;
}
}
return err;
}
static int fuse_readlink_page(struct inode *inode, struct page *page)
{
struct fuse_conn *fc = get_fuse_conn(inode);
struct fuse_page_desc desc = { .length = PAGE_SIZE - 1 };
struct fuse_args_pages ap = {
.num_pages = 1,
.pages = &page,
.descs = &desc,
};
char *link;
ssize_t res;
ap.args.opcode = FUSE_READLINK;
ap.args.nodeid = get_node_id(inode);
ap.args.out_pages = true;
ap.args.out_argvar = true;
ap.args.page_zeroing = true;
ap.args.out_numargs = 1;
ap.args.out_args[0].size = desc.length;
res = fuse_simple_request(fc, &ap.args);
fuse_invalidate_atime(inode);
if (res < 0)
return res;
if (WARN_ON(res >= PAGE_SIZE))
return -EIO;
link = page_address(page);
link[res] = '\0';
return 0;
}
static const char *fuse_get_link(struct dentry *dentry, struct inode *inode,
struct delayed_call *callback)
{
struct fuse_conn *fc = get_fuse_conn(inode);
struct page *page;
int err;
err = -EIO;
if (is_bad_inode(inode))
goto out_err;
if (fc->cache_symlinks)
return page_get_link(dentry, inode, callback);
err = -ECHILD;
if (!dentry)
goto out_err;
page = alloc_page(GFP_KERNEL);
err = -ENOMEM;
if (!page)
goto out_err;
err = fuse_readlink_page(inode, page);
if (err) {
__free_page(page);
goto out_err;
}
set_delayed_call(callback, page_put_link, page);
return page_address(page);
out_err:
return ERR_PTR(err);
}
static int fuse_dir_open(struct inode *inode, struct file *file)
{
return fuse_open_common(inode, file, true);
}
static int fuse_dir_release(struct inode *inode, struct file *file)
{
fuse_release_common(file, true);
return 0;
}
static int fuse_dir_fsync(struct file *file, loff_t start, loff_t end,
int datasync)
{
struct inode *inode = file->f_mapping->host;
struct fuse_conn *fc = get_fuse_conn(inode);
int err;
if (is_bad_inode(inode))
return -EIO;
if (fc->no_fsyncdir)
return 0;
inode_lock(inode);
err = fuse_fsync_common(file, start, end, datasync, FUSE_FSYNCDIR);
if (err == -ENOSYS) {
fc->no_fsyncdir = 1;
err = 0;
}
inode_unlock(inode);
return err;
}
static long fuse_dir_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
struct fuse_conn *fc = get_fuse_conn(file->f_mapping->host);
/* FUSE_IOCTL_DIR only supported for API version >= 7.18 */
if (fc->minor < 18)
return -ENOTTY;
return fuse_ioctl_common(file, cmd, arg, FUSE_IOCTL_DIR);
}
static long fuse_dir_compat_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
struct fuse_conn *fc = get_fuse_conn(file->f_mapping->host);
if (fc->minor < 18)
return -ENOTTY;
return fuse_ioctl_common(file, cmd, arg,
FUSE_IOCTL_COMPAT | FUSE_IOCTL_DIR);
}
static bool update_mtime(unsigned ivalid, bool trust_local_mtime)
{
/* Always update if mtime is explicitly set */
if (ivalid & ATTR_MTIME_SET)
return true;
/* Or if kernel i_mtime is the official one */
if (trust_local_mtime)
return true;
/* If it's an open(O_TRUNC) or an ftruncate(), don't update */
if ((ivalid & ATTR_SIZE) && (ivalid & (ATTR_OPEN | ATTR_FILE)))
return false;
/* In all other cases update */
return true;
}
fuse: Support fuse filesystems outside of init_user_ns In order to support mounts from namespaces other than init_user_ns, fuse must translate uids and gids to/from the userns of the process servicing requests on /dev/fuse. This patch does that, with a couple of restrictions on the namespace: - The userns for the fuse connection is fixed to the namespace from which /dev/fuse is opened. - The namespace must be the same as s_user_ns. These restrictions simplify the implementation by avoiding the need to pass around userns references and by allowing fuse to rely on the checks in setattr_prepare for ownership changes. Either restriction could be relaxed in the future if needed. For cuse the userns used is the opener of /dev/cuse. Semantically the cuse support does not appear safe for unprivileged users. Practically the permissions on /dev/cuse only make it accessible to the global root user. If something slips through the cracks in a user namespace the only users who will be able to use the cuse device are those users mapped into the user namespace. Translation in the posix acl is updated to use the uuser namespace of the filesystem. Avoiding cases which might bypass this translation is handled in a following change. This change is stronlgy based on a similar change from Seth Forshee and Dongsu Park. Cc: Seth Forshee <seth.forshee@canonical.com> Cc: Dongsu Park <dongsu@kinvolk.io> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
2018-02-21 17:18:07 +00:00
static void iattr_to_fattr(struct fuse_conn *fc, struct iattr *iattr,
struct fuse_setattr_in *arg, bool trust_local_cmtime)
{
unsigned ivalid = iattr->ia_valid;
if (ivalid & ATTR_MODE)
arg->valid |= FATTR_MODE, arg->mode = iattr->ia_mode;
if (ivalid & ATTR_UID)
fuse: Support fuse filesystems outside of init_user_ns In order to support mounts from namespaces other than init_user_ns, fuse must translate uids and gids to/from the userns of the process servicing requests on /dev/fuse. This patch does that, with a couple of restrictions on the namespace: - The userns for the fuse connection is fixed to the namespace from which /dev/fuse is opened. - The namespace must be the same as s_user_ns. These restrictions simplify the implementation by avoiding the need to pass around userns references and by allowing fuse to rely on the checks in setattr_prepare for ownership changes. Either restriction could be relaxed in the future if needed. For cuse the userns used is the opener of /dev/cuse. Semantically the cuse support does not appear safe for unprivileged users. Practically the permissions on /dev/cuse only make it accessible to the global root user. If something slips through the cracks in a user namespace the only users who will be able to use the cuse device are those users mapped into the user namespace. Translation in the posix acl is updated to use the uuser namespace of the filesystem. Avoiding cases which might bypass this translation is handled in a following change. This change is stronlgy based on a similar change from Seth Forshee and Dongsu Park. Cc: Seth Forshee <seth.forshee@canonical.com> Cc: Dongsu Park <dongsu@kinvolk.io> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
2018-02-21 17:18:07 +00:00
arg->valid |= FATTR_UID, arg->uid = from_kuid(fc->user_ns, iattr->ia_uid);
if (ivalid & ATTR_GID)
fuse: Support fuse filesystems outside of init_user_ns In order to support mounts from namespaces other than init_user_ns, fuse must translate uids and gids to/from the userns of the process servicing requests on /dev/fuse. This patch does that, with a couple of restrictions on the namespace: - The userns for the fuse connection is fixed to the namespace from which /dev/fuse is opened. - The namespace must be the same as s_user_ns. These restrictions simplify the implementation by avoiding the need to pass around userns references and by allowing fuse to rely on the checks in setattr_prepare for ownership changes. Either restriction could be relaxed in the future if needed. For cuse the userns used is the opener of /dev/cuse. Semantically the cuse support does not appear safe for unprivileged users. Practically the permissions on /dev/cuse only make it accessible to the global root user. If something slips through the cracks in a user namespace the only users who will be able to use the cuse device are those users mapped into the user namespace. Translation in the posix acl is updated to use the uuser namespace of the filesystem. Avoiding cases which might bypass this translation is handled in a following change. This change is stronlgy based on a similar change from Seth Forshee and Dongsu Park. Cc: Seth Forshee <seth.forshee@canonical.com> Cc: Dongsu Park <dongsu@kinvolk.io> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
2018-02-21 17:18:07 +00:00
arg->valid |= FATTR_GID, arg->gid = from_kgid(fc->user_ns, iattr->ia_gid);
if (ivalid & ATTR_SIZE)
arg->valid |= FATTR_SIZE, arg->size = iattr->ia_size;
if (ivalid & ATTR_ATIME) {
arg->valid |= FATTR_ATIME;
arg->atime = iattr->ia_atime.tv_sec;
arg->atimensec = iattr->ia_atime.tv_nsec;
if (!(ivalid & ATTR_ATIME_SET))
arg->valid |= FATTR_ATIME_NOW;
}
if ((ivalid & ATTR_MTIME) && update_mtime(ivalid, trust_local_cmtime)) {
arg->valid |= FATTR_MTIME;
arg->mtime = iattr->ia_mtime.tv_sec;
arg->mtimensec = iattr->ia_mtime.tv_nsec;
if (!(ivalid & ATTR_MTIME_SET) && !trust_local_cmtime)
arg->valid |= FATTR_MTIME_NOW;
}
if ((ivalid & ATTR_CTIME) && trust_local_cmtime) {
arg->valid |= FATTR_CTIME;
arg->ctime = iattr->ia_ctime.tv_sec;
arg->ctimensec = iattr->ia_ctime.tv_nsec;
}
}
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
/*
* Prevent concurrent writepages on inode
*
* This is done by adding a negative bias to the inode write counter
* and waiting for all pending writes to finish.
*/
void fuse_set_nowrite(struct inode *inode)
{
struct fuse_inode *fi = get_fuse_inode(inode);
BUG_ON(!inode_is_locked(inode));
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
spin_lock(&fi->lock);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
BUG_ON(fi->writectr < 0);
fi->writectr += FUSE_NOWRITE;
spin_unlock(&fi->lock);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
wait_event(fi->page_waitq, fi->writectr == FUSE_NOWRITE);
}
/*
* Allow writepages on inode
*
* Remove the bias from the writecounter and send any queued
* writepages.
*/
static void __fuse_release_nowrite(struct inode *inode)
{
struct fuse_inode *fi = get_fuse_inode(inode);
BUG_ON(fi->writectr != FUSE_NOWRITE);
fi->writectr = 0;
fuse_flush_writepages(inode);
}
void fuse_release_nowrite(struct inode *inode)
{
struct fuse_inode *fi = get_fuse_inode(inode);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
spin_lock(&fi->lock);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
__fuse_release_nowrite(inode);
spin_unlock(&fi->lock);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
}
static void fuse_setattr_fill(struct fuse_conn *fc, struct fuse_args *args,
struct inode *inode,
struct fuse_setattr_in *inarg_p,
struct fuse_attr_out *outarg_p)
{
args->opcode = FUSE_SETATTR;
args->nodeid = get_node_id(inode);
args->in_numargs = 1;
args->in_args[0].size = sizeof(*inarg_p);
args->in_args[0].value = inarg_p;
args->out_numargs = 1;
args->out_args[0].size = sizeof(*outarg_p);
args->out_args[0].value = outarg_p;
}
/*
* Flush inode->i_mtime to the server
*/
int fuse_flush_times(struct inode *inode, struct fuse_file *ff)
{
struct fuse_conn *fc = get_fuse_conn(inode);
FUSE_ARGS(args);
struct fuse_setattr_in inarg;
struct fuse_attr_out outarg;
memset(&inarg, 0, sizeof(inarg));
memset(&outarg, 0, sizeof(outarg));
inarg.valid = FATTR_MTIME;
inarg.mtime = inode->i_mtime.tv_sec;
inarg.mtimensec = inode->i_mtime.tv_nsec;
if (fc->minor >= 23) {
inarg.valid |= FATTR_CTIME;
inarg.ctime = inode->i_ctime.tv_sec;
inarg.ctimensec = inode->i_ctime.tv_nsec;
}
if (ff) {
inarg.valid |= FATTR_FH;
inarg.fh = ff->fh;
}
fuse_setattr_fill(fc, &args, inode, &inarg, &outarg);
return fuse_simple_request(fc, &args);
}
/*
* Set attributes, and at the same time refresh them.
*
* Truncation is slightly complicated, because the 'truncate' request
* may fail, in which case we don't want to touch the mapping.
* vmtruncate() doesn't allow for this case, so do the rlimit checking
* and the actual truncation by hand.
*/
int fuse_do_setattr(struct dentry *dentry, struct iattr *attr,
struct file *file)
{
struct inode *inode = d_inode(dentry);
struct fuse_conn *fc = get_fuse_conn(inode);
fuse: hotfix truncate_pagecache() issue The way how fuse calls truncate_pagecache() from fuse_change_attributes() is completely wrong. Because, w/o i_mutex held, we never sure whether 'oldsize' and 'attr->size' are valid by the time of execution of truncate_pagecache(inode, oldsize, attr->size). In fact, as soon as we released fc->lock in the middle of fuse_change_attributes(), we completely loose control of actions which may happen with given inode until we reach truncate_pagecache. The list of potentially dangerous actions includes mmap-ed reads and writes, ftruncate(2) and write(2) extending file size. The typical outcome of doing truncate_pagecache() with outdated arguments is data corruption from user point of view. This is (in some sense) acceptable in cases when the issue is triggered by a change of the file on the server (i.e. externally wrt fuse operation), but it is absolutely intolerable in scenarios when a single fuse client modifies a file without any external intervention. A real life case I discovered by fsx-linux looked like this: 1. Shrinking ftruncate(2) comes to fuse_do_setattr(). The latter sends FUSE_SETATTR to the server synchronously, but before getting fc->lock ... 2. fuse_dentry_revalidate() is asynchronously called. It sends FUSE_LOOKUP to the server synchronously, then calls fuse_change_attributes(). The latter updates i_size, releases fc->lock, but before comparing oldsize vs attr->size.. 3. fuse_do_setattr() from the first step proceeds by acquiring fc->lock and updating attributes and i_size, but now oldsize is equal to outarg.attr.size because i_size has just been updated (step 2). Hence, fuse_do_setattr() returns w/o calling truncate_pagecache(). 4. As soon as ftruncate(2) completes, the user extends file size by write(2) making a hole in the middle of file, then reads data from the hole either by read(2) or mmap-ed read. The user expects to get zero data from the hole, but gets stale data because truncate_pagecache() is not executed yet. The scenario above illustrates one side of the problem: not truncating the page cache even though we should. Another side corresponds to truncating page cache too late, when the state of inode changed significantly. Theoretically, the following is possible: 1. As in the previous scenario fuse_dentry_revalidate() discovered that i_size changed (due to our own fuse_do_setattr()) and is going to call truncate_pagecache() for some 'new_size' it believes valid right now. But by the time that particular truncate_pagecache() is called ... 2. fuse_do_setattr() returns (either having called truncate_pagecache() or not -- it doesn't matter). 3. The file is extended either by write(2) or ftruncate(2) or fallocate(2). 4. mmap-ed write makes a page in the extended region dirty. The result will be the lost of data user wrote on the fourth step. The patch is a hotfix resolving the issue in a simplistic way: let's skip dangerous i_size update and truncate_pagecache if an operation changing file size is in progress. This simplistic approach looks correct for the cases w/o external changes. And to handle them properly, more sophisticated and intrusive techniques (e.g. NFS-like one) would be required. I'd like to postpone it until the issue is well discussed on the mailing list(s). Changed in v2: - improved patch description to cover both sides of the issue. Signed-off-by: Maxim Patlasov <mpatlasov@parallels.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: stable@vger.kernel.org
2013-08-30 13:06:04 +00:00
struct fuse_inode *fi = get_fuse_inode(inode);
FUSE_ARGS(args);
struct fuse_setattr_in inarg;
struct fuse_attr_out outarg;
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
bool is_truncate = false;
bool is_wb = fc->writeback_cache;
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
loff_t oldsize;
int err;
bool trust_local_cmtime = is_wb && S_ISREG(inode->i_mode);
if (!fc->default_permissions)
attr->ia_valid |= ATTR_FORCE;
err = setattr_prepare(dentry, attr);
if (err)
return err;
if (attr->ia_valid & ATTR_OPEN) {
/* This is coming from open(..., ... | O_TRUNC); */
WARN_ON(!(attr->ia_valid & ATTR_SIZE));
WARN_ON(attr->ia_size != 0);
if (fc->atomic_o_trunc) {
/*
* No need to send request to userspace, since actual
* truncation has already been done by OPEN. But still
* need to truncate page cache.
*/
i_size_write(inode, 0);
truncate_pagecache(inode, 0);
return 0;
}
file = NULL;
}
if (attr->ia_valid & ATTR_SIZE) {
if (WARN_ON(!S_ISREG(inode->i_mode)))
return -EIO;
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
is_truncate = true;
}
/* Flush dirty data/metadata before non-truncate SETATTR */
if (is_wb && S_ISREG(inode->i_mode) &&
attr->ia_valid &
(ATTR_MODE | ATTR_UID | ATTR_GID | ATTR_MTIME_SET |
ATTR_TIMES_SET)) {
err = write_inode_now(inode, true);
if (err)
return err;
fuse_set_nowrite(inode);
fuse_release_nowrite(inode);
}
fuse: hotfix truncate_pagecache() issue The way how fuse calls truncate_pagecache() from fuse_change_attributes() is completely wrong. Because, w/o i_mutex held, we never sure whether 'oldsize' and 'attr->size' are valid by the time of execution of truncate_pagecache(inode, oldsize, attr->size). In fact, as soon as we released fc->lock in the middle of fuse_change_attributes(), we completely loose control of actions which may happen with given inode until we reach truncate_pagecache. The list of potentially dangerous actions includes mmap-ed reads and writes, ftruncate(2) and write(2) extending file size. The typical outcome of doing truncate_pagecache() with outdated arguments is data corruption from user point of view. This is (in some sense) acceptable in cases when the issue is triggered by a change of the file on the server (i.e. externally wrt fuse operation), but it is absolutely intolerable in scenarios when a single fuse client modifies a file without any external intervention. A real life case I discovered by fsx-linux looked like this: 1. Shrinking ftruncate(2) comes to fuse_do_setattr(). The latter sends FUSE_SETATTR to the server synchronously, but before getting fc->lock ... 2. fuse_dentry_revalidate() is asynchronously called. It sends FUSE_LOOKUP to the server synchronously, then calls fuse_change_attributes(). The latter updates i_size, releases fc->lock, but before comparing oldsize vs attr->size.. 3. fuse_do_setattr() from the first step proceeds by acquiring fc->lock and updating attributes and i_size, but now oldsize is equal to outarg.attr.size because i_size has just been updated (step 2). Hence, fuse_do_setattr() returns w/o calling truncate_pagecache(). 4. As soon as ftruncate(2) completes, the user extends file size by write(2) making a hole in the middle of file, then reads data from the hole either by read(2) or mmap-ed read. The user expects to get zero data from the hole, but gets stale data because truncate_pagecache() is not executed yet. The scenario above illustrates one side of the problem: not truncating the page cache even though we should. Another side corresponds to truncating page cache too late, when the state of inode changed significantly. Theoretically, the following is possible: 1. As in the previous scenario fuse_dentry_revalidate() discovered that i_size changed (due to our own fuse_do_setattr()) and is going to call truncate_pagecache() for some 'new_size' it believes valid right now. But by the time that particular truncate_pagecache() is called ... 2. fuse_do_setattr() returns (either having called truncate_pagecache() or not -- it doesn't matter). 3. The file is extended either by write(2) or ftruncate(2) or fallocate(2). 4. mmap-ed write makes a page in the extended region dirty. The result will be the lost of data user wrote on the fourth step. The patch is a hotfix resolving the issue in a simplistic way: let's skip dangerous i_size update and truncate_pagecache if an operation changing file size is in progress. This simplistic approach looks correct for the cases w/o external changes. And to handle them properly, more sophisticated and intrusive techniques (e.g. NFS-like one) would be required. I'd like to postpone it until the issue is well discussed on the mailing list(s). Changed in v2: - improved patch description to cover both sides of the issue. Signed-off-by: Maxim Patlasov <mpatlasov@parallels.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: stable@vger.kernel.org
2013-08-30 13:06:04 +00:00
if (is_truncate) {
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
fuse_set_nowrite(inode);
fuse: hotfix truncate_pagecache() issue The way how fuse calls truncate_pagecache() from fuse_change_attributes() is completely wrong. Because, w/o i_mutex held, we never sure whether 'oldsize' and 'attr->size' are valid by the time of execution of truncate_pagecache(inode, oldsize, attr->size). In fact, as soon as we released fc->lock in the middle of fuse_change_attributes(), we completely loose control of actions which may happen with given inode until we reach truncate_pagecache. The list of potentially dangerous actions includes mmap-ed reads and writes, ftruncate(2) and write(2) extending file size. The typical outcome of doing truncate_pagecache() with outdated arguments is data corruption from user point of view. This is (in some sense) acceptable in cases when the issue is triggered by a change of the file on the server (i.e. externally wrt fuse operation), but it is absolutely intolerable in scenarios when a single fuse client modifies a file without any external intervention. A real life case I discovered by fsx-linux looked like this: 1. Shrinking ftruncate(2) comes to fuse_do_setattr(). The latter sends FUSE_SETATTR to the server synchronously, but before getting fc->lock ... 2. fuse_dentry_revalidate() is asynchronously called. It sends FUSE_LOOKUP to the server synchronously, then calls fuse_change_attributes(). The latter updates i_size, releases fc->lock, but before comparing oldsize vs attr->size.. 3. fuse_do_setattr() from the first step proceeds by acquiring fc->lock and updating attributes and i_size, but now oldsize is equal to outarg.attr.size because i_size has just been updated (step 2). Hence, fuse_do_setattr() returns w/o calling truncate_pagecache(). 4. As soon as ftruncate(2) completes, the user extends file size by write(2) making a hole in the middle of file, then reads data from the hole either by read(2) or mmap-ed read. The user expects to get zero data from the hole, but gets stale data because truncate_pagecache() is not executed yet. The scenario above illustrates one side of the problem: not truncating the page cache even though we should. Another side corresponds to truncating page cache too late, when the state of inode changed significantly. Theoretically, the following is possible: 1. As in the previous scenario fuse_dentry_revalidate() discovered that i_size changed (due to our own fuse_do_setattr()) and is going to call truncate_pagecache() for some 'new_size' it believes valid right now. But by the time that particular truncate_pagecache() is called ... 2. fuse_do_setattr() returns (either having called truncate_pagecache() or not -- it doesn't matter). 3. The file is extended either by write(2) or ftruncate(2) or fallocate(2). 4. mmap-ed write makes a page in the extended region dirty. The result will be the lost of data user wrote on the fourth step. The patch is a hotfix resolving the issue in a simplistic way: let's skip dangerous i_size update and truncate_pagecache if an operation changing file size is in progress. This simplistic approach looks correct for the cases w/o external changes. And to handle them properly, more sophisticated and intrusive techniques (e.g. NFS-like one) would be required. I'd like to postpone it until the issue is well discussed on the mailing list(s). Changed in v2: - improved patch description to cover both sides of the issue. Signed-off-by: Maxim Patlasov <mpatlasov@parallels.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: stable@vger.kernel.org
2013-08-30 13:06:04 +00:00
set_bit(FUSE_I_SIZE_UNSTABLE, &fi->state);
if (trust_local_cmtime && attr->ia_size != inode->i_size)
attr->ia_valid |= ATTR_MTIME | ATTR_CTIME;
fuse: hotfix truncate_pagecache() issue The way how fuse calls truncate_pagecache() from fuse_change_attributes() is completely wrong. Because, w/o i_mutex held, we never sure whether 'oldsize' and 'attr->size' are valid by the time of execution of truncate_pagecache(inode, oldsize, attr->size). In fact, as soon as we released fc->lock in the middle of fuse_change_attributes(), we completely loose control of actions which may happen with given inode until we reach truncate_pagecache. The list of potentially dangerous actions includes mmap-ed reads and writes, ftruncate(2) and write(2) extending file size. The typical outcome of doing truncate_pagecache() with outdated arguments is data corruption from user point of view. This is (in some sense) acceptable in cases when the issue is triggered by a change of the file on the server (i.e. externally wrt fuse operation), but it is absolutely intolerable in scenarios when a single fuse client modifies a file without any external intervention. A real life case I discovered by fsx-linux looked like this: 1. Shrinking ftruncate(2) comes to fuse_do_setattr(). The latter sends FUSE_SETATTR to the server synchronously, but before getting fc->lock ... 2. fuse_dentry_revalidate() is asynchronously called. It sends FUSE_LOOKUP to the server synchronously, then calls fuse_change_attributes(). The latter updates i_size, releases fc->lock, but before comparing oldsize vs attr->size.. 3. fuse_do_setattr() from the first step proceeds by acquiring fc->lock and updating attributes and i_size, but now oldsize is equal to outarg.attr.size because i_size has just been updated (step 2). Hence, fuse_do_setattr() returns w/o calling truncate_pagecache(). 4. As soon as ftruncate(2) completes, the user extends file size by write(2) making a hole in the middle of file, then reads data from the hole either by read(2) or mmap-ed read. The user expects to get zero data from the hole, but gets stale data because truncate_pagecache() is not executed yet. The scenario above illustrates one side of the problem: not truncating the page cache even though we should. Another side corresponds to truncating page cache too late, when the state of inode changed significantly. Theoretically, the following is possible: 1. As in the previous scenario fuse_dentry_revalidate() discovered that i_size changed (due to our own fuse_do_setattr()) and is going to call truncate_pagecache() for some 'new_size' it believes valid right now. But by the time that particular truncate_pagecache() is called ... 2. fuse_do_setattr() returns (either having called truncate_pagecache() or not -- it doesn't matter). 3. The file is extended either by write(2) or ftruncate(2) or fallocate(2). 4. mmap-ed write makes a page in the extended region dirty. The result will be the lost of data user wrote on the fourth step. The patch is a hotfix resolving the issue in a simplistic way: let's skip dangerous i_size update and truncate_pagecache if an operation changing file size is in progress. This simplistic approach looks correct for the cases w/o external changes. And to handle them properly, more sophisticated and intrusive techniques (e.g. NFS-like one) would be required. I'd like to postpone it until the issue is well discussed on the mailing list(s). Changed in v2: - improved patch description to cover both sides of the issue. Signed-off-by: Maxim Patlasov <mpatlasov@parallels.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: stable@vger.kernel.org
2013-08-30 13:06:04 +00:00
}
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
memset(&inarg, 0, sizeof(inarg));
memset(&outarg, 0, sizeof(outarg));
fuse: Support fuse filesystems outside of init_user_ns In order to support mounts from namespaces other than init_user_ns, fuse must translate uids and gids to/from the userns of the process servicing requests on /dev/fuse. This patch does that, with a couple of restrictions on the namespace: - The userns for the fuse connection is fixed to the namespace from which /dev/fuse is opened. - The namespace must be the same as s_user_ns. These restrictions simplify the implementation by avoiding the need to pass around userns references and by allowing fuse to rely on the checks in setattr_prepare for ownership changes. Either restriction could be relaxed in the future if needed. For cuse the userns used is the opener of /dev/cuse. Semantically the cuse support does not appear safe for unprivileged users. Practically the permissions on /dev/cuse only make it accessible to the global root user. If something slips through the cracks in a user namespace the only users who will be able to use the cuse device are those users mapped into the user namespace. Translation in the posix acl is updated to use the uuser namespace of the filesystem. Avoiding cases which might bypass this translation is handled in a following change. This change is stronlgy based on a similar change from Seth Forshee and Dongsu Park. Cc: Seth Forshee <seth.forshee@canonical.com> Cc: Dongsu Park <dongsu@kinvolk.io> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
2018-02-21 17:18:07 +00:00
iattr_to_fattr(fc, attr, &inarg, trust_local_cmtime);
if (file) {
struct fuse_file *ff = file->private_data;
inarg.valid |= FATTR_FH;
inarg.fh = ff->fh;
}
if (attr->ia_valid & ATTR_SIZE) {
/* For mandatory locking in truncate */
inarg.valid |= FATTR_LOCKOWNER;
inarg.lock_owner = fuse_lock_owner_id(fc, current->files);
}
fuse_setattr_fill(fc, &args, inode, &inarg, &outarg);
err = fuse_simple_request(fc, &args);
if (err) {
if (err == -EINTR)
fuse_invalidate_attr(inode);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
goto error;
}
if (fuse_invalid_attr(&outarg.attr) ||
(inode->i_mode ^ outarg.attr.mode) & S_IFMT) {
make_bad_inode(inode);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
err = -EIO;
goto error;
}
spin_lock(&fi->lock);
/* the kernel maintains i_mtime locally */
if (trust_local_cmtime) {
if (attr->ia_valid & ATTR_MTIME)
inode->i_mtime = attr->ia_mtime;
if (attr->ia_valid & ATTR_CTIME)
inode->i_ctime = attr->ia_ctime;
/* FIXME: clear I_DIRTY_SYNC? */
}
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
fuse_change_attributes_common(inode, &outarg.attr,
attr_timeout(&outarg));
oldsize = inode->i_size;
/* see the comment in fuse_change_attributes() */
if (!is_wb || is_truncate || !S_ISREG(inode->i_mode))
i_size_write(inode, outarg.attr.size);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
if (is_truncate) {
/* NOTE: this may release/reacquire fi->lock */
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
__fuse_release_nowrite(inode);
}
spin_unlock(&fi->lock);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
/*
* Only call invalidate_inode_pages2() after removing
* FUSE_NOWRITE, otherwise fuse_launder_page() would deadlock.
*/
if ((is_truncate || !is_wb) &&
S_ISREG(inode->i_mode) && oldsize != outarg.attr.size) {
truncate_pagecache(inode, outarg.attr.size);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
invalidate_inode_pages2(inode->i_mapping);
}
fuse: hotfix truncate_pagecache() issue The way how fuse calls truncate_pagecache() from fuse_change_attributes() is completely wrong. Because, w/o i_mutex held, we never sure whether 'oldsize' and 'attr->size' are valid by the time of execution of truncate_pagecache(inode, oldsize, attr->size). In fact, as soon as we released fc->lock in the middle of fuse_change_attributes(), we completely loose control of actions which may happen with given inode until we reach truncate_pagecache. The list of potentially dangerous actions includes mmap-ed reads and writes, ftruncate(2) and write(2) extending file size. The typical outcome of doing truncate_pagecache() with outdated arguments is data corruption from user point of view. This is (in some sense) acceptable in cases when the issue is triggered by a change of the file on the server (i.e. externally wrt fuse operation), but it is absolutely intolerable in scenarios when a single fuse client modifies a file without any external intervention. A real life case I discovered by fsx-linux looked like this: 1. Shrinking ftruncate(2) comes to fuse_do_setattr(). The latter sends FUSE_SETATTR to the server synchronously, but before getting fc->lock ... 2. fuse_dentry_revalidate() is asynchronously called. It sends FUSE_LOOKUP to the server synchronously, then calls fuse_change_attributes(). The latter updates i_size, releases fc->lock, but before comparing oldsize vs attr->size.. 3. fuse_do_setattr() from the first step proceeds by acquiring fc->lock and updating attributes and i_size, but now oldsize is equal to outarg.attr.size because i_size has just been updated (step 2). Hence, fuse_do_setattr() returns w/o calling truncate_pagecache(). 4. As soon as ftruncate(2) completes, the user extends file size by write(2) making a hole in the middle of file, then reads data from the hole either by read(2) or mmap-ed read. The user expects to get zero data from the hole, but gets stale data because truncate_pagecache() is not executed yet. The scenario above illustrates one side of the problem: not truncating the page cache even though we should. Another side corresponds to truncating page cache too late, when the state of inode changed significantly. Theoretically, the following is possible: 1. As in the previous scenario fuse_dentry_revalidate() discovered that i_size changed (due to our own fuse_do_setattr()) and is going to call truncate_pagecache() for some 'new_size' it believes valid right now. But by the time that particular truncate_pagecache() is called ... 2. fuse_do_setattr() returns (either having called truncate_pagecache() or not -- it doesn't matter). 3. The file is extended either by write(2) or ftruncate(2) or fallocate(2). 4. mmap-ed write makes a page in the extended region dirty. The result will be the lost of data user wrote on the fourth step. The patch is a hotfix resolving the issue in a simplistic way: let's skip dangerous i_size update and truncate_pagecache if an operation changing file size is in progress. This simplistic approach looks correct for the cases w/o external changes. And to handle them properly, more sophisticated and intrusive techniques (e.g. NFS-like one) would be required. I'd like to postpone it until the issue is well discussed on the mailing list(s). Changed in v2: - improved patch description to cover both sides of the issue. Signed-off-by: Maxim Patlasov <mpatlasov@parallels.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: stable@vger.kernel.org
2013-08-30 13:06:04 +00:00
clear_bit(FUSE_I_SIZE_UNSTABLE, &fi->state);
return 0;
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
error:
if (is_truncate)
fuse_release_nowrite(inode);
fuse: hotfix truncate_pagecache() issue The way how fuse calls truncate_pagecache() from fuse_change_attributes() is completely wrong. Because, w/o i_mutex held, we never sure whether 'oldsize' and 'attr->size' are valid by the time of execution of truncate_pagecache(inode, oldsize, attr->size). In fact, as soon as we released fc->lock in the middle of fuse_change_attributes(), we completely loose control of actions which may happen with given inode until we reach truncate_pagecache. The list of potentially dangerous actions includes mmap-ed reads and writes, ftruncate(2) and write(2) extending file size. The typical outcome of doing truncate_pagecache() with outdated arguments is data corruption from user point of view. This is (in some sense) acceptable in cases when the issue is triggered by a change of the file on the server (i.e. externally wrt fuse operation), but it is absolutely intolerable in scenarios when a single fuse client modifies a file without any external intervention. A real life case I discovered by fsx-linux looked like this: 1. Shrinking ftruncate(2) comes to fuse_do_setattr(). The latter sends FUSE_SETATTR to the server synchronously, but before getting fc->lock ... 2. fuse_dentry_revalidate() is asynchronously called. It sends FUSE_LOOKUP to the server synchronously, then calls fuse_change_attributes(). The latter updates i_size, releases fc->lock, but before comparing oldsize vs attr->size.. 3. fuse_do_setattr() from the first step proceeds by acquiring fc->lock and updating attributes and i_size, but now oldsize is equal to outarg.attr.size because i_size has just been updated (step 2). Hence, fuse_do_setattr() returns w/o calling truncate_pagecache(). 4. As soon as ftruncate(2) completes, the user extends file size by write(2) making a hole in the middle of file, then reads data from the hole either by read(2) or mmap-ed read. The user expects to get zero data from the hole, but gets stale data because truncate_pagecache() is not executed yet. The scenario above illustrates one side of the problem: not truncating the page cache even though we should. Another side corresponds to truncating page cache too late, when the state of inode changed significantly. Theoretically, the following is possible: 1. As in the previous scenario fuse_dentry_revalidate() discovered that i_size changed (due to our own fuse_do_setattr()) and is going to call truncate_pagecache() for some 'new_size' it believes valid right now. But by the time that particular truncate_pagecache() is called ... 2. fuse_do_setattr() returns (either having called truncate_pagecache() or not -- it doesn't matter). 3. The file is extended either by write(2) or ftruncate(2) or fallocate(2). 4. mmap-ed write makes a page in the extended region dirty. The result will be the lost of data user wrote on the fourth step. The patch is a hotfix resolving the issue in a simplistic way: let's skip dangerous i_size update and truncate_pagecache if an operation changing file size is in progress. This simplistic approach looks correct for the cases w/o external changes. And to handle them properly, more sophisticated and intrusive techniques (e.g. NFS-like one) would be required. I'd like to postpone it until the issue is well discussed on the mailing list(s). Changed in v2: - improved patch description to cover both sides of the issue. Signed-off-by: Maxim Patlasov <mpatlasov@parallels.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: stable@vger.kernel.org
2013-08-30 13:06:04 +00:00
clear_bit(FUSE_I_SIZE_UNSTABLE, &fi->state);
fuse: support writable mmap Quoting Linus (3 years ago, FUSE inclusion discussions): "User-space filesystems are hard to get right. I'd claim that they are almost impossible, unless you limit them somehow (shared writable mappings are the nastiest part - if you don't have those, you can reasonably limit your problems by limiting the number of dirty pages you accept through normal "write()" calls)." Instead of attempting the impossible, I've just waited for the dirty page accounting infrastructure to materialize (thanks to Peter Zijlstra and others). This nicely solved the biggest problem: limiting the number of pages used for write caching. Some small details remained, however, which this largish patch attempts to address. It provides a page writeback implementation for fuse, which is completely safe against VM related deadlocks. Performance may not be very good for certain usage patterns, but generally it should be acceptable. It has been tested extensively with fsx-linux and bash-shared-mapping. Fuse page writeback design -------------------------- fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM. It copies the contents of the original page, and queues a WRITE request to the userspace filesystem using this temp page. The writeback is finished instantly from the MM's point of view: the page is removed from the radix trees, and the PageDirty and PageWriteback flags are cleared. For the duration of the actual write, the NR_WRITEBACK_TEMP counter is incremented. The per-bdi writeback count is not decremented until the actual write completes. On dirtying the page, fuse waits for a previous write to finish before proceeding. This makes sure, there can only be one temporary page used at a time for one cached page. This approach is wasteful in both memory and CPU bandwidth, so why is this complication needed? The basic problem is that there can be no guarantee about the time in which the userspace filesystem will complete a write. It may be buggy or even malicious, and fail to complete WRITE requests. We don't want unrelated parts of the system to grind to a halt in such cases. Also a filesystem may need additional resources (particularly memory) to complete a WRITE request. There's a great danger of a deadlock if that allocation may wait for the writepage to finish. Currently there are several cases where the kernel can block on page writeback: - allocation order is larger than PAGE_ALLOC_COSTLY_ORDER - page migration - throttle_vm_writeout (through NR_WRITEBACK) - sync(2) Of course in some cases (fsync, msync) we explicitly want to allow blocking. So for these cases new code has to be added to fuse, since the VM is not tracking writeback pages for us any more. As an extra safetly measure, the maximum dirty ratio allocated to a single fuse filesystem is set to 1% by default. This way one (or several) buggy or malicious fuse filesystems cannot slow down the rest of the system by hogging dirty memory. With appropriate privileges, this limit can be raised through '/sys/class/bdi/<bdi>/max_ratio'. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-30 07:54:41 +00:00
return err;
}
static int fuse_setattr(struct dentry *entry, struct iattr *attr)
{
struct inode *inode = d_inode(entry);
struct fuse_conn *fc = get_fuse_conn(inode);
struct file *file = (attr->ia_valid & ATTR_FILE) ? attr->ia_file : NULL;
int ret;
if (!fuse_allow_current_process(get_fuse_conn(inode)))
return -EACCES;
if (attr->ia_valid & (ATTR_KILL_SUID | ATTR_KILL_SGID)) {
attr->ia_valid &= ~(ATTR_KILL_SUID | ATTR_KILL_SGID |
ATTR_MODE);
/*
* The only sane way to reliably kill suid/sgid is to do it in
* the userspace filesystem
*
* This should be done on write(), truncate() and chown().
*/
if (!fc->handle_killpriv) {
/*
* ia_mode calculation may have used stale i_mode.
* Refresh and recalculate.
*/
ret = fuse_do_getattr(inode, NULL, file);
if (ret)
return ret;
attr->ia_mode = inode->i_mode;
if (inode->i_mode & S_ISUID) {
attr->ia_valid |= ATTR_MODE;
attr->ia_mode &= ~S_ISUID;
}
if ((inode->i_mode & (S_ISGID | S_IXGRP)) == (S_ISGID | S_IXGRP)) {
attr->ia_valid |= ATTR_MODE;
attr->ia_mode &= ~S_ISGID;
}
}
}
if (!attr->ia_valid)
return 0;
ret = fuse_do_setattr(entry, attr, file);
if (!ret) {
/*
* If filesystem supports acls it may have updated acl xattrs in
* the filesystem, so forget cached acls for the inode.
*/
if (fc->posix_acl)
forget_all_cached_acls(inode);
/* Directory mode changed, may need to revalidate access */
if (d_is_dir(entry) && (attr->ia_valid & ATTR_MODE))
fuse_invalidate_entry_cache(entry);
}
return ret;
}
statx: Add a system call to make enhanced file info available Add a system call to make extended file information available, including file creation and some attribute flags where available through the underlying filesystem. The getattr inode operation is altered to take two additional arguments: a u32 request_mask and an unsigned int flags that indicate the synchronisation mode. This change is propagated to the vfs_getattr*() function. Functions like vfs_stat() are now inline wrappers around new functions vfs_statx() and vfs_statx_fd() to reduce stack usage. ======== OVERVIEW ======== The idea was initially proposed as a set of xattrs that could be retrieved with getxattr(), but the general preference proved to be for a new syscall with an extended stat structure. A number of requests were gathered for features to be included. The following have been included: (1) Make the fields a consistent size on all arches and make them large. (2) Spare space, request flags and information flags are provided for future expansion. (3) Better support for the y2038 problem [Arnd Bergmann] (tv_sec is an __s64). (4) Creation time: The SMB protocol carries the creation time, which could be exported by Samba, which will in turn help CIFS make use of FS-Cache as that can be used for coherency data (stx_btime). This is also specified in NFSv4 as a recommended attribute and could be exported by NFSD [Steve French]. (5) Lightweight stat: Ask for just those details of interest, and allow a netfs (such as NFS) to approximate anything not of interest, possibly without going to the server [Trond Myklebust, Ulrich Drepper, Andreas Dilger] (AT_STATX_DONT_SYNC). (6) Heavyweight stat: Force a netfs to go to the server, even if it thinks its cached attributes are up to date [Trond Myklebust] (AT_STATX_FORCE_SYNC). And the following have been left out for future extension: (7) Data version number: Could be used by userspace NFS servers [Aneesh Kumar]. Can also be used to modify fill_post_wcc() in NFSD which retrieves i_version directly, but has just called vfs_getattr(). It could get it from the kstat struct if it used vfs_xgetattr() instead. (There's disagreement on the exact semantics of a single field, since not all filesystems do this the same way). (8) BSD stat compatibility: Including more fields from the BSD stat such as creation time (st_btime) and inode generation number (st_gen) [Jeremy Allison, Bernd Schubert]. (9) Inode generation number: Useful for FUSE and userspace NFS servers [Bernd Schubert]. (This was asked for but later deemed unnecessary with the open-by-handle capability available and caused disagreement as to whether it's a security hole or not). (10) Extra coherency data may be useful in making backups [Andreas Dilger]. (No particular data were offered, but things like last backup timestamp, the data version number and the DOS archive bit would come into this category). (11) Allow the filesystem to indicate what it can/cannot provide: A filesystem can now say it doesn't support a standard stat feature if that isn't available, so if, for instance, inode numbers or UIDs don't exist or are fabricated locally... (This requires a separate system call - I have an fsinfo() call idea for this). (12) Store a 16-byte volume ID in the superblock that can be returned in struct xstat [Steve French]. (Deferred to fsinfo). (13) Include granularity fields in the time data to indicate the granularity of each of the times (NFSv4 time_delta) [Steve French]. (Deferred to fsinfo). (14) FS_IOC_GETFLAGS value. These could be translated to BSD's st_flags. Note that the Linux IOC flags are a mess and filesystems such as Ext4 define flags that aren't in linux/fs.h, so translation in the kernel may be a necessity (or, possibly, we provide the filesystem type too). (Some attributes are made available in stx_attributes, but the general feeling was that the IOC flags were to ext[234]-specific and shouldn't be exposed through statx this way). (15) Mask of features available on file (eg: ACLs, seclabel) [Brad Boyer, Michael Kerrisk]. (Deferred, probably to fsinfo. Finding out if there's an ACL or seclabal might require extra filesystem operations). (16) Femtosecond-resolution timestamps [Dave Chinner]. (A __reserved field has been left in the statx_timestamp struct for this - if there proves to be a need). (17) A set multiple attributes syscall to go with this. =============== NEW SYSTEM CALL =============== The new system call is: int ret = statx(int dfd, const char *filename, unsigned int flags, unsigned int mask, struct statx *buffer); The dfd, filename and flags parameters indicate the file to query, in a similar way to fstatat(). There is no equivalent of lstat() as that can be emulated with statx() by passing AT_SYMLINK_NOFOLLOW in flags. There is also no equivalent of fstat() as that can be emulated by passing a NULL filename to statx() with the fd of interest in dfd. Whether or not statx() synchronises the attributes with the backing store can be controlled by OR'ing a value into the flags argument (this typically only affects network filesystems): (1) AT_STATX_SYNC_AS_STAT tells statx() to behave as stat() does in this respect. (2) AT_STATX_FORCE_SYNC will require a network filesystem to synchronise its attributes with the server - which might require data writeback to occur to get the timestamps correct. (3) AT_STATX_DONT_SYNC will suppress synchronisation with the server in a network filesystem. The resulting values should be considered approximate. mask is a bitmask indicating the fields in struct statx that are of interest to the caller. The user should set this to STATX_BASIC_STATS to get the basic set returned by stat(). It should be noted that asking for more information may entail extra I/O operations. buffer points to the destination for the data. This must be 256 bytes in size. ====================== MAIN ATTRIBUTES RECORD ====================== The following structures are defined in which to return the main attribute set: struct statx_timestamp { __s64 tv_sec; __s32 tv_nsec; __s32 __reserved; }; struct statx { __u32 stx_mask; __u32 stx_blksize; __u64 stx_attributes; __u32 stx_nlink; __u32 stx_uid; __u32 stx_gid; __u16 stx_mode; __u16 __spare0[1]; __u64 stx_ino; __u64 stx_size; __u64 stx_blocks; __u64 __spare1[1]; struct statx_timestamp stx_atime; struct statx_timestamp stx_btime; struct statx_timestamp stx_ctime; struct statx_timestamp stx_mtime; __u32 stx_rdev_major; __u32 stx_rdev_minor; __u32 stx_dev_major; __u32 stx_dev_minor; __u64 __spare2[14]; }; The defined bits in request_mask and stx_mask are: STATX_TYPE Want/got stx_mode & S_IFMT STATX_MODE Want/got stx_mode & ~S_IFMT STATX_NLINK Want/got stx_nlink STATX_UID Want/got stx_uid STATX_GID Want/got stx_gid STATX_ATIME Want/got stx_atime{,_ns} STATX_MTIME Want/got stx_mtime{,_ns} STATX_CTIME Want/got stx_ctime{,_ns} STATX_INO Want/got stx_ino STATX_SIZE Want/got stx_size STATX_BLOCKS Want/got stx_blocks STATX_BASIC_STATS [The stuff in the normal stat struct] STATX_BTIME Want/got stx_btime{,_ns} STATX_ALL [All currently available stuff] stx_btime is the file creation time, stx_mask is a bitmask indicating the data provided and __spares*[] are where as-yet undefined fields can be placed. Time fields are structures with separate seconds and nanoseconds fields plus a reserved field in case we want to add even finer resolution. Note that times will be negative if before 1970; in such a case, the nanosecond fields will also be negative if not zero. The bits defined in the stx_attributes field convey information about a file, how it is accessed, where it is and what it does. The following attributes map to FS_*_FL flags and are the same numerical value: STATX_ATTR_COMPRESSED File is compressed by the fs STATX_ATTR_IMMUTABLE File is marked immutable STATX_ATTR_APPEND File is append-only STATX_ATTR_NODUMP File is not to be dumped STATX_ATTR_ENCRYPTED File requires key to decrypt in fs Within the kernel, the supported flags are listed by: KSTAT_ATTR_FS_IOC_FLAGS [Are any other IOC flags of sufficient general interest to be exposed through this interface?] New flags include: STATX_ATTR_AUTOMOUNT Object is an automount trigger These are for the use of GUI tools that might want to mark files specially, depending on what they are. Fields in struct statx come in a number of classes: (0) stx_dev_*, stx_blksize. These are local system information and are always available. (1) stx_mode, stx_nlinks, stx_uid, stx_gid, stx_[amc]time, stx_ino, stx_size, stx_blocks. These will be returned whether the caller asks for them or not. The corresponding bits in stx_mask will be set to indicate whether they actually have valid values. If the caller didn't ask for them, then they may be approximated. For example, NFS won't waste any time updating them from the server, unless as a byproduct of updating something requested. If the values don't actually exist for the underlying object (such as UID or GID on a DOS file), then the bit won't be set in the stx_mask, even if the caller asked for the value. In such a case, the returned value will be a fabrication. Note that there are instances where the type might not be valid, for instance Windows reparse points. (2) stx_rdev_*. This will be set only if stx_mode indicates we're looking at a blockdev or a chardev, otherwise will be 0. (3) stx_btime. Similar to (1), except this will be set to 0 if it doesn't exist. ======= TESTING ======= The following test program can be used to test the statx system call: samples/statx/test-statx.c Just compile and run, passing it paths to the files you want to examine. The file is built automatically if CONFIG_SAMPLES is enabled. Here's some example output. Firstly, an NFS directory that crosses to another FSID. Note that the AUTOMOUNT attribute is set because transiting this directory will cause d_automount to be invoked by the VFS. [root@andromeda ~]# /tmp/test-statx -A /warthog/data statx(/warthog/data) = 0 results=7ff Size: 4096 Blocks: 8 IO Block: 1048576 directory Device: 00:26 Inode: 1703937 Links: 125 Access: (3777/drwxrwxrwx) Uid: 0 Gid: 4041 Access: 2016-11-24 09:02:12.219699527+0000 Modify: 2016-11-17 10:44:36.225653653+0000 Change: 2016-11-17 10:44:36.225653653+0000 Attributes: 0000000000001000 (-------- -------- -------- -------- -------- -------- ---m---- --------) Secondly, the result of automounting on that directory. [root@andromeda ~]# /tmp/test-statx /warthog/data statx(/warthog/data) = 0 results=7ff Size: 4096 Blocks: 8 IO Block: 1048576 directory Device: 00:27 Inode: 2 Links: 125 Access: (3777/drwxrwxrwx) Uid: 0 Gid: 4041 Access: 2016-11-24 09:02:12.219699527+0000 Modify: 2016-11-17 10:44:36.225653653+0000 Change: 2016-11-17 10:44:36.225653653+0000 Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2017-01-31 16:46:22 +00:00
static int fuse_getattr(const struct path *path, struct kstat *stat,
u32 request_mask, unsigned int flags)
{
statx: Add a system call to make enhanced file info available Add a system call to make extended file information available, including file creation and some attribute flags where available through the underlying filesystem. The getattr inode operation is altered to take two additional arguments: a u32 request_mask and an unsigned int flags that indicate the synchronisation mode. This change is propagated to the vfs_getattr*() function. Functions like vfs_stat() are now inline wrappers around new functions vfs_statx() and vfs_statx_fd() to reduce stack usage. ======== OVERVIEW ======== The idea was initially proposed as a set of xattrs that could be retrieved with getxattr(), but the general preference proved to be for a new syscall with an extended stat structure. A number of requests were gathered for features to be included. The following have been included: (1) Make the fields a consistent size on all arches and make them large. (2) Spare space, request flags and information flags are provided for future expansion. (3) Better support for the y2038 problem [Arnd Bergmann] (tv_sec is an __s64). (4) Creation time: The SMB protocol carries the creation time, which could be exported by Samba, which will in turn help CIFS make use of FS-Cache as that can be used for coherency data (stx_btime). This is also specified in NFSv4 as a recommended attribute and could be exported by NFSD [Steve French]. (5) Lightweight stat: Ask for just those details of interest, and allow a netfs (such as NFS) to approximate anything not of interest, possibly without going to the server [Trond Myklebust, Ulrich Drepper, Andreas Dilger] (AT_STATX_DONT_SYNC). (6) Heavyweight stat: Force a netfs to go to the server, even if it thinks its cached attributes are up to date [Trond Myklebust] (AT_STATX_FORCE_SYNC). And the following have been left out for future extension: (7) Data version number: Could be used by userspace NFS servers [Aneesh Kumar]. Can also be used to modify fill_post_wcc() in NFSD which retrieves i_version directly, but has just called vfs_getattr(). It could get it from the kstat struct if it used vfs_xgetattr() instead. (There's disagreement on the exact semantics of a single field, since not all filesystems do this the same way). (8) BSD stat compatibility: Including more fields from the BSD stat such as creation time (st_btime) and inode generation number (st_gen) [Jeremy Allison, Bernd Schubert]. (9) Inode generation number: Useful for FUSE and userspace NFS servers [Bernd Schubert]. (This was asked for but later deemed unnecessary with the open-by-handle capability available and caused disagreement as to whether it's a security hole or not). (10) Extra coherency data may be useful in making backups [Andreas Dilger]. (No particular data were offered, but things like last backup timestamp, the data version number and the DOS archive bit would come into this category). (11) Allow the filesystem to indicate what it can/cannot provide: A filesystem can now say it doesn't support a standard stat feature if that isn't available, so if, for instance, inode numbers or UIDs don't exist or are fabricated locally... (This requires a separate system call - I have an fsinfo() call idea for this). (12) Store a 16-byte volume ID in the superblock that can be returned in struct xstat [Steve French]. (Deferred to fsinfo). (13) Include granularity fields in the time data to indicate the granularity of each of the times (NFSv4 time_delta) [Steve French]. (Deferred to fsinfo). (14) FS_IOC_GETFLAGS value. These could be translated to BSD's st_flags. Note that the Linux IOC flags are a mess and filesystems such as Ext4 define flags that aren't in linux/fs.h, so translation in the kernel may be a necessity (or, possibly, we provide the filesystem type too). (Some attributes are made available in stx_attributes, but the general feeling was that the IOC flags were to ext[234]-specific and shouldn't be exposed through statx this way). (15) Mask of features available on file (eg: ACLs, seclabel) [Brad Boyer, Michael Kerrisk]. (Deferred, probably to fsinfo. Finding out if there's an ACL or seclabal might require extra filesystem operations). (16) Femtosecond-resolution timestamps [Dave Chinner]. (A __reserved field has been left in the statx_timestamp struct for this - if there proves to be a need). (17) A set multiple attributes syscall to go with this. =============== NEW SYSTEM CALL =============== The new system call is: int ret = statx(int dfd, const char *filename, unsigned int flags, unsigned int mask, struct statx *buffer); The dfd, filename and flags parameters indicate the file to query, in a similar way to fstatat(). There is no equivalent of lstat() as that can be emulated with statx() by passing AT_SYMLINK_NOFOLLOW in flags. There is also no equivalent of fstat() as that can be emulated by passing a NULL filename to statx() with the fd of interest in dfd. Whether or not statx() synchronises the attributes with the backing store can be controlled by OR'ing a value into the flags argument (this typically only affects network filesystems): (1) AT_STATX_SYNC_AS_STAT tells statx() to behave as stat() does in this respect. (2) AT_STATX_FORCE_SYNC will require a network filesystem to synchronise its attributes with the server - which might require data writeback to occur to get the timestamps correct. (3) AT_STATX_DONT_SYNC will suppress synchronisation with the server in a network filesystem. The resulting values should be considered approximate. mask is a bitmask indicating the fields in struct statx that are of interest to the caller. The user should set this to STATX_BASIC_STATS to get the basic set returned by stat(). It should be noted that asking for more information may entail extra I/O operations. buffer points to the destination for the data. This must be 256 bytes in size. ====================== MAIN ATTRIBUTES RECORD ====================== The following structures are defined in which to return the main attribute set: struct statx_timestamp { __s64 tv_sec; __s32 tv_nsec; __s32 __reserved; }; struct statx { __u32 stx_mask; __u32 stx_blksize; __u64 stx_attributes; __u32 stx_nlink; __u32 stx_uid; __u32 stx_gid; __u16 stx_mode; __u16 __spare0[1]; __u64 stx_ino; __u64 stx_size; __u64 stx_blocks; __u64 __spare1[1]; struct statx_timestamp stx_atime; struct statx_timestamp stx_btime; struct statx_timestamp stx_ctime; struct statx_timestamp stx_mtime; __u32 stx_rdev_major; __u32 stx_rdev_minor; __u32 stx_dev_major; __u32 stx_dev_minor; __u64 __spare2[14]; }; The defined bits in request_mask and stx_mask are: STATX_TYPE Want/got stx_mode & S_IFMT STATX_MODE Want/got stx_mode & ~S_IFMT STATX_NLINK Want/got stx_nlink STATX_UID Want/got stx_uid STATX_GID Want/got stx_gid STATX_ATIME Want/got stx_atime{,_ns} STATX_MTIME Want/got stx_mtime{,_ns} STATX_CTIME Want/got stx_ctime{,_ns} STATX_INO Want/got stx_ino STATX_SIZE Want/got stx_size STATX_BLOCKS Want/got stx_blocks STATX_BASIC_STATS [The stuff in the normal stat struct] STATX_BTIME Want/got stx_btime{,_ns} STATX_ALL [All currently available stuff] stx_btime is the file creation time, stx_mask is a bitmask indicating the data provided and __spares*[] are where as-yet undefined fields can be placed. Time fields are structures with separate seconds and nanoseconds fields plus a reserved field in case we want to add even finer resolution. Note that times will be negative if before 1970; in such a case, the nanosecond fields will also be negative if not zero. The bits defined in the stx_attributes field convey information about a file, how it is accessed, where it is and what it does. The following attributes map to FS_*_FL flags and are the same numerical value: STATX_ATTR_COMPRESSED File is compressed by the fs STATX_ATTR_IMMUTABLE File is marked immutable STATX_ATTR_APPEND File is append-only STATX_ATTR_NODUMP File is not to be dumped STATX_ATTR_ENCRYPTED File requires key to decrypt in fs Within the kernel, the supported flags are listed by: KSTAT_ATTR_FS_IOC_FLAGS [Are any other IOC flags of sufficient general interest to be exposed through this interface?] New flags include: STATX_ATTR_AUTOMOUNT Object is an automount trigger These are for the use of GUI tools that might want to mark files specially, depending on what they are. Fields in struct statx come in a number of classes: (0) stx_dev_*, stx_blksize. These are local system information and are always available. (1) stx_mode, stx_nlinks, stx_uid, stx_gid, stx_[amc]time, stx_ino, stx_size, stx_blocks. These will be returned whether the caller asks for them or not. The corresponding bits in stx_mask will be set to indicate whether they actually have valid values. If the caller didn't ask for them, then they may be approximated. For example, NFS won't waste any time updating them from the server, unless as a byproduct of updating something requested. If the values don't actually exist for the underlying object (such as UID or GID on a DOS file), then the bit won't be set in the stx_mask, even if the caller asked for the value. In such a case, the returned value will be a fabrication. Note that there are instances where the type might not be valid, for instance Windows reparse points. (2) stx_rdev_*. This will be set only if stx_mode indicates we're looking at a blockdev or a chardev, otherwise will be 0. (3) stx_btime. Similar to (1), except this will be set to 0 if it doesn't exist. ======= TESTING ======= The following test program can be used to test the statx system call: samples/statx/test-statx.c Just compile and run, passing it paths to the files you want to examine. The file is built automatically if CONFIG_SAMPLES is enabled. Here's some example output. Firstly, an NFS directory that crosses to another FSID. Note that the AUTOMOUNT attribute is set because transiting this directory will cause d_automount to be invoked by the VFS. [root@andromeda ~]# /tmp/test-statx -A /warthog/data statx(/warthog/data) = 0 results=7ff Size: 4096 Blocks: 8 IO Block: 1048576 directory Device: 00:26 Inode: 1703937 Links: 125 Access: (3777/drwxrwxrwx) Uid: 0 Gid: 4041 Access: 2016-11-24 09:02:12.219699527+0000 Modify: 2016-11-17 10:44:36.225653653+0000 Change: 2016-11-17 10:44:36.225653653+0000 Attributes: 0000000000001000 (-------- -------- -------- -------- -------- -------- ---m---- --------) Secondly, the result of automounting on that directory. [root@andromeda ~]# /tmp/test-statx /warthog/data statx(/warthog/data) = 0 results=7ff Size: 4096 Blocks: 8 IO Block: 1048576 directory Device: 00:27 Inode: 2 Links: 125 Access: (3777/drwxrwxrwx) Uid: 0 Gid: 4041 Access: 2016-11-24 09:02:12.219699527+0000 Modify: 2016-11-17 10:44:36.225653653+0000 Change: 2016-11-17 10:44:36.225653653+0000 Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2017-01-31 16:46:22 +00:00
struct inode *inode = d_inode(path->dentry);
struct fuse_conn *fc = get_fuse_conn(inode);
if (!fuse_allow_current_process(fc)) {
if (!request_mask) {
/*
* If user explicitly requested *nothing* then don't
* error out, but return st_dev only.
*/
stat->result_mask = 0;
stat->dev = inode->i_sb->s_dev;
return 0;
}
return -EACCES;
}
return fuse_update_get_attr(inode, NULL, stat, request_mask, flags);
}
static const struct inode_operations fuse_dir_inode_operations = {
.lookup = fuse_lookup,
.mkdir = fuse_mkdir,
.symlink = fuse_symlink,
.unlink = fuse_unlink,
.rmdir = fuse_rmdir,
.rename = fuse_rename2,
.link = fuse_link,
.setattr = fuse_setattr,
.create = fuse_create,
.atomic_open = fuse_atomic_open,
.mknod = fuse_mknod,
.permission = fuse_permission,
.getattr = fuse_getattr,
.listxattr = fuse_listxattr,
.get_acl = fuse_get_acl,
.set_acl = fuse_set_acl,
};
static const struct file_operations fuse_dir_operations = {
.llseek = generic_file_llseek,
.read = generic_read_dir,
.iterate_shared = fuse_readdir,
.open = fuse_dir_open,
.release = fuse_dir_release,
.fsync = fuse_dir_fsync,
.unlocked_ioctl = fuse_dir_ioctl,
.compat_ioctl = fuse_dir_compat_ioctl,
};
static const struct inode_operations fuse_common_inode_operations = {
.setattr = fuse_setattr,
.permission = fuse_permission,
.getattr = fuse_getattr,
.listxattr = fuse_listxattr,
.get_acl = fuse_get_acl,
.set_acl = fuse_set_acl,
};
static const struct inode_operations fuse_symlink_inode_operations = {
.setattr = fuse_setattr,
.get_link = fuse_get_link,
.getattr = fuse_getattr,
.listxattr = fuse_listxattr,
};
void fuse_init_common(struct inode *inode)
{
inode->i_op = &fuse_common_inode_operations;
}
void fuse_init_dir(struct inode *inode)
{
struct fuse_inode *fi = get_fuse_inode(inode);
inode->i_op = &fuse_dir_inode_operations;
inode->i_fop = &fuse_dir_operations;
spin_lock_init(&fi->rdc.lock);
fi->rdc.cached = false;
fi->rdc.size = 0;
fi->rdc.pos = 0;
fi->rdc.version = 0;
}
static int fuse_symlink_readpage(struct file *null, struct page *page)
{
int err = fuse_readlink_page(page->mapping->host, page);
if (!err)
SetPageUptodate(page);
unlock_page(page);
return err;
}
static const struct address_space_operations fuse_symlink_aops = {
.readpage = fuse_symlink_readpage,
};
void fuse_init_symlink(struct inode *inode)
{
inode->i_op = &fuse_symlink_inode_operations;
inode->i_data.a_ops = &fuse_symlink_aops;
inode_nohighmem(inode);
}