aboutsummaryrefslogtreecommitdiffstats
path: root/fs
diff options
context:
space:
mode:
Diffstat (limited to 'fs')
-rw-r--r--fs/9p/vfs_file.c6
-rw-r--r--fs/ceph/export.c10
-rw-r--r--fs/dax.c7
-rw-r--r--fs/fuse/dir.c4
-rw-r--r--fs/fuse/fuse_i.h9
-rw-r--r--fs/fuse/inode.c19
-rw-r--r--fs/libfs.c109
-rw-r--r--fs/lockd/svc.c13
-rw-r--r--fs/locks.c2
-rw-r--r--fs/namespace.c1
-rw-r--r--fs/overlayfs/inode.c29
-rw-r--r--fs/overlayfs/super.c12
12 files changed, 159 insertions, 62 deletions
diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index b84c291ba1eb..d7b78d531e63 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -74,7 +74,7 @@ int v9fs_file_open(struct inode *inode, struct file *file)
v9fs_proto_dotu(v9ses));
fid = file->private_data;
if (!fid) {
- fid = v9fs_fid_clone(file->f_path.dentry);
+ fid = v9fs_fid_clone(file_dentry(file));
if (IS_ERR(fid))
return PTR_ERR(fid);
@@ -100,7 +100,7 @@ int v9fs_file_open(struct inode *inode, struct file *file)
* because we want write after unlink usecase
* to work.
*/
- fid = v9fs_writeback_fid(file->f_path.dentry);
+ fid = v9fs_writeback_fid(file_dentry(file));
if (IS_ERR(fid)) {
err = PTR_ERR(fid);
mutex_unlock(&v9inode->v_mutex);
@@ -516,7 +516,7 @@ v9fs_mmap_file_mmap(struct file *filp, struct vm_area_struct *vma)
* because we want write after unlink usecase
* to work.
*/
- fid = v9fs_writeback_fid(filp->f_path.dentry);
+ fid = v9fs_writeback_fid(file_dentry(filp));
if (IS_ERR(fid)) {
retval = PTR_ERR(fid);
mutex_unlock(&v9inode->v_mutex);
diff --git a/fs/ceph/export.c b/fs/ceph/export.c
index 6e72c98162d5..1780218a48f0 100644
--- a/fs/ceph/export.c
+++ b/fs/ceph/export.c
@@ -95,10 +95,8 @@ static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino)
}
dentry = d_obtain_alias(inode);
- if (IS_ERR(dentry)) {
- iput(inode);
+ if (IS_ERR(dentry))
return dentry;
- }
err = ceph_init_dentry(dentry);
if (err < 0) {
dput(dentry);
@@ -167,10 +165,8 @@ static struct dentry *__get_parent(struct super_block *sb,
return ERR_PTR(-ENOENT);
dentry = d_obtain_alias(inode);
- if (IS_ERR(dentry)) {
- iput(inode);
+ if (IS_ERR(dentry))
return dentry;
- }
err = ceph_init_dentry(dentry);
if (err < 0) {
dput(dentry);
@@ -210,7 +206,7 @@ static struct dentry *ceph_fh_to_parent(struct super_block *sb,
dout("fh_to_parent %llx\n", cfh->parent_ino);
dentry = __get_parent(sb, NULL, cfh->ino);
- if (IS_ERR(dentry) && PTR_ERR(dentry) == -ENOENT)
+ if (unlikely(dentry == ERR_PTR(-ENOENT)))
dentry = __fh_to_dentry(sb, cfh->parent_ino);
return dentry;
}
diff --git a/fs/dax.c b/fs/dax.c
index 761495bf5eb9..e207f8f9b700 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -208,7 +208,12 @@ static ssize_t dax_io(struct inode *inode, struct iov_iter *iter,
dax.addr += first;
size = map_len - first;
}
- max = min(pos + size, end);
+ /*
+ * pos + size is one past the last offset for IO,
+ * so pos + size can overflow loff_t at extreme offsets.
+ * Cast to u64 to catch this and get the true minimum.
+ */
+ max = min_t(u64, pos + size, end);
}
if (iov_iter_rw(iter) == WRITE) {
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index ccd4971cc6c1..264f07c7754e 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -341,8 +341,10 @@ static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry,
struct dentry *newent;
bool outarg_valid = true;
+ fuse_lock_inode(dir);
err = fuse_lookup_name(dir->i_sb, get_node_id(dir), &entry->d_name,
&outarg, &inode);
+ fuse_unlock_inode(dir);
if (err == -ENOENT) {
outarg_valid = false;
err = 0;
@@ -1341,7 +1343,9 @@ static int fuse_readdir(struct file *file, struct dir_context *ctx)
fuse_read_fill(req, file, ctx->pos, PAGE_SIZE,
FUSE_READDIR);
}
+ fuse_lock_inode(inode);
fuse_request_send(fc, req);
+ fuse_unlock_inode(inode);
nbytes = req->out.args[0].size;
err = req->out.h.error;
fuse_put_request(fc, req);
diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index eddbe02c4028..929c383432b0 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -110,6 +110,9 @@ struct fuse_inode {
/** Miscellaneous bits describing inode state */
unsigned long state;
+
+ /** Lock for serializing lookup and readdir for back compatibility*/
+ struct mutex mutex;
};
/** FUSE inode state bits */
@@ -540,6 +543,9 @@ struct fuse_conn {
/** write-back cache policy (default is write-through) */
unsigned writeback_cache:1;
+ /** allow parallel lookups and readdir (default is serialized) */
+ unsigned parallel_dirops:1;
+
/*
* The following bitfields are only for optimization purposes
* and hence races in setting them will not cause malfunction
@@ -956,4 +962,7 @@ int fuse_do_setattr(struct inode *inode, struct iattr *attr,
void fuse_set_initialized(struct fuse_conn *fc);
+void fuse_unlock_inode(struct inode *inode);
+void fuse_lock_inode(struct inode *inode);
+
#endif /* _FS_FUSE_I_H */
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index 1ce67668a8e1..9961d8432ce3 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -97,6 +97,7 @@ static struct inode *fuse_alloc_inode(struct super_block *sb)
INIT_LIST_HEAD(&fi->queued_writes);
INIT_LIST_HEAD(&fi->writepages);
init_waitqueue_head(&fi->page_waitq);
+ mutex_init(&fi->mutex);
fi->forget = fuse_alloc_forget();
if (!fi->forget) {
kmem_cache_free(fuse_inode_cachep, inode);
@@ -117,6 +118,7 @@ static void fuse_destroy_inode(struct inode *inode)
struct fuse_inode *fi = get_fuse_inode(inode);
BUG_ON(!list_empty(&fi->write_files));
BUG_ON(!list_empty(&fi->queued_writes));
+ mutex_destroy(&fi->mutex);
kfree(fi->forget);
call_rcu(&inode->i_rcu, fuse_i_callback);
}
@@ -351,6 +353,18 @@ int fuse_reverse_inval_inode(struct super_block *sb, u64 nodeid,
return 0;
}
+void fuse_lock_inode(struct inode *inode)
+{
+ if (!get_fuse_conn(inode)->parallel_dirops)
+ mutex_lock(&get_fuse_inode(inode)->mutex);
+}
+
+void fuse_unlock_inode(struct inode *inode)
+{
+ if (!get_fuse_conn(inode)->parallel_dirops)
+ mutex_unlock(&get_fuse_inode(inode)->mutex);
+}
+
static void fuse_umount_begin(struct super_block *sb)
{
fuse_abort_conn(get_fuse_conn_super(sb));
@@ -898,6 +912,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req)
fc->async_dio = 1;
if (arg->flags & FUSE_WRITEBACK_CACHE)
fc->writeback_cache = 1;
+ if (arg->flags & FUSE_PARALLEL_DIROPS)
+ fc->parallel_dirops = 1;
if (arg->time_gran && arg->time_gran <= 1000000000)
fc->sb->s_time_gran = arg->time_gran;
} else {
@@ -928,7 +944,8 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req)
FUSE_SPLICE_WRITE | FUSE_SPLICE_MOVE | FUSE_SPLICE_READ |
FUSE_FLOCK_LOCKS | FUSE_IOCTL_DIR | FUSE_AUTO_INVAL_DATA |
FUSE_DO_READDIRPLUS | FUSE_READDIRPLUS_AUTO | FUSE_ASYNC_DIO |
- FUSE_WRITEBACK_CACHE | FUSE_NO_OPEN_SUPPORT;
+ FUSE_WRITEBACK_CACHE | FUSE_NO_OPEN_SUPPORT |
+ FUSE_PARALLEL_DIROPS;
req->in.h.opcode = FUSE_INIT;
req->in.numargs = 1;
req->in.args[0].size = sizeof(*arg);
diff --git a/fs/libfs.c b/fs/libfs.c
index cedeacbae303..74dc8b9e7f53 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -84,6 +84,61 @@ int dcache_dir_close(struct inode *inode, struct file *file)
}
EXPORT_SYMBOL(dcache_dir_close);
+/* parent is locked at least shared */
+static struct dentry *next_positive(struct dentry *parent,
+ struct list_head *from,
+ int count)
+{
+ unsigned *seq = &parent->d_inode->i_dir_seq, n;
+ struct dentry *res;
+ struct list_head *p;
+ bool skipped;
+ int i;
+
+retry:
+ i = count;
+ skipped = false;
+ n = smp_load_acquire(seq) & ~1;
+ res = NULL;
+ rcu_read_lock();
+ for (p = from->next; p != &parent->d_subdirs; p = p->next) {
+ struct dentry *d = list_entry(p, struct dentry, d_child);
+ if (!simple_positive(d)) {
+ skipped = true;
+ } else if (!--i) {
+ res = d;
+ break;
+ }
+ }
+ rcu_read_unlock();
+ if (skipped) {
+ smp_rmb();
+ if (unlikely(*seq != n))
+ goto retry;
+ }
+ return res;
+}
+
+static void move_cursor(struct dentry *cursor, struct list_head *after)
+{
+ struct dentry *parent = cursor->d_parent;
+ unsigned n, *seq = &parent->d_inode->i_dir_seq;
+ spin_lock(&parent->d_lock);
+ for (;;) {
+ n = *seq;
+ if (!(n & 1) && cmpxchg(seq, n, n + 1) == n)
+ break;
+ cpu_relax();
+ }
+ __list_del(cursor->d_child.prev, cursor->d_child.next);
+ if (after)
+ list_add(&cursor->d_child, after);
+ else
+ list_add_tail(&cursor->d_child, &parent->d_subdirs);
+ smp_store_release(seq, n + 2);
+ spin_unlock(&parent->d_lock);
+}
+
loff_t dcache_dir_lseek(struct file *file, loff_t offset, int whence)
{
struct dentry *dentry = file->f_path.dentry;
@@ -99,25 +154,14 @@ loff_t dcache_dir_lseek(struct file *file, loff_t offset, int whence)
if (offset != file->f_pos) {
file->f_pos = offset;
if (file->f_pos >= 2) {
- struct list_head *p;
struct dentry *cursor = file->private_data;
+ struct dentry *to;
loff_t n = file->f_pos - 2;
- spin_lock(&dentry->d_lock);
- /* d_lock not required for cursor */
- list_del(&cursor->d_child);
- p = dentry->d_subdirs.next;
- while (n && p != &dentry->d_subdirs) {
- struct dentry *next;
- next = list_entry(p, struct dentry, d_child);
- spin_lock_nested(&next->d_lock, DENTRY_D_LOCK_NESTED);
- if (simple_positive(next))
- n--;
- spin_unlock(&next->d_lock);
- p = p->next;
- }
- list_add_tail(&cursor->d_child, p);
- spin_unlock(&dentry->d_lock);
+ inode_lock_shared(dentry->d_inode);
+ to = next_positive(dentry, &dentry->d_subdirs, n);
+ move_cursor(cursor, to ? &to->d_child : NULL);
+ inode_unlock_shared(dentry->d_inode);
}
}
return offset;
@@ -140,36 +184,25 @@ int dcache_readdir(struct file *file, struct dir_context *ctx)
{
struct dentry *dentry = file->f_path.dentry;
struct dentry *cursor = file->private_data;
- struct list_head *p, *q = &cursor->d_child;
+ struct list_head *p = &cursor->d_child;
+ struct dentry *next;
+ bool moved = false;
if (!dir_emit_dots(file, ctx))
return 0;
- spin_lock(&dentry->d_lock);
- if (ctx->pos == 2)
- list_move(q, &dentry->d_subdirs);
- for (p = q->next; p != &dentry->d_subdirs; p = p->next) {
- struct dentry *next = list_entry(p, struct dentry, d_child);
- spin_lock_nested(&next->d_lock, DENTRY_D_LOCK_NESTED);
- if (!simple_positive(next)) {
- spin_unlock(&next->d_lock);
- continue;
- }
-
- spin_unlock(&next->d_lock);
- spin_unlock(&dentry->d_lock);
+ if (ctx->pos == 2)
+ p = &dentry->d_subdirs;
+ while ((next = next_positive(dentry, p, 1)) != NULL) {
if (!dir_emit(ctx, next->d_name.name, next->d_name.len,
d_inode(next)->i_ino, dt_type(d_inode(next))))
- return 0;
- spin_lock(&dentry->d_lock);
- spin_lock_nested(&next->d_lock, DENTRY_D_LOCK_NESTED);
- /* next is still alive */
- list_move(q, p);
- spin_unlock(&next->d_lock);
- p = q;
+ break;
+ moved = true;
+ p = &next->d_child;
ctx->pos++;
}
- spin_unlock(&dentry->d_lock);
+ if (moved)
+ move_cursor(cursor, p);
return 0;
}
EXPORT_SYMBOL(dcache_readdir);
diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c
index 154a107cd376..fc4084ef4736 100644
--- a/fs/lockd/svc.c
+++ b/fs/lockd/svc.c
@@ -335,12 +335,17 @@ static struct notifier_block lockd_inet6addr_notifier = {
};
#endif
-static void lockd_svc_exit_thread(void)
+static void lockd_unregister_notifiers(void)
{
unregister_inetaddr_notifier(&lockd_inetaddr_notifier);
#if IS_ENABLED(CONFIG_IPV6)
unregister_inet6addr_notifier(&lockd_inet6addr_notifier);
#endif
+}
+
+static void lockd_svc_exit_thread(void)
+{
+ lockd_unregister_notifiers();
svc_exit_thread(nlmsvc_rqst);
}
@@ -462,7 +467,7 @@ int lockd_up(struct net *net)
* Note: svc_serv structures have an initial use count of 1,
* so we exit through here on both success and failure.
*/
-err_net:
+err_put:
svc_destroy(serv);
err_create:
mutex_unlock(&nlmsvc_mutex);
@@ -470,7 +475,9 @@ err_create:
err_start:
lockd_down_net(serv, net);
- goto err_net;
+err_net:
+ lockd_unregister_notifiers();
+ goto err_put;
}
EXPORT_SYMBOL_GPL(lockd_up);
diff --git a/fs/locks.c b/fs/locks.c
index 7c5f91be9b65..ee1b15f6fc13 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -1628,7 +1628,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr
{
struct file_lock *fl, *my_fl = NULL, *lease;
struct dentry *dentry = filp->f_path.dentry;
- struct inode *inode = dentry->d_inode;
+ struct inode *inode = file_inode(filp);
struct file_lock_context *ctx;
bool is_deleg = (*flp)->fl_flags & FL_DELEG;
int error;
diff --git a/fs/namespace.c b/fs/namespace.c
index 783004af5707..419f746d851d 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1562,6 +1562,7 @@ void __detach_mounts(struct dentry *dentry)
goto out_unlock;
lock_mount_hash();
+ event++;
while (!hlist_empty(&mp->m_list)) {
mnt = hlist_entry(mp->m_list.first, struct mount, mnt_mp_list);
if (mnt->mnt.mnt_flags & MNT_UMOUNT) {
diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c
index 1dbeab6cf96e..c831c2e5f803 100644
--- a/fs/overlayfs/inode.c
+++ b/fs/overlayfs/inode.c
@@ -59,16 +59,37 @@ int ovl_setattr(struct dentry *dentry, struct iattr *attr)
if (err)
goto out;
+ if (attr->ia_valid & ATTR_SIZE) {
+ struct inode *realinode = d_inode(ovl_dentry_real(dentry));
+
+ err = -ETXTBSY;
+ if (atomic_read(&realinode->i_writecount) < 0)
+ goto out_drop_write;
+ }
+
err = ovl_copy_up(dentry);
if (!err) {
+ struct inode *winode = NULL;
+
upperdentry = ovl_dentry_upper(dentry);
+ if (attr->ia_valid & ATTR_SIZE) {
+ winode = d_inode(upperdentry);
+ err = get_write_access(winode);
+ if (err)
+ goto out_drop_write;
+ }
+
inode_lock(upperdentry->d_inode);
err = notify_change(upperdentry, attr, NULL);
if (!err)
ovl_copyattr(upperdentry->d_inode, dentry->d_inode);
inode_unlock(upperdentry->d_inode);
+
+ if (winode)
+ put_write_access(winode);
}
+out_drop_write:
ovl_drop_write(dentry);
out:
return err;
@@ -121,16 +142,18 @@ int ovl_permission(struct inode *inode, int mask)
err = vfs_getattr(&realpath, &stat);
if (err)
- return err;
+ goto out_dput;
+ err = -ESTALE;
if ((stat.mode ^ inode->i_mode) & S_IFMT)
- return -ESTALE;
+ goto out_dput;
inode->i_mode = stat.mode;
inode->i_uid = stat.uid;
inode->i_gid = stat.gid;
- return generic_permission(inode, mask);
+ err = generic_permission(inode, mask);
+ goto out_dput;
}
/* Careful in RCU walk mode */
diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index ce02f46029da..9a7693d5f8ff 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -1082,11 +1082,13 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent)
if (err < 0)
goto out_put_workdir;
- if (!err) {
- pr_err("overlayfs: upper fs needs to support d_type.\n");
- err = -EINVAL;
- goto out_put_workdir;
- }
+ /*
+ * We allowed this configuration and don't want to
+ * break users over kernel upgrade. So warn instead
+ * of erroring out.
+ */
+ if (!err)
+ pr_warn("overlayfs: upper fs needs to support d_type.\n");
}
}
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.