Resguardo inicial

linux-7.2.3/fs/fuse/Kconfig

@@ -0,0 +1,78 @@
+# SPDX-License-Identifier: GPL-2.0-only
+config FUSE_FS
+	tristate "FUSE (Filesystem in Userspace) support"
+	select FS_POSIX_ACL
+	select FS_IOMAP
+	help
+	  With FUSE it is possible to implement a fully functional filesystem
+	  in a userspace program.
+
+	  There's also a companion library: libfuse2.  This library is available
+	  from the FUSE homepage:
+	  <https://github.com/libfuse/>
+	  although chances are your distribution already has that library
+	  installed if you've installed the "fuse" package itself.
+
+	  See <file:Documentation/filesystems/fuse/fuse.rst> for more information.
+	  See <file:Documentation/Changes> for needed library/utility version.
+
+	  If you want to develop a userspace FS, or if you want to use
+	  a filesystem based on FUSE, answer Y or M.
+
+config CUSE
+	tristate "Character device in Userspace support"
+	depends on FUSE_FS
+	help
+	  This FUSE extension allows character devices to be
+	  implemented in userspace.
+
+	  If you want to develop or use a userspace character device
+	  based on CUSE, answer Y or M.
+
+config VIRTIO_FS
+	tristate "Virtio Filesystem"
+	depends on FUSE_FS
+	select VIRTIO
+	help
+	  The Virtio Filesystem allows guests to mount file systems from the
+	  host.
+
+	  If you want to share files between guests or with the host, answer Y
+	  or M.
+
+config FUSE_DAX
+	bool "Virtio Filesystem Direct Host Memory Access support"
+	default y
+	select INTERVAL_TREE
+	depends on VIRTIO_FS
+	depends on FS_DAX
+	depends on DAX
+	help
+	  This allows bypassing guest page cache and allows mapping host page
+	  cache directly in guest address space.
+
+	  If you want to allow mounting a Virtio Filesystem with the "dax"
+	  option, answer Y.
+
+config FUSE_PASSTHROUGH
+	bool "FUSE passthrough operations support"
+	default y
+	depends on FUSE_FS
+	select FS_STACK
+	help
+	  This allows bypassing FUSE server by mapping specific FUSE operations
+	  to be performed directly on a backing file.
+
+	  If you want to allow passthrough operations, answer Y.
+
+config FUSE_IO_URING
+	bool "FUSE communication over io-uring"
+	default y
+	depends on FUSE_FS
+	depends on IO_URING
+	help
+	  This allows sending FUSE requests over the io-uring interface and
+          also adds request core affinity.
+
+	  If you want to allow fuse server/client communication through io-uring,
+	  answer Y

linux-7.2.3/fs/fuse/Makefile

@@ -0,0 +1,22 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for the FUSE filesystem.
+#
+
+# Needed for trace events
+ccflags-y = -I$(src)
+
+obj-$(CONFIG_FUSE_FS) += fuse.o
+obj-$(CONFIG_CUSE) += cuse.o
+obj-$(CONFIG_VIRTIO_FS) += virtiofs.o
+
+fuse-y := trace.o	# put trace.o first so we see ftrace errors sooner
+fuse-y += dev.o dir.o file.o inode.o control.o xattr.o acl.o readdir.o ioctl.o req_timeout.o req.o
+fuse-y += poll.o notify.o
+fuse-y += iomode.o
+fuse-$(CONFIG_FUSE_DAX) += dax.o
+fuse-$(CONFIG_FUSE_PASSTHROUGH) += passthrough.o backing.o
+fuse-$(CONFIG_SYSCTL) += sysctl.o
+fuse-$(CONFIG_FUSE_IO_URING) += dev_uring.o
+
+virtiofs-y := virtio_fs.o

linux-7.2.3/fs/fuse/acl.c

@@ -0,0 +1,161 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * FUSE: Filesystem in Userspace
+ * Copyright (C) 2016 Canonical Ltd. <seth.forshee@canonical.com>
+ */
+
+#include "fuse_i.h"
+
+#include <linux/posix_acl.h>
+#include <linux/posix_acl_xattr.h>
+
+static struct posix_acl *__fuse_get_acl(struct fuse_conn *fc,
+					struct inode *inode, int type, bool rcu)
+{
+	int size;
+	const char *name;
+	void *value = NULL;
+	struct posix_acl *acl;
+
+	if (rcu)
+		return ERR_PTR(-ECHILD);
+
+	if (fuse_is_bad(inode))
+		return ERR_PTR(-EIO);
+
+	if (fc->no_getxattr)
+		return NULL;
+
+	if (type == ACL_TYPE_ACCESS)
+		name = XATTR_NAME_POSIX_ACL_ACCESS;
+	else if (type == ACL_TYPE_DEFAULT)
+		name = XATTR_NAME_POSIX_ACL_DEFAULT;
+	else
+		return ERR_PTR(-EOPNOTSUPP);
+
+	value = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!value)
+		return ERR_PTR(-ENOMEM);
+	size = fuse_getxattr(inode, name, value, PAGE_SIZE);
+	if (size > 0)
+		acl = posix_acl_from_xattr(fc->user_ns, value, size);
+	else if ((size == 0) || (size == -ENODATA) ||
+		 (size == -EOPNOTSUPP && fc->no_getxattr))
+		acl = NULL;
+	else if (size == -ERANGE)
+		acl = ERR_PTR(-E2BIG);
+	else
+		acl = ERR_PTR(size);
+
+	kfree(value);
+	return acl;
+}
+
+static inline bool fuse_no_acl(const struct fuse_conn *fc,
+			       const struct inode *inode)
+{
+	/*
+	 * Refuse interacting with POSIX ACLs for daemons that
+	 * don't support FUSE_POSIX_ACL and are not mounted on
+	 * the host to retain backwards compatibility.
+	 */
+	return !fc->posix_acl && (i_user_ns(inode) != &init_user_ns);
+}
+
+struct posix_acl *fuse_get_acl(struct mnt_idmap *idmap,
+			       struct dentry *dentry, int type)
+{
+	struct inode *inode = d_inode(dentry);
+	struct fuse_conn *fc = get_fuse_conn(inode);
+
+	if (fuse_no_acl(fc, inode))
+		return ERR_PTR(-EOPNOTSUPP);
+
+	return __fuse_get_acl(fc, inode, type, false);
+}
+
+struct posix_acl *fuse_get_inode_acl(struct inode *inode, int type, bool rcu)
+{
+	struct fuse_conn *fc = get_fuse_conn(inode);
+
+	/*
+	 * FUSE daemons before FUSE_POSIX_ACL was introduced could get and set
+	 * POSIX ACLs without them being used for permission checking by the
+	 * vfs. Retain that behavior for backwards compatibility as there are
+	 * filesystems that do all permission checking for acls in the daemon
+	 * and not in the kernel.
+	 */
+	if (!fc->posix_acl)
+		return NULL;
+	return __fuse_get_acl(fc,  inode, type, rcu);
+}
+
+int fuse_set_acl(struct mnt_idmap *idmap, struct dentry *dentry,
+		 struct posix_acl *acl, int type)
+{
+	struct inode *inode = d_inode(dentry);
+	struct fuse_conn *fc = get_fuse_conn(inode);
+	const char *name;
+	int ret;
+
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	if (fc->no_setxattr || fuse_no_acl(fc, inode))
+		return -EOPNOTSUPP;
+
+	if (type == ACL_TYPE_ACCESS)
+		name = XATTR_NAME_POSIX_ACL_ACCESS;
+	else if (type == ACL_TYPE_DEFAULT)
+		name = XATTR_NAME_POSIX_ACL_DEFAULT;
+	else
+		return -EINVAL;
+
+	if (acl) {
+		unsigned int extra_flags = 0;
+		/*
+		 * Fuse userspace is responsible for updating access
+		 * permissions in the inode, if needed. fuse_setxattr
+		 * invalidates the inode attributes, which will force
+		 * them to be refreshed the next time they are used,
+		 * and it also updates i_ctime.
+		 */
+		size_t size;
+		void *value;
+
+		value = posix_acl_to_xattr(fc->user_ns, acl, &size, GFP_KERNEL);
+		if (!value)
+			return -ENOMEM;
+
+		if (size > PAGE_SIZE) {
+			kfree(value);
+			return -E2BIG;
+		}
+
+		/*
+		 * Fuse daemons without FUSE_POSIX_ACL never changed the passed
+		 * through POSIX ACLs. Such daemons don't expect setgid bits to
+		 * be stripped.
+		 */
+		if (fc->posix_acl &&
+		    !in_group_or_capable(idmap, inode,
+					 i_gid_into_vfsgid(idmap, inode)))
+			extra_flags |= FUSE_SETXATTR_ACL_KILL_SGID;
+
+		ret = fuse_setxattr(inode, name, value, size, 0, extra_flags);
+		kfree(value);
+	} else {
+		ret = fuse_removexattr(inode, name);
+	}
+
+	if (fc->posix_acl) {
+		/*
+		 * Fuse daemons without FUSE_POSIX_ACL never cached POSIX ACLs
+		 * and didn't invalidate attributes. Retain that behavior.
+		 */
+		forget_all_cached_acls(inode);
+		fuse_invalidate_attr(inode);
+	}
+
+	return ret;
+}

linux-7.2.3/fs/fuse/args.h

@@ -0,0 +1,65 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef _FS_FUSE_ARGS_H
+#define _FS_FUSE_ARGS_H
+
+#include <linux/types.h>
+
+struct fuse_mount;
+
+/** One input argument of a request */
+struct fuse_in_arg {
+	unsigned size;
+	const void *value;
+};
+
+/** One output argument of a request */
+struct fuse_arg {
+	unsigned size;
+	void *value;
+};
+
+struct fuse_args {
+	u64 nodeid;
+	u32 opcode;
+	u32 uid;
+	u32 gid;
+	u32 pid;
+	u8 in_numargs;
+	u8 out_numargs;
+	u8 ext_idx;
+	bool force:1;
+	bool noreply:1;
+	bool nocreds:1;
+	bool in_pages:1;
+	bool out_pages:1;
+	bool user_pages:1;
+	bool out_argvar:1;
+	bool page_zeroing:1;
+	bool page_replace:1;
+	bool may_block:1;
+	bool is_ext:1;
+	bool is_pinned:1;
+	bool invalidate_vmap:1;
+	bool abort_on_kill:1;
+	struct fuse_in_arg in_args[4];
+	struct fuse_arg out_args[2];
+	void (*end)(struct fuse_args *args, int error);
+	/* Used for kvec iter backed by vmalloc address */
+	void *vmap_base;
+};
+
+/** FUSE folio descriptor */
+struct fuse_folio_desc {
+	unsigned int length;
+	unsigned int offset;
+};
+
+struct fuse_args_pages {
+	struct fuse_args args;
+	struct folio **folios;
+	struct fuse_folio_desc *descs;
+	unsigned int num_folios;
+};
+
+#endif /* _FS_FUSE_ARGS_H */

linux-7.2.3/fs/fuse/backing.c

@@ -0,0 +1,180 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * FUSE passthrough to backing file.
+ *
+ * Copyright (c) 2023 CTERA Networks.
+ */
+
+#include "dev.h"
+#include "fuse_i.h"
+
+#include <linux/file.h>
+
+struct fuse_backing *fuse_backing_get(struct fuse_backing *fb)
+{
+	if (fb && refcount_inc_not_zero(&fb->count))
+		return fb;
+	return NULL;
+}
+
+static void fuse_backing_free(struct fuse_backing *fb)
+{
+	pr_debug("%s: fb=0x%p\n", __func__, fb);
+
+	if (fb->file)
+		fput(fb->file);
+	put_cred(fb->cred);
+	kfree_rcu(fb, rcu);
+}
+
+void fuse_backing_put(struct fuse_backing *fb)
+{
+	if (fb && refcount_dec_and_test(&fb->count))
+		fuse_backing_free(fb);
+}
+
+void fuse_backing_files_init(struct fuse_conn *fc)
+{
+	idr_init(&fc->backing_files_map);
+}
+
+static int fuse_backing_id_alloc(struct fuse_conn *fc, struct fuse_backing *fb)
+{
+	int id;
+
+	idr_preload(GFP_KERNEL);
+	spin_lock(&fc->lock);
+	/* FIXME: xarray might be space inefficient */
+	id = idr_alloc_cyclic(&fc->backing_files_map, fb, 1, 0, GFP_ATOMIC);
+	spin_unlock(&fc->lock);
+	idr_preload_end();
+
+	WARN_ON_ONCE(id == 0);
+	return id;
+}
+
+static struct fuse_backing *fuse_backing_id_remove(struct fuse_conn *fc,
+						   int id)
+{
+	struct fuse_backing *fb;
+
+	spin_lock(&fc->lock);
+	fb = idr_remove(&fc->backing_files_map, id);
+	spin_unlock(&fc->lock);
+
+	return fb;
+}
+
+static int fuse_backing_id_free(int id, void *p, void *data)
+{
+	struct fuse_backing *fb = p;
+
+	WARN_ON_ONCE(refcount_read(&fb->count) != 1);
+	fuse_backing_free(fb);
+	return 0;
+}
+
+void fuse_backing_files_free(struct fuse_conn *fc)
+{
+	idr_for_each(&fc->backing_files_map, fuse_backing_id_free, NULL);
+	idr_destroy(&fc->backing_files_map);
+}
+
+int fuse_backing_open(struct fuse_conn *fc, struct fuse_backing_map *map)
+{
+	struct file *file;
+	struct super_block *backing_sb;
+	struct fuse_backing *fb = NULL;
+	int res;
+
+	pr_debug("%s: fd=%d flags=0x%x\n", __func__, map->fd, map->flags);
+
+	/* TODO: relax CAP_SYS_ADMIN once backing files are visible to lsof */
+	res = -EPERM;
+	if (!fc->passthrough || !capable(CAP_SYS_ADMIN))
+		goto out;
+
+	res = -EINVAL;
+	if (map->flags || map->padding)
+		goto out;
+
+	file = fget_raw(map->fd);
+	res = -EBADF;
+	if (!file)
+		goto out;
+
+	/* read/write/splice/mmap passthrough only relevant for regular files */
+	res = d_is_dir(file->f_path.dentry) ? -EISDIR : -EINVAL;
+	if (!d_is_reg(file->f_path.dentry))
+		goto out_fput;
+
+	backing_sb = file_inode(file)->i_sb;
+	res = -ELOOP;
+	if (backing_sb->s_stack_depth >= fc->max_stack_depth)
+		goto out_fput;
+
+	fb = kmalloc_obj(struct fuse_backing);
+	res = -ENOMEM;
+	if (!fb)
+		goto out_fput;
+
+	fb->file = file;
+	fb->cred = get_current_cred();
+	refcount_set(&fb->count, 1);
+
+	res = fuse_backing_id_alloc(fc, fb);
+	if (res < 0) {
+		fuse_backing_free(fb);
+		fb = NULL;
+	}
+
+out:
+	pr_debug("%s: fb=0x%p, ret=%i\n", __func__, fb, res);
+
+	return res;
+
+out_fput:
+	fput(file);
+	goto out;
+}
+
+int fuse_backing_close(struct fuse_conn *fc, int backing_id)
+{
+	struct fuse_backing *fb = NULL;
+	int err;
+
+	pr_debug("%s: backing_id=%d\n", __func__, backing_id);
+
+	/* TODO: relax CAP_SYS_ADMIN once backing files are visible to lsof */
+	err = -EPERM;
+	if (!fc->passthrough || !capable(CAP_SYS_ADMIN))
+		goto out;
+
+	err = -EINVAL;
+	if (backing_id <= 0)
+		goto out;
+
+	err = -ENOENT;
+	fb = fuse_backing_id_remove(fc, backing_id);
+	if (!fb)
+		goto out;
+
+	fuse_backing_put(fb);
+	err = 0;
+out:
+	pr_debug("%s: fb=0x%p, err=%i\n", __func__, fb, err);
+
+	return err;
+}
+
+struct fuse_backing *fuse_backing_lookup(struct fuse_conn *fc, int backing_id)
+{
+	struct fuse_backing *fb;
+
+	rcu_read_lock();
+	fb = idr_find(&fc->backing_files_map, backing_id);
+	fb = fuse_backing_get(fb);
+	rcu_read_unlock();
+
+	return fb;
+}

linux-7.2.3/fs/fuse/control.c

@@ -0,0 +1,366 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+  FUSE: Filesystem in Userspace
+  Copyright (C) 2001-2008  Miklos Szeredi <miklos@szeredi.hu>
+*/
+
+#include "fuse_i.h"
+#include "dev.h"
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/fs_context.h>
+#include <linux/namei.h>
+
+#define FUSE_CTL_SUPER_MAGIC 0x65735543
+
+/*
+ * This is non-NULL when the single instance of the control filesystem
+ * exists.  Protected by fuse_mutex
+ */
+static struct super_block *fuse_control_sb;
+
+static struct fuse_conn *fuse_ctl_file_conn_get(struct file *file)
+{
+	struct fuse_conn *fc;
+	mutex_lock(&fuse_mutex);
+	fc = file_inode(file)->i_private;
+	if (fc)
+		fc = fuse_conn_get(fc);
+	mutex_unlock(&fuse_mutex);
+	return fc;
+}
+
+static ssize_t fuse_conn_abort_write(struct file *file, const char __user *buf,
+				     size_t count, loff_t *ppos)
+{
+	struct fuse_conn *fc = fuse_ctl_file_conn_get(file);
+	if (fc) {
+		fuse_chan_abort(fc->chan, fc->abort_err);
+		fuse_conn_put(fc);
+	}
+	return count;
+}
+
+static ssize_t fuse_conn_waiting_read(struct file *file, char __user *buf,
+				      size_t len, loff_t *ppos)
+{
+	char tmp[32];
+	size_t size;
+
+	if (!*ppos) {
+		long value;
+		struct fuse_conn *fc = fuse_ctl_file_conn_get(file);
+		if (!fc)
+			return 0;
+
+		value = fuse_chan_num_waiting(fc->chan);
+		file->private_data = (void *)value;
+		fuse_conn_put(fc);
+	}
+	size = sprintf(tmp, "%ld\n", (long)file->private_data);
+	return simple_read_from_buffer(buf, len, ppos, tmp, size);
+}
+
+static ssize_t fuse_conn_limit_read(struct file *file, char __user *buf,
+				    size_t len, loff_t *ppos, unsigned val)
+{
+	char tmp[32];
+	size_t size = sprintf(tmp, "%u\n", val);
+
+	return simple_read_from_buffer(buf, len, ppos, tmp, size);
+}
+
+static ssize_t fuse_conn_limit_write(struct file *file, const char __user *buf,
+				     size_t count, loff_t *ppos, unsigned *val,
+				     unsigned global_limit)
+{
+	unsigned long t;
+	unsigned limit = (1 << 16) - 1;
+	int err;
+
+	if (*ppos)
+		return -EINVAL;
+
+	err = kstrtoul_from_user(buf, count, 0, &t);
+	if (err)
+		return err;
+
+	if (!capable(CAP_SYS_ADMIN))
+		limit = min(limit, global_limit);
+
+	if (t > limit)
+		return -EINVAL;
+
+	*val = t;
+
+	return count;
+}
+
+static ssize_t fuse_conn_max_background_read(struct file *file,
+					     char __user *buf, size_t len,
+					     loff_t *ppos)
+{
+	struct fuse_conn *fc;
+	unsigned val;
+
+	fc = fuse_ctl_file_conn_get(file);
+	if (!fc)
+		return 0;
+
+	val = fuse_chan_max_background(fc->chan);
+	fuse_conn_put(fc);
+
+	return fuse_conn_limit_read(file, buf, len, ppos, val);
+}
+
+static ssize_t fuse_conn_max_background_write(struct file *file,
+					      const char __user *buf,
+					      size_t count, loff_t *ppos)
+{
+	unsigned int val = 0;
+	ssize_t ret;
+
+	ret = fuse_conn_limit_write(file, buf, count, ppos, &val,
+				    max_user_bgreq);
+	if (ret > 0) {
+		struct fuse_conn *fc = fuse_ctl_file_conn_get(file);
+		if (fc) {
+			fuse_chan_max_background_set(fc->chan, val);
+			fuse_conn_put(fc);
+		}
+	}
+
+	return ret;
+}
+
+static ssize_t fuse_conn_congestion_threshold_read(struct file *file,
+						   char __user *buf, size_t len,
+						   loff_t *ppos)
+{
+	struct fuse_conn *fc;
+	unsigned val;
+
+	fc = fuse_ctl_file_conn_get(file);
+	if (!fc)
+		return 0;
+
+	val = READ_ONCE(fc->congestion_threshold);
+	fuse_conn_put(fc);
+
+	return fuse_conn_limit_read(file, buf, len, ppos, val);
+}
+
+static ssize_t fuse_conn_congestion_threshold_write(struct file *file,
+						    const char __user *buf,
+						    size_t count, loff_t *ppos)
+{
+	unsigned int val = 0;
+	struct fuse_conn *fc;
+	ssize_t ret;
+
+	ret = fuse_conn_limit_write(file, buf, count, ppos, &val,
+				    max_user_congthresh);
+	if (ret <= 0)
+		goto out;
+	fc = fuse_ctl_file_conn_get(file);
+	if (!fc)
+		goto out;
+
+	WRITE_ONCE(fc->congestion_threshold, val);
+	fuse_conn_put(fc);
+out:
+	return ret;
+}
+
+static const struct file_operations fuse_ctl_abort_ops = {
+	.open = nonseekable_open,
+	.write = fuse_conn_abort_write,
+};
+
+static const struct file_operations fuse_ctl_waiting_ops = {
+	.open = nonseekable_open,
+	.read = fuse_conn_waiting_read,
+};
+
+static const struct file_operations fuse_conn_max_background_ops = {
+	.open = nonseekable_open,
+	.read = fuse_conn_max_background_read,
+	.write = fuse_conn_max_background_write,
+};
+
+static const struct file_operations fuse_conn_congestion_threshold_ops = {
+	.open = nonseekable_open,
+	.read = fuse_conn_congestion_threshold_read,
+	.write = fuse_conn_congestion_threshold_write,
+};
+
+static struct dentry *fuse_ctl_add_dentry(struct dentry *parent,
+					  struct fuse_conn *fc,
+					  const char *name, int mode,
+					  const struct inode_operations *iop,
+					  const struct file_operations *fop)
+{
+	struct dentry *dentry;
+	struct inode *inode;
+
+	dentry = d_alloc_name(parent, name);
+	if (!dentry)
+		return NULL;
+
+	inode = new_inode(fuse_control_sb);
+	if (!inode) {
+		dput(dentry);
+		return NULL;
+	}
+
+	inode->i_ino = get_next_ino();
+	inode->i_mode = mode;
+	inode->i_uid = fc->user_id;
+	inode->i_gid = fc->group_id;
+	simple_inode_init_ts(inode);
+	/* setting ->i_op to NULL is not allowed */
+	if (iop)
+		inode->i_op = iop;
+	inode->i_fop = fop;
+	if (S_ISDIR(mode)) {
+		inc_nlink(d_inode(parent));
+		inc_nlink(inode);
+	}
+	inode->i_private = fc;
+	d_make_persistent(dentry, inode);
+	dput(dentry);
+
+	/*
+	 * We are returning a borrowed reference here - it's only good while
+	 * fuse_mutex is held.  Actually it's d_make_persistent() return
+	 * value...
+	 */
+	return dentry;
+}
+
+/*
+ * Add a connection to the control filesystem (if it exists).  Caller
+ * must hold fuse_mutex
+ */
+int fuse_ctl_add_conn(struct fuse_conn *fc)
+{
+	struct dentry *parent;
+	char name[32];
+
+	if (!fuse_control_sb || fc->no_control)
+		return 0;
+
+	parent = fuse_control_sb->s_root;
+	sprintf(name, "%u", fc->dev);
+	parent = fuse_ctl_add_dentry(parent, fc, name, S_IFDIR | 0500,
+				     &simple_dir_inode_operations,
+				     &simple_dir_operations);
+	if (!parent)
+		goto err;
+
+	if (!fuse_ctl_add_dentry(parent, fc, "waiting", S_IFREG | 0400,
+				 NULL, &fuse_ctl_waiting_ops) ||
+	    !fuse_ctl_add_dentry(parent, fc, "abort", S_IFREG | 0200,
+				 NULL, &fuse_ctl_abort_ops) ||
+	    !fuse_ctl_add_dentry(parent, fc, "max_background", S_IFREG | 0600,
+				 NULL, &fuse_conn_max_background_ops) ||
+	    !fuse_ctl_add_dentry(parent, fc, "congestion_threshold",
+				 S_IFREG | 0600, NULL,
+				 &fuse_conn_congestion_threshold_ops))
+		goto err;
+
+	return 0;
+
+ err:
+	fuse_ctl_remove_conn(fc);
+	return -ENOMEM;
+}
+
+static void remove_one(struct dentry *dentry)
+{
+	d_inode(dentry)->i_private = NULL;
+}
+
+/*
+ * Remove a connection from the control filesystem (if it exists).
+ * Caller must hold fuse_mutex
+ */
+void fuse_ctl_remove_conn(struct fuse_conn *fc)
+{
+	char name[32];
+
+	if (!fuse_control_sb || fc->no_control)
+		return;
+
+	sprintf(name, "%u", fc->dev);
+	simple_remove_by_name(fuse_control_sb->s_root, name, remove_one);
+}
+
+static int fuse_ctl_fill_super(struct super_block *sb, struct fs_context *fsc)
+{
+	static const struct tree_descr empty_descr = {""};
+	struct fuse_conn *fc;
+	int err;
+
+	err = simple_fill_super(sb, FUSE_CTL_SUPER_MAGIC, &empty_descr);
+	if (err)
+		return err;
+
+	mutex_lock(&fuse_mutex);
+	BUG_ON(fuse_control_sb);
+	fuse_control_sb = sb;
+	list_for_each_entry(fc, &fuse_conn_list, entry) {
+		err = fuse_ctl_add_conn(fc);
+		if (err) {
+			fuse_control_sb = NULL;
+			mutex_unlock(&fuse_mutex);
+			return err;
+		}
+	}
+	mutex_unlock(&fuse_mutex);
+
+	return 0;
+}
+
+static int fuse_ctl_get_tree(struct fs_context *fsc)
+{
+	return get_tree_single(fsc, fuse_ctl_fill_super);
+}
+
+static const struct fs_context_operations fuse_ctl_context_ops = {
+	.get_tree	= fuse_ctl_get_tree,
+};
+
+static int fuse_ctl_init_fs_context(struct fs_context *fsc)
+{
+	fsc->ops = &fuse_ctl_context_ops;
+	return 0;
+}
+
+static void fuse_ctl_kill_sb(struct super_block *sb)
+{
+	mutex_lock(&fuse_mutex);
+	fuse_control_sb = NULL;
+	mutex_unlock(&fuse_mutex);
+
+	kill_anon_super(sb);
+}
+
+static struct file_system_type fuse_ctl_fs_type = {
+	.owner		= THIS_MODULE,
+	.name		= "fusectl",
+	.init_fs_context = fuse_ctl_init_fs_context,
+	.kill_sb	= fuse_ctl_kill_sb,
+};
+MODULE_ALIAS_FS("fusectl");
+
+int __init fuse_ctl_init(void)
+{
+	return register_filesystem(&fuse_ctl_fs_type);
+}
+
+void __exit fuse_ctl_cleanup(void)
+{
+	unregister_filesystem(&fuse_ctl_fs_type);
+}

linux-7.2.3/fs/fuse/cuse.c

@@ -0,0 +1,663 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * CUSE: Character device in Userspace
+ *
+ * Copyright (C) 2008-2009  SUSE Linux Products GmbH
+ * Copyright (C) 2008-2009  Tejun Heo <tj@kernel.org>
+ *
+ * CUSE enables character devices to be implemented from userland much
+ * like FUSE allows filesystems.  On initialization /dev/cuse is
+ * created.  By opening the file and replying to the CUSE_INIT request
+ * userland CUSE server can create a character device.  After that the
+ * operation is very similar to FUSE.
+ *
+ * A CUSE instance involves the following objects.
+ *
+ * cuse_conn	: contains fuse_conn and serves as bonding structure
+ * channel	: file handle connected to the userland CUSE server
+ * cdev		: the implemented character device
+ * dev		: generic device for cdev
+ *
+ * Note that 'channel' is what 'dev' is in FUSE.  As CUSE deals with
+ * devices, it's called 'channel' to reduce confusion.
+ *
+ * channel determines when the character device dies.  When channel is
+ * closed, everything begins to destruct.  The cuse_conn is taken off
+ * the lookup table preventing further access from cdev, cdev and
+ * generic device are removed and the base reference of cuse_conn is
+ * put.
+ *
+ * On each open, the matching cuse_conn is looked up and if found an
+ * additional reference is taken which is released when the file is
+ * closed.
+ */
+
+#define pr_fmt(fmt) "CUSE: " fmt
+
+#include <linux/fuse.h>
+#include <linux/cdev.h>
+#include <linux/device.h>
+#include <linux/file.h>
+#include <linux/fs.h>
+#include <linux/kdev_t.h>
+#include <linux/kthread.h>
+#include <linux/list.h>
+#include <linux/magic.h>
+#include <linux/miscdevice.h>
+#include <linux/mutex.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/module.h>
+#include <linux/uio.h>
+#include <linux/user_namespace.h>
+
+#include "dev.h"
+#include "fuse_i.h"
+#include "fuse_dev_i.h"
+
+#define CUSE_CONNTBL_LEN	64
+
+struct cuse_conn {
+	struct list_head	list;	/* linked on cuse_conntbl */
+	struct fuse_mount	fm;	/* Dummy mount referencing fc */
+	struct fuse_conn	fc;	/* fuse connection */
+	struct cdev		*cdev;	/* associated character device */
+	struct device		*dev;	/* device representing @cdev */
+
+	/* init parameters, set once during initialization */
+	bool			unrestricted_ioctl;
+};
+
+static DEFINE_MUTEX(cuse_lock);		/* protects registration */
+static struct list_head cuse_conntbl[CUSE_CONNTBL_LEN];
+static struct class *cuse_class;
+
+static struct cuse_conn *fc_to_cc(struct fuse_conn *fc)
+{
+	return container_of(fc, struct cuse_conn, fc);
+}
+
+static struct list_head *cuse_conntbl_head(dev_t devt)
+{
+	return &cuse_conntbl[(MAJOR(devt) + MINOR(devt)) % CUSE_CONNTBL_LEN];
+}
+
+
+/**************************************************************************
+ * CUSE frontend operations
+ *
+ * These are file operations for the character device.
+ *
+ * On open, CUSE opens a file from the FUSE mnt and stores it to
+ * private_data of the open file.  All other ops call FUSE ops on the
+ * FUSE file.
+ */
+
+static ssize_t cuse_read_iter(struct kiocb *kiocb, struct iov_iter *to)
+{
+	struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(kiocb);
+	loff_t pos = 0;
+
+	return fuse_direct_io(&io, to, &pos, FUSE_DIO_CUSE);
+}
+
+static ssize_t cuse_write_iter(struct kiocb *kiocb, struct iov_iter *from)
+{
+	struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(kiocb);
+	loff_t pos = 0;
+	/*
+	 * No locking or generic_write_checks(), the server is
+	 * responsible for locking and sanity checks.
+	 */
+	return fuse_direct_io(&io, from, &pos,
+			      FUSE_DIO_WRITE | FUSE_DIO_CUSE);
+}
+
+static int cuse_open(struct inode *inode, struct file *file)
+{
+	dev_t devt = inode->i_cdev->dev;
+	struct cuse_conn *cc = NULL, *pos;
+	int rc;
+
+	/* look up and get the connection */
+	mutex_lock(&cuse_lock);
+	list_for_each_entry(pos, cuse_conntbl_head(devt), list)
+		if (pos->dev->devt == devt) {
+			fuse_conn_get(&pos->fc);
+			cc = pos;
+			break;
+		}
+	mutex_unlock(&cuse_lock);
+
+	/* dead? */
+	if (!cc)
+		return -ENODEV;
+
+	/*
+	 * Generic permission check is already done against the chrdev
+	 * file, proceed to open.
+	 */
+	rc = fuse_do_open(&cc->fm, 0, file, 0);
+	if (rc)
+		fuse_conn_put(&cc->fc);
+	return rc;
+}
+
+static int cuse_release(struct inode *inode, struct file *file)
+{
+	struct fuse_file *ff = file->private_data;
+	struct fuse_mount *fm = ff->fm;
+
+	fuse_sync_release(NULL, ff, file->f_flags);
+	fuse_conn_put(fm->fc);
+
+	return 0;
+}
+
+static long cuse_file_ioctl(struct file *file, unsigned int cmd,
+			    unsigned long arg)
+{
+	struct fuse_file *ff = file->private_data;
+	struct cuse_conn *cc = fc_to_cc(ff->fm->fc);
+	unsigned int flags = 0;
+
+	if (cc->unrestricted_ioctl)
+		flags |= FUSE_IOCTL_UNRESTRICTED;
+
+	return fuse_do_ioctl(file, cmd, arg, flags);
+}
+
+static long cuse_file_compat_ioctl(struct file *file, unsigned int cmd,
+				   unsigned long arg)
+{
+	struct fuse_file *ff = file->private_data;
+	struct cuse_conn *cc = fc_to_cc(ff->fm->fc);
+	unsigned int flags = FUSE_IOCTL_COMPAT;
+
+	if (cc->unrestricted_ioctl)
+		flags |= FUSE_IOCTL_UNRESTRICTED;
+
+	return fuse_do_ioctl(file, cmd, arg, flags);
+}
+
+static const struct file_operations cuse_frontend_fops = {
+	.owner			= THIS_MODULE,
+	.read_iter		= cuse_read_iter,
+	.write_iter		= cuse_write_iter,
+	.open			= cuse_open,
+	.release		= cuse_release,
+	.unlocked_ioctl		= cuse_file_ioctl,
+	.compat_ioctl		= cuse_file_compat_ioctl,
+	.poll			= fuse_file_poll,
+	.llseek		= noop_llseek,
+};
+
+
+/**************************************************************************
+ * CUSE channel initialization and destruction
+ */
+
+struct cuse_devinfo {
+	const char		*name;
+};
+
+/**
+ * cuse_parse_one - parse one key=value pair
+ * @pp: i/o parameter for the current position
+ * @end: points to one past the end of the packed string
+ * @keyp: out parameter for key
+ * @valp: out parameter for value
+ *
+ * *@pp points to packed strings - "key0=val0\0key1=val1\0" which ends
+ * at @end - 1.  This function parses one pair and set *@keyp to the
+ * start of the key and *@valp to the start of the value.  Note that
+ * the original string is modified such that the key string is
+ * terminated with '\0'.  *@pp is updated to point to the next string.
+ *
+ * RETURNS:
+ * 1 on successful parse, 0 on EOF, -errno on failure.
+ */
+static int cuse_parse_one(char **pp, char *end, char **keyp, char **valp)
+{
+	char *p = *pp;
+	char *key, *val;
+
+	while (p < end && *p == '\0')
+		p++;
+	if (p == end)
+		return 0;
+
+	if (end[-1] != '\0') {
+		pr_err("info not properly terminated\n");
+		return -EINVAL;
+	}
+
+	key = val = p;
+	p += strlen(p);
+
+	if (valp) {
+		strsep(&val, "=");
+		if (!val)
+			val = key + strlen(key);
+		key = strstrip(key);
+		val = strstrip(val);
+	} else
+		key = strstrip(key);
+
+	if (!strlen(key)) {
+		pr_err("zero length info key specified\n");
+		return -EINVAL;
+	}
+
+	*pp = p;
+	*keyp = key;
+	if (valp)
+		*valp = val;
+
+	return 1;
+}
+
+/**
+ * cuse_parse_devinfo - parse device info
+ * @p: device info string
+ * @len: length of device info string
+ * @devinfo: out parameter for parsed device info
+ *
+ * Parse @p to extract device info and store it into @devinfo.  String
+ * pointed to by @p is modified by parsing and @devinfo points into
+ * them, so @p shouldn't be freed while @devinfo is in use.
+ *
+ * RETURNS:
+ * 0 on success, -errno on failure.
+ */
+static int cuse_parse_devinfo(char *p, size_t len, struct cuse_devinfo *devinfo)
+{
+	char *end = p + len;
+	char *key, *val;
+	int rc;
+
+	while (true) {
+		rc = cuse_parse_one(&p, end, &key, &val);
+		if (rc < 0)
+			return rc;
+		if (!rc)
+			break;
+		if (strcmp(key, "DEVNAME") == 0)
+			devinfo->name = val;
+		else
+			pr_warn("unknown device info \"%s\"\n", key);
+	}
+
+	if (!devinfo->name || !strlen(devinfo->name)) {
+		pr_err("DEVNAME unspecified\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void cuse_gendev_release(struct device *dev)
+{
+	kfree(dev);
+}
+
+struct cuse_init_args {
+	struct fuse_args_pages ap;
+	struct cuse_init_in in;
+	struct cuse_init_out out;
+	struct folio *folio;
+	struct fuse_folio_desc desc;
+	struct fuse_conn *fc;
+};
+
+/**
+ * cuse_process_init_reply - finish initializing CUSE channel
+ *
+ * @fm: The fuse mount information containing the CUSE connection.
+ * @args: The arguments passed to the init reply.
+ * @error: The error code signifying if any error occurred during the process.
+ *
+ * This function creates the character device and sets up all the
+ * required data structures for it.  Please read the comment at the
+ * top of this file for high level overview.
+ */
+static void cuse_process_init_reply(struct fuse_args *args, int error)
+{
+	struct cuse_init_args *ia = container_of(args, typeof(*ia), ap.args);
+	struct fuse_conn *fc = ia->fc;
+	struct fuse_args_pages *ap = &ia->ap;
+	struct cuse_conn *cc = fc_to_cc(fc), *pos;
+	struct cuse_init_out *arg = &ia->out;
+	struct folio *folio = ap->folios[0];
+	struct cuse_devinfo devinfo = { };
+	struct device *dev;
+	struct cdev *cdev;
+	dev_t devt;
+	int rc, i;
+
+	if (error || arg->major != FUSE_KERNEL_VERSION || arg->minor < 11)
+		goto err;
+
+	fc->minor = arg->minor;
+	fc->max_read = max_t(unsigned, arg->max_read, 4096);
+	fc->max_write = max_t(unsigned, arg->max_write, 4096);
+
+	/* parse init reply */
+	cc->unrestricted_ioctl = arg->flags & CUSE_UNRESTRICTED_IOCTL;
+
+	rc = cuse_parse_devinfo(folio_address(folio), ap->args.out_args[1].size,
+				&devinfo);
+	if (rc)
+		goto err;
+
+	/* determine and reserve devt */
+	devt = MKDEV(arg->dev_major, arg->dev_minor);
+	if (!MAJOR(devt))
+		rc = alloc_chrdev_region(&devt, MINOR(devt), 1, devinfo.name);
+	else
+		rc = register_chrdev_region(devt, 1, devinfo.name);
+	if (rc) {
+		pr_err("failed to register chrdev region\n");
+		goto err;
+	}
+
+	/* devt determined, create device */
+	rc = -ENOMEM;
+	dev = kzalloc_obj(*dev);
+	if (!dev)
+		goto err_region;
+
+	device_initialize(dev);
+	dev_set_uevent_suppress(dev, 1);
+	dev->class = cuse_class;
+	dev->devt = devt;
+	dev->release = cuse_gendev_release;
+	dev_set_drvdata(dev, cc);
+	dev_set_name(dev, "%s", devinfo.name);
+
+	mutex_lock(&cuse_lock);
+
+	/* make sure the device-name is unique */
+	for (i = 0; i < CUSE_CONNTBL_LEN; ++i) {
+		list_for_each_entry(pos, &cuse_conntbl[i], list)
+			if (!strcmp(dev_name(pos->dev), dev_name(dev)))
+				goto err_unlock;
+	}
+
+	rc = device_add(dev);
+	if (rc)
+		goto err_unlock;
+
+	/* register cdev */
+	rc = -ENOMEM;
+	cdev = cdev_alloc();
+	if (!cdev)
+		goto err_dev;
+
+	cdev->owner = THIS_MODULE;
+	cdev->ops = &cuse_frontend_fops;
+
+	rc = cdev_add(cdev, devt, 1);
+	if (rc)
+		goto err_cdev;
+
+	cc->dev = dev;
+	cc->cdev = cdev;
+
+	/* make the device available */
+	list_add(&cc->list, cuse_conntbl_head(devt));
+	mutex_unlock(&cuse_lock);
+
+	/* announce device availability */
+	dev_set_uevent_suppress(dev, 0);
+	kobject_uevent(&dev->kobj, KOBJ_ADD);
+out:
+	kfree(ia);
+	folio_put(folio);
+	return;
+
+err_cdev:
+	cdev_del(cdev);
+err_dev:
+	device_del(dev);
+err_unlock:
+	mutex_unlock(&cuse_lock);
+	put_device(dev);
+err_region:
+	unregister_chrdev_region(devt, 1);
+err:
+	fuse_chan_abort(fc->chan, false);
+	goto out;
+}
+
+static int cuse_send_init(struct cuse_conn *cc)
+{
+	int rc;
+	struct folio *folio;
+	struct fuse_mount *fm = &cc->fm;
+	struct cuse_init_args *ia;
+	struct fuse_args_pages *ap;
+
+	BUILD_BUG_ON(CUSE_INIT_INFO_MAX > PAGE_SIZE);
+
+	rc = -ENOMEM;
+
+	folio = folio_alloc(GFP_KERNEL | __GFP_ZERO, 0);
+	if (!folio)
+		goto err;
+
+	ia = kzalloc_obj(*ia);
+	if (!ia)
+		goto err_free_folio;
+
+	ap = &ia->ap;
+	ia->in.major = FUSE_KERNEL_VERSION;
+	ia->in.minor = FUSE_KERNEL_MINOR_VERSION;
+	ia->in.flags |= CUSE_UNRESTRICTED_IOCTL;
+	ap->args.opcode = CUSE_INIT;
+	ap->args.in_numargs = 1;
+	ap->args.in_args[0].size = sizeof(ia->in);
+	ap->args.in_args[0].value = &ia->in;
+	ap->args.out_numargs = 2;
+	ap->args.out_args[0].size = sizeof(ia->out);
+	ap->args.out_args[0].value = &ia->out;
+	ap->args.out_args[1].size = CUSE_INIT_INFO_MAX;
+	ap->args.out_argvar = true;
+	ap->args.out_pages = true;
+	ap->num_folios = 1;
+	ap->folios = &ia->folio;
+	ap->descs = &ia->desc;
+	ia->folio = folio;
+	ia->desc.length = ap->args.out_args[1].size;
+	ia->fc = &cc->fc;
+	ap->args.end = cuse_process_init_reply;
+
+	rc = fuse_simple_background(fm, &ap->args, GFP_KERNEL);
+	if (rc) {
+		kfree(ia);
+err_free_folio:
+		folio_put(folio);
+	}
+err:
+	return rc;
+}
+
+static void cuse_fc_release(struct fuse_conn *fc)
+{
+	kfree(fc_to_cc(fc));
+}
+
+/**
+ * cuse_channel_open - open method for /dev/cuse
+ * @inode: inode for /dev/cuse
+ * @file: file struct being opened
+ *
+ * Userland CUSE server can create a CUSE device by opening /dev/cuse
+ * and replying to the initialization request kernel sends.  This
+ * function is responsible for handling CUSE device initialization.
+ * Because the fd opened by this function is used during
+ * initialization, this function only creates cuse_conn and sends
+ * init.  The rest is delegated to a kthread.
+ *
+ * RETURNS:
+ * 0 on success, -errno on failure.
+ */
+static int cuse_channel_open(struct inode *inode, struct file *file)
+{
+	struct fuse_dev *fud;
+	struct cuse_conn *cc;
+	struct fuse_chan *fch __free(fuse_chan_free) = fuse_dev_chan_new();
+	int rc;
+
+	if (!fch)
+		return -ENOMEM;
+
+	/* set up cuse_conn */
+	cc = kzalloc_obj(*cc);
+	if (!cc)
+		return -ENOMEM;
+
+	/*
+	 * Limit the cuse channel to requests that can
+	 * be represented in file->f_cred->user_ns.
+	 */
+	fuse_conn_init(&cc->fc, &cc->fm, file->f_cred->user_ns, no_free_ptr(fch));
+	cc->fc.release = cuse_fc_release;
+	fud = fuse_dev_alloc_install(cc->fc.chan);
+	fuse_conn_put(&cc->fc);
+	if (!fud)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&cc->list);
+
+	cc->fc.chan->initialized = 1;
+	rc = cuse_send_init(cc);
+	if (rc) {
+		fuse_dev_put(fud);
+		return rc;
+	}
+	file->private_data = fud;
+
+	return 0;
+}
+
+/**
+ * cuse_channel_release - release method for /dev/cuse
+ * @inode: inode for /dev/cuse
+ * @file: file struct being closed
+ *
+ * Disconnect the channel, deregister CUSE device and initiate
+ * destruction by putting the default reference.
+ *
+ * RETURNS:
+ * 0 on success, -errno on failure.
+ */
+static int cuse_channel_release(struct inode *inode, struct file *file)
+{
+	struct fuse_dev *fud = __fuse_get_dev(file);
+	struct cuse_conn *cc = fc_to_cc(fud->chan->conn);
+
+	/* remove from the conntbl, no more access from this point on */
+	mutex_lock(&cuse_lock);
+	list_del_init(&cc->list);
+	mutex_unlock(&cuse_lock);
+
+	/* remove device */
+	if (cc->dev)
+		device_unregister(cc->dev);
+	if (cc->cdev) {
+		unregister_chrdev_region(cc->cdev->dev, 1);
+		cdev_del(cc->cdev);
+	}
+
+	return fuse_dev_release(inode, file);
+}
+
+static struct file_operations cuse_channel_fops; /* initialized during init */
+
+
+/**************************************************************************
+ * Misc stuff and module initializatiion
+ *
+ * CUSE exports the same set of attributes to sysfs as fusectl.
+ */
+
+static ssize_t cuse_class_waiting_show(struct device *dev,
+				       struct device_attribute *attr, char *buf)
+{
+	struct cuse_conn *cc = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%d\n", atomic_read(&cc->fc.chan->num_waiting));
+}
+static DEVICE_ATTR(waiting, 0400, cuse_class_waiting_show, NULL);
+
+static ssize_t cuse_class_abort_store(struct device *dev,
+				      struct device_attribute *attr,
+				      const char *buf, size_t count)
+{
+	struct cuse_conn *cc = dev_get_drvdata(dev);
+
+	fuse_chan_abort(cc->fc.chan, false);
+	return count;
+}
+static DEVICE_ATTR(abort, 0200, NULL, cuse_class_abort_store);
+
+static struct attribute *cuse_class_dev_attrs[] = {
+	&dev_attr_waiting.attr,
+	&dev_attr_abort.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(cuse_class_dev);
+
+static struct miscdevice cuse_miscdev = {
+	.minor		= CUSE_MINOR,
+	.name		= "cuse",
+	.fops		= &cuse_channel_fops,
+};
+
+MODULE_ALIAS_MISCDEV(CUSE_MINOR);
+MODULE_ALIAS("devname:cuse");
+
+static int __init cuse_init(void)
+{
+	int i, rc;
+
+	/* init conntbl */
+	for (i = 0; i < CUSE_CONNTBL_LEN; i++)
+		INIT_LIST_HEAD(&cuse_conntbl[i]);
+
+	/* inherit and extend fuse_dev_operations */
+	cuse_channel_fops		= fuse_dev_operations;
+	cuse_channel_fops.owner		= THIS_MODULE;
+	cuse_channel_fops.open		= cuse_channel_open;
+	cuse_channel_fops.release	= cuse_channel_release;
+	/* CUSE is not prepared for FUSE_DEV_IOC_CLONE */
+	cuse_channel_fops.unlocked_ioctl	= NULL;
+
+	cuse_class = class_create("cuse");
+	if (IS_ERR(cuse_class))
+		return PTR_ERR(cuse_class);
+
+	cuse_class->dev_groups = cuse_class_dev_groups;
+
+	rc = misc_register(&cuse_miscdev);
+	if (rc) {
+		class_destroy(cuse_class);
+		return rc;
+	}
+
+	return 0;
+}
+
+static void __exit cuse_exit(void)
+{
+	misc_deregister(&cuse_miscdev);
+	class_destroy(cuse_class);
+}
+
+module_init(cuse_init);
+module_exit(cuse_exit);
+
+MODULE_AUTHOR("Tejun Heo <tj@kernel.org>");
+MODULE_DESCRIPTION("Character device in Userspace");
+MODULE_LICENSE("GPL");

linux-7.2.3/fs/fuse/dev.h

@@ -0,0 +1,110 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef _FS_FUSE_DEV_H
+#define _FS_FUSE_DEV_H
+
+#include <linux/cleanup.h>
+
+/** Maximum number of outstanding background requests */
+#define FUSE_DEFAULT_MAX_BACKGROUND 12
+
+struct fuse_conn;
+struct fuse_chan;
+struct fuse_dev;
+struct fuse_args;
+struct fuse_copy_state;
+struct fuse_backing_map;
+struct file;
+struct folio;
+enum fuse_notify_code;
+
+struct fuse_chan_param {
+	unsigned int minor;
+	unsigned int max_write;
+	unsigned int max_pages;
+};
+
+struct fuse_chan *fuse_chan_new(void);
+struct fuse_chan *fuse_dev_chan_new(void);
+void fuse_chan_release(struct fuse_chan *fch);
+void fuse_chan_free(struct fuse_chan *fch);
+unsigned int fuse_chan_num_background(struct fuse_chan *fch);
+unsigned int fuse_chan_max_background(struct fuse_chan *fch);
+void fuse_chan_max_background_set(struct fuse_chan *fch, unsigned int val);
+unsigned int fuse_chan_num_waiting(struct fuse_chan *fch);
+void fuse_chan_set_fc(struct fuse_chan *fch, struct fuse_conn *fc);
+void fuse_chan_set_initialized(struct fuse_chan *fch, struct fuse_chan_param *param);
+void fuse_chan_io_uring_enable(struct fuse_chan *fch);
+ssize_t fuse_chan_send(struct fuse_chan *fch, struct fuse_args *args);
+int fuse_chan_send_bg(struct fuse_chan *fch, struct fuse_args *args, gfp_t gfp_flags);
+int fuse_chan_send_notify_reply(struct fuse_chan *fch, struct fuse_args *args, u64 unique);
+void fuse_chan_resend(struct fuse_chan *fch);
+
+struct fuse_forget_link *fuse_alloc_forget(void);
+void fuse_chan_queue_forget(struct fuse_chan *fch, struct fuse_forget_link *forget,
+			    u64 nodeid, u64 nlookup);
+
+DEFINE_FREE(fuse_chan_free, struct fuse_chan *, if (_T) fuse_chan_free(_T))
+
+/**
+ * Initialize the client device
+ */
+int fuse_dev_init(void);
+
+/**
+ * Cleanup the client device
+ */
+void fuse_dev_cleanup(void);
+
+void fuse_dev_install(struct fuse_dev *fud, struct fuse_chan *fch);
+bool fuse_dev_verify(struct fuse_dev *fud, struct fuse_chan *fch);
+void fuse_dev_put(struct fuse_dev *fud);
+bool fuse_dev_is_installed(struct fuse_dev *fud);
+bool fuse_dev_is_sync_init(struct fuse_dev *fud);
+struct fuse_dev *fuse_dev_grab(struct file *file);
+
+void fuse_init_server_timeout(struct fuse_chan *fch, unsigned int timeout);
+
+/* Abort all requests */
+void fuse_chan_abort(struct fuse_chan *fch, bool abort_with_err);
+void fuse_chan_wait_aborted(struct fuse_chan *fch);
+
+/**
+ * Acquire reference to fuse_conn
+ */
+struct fuse_conn *fuse_conn_get(struct fuse_conn *fc);
+
+/**
+ * Release reference to fuse_conn
+ */
+void fuse_conn_put(struct fuse_conn *fc);
+
+dev_t fuse_conn_get_id(struct fuse_conn *fc);
+
+void fuse_end_polls(struct fuse_conn *fc);
+int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code,
+		unsigned int size, struct fuse_copy_state *cs);
+
+int fuse_backing_open(struct fuse_conn *fc, struct fuse_backing_map *map);
+int fuse_backing_close(struct fuse_conn *fc, int backing_id);
+
+int fuse_copy_one(struct fuse_copy_state *cs, void *val, unsigned size);
+int fuse_copy_folio(struct fuse_copy_state *cs, struct folio **foliop,
+		    unsigned offset, unsigned count, int zeroing);
+void fuse_copy_finish(struct fuse_copy_state *cs);
+
+#ifdef CONFIG_FUSE_IO_URING
+bool fuse_uring_enabled(void);
+void fuse_uring_destruct(struct fuse_chan *fch);
+#else /* CONFIG_FUSE_IO_URING */
+static inline bool fuse_uring_enabled(void)
+{
+	return false;
+}
+
+static inline void fuse_uring_destruct(struct fuse_chan *fch)
+{
+}
+#endif /* CONFIG_FUSE_IO_URING */
+
+#endif /* _FS_FUSE_DEV_H */

linux-7.2.3/fs/fuse/dev_uring_i.h

@@ -0,0 +1,202 @@
+/* SPDX-License-Identifier: GPL-2.0
+ *
+ * FUSE: Filesystem in Userspace
+ * Copyright (c) 2023-2024 DataDirect Networks.
+ */
+
+#ifndef _FS_FUSE_DEV_URING_I_H
+#define _FS_FUSE_DEV_URING_I_H
+
+#include "fuse_dev_i.h"
+
+#ifdef CONFIG_FUSE_IO_URING
+
+#define FUSE_URING_TEARDOWN_TIMEOUT (5 * HZ)
+#define FUSE_URING_TEARDOWN_INTERVAL (HZ/20)
+
+enum fuse_ring_req_state {
+	FRRS_INVALID = 0,
+
+	/* The ring entry received from userspace and it is being processed */
+	FRRS_COMMIT,
+
+	/* The ring entry is waiting for new fuse requests */
+	FRRS_AVAILABLE,
+
+	/* The ring entry got assigned a fuse req */
+	FRRS_FUSE_REQ,
+
+	/* The ring entry is in or on the way to user space */
+	FRRS_USERSPACE,
+
+	/* The ring entry is in teardown */
+	FRRS_TEARDOWN,
+
+	/* The ring entry is released, but not freed yet */
+	FRRS_RELEASED,
+};
+
+/** A fuse ring entry, part of the ring queue */
+struct fuse_ring_ent {
+	/* userspace buffer */
+	struct fuse_uring_req_header __user *headers;
+	void __user *payload;
+
+	/* the ring queue that owns the request */
+	struct fuse_ring_queue *queue;
+
+	/* fields below are protected by queue->lock */
+
+	struct io_uring_cmd *cmd;
+
+	struct list_head list;
+
+	enum fuse_ring_req_state state;
+
+	struct fuse_req *fuse_req;
+};
+
+struct fuse_ring_queue {
+	/*
+	 * back pointer to the main fuse uring structure that holds this
+	 * queue
+	 */
+	struct fuse_ring *ring;
+
+	/* queue id, corresponds to the cpu core */
+	unsigned int qid;
+
+	/*
+	 * queue lock, taken when any value in the queue changes _and_ also
+	 * a ring entry state changes.
+	 */
+	spinlock_t lock;
+
+	/* available ring entries (struct fuse_ring_ent) */
+	struct list_head ent_avail_queue;
+
+	/*
+	 * entries in the process of being committed or in the process
+	 * to be sent to userspace
+	 */
+	struct list_head ent_w_req_queue;
+	struct list_head ent_commit_queue;
+
+	/* entries in userspace */
+	struct list_head ent_in_userspace;
+
+	/* entries that are released */
+	struct list_head ent_released;
+
+	/* fuse requests waiting for an entry slot */
+	struct list_head fuse_req_queue;
+
+	/* background fuse requests */
+	struct list_head fuse_req_bg_queue;
+
+	struct fuse_pqueue fpq;
+
+	unsigned int active_background;
+
+	bool stopped;
+};
+
+/*
+ * Describes if uring is for communication and holds alls the data needed
+ * for uring communication
+ */
+struct fuse_ring {
+	/* back pointer */
+	struct fuse_chan *chan;
+
+	/* number of ring queues */
+	size_t nr_queues;
+
+	/* maximum payload/arg size */
+	size_t max_payload_sz;
+
+	struct fuse_ring_queue **queues;
+
+	/*
+	 * Log ring entry states on stop when entries cannot be released
+	 */
+	unsigned int stop_debug_log : 1;
+
+	wait_queue_head_t stop_waitq;
+
+	/* async tear down */
+	struct delayed_work async_teardown_work;
+
+	/* log */
+	unsigned long teardown_time;
+
+	atomic_t queue_refs;
+
+	bool ready;
+};
+
+void fuse_uring_stop_queues(struct fuse_ring *ring);
+void fuse_uring_abort_end_requests(struct fuse_ring *ring);
+int fuse_uring_cmd(struct io_uring_cmd *cmd, unsigned int issue_flags);
+void fuse_uring_queue_fuse_req(struct fuse_iqueue *fiq, struct fuse_req *req);
+bool fuse_uring_queue_bq_req(struct fuse_req *req);
+bool fuse_uring_remove_pending_req(struct fuse_req *req);
+bool fuse_uring_request_expired(struct fuse_chan *fch);
+
+static inline void fuse_uring_abort(struct fuse_chan *fch)
+{
+	struct fuse_ring *ring = fch->ring;
+
+	if (ring == NULL)
+		return;
+
+	fuse_uring_abort_end_requests(ring);
+
+	if (atomic_read(&ring->queue_refs) > 0)
+		fuse_uring_stop_queues(ring);
+}
+
+static inline void fuse_uring_wait_stopped_queues(struct fuse_chan *fch)
+{
+	struct fuse_ring *ring = fch->ring;
+
+	if (ring)
+		wait_event(ring->stop_waitq,
+			   atomic_read(&ring->queue_refs) == 0);
+}
+
+static inline bool fuse_uring_ready(struct fuse_chan *fch)
+{
+	struct fuse_ring *ring = READ_ONCE(fch->ring);
+
+	return ring && smp_load_acquire(&ring->ready);
+}
+
+#else /* CONFIG_FUSE_IO_URING */
+
+static inline void fuse_uring_abort(struct fuse_chan *fch)
+{
+}
+
+static inline void fuse_uring_wait_stopped_queues(struct fuse_chan *fch)
+{
+}
+
+static inline bool fuse_uring_ready(struct fuse_chan *fch)
+{
+	return false;
+}
+
+static inline bool fuse_uring_remove_pending_req(struct fuse_req *req)
+{
+	return false;
+}
+
+static inline bool fuse_uring_request_expired(struct fuse_chan *fch)
+{
+	return false;
+}
+
+#endif /* CONFIG_FUSE_IO_URING */
+
+#endif /* _FS_FUSE_DEV_URING_I_H */

linux-7.2.3/fs/fuse/fuse_dev_i.h

@@ -0,0 +1,422 @@
+/* SPDX-License-Identifier: GPL-2.0
+ *
+ * FUSE: Filesystem in Userspace
+ * Copyright (C) 2001-2008  Miklos Szeredi <miklos@szeredi.hu>
+ */
+#ifndef _FS_FUSE_DEV_I_H
+#define _FS_FUSE_DEV_I_H
+
+#include <linux/fuse.h>
+#include <linux/types.h>
+#include <linux/refcount.h>
+#include <linux/wait.h>
+#include <linux/workqueue.h>
+#include <linux/fs.h>
+
+/* Ordinary requests have even IDs, while interrupts IDs are odd */
+#define FUSE_INT_REQ_BIT (1ULL << 0)
+#define FUSE_REQ_ID_STEP (1ULL << 1)
+
+struct fuse_arg;
+struct fuse_args;
+struct fuse_pqueue;
+struct fuse_iqueue;
+
+/**
+ * enum fuse_req_flag - Request flags
+ *
+ * @FR_ISREPLY:		set if the request has reply
+ * @FR_FORCE:		force sending of the request even if interrupted
+ * @FR_BACKGROUND:	request is sent in the background
+ * @FR_WAITING:		request is counted as "waiting"
+ * @FR_ABORTED:		the request was aborted
+ * @FR_INTERRUPTED:	the request has been interrupted
+ * @FR_LOCKED:		data is being copied to/from the request
+ * @FR_PENDING:		request is not yet in userspace
+ * @FR_SENT:		request is in userspace, waiting for an answer
+ * @FR_FINISHED:	request is finished
+ * @FR_PRIVATE:		request is on private list
+ * @FR_ASYNC:		request is asynchronous
+ * @FR_URING:		request is handled through fuse-io-uring
+ */
+enum fuse_req_flag {
+	FR_ISREPLY,
+	FR_FORCE,
+	FR_BACKGROUND,
+	FR_WAITING,
+	FR_ABORTED,
+	FR_INTERRUPTED,
+	FR_LOCKED,
+	FR_PENDING,
+	FR_SENT,
+	FR_FINISHED,
+	FR_PRIVATE,
+	FR_ASYNC,
+	FR_URING,
+};
+
+/**
+ * struct fuse_req - A request to the client
+ *
+ * .waitq.lock protects the following fields:
+ *   - FR_ABORTED
+ *   - FR_LOCKED (may also be modified under fpq->lock, tested under both)
+ */
+struct fuse_req {
+	/**
+	 * @list: This can be on either pending processing or io lists in
+	 * fuse_conn
+	 */
+	struct list_head list;
+
+	/** @intr_entry: Entry on the interrupts list  */
+	struct list_head intr_entry;
+
+	/** @args: Input/output arguments */
+	struct fuse_args *args;
+
+	/** @count: refcount */
+	refcount_t count;
+
+	/** @flags: Request flags, updated with test/set/clear_bit() */
+	unsigned long flags;
+
+	/** @in: The request input header */
+	struct {
+		/** @in.h: The request input header */
+		struct fuse_in_header h;
+	} in;
+
+	/** @out: The request output header */
+	struct {
+		/** @out.h: The request output header */
+		struct fuse_out_header h;
+	} out;
+
+	/** @waitq: Used to wake up the task waiting for completion of request */
+	wait_queue_head_t waitq;
+
+#if IS_ENABLED(CONFIG_VIRTIO_FS)
+	/**
+	 * @argbuf: virtio-fs's physically contiguous buffer for in and out
+	 * args
+	 */
+	void *argbuf;
+#endif
+
+	/** @chan: fuse_chan this request belongs to */
+	struct fuse_chan *chan;
+
+#ifdef CONFIG_FUSE_IO_URING
+	void *ring_entry;
+	void *ring_queue;
+#endif
+	/** @create_time: When (in jiffies) the request was created */
+	unsigned long create_time;
+};
+
+/* One forget request */
+struct fuse_forget_link {
+	struct fuse_forget_one forget_one;
+	struct fuse_forget_link *next;
+};
+
+/**
+ * struct fuse_iqueue_ops - Input queue callbacks
+ *
+ * Input queue signalling is device-specific.  For example, the /dev/fuse file
+ * uses fiq->waitq and fasync to wake processes that are waiting on queue
+ * readiness.  These callbacks allow other device types to respond to input
+ * queue activity.
+ */
+struct fuse_iqueue_ops {
+	/**
+	 * @send_forget: Send one forget
+	 */
+	void (*send_forget)(struct fuse_iqueue *fiq, struct fuse_forget_link *link);
+
+	/**
+	 * @send_interrupt: Send interrupt for request
+	 */
+	void (*send_interrupt)(struct fuse_iqueue *fiq, struct fuse_req *req);
+
+	/**
+	 * @send_req: Send one request
+	 */
+	void (*send_req)(struct fuse_iqueue *fiq, struct fuse_req *req);
+
+	/**
+	 * @release: Clean up when fuse_iqueue is destroyed
+	 */
+	void (*release)(struct fuse_iqueue *fiq);
+};
+
+struct fuse_iqueue {
+	/** Connection established */
+	unsigned connected;
+
+	/** Lock protecting accesses to members of this structure */
+	spinlock_t lock;
+
+	/** Readers of the connection are waiting on this */
+	wait_queue_head_t waitq;
+
+	/** The next unique request id */
+	u64 reqctr;
+
+	/** The list of pending requests */
+	struct list_head pending;
+
+	/** Pending interrupts */
+	struct list_head interrupts;
+
+	/** Queue of pending forgets */
+	struct fuse_forget_link forget_list_head;
+	struct fuse_forget_link *forget_list_tail;
+
+	/** Batching of FORGET requests (positive indicates FORGET batch) */
+	int forget_batch;
+
+	/** O_ASYNC requests */
+	struct fasync_struct *fasync;
+
+	/** Device-specific callbacks */
+	const struct fuse_iqueue_ops *ops;
+
+	/** Device-specific state */
+	void *priv;
+};
+
+struct fuse_chan {
+	/** Lock protecting:
+	    - devices
+	    - connected
+	    - ring
+	    - ring->queues[qid]
+	 */
+	spinlock_t lock;
+
+	/* back pointer: fc->chan->conn == fc */
+	struct fuse_conn *conn;
+
+	/** Input queue */
+	struct fuse_iqueue iq;
+
+	/** List of device instances belonging to this connection */
+	struct list_head devices;
+
+	/** Maximum number of outstanding background requests */
+	unsigned max_background;
+
+	/** Number of requests currently in the background */
+	unsigned num_background;
+
+	/** Number of background requests currently queued for userspace */
+	unsigned active_background;
+
+	/** The list of background requests set aside for later queuing */
+	struct list_head bg_queue;
+
+	/** Protects: max_background, num_background, active_background, bg_queue, blocked */
+	spinlock_t bg_lock;
+
+	/** Flag indicating that INIT reply has been received. Allocating
+	 * any fuse request will be suspended until the flag is set */
+	int initialized;
+
+	/** Flag indicating if connection is blocked.  This will be
+	    the case before the INIT reply is received, and if there
+	    are too many outstading backgrounds requests */
+	int blocked;
+
+	/** waitq for blocked connection */
+	wait_queue_head_t blocked_waitq;
+
+	/** Connection established, cleared on umount, connection
+	    abort and device release */
+	unsigned connected;
+
+	/** The number of requests waiting for completion */
+	atomic_t num_waiting;
+
+	/** Is interrupt not implemented by fs? */
+	bool no_interrupt;
+
+	/* Use io_uring for communication */
+	unsigned int io_uring;
+
+	/* Negotiated minor version */
+	unsigned int minor;
+
+	/* Maximum write size */
+	unsigned int max_write;
+
+	/* Maximum number of pages that can be used in a single request */
+	unsigned int max_pages;
+
+	/* Before being installed into fud, contains the preallocated pq array*/
+	struct list_head *pq_prealloc;
+
+	/** Connection aborted via sysfs, respond with ECONNABORTED on device I/O */
+	bool abort_with_err;
+
+#ifdef CONFIG_FUSE_IO_URING
+	/**  uring connection information*/
+	struct fuse_ring *ring;
+#endif
+
+	/** Only used if the connection opts into request timeouts */
+	struct {
+		/* Worker for checking if any requests have timed out */
+		struct delayed_work work;
+
+		/* Request timeout (in jiffies). 0 = no timeout */
+		unsigned int req_timeout;
+	} timeout;
+};
+
+#define FUSE_PQ_HASH_BITS 8
+#define FUSE_PQ_HASH_SIZE (1 << FUSE_PQ_HASH_BITS)
+
+struct fuse_pqueue {
+	/** Connection established */
+	unsigned connected;
+
+	/** Lock protecting accessess to  members of this structure */
+	spinlock_t lock;
+
+	/** Hash table of requests being processed */
+	struct list_head *processing;
+
+	/** The list of requests under I/O */
+	struct list_head io;
+};
+
+/**
+ * struct fuse_dev - Fuse device instance
+ */
+struct fuse_dev {
+	/** @ref: Reference count of this object */
+	refcount_t ref;
+
+	/** @sync_init: Issue FUSE_INIT synchronously */
+	bool sync_init;
+
+	/** @chan: Fuse channel for this device */
+	struct fuse_chan *chan;
+
+	/** @pq: Processing queue */
+	struct fuse_pqueue pq;
+
+	/** @entry: list entry on fch->devices */
+	struct list_head entry;
+};
+
+struct fuse_copy_state {
+	struct fuse_req *req;
+	struct iov_iter *iter;
+	struct pipe_buffer *pipebufs;
+	struct pipe_buffer *currbuf;
+	struct pipe_inode_info *pipe;
+	unsigned long nr_segs;
+	struct page *pg;
+	unsigned int len;
+	unsigned int offset;
+	bool write:1;
+	bool move_folios:1;
+	bool is_uring:1;
+	struct {
+		unsigned int copied_sz; /* copied size into the user buffer */
+	} ring;
+};
+
+/* fud->chan gets assigned to this value when /dev/fuse is closed */
+#define FUSE_DEV_CHAN_DISCONNECTED ((struct fuse_chan *) 1)
+
+/*
+ * Lockless access is OK, because fud->chan is set once during mount and is valid
+ * until the file is released.
+ *
+ * fud->chan is set to FUSE_DEV_CHAN_DISCONNECTED only after the containing file is
+ * released, so result is safe to dereference in most cases.  Exceptions are:
+ * fuse_dev_put() and fuse_fill_super_common().
+ */
+static inline struct fuse_chan *fuse_dev_chan_get(struct fuse_dev *fud)
+{
+	/* Pairs with xchg() in fuse_dev_install() */
+	return smp_load_acquire(&fud->chan);
+}
+
+static inline struct fuse_dev *fuse_file_to_fud(struct file *file)
+{
+	return file->private_data;
+}
+
+static inline struct fuse_dev *__fuse_get_dev(struct file *file)
+{
+	struct fuse_dev *fud = fuse_file_to_fud(file);
+
+	if (!fuse_dev_chan_get(fud))
+		return NULL;
+
+	return fud;
+}
+
+void fuse_iqueue_init(struct fuse_iqueue *fiq, const struct fuse_iqueue_ops *ops, void *priv);
+
+struct fuse_dev *fuse_get_dev(struct file *file);
+
+unsigned int fuse_req_hash(u64 unique);
+struct fuse_req *fuse_request_find(struct fuse_pqueue *fpq, u64 unique);
+
+void fuse_dev_end_requests(struct list_head *head);
+void fuse_request_bg_finish(struct fuse_chan *fch, struct fuse_req *req);
+
+void fuse_copy_init(struct fuse_copy_state *cs, bool write,
+			   struct iov_iter *iter);
+/*
+ * Return the number of bytes in an arguments list
+ */
+unsigned int fuse_len_args(unsigned int numargs, struct fuse_arg *args);
+
+int fuse_copy_args(struct fuse_copy_state *cs, unsigned int numargs,
+		   unsigned int argpages, struct fuse_arg *args,
+		   int zeroing);
+int fuse_copy_out_args(struct fuse_copy_state *cs, struct fuse_args *args,
+		       unsigned int nbytes);
+void fuse_dev_queue_forget(struct fuse_iqueue *fiq,
+			   struct fuse_forget_link *forget);
+void fuse_dev_queue_interrupt(struct fuse_iqueue *fiq, struct fuse_req *req);
+bool fuse_remove_pending_req(struct fuse_req *req, spinlock_t *lock);
+
+bool fuse_request_expired(struct fuse_chan *fch, struct list_head *list);
+
+/*
+ * Assign a unique id to a fuse request
+ */
+void fuse_request_assign_unique(struct fuse_iqueue *fiq, struct fuse_req *req);
+
+/*
+ * Get the next unique ID for a request
+ */
+u64 fuse_get_unique(struct fuse_iqueue *fiq);
+
+struct fuse_dev *fuse_dev_alloc_install(struct fuse_chan *fch);
+struct fuse_dev *fuse_dev_alloc(void);
+
+int fuse_dev_release(struct inode *inode, struct file *file);
+
+struct list_head *fuse_pqueue_alloc(void);
+
+/*
+ * Initialize the fuse processing queue
+ */
+void fuse_pqueue_init(struct fuse_pqueue *fpq);
+
+/*
+ * End a finished request
+ */
+void fuse_request_end(struct fuse_req *req);
+
+#endif
+

linux-7.2.3/fs/fuse/fuse_trace.h

@@ -0,0 +1,154 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM fuse
+
+#if !defined(_TRACE_FUSE_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_FUSE_H
+
+#include <linux/tracepoint.h>
+
+#define OPCODES							\
+	EM( FUSE_LOOKUP,		"FUSE_LOOKUP")		\
+	EM( FUSE_FORGET,		"FUSE_FORGET")		\
+	EM( FUSE_GETATTR,		"FUSE_GETATTR")		\
+	EM( FUSE_SETATTR,		"FUSE_SETATTR")		\
+	EM( FUSE_READLINK,		"FUSE_READLINK")	\
+	EM( FUSE_SYMLINK,		"FUSE_SYMLINK")		\
+	EM( FUSE_MKNOD,			"FUSE_MKNOD")		\
+	EM( FUSE_MKDIR,			"FUSE_MKDIR")		\
+	EM( FUSE_UNLINK,		"FUSE_UNLINK")		\
+	EM( FUSE_RMDIR,			"FUSE_RMDIR")		\
+	EM( FUSE_RENAME,		"FUSE_RENAME")		\
+	EM( FUSE_LINK,			"FUSE_LINK")		\
+	EM( FUSE_OPEN,			"FUSE_OPEN")		\
+	EM( FUSE_READ,			"FUSE_READ")		\
+	EM( FUSE_WRITE,			"FUSE_WRITE")		\
+	EM( FUSE_STATFS,		"FUSE_STATFS")		\
+	EM( FUSE_RELEASE,		"FUSE_RELEASE")		\
+	EM( FUSE_FSYNC,			"FUSE_FSYNC")		\
+	EM( FUSE_SETXATTR,		"FUSE_SETXATTR")	\
+	EM( FUSE_GETXATTR,		"FUSE_GETXATTR")	\
+	EM( FUSE_LISTXATTR,		"FUSE_LISTXATTR")	\
+	EM( FUSE_REMOVEXATTR,		"FUSE_REMOVEXATTR")	\
+	EM( FUSE_FLUSH,			"FUSE_FLUSH")		\
+	EM( FUSE_INIT,			"FUSE_INIT")		\
+	EM( FUSE_OPENDIR,		"FUSE_OPENDIR")		\
+	EM( FUSE_READDIR,		"FUSE_READDIR")		\
+	EM( FUSE_RELEASEDIR,		"FUSE_RELEASEDIR")	\
+	EM( FUSE_FSYNCDIR,		"FUSE_FSYNCDIR")	\
+	EM( FUSE_GETLK,			"FUSE_GETLK")		\
+	EM( FUSE_SETLK,			"FUSE_SETLK")		\
+	EM( FUSE_SETLKW,		"FUSE_SETLKW")		\
+	EM( FUSE_ACCESS,		"FUSE_ACCESS")		\
+	EM( FUSE_CREATE,		"FUSE_CREATE")		\
+	EM( FUSE_INTERRUPT,		"FUSE_INTERRUPT")	\
+	EM( FUSE_BMAP,			"FUSE_BMAP")		\
+	EM( FUSE_DESTROY,		"FUSE_DESTROY")		\
+	EM( FUSE_IOCTL,			"FUSE_IOCTL")		\
+	EM( FUSE_POLL,			"FUSE_POLL")		\
+	EM( FUSE_NOTIFY_REPLY,		"FUSE_NOTIFY_REPLY")	\
+	EM( FUSE_BATCH_FORGET,		"FUSE_BATCH_FORGET")	\
+	EM( FUSE_FALLOCATE,		"FUSE_FALLOCATE")	\
+	EM( FUSE_READDIRPLUS,		"FUSE_READDIRPLUS")	\
+	EM( FUSE_RENAME2,		"FUSE_RENAME2")		\
+	EM( FUSE_LSEEK,			"FUSE_LSEEK")		\
+	EM( FUSE_COPY_FILE_RANGE,	"FUSE_COPY_FILE_RANGE")	\
+	EM( FUSE_SETUPMAPPING,		"FUSE_SETUPMAPPING")	\
+	EM( FUSE_REMOVEMAPPING,		"FUSE_REMOVEMAPPING")	\
+	EM( FUSE_SYNCFS,		"FUSE_SYNCFS")		\
+	EM( FUSE_TMPFILE,		"FUSE_TMPFILE")		\
+	EM( FUSE_STATX,			"FUSE_STATX")		\
+	EMe(CUSE_INIT,			"CUSE_INIT")
+
+/*
+ * This will turn the above table into TRACE_DEFINE_ENUM() for each of the
+ * entries.
+ */
+#undef EM
+#undef EMe
+#define EM(a, b)	TRACE_DEFINE_ENUM(a);
+#define EMe(a, b)	TRACE_DEFINE_ENUM(a);
+
+OPCODES
+
+/* Now we redfine it with the table that __print_symbolic needs. */
+#undef EM
+#undef EMe
+#define EM(a, b)	{a, b},
+#define EMe(a, b)	{a, b}
+
+TRACE_EVENT(fuse_request_send,
+	TP_PROTO(const struct fuse_req *req),
+
+	TP_ARGS(req),
+
+	TP_STRUCT__entry(
+		__field(dev_t,			connection)
+		__field(uint64_t,		unique)
+		__field(enum fuse_opcode,	opcode)
+		__field(uint32_t,		len)
+	),
+
+	TP_fast_assign(
+		__entry->connection	=	req->chan->conn->dev;
+		__entry->unique		=	req->in.h.unique;
+		__entry->opcode		=	req->in.h.opcode;
+		__entry->len		=	req->in.h.len;
+	),
+
+	TP_printk("connection %u req %llu opcode %u (%s) len %u ",
+		  __entry->connection, __entry->unique, __entry->opcode,
+		  __print_symbolic(__entry->opcode, OPCODES), __entry->len)
+);
+
+TRACE_EVENT(fuse_request_sent,
+	TP_PROTO(const struct fuse_req *req),
+
+	TP_ARGS(req),
+
+	TP_STRUCT__entry(
+		__field(dev_t,			connection)
+		__field(uint64_t,		unique)
+		__field(enum fuse_opcode,	opcode)
+	),
+
+	TP_fast_assign(
+		__entry->connection	=	req->chan->conn->dev;
+		__entry->unique		=	req->in.h.unique;
+		__entry->opcode		=	req->in.h.opcode;
+	),
+
+	TP_printk("connection %u req %llu opcode %u (%s)",
+		  __entry->connection, __entry->unique, __entry->opcode,
+		  __print_symbolic(__entry->opcode, OPCODES))
+);
+
+TRACE_EVENT(fuse_request_end,
+	TP_PROTO(const struct fuse_req *req),
+
+	TP_ARGS(req),
+
+	TP_STRUCT__entry(
+		__field(dev_t,		connection)
+		__field(uint64_t,	unique)
+		__field(uint32_t,	len)
+		__field(int32_t,	error)
+	),
+
+	TP_fast_assign(
+		__entry->connection	=	req->chan->conn->dev;
+		__entry->unique		=	req->in.h.unique;
+		__entry->len		=	req->out.h.len;
+		__entry->error		=	req->out.h.error;
+	),
+
+	TP_printk("connection %u req %llu len %u error %d", __entry->connection,
+		  __entry->unique, __entry->len, __entry->error)
+);
+
+#endif /* _TRACE_FUSE_H */
+
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#define TRACE_INCLUDE_FILE fuse_trace
+#include <trace/define_trace.h>

linux-7.2.3/fs/fuse/ioctl.c

@@ -0,0 +1,577 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2017 Red Hat, Inc.
+ */
+
+#include "fuse_i.h"
+
+#include <linux/uio.h>
+#include <linux/compat.h>
+#include <linux/fileattr.h>
+#include <linux/fsverity.h>
+
+#include <linux/slab.h>
+#define FUSE_VERITY_ENABLE_ARG_MAX_PAGES 256
+
+static ssize_t fuse_send_ioctl(struct fuse_mount *fm, struct fuse_args *args,
+			       struct fuse_ioctl_out *outarg)
+{
+	ssize_t ret;
+
+	args->out_args[0].size = sizeof(*outarg);
+	args->out_args[0].value = outarg;
+
+	ret = fuse_simple_request(fm, args);
+
+	/* Translate ENOSYS, which shouldn't be returned from fs */
+	if (ret == -ENOSYS)
+		ret = -ENOTTY;
+
+	if (ret >= 0 && outarg->result == -ENOSYS)
+		outarg->result = -ENOTTY;
+
+	return ret;
+}
+
+/*
+ * CUSE servers compiled on 32bit broke on 64bit kernels because the
+ * ABI was defined to be 'struct iovec' which is different on 32bit
+ * and 64bit.  Fortunately we can determine which structure the server
+ * used from the size of the reply.
+ */
+static int fuse_copy_ioctl_iovec_old(struct iovec *dst, void *src,
+				     size_t transferred, unsigned count,
+				     bool is_compat)
+{
+#ifdef CONFIG_COMPAT
+	if (count * sizeof(struct compat_iovec) == transferred) {
+		struct compat_iovec *ciov = src;
+		unsigned i;
+
+		/*
+		 * With this interface a 32bit server cannot support
+		 * non-compat (i.e. ones coming from 64bit apps) ioctl
+		 * requests
+		 */
+		if (!is_compat)
+			return -EINVAL;
+
+		for (i = 0; i < count; i++) {
+			dst[i].iov_base = compat_ptr(ciov[i].iov_base);
+			dst[i].iov_len = ciov[i].iov_len;
+		}
+		return 0;
+	}
+#endif
+
+	if (count * sizeof(struct iovec) != transferred)
+		return -EIO;
+
+	memcpy(dst, src, transferred);
+	return 0;
+}
+
+/* Make sure iov_length() won't overflow */
+static int fuse_verify_ioctl_iov(struct fuse_conn *fc, struct iovec *iov,
+				 size_t count)
+{
+	size_t n;
+	u32 max = fc->max_pages << PAGE_SHIFT;
+
+	for (n = 0; n < count; n++, iov++) {
+		if (iov->iov_len > (size_t) max)
+			return -ENOMEM;
+		max -= iov->iov_len;
+	}
+	return 0;
+}
+
+static int fuse_copy_ioctl_iovec(struct fuse_conn *fc, struct iovec *dst,
+				 void *src, size_t transferred, unsigned count,
+				 bool is_compat)
+{
+	unsigned i;
+	struct fuse_ioctl_iovec *fiov = src;
+
+	if (fc->minor < 16) {
+		return fuse_copy_ioctl_iovec_old(dst, src, transferred,
+						 count, is_compat);
+	}
+
+	if (count * sizeof(struct fuse_ioctl_iovec) != transferred)
+		return -EIO;
+
+	for (i = 0; i < count; i++) {
+		/* Did the server supply an inappropriate value? */
+		if (fiov[i].base != (unsigned long) fiov[i].base ||
+		    fiov[i].len != (unsigned long) fiov[i].len)
+			return -EIO;
+
+		dst[i].iov_base = (void __user *) (unsigned long) fiov[i].base;
+		dst[i].iov_len = (size_t) fiov[i].len;
+
+#ifdef CONFIG_COMPAT
+		if (is_compat &&
+		    (ptr_to_compat(dst[i].iov_base) != fiov[i].base ||
+		     (compat_size_t) dst[i].iov_len != fiov[i].len))
+			return -EIO;
+#endif
+	}
+
+	return 0;
+}
+
+/* For fs-verity, determine iov lengths from input */
+static int fuse_setup_measure_verity(unsigned long arg, struct iovec *iov)
+{
+	__u16 digest_size;
+	struct fsverity_digest __user *uarg = (void __user *)arg;
+
+	if (copy_from_user(&digest_size, &uarg->digest_size, sizeof(digest_size)))
+		return -EFAULT;
+
+	if (digest_size > SIZE_MAX - sizeof(struct fsverity_digest))
+		return -EINVAL;
+
+	iov->iov_len = sizeof(struct fsverity_digest) + digest_size;
+
+	return 0;
+}
+
+static int fuse_setup_enable_verity(unsigned long arg, struct iovec *iov,
+				    unsigned int *in_iovs)
+{
+	struct fsverity_enable_arg enable;
+	struct fsverity_enable_arg __user *uarg = (void __user *)arg;
+	const __u32 max_buffer_len = FUSE_VERITY_ENABLE_ARG_MAX_PAGES * PAGE_SIZE;
+
+	if (copy_from_user(&enable, uarg, sizeof(enable)))
+		return -EFAULT;
+
+	if (enable.salt_size > max_buffer_len || enable.sig_size > max_buffer_len)
+		return -ENOMEM;
+
+	if (enable.salt_size > 0) {
+		iov++;
+		(*in_iovs)++;
+
+		iov->iov_base = u64_to_user_ptr(enable.salt_ptr);
+		iov->iov_len = enable.salt_size;
+	}
+
+	if (enable.sig_size > 0) {
+		iov++;
+		(*in_iovs)++;
+
+		iov->iov_base = u64_to_user_ptr(enable.sig_ptr);
+		iov->iov_len = enable.sig_size;
+	}
+	return 0;
+}
+
+/*
+ * For ioctls, there is no generic way to determine how much memory
+ * needs to be read and/or written.  Furthermore, ioctls are allowed
+ * to dereference the passed pointer, so the parameter requires deep
+ * copying but FUSE has no idea whatsoever about what to copy in or
+ * out.
+ *
+ * This is solved by allowing FUSE server to retry ioctl with
+ * necessary in/out iovecs.  Let's assume the ioctl implementation
+ * needs to read in the following structure.
+ *
+ * struct a {
+ *	char	*buf;
+ *	size_t	buflen;
+ * }
+ *
+ * On the first callout to FUSE server, inarg->in_size and
+ * inarg->out_size will be NULL; then, the server completes the ioctl
+ * with FUSE_IOCTL_RETRY set in out->flags, out->in_iovs set to 1 and
+ * the actual iov array to
+ *
+ * { { .iov_base = inarg.arg,	.iov_len = sizeof(struct a) } }
+ *
+ * which tells FUSE to copy in the requested area and retry the ioctl.
+ * On the second round, the server has access to the structure and
+ * from that it can tell what to look for next, so on the invocation,
+ * it sets FUSE_IOCTL_RETRY, out->in_iovs to 2 and iov array to
+ *
+ * { { .iov_base = inarg.arg,	.iov_len = sizeof(struct a)	},
+ *   { .iov_base = a.buf,	.iov_len = a.buflen		} }
+ *
+ * FUSE will copy both struct a and the pointed buffer from the
+ * process doing the ioctl and retry ioctl with both struct a and the
+ * buffer.
+ *
+ * This time, FUSE server has everything it needs and completes ioctl
+ * without FUSE_IOCTL_RETRY which finishes the ioctl call.
+ *
+ * Copying data out works the same way.
+ *
+ * Note that if FUSE_IOCTL_UNRESTRICTED is clear, the kernel
+ * automatically initializes in and out iovs by decoding @cmd with
+ * _IOC_* macros and the server is not allowed to request RETRY.  This
+ * limits ioctl data transfers to well-formed ioctls and is the forced
+ * behavior for all FUSE servers.
+ */
+long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg,
+		   unsigned int flags)
+{
+	struct fuse_file *ff = file->private_data;
+	struct fuse_mount *fm = ff->fm;
+	struct fuse_ioctl_in inarg = {
+		.fh = ff->fh,
+		.cmd = cmd,
+		.arg = arg,
+		.flags = flags
+	};
+	struct fuse_ioctl_out outarg;
+	struct iovec *iov_page = NULL;
+	struct iovec *in_iov = NULL, *out_iov = NULL;
+	unsigned int in_iovs = 0, out_iovs = 0, max_pages;
+	size_t in_size, out_size, c;
+	ssize_t transferred;
+	int err, i;
+	struct iov_iter ii;
+	struct fuse_args_pages ap = {};
+
+#if BITS_PER_LONG == 32
+	inarg.flags |= FUSE_IOCTL_32BIT;
+#else
+	if (flags & FUSE_IOCTL_COMPAT) {
+		inarg.flags |= FUSE_IOCTL_32BIT;
+#ifdef CONFIG_X86_X32_ABI
+		if (in_x32_syscall())
+			inarg.flags |= FUSE_IOCTL_COMPAT_X32;
+#endif
+	}
+#endif
+
+	/* assume all the iovs returned by client always fits in a page */
+	BUILD_BUG_ON(sizeof(struct fuse_ioctl_iovec) * FUSE_IOCTL_MAX_IOV > PAGE_SIZE);
+
+	err = -ENOMEM;
+	ap.folios = fuse_folios_alloc(fm->fc->max_pages, GFP_KERNEL, &ap.descs);
+	iov_page = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!ap.folios || !iov_page)
+		goto out;
+
+	fuse_folio_descs_length_init(ap.descs, 0, fm->fc->max_pages);
+
+	/*
+	 * If restricted, initialize IO parameters as encoded in @cmd.
+	 * RETRY from server is not allowed.
+	 */
+	if (!(flags & FUSE_IOCTL_UNRESTRICTED)) {
+		struct iovec *iov = iov_page;
+
+		iov->iov_base = (void __user *)arg;
+		iov->iov_len = _IOC_SIZE(cmd);
+
+		if (_IOC_DIR(cmd) & _IOC_WRITE) {
+			in_iov = iov;
+			in_iovs = 1;
+		}
+
+		if (_IOC_DIR(cmd) & _IOC_READ) {
+			out_iov = iov;
+			out_iovs = 1;
+		}
+
+		err = 0;
+		switch (cmd) {
+		case FS_IOC_MEASURE_VERITY:
+			err = fuse_setup_measure_verity(arg, iov);
+			break;
+		case FS_IOC_ENABLE_VERITY:
+			err = fuse_setup_enable_verity(arg, iov, &in_iovs);
+			break;
+		}
+		if (err)
+			goto out;
+	}
+
+ retry:
+	inarg.in_size = in_size = iov_length(in_iov, in_iovs);
+	inarg.out_size = out_size = iov_length(out_iov, out_iovs);
+
+	/*
+	 * Out data can be used either for actual out data or iovs,
+	 * make sure there always is at least one page.
+	 */
+	out_size = max_t(size_t, out_size, PAGE_SIZE);
+	max_pages = DIV_ROUND_UP(max(in_size, out_size), PAGE_SIZE);
+
+	/* make sure there are enough buffer pages and init request with them */
+	err = -ENOMEM;
+	if (max_pages > fm->fc->max_pages)
+		goto out;
+	while (ap.num_folios < max_pages) {
+		ap.folios[ap.num_folios] = folio_alloc(GFP_KERNEL | __GFP_HIGHMEM, 0);
+		if (!ap.folios[ap.num_folios])
+			goto out;
+		ap.num_folios++;
+	}
+
+	/* okay, let's send it to the client */
+	ap.args.opcode = FUSE_IOCTL;
+	ap.args.nodeid = ff->nodeid;
+	ap.args.in_numargs = 1;
+	ap.args.in_args[0].size = sizeof(inarg);
+	ap.args.in_args[0].value = &inarg;
+	if (in_size) {
+		ap.args.in_numargs++;
+		ap.args.in_args[1].size = in_size;
+		ap.args.in_pages = true;
+
+		err = -EFAULT;
+		iov_iter_init(&ii, ITER_SOURCE, in_iov, in_iovs, in_size);
+		for (i = 0; iov_iter_count(&ii) && !WARN_ON(i >= ap.num_folios); i++) {
+			c = copy_folio_from_iter(ap.folios[i], 0, PAGE_SIZE, &ii);
+			if (c != PAGE_SIZE && iov_iter_count(&ii))
+				goto out;
+		}
+	}
+
+	ap.args.out_numargs = 2;
+	ap.args.out_args[1].size = out_size;
+	ap.args.out_pages = true;
+	ap.args.out_argvar = true;
+
+	transferred = fuse_send_ioctl(fm, &ap.args, &outarg);
+	err = transferred;
+	if (transferred < 0)
+		goto out;
+
+	/* did it ask for retry? */
+	if (outarg.flags & FUSE_IOCTL_RETRY) {
+		void *vaddr;
+
+		/* no retry if in restricted mode */
+		err = -EIO;
+		if (!(flags & FUSE_IOCTL_UNRESTRICTED))
+			goto out;
+
+		in_iovs = outarg.in_iovs;
+		out_iovs = outarg.out_iovs;
+
+		/*
+		 * Make sure things are in boundary, separate checks
+		 * are to protect against overflow.
+		 */
+		err = -ENOMEM;
+		if (in_iovs > FUSE_IOCTL_MAX_IOV ||
+		    out_iovs > FUSE_IOCTL_MAX_IOV ||
+		    in_iovs + out_iovs > FUSE_IOCTL_MAX_IOV)
+			goto out;
+
+		vaddr = kmap_local_folio(ap.folios[0], 0);
+		err = fuse_copy_ioctl_iovec(fm->fc, iov_page, vaddr,
+					    transferred, in_iovs + out_iovs,
+					    (flags & FUSE_IOCTL_COMPAT) != 0);
+		kunmap_local(vaddr);
+		if (err)
+			goto out;
+
+		in_iov = iov_page;
+		out_iov = in_iov + in_iovs;
+
+		err = fuse_verify_ioctl_iov(fm->fc, in_iov, in_iovs);
+		if (err)
+			goto out;
+
+		err = fuse_verify_ioctl_iov(fm->fc, out_iov, out_iovs);
+		if (err)
+			goto out;
+
+		goto retry;
+	}
+
+	err = -EIO;
+	if (transferred > inarg.out_size)
+		goto out;
+
+	err = -EFAULT;
+	iov_iter_init(&ii, ITER_DEST, out_iov, out_iovs, transferred);
+	for (i = 0; iov_iter_count(&ii) && !WARN_ON(i >= ap.num_folios); i++) {
+		c = copy_folio_to_iter(ap.folios[i], 0, PAGE_SIZE, &ii);
+		if (c != PAGE_SIZE && iov_iter_count(&ii))
+			goto out;
+	}
+	err = 0;
+ out:
+	kfree(iov_page);
+	while (ap.num_folios)
+		folio_put(ap.folios[--ap.num_folios]);
+	kfree(ap.folios);
+
+	return err ? err : outarg.result;
+}
+EXPORT_SYMBOL_GPL(fuse_do_ioctl);
+
+long fuse_ioctl_common(struct file *file, unsigned int cmd,
+		       unsigned long arg, unsigned int flags)
+{
+	struct inode *inode = file_inode(file);
+	struct fuse_conn *fc = get_fuse_conn(inode);
+
+	if (!fuse_allow_current_process(fc))
+		return -EACCES;
+
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	return fuse_do_ioctl(file, cmd, arg, flags);
+}
+
+long fuse_file_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+	return fuse_ioctl_common(file, cmd, arg, 0);
+}
+
+long fuse_file_compat_ioctl(struct file *file, unsigned int cmd,
+			    unsigned long arg)
+{
+	return fuse_ioctl_common(file, cmd, arg, FUSE_IOCTL_COMPAT);
+}
+
+static int fuse_priv_ioctl(struct inode *inode, struct fuse_file *ff,
+			   unsigned int cmd, void *ptr, size_t size)
+{
+	struct fuse_mount *fm = ff->fm;
+	struct fuse_ioctl_in inarg;
+	struct fuse_ioctl_out outarg;
+	FUSE_ARGS(args);
+	int err;
+
+	memset(&inarg, 0, sizeof(inarg));
+	inarg.fh = ff->fh;
+	inarg.cmd = cmd;
+
+#if BITS_PER_LONG == 32
+	inarg.flags |= FUSE_IOCTL_32BIT;
+#endif
+	if (S_ISDIR(inode->i_mode))
+		inarg.flags |= FUSE_IOCTL_DIR;
+
+	if (_IOC_DIR(cmd) & _IOC_READ)
+		inarg.out_size = size;
+	if (_IOC_DIR(cmd) & _IOC_WRITE)
+		inarg.in_size = size;
+
+	args.opcode = FUSE_IOCTL;
+	args.nodeid = ff->nodeid;
+	args.in_numargs = 2;
+	args.in_args[0].size = sizeof(inarg);
+	args.in_args[0].value = &inarg;
+	args.in_args[1].size = inarg.in_size;
+	args.in_args[1].value = ptr;
+	args.out_numargs = 2;
+	args.out_args[1].size = inarg.out_size;
+	args.out_args[1].value = ptr;
+
+	err = fuse_send_ioctl(fm, &args, &outarg);
+	if (!err) {
+		if (outarg.result < 0)
+			err = outarg.result;
+		else if (outarg.flags & FUSE_IOCTL_RETRY)
+			err = -EIO;
+	}
+	return err;
+}
+
+static struct fuse_file *fuse_priv_ioctl_prepare(struct inode *inode)
+{
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	bool isdir = S_ISDIR(inode->i_mode);
+
+	if (!fuse_allow_current_process(fm->fc))
+		return ERR_PTR(-EACCES);
+
+	if (fuse_is_bad(inode))
+		return ERR_PTR(-EIO);
+
+	if (!S_ISREG(inode->i_mode) && !isdir)
+		return ERR_PTR(-ENOTTY);
+
+	return fuse_file_open(fm, get_node_id(inode), O_RDONLY, isdir);
+}
+
+static void fuse_priv_ioctl_cleanup(struct inode *inode, struct fuse_file *ff)
+{
+	fuse_file_release(inode, ff, O_RDONLY, NULL, S_ISDIR(inode->i_mode));
+}
+
+int fuse_fileattr_get(struct dentry *dentry, struct file_kattr *fa)
+{
+	struct inode *inode = d_inode(dentry);
+	struct fuse_file *ff;
+	unsigned int flags;
+	struct fsxattr xfa;
+	int err;
+
+	ff = fuse_priv_ioctl_prepare(inode);
+	if (IS_ERR(ff))
+		return PTR_ERR(ff);
+
+	if (fa->flags_valid) {
+		err = fuse_priv_ioctl(inode, ff, FS_IOC_GETFLAGS,
+				      &flags, sizeof(flags));
+		if (err)
+			goto cleanup;
+
+		fileattr_fill_flags(fa, flags);
+	} else {
+		err = fuse_priv_ioctl(inode, ff, FS_IOC_FSGETXATTR,
+				      &xfa, sizeof(xfa));
+		if (err)
+			goto cleanup;
+
+		fileattr_fill_xflags(fa, xfa.fsx_xflags);
+		fa->fsx_extsize = xfa.fsx_extsize;
+		fa->fsx_nextents = xfa.fsx_nextents;
+		fa->fsx_projid = xfa.fsx_projid;
+		fa->fsx_cowextsize = xfa.fsx_cowextsize;
+	}
+cleanup:
+	fuse_priv_ioctl_cleanup(inode, ff);
+
+	return err;
+}
+
+int fuse_fileattr_set(struct mnt_idmap *idmap,
+		      struct dentry *dentry, struct file_kattr *fa)
+{
+	struct inode *inode = d_inode(dentry);
+	struct fuse_file *ff;
+	unsigned int flags = fa->flags;
+	struct fsxattr xfa;
+	int err;
+
+	ff = fuse_priv_ioctl_prepare(inode);
+	if (IS_ERR(ff))
+		return PTR_ERR(ff);
+
+	if (fa->flags_valid) {
+		err = fuse_priv_ioctl(inode, ff, FS_IOC_SETFLAGS,
+				      &flags, sizeof(flags));
+		if (err)
+			goto cleanup;
+	} else {
+		memset(&xfa, 0, sizeof(xfa));
+		xfa.fsx_xflags = fa->fsx_xflags;
+		xfa.fsx_extsize = fa->fsx_extsize;
+		xfa.fsx_nextents = fa->fsx_nextents;
+		xfa.fsx_projid = fa->fsx_projid;
+		xfa.fsx_cowextsize = fa->fsx_cowextsize;
+
+		err = fuse_priv_ioctl(inode, ff, FS_IOC_FSSETXATTR,
+				      &xfa, sizeof(xfa));
+	}
+
+cleanup:
+	fuse_priv_ioctl_cleanup(inode, ff);
+
+	return err;
+}

linux-7.2.3/fs/fuse/iomode.c

@@ -0,0 +1,275 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * FUSE inode io modes.
+ *
+ * Copyright (c) 2024 CTERA Networks.
+ */
+
+#include "fuse_i.h"
+
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/file.h>
+#include <linux/fs.h>
+
+/*
+ * Return true if need to wait for new opens in caching mode.
+ */
+static inline bool fuse_is_io_cache_wait(struct fuse_inode *fi)
+{
+	return READ_ONCE(fi->iocachectr) < 0 && !fuse_inode_backing(fi);
+}
+
+/*
+ * Called on cached file open() and on first mmap() of direct_io file.
+ * Takes cached_io inode mode reference to be dropped on file release.
+ *
+ * Blocks new parallel dio writes and waits for the in-progress parallel dio
+ * writes to complete.
+ */
+int fuse_file_cached_io_open(struct inode *inode, struct fuse_file *ff)
+{
+	struct fuse_inode *fi = get_fuse_inode(inode);
+
+	/* There are no io modes if server does not implement open */
+	if (!ff->args)
+		return 0;
+
+	spin_lock(&fi->lock);
+	/*
+	 * Setting the bit advises new direct-io writes to use an exclusive
+	 * lock - without it the wait below might be forever.
+	 */
+	while (fuse_is_io_cache_wait(fi)) {
+		set_bit(FUSE_I_CACHE_IO_MODE, &fi->state);
+		spin_unlock(&fi->lock);
+		wait_event(fi->direct_io_waitq, !fuse_is_io_cache_wait(fi));
+		spin_lock(&fi->lock);
+	}
+
+	/*
+	 * Check if inode entered passthrough io mode while waiting for parallel
+	 * dio write completion.
+	 */
+	if (fuse_inode_backing(fi)) {
+		clear_bit(FUSE_I_CACHE_IO_MODE, &fi->state);
+		spin_unlock(&fi->lock);
+		return -ETXTBSY;
+	}
+
+	WARN_ON(ff->iomode == IOM_UNCACHED);
+	if (ff->iomode == IOM_NONE) {
+		ff->iomode = IOM_CACHED;
+		if (fi->iocachectr == 0)
+			set_bit(FUSE_I_CACHE_IO_MODE, &fi->state);
+		fi->iocachectr++;
+	}
+	spin_unlock(&fi->lock);
+	return 0;
+}
+
+static void fuse_file_cached_io_release(struct fuse_file *ff,
+					struct fuse_inode *fi)
+{
+	spin_lock(&fi->lock);
+	WARN_ON(fi->iocachectr <= 0);
+	WARN_ON(ff->iomode != IOM_CACHED);
+	ff->iomode = IOM_NONE;
+	fi->iocachectr--;
+	if (fi->iocachectr == 0)
+		clear_bit(FUSE_I_CACHE_IO_MODE, &fi->state);
+	spin_unlock(&fi->lock);
+}
+
+/* Start strictly uncached io mode where cache access is not allowed */
+int fuse_inode_uncached_io_start(struct fuse_inode *fi, struct fuse_backing *fb)
+{
+	struct fuse_backing *oldfb;
+	int err = 0;
+
+	spin_lock(&fi->lock);
+	/* deny conflicting backing files on same fuse inode */
+	oldfb = fuse_inode_backing(fi);
+	if (fb && oldfb && oldfb != fb) {
+		err = -EBUSY;
+		goto unlock;
+	}
+	if (fi->iocachectr > 0) {
+		err = -ETXTBSY;
+		goto unlock;
+	}
+	fi->iocachectr--;
+
+	/* fuse inode holds a single refcount of backing file */
+	if (fb && !oldfb) {
+		oldfb = fuse_inode_backing_set(fi, fb);
+		WARN_ON_ONCE(oldfb != NULL);
+	} else {
+		fuse_backing_put(fb);
+	}
+unlock:
+	spin_unlock(&fi->lock);
+	return err;
+}
+
+/* Takes uncached_io inode mode reference to be dropped on file release */
+static int fuse_file_uncached_io_open(struct inode *inode,
+				      struct fuse_file *ff,
+				      struct fuse_backing *fb)
+{
+	struct fuse_inode *fi = get_fuse_inode(inode);
+	int err;
+
+	err = fuse_inode_uncached_io_start(fi, fb);
+	if (err)
+		return err;
+
+	WARN_ON(ff->iomode != IOM_NONE);
+	ff->iomode = IOM_UNCACHED;
+	return 0;
+}
+
+void fuse_inode_uncached_io_end(struct fuse_inode *fi)
+{
+	struct fuse_backing *oldfb = NULL;
+
+	spin_lock(&fi->lock);
+	WARN_ON(fi->iocachectr >= 0);
+	fi->iocachectr++;
+	if (!fi->iocachectr) {
+		wake_up(&fi->direct_io_waitq);
+		oldfb = fuse_inode_backing_set(fi, NULL);
+	}
+	spin_unlock(&fi->lock);
+	if (oldfb)
+		fuse_backing_put(oldfb);
+}
+
+/* Drop uncached_io reference from passthrough open */
+static void fuse_file_uncached_io_release(struct fuse_file *ff,
+					  struct fuse_inode *fi)
+{
+	WARN_ON(ff->iomode != IOM_UNCACHED);
+	ff->iomode = IOM_NONE;
+	fuse_inode_uncached_io_end(fi);
+}
+
+/*
+ * Open flags that are allowed in combination with FOPEN_PASSTHROUGH.
+ * A combination of FOPEN_PASSTHROUGH and FOPEN_DIRECT_IO means that read/write
+ * operations go directly to the server, but mmap is done on the backing file.
+ * FOPEN_PASSTHROUGH mode should not co-exist with any users of the fuse inode
+ * page cache, so FOPEN_KEEP_CACHE is a strange and undesired combination.
+ */
+#define FOPEN_PASSTHROUGH_MASK \
+	(FOPEN_PASSTHROUGH | FOPEN_DIRECT_IO | FOPEN_PARALLEL_DIRECT_WRITES | \
+	 FOPEN_NOFLUSH)
+
+static int fuse_file_passthrough_open(struct inode *inode, struct file *file)
+{
+	struct fuse_file *ff = file->private_data;
+	struct fuse_conn *fc = get_fuse_conn(inode);
+	struct fuse_backing *fb;
+	int err;
+
+	/* Check allowed conditions for file open in passthrough mode */
+	if (!IS_ENABLED(CONFIG_FUSE_PASSTHROUGH) || !fc->passthrough ||
+	    (ff->open_flags & ~FOPEN_PASSTHROUGH_MASK))
+		return -EINVAL;
+
+	fb = fuse_passthrough_open(file, ff->args->open_outarg.backing_id);
+	if (IS_ERR(fb))
+		return PTR_ERR(fb);
+
+	/* First passthrough file open denies caching inode io mode */
+	err = fuse_file_uncached_io_open(inode, ff, fb);
+	if (!err)
+		return 0;
+
+	fuse_passthrough_release(ff, fb);
+	fuse_backing_put(fb);
+
+	return err;
+}
+
+/* Request access to submit new io to inode via open file */
+int fuse_file_io_open(struct file *file, struct inode *inode)
+{
+	struct fuse_file *ff = file->private_data;
+	struct fuse_inode *fi = get_fuse_inode(inode);
+	int err;
+
+	/*
+	 * io modes are not relevant with DAX and with server that does not
+	 * implement open.
+	 */
+	if (FUSE_IS_DAX(inode) || !ff->args)
+		return 0;
+
+	/*
+	 * Server is expected to use FOPEN_PASSTHROUGH for all opens of an inode
+	 * which is already open for passthrough.
+	 */
+	err = -EINVAL;
+	if (fuse_inode_backing(fi) && !(ff->open_flags & FOPEN_PASSTHROUGH))
+		goto fail;
+
+	/*
+	 * FOPEN_PARALLEL_DIRECT_WRITES requires FOPEN_DIRECT_IO.
+	 */
+	if (!(ff->open_flags & FOPEN_DIRECT_IO))
+		ff->open_flags &= ~FOPEN_PARALLEL_DIRECT_WRITES;
+
+	/*
+	 * First passthrough file open denies caching inode io mode.
+	 * First caching file open enters caching inode io mode.
+	 *
+	 * Note that if user opens a file open with O_DIRECT, but server did
+	 * not specify FOPEN_DIRECT_IO, a later fcntl() could remove O_DIRECT,
+	 * so we put the inode in caching mode to prevent parallel dio.
+	 */
+	if ((ff->open_flags & FOPEN_DIRECT_IO) &&
+	    !(ff->open_flags & FOPEN_PASSTHROUGH))
+		return 0;
+
+	if (ff->open_flags & FOPEN_PASSTHROUGH)
+		err = fuse_file_passthrough_open(inode, file);
+	else
+		err = fuse_file_cached_io_open(inode, ff);
+	if (err)
+		goto fail;
+
+	return 0;
+
+fail:
+	pr_debug("failed to open file in requested io mode (open_flags=0x%x, err=%i).\n",
+		 ff->open_flags, err);
+	/*
+	 * The file open mode determines the inode io mode.
+	 * Using incorrect open mode is a server mistake, which results in
+	 * user visible failure of open() with EIO error.
+	 */
+	return -EIO;
+}
+
+/* No more pending io and no new io possible to inode via open/mmapped file */
+void fuse_file_io_release(struct fuse_file *ff, struct inode *inode)
+{
+	struct fuse_inode *fi = get_fuse_inode(inode);
+
+	/*
+	 * Last passthrough file close allows caching inode io mode.
+	 * Last caching file close exits caching inode io mode.
+	 */
+	switch (ff->iomode) {
+	case IOM_NONE:
+		/* Nothing to do */
+		break;
+	case IOM_UNCACHED:
+		fuse_file_uncached_io_release(ff, fi);
+		break;
+	case IOM_CACHED:
+		fuse_file_cached_io_release(ff, fi);
+		break;
+	}
+}

linux-7.2.3/fs/fuse/notify.c

@@ -0,0 +1,444 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include "dev.h"
+#include "fuse_i.h"
+#include <linux/pagemap.h>
+
+static int fuse_notify_poll(struct fuse_conn *fc, unsigned int size,
+			    struct fuse_copy_state *cs)
+{
+	struct fuse_notify_poll_wakeup_out outarg;
+	int err;
+
+	if (size != sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+
+	fuse_copy_finish(cs);
+	return fuse_notify_poll_wakeup(fc, &outarg);
+}
+
+static int fuse_notify_inval_inode(struct fuse_conn *fc, unsigned int size,
+				   struct fuse_copy_state *cs)
+{
+	struct fuse_notify_inval_inode_out outarg;
+	int err;
+
+	if (size != sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+	fuse_copy_finish(cs);
+
+	down_read(&fc->killsb);
+	err = fuse_reverse_inval_inode(fc, outarg.ino,
+				       outarg.off, outarg.len);
+	up_read(&fc->killsb);
+	return err;
+}
+
+static int fuse_notify_inval_entry(struct fuse_conn *fc, unsigned int size,
+				   struct fuse_copy_state *cs)
+{
+	struct fuse_notify_inval_entry_out outarg;
+	int err;
+	char *buf;
+	struct qstr name;
+
+	if (size < sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+
+	if (outarg.namelen > fc->name_max)
+		return -ENAMETOOLONG;
+
+	err = -EINVAL;
+	if (size != sizeof(outarg) + outarg.namelen + 1)
+		return -EINVAL;
+
+	buf = kzalloc(outarg.namelen + 1, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	name.name = buf;
+	name.len = outarg.namelen;
+	err = fuse_copy_one(cs, buf, outarg.namelen + 1);
+	if (err)
+		goto err;
+	fuse_copy_finish(cs);
+	buf[outarg.namelen] = 0;
+
+	down_read(&fc->killsb);
+	err = fuse_reverse_inval_entry(fc, outarg.parent, 0, &name, outarg.flags);
+	up_read(&fc->killsb);
+err:
+	kfree(buf);
+	return err;
+}
+
+static int fuse_notify_delete(struct fuse_conn *fc, unsigned int size,
+			      struct fuse_copy_state *cs)
+{
+	struct fuse_notify_delete_out outarg;
+	int err;
+	char *buf;
+	struct qstr name;
+
+	if (size < sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+
+	if (outarg.namelen > fc->name_max)
+		return -ENAMETOOLONG;
+
+	if (size != sizeof(outarg) + outarg.namelen + 1)
+		return -EINVAL;
+
+	buf = kzalloc(outarg.namelen + 1, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	name.name = buf;
+	name.len = outarg.namelen;
+	err = fuse_copy_one(cs, buf, outarg.namelen + 1);
+	if (err)
+		goto err;
+	fuse_copy_finish(cs);
+	buf[outarg.namelen] = 0;
+
+	down_read(&fc->killsb);
+	err = fuse_reverse_inval_entry(fc, outarg.parent, outarg.child, &name, 0);
+	up_read(&fc->killsb);
+err:
+	kfree(buf);
+	return err;
+}
+
+static int fuse_notify_store(struct fuse_conn *fc, unsigned int size,
+			     struct fuse_copy_state *cs)
+{
+	struct fuse_notify_store_out outarg;
+	struct inode *inode;
+	struct address_space *mapping;
+	u64 nodeid;
+	int err;
+	unsigned int num;
+	loff_t file_size;
+	loff_t pos;
+	loff_t end;
+
+	if (size < sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+
+	if (size - sizeof(outarg) != outarg.size)
+		return -EINVAL;
+
+	if (outarg.offset >= MAX_LFS_FILESIZE)
+		return -EINVAL;
+
+	nodeid = outarg.nodeid;
+	pos = outarg.offset;
+	num = min(outarg.size, MAX_LFS_FILESIZE - pos);
+
+	down_read(&fc->killsb);
+
+	err = -ENOENT;
+	inode = fuse_ilookup(fc, nodeid,  NULL);
+	if (!inode)
+		goto out_up_killsb;
+	if (!S_ISREG(inode->i_mode)) {
+		err = -EINVAL;
+		goto out_iput;
+	}
+
+	mapping = inode->i_mapping;
+	file_size = i_size_read(inode);
+	end = pos + num;
+	if (end > file_size) {
+		file_size = end;
+		fuse_write_update_attr(inode, file_size, num);
+	}
+
+	while (num) {
+		struct folio *folio;
+		unsigned int folio_offset;
+		unsigned int nr_bytes;
+		pgoff_t index = pos >> PAGE_SHIFT;
+
+		folio = filemap_grab_folio(mapping, index);
+		err = PTR_ERR(folio);
+		if (IS_ERR(folio))
+			goto out_iput;
+
+		folio_offset = offset_in_folio(folio, pos);
+		nr_bytes = min(num, folio_size(folio) - folio_offset);
+
+		err = fuse_copy_folio(cs, &folio, folio_offset, nr_bytes, 0);
+		if (!folio_test_uptodate(folio) && !err && folio_offset == 0 &&
+		    (nr_bytes == folio_size(folio) || file_size == end)) {
+			folio_zero_segment(folio, nr_bytes, folio_size(folio));
+			folio_mark_uptodate(folio);
+		}
+		folio_unlock(folio);
+		folio_put(folio);
+
+		if (err)
+			goto out_iput;
+
+		pos += nr_bytes;
+		num -= nr_bytes;
+	}
+
+	err = 0;
+
+out_iput:
+	iput(inode);
+out_up_killsb:
+	up_read(&fc->killsb);
+	return err;
+}
+
+struct fuse_retrieve_args {
+	struct fuse_args_pages ap;
+	struct fuse_notify_retrieve_in inarg;
+};
+
+static void fuse_retrieve_end(struct fuse_args *args, int error)
+{
+	struct fuse_retrieve_args *ra =
+		container_of(args, typeof(*ra), ap.args);
+
+	release_pages(ra->ap.folios, ra->ap.num_folios);
+	kfree(ra);
+}
+
+static int fuse_retrieve(struct fuse_mount *fm, struct inode *inode,
+			 struct fuse_notify_retrieve_out *outarg)
+{
+	int err;
+	struct address_space *mapping = inode->i_mapping;
+	loff_t file_size;
+	unsigned int num;
+	unsigned int offset;
+	size_t total_len = 0;
+	unsigned int num_pages;
+	struct fuse_conn *fc = fm->fc;
+	struct fuse_retrieve_args *ra;
+	size_t args_size = sizeof(*ra);
+	struct fuse_args_pages *ap;
+	struct fuse_args *args;
+	loff_t pos = outarg->offset;
+
+	offset = offset_in_page(pos);
+	file_size = i_size_read(inode);
+
+	num = min(outarg->size, fc->max_write);
+	if (pos > file_size)
+		num = 0;
+	else if (num > file_size - pos)
+		num = file_size - pos;
+
+	num_pages = DIV_ROUND_UP(num + offset, PAGE_SIZE);
+	num_pages = min(num_pages, fc->max_pages);
+	num = min(num, num_pages << PAGE_SHIFT);
+
+	args_size += num_pages * (sizeof(ap->folios[0]) + sizeof(ap->descs[0]));
+
+	ra = kzalloc(args_size, GFP_KERNEL);
+	if (!ra)
+		return -ENOMEM;
+
+	ap = &ra->ap;
+	ap->folios = (void *) (ra + 1);
+	ap->descs = (void *) (ap->folios + num_pages);
+
+	args = &ap->args;
+	args->nodeid = outarg->nodeid;
+	args->opcode = FUSE_NOTIFY_REPLY;
+	args->in_numargs = 3;
+	args->in_pages = true;
+	args->end = fuse_retrieve_end;
+
+	while (num && ap->num_folios < num_pages) {
+		struct folio *folio;
+		unsigned int folio_offset;
+		unsigned int nr_bytes;
+		pgoff_t index = pos >> PAGE_SHIFT;
+
+		folio = filemap_get_folio(mapping, index);
+		if (IS_ERR(folio))
+			break;
+		if (!folio_test_uptodate(folio)) {
+			folio_put(folio);
+			break;
+		}
+
+		folio_offset = offset_in_folio(folio, pos);
+		nr_bytes = min(folio_size(folio) - folio_offset, num);
+
+		ap->folios[ap->num_folios] = folio;
+		ap->descs[ap->num_folios].offset = folio_offset;
+		ap->descs[ap->num_folios].length = nr_bytes;
+		ap->num_folios++;
+
+		pos += nr_bytes;
+		num -= nr_bytes;
+		total_len += nr_bytes;
+	}
+	ra->inarg.offset = outarg->offset;
+	ra->inarg.size = total_len;
+	fuse_set_zero_arg0(args);
+	args->in_args[1].size = sizeof(ra->inarg);
+	args->in_args[1].value = &ra->inarg;
+	args->in_args[2].size = total_len;
+
+	err = fuse_simple_notify_reply(fm, args, outarg->notify_unique);
+	if (err)
+		fuse_retrieve_end(args, err);
+
+	return err;
+}
+
+static int fuse_notify_retrieve(struct fuse_conn *fc, unsigned int size,
+				struct fuse_copy_state *cs)
+{
+	struct fuse_notify_retrieve_out outarg;
+	struct fuse_mount *fm;
+	struct inode *inode;
+	u64 nodeid;
+	int err;
+
+	if (size != sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+
+	fuse_copy_finish(cs);
+
+	if (outarg.offset >= MAX_LFS_FILESIZE)
+		return -EINVAL;
+
+	down_read(&fc->killsb);
+	err = -ENOENT;
+	nodeid = outarg.nodeid;
+
+	inode = fuse_ilookup(fc, nodeid, &fm);
+	if (inode) {
+		err = -EINVAL;
+		if (S_ISREG(inode->i_mode))
+			err = fuse_retrieve(fm, inode, &outarg);
+		iput(inode);
+	}
+	up_read(&fc->killsb);
+
+	return err;
+}
+
+static int fuse_notify_resend(struct fuse_conn *fc)
+{
+	fuse_chan_resend(fc->chan);
+	return 0;
+}
+
+/*
+ * Increments the fuse connection epoch.  This will cause dentries and
+ * readdir caches from previous epochs to be invalidated.  Additionally,
+ * if inval_wq is set, a work queue is scheduled to trigger the invalidation.
+ */
+static int fuse_notify_inc_epoch(struct fuse_conn *fc)
+{
+	atomic_inc(&fc->epoch);
+	if (inval_wq)
+		schedule_work(&fc->epoch_work);
+
+	return 0;
+}
+
+static int fuse_notify_prune(struct fuse_conn *fc, unsigned int size,
+			     struct fuse_copy_state *cs)
+{
+	struct fuse_notify_prune_out outarg;
+	const unsigned int batch = 512;
+	u64 *nodeids __free(kfree) = kmalloc(sizeof(u64) * batch, GFP_KERNEL);
+	unsigned int num, i;
+	int err;
+
+	if (!nodeids)
+		return -ENOMEM;
+
+	if (size < sizeof(outarg))
+		return -EINVAL;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		return err;
+
+	if (size - sizeof(outarg) != array_size(outarg.count, sizeof(u64)))
+		return -EINVAL;
+
+	for (; outarg.count; outarg.count -= num) {
+		num = min(batch, outarg.count);
+		err = fuse_copy_one(cs, nodeids, num * sizeof(u64));
+		if (err)
+			return err;
+
+		scoped_guard(rwsem_read, &fc->killsb) {
+			for (i = 0; i < num; i++)
+				fuse_try_prune_one_inode(fc, nodeids[i]);
+		}
+	}
+	return 0;
+}
+
+int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code,
+		unsigned int size, struct fuse_copy_state *cs)
+{
+	switch (code) {
+	case FUSE_NOTIFY_POLL:
+		return fuse_notify_poll(fc, size, cs);
+
+	case FUSE_NOTIFY_INVAL_INODE:
+		return fuse_notify_inval_inode(fc, size, cs);
+
+	case FUSE_NOTIFY_INVAL_ENTRY:
+		return fuse_notify_inval_entry(fc, size, cs);
+
+	case FUSE_NOTIFY_STORE:
+		return fuse_notify_store(fc, size, cs);
+
+	case FUSE_NOTIFY_RETRIEVE:
+		return fuse_notify_retrieve(fc, size, cs);
+
+	case FUSE_NOTIFY_DELETE:
+		return fuse_notify_delete(fc, size, cs);
+
+	case FUSE_NOTIFY_RESEND:
+		return fuse_notify_resend(fc);
+
+	case FUSE_NOTIFY_INC_EPOCH:
+		return fuse_notify_inc_epoch(fc);
+
+	case FUSE_NOTIFY_PRUNE:
+		return fuse_notify_prune(fc, size, cs);
+
+	default:
+		return -EINVAL;
+	}
+}

linux-7.2.3/fs/fuse/passthrough.c

@@ -0,0 +1,197 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * FUSE passthrough to backing file.
+ *
+ * Copyright (c) 2023 CTERA Networks.
+ */
+
+#include "fuse_i.h"
+
+#include <linux/file.h>
+#include <linux/backing-file.h>
+#include <linux/splice.h>
+
+static void fuse_file_accessed(struct file *file)
+{
+	struct inode *inode = file_inode(file);
+
+	fuse_invalidate_atime(inode);
+}
+
+static void fuse_passthrough_end_write(struct kiocb *iocb, ssize_t ret)
+{
+	struct inode *inode = file_inode(iocb->ki_filp);
+
+	fuse_write_update_attr(inode, iocb->ki_pos, ret);
+}
+
+ssize_t fuse_passthrough_read_iter(struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct file *file = iocb->ki_filp;
+	struct fuse_file *ff = file->private_data;
+	struct file *backing_file = fuse_file_passthrough(ff);
+	size_t count = iov_iter_count(iter);
+	ssize_t ret;
+	struct backing_file_ctx ctx = {
+		.cred = ff->cred,
+		.accessed = fuse_file_accessed,
+	};
+
+
+	pr_debug("%s: backing_file=0x%p, pos=%lld, len=%zu\n", __func__,
+		 backing_file, iocb->ki_pos, count);
+
+	if (!count)
+		return 0;
+
+	ret = backing_file_read_iter(backing_file, iter, iocb, iocb->ki_flags,
+				     &ctx);
+
+	return ret;
+}
+
+ssize_t fuse_passthrough_write_iter(struct kiocb *iocb,
+				    struct iov_iter *iter)
+{
+	struct file *file = iocb->ki_filp;
+	struct inode *inode = file_inode(file);
+	struct fuse_file *ff = file->private_data;
+	struct file *backing_file = fuse_file_passthrough(ff);
+	size_t count = iov_iter_count(iter);
+	ssize_t ret;
+	struct backing_file_ctx ctx = {
+		.cred = ff->cred,
+		.end_write = fuse_passthrough_end_write,
+	};
+
+	pr_debug("%s: backing_file=0x%p, pos=%lld, len=%zu\n", __func__,
+		 backing_file, iocb->ki_pos, count);
+
+	if (!count)
+		return 0;
+
+	inode_lock(inode);
+	ret = backing_file_write_iter(backing_file, iter, iocb, iocb->ki_flags,
+				      &ctx);
+	inode_unlock(inode);
+
+	return ret;
+}
+
+ssize_t fuse_passthrough_splice_read(struct file *in, loff_t *ppos,
+				     struct pipe_inode_info *pipe,
+				     size_t len, unsigned int flags)
+{
+	struct fuse_file *ff = in->private_data;
+	struct file *backing_file = fuse_file_passthrough(ff);
+	struct backing_file_ctx ctx = {
+		.cred = ff->cred,
+		.accessed = fuse_file_accessed,
+	};
+	struct kiocb iocb;
+	ssize_t ret;
+
+	pr_debug("%s: backing_file=0x%p, pos=%lld, len=%zu, flags=0x%x\n", __func__,
+		 backing_file, *ppos, len, flags);
+
+	init_sync_kiocb(&iocb, in);
+	iocb.ki_pos = *ppos;
+	ret = backing_file_splice_read(backing_file, &iocb, pipe, len, flags, &ctx);
+	*ppos = iocb.ki_pos;
+
+	return ret;
+}
+
+ssize_t fuse_passthrough_splice_write(struct pipe_inode_info *pipe,
+				      struct file *out, loff_t *ppos,
+				      size_t len, unsigned int flags)
+{
+	struct fuse_file *ff = out->private_data;
+	struct file *backing_file = fuse_file_passthrough(ff);
+	struct inode *inode = file_inode(out);
+	ssize_t ret;
+	struct backing_file_ctx ctx = {
+		.cred = ff->cred,
+		.end_write = fuse_passthrough_end_write,
+	};
+	struct kiocb iocb;
+
+	pr_debug("%s: backing_file=0x%p, pos=%lld, len=%zu, flags=0x%x\n", __func__,
+		 backing_file, *ppos, len, flags);
+
+	inode_lock(inode);
+	init_sync_kiocb(&iocb, out);
+	iocb.ki_pos = *ppos;
+	ret = backing_file_splice_write(pipe, backing_file, &iocb, len, flags, &ctx);
+	*ppos = iocb.ki_pos;
+	inode_unlock(inode);
+
+	return ret;
+}
+
+ssize_t fuse_passthrough_mmap(struct file *file, struct vm_area_struct *vma)
+{
+	struct fuse_file *ff = file->private_data;
+	struct file *backing_file = fuse_file_passthrough(ff);
+	struct backing_file_ctx ctx = {
+		.cred = ff->cred,
+		.accessed = fuse_file_accessed,
+	};
+
+	pr_debug("%s: backing_file=0x%p, start=%lu, end=%lu\n", __func__,
+		 backing_file, vma->vm_start, vma->vm_end);
+
+	return backing_file_mmap(backing_file, vma, &ctx);
+}
+
+/*
+ * Setup passthrough to a backing file.
+ *
+ * Returns an fb object with elevated refcount to be stored in fuse inode.
+ */
+struct fuse_backing *fuse_passthrough_open(struct file *file, int backing_id)
+{
+	struct fuse_file *ff = file->private_data;
+	struct fuse_conn *fc = ff->fm->fc;
+	struct fuse_backing *fb = NULL;
+	struct file *backing_file;
+	int err;
+
+	err = -EINVAL;
+	if (backing_id <= 0)
+		goto out;
+
+	err = -ENOENT;
+	fb = fuse_backing_lookup(fc, backing_id);
+	if (!fb)
+		goto out;
+
+	/* Allocate backing file per fuse file to store fuse path */
+	backing_file = backing_file_open(file, file->f_flags,
+					 &fb->file->f_path, fb->cred);
+	err = PTR_ERR(backing_file);
+	if (IS_ERR(backing_file)) {
+		fuse_backing_put(fb);
+		goto out;
+	}
+
+	err = 0;
+	ff->passthrough = backing_file;
+	ff->cred = get_cred(fb->cred);
+out:
+	pr_debug("%s: backing_id=%d, fb=0x%p, backing_file=0x%p, err=%i\n", __func__,
+		 backing_id, fb, ff->passthrough, err);
+
+	return err ? ERR_PTR(err) : fb;
+}
+
+void fuse_passthrough_release(struct fuse_file *ff, struct fuse_backing *fb)
+{
+	pr_debug("%s: fb=0x%p, backing_file=0x%p\n", __func__,
+		 fb, ff->passthrough);
+
+	fput(ff->passthrough);
+	ff->passthrough = NULL;
+	put_cred(ff->cred);
+	ff->cred = NULL;
+}

linux-7.2.3/fs/fuse/poll.c

@@ -0,0 +1,141 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include "dev.h"
+#include "fuse_i.h"
+
+void fuse_end_polls(struct fuse_conn *fc)
+{
+	struct rb_node *p;
+
+	spin_lock(&fc->lock);
+	p = rb_first(&fc->polled_files);
+
+	while (p) {
+		struct fuse_file *ff;
+		ff = rb_entry(p, struct fuse_file, polled_node);
+		wake_up_interruptible_all(&ff->poll_wait);
+
+		p = rb_next(p);
+	}
+	spin_unlock(&fc->lock);
+}
+
+/*
+ * All files which have been polled are linked to RB tree
+ * fuse_conn->polled_files which is indexed by kh.  Walk the tree and
+ * find the matching one.
+ */
+static struct rb_node **fuse_find_polled_node(struct fuse_conn *fc, u64 kh,
+					      struct rb_node **parent_out)
+{
+	struct rb_node **link = &fc->polled_files.rb_node;
+	struct rb_node *last = NULL;
+
+	while (*link) {
+		struct fuse_file *ff;
+
+		last = *link;
+		ff = rb_entry(last, struct fuse_file, polled_node);
+
+		if (kh < ff->kh)
+			link = &last->rb_left;
+		else if (kh > ff->kh)
+			link = &last->rb_right;
+		else
+			return link;
+	}
+
+	if (parent_out)
+		*parent_out = last;
+	return link;
+}
+
+/*
+ * The file is about to be polled.  Make sure it's on the polled_files
+ * RB tree.  Note that files once added to the polled_files tree are
+ * not removed before the file is released.  This is because a file
+ * polled once is likely to be polled again.
+ */
+static void fuse_register_polled_file(struct fuse_conn *fc,
+				      struct fuse_file *ff)
+{
+	spin_lock(&fc->lock);
+	if (RB_EMPTY_NODE(&ff->polled_node)) {
+		struct rb_node **link, *parent;
+
+		link = fuse_find_polled_node(fc, ff->kh, &parent);
+		BUG_ON(*link);
+		rb_link_node(&ff->polled_node, parent, link);
+		rb_insert_color(&ff->polled_node, &fc->polled_files);
+	}
+	spin_unlock(&fc->lock);
+}
+
+__poll_t fuse_file_poll(struct file *file, poll_table *wait)
+{
+	struct fuse_file *ff = file->private_data;
+	struct fuse_mount *fm = ff->fm;
+	struct fuse_poll_in inarg = { .fh = ff->fh, .kh = ff->kh };
+	struct fuse_poll_out outarg;
+	FUSE_ARGS(args);
+	int err;
+
+	if (fm->fc->no_poll)
+		return DEFAULT_POLLMASK;
+
+	poll_wait(file, &ff->poll_wait, wait);
+	inarg.events = mangle_poll(poll_requested_events(wait));
+
+	/*
+	 * Ask for notification iff there's someone waiting for it.
+	 * The client may ignore the flag and always notify.
+	 */
+	if (waitqueue_active(&ff->poll_wait)) {
+		inarg.flags |= FUSE_POLL_SCHEDULE_NOTIFY;
+		fuse_register_polled_file(fm->fc, ff);
+	}
+
+	args.opcode = FUSE_POLL;
+	args.nodeid = ff->nodeid;
+	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(fm, &args);
+
+	if (!err)
+		return demangle_poll(outarg.revents);
+	if (err == -ENOSYS) {
+		fm->fc->no_poll = 1;
+		return DEFAULT_POLLMASK;
+	}
+	return EPOLLERR;
+}
+EXPORT_SYMBOL_GPL(fuse_file_poll);
+
+/*
+ * This is called from fuse_handle_notify() on FUSE_NOTIFY_POLL and
+ * wakes up the poll waiters.
+ */
+int fuse_notify_poll_wakeup(struct fuse_conn *fc,
+			    struct fuse_notify_poll_wakeup_out *outarg)
+{
+	u64 kh = outarg->kh;
+	struct rb_node **link;
+
+	spin_lock(&fc->lock);
+
+	link = fuse_find_polled_node(fc, kh, NULL);
+	if (*link) {
+		struct fuse_file *ff;
+
+		ff = rb_entry(*link, struct fuse_file, polled_node);
+		wake_up_interruptible_sync(&ff->poll_wait);
+	}
+
+	spin_unlock(&fc->lock);
+	return 0;
+}
+

linux-7.2.3/fs/fuse/readdir.c

@@ -0,0 +1,656 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+  FUSE: Filesystem in Userspace
+  Copyright (C) 2001-2018  Miklos Szeredi <miklos@szeredi.hu>
+*/
+
+
+#include "fuse_i.h"
+#include <linux/iversion.h>
+#include <linux/posix_acl.h>
+#include <linux/pagemap.h>
+#include <linux/highmem.h>
+#include <linux/vmalloc.h>
+
+static bool fuse_use_readdirplus(struct inode *dir, struct dir_context *ctx)
+{
+	struct fuse_conn *fc = get_fuse_conn(dir);
+	struct fuse_inode *fi = get_fuse_inode(dir);
+
+	if (!fc->do_readdirplus)
+		return false;
+	if (!fc->readdirplus_auto)
+		return true;
+	if (test_and_clear_bit(FUSE_I_ADVISE_RDPLUS, &fi->state))
+		return true;
+	if (ctx->pos == 0)
+		return true;
+	return false;
+}
+
+static void fuse_add_dirent_to_cache(struct file *file,
+				     struct fuse_dirent *dirent, loff_t pos)
+{
+	struct fuse_inode *fi = get_fuse_inode(file_inode(file));
+	size_t reclen = FUSE_DIRENT_SIZE(dirent);
+	pgoff_t index;
+	struct page *page;
+	loff_t size;
+	u64 version;
+	unsigned int offset;
+	void *addr;
+
+	/* Dirent doesn't fit in readdir cache page?  Skip caching. */
+	if (reclen > PAGE_SIZE)
+		return;
+
+	spin_lock(&fi->rdc.lock);
+	/*
+	 * Is cache already completed?  Or this entry does not go at the end of
+	 * cache?
+	 */
+	if (fi->rdc.cached || pos != fi->rdc.pos) {
+		spin_unlock(&fi->rdc.lock);
+		return;
+	}
+	version = fi->rdc.version;
+	size = fi->rdc.size;
+	offset = offset_in_page(size);
+	index = size >> PAGE_SHIFT;
+	/* Dirent doesn't fit in current page?  Jump to next page. */
+	if (offset + reclen > PAGE_SIZE) {
+		index++;
+		offset = 0;
+	}
+	spin_unlock(&fi->rdc.lock);
+
+	if (offset) {
+		page = find_lock_page(file->f_mapping, index);
+	} else {
+		page = find_or_create_page(file->f_mapping, index,
+					   mapping_gfp_mask(file->f_mapping));
+	}
+	if (!page)
+		return;
+
+	spin_lock(&fi->rdc.lock);
+	/* Raced with another readdir */
+	if (fi->rdc.version != version || fi->rdc.size != size ||
+	    WARN_ON(fi->rdc.pos != pos))
+		goto unlock;
+
+	addr = kmap_local_page(page);
+	if (!offset) {
+		clear_page(addr);
+		SetPageUptodate(page);
+	}
+	memcpy(addr + offset, dirent, reclen);
+	kunmap_local(addr);
+	fi->rdc.size = (index << PAGE_SHIFT) + offset + reclen;
+	fi->rdc.pos = dirent->off;
+unlock:
+	spin_unlock(&fi->rdc.lock);
+	unlock_page(page);
+	put_page(page);
+}
+
+static void fuse_readdir_cache_end(struct file *file, loff_t pos)
+{
+	struct fuse_inode *fi = get_fuse_inode(file_inode(file));
+	loff_t end;
+
+	spin_lock(&fi->rdc.lock);
+	/* does cache end position match current position? */
+	if (fi->rdc.pos != pos) {
+		spin_unlock(&fi->rdc.lock);
+		return;
+	}
+
+	fi->rdc.cached = true;
+	end = ALIGN(fi->rdc.size, PAGE_SIZE);
+	spin_unlock(&fi->rdc.lock);
+
+	/* truncate unused tail of cache */
+	truncate_inode_pages(file->f_mapping, end);
+}
+
+static bool fuse_emit(struct file *file, struct dir_context *ctx,
+		      struct fuse_dirent *dirent)
+{
+	struct fuse_file *ff = file->private_data;
+
+	if (ff->open_flags & FOPEN_CACHE_DIR)
+		fuse_add_dirent_to_cache(file, dirent, ctx->pos);
+
+	return dir_emit(ctx, dirent->name, dirent->namelen, dirent->ino,
+			dirent->type | FILLDIR_FLAG_NOINTR);
+}
+
+static int parse_dirfile(char *buf, size_t nbytes, struct file *file,
+			 struct dir_context *ctx)
+{
+	while (nbytes >= FUSE_NAME_OFFSET) {
+		struct fuse_dirent *dirent = (struct fuse_dirent *) buf;
+		size_t reclen = FUSE_DIRENT_SIZE(dirent);
+		if (!dirent->namelen || dirent->namelen > FUSE_NAME_MAX)
+			return -EIO;
+		if (reclen > nbytes)
+			break;
+		if (memchr(dirent->name, '/', dirent->namelen) != NULL)
+			return -EIO;
+
+		if (!fuse_emit(file, ctx, dirent))
+			break;
+
+		buf += reclen;
+		nbytes -= reclen;
+		ctx->pos = dirent->off;
+	}
+
+	return 0;
+}
+
+static int fuse_direntplus_link(struct file *file,
+				struct fuse_direntplus *direntplus,
+				u64 attr_version, u64 evict_ctr)
+{
+	struct fuse_entry_out *o = &direntplus->entry_out;
+	struct fuse_dirent *dirent = &direntplus->dirent;
+	struct dentry *parent = file->f_path.dentry;
+	struct qstr name = QSTR_INIT(dirent->name, dirent->namelen);
+	struct dentry *dentry;
+	struct dentry *alias;
+	struct inode *dir = d_inode(parent);
+	struct fuse_conn *fc;
+	struct inode *inode;
+	int epoch;
+
+	if (!o->nodeid) {
+		/*
+		 * Unlike in the case of fuse_lookup, zero nodeid does not mean
+		 * ENOENT. Instead, it only means the userspace filesystem did
+		 * not want to return attributes/handle for this entry.
+		 *
+		 * So do nothing.
+		 */
+		return 0;
+	}
+
+	if (name.name[0] == '.') {
+		/*
+		 * We could potentially refresh the attributes of the directory
+		 * and its parent?
+		 */
+		if (name.len == 1)
+			return 0;
+		if (name.name[1] == '.' && name.len == 2)
+			return 0;
+	}
+
+	if (invalid_nodeid(o->nodeid))
+		return -EIO;
+	if (fuse_invalid_attr(&o->attr))
+		return -EIO;
+
+	fc = get_fuse_conn(dir);
+	epoch = atomic_read(&fc->epoch);
+
+	name.hash = full_name_hash(parent, name.name, name.len);
+	dentry = d_lookup(parent, &name);
+	if (!dentry) {
+retry:
+		dentry = d_alloc_parallel(parent, &name);
+		if (IS_ERR(dentry))
+			return PTR_ERR(dentry);
+	}
+	if (!d_in_lookup(dentry)) {
+		struct fuse_inode *fi;
+		inode = d_inode(dentry);
+		if (inode && get_node_id(inode) != o->nodeid)
+			inode = NULL;
+		if (!inode ||
+		    fuse_stale_inode(inode, o->generation, &o->attr)) {
+			if (inode)
+				fuse_make_bad(inode);
+			d_invalidate(dentry);
+			dput(dentry);
+			goto retry;
+		}
+		if (fuse_is_bad(inode)) {
+			dput(dentry);
+			return -EIO;
+		}
+
+		fi = get_fuse_inode(inode);
+		spin_lock(&fi->lock);
+		fi->nlookup++;
+		spin_unlock(&fi->lock);
+
+		forget_all_cached_acls(inode);
+		fuse_change_attributes(inode, &o->attr, NULL,
+				       ATTR_TIMEOUT(o),
+				       attr_version);
+		/*
+		 * The other branch comes via fuse_iget()
+		 * which bumps nlookup inside
+		 */
+	} else {
+		inode = fuse_iget(dir->i_sb, o->nodeid, o->generation,
+				  &o->attr, ATTR_TIMEOUT(o),
+				  attr_version, evict_ctr);
+		if (!inode)
+			inode = ERR_PTR(-ENOMEM);
+
+		alias = d_splice_alias(inode, dentry);
+		d_lookup_done(dentry);
+		if (alias) {
+			dput(dentry);
+			dentry = alias;
+		}
+		if (IS_ERR(dentry)) {
+			if (!IS_ERR(inode)) {
+				struct fuse_inode *fi = get_fuse_inode(inode);
+
+				spin_lock(&fi->lock);
+				fi->nlookup--;
+				spin_unlock(&fi->lock);
+			}
+			return PTR_ERR(dentry);
+		}
+	}
+	if (fc->readdirplus_auto)
+		set_bit(FUSE_I_INIT_RDPLUS, &get_fuse_inode(inode)->state);
+	dentry->d_time = epoch;
+	fuse_change_entry_timeout(dentry, o);
+
+	dput(dentry);
+	return 0;
+}
+
+static void fuse_force_forget(struct file *file, u64 nodeid)
+{
+	struct inode *inode = file_inode(file);
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	struct fuse_forget_in inarg;
+	FUSE_ARGS(args);
+
+	memset(&inarg, 0, sizeof(inarg));
+	inarg.nlookup = 1;
+	args.opcode = FUSE_FORGET;
+	args.nodeid = nodeid;
+	args.in_numargs = 1;
+	args.in_args[0].size = sizeof(inarg);
+	args.in_args[0].value = &inarg;
+	args.force = true;
+	args.noreply = true;
+
+	fuse_simple_request(fm, &args);
+	/* ignore errors */
+}
+
+static int parse_dirplusfile(char *buf, size_t nbytes, struct file *file,
+			     struct dir_context *ctx, u64 attr_version,
+			     u64 evict_ctr)
+{
+	struct fuse_direntplus *direntplus;
+	struct fuse_dirent *dirent;
+	size_t reclen;
+	int over = 0;
+	int ret;
+
+	while (nbytes >= FUSE_NAME_OFFSET_DIRENTPLUS) {
+		direntplus = (struct fuse_direntplus *) buf;
+		dirent = &direntplus->dirent;
+		reclen = FUSE_DIRENTPLUS_SIZE(direntplus);
+
+		if (!dirent->namelen || dirent->namelen > FUSE_NAME_MAX)
+			return -EIO;
+		if (reclen > nbytes)
+			break;
+		if (memchr(dirent->name, '/', dirent->namelen) != NULL)
+			return -EIO;
+
+		if (!over) {
+			/* We fill entries into dstbuf only as much as
+			   it can hold. But we still continue iterating
+			   over remaining entries to link them. If not,
+			   we need to send a FORGET for each of those
+			   which we did not link.
+			*/
+			over = !fuse_emit(file, ctx, dirent);
+			if (!over)
+				ctx->pos = dirent->off;
+		}
+
+		buf += reclen;
+		nbytes -= reclen;
+
+		ret = fuse_direntplus_link(file, direntplus, attr_version, evict_ctr);
+		if (ret)
+			fuse_force_forget(file, direntplus->entry_out.nodeid);
+	}
+
+	return 0;
+}
+
+static struct page **fuse_readdir_alloc_buf(struct fuse_args_pages *ap, size_t *bufsize)
+{
+	unsigned int i, nr_alloc, nr_pages = DIV_ROUND_UP(*bufsize, PAGE_SIZE);
+	struct page **pages = kcalloc(nr_pages, sizeof(*pages), GFP_KERNEL);
+
+	if (!pages)
+		return NULL;
+
+	nr_alloc = alloc_pages_bulk(GFP_KERNEL, nr_pages, pages);
+	if (!nr_alloc)
+		goto free_array;
+
+	if (nr_alloc < nr_pages) {
+		nr_pages = nr_alloc;
+		*bufsize = (size_t) nr_pages << PAGE_SHIFT;
+	}
+
+	ap->folios = fuse_folios_alloc(nr_pages, GFP_KERNEL, &ap->descs);
+	if (!ap->folios)
+		goto release_pages;
+
+	for (i = 0; i < nr_pages; i++) {
+		ap->folios[i] = page_folio(pages[i]);
+		ap->descs[i].length = min_t(size_t, *bufsize - (size_t)i * PAGE_SIZE, PAGE_SIZE);
+	}
+	ap->num_folios = nr_pages;
+	ap->args.out_pages = true;
+
+	return pages;
+
+release_pages:
+	release_pages(pages, nr_pages);
+free_array:
+	kfree(pages);
+	return NULL;
+}
+
+static int fuse_readdir_uncached(struct file *file, struct dir_context *ctx)
+{
+	int plus;
+	ssize_t res;
+	struct inode *inode = file_inode(file);
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	struct fuse_conn *fc = fm->fc;
+	struct fuse_io_args ia = {};
+	struct fuse_args_pages *ap = &ia.ap;
+	void *buf;
+	size_t bufsize = clamp((unsigned int) ctx->count, PAGE_SIZE, fc->max_pages << PAGE_SHIFT);
+	u64 attr_version = 0, evict_ctr = 0;
+	bool locked;
+	struct page **pages = fuse_readdir_alloc_buf(ap, &bufsize);
+
+	if (!pages)
+		return -ENOMEM;
+
+	plus = fuse_use_readdirplus(inode, ctx);
+	if (plus) {
+		attr_version = fuse_get_attr_version(fm->fc);
+		evict_ctr = fuse_get_evict_ctr(fm->fc);
+		fuse_read_args_fill(&ia, file, ctx->pos, bufsize, FUSE_READDIRPLUS);
+	} else {
+		fuse_read_args_fill(&ia, file, ctx->pos, bufsize, FUSE_READDIR);
+	}
+	locked = fuse_lock_inode(inode);
+	res = fuse_simple_request(fm, &ap->args);
+	fuse_unlock_inode(inode, locked);
+	if (res < 0)
+		goto out;
+
+	if (!res) {
+		struct fuse_file *ff = file->private_data;
+
+		if (ff->open_flags & FOPEN_CACHE_DIR)
+			fuse_readdir_cache_end(file, ctx->pos);
+		goto out;
+	}
+
+	buf = vm_map_ram(pages, ap->num_folios, -1);
+	if (!buf) {
+		res = -ENOMEM;
+	} else {
+		if (plus)
+			res = parse_dirplusfile(buf, res, file, ctx, attr_version, evict_ctr);
+		else
+			res = parse_dirfile(buf, res, file, ctx);
+
+		vm_unmap_ram(buf, ap->num_folios);
+	}
+out:
+	kfree(ap->folios);
+	release_pages(pages, ap->num_folios);
+	kfree(pages);
+
+	fuse_invalidate_atime(inode);
+
+	return res;
+}
+
+enum fuse_parse_result {
+	FOUND_ERR = -1,
+	FOUND_NONE = 0,
+	FOUND_SOME,
+	FOUND_ALL,
+};
+
+static enum fuse_parse_result fuse_parse_cache(struct fuse_file *ff,
+					       void *addr, unsigned int size,
+					       struct dir_context *ctx)
+{
+	unsigned int offset = offset_in_page(ff->readdir.cache_off);
+	enum fuse_parse_result res = FOUND_NONE;
+
+	WARN_ON(offset >= size);
+
+	for (;;) {
+		struct fuse_dirent *dirent = addr + offset;
+		unsigned int nbytes = size - offset;
+		size_t reclen;
+
+		if (nbytes < FUSE_NAME_OFFSET || !dirent->namelen)
+			break;
+
+		reclen = FUSE_DIRENT_SIZE(dirent); /* derefs ->namelen */
+
+		if (WARN_ON(dirent->namelen > FUSE_NAME_MAX))
+			return FOUND_ERR;
+		if (WARN_ON(reclen > nbytes))
+			return FOUND_ERR;
+		if (WARN_ON(memchr(dirent->name, '/', dirent->namelen) != NULL))
+			return FOUND_ERR;
+
+		if (ff->readdir.pos == ctx->pos) {
+			res = FOUND_SOME;
+			if (!dir_emit(ctx, dirent->name, dirent->namelen,
+				      dirent->ino, dirent->type | FILLDIR_FLAG_NOINTR))
+				return FOUND_ALL;
+			ctx->pos = dirent->off;
+		}
+		ff->readdir.pos = dirent->off;
+		ff->readdir.cache_off += reclen;
+
+		offset += reclen;
+	}
+
+	return res;
+}
+
+static void fuse_rdc_reset(struct inode *inode)
+{
+	struct fuse_inode *fi = get_fuse_inode(inode);
+
+	fi->rdc.cached = false;
+	fi->rdc.version++;
+	fi->rdc.size = 0;
+	fi->rdc.pos = 0;
+	fi->rdc.epoch = 0;
+}
+
+#define UNCACHED 1
+
+static int fuse_readdir_cached(struct file *file, struct dir_context *ctx)
+{
+	struct fuse_file *ff = file->private_data;
+	struct inode *inode = file_inode(file);
+	struct fuse_conn *fc = get_fuse_conn(inode);
+	struct fuse_inode *fi = get_fuse_inode(inode);
+	enum fuse_parse_result res;
+	pgoff_t index;
+	unsigned int size;
+	struct page *page;
+	void *addr;
+
+	/* Seeked?  If so, reset the cache stream */
+	if (ff->readdir.pos != ctx->pos) {
+		ff->readdir.pos = 0;
+		ff->readdir.cache_off = 0;
+	}
+
+	/*
+	 * We're just about to start reading into the cache or reading the
+	 * cache; both cases require an up-to-date mtime value.
+	 */
+	if (!ctx->pos && fc->auto_inval_data) {
+		int err = fuse_update_attributes(inode, file, STATX_MTIME);
+
+		if (err)
+			return err;
+	}
+
+retry:
+	spin_lock(&fi->rdc.lock);
+retry_locked:
+	if (!fi->rdc.cached) {
+		/* Starting cache? Set cache mtime. */
+		if (!ctx->pos && !fi->rdc.size) {
+			fi->rdc.mtime = inode_get_mtime(inode);
+			fi->rdc.iversion = inode_query_iversion(inode);
+			fi->rdc.epoch = atomic_read(&fc->epoch);
+		}
+		spin_unlock(&fi->rdc.lock);
+		return UNCACHED;
+	}
+	/*
+	 * When at the beginning of the directory (i.e. just after opendir(3) or
+	 * rewinddir(3)), then need to check whether directory contents have
+	 * changed, and reset the cache if so.
+	 */
+	if (!ctx->pos) {
+		struct timespec64 mtime = inode_get_mtime(inode);
+
+		if (inode_peek_iversion(inode) != fi->rdc.iversion ||
+		    !timespec64_equal(&fi->rdc.mtime, &mtime) ||
+		    fi->rdc.epoch != atomic_read(&fc->epoch)) {
+			fuse_rdc_reset(inode);
+			goto retry_locked;
+		}
+	}
+
+	/*
+	 * If cache version changed since the last getdents() call, then reset
+	 * the cache stream.
+	 */
+	if (ff->readdir.version != fi->rdc.version) {
+		ff->readdir.pos = 0;
+		ff->readdir.cache_off = 0;
+	}
+	/*
+	 * If at the beginning of the cache, than reset version to
+	 * current.
+	 */
+	if (ff->readdir.pos == 0)
+		ff->readdir.version = fi->rdc.version;
+
+	WARN_ON(fi->rdc.size < ff->readdir.cache_off);
+
+	index = ff->readdir.cache_off >> PAGE_SHIFT;
+
+	if (index == (fi->rdc.size >> PAGE_SHIFT))
+		size = offset_in_page(fi->rdc.size);
+	else
+		size = PAGE_SIZE;
+	spin_unlock(&fi->rdc.lock);
+
+	/* EOF? */
+	if (offset_in_page(ff->readdir.cache_off) == size)
+		return 0;
+
+	page = find_get_page_flags(file->f_mapping, index,
+				   FGP_ACCESSED | FGP_LOCK);
+	/* Page gone missing, then re-added to cache, but not initialized? */
+	if (page && !PageUptodate(page)) {
+		unlock_page(page);
+		put_page(page);
+		page = NULL;
+	}
+	spin_lock(&fi->rdc.lock);
+	if (!page) {
+		/*
+		 * Uh-oh: page gone missing, cache is useless
+		 */
+		if (fi->rdc.version == ff->readdir.version)
+			fuse_rdc_reset(inode);
+		goto retry_locked;
+	}
+
+	/* Make sure it's still the same version after getting the page. */
+	if (ff->readdir.version != fi->rdc.version) {
+		spin_unlock(&fi->rdc.lock);
+		unlock_page(page);
+		put_page(page);
+		goto retry;
+	}
+	spin_unlock(&fi->rdc.lock);
+
+	/*
+	 * Contents of the page are now protected against changing by holding
+	 * the page lock.
+	 */
+	addr = kmap_local_page(page);
+	res = fuse_parse_cache(ff, addr, size, ctx);
+	kunmap_local(addr);
+	unlock_page(page);
+	put_page(page);
+
+	if (res == FOUND_ERR)
+		return -EIO;
+
+	if (res == FOUND_ALL)
+		return 0;
+
+	if (size == PAGE_SIZE) {
+		/* We hit end of page: skip to next page. */
+		ff->readdir.cache_off = ALIGN(ff->readdir.cache_off, PAGE_SIZE);
+		goto retry;
+	}
+
+	/*
+	 * End of cache reached.  If found position, then we are done, otherwise
+	 * need to fall back to uncached, since the position we were looking for
+	 * wasn't in the cache.
+	 */
+	return res == FOUND_SOME ? 0 : UNCACHED;
+}
+
+int fuse_readdir(struct file *file, struct dir_context *ctx)
+{
+	struct fuse_file *ff = file->private_data;
+	struct inode *inode = file_inode(file);
+	int err;
+
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	err = UNCACHED;
+	if (ff->open_flags & FOPEN_CACHE_DIR)
+		err = fuse_readdir_cached(file, ctx);
+	if (err == UNCACHED)
+		err = fuse_readdir_uncached(file, ctx);
+
+	return err;
+}

linux-7.2.3/fs/fuse/req.c

@@ -0,0 +1,99 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include "dev.h"
+#include "fuse_i.h"
+
+static int fuse_fill_creds(struct fuse_mount *fm, struct fuse_args *args, struct mnt_idmap *idmap)
+{
+	struct fuse_conn *fc = fm->fc;
+	bool no_idmap = !fm->sb || (fm->sb->s_iflags & SB_I_NOIDMAP);
+	kuid_t fsuid = mapped_fsuid(idmap, fc->user_ns);
+	kgid_t fsgid = mapped_fsgid(idmap, fc->user_ns);
+
+	args->pid = pid_nr_ns(task_pid(current), fc->pid_ns);
+
+	if (args->force) {
+		if (args->nocreds)
+			return 0;
+
+		if (no_idmap) {
+			args->uid = from_kuid_munged(fc->user_ns, current_fsuid());
+			args->gid = from_kgid_munged(fc->user_ns, current_fsgid());
+		} else {
+			args->uid = FUSE_INVALID_UIDGID;
+			args->gid = FUSE_INVALID_UIDGID;
+		}
+		return 0;
+	}
+
+	WARN_ON(args->nocreds);
+	/*
+	 * Keep the old behavior when idmappings support was not
+	 * declared by a FUSE server.
+	 *
+	 * For those FUSE servers who support idmapped mounts, we send UID/GID
+	 * only along with "inode creation" fuse requests, otherwise idmap ==
+	 * &invalid_mnt_idmap and req->in.h.{u,g}id will be equal to
+	 * FUSE_INVALID_UIDGID.
+	 */
+	if (no_idmap) {
+		fsuid = current_fsuid();
+		fsgid = current_fsgid();
+	}
+	args->uid = from_kuid(fc->user_ns, fsuid);
+	args->gid = from_kgid(fc->user_ns, fsgid);
+
+	if (no_idmap && unlikely(args->uid == ((uid_t)-1) || args->gid == ((gid_t)-1)))
+		return -EOVERFLOW;
+
+	return 0;
+}
+
+static int fuse_req_prep(struct fuse_mount *fm, struct fuse_args *args, struct mnt_idmap *idmap)
+{
+	if (!args->force && fm->fc->conn_error)
+		return -ECONNREFUSED;
+
+	return fuse_fill_creds(fm, args, idmap);
+}
+
+ssize_t __fuse_simple_request(struct mnt_idmap *idmap, struct fuse_mount *fm,
+			      struct fuse_args *args)
+{
+	struct fuse_conn *fc = fm->fc;
+	int err = fuse_req_prep(fm, args, idmap);
+
+	if (err)
+		return err;
+
+	return fuse_chan_send(fc->chan, args);
+}
+
+int fuse_simple_background(struct fuse_mount *fm, struct fuse_args *args, gfp_t gfp_flags)
+{
+	struct fuse_conn *fc = fm->fc;
+	int err;
+
+	WARN_ON(args->force && !args->nocreds);
+
+	err = fuse_req_prep(fm, args, &invalid_mnt_idmap);
+	if (err)
+		return err;
+
+	return fuse_chan_send_bg(fc->chan, args, gfp_flags);
+}
+EXPORT_SYMBOL_GPL(fuse_simple_background);
+
+int fuse_simple_notify_reply(struct fuse_mount *fm, struct fuse_args *args, u64 unique)
+{
+	struct fuse_conn *fc = fm->fc;
+	int err;
+
+	WARN_ON(args->force && !args->nocreds);
+
+	err = fuse_req_prep(fm, args, &invalid_mnt_idmap);
+	if (err)
+		return err;
+
+	return fuse_chan_send_notify_reply(fc->chan, args, unique);
+}

linux-7.2.3/fs/fuse/req_timeout.c

@@ -0,0 +1,148 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include "dev.h"
+#include "sysctl.h"
+#include "fuse_dev_i.h"
+#include "dev_uring_i.h"
+
+/* Frequency (in seconds) of request timeout checks, if opted into */
+#define FUSE_TIMEOUT_TIMER_FREQ 15
+
+/* Frequency (in jiffies) of request timeout checks, if opted into */
+static const unsigned long fuse_timeout_timer_freq =
+	secs_to_jiffies(FUSE_TIMEOUT_TIMER_FREQ);
+
+/*
+ * Default timeout (in seconds) for the server to reply to a request
+ * before the connection is aborted, if no timeout was specified on mount.
+ *
+ * Exported via sysctl
+ */
+unsigned int fuse_default_req_timeout;
+
+/*
+ * Max timeout (in seconds) for the server to reply to a request before
+ * the connection is aborted.
+ *
+ * Exported via sysctl
+ */
+unsigned int fuse_max_req_timeout;
+
+bool fuse_request_expired(struct fuse_chan *fch, struct list_head *list)
+{
+	struct fuse_req *req;
+
+	req = list_first_entry_or_null(list, struct fuse_req, list);
+	if (!req)
+		return false;
+	return time_is_before_jiffies(req->create_time + fch->timeout.req_timeout);
+}
+
+static bool fuse_fpq_processing_expired(struct fuse_chan *fch, struct list_head *processing)
+{
+	int i;
+
+	for (i = 0; i < FUSE_PQ_HASH_SIZE; i++)
+		if (fuse_request_expired(fch, &processing[i]))
+			return true;
+
+	return false;
+}
+
+/*
+ * Check if any requests aren't being completed by the time the request timeout
+ * elapses. To do so, we:
+ * - check the fiq pending list
+ * - check the bg queue
+ * - check the fpq io and processing lists
+ *
+ * To make this fast, we only check against the head request on each list since
+ * these are generally queued in order of creation time (eg newer requests get
+ * queued to the tail). We might miss a few edge cases (eg requests transitioning
+ * between lists, re-sent requests at the head of the pending list having a
+ * later creation time than other requests on that list, etc.) but that is fine
+ * since if the request never gets fulfilled, it will eventually be caught.
+ */
+static void fuse_check_timeout(struct work_struct *work)
+{
+	struct delayed_work *dwork = to_delayed_work(work);
+	struct fuse_chan *fch = container_of(dwork, struct fuse_chan, timeout.work);
+	struct fuse_iqueue *fiq = &fch->iq;
+	struct fuse_dev *fud;
+	struct fuse_pqueue *fpq;
+	bool expired = false;
+
+	if (!atomic_read(&fch->num_waiting))
+		goto out;
+
+	spin_lock(&fiq->lock);
+	expired = fuse_request_expired(fch, &fiq->pending);
+	spin_unlock(&fiq->lock);
+	if (expired)
+		goto chan_abort;
+
+	spin_lock(&fch->bg_lock);
+	expired = fuse_request_expired(fch, &fch->bg_queue);
+	spin_unlock(&fch->bg_lock);
+	if (expired)
+		goto chan_abort;
+
+	spin_lock(&fch->lock);
+	if (!fch->connected) {
+		spin_unlock(&fch->lock);
+		return;
+	}
+	list_for_each_entry(fud, &fch->devices, entry) {
+		fpq = &fud->pq;
+		spin_lock(&fpq->lock);
+		if (fuse_request_expired(fch, &fpq->io) ||
+		    fuse_fpq_processing_expired(fch, fpq->processing)) {
+			spin_unlock(&fpq->lock);
+			spin_unlock(&fch->lock);
+			goto chan_abort;
+		}
+
+		spin_unlock(&fpq->lock);
+	}
+	spin_unlock(&fch->lock);
+
+	if (fuse_uring_request_expired(fch))
+		goto chan_abort;
+
+out:
+	queue_delayed_work(system_percpu_wq, &fch->timeout.work,
+			   fuse_timeout_timer_freq);
+	return;
+
+chan_abort:
+	fuse_chan_abort(fch, false);
+}
+
+static void set_request_timeout(struct fuse_chan *fch, unsigned int timeout)
+{
+	fch->timeout.req_timeout = secs_to_jiffies(timeout);
+	INIT_DELAYED_WORK(&fch->timeout.work, fuse_check_timeout);
+	queue_delayed_work(system_percpu_wq, &fch->timeout.work,
+			   fuse_timeout_timer_freq);
+}
+
+void fuse_init_server_timeout(struct fuse_chan *fch, unsigned int timeout)
+{
+	if (!timeout && !fuse_max_req_timeout && !fuse_default_req_timeout)
+		return;
+
+	if (!timeout)
+		timeout = fuse_default_req_timeout;
+
+	if (fuse_max_req_timeout) {
+		if (timeout)
+			timeout = min(fuse_max_req_timeout, timeout);
+		else
+			timeout = fuse_max_req_timeout;
+	}
+
+	timeout = max(FUSE_TIMEOUT_TIMER_FREQ, timeout);
+
+	set_request_timeout(fch, timeout);
+}
+

linux-7.2.3/fs/fuse/sysctl.c

@@ -0,0 +1,65 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * linux/fs/fuse/fuse_sysctl.c
+ *
+ * Sysctl interface to fuse parameters
+ */
+#include <linux/sysctl.h>
+
+#include "sysctl.h"
+#include "fuse_i.h"
+
+static struct ctl_table_header *fuse_table_header;
+
+/* Bound by fuse_init_out max_pages, which is a u16 */
+static unsigned int sysctl_fuse_max_pages_limit = 65535;
+
+/*
+ * fuse_init_out request timeouts are u16.
+ * This goes up to ~18 hours, which is plenty for a timeout.
+ */
+static unsigned int sysctl_fuse_req_timeout_limit = 65535;
+
+static const struct ctl_table fuse_sysctl_table[] = {
+	{
+		.procname	= "max_pages_limit",
+		.data		= &fuse_max_pages_limit,
+		.maxlen		= sizeof(fuse_max_pages_limit),
+		.mode		= 0644,
+		.proc_handler	= proc_douintvec_minmax,
+		.extra1		= SYSCTL_ONE,
+		.extra2		= &sysctl_fuse_max_pages_limit,
+	},
+	{
+		.procname	= "default_request_timeout",
+		.data		= &fuse_default_req_timeout,
+		.maxlen		= sizeof(fuse_default_req_timeout),
+		.mode		= 0644,
+		.proc_handler	= proc_douintvec_minmax,
+		.extra1		= SYSCTL_ZERO,
+		.extra2		= &sysctl_fuse_req_timeout_limit,
+	},
+	{
+		.procname	= "max_request_timeout",
+		.data		= &fuse_max_req_timeout,
+		.maxlen		= sizeof(fuse_max_req_timeout),
+		.mode		= 0644,
+		.proc_handler	= proc_douintvec_minmax,
+		.extra1		= SYSCTL_ZERO,
+		.extra2		= &sysctl_fuse_req_timeout_limit,
+	},
+};
+
+int fuse_sysctl_register(void)
+{
+	fuse_table_header = register_sysctl("fs/fuse", fuse_sysctl_table);
+	if (!fuse_table_header)
+		return -ENOMEM;
+	return 0;
+}
+
+void fuse_sysctl_unregister(void)
+{
+	unregister_sysctl_table(fuse_table_header);
+	fuse_table_header = NULL;
+}

linux-7.2.3/fs/fuse/sysctl.h

@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef _FS_FUSE_SYSCTL_H
+#define _FS_FUSE_SYSCTL_H
+
+extern unsigned int fuse_default_req_timeout;
+extern unsigned int fuse_max_req_timeout;
+
+#endif /* _FS_FUSE_SYSCTL_H */

linux-7.2.3/fs/fuse/trace.c

@@ -0,0 +1,13 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2025 Oracle.  All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "dev_uring_i.h"
+#include "fuse_i.h"
+#include "fuse_dev_i.h"
+
+#include <linux/pagemap.h>
+
+#define CREATE_TRACE_POINTS
+#include "fuse_trace.h"

linux-7.2.3/fs/fuse/xattr.c

@@ -0,0 +1,214 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * FUSE: Filesystem in Userspace
+ * Copyright (C) 2001-2016  Miklos Szeredi <miklos@szeredi.hu>
+ */
+
+#include "fuse_i.h"
+
+#include <linux/xattr.h>
+#include <linux/posix_acl_xattr.h>
+
+int fuse_setxattr(struct inode *inode, const char *name, const void *value,
+		  size_t size, int flags, unsigned int extra_flags)
+{
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	FUSE_ARGS(args);
+	struct fuse_setxattr_in inarg;
+	int err;
+
+	if (fm->fc->no_setxattr)
+		return -EOPNOTSUPP;
+
+	memset(&inarg, 0, sizeof(inarg));
+	inarg.size = size;
+	inarg.flags = flags;
+	inarg.setxattr_flags = extra_flags;
+
+	args.opcode = FUSE_SETXATTR;
+	args.nodeid = get_node_id(inode);
+	args.in_numargs = 3;
+	args.in_args[0].size = fm->fc->setxattr_ext ?
+		sizeof(inarg) : FUSE_COMPAT_SETXATTR_IN_SIZE;
+	args.in_args[0].value = &inarg;
+	args.in_args[1].size = strlen(name) + 1;
+	args.in_args[1].value = name;
+	args.in_args[2].size = size;
+	args.in_args[2].value = value;
+	err = fuse_simple_request(fm, &args);
+	if (err == -ENOSYS) {
+		fm->fc->no_setxattr = 1;
+		err = -EOPNOTSUPP;
+	}
+	if (!err)
+		fuse_update_ctime(inode);
+
+	return err;
+}
+
+ssize_t fuse_getxattr(struct inode *inode, const char *name, void *value,
+		      size_t size)
+{
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	FUSE_ARGS(args);
+	struct fuse_getxattr_in inarg;
+	struct fuse_getxattr_out outarg;
+	ssize_t ret;
+
+	if (fm->fc->no_getxattr)
+		return -EOPNOTSUPP;
+
+	memset(&inarg, 0, sizeof(inarg));
+	inarg.size = size;
+	args.opcode = FUSE_GETXATTR;
+	args.nodeid = get_node_id(inode);
+	args.in_numargs = 2;
+	args.in_args[0].size = sizeof(inarg);
+	args.in_args[0].value = &inarg;
+	args.in_args[1].size = strlen(name) + 1;
+	args.in_args[1].value = name;
+	/* This is really two different operations rolled into one */
+	args.out_numargs = 1;
+	if (size) {
+		args.out_argvar = true;
+		args.out_args[0].size = size;
+		args.out_args[0].value = value;
+	} else {
+		args.out_args[0].size = sizeof(outarg);
+		args.out_args[0].value = &outarg;
+	}
+	ret = fuse_simple_request(fm, &args);
+	if (!ret && !size)
+		ret = min_t(size_t, outarg.size, XATTR_SIZE_MAX);
+	if (ret == -ENOSYS) {
+		fm->fc->no_getxattr = 1;
+		ret = -EOPNOTSUPP;
+	}
+	return ret;
+}
+
+static int fuse_verify_xattr_list(char *list, size_t size)
+{
+	size_t origsize = size;
+
+	while (size) {
+		size_t thislen = strnlen(list, size);
+
+		if (!thislen || thislen == size)
+			return -EIO;
+
+		size -= thislen + 1;
+		list += thislen + 1;
+	}
+
+	return origsize;
+}
+
+ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size)
+{
+	struct inode *inode = d_inode(entry);
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	FUSE_ARGS(args);
+	struct fuse_getxattr_in inarg;
+	struct fuse_getxattr_out outarg;
+	ssize_t ret;
+
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	if (!fuse_allow_current_process(fm->fc))
+		return -EACCES;
+
+	if (fm->fc->no_listxattr)
+		return -EOPNOTSUPP;
+
+	memset(&inarg, 0, sizeof(inarg));
+	inarg.size = size;
+	args.opcode = FUSE_LISTXATTR;
+	args.nodeid = get_node_id(inode);
+	args.in_numargs = 1;
+	args.in_args[0].size = sizeof(inarg);
+	args.in_args[0].value = &inarg;
+	/* This is really two different operations rolled into one */
+	args.out_numargs = 1;
+	if (size) {
+		args.out_argvar = true;
+		args.out_args[0].size = size;
+		args.out_args[0].value = list;
+	} else {
+		args.out_args[0].size = sizeof(outarg);
+		args.out_args[0].value = &outarg;
+	}
+	ret = fuse_simple_request(fm, &args);
+	if (!ret && !size)
+		ret = min_t(size_t, outarg.size, XATTR_LIST_MAX);
+	if (ret > 0 && size)
+		ret = fuse_verify_xattr_list(list, ret);
+	if (ret == -ENOSYS) {
+		fm->fc->no_listxattr = 1;
+		ret = -EOPNOTSUPP;
+	}
+	return ret;
+}
+
+int fuse_removexattr(struct inode *inode, const char *name)
+{
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	FUSE_ARGS(args);
+	int err;
+
+	if (fm->fc->no_removexattr)
+		return -EOPNOTSUPP;
+
+	args.opcode = FUSE_REMOVEXATTR;
+	args.nodeid = get_node_id(inode);
+	args.in_numargs = 2;
+	fuse_set_zero_arg0(&args);
+	args.in_args[1].size = strlen(name) + 1;
+	args.in_args[1].value = name;
+	err = fuse_simple_request(fm, &args);
+	if (err == -ENOSYS) {
+		fm->fc->no_removexattr = 1;
+		err = -EOPNOTSUPP;
+	}
+	if (!err)
+		fuse_update_ctime(inode);
+
+	return err;
+}
+
+static int fuse_xattr_get(const struct xattr_handler *handler,
+			 struct dentry *dentry, struct inode *inode,
+			 const char *name, void *value, size_t size)
+{
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	return fuse_getxattr(inode, name, value, size);
+}
+
+static int fuse_xattr_set(const struct xattr_handler *handler,
+			  struct mnt_idmap *idmap,
+			  struct dentry *dentry, struct inode *inode,
+			  const char *name, const void *value, size_t size,
+			  int flags)
+{
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	if (!value)
+		return fuse_removexattr(inode, name);
+
+	return fuse_setxattr(inode, name, value, size, flags, 0);
+}
+
+static const struct xattr_handler fuse_xattr_handler = {
+	.prefix = "",
+	.get    = fuse_xattr_get,
+	.set    = fuse_xattr_set,
+};
+
+const struct xattr_handler * const fuse_xattr_handlers[] = {
+	&fuse_xattr_handler,
+	NULL
+};