[PATCH 1/2] fuse: optimize struct fuse_conn fields

[PATCH 1/2] fuse: optimize struct fuse_conn fields

[Joanne Koong]

Use a bitfield for tracking initialized, blocked, aborted, and io_uring
state of the fuse connection. Track connected state using a bool instead
of an unsigned.

On a 64-bit system, this shaves off 16 bytes from the size of struct
fuse_conn.

No functional changes.

Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
---
 fs/fuse/fuse_i.h | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index b54f4f57789f..6aecada8aadd 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -690,24 +690,24 @@ struct fuse_conn {
 	 * 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;
+	bool connected;
+
+	/** Flag indicating that INIT reply has been received. Allocating
+	 * any fuse request will be suspended until the flag is set */
+	int initialized:1;
+
+	/** Flag indicating if connection is blocked.  This will be
+	    the case before the INIT reply is received, and if there
+	    are too many outstanding backgrounds requests */
+	int blocked:1;
 
 	/** Connection aborted via sysfs */
-	bool aborted;
+	bool aborted:1;
 
 	/** Connection failed (version mismatch).  Cannot race with
 	    setting other bitfields since it is only set once in INIT
@@ -896,7 +896,7 @@ struct fuse_conn {
 	unsigned int no_link:1;
 
 	/* Use io_uring for communication */
-	unsigned int io_uring;
+	unsigned int io_uring:1;
 
 	/** Maximum stack depth for passthrough backing files */
 	int max_stack_depth;
-- 
2.47.1

[PATCH 2/2] fuse: clean up struct fuse_conn bitfields

[Joanne Koong]

Use boolean bitfields to improve clarity/readability and to be
consistent with bitfields in other fuse structs (eg fuse_fs_context,
fuse_args). Set bitfield values with true/false to improve readability.

No functional changes.

Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
---
 fs/fuse/cuse.c      |   2 +-
 fs/fuse/dev.c       |  10 ++--
 fs/fuse/dev_uring.c |   4 +-
 fs/fuse/dir.c       |  16 +++---
 fs/fuse/file.c      |  18 +++----
 fs/fuse/fuse_i.h    | 122 ++++++++++++++++++++++----------------------
 fs/fuse/inode.c     |  76 +++++++++++++--------------
 fs/fuse/virtio_fs.c |   4 +-
 fs/fuse/xattr.c     |   8 +--
 9 files changed, 130 insertions(+), 130 deletions(-)

diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c
index b39844d75a80..d5f926b4547c 100644
--- a/fs/fuse/cuse.c
+++ b/fs/fuse/cuse.c
@@ -523,7 +523,7 @@ static int cuse_channel_open(struct inode *inode, struct file *file)
 
 	INIT_LIST_HEAD(&cc->list);
 
-	cc->fc.initialized = 1;
+	cc->fc.initialized = true;
 	rc = cuse_send_init(cc);
 	if (rc) {
 		fuse_dev_free(fud);
diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index 155bb6aeaef5..67d07b4c778a 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -166,7 +166,7 @@ void fuse_set_initialized(struct fuse_conn *fc)
 {
 	/* Make sure stores before this are seen on another CPU */
 	smp_wmb();
-	fc->initialized = 1;
+	fc->initialized = true;
 }
 
 static bool fuse_block_alloc(struct fuse_conn *fc, bool for_background)
@@ -460,7 +460,7 @@ void fuse_request_end(struct fuse_req *req)
 		spin_lock(&fc->bg_lock);
 		clear_bit(FR_BACKGROUND, &req->flags);
 		if (fc->num_background == fc->max_background) {
-			fc->blocked = 0;
+			fc->blocked = false;
 			wake_up(&fc->blocked_waitq);
 		} else if (!fc->blocked) {
 			/*
@@ -720,7 +720,7 @@ static int fuse_request_queue_background(struct fuse_req *req)
 	if (likely(fc->connected)) {
 		fc->num_background++;
 		if (fc->num_background == fc->max_background)
-			fc->blocked = 1;
+			fc->blocked = true;
 		list_add_tail(&req->list, &fc->bg_queue);
 		flush_bg_queue(fc);
 		queued = true;
@@ -2173,7 +2173,7 @@ static ssize_t fuse_dev_do_write(struct fuse_dev *fud,
 		if (nbytes != sizeof(struct fuse_out_header))
 			err = -EINVAL;
 		else if (oh.error == -ENOSYS)
-			fc->no_interrupt = 1;
+			fc->no_interrupt = true;
 		else if (oh.error == -EAGAIN)
 			err = queue_interrupt(req);
 
@@ -2435,7 +2435,7 @@ void fuse_abort_conn(struct fuse_conn *fc)
 			spin_unlock(&fpq->lock);
 		}
 		spin_lock(&fc->bg_lock);
-		fc->blocked = 0;
+		fc->blocked = false;
 		fc->max_background = UINT_MAX;
 		flush_bg_queue(fc);
 		spin_unlock(&fc->bg_lock);
diff --git a/fs/fuse/dev_uring.c b/fs/fuse/dev_uring.c
index 249b210becb1..ef470c4a9261 100644
--- a/fs/fuse/dev_uring.c
+++ b/fs/fuse/dev_uring.c
@@ -1169,7 +1169,7 @@ int fuse_uring_cmd(struct io_uring_cmd *cmd, unsigned int issue_flags)
 		if (err) {
 			pr_info_once("FUSE_IO_URING_CMD_REGISTER failed err=%d\n",
 				     err);
-			fc->io_uring = 0;
+			fc->io_uring = false;
 			wake_up_all(&fc->blocked_waitq);
 			return err;
 		}
@@ -1325,7 +1325,7 @@ bool fuse_uring_queue_bq_req(struct fuse_req *req)
 	spin_lock(&fc->bg_lock);
 	fc->num_background++;
 	if (fc->num_background == fc->max_background)
-		fc->blocked = 1;
+		fc->blocked = true;
 	fuse_uring_flush_bg(queue);
 	spin_unlock(&fc->bg_lock);
 
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 1fb0b15a6088..25e06c73f83b 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -774,7 +774,7 @@ static int fuse_atomic_open(struct inode *dir, struct dentry *entry,
 
 	err = fuse_create_open(idmap, dir, entry, file, flags, mode, FUSE_CREATE);
 	if (err == -ENOSYS) {
-		fc->no_create = 1;
+		fc->no_create = true;
 		goto mknod;
 	} else if (err == -EEXIST)
 		fuse_invalidate_entry(entry);
@@ -923,7 +923,7 @@ static int fuse_tmpfile(struct mnt_idmap *idmap, struct inode *dir,
 	err = fuse_create_open(idmap, dir, file->f_path.dentry, file,
 			       file->f_flags, mode, FUSE_TMPFILE);
 	if (err == -ENOSYS) {
-		fc->no_tmpfile = 1;
+		fc->no_tmpfile = true;
 		err = -EOPNOTSUPP;
 	}
 	return err;
@@ -1133,7 +1133,7 @@ static int fuse_rename2(struct mnt_idmap *idmap, struct inode *olddir,
 					 FUSE_RENAME2,
 					 sizeof(struct fuse_rename2_in));
 		if (err == -ENOSYS) {
-			fc->no_rename2 = 1;
+			fc->no_rename2 = true;
 			err = -EINVAL;
 		}
 	} else {
@@ -1172,7 +1172,7 @@ static int fuse_link(struct dentry *entry, struct inode *newdir,
 		fuse_invalidate_attr(inode);
 
 	if (err == -ENOSYS)
-		fm->fc->no_link = 1;
+		fm->fc->no_link = true;
 out:
 	if (fm->fc->no_link)
 		return -EPERM;
@@ -1376,7 +1376,7 @@ static int fuse_update_get_attr(struct mnt_idmap *idmap, struct inode *inode,
 		if (!fc->no_statx && (request_mask & ~STATX_BASIC_STATS)) {
 			err = fuse_do_statx(idmap, inode, file, stat);
 			if (err == -ENOSYS) {
-				fc->no_statx = 1;
+				fc->no_statx = true;
 				err = 0;
 				goto retry;
 			}
@@ -1521,7 +1521,7 @@ static int fuse_access(struct inode *inode, int mask)
 	/*
 	 * We should not send FUSE_ACCESS to the userspace
 	 * when idmapped mounts are enabled as for this case
-	 * we have fc->default_permissions = 1 and access
+	 * we have fc->default_permissions = true and access
 	 * permission checks are done on the kernel side.
 	 */
 	WARN_ON_ONCE(!(fm->sb->s_iflags & SB_I_NOIDMAP));
@@ -1538,7 +1538,7 @@ static int fuse_access(struct inode *inode, int mask)
 	args.in_args[0].value = &inarg;
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fm->fc->no_access = 1;
+		fm->fc->no_access = true;
 		err = 0;
 	}
 	return err;
@@ -1754,7 +1754,7 @@ static int fuse_dir_fsync(struct file *file, loff_t start, loff_t end,
 	inode_lock(inode);
 	err = fuse_fsync_common(file, start, end, datasync, FUSE_FSYNCDIR);
 	if (err == -ENOSYS) {
-		fc->no_fsyncdir = 1;
+		fc->no_fsyncdir = true;
 		err = 0;
 	}
 	inode_unlock(inode);
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index e203dd4fcc0f..f8ca57cebc3b 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -155,9 +155,9 @@ struct fuse_file *fuse_file_open(struct fuse_mount *fm, u64 nodeid,
 			kfree(ff->args);
 			ff->args = NULL;
 			if (isdir)
-				fc->no_opendir = 1;
+				fc->no_opendir = true;
 			else
-				fc->no_open = 1;
+				fc->no_open = true;
 		}
 	}
 
@@ -474,7 +474,7 @@ static int fuse_flush(struct file *file, fl_owner_t id)
 
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fm->fc->no_flush = 1;
+		fm->fc->no_flush = true;
 		err = 0;
 	}
 
@@ -549,7 +549,7 @@ static int fuse_fsync(struct file *file, loff_t start, loff_t end,
 
 	err = fuse_fsync_common(file, start, end, datasync, FUSE_FSYNC);
 	if (err == -ENOSYS) {
-		fc->no_fsync = 1;
+		fc->no_fsync = true;
 		err = 0;
 	}
 out:
@@ -2548,7 +2548,7 @@ static sector_t fuse_bmap(struct address_space *mapping, sector_t block)
 	args.out_args[0].value = &outarg;
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS)
-		fm->fc->no_bmap = 1;
+		fm->fc->no_bmap = true;
 
 	return err ? 0 : outarg.block;
 }
@@ -2581,7 +2581,7 @@ static loff_t fuse_lseek(struct file *file, loff_t offset, int whence)
 	err = fuse_simple_request(fm, &args);
 	if (err) {
 		if (err == -ENOSYS) {
-			fm->fc->no_lseek = 1;
+			fm->fc->no_lseek = true;
 			goto fallback;
 		}
 		return err;
@@ -2716,7 +2716,7 @@ __poll_t fuse_file_poll(struct file *file, poll_table *wait)
 	if (!err)
 		return demangle_poll(outarg.revents);
 	if (err == -ENOSYS) {
-		fm->fc->no_poll = 1;
+		fm->fc->no_poll = true;
 		return DEFAULT_POLLMASK;
 	}
 	return EPOLLERR;
@@ -2935,7 +2935,7 @@ static long fuse_file_fallocate(struct file *file, int mode, loff_t offset,
 	args.in_args[0].value = &inarg;
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fm->fc->no_fallocate = 1;
+		fm->fc->no_fallocate = true;
 		err = -EOPNOTSUPP;
 	}
 	if (err)
@@ -3047,7 +3047,7 @@ static ssize_t __fuse_copy_file_range(struct file *file_in, loff_t pos_in,
 	args.out_args[0].value = &outarg;
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fc->no_copy_file_range = 1;
+		fc->no_copy_file_range = true;
 		err = -EOPNOTSUPP;
 	}
 	if (err)
diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index 6aecada8aadd..3d5289cb82a5 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -699,12 +699,12 @@ struct fuse_conn {
 
 	/** Flag indicating that INIT reply has been received. Allocating
 	 * any fuse request will be suspended until the flag is set */
-	int initialized:1;
+	bool initialized:1;
 
 	/** Flag indicating if connection is blocked.  This will be
 	    the case before the INIT reply is received, and if there
 	    are too many outstanding backgrounds requests */
-	int blocked:1;
+	bool blocked:1;
 
 	/** Connection aborted via sysfs */
 	bool aborted:1;
@@ -712,37 +712,37 @@ struct fuse_conn {
 	/** Connection failed (version mismatch).  Cannot race with
 	    setting other bitfields since it is only set once in INIT
 	    reply, before any other request, and never cleared */
-	unsigned conn_error:1;
+	bool conn_error:1;
 
 	/** Connection successful.  Only set in INIT */
-	unsigned conn_init:1;
+	bool conn_init:1;
 
 	/** Do readahead asynchronously?  Only set in INIT */
-	unsigned async_read:1;
+	bool async_read:1;
 
 	/** Return an unique read error after abort.  Only set in INIT */
-	unsigned abort_err:1;
+	bool abort_err:1;
 
 	/** Do not send separate SETATTR request before open(O_TRUNC)  */
-	unsigned atomic_o_trunc:1;
+	bool atomic_o_trunc:1;
 
 	/** Filesystem supports NFS exporting.  Only set in INIT */
-	unsigned export_support:1;
+	bool export_support:1;
 
 	/** write-back cache policy (default is write-through) */
-	unsigned writeback_cache:1;
+	bool writeback_cache:1;
 
 	/** allow parallel lookups and readdir (default is serialized) */
-	unsigned parallel_dirops:1;
+	bool parallel_dirops:1;
 
 	/** handle fs handles killing suid/sgid/cap on write/chown/trunc */
-	unsigned handle_killpriv:1;
+	bool handle_killpriv:1;
 
 	/** cache READLINK responses in page cache */
-	unsigned cache_symlinks:1;
+	bool cache_symlinks:1;
 
 	/* show legacy mount options */
-	unsigned int legacy_opts_show:1;
+	bool legacy_opts_show:1;
 
 	/*
 	 * fs kills suid/sgid/cap on write/chown/trunc. suid is killed on
@@ -750,7 +750,7 @@ struct fuse_conn {
 	 * on write/truncate only if caller did not have CAP_FSETID as well as
 	 * file has group execute permission.
 	 */
-	unsigned handle_killpriv_v2:1;
+	bool handle_killpriv_v2:1;
 
 	/*
 	 * The following bitfields are only for optimization purposes
@@ -758,145 +758,145 @@ struct fuse_conn {
 	 */
 
 	/** Is open/release not implemented by fs? */
-	unsigned no_open:1;
+	bool no_open:1;
 
 	/** Is opendir/releasedir not implemented by fs? */
-	unsigned no_opendir:1;
+	bool no_opendir:1;
 
 	/** Is fsync not implemented by fs? */
-	unsigned no_fsync:1;
+	bool no_fsync:1;
 
 	/** Is fsyncdir not implemented by fs? */
-	unsigned no_fsyncdir:1;
+	bool no_fsyncdir:1;
 
 	/** Is flush not implemented by fs? */
-	unsigned no_flush:1;
+	bool no_flush:1;
 
 	/** Is setxattr not implemented by fs? */
-	unsigned no_setxattr:1;
+	bool no_setxattr:1;
 
 	/** Does file server support extended setxattr */
-	unsigned setxattr_ext:1;
+	bool setxattr_ext:1;
 
 	/** Is getxattr not implemented by fs? */
-	unsigned no_getxattr:1;
+	bool no_getxattr:1;
 
 	/** Is listxattr not implemented by fs? */
-	unsigned no_listxattr:1;
+	bool no_listxattr:1;
 
 	/** Is removexattr not implemented by fs? */
-	unsigned no_removexattr:1;
+	bool no_removexattr:1;
 
 	/** Are posix file locking primitives not implemented by fs? */
-	unsigned no_lock:1;
+	bool no_lock:1;
 
 	/** Is access not implemented by fs? */
-	unsigned no_access:1;
+	bool no_access:1;
 
 	/** Is create not implemented by fs? */
-	unsigned no_create:1;
+	bool no_create:1;
 
 	/** Is interrupt not implemented by fs? */
-	unsigned no_interrupt:1;
+	bool no_interrupt:1;
 
 	/** Is bmap not implemented by fs? */
-	unsigned no_bmap:1;
+	bool no_bmap:1;
 
 	/** Is poll not implemented by fs? */
-	unsigned no_poll:1;
+	bool no_poll:1;
 
 	/** Do multi-page cached writes */
-	unsigned big_writes:1;
+	bool big_writes:1;
 
 	/** Don't apply umask to creation modes */
-	unsigned dont_mask:1;
+	bool dont_mask:1;
 
 	/** Are BSD file locking primitives not implemented by fs? */
-	unsigned no_flock:1;
+	bool no_flock:1;
 
 	/** Is fallocate not implemented by fs? */
-	unsigned no_fallocate:1;
+	bool no_fallocate:1;
 
 	/** Is rename with flags implemented by fs? */
-	unsigned no_rename2:1;
+	bool no_rename2:1;
 
 	/** Use enhanced/automatic page cache invalidation. */
-	unsigned auto_inval_data:1;
+	bool auto_inval_data:1;
 
 	/** Filesystem is fully responsible for page cache invalidation. */
-	unsigned explicit_inval_data:1;
+	bool explicit_inval_data:1;
 
 	/** Does the filesystem support readdirplus? */
-	unsigned do_readdirplus:1;
+	bool do_readdirplus:1;
 
 	/** Does the filesystem want adaptive readdirplus? */
-	unsigned readdirplus_auto:1;
+	bool readdirplus_auto:1;
 
 	/** Does the filesystem support asynchronous direct-IO submission? */
-	unsigned async_dio:1;
+	bool async_dio:1;
 
 	/** Is lseek not implemented by fs? */
-	unsigned no_lseek:1;
+	bool no_lseek:1;
 
 	/** Does the filesystem support posix acls? */
-	unsigned posix_acl:1;
+	bool posix_acl:1;
 
 	/** Check permissions based on the file mode or not? */
-	unsigned default_permissions:1;
+	bool default_permissions:1;
 
 	/** Allow other than the mounter user to access the filesystem ? */
-	unsigned allow_other:1;
+	bool allow_other:1;
 
 	/** Does the filesystem support copy_file_range? */
-	unsigned no_copy_file_range:1;
+	bool no_copy_file_range:1;
 
 	/* Send DESTROY request */
-	unsigned int destroy:1;
+	bool destroy:1;
 
 	/* Delete dentries that have gone stale */
-	unsigned int delete_stale:1;
+	bool delete_stale:1;
 
 	/** Do not create entry in fusectl fs */
-	unsigned int no_control:1;
+	bool no_control:1;
 
 	/** Do not allow MNT_FORCE umount */
-	unsigned int no_force_umount:1;
+	bool no_force_umount:1;
 
 	/* Auto-mount submounts announced by the server */
-	unsigned int auto_submounts:1;
+	bool auto_submounts:1;
 
 	/* Propagate syncfs() to server */
-	unsigned int sync_fs:1;
+	bool sync_fs:1;
 
 	/* Initialize security xattrs when creating a new inode */
-	unsigned int init_security:1;
+	bool init_security:1;
 
 	/* Add supplementary group info when creating a new inode */
-	unsigned int create_supp_group:1;
+	bool create_supp_group:1;
 
 	/* Does the filesystem support per inode DAX? */
-	unsigned int inode_dax:1;
+	bool inode_dax:1;
 
 	/* Is tmpfile not implemented by fs? */
-	unsigned int no_tmpfile:1;
+	bool no_tmpfile:1;
 
 	/* Relax restrictions to allow shared mmap in FOPEN_DIRECT_IO mode */
-	unsigned int direct_io_allow_mmap:1;
+	bool direct_io_allow_mmap:1;
 
 	/* Is statx not implemented by fs? */
-	unsigned int no_statx:1;
+	bool no_statx:1;
 
 	/** Passthrough support for read/write IO */
-	unsigned int passthrough:1;
+	bool passthrough:1;
 
 	/* Use pages instead of pointer for kernel I/O */
-	unsigned int use_pages_for_kvec_io:1;
+	bool use_pages_for_kvec_io:1;
 
 	/* Is link not implemented by fs? */
-	unsigned int no_link:1;
+	bool no_link:1;
 
 	/* Use io_uring for communication */
-	unsigned int io_uring:1;
+	bool io_uring:1;
 
 	/** Maximum stack depth for passthrough backing files */
 	int max_stack_depth;
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index bfe8d8af46f3..43b6643635ee 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -751,7 +751,7 @@ static int fuse_sync_fs(struct super_block *sb, int wait)
 
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fc->sync_fs = 0;
+		fc->sync_fs = false;
 		err = 0;
 	}
 
@@ -973,9 +973,9 @@ void fuse_conn_init(struct fuse_conn *fc, struct fuse_mount *fm,
 	fc->congestion_threshold = FUSE_DEFAULT_CONGESTION_THRESHOLD;
 	atomic64_set(&fc->khctr, 0);
 	fc->polled_files = RB_ROOT;
-	fc->blocked = 0;
-	fc->initialized = 0;
-	fc->connected = 1;
+	fc->blocked = false;
+	fc->initialized = false;
+	fc->connected = true;
 	atomic64_set(&fc->attr_version, 1);
 	atomic64_set(&fc->evict_ctr, 1);
 	get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key));
@@ -1323,54 +1323,54 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
 
 			ra_pages = arg->max_readahead / PAGE_SIZE;
 			if (flags & FUSE_ASYNC_READ)
-				fc->async_read = 1;
+				fc->async_read = true;
 			if (!(flags & FUSE_POSIX_LOCKS))
-				fc->no_lock = 1;
+				fc->no_lock = true;
 			if (arg->minor >= 17) {
 				if (!(flags & FUSE_FLOCK_LOCKS))
-					fc->no_flock = 1;
+					fc->no_flock = true;
 			} else {
 				if (!(flags & FUSE_POSIX_LOCKS))
-					fc->no_flock = 1;
+					fc->no_flock = true;
 			}
 			if (flags & FUSE_ATOMIC_O_TRUNC)
-				fc->atomic_o_trunc = 1;
+				fc->atomic_o_trunc = true;
 			if (arg->minor >= 9) {
 				/* LOOKUP has dependency on proto version */
 				if (flags & FUSE_EXPORT_SUPPORT)
-					fc->export_support = 1;
+					fc->export_support = true;
 			}
 			if (flags & FUSE_BIG_WRITES)
-				fc->big_writes = 1;
+				fc->big_writes = true;
 			if (flags & FUSE_DONT_MASK)
-				fc->dont_mask = 1;
+				fc->dont_mask = true;
 			if (flags & FUSE_AUTO_INVAL_DATA)
-				fc->auto_inval_data = 1;
+				fc->auto_inval_data = true;
 			else if (flags & FUSE_EXPLICIT_INVAL_DATA)
-				fc->explicit_inval_data = 1;
+				fc->explicit_inval_data = true;
 			if (flags & FUSE_DO_READDIRPLUS) {
-				fc->do_readdirplus = 1;
+				fc->do_readdirplus = true;
 				if (flags & FUSE_READDIRPLUS_AUTO)
-					fc->readdirplus_auto = 1;
+					fc->readdirplus_auto = true;
 			}
 			if (flags & FUSE_ASYNC_DIO)
-				fc->async_dio = 1;
+				fc->async_dio = true;
 			if (flags & FUSE_WRITEBACK_CACHE)
-				fc->writeback_cache = 1;
+				fc->writeback_cache = true;
 			if (flags & FUSE_PARALLEL_DIROPS)
-				fc->parallel_dirops = 1;
+				fc->parallel_dirops = true;
 			if (flags & FUSE_HANDLE_KILLPRIV)
-				fc->handle_killpriv = 1;
+				fc->handle_killpriv = true;
 			if (arg->time_gran && arg->time_gran <= 1000000000)
 				fm->sb->s_time_gran = arg->time_gran;
 			if ((flags & FUSE_POSIX_ACL)) {
-				fc->default_permissions = 1;
-				fc->posix_acl = 1;
+				fc->default_permissions = true;
+				fc->posix_acl = true;
 			}
 			if (flags & FUSE_CACHE_SYMLINKS)
-				fc->cache_symlinks = 1;
+				fc->cache_symlinks = true;
 			if (flags & FUSE_ABORT_ERROR)
-				fc->abort_err = 1;
+				fc->abort_err = true;
 			if (flags & FUSE_MAX_PAGES) {
 				fc->max_pages =
 					min_t(unsigned int, fc->max_pages_limit,
@@ -1389,20 +1389,20 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
 					ok = false;
 				}
 				if (flags & FUSE_HAS_INODE_DAX)
-					fc->inode_dax = 1;
+					fc->inode_dax = true;
 			}
 			if (flags & FUSE_HANDLE_KILLPRIV_V2) {
-				fc->handle_killpriv_v2 = 1;
+				fc->handle_killpriv_v2 = true;
 				fm->sb->s_flags |= SB_NOSEC;
 			}
 			if (flags & FUSE_SETXATTR_EXT)
-				fc->setxattr_ext = 1;
+				fc->setxattr_ext = true;
 			if (flags & FUSE_SECURITY_CTX)
-				fc->init_security = 1;
+				fc->init_security = true;
 			if (flags & FUSE_CREATE_SUPP_GROUP)
-				fc->create_supp_group = 1;
+				fc->create_supp_group = true;
 			if (flags & FUSE_DIRECT_IO_ALLOW_MMAP)
-				fc->direct_io_allow_mmap = 1;
+				fc->direct_io_allow_mmap = true;
 			/*
 			 * max_stack_depth is the max stack depth of FUSE fs,
 			 * so it has to be at least 1 to support passthrough
@@ -1422,7 +1422,7 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
 			    arg->max_stack_depth > 0 &&
 			    arg->max_stack_depth <= FILESYSTEM_MAX_STACK_DEPTH &&
 			    !(flags & FUSE_WRITEBACK_CACHE))  {
-				fc->passthrough = 1;
+				fc->passthrough = true;
 				fc->max_stack_depth = arg->max_stack_depth;
 				fm->sb->s_stack_depth = arg->max_stack_depth;
 			}
@@ -1435,14 +1435,14 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
 					ok = false;
 			}
 			if (flags & FUSE_OVER_IO_URING && fuse_uring_enabled())
-				fc->io_uring = 1;
+				fc->io_uring = true;
 
 			if (flags & FUSE_REQUEST_TIMEOUT)
 				timeout = arg->request_timeout;
 		} else {
 			ra_pages = fc->max_read / PAGE_SIZE;
-			fc->no_lock = 1;
-			fc->no_flock = 1;
+			fc->no_lock = true;
+			fc->no_flock = true;
 		}
 
 		init_server_timeout(fc, timeout);
@@ -1452,13 +1452,13 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
 		fc->minor = arg->minor;
 		fc->max_write = arg->minor < 5 ? 4096 : arg->max_write;
 		fc->max_write = max_t(unsigned, 4096, fc->max_write);
-		fc->conn_init = 1;
+		fc->conn_init = true;
 	}
 	kfree(ia);
 
 	if (!ok) {
-		fc->conn_init = 0;
-		fc->conn_error = 1;
+		fc->conn_init = false;
+		fc->conn_error = true;
 	}
 
 	fuse_set_initialized(fc);
@@ -1835,7 +1835,7 @@ int fuse_fill_super_common(struct super_block *sb, struct fuse_fs_context *ctx)
 
 	/* Handle umasking inside the fuse code */
 	if (sb->s_flags & SB_POSIXACL)
-		fc->dont_mask = 1;
+		fc->dont_mask = true;
 	sb->s_flags |= SB_POSIXACL;
 
 	fc->default_permissions = ctx->default_permissions;
diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
index 2c7b24cb67ad..d7b0f7f288e3 100644
--- a/fs/fuse/virtio_fs.c
+++ b/fs/fuse/virtio_fs.c
@@ -1531,8 +1531,8 @@ static const struct fuse_iqueue_ops virtio_fs_fiq_ops = {
 static inline void virtio_fs_ctx_set_defaults(struct fuse_fs_context *ctx)
 {
 	ctx->rootmode = S_IFDIR;
-	ctx->default_permissions = 1;
-	ctx->allow_other = 1;
+	ctx->default_permissions = true;
+	ctx->allow_other = true;
 	ctx->max_read = UINT_MAX;
 	ctx->blksize = 512;
 	ctx->destroy = true;
diff --git a/fs/fuse/xattr.c b/fs/fuse/xattr.c
index 93dfb06b6cea..73100df1b24b 100644
--- a/fs/fuse/xattr.c
+++ b/fs/fuse/xattr.c
@@ -39,7 +39,7 @@ int fuse_setxattr(struct inode *inode, const char *name, const void *value,
 	args.in_args[2].value = value;
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fm->fc->no_setxattr = 1;
+		fm->fc->no_setxattr = true;
 		err = -EOPNOTSUPP;
 	}
 	if (!err)
@@ -83,7 +83,7 @@ ssize_t fuse_getxattr(struct inode *inode, const char *name, void *value,
 	if (!ret && !size)
 		ret = min_t(size_t, outarg.size, XATTR_SIZE_MAX);
 	if (ret == -ENOSYS) {
-		fm->fc->no_getxattr = 1;
+		fm->fc->no_getxattr = true;
 		ret = -EOPNOTSUPP;
 	}
 	return ret;
@@ -147,7 +147,7 @@ ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size)
 	if (ret > 0 && size)
 		ret = fuse_verify_xattr_list(list, ret);
 	if (ret == -ENOSYS) {
-		fm->fc->no_listxattr = 1;
+		fm->fc->no_listxattr = true;
 		ret = -EOPNOTSUPP;
 	}
 	return ret;
@@ -170,7 +170,7 @@ int fuse_removexattr(struct inode *inode, const char *name)
 	args.in_args[1].value = name;
 	err = fuse_simple_request(fm, &args);
 	if (err == -ENOSYS) {
-		fm->fc->no_removexattr = 1;
+		fm->fc->no_removexattr = true;
 		err = -EOPNOTSUPP;
 	}
 	if (!err)
-- 
2.47.1

Re: [PATCH 1/2] fuse: optimize struct fuse_conn fields

[Bernd Schubert]

On 4/18/25 23:06, Joanne Koong wrote:
> Use a bitfield for tracking initialized, blocked, aborted, and io_uring
> state of the fuse connection. Track connected state using a bool instead
> of an unsigned.
> 
> On a 64-bit system, this shaves off 16 bytes from the size of struct
> fuse_conn.
> 
> No functional changes.
> 
> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
> ---
>  fs/fuse/fuse_i.h | 24 ++++++++++++------------
>  1 file changed, 12 insertions(+), 12 deletions(-)
> 
> diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
> index b54f4f57789f..6aecada8aadd 100644
> --- a/fs/fuse/fuse_i.h
> +++ b/fs/fuse/fuse_i.h
> @@ -690,24 +690,24 @@ struct fuse_conn {
>  	 * 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;
> +	bool connected;
> +
> +	/** Flag indicating that INIT reply has been received. Allocating
> +	 * any fuse request will be suspended until the flag is set */
> +	int initialized:1;
> +
> +	/** Flag indicating if connection is blocked.  This will be
> +	    the case before the INIT reply is received, and if there
> +	    are too many outstanding backgrounds requests */
> +	int blocked:1;
>  
>  	/** Connection aborted via sysfs */
> -	bool aborted;
> +	bool aborted:1;
>  
>  	/** Connection failed (version mismatch).  Cannot race with
>  	    setting other bitfields since it is only set once in INIT
> @@ -896,7 +896,7 @@ struct fuse_conn {
>  	unsigned int no_link:1;
>  
>  	/* Use io_uring for communication */
> -	unsigned int io_uring;
> +	unsigned int io_uring:1;

Ah yes, I forgotten to change it. There had been io-uring patch versions
that were using READ_ONCE and then I forgot to update it.



Reviewed-by: Bernd Schubert <bschubert@ddn.com>

Re: [PATCH 2/2] fuse: clean up struct fuse_conn bitfields

[Bernd Schubert]

On 4/18/25 23:06, Joanne Koong wrote:
> Use boolean bitfields to improve clarity/readability and to be
> consistent with bitfields in other fuse structs (eg fuse_fs_context,
> fuse_args). Set bitfield values with true/false to improve readability.
> 
> No functional changes.
> 
> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>


Reviewed-by: Bernd Schubert <bschubert@ddn.com>

> ---
>  fs/fuse/cuse.c      |   2 +-
>  fs/fuse/dev.c       |  10 ++--
>  fs/fuse/dev_uring.c |   4 +-
>  fs/fuse/dir.c       |  16 +++---
>  fs/fuse/file.c      |  18 +++----
>  fs/fuse/fuse_i.h    | 122 ++++++++++++++++++++++----------------------
>  fs/fuse/inode.c     |  76 +++++++++++++--------------
>  fs/fuse/virtio_fs.c |   4 +-
>  fs/fuse/xattr.c     |   8 +--
>  9 files changed, 130 insertions(+), 130 deletions(-)
> 
> diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c
> index b39844d75a80..d5f926b4547c 100644
> --- a/fs/fuse/cuse.c
> +++ b/fs/fuse/cuse.c
> @@ -523,7 +523,7 @@ static int cuse_channel_open(struct inode *inode, struct file *file)
>  
>  	INIT_LIST_HEAD(&cc->list);
>  
> -	cc->fc.initialized = 1;
> +	cc->fc.initialized = true;
>  	rc = cuse_send_init(cc);
>  	if (rc) {
>  		fuse_dev_free(fud);
> diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
> index 155bb6aeaef5..67d07b4c778a 100644
> --- a/fs/fuse/dev.c
> +++ b/fs/fuse/dev.c
> @@ -166,7 +166,7 @@ void fuse_set_initialized(struct fuse_conn *fc)
>  {
>  	/* Make sure stores before this are seen on another CPU */
>  	smp_wmb();
> -	fc->initialized = 1;
> +	fc->initialized = true;
>  }
>  
>  static bool fuse_block_alloc(struct fuse_conn *fc, bool for_background)
> @@ -460,7 +460,7 @@ void fuse_request_end(struct fuse_req *req)
>  		spin_lock(&fc->bg_lock);
>  		clear_bit(FR_BACKGROUND, &req->flags);
>  		if (fc->num_background == fc->max_background) {
> -			fc->blocked = 0;
> +			fc->blocked = false;
>  			wake_up(&fc->blocked_waitq);
>  		} else if (!fc->blocked) {
>  			/*
> @@ -720,7 +720,7 @@ static int fuse_request_queue_background(struct fuse_req *req)
>  	if (likely(fc->connected)) {
>  		fc->num_background++;
>  		if (fc->num_background == fc->max_background)
> -			fc->blocked = 1;
> +			fc->blocked = true;
>  		list_add_tail(&req->list, &fc->bg_queue);
>  		flush_bg_queue(fc);
>  		queued = true;
> @@ -2173,7 +2173,7 @@ static ssize_t fuse_dev_do_write(struct fuse_dev *fud,
>  		if (nbytes != sizeof(struct fuse_out_header))
>  			err = -EINVAL;
>  		else if (oh.error == -ENOSYS)
> -			fc->no_interrupt = 1;
> +			fc->no_interrupt = true;
>  		else if (oh.error == -EAGAIN)
>  			err = queue_interrupt(req);
>  
> @@ -2435,7 +2435,7 @@ void fuse_abort_conn(struct fuse_conn *fc)
>  			spin_unlock(&fpq->lock);
>  		}
>  		spin_lock(&fc->bg_lock);
> -		fc->blocked = 0;
> +		fc->blocked = false;
>  		fc->max_background = UINT_MAX;
>  		flush_bg_queue(fc);
>  		spin_unlock(&fc->bg_lock);
> diff --git a/fs/fuse/dev_uring.c b/fs/fuse/dev_uring.c
> index 249b210becb1..ef470c4a9261 100644
> --- a/fs/fuse/dev_uring.c
> +++ b/fs/fuse/dev_uring.c
> @@ -1169,7 +1169,7 @@ int fuse_uring_cmd(struct io_uring_cmd *cmd, unsigned int issue_flags)
>  		if (err) {
>  			pr_info_once("FUSE_IO_URING_CMD_REGISTER failed err=%d\n",
>  				     err);
> -			fc->io_uring = 0;
> +			fc->io_uring = false;
>  			wake_up_all(&fc->blocked_waitq);
>  			return err;
>  		}
> @@ -1325,7 +1325,7 @@ bool fuse_uring_queue_bq_req(struct fuse_req *req)
>  	spin_lock(&fc->bg_lock);
>  	fc->num_background++;
>  	if (fc->num_background == fc->max_background)
> -		fc->blocked = 1;
> +		fc->blocked = true;
>  	fuse_uring_flush_bg(queue);
>  	spin_unlock(&fc->bg_lock);
>  
> diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
> index 1fb0b15a6088..25e06c73f83b 100644
> --- a/fs/fuse/dir.c
> +++ b/fs/fuse/dir.c
> @@ -774,7 +774,7 @@ static int fuse_atomic_open(struct inode *dir, struct dentry *entry,
>  
>  	err = fuse_create_open(idmap, dir, entry, file, flags, mode, FUSE_CREATE);
>  	if (err == -ENOSYS) {
> -		fc->no_create = 1;
> +		fc->no_create = true;
>  		goto mknod;
>  	} else if (err == -EEXIST)
>  		fuse_invalidate_entry(entry);
> @@ -923,7 +923,7 @@ static int fuse_tmpfile(struct mnt_idmap *idmap, struct inode *dir,
>  	err = fuse_create_open(idmap, dir, file->f_path.dentry, file,
>  			       file->f_flags, mode, FUSE_TMPFILE);
>  	if (err == -ENOSYS) {
> -		fc->no_tmpfile = 1;
> +		fc->no_tmpfile = true;
>  		err = -EOPNOTSUPP;
>  	}
>  	return err;
> @@ -1133,7 +1133,7 @@ static int fuse_rename2(struct mnt_idmap *idmap, struct inode *olddir,
>  					 FUSE_RENAME2,
>  					 sizeof(struct fuse_rename2_in));
>  		if (err == -ENOSYS) {
> -			fc->no_rename2 = 1;
> +			fc->no_rename2 = true;
>  			err = -EINVAL;
>  		}
>  	} else {
> @@ -1172,7 +1172,7 @@ static int fuse_link(struct dentry *entry, struct inode *newdir,
>  		fuse_invalidate_attr(inode);
>  
>  	if (err == -ENOSYS)
> -		fm->fc->no_link = 1;
> +		fm->fc->no_link = true;
>  out:
>  	if (fm->fc->no_link)
>  		return -EPERM;
> @@ -1376,7 +1376,7 @@ static int fuse_update_get_attr(struct mnt_idmap *idmap, struct inode *inode,
>  		if (!fc->no_statx && (request_mask & ~STATX_BASIC_STATS)) {
>  			err = fuse_do_statx(idmap, inode, file, stat);
>  			if (err == -ENOSYS) {
> -				fc->no_statx = 1;
> +				fc->no_statx = true;
>  				err = 0;
>  				goto retry;
>  			}
> @@ -1521,7 +1521,7 @@ static int fuse_access(struct inode *inode, int mask)
>  	/*
>  	 * We should not send FUSE_ACCESS to the userspace
>  	 * when idmapped mounts are enabled as for this case
> -	 * we have fc->default_permissions = 1 and access
> +	 * we have fc->default_permissions = true and access
>  	 * permission checks are done on the kernel side.
>  	 */
>  	WARN_ON_ONCE(!(fm->sb->s_iflags & SB_I_NOIDMAP));
> @@ -1538,7 +1538,7 @@ static int fuse_access(struct inode *inode, int mask)
>  	args.in_args[0].value = &inarg;
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fm->fc->no_access = 1;
> +		fm->fc->no_access = true;
>  		err = 0;
>  	}
>  	return err;
> @@ -1754,7 +1754,7 @@ static int fuse_dir_fsync(struct file *file, loff_t start, loff_t end,
>  	inode_lock(inode);
>  	err = fuse_fsync_common(file, start, end, datasync, FUSE_FSYNCDIR);
>  	if (err == -ENOSYS) {
> -		fc->no_fsyncdir = 1;
> +		fc->no_fsyncdir = true;
>  		err = 0;
>  	}
>  	inode_unlock(inode);
> diff --git a/fs/fuse/file.c b/fs/fuse/file.c
> index e203dd4fcc0f..f8ca57cebc3b 100644
> --- a/fs/fuse/file.c
> +++ b/fs/fuse/file.c
> @@ -155,9 +155,9 @@ struct fuse_file *fuse_file_open(struct fuse_mount *fm, u64 nodeid,
>  			kfree(ff->args);
>  			ff->args = NULL;
>  			if (isdir)
> -				fc->no_opendir = 1;
> +				fc->no_opendir = true;
>  			else
> -				fc->no_open = 1;
> +				fc->no_open = true;
>  		}
>  	}
>  
> @@ -474,7 +474,7 @@ static int fuse_flush(struct file *file, fl_owner_t id)
>  
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fm->fc->no_flush = 1;
> +		fm->fc->no_flush = true;
>  		err = 0;
>  	}
>  
> @@ -549,7 +549,7 @@ static int fuse_fsync(struct file *file, loff_t start, loff_t end,
>  
>  	err = fuse_fsync_common(file, start, end, datasync, FUSE_FSYNC);
>  	if (err == -ENOSYS) {
> -		fc->no_fsync = 1;
> +		fc->no_fsync = true;
>  		err = 0;
>  	}
>  out:
> @@ -2548,7 +2548,7 @@ static sector_t fuse_bmap(struct address_space *mapping, sector_t block)
>  	args.out_args[0].value = &outarg;
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS)
> -		fm->fc->no_bmap = 1;
> +		fm->fc->no_bmap = true;
>  
>  	return err ? 0 : outarg.block;
>  }
> @@ -2581,7 +2581,7 @@ static loff_t fuse_lseek(struct file *file, loff_t offset, int whence)
>  	err = fuse_simple_request(fm, &args);
>  	if (err) {
>  		if (err == -ENOSYS) {
> -			fm->fc->no_lseek = 1;
> +			fm->fc->no_lseek = true;
>  			goto fallback;
>  		}
>  		return err;
> @@ -2716,7 +2716,7 @@ __poll_t fuse_file_poll(struct file *file, poll_table *wait)
>  	if (!err)
>  		return demangle_poll(outarg.revents);
>  	if (err == -ENOSYS) {
> -		fm->fc->no_poll = 1;
> +		fm->fc->no_poll = true;
>  		return DEFAULT_POLLMASK;
>  	}
>  	return EPOLLERR;
> @@ -2935,7 +2935,7 @@ static long fuse_file_fallocate(struct file *file, int mode, loff_t offset,
>  	args.in_args[0].value = &inarg;
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fm->fc->no_fallocate = 1;
> +		fm->fc->no_fallocate = true;
>  		err = -EOPNOTSUPP;
>  	}
>  	if (err)
> @@ -3047,7 +3047,7 @@ static ssize_t __fuse_copy_file_range(struct file *file_in, loff_t pos_in,
>  	args.out_args[0].value = &outarg;
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fc->no_copy_file_range = 1;
> +		fc->no_copy_file_range = true;
>  		err = -EOPNOTSUPP;
>  	}
>  	if (err)
> diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
> index 6aecada8aadd..3d5289cb82a5 100644
> --- a/fs/fuse/fuse_i.h
> +++ b/fs/fuse/fuse_i.h
> @@ -699,12 +699,12 @@ struct fuse_conn {
>  
>  	/** Flag indicating that INIT reply has been received. Allocating
>  	 * any fuse request will be suspended until the flag is set */
> -	int initialized:1;
> +	bool initialized:1;
>  
>  	/** Flag indicating if connection is blocked.  This will be
>  	    the case before the INIT reply is received, and if there
>  	    are too many outstanding backgrounds requests */
> -	int blocked:1;
> +	bool blocked:1;
>  
>  	/** Connection aborted via sysfs */
>  	bool aborted:1;
> @@ -712,37 +712,37 @@ struct fuse_conn {
>  	/** Connection failed (version mismatch).  Cannot race with
>  	    setting other bitfields since it is only set once in INIT
>  	    reply, before any other request, and never cleared */
> -	unsigned conn_error:1;
> +	bool conn_error:1;
>  
>  	/** Connection successful.  Only set in INIT */
> -	unsigned conn_init:1;
> +	bool conn_init:1;
>  
>  	/** Do readahead asynchronously?  Only set in INIT */
> -	unsigned async_read:1;
> +	bool async_read:1;
>  
>  	/** Return an unique read error after abort.  Only set in INIT */
> -	unsigned abort_err:1;
> +	bool abort_err:1;
>  
>  	/** Do not send separate SETATTR request before open(O_TRUNC)  */
> -	unsigned atomic_o_trunc:1;
> +	bool atomic_o_trunc:1;
>  
>  	/** Filesystem supports NFS exporting.  Only set in INIT */
> -	unsigned export_support:1;
> +	bool export_support:1;
>  
>  	/** write-back cache policy (default is write-through) */
> -	unsigned writeback_cache:1;
> +	bool writeback_cache:1;
>  
>  	/** allow parallel lookups and readdir (default is serialized) */
> -	unsigned parallel_dirops:1;
> +	bool parallel_dirops:1;
>  
>  	/** handle fs handles killing suid/sgid/cap on write/chown/trunc */
> -	unsigned handle_killpriv:1;
> +	bool handle_killpriv:1;
>  
>  	/** cache READLINK responses in page cache */
> -	unsigned cache_symlinks:1;
> +	bool cache_symlinks:1;
>  
>  	/* show legacy mount options */
> -	unsigned int legacy_opts_show:1;
> +	bool legacy_opts_show:1;
>  
>  	/*
>  	 * fs kills suid/sgid/cap on write/chown/trunc. suid is killed on
> @@ -750,7 +750,7 @@ struct fuse_conn {
>  	 * on write/truncate only if caller did not have CAP_FSETID as well as
>  	 * file has group execute permission.
>  	 */
> -	unsigned handle_killpriv_v2:1;
> +	bool handle_killpriv_v2:1;
>  
>  	/*
>  	 * The following bitfields are only for optimization purposes
> @@ -758,145 +758,145 @@ struct fuse_conn {
>  	 */
>  
>  	/** Is open/release not implemented by fs? */
> -	unsigned no_open:1;
> +	bool no_open:1;
>  
>  	/** Is opendir/releasedir not implemented by fs? */
> -	unsigned no_opendir:1;
> +	bool no_opendir:1;
>  
>  	/** Is fsync not implemented by fs? */
> -	unsigned no_fsync:1;
> +	bool no_fsync:1;
>  
>  	/** Is fsyncdir not implemented by fs? */
> -	unsigned no_fsyncdir:1;
> +	bool no_fsyncdir:1;
>  
>  	/** Is flush not implemented by fs? */
> -	unsigned no_flush:1;
> +	bool no_flush:1;
>  
>  	/** Is setxattr not implemented by fs? */
> -	unsigned no_setxattr:1;
> +	bool no_setxattr:1;
>  
>  	/** Does file server support extended setxattr */
> -	unsigned setxattr_ext:1;
> +	bool setxattr_ext:1;
>  
>  	/** Is getxattr not implemented by fs? */
> -	unsigned no_getxattr:1;
> +	bool no_getxattr:1;
>  
>  	/** Is listxattr not implemented by fs? */
> -	unsigned no_listxattr:1;
> +	bool no_listxattr:1;
>  
>  	/** Is removexattr not implemented by fs? */
> -	unsigned no_removexattr:1;
> +	bool no_removexattr:1;
>  
>  	/** Are posix file locking primitives not implemented by fs? */
> -	unsigned no_lock:1;
> +	bool no_lock:1;
>  
>  	/** Is access not implemented by fs? */
> -	unsigned no_access:1;
> +	bool no_access:1;
>  
>  	/** Is create not implemented by fs? */
> -	unsigned no_create:1;
> +	bool no_create:1;
>  
>  	/** Is interrupt not implemented by fs? */
> -	unsigned no_interrupt:1;
> +	bool no_interrupt:1;
>  
>  	/** Is bmap not implemented by fs? */
> -	unsigned no_bmap:1;
> +	bool no_bmap:1;
>  
>  	/** Is poll not implemented by fs? */
> -	unsigned no_poll:1;
> +	bool no_poll:1;
>  
>  	/** Do multi-page cached writes */
> -	unsigned big_writes:1;
> +	bool big_writes:1;
>  
>  	/** Don't apply umask to creation modes */
> -	unsigned dont_mask:1;
> +	bool dont_mask:1;
>  
>  	/** Are BSD file locking primitives not implemented by fs? */
> -	unsigned no_flock:1;
> +	bool no_flock:1;
>  
>  	/** Is fallocate not implemented by fs? */
> -	unsigned no_fallocate:1;
> +	bool no_fallocate:1;
>  
>  	/** Is rename with flags implemented by fs? */
> -	unsigned no_rename2:1;
> +	bool no_rename2:1;
>  
>  	/** Use enhanced/automatic page cache invalidation. */
> -	unsigned auto_inval_data:1;
> +	bool auto_inval_data:1;
>  
>  	/** Filesystem is fully responsible for page cache invalidation. */
> -	unsigned explicit_inval_data:1;
> +	bool explicit_inval_data:1;
>  
>  	/** Does the filesystem support readdirplus? */
> -	unsigned do_readdirplus:1;
> +	bool do_readdirplus:1;
>  
>  	/** Does the filesystem want adaptive readdirplus? */
> -	unsigned readdirplus_auto:1;
> +	bool readdirplus_auto:1;
>  
>  	/** Does the filesystem support asynchronous direct-IO submission? */
> -	unsigned async_dio:1;
> +	bool async_dio:1;
>  
>  	/** Is lseek not implemented by fs? */
> -	unsigned no_lseek:1;
> +	bool no_lseek:1;
>  
>  	/** Does the filesystem support posix acls? */
> -	unsigned posix_acl:1;
> +	bool posix_acl:1;
>  
>  	/** Check permissions based on the file mode or not? */
> -	unsigned default_permissions:1;
> +	bool default_permissions:1;
>  
>  	/** Allow other than the mounter user to access the filesystem ? */
> -	unsigned allow_other:1;
> +	bool allow_other:1;
>  
>  	/** Does the filesystem support copy_file_range? */
> -	unsigned no_copy_file_range:1;
> +	bool no_copy_file_range:1;
>  
>  	/* Send DESTROY request */
> -	unsigned int destroy:1;
> +	bool destroy:1;
>  
>  	/* Delete dentries that have gone stale */
> -	unsigned int delete_stale:1;
> +	bool delete_stale:1;
>  
>  	/** Do not create entry in fusectl fs */
> -	unsigned int no_control:1;
> +	bool no_control:1;
>  
>  	/** Do not allow MNT_FORCE umount */
> -	unsigned int no_force_umount:1;
> +	bool no_force_umount:1;
>  
>  	/* Auto-mount submounts announced by the server */
> -	unsigned int auto_submounts:1;
> +	bool auto_submounts:1;
>  
>  	/* Propagate syncfs() to server */
> -	unsigned int sync_fs:1;
> +	bool sync_fs:1;
>  
>  	/* Initialize security xattrs when creating a new inode */
> -	unsigned int init_security:1;
> +	bool init_security:1;
>  
>  	/* Add supplementary group info when creating a new inode */
> -	unsigned int create_supp_group:1;
> +	bool create_supp_group:1;
>  
>  	/* Does the filesystem support per inode DAX? */
> -	unsigned int inode_dax:1;
> +	bool inode_dax:1;
>  
>  	/* Is tmpfile not implemented by fs? */
> -	unsigned int no_tmpfile:1;
> +	bool no_tmpfile:1;
>  
>  	/* Relax restrictions to allow shared mmap in FOPEN_DIRECT_IO mode */
> -	unsigned int direct_io_allow_mmap:1;
> +	bool direct_io_allow_mmap:1;
>  
>  	/* Is statx not implemented by fs? */
> -	unsigned int no_statx:1;
> +	bool no_statx:1;
>  
>  	/** Passthrough support for read/write IO */
> -	unsigned int passthrough:1;
> +	bool passthrough:1;
>  
>  	/* Use pages instead of pointer for kernel I/O */
> -	unsigned int use_pages_for_kvec_io:1;
> +	bool use_pages_for_kvec_io:1;
>  
>  	/* Is link not implemented by fs? */
> -	unsigned int no_link:1;
> +	bool no_link:1;
>  
>  	/* Use io_uring for communication */
> -	unsigned int io_uring:1;
> +	bool io_uring:1;
>  
>  	/** Maximum stack depth for passthrough backing files */
>  	int max_stack_depth;
> diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
> index bfe8d8af46f3..43b6643635ee 100644
> --- a/fs/fuse/inode.c
> +++ b/fs/fuse/inode.c
> @@ -751,7 +751,7 @@ static int fuse_sync_fs(struct super_block *sb, int wait)
>  
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fc->sync_fs = 0;
> +		fc->sync_fs = false;
>  		err = 0;
>  	}
>  
> @@ -973,9 +973,9 @@ void fuse_conn_init(struct fuse_conn *fc, struct fuse_mount *fm,
>  	fc->congestion_threshold = FUSE_DEFAULT_CONGESTION_THRESHOLD;
>  	atomic64_set(&fc->khctr, 0);
>  	fc->polled_files = RB_ROOT;
> -	fc->blocked = 0;
> -	fc->initialized = 0;
> -	fc->connected = 1;
> +	fc->blocked = false;
> +	fc->initialized = false;
> +	fc->connected = true;
>  	atomic64_set(&fc->attr_version, 1);
>  	atomic64_set(&fc->evict_ctr, 1);
>  	get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key));
> @@ -1323,54 +1323,54 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
>  
>  			ra_pages = arg->max_readahead / PAGE_SIZE;
>  			if (flags & FUSE_ASYNC_READ)
> -				fc->async_read = 1;
> +				fc->async_read = true;
>  			if (!(flags & FUSE_POSIX_LOCKS))
> -				fc->no_lock = 1;
> +				fc->no_lock = true;
>  			if (arg->minor >= 17) {
>  				if (!(flags & FUSE_FLOCK_LOCKS))
> -					fc->no_flock = 1;
> +					fc->no_flock = true;
>  			} else {
>  				if (!(flags & FUSE_POSIX_LOCKS))
> -					fc->no_flock = 1;
> +					fc->no_flock = true;
>  			}
>  			if (flags & FUSE_ATOMIC_O_TRUNC)
> -				fc->atomic_o_trunc = 1;
> +				fc->atomic_o_trunc = true;
>  			if (arg->minor >= 9) {
>  				/* LOOKUP has dependency on proto version */
>  				if (flags & FUSE_EXPORT_SUPPORT)
> -					fc->export_support = 1;
> +					fc->export_support = true;
>  			}
>  			if (flags & FUSE_BIG_WRITES)
> -				fc->big_writes = 1;
> +				fc->big_writes = true;
>  			if (flags & FUSE_DONT_MASK)
> -				fc->dont_mask = 1;
> +				fc->dont_mask = true;
>  			if (flags & FUSE_AUTO_INVAL_DATA)
> -				fc->auto_inval_data = 1;
> +				fc->auto_inval_data = true;
>  			else if (flags & FUSE_EXPLICIT_INVAL_DATA)
> -				fc->explicit_inval_data = 1;
> +				fc->explicit_inval_data = true;
>  			if (flags & FUSE_DO_READDIRPLUS) {
> -				fc->do_readdirplus = 1;
> +				fc->do_readdirplus = true;
>  				if (flags & FUSE_READDIRPLUS_AUTO)
> -					fc->readdirplus_auto = 1;
> +					fc->readdirplus_auto = true;
>  			}
>  			if (flags & FUSE_ASYNC_DIO)
> -				fc->async_dio = 1;
> +				fc->async_dio = true;
>  			if (flags & FUSE_WRITEBACK_CACHE)
> -				fc->writeback_cache = 1;
> +				fc->writeback_cache = true;
>  			if (flags & FUSE_PARALLEL_DIROPS)
> -				fc->parallel_dirops = 1;
> +				fc->parallel_dirops = true;
>  			if (flags & FUSE_HANDLE_KILLPRIV)
> -				fc->handle_killpriv = 1;
> +				fc->handle_killpriv = true;
>  			if (arg->time_gran && arg->time_gran <= 1000000000)
>  				fm->sb->s_time_gran = arg->time_gran;
>  			if ((flags & FUSE_POSIX_ACL)) {
> -				fc->default_permissions = 1;
> -				fc->posix_acl = 1;
> +				fc->default_permissions = true;
> +				fc->posix_acl = true;
>  			}
>  			if (flags & FUSE_CACHE_SYMLINKS)
> -				fc->cache_symlinks = 1;
> +				fc->cache_symlinks = true;
>  			if (flags & FUSE_ABORT_ERROR)
> -				fc->abort_err = 1;
> +				fc->abort_err = true;
>  			if (flags & FUSE_MAX_PAGES) {
>  				fc->max_pages =
>  					min_t(unsigned int, fc->max_pages_limit,
> @@ -1389,20 +1389,20 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
>  					ok = false;
>  				}
>  				if (flags & FUSE_HAS_INODE_DAX)
> -					fc->inode_dax = 1;
> +					fc->inode_dax = true;
>  			}
>  			if (flags & FUSE_HANDLE_KILLPRIV_V2) {
> -				fc->handle_killpriv_v2 = 1;
> +				fc->handle_killpriv_v2 = true;
>  				fm->sb->s_flags |= SB_NOSEC;
>  			}
>  			if (flags & FUSE_SETXATTR_EXT)
> -				fc->setxattr_ext = 1;
> +				fc->setxattr_ext = true;
>  			if (flags & FUSE_SECURITY_CTX)
> -				fc->init_security = 1;
> +				fc->init_security = true;
>  			if (flags & FUSE_CREATE_SUPP_GROUP)
> -				fc->create_supp_group = 1;
> +				fc->create_supp_group = true;
>  			if (flags & FUSE_DIRECT_IO_ALLOW_MMAP)
> -				fc->direct_io_allow_mmap = 1;
> +				fc->direct_io_allow_mmap = true;
>  			/*
>  			 * max_stack_depth is the max stack depth of FUSE fs,
>  			 * so it has to be at least 1 to support passthrough
> @@ -1422,7 +1422,7 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
>  			    arg->max_stack_depth > 0 &&
>  			    arg->max_stack_depth <= FILESYSTEM_MAX_STACK_DEPTH &&
>  			    !(flags & FUSE_WRITEBACK_CACHE))  {
> -				fc->passthrough = 1;
> +				fc->passthrough = true;
>  				fc->max_stack_depth = arg->max_stack_depth;
>  				fm->sb->s_stack_depth = arg->max_stack_depth;
>  			}
> @@ -1435,14 +1435,14 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
>  					ok = false;
>  			}
>  			if (flags & FUSE_OVER_IO_URING && fuse_uring_enabled())
> -				fc->io_uring = 1;
> +				fc->io_uring = true;
>  
>  			if (flags & FUSE_REQUEST_TIMEOUT)
>  				timeout = arg->request_timeout;
>  		} else {
>  			ra_pages = fc->max_read / PAGE_SIZE;
> -			fc->no_lock = 1;
> -			fc->no_flock = 1;
> +			fc->no_lock = true;
> +			fc->no_flock = true;
>  		}
>  
>  		init_server_timeout(fc, timeout);
> @@ -1452,13 +1452,13 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
>  		fc->minor = arg->minor;
>  		fc->max_write = arg->minor < 5 ? 4096 : arg->max_write;
>  		fc->max_write = max_t(unsigned, 4096, fc->max_write);
> -		fc->conn_init = 1;
> +		fc->conn_init = true;
>  	}
>  	kfree(ia);
>  
>  	if (!ok) {
> -		fc->conn_init = 0;
> -		fc->conn_error = 1;
> +		fc->conn_init = false;
> +		fc->conn_error = true;
>  	}
>  
>  	fuse_set_initialized(fc);
> @@ -1835,7 +1835,7 @@ int fuse_fill_super_common(struct super_block *sb, struct fuse_fs_context *ctx)
>  
>  	/* Handle umasking inside the fuse code */
>  	if (sb->s_flags & SB_POSIXACL)
> -		fc->dont_mask = 1;
> +		fc->dont_mask = true;
>  	sb->s_flags |= SB_POSIXACL;
>  
>  	fc->default_permissions = ctx->default_permissions;
> diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
> index 2c7b24cb67ad..d7b0f7f288e3 100644
> --- a/fs/fuse/virtio_fs.c
> +++ b/fs/fuse/virtio_fs.c
> @@ -1531,8 +1531,8 @@ static const struct fuse_iqueue_ops virtio_fs_fiq_ops = {
>  static inline void virtio_fs_ctx_set_defaults(struct fuse_fs_context *ctx)
>  {
>  	ctx->rootmode = S_IFDIR;
> -	ctx->default_permissions = 1;
> -	ctx->allow_other = 1;
> +	ctx->default_permissions = true;
> +	ctx->allow_other = true;
>  	ctx->max_read = UINT_MAX;
>  	ctx->blksize = 512;
>  	ctx->destroy = true;
> diff --git a/fs/fuse/xattr.c b/fs/fuse/xattr.c
> index 93dfb06b6cea..73100df1b24b 100644
> --- a/fs/fuse/xattr.c
> +++ b/fs/fuse/xattr.c
> @@ -39,7 +39,7 @@ int fuse_setxattr(struct inode *inode, const char *name, const void *value,
>  	args.in_args[2].value = value;
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fm->fc->no_setxattr = 1;
> +		fm->fc->no_setxattr = true;
>  		err = -EOPNOTSUPP;
>  	}
>  	if (!err)
> @@ -83,7 +83,7 @@ ssize_t fuse_getxattr(struct inode *inode, const char *name, void *value,
>  	if (!ret && !size)
>  		ret = min_t(size_t, outarg.size, XATTR_SIZE_MAX);
>  	if (ret == -ENOSYS) {
> -		fm->fc->no_getxattr = 1;
> +		fm->fc->no_getxattr = true;
>  		ret = -EOPNOTSUPP;
>  	}
>  	return ret;
> @@ -147,7 +147,7 @@ ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size)
>  	if (ret > 0 && size)
>  		ret = fuse_verify_xattr_list(list, ret);
>  	if (ret == -ENOSYS) {
> -		fm->fc->no_listxattr = 1;
> +		fm->fc->no_listxattr = true;
>  		ret = -EOPNOTSUPP;
>  	}
>  	return ret;
> @@ -170,7 +170,7 @@ int fuse_removexattr(struct inode *inode, const char *name)
>  	args.in_args[1].value = name;
>  	err = fuse_simple_request(fm, &args);
>  	if (err == -ENOSYS) {
> -		fm->fc->no_removexattr = 1;
> +		fm->fc->no_removexattr = true;
>  		err = -EOPNOTSUPP;
>  	}
>  	if (!err)

Re: [PATCH 1/2] fuse: optimize struct fuse_conn fields

[Jingbo Xu]

On 4/19/25 5:06 AM, Joanne Koong wrote:
> Use a bitfield for tracking initialized, blocked, aborted, and io_uring
> state of the fuse connection. Track connected state using a bool instead
> of an unsigned.
> 
> On a 64-bit system, this shaves off 16 bytes from the size of struct
> fuse_conn.
> 
> No functional changes.
> 
> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
> ---
>  fs/fuse/fuse_i.h | 24 ++++++++++++------------
>  1 file changed, 12 insertions(+), 12 deletions(-)
> 
> diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
> index b54f4f57789f..6aecada8aadd 100644
> --- a/fs/fuse/fuse_i.h
> +++ b/fs/fuse/fuse_i.h
> @@ -690,24 +690,24 @@ struct fuse_conn {
>  	 * 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;
> +	bool connected;

Why not also convert connected to bitfield?


-- 
Thanks,
Jingbo

Re: [PATCH 1/2] fuse: optimize struct fuse_conn fields

[Joanne Koong]

On Mon, Apr 21, 2025 at 7:07 PM Jingbo Xu <jefflexu@linux.alibaba.com> wrote:
>
>
>
> On 4/19/25 5:06 AM, Joanne Koong wrote:
> > Use a bitfield for tracking initialized, blocked, aborted, and io_uring
> > state of the fuse connection. Track connected state using a bool instead
> > of an unsigned.
> >
> > On a 64-bit system, this shaves off 16 bytes from the size of struct
> > fuse_conn.
> >
> > No functional changes.
> >
> > Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
> > ---
> >  fs/fuse/fuse_i.h | 24 ++++++++++++------------
> >  1 file changed, 12 insertions(+), 12 deletions(-)
> >
> > diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
> > index b54f4f57789f..6aecada8aadd 100644
> > --- a/fs/fuse/fuse_i.h
> > +++ b/fs/fuse/fuse_i.h
> > @@ -690,24 +690,24 @@ struct fuse_conn {
> >        * 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;
> > +     bool connected;
>
> Why not also convert connected to bitfield?

fuse_drop_waiting() checks the connected state locklessly through
READ_ONCE(fc->connected). The smallest size READ_ONCE supports is a
byte, I don't think it works on bitfields.


Thanks,
Joanne
>
>
> --
> Thanks,
> Jingbo

Re: [PATCH 1/2] fuse: optimize struct fuse_conn fields

[Jingbo Xu]

On 4/23/25 12:25 AM, Joanne Koong wrote:
> On Mon, Apr 21, 2025 at 7:07 PM Jingbo Xu <jefflexu@linux.alibaba.com> wrote:
>>
>>
>>
>> On 4/19/25 5:06 AM, Joanne Koong wrote:
>>> Use a bitfield for tracking initialized, blocked, aborted, and io_uring
>>> state of the fuse connection. Track connected state using a bool instead
>>> of an unsigned.
>>>
>>> On a 64-bit system, this shaves off 16 bytes from the size of struct
>>> fuse_conn.
>>>
>>> No functional changes.
>>>
>>> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
>>> ---
>>>  fs/fuse/fuse_i.h | 24 ++++++++++++------------
>>>  1 file changed, 12 insertions(+), 12 deletions(-)
>>>
>>> diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
>>> index b54f4f57789f..6aecada8aadd 100644
>>> --- a/fs/fuse/fuse_i.h
>>> +++ b/fs/fuse/fuse_i.h
>>> @@ -690,24 +690,24 @@ struct fuse_conn {
>>>        * 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;
>>> +     bool connected;
>>
>> Why not also convert connected to bitfield?
> 
> fuse_drop_waiting() checks the connected state locklessly through
> READ_ONCE(fc->connected). The smallest size READ_ONCE supports is a
> byte, I don't think it works on bitfields.

Okay, that makes sense.  Thanks.

-- 
Thanks,
Jingbo

Re: [PATCH 1/2] fuse: optimize struct fuse_conn fields

[Miklos Szeredi]

On Fri, 18 Apr 2025 at 23:06, Joanne Koong <joannelkoong@gmail.com> wrote:
>
> Use a bitfield for tracking initialized, blocked, aborted, and io_uring
> state of the fuse connection. Track connected state using a bool instead
> of an unsigned.
>
> On a 64-bit system, this shaves off 16 bytes from the size of struct
> fuse_conn.
>
> No functional changes.

Not sure about that.

AFAIK aligned int or long is supposed to be independent from
neighboring fields on all architectures.  But that's definitely not
true of bitfields and I'm not sure about bool.  Maybe
READ_ONCE()/WRITE_ONCE() make accessing bool safe, but I haven't found
any documentation about that.

Previous rule about bitfields in fuse_conn have been that they are either

 - only set at INIT reply time, or
 - losing a setting due to a race is a non-issue

The new ones are not so clear cut, so it definitely needs some more
explanation why they are safe (if they are safe).

Thanks,
Miklos

Re: [PATCH 1/2] fuse: optimize struct fuse_conn fields

[Joanne Koong]

On Tue, May 6, 2025 at 4:09 AM Miklos Szeredi <miklos@szeredi.hu> wrote:
>
> On Fri, 18 Apr 2025 at 23:06, Joanne Koong <joannelkoong@gmail.com> wrote:
> >
> > Use a bitfield for tracking initialized, blocked, aborted, and io_uring
> > state of the fuse connection. Track connected state using a bool instead
> > of an unsigned.
> >
> > On a 64-bit system, this shaves off 16 bytes from the size of struct
> > fuse_conn.
> >
> > No functional changes.
>
> Not sure about that.
>
> AFAIK aligned int or long is supposed to be independent from
> neighboring fields on all architectures.  But that's definitely not
> true of bitfields and I'm not sure about bool.  Maybe
> READ_ONCE()/WRITE_ONCE() make accessing bool safe, but I haven't found
> any documentation about that.

Ohh interesting. That's a great point about race conditions from
modifying values that might be packed together. I will drop this patch
then.

Thanks,
Joanne
>
> Previous rule about bitfields in fuse_conn have been that they are either
>
>  - only set at INIT reply time, or
>  - losing a setting due to a race is a non-issue
>
> The new ones are not so clear cut, so it definitely needs some more
> explanation why they are safe (if they are safe).
>
> Thanks,
> Miklos