fuse update for 6.3

-----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQSQHSd0lITzzeNWNm3h3BK/laaZPAUCY/ysqAAKCRDh3BK/laaZ
 PLSfAP9c4z/KOZj/Am8nQ0mqI8Ss0Ei+hyu8Vow6NoyJkR4NvwD/SxEeI2+rpXj7
 TGOA+6k4chLc7QIFBsIwGvQgeXld1gA=
 =s4b/
 -----END PGP SIGNATURE-----

Merge tag 'fuse-update-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse

Pull fuse updates from Miklos Szeredi:

 - Fix regression in fileattr permission checking

 - Fix possible hang during PID namespace destruction

 - Add generic support for request extensions

 - Add supplementary group list extension

 - Add limited support for supplying supplementary groups in create
   requests

 - Documentation fixes

* tag 'fuse-update-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
  fuse: add inode/permission checks to fileattr_get/fileattr_set
  fuse: fix all W=1 kernel-doc warnings
  fuse: in fuse_flush only wait if someone wants the return code
  fuse: optional supplementary group in create requests
  fuse: add request extension

fs/fuse/cuse.c

@@ -256,7 +256,7 @@ static int cuse_parse_one(char **pp, char *end, char **keyp, char **valp)
 }
 
 /**
- * cuse_parse_dev_info - parse device info
+ * cuse_parse_devinfo - parse device info
  * @p: device info string
  * @len: length of device info string
  * @devinfo: out parameter for parsed device info

fs/fuse/dev.c

@@ -204,7 +204,7 @@ static unsigned int fuse_req_hash(u64 unique)
 	return hash_long(unique & ~FUSE_INT_REQ_BIT, FUSE_PQ_HASH_BITS);
 }
 
-/**
+/*
  * A new request is available, wake fiq->waitq
  */
 static void fuse_dev_wake_and_unlock(struct fuse_iqueue *fiq)
@@ -476,6 +476,8 @@ static void fuse_args_to_req(struct fuse_req *req, struct fuse_args *args)
 	req->in.h.opcode = args->opcode;
 	req->in.h.nodeid = args->nodeid;
 	req->args = args;
+	if (args->is_ext)
+		req->in.h.total_extlen = args->in_args[args->ext_idx].size / 8;
 	if (args->end)
 		__set_bit(FR_ASYNC, &req->flags);
 }

fs/fuse/dir.c

@@ -145,7 +145,7 @@ static void fuse_dir_changed(struct inode *dir)
 	inode_maybe_inc_iversion(dir, false);
 }
 
-/**
+/*
  * Mark the attributes as stale due to an atime change.  Avoid the invalidate if
  * atime is not used.
  */
@@ -466,7 +466,7 @@ static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry,
 }
 
 static int get_security_context(struct dentry *entry, umode_t mode,
-				void **security_ctx, u32 *security_ctxlen)
+				struct fuse_in_arg *ext)
 {
 	struct fuse_secctx *fctx;
 	struct fuse_secctx_header *header;
@@ -513,14 +513,100 @@ static int get_security_context(struct dentry *entry, umode_t mode,
 
 		memcpy(ptr, ctx, ctxlen);
 	}
-	*security_ctxlen = total_len;
-	*security_ctx = header;
+	ext->size = total_len;
+	ext->value = header;
 	err = 0;
 out_err:
 	kfree(ctx);
 	return err;
 }
 
+static void *extend_arg(struct fuse_in_arg *buf, u32 bytes)
+{
+	void *p;
+	u32 newlen = buf->size + bytes;
+
+	p = krealloc(buf->value, newlen, GFP_KERNEL);
+	if (!p) {
+		kfree(buf->value);
+		buf->size = 0;
+		buf->value = NULL;
+		return NULL;
+	}
+
+	memset(p + buf->size, 0, bytes);
+	buf->value = p;
+	buf->size = newlen;
+
+	return p + newlen - bytes;
+}
+
+static u32 fuse_ext_size(size_t size)
+{
+	return FUSE_REC_ALIGN(sizeof(struct fuse_ext_header) + size);
+}
+
+/*
+ * This adds just a single supplementary group that matches the parent's group.
+ */
+static int get_create_supp_group(struct inode *dir, struct fuse_in_arg *ext)
+{
+	struct fuse_conn *fc = get_fuse_conn(dir);
+	struct fuse_ext_header *xh;
+	struct fuse_supp_groups *sg;
+	kgid_t kgid = dir->i_gid;
+	gid_t parent_gid = from_kgid(fc->user_ns, kgid);
+	u32 sg_len = fuse_ext_size(sizeof(*sg) + sizeof(sg->groups[0]));
+
+	if (parent_gid == (gid_t) -1 || gid_eq(kgid, current_fsgid()) ||
+	    !in_group_p(kgid))
+		return 0;
+
+	xh = extend_arg(ext, sg_len);
+	if (!xh)
+		return -ENOMEM;
+
+	xh->size = sg_len;
+	xh->type = FUSE_EXT_GROUPS;
+
+	sg = (struct fuse_supp_groups *) &xh[1];
+	sg->nr_groups = 1;
+	sg->groups[0] = parent_gid;
+
+	return 0;
+}
+
+static int get_create_ext(struct fuse_args *args,
+			  struct inode *dir, struct dentry *dentry,
+			  umode_t mode)
+{
+	struct fuse_conn *fc = get_fuse_conn_super(dentry->d_sb);
+	struct fuse_in_arg ext = { .size = 0, .value = NULL };
+	int err = 0;
+
+	if (fc->init_security)
+		err = get_security_context(dentry, mode, &ext);
+	if (!err && fc->create_supp_group)
+		err = get_create_supp_group(dir, &ext);
+
+	if (!err && ext.size) {
+		WARN_ON(args->in_numargs >= ARRAY_SIZE(args->in_args));
+		args->is_ext = true;
+		args->ext_idx = args->in_numargs++;
+		args->in_args[args->ext_idx] = ext;
+	} else {
+		kfree(ext.value);
+	}
+
+	return err;
+}
+
+static void free_ext_value(struct fuse_args *args)
+{
+	if (args->is_ext)
+		kfree(args->in_args[args->ext_idx].value);
+}
+
 /*
  * Atomic create+open operation
  *
@@ -541,8 +627,6 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry,
 	struct fuse_entry_out outentry;
 	struct fuse_inode *fi;
 	struct fuse_file *ff;
-	void *security_ctx = NULL;
-	u32 security_ctxlen;
 	bool trunc = flags & O_TRUNC;
 
 	/* Userspace expects S_IFREG in create mode */
@@ -586,19 +670,12 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry,
 	args.out_args[1].size = sizeof(outopen);
 	args.out_args[1].value = &outopen;
 
-	if (fm->fc->init_security) {
-		err = get_security_context(entry, mode, &security_ctx,
-					   &security_ctxlen);
-		if (err)
-			goto out_put_forget_req;
-
-		args.in_numargs = 3;
-		args.in_args[2].size = security_ctxlen;
-		args.in_args[2].value = security_ctx;
-	}
+	err = get_create_ext(&args, dir, entry, mode);
+	if (err)
+		goto out_put_forget_req;
 
 	err = fuse_simple_request(fm, &args);
-	kfree(security_ctx);
+	free_ext_value(&args);
 	if (err)
 		goto out_free_ff;
 
@@ -705,8 +782,6 @@ static int create_new_entry(struct fuse_mount *fm, struct fuse_args *args,
 	struct dentry *d;
 	int err;
 	struct fuse_forget_link *forget;
-	void *security_ctx = NULL;
-	u32 security_ctxlen;
 
 	if (fuse_is_bad(dir))
 		return -EIO;
@@ -721,21 +796,14 @@ static int create_new_entry(struct fuse_mount *fm, struct fuse_args *args,
 	args->out_args[0].size = sizeof(outarg);
 	args->out_args[0].value = &outarg;
 
-	if (fm->fc->init_security && args->opcode != FUSE_LINK) {
-		err = get_security_context(entry, mode, &security_ctx,
-					   &security_ctxlen);
+	if (args->opcode != FUSE_LINK) {
+		err = get_create_ext(args, dir, entry, mode);
 		if (err)
 			goto out_put_forget_req;
-
-		BUG_ON(args->in_numargs != 2);
-
-		args->in_numargs = 3;
-		args->in_args[2].size = security_ctxlen;
-		args->in_args[2].value = security_ctx;
 	}
 
 	err = fuse_simple_request(fm, args);
-	kfree(security_ctx);
+	free_ext_value(args);
 	if (err)
 		goto out_put_forget_req;
 

fs/fuse/file.c

@@ -19,6 +19,7 @@
 #include <linux/uio.h>
 #include <linux/fs.h>
 #include <linux/filelock.h>
+#include <linux/file.h>
 
 static int fuse_send_open(struct fuse_mount *fm, u64 nodeid,
 			  unsigned int open_flags, int opcode,
@@ -478,48 +479,36 @@ static void fuse_sync_writes(struct inode *inode)
 	fuse_release_nowrite(inode);
 }
 
-static int fuse_flush(struct file *file, fl_owner_t id)
-{
-	struct inode *inode = file_inode(file);
-	struct fuse_mount *fm = get_fuse_mount(inode);
-	struct fuse_file *ff = file->private_data;
+struct fuse_flush_args {
+	struct fuse_args args;
 	struct fuse_flush_in inarg;
-	FUSE_ARGS(args);
+	struct work_struct work;
+	struct file *file;
+};
+
+static int fuse_do_flush(struct fuse_flush_args *fa)
+{
 	int err;
-
-	if (fuse_is_bad(inode))
-		return -EIO;
-
-	if (ff->open_flags & FOPEN_NOFLUSH && !fm->fc->writeback_cache)
-		return 0;
+	struct inode *inode = file_inode(fa->file);
+	struct fuse_mount *fm = get_fuse_mount(inode);
 
 	err = write_inode_now(inode, 1);
 	if (err)
-		return err;
+		goto out;
 
 	inode_lock(inode);
 	fuse_sync_writes(inode);
 	inode_unlock(inode);
 
-	err = filemap_check_errors(file->f_mapping);
+	err = filemap_check_errors(fa->file->f_mapping);
 	if (err)
-		return err;
+		goto out;
 
 	err = 0;
 	if (fm->fc->no_flush)
 		goto inval_attr_out;
 
-	memset(&inarg, 0, sizeof(inarg));
-	inarg.fh = ff->fh;
-	inarg.lock_owner = fuse_lock_owner_id(fm->fc, id);
-	args.opcode = FUSE_FLUSH;
-	args.nodeid = get_node_id(inode);
-	args.in_numargs = 1;
-	args.in_args[0].size = sizeof(inarg);
-	args.in_args[0].value = &inarg;
-	args.force = true;
-
-	err = fuse_simple_request(fm, &args);
+	err = fuse_simple_request(fm, &fa->args);
 	if (err == -ENOSYS) {
 		fm->fc->no_flush = 1;
 		err = 0;
@@ -532,9 +521,57 @@ static int fuse_flush(struct file *file, fl_owner_t id)
 	 */
 	if (!err && fm->fc->writeback_cache)
 		fuse_invalidate_attr_mask(inode, STATX_BLOCKS);
+
+out:
+	fput(fa->file);
+	kfree(fa);
 	return err;
 }
 
+static void fuse_flush_async(struct work_struct *work)
+{
+	struct fuse_flush_args *fa = container_of(work, typeof(*fa), work);
+
+	fuse_do_flush(fa);
+}
+
+static int fuse_flush(struct file *file, fl_owner_t id)
+{
+	struct fuse_flush_args *fa;
+	struct inode *inode = file_inode(file);
+	struct fuse_mount *fm = get_fuse_mount(inode);
+	struct fuse_file *ff = file->private_data;
+
+	if (fuse_is_bad(inode))
+		return -EIO;
+
+	if (ff->open_flags & FOPEN_NOFLUSH && !fm->fc->writeback_cache)
+		return 0;
+
+	fa = kzalloc(sizeof(*fa), GFP_KERNEL);
+	if (!fa)
+		return -ENOMEM;
+
+	fa->inarg.fh = ff->fh;
+	fa->inarg.lock_owner = fuse_lock_owner_id(fm->fc, id);
+	fa->args.opcode = FUSE_FLUSH;
+	fa->args.nodeid = get_node_id(inode);
+	fa->args.in_numargs = 1;
+	fa->args.in_args[0].size = sizeof(fa->inarg);
+	fa->args.in_args[0].value = &fa->inarg;
+	fa->args.force = true;
+	fa->file = get_file(file);
+
+	/* Don't wait if the task is exiting */
+	if (current->flags & PF_EXITING) {
+		INIT_WORK(&fa->work, fuse_flush_async);
+		schedule_work(&fa->work);
+		return 0;
+	}
+
+	return fuse_do_flush(fa);
+}
+
 int fuse_fsync_common(struct file *file, loff_t start, loff_t end,
 		      int datasync, int opcode)
 {
@@ -653,7 +690,7 @@ static ssize_t fuse_get_res_by_io(struct fuse_io_priv *io)
 	return io->bytes < 0 ? io->size : io->bytes;
 }
 
-/**
+/*
  * In case of short read, the caller sets 'pos' to the position of
  * actual end of fuse request in IO request. Otherwise, if bytes_requested
  * == bytes_transferred or rw == WRITE, the caller sets 'pos' to -1.

fs/fuse/fuse_i.h

@@ -249,8 +249,9 @@ struct fuse_page_desc {
 struct fuse_args {
 	uint64_t nodeid;
 	uint32_t opcode;
-	unsigned short in_numargs;
-	unsigned short out_numargs;
+	uint8_t in_numargs;
+	uint8_t out_numargs;
+	uint8_t ext_idx;
 	bool force:1;
 	bool noreply:1;
 	bool nocreds:1;
@@ -261,6 +262,7 @@ struct fuse_args {
 	bool page_zeroing:1;
 	bool page_replace:1;
 	bool may_block:1;
+	bool is_ext:1;
 	struct fuse_in_arg in_args[3];
 	struct fuse_arg out_args[2];
 	void (*end)(struct fuse_mount *fm, struct fuse_args *args, int error);
@@ -781,6 +783,9 @@ struct fuse_conn {
 	/* Initialize security xattrs when creating a new inode */
 	unsigned int init_security:1;
 
+	/* Add supplementary group info when creating a new inode */
+	unsigned int create_supp_group:1;
+
 	/* Does the filesystem support per inode DAX? */
 	unsigned int inode_dax:1;
 

fs/fuse/inode.c

@@ -1207,6 +1207,8 @@ static void process_init_reply(struct fuse_mount *fm, struct fuse_args *args,
 				fc->setxattr_ext = 1;
 			if (flags & FUSE_SECURITY_CTX)
 				fc->init_security = 1;
+			if (flags & FUSE_CREATE_SUPP_GROUP)
+				fc->create_supp_group = 1;
 		} else {
 			ra_pages = fc->max_read / PAGE_SIZE;
 			fc->no_lock = 1;
@@ -1252,7 +1254,7 @@ void fuse_send_init(struct fuse_mount *fm)
 		FUSE_ABORT_ERROR | FUSE_MAX_PAGES | FUSE_CACHE_SYMLINKS |
 		FUSE_NO_OPENDIR_SUPPORT | FUSE_EXPLICIT_INVAL_DATA |
 		FUSE_HANDLE_KILLPRIV_V2 | FUSE_SETXATTR_EXT | FUSE_INIT_EXT |
-		FUSE_SECURITY_CTX;
+		FUSE_SECURITY_CTX | FUSE_CREATE_SUPP_GROUP;
 #ifdef CONFIG_FUSE_DAX
 	if (fm->fc->dax)
 		flags |= FUSE_MAP_ALIGNMENT;

fs/fuse/ioctl.c

@@ -419,6 +419,12 @@ 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);
 

include/uapi/linux/fuse.h

@@ -201,6 +201,11 @@
  *  7.38
  *  - add FUSE_EXPIRE_ONLY flag to fuse_notify_inval_entry
  *  - add FOPEN_PARALLEL_DIRECT_WRITES
+ *  - add total_extlen to fuse_in_header
+ *  - add FUSE_MAX_NR_SECCTX
+ *  - add extension header
+ *  - add FUSE_EXT_GROUPS
+ *  - add FUSE_CREATE_SUPP_GROUP
  */
 
 #ifndef _LINUX_FUSE_H
@@ -362,6 +367,8 @@ struct fuse_file_lock {
  * FUSE_SECURITY_CTX:	add security context to create, mkdir, symlink, and
  *			mknod
  * FUSE_HAS_INODE_DAX:  use per inode DAX
+ * FUSE_CREATE_SUPP_GROUP: add supplementary group info to create, mkdir,
+ *			symlink and mknod (single group that matches parent)
  */
 #define FUSE_ASYNC_READ		(1 << 0)
 #define FUSE_POSIX_LOCKS	(1 << 1)
@@ -398,6 +405,7 @@ struct fuse_file_lock {
 /* bits 32..63 get shifted down 32 bits into the flags2 field */
 #define FUSE_SECURITY_CTX	(1ULL << 32)
 #define FUSE_HAS_INODE_DAX	(1ULL << 33)
+#define FUSE_CREATE_SUPP_GROUP	(1ULL << 34)
 
 /**
  * CUSE INIT request/reply flags
@@ -503,6 +511,17 @@ struct fuse_file_lock {
  */
 #define FUSE_EXPIRE_ONLY		(1 << 0)
 
+/**
+ * extension type
+ * FUSE_MAX_NR_SECCTX: maximum value of &fuse_secctx_header.nr_secctx
+ * FUSE_EXT_GROUPS: &fuse_supp_groups extension
+ */
+enum fuse_ext_type {
+	/* Types 0..31 are reserved for fuse_secctx_header */
+	FUSE_MAX_NR_SECCTX	= 31,
+	FUSE_EXT_GROUPS		= 32,
+};
+
 enum fuse_opcode {
 	FUSE_LOOKUP		= 1,
 	FUSE_FORGET		= 2,  /* no reply */
@@ -886,7 +905,8 @@ struct fuse_in_header {
 	uint32_t	uid;
 	uint32_t	gid;
 	uint32_t	pid;
-	uint32_t	padding;
+	uint16_t	total_extlen; /* length of extensions in 8byte units */
+	uint16_t	padding;
 };
 
 struct fuse_out_header {
@@ -1047,4 +1067,27 @@ struct fuse_secctx_header {
 	uint32_t	nr_secctx;
 };
 
+/**
+ * struct fuse_ext_header - extension header
+ * @size: total size of this extension including this header
+ * @type: type of extension
+ *
+ * This is made compatible with fuse_secctx_header by using type values >
+ * FUSE_MAX_NR_SECCTX
+ */
+struct fuse_ext_header {
+	uint32_t	size;
+	uint32_t	type;
+};
+
+/**
+ * struct fuse_supp_groups - Supplementary group extension
+ * @nr_groups: number of supplementary groups
+ * @groups: flexible array of group IDs
+ */
+struct fuse_supp_groups {
+	uint32_t	nr_groups;
+	uint32_t	groups[];
+};
+
 #endif /* _LINUX_FUSE_H */