22 files changed, 145 insertions, 60 deletions

diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 4a2b39d9a61a..bdcffb04513f 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -2019,9 +2019,10 @@ unsigned ceph_dentry_hash(struct inode *dir, struct dentry *dn)
 	}
 }
 
+WRAP_DIR_ITER(ceph_readdir) // FIXME!
 const struct file_operations ceph_dir_fops = {
 	.read = ceph_read_dir,
-	.iterate = ceph_readdir,
+	.iterate_shared = shared_ceph_readdir,
 	.llseek = ceph_dir_llseek,
 	.open = ceph_open,
 	.release = ceph_release,
@@ -2033,7 +2034,7 @@ const struct file_operations ceph_dir_fops = {
 };
 
 const struct file_operations ceph_snapdir_fops = {
-	.iterate = ceph_readdir,
+	.iterate_shared = shared_ceph_readdir,
 	.llseek = ceph_dir_llseek,
 	.open = ceph_open,
 	.release = ceph_release,
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 66048a86c480..5fb367b1d4b0 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -4764,7 +4764,7 @@ static void delayed_work(struct work_struct *work)
 
 	dout("mdsc delayed_work\n");
 
-	if (mdsc->stopping)
+	if (mdsc->stopping >= CEPH_MDSC_STOPPING_FLUSHED)
 		return;
 
 	mutex_lock(&mdsc->mutex);
@@ -4943,7 +4943,7 @@ void send_flush_mdlog(struct ceph_mds_session *s)
 void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
 {
 	dout("pre_umount\n");
-	mdsc->stopping = 1;
+	mdsc->stopping = CEPH_MDSC_STOPPING_BEGIN;
 
 	ceph_mdsc_iterate_sessions(mdsc, send_flush_mdlog, true);
 	ceph_mdsc_iterate_sessions(mdsc, lock_unlock_session, false);
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index 724307ff89cd..86d2965e68a1 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -380,6 +380,11 @@ struct cap_wait {
 	int			want;
 };
 
+enum {
+       CEPH_MDSC_STOPPING_BEGIN = 1,
+       CEPH_MDSC_STOPPING_FLUSHED = 2,
+};
+
 /*
  * mds client state
  */
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index 3fc48b43cab0..a5f52013314d 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -1374,6 +1374,16 @@ static void ceph_kill_sb(struct super_block *s)
 	ceph_mdsc_pre_umount(fsc->mdsc);
 	flush_fs_workqueues(fsc);
 
+	/*
+	 * Though the kill_anon_super() will finally trigger the
+	 * sync_filesystem() anyway, we still need to do it here
+	 * and then bump the stage of shutdown to stop the work
+	 * queue as earlier as possible.
+	 */
+	sync_filesystem(s);
+
+	fsc->mdsc->stopping = CEPH_MDSC_STOPPING_FLUSHED;
+
 	kill_anon_super(s);
 
 	fsc->client->extra_mon_dispatch = NULL;
diff --git a/fs/coda/dir.c b/fs/coda/dir.c
index 8450b1bd354b..1b960de2bf39 100644
--- a/fs/coda/dir.c
+++ b/fs/coda/dir.c
@@ -429,21 +429,14 @@ static int coda_readdir(struct file *coda_file, struct dir_context *ctx)
 	cfi = coda_ftoc(coda_file);
 	host_file = cfi->cfi_container;
 
-	if (host_file->f_op->iterate || host_file->f_op->iterate_shared) {
+	if (host_file->f_op->iterate_shared) {
 		struct inode *host_inode = file_inode(host_file);
 		ret = -ENOENT;
 		if (!IS_DEADDIR(host_inode)) {
-			if (host_file->f_op->iterate_shared) {
-				inode_lock_shared(host_inode);
-				ret = host_file->f_op->iterate_shared(host_file, ctx);
-				file_accessed(host_file);
-				inode_unlock_shared(host_inode);
-			} else {
-				inode_lock(host_inode);
-				ret = host_file->f_op->iterate(host_file, ctx);
-				file_accessed(host_file);
-				inode_unlock(host_inode);
-			}
+			inode_lock_shared(host_inode);
+			ret = host_file->f_op->iterate_shared(host_file, ctx);
+			file_accessed(host_file);
+			inode_unlock_shared(host_inode);
 		}
 		return ret;
 	}
@@ -585,10 +578,11 @@ const struct inode_operations coda_dir_inode_operations = {
 	.setattr	= coda_setattr,
 };
 
+WRAP_DIR_ITER(coda_readdir) // FIXME!
 const struct file_operations coda_dir_operations = {
 	.llseek		= generic_file_llseek,
 	.read		= generic_read_dir,
-	.iterate	= coda_readdir,
+	.iterate_shared	= shared_coda_readdir,
 	.open		= coda_open,
 	.release	= coda_release,
 	.fsync		= coda_fsync,
diff --git a/fs/exfat/dir.c b/fs/exfat/dir.c
index 598081d0d059..e1586bba6d86 100644
--- a/fs/exfat/dir.c
+++ b/fs/exfat/dir.c
@@ -306,10 +306,11 @@ out:
 	return err;
 }
 
+WRAP_DIR_ITER(exfat_iterate) // FIXME!
 const struct file_operations exfat_dir_operations = {
 	.llseek		= generic_file_llseek,
 	.read		= generic_read_dir,
-	.iterate	= exfat_iterate,
+	.iterate_shared	= shared_exfat_iterate,
 	.unlocked_ioctl = exfat_ioctl,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = exfat_compat_ioctl,
diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c
index 40e624cf7e92..d1dbe47c7975 100644
--- a/fs/exportfs/expfs.c
+++ b/fs/exportfs/expfs.c
@@ -315,7 +315,7 @@ static int get_name(const struct path *path, char *name, struct dentry *child)
 		goto out;
 
 	error = -EINVAL;
-	if (!file->f_op->iterate && !file->f_op->iterate_shared)
+	if (!file->f_op->iterate_shared)
 		goto out_close;
 
 	buffer.sequence = 0;
diff --git a/fs/file.c b/fs/file.c
index 35c62b54c9d6..3fd003a8604f 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -1036,12 +1036,28 @@ unsigned long __fdget_raw(unsigned int fd)
 	return __fget_light(fd, 0);
 }
 
+/*
+ * Try to avoid f_pos locking. We only need it if the
+ * file is marked for FMODE_ATOMIC_POS, and it can be
+ * accessed multiple ways.
+ *
+ * Always do it for directories, because pidfd_getfd()
+ * can make a file accessible even if it otherwise would
+ * not be, and for directories this is a correctness
+ * issue, not a "POSIX requirement".
+ */
+static inline bool file_needs_f_pos_lock(struct file *file)
+{
+	return (file->f_mode & FMODE_ATOMIC_POS) &&
+		(file_count(file) > 1 || file->f_op->iterate_shared);
+}
+
 unsigned long __fdget_pos(unsigned int fd)
 {
 	unsigned long v = __fdget(fd);
 	struct file *file = (struct file *)(v & ~3);
 
-	if (file && (file->f_mode & FMODE_ATOMIC_POS)) {
+	if (file && file_needs_f_pos_lock(file)) {
 		v |= FDPUT_POS_UNLOCK;
 		mutex_lock(&file->f_pos_lock);
 	}
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 1bf3c4453516..b43fa8b8fc05 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -1578,7 +1578,7 @@ const struct file_operations gfs2_file_fops = {
 	.fsync		= gfs2_fsync,
 	.lock		= gfs2_lock,
 	.flock		= gfs2_flock,
-	.splice_read	= filemap_splice_read,
+	.splice_read	= copy_splice_read,
 	.splice_write	= gfs2_file_splice_write,
 	.setlease	= simple_nosetlease,
 	.fallocate	= gfs2_fallocate,
@@ -1609,7 +1609,7 @@ const struct file_operations gfs2_file_fops_nolock = {
 	.open		= gfs2_open,
 	.release	= gfs2_release,
 	.fsync		= gfs2_fsync,
-	.splice_read	= filemap_splice_read,
+	.splice_read	= copy_splice_read,
 	.splice_write	= gfs2_file_splice_write,
 	.setlease	= generic_setlease,
 	.fallocate	= gfs2_fallocate,
diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index ec1631257978..7e835be7032d 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -230,9 +230,11 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 {
 
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
+	struct super_block *sb = sdp->sd_vfs;
 	struct gfs2_bufdata *bd;
 	struct gfs2_meta_header *mh;
 	struct gfs2_trans *tr = current->journal_info;
+	bool withdraw = false;
 
 	lock_buffer(bh);
 	if (buffer_pinned(bh)) {
@@ -266,13 +268,15 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 		       (unsigned long long)bd->bd_bh->b_blocknr);
 		BUG();
 	}
-	if (unlikely(test_bit(SDF_FROZEN, &sdp->sd_flags))) {
-		fs_info(sdp, "GFS2:adding buf while frozen\n");
-		gfs2_assert_withdraw(sdp, 0);
-	}
 	if (unlikely(gfs2_withdrawn(sdp))) {
 		fs_info(sdp, "GFS2:adding buf while withdrawn! 0x%llx\n",
 			(unsigned long long)bd->bd_bh->b_blocknr);
+		goto out_unlock;
+	}
+	if (unlikely(sb->s_writers.frozen == SB_FREEZE_COMPLETE)) {
+		fs_info(sdp, "GFS2:adding buf while frozen\n");
+		withdraw = true;
+		goto out_unlock;
 	}
 	gfs2_pin(sdp, bd->bd_bh);
 	mh->__pad0 = cpu_to_be64(0);
@@ -281,6 +285,8 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 	tr->tr_num_buf_new++;
 out_unlock:
 	gfs2_log_unlock(sdp);
+	if (withdraw)
+		gfs2_assert_withdraw(sdp, 0);
 out:
 	unlock_buffer(bh);
 }
diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index 9b030297aa64..e98ddb2b1cf2 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -1535,9 +1535,10 @@ const struct inode_operations jfs_dir_inode_operations = {
 #endif
 };
 
+WRAP_DIR_ITER(jfs_readdir) // FIXME!
 const struct file_operations jfs_dir_operations = {
 	.read		= generic_read_dir,
-	.iterate	= jfs_readdir,
+	.iterate_shared	= shared_jfs_readdir,
 	.fsync		= jfs_fsync,
 	.unlocked_ioctl = jfs_ioctl,
 	.compat_ioctl	= compat_ptr_ioctl,
diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c
index 518c3a21a556..4596c90e7b7c 100644
--- a/fs/ntfs/dir.c
+++ b/fs/ntfs/dir.c
@@ -1525,10 +1525,11 @@ static int ntfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
 
 #endif /* NTFS_RW */
 
+WRAP_DIR_ITER(ntfs_readdir) // FIXME!
 const struct file_operations ntfs_dir_ops = {
 	.llseek		= generic_file_llseek,	/* Seek inside directory. */
 	.read		= generic_read_dir,	/* Return -EISDIR. */
-	.iterate	= ntfs_readdir,		/* Read directory contents. */
+	.iterate_shared	= shared_ntfs_readdir,	/* Read directory contents. */
 #ifdef NTFS_RW
 	.fsync		= ntfs_dir_fsync,	/* Sync a directory to disk. */
 #endif /* NTFS_RW */
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 91a194596552..bf2c17ea96a0 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2793,10 +2793,11 @@ const struct file_operations ocfs2_fops = {
 	.remap_file_range = ocfs2_remap_file_range,
 };
 
+WRAP_DIR_ITER(ocfs2_readdir) // FIXME!
 const struct file_operations ocfs2_dops = {
 	.llseek		= generic_file_llseek,
 	.read		= generic_read_dir,
-	.iterate	= ocfs2_readdir,
+	.iterate_shared	= shared_ocfs2_readdir,
 	.fsync		= ocfs2_sync_file,
 	.release	= ocfs2_dir_release,
 	.open		= ocfs2_dir_open,
@@ -2842,7 +2843,7 @@ const struct file_operations ocfs2_fops_no_plocks = {
 const struct file_operations ocfs2_dops_no_plocks = {
 	.llseek		= generic_file_llseek,
 	.read		= generic_read_dir,
-	.iterate	= ocfs2_readdir,
+	.iterate_shared	= shared_ocfs2_readdir,
 	.fsync		= ocfs2_sync_file,
 	.release	= ocfs2_dir_release,
 	.open		= ocfs2_dir_open,
diff --git a/fs/open.c b/fs/open.c
index 0c55c8e7f837..e6ead0f19964 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -1322,7 +1322,7 @@ inline int build_open_flags(const struct open_how *how, struct open_flags *op)
 		lookup_flags |= LOOKUP_IN_ROOT;
 	if (how->resolve & RESOLVE_CACHED) {
 		/* Don't bother even trying for create/truncate/tmpfile open */
-		if (flags & (O_TRUNC | O_CREAT | O_TMPFILE))
+		if (flags & (O_TRUNC | O_CREAT | __O_TMPFILE))
 			return -EAGAIN;
 		lookup_flags |= LOOKUP_CACHED;
 	}
diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c
index ee5c4736480f..de39e067ae65 100644
--- a/fs/overlayfs/readdir.c
+++ b/fs/overlayfs/readdir.c
@@ -954,10 +954,11 @@ static int ovl_dir_open(struct inode *inode, struct file *file)
 	return 0;
 }
 
+WRAP_DIR_ITER(ovl_iterate) // FIXME!
 const struct file_operations ovl_dir_operations = {
 	.read		= generic_read_dir,
 	.open		= ovl_dir_open,
-	.iterate	= ovl_iterate,
+	.iterate_shared	= shared_ovl_iterate,
 	.llseek		= ovl_dir_llseek,
 	.fsync		= ovl_dir_fsync,
 	.release	= ovl_dir_release,
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 05452c3b9872..9df3f4839662 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2817,7 +2817,7 @@ static int proc_##LSM##_attr_dir_iterate(struct file *filp, \
 \
 static const struct file_operations proc_##LSM##_attr_dir_ops = { \
 	.read		= generic_read_dir, \
-	.iterate	= proc_##LSM##_attr_dir_iterate, \
+	.iterate_shared	= proc_##LSM##_attr_dir_iterate, \
 	.llseek		= default_llseek, \
 }; \
 \
diff --git a/fs/readdir.c b/fs/readdir.c
index b264ce60114d..c8c46e294431 100644
--- a/fs/readdir.c
+++ b/fs/readdir.c
@@ -25,6 +25,53 @@
 #include <asm/unaligned.h>
 
 /*
+ * Some filesystems were never converted to '->iterate_shared()'
+ * and their directory iterators want the inode lock held for
+ * writing. This wrapper allows for converting from the shared
+ * semantics to the exclusive inode use.
+ */
+int wrap_directory_iterator(struct file *file,
+			    struct dir_context *ctx,
+			    int (*iter)(struct file *, struct dir_context *))
+{
+	struct inode *inode = file_inode(file);
+	int ret;
+
+	/*
+	 * We'd love to have an 'inode_upgrade_trylock()' operation,
+	 * see the comment in mmap_upgrade_trylock() in mm/memory.c.
+	 *
+	 * But considering this is for "filesystems that never got
+	 * converted", it really doesn't matter.
+	 *
+	 * Also note that since we have to return with the lock held
+	 * for reading, we can't use the "killable()" locking here,
+	 * since we do need to get the lock even if we're dying.
+	 *
+	 * We could do the write part killably and then get the read
+	 * lock unconditionally if it mattered, but see above on why
+	 * this does the very simplistic conversion.
+	 */
+	up_read(&inode->i_rwsem);
+	down_write(&inode->i_rwsem);
+
+	/*
+	 * Since we dropped the inode lock, we should do the
+	 * DEADDIR test again. See 'iterate_dir()' below.
+	 *
+	 * Note that we don't need to re-do the f_pos games,
+	 * since the file must be locked wrt f_pos anyway.
+	 */
+	ret = -ENOENT;
+	if (!IS_DEADDIR(inode))
+		ret = iter(file, ctx);
+
+	downgrade_write(&inode->i_rwsem);
+	return ret;
+}
+EXPORT_SYMBOL(wrap_directory_iterator);
+
+/*
  * Note the "unsafe_put_user() semantics: we goto a
  * label for errors.
  */
@@ -40,39 +87,28 @@
 int iterate_dir(struct file *file, struct dir_context *ctx)
 {
 	struct inode *inode = file_inode(file);
-	bool shared = false;
 	int res = -ENOTDIR;
-	if (file->f_op->iterate_shared)
-		shared = true;
-	else if (!file->f_op->iterate)
+
+	if (!file->f_op->iterate_shared)
 		goto out;
 
 	res = security_file_permission(file, MAY_READ);
 	if (res)
 		goto out;
 
-	if (shared)
-		res = down_read_killable(&inode->i_rwsem);
-	else
-		res = down_write_killable(&inode->i_rwsem);
+	res = down_read_killable(&inode->i_rwsem);
 	if (res)
 		goto out;
 
 	res = -ENOENT;
 	if (!IS_DEADDIR(inode)) {
 		ctx->pos = file->f_pos;
-		if (shared)
-			res = file->f_op->iterate_shared(file, ctx);
-		else
-			res = file->f_op->iterate(file, ctx);
+		res = file->f_op->iterate_shared(file, ctx);
 		file->f_pos = ctx->pos;
 		fsnotify_access(file);
 		file_accessed(file);
 	}
-	if (shared)
-		inode_unlock_shared(inode);
-	else
-		inode_unlock(inode);
+	inode_unlock_shared(inode);
 out:
 	return res;
 }
diff --git a/fs/smb/client/dfs.c b/fs/smb/client/dfs.c
index df3fd3b720da..ee772c3d9f00 100644
--- a/fs/smb/client/dfs.c
+++ b/fs/smb/client/dfs.c
@@ -177,8 +177,12 @@ static int __dfs_mount_share(struct cifs_mount_ctx *mnt_ctx)
 		struct dfs_cache_tgt_list tl = DFS_CACHE_TGT_LIST_INIT(tl);
 
 		rc = dfs_get_referral(mnt_ctx, ref_path + 1, NULL, &tl);
-		if (rc)
+		if (rc) {
+			rc = cifs_mount_get_tcon(mnt_ctx);
+			if (!rc)
+				rc = cifs_is_path_remote(mnt_ctx);
 			break;
+		}
 
 		tit = dfs_cache_get_tgt_iterator(&tl);
 		if (!tit) {
diff --git a/fs/smb/server/smb2misc.c b/fs/smb/server/smb2misc.c
index 33b7e6c4ceff..e881df1d10cb 100644
--- a/fs/smb/server/smb2misc.c
+++ b/fs/smb/server/smb2misc.c
@@ -380,13 +380,13 @@ int ksmbd_smb2_check_message(struct ksmbd_work *work)
 	}
 
 	if (smb2_req_struct_sizes[command] != pdu->StructureSize2) {
-		if (command == SMB2_OPLOCK_BREAK_HE &&
-		    le16_to_cpu(pdu->StructureSize2) != OP_BREAK_STRUCT_SIZE_20 &&
-		    le16_to_cpu(pdu->StructureSize2) != OP_BREAK_STRUCT_SIZE_21) {
+		if (!(command == SMB2_OPLOCK_BREAK_HE &&
+		    (le16_to_cpu(pdu->StructureSize2) == OP_BREAK_STRUCT_SIZE_20 ||
+		    le16_to_cpu(pdu->StructureSize2) == OP_BREAK_STRUCT_SIZE_21))) {
 			/* special case for SMB2.1 lease break message */
 			ksmbd_debug(SMB,
-				    "Illegal request size %d for oplock break\n",
-				    le16_to_cpu(pdu->StructureSize2));
+				"Illegal request size %u for command %d\n",
+				le16_to_cpu(pdu->StructureSize2), command);
 			return 1;
 		}
 	}
diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c
index 9849d7489345..7cc1b0c47d0a 100644
--- a/fs/smb/server/smb2pdu.c
+++ b/fs/smb/server/smb2pdu.c
@@ -2324,9 +2324,16 @@ next:
 			break;
 		buf_len -= next;
 		eabuf = (struct smb2_ea_info *)((char *)eabuf + next);
-		if (next < (u32)eabuf->EaNameLength + le16_to_cpu(eabuf->EaValueLength))
+		if (buf_len < sizeof(struct smb2_ea_info)) {
+			rc = -EINVAL;
 			break;
+		}
 
+		if (buf_len < sizeof(struct smb2_ea_info) + eabuf->EaNameLength +
+				le16_to_cpu(eabuf->EaValueLength)) {
+			rc = -EINVAL;
+			break;
+		}
 	} while (next != 0);
 
 	kfree(attr_name);
diff --git a/fs/vboxsf/dir.c b/fs/vboxsf/dir.c
index 075f15c43c78..5f1a14d5b927 100644
--- a/fs/vboxsf/dir.c
+++ b/fs/vboxsf/dir.c
@@ -179,9 +179,10 @@ static int vboxsf_dir_iterate(struct file *dir, struct dir_context *ctx)
 	return 0;
 }
 
+WRAP_DIR_ITER(vboxsf_dir_iterate) // FIXME!
 const struct file_operations vboxsf_dir_fops = {
 	.open = vboxsf_dir_open,
-	.iterate = vboxsf_dir_iterate,
+	.iterate_shared = shared_vboxsf_dir_iterate,
 	.release = vboxsf_dir_release,
 	.read = generic_read_dir,
 	.llseek = generic_file_llseek,
diff --git a/fs/vboxsf/shfl_hostintf.h b/fs/vboxsf/shfl_hostintf.h
index aca829062c12..069a019c9247 100644
--- a/fs/vboxsf/shfl_hostintf.h
+++ b/fs/vboxsf/shfl_hostintf.h
@@ -68,9 +68,9 @@ struct shfl_string {
 
 	/** UTF-8 or UTF-16 string. Nul terminated. */
 	union {
-		u8 utf8[2];
-		u16 utf16[1];
-		u16 ucs2[1]; /* misnomer, use utf16. */
+		u8 legacy_padding[2];
+		DECLARE_FLEX_ARRAY(u8, utf8);
+		DECLARE_FLEX_ARRAY(u16, utf16);
 	} string;
 };
 VMMDEV_ASSERT_SIZE(shfl_string, 6);