cifs: share dfs connections and supers

When matching DFS superblocks we can't rely on either the server's
address or tcon's UNC name from mount(2) as the existing servers and
tcons might be connected to somewhere else.  Instead, check if
superblock is dfs, and if so, match its original source pathname with
the new mount's source pathname.

For DFS connections, instead of checking server's address, match its
referral path as it could be connected to different targets.
Signed-off-by: Paulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: Steve French <stfrench@microsoft.com>

fs/cifs/cifs_debug.c

@@ -372,6 +372,14 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v)
 		seq_printf(m, "\nIn Send: %d In MaxReq Wait: %d",
 				atomic_read(&server->in_send),
 				atomic_read(&server->num_waiters));
+		if (IS_ENABLED(CONFIG_CIFS_DFS_UPCALL)) {
+			if (server->origin_fullpath)
+				seq_printf(m, "\nDFS origin full path: %s",
+					   server->origin_fullpath);
+			if (server->leaf_fullpath)
+				seq_printf(m, "\nDFS leaf full path:   %s",
+					   server->leaf_fullpath);
+		}
 
 		seq_printf(m, "\n\n\tSessions: ");
 		i = 0;

fs/cifs/cifsglob.h

@@ -738,8 +738,6 @@ struct TCP_Server_Info {
 	bool use_swn_dstaddr;
 	struct sockaddr_storage swn_dstaddr;
 #endif
-#ifdef CONFIG_CIFS_DFS_UPCALL
-	bool is_dfs_conn; /* if a dfs connection */
 	struct mutex refpath_lock; /* protects leaf_fullpath */
 	/*
 	 * Canonical DFS full paths that were used to chase referrals in mount and reconnect.
@@ -753,7 +751,6 @@ struct TCP_Server_Info {
 	 * format: \\HOST\SHARE\[OPTIONAL PATH]
 	 */
 	char *origin_fullpath, *leaf_fullpath, *current_fullpath;
-#endif
 };
 
 static inline bool is_smb1(struct TCP_Server_Info *server)
@@ -1767,11 +1764,9 @@ struct cifs_mount_ctx {
 	struct TCP_Server_Info *server;
 	struct cifs_ses *ses;
 	struct cifs_tcon *tcon;
-#ifdef CONFIG_CIFS_DFS_UPCALL
 	struct cifs_ses *root_ses;
 	uuid_t mount_id;
 	char *origin_fullpath, *leaf_fullpath;
-#endif
 };
 
 static inline void free_dfs_info_param(struct dfs_info3_param *param)

fs/cifs/cifsproto.h

@@ -242,7 +242,9 @@ extern int cifs_read_page_from_socket(struct TCP_Server_Info *server,
 					unsigned int page_offset,
 					unsigned int to_read);
 extern int cifs_setup_cifs_sb(struct cifs_sb_info *cifs_sb);
+void cifs_mount_put_conns(struct cifs_mount_ctx *mnt_ctx);
 int cifs_mount_get_session(struct cifs_mount_ctx *mnt_ctx);
+int cifs_is_path_remote(struct cifs_mount_ctx *mnt_ctx);
 int cifs_mount_get_tcon(struct cifs_mount_ctx *mnt_ctx);
 extern int cifs_match_super(struct super_block *, void *);
 extern int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb3_fs_context *ctx);

fs/cifs/connect.c

@@ -546,16 +546,8 @@ static int reconnect_dfs_server(struct TCP_Server_Info *server)
 
 int cifs_reconnect(struct TCP_Server_Info *server, bool mark_smb_session)
 {
-	/* If tcp session is not an dfs connection, then reconnect to last target server */
-	spin_lock(&server->srv_lock);
-	if (!server->is_dfs_conn) {
-		spin_unlock(&server->srv_lock);
-		return __cifs_reconnect(server, mark_smb_session);
-	}
-	spin_unlock(&server->srv_lock);
-
 	mutex_lock(&server->refpath_lock);
-	if (!server->origin_fullpath || !server->leaf_fullpath) {
+	if (!server->leaf_fullpath) {
 		mutex_unlock(&server->refpath_lock);
 		return __cifs_reconnect(server, mark_smb_session);
 	}
@@ -1367,9 +1359,7 @@ match_port(struct TCP_Server_Info *server, struct sockaddr *addr)
 	return port == *sport;
 }
 
-static bool
-match_address(struct TCP_Server_Info *server, struct sockaddr *addr,
-	      struct sockaddr *srcaddr)
+static bool match_server_address(struct TCP_Server_Info *server, struct sockaddr *addr)
 {
 	switch (addr->sa_family) {
 	case AF_INET: {
@@ -1398,9 +1388,6 @@ match_address(struct TCP_Server_Info *server, struct sockaddr *addr,
 		return false; /* don't expect to be here */
 	}
 
-	if (!cifs_match_ipaddr(srcaddr, (struct sockaddr *)&server->srcaddr))
-		return false;
-
 	return true;
 }
 
@@ -1428,7 +1415,8 @@ match_security(struct TCP_Server_Info *server, struct smb3_fs_context *ctx)
 }
 
 /* this function must be called with srv_lock held */
-static int match_server(struct TCP_Server_Info *server, struct smb3_fs_context *ctx)
+static int match_server(struct TCP_Server_Info *server, struct smb3_fs_context *ctx,
+			bool dfs_super_cmp)
 {
 	struct sockaddr *addr = (struct sockaddr *)&ctx->dstaddr;
 
@@ -1453,15 +1441,30 @@ static int match_server(struct TCP_Server_Info *server, struct smb3_fs_context *
 	if (!net_eq(cifs_net_ns(server), current->nsproxy->net_ns))
 		return 0;
 
-	if (strcasecmp(server->hostname, ctx->server_hostname))
+	if (!cifs_match_ipaddr((struct sockaddr *)&ctx->srcaddr,
+			       (struct sockaddr *)&server->srcaddr))
 		return 0;
-
-	if (!match_address(server, addr,
-			   (struct sockaddr *)&ctx->srcaddr))
+	/*
+	 * When matching DFS superblocks, we only check for original source pathname as the
+	 * currently connected target might be different than the one parsed earlier in i.e.
+	 * mount.cifs(8).
+	 */
+	if (dfs_super_cmp) {
+		if (!ctx->source || !server->origin_fullpath ||
+		    strcasecmp(server->origin_fullpath, ctx->source))
 			return 0;
-
-	if (!match_port(server, addr))
+	} else {
+		/* Skip addr, hostname and port matching for DFS connections */
+		if (server->leaf_fullpath) {
+			if (!ctx->leaf_fullpath ||
+			    strcasecmp(server->leaf_fullpath, ctx->leaf_fullpath))
+				return 0;
+		} else if (strcasecmp(server->hostname, ctx->server_hostname) ||
+			   !match_server_address(server, addr) ||
+			   !match_port(server, addr)) {
 			return 0;
+		}
+	}
 
 	if (!match_security(server, ctx))
 		return 0;
@@ -1489,23 +1492,11 @@ cifs_find_tcp_session(struct smb3_fs_context *ctx)
 	spin_lock(&cifs_tcp_ses_lock);
 	list_for_each_entry(server, &cifs_tcp_ses_list, tcp_ses_list) {
 		spin_lock(&server->srv_lock);
-#ifdef CONFIG_CIFS_DFS_UPCALL
-		/*
-		 * DFS failover implementation in cifs_reconnect() requires unique tcp sessions for
-		 * DFS connections to do failover properly, so avoid sharing them with regular
-		 * shares or even links that may connect to same server but having completely
-		 * different failover targets.
-		 */
-		if (server->is_dfs_conn) {
-			spin_unlock(&server->srv_lock);
-			continue;
-		}
-#endif
 		/*
 		 * Skip ses channels since they're only handled in lower layers
 		 * (e.g. cifs_send_recv).
 		 */
-		if (CIFS_SERVER_IS_CHAN(server) || !match_server(server, ctx)) {
+		if (CIFS_SERVER_IS_CHAN(server) || !match_server(server, ctx, false)) {
 			spin_unlock(&server->srv_lock);
 			continue;
 		}
@@ -1600,6 +1591,15 @@ cifs_get_tcp_session(struct smb3_fs_context *ctx,
 		goto out_err;
 	}
 
+	if (ctx->leaf_fullpath) {
+		tcp_ses->leaf_fullpath = kstrdup(ctx->leaf_fullpath, GFP_KERNEL);
+		if (!tcp_ses->leaf_fullpath) {
+			rc = -ENOMEM;
+			goto out_err;
+		}
+		tcp_ses->current_fullpath = tcp_ses->leaf_fullpath;
+	}
+
 	if (ctx->nosharesock)
 		tcp_ses->nosharesock = true;
 
@@ -1748,6 +1748,7 @@ cifs_get_tcp_session(struct smb3_fs_context *ctx,
 		if (CIFS_SERVER_IS_CHAN(tcp_ses))
 			cifs_put_tcp_session(tcp_ses->primary_server, false);
 		kfree(tcp_ses->hostname);
+		kfree(tcp_ses->leaf_fullpath);
 		if (tcp_ses->ssocket)
 			sock_release(tcp_ses->ssocket);
 		kfree(tcp_ses);
@@ -2277,11 +2278,12 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb3_fs_context *ctx)
 }
 
 /* this function must be called with tc_lock held */
-static int match_tcon(struct cifs_tcon *tcon, struct smb3_fs_context *ctx)
+static int match_tcon(struct cifs_tcon *tcon, struct smb3_fs_context *ctx, bool dfs_super_cmp)
 {
 	if (tcon->status == TID_EXITING)
 		return 0;
-	if (strncmp(tcon->tree_name, ctx->UNC, MAX_TREE_SIZE))
+	/* Skip UNC validation when matching DFS superblocks */
+	if (!dfs_super_cmp && strncmp(tcon->tree_name, ctx->UNC, MAX_TREE_SIZE))
 		return 0;
 	if (tcon->seal != ctx->seal)
 		return 0;
@@ -2304,7 +2306,7 @@ cifs_find_tcon(struct cifs_ses *ses, struct smb3_fs_context *ctx)
 	spin_lock(&cifs_tcp_ses_lock);
 	list_for_each_entry(tcon, &ses->tcon_list, tcon_list) {
 		spin_lock(&tcon->tc_lock);
-		if (!match_tcon(tcon, ctx)) {
+		if (!match_tcon(tcon, ctx, false)) {
 			spin_unlock(&tcon->tc_lock);
 			continue;
 		}
@@ -2699,6 +2701,7 @@ cifs_match_super(struct super_block *sb, void *data)
 	struct cifs_ses *ses;
 	struct cifs_tcon *tcon;
 	struct tcon_link *tlink;
+	bool dfs_super_cmp;
 	int rc = 0;
 
 	spin_lock(&cifs_tcp_ses_lock);
@@ -2713,14 +2716,16 @@ cifs_match_super(struct super_block *sb, void *data)
 	ses = tcon->ses;
 	tcp_srv = ses->server;
 
+	dfs_super_cmp = IS_ENABLED(CONFIG_CIFS_DFS_UPCALL) && tcp_srv->origin_fullpath;
+
 	ctx = mnt_data->ctx;
 
 	spin_lock(&tcp_srv->srv_lock);
 	spin_lock(&ses->ses_lock);
 	spin_lock(&tcon->tc_lock);
-	if (!match_server(tcp_srv, ctx) ||
+	if (!match_server(tcp_srv, ctx, dfs_super_cmp) ||
 	    !match_session(ses, ctx) ||
-	    !match_tcon(tcon, ctx) ||
+	    !match_tcon(tcon, ctx, dfs_super_cmp) ||
 	    !match_prepath(sb, mnt_data)) {
 		rc = 0;
 		goto out;
@@ -3177,7 +3182,7 @@ int cifs_setup_cifs_sb(struct cifs_sb_info *cifs_sb)
 }
 
 /* Release all succeed connections */
-static inline void mount_put_conns(struct cifs_mount_ctx *mnt_ctx)
+void cifs_mount_put_conns(struct cifs_mount_ctx *mnt_ctx)
 {
 	int rc = 0;
 
@@ -3327,18 +3332,6 @@ int cifs_mount_get_tcon(struct cifs_mount_ctx *mnt_ctx)
 	return rc;
 }
 
-/* Get connections for tcp, ses and tcon */
-static int mount_get_conns(struct cifs_mount_ctx *mnt_ctx)
-{
-	int rc;
-
-	rc = cifs_mount_get_session(mnt_ctx);
-	if (rc)
-		return rc;
-
-	return cifs_mount_get_tcon(mnt_ctx);
-}
-
 static int mount_setup_tlink(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses,
 			     struct cifs_tcon *tcon)
 {
@@ -3365,59 +3358,6 @@ static int mount_setup_tlink(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses,
 	return 0;
 }
 
-#ifdef CONFIG_CIFS_DFS_UPCALL
-/* Get unique dfs connections */
-static int mount_get_dfs_conns(struct cifs_mount_ctx *mnt_ctx)
-{
-	int rc;
-
-	mnt_ctx->fs_ctx->nosharesock = true;
-	rc = mount_get_conns(mnt_ctx);
-	if (mnt_ctx->server) {
-		cifs_dbg(FYI, "%s: marking tcp session as a dfs connection\n", __func__);
-		spin_lock(&mnt_ctx->server->srv_lock);
-		mnt_ctx->server->is_dfs_conn = true;
-		spin_unlock(&mnt_ctx->server->srv_lock);
-	}
-	return rc;
-}
-
-/*
- * cifs_build_path_to_root returns full path to root when we do not have an
- * existing connection (tcon)
- */
-static char *
-build_unc_path_to_root(const struct smb3_fs_context *ctx,
-		       const struct cifs_sb_info *cifs_sb, bool useppath)
-{
-	char *full_path, *pos;
-	unsigned int pplen = useppath && ctx->prepath ?
-		strlen(ctx->prepath) + 1 : 0;
-	unsigned int unc_len = strnlen(ctx->UNC, MAX_TREE_SIZE + 1);
-
-	if (unc_len > MAX_TREE_SIZE)
-		return ERR_PTR(-EINVAL);
-
-	full_path = kmalloc(unc_len + pplen + 1, GFP_KERNEL);
-	if (full_path == NULL)
-		return ERR_PTR(-ENOMEM);
-
-	memcpy(full_path, ctx->UNC, unc_len);
-	pos = full_path + unc_len;
-
-	if (pplen) {
-		*pos = CIFS_DIR_SEP(cifs_sb);
-		memcpy(pos + 1, ctx->prepath, pplen);
-		pos += pplen;
-	}
-
-	*pos = '\0'; /* add trailing null */
-	convert_delimiter(full_path, CIFS_DIR_SEP(cifs_sb));
-	cifs_dbg(FYI, "%s: full_path=%s\n", __func__, full_path);
-	return full_path;
-}
-#endif
-
 static int
 cifs_are_all_path_components_accessible(struct TCP_Server_Info *server,
 					unsigned int xid,
@@ -3470,7 +3410,7 @@ cifs_are_all_path_components_accessible(struct TCP_Server_Info *server,
  *
  * Return -EREMOTE if it is, otherwise 0 or -errno.
  */
-static int is_path_remote(struct cifs_mount_ctx *mnt_ctx)
+int cifs_is_path_remote(struct cifs_mount_ctx *mnt_ctx)
 {
 	int rc;
 	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
@@ -3514,250 +3454,19 @@ static int is_path_remote(struct cifs_mount_ctx *mnt_ctx)
 }
 
 #ifdef CONFIG_CIFS_DFS_UPCALL
-static void set_root_ses(struct cifs_mount_ctx *mnt_ctx)
-{
-	if (mnt_ctx->ses) {
-		spin_lock(&cifs_tcp_ses_lock);
-		mnt_ctx->ses->ses_count++;
-		spin_unlock(&cifs_tcp_ses_lock);
-		dfs_cache_add_refsrv_session(&mnt_ctx->mount_id, mnt_ctx->ses);
-	}
-	mnt_ctx->root_ses = mnt_ctx->ses;
-}
-
-static int is_dfs_mount(struct cifs_mount_ctx *mnt_ctx, bool *isdfs,
-			struct dfs_cache_tgt_list *root_tl)
-{
-	int rc;
-	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
-	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
-
-	*isdfs = true;
-
-	rc = mount_get_conns(mnt_ctx);
-	/*
-	 * If called with 'nodfs' mount option, then skip DFS resolving.  Otherwise unconditionally
-	 * try to get an DFS referral (even cached) to determine whether it is an DFS mount.
-	 *
-	 * Skip prefix path to provide support for DFS referrals from w2k8 servers which don't seem
-	 * to respond with PATH_NOT_COVERED to requests that include the prefix.
-	 */
-	if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_DFS) ||
-	    dfs_cache_find(mnt_ctx->xid, mnt_ctx->ses, cifs_sb->local_nls, cifs_remap(cifs_sb),
-			   ctx->UNC + 1, NULL, root_tl)) {
-		if (rc)
-			return rc;
-		/* Check if it is fully accessible and then mount it */
-		rc = is_path_remote(mnt_ctx);
-		if (!rc)
-			*isdfs = false;
-		else if (rc != -EREMOTE)
-			return rc;
-	}
-	return 0;
-}
-
-static int connect_dfs_target(struct cifs_mount_ctx *mnt_ctx, const char *full_path,
-			      const char *ref_path, struct dfs_cache_tgt_iterator *tit)
-{
-	int rc;
-	struct dfs_info3_param ref = {};
-	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
-
-	cifs_dbg(FYI, "%s: full_path=%s ref_path=%s target=%s\n", __func__, full_path, ref_path,
-		 dfs_cache_get_tgt_name(tit));
-
-	rc = dfs_cache_get_tgt_referral(ref_path, tit, &ref);
-	if (rc)
-		goto out;
-
-	rc = dfs_parse_target_referral(full_path + 1, &ref, mnt_ctx->fs_ctx);
-	if (rc)
-		goto out;
-
-	/* XXX: maybe check if we were actually redirected and avoid reconnecting? */
-	mount_put_conns(mnt_ctx);
-	rc = mount_get_dfs_conns(mnt_ctx);
-
-	if (!rc) {
-		if (cifs_is_referral_server(mnt_ctx->tcon, &ref))
-			set_root_ses(mnt_ctx);
-		rc = dfs_cache_update_tgthint(mnt_ctx->xid, mnt_ctx->root_ses, cifs_sb->local_nls,
-					      cifs_remap(cifs_sb), ref_path, tit);
-	}
-
-out:
-	free_dfs_info_param(&ref);
-	return rc;
-}
-
-static int connect_dfs_root(struct cifs_mount_ctx *mnt_ctx, struct dfs_cache_tgt_list *root_tl)
-{
-	int rc;
-	char *full_path;
-	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
-	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
-	struct dfs_cache_tgt_iterator *tit;
-
-	/* Put initial connections as they might be shared with other mounts.  We need unique dfs
-	 * connections per mount to properly failover, so mount_get_dfs_conns() must be used from
-	 * now on.
-	 */
-	mount_put_conns(mnt_ctx);
-	mount_get_dfs_conns(mnt_ctx);
-	set_root_ses(mnt_ctx);
-
-	full_path = build_unc_path_to_root(ctx, cifs_sb, true);
-	if (IS_ERR(full_path))
-		return PTR_ERR(full_path);
-
-	mnt_ctx->origin_fullpath = dfs_cache_canonical_path(ctx->UNC, cifs_sb->local_nls,
-							    cifs_remap(cifs_sb));
-	if (IS_ERR(mnt_ctx->origin_fullpath)) {
-		rc = PTR_ERR(mnt_ctx->origin_fullpath);
-		mnt_ctx->origin_fullpath = NULL;
-		goto out;
-	}
-
-	/* Try all dfs root targets */
-	for (rc = -ENOENT, tit = dfs_cache_get_tgt_iterator(root_tl);
-	     tit; tit = dfs_cache_get_next_tgt(root_tl, tit)) {
-		rc = connect_dfs_target(mnt_ctx, full_path, mnt_ctx->origin_fullpath + 1, tit);
-		if (!rc) {
-			mnt_ctx->leaf_fullpath = kstrdup(mnt_ctx->origin_fullpath, GFP_KERNEL);
-			if (!mnt_ctx->leaf_fullpath)
-				rc = -ENOMEM;
-			break;
-		}
-	}
-
-out:
-	kfree(full_path);
-	return rc;
-}
-
-static int __follow_dfs_link(struct cifs_mount_ctx *mnt_ctx)
-{
-	int rc;
-	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
-	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
-	char *full_path;
-	struct dfs_cache_tgt_list tl = DFS_CACHE_TGT_LIST_INIT(tl);
-	struct dfs_cache_tgt_iterator *tit;
-
-	full_path = build_unc_path_to_root(ctx, cifs_sb, true);
-	if (IS_ERR(full_path))
-		return PTR_ERR(full_path);
-
-	kfree(mnt_ctx->leaf_fullpath);
-	mnt_ctx->leaf_fullpath = dfs_cache_canonical_path(full_path, cifs_sb->local_nls,
-							  cifs_remap(cifs_sb));
-	if (IS_ERR(mnt_ctx->leaf_fullpath)) {
-		rc = PTR_ERR(mnt_ctx->leaf_fullpath);
-		mnt_ctx->leaf_fullpath = NULL;
-		goto out;
-	}
-
-	/* Get referral from dfs link */
-	rc = dfs_cache_find(mnt_ctx->xid, mnt_ctx->root_ses, cifs_sb->local_nls,
-			    cifs_remap(cifs_sb), mnt_ctx->leaf_fullpath + 1, NULL, &tl);
-	if (rc)
-		goto out;
-
-	/* Try all dfs link targets.  If an I/O fails from currently connected DFS target with an
-	 * error other than STATUS_PATH_NOT_COVERED (-EREMOTE), then retry it from other targets as
-	 * specified in MS-DFSC "3.1.5.2 I/O Operation to Target Fails with an Error Other Than
-	 * STATUS_PATH_NOT_COVERED."
-	 */
-	for (rc = -ENOENT, tit = dfs_cache_get_tgt_iterator(&tl);
-	     tit; tit = dfs_cache_get_next_tgt(&tl, tit)) {
-		rc = connect_dfs_target(mnt_ctx, full_path, mnt_ctx->leaf_fullpath + 1, tit);
-		if (!rc) {
-			rc = is_path_remote(mnt_ctx);
-			if (!rc || rc == -EREMOTE)
-				break;
-		}
-	}
-
-out:
-	kfree(full_path);
-	dfs_cache_free_tgts(&tl);
-	return rc;
-}
-
-static int follow_dfs_link(struct cifs_mount_ctx *mnt_ctx)
-{
-	int rc;
-	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
-	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
-	char *full_path;
-	int num_links = 0;
-
-	full_path = build_unc_path_to_root(ctx, cifs_sb, true);
-	if (IS_ERR(full_path))
-		return PTR_ERR(full_path);
-
-	kfree(mnt_ctx->origin_fullpath);
-	mnt_ctx->origin_fullpath = dfs_cache_canonical_path(full_path, cifs_sb->local_nls,
-							    cifs_remap(cifs_sb));
-	kfree(full_path);
-
-	if (IS_ERR(mnt_ctx->origin_fullpath)) {
-		rc = PTR_ERR(mnt_ctx->origin_fullpath);
-		mnt_ctx->origin_fullpath = NULL;
-		return rc;
-	}
-
-	do {
-		rc = __follow_dfs_link(mnt_ctx);
-		if (!rc || rc != -EREMOTE)
-			break;
-	} while (rc = -ELOOP, ++num_links < MAX_NESTED_LINKS);
-
-	return rc;
-}
-
-/* Set up DFS referral paths for failover */
-static void setup_server_referral_paths(struct cifs_mount_ctx *mnt_ctx)
-{
-	struct TCP_Server_Info *server = mnt_ctx->server;
-
-	mutex_lock(&server->refpath_lock);
-	server->origin_fullpath = mnt_ctx->origin_fullpath;
-	server->leaf_fullpath = mnt_ctx->leaf_fullpath;
-	server->current_fullpath = mnt_ctx->leaf_fullpath;
-	mutex_unlock(&server->refpath_lock);
-	mnt_ctx->origin_fullpath = mnt_ctx->leaf_fullpath = NULL;
-}
-
 int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb3_fs_context *ctx)
 {
-	int rc;
 	struct cifs_mount_ctx mnt_ctx = { .cifs_sb = cifs_sb, .fs_ctx = ctx, };
-	struct dfs_cache_tgt_list tl = DFS_CACHE_TGT_LIST_INIT(tl);
 	bool isdfs;
+	int rc;
 
-	rc = is_dfs_mount(&mnt_ctx, &isdfs, &tl);
+	uuid_gen(&mnt_ctx.mount_id);
+	rc = dfs_mount_share(&mnt_ctx, &isdfs);
 	if (rc)
 		goto error;
 	if (!isdfs)
 		goto out;
 
-	/* proceed as DFS mount */
-	uuid_gen(&mnt_ctx.mount_id);
-	rc = connect_dfs_root(&mnt_ctx, &tl);
-	dfs_cache_free_tgts(&tl);
-
-	if (rc)
-		goto error;
-
-	rc = is_path_remote(&mnt_ctx);
-	if (rc)
-		rc = follow_dfs_link(&mnt_ctx);
-	if (rc)
-		goto error;
-
-	setup_server_referral_paths(&mnt_ctx);
 	/*
 	 * After reconnecting to a different server, unique ids won't match anymore, so we disable
 	 * serverino. This prevents dentry revalidation to think the dentry are stale (ESTALE).
@@ -3786,7 +3495,7 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb3_fs_context *ctx)
 	dfs_cache_put_refsrv_sessions(&mnt_ctx.mount_id);
 	kfree(mnt_ctx.origin_fullpath);
 	kfree(mnt_ctx.leaf_fullpath);
-	mount_put_conns(&mnt_ctx);
+	cifs_mount_put_conns(&mnt_ctx);
 	return rc;
 }
 #else
@@ -3795,17 +3504,19 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb3_fs_context *ctx)
 	int rc = 0;
 	struct cifs_mount_ctx mnt_ctx = { .cifs_sb = cifs_sb, .fs_ctx = ctx, };
 
-	rc = mount_get_conns(&mnt_ctx);
+	rc = cifs_mount_get_session(&mnt_ctx);
+	if (rc)
+		goto error;
+
+	rc = cifs_mount_get_tcon(&mnt_ctx);
 	if (rc)
 		goto error;
 
-	if (mnt_ctx.tcon) {
-		rc = is_path_remote(&mnt_ctx);
+	rc = cifs_is_path_remote(&mnt_ctx);
 	if (rc == -EREMOTE)
 		rc = -EOPNOTSUPP;
 	if (rc)
 		goto error;
-	}
 
 	rc = mount_setup_tlink(cifs_sb, mnt_ctx.ses, mnt_ctx.tcon);
 	if (rc)
@@ -3815,7 +3526,7 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb3_fs_context *ctx)
 	return rc;
 
 error:
-	mount_put_conns(&mnt_ctx);
+	cifs_mount_put_conns(&mnt_ctx);
 	return rc;
 }
 #endif

fs/cifs/dfs.c

@@ -3,6 +3,7 @@
  * Copyright (c) 2022 Paulo Alcantara <palcantara@suse.de>
  */
 
+#include <linux/namei.h>
 #include "cifsproto.h"
 #include "cifs_debug.h"
 #include "dns_resolve.h"
@@ -52,3 +53,228 @@ int dfs_parse_target_referral(const char *full_path, const struct dfs_info3_para
 	kfree(path);
 	return rc;
 }
+
+/*
+ * cifs_build_path_to_root returns full path to root when we do not have an
+ * existing connection (tcon)
+ */
+static char *build_unc_path_to_root(const struct smb3_fs_context *ctx,
+				    const struct cifs_sb_info *cifs_sb, bool useppath)
+{
+	char *full_path, *pos;
+	unsigned int pplen = useppath && ctx->prepath ? strlen(ctx->prepath) + 1 : 0;
+	unsigned int unc_len = strnlen(ctx->UNC, MAX_TREE_SIZE + 1);
+
+	if (unc_len > MAX_TREE_SIZE)
+		return ERR_PTR(-EINVAL);
+
+	full_path = kmalloc(unc_len + pplen + 1, GFP_KERNEL);
+	if (full_path == NULL)
+		return ERR_PTR(-ENOMEM);
+
+	memcpy(full_path, ctx->UNC, unc_len);
+	pos = full_path + unc_len;
+
+	if (pplen) {
+		*pos = CIFS_DIR_SEP(cifs_sb);
+		memcpy(pos + 1, ctx->prepath, pplen);
+		pos += pplen;
+	}
+
+	*pos = '\0'; /* add trailing null */
+	convert_delimiter(full_path, CIFS_DIR_SEP(cifs_sb));
+	cifs_dbg(FYI, "%s: full_path=%s\n", __func__, full_path);
+	return full_path;
+}
+
+static int get_session(struct cifs_mount_ctx *mnt_ctx, const char *full_path)
+{
+	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
+	int rc;
+
+	ctx->leaf_fullpath = (char *)full_path;
+	rc = cifs_mount_get_session(mnt_ctx);
+	ctx->leaf_fullpath = NULL;
+
+	return rc;
+}
+
+static void set_root_ses(struct cifs_mount_ctx *mnt_ctx)
+{
+	if (mnt_ctx->ses) {
+		spin_lock(&cifs_tcp_ses_lock);
+		mnt_ctx->ses->ses_count++;
+		spin_unlock(&cifs_tcp_ses_lock);
+		dfs_cache_add_refsrv_session(&mnt_ctx->mount_id, mnt_ctx->ses);
+	}
+	mnt_ctx->root_ses = mnt_ctx->ses;
+}
+
+static int get_dfs_conn(struct cifs_mount_ctx *mnt_ctx, const char *ref_path, const char *full_path,
+			const struct dfs_cache_tgt_iterator *tit)
+{
+	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
+	struct dfs_info3_param ref = {};
+	int rc;
+
+	rc = dfs_cache_get_tgt_referral(ref_path + 1, tit, &ref);
+	if (rc)
+		return rc;
+
+	rc = dfs_parse_target_referral(full_path + 1, &ref, ctx);
+	if (rc)
+		goto out;
+
+	cifs_mount_put_conns(mnt_ctx);
+	rc = get_session(mnt_ctx, ref_path);
+	if (rc)
+		goto out;
+
+	if (ref.flags & DFSREF_REFERRAL_SERVER)
+		set_root_ses(mnt_ctx);
+
+	rc = -EREMOTE;
+	if (ref.flags & DFSREF_STORAGE_SERVER) {
+		rc = cifs_mount_get_tcon(mnt_ctx);
+		if (rc)
+			goto out;
+
+		/* some servers may not advertise referral capability under ref.flags */
+		if (!(ref.flags & DFSREF_REFERRAL_SERVER) &&
+		    is_tcon_dfs(mnt_ctx->tcon))
+			set_root_ses(mnt_ctx);
+
+		rc = cifs_is_path_remote(mnt_ctx);
+	}
+
+out:
+	free_dfs_info_param(&ref);
+	return rc;
+}
+
+static int __dfs_mount_share(struct cifs_mount_ctx *mnt_ctx)
+{
+	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
+	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
+	char *ref_path = NULL, *full_path = NULL;
+	struct dfs_cache_tgt_iterator *tit;
+	struct TCP_Server_Info *server;
+	char *origin_fullpath = NULL;
+	int num_links = 0;
+	int rc;
+
+	ref_path = dfs_get_path(cifs_sb, ctx->UNC);
+	if (IS_ERR(ref_path))
+		return PTR_ERR(ref_path);
+
+	full_path = build_unc_path_to_root(ctx, cifs_sb, true);
+	if (IS_ERR(full_path)) {
+		rc = PTR_ERR(full_path);
+		full_path = NULL;
+		goto out;
+	}
+
+	origin_fullpath = kstrdup(full_path, GFP_KERNEL);
+	if (!origin_fullpath) {
+		rc = -ENOMEM;
+		goto out;
+	}
+
+	do {
+		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)
+			break;
+
+		tit = dfs_cache_get_tgt_iterator(&tl);
+		if (!tit) {
+			cifs_dbg(VFS, "%s: dfs referral (%s) with no targets\n", __func__,
+				 ref_path + 1);
+			rc = -ENOENT;
+			dfs_cache_free_tgts(&tl);
+			break;
+		}
+
+		do {
+			rc = get_dfs_conn(mnt_ctx, ref_path, full_path, tit);
+			if (!rc)
+				break;
+			if (rc == -EREMOTE) {
+				if (++num_links > MAX_NESTED_LINKS) {
+					rc = -ELOOP;
+					break;
+				}
+				kfree(ref_path);
+				kfree(full_path);
+				ref_path = full_path = NULL;
+
+				full_path = build_unc_path_to_root(ctx, cifs_sb, true);
+				if (IS_ERR(full_path)) {
+					rc = PTR_ERR(full_path);
+					full_path = NULL;
+				} else {
+					ref_path = dfs_get_path(cifs_sb, full_path);
+					if (IS_ERR(ref_path)) {
+						rc = PTR_ERR(ref_path);
+						ref_path = NULL;
+					}
+				}
+				break;
+			}
+		} while ((tit = dfs_cache_get_next_tgt(&tl, tit)));
+		dfs_cache_free_tgts(&tl);
+	} while (rc == -EREMOTE);
+
+	if (!rc) {
+		server = mnt_ctx->server;
+
+		mutex_lock(&server->refpath_lock);
+		server->origin_fullpath = origin_fullpath;
+		server->current_fullpath = server->leaf_fullpath;
+		mutex_unlock(&server->refpath_lock);
+		origin_fullpath = NULL;
+	}
+
+out:
+	kfree(origin_fullpath);
+	kfree(ref_path);
+	kfree(full_path);
+	return rc;
+}
+
+int dfs_mount_share(struct cifs_mount_ctx *mnt_ctx, bool *isdfs)
+{
+	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
+	struct smb3_fs_context *ctx = mnt_ctx->fs_ctx;
+	int rc;
+
+	*isdfs = false;
+
+	rc = get_session(mnt_ctx, NULL);
+	if (rc)
+		return rc;
+	mnt_ctx->root_ses = mnt_ctx->ses;
+	/*
+	 * If called with 'nodfs' mount option, then skip DFS resolving.  Otherwise unconditionally
+	 * try to get an DFS referral (even cached) to determine whether it is an DFS mount.
+	 *
+	 * Skip prefix path to provide support for DFS referrals from w2k8 servers which don't seem
+	 * to respond with PATH_NOT_COVERED to requests that include the prefix.
+	 */
+	if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_DFS) ||
+	    dfs_get_referral(mnt_ctx, ctx->UNC + 1, NULL, NULL)) {
+		rc = cifs_mount_get_tcon(mnt_ctx);
+		if (rc)
+			return rc;
+
+		rc = cifs_is_path_remote(mnt_ctx);
+		if (!rc || rc != -EREMOTE)
+			return rc;
+	}
+
+	*isdfs = true;
+	set_root_ses(mnt_ctx);
+
+	return __dfs_mount_share(mnt_ctx);
+}

fs/cifs/dfs.h

@@ -8,9 +8,24 @@
 
 #include "cifsglob.h"
 #include "fs_context.h"
+#include "cifs_unicode.h"
 
 int dfs_parse_target_referral(const char *full_path, const struct dfs_info3_param *ref,
 			      struct smb3_fs_context *ctx);
+int dfs_mount_share(struct cifs_mount_ctx *mnt_ctx, bool *isdfs);
 
+static inline char *dfs_get_path(struct cifs_sb_info *cifs_sb, const char *path)
+{
+	return dfs_cache_canonical_path(path, cifs_sb->local_nls, cifs_remap(cifs_sb));
+}
+
+static inline int dfs_get_referral(struct cifs_mount_ctx *mnt_ctx, const char *path,
+				   struct dfs_info3_param *ref, struct dfs_cache_tgt_list *tl)
+{
+	struct cifs_sb_info *cifs_sb = mnt_ctx->cifs_sb;
+
+	return dfs_cache_find(mnt_ctx->xid, mnt_ctx->root_ses, cifs_sb->local_nls,
+			      cifs_remap(cifs_sb), path, ref, tl);
+}
 
 #endif /* _CIFS_DFS_H */

fs/cifs/dfs_cache.c

@@ -1519,12 +1519,8 @@ static void refresh_mounts(struct cifs_ses **sessions)
 
 	spin_lock(&cifs_tcp_ses_lock);
 	list_for_each_entry(server, &cifs_tcp_ses_list, tcp_ses_list) {
-		spin_lock(&server->srv_lock);
-		if (!server->is_dfs_conn) {
-			spin_unlock(&server->srv_lock);
+		if (!server->leaf_fullpath)
 			continue;
-		}
-		spin_unlock(&server->srv_lock);
 
 		list_for_each_entry(ses, &server->smb_ses_list, smb_ses_list) {
 			list_for_each_entry(tcon, &ses->tcon_list, tcon_list) {
@@ -1545,12 +1541,8 @@ static void refresh_mounts(struct cifs_ses **sessions)
 		list_del_init(&tcon->ulist);
 
 		mutex_lock(&server->refpath_lock);
-		if (server->origin_fullpath) {
-			if (server->leaf_fullpath && strcasecmp(server->leaf_fullpath,
-								server->origin_fullpath))
+		if (server->leaf_fullpath)
 			__refresh_tcon(server->leaf_fullpath + 1, sessions, tcon, false);
-			__refresh_tcon(server->origin_fullpath + 1, sessions, tcon, false);
-		}
 		mutex_unlock(&server->refpath_lock);
 
 		cifs_put_tcon(tcon);

fs/cifs/fs_context.c

@@ -316,6 +316,7 @@ smb3_fs_context_dup(struct smb3_fs_context *new_ctx, struct smb3_fs_context *ctx
 	new_ctx->UNC = NULL;
 	new_ctx->source = NULL;
 	new_ctx->iocharset = NULL;
+	new_ctx->leaf_fullpath = NULL;
 	/*
 	 * Make sure to stay in sync with smb3_cleanup_fs_context_contents()
 	 */
@@ -328,6 +329,7 @@ smb3_fs_context_dup(struct smb3_fs_context *new_ctx, struct smb3_fs_context *ctx
 	DUP_CTX_STR(domainname);
 	DUP_CTX_STR(nodename);
 	DUP_CTX_STR(iocharset);
+	DUP_CTX_STR(leaf_fullpath);
 
 	return 0;
 }
@@ -1592,6 +1594,8 @@ smb3_cleanup_fs_context_contents(struct smb3_fs_context *ctx)
 	ctx->iocharset = NULL;
 	kfree(ctx->prepath);
 	ctx->prepath = NULL;
+	kfree(ctx->leaf_fullpath);
+	ctx->leaf_fullpath = NULL;
 }
 
 void

fs/cifs/fs_context.h

@@ -264,6 +264,7 @@ struct smb3_fs_context {
 	__u16 compression; /* compression algorithm 0xFFFF default 0=disabled */
 	bool rootfs:1; /* if it's a SMB root file system */
 	bool witness:1; /* use witness protocol */
+	char *leaf_fullpath;
 };
 
 extern const struct fs_parameter_spec smb3_fs_parameters[];