[PATCH AUTOSEL 6.7 28/44] cifs: make sure that channel scaling is done only once

[PATCH AUTOSEL 6.7 28/44] cifs: make sure that channel scaling is done only once

[Sasha Levin]

From: Shyam Prasad N <sprasad@microsoft.com>

[ Upstream commit ee36a3b345c433a846effcdcfba437c2298eeda5 ]

Following a successful cifs_tree_connect, we have the code
to scale up/down the number of channels in the session.
However, it is not protected by a lock today.

As a result, this code can be executed by several processes
that select the same channel. The core functions handle this
well, as they pick chan_lock. However, we've seen cases where
smb2_reconnect throws some warnings.

To fix that, this change introduces a flags bitmap inside the
cifs_ses structure. A new flag type is used to ensure that
only one process enters this section at any time.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/smb/client/cifsglob.h |  3 +++
 fs/smb/client/smb2pdu.c  | 18 +++++++++++++++++-
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/fs/smb/client/cifsglob.h b/fs/smb/client/cifsglob.h
index 5e32c79f03a7..75ca732a1679 100644
--- a/fs/smb/client/cifsglob.h
+++ b/fs/smb/client/cifsglob.h
@@ -1004,6 +1004,8 @@ struct cifs_chan {
 	__u8 signkey[SMB3_SIGN_KEY_SIZE];
 };
 
+#define CIFS_SES_FLAG_SCALE_CHANNELS (0x1)
+
 /*
  * Session structure.  One of these for each uid session with a particular host
  */
@@ -1036,6 +1038,7 @@ struct cifs_ses {
 	enum securityEnum sectype; /* what security flavor was specified? */
 	bool sign;		/* is signing required? */
 	bool domainAuto:1;
+	unsigned int flags;
 	__u16 session_flags;
 	__u8 smb3signingkey[SMB3_SIGN_KEY_SIZE];
 	__u8 smb3encryptionkey[SMB3_ENC_DEC_KEY_SIZE];
diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c
index f5006aa97f5b..ab8765113392 100644
--- a/fs/smb/client/smb2pdu.c
+++ b/fs/smb/client/smb2pdu.c
@@ -399,6 +399,15 @@ smb2_reconnect(__le16 smb2_command, struct cifs_tcon *tcon,
 		goto out;
 	}
 
+	spin_lock(&ses->ses_lock);
+	if (ses->flags & CIFS_SES_FLAG_SCALE_CHANNELS) {
+		spin_unlock(&ses->ses_lock);
+		mutex_unlock(&ses->session_mutex);
+		goto skip_add_channels;
+	}
+	ses->flags |= CIFS_SES_FLAG_SCALE_CHANNELS;
+	spin_unlock(&ses->ses_lock);
+
 	if (!rc &&
 	    (server->capabilities & SMB2_GLOBAL_CAP_MULTI_CHANNEL)) {
 		mutex_unlock(&ses->session_mutex);
@@ -428,15 +437,22 @@ smb2_reconnect(__le16 smb2_command, struct cifs_tcon *tcon,
 		if (ses->chan_max > ses->chan_count &&
 		    ses->iface_count &&
 		    !SERVER_IS_CHAN(server)) {
-			if (ses->chan_count == 1)
+			if (ses->chan_count == 1) {
 				cifs_server_dbg(VFS, "supports multichannel now\n");
+				queue_delayed_work(cifsiod_wq, &tcon->query_interfaces,
+						 (SMB_INTERFACE_POLL_INTERVAL * HZ));
+			}
 
 			cifs_try_adding_channels(ses);
 		}
 	} else {
 		mutex_unlock(&ses->session_mutex);
 	}
+
 skip_add_channels:
+	spin_lock(&ses->ses_lock);
+	ses->flags &= ~CIFS_SES_FLAG_SCALE_CHANNELS;
+	spin_unlock(&ses->ses_lock);
 
 	if (smb2_command != SMB2_INTERNAL_CMD)
 		mod_delayed_work(cifsiod_wq, &server->reconnect, 0);
-- 
2.43.0

[PATCH AUTOSEL 6.7 41/44] cifs: do not search for channel if server is terminating

[Sasha Levin]

From: Shyam Prasad N <sprasad@microsoft.com>

[ Upstream commit 88675b22d34e6e815ad4bde09c590ccb2d50c59d ]

In order to scale down the channels, the following sequence
of operations happen:
1. server struct is marked for terminate
2. the channel is deallocated in the ses->chans array
3. at a later point the cifsd thread actually terminates the server

Between 2 and 3, there can be calls to find the channel for
a server struct. When that happens, there can be an ugly warning
that's logged. But this is expected.

So this change does two things:
1. in cifs_ses_get_chan_index, if server->terminate is set, return
2. always make sure server->terminate is set with chan_lock held

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/smb/client/sess.c    | 4 ++++
 fs/smb/client/smb2pdu.c | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/fs/smb/client/sess.c b/fs/smb/client/sess.c
index 2d3b332a79a1..b503a4833ae0 100644
--- a/fs/smb/client/sess.c
+++ b/fs/smb/client/sess.c
@@ -75,6 +75,10 @@ cifs_ses_get_chan_index(struct cifs_ses *ses,
 {
 	unsigned int i;
 
+	/* if the channel is waiting for termination */
+	if (server->terminate)
+		return CIFS_INVAL_CHAN_INDEX;
+
 	for (i = 0; i < ses->chan_count; i++) {
 		if (ses->chans[i].server == server)
 			return i;
diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c
index ab8765113392..bd2822ad9b22 100644
--- a/fs/smb/client/smb2pdu.c
+++ b/fs/smb/client/smb2pdu.c
@@ -178,6 +178,7 @@ cifs_chan_skip_or_disable(struct cifs_ses *ses,
 		}
 
 		ses->chans[chan_index].server = NULL;
+		server->terminate = true;
 		spin_unlock(&ses->chan_lock);
 
 		/*
@@ -188,7 +189,6 @@ cifs_chan_skip_or_disable(struct cifs_ses *ses,
 		 */
 		cifs_put_tcp_session(server, from_reconnect);
 
-		server->terminate = true;
 		cifs_signal_cifsd_for_reconnect(server, false);
 
 		/* mark primary server as needing reconnect */
-- 
2.43.0

[PATCH AUTOSEL 6.7 42/44] smb: client: increase number of PDUs allowed in a compound request

[Sasha Levin]

From: Paulo Alcantara <pc@manguebit.com>

[ Upstream commit 11d4d1dba3315f73d2d1d386f5bf4811a8241d45 ]

With the introduction of SMB2_OP_QUERY_WSL_EA, the client may now send
5 commands in a single compound request in order to query xattrs from
potential WSL reparse points, which should be fine as we currently
allow up to 5 PDUs in a single compound request.  However, if
encryption is enabled (e.g. 'seal' mount option) or enforced by the
server, current MAX_COMPOUND(5) won't be enough as we require an extra
PDU for the transform header.

Fix this by increasing MAX_COMPOUND to 7 and, while we're at it, add
an WARN_ON_ONCE() and return -EIO instead of -ENOMEM in case we
attempt to send a compound request that couldn't include the extra
transform header.

Signed-off-by: Paulo Alcantara <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/smb/client/cifsglob.h  | 2 +-
 fs/smb/client/transport.c | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/smb/client/cifsglob.h b/fs/smb/client/cifsglob.h
index 75ca732a1679..b65ce8d13143 100644
--- a/fs/smb/client/cifsglob.h
+++ b/fs/smb/client/cifsglob.h
@@ -82,7 +82,7 @@
 #define SMB_INTERFACE_POLL_INTERVAL	600
 
 /* maximum number of PDUs in one compound */
-#define MAX_COMPOUND 5
+#define MAX_COMPOUND 7
 
 /*
  * Default number of credits to keep available for SMB3.
diff --git a/fs/smb/client/transport.c b/fs/smb/client/transport.c
index 4f717ad7c21b..2d9146600371 100644
--- a/fs/smb/client/transport.c
+++ b/fs/smb/client/transport.c
@@ -428,8 +428,8 @@ smb_send_rqst(struct TCP_Server_Info *server, int num_rqst,
 	if (!(flags & CIFS_TRANSFORM_REQ))
 		return __smb_send_rqst(server, num_rqst, rqst);
 
-	if (num_rqst > MAX_COMPOUND - 1)
-		return -ENOMEM;
+	if (WARN_ON_ONCE(num_rqst > MAX_COMPOUND - 1))
+		return -EIO;
 
 	if (!server->ops->init_transform_rq) {
 		cifs_server_dbg(VFS, "Encryption requested but transform callback is missing\n");
-- 
2.43.0