Merge tag '5.2-smb3' of git://git.samba.org/sfrench/cifs-2.6

Pull cifs fixes from Steve French:
 "CIFS/SMB3 changes:

   - three fixes for stable

   - add fiemap support

   - improve zero-range support

   - various RDMA (smb direct fixes)

  I have an additional set of fixes (for improved handling of sparse
  files, mode bits, POSIX extensions) that are still being tested that
  are not included in this pull request but I expect to send in the next
  week"

* tag '5.2-smb3' of git://git.samba.org/sfrench/cifs-2.6: (29 commits)
  cifs: update module internal version number
  SMB3: Clean up query symlink when reparse point
  cifs: fix strcat buffer overflow and reduce raciness in smb21_set_oplock_level()
  Negotiate and save preferred compression algorithms
  cifs: rename and clarify CIFS_ASYNC_OP and CIFS_NO_RESP
  cifs: fix credits leak for SMB1 oplock breaks
  smb3: Add protocol structs for change notify support
  cifs: fix smb3_zero_range for Azure
  cifs: zero-range does not require the file is sparse
  Add new flag on SMB3.1.1 read
  cifs: add fiemap support
  SMB3: Add defines for new negotiate contexts
  cifs: fix bi-directional fsctl passthrough calls
  cifs: smbd: take an array of reqeusts when sending upper layer data
  SMB3: Add handling for different FSCTL access flags
  cifs: Add support for FSCTL passthrough that write data to the server
  cifs: remove superfluous inode_lock in cifs_{strict_}fsync
  cifs: Call MID callback before destroying transport
  cifs: smbd: Retry on memory registration failure
  cifs: smbd: Indicate to retry on transport sending failure
  ...

fs/cifs/cifs_debug.c

@@ -312,12 +312,10 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v)
 			atomic_read(&server->smbd_conn->send_credits),
 			atomic_read(&server->smbd_conn->receive_credits),
 			server->smbd_conn->receive_credit_target);
-		seq_printf(m, "\nPending send_pending: %x send_payload_pending:"
-			" %x smbd_send_pending: %x smbd_recv_pending: %x",
+		seq_printf(m, "\nPending send_pending: %x "
+			"send_payload_pending: %x",
 			atomic_read(&server->smbd_conn->send_pending),
-			atomic_read(&server->smbd_conn->send_payload_pending),
-			server->smbd_conn->smbd_send_pending,
-			server->smbd_conn->smbd_recv_pending);
+			atomic_read(&server->smbd_conn->send_payload_pending));
 		seq_printf(m, "\nReceive buffers count_receive_queue: %x "
 			"count_empty_packet_queue: %x",
 			server->smbd_conn->count_receive_queue,
@@ -334,6 +332,12 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v)
 #endif
 		seq_printf(m, "\nNumber of credits: %d Dialect 0x%x",
 			server->credits,  server->dialect);
+		if (server->compress_algorithm == SMB3_COMPRESS_LZNT1)
+			seq_printf(m, " COMPRESS_LZNT1");
+		else if (server->compress_algorithm == SMB3_COMPRESS_LZ77)
+			seq_printf(m, " COMPRESS_LZ77");
+		else if (server->compress_algorithm == SMB3_COMPRESS_LZ77_HUFF)
+			seq_printf(m, " COMPRESS_LZ77_HUFF");
 		if (server->sign)
 			seq_printf(m, " signed");
 		if (server->posix_ext_supported)
@@ -462,8 +466,13 @@ static ssize_t cifs_stats_proc_write(struct file *file,
 			server = list_entry(tmp1, struct TCP_Server_Info,
 					    tcp_ses_list);
 #ifdef CONFIG_CIFS_STATS2
-			for (i = 0; i < NUMBER_OF_SMB2_COMMANDS; i++)
+			for (i = 0; i < NUMBER_OF_SMB2_COMMANDS; i++) {
+				atomic_set(&server->num_cmds[i], 0);
 				atomic_set(&server->smb2slowcmd[i], 0);
+				server->time_per_cmd[i] = 0;
+				server->slowest_cmd[i] = 0;
+				server->fastest_cmd[0] = 0;
+			}
 #endif /* CONFIG_CIFS_STATS2 */
 			list_for_each(tmp2, &server->smb_ses_list) {
 				ses = list_entry(tmp2, struct cifs_ses,
@@ -531,9 +540,19 @@ static int cifs_stats_proc_show(struct seq_file *m, void *v)
 		server = list_entry(tmp1, struct TCP_Server_Info,
 				    tcp_ses_list);
 #ifdef CONFIG_CIFS_STATS2
+		seq_puts(m, "\nTotal time spent processing by command. Time ");
+		seq_printf(m, "units are jiffies (%d per second)\n", HZ);
+		seq_puts(m, "  SMB3 CMD\tNumber\tTotal Time\tFastest\tSlowest\n");
+		seq_puts(m, "  --------\t------\t----------\t-------\t-------\n");
+		for (j = 0; j < NUMBER_OF_SMB2_COMMANDS; j++)
+			seq_printf(m, "  %d\t\t%d\t%llu\t\t%u\t%u\n", j,
+				atomic_read(&server->num_cmds[j]),
+				server->time_per_cmd[j],
+				server->fastest_cmd[j],
+				server->slowest_cmd[j]);
 		for (j = 0; j < NUMBER_OF_SMB2_COMMANDS; j++)
 			if (atomic_read(&server->smb2slowcmd[j]))
-				seq_printf(m, "%d slow responses from %s for command %d\n",
+				seq_printf(m, "  %d slow responses from %s for command %d\n",
 					atomic_read(&server->smb2slowcmd[j]),
 					server->hostname, j);
 #endif /* STATS2 */

fs/cifs/cifsfs.c

@@ -483,6 +483,8 @@ cifs_show_options(struct seq_file *s, struct dentry *root)
 		seq_puts(s, ",seal");
 	if (tcon->nocase)
 		seq_puts(s, ",nocase");
+	if (tcon->local_lease)
+		seq_puts(s, ",locallease");
 	if (tcon->retry)
 		seq_puts(s, ",hard");
 	else
@@ -984,6 +986,7 @@ const struct inode_operations cifs_file_inode_ops = {
 	.getattr = cifs_getattr,
 	.permission = cifs_permission,
 	.listxattr = cifs_listxattr,
+	.fiemap = cifs_fiemap,
 };
 
 const struct inode_operations cifs_symlink_inode_ops = {

fs/cifs/cifsfs.h

@@ -84,6 +84,8 @@ extern int cifs_revalidate_mapping(struct inode *inode);
 extern int cifs_zap_mapping(struct inode *inode);
 extern int cifs_getattr(const struct path *, struct kstat *, u32, unsigned int);
 extern int cifs_setattr(struct dentry *, struct iattr *);
+extern int cifs_fiemap(struct inode *, struct fiemap_extent_info *, u64 start,
+		       u64 len);
 
 extern const struct inode_operations cifs_file_inode_ops;
 extern const struct inode_operations cifs_symlink_inode_ops;
@@ -150,5 +152,5 @@ extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg);
 extern const struct export_operations cifs_export_ops;
 #endif /* CONFIG_CIFS_NFSD_EXPORT */
 
-#define CIFS_VERSION   "2.19"
+#define CIFS_VERSION   "2.20"
 #endif				/* _CIFSFS_H */

fs/cifs/cifsglob.h

@@ -355,7 +355,8 @@ struct smb_version_operations {
 			       struct cifs_sb_info *);
 	/* query symlink target */
 	int (*query_symlink)(const unsigned int, struct cifs_tcon *,
-			     const char *, char **, struct cifs_sb_info *);
+			     struct cifs_sb_info *, const char *,
+			     char **, bool);
 	/* open a file for non-posix mounts */
 	int (*open)(const unsigned int, struct cifs_open_parms *,
 		    __u32 *, FILE_ALL_INFO *);
@@ -493,6 +494,9 @@ struct smb_version_operations {
 			 char *full_path,
 			 umode_t mode,
 			 dev_t device_number);
+	/* version specific fiemap implementation */
+	int (*fiemap)(struct cifs_tcon *tcon, struct cifsFileInfo *,
+		      struct fiemap_extent_info *, u64, u64);
 };
 
 struct smb_version_values {
@@ -596,6 +600,10 @@ struct smb_vol {
 	unsigned int max_credits; /* smb3 max_credits 10 < credits < 60000 */
 };
 
+/**
+ * CIFS superblock mount flags (mnt_cifs_flags) to consider when
+ * trying to reuse existing superblock for a new mount
+ */
 #define CIFS_MOUNT_MASK (CIFS_MOUNT_NO_PERM | CIFS_MOUNT_SET_UID | \
 			 CIFS_MOUNT_SERVER_INUM | CIFS_MOUNT_DIRECT_IO | \
 			 CIFS_MOUNT_NO_XATTR | CIFS_MOUNT_MAP_SPECIAL_CHR | \
@@ -606,8 +614,13 @@ struct smb_vol {
 			 CIFS_MOUNT_NOPOSIXBRL | CIFS_MOUNT_NOSSYNC | \
 			 CIFS_MOUNT_FSCACHE | CIFS_MOUNT_MF_SYMLINKS | \
 			 CIFS_MOUNT_MULTIUSER | CIFS_MOUNT_STRICT_IO | \
-			 CIFS_MOUNT_CIFS_BACKUPUID | CIFS_MOUNT_CIFS_BACKUPGID)
+			 CIFS_MOUNT_CIFS_BACKUPUID | CIFS_MOUNT_CIFS_BACKUPGID | \
+			 CIFS_MOUNT_NO_DFS)
 
+/**
+ * Generic VFS superblock mount flags (s_flags) to consider when
+ * trying to reuse existing superblock for a new mount
+ */
 #define CIFS_MS_MASK (SB_RDONLY | SB_MANDLOCK | SB_NOEXEC | SB_NOSUID | \
 		      SB_NODEV | SB_SYNCHRONOUS)
 
@@ -714,10 +727,15 @@ struct TCP_Server_Info {
 #ifdef CONFIG_CIFS_STATS2
 	atomic_t in_send; /* requests trying to send */
 	atomic_t num_waiters;   /* blocked waiting to get in sendrecv */
+	atomic_t num_cmds[NUMBER_OF_SMB2_COMMANDS]; /* total requests by cmd */
 	atomic_t smb2slowcmd[NUMBER_OF_SMB2_COMMANDS]; /* count resps > 1 sec */
+	__u64 time_per_cmd[NUMBER_OF_SMB2_COMMANDS]; /* total time per cmd */
+	__u32 slowest_cmd[NUMBER_OF_SMB2_COMMANDS];
+	__u32 fastest_cmd[NUMBER_OF_SMB2_COMMANDS];
 #endif /* STATS2 */
 	unsigned int	max_read;
 	unsigned int	max_write;
+	__le16	compress_algorithm;
 	__le16	cipher_type;
 	 /* save initital negprot hash */
 	__u8	preauth_sha_hash[SMB2_PREAUTH_HASH_SIZE];
@@ -1673,11 +1691,11 @@ static inline bool is_retryable_error(int error)
 
 /* Type of Request to SendReceive2 */
 #define   CIFS_BLOCKING_OP      1    /* operation can block */
-#define   CIFS_ASYNC_OP         2    /* do not wait for response */
+#define   CIFS_NON_BLOCKING     2    /* do not block waiting for credits */
 #define   CIFS_TIMEOUT_MASK 0x003    /* only one of above set in req */
 #define   CIFS_LOG_ERROR    0x010    /* log NT STATUS if non-zero */
 #define   CIFS_LARGE_BUF_OP 0x020    /* large request buffer */
-#define   CIFS_NO_RESP      0x040    /* no response buffer required */
+#define   CIFS_NO_RSP_BUF   0x040    /* no response buffer required */
 
 /* Type of request operation */
 #define   CIFS_ECHO_OP      0x080    /* echo request */
@@ -1687,6 +1705,7 @@ static inline bool is_retryable_error(int error)
 
 #define   CIFS_HAS_CREDITS 0x0400    /* already has credits */
 #define   CIFS_TRANSFORM_REQ 0x0800    /* transform request before sending */
+#define   CIFS_NO_SRV_RSP    0x1000    /* there is no server response */
 
 /* Security Flags: indicate type of session setup needed */
 #define   CIFSSEC_MAY_SIGN	0x00001

fs/cifs/cifsproto.h

@@ -526,12 +526,21 @@ extern int E_md4hash(const unsigned char *passwd, unsigned char *p16,
 			const struct nls_table *codepage);
 extern int SMBencrypt(unsigned char *passwd, const unsigned char *c8,
 			unsigned char *p24);
+
+extern int
+cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data,
+		       const char *devname, bool is_smb3);
 extern void
 cifs_cleanup_volume_info_contents(struct smb_vol *volume_info);
 
 extern struct TCP_Server_Info *
 cifs_find_tcp_session(struct smb_vol *vol);
 
+extern void cifs_put_smb_ses(struct cifs_ses *ses);
+
+extern struct cifs_ses *
+cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info);
+
 void cifs_readdata_release(struct kref *refcount);
 int cifs_async_readv(struct cifs_readdata *rdata);
 int cifs_readv_receive(struct TCP_Server_Info *server, struct mid_q_entry *mid);

fs/cifs/cifssmb.c

@@ -860,7 +860,7 @@ CIFSSMBEcho(struct TCP_Server_Info *server)
 	iov[1].iov_base = (char *)smb + 4;
 
 	rc = cifs_call_async(server, &rqst, NULL, cifs_echo_callback, NULL,
-			     server, CIFS_ASYNC_OP | CIFS_ECHO_OP, NULL);
+			     server, CIFS_NON_BLOCKING | CIFS_ECHO_OP, NULL);
 	if (rc)
 		cifs_dbg(FYI, "Echo request failed: %d\n", rc);
 
@@ -2508,8 +2508,8 @@ int cifs_lockv(const unsigned int xid, struct cifs_tcon *tcon,
 	iov[1].iov_len = (num_unlock + num_lock) * sizeof(LOCKING_ANDX_RANGE);
 
 	cifs_stats_inc(&tcon->stats.cifs_stats.num_locks);
-	rc = SendReceive2(xid, tcon->ses, iov, 2, &resp_buf_type, CIFS_NO_RESP,
-			  &rsp_iov);
+	rc = SendReceive2(xid, tcon->ses, iov, 2, &resp_buf_type,
+			  CIFS_NO_RSP_BUF, &rsp_iov);
 	cifs_small_buf_release(pSMB);
 	if (rc)
 		cifs_dbg(FYI, "Send error in cifs_lockv = %d\n", rc);
@@ -2540,7 +2540,7 @@ CIFSSMBLock(const unsigned int xid, struct cifs_tcon *tcon,
 
 	if (lockType == LOCKING_ANDX_OPLOCK_RELEASE) {
 		/* no response expected */
-		flags = CIFS_ASYNC_OP | CIFS_OBREAK_OP;
+		flags = CIFS_NO_SRV_RSP | CIFS_NON_BLOCKING | CIFS_OBREAK_OP;
 		pSMB->Timeout = 0;
 	} else if (waitFlag) {
 		flags = CIFS_BLOCKING_OP; /* blocking operation, no timeout */
@@ -6567,93 +6567,3 @@ CIFSSMBSetEA(const unsigned int xid, struct cifs_tcon *tcon,
 	return rc;
 }
 #endif
-
-#ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* BB unused temporarily */
-/*
- *	Years ago the kernel added a "dnotify" function for Samba server,
- *	to allow network clients (such as Windows) to display updated
- *	lists of files in directory listings automatically when
- *	files are added by one user when another user has the
- *	same directory open on their desktop.  The Linux cifs kernel
- *	client hooked into the kernel side of this interface for
- *	the same reason, but ironically when the VFS moved from
- *	"dnotify" to "inotify" it became harder to plug in Linux
- *	network file system clients (the most obvious use case
- *	for notify interfaces is when multiple users can update
- *	the contents of the same directory - exactly what network
- *	file systems can do) although the server (Samba) could
- *	still use it.  For the short term we leave the worker
- *	function ifdeffed out (below) until inotify is fixed
- *	in the VFS to make it easier to plug in network file
- *	system clients.  If inotify turns out to be permanently
- *	incompatible for network fs clients, we could instead simply
- *	expose this config flag by adding a future cifs (and smb2) notify ioctl.
- */
-int CIFSSMBNotify(const unsigned int xid, struct cifs_tcon *tcon,
-		  const int notify_subdirs, const __u16 netfid,
-		  __u32 filter, struct file *pfile, int multishot,
-		  const struct nls_table *nls_codepage)
-{
-	int rc = 0;
-	struct smb_com_transaction_change_notify_req *pSMB = NULL;
-	struct smb_com_ntransaction_change_notify_rsp *pSMBr = NULL;
-	struct dir_notify_req *dnotify_req;
-	int bytes_returned;
-
-	cifs_dbg(FYI, "In CIFSSMBNotify for file handle %d\n", (int)netfid);
-	rc = smb_init(SMB_COM_NT_TRANSACT, 23, tcon, (void **) &pSMB,
-		      (void **) &pSMBr);
-	if (rc)
-		return rc;
-
-	pSMB->TotalParameterCount = 0 ;
-	pSMB->TotalDataCount = 0;
-	pSMB->MaxParameterCount = cpu_to_le32(2);
-	pSMB->MaxDataCount = cpu_to_le32(CIFSMaxBufSize & 0xFFFFFF00);
-	pSMB->MaxSetupCount = 4;
-	pSMB->Reserved = 0;
-	pSMB->ParameterOffset = 0;
-	pSMB->DataCount = 0;
-	pSMB->DataOffset = 0;
-	pSMB->SetupCount = 4; /* single byte does not need le conversion */
-	pSMB->SubCommand = cpu_to_le16(NT_TRANSACT_NOTIFY_CHANGE);
-	pSMB->ParameterCount = pSMB->TotalParameterCount;
-	if (notify_subdirs)
-		pSMB->WatchTree = 1; /* one byte - no le conversion needed */
-	pSMB->Reserved2 = 0;
-	pSMB->CompletionFilter = cpu_to_le32(filter);
-	pSMB->Fid = netfid; /* file handle always le */
-	pSMB->ByteCount = 0;
-
-	rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB,
-			 (struct smb_hdr *)pSMBr, &bytes_returned,
-			 CIFS_ASYNC_OP);
-	if (rc) {
-		cifs_dbg(FYI, "Error in Notify = %d\n", rc);
-	} else {
-		/* Add file to outstanding requests */
-		/* BB change to kmem cache alloc */
-		dnotify_req = kmalloc(
-						sizeof(struct dir_notify_req),
-						 GFP_KERNEL);
-		if (dnotify_req) {
-			dnotify_req->Pid = pSMB->hdr.Pid;
-			dnotify_req->PidHigh = pSMB->hdr.PidHigh;
-			dnotify_req->Mid = pSMB->hdr.Mid;
-			dnotify_req->Tid = pSMB->hdr.Tid;
-			dnotify_req->Uid = pSMB->hdr.Uid;
-			dnotify_req->netfid = netfid;
-			dnotify_req->pfile = pfile;
-			dnotify_req->filter = filter;
-			dnotify_req->multishot = multishot;
-			spin_lock(&GlobalMid_Lock);
-			list_add_tail(&dnotify_req->lhead,
-					&GlobalDnotifyReqList);
-			spin_unlock(&GlobalMid_Lock);
-		} else
-			rc = -ENOMEM;
-	}
-	cifs_buf_release(pSMB);
-	return rc;
-}
-#endif /* was needed for dnotify, and will be needed for inotify when VFS fix */

fs/cifs/connect.c

@@ -323,8 +323,6 @@ static int ip_connect(struct TCP_Server_Info *server);
 static int generic_ip_connect(struct TCP_Server_Info *server);
 static void tlink_rb_insert(struct rb_root *root, struct tcon_link *new_tlink);
 static void cifs_prune_tlinks(struct work_struct *work);
-static int cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data,
-					const char *devname, bool is_smb3);
 static char *extract_hostname(const char *unc);
 
 /*
@@ -530,21 +528,6 @@ cifs_reconnect(struct TCP_Server_Info *server)
 	/* do not want to be sending data on a socket we are freeing */
 	cifs_dbg(FYI, "%s: tearing down socket\n", __func__);
 	mutex_lock(&server->srv_mutex);
-	if (server->ssocket) {
-		cifs_dbg(FYI, "State: 0x%x Flags: 0x%lx\n",
-			 server->ssocket->state, server->ssocket->flags);
-		kernel_sock_shutdown(server->ssocket, SHUT_WR);
-		cifs_dbg(FYI, "Post shutdown state: 0x%x Flags: 0x%lx\n",
-			 server->ssocket->state, server->ssocket->flags);
-		sock_release(server->ssocket);
-		server->ssocket = NULL;
-	}
-	server->sequence_number = 0;
-	server->session_estab = false;
-	kfree(server->session_key.response);
-	server->session_key.response = NULL;
-	server->session_key.len = 0;
-	server->lstrp = jiffies;
 
 	/* mark submitted MIDs for retry and issue callback */
 	INIT_LIST_HEAD(&retry_list);
@@ -557,7 +540,6 @@ cifs_reconnect(struct TCP_Server_Info *server)
 		list_move(&mid_entry->qhead, &retry_list);
 	}
 	spin_unlock(&GlobalMid_Lock);
-	mutex_unlock(&server->srv_mutex);
 
 	cifs_dbg(FYI, "%s: issuing mid callbacks\n", __func__);
 	list_for_each_safe(tmp, tmp2, &retry_list) {
@@ -566,6 +548,25 @@ cifs_reconnect(struct TCP_Server_Info *server)
 		mid_entry->callback(mid_entry);
 	}
 
+	if (server->ssocket) {
+		cifs_dbg(FYI, "State: 0x%x Flags: 0x%lx\n",
+			 server->ssocket->state, server->ssocket->flags);
+		kernel_sock_shutdown(server->ssocket, SHUT_WR);
+		cifs_dbg(FYI, "Post shutdown state: 0x%x Flags: 0x%lx\n",
+			 server->ssocket->state, server->ssocket->flags);
+		sock_release(server->ssocket);
+		server->ssocket = NULL;
+	} else if (cifs_rdma_enabled(server))
+		smbd_destroy(server);
+	server->sequence_number = 0;
+	server->session_estab = false;
+	kfree(server->session_key.response);
+	server->session_key.response = NULL;
+	server->session_key.len = 0;
+	server->lstrp = jiffies;
+
+	mutex_unlock(&server->srv_mutex);
+
 	do {
 		try_to_freeze();
 
@@ -931,10 +932,8 @@ static void clean_demultiplex_info(struct TCP_Server_Info *server)
 	wake_up_all(&server->request_q);
 	/* give those requests time to exit */
 	msleep(125);
-	if (cifs_rdma_enabled(server) && server->smbd_conn) {
-		smbd_destroy(server->smbd_conn);
-		server->smbd_conn = NULL;
-	}
+	if (cifs_rdma_enabled(server))
+		smbd_destroy(server);
 	if (server->ssocket) {
 		sock_release(server->ssocket);
 		server->ssocket = NULL;
@@ -2904,8 +2903,7 @@ cifs_find_smb_ses(struct TCP_Server_Info *server, struct smb_vol *vol)
 	return NULL;
 }
 
-static void
-cifs_put_smb_ses(struct cifs_ses *ses)
+void cifs_put_smb_ses(struct cifs_ses *ses)
 {
 	unsigned int rc, xid;
 	struct TCP_Server_Info *server = ses->server;
@@ -3082,7 +3080,7 @@ cifs_set_cifscreds(struct smb_vol *vol __attribute__((unused)),
  * already got a server reference (server refcount +1). See
  * cifs_get_tcon() for refcount explanations.
  */
-static struct cifs_ses *
+struct cifs_ses *
 cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info)
 {
 	int rc = -ENOMEM;
@@ -4389,7 +4387,7 @@ static int mount_do_dfs_failover(const char *path,
 }
 #endif
 
-static int
+int
 cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data,
 			const char *devname, bool is_smb3)
 {
@@ -4543,7 +4541,7 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb_vol *vol)
 	struct cifs_tcon *tcon = NULL;
 	struct TCP_Server_Info *server;
 	char *root_path = NULL, *full_path = NULL;
-	char *old_mountdata;
+	char *old_mountdata, *origin_mountdata = NULL;
 	int count;
 
 	rc = mount_get_conns(vol, cifs_sb, &xid, &server, &ses, &tcon);
@@ -4602,6 +4600,14 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb_vol *vol)
 		goto error;
 	}
 
+	/* Save DFS root volume information for DFS refresh worker */
+	origin_mountdata = kstrndup(cifs_sb->mountdata,
+				    strlen(cifs_sb->mountdata), GFP_KERNEL);
+	if (!origin_mountdata) {
+		rc = -ENOMEM;
+		goto error;
+	}
+
 	if (cifs_sb->mountdata != old_mountdata) {
 		/* If we were redirected, reconnect to new target server */
 		mount_put_conns(cifs_sb, xid, server, ses, tcon);
@@ -4710,7 +4716,7 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb_vol *vol)
 	}
 	spin_unlock(&cifs_tcp_ses_lock);
 
-	rc = dfs_cache_add_vol(vol, cifs_sb->origin_fullpath);
+	rc = dfs_cache_add_vol(origin_mountdata, vol, cifs_sb->origin_fullpath);
 	if (rc) {
 		kfree(cifs_sb->origin_fullpath);
 		goto error;
@@ -4728,6 +4734,7 @@ int cifs_mount(struct cifs_sb_info *cifs_sb, struct smb_vol *vol)
 error:
 	kfree(full_path);
 	kfree(root_path);
+	kfree(origin_mountdata);
 	mount_put_conns(cifs_sb, xid, server, ses, tcon);
 	return rc;
 }

fs/cifs/dfs_cache.c

@@ -2,7 +2,7 @@
 /*
  * DFS referral cache routines
  *
- * Copyright (c) 2018 Paulo Alcantara <palcantara@suse.de>
+ * Copyright (c) 2018-2019 Paulo Alcantara <palcantara@suse.de>
  */
 
 #include <linux/rcupdate.h>
@@ -52,6 +52,7 @@ static struct kmem_cache *dfs_cache_slab __read_mostly;
 struct dfs_cache_vol_info {
 	char *vi_fullpath;
 	struct smb_vol vi_vol;
+	char *vi_mntdata;
 	struct list_head vi_list;
 };
 
@@ -529,6 +530,7 @@ static inline void free_vol(struct dfs_cache_vol_info *vi)
 {
 	list_del(&vi->vi_list);
 	kfree(vi->vi_fullpath);
+	kfree(vi->vi_mntdata);
 	cifs_cleanup_volume_info_contents(&vi->vi_vol);
 	kfree(vi);
 }
@@ -1139,17 +1141,18 @@ static int dup_vol(struct smb_vol *vol, struct smb_vol *new)
  * dfs_cache_add_vol - add a cifs volume during mount() that will be handled by
  * DFS cache refresh worker.
  *
+ * @mntdata: mount data.
  * @vol: cifs volume.
  * @fullpath: origin full path.
  *
  * Return zero if volume was set up correctly, otherwise non-zero.
  */
-int dfs_cache_add_vol(struct smb_vol *vol, const char *fullpath)
+int dfs_cache_add_vol(char *mntdata, struct smb_vol *vol, const char *fullpath)
 {
 	int rc;
 	struct dfs_cache_vol_info *vi;
 
-	if (!vol || !fullpath)
+	if (!vol || !fullpath || !mntdata)
 		return -EINVAL;
 
 	cifs_dbg(FYI, "%s: fullpath: %s\n", __func__, fullpath);
@@ -1168,6 +1171,8 @@ int dfs_cache_add_vol(struct smb_vol *vol, const char *fullpath)
 	if (rc)
 		goto err_free_fullpath;
 
+	vi->vi_mntdata = mntdata;
+
 	mutex_lock(&dfs_cache.dc_lock);
 	list_add_tail(&vi->vi_list, &dfs_cache.dc_vol_list);
 	mutex_unlock(&dfs_cache.dc_lock);
@@ -1275,8 +1280,102 @@ static void get_tcons(struct TCP_Server_Info *server, struct list_head *head)
 	spin_unlock(&cifs_tcp_ses_lock);
 }
 
+static inline bool is_dfs_link(const char *path)
+{
+	char *s;
+
+	s = strchr(path + 1, '\\');
+	if (!s)
+		return false;
+	return !!strchr(s + 1, '\\');
+}
+
+static inline char *get_dfs_root(const char *path)
+{
+	char *s, *npath;
+
+	s = strchr(path + 1, '\\');
+	if (!s)
+		return ERR_PTR(-EINVAL);
+
+	s = strchr(s + 1, '\\');
+	if (!s)
+		return ERR_PTR(-EINVAL);
+
+	npath = kstrndup(path, s - path, GFP_KERNEL);
+	if (!npath)
+		return ERR_PTR(-ENOMEM);
+
+	return npath;
+}
+
+/* Find root SMB session out of a DFS link path */
+static struct cifs_ses *find_root_ses(struct dfs_cache_vol_info *vi,
+				      struct cifs_tcon *tcon, const char *path)
+{
+	char *rpath;
+	int rc;
+	struct dfs_info3_param ref = {0};
+	char *mdata = NULL, *devname = NULL;
+	bool is_smb3 = tcon->ses->server->vals->header_preamble_size == 0;
+	struct TCP_Server_Info *server;
+	struct cifs_ses *ses;
+	struct smb_vol vol;
+
+	rpath = get_dfs_root(path);
+	if (IS_ERR(rpath))
+		return ERR_CAST(rpath);
+
+	memset(&vol, 0, sizeof(vol));
+
+	rc = dfs_cache_noreq_find(rpath, &ref, NULL);
+	if (rc) {
+		ses = ERR_PTR(rc);
+		goto out;
+	}
+
+	mdata = cifs_compose_mount_options(vi->vi_mntdata, rpath, &ref,
+					   &devname);
+	free_dfs_info_param(&ref);
+
+	if (IS_ERR(mdata)) {
+		ses = ERR_CAST(mdata);
+		mdata = NULL;
+		goto out;
+	}
+
+	rc = cifs_setup_volume_info(&vol, mdata, devname, is_smb3);
+	kfree(devname);
+
+	if (rc) {
+		ses = ERR_PTR(rc);
+		goto out;
+	}
+
+	server = cifs_find_tcp_session(&vol);
+	if (IS_ERR_OR_NULL(server)) {
+		ses = ERR_PTR(-EHOSTDOWN);
+		goto out;
+	}
+	if (server->tcpStatus != CifsGood) {
+		cifs_put_tcp_session(server, 0);
+		ses = ERR_PTR(-EHOSTDOWN);
+		goto out;
+	}
+
+	ses = cifs_get_smb_ses(server, &vol);
+
+out:
+	cifs_cleanup_volume_info_contents(&vol);
+	kfree(mdata);
+	kfree(rpath);
+
+	return ses;
+}
+
 /* Refresh DFS cache entry from a given tcon */
-static void do_refresh_tcon(struct dfs_cache *dc, struct cifs_tcon *tcon)
+static void do_refresh_tcon(struct dfs_cache *dc, struct dfs_cache_vol_info *vi,
+			    struct cifs_tcon *tcon)
 {
 	int rc = 0;
 	unsigned int xid;
@@ -1285,6 +1384,7 @@ static void do_refresh_tcon(struct dfs_cache *dc, struct cifs_tcon *tcon)
 	struct dfs_cache_entry *ce;
 	struct dfs_info3_param *refs = NULL;
 	int numrefs = 0;
+	struct cifs_ses *root_ses = NULL, *ses;
 
 	xid = get_xid();
 
@@ -1306,13 +1406,24 @@ static void do_refresh_tcon(struct dfs_cache *dc, struct cifs_tcon *tcon)
 	if (!cache_entry_expired(ce))
 		goto out;
 
-	if (unlikely(!tcon->ses->server->ops->get_dfs_refer)) {
+	/* If it's a DFS Link, then use root SMB session for refreshing it */
+	if (is_dfs_link(npath)) {
+		ses = root_ses = find_root_ses(vi, tcon, npath);
+		if (IS_ERR(ses)) {
+			rc = PTR_ERR(ses);
+			root_ses = NULL;
+			goto out;
+		}
+	} else {
+		ses = tcon->ses;
+	}
+
+	if (unlikely(!ses->server->ops->get_dfs_refer)) {
 		rc = -EOPNOTSUPP;
 	} else {
-		rc = tcon->ses->server->ops->get_dfs_refer(xid, tcon->ses, path,
-							   &refs, &numrefs,
-							   dc->dc_nlsc,
-							   tcon->remap);
+		rc = ses->server->ops->get_dfs_refer(xid, ses, path, &refs,
+						     &numrefs, dc->dc_nlsc,
+						     tcon->remap);
 		if (!rc) {
 			mutex_lock(&dfs_cache_list_lock);
 			ce = __update_cache_entry(npath, refs, numrefs);
@@ -1323,9 +1434,11 @@ static void do_refresh_tcon(struct dfs_cache *dc, struct cifs_tcon *tcon)
 				rc = PTR_ERR(ce);
 		}
 	}
-	if (rc)
-		cifs_dbg(FYI, "%s: failed to update expired entry\n", __func__);
+
 out:
+	if (root_ses)
+		cifs_put_smb_ses(root_ses);
+
 	free_xid(xid);
 	free_normalized_path(path, npath);
 }
@@ -1333,9 +1446,6 @@ static void do_refresh_tcon(struct dfs_cache *dc, struct cifs_tcon *tcon)
 /*
  * Worker that will refresh DFS cache based on lowest TTL value from a DFS
  * referral.
- *
- * FIXME: ensure that all requests are sent to DFS root for refreshing the
- * cache.
  */
 static void refresh_cache_worker(struct work_struct *work)
 {
@@ -1356,7 +1466,7 @@ static void refresh_cache_worker(struct work_struct *work)
 			goto next;
 		get_tcons(server, &list);
 		list_for_each_entry_safe(tcon, ntcon, &list, ulist) {
-			do_refresh_tcon(dc, tcon);
+			do_refresh_tcon(dc, vi, tcon);
 			list_del_init(&tcon->ulist);
 			cifs_put_tcon(tcon);
 		}

fs/cifs/dfs_cache.h

@@ -2,7 +2,7 @@
 /*
  * DFS referral cache routines
  *
- * Copyright (c) 2018 Paulo Alcantara <palcantara@suse.de>
+ * Copyright (c) 2018-2019 Paulo Alcantara <palcantara@suse.de>
  */
 
 #ifndef _CIFS_DFS_CACHE_H
@@ -43,7 +43,8 @@ dfs_cache_noreq_update_tgthint(const char *path,
 extern int dfs_cache_get_tgt_referral(const char *path,
 				      const struct dfs_cache_tgt_iterator *it,
 				      struct dfs_info3_param *ref);
-extern int dfs_cache_add_vol(struct smb_vol *vol, const char *fullpath);
+extern int dfs_cache_add_vol(char *mntdata, struct smb_vol *vol,
+			const char *fullpath);
 extern int dfs_cache_update_vol(const char *fullpath,
 				struct TCP_Server_Info *server);
 extern void dfs_cache_del_vol(const char *fullpath);

fs/cifs/file.c

@@ -2443,7 +2443,6 @@ int cifs_strict_fsync(struct file *file, loff_t start, loff_t end,
 	rc = file_write_and_wait_range(file, start, end);
 	if (rc)
 		return rc;
-	inode_lock(inode);
 
 	xid = get_xid();
 
@@ -2468,7 +2467,6 @@ int cifs_strict_fsync(struct file *file, loff_t start, loff_t end,
 	}
 
 	free_xid(xid);
-	inode_unlock(inode);
 	return rc;
 }
 
@@ -2480,12 +2478,10 @@ int cifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 	struct TCP_Server_Info *server;
 	struct cifsFileInfo *smbfile = file->private_data;
 	struct cifs_sb_info *cifs_sb = CIFS_FILE_SB(file);
-	struct inode *inode = file->f_mapping->host;
 
 	rc = file_write_and_wait_range(file, start, end);
 	if (rc)
 		return rc;
-	inode_lock(inode);
 
 	xid = get_xid();
 
@@ -2502,7 +2498,6 @@ int cifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 	}
 
 	free_xid(xid);
-	inode_unlock(inode);
 	return rc;
 }
 

fs/cifs/inode.c

@@ -2116,6 +2116,43 @@ int cifs_getattr(const struct path *path, struct kstat *stat,
 	return rc;
 }
 
+int cifs_fiemap(struct inode *inode, struct fiemap_extent_info *fei, u64 start,
+		u64 len)
+{
+	struct cifsInodeInfo *cifs_i = CIFS_I(inode);
+	struct cifs_sb_info *cifs_sb = CIFS_SB(cifs_i->vfs_inode.i_sb);
+	struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb);
+	struct TCP_Server_Info *server = tcon->ses->server;
+	struct cifsFileInfo *cfile;
+	int rc;
+
+	/*
+	 * We need to be sure that all dirty pages are written as they
+	 * might fill holes on the server.
+	 */
+	if (!CIFS_CACHE_READ(CIFS_I(inode)) && inode->i_mapping &&
+	    inode->i_mapping->nrpages != 0) {
+		rc = filemap_fdatawait(inode->i_mapping);
+		if (rc) {
+			mapping_set_error(inode->i_mapping, rc);
+			return rc;
+		}
+	}
+
+	cfile = find_readable_file(cifs_i, false);
+	if (cfile == NULL)
+		return -EINVAL;
+
+	if (server->ops->fiemap) {
+		rc = server->ops->fiemap(tcon, cfile, fei, start, len);
+		cifsFileInfo_put(cfile);
+		return rc;
+	}
+
+	cifsFileInfo_put(cfile);
+	return -ENOTSUPP;
+}
+
 static int cifs_truncate_page(struct address_space *mapping, loff_t from)
 {
 	pgoff_t index = from >> PAGE_SHIFT;

fs/cifs/link.c

@@ -648,9 +648,16 @@ cifs_get_link(struct dentry *direntry, struct inode *inode,
 		rc = query_mf_symlink(xid, tcon, cifs_sb, full_path,
 				      &target_path);
 
-	if (rc != 0 && server->ops->query_symlink)
-		rc = server->ops->query_symlink(xid, tcon, full_path,
-						&target_path, cifs_sb);
+	if (rc != 0 && server->ops->query_symlink) {
+		struct cifsInodeInfo *cifsi = CIFS_I(inode);
+		bool reparse_point = false;
+
+		if (cifsi->cifsAttrs & ATTR_REPARSE)
+			reparse_point = true;
+
+		rc = server->ops->query_symlink(xid, tcon, cifs_sb, full_path,
+						&target_path, reparse_point);
+	}
 
 	kfree(full_path);
 	free_xid(xid);

fs/cifs/smb1ops.c

@@ -950,8 +950,8 @@ cifs_unix_dfs_readlink(const unsigned int xid, struct cifs_tcon *tcon,
 
 static int
 cifs_query_symlink(const unsigned int xid, struct cifs_tcon *tcon,
-		   const char *full_path, char **target_path,
-		   struct cifs_sb_info *cifs_sb)
+		   struct cifs_sb_info *cifs_sb, const char *full_path,
+		   char **target_path, bool is_reparse_point)
 {
 	int rc;
 	int oplock = 0;
@@ -960,6 +960,11 @@ cifs_query_symlink(const unsigned int xid, struct cifs_tcon *tcon,
 
 	cifs_dbg(FYI, "%s: path: %s\n", __func__, full_path);
 
+	if (is_reparse_point) {
+		cifs_dbg(VFS, "reparse points not handled for SMB1 symlinks\n");
+		return -EOPNOTSUPP;
+	}
+
 	/* Check for unix extensions */
 	if (cap_unix(tcon->ses)) {
 		rc = CIFSSMBUnixQuerySymLink(xid, tcon, full_path, target_path,

fs/cifs/smb2ops.c

@@ -1382,6 +1382,26 @@ smb2_ioctl_query_info(const unsigned int xid,
 	oparms.fid = &fid;
 	oparms.reconnect = false;
 
+	/*
+	 * FSCTL codes encode the special access they need in the fsctl code.
+	 */
+	if (qi.flags & PASSTHRU_FSCTL) {
+		switch (qi.info_type & FSCTL_DEVICE_ACCESS_MASK) {
+		case FSCTL_DEVICE_ACCESS_FILE_READ_WRITE_ACCESS:
+			oparms.desired_access = FILE_READ_DATA | FILE_WRITE_DATA | FILE_READ_ATTRIBUTES | SYNCHRONIZE;
+			break;
+		case FSCTL_DEVICE_ACCESS_FILE_ANY_ACCESS:
+			oparms.desired_access = GENERIC_ALL;
+			break;
+		case FSCTL_DEVICE_ACCESS_FILE_READ_ACCESS:
+			oparms.desired_access = GENERIC_READ;
+			break;
+		case FSCTL_DEVICE_ACCESS_FILE_WRITE_ACCESS:
+			oparms.desired_access = GENERIC_WRITE;
+			break;
+		}
+	}
+
 	rc = SMB2_open_init(tcon, &rqst[0], &oplock, &oparms, path);
 	if (rc)
 		goto iqinf_exit;
@@ -1399,8 +1419,9 @@ smb2_ioctl_query_info(const unsigned int xid,
 
 			rc = SMB2_ioctl_init(tcon, &rqst[1],
 					     COMPOUND_FID, COMPOUND_FID,
-					     qi.info_type, true, NULL,
-					     0, CIFSMaxBufSize);
+					     qi.info_type, true, buffer,
+					     qi.output_buffer_length,
+					     CIFSMaxBufSize);
 		}
 	} else if (qi.flags == PASSTHRU_QUERY_INFO) {
 		memset(&qi_iov, 0, sizeof(qi_iov));
@@ -1441,12 +1462,19 @@ smb2_ioctl_query_info(const unsigned int xid,
 		io_rsp = (struct smb2_ioctl_rsp *)rsp_iov[1].iov_base;
 		if (le32_to_cpu(io_rsp->OutputCount) < qi.input_buffer_length)
 			qi.input_buffer_length = le32_to_cpu(io_rsp->OutputCount);
+		if (qi.input_buffer_length > 0 &&
+		    le32_to_cpu(io_rsp->OutputOffset) + qi.input_buffer_length > rsp_iov[1].iov_len) {
+			rc = -EFAULT;
+			goto iqinf_exit;
+		}
 		if (copy_to_user(&pqi->input_buffer_length, &qi.input_buffer_length,
 				 sizeof(qi.input_buffer_length))) {
 			rc = -EFAULT;
 			goto iqinf_exit;
 		}
-		if (copy_to_user(pqi + 1, &io_rsp[1], qi.input_buffer_length)) {
+		if (copy_to_user((void __user *)pqi + sizeof(struct smb_query_info),
+				 (const void *)io_rsp + le32_to_cpu(io_rsp->OutputOffset),
+				 qi.input_buffer_length)) {
 			rc = -EFAULT;
 			goto iqinf_exit;
 		}
@@ -1821,6 +1849,14 @@ smb3_enum_snapshots(const unsigned int xid, struct cifs_tcon *tcon,
 	u32 max_response_size;
 	struct smb_snapshot_array snapshot_in;
 
+	/*
+	 * On the first query to enumerate the list of snapshots available
+	 * for this volume the buffer begins with 0 (number of snapshots
+	 * which can be returned is zero since at that point we do not know
+	 * how big the buffer needs to be). On the second query,
+	 * it (ret_data_len) is set to number of snapshots so we can
+	 * know to set the maximum response size larger (see below).
+	 */
 	if (get_user(ret_data_len, (unsigned int __user *)ioc_buf))
 		return -EFAULT;
 
@@ -2354,46 +2390,129 @@ smb2_get_dfs_refer(const unsigned int xid, struct cifs_ses *ses,
 
 static int
 smb2_query_symlink(const unsigned int xid, struct cifs_tcon *tcon,
-		   const char *full_path, char **target_path,
-		   struct cifs_sb_info *cifs_sb)
+		   struct cifs_sb_info *cifs_sb, const char *full_path,
+		   char **target_path, bool is_reparse_point)
 {
 	int rc;
-	__le16 *utf16_path;
+	__le16 *utf16_path = NULL;
 	__u8 oplock = SMB2_OPLOCK_LEVEL_NONE;
 	struct cifs_open_parms oparms;
 	struct cifs_fid fid;
 	struct kvec err_iov = {NULL, 0};
 	struct smb2_err_rsp *err_buf = NULL;
-	int resp_buftype;
 	struct smb2_symlink_err_rsp *symlink;
 	unsigned int sub_len;
 	unsigned int sub_offset;
 	unsigned int print_len;
 	unsigned int print_offset;
+	int flags = 0;
+	struct smb_rqst rqst[3];
+	int resp_buftype[3];
+	struct kvec rsp_iov[3];
+	struct kvec open_iov[SMB2_CREATE_IOV_SIZE];
+	struct kvec io_iov[SMB2_IOCTL_IOV_SIZE];
+	struct kvec close_iov[1];
+	struct smb2_create_rsp *create_rsp;
+	struct smb2_ioctl_rsp *ioctl_rsp;
+	char *ioctl_buf;
+	u32 plen;
 
 	cifs_dbg(FYI, "%s: path: %s\n", __func__, full_path);
 
+	if (smb3_encryption_required(tcon))
+		flags |= CIFS_TRANSFORM_REQ;
+
+	memset(rqst, 0, sizeof(rqst));
+	resp_buftype[0] = resp_buftype[1] = resp_buftype[2] = CIFS_NO_BUFFER;
+	memset(rsp_iov, 0, sizeof(rsp_iov));
+
 	utf16_path = cifs_convert_path_to_utf16(full_path, cifs_sb);
 	if (!utf16_path)
 		return -ENOMEM;
 
+	/* Open */
+	memset(&open_iov, 0, sizeof(open_iov));
+	rqst[0].rq_iov = open_iov;
+	rqst[0].rq_nvec = SMB2_CREATE_IOV_SIZE;
+
+	memset(&oparms, 0, sizeof(oparms));
 	oparms.tcon = tcon;
 	oparms.desired_access = FILE_READ_ATTRIBUTES;
 	oparms.disposition = FILE_OPEN;
+
 	if (backup_cred(cifs_sb))
 		oparms.create_options = CREATE_OPEN_BACKUP_INTENT;
 	else
 		oparms.create_options = 0;
+	if (is_reparse_point)
+		oparms.create_options = OPEN_REPARSE_POINT;
+
 	oparms.fid = &fid;
 	oparms.reconnect = false;
 
-	rc = SMB2_open(xid, &oparms, utf16_path, &oplock, NULL, &err_iov,
-		       &resp_buftype);
-	if (!rc)
-		SMB2_close(xid, tcon, fid.persistent_fid, fid.volatile_fid);
+	rc = SMB2_open_init(tcon, &rqst[0], &oplock, &oparms, utf16_path);
+	if (rc)
+		goto querty_exit;
+	smb2_set_next_command(tcon, &rqst[0]);
+
+
+	/* IOCTL */
+	memset(&io_iov, 0, sizeof(io_iov));
+	rqst[1].rq_iov = io_iov;
+	rqst[1].rq_nvec = SMB2_IOCTL_IOV_SIZE;
+
+	rc = SMB2_ioctl_init(tcon, &rqst[1], fid.persistent_fid,
+			     fid.volatile_fid, FSCTL_GET_REPARSE_POINT,
+			     true /* is_fctl */, NULL, 0, CIFSMaxBufSize);
+	if (rc)
+		goto querty_exit;
+
+	smb2_set_next_command(tcon, &rqst[1]);
+	smb2_set_related(&rqst[1]);
+
+
+	/* Close */
+	memset(&close_iov, 0, sizeof(close_iov));
+	rqst[2].rq_iov = close_iov;
+	rqst[2].rq_nvec = 1;
+
+	rc = SMB2_close_init(tcon, &rqst[2], COMPOUND_FID, COMPOUND_FID);
+	if (rc)
+		goto querty_exit;
+
+	smb2_set_related(&rqst[2]);
+
+	rc = compound_send_recv(xid, tcon->ses, flags, 3, rqst,
+				resp_buftype, rsp_iov);
+
+	create_rsp = rsp_iov[0].iov_base;
+	if (create_rsp && create_rsp->sync_hdr.Status)
+		err_iov = rsp_iov[0];
+	ioctl_rsp = rsp_iov[1].iov_base;
+
+	/*
+	 * Open was successful and we got an ioctl response.
+	 */
+	if ((rc == 0) && (is_reparse_point)) {
+		/* See MS-FSCC 2.3.23 */
+
+		ioctl_buf = (char *)ioctl_rsp + le32_to_cpu(ioctl_rsp->OutputOffset);
+		plen = le32_to_cpu(ioctl_rsp->OutputCount);
+
+		if (plen + le32_to_cpu(ioctl_rsp->OutputOffset) >
+		    rsp_iov[1].iov_len) {
+			cifs_dbg(VFS, "srv returned invalid ioctl length: %d\n", plen);
+			rc = -EIO;
+			goto querty_exit;
+		}
+
+		/* Do stuff with ioctl_buf/plen */
+		goto querty_exit;
+	}
+
 	if (!rc || !err_iov.iov_base) {
 		rc = -ENOENT;
-		goto free_path;
+		goto querty_exit;
 	}
 
 	err_buf = err_iov.iov_base;
@@ -2433,9 +2552,14 @@ smb2_query_symlink(const unsigned int xid, struct cifs_tcon *tcon,
 	cifs_dbg(FYI, "%s: target path: %s\n", __func__, *target_path);
 
  querty_exit:
-	free_rsp_buf(resp_buftype, err_buf);
- free_path:
+	cifs_dbg(FYI, "query symlink rc %d\n", rc);
 	kfree(utf16_path);
+	SMB2_open_free(&rqst[0]);
+	SMB2_ioctl_free(&rqst[1]);
+	SMB2_close_free(&rqst[2]);
+	free_rsp_buf(resp_buftype[0], rsp_iov[0].iov_base);
+	free_rsp_buf(resp_buftype[1], rsp_iov[1].iov_base);
+	free_rsp_buf(resp_buftype[2], rsp_iov[2].iov_base);
 	return rc;
 }
 
@@ -2612,16 +2736,8 @@ static long smb3_zero_range(struct file *file, struct cifs_tcon *tcon,
 	struct cifsInodeInfo *cifsi;
 	struct cifsFileInfo *cfile = file->private_data;
 	struct file_zero_data_information fsctl_buf;
-	struct smb_rqst rqst[2];
-	int resp_buftype[2];
-	struct kvec rsp_iov[2];
-	struct kvec io_iov[SMB2_IOCTL_IOV_SIZE];
-	struct kvec si_iov[1];
-	unsigned int size[1];
-	void *data[1];
 	long rc;
 	unsigned int xid;
-	int num = 0, flags = 0;
 	__le64 eof;
 
 	xid = get_xid();
@@ -2643,39 +2759,16 @@ static long smb3_zero_range(struct file *file, struct cifs_tcon *tcon,
 			return rc;
 		}
 
-	/*
-	 * Must check if file sparse since fallocate -z (zero range) assumes
-	 * non-sparse allocation
-	 */
-	if (!(cifsi->cifsAttrs & FILE_ATTRIBUTE_SPARSE_FILE)) {
-		rc = -EOPNOTSUPP;
-		trace_smb3_zero_err(xid, cfile->fid.persistent_fid, tcon->tid,
-			      ses->Suid, offset, len, rc);
-		free_xid(xid);
-		return rc;
-	}
-
 	cifs_dbg(FYI, "offset %lld len %lld", offset, len);
 
 	fsctl_buf.FileOffset = cpu_to_le64(offset);
 	fsctl_buf.BeyondFinalZero = cpu_to_le64(offset + len);
 
-	if (smb3_encryption_required(tcon))
-		flags |= CIFS_TRANSFORM_REQ;
-
-	memset(rqst, 0, sizeof(rqst));
-	resp_buftype[0] = resp_buftype[1] = CIFS_NO_BUFFER;
-	memset(rsp_iov, 0, sizeof(rsp_iov));
-
-
-	memset(&io_iov, 0, sizeof(io_iov));
-	rqst[num].rq_iov = io_iov;
-	rqst[num].rq_nvec = SMB2_IOCTL_IOV_SIZE;
-	rc = SMB2_ioctl_init(tcon, &rqst[num++], cfile->fid.persistent_fid,
-			     cfile->fid.volatile_fid, FSCTL_SET_ZERO_DATA,
-			     true /* is_fctl */, (char *)&fsctl_buf,
-			     sizeof(struct file_zero_data_information),
-			     CIFSMaxBufSize);
+	rc = SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid,
+			cfile->fid.volatile_fid, FSCTL_SET_ZERO_DATA, true,
+			(char *)&fsctl_buf,
+			sizeof(struct file_zero_data_information),
+			0, NULL, NULL);
 	if (rc)
 		goto zero_range_exit;
 
@@ -2683,33 +2776,12 @@ static long smb3_zero_range(struct file *file, struct cifs_tcon *tcon,
 	 * do we also need to change the size of the file?
 	 */
 	if (keep_size == false && i_size_read(inode) < offset + len) {
-		smb2_set_next_command(tcon, &rqst[0]);
-
-		memset(&si_iov, 0, sizeof(si_iov));
-		rqst[num].rq_iov = si_iov;
-		rqst[num].rq_nvec = 1;
-
 		eof = cpu_to_le64(offset + len);
-		size[0] = 8; /* sizeof __le64 */
-		data[0] = &eof;
-
-		rc = SMB2_set_info_init(tcon, &rqst[num++],
-					cfile->fid.persistent_fid,
-					cfile->fid.persistent_fid,
-					current->tgid,
-					FILE_END_OF_FILE_INFORMATION,
-					SMB2_O_INFO_FILE, 0, data, size);
-		smb2_set_related(&rqst[1]);
+		rc = SMB2_set_eof(xid, tcon, cfile->fid.persistent_fid,
+				  cfile->fid.volatile_fid, cfile->pid, &eof);
 	}
 
-	rc = compound_send_recv(xid, ses, flags, num, rqst,
-				resp_buftype, rsp_iov);
-
  zero_range_exit:
-	SMB2_ioctl_free(&rqst[0]);
-	SMB2_set_info_free(&rqst[1]);
-	free_rsp_buf(resp_buftype[0], rsp_iov[0].iov_base);
-	free_rsp_buf(resp_buftype[1], rsp_iov[1].iov_base);
 	free_xid(xid);
 	if (rc)
 		trace_smb3_zero_err(xid, cfile->fid.persistent_fid, tcon->tid,
@@ -2850,6 +2922,79 @@ static long smb3_simple_falloc(struct file *file, struct cifs_tcon *tcon,
 	return rc;
 }
 
+static int smb3_fiemap(struct cifs_tcon *tcon,
+		       struct cifsFileInfo *cfile,
+		       struct fiemap_extent_info *fei, u64 start, u64 len)
+{
+	unsigned int xid;
+	struct file_allocated_range_buffer in_data, *out_data;
+	u32 out_data_len;
+	int i, num, rc, flags, last_blob;
+	u64 next;
+
+	if (fiemap_check_flags(fei, FIEMAP_FLAG_SYNC))
+		return -EBADR;
+
+	xid = get_xid();
+ again:
+	in_data.file_offset = cpu_to_le64(start);
+	in_data.length = cpu_to_le64(len);
+
+	rc = SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid,
+			cfile->fid.volatile_fid,
+			FSCTL_QUERY_ALLOCATED_RANGES, true,
+			(char *)&in_data, sizeof(in_data),
+			1024 * sizeof(struct file_allocated_range_buffer),
+			(char **)&out_data, &out_data_len);
+	if (rc == -E2BIG) {
+		last_blob = 0;
+		rc = 0;
+	} else
+		last_blob = 1;
+	if (rc)
+		goto out;
+
+	if (out_data_len < sizeof(struct file_allocated_range_buffer)) {
+		rc = -EINVAL;
+		goto out;
+	}
+	if (out_data_len % sizeof(struct file_allocated_range_buffer)) {
+		rc = -EINVAL;
+		goto out;
+	}
+
+	num = out_data_len / sizeof(struct file_allocated_range_buffer);
+	for (i = 0; i < num; i++) {
+		flags = 0;
+		if (i == num - 1 && last_blob)
+			flags |= FIEMAP_EXTENT_LAST;
+
+		rc = fiemap_fill_next_extent(fei,
+				le64_to_cpu(out_data[i].file_offset),
+				le64_to_cpu(out_data[i].file_offset),
+				le64_to_cpu(out_data[i].length),
+				flags);
+		if (rc < 0)
+			goto out;
+		if (rc == 1) {
+			rc = 0;
+			goto out;
+		}
+	}
+
+	if (!last_blob) {
+		next = le64_to_cpu(out_data[num - 1].file_offset) +
+		  le64_to_cpu(out_data[num - 1].length);
+		len = len - (next - start);
+		start = next;
+		goto again;
+	}
+
+ out:
+	free_xid(xid);
+	kfree(out_data);
+	return rc;
+}
 
 static long smb3_fallocate(struct file *file, struct cifs_tcon *tcon, int mode,
 			   loff_t off, loff_t len)
@@ -2917,26 +3062,28 @@ smb21_set_oplock_level(struct cifsInodeInfo *cinode, __u32 oplock,
 		       unsigned int epoch, bool *purge_cache)
 {
 	char message[5] = {0};
+	unsigned int new_oplock = 0;
 
 	oplock &= 0xFF;
 	if (oplock == SMB2_OPLOCK_LEVEL_NOCHANGE)
 		return;
 
-	cinode->oplock = 0;
 	if (oplock & SMB2_LEASE_READ_CACHING_HE) {
-		cinode->oplock |= CIFS_CACHE_READ_FLG;
+		new_oplock |= CIFS_CACHE_READ_FLG;
 		strcat(message, "R");
 	}
 	if (oplock & SMB2_LEASE_HANDLE_CACHING_HE) {
-		cinode->oplock |= CIFS_CACHE_HANDLE_FLG;
+		new_oplock |= CIFS_CACHE_HANDLE_FLG;
 		strcat(message, "H");
 	}
 	if (oplock & SMB2_LEASE_WRITE_CACHING_HE) {
-		cinode->oplock |= CIFS_CACHE_WRITE_FLG;
+		new_oplock |= CIFS_CACHE_WRITE_FLG;
 		strcat(message, "W");
 	}
-	if (!cinode->oplock)
-		strcat(message, "None");
+	if (!new_oplock)
+		strncpy(message, "None", sizeof(message));
+
+	cinode->oplock = new_oplock;
 	cifs_dbg(FYI, "%s Lease granted on inode %p\n", message,
 		 &cinode->vfs_inode);
 }
@@ -4018,6 +4165,7 @@ struct smb_version_operations smb20_operations = {
 	.next_header = smb2_next_header,
 	.ioctl_query_info = smb2_ioctl_query_info,
 	.make_node = smb2_make_node,
+	.fiemap = smb3_fiemap,
 };
 
 struct smb_version_operations smb21_operations = {
@@ -4117,6 +4265,7 @@ struct smb_version_operations smb21_operations = {
 	.next_header = smb2_next_header,
 	.ioctl_query_info = smb2_ioctl_query_info,
 	.make_node = smb2_make_node,
+	.fiemap = smb3_fiemap,
 };
 
 struct smb_version_operations smb30_operations = {
@@ -4225,6 +4374,7 @@ struct smb_version_operations smb30_operations = {
 	.next_header = smb2_next_header,
 	.ioctl_query_info = smb2_ioctl_query_info,
 	.make_node = smb2_make_node,
+	.fiemap = smb3_fiemap,
 };
 
 struct smb_version_operations smb311_operations = {
@@ -4334,6 +4484,7 @@ struct smb_version_operations smb311_operations = {
 	.next_header = smb2_next_header,
 	.ioctl_query_info = smb2_ioctl_query_info,
 	.make_node = smb2_make_node,
+	.fiemap = smb3_fiemap,
 };
 
 struct smb_version_values smb20_values = {

fs/cifs/smb2pdu.c

@@ -459,10 +459,7 @@ smb2_plain_req_init(__le16 smb2_command, struct cifs_tcon *tcon,
 	return rc;
 }
 
-
-#define SMB2_PREAUTH_INTEGRITY_CAPABILITIES	cpu_to_le16(1)
-#define SMB2_ENCRYPTION_CAPABILITIES		cpu_to_le16(2)
-#define SMB2_POSIX_EXTENSIONS_AVAILABLE		cpu_to_le16(0x100)
+/* For explanation of negotiate contexts see MS-SMB2 section 2.2.3.1 */
 
 static void
 build_preauth_ctxt(struct smb2_preauth_neg_context *pneg_ctxt)
@@ -475,6 +472,19 @@ build_preauth_ctxt(struct smb2_preauth_neg_context *pneg_ctxt)
 	pneg_ctxt->HashAlgorithms = SMB2_PREAUTH_INTEGRITY_SHA512;
 }
 
+static void
+build_compression_ctxt(struct smb2_compression_capabilities_context *pneg_ctxt)
+{
+	pneg_ctxt->ContextType = SMB2_COMPRESSION_CAPABILITIES;
+	pneg_ctxt->DataLength =
+		cpu_to_le16(sizeof(struct smb2_compression_capabilities_context)
+			  - sizeof(struct smb2_neg_context));
+	pneg_ctxt->CompressionAlgorithmCount = cpu_to_le16(3);
+	pneg_ctxt->CompressionAlgorithms[0] = SMB3_COMPRESS_LZ77;
+	pneg_ctxt->CompressionAlgorithms[1] = SMB3_COMPRESS_LZ77_HUFF;
+	pneg_ctxt->CompressionAlgorithms[2] = SMB3_COMPRESS_LZNT1;
+}
+
 static void
 build_encrypt_ctxt(struct smb2_encryption_neg_context *pneg_ctxt)
 {
@@ -541,10 +551,17 @@ assemble_neg_contexts(struct smb2_negotiate_req *req,
 	*total_len += ctxt_len;
 	pneg_ctxt += ctxt_len;
 
+	build_compression_ctxt((struct smb2_compression_capabilities_context *)
+				pneg_ctxt);
+	ctxt_len = DIV_ROUND_UP(
+		sizeof(struct smb2_compression_capabilities_context), 8) * 8;
+	*total_len += ctxt_len;
+	pneg_ctxt += ctxt_len;
+
 	build_posix_ctxt((struct smb2_posix_neg_context *)pneg_ctxt);
 	*total_len += sizeof(struct smb2_posix_neg_context);
 
-	req->NegotiateContextCount = cpu_to_le16(3);
+	req->NegotiateContextCount = cpu_to_le16(4);
 }
 
 static void decode_preauth_context(struct smb2_preauth_neg_context *ctxt)
@@ -562,6 +579,27 @@ static void decode_preauth_context(struct smb2_preauth_neg_context *ctxt)
 		printk_once(KERN_WARNING "unknown SMB3 hash algorithm\n");
 }
 
+static void decode_compress_ctx(struct TCP_Server_Info *server,
+			 struct smb2_compression_capabilities_context *ctxt)
+{
+	unsigned int len = le16_to_cpu(ctxt->DataLength);
+
+	/* sizeof compress context is a one element compression capbility struct */
+	if (len < 10) {
+		printk_once(KERN_WARNING "server sent bad compression cntxt\n");
+		return;
+	}
+	if (le16_to_cpu(ctxt->CompressionAlgorithmCount) != 1) {
+		printk_once(KERN_WARNING "illegal SMB3 compress algorithm count\n");
+		return;
+	}
+	if (le16_to_cpu(ctxt->CompressionAlgorithms[0]) > 3) {
+		printk_once(KERN_WARNING "unknown compression algorithm\n");
+		return;
+	}
+	server->compress_algorithm = ctxt->CompressionAlgorithms[0];
+}
+
 static int decode_encrypt_ctx(struct TCP_Server_Info *server,
 			      struct smb2_encryption_neg_context *ctxt)
 {
@@ -626,6 +664,9 @@ static int smb311_decode_neg_context(struct smb2_negotiate_rsp *rsp,
 		else if (pctx->ContextType == SMB2_ENCRYPTION_CAPABILITIES)
 			rc = decode_encrypt_ctx(server,
 				(struct smb2_encryption_neg_context *)pctx);
+		else if (pctx->ContextType == SMB2_COMPRESSION_CAPABILITIES)
+			decode_compress_ctx(server,
+				(struct smb2_compression_capabilities_context *)pctx);
 		else if (pctx->ContextType == SMB2_POSIX_EXTENSIONS_AVAILABLE)
 			server->posix_ext_supported = true;
 		else
@@ -1541,7 +1582,7 @@ SMB2_logoff(const unsigned int xid, struct cifs_ses *ses)
 	else if (server->sign)
 		req->sync_hdr.Flags |= SMB2_FLAGS_SIGNED;
 
-	flags |= CIFS_NO_RESP;
+	flags |= CIFS_NO_RSP_BUF;
 
 	iov[0].iov_base = (char *)req;
 	iov[0].iov_len = total_len;
@@ -1742,7 +1783,7 @@ SMB2_tdis(const unsigned int xid, struct cifs_tcon *tcon)
 	if (smb3_encryption_required(tcon))
 		flags |= CIFS_TRANSFORM_REQ;
 
-	flags |= CIFS_NO_RESP;
+	flags |= CIFS_NO_RSP_BUF;
 
 	iov[0].iov_base = (char *)req;
 	iov[0].iov_len = total_len;
@@ -2625,7 +2666,7 @@ SMB2_ioctl(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_fid,
 		trace_smb3_fsctl_err(xid, persistent_fid, tcon->tid,
 				ses->Suid, 0, opcode, rc);
 
-	if ((rc != 0) && (rc != -EINVAL)) {
+	if ((rc != 0) && (rc != -EINVAL) && (rc != -E2BIG)) {
 		cifs_stats_fail_inc(tcon, SMB2_IOCTL_HE);
 		goto ioctl_exit;
 	} else if (rc == -EINVAL) {
@@ -2634,6 +2675,11 @@ SMB2_ioctl(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_fid,
 			cifs_stats_fail_inc(tcon, SMB2_IOCTL_HE);
 			goto ioctl_exit;
 		}
+	} else if (rc == -E2BIG) {
+		if (opcode != FSCTL_QUERY_ALLOCATED_RANGES) {
+			cifs_stats_fail_inc(tcon, SMB2_IOCTL_HE);
+			goto ioctl_exit;
+		}
 	}
 
 	/* check if caller wants to look at return data or just return rc */
@@ -3223,7 +3269,7 @@ smb2_new_read_req(void **buf, unsigned int *total_len,
 				rdata->nr_pages, rdata->page_offset,
 				rdata->tailsz, true, need_invalidate);
 		if (!rdata->mr)
-			return -ENOBUFS;
+			return -EAGAIN;
 
 		req->Channel = SMB2_CHANNEL_RDMA_V1_INVALIDATE;
 		if (need_invalidate)
@@ -3628,7 +3674,7 @@ smb2_async_writev(struct cifs_writedata *wdata,
 				wdata->nr_pages, wdata->page_offset,
 				wdata->tailsz, false, need_invalidate);
 		if (!wdata->mr) {
-			rc = -ENOBUFS;
+			rc = -EAGAIN;
 			goto async_writev_out;
 		}
 		req->Length = 0;
@@ -4164,7 +4210,7 @@ SMB2_oplock_break(const unsigned int xid, struct cifs_tcon *tcon,
 	req->OplockLevel = oplock_level;
 	req->sync_hdr.CreditRequest = cpu_to_le16(1);
 
-	flags |= CIFS_NO_RESP;
+	flags |= CIFS_NO_RSP_BUF;
 
 	iov[0].iov_base = (char *)req;
 	iov[0].iov_len = total_len;
@@ -4438,7 +4484,7 @@ smb2_lockv(const unsigned int xid, struct cifs_tcon *tcon,
 	struct kvec rsp_iov;
 	int resp_buf_type;
 	unsigned int count;
-	int flags = CIFS_NO_RESP;
+	int flags = CIFS_NO_RSP_BUF;
 	unsigned int total_len;
 
 	cifs_dbg(FYI, "smb2_lockv num lock %d\n", num_lock);
@@ -4531,7 +4577,7 @@ SMB2_lease_break(const unsigned int xid, struct cifs_tcon *tcon,
 	memcpy(req->LeaseKey, lease_key, 16);
 	req->LeaseState = lease_state;
 
-	flags |= CIFS_NO_RESP;
+	flags |= CIFS_NO_RSP_BUF;
 
 	iov[0].iov_base = (char *)req;
 	iov[0].iov_len = total_len;

fs/cifs/smb2pdu.h

@@ -251,6 +251,14 @@ struct smb2_negotiate_req {
 #define SMB2_NT_FIND			0x00100000
 #define SMB2_LARGE_FILES		0x00200000
 
+
+/* Negotiate Contexts - ContextTypes. See MS-SMB2 section 2.2.3.1 for details */
+#define SMB2_PREAUTH_INTEGRITY_CAPABILITIES	cpu_to_le16(1)
+#define SMB2_ENCRYPTION_CAPABILITIES		cpu_to_le16(2)
+#define SMB2_COMPRESSION_CAPABILITIES		cpu_to_le16(3)
+#define SMB2_NETNAME_NEGOTIATE_CONTEXT_ID	cpu_to_le16(5)
+#define SMB2_POSIX_EXTENSIONS_AVAILABLE		cpu_to_le16(0x100)
+
 struct smb2_neg_context {
 	__le16	ContextType;
 	__le16	DataLength;
@@ -288,6 +296,27 @@ struct smb2_encryption_neg_context {
 	__le16	Ciphers[1]; /* Ciphers[0] since only one used now */
 } __packed;
 
+/* See MS-SMB2 2.2.3.1.3 */
+#define SMB3_COMPRESS_NONE	cpu_to_le16(0x0000)
+#define SMB3_COMPRESS_LZNT1	cpu_to_le16(0x0001)
+#define SMB3_COMPRESS_LZ77	cpu_to_le16(0x0002)
+#define SMB3_COMPRESS_LZ77_HUFF	cpu_to_le16(0x0003)
+
+struct smb2_compression_capabilities_context {
+	__le16	ContextType; /* 3 */
+	__le16  DataLength;
+	__u32	Reserved;
+	__le16	CompressionAlgorithmCount;
+	__u16	Padding;
+	__u32	Reserved1;
+	__le16	CompressionAlgorithms[3];
+} __packed;
+
+/*
+ * For smb2_netname_negotiate_context_id See MS-SMB2 2.2.3.1.4.
+ * Its struct simply contains NetName, an array of Unicode characters
+ */
+
 #define POSIX_CTXT_DATA_LEN	16
 struct smb2_posix_neg_context {
 	__le16	ContextType; /* 0x100 */
@@ -842,6 +871,11 @@ struct fsctl_get_integrity_information_rsp {
 	__le32	ClusterSizeInBytes;
 } __packed;
 
+struct file_allocated_range_buffer {
+	__le64	file_offset;
+	__le64	length;
+} __packed;
+
 /* Integrity ChecksumAlgorithm choices for above */
 #define	CHECKSUM_TYPE_NONE	0x0000
 #define	CHECKSUM_TYPE_CRC64	0x0002
@@ -1047,6 +1081,7 @@ struct smb2_flush_rsp {
 
 /* For read request Flags field below, following flag is defined for SMB3.02 */
 #define SMB2_READFLAG_READ_UNBUFFERED	0x01
+#define SMB2_READFLAG_REQUEST_COMPRESSED 0x02 /* See MS-SMB2 2.2.19 */
 
 /* Channel field for read and write: exactly one of following flags can be set*/
 #define SMB2_CHANNEL_NONE	cpu_to_le32(0x00000000)
@@ -1113,6 +1148,42 @@ struct smb2_write_rsp {
 	__u8   Buffer[1];
 } __packed;
 
+/* notify flags */
+#define SMB2_WATCH_TREE			0x0001
+
+/* notify completion filter flags. See MS-FSCC 2.6 and MS-SMB2 2.2.35 */
+#define FILE_NOTIFY_CHANGE_FILE_NAME		0x00000001
+#define FILE_NOTIFY_CHANGE_DIR_NAME		0x00000002
+#define FILE_NOTIFY_CHANGE_ATTRIBUTES		0x00000004
+#define FILE_NOTIFY_CHANGE_SIZE			0x00000008
+#define FILE_NOTIFY_CHANGE_LAST_WRITE		0x00000010
+#define FILE_NOTIFY_CHANGE_LAST_ACCESS		0x00000020
+#define FILE_NOTIFY_CHANGE_CREATION		0x00000040
+#define FILE_NOTIFY_CHANGE_EA			0x00000080
+#define FILE_NOTIFY_CHANGE_SECURITY		0x00000100
+#define FILE_NOTIFY_CHANGE_STREAM_NAME		0x00000200
+#define FILE_NOTIFY_CHANGE_STREAM_SIZE		0x00000400
+#define FILE_NOTIFY_CHANGE_STREAM_WRITE		0x00000800
+
+struct smb2_change_notify_req {
+	struct smb2_sync_hdr sync_hdr;
+	__le16	StructureSize;
+	__le16	Flags;
+	__le32	OutputBufferLength;
+	__u64	PersistentFileId; /* opaque endianness */
+	__u64	VolatileFileId; /* opaque endianness */
+	__le32	CompletionFilter;
+	__u32	Reserved;
+} __packed;
+
+struct smb2_change_notify_rsp {
+	struct smb2_sync_hdr sync_hdr;
+	__le16	StructureSize;  /* Must be 9 */
+	__le16	OutputBufferOffset;
+	__le32	OutputBufferLength;
+	__u8	Buffer[1]; /* array of file notify structs */
+} __packed;
+
 #define SMB2_LOCKFLAG_SHARED_LOCK	0x0001
 #define SMB2_LOCKFLAG_EXCLUSIVE_LOCK	0x0002
 #define SMB2_LOCKFLAG_UNLOCK		0x0004

fs/cifs/smbdirect.c

@@ -164,95 +164,6 @@ do {									\
 #define log_rdma_mr(level, fmt, args...) \
 		log_rdma(level, LOG_RDMA_MR, fmt, ##args)
 
-/*
- * Destroy the transport and related RDMA and memory resources
- * Need to go through all the pending counters and make sure on one is using
- * the transport while it is destroyed
- */
-static void smbd_destroy_rdma_work(struct work_struct *work)
-{
-	struct smbd_response *response;
-	struct smbd_connection *info =
-		container_of(work, struct smbd_connection, destroy_work);
-	unsigned long flags;
-
-	log_rdma_event(INFO, "destroying qp\n");
-	ib_drain_qp(info->id->qp);
-	rdma_destroy_qp(info->id);
-
-	/* Unblock all I/O waiting on the send queue */
-	wake_up_interruptible_all(&info->wait_send_queue);
-
-	log_rdma_event(INFO, "cancelling idle timer\n");
-	cancel_delayed_work_sync(&info->idle_timer_work);
-	log_rdma_event(INFO, "cancelling send immediate work\n");
-	cancel_delayed_work_sync(&info->send_immediate_work);
-
-	log_rdma_event(INFO, "wait for all send to finish\n");
-	wait_event(info->wait_smbd_send_pending,
-		info->smbd_send_pending == 0);
-
-	log_rdma_event(INFO, "wait for all recv to finish\n");
-	wake_up_interruptible(&info->wait_reassembly_queue);
-	wait_event(info->wait_smbd_recv_pending,
-		info->smbd_recv_pending == 0);
-
-	log_rdma_event(INFO, "wait for all send posted to IB to finish\n");
-	wait_event(info->wait_send_pending,
-		atomic_read(&info->send_pending) == 0);
-	wait_event(info->wait_send_payload_pending,
-		atomic_read(&info->send_payload_pending) == 0);
-
-	log_rdma_event(INFO, "freeing mr list\n");
-	wake_up_interruptible_all(&info->wait_mr);
-	wait_event(info->wait_for_mr_cleanup,
-		atomic_read(&info->mr_used_count) == 0);
-	destroy_mr_list(info);
-
-	/* It's not posssible for upper layer to get to reassembly */
-	log_rdma_event(INFO, "drain the reassembly queue\n");
-	do {
-		spin_lock_irqsave(&info->reassembly_queue_lock, flags);
-		response = _get_first_reassembly(info);
-		if (response) {
-			list_del(&response->list);
-			spin_unlock_irqrestore(
-				&info->reassembly_queue_lock, flags);
-			put_receive_buffer(info, response);
-		} else
-			spin_unlock_irqrestore(&info->reassembly_queue_lock, flags);
-	} while (response);
-
-	info->reassembly_data_length = 0;
-
-	log_rdma_event(INFO, "free receive buffers\n");
-	wait_event(info->wait_receive_queues,
-		info->count_receive_queue + info->count_empty_packet_queue
-			== info->receive_credit_max);
-	destroy_receive_buffers(info);
-
-	ib_free_cq(info->send_cq);
-	ib_free_cq(info->recv_cq);
-	ib_dealloc_pd(info->pd);
-	rdma_destroy_id(info->id);
-
-	/* free mempools */
-	mempool_destroy(info->request_mempool);
-	kmem_cache_destroy(info->request_cache);
-
-	mempool_destroy(info->response_mempool);
-	kmem_cache_destroy(info->response_cache);
-
-	info->transport_status = SMBD_DESTROYED;
-	wake_up_all(&info->wait_destroy);
-}
-
-static int smbd_process_disconnected(struct smbd_connection *info)
-{
-	schedule_work(&info->destroy_work);
-	return 0;
-}
-
 static void smbd_disconnect_rdma_work(struct work_struct *work)
 {
 	struct smbd_connection *info =
@@ -319,7 +230,9 @@ static int smbd_conn_upcall(
 		}
 
 		info->transport_status = SMBD_DISCONNECTED;
-		smbd_process_disconnected(info);
+		wake_up_interruptible(&info->disconn_wait);
+		wake_up_interruptible(&info->wait_reassembly_queue);
+		wake_up_interruptible_all(&info->wait_send_queue);
 		break;
 
 	default:
@@ -940,7 +853,7 @@ static int smbd_create_header(struct smbd_connection *info,
 
 	if (info->transport_status != SMBD_CONNECTED) {
 		log_outgoing(ERR, "disconnected not sending\n");
-		return -ENOENT;
+		return -EAGAIN;
 	}
 	atomic_dec(&info->send_credits);
 
@@ -1066,6 +979,7 @@ static int smbd_post_send(struct smbd_connection *info,
 				wake_up(&info->wait_send_pending);
 		}
 		smbd_disconnect_rdma_connection(info);
+		rc = -EAGAIN;
 	} else
 		/* Reset timer for idle connection after packet is sent */
 		mod_delayed_work(info->workqueue, &info->idle_timer_work,
@@ -1478,17 +1392,97 @@ static void idle_connection_timer(struct work_struct *work)
 			info->keep_alive_interval*HZ);
 }
 
-/* Destroy this SMBD connection, called from upper layer */
-void smbd_destroy(struct smbd_connection *info)
+/*
+ * Destroy the transport and related RDMA and memory resources
+ * Need to go through all the pending counters and make sure on one is using
+ * the transport while it is destroyed
+ */
+void smbd_destroy(struct TCP_Server_Info *server)
 {
+	struct smbd_connection *info = server->smbd_conn;
+	struct smbd_response *response;
+	unsigned long flags;
+
+	if (!info) {
+		log_rdma_event(INFO, "rdma session already destroyed\n");
+		return;
+	}
+
 	log_rdma_event(INFO, "destroying rdma session\n");
+	if (info->transport_status != SMBD_DISCONNECTED) {
+		rdma_disconnect(server->smbd_conn->id);
+		log_rdma_event(INFO, "wait for transport being disconnected\n");
+		wait_event_interruptible(
+			info->disconn_wait,
+			info->transport_status == SMBD_DISCONNECTED);
+	}
 
-	/* Kick off the disconnection process */
-	smbd_disconnect_rdma_connection(info);
+	log_rdma_event(INFO, "destroying qp\n");
+	ib_drain_qp(info->id->qp);
+	rdma_destroy_qp(info->id);
+
+	log_rdma_event(INFO, "cancelling idle timer\n");
+	cancel_delayed_work_sync(&info->idle_timer_work);
+	log_rdma_event(INFO, "cancelling send immediate work\n");
+	cancel_delayed_work_sync(&info->send_immediate_work);
+
+	log_rdma_event(INFO, "wait for all send posted to IB to finish\n");
+	wait_event(info->wait_send_pending,
+		atomic_read(&info->send_pending) == 0);
+	wait_event(info->wait_send_payload_pending,
+		atomic_read(&info->send_payload_pending) == 0);
+
+	/* It's not posssible for upper layer to get to reassembly */
+	log_rdma_event(INFO, "drain the reassembly queue\n");
+	do {
+		spin_lock_irqsave(&info->reassembly_queue_lock, flags);
+		response = _get_first_reassembly(info);
+		if (response) {
+			list_del(&response->list);
+			spin_unlock_irqrestore(
+				&info->reassembly_queue_lock, flags);
+			put_receive_buffer(info, response);
+		} else
+			spin_unlock_irqrestore(
+				&info->reassembly_queue_lock, flags);
+	} while (response);
+	info->reassembly_data_length = 0;
+
+	log_rdma_event(INFO, "free receive buffers\n");
+	wait_event(info->wait_receive_queues,
+		info->count_receive_queue + info->count_empty_packet_queue
+			== info->receive_credit_max);
+	destroy_receive_buffers(info);
+
+	/*
+	 * For performance reasons, memory registration and deregistration
+	 * are not locked by srv_mutex. It is possible some processes are
+	 * blocked on transport srv_mutex while holding memory registration.
+	 * Release the transport srv_mutex to allow them to hit the failure
+	 * path when sending data, and then release memory registartions.
+	 */
+	log_rdma_event(INFO, "freeing mr list\n");
+	wake_up_interruptible_all(&info->wait_mr);
+	while (atomic_read(&info->mr_used_count)) {
+		mutex_unlock(&server->srv_mutex);
+		msleep(1000);
+		mutex_lock(&server->srv_mutex);
+	}
+	destroy_mr_list(info);
+
+	ib_free_cq(info->send_cq);
+	ib_free_cq(info->recv_cq);
+	ib_dealloc_pd(info->pd);
+	rdma_destroy_id(info->id);
+
+	/* free mempools */
+	mempool_destroy(info->request_mempool);
+	kmem_cache_destroy(info->request_cache);
 
-	log_rdma_event(INFO, "wait for transport being destroyed\n");
-	wait_event(info->wait_destroy,
-		info->transport_status == SMBD_DESTROYED);
+	mempool_destroy(info->response_mempool);
+	kmem_cache_destroy(info->response_cache);
+
+	info->transport_status = SMBD_DESTROYED;
 
 	destroy_workqueue(info->workqueue);
 	kfree(info);
@@ -1513,17 +1507,9 @@ int smbd_reconnect(struct TCP_Server_Info *server)
 	 */
 	if (server->smbd_conn->transport_status == SMBD_CONNECTED) {
 		log_rdma_event(INFO, "disconnecting transport\n");
-		smbd_disconnect_rdma_connection(server->smbd_conn);
+		smbd_destroy(server);
 	}
 
-	/* wait until the transport is destroyed */
-	if (!wait_event_timeout(server->smbd_conn->wait_destroy,
-		server->smbd_conn->transport_status == SMBD_DESTROYED, 5*HZ))
-		return -EAGAIN;
-
-	destroy_workqueue(server->smbd_conn->workqueue);
-	kfree(server->smbd_conn);
-
 create_conn:
 	log_rdma_event(INFO, "creating rdma session\n");
 	server->smbd_conn = smbd_get_connection(
@@ -1739,12 +1725,13 @@ static struct smbd_connection *_smbd_get_connection(
 	conn_param.retry_count = SMBD_CM_RETRY;
 	conn_param.rnr_retry_count = SMBD_CM_RNR_RETRY;
 	conn_param.flow_control = 0;
-	init_waitqueue_head(&info->wait_destroy);
 
 	log_rdma_event(INFO, "connecting to IP %pI4 port %d\n",
 		&addr_in->sin_addr, port);
 
 	init_waitqueue_head(&info->conn_wait);
+	init_waitqueue_head(&info->disconn_wait);
+	init_waitqueue_head(&info->wait_reassembly_queue);
 	rc = rdma_connect(info->id, &conn_param);
 	if (rc) {
 		log_rdma_event(ERR, "rdma_connect() failed with %i\n", rc);
@@ -1768,19 +1755,11 @@ static struct smbd_connection *_smbd_get_connection(
 	}
 
 	init_waitqueue_head(&info->wait_send_queue);
-	init_waitqueue_head(&info->wait_reassembly_queue);
-
 	INIT_DELAYED_WORK(&info->idle_timer_work, idle_connection_timer);
 	INIT_DELAYED_WORK(&info->send_immediate_work, send_immediate_work);
 	queue_delayed_work(info->workqueue, &info->idle_timer_work,
 		info->keep_alive_interval*HZ);
 
-	init_waitqueue_head(&info->wait_smbd_send_pending);
-	info->smbd_send_pending = 0;
-
-	init_waitqueue_head(&info->wait_smbd_recv_pending);
-	info->smbd_recv_pending = 0;
-
 	init_waitqueue_head(&info->wait_send_pending);
 	atomic_set(&info->send_pending, 0);
 
@@ -1788,7 +1767,6 @@ static struct smbd_connection *_smbd_get_connection(
 	atomic_set(&info->send_payload_pending, 0);
 
 	INIT_WORK(&info->disconnect_work, smbd_disconnect_rdma_work);
-	INIT_WORK(&info->destroy_work, smbd_destroy_rdma_work);
 	INIT_WORK(&info->recv_done_work, smbd_recv_done_work);
 	INIT_WORK(&info->post_send_credits_work, smbd_post_send_credits);
 	info->new_credits_offered = 0;
@@ -1810,7 +1788,7 @@ static struct smbd_connection *_smbd_get_connection(
 
 allocate_mr_failed:
 	/* At this point, need to a full transport shutdown */
-	smbd_destroy(info);
+	smbd_destroy(server);
 	return NULL;
 
 negotiation_failed:
@@ -1882,11 +1860,6 @@ static int smbd_recv_buf(struct smbd_connection *info, char *buf,
 	int rc;
 
 again:
-	if (info->transport_status != SMBD_CONNECTED) {
-		log_read(ERR, "disconnected\n");
-		return -ENODEV;
-	}
-
 	/*
 	 * No need to hold the reassembly queue lock all the time as we are
 	 * the only one reading from the front of the queue. The transport
@@ -2000,7 +1973,12 @@ static int smbd_recv_buf(struct smbd_connection *info, char *buf,
 			info->transport_status != SMBD_CONNECTED);
 	/* Don't return any data if interrupted */
 	if (rc)
-		return -ENODEV;
+		return rc;
+
+	if (info->transport_status != SMBD_CONNECTED) {
+		log_read(ERR, "disconnected\n");
+		return 0;
+	}
 
 	goto again;
 }
@@ -2052,8 +2030,6 @@ int smbd_recv(struct smbd_connection *info, struct msghdr *msg)
 	unsigned int to_read, page_offset;
 	int rc;
 
-	info->smbd_recv_pending++;
-
 	if (iov_iter_rw(&msg->msg_iter) == WRITE) {
 		/* It's a bug in upper layer to get there */
 		cifs_dbg(VFS, "CIFS: invalid msg iter dir %u\n",
@@ -2084,9 +2060,6 @@ int smbd_recv(struct smbd_connection *info, struct msghdr *msg)
 	}
 
 out:
-	info->smbd_recv_pending--;
-	wake_up(&info->wait_smbd_recv_pending);
-
 	/* SMBDirect will read it all or nothing */
 	if (rc > 0)
 		msg->msg_iter.count = 0;
@@ -2099,7 +2072,8 @@ int smbd_recv(struct smbd_connection *info, struct msghdr *msg)
  * rqst: the data to write
  * return value: 0 if successfully write, otherwise error code
  */
-int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst)
+int smbd_send(struct TCP_Server_Info *server,
+	int num_rqst, struct smb_rqst *rqst_array)
 {
 	struct smbd_connection *info = server->smbd_conn;
 	struct kvec vec;
@@ -2111,54 +2085,49 @@ int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst)
 		info->max_send_size - sizeof(struct smbd_data_transfer);
 	struct kvec *iov;
 	int rc;
+	struct smb_rqst *rqst;
+	int rqst_idx;
 
-	info->smbd_send_pending++;
 	if (info->transport_status != SMBD_CONNECTED) {
-		rc = -ENODEV;
+		rc = -EAGAIN;
 		goto done;
 	}
 
-	/*
-	 * Skip the RFC1002 length defined in MS-SMB2 section 2.1
-	 * It is used only for TCP transport in the iov[0]
-	 * In future we may want to add a transport layer under protocol
-	 * layer so this will only be issued to TCP transport
-	 */
-
-	if (rqst->rq_iov[0].iov_len != 4) {
-		log_write(ERR, "expected the pdu length in 1st iov, but got %zu\n", rqst->rq_iov[0].iov_len);
-		return -EINVAL;
-	}
-
 	/*
 	 * Add in the page array if there is one. The caller needs to set
 	 * rq_tailsz to PAGE_SIZE when the buffer has multiple pages and
 	 * ends at page boundary
 	 */
-	buflen = smb_rqst_len(server, rqst);
+	remaining_data_length = 0;
+	for (i = 0; i < num_rqst; i++)
+		remaining_data_length += smb_rqst_len(server, &rqst_array[i]);
 
-	if (buflen + sizeof(struct smbd_data_transfer) >
+	if (remaining_data_length + sizeof(struct smbd_data_transfer) >
 		info->max_fragmented_send_size) {
 		log_write(ERR, "payload size %d > max size %d\n",
-			buflen, info->max_fragmented_send_size);
+			remaining_data_length, info->max_fragmented_send_size);
 		rc = -EINVAL;
 		goto done;
 	}
 
-	iov = &rqst->rq_iov[1];
+	rqst_idx = 0;
+
+next_rqst:
+	rqst = &rqst_array[rqst_idx];
+	iov = rqst->rq_iov;
 
-	cifs_dbg(FYI, "Sending smb (RDMA): smb_len=%u\n", buflen);
-	for (i = 0; i < rqst->rq_nvec-1; i++)
+	cifs_dbg(FYI, "Sending smb (RDMA): idx=%d smb_len=%lu\n",
+		rqst_idx, smb_rqst_len(server, rqst));
+	for (i = 0; i < rqst->rq_nvec; i++)
 		dump_smb(iov[i].iov_base, iov[i].iov_len);
 
-	remaining_data_length = buflen;
 
-	log_write(INFO, "rqst->rq_nvec=%d rqst->rq_npages=%d rq_pagesz=%d "
-		"rq_tailsz=%d buflen=%d\n",
-		rqst->rq_nvec, rqst->rq_npages, rqst->rq_pagesz,
-		rqst->rq_tailsz, buflen);
+	log_write(INFO, "rqst_idx=%d nvec=%d rqst->rq_npages=%d rq_pagesz=%d "
+		"rq_tailsz=%d buflen=%lu\n",
+		rqst_idx, rqst->rq_nvec, rqst->rq_npages, rqst->rq_pagesz,
+		rqst->rq_tailsz, smb_rqst_len(server, rqst));
 
-	start = i = iov[0].iov_len ? 0 : 1;
+	start = i = 0;
 	buflen = 0;
 	while (true) {
 		buflen += iov[i].iov_len;
@@ -2206,14 +2175,14 @@ int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst)
 						goto done;
 				}
 				i++;
-				if (i == rqst->rq_nvec-1)
+				if (i == rqst->rq_nvec)
 					break;
 			}
 			start = i;
 			buflen = 0;
 		} else {
 			i++;
-			if (i == rqst->rq_nvec-1) {
+			if (i == rqst->rq_nvec) {
 				/* send out all remaining vecs */
 				remaining_data_length -= buflen;
 				log_write(INFO,
@@ -2257,6 +2226,10 @@ int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst)
 		}
 	}
 
+	rqst_idx++;
+	if (rqst_idx < num_rqst)
+		goto next_rqst;
+
 done:
 	/*
 	 * As an optimization, we don't wait for individual I/O to finish
@@ -2268,9 +2241,6 @@ int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst)
 	wait_event(info->wait_send_payload_pending,
 		atomic_read(&info->send_payload_pending) == 0);
 
-	info->smbd_send_pending--;
-	wake_up(&info->wait_smbd_send_pending);
-
 	return rc;
 }
 

fs/cifs/smbdirect.h

@@ -70,12 +70,11 @@ struct smbd_connection {
 	int ri_rc;
 	struct completion ri_done;
 	wait_queue_head_t conn_wait;
-	wait_queue_head_t wait_destroy;
+	wait_queue_head_t disconn_wait;
 
 	struct completion negotiate_completion;
 	bool negotiate_done;
 
-	struct work_struct destroy_work;
 	struct work_struct disconnect_work;
 	struct work_struct recv_done_work;
 	struct work_struct post_send_credits_work;
@@ -123,13 +122,6 @@ struct smbd_connection {
 	wait_queue_head_t wait_for_mr_cleanup;
 
 	/* Activity accoutning */
-	/* Pending reqeusts issued from upper layer */
-	int smbd_send_pending;
-	wait_queue_head_t wait_smbd_send_pending;
-
-	int smbd_recv_pending;
-	wait_queue_head_t wait_smbd_recv_pending;
-
 	atomic_t send_pending;
 	wait_queue_head_t wait_send_pending;
 	atomic_t send_payload_pending;
@@ -288,11 +280,12 @@ struct smbd_connection *smbd_get_connection(
 /* Reconnect SMBDirect session */
 int smbd_reconnect(struct TCP_Server_Info *server);
 /* Destroy SMBDirect session */
-void smbd_destroy(struct smbd_connection *info);
+void smbd_destroy(struct TCP_Server_Info *server);
 
 /* Interface for carrying upper layer I/O through send/recv */
 int smbd_recv(struct smbd_connection *info, struct msghdr *msg);
-int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst);
+int smbd_send(struct TCP_Server_Info *server,
+	int num_rqst, struct smb_rqst *rqst);
 
 enum mr_state {
 	MR_READY,
@@ -330,9 +323,9 @@ struct smbd_connection {};
 static inline void *smbd_get_connection(
 	struct TCP_Server_Info *server, struct sockaddr *dstaddr) {return NULL;}
 static inline int smbd_reconnect(struct TCP_Server_Info *server) {return -1; }
-static inline void smbd_destroy(struct smbd_connection *info) {}
+static inline void smbd_destroy(struct TCP_Server_Info *server) {}
 static inline int smbd_recv(struct smbd_connection *info, struct msghdr *msg) {return -1; }
-static inline int smbd_send(struct TCP_Server_Info *server, struct smb_rqst *rqst) {return -1; }
+static inline int smbd_send(struct TCP_Server_Info *server, int num_rqst, struct smb_rqst *rqst) {return -1; }
 #endif
 
 #endif

fs/cifs/smbfsctl.h

@@ -35,6 +35,33 @@
  * below). Additional detail on less common ones can be found in MS-FSCC
  * section 2.3.
  */
+
+/*
+ * FSCTL values are 32 bits and are constructed as
+ * <device 16bits> <access 2bits> <function 12bits> <method 2bits>
+ */
+/* Device */
+#define FSCTL_DEVICE_DFS                 (0x0006 << 16)
+#define FSCTL_DEVICE_FILE_SYSTEM         (0x0009 << 16)
+#define FSCTL_DEVICE_NAMED_PIPE          (0x0011 << 16)
+#define FSCTL_DEVICE_NETWORK_FILE_SYSTEM (0x0014 << 16)
+#define FSCTL_DEVICE_MASK                0xffff0000
+/* Access */
+#define FSCTL_DEVICE_ACCESS_FILE_ANY_ACCESS        (0x00 << 14)
+#define FSCTL_DEVICE_ACCESS_FILE_READ_ACCESS       (0x01 << 14)
+#define FSCTL_DEVICE_ACCESS_FILE_WRITE_ACCESS      (0x02 << 14)
+#define FSCTL_DEVICE_ACCESS_FILE_READ_WRITE_ACCESS (0x03 << 14)
+#define FSCTL_DEVICE_ACCESS_MASK                   0x0000c000
+/* Function */
+#define FSCTL_DEVICE_FUNCTION_MASK       0x00003ffc
+/* Method */
+#define FSCTL_DEVICE_METHOD_BUFFERED   0x00
+#define FSCTL_DEVICE_METHOD_IN_DIRECT  0x01
+#define FSCTL_DEVICE_METHOD_OUT_DIRECT 0x02
+#define FSCTL_DEVICE_METHOD_NEITHER    0x03
+#define FSCTL_DEVICE_METHOD_MASK       0x00000003
+
+
 #define FSCTL_DFS_GET_REFERRALS      0x00060194
 #define FSCTL_DFS_GET_REFERRALS_EX   0x000601B0
 #define FSCTL_REQUEST_OPLOCK_LEVEL_1 0x00090000
@@ -76,7 +103,7 @@
 #define FSCTL_SET_ZERO_ON_DEALLOC    0x00090194 /* BB add struct */
 #define FSCTL_SET_SHORT_NAME_BEHAVIOR 0x000901B4 /* BB add struct */
 #define FSCTL_GET_INTEGRITY_INFORMATION 0x0009027C
-#define FSCTL_QUERY_ALLOCATED_RANGES 0x000940CF /* BB add struct */
+#define FSCTL_QUERY_ALLOCATED_RANGES 0x000940CF
 #define FSCTL_SET_DEFECT_MANAGEMENT  0x00098134 /* BB add struct */
 #define FSCTL_FILE_LEVEL_TRIM        0x00098208 /* BB add struct */
 #define FSCTL_DUPLICATE_EXTENTS_TO_FILE 0x00098344

fs/cifs/transport.c

@@ -104,7 +104,10 @@ DeleteMidQEntry(struct mid_q_entry *midEntry)
 {
 #ifdef CONFIG_CIFS_STATS2
 	__le16 command = midEntry->server->vals->lock_cmd;
+	__u16 smb_cmd = le16_to_cpu(midEntry->command);
 	unsigned long now;
+	unsigned long roundtrip_time;
+	struct TCP_Server_Info *server = midEntry->server;
 #endif
 	midEntry->mid_state = MID_FREE;
 	atomic_dec(&midCount);
@@ -114,6 +117,23 @@ DeleteMidQEntry(struct mid_q_entry *midEntry)
 		cifs_small_buf_release(midEntry->resp_buf);
 #ifdef CONFIG_CIFS_STATS2
 	now = jiffies;
+	if (now < midEntry->when_alloc)
+		cifs_dbg(VFS, "invalid mid allocation time\n");
+	roundtrip_time = now - midEntry->when_alloc;
+
+	if (smb_cmd < NUMBER_OF_SMB2_COMMANDS) {
+		if (atomic_read(&server->num_cmds[smb_cmd]) == 0) {
+			server->slowest_cmd[smb_cmd] = roundtrip_time;
+			server->fastest_cmd[smb_cmd] = roundtrip_time;
+		} else {
+			if (server->slowest_cmd[smb_cmd] < roundtrip_time)
+				server->slowest_cmd[smb_cmd] = roundtrip_time;
+			else if (server->fastest_cmd[smb_cmd] > roundtrip_time)
+				server->fastest_cmd[smb_cmd] = roundtrip_time;
+		}
+		cifs_stats_inc(&server->num_cmds[smb_cmd]);
+		server->time_per_cmd[smb_cmd] += roundtrip_time;
+	}
 	/*
 	 * commands taking longer than one second (default) can be indications
 	 * that something is wrong, unless it is quite a slow link or a very
@@ -131,11 +151,10 @@ DeleteMidQEntry(struct mid_q_entry *midEntry)
 		 * smb2slowcmd[NUMBER_OF_SMB2_COMMANDS] counts by command
 		 * NB: le16_to_cpu returns unsigned so can not be negative below
 		 */
-		if (le16_to_cpu(midEntry->command) < NUMBER_OF_SMB2_COMMANDS)
-			cifs_stats_inc(&midEntry->server->smb2slowcmd[le16_to_cpu(midEntry->command)]);
+		if (smb_cmd < NUMBER_OF_SMB2_COMMANDS)
+			cifs_stats_inc(&server->smb2slowcmd[smb_cmd]);
 
-		trace_smb3_slow_rsp(le16_to_cpu(midEntry->command),
-			       midEntry->mid, midEntry->pid,
+		trace_smb3_slow_rsp(smb_cmd, midEntry->mid, midEntry->pid,
 			       midEntry->when_sent, midEntry->when_received);
 		if (cifsFYI & CIFS_TIMER) {
 			pr_debug(" CIFS slow rsp: cmd %d mid %llu",
@@ -300,7 +319,7 @@ __smb_send_rqst(struct TCP_Server_Info *server, int num_rqst,
 	__be32 rfc1002_marker;
 
 	if (cifs_rdma_enabled(server) && server->smbd_conn) {
-		rc = smbd_send(server, rqst);
+		rc = smbd_send(server, num_rqst, rqst);
 		goto smbd_done;
 	}
 
@@ -510,7 +529,7 @@ wait_for_free_credits(struct TCP_Server_Info *server, const int num_credits,
 		return -EAGAIN;
 
 	spin_lock(&server->req_lock);
-	if ((flags & CIFS_TIMEOUT_MASK) == CIFS_ASYNC_OP) {
+	if ((flags & CIFS_TIMEOUT_MASK) == CIFS_NON_BLOCKING) {
 		/* oplock breaks must not be held up */
 		server->in_flight++;
 		*credits -= 1;
@@ -819,7 +838,7 @@ SendReceiveNoRsp(const unsigned int xid, struct cifs_ses *ses,
 
 	iov[0].iov_base = in_buf;
 	iov[0].iov_len = get_rfc1002_length(in_buf) + 4;
-	flags |= CIFS_NO_RESP;
+	flags |= CIFS_NO_RSP_BUF;
 	rc = SendReceive2(xid, ses, iov, 1, &resp_buf_type, flags, &rsp_iov);
 	cifs_dbg(NOISY, "SendRcvNoRsp flags %d rc %d\n", flags, rc);
 
@@ -1054,8 +1073,11 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses,
 
 	mutex_unlock(&ses->server->srv_mutex);
 
-	if (rc < 0) {
-		/* Sending failed for some reason - return credits back */
+	/*
+	 * If sending failed for some reason or it is an oplock break that we
+	 * will not receive a response to - return credits back
+	 */
+	if (rc < 0 || (flags & CIFS_NO_SRV_RSP)) {
 		for (i = 0; i < num_rqst; i++)
 			add_credits(ses->server, &credits[i], optype);
 		goto out;
@@ -1076,9 +1098,6 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses,
 		smb311_update_preauth_hash(ses, rqst[0].rq_iov,
 					   rqst[0].rq_nvec);
 
-	if ((flags & CIFS_TIMEOUT_MASK) == CIFS_ASYNC_OP)
-		goto out;
-
 	for (i = 0; i < num_rqst; i++) {
 		rc = wait_for_response(ses->server, midQ[i]);
 		if (rc != 0)
@@ -1132,7 +1151,7 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses,
 						     flags & CIFS_LOG_ERROR);
 
 		/* mark it so buf will not be freed by cifs_delete_mid */
-		if ((flags & CIFS_NO_RESP) == 0)
+		if ((flags & CIFS_NO_RSP_BUF) == 0)
 			midQ[i]->resp_buf = NULL;
 
 	}
@@ -1283,9 +1302,6 @@ SendReceive(const unsigned int xid, struct cifs_ses *ses,
 	if (rc < 0)
 		goto out;
 
-	if ((flags & CIFS_TIMEOUT_MASK) == CIFS_ASYNC_OP)
-		goto out;
-
 	rc = wait_for_response(ses->server, midQ);
 	if (rc != 0) {
 		send_cancel(ses->server, &rqst, midQ);