net/sunrpc/xprtsock.c

   1 /*
   2  * linux/net/sunrpc/xprtsock.c
   3  *
   4  * Client-side transport implementation for sockets.
   5  *
   6  * TCP callback races fixes (C) 1998 Red Hat
   7  * TCP send fixes (C) 1998 Red Hat
   8  * TCP NFS related read + write fixes
   9  *  (C) 1999 Dave Airlie, University of Limerick, Ireland <airlied@linux.ie>
  10  *
  11  * Rewrite of larges part of the code in order to stabilize TCP stuff.
  12  * Fix behaviour when socket buffer is full.
  13  *  (C) 1999 Trond Myklebust <trond.myklebust@fys.uio.no>
  14  *
  15  * IP socket transport implementation, (C) 2005 Chuck Lever <cel@netapp.com>
  16  *
  17  * IPv6 support contributed by Gilles Quillard, Bull Open Source, 2005.
  18  *   <gilles.quillard@bull.net>
  19  */
  20
  21 #include <linux/types.h>
  22 #include <linux/string.h>
  23 #include <linux/slab.h>
  24 #include <linux/module.h>
  25 #include <linux/capability.h>
  26 #include <linux/pagemap.h>
  27 #include <linux/errno.h>
  28 #include <linux/socket.h>
  29 #include <linux/in.h>
  30 #include <linux/net.h>
  31 #include <linux/mm.h>
  32 #include <linux/un.h>
  33 #include <linux/udp.h>
  34 #include <linux/tcp.h>
  35 #include <linux/sunrpc/clnt.h>
  36 #include <linux/sunrpc/addr.h>
  37 #include <linux/sunrpc/sched.h>
  38 #include <linux/sunrpc/svcsock.h>
  39 #include <linux/sunrpc/xprtsock.h>
  40 #include <linux/file.h>
  41 #ifdef CONFIG_SUNRPC_BACKCHANNEL
  42 #include <linux/sunrpc/bc_xprt.h>
  43 #endif
  44
  45 #include <net/sock.h>
  46 #include <net/checksum.h>
  47 #include <net/udp.h>
  48 #include <net/tcp.h>
  49
  50 #include <trace/events/sunrpc.h>
  51
  52 #include "sunrpc.h"
  53
  54 static void xs_close(struct rpc_xprt *xprt);
  55
  56 /*
  57  * xprtsock tunables
  58  */
  59 static unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE;
  60 static unsigned int xprt_tcp_slot_table_entries = RPC_MIN_SLOT_TABLE;
  61 static unsigned int xprt_max_tcp_slot_table_entries = RPC_MAX_SLOT_TABLE;
  62
  63 static unsigned int xprt_min_resvport = RPC_DEF_MIN_RESVPORT;
  64 static unsigned int xprt_max_resvport = RPC_DEF_MAX_RESVPORT;
  65
  66 #define XS_TCP_LINGER_TO        (15U * HZ)
  67 static unsigned int xs_tcp_fin_timeout __read_mostly = XS_TCP_LINGER_TO;
  68
  69 /*
  70  * We can register our own files under /proc/sys/sunrpc by
  71  * calling register_sysctl_table() again.  The files in that
  72  * directory become the union of all files registered there.
  73  *
  74  * We simply need to make sure that we don't collide with
  75  * someone else's file names!
  76  */
  77
  78 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
  79
  80 static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE;
  81 static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE;
  82 static unsigned int max_tcp_slot_table_limit = RPC_MAX_SLOT_TABLE_LIMIT;
  83 static unsigned int xprt_min_resvport_limit = RPC_MIN_RESVPORT;
  84 static unsigned int xprt_max_resvport_limit = RPC_MAX_RESVPORT;
  85
  86 static struct ctl_table_header *sunrpc_table_header;
  87
  88 /*
  89  * FIXME: changing the UDP slot table size should also resize the UDP
  90  *        socket buffers for existing UDP transports
  91  */
  92 static struct ctl_table xs_tunables_table[] = {
  93         {
  94                 .procname       = "udp_slot_table_entries",
  95                 .data           = &xprt_udp_slot_table_entries,
  96                 .maxlen         = sizeof(unsigned int),
  97                 .mode           = 0644,
  98                 .proc_handler   = proc_dointvec_minmax,
  99                 .extra1         = &min_slot_table_size,
 100                 .extra2         = &max_slot_table_size
 101         },
 102         {
 103                 .procname       = "tcp_slot_table_entries",
 104                 .data           = &xprt_tcp_slot_table_entries,
 105                 .maxlen         = sizeof(unsigned int),
 106                 .mode           = 0644,
 107                 .proc_handler   = proc_dointvec_minmax,
 108                 .extra1         = &min_slot_table_size,
 109                 .extra2         = &max_slot_table_size
 110         },
 111         {
 112                 .procname       = "tcp_max_slot_table_entries",
 113                 .data           = &xprt_max_tcp_slot_table_entries,
 114                 .maxlen         = sizeof(unsigned int),
 115                 .mode           = 0644,
 116                 .proc_handler   = proc_dointvec_minmax,
 117                 .extra1         = &min_slot_table_size,
 118                 .extra2         = &max_tcp_slot_table_limit
 119         },
 120         {
 121                 .procname       = "min_resvport",
 122                 .data           = &xprt_min_resvport,
 123                 .maxlen         = sizeof(unsigned int),
 124                 .mode           = 0644,
 125                 .proc_handler   = proc_dointvec_minmax,
 126                 .extra1         = &xprt_min_resvport_limit,
 127                 .extra2         = &xprt_max_resvport_limit
 128         },
 129         {
 130                 .procname       = "max_resvport",
 131                 .data           = &xprt_max_resvport,
 132                 .maxlen         = sizeof(unsigned int),
 133                 .mode           = 0644,
 134                 .proc_handler   = proc_dointvec_minmax,
 135                 .extra1         = &xprt_min_resvport_limit,
 136                 .extra2         = &xprt_max_resvport_limit
 137         },
 138         {
 139                 .procname       = "tcp_fin_timeout",
 140                 .data           = &xs_tcp_fin_timeout,
 141                 .maxlen         = sizeof(xs_tcp_fin_timeout),
 142                 .mode           = 0644,
 143                 .proc_handler   = proc_dointvec_jiffies,
 144         },
 145         { },
 146 };
 147
 148 static struct ctl_table sunrpc_table[] = {
 149         {
 150                 .procname       = "sunrpc",
 151                 .mode           = 0555,
 152                 .child          = xs_tunables_table
 153         },
 154         { },
 155 };
 156
 157 #endif
 158
 159 /*
 160  * Wait duration for a reply from the RPC portmapper.
 161  */
 162 #define XS_BIND_TO              (60U * HZ)
 163
 164 /*
 165  * Delay if a UDP socket connect error occurs.  This is most likely some
 166  * kind of resource problem on the local host.
 167  */
 168 #define XS_UDP_REEST_TO         (2U * HZ)
 169
 170 /*
 171  * The reestablish timeout allows clients to delay for a bit before attempting
 172  * to reconnect to a server that just dropped our connection.
 173  *
 174  * We implement an exponential backoff when trying to reestablish a TCP
 175  * transport connection with the server.  Some servers like to drop a TCP
 176  * connection when they are overworked, so we start with a short timeout and
 177  * increase over time if the server is down or not responding.
 178  */
 179 #define XS_TCP_INIT_REEST_TO    (3U * HZ)
 180 #define XS_TCP_MAX_REEST_TO     (5U * 60 * HZ)
 181
 182 /*
 183  * TCP idle timeout; client drops the transport socket if it is idle
 184  * for this long.  Note that we also timeout UDP sockets to prevent
 185  * holding port numbers when there is no RPC traffic.
 186  */
 187 #define XS_IDLE_DISC_TO         (5U * 60 * HZ)
 188
 189 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 190 # undef  RPC_DEBUG_DATA
 191 # define RPCDBG_FACILITY        RPCDBG_TRANS
 192 #endif
 193
 194 #ifdef RPC_DEBUG_DATA
 195 static void xs_pktdump(char *msg, u32 *packet, unsigned int count)
 196 {
 197         u8 *buf = (u8 *) packet;
 198         int j;
 199
 200         dprintk("RPC:       %s\n", msg);
 201         for (j = 0; j < count && j < 128; j += 4) {
 202                 if (!(j & 31)) {
 203                         if (j)
 204                                 dprintk("\n");
 205                         dprintk("0x%04x ", j);
 206                 }
 207                 dprintk("%02x%02x%02x%02x ",
 208                         buf[j], buf[j+1], buf[j+2], buf[j+3]);
 209         }
 210         dprintk("\n");
 211 }
 212 #else
 213 static inline void xs_pktdump(char *msg, u32 *packet, unsigned int count)
 214 {
 215         /* NOP */
 216 }
 217 #endif
 218
 219 static inline struct rpc_xprt *xprt_from_sock(struct sock *sk)
 220 {
 221         return (struct rpc_xprt *) sk->sk_user_data;
 222 }
 223
 224 static inline struct sockaddr *xs_addr(struct rpc_xprt *xprt)
 225 {
 226         return (struct sockaddr *) &xprt->addr;
 227 }
 228
 229 static inline struct sockaddr_un *xs_addr_un(struct rpc_xprt *xprt)
 230 {
 231         return (struct sockaddr_un *) &xprt->addr;
 232 }
 233
 234 static inline struct sockaddr_in *xs_addr_in(struct rpc_xprt *xprt)
 235 {
 236         return (struct sockaddr_in *) &xprt->addr;
 237 }
 238
 239 static inline struct sockaddr_in6 *xs_addr_in6(struct rpc_xprt *xprt)
 240 {
 241         return (struct sockaddr_in6 *) &xprt->addr;
 242 }
 243
 244 static void xs_format_common_peer_addresses(struct rpc_xprt *xprt)
 245 {
 246         struct sockaddr *sap = xs_addr(xprt);
 247         struct sockaddr_in6 *sin6;
 248         struct sockaddr_in *sin;
 249         struct sockaddr_un *sun;
 250         char buf[128];
 251
 252         switch (sap->sa_family) {
 253         case AF_LOCAL:
 254                 sun = xs_addr_un(xprt);
 255                 strlcpy(buf, sun->sun_path, sizeof(buf));
 256                 xprt->address_strings[RPC_DISPLAY_ADDR] =
 257                                                 kstrdup(buf, GFP_KERNEL);
 258                 break;
 259         case AF_INET:
 260                 (void)rpc_ntop(sap, buf, sizeof(buf));
 261                 xprt->address_strings[RPC_DISPLAY_ADDR] =
 262                                                 kstrdup(buf, GFP_KERNEL);
 263                 sin = xs_addr_in(xprt);
 264                 snprintf(buf, sizeof(buf), "%08x", ntohl(sin->sin_addr.s_addr));
 265                 break;
 266         case AF_INET6:
 267                 (void)rpc_ntop(sap, buf, sizeof(buf));
 268                 xprt->address_strings[RPC_DISPLAY_ADDR] =
 269                                                 kstrdup(buf, GFP_KERNEL);
 270                 sin6 = xs_addr_in6(xprt);
 271                 snprintf(buf, sizeof(buf), "%pi6", &sin6->sin6_addr);
 272                 break;
 273         default:
 274                 BUG();
 275         }
 276
 277         xprt->address_strings[RPC_DISPLAY_HEX_ADDR] = kstrdup(buf, GFP_KERNEL);
 278 }
 279
 280 static void xs_format_common_peer_ports(struct rpc_xprt *xprt)
 281 {
 282         struct sockaddr *sap = xs_addr(xprt);
 283         char buf[128];
 284
 285         snprintf(buf, sizeof(buf), "%u", rpc_get_port(sap));
 286         xprt->address_strings[RPC_DISPLAY_PORT] = kstrdup(buf, GFP_KERNEL);
 287
 288         snprintf(buf, sizeof(buf), "%4hx", rpc_get_port(sap));
 289         xprt->address_strings[RPC_DISPLAY_HEX_PORT] = kstrdup(buf, GFP_KERNEL);
 290 }
 291
 292 static void xs_format_peer_addresses(struct rpc_xprt *xprt,
 293                                      const char *protocol,
 294                                      const char *netid)
 295 {
 296         xprt->address_strings[RPC_DISPLAY_PROTO] = protocol;
 297         xprt->address_strings[RPC_DISPLAY_NETID] = netid;
 298         xs_format_common_peer_addresses(xprt);
 299         xs_format_common_peer_ports(xprt);
 300 }
 301
 302 static void xs_update_peer_port(struct rpc_xprt *xprt)
 303 {
 304         kfree(xprt->address_strings[RPC_DISPLAY_HEX_PORT]);
 305         kfree(xprt->address_strings[RPC_DISPLAY_PORT]);
 306
 307         xs_format_common_peer_ports(xprt);
 308 }
 309
 310 static void xs_free_peer_addresses(struct rpc_xprt *xprt)
 311 {
 312         unsigned int i;
 313
 314         for (i = 0; i < RPC_DISPLAY_MAX; i++)
 315                 switch (i) {
 316                 case RPC_DISPLAY_PROTO:
 317                 case RPC_DISPLAY_NETID:
 318                         continue;
 319                 default:
 320                         kfree(xprt->address_strings[i]);
 321                 }
 322 }
 323
 324 #define XS_SENDMSG_FLAGS        (MSG_DONTWAIT | MSG_NOSIGNAL)
 325
 326 static int xs_send_kvec(struct socket *sock, struct sockaddr *addr, int addrlen, struct kvec *vec, unsigned int base, int more)
 327 {
 328         struct msghdr msg = {
 329                 .msg_name       = addr,
 330                 .msg_namelen    = addrlen,
 331                 .msg_flags      = XS_SENDMSG_FLAGS | (more ? MSG_MORE : 0),
 332         };
 333         struct kvec iov = {
 334                 .iov_base       = vec->iov_base + base,
 335                 .iov_len        = vec->iov_len - base,
 336         };
 337
 338         if (iov.iov_len != 0)
 339                 return kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len);
 340         return kernel_sendmsg(sock, &msg, NULL, 0, 0);
 341 }
 342
 343 static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned int base, int more, bool zerocopy, int *sent_p)
 344 {
 345         ssize_t (*do_sendpage)(struct socket *sock, struct page *page,
 346                         int offset, size_t size, int flags);
 347         struct page **ppage;
 348         unsigned int remainder;
 349         int err;
 350
 351         remainder = xdr->page_len - base;
 352         base += xdr->page_base;
 353         ppage = xdr->pages + (base >> PAGE_SHIFT);
 354         base &= ~PAGE_MASK;
 355         do_sendpage = sock->ops->sendpage;
 356         if (!zerocopy)
 357                 do_sendpage = sock_no_sendpage;
 358         for(;;) {
 359                 unsigned int len = min_t(unsigned int, PAGE_SIZE - base, remainder);
 360                 int flags = XS_SENDMSG_FLAGS;
 361
 362                 remainder -= len;
 363                 if (remainder != 0 || more)
 364                         flags |= MSG_MORE;
 365                 err = do_sendpage(sock, *ppage, base, len, flags);
 366                 if (remainder == 0 || err != len)
 367                         break;
 368                 *sent_p += err;
 369                 ppage++;
 370                 base = 0;
 371         }
 372         if (err > 0) {
 373                 *sent_p += err;
 374                 err = 0;
 375         }
 376         return err;
 377 }
 378
 379 /**
 380  * xs_sendpages - write pages directly to a socket
 381  * @sock: socket to send on
 382  * @addr: UDP only -- address of destination
 383  * @addrlen: UDP only -- length of destination address
 384  * @xdr: buffer containing this request
 385  * @base: starting position in the buffer
 386  * @zerocopy: true if it is safe to use sendpage()
 387  * @sent_p: return the total number of bytes successfully queued for sending
 388  *
 389  */
 390 static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, bool zerocopy, int *sent_p)
 391 {
 392         unsigned int remainder = xdr->len - base;
 393         int err = 0;
 394         int sent = 0;
 395
 396         if (unlikely(!sock))
 397                 return -ENOTSOCK;
 398
 399         clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags);
 400         if (base != 0) {
 401                 addr = NULL;
 402                 addrlen = 0;
 403         }
 404
 405         if (base < xdr->head[0].iov_len || addr != NULL) {
 406                 unsigned int len = xdr->head[0].iov_len - base;
 407                 remainder -= len;
 408                 err = xs_send_kvec(sock, addr, addrlen, &xdr->head[0], base, remainder != 0);
 409                 if (remainder == 0 || err != len)
 410                         goto out;
 411                 *sent_p += err;
 412                 base = 0;
 413         } else
 414                 base -= xdr->head[0].iov_len;
 415
 416         if (base < xdr->page_len) {
 417                 unsigned int len = xdr->page_len - base;
 418                 remainder -= len;
 419                 err = xs_send_pagedata(sock, xdr, base, remainder != 0, zerocopy, &sent);
 420                 *sent_p += sent;
 421                 if (remainder == 0 || sent != len)
 422                         goto out;
 423                 base = 0;
 424         } else
 425                 base -= xdr->page_len;
 426
 427         if (base >= xdr->tail[0].iov_len)
 428                 return 0;
 429         err = xs_send_kvec(sock, NULL, 0, &xdr->tail[0], base, 0);
 430 out:
 431         if (err > 0) {
 432                 *sent_p += err;
 433                 err = 0;
 434         }
 435         return err;
 436 }
 437
 438 static void xs_nospace_callback(struct rpc_task *task)
 439 {
 440         struct sock_xprt *transport = container_of(task->tk_rqstp->rq_xprt, struct sock_xprt, xprt);
 441
 442         transport->inet->sk_write_pending--;
 443         clear_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags);
 444 }
 445
 446 /**
 447  * xs_nospace - place task on wait queue if transmit was incomplete
 448  * @task: task to put to sleep
 449  *
 450  */
 451 static int xs_nospace(struct rpc_task *task)
 452 {
 453         struct rpc_rqst *req = task->tk_rqstp;
 454         struct rpc_xprt *xprt = req->rq_xprt;
 455         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
 456         struct sock *sk = transport->inet;
 457         int ret = -EAGAIN;
 458
 459         dprintk("RPC: %5u xmit incomplete (%u left of %u)\n",
 460                         task->tk_pid, req->rq_slen - req->rq_bytes_sent,
 461                         req->rq_slen);
 462
 463         /* Protect against races with write_space */
 464         spin_lock_bh(&xprt->transport_lock);
 465
 466         /* Don't race with disconnect */
 467         if (xprt_connected(xprt)) {
 468                 if (test_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags)) {
 469                         /*
 470                          * Notify TCP that we're limited by the application
 471                          * window size
 472                          */
 473                         set_bit(SOCK_NOSPACE, &transport->sock->flags);
 474                         sk->sk_write_pending++;
 475                         /* ...and wait for more buffer space */
 476                         xprt_wait_for_buffer_space(task, xs_nospace_callback);
 477                 }
 478         } else {
 479                 clear_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags);
 480                 ret = -ENOTCONN;
 481         }
 482
 483         spin_unlock_bh(&xprt->transport_lock);
 484
 485         /* Race breaker in case memory is freed before above code is called */
 486         sk->sk_write_space(sk);
 487         return ret;
 488 }
 489
 490 /*
 491  * Construct a stream transport record marker in @buf.
 492  */
 493 static inline void xs_encode_stream_record_marker(struct xdr_buf *buf)
 494 {
 495         u32 reclen = buf->len - sizeof(rpc_fraghdr);
 496         rpc_fraghdr *base = buf->head[0].iov_base;
 497         *base = cpu_to_be32(RPC_LAST_STREAM_FRAGMENT | reclen);
 498 }
 499
 500 /**
 501  * xs_local_send_request - write an RPC request to an AF_LOCAL socket
 502  * @task: RPC task that manages the state of an RPC request
 503  *
 504  * Return values:
 505  *        0:    The request has been sent
 506  *   EAGAIN:    The socket was blocked, please call again later to
 507  *              complete the request
 508  * ENOTCONN:    Caller needs to invoke connect logic then call again
 509  *    other:    Some other error occured, the request was not sent
 510  */
 511 static int xs_local_send_request(struct rpc_task *task)
 512 {
 513         struct rpc_rqst *req = task->tk_rqstp;
 514         struct rpc_xprt *xprt = req->rq_xprt;
 515         struct sock_xprt *transport =
 516                                 container_of(xprt, struct sock_xprt, xprt);
 517         struct xdr_buf *xdr = &req->rq_snd_buf;
 518         int status;
 519         int sent = 0;
 520
 521         xs_encode_stream_record_marker(&req->rq_snd_buf);
 522
 523         xs_pktdump("packet data:",
 524                         req->rq_svec->iov_base, req->rq_svec->iov_len);
 525
 526         status = xs_sendpages(transport->sock, NULL, 0, xdr, req->rq_bytes_sent,
 527                               true, &sent);
 528         dprintk("RPC:       %s(%u) = %d\n",
 529                         __func__, xdr->len - req->rq_bytes_sent, status);
 530         if (likely(sent > 0) || status == 0) {
 531                 req->rq_bytes_sent += sent;
 532                 req->rq_xmit_bytes_sent += sent;
 533                 if (likely(req->rq_bytes_sent >= req->rq_slen)) {
 534                         req->rq_bytes_sent = 0;
 535                         return 0;
 536                 }
 537                 status = -EAGAIN;
 538         }
 539
 540         switch (status) {
 541         case -ENOBUFS:
 542         case -EAGAIN:
 543                 status = xs_nospace(task);
 544                 break;
 545         default:
 546                 dprintk("RPC:       sendmsg returned unrecognized error %d\n",
 547                         -status);
 548         case -EPIPE:
 549                 xs_close(xprt);
 550                 status = -ENOTCONN;
 551         }
 552
 553         return status;
 554 }
 555
 556 /**
 557  * xs_udp_send_request - write an RPC request to a UDP socket
 558  * @task: address of RPC task that manages the state of an RPC request
 559  *
 560  * Return values:
 561  *        0:    The request has been sent
 562  *   EAGAIN:    The socket was blocked, please call again later to
 563  *              complete the request
 564  * ENOTCONN:    Caller needs to invoke connect logic then call again
 565  *    other:    Some other error occurred, the request was not sent
 566  */
 567 static int xs_udp_send_request(struct rpc_task *task)
 568 {
 569         struct rpc_rqst *req = task->tk_rqstp;
 570         struct rpc_xprt *xprt = req->rq_xprt;
 571         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
 572         struct xdr_buf *xdr = &req->rq_snd_buf;
 573         int sent = 0;
 574         int status;
 575
 576         xs_pktdump("packet data:",
 577                                 req->rq_svec->iov_base,
 578                                 req->rq_svec->iov_len);
 579
 580         if (!xprt_bound(xprt))
 581                 return -ENOTCONN;
 582         status = xs_sendpages(transport->sock, xs_addr(xprt), xprt->addrlen,
 583                               xdr, req->rq_bytes_sent, true, &sent);
 584
 585         dprintk("RPC:       xs_udp_send_request(%u) = %d\n",
 586                         xdr->len - req->rq_bytes_sent, status);
 587
 588         /* firewall is blocking us, don't return -EAGAIN or we end up looping */
 589         if (status == -EPERM)
 590                 goto process_status;
 591
 592         if (sent > 0 || status == 0) {
 593                 req->rq_xmit_bytes_sent += sent;
 594                 if (sent >= req->rq_slen)
 595                         return 0;
 596                 /* Still some bytes left; set up for a retry later. */
 597                 status = -EAGAIN;
 598         }
 599
 600 process_status:
 601         switch (status) {
 602         case -ENOTSOCK:
 603                 status = -ENOTCONN;
 604                 /* Should we call xs_close() here? */
 605                 break;
 606         case -EAGAIN:
 607                 status = xs_nospace(task);
 608                 break;
 609         default:
 610                 dprintk("RPC:       sendmsg returned unrecognized error %d\n",
 611                         -status);
 612         case -ENETUNREACH:
 613         case -ENOBUFS:
 614         case -EPIPE:
 615         case -ECONNREFUSED:
 616         case -EPERM:
 617                 /* When the server has died, an ICMP port unreachable message
 618                  * prompts ECONNREFUSED. */
 619                 clear_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags);
 620         }
 621
 622         return status;
 623 }
 624
 625 /**
 626  * xs_tcp_shutdown - gracefully shut down a TCP socket
 627  * @xprt: transport
 628  *
 629  * Initiates a graceful shutdown of the TCP socket by calling the
 630  * equivalent of shutdown(SHUT_WR);
 631  */
 632 static void xs_tcp_shutdown(struct rpc_xprt *xprt)
 633 {
 634         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
 635         struct socket *sock = transport->sock;
 636
 637         if (sock != NULL) {
 638                 kernel_sock_shutdown(sock, SHUT_WR);
 639                 trace_rpc_socket_shutdown(xprt, sock);
 640         }
 641 }
 642
 643 /**
 644  * xs_tcp_send_request - write an RPC request to a TCP socket
 645  * @task: address of RPC task that manages the state of an RPC request
 646  *
 647  * Return values:
 648  *        0:    The request has been sent
 649  *   EAGAIN:    The socket was blocked, please call again later to
 650  *              complete the request
 651  * ENOTCONN:    Caller needs to invoke connect logic then call again
 652  *    other:    Some other error occurred, the request was not sent
 653  *
 654  * XXX: In the case of soft timeouts, should we eventually give up
 655  *      if sendmsg is not able to make progress?
 656  */
 657 static int xs_tcp_send_request(struct rpc_task *task)
 658 {
 659         struct rpc_rqst *req = task->tk_rqstp;
 660         struct rpc_xprt *xprt = req->rq_xprt;
 661         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
 662         struct xdr_buf *xdr = &req->rq_snd_buf;
 663         bool zerocopy = true;
 664         int status;
 665         int sent;
 666
 667         xs_encode_stream_record_marker(&req->rq_snd_buf);
 668
 669         xs_pktdump("packet data:",
 670                                 req->rq_svec->iov_base,
 671                                 req->rq_svec->iov_len);
 672         /* Don't use zero copy if this is a resend. If the RPC call
 673          * completes while the socket holds a reference to the pages,
 674          * then we may end up resending corrupted data.
 675          */
 676         if (task->tk_flags & RPC_TASK_SENT)
 677                 zerocopy = false;
 678
 679         /* Continue transmitting the packet/record. We must be careful
 680          * to cope with writespace callbacks arriving _after_ we have
 681          * called sendmsg(). */
 682         while (1) {
 683                 sent = 0;
 684                 status = xs_sendpages(transport->sock, NULL, 0, xdr,
 685                                       req->rq_bytes_sent, zerocopy, &sent);
 686
 687                 dprintk("RPC:       xs_tcp_send_request(%u) = %d\n",
 688                                 xdr->len - req->rq_bytes_sent, status);
 689
 690                 if (unlikely(sent == 0 && status < 0))
 691                         break;
 692
 693                 /* If we've sent the entire packet, immediately
 694                  * reset the count of bytes sent. */
 695                 req->rq_bytes_sent += sent;
 696                 req->rq_xmit_bytes_sent += sent;
 697                 if (likely(req->rq_bytes_sent >= req->rq_slen)) {
 698                         req->rq_bytes_sent = 0;
 699                         return 0;
 700                 }
 701
 702                 if (sent != 0)
 703                         continue;
 704                 status = -EAGAIN;
 705                 break;
 706         }
 707
 708         switch (status) {
 709         case -ENOTSOCK:
 710                 status = -ENOTCONN;
 711                 /* Should we call xs_close() here? */
 712                 break;
 713         case -ENOBUFS:
 714         case -EAGAIN:
 715                 status = xs_nospace(task);
 716                 break;
 717         default:
 718                 dprintk("RPC:       sendmsg returned unrecognized error %d\n",
 719                         -status);
 720         case -ECONNRESET:
 721                 xs_tcp_shutdown(xprt);
 722         case -ECONNREFUSED:
 723         case -ENOTCONN:
 724         case -EADDRINUSE:
 725         case -EPIPE:
 726                 clear_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags);
 727         }
 728
 729         return status;
 730 }
 731
 732 /**
 733  * xs_tcp_release_xprt - clean up after a tcp transmission
 734  * @xprt: transport
 735  * @task: rpc task
 736  *
 737  * This cleans up if an error causes us to abort the transmission of a request.
 738  * In this case, the socket may need to be reset in order to avoid confusing
 739  * the server.
 740  */
 741 static void xs_tcp_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 742 {
 743         struct rpc_rqst *req;
 744
 745         if (task != xprt->snd_task)
 746                 return;
 747         if (task == NULL)
 748                 goto out_release;
 749         req = task->tk_rqstp;
 750         if (req == NULL)
 751                 goto out_release;
 752         if (req->rq_bytes_sent == 0)
 753                 goto out_release;
 754         if (req->rq_bytes_sent == req->rq_snd_buf.len)
 755                 goto out_release;
 756         set_bit(XPRT_CLOSE_WAIT, &xprt->state);
 757 out_release:
 758         xprt_release_xprt(xprt, task);
 759 }
 760
 761 static void xs_save_old_callbacks(struct sock_xprt *transport, struct sock *sk)
 762 {
 763         transport->old_data_ready = sk->sk_data_ready;
 764         transport->old_state_change = sk->sk_state_change;
 765         transport->old_write_space = sk->sk_write_space;
 766         transport->old_error_report = sk->sk_error_report;
 767 }
 768
 769 static void xs_restore_old_callbacks(struct sock_xprt *transport, struct sock *sk)
 770 {
 771         sk->sk_data_ready = transport->old_data_ready;
 772         sk->sk_state_change = transport->old_state_change;
 773         sk->sk_write_space = transport->old_write_space;
 774         sk->sk_error_report = transport->old_error_report;
 775 }
 776
 777 /**
 778  * xs_error_report - callback to handle TCP socket state errors
 779  * @sk: socket
 780  *
 781  * Note: we don't call sock_error() since there may be a rpc_task
 782  * using the socket, and so we don't want to clear sk->sk_err.
 783  */
 784 static void xs_error_report(struct sock *sk)
 785 {
 786         struct rpc_xprt *xprt;
 787         int err;
 788
 789         read_lock_bh(&sk->sk_callback_lock);
 790         if (!(xprt = xprt_from_sock(sk)))
 791                 goto out;
 792
 793         err = -sk->sk_err;
 794         if (err == 0)
 795                 goto out;
 796         dprintk("RPC:       xs_error_report client %p, error=%d...\n",
 797                         xprt, -err);
 798         trace_rpc_socket_error(xprt, sk->sk_socket, err);
 799         xprt_wake_pending_tasks(xprt, err);
 800  out:
 801         read_unlock_bh(&sk->sk_callback_lock);
 802 }
 803
 804 static void xs_sock_reset_connection_flags(struct rpc_xprt *xprt)
 805 {
 806         smp_mb__before_atomic();
 807         clear_bit(XPRT_CONNECTION_CLOSE, &xprt->state);
 808         clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
 809         clear_bit(XPRT_CLOSING, &xprt->state);
 810         smp_mb__after_atomic();
 811 }
 812
 813 static void xs_reset_transport(struct sock_xprt *transport)
 814 {
 815         struct socket *sock = transport->sock;
 816         struct sock *sk = transport->inet;
 817         struct rpc_xprt *xprt = &transport->xprt;
 818
 819         if (sk == NULL)
 820                 return;
 821
 822         write_lock_bh(&sk->sk_callback_lock);
 823         transport->inet = NULL;
 824         transport->sock = NULL;
 825
 826         sk->sk_user_data = NULL;
 827
 828         xs_restore_old_callbacks(transport, sk);
 829         write_unlock_bh(&sk->sk_callback_lock);
 830         xs_sock_reset_connection_flags(xprt);
 831
 832         trace_rpc_socket_close(xprt, sock);
 833         sock_release(sock);
 834 }
 835
 836 /**
 837  * xs_close - close a socket
 838  * @xprt: transport
 839  *
 840  * This is used when all requests are complete; ie, no DRC state remains
 841  * on the server we want to save.
 842  *
 843  * The caller _must_ be holding XPRT_LOCKED in order to avoid issues with
 844  * xs_reset_transport() zeroing the socket from underneath a writer.
 845  */
 846 static void xs_close(struct rpc_xprt *xprt)
 847 {
 848         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
 849
 850         dprintk("RPC:       xs_close xprt %p\n", xprt);
 851
 852         xs_reset_transport(transport);
 853         xprt->reestablish_timeout = 0;
 854
 855         xprt_disconnect_done(xprt);
 856 }
 857
 858 static void xs_tcp_close(struct rpc_xprt *xprt)
 859 {
 860         if (test_and_clear_bit(XPRT_CONNECTION_CLOSE, &xprt->state))
 861                 xs_close(xprt);
 862         else
 863                 xs_tcp_shutdown(xprt);
 864 }
 865
 866 static void xs_xprt_free(struct rpc_xprt *xprt)
 867 {
 868         xs_free_peer_addresses(xprt);
 869         xprt_free(xprt);
 870 }
 871
 872 /**
 873  * xs_destroy - prepare to shutdown a transport
 874  * @xprt: doomed transport
 875  *
 876  */
 877 static void xs_destroy(struct rpc_xprt *xprt)
 878 {
 879         dprintk("RPC:       xs_destroy xprt %p\n", xprt);
 880
 881         xs_close(xprt);
 882         xs_xprt_free(xprt);
 883         module_put(THIS_MODULE);
 884 }
 885
 886 static int xs_local_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb)
 887 {
 888         struct xdr_skb_reader desc = {
 889                 .skb            = skb,
 890                 .offset         = sizeof(rpc_fraghdr),
 891                 .count          = skb->len - sizeof(rpc_fraghdr),
 892         };
 893
 894         if (xdr_partial_copy_from_skb(xdr, 0, &desc, xdr_skb_read_bits) < 0)
 895                 return -1;
 896         if (desc.count)
 897                 return -1;
 898         return 0;
 899 }
 900
 901 /**
 902  * xs_local_data_ready - "data ready" callback for AF_LOCAL sockets
 903  * @sk: socket with data to read
 904  * @len: how much data to read
 905  *
 906  * Currently this assumes we can read the whole reply in a single gulp.
 907  */
 908 static void xs_local_data_ready(struct sock *sk)
 909 {
 910         struct rpc_task *task;
 911         struct rpc_xprt *xprt;
 912         struct rpc_rqst *rovr;
 913         struct sk_buff *skb;
 914         int err, repsize, copied;
 915         u32 _xid;
 916         __be32 *xp;
 917
 918         read_lock_bh(&sk->sk_callback_lock);
 919         dprintk("RPC:       %s...\n", __func__);
 920         xprt = xprt_from_sock(sk);
 921         if (xprt == NULL)
 922                 goto out;
 923
 924         skb = skb_recv_datagram(sk, 0, 1, &err);
 925         if (skb == NULL)
 926                 goto out;
 927
 928         repsize = skb->len - sizeof(rpc_fraghdr);
 929         if (repsize < 4) {
 930                 dprintk("RPC:       impossible RPC reply size %d\n", repsize);
 931                 goto dropit;
 932         }
 933
 934         /* Copy the XID from the skb... */
 935         xp = skb_header_pointer(skb, sizeof(rpc_fraghdr), sizeof(_xid), &_xid);
 936         if (xp == NULL)
 937                 goto dropit;
 938
 939         /* Look up and lock the request corresponding to the given XID */
 940         spin_lock(&xprt->transport_lock);
 941         rovr = xprt_lookup_rqst(xprt, *xp);
 942         if (!rovr)
 943                 goto out_unlock;
 944         task = rovr->rq_task;
 945
 946         copied = rovr->rq_private_buf.buflen;
 947         if (copied > repsize)
 948                 copied = repsize;
 949
 950         if (xs_local_copy_to_xdr(&rovr->rq_private_buf, skb)) {
 951                 dprintk("RPC:       sk_buff copy failed\n");
 952                 goto out_unlock;
 953         }
 954
 955         xprt_complete_rqst(task, copied);
 956
 957  out_unlock:
 958         spin_unlock(&xprt->transport_lock);
 959  dropit:
 960         skb_free_datagram(sk, skb);
 961  out:
 962         read_unlock_bh(&sk->sk_callback_lock);
 963 }
 964
 965 /**
 966  * xs_udp_data_ready - "data ready" callback for UDP sockets
 967  * @sk: socket with data to read
 968  * @len: how much data to read
 969  *
 970  */
 971 static void xs_udp_data_ready(struct sock *sk)
 972 {
 973         struct rpc_task *task;
 974         struct rpc_xprt *xprt;
 975         struct rpc_rqst *rovr;
 976         struct sk_buff *skb;
 977         int err, repsize, copied;
 978         u32 _xid;
 979         __be32 *xp;
 980
 981         read_lock_bh(&sk->sk_callback_lock);
 982         dprintk("RPC:       xs_udp_data_ready...\n");
 983         if (!(xprt = xprt_from_sock(sk)))
 984                 goto out;
 985
 986         if ((skb = skb_recv_datagram(sk, 0, 1, &err)) == NULL)
 987                 goto out;
 988
 989         repsize = skb->len - sizeof(struct udphdr);
 990         if (repsize < 4) {
 991                 dprintk("RPC:       impossible RPC reply size %d!\n", repsize);
 992                 goto dropit;
 993         }
 994
 995         /* Copy the XID from the skb... */
 996         xp = skb_header_pointer(skb, sizeof(struct udphdr),
 997                                 sizeof(_xid), &_xid);
 998         if (xp == NULL)
 999                 goto dropit;
1000
1001         /* Look up and lock the request corresponding to the given XID */
1002         spin_lock(&xprt->transport_lock);
1003         rovr = xprt_lookup_rqst(xprt, *xp);
1004         if (!rovr)
1005                 goto out_unlock;
1006         task = rovr->rq_task;
1007
1008         if ((copied = rovr->rq_private_buf.buflen) > repsize)
1009                 copied = repsize;
1010
1011         /* Suck it into the iovec, verify checksum if not done by hw. */
1012         if (csum_partial_copy_to_xdr(&rovr->rq_private_buf, skb)) {
1013                 UDPX_INC_STATS_BH(sk, UDP_MIB_INERRORS);
1014                 goto out_unlock;
1015         }
1016
1017         UDPX_INC_STATS_BH(sk, UDP_MIB_INDATAGRAMS);
1018
1019         xprt_adjust_cwnd(xprt, task, copied);
1020         xprt_complete_rqst(task, copied);
1021
1022  out_unlock:
1023         spin_unlock(&xprt->transport_lock);
1024  dropit:
1025         skb_free_datagram(sk, skb);
1026  out:
1027         read_unlock_bh(&sk->sk_callback_lock);
1028 }
1029
1030 /*
1031  * Helper function to force a TCP close if the server is sending
1032  * junk and/or it has put us in CLOSE_WAIT
1033  */
1034 static void xs_tcp_force_close(struct rpc_xprt *xprt)
1035 {
1036         set_bit(XPRT_CONNECTION_CLOSE, &xprt->state);
1037         xprt_force_disconnect(xprt);
1038 }
1039
1040 static inline void xs_tcp_read_fraghdr(struct rpc_xprt *xprt, struct xdr_skb_reader *desc)
1041 {
1042         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1043         size_t len, used;
1044         char *p;
1045
1046         p = ((char *) &transport->tcp_fraghdr) + transport->tcp_offset;
1047         len = sizeof(transport->tcp_fraghdr) - transport->tcp_offset;
1048         used = xdr_skb_read_bits(desc, p, len);
1049         transport->tcp_offset += used;
1050         if (used != len)
1051                 return;
1052
1053         transport->tcp_reclen = ntohl(transport->tcp_fraghdr);
1054         if (transport->tcp_reclen & RPC_LAST_STREAM_FRAGMENT)
1055                 transport->tcp_flags |= TCP_RCV_LAST_FRAG;
1056         else
1057                 transport->tcp_flags &= ~TCP_RCV_LAST_FRAG;
1058         transport->tcp_reclen &= RPC_FRAGMENT_SIZE_MASK;
1059
1060         transport->tcp_flags &= ~TCP_RCV_COPY_FRAGHDR;
1061         transport->tcp_offset = 0;
1062
1063         /* Sanity check of the record length */
1064         if (unlikely(transport->tcp_reclen < 8)) {
1065                 dprintk("RPC:       invalid TCP record fragment length\n");
1066                 xs_tcp_force_close(xprt);
1067                 return;
1068         }
1069         dprintk("RPC:       reading TCP record fragment of length %d\n",
1070                         transport->tcp_reclen);
1071 }
1072
1073 static void xs_tcp_check_fraghdr(struct sock_xprt *transport)
1074 {
1075         if (transport->tcp_offset == transport->tcp_reclen) {
1076                 transport->tcp_flags |= TCP_RCV_COPY_FRAGHDR;
1077                 transport->tcp_offset = 0;
1078                 if (transport->tcp_flags & TCP_RCV_LAST_FRAG) {
1079                         transport->tcp_flags &= ~TCP_RCV_COPY_DATA;
1080                         transport->tcp_flags |= TCP_RCV_COPY_XID;
1081                         transport->tcp_copied = 0;
1082                 }
1083         }
1084 }
1085
1086 static inline void xs_tcp_read_xid(struct sock_xprt *transport, struct xdr_skb_reader *desc)
1087 {
1088         size_t len, used;
1089         char *p;
1090
1091         len = sizeof(transport->tcp_xid) - transport->tcp_offset;
1092         dprintk("RPC:       reading XID (%Zu bytes)\n", len);
1093         p = ((char *) &transport->tcp_xid) + transport->tcp_offset;
1094         used = xdr_skb_read_bits(desc, p, len);
1095         transport->tcp_offset += used;
1096         if (used != len)
1097                 return;
1098         transport->tcp_flags &= ~TCP_RCV_COPY_XID;
1099         transport->tcp_flags |= TCP_RCV_READ_CALLDIR;
1100         transport->tcp_copied = 4;
1101         dprintk("RPC:       reading %s XID %08x\n",
1102                         (transport->tcp_flags & TCP_RPC_REPLY) ? "reply for"
1103                                                               : "request with",
1104                         ntohl(transport->tcp_xid));
1105         xs_tcp_check_fraghdr(transport);
1106 }
1107
1108 static inline void xs_tcp_read_calldir(struct sock_xprt *transport,
1109                                        struct xdr_skb_reader *desc)
1110 {
1111         size_t len, used;
1112         u32 offset;
1113         char *p;
1114
1115         /*
1116          * We want transport->tcp_offset to be 8 at the end of this routine
1117          * (4 bytes for the xid and 4 bytes for the call/reply flag).
1118          * When this function is called for the first time,
1119          * transport->tcp_offset is 4 (after having already read the xid).
1120          */
1121         offset = transport->tcp_offset - sizeof(transport->tcp_xid);
1122         len = sizeof(transport->tcp_calldir) - offset;
1123         dprintk("RPC:       reading CALL/REPLY flag (%Zu bytes)\n", len);
1124         p = ((char *) &transport->tcp_calldir) + offset;
1125         used = xdr_skb_read_bits(desc, p, len);
1126         transport->tcp_offset += used;
1127         if (used != len)
1128                 return;
1129         transport->tcp_flags &= ~TCP_RCV_READ_CALLDIR;
1130         /*
1131          * We don't yet have the XDR buffer, so we will write the calldir
1132          * out after we get the buffer from the 'struct rpc_rqst'
1133          */
1134         switch (ntohl(transport->tcp_calldir)) {
1135         case RPC_REPLY:
1136                 transport->tcp_flags |= TCP_RCV_COPY_CALLDIR;
1137                 transport->tcp_flags |= TCP_RCV_COPY_DATA;
1138                 transport->tcp_flags |= TCP_RPC_REPLY;
1139                 break;
1140         case RPC_CALL:
1141                 transport->tcp_flags |= TCP_RCV_COPY_CALLDIR;
1142                 transport->tcp_flags |= TCP_RCV_COPY_DATA;
1143                 transport->tcp_flags &= ~TCP_RPC_REPLY;
1144                 break;
1145         default:
1146                 dprintk("RPC:       invalid request message type\n");
1147                 xs_tcp_force_close(&transport->xprt);
1148         }
1149         xs_tcp_check_fraghdr(transport);
1150 }
1151
1152 static inline void xs_tcp_read_common(struct rpc_xprt *xprt,
1153                                      struct xdr_skb_reader *desc,
1154                                      struct rpc_rqst *req)
1155 {
1156         struct sock_xprt *transport =
1157                                 container_of(xprt, struct sock_xprt, xprt);
1158         struct xdr_buf *rcvbuf;
1159         size_t len;
1160         ssize_t r;
1161
1162         rcvbuf = &req->rq_private_buf;
1163
1164         if (transport->tcp_flags & TCP_RCV_COPY_CALLDIR) {
1165                 /*
1166                  * Save the RPC direction in the XDR buffer
1167                  */
1168                 memcpy(rcvbuf->head[0].iov_base + transport->tcp_copied,
1169                         &transport->tcp_calldir,
1170                         sizeof(transport->tcp_calldir));
1171                 transport->tcp_copied += sizeof(transport->tcp_calldir);
1172                 transport->tcp_flags &= ~TCP_RCV_COPY_CALLDIR;
1173         }
1174
1175         len = desc->count;
1176         if (len > transport->tcp_reclen - transport->tcp_offset) {
1177                 struct xdr_skb_reader my_desc;
1178
1179                 len = transport->tcp_reclen - transport->tcp_offset;
1180                 memcpy(&my_desc, desc, sizeof(my_desc));
1181                 my_desc.count = len;
1182                 r = xdr_partial_copy_from_skb(rcvbuf, transport->tcp_copied,
1183                                           &my_desc, xdr_skb_read_bits);
1184                 desc->count -= r;
1185                 desc->offset += r;
1186         } else
1187                 r = xdr_partial_copy_from_skb(rcvbuf, transport->tcp_copied,
1188                                           desc, xdr_skb_read_bits);
1189
1190         if (r > 0) {
1191                 transport->tcp_copied += r;
1192                 transport->tcp_offset += r;
1193         }
1194         if (r != len) {
1195                 /* Error when copying to the receive buffer,
1196                  * usually because we weren't able to allocate
1197                  * additional buffer pages. All we can do now
1198                  * is turn off TCP_RCV_COPY_DATA, so the request
1199                  * will not receive any additional updates,
1200                  * and time out.
1201                  * Any remaining data from this record will
1202                  * be discarded.
1203                  */
1204                 transport->tcp_flags &= ~TCP_RCV_COPY_DATA;
1205                 dprintk("RPC:       XID %08x truncated request\n",
1206                                 ntohl(transport->tcp_xid));
1207                 dprintk("RPC:       xprt = %p, tcp_copied = %lu, "
1208                                 "tcp_offset = %u, tcp_reclen = %u\n",
1209                                 xprt, transport->tcp_copied,
1210                                 transport->tcp_offset, transport->tcp_reclen);
1211                 return;
1212         }
1213
1214         dprintk("RPC:       XID %08x read %Zd bytes\n",
1215                         ntohl(transport->tcp_xid), r);
1216         dprintk("RPC:       xprt = %p, tcp_copied = %lu, tcp_offset = %u, "
1217                         "tcp_reclen = %u\n", xprt, transport->tcp_copied,
1218                         transport->tcp_offset, transport->tcp_reclen);
1219
1220         if (transport->tcp_copied == req->rq_private_buf.buflen)
1221                 transport->tcp_flags &= ~TCP_RCV_COPY_DATA;
1222         else if (transport->tcp_offset == transport->tcp_reclen) {
1223                 if (transport->tcp_flags & TCP_RCV_LAST_FRAG)
1224                         transport->tcp_flags &= ~TCP_RCV_COPY_DATA;
1225         }
1226 }
1227
1228 /*
1229  * Finds the request corresponding to the RPC xid and invokes the common
1230  * tcp read code to read the data.
1231  */
1232 static inline int xs_tcp_read_reply(struct rpc_xprt *xprt,
1233                                     struct xdr_skb_reader *desc)
1234 {
1235         struct sock_xprt *transport =
1236                                 container_of(xprt, struct sock_xprt, xprt);
1237         struct rpc_rqst *req;
1238
1239         dprintk("RPC:       read reply XID %08x\n", ntohl(transport->tcp_xid));
1240
1241         /* Find and lock the request corresponding to this xid */
1242         spin_lock(&xprt->transport_lock);
1243         req = xprt_lookup_rqst(xprt, transport->tcp_xid);
1244         if (!req) {
1245                 dprintk("RPC:       XID %08x request not found!\n",
1246                                 ntohl(transport->tcp_xid));
1247                 spin_unlock(&xprt->transport_lock);
1248                 return -1;
1249         }
1250
1251         xs_tcp_read_common(xprt, desc, req);
1252
1253         if (!(transport->tcp_flags & TCP_RCV_COPY_DATA))
1254                 xprt_complete_rqst(req->rq_task, transport->tcp_copied);
1255
1256         spin_unlock(&xprt->transport_lock);
1257         return 0;
1258 }
1259
1260 #if defined(CONFIG_SUNRPC_BACKCHANNEL)
1261 /*
1262  * Obtains an rpc_rqst previously allocated and invokes the common
1263  * tcp read code to read the data.  The result is placed in the callback
1264  * queue.
1265  * If we're unable to obtain the rpc_rqst we schedule the closing of the
1266  * connection and return -1.
1267  */
1268 static int xs_tcp_read_callback(struct rpc_xprt *xprt,
1269                                        struct xdr_skb_reader *desc)
1270 {
1271         struct sock_xprt *transport =
1272                                 container_of(xprt, struct sock_xprt, xprt);
1273         struct rpc_rqst *req;
1274
1275         /* Look up and lock the request corresponding to the given XID */
1276         spin_lock(&xprt->transport_lock);
1277         req = xprt_lookup_bc_request(xprt, transport->tcp_xid);
1278         if (req == NULL) {
1279                 spin_unlock(&xprt->transport_lock);
1280                 printk(KERN_WARNING "Callback slot table overflowed\n");
1281                 xprt_force_disconnect(xprt);
1282                 return -1;
1283         }
1284
1285         dprintk("RPC:       read callback  XID %08x\n", ntohl(req->rq_xid));
1286         xs_tcp_read_common(xprt, desc, req);
1287
1288         if (!(transport->tcp_flags & TCP_RCV_COPY_DATA))
1289                 xprt_complete_bc_request(req, transport->tcp_copied);
1290         spin_unlock(&xprt->transport_lock);
1291
1292         return 0;
1293 }
1294
1295 static inline int _xs_tcp_read_data(struct rpc_xprt *xprt,
1296                                         struct xdr_skb_reader *desc)
1297 {
1298         struct sock_xprt *transport =
1299                                 container_of(xprt, struct sock_xprt, xprt);
1300
1301         return (transport->tcp_flags & TCP_RPC_REPLY) ?
1302                 xs_tcp_read_reply(xprt, desc) :
1303                 xs_tcp_read_callback(xprt, desc);
1304 }
1305 #else
1306 static inline int _xs_tcp_read_data(struct rpc_xprt *xprt,
1307                                         struct xdr_skb_reader *desc)
1308 {
1309         return xs_tcp_read_reply(xprt, desc);
1310 }
1311 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
1312
1313 /*
1314  * Read data off the transport.  This can be either an RPC_CALL or an
1315  * RPC_REPLY.  Relay the processing to helper functions.
1316  */
1317 static void xs_tcp_read_data(struct rpc_xprt *xprt,
1318                                     struct xdr_skb_reader *desc)
1319 {
1320         struct sock_xprt *transport =
1321                                 container_of(xprt, struct sock_xprt, xprt);
1322
1323         if (_xs_tcp_read_data(xprt, desc) == 0)
1324                 xs_tcp_check_fraghdr(transport);
1325         else {
1326                 /*
1327                  * The transport_lock protects the request handling.
1328                  * There's no need to hold it to update the tcp_flags.
1329                  */
1330                 transport->tcp_flags &= ~TCP_RCV_COPY_DATA;
1331         }
1332 }
1333
1334 static inline void xs_tcp_read_discard(struct sock_xprt *transport, struct xdr_skb_reader *desc)
1335 {
1336         size_t len;
1337
1338         len = transport->tcp_reclen - transport->tcp_offset;
1339         if (len > desc->count)
1340                 len = desc->count;
1341         desc->count -= len;
1342         desc->offset += len;
1343         transport->tcp_offset += len;
1344         dprintk("RPC:       discarded %Zu bytes\n", len);
1345         xs_tcp_check_fraghdr(transport);
1346 }
1347
1348 static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, unsigned int offset, size_t len)
1349 {
1350         struct rpc_xprt *xprt = rd_desc->arg.data;
1351         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1352         struct xdr_skb_reader desc = {
1353                 .skb    = skb,
1354                 .offset = offset,
1355                 .count  = len,
1356         };
1357
1358         dprintk("RPC:       xs_tcp_data_recv started\n");
1359         do {
1360                 trace_xs_tcp_data_recv(transport);
1361                 /* Read in a new fragment marker if necessary */
1362                 /* Can we ever really expect to get completely empty fragments? */
1363                 if (transport->tcp_flags & TCP_RCV_COPY_FRAGHDR) {
1364                         xs_tcp_read_fraghdr(xprt, &desc);
1365                         continue;
1366                 }
1367                 /* Read in the xid if necessary */
1368                 if (transport->tcp_flags & TCP_RCV_COPY_XID) {
1369                         xs_tcp_read_xid(transport, &desc);
1370                         continue;
1371                 }
1372                 /* Read in the call/reply flag */
1373                 if (transport->tcp_flags & TCP_RCV_READ_CALLDIR) {
1374                         xs_tcp_read_calldir(transport, &desc);
1375                         continue;
1376                 }
1377                 /* Read in the request data */
1378                 if (transport->tcp_flags & TCP_RCV_COPY_DATA) {
1379                         xs_tcp_read_data(xprt, &desc);
1380                         continue;
1381                 }
1382                 /* Skip over any trailing bytes on short reads */
1383                 xs_tcp_read_discard(transport, &desc);
1384         } while (desc.count);
1385         trace_xs_tcp_data_recv(transport);
1386         dprintk("RPC:       xs_tcp_data_recv done\n");
1387         return len - desc.count;
1388 }
1389
1390 /**
1391  * xs_tcp_data_ready - "data ready" callback for TCP sockets
1392  * @sk: socket with data to read
1393  * @bytes: how much data to read
1394  *
1395  */
1396 static void xs_tcp_data_ready(struct sock *sk)
1397 {
1398         struct rpc_xprt *xprt;
1399         read_descriptor_t rd_desc;
1400         int read;
1401         unsigned long total = 0;
1402
1403         dprintk("RPC:       xs_tcp_data_ready...\n");
1404
1405         read_lock_bh(&sk->sk_callback_lock);
1406         if (!(xprt = xprt_from_sock(sk))) {
1407                 read = 0;
1408                 goto out;
1409         }
1410         /* Any data means we had a useful conversation, so
1411          * the we don't need to delay the next reconnect
1412          */
1413         if (xprt->reestablish_timeout)
1414                 xprt->reestablish_timeout = 0;
1415
1416         /* We use rd_desc to pass struct xprt to xs_tcp_data_recv */
1417         rd_desc.arg.data = xprt;
1418         do {
1419                 rd_desc.count = 65536;
1420                 read = tcp_read_sock(sk, &rd_desc, xs_tcp_data_recv);
1421                 if (read > 0)
1422                         total += read;
1423         } while (read > 0);
1424 out:
1425         trace_xs_tcp_data_ready(xprt, read, total);
1426         read_unlock_bh(&sk->sk_callback_lock);
1427 }
1428
1429 static void xs_sock_mark_closed(struct rpc_xprt *xprt)
1430 {
1431         xs_sock_reset_connection_flags(xprt);
1432         /* Mark transport as closed and wake up all pending tasks */
1433         xprt_disconnect_done(xprt);
1434 }
1435
1436 /**
1437  * xs_tcp_state_change - callback to handle TCP socket state changes
1438  * @sk: socket whose state has changed
1439  *
1440  */
1441 static void xs_tcp_state_change(struct sock *sk)
1442 {
1443         struct rpc_xprt *xprt;
1444
1445         read_lock_bh(&sk->sk_callback_lock);
1446         if (!(xprt = xprt_from_sock(sk)))
1447                 goto out;
1448         dprintk("RPC:       xs_tcp_state_change client %p...\n", xprt);
1449         dprintk("RPC:       state %x conn %d dead %d zapped %d sk_shutdown %d\n",
1450                         sk->sk_state, xprt_connected(xprt),
1451                         sock_flag(sk, SOCK_DEAD),
1452                         sock_flag(sk, SOCK_ZAPPED),
1453                         sk->sk_shutdown);
1454
1455         trace_rpc_socket_state_change(xprt, sk->sk_socket);
1456         switch (sk->sk_state) {
1457         case TCP_ESTABLISHED:
1458                 spin_lock(&xprt->transport_lock);
1459                 if (!xprt_test_and_set_connected(xprt)) {
1460                         struct sock_xprt *transport = container_of(xprt,
1461                                         struct sock_xprt, xprt);
1462
1463                         /* Reset TCP record info */
1464                         transport->tcp_offset = 0;
1465                         transport->tcp_reclen = 0;
1466                         transport->tcp_copied = 0;
1467                         transport->tcp_flags =
1468                                 TCP_RCV_COPY_FRAGHDR | TCP_RCV_COPY_XID;
1469                         xprt->connect_cookie++;
1470
1471                         xprt_wake_pending_tasks(xprt, -EAGAIN);
1472                 }
1473                 spin_unlock(&xprt->transport_lock);
1474                 break;
1475         case TCP_FIN_WAIT1:
1476                 /* The client initiated a shutdown of the socket */
1477                 xprt->connect_cookie++;
1478                 xprt->reestablish_timeout = 0;
1479                 set_bit(XPRT_CLOSING, &xprt->state);
1480                 smp_mb__before_atomic();
1481                 clear_bit(XPRT_CONNECTED, &xprt->state);
1482                 clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
1483                 smp_mb__after_atomic();
1484                 break;
1485         case TCP_CLOSE_WAIT:
1486                 /* The server initiated a shutdown of the socket */
1487                 xprt->connect_cookie++;
1488                 clear_bit(XPRT_CONNECTED, &xprt->state);
1489                 xs_tcp_force_close(xprt);
1490         case TCP_CLOSING:
1491                 /*
1492                  * If the server closed down the connection, make sure that
1493                  * we back off before reconnecting
1494                  */
1495                 if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
1496                         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
1497                 break;
1498         case TCP_LAST_ACK:
1499                 set_bit(XPRT_CLOSING, &xprt->state);
1500                 smp_mb__before_atomic();
1501                 clear_bit(XPRT_CONNECTED, &xprt->state);
1502                 smp_mb__after_atomic();
1503                 break;
1504         case TCP_CLOSE:
1505                 xs_sock_mark_closed(xprt);
1506         }
1507  out:
1508         read_unlock_bh(&sk->sk_callback_lock);
1509 }
1510
1511 static void xs_write_space(struct sock *sk)
1512 {
1513         struct socket *sock;
1514         struct rpc_xprt *xprt;
1515
1516         if (unlikely(!(sock = sk->sk_socket)))
1517                 return;
1518         clear_bit(SOCK_NOSPACE, &sock->flags);
1519
1520         if (unlikely(!(xprt = xprt_from_sock(sk))))
1521                 return;
1522         if (test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags) == 0)
1523                 return;
1524
1525         xprt_write_space(xprt);
1526 }
1527
1528 /**
1529  * xs_udp_write_space - callback invoked when socket buffer space
1530  *                             becomes available
1531  * @sk: socket whose state has changed
1532  *
1533  * Called when more output buffer space is available for this socket.
1534  * We try not to wake our writers until they can make "significant"
1535  * progress, otherwise we'll waste resources thrashing kernel_sendmsg
1536  * with a bunch of small requests.
1537  */
1538 static void xs_udp_write_space(struct sock *sk)
1539 {
1540         read_lock_bh(&sk->sk_callback_lock);
1541
1542         /* from net/core/sock.c:sock_def_write_space */
1543         if (sock_writeable(sk))
1544                 xs_write_space(sk);
1545
1546         read_unlock_bh(&sk->sk_callback_lock);
1547 }
1548
1549 /**
1550  * xs_tcp_write_space - callback invoked when socket buffer space
1551  *                             becomes available
1552  * @sk: socket whose state has changed
1553  *
1554  * Called when more output buffer space is available for this socket.
1555  * We try not to wake our writers until they can make "significant"
1556  * progress, otherwise we'll waste resources thrashing kernel_sendmsg
1557  * with a bunch of small requests.
1558  */
1559 static void xs_tcp_write_space(struct sock *sk)
1560 {
1561         read_lock_bh(&sk->sk_callback_lock);
1562
1563         /* from net/core/stream.c:sk_stream_write_space */
1564         if (sk_stream_is_writeable(sk))
1565                 xs_write_space(sk);
1566
1567         read_unlock_bh(&sk->sk_callback_lock);
1568 }
1569
1570 static void xs_udp_do_set_buffer_size(struct rpc_xprt *xprt)
1571 {
1572         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1573         struct sock *sk = transport->inet;
1574
1575         if (transport->rcvsize) {
1576                 sk->sk_userlocks |= SOCK_RCVBUF_LOCK;
1577                 sk->sk_rcvbuf = transport->rcvsize * xprt->max_reqs * 2;
1578         }
1579         if (transport->sndsize) {
1580                 sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
1581                 sk->sk_sndbuf = transport->sndsize * xprt->max_reqs * 2;
1582                 sk->sk_write_space(sk);
1583         }
1584 }
1585
1586 /**
1587  * xs_udp_set_buffer_size - set send and receive limits
1588  * @xprt: generic transport
1589  * @sndsize: requested size of send buffer, in bytes
1590  * @rcvsize: requested size of receive buffer, in bytes
1591  *
1592  * Set socket send and receive buffer size limits.
1593  */
1594 static void xs_udp_set_buffer_size(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize)
1595 {
1596         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1597
1598         transport->sndsize = 0;
1599         if (sndsize)
1600                 transport->sndsize = sndsize + 1024;
1601         transport->rcvsize = 0;
1602         if (rcvsize)
1603                 transport->rcvsize = rcvsize + 1024;
1604
1605         xs_udp_do_set_buffer_size(xprt);
1606 }
1607
1608 /**
1609  * xs_udp_timer - called when a retransmit timeout occurs on a UDP transport
1610  * @task: task that timed out
1611  *
1612  * Adjust the congestion window after a retransmit timeout has occurred.
1613  */
1614 static void xs_udp_timer(struct rpc_xprt *xprt, struct rpc_task *task)
1615 {
1616         xprt_adjust_cwnd(xprt, task, -ETIMEDOUT);
1617 }
1618
1619 static unsigned short xs_get_random_port(void)
1620 {
1621         unsigned short range = xprt_max_resvport - xprt_min_resvport;
1622         unsigned short rand = (unsigned short) prandom_u32() % range;
1623         return rand + xprt_min_resvport;
1624 }
1625
1626 /**
1627  * xs_set_reuseaddr_port - set the socket's port and address reuse options
1628  * @sock: socket
1629  *
1630  * Note that this function has to be called on all sockets that share the
1631  * same port, and it must be called before binding.
1632  */
1633 static void xs_sock_set_reuseport(struct socket *sock)
1634 {
1635         char opt = 1;
1636
1637         kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEPORT, &opt, sizeof(opt));
1638 }
1639
1640 static unsigned short xs_sock_getport(struct socket *sock)
1641 {
1642         struct sockaddr_storage buf;
1643         int buflen;
1644         unsigned short port = 0;
1645
1646         if (kernel_getsockname(sock, (struct sockaddr *)&buf, &buflen) < 0)
1647                 goto out;
1648         switch (buf.ss_family) {
1649         case AF_INET6:
1650                 port = ntohs(((struct sockaddr_in6 *)&buf)->sin6_port);
1651                 break;
1652         case AF_INET:
1653                 port = ntohs(((struct sockaddr_in *)&buf)->sin_port);
1654         }
1655 out:
1656         return port;
1657 }
1658
1659 /**
1660  * xs_set_port - reset the port number in the remote endpoint address
1661  * @xprt: generic transport
1662  * @port: new port number
1663  *
1664  */
1665 static void xs_set_port(struct rpc_xprt *xprt, unsigned short port)
1666 {
1667         dprintk("RPC:       setting port for xprt %p to %u\n", xprt, port);
1668
1669         rpc_set_port(xs_addr(xprt), port);
1670         xs_update_peer_port(xprt);
1671 }
1672
1673 static void xs_set_srcport(struct sock_xprt *transport, struct socket *sock)
1674 {
1675         if (transport->srcport == 0)
1676                 transport->srcport = xs_sock_getport(sock);
1677 }
1678
1679 static unsigned short xs_get_srcport(struct sock_xprt *transport)
1680 {
1681         unsigned short port = transport->srcport;
1682
1683         if (port == 0 && transport->xprt.resvport)
1684                 port = xs_get_random_port();
1685         return port;
1686 }
1687
1688 static unsigned short xs_next_srcport(struct sock_xprt *transport, unsigned short port)
1689 {
1690         if (transport->srcport != 0)
1691                 transport->srcport = 0;
1692         if (!transport->xprt.resvport)
1693                 return 0;
1694         if (port <= xprt_min_resvport || port > xprt_max_resvport)
1695                 return xprt_max_resvport;
1696         return --port;
1697 }
1698 static int xs_bind(struct sock_xprt *transport, struct socket *sock)
1699 {
1700         struct sockaddr_storage myaddr;
1701         int err, nloop = 0;
1702         unsigned short port = xs_get_srcport(transport);
1703         unsigned short last;
1704
1705         /*
1706          * If we are asking for any ephemeral port (i.e. port == 0 &&
1707          * transport->xprt.resvport == 0), don't bind.  Let the local
1708          * port selection happen implicitly when the socket is used
1709          * (for example at connect time).
1710          *
1711          * This ensures that we can continue to establish TCP
1712          * connections even when all local ephemeral ports are already
1713          * a part of some TCP connection.  This makes no difference
1714          * for UDP sockets, but also doens't harm them.
1715          *
1716          * If we're asking for any reserved port (i.e. port == 0 &&
1717          * transport->xprt.resvport == 1) xs_get_srcport above will
1718          * ensure that port is non-zero and we will bind as needed.
1719          */
1720         if (port == 0)
1721                 return 0;
1722
1723         memcpy(&myaddr, &transport->srcaddr, transport->xprt.addrlen);
1724         do {
1725                 rpc_set_port((struct sockaddr *)&myaddr, port);
1726                 err = kernel_bind(sock, (struct sockaddr *)&myaddr,
1727                                 transport->xprt.addrlen);
1728                 if (err == 0) {
1729                         transport->srcport = port;
1730                         break;
1731                 }
1732                 last = port;
1733                 port = xs_next_srcport(transport, port);
1734                 if (port > last)
1735                         nloop++;
1736         } while (err == -EADDRINUSE && nloop != 2);
1737
1738         if (myaddr.ss_family == AF_INET)
1739                 dprintk("RPC:       %s %pI4:%u: %s (%d)\n", __func__,
1740                                 &((struct sockaddr_in *)&myaddr)->sin_addr,
1741                                 port, err ? "failed" : "ok", err);
1742         else
1743                 dprintk("RPC:       %s %pI6:%u: %s (%d)\n", __func__,
1744                                 &((struct sockaddr_in6 *)&myaddr)->sin6_addr,
1745                                 port, err ? "failed" : "ok", err);
1746         return err;
1747 }
1748
1749 /*
1750  * We don't support autobind on AF_LOCAL sockets
1751  */
1752 static void xs_local_rpcbind(struct rpc_task *task)
1753 {
1754         rcu_read_lock();
1755         xprt_set_bound(rcu_dereference(task->tk_client->cl_xprt));
1756         rcu_read_unlock();
1757 }
1758
1759 static void xs_local_set_port(struct rpc_xprt *xprt, unsigned short port)
1760 {
1761 }
1762
1763 #ifdef CONFIG_DEBUG_LOCK_ALLOC
1764 static struct lock_class_key xs_key[2];
1765 static struct lock_class_key xs_slock_key[2];
1766
1767 static inline void xs_reclassify_socketu(struct socket *sock)
1768 {
1769         struct sock *sk = sock->sk;
1770
1771         sock_lock_init_class_and_name(sk, "slock-AF_LOCAL-RPC",
1772                 &xs_slock_key[1], "sk_lock-AF_LOCAL-RPC", &xs_key[1]);
1773 }
1774
1775 static inline void xs_reclassify_socket4(struct socket *sock)
1776 {
1777         struct sock *sk = sock->sk;
1778
1779         sock_lock_init_class_and_name(sk, "slock-AF_INET-RPC",
1780                 &xs_slock_key[0], "sk_lock-AF_INET-RPC", &xs_key[0]);
1781 }
1782
1783 static inline void xs_reclassify_socket6(struct socket *sock)
1784 {
1785         struct sock *sk = sock->sk;
1786
1787         sock_lock_init_class_and_name(sk, "slock-AF_INET6-RPC",
1788                 &xs_slock_key[1], "sk_lock-AF_INET6-RPC", &xs_key[1]);
1789 }
1790
1791 static inline void xs_reclassify_socket(int family, struct socket *sock)
1792 {
1793         WARN_ON_ONCE(sock_owned_by_user(sock->sk));
1794         if (sock_owned_by_user(sock->sk))
1795                 return;
1796
1797         switch (family) {
1798         case AF_LOCAL:
1799                 xs_reclassify_socketu(sock);
1800                 break;
1801         case AF_INET:
1802                 xs_reclassify_socket4(sock);
1803                 break;
1804         case AF_INET6:
1805                 xs_reclassify_socket6(sock);
1806                 break;
1807         }
1808 }
1809 #else
1810 static inline void xs_reclassify_socketu(struct socket *sock)
1811 {
1812 }
1813
1814 static inline void xs_reclassify_socket4(struct socket *sock)
1815 {
1816 }
1817
1818 static inline void xs_reclassify_socket6(struct socket *sock)
1819 {
1820 }
1821
1822 static inline void xs_reclassify_socket(int family, struct socket *sock)
1823 {
1824 }
1825 #endif
1826
1827 static void xs_dummy_setup_socket(struct work_struct *work)
1828 {
1829 }
1830
1831 static struct socket *xs_create_sock(struct rpc_xprt *xprt,
1832                 struct sock_xprt *transport, int family, int type,
1833                 int protocol, bool reuseport)
1834 {
1835         struct socket *sock;
1836         int err;
1837
1838         err = __sock_create(xprt->xprt_net, family, type, protocol, &sock, 1);
1839         if (err < 0) {
1840                 dprintk("RPC:       can't create %d transport socket (%d).\n",
1841                                 protocol, -err);
1842                 goto out;
1843         }
1844         xs_reclassify_socket(family, sock);
1845
1846         if (reuseport)
1847                 xs_sock_set_reuseport(sock);
1848
1849         err = xs_bind(transport, sock);
1850         if (err) {
1851                 sock_release(sock);
1852                 goto out;
1853         }
1854
1855         return sock;
1856 out:
1857         return ERR_PTR(err);
1858 }
1859
1860 static int xs_local_finish_connecting(struct rpc_xprt *xprt,
1861                                       struct socket *sock)
1862 {
1863         struct sock_xprt *transport = container_of(xprt, struct sock_xprt,
1864                                                                         xprt);
1865
1866         if (!transport->inet) {
1867                 struct sock *sk = sock->sk;
1868
1869                 write_lock_bh(&sk->sk_callback_lock);
1870
1871                 xs_save_old_callbacks(transport, sk);
1872
1873                 sk->sk_user_data = xprt;
1874                 sk->sk_data_ready = xs_local_data_ready;
1875                 sk->sk_write_space = xs_udp_write_space;
1876                 sk->sk_error_report = xs_error_report;
1877                 sk->sk_allocation = GFP_ATOMIC;
1878
1879                 xprt_clear_connected(xprt);
1880
1881                 /* Reset to new socket */
1882                 transport->sock = sock;
1883                 transport->inet = sk;
1884
1885                 write_unlock_bh(&sk->sk_callback_lock);
1886         }
1887
1888         /* Tell the socket layer to start connecting... */
1889         xprt->stat.connect_count++;
1890         xprt->stat.connect_start = jiffies;
1891         return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, 0);
1892 }
1893
1894 /**
1895  * xs_local_setup_socket - create AF_LOCAL socket, connect to a local endpoint
1896  * @xprt: RPC transport to connect
1897  * @transport: socket transport to connect
1898  * @create_sock: function to create a socket of the correct type
1899  */
1900 static int xs_local_setup_socket(struct sock_xprt *transport)
1901 {
1902         struct rpc_xprt *xprt = &transport->xprt;
1903         struct socket *sock;
1904         int status = -EIO;
1905
1906         status = __sock_create(xprt->xprt_net, AF_LOCAL,
1907                                         SOCK_STREAM, 0, &sock, 1);
1908         if (status < 0) {
1909                 dprintk("RPC:       can't create AF_LOCAL "
1910                         "transport socket (%d).\n", -status);
1911                 goto out;
1912         }
1913         xs_reclassify_socketu(sock);
1914
1915         dprintk("RPC:       worker connecting xprt %p via AF_LOCAL to %s\n",
1916                         xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1917
1918         status = xs_local_finish_connecting(xprt, sock);
1919         trace_rpc_socket_connect(xprt, sock, status);
1920         switch (status) {
1921         case 0:
1922                 dprintk("RPC:       xprt %p connected to %s\n",
1923                                 xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1924                 xprt_set_connected(xprt);
1925         case -ENOBUFS:
1926                 break;
1927         case -ENOENT:
1928                 dprintk("RPC:       xprt %p: socket %s does not exist\n",
1929                                 xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1930                 break;
1931         case -ECONNREFUSED:
1932                 dprintk("RPC:       xprt %p: connection refused for %s\n",
1933                                 xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1934                 break;
1935         default:
1936                 printk(KERN_ERR "%s: unhandled error (%d) connecting to %s\n",
1937                                 __func__, -status,
1938                                 xprt->address_strings[RPC_DISPLAY_ADDR]);
1939         }
1940
1941 out:
1942         xprt_clear_connecting(xprt);
1943         xprt_wake_pending_tasks(xprt, status);
1944         return status;
1945 }
1946
1947 static void xs_local_connect(struct rpc_xprt *xprt, struct rpc_task *task)
1948 {
1949         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1950         int ret;
1951
1952          if (RPC_IS_ASYNC(task)) {
1953                 /*
1954                  * We want the AF_LOCAL connect to be resolved in the
1955                  * filesystem namespace of the process making the rpc
1956                  * call.  Thus we connect synchronously.
1957                  *
1958                  * If we want to support asynchronous AF_LOCAL calls,
1959                  * we'll need to figure out how to pass a namespace to
1960                  * connect.
1961                  */
1962                 rpc_exit(task, -ENOTCONN);
1963                 return;
1964         }
1965         ret = xs_local_setup_socket(transport);
1966         if (ret && !RPC_IS_SOFTCONN(task))
1967                 msleep_interruptible(15000);
1968 }
1969
1970 #ifdef CONFIG_SUNRPC_SWAP
1971 static void xs_set_memalloc(struct rpc_xprt *xprt)
1972 {
1973         struct sock_xprt *transport = container_of(xprt, struct sock_xprt,
1974                         xprt);
1975
1976         if (xprt->swapper)
1977                 sk_set_memalloc(transport->inet);
1978 }
1979
1980 /**
1981  * xs_swapper - Tag this transport as being used for swap.
1982  * @xprt: transport to tag
1983  * @enable: enable/disable
1984  *
1985  */
1986 int xs_swapper(struct rpc_xprt *xprt, int enable)
1987 {
1988         struct sock_xprt *transport = container_of(xprt, struct sock_xprt,
1989                         xprt);
1990         int err = 0;
1991
1992         if (enable) {
1993                 xprt->swapper++;
1994                 xs_set_memalloc(xprt);
1995         } else if (xprt->swapper) {
1996                 xprt->swapper--;
1997                 sk_clear_memalloc(transport->inet);
1998         }
1999
2000         return err;
2001 }
2002 EXPORT_SYMBOL_GPL(xs_swapper);
2003 #else
2004 static void xs_set_memalloc(struct rpc_xprt *xprt)
2005 {
2006 }
2007 #endif
2008
2009 static void xs_udp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
2010 {
2011         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2012
2013         if (!transport->inet) {
2014                 struct sock *sk = sock->sk;
2015
2016                 write_lock_bh(&sk->sk_callback_lock);
2017
2018                 xs_save_old_callbacks(transport, sk);
2019
2020                 sk->sk_user_data = xprt;
2021                 sk->sk_data_ready = xs_udp_data_ready;
2022                 sk->sk_write_space = xs_udp_write_space;
2023                 sk->sk_allocation = GFP_ATOMIC;
2024
2025                 xprt_set_connected(xprt);
2026
2027                 /* Reset to new socket */
2028                 transport->sock = sock;
2029                 transport->inet = sk;
2030
2031                 xs_set_memalloc(xprt);
2032
2033                 write_unlock_bh(&sk->sk_callback_lock);
2034         }
2035         xs_udp_do_set_buffer_size(xprt);
2036 }
2037
2038 static void xs_udp_setup_socket(struct work_struct *work)
2039 {
2040         struct sock_xprt *transport =
2041                 container_of(work, struct sock_xprt, connect_worker.work);
2042         struct rpc_xprt *xprt = &transport->xprt;
2043         struct socket *sock = transport->sock;
2044         int status = -EIO;
2045
2046         sock = xs_create_sock(xprt, transport,
2047                         xs_addr(xprt)->sa_family, SOCK_DGRAM,
2048                         IPPROTO_UDP, false);
2049         if (IS_ERR(sock))
2050                 goto out;
2051
2052         dprintk("RPC:       worker connecting xprt %p via %s to "
2053                                 "%s (port %s)\n", xprt,
2054                         xprt->address_strings[RPC_DISPLAY_PROTO],
2055                         xprt->address_strings[RPC_DISPLAY_ADDR],
2056                         xprt->address_strings[RPC_DISPLAY_PORT]);
2057
2058         xs_udp_finish_connecting(xprt, sock);
2059         trace_rpc_socket_connect(xprt, sock, 0);
2060         status = 0;
2061 out:
2062         xprt_unlock_connect(xprt, transport);
2063         xprt_clear_connecting(xprt);
2064         xprt_wake_pending_tasks(xprt, status);
2065 }
2066
2067 static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
2068 {
2069         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2070         int ret = -ENOTCONN;
2071
2072         if (!transport->inet) {
2073                 struct sock *sk = sock->sk;
2074                 unsigned int keepidle = xprt->timeout->to_initval / HZ;
2075                 unsigned int keepcnt = xprt->timeout->to_retries + 1;
2076                 unsigned int opt_on = 1;
2077
2078                 /* TCP Keepalive options */
2079                 kernel_setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE,
2080                                 (char *)&opt_on, sizeof(opt_on));
2081                 kernel_setsockopt(sock, SOL_TCP, TCP_KEEPIDLE,
2082                                 (char *)&keepidle, sizeof(keepidle));
2083                 kernel_setsockopt(sock, SOL_TCP, TCP_KEEPINTVL,
2084                                 (char *)&keepidle, sizeof(keepidle));
2085                 kernel_setsockopt(sock, SOL_TCP, TCP_KEEPCNT,
2086                                 (char *)&keepcnt, sizeof(keepcnt));
2087
2088                 write_lock_bh(&sk->sk_callback_lock);
2089
2090                 xs_save_old_callbacks(transport, sk);
2091
2092                 sk->sk_user_data = xprt;
2093                 sk->sk_data_ready = xs_tcp_data_ready;
2094                 sk->sk_state_change = xs_tcp_state_change;
2095                 sk->sk_write_space = xs_tcp_write_space;
2096                 sk->sk_error_report = xs_error_report;
2097                 sk->sk_allocation = GFP_ATOMIC;
2098
2099                 /* socket options */
2100                 sock_reset_flag(sk, SOCK_LINGER);
2101                 tcp_sk(sk)->nonagle |= TCP_NAGLE_OFF;
2102
2103                 xprt_clear_connected(xprt);
2104
2105                 /* Reset to new socket */
2106                 transport->sock = sock;
2107                 transport->inet = sk;
2108
2109                 write_unlock_bh(&sk->sk_callback_lock);
2110         }
2111
2112         if (!xprt_bound(xprt))
2113                 goto out;
2114
2115         xs_set_memalloc(xprt);
2116
2117         /* Tell the socket layer to start connecting... */
2118         xprt->stat.connect_count++;
2119         xprt->stat.connect_start = jiffies;
2120         ret = kernel_connect(sock, xs_addr(xprt), xprt->addrlen, O_NONBLOCK);
2121         switch (ret) {
2122         case 0:
2123                 xs_set_srcport(transport, sock);
2124         case -EINPROGRESS:
2125                 /* SYN_SENT! */
2126                 if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
2127                         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
2128         }
2129 out:
2130         return ret;
2131 }
2132
2133 /**
2134  * xs_tcp_setup_socket - create a TCP socket and connect to a remote endpoint
2135  * @xprt: RPC transport to connect
2136  * @transport: socket transport to connect
2137  * @create_sock: function to create a socket of the correct type
2138  *
2139  * Invoked by a work queue tasklet.
2140  */
2141 static void xs_tcp_setup_socket(struct work_struct *work)
2142 {
2143         struct sock_xprt *transport =
2144                 container_of(work, struct sock_xprt, connect_worker.work);
2145         struct socket *sock = transport->sock;
2146         struct rpc_xprt *xprt = &transport->xprt;
2147         int status = -EIO;
2148
2149         if (!sock) {
2150                 sock = xs_create_sock(xprt, transport,
2151                                 xs_addr(xprt)->sa_family, SOCK_STREAM,
2152                                 IPPROTO_TCP, true);
2153                 if (IS_ERR(sock)) {
2154                         status = PTR_ERR(sock);
2155                         goto out;
2156                 }
2157         }
2158
2159         dprintk("RPC:       worker connecting xprt %p via %s to "
2160                                 "%s (port %s)\n", xprt,
2161                         xprt->address_strings[RPC_DISPLAY_PROTO],
2162                         xprt->address_strings[RPC_DISPLAY_ADDR],
2163                         xprt->address_strings[RPC_DISPLAY_PORT]);
2164
2165         status = xs_tcp_finish_connecting(xprt, sock);
2166         trace_rpc_socket_connect(xprt, sock, status);
2167         dprintk("RPC:       %p connect status %d connected %d sock state %d\n",
2168                         xprt, -status, xprt_connected(xprt),
2169                         sock->sk->sk_state);
2170         switch (status) {
2171         default:
2172                 printk("%s: connect returned unhandled error %d\n",
2173                         __func__, status);
2174         case -EADDRNOTAVAIL:
2175                 /* We're probably in TIME_WAIT. Get rid of existing socket,
2176                  * and retry
2177                  */
2178                 xs_tcp_force_close(xprt);
2179                 break;
2180         case 0:
2181         case -EINPROGRESS:
2182         case -EALREADY:
2183                 xprt_unlock_connect(xprt, transport);
2184                 xprt_clear_connecting(xprt);
2185                 return;
2186         case -EINVAL:
2187                 /* Happens, for instance, if the user specified a link
2188                  * local IPv6 address without a scope-id.
2189                  */
2190         case -ECONNREFUSED:
2191         case -ECONNRESET:
2192         case -ENETUNREACH:
2193         case -EADDRINUSE:
2194         case -ENOBUFS:
2195                 /* retry with existing socket, after a delay */
2196                 xs_tcp_force_close(xprt);
2197                 goto out;
2198         }
2199         status = -EAGAIN;
2200 out:
2201         xprt_unlock_connect(xprt, transport);
2202         xprt_clear_connecting(xprt);
2203         xprt_wake_pending_tasks(xprt, status);
2204 }
2205
2206 /**
2207  * xs_connect - connect a socket to a remote endpoint
2208  * @xprt: pointer to transport structure
2209  * @task: address of RPC task that manages state of connect request
2210  *
2211  * TCP: If the remote end dropped the connection, delay reconnecting.
2212  *
2213  * UDP socket connects are synchronous, but we use a work queue anyway
2214  * to guarantee that even unprivileged user processes can set up a
2215  * socket on a privileged port.
2216  *
2217  * If a UDP socket connect fails, the delay behavior here prevents
2218  * retry floods (hard mounts).
2219  */
2220 static void xs_connect(struct rpc_xprt *xprt, struct rpc_task *task)
2221 {
2222         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2223
2224         WARN_ON_ONCE(!xprt_lock_connect(xprt, task, transport));
2225
2226         /* Start by resetting any existing state */
2227         xs_reset_transport(transport);
2228
2229         if (transport->sock != NULL && !RPC_IS_SOFTCONN(task)) {
2230                 dprintk("RPC:       xs_connect delayed xprt %p for %lu "
2231                                 "seconds\n",
2232                                 xprt, xprt->reestablish_timeout / HZ);
2233                 queue_delayed_work(rpciod_workqueue,
2234                                    &transport->connect_worker,
2235                                    xprt->reestablish_timeout);
2236                 xprt->reestablish_timeout <<= 1;
2237                 if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
2238                         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
2239                 if (xprt->reestablish_timeout > XS_TCP_MAX_REEST_TO)
2240                         xprt->reestablish_timeout = XS_TCP_MAX_REEST_TO;
2241         } else {
2242                 dprintk("RPC:       xs_connect scheduled xprt %p\n", xprt);
2243                 queue_delayed_work(rpciod_workqueue,
2244                                    &transport->connect_worker, 0);
2245         }
2246 }
2247
2248 /**
2249  * xs_local_print_stats - display AF_LOCAL socket-specifc stats
2250  * @xprt: rpc_xprt struct containing statistics
2251  * @seq: output file
2252  *
2253  */
2254 static void xs_local_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
2255 {
2256         long idle_time = 0;
2257
2258         if (xprt_connected(xprt))
2259                 idle_time = (long)(jiffies - xprt->last_used) / HZ;
2260
2261         seq_printf(seq, "\txprt:\tlocal %lu %lu %lu %ld %lu %lu %lu "
2262                         "%llu %llu %lu %llu %llu\n",
2263                         xprt->stat.bind_count,
2264                         xprt->stat.connect_count,
2265                         xprt->stat.connect_time,
2266                         idle_time,
2267                         xprt->stat.sends,
2268                         xprt->stat.recvs,
2269                         xprt->stat.bad_xids,
2270                         xprt->stat.req_u,
2271                         xprt->stat.bklog_u,
2272                         xprt->stat.max_slots,
2273                         xprt->stat.sending_u,
2274                         xprt->stat.pending_u);
2275 }
2276
2277 /**
2278  * xs_udp_print_stats - display UDP socket-specifc stats
2279  * @xprt: rpc_xprt struct containing statistics
2280  * @seq: output file
2281  *
2282  */
2283 static void xs_udp_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
2284 {
2285         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2286
2287         seq_printf(seq, "\txprt:\tudp %u %lu %lu %lu %lu %llu %llu "
2288                         "%lu %llu %llu\n",
2289                         transport->srcport,
2290                         xprt->stat.bind_count,
2291                         xprt->stat.sends,
2292                         xprt->stat.recvs,
2293                         xprt->stat.bad_xids,
2294                         xprt->stat.req_u,
2295                         xprt->stat.bklog_u,
2296                         xprt->stat.max_slots,
2297                         xprt->stat.sending_u,
2298                         xprt->stat.pending_u);
2299 }
2300
2301 /**
2302  * xs_tcp_print_stats - display TCP socket-specifc stats
2303  * @xprt: rpc_xprt struct containing statistics
2304  * @seq: output file
2305  *
2306  */
2307 static void xs_tcp_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
2308 {
2309         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2310         long idle_time = 0;
2311
2312         if (xprt_connected(xprt))
2313                 idle_time = (long)(jiffies - xprt->last_used) / HZ;
2314
2315         seq_printf(seq, "\txprt:\ttcp %u %lu %lu %lu %ld %lu %lu %lu "
2316                         "%llu %llu %lu %llu %llu\n",
2317                         transport->srcport,
2318                         xprt->stat.bind_count,
2319                         xprt->stat.connect_count,
2320                         xprt->stat.connect_time,
2321                         idle_time,
2322                         xprt->stat.sends,
2323                         xprt->stat.recvs,
2324                         xprt->stat.bad_xids,
2325                         xprt->stat.req_u,
2326                         xprt->stat.bklog_u,
2327                         xprt->stat.max_slots,
2328                         xprt->stat.sending_u,
2329                         xprt->stat.pending_u);
2330 }
2331
2332 /*
2333  * Allocate a bunch of pages for a scratch buffer for the rpc code. The reason
2334  * we allocate pages instead doing a kmalloc like rpc_malloc is because we want
2335  * to use the server side send routines.
2336  */
2337 static void *bc_malloc(struct rpc_task *task, size_t size)
2338 {
2339         struct page *page;
2340         struct rpc_buffer *buf;
2341
2342         WARN_ON_ONCE(size > PAGE_SIZE - sizeof(struct rpc_buffer));
2343         if (size > PAGE_SIZE - sizeof(struct rpc_buffer))
2344                 return NULL;
2345
2346         page = alloc_page(GFP_KERNEL);
2347         if (!page)
2348                 return NULL;
2349
2350         buf = page_address(page);
2351         buf->len = PAGE_SIZE;
2352
2353         return buf->data;
2354 }
2355
2356 /*
2357  * Free the space allocated in the bc_alloc routine
2358  */
2359 static void bc_free(void *buffer)
2360 {
2361         struct rpc_buffer *buf;
2362
2363         if (!buffer)
2364                 return;
2365
2366         buf = container_of(buffer, struct rpc_buffer, data);
2367         free_page((unsigned long)buf);
2368 }
2369
2370 /*
2371  * Use the svc_sock to send the callback. Must be called with svsk->sk_mutex
2372  * held. Borrows heavily from svc_tcp_sendto and xs_tcp_send_request.
2373  */
2374 static int bc_sendto(struct rpc_rqst *req)
2375 {
2376         int len;
2377         struct xdr_buf *xbufp = &req->rq_snd_buf;
2378         struct rpc_xprt *xprt = req->rq_xprt;
2379         struct sock_xprt *transport =
2380                                 container_of(xprt, struct sock_xprt, xprt);
2381         struct socket *sock = transport->sock;
2382         unsigned long headoff;
2383         unsigned long tailoff;
2384
2385         xs_encode_stream_record_marker(xbufp);
2386
2387         tailoff = (unsigned long)xbufp->tail[0].iov_base & ~PAGE_MASK;
2388         headoff = (unsigned long)xbufp->head[0].iov_base & ~PAGE_MASK;
2389         len = svc_send_common(sock, xbufp,
2390                               virt_to_page(xbufp->head[0].iov_base), headoff,
2391                               xbufp->tail[0].iov_base, tailoff);
2392
2393         if (len != xbufp->len) {
2394                 printk(KERN_NOTICE "Error sending entire callback!\n");
2395                 len = -EAGAIN;
2396         }
2397
2398         return len;
2399 }
2400
2401 /*
2402  * The send routine. Borrows from svc_send
2403  */
2404 static int bc_send_request(struct rpc_task *task)
2405 {
2406         struct rpc_rqst *req = task->tk_rqstp;
2407         struct svc_xprt *xprt;
2408         u32                     len;
2409
2410         dprintk("sending request with xid: %08x\n", ntohl(req->rq_xid));
2411         /*
2412          * Get the server socket associated with this callback xprt
2413          */
2414         xprt = req->rq_xprt->bc_xprt;
2415
2416         /*
2417          * Grab the mutex to serialize data as the connection is shared
2418          * with the fore channel
2419          */
2420         if (!mutex_trylock(&xprt->xpt_mutex)) {
2421                 rpc_sleep_on(&xprt->xpt_bc_pending, task, NULL);
2422                 if (!mutex_trylock(&xprt->xpt_mutex))
2423                         return -EAGAIN;
2424                 rpc_wake_up_queued_task(&xprt->xpt_bc_pending, task);
2425         }
2426         if (test_bit(XPT_DEAD, &xprt->xpt_flags))
2427                 len = -ENOTCONN;
2428         else
2429                 len = bc_sendto(req);
2430         mutex_unlock(&xprt->xpt_mutex);
2431
2432         if (len > 0)
2433                 len = 0;
2434
2435         return len;
2436 }
2437
2438 /*
2439  * The close routine. Since this is client initiated, we do nothing
2440  */
2441
2442 static void bc_close(struct rpc_xprt *xprt)
2443 {
2444 }
2445
2446 /*
2447  * The xprt destroy routine. Again, because this connection is client
2448  * initiated, we do nothing
2449  */
2450
2451 static void bc_destroy(struct rpc_xprt *xprt)
2452 {
2453         dprintk("RPC:       bc_destroy xprt %p\n", xprt);
2454
2455         xs_xprt_free(xprt);
2456         module_put(THIS_MODULE);
2457 }
2458
2459 static struct rpc_xprt_ops xs_local_ops = {
2460         .reserve_xprt           = xprt_reserve_xprt,
2461         .release_xprt           = xs_tcp_release_xprt,
2462         .alloc_slot             = xprt_alloc_slot,
2463         .rpcbind                = xs_local_rpcbind,
2464         .set_port               = xs_local_set_port,
2465         .connect                = xs_local_connect,
2466         .buf_alloc              = rpc_malloc,
2467         .buf_free               = rpc_free,
2468         .send_request           = xs_local_send_request,
2469         .set_retrans_timeout    = xprt_set_retrans_timeout_def,
2470         .close                  = xs_close,
2471         .destroy                = xs_destroy,
2472         .print_stats            = xs_local_print_stats,
2473 };
2474
2475 static struct rpc_xprt_ops xs_udp_ops = {
2476         .set_buffer_size        = xs_udp_set_buffer_size,
2477         .reserve_xprt           = xprt_reserve_xprt_cong,
2478         .release_xprt           = xprt_release_xprt_cong,
2479         .alloc_slot             = xprt_alloc_slot,
2480         .rpcbind                = rpcb_getport_async,
2481         .set_port               = xs_set_port,
2482         .connect                = xs_connect,
2483         .buf_alloc              = rpc_malloc,
2484         .buf_free               = rpc_free,
2485         .send_request           = xs_udp_send_request,
2486         .set_retrans_timeout    = xprt_set_retrans_timeout_rtt,
2487         .timer                  = xs_udp_timer,
2488         .release_request        = xprt_release_rqst_cong,
2489         .close                  = xs_close,
2490         .destroy                = xs_destroy,
2491         .print_stats            = xs_udp_print_stats,
2492 };
2493
2494 static struct rpc_xprt_ops xs_tcp_ops = {
2495         .reserve_xprt           = xprt_reserve_xprt,
2496         .release_xprt           = xs_tcp_release_xprt,
2497         .alloc_slot             = xprt_lock_and_alloc_slot,
2498         .rpcbind                = rpcb_getport_async,
2499         .set_port               = xs_set_port,
2500         .connect                = xs_connect,
2501         .buf_alloc              = rpc_malloc,
2502         .buf_free               = rpc_free,
2503         .send_request           = xs_tcp_send_request,
2504         .set_retrans_timeout    = xprt_set_retrans_timeout_def,
2505         .close                  = xs_tcp_close,
2506         .destroy                = xs_destroy,
2507         .print_stats            = xs_tcp_print_stats,
2508 };
2509
2510 /*
2511  * The rpc_xprt_ops for the server backchannel
2512  */
2513
2514 static struct rpc_xprt_ops bc_tcp_ops = {
2515         .reserve_xprt           = xprt_reserve_xprt,
2516         .release_xprt           = xprt_release_xprt,
2517         .alloc_slot             = xprt_alloc_slot,
2518         .buf_alloc              = bc_malloc,
2519         .buf_free               = bc_free,
2520         .send_request           = bc_send_request,
2521         .set_retrans_timeout    = xprt_set_retrans_timeout_def,
2522         .close                  = bc_close,
2523         .destroy                = bc_destroy,
2524         .print_stats            = xs_tcp_print_stats,
2525 };
2526
2527 static int xs_init_anyaddr(const int family, struct sockaddr *sap)
2528 {
2529         static const struct sockaddr_in sin = {
2530                 .sin_family             = AF_INET,
2531                 .sin_addr.s_addr        = htonl(INADDR_ANY),
2532         };
2533         static const struct sockaddr_in6 sin6 = {
2534                 .sin6_family            = AF_INET6,
2535                 .sin6_addr              = IN6ADDR_ANY_INIT,
2536         };
2537
2538         switch (family) {
2539         case AF_LOCAL:
2540                 break;
2541         case AF_INET:
2542                 memcpy(sap, &sin, sizeof(sin));
2543                 break;
2544         case AF_INET6:
2545                 memcpy(sap, &sin6, sizeof(sin6));
2546                 break;
2547         default:
2548                 dprintk("RPC:       %s: Bad address family\n", __func__);
2549                 return -EAFNOSUPPORT;
2550         }
2551         return 0;
2552 }
2553
2554 static struct rpc_xprt *xs_setup_xprt(struct xprt_create *args,
2555                                       unsigned int slot_table_size,
2556                                       unsigned int max_slot_table_size)
2557 {
2558         struct rpc_xprt *xprt;
2559         struct sock_xprt *new;
2560
2561         if (args->addrlen > sizeof(xprt->addr)) {
2562                 dprintk("RPC:       xs_setup_xprt: address too large\n");
2563                 return ERR_PTR(-EBADF);
2564         }
2565
2566         xprt = xprt_alloc(args->net, sizeof(*new), slot_table_size,
2567                         max_slot_table_size);
2568         if (xprt == NULL) {
2569                 dprintk("RPC:       xs_setup_xprt: couldn't allocate "
2570                                 "rpc_xprt\n");
2571                 return ERR_PTR(-ENOMEM);
2572         }
2573
2574         new = container_of(xprt, struct sock_xprt, xprt);
2575         memcpy(&xprt->addr, args->dstaddr, args->addrlen);
2576         xprt->addrlen = args->addrlen;
2577         if (args->srcaddr)
2578                 memcpy(&new->srcaddr, args->srcaddr, args->addrlen);
2579         else {
2580                 int err;
2581                 err = xs_init_anyaddr(args->dstaddr->sa_family,
2582                                         (struct sockaddr *)&new->srcaddr);
2583                 if (err != 0) {
2584                         xprt_free(xprt);
2585                         return ERR_PTR(err);
2586                 }
2587         }
2588
2589         return xprt;
2590 }
2591
2592 static const struct rpc_timeout xs_local_default_timeout = {
2593         .to_initval = 10 * HZ,
2594         .to_maxval = 10 * HZ,
2595         .to_retries = 2,
2596 };
2597
2598 /**
2599  * xs_setup_local - Set up transport to use an AF_LOCAL socket
2600  * @args: rpc transport creation arguments
2601  *
2602  * AF_LOCAL is a "tpi_cots_ord" transport, just like TCP
2603  */
2604 static struct rpc_xprt *xs_setup_local(struct xprt_create *args)
2605 {
2606         struct sockaddr_un *sun = (struct sockaddr_un *)args->dstaddr;
2607         struct sock_xprt *transport;
2608         struct rpc_xprt *xprt;
2609         struct rpc_xprt *ret;
2610
2611         xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
2612                         xprt_max_tcp_slot_table_entries);
2613         if (IS_ERR(xprt))
2614                 return xprt;
2615         transport = container_of(xprt, struct sock_xprt, xprt);
2616
2617         xprt->prot = 0;
2618         xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32);
2619         xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
2620
2621         xprt->bind_timeout = XS_BIND_TO;
2622         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
2623         xprt->idle_timeout = XS_IDLE_DISC_TO;
2624
2625         xprt->ops = &xs_local_ops;
2626         xprt->timeout = &xs_local_default_timeout;
2627
2628         INIT_DELAYED_WORK(&transport->connect_worker,
2629                         xs_dummy_setup_socket);
2630
2631         switch (sun->sun_family) {
2632         case AF_LOCAL:
2633                 if (sun->sun_path[0] != '/') {
2634                         dprintk("RPC:       bad AF_LOCAL address: %s\n",
2635                                         sun->sun_path);
2636                         ret = ERR_PTR(-EINVAL);
2637                         goto out_err;
2638                 }
2639                 xprt_set_bound(xprt);
2640                 xs_format_peer_addresses(xprt, "local", RPCBIND_NETID_LOCAL);
2641                 ret = ERR_PTR(xs_local_setup_socket(transport));
2642                 if (ret)
2643                         goto out_err;
2644                 break;
2645         default:
2646                 ret = ERR_PTR(-EAFNOSUPPORT);
2647                 goto out_err;
2648         }
2649
2650         dprintk("RPC:       set up xprt to %s via AF_LOCAL\n",
2651                         xprt->address_strings[RPC_DISPLAY_ADDR]);
2652
2653         if (try_module_get(THIS_MODULE))
2654                 return xprt;
2655         ret = ERR_PTR(-EINVAL);
2656 out_err:
2657         xs_xprt_free(xprt);
2658         return ret;
2659 }
2660
2661 static const struct rpc_timeout xs_udp_default_timeout = {
2662         .to_initval = 5 * HZ,
2663         .to_maxval = 30 * HZ,
2664         .to_increment = 5 * HZ,
2665         .to_retries = 5,
2666 };
2667
2668 /**
2669  * xs_setup_udp - Set up transport to use a UDP socket
2670  * @args: rpc transport creation arguments
2671  *
2672  */
2673 static struct rpc_xprt *xs_setup_udp(struct xprt_create *args)
2674 {
2675         struct sockaddr *addr = args->dstaddr;
2676         struct rpc_xprt *xprt;
2677         struct sock_xprt *transport;
2678         struct rpc_xprt *ret;
2679
2680         xprt = xs_setup_xprt(args, xprt_udp_slot_table_entries,
2681                         xprt_udp_slot_table_entries);
2682         if (IS_ERR(xprt))
2683                 return xprt;
2684         transport = container_of(xprt, struct sock_xprt, xprt);
2685
2686         xprt->prot = IPPROTO_UDP;
2687         xprt->tsh_size = 0;
2688         /* XXX: header size can vary due to auth type, IPv6, etc. */
2689         xprt->max_payload = (1U << 16) - (MAX_HEADER << 3);
2690
2691         xprt->bind_timeout = XS_BIND_TO;
2692         xprt->reestablish_timeout = XS_UDP_REEST_TO;
2693         xprt->idle_timeout = XS_IDLE_DISC_TO;
2694
2695         xprt->ops = &xs_udp_ops;
2696
2697         xprt->timeout = &xs_udp_default_timeout;
2698
2699         switch (addr->sa_family) {
2700         case AF_INET:
2701                 if (((struct sockaddr_in *)addr)->sin_port != htons(0))
2702                         xprt_set_bound(xprt);
2703
2704                 INIT_DELAYED_WORK(&transport->connect_worker,
2705                                         xs_udp_setup_socket);
2706                 xs_format_peer_addresses(xprt, "udp", RPCBIND_NETID_UDP);
2707                 break;
2708         case AF_INET6:
2709                 if (((struct sockaddr_in6 *)addr)->sin6_port != htons(0))
2710                         xprt_set_bound(xprt);
2711
2712                 INIT_DELAYED_WORK(&transport->connect_worker,
2713                                         xs_udp_setup_socket);
2714                 xs_format_peer_addresses(xprt, "udp", RPCBIND_NETID_UDP6);
2715                 break;
2716         default:
2717                 ret = ERR_PTR(-EAFNOSUPPORT);
2718                 goto out_err;
2719         }
2720
2721         if (xprt_bound(xprt))
2722                 dprintk("RPC:       set up xprt to %s (port %s) via %s\n",
2723                                 xprt->address_strings[RPC_DISPLAY_ADDR],
2724                                 xprt->address_strings[RPC_DISPLAY_PORT],
2725                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
2726         else
2727                 dprintk("RPC:       set up xprt to %s (autobind) via %s\n",
2728                                 xprt->address_strings[RPC_DISPLAY_ADDR],
2729                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
2730
2731         if (try_module_get(THIS_MODULE))
2732                 return xprt;
2733         ret = ERR_PTR(-EINVAL);
2734 out_err:
2735         xs_xprt_free(xprt);
2736         return ret;
2737 }
2738
2739 static const struct rpc_timeout xs_tcp_default_timeout = {
2740         .to_initval = 60 * HZ,
2741         .to_maxval = 60 * HZ,
2742         .to_retries = 2,
2743 };
2744
2745 /**
2746  * xs_setup_tcp - Set up transport to use a TCP socket
2747  * @args: rpc transport creation arguments
2748  *
2749  */
2750 static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args)
2751 {
2752         struct sockaddr *addr = args->dstaddr;
2753         struct rpc_xprt *xprt;
2754         struct sock_xprt *transport;
2755         struct rpc_xprt *ret;
2756         unsigned int max_slot_table_size = xprt_max_tcp_slot_table_entries;
2757
2758         if (args->flags & XPRT_CREATE_INFINITE_SLOTS)
2759                 max_slot_table_size = RPC_MAX_SLOT_TABLE_LIMIT;
2760
2761         xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
2762                         max_slot_table_size);
2763         if (IS_ERR(xprt))
2764                 return xprt;
2765         transport = container_of(xprt, struct sock_xprt, xprt);
2766
2767         xprt->prot = IPPROTO_TCP;
2768         xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32);
2769         xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
2770
2771         xprt->bind_timeout = XS_BIND_TO;
2772         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
2773         xprt->idle_timeout = XS_IDLE_DISC_TO;
2774
2775         xprt->ops = &xs_tcp_ops;
2776         xprt->timeout = &xs_tcp_default_timeout;
2777
2778         switch (addr->sa_family) {
2779         case AF_INET:
2780                 if (((struct sockaddr_in *)addr)->sin_port != htons(0))
2781                         xprt_set_bound(xprt);
2782
2783                 INIT_DELAYED_WORK(&transport->connect_worker,
2784                                         xs_tcp_setup_socket);
2785                 xs_format_peer_addresses(xprt, "tcp", RPCBIND_NETID_TCP);
2786                 break;
2787         case AF_INET6:
2788                 if (((struct sockaddr_in6 *)addr)->sin6_port != htons(0))
2789                         xprt_set_bound(xprt);
2790
2791                 INIT_DELAYED_WORK(&transport->connect_worker,
2792                                         xs_tcp_setup_socket);
2793                 xs_format_peer_addresses(xprt, "tcp", RPCBIND_NETID_TCP6);
2794                 break;
2795         default:
2796                 ret = ERR_PTR(-EAFNOSUPPORT);
2797                 goto out_err;
2798         }
2799
2800         if (xprt_bound(xprt))
2801                 dprintk("RPC:       set up xprt to %s (port %s) via %s\n",
2802                                 xprt->address_strings[RPC_DISPLAY_ADDR],
2803                                 xprt->address_strings[RPC_DISPLAY_PORT],
2804                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
2805         else
2806                 dprintk("RPC:       set up xprt to %s (autobind) via %s\n",
2807                                 xprt->address_strings[RPC_DISPLAY_ADDR],
2808                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
2809
2810         if (try_module_get(THIS_MODULE))
2811                 return xprt;
2812         ret = ERR_PTR(-EINVAL);
2813 out_err:
2814         xs_xprt_free(xprt);
2815         return ret;
2816 }
2817
2818 /**
2819  * xs_setup_bc_tcp - Set up transport to use a TCP backchannel socket
2820  * @args: rpc transport creation arguments
2821  *
2822  */
2823 static struct rpc_xprt *xs_setup_bc_tcp(struct xprt_create *args)
2824 {
2825         struct sockaddr *addr = args->dstaddr;
2826         struct rpc_xprt *xprt;
2827         struct sock_xprt *transport;
2828         struct svc_sock *bc_sock;
2829         struct rpc_xprt *ret;
2830
2831         xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
2832                         xprt_tcp_slot_table_entries);
2833         if (IS_ERR(xprt))
2834                 return xprt;
2835         transport = container_of(xprt, struct sock_xprt, xprt);
2836
2837         xprt->prot = IPPROTO_TCP;
2838         xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32);
2839         xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
2840         xprt->timeout = &xs_tcp_default_timeout;
2841
2842         /* backchannel */
2843         xprt_set_bound(xprt);
2844         xprt->bind_timeout = 0;
2845         xprt->reestablish_timeout = 0;
2846         xprt->idle_timeout = 0;
2847
2848         xprt->ops = &bc_tcp_ops;
2849
2850         switch (addr->sa_family) {
2851         case AF_INET:
2852                 xs_format_peer_addresses(xprt, "tcp",
2853                                          RPCBIND_NETID_TCP);
2854                 break;
2855         case AF_INET6:
2856                 xs_format_peer_addresses(xprt, "tcp",
2857                                    RPCBIND_NETID_TCP6);
2858                 break;
2859         default:
2860                 ret = ERR_PTR(-EAFNOSUPPORT);
2861                 goto out_err;
2862         }
2863
2864         dprintk("RPC:       set up xprt to %s (port %s) via %s\n",
2865                         xprt->address_strings[RPC_DISPLAY_ADDR],
2866                         xprt->address_strings[RPC_DISPLAY_PORT],
2867                         xprt->address_strings[RPC_DISPLAY_PROTO]);
2868
2869         /*
2870          * Once we've associated a backchannel xprt with a connection,
2871          * we want to keep it around as long as the connection lasts,
2872          * in case we need to start using it for a backchannel again;
2873          * this reference won't be dropped until bc_xprt is destroyed.
2874          */
2875         xprt_get(xprt);
2876         args->bc_xprt->xpt_bc_xprt = xprt;
2877         xprt->bc_xprt = args->bc_xprt;
2878         bc_sock = container_of(args->bc_xprt, struct svc_sock, sk_xprt);
2879         transport->sock = bc_sock->sk_sock;
2880         transport->inet = bc_sock->sk_sk;
2881
2882         /*
2883          * Since we don't want connections for the backchannel, we set
2884          * the xprt status to connected
2885          */
2886         xprt_set_connected(xprt);
2887
2888         if (try_module_get(THIS_MODULE))
2889                 return xprt;
2890
2891         args->bc_xprt->xpt_bc_xprt = NULL;
2892         xprt_put(xprt);
2893         ret = ERR_PTR(-EINVAL);
2894 out_err:
2895         xs_xprt_free(xprt);
2896         return ret;
2897 }
2898
2899 static struct xprt_class        xs_local_transport = {
2900         .list           = LIST_HEAD_INIT(xs_local_transport.list),
2901         .name           = "named UNIX socket",
2902         .owner          = THIS_MODULE,
2903         .ident          = XPRT_TRANSPORT_LOCAL,
2904         .setup          = xs_setup_local,
2905 };
2906
2907 static struct xprt_class        xs_udp_transport = {
2908         .list           = LIST_HEAD_INIT(xs_udp_transport.list),
2909         .name           = "udp",
2910         .owner          = THIS_MODULE,
2911         .ident          = XPRT_TRANSPORT_UDP,
2912         .setup          = xs_setup_udp,
2913 };
2914
2915 static struct xprt_class        xs_tcp_transport = {
2916         .list           = LIST_HEAD_INIT(xs_tcp_transport.list),
2917         .name           = "tcp",
2918         .owner          = THIS_MODULE,
2919         .ident          = XPRT_TRANSPORT_TCP,
2920         .setup          = xs_setup_tcp,
2921 };
2922
2923 static struct xprt_class        xs_bc_tcp_transport = {
2924         .list           = LIST_HEAD_INIT(xs_bc_tcp_transport.list),
2925         .name           = "tcp NFSv4.1 backchannel",
2926         .owner          = THIS_MODULE,
2927         .ident          = XPRT_TRANSPORT_BC_TCP,
2928         .setup          = xs_setup_bc_tcp,
2929 };
2930
2931 /**
2932  * init_socket_xprt - set up xprtsock's sysctls, register with RPC client
2933  *
2934  */
2935 int init_socket_xprt(void)
2936 {
2937 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
2938         if (!sunrpc_table_header)
2939                 sunrpc_table_header = register_sysctl_table(sunrpc_table);
2940 #endif
2941
2942         xprt_register_transport(&xs_local_transport);
2943         xprt_register_transport(&xs_udp_transport);
2944         xprt_register_transport(&xs_tcp_transport);
2945         xprt_register_transport(&xs_bc_tcp_transport);
2946
2947         return 0;
2948 }
2949
2950 /**
2951  * cleanup_socket_xprt - remove xprtsock's sysctls, unregister
2952  *
2953  */
2954 void cleanup_socket_xprt(void)
2955 {
2956 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
2957         if (sunrpc_table_header) {
2958                 unregister_sysctl_table(sunrpc_table_header);
2959                 sunrpc_table_header = NULL;
2960         }
2961 #endif
2962
2963         xprt_unregister_transport(&xs_local_transport);
2964         xprt_unregister_transport(&xs_udp_transport);
2965         xprt_unregister_transport(&xs_tcp_transport);
2966         xprt_unregister_transport(&xs_bc_tcp_transport);
2967 }
2968
2969 static int param_set_uint_minmax(const char *val,
2970                 const struct kernel_param *kp,
2971                 unsigned int min, unsigned int max)
2972 {
2973         unsigned int num;
2974         int ret;
2975
2976         if (!val)
2977                 return -EINVAL;
2978         ret = kstrtouint(val, 0, &num);
2979         if (ret == -EINVAL || num < min || num > max)
2980                 return -EINVAL;
2981         *((unsigned int *)kp->arg) = num;
2982         return 0;
2983 }
2984
2985 static int param_set_portnr(const char *val, const struct kernel_param *kp)
2986 {
2987         return param_set_uint_minmax(val, kp,
2988                         RPC_MIN_RESVPORT,
2989                         RPC_MAX_RESVPORT);
2990 }
2991
2992 static struct kernel_param_ops param_ops_portnr = {
2993         .set = param_set_portnr,
2994         .get = param_get_uint,
2995 };
2996
2997 #define param_check_portnr(name, p) \
2998         __param_check(name, p, unsigned int);
2999
3000 module_param_named(min_resvport, xprt_min_resvport, portnr, 0644);
3001 module_param_named(max_resvport, xprt_max_resvport, portnr, 0644);
3002
3003 static int param_set_slot_table_size(const char *val,
3004                                      const struct kernel_param *kp)
3005 {
3006         return param_set_uint_minmax(val, kp,
3007                         RPC_MIN_SLOT_TABLE,
3008                         RPC_MAX_SLOT_TABLE);
3009 }
3010
3011 static struct kernel_param_ops param_ops_slot_table_size = {
3012         .set = param_set_slot_table_size,
3013         .get = param_get_uint,
3014 };
3015
3016 #define param_check_slot_table_size(name, p) \
3017         __param_check(name, p, unsigned int);
3018
3019 static int param_set_max_slot_table_size(const char *val,
3020                                      const struct kernel_param *kp)
3021 {
3022         return param_set_uint_minmax(val, kp,
3023                         RPC_MIN_SLOT_TABLE,
3024                         RPC_MAX_SLOT_TABLE_LIMIT);
3025 }
3026
3027 static struct kernel_param_ops param_ops_max_slot_table_size = {
3028         .set = param_set_max_slot_table_size,
3029         .get = param_get_uint,
3030 };
3031
3032 #define param_check_max_slot_table_size(name, p) \
3033         __param_check(name, p, unsigned int);
3034
3035 module_param_named(tcp_slot_table_entries, xprt_tcp_slot_table_entries,
3036                    slot_table_size, 0644);
3037 module_param_named(tcp_max_slot_table_entries, xprt_max_tcp_slot_table_entries,
3038                    max_slot_table_size, 0644);
3039 module_param_named(udp_slot_table_entries, xprt_udp_slot_table_entries,
3040                    slot_table_size, 0644);
3041