net/sunrpc/xprtsock.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 /*
   3  * linux/net/sunrpc/xprtsock.c
   4  *
   5  * Client-side transport implementation for sockets.
   6  *
   7  * TCP callback races fixes (C) 1998 Red Hat
   8  * TCP send fixes (C) 1998 Red Hat
   9  * TCP NFS related read + write fixes
  10  *  (C) 1999 Dave Airlie, University of Limerick, Ireland <airlied@linux.ie>
  11  *
  12  * Rewrite of larges part of the code in order to stabilize TCP stuff.
  13  * Fix behaviour when socket buffer is full.
  14  *  (C) 1999 Trond Myklebust <trond.myklebust@fys.uio.no>
  15  *
  16  * IP socket transport implementation, (C) 2005 Chuck Lever <cel@netapp.com>
  17  *
  18  * IPv6 support contributed by Gilles Quillard, Bull Open Source, 2005.
  19  *   <gilles.quillard@bull.net>
  20  */
  21
  22 #include <linux/types.h>
  23 #include <linux/string.h>
  24 #include <linux/slab.h>
  25 #include <linux/module.h>
  26 #include <linux/capability.h>
  27 #include <linux/pagemap.h>
  28 #include <linux/errno.h>
  29 #include <linux/socket.h>
  30 #include <linux/in.h>
  31 #include <linux/net.h>
  32 #include <linux/mm.h>
  33 #include <linux/un.h>
  34 #include <linux/udp.h>
  35 #include <linux/tcp.h>
  36 #include <linux/sunrpc/clnt.h>
  37 #include <linux/sunrpc/addr.h>
  38 #include <linux/sunrpc/sched.h>
  39 #include <linux/sunrpc/svcsock.h>
  40 #include <linux/sunrpc/xprtsock.h>
  41 #include <linux/file.h>
  42 #ifdef CONFIG_SUNRPC_BACKCHANNEL
  43 #include <linux/sunrpc/bc_xprt.h>
  44 #endif
  45
  46 #include <net/sock.h>
  47 #include <net/checksum.h>
  48 #include <net/udp.h>
  49 #include <net/tcp.h>
  50 #include <linux/bvec.h>
  51 #include <linux/highmem.h>
  52 #include <linux/uio.h>
  53 #include <linux/sched/mm.h>
  54
  55 #include <trace/events/sunrpc.h>
  56
  57 #include "socklib.h"
  58 #include "sunrpc.h"
  59
  60 static void xs_close(struct rpc_xprt *xprt);
  61 static void xs_set_srcport(struct sock_xprt *transport, struct socket *sock);
  62 static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt,
  63                 struct socket *sock);
  64
  65 /*
  66  * xprtsock tunables
  67  */
  68 static unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE;
  69 static unsigned int xprt_tcp_slot_table_entries = RPC_MIN_SLOT_TABLE;
  70 static unsigned int xprt_max_tcp_slot_table_entries = RPC_MAX_SLOT_TABLE;
  71
  72 static unsigned int xprt_min_resvport = RPC_DEF_MIN_RESVPORT;
  73 static unsigned int xprt_max_resvport = RPC_DEF_MAX_RESVPORT;
  74
  75 #define XS_TCP_LINGER_TO        (15U * HZ)
  76 static unsigned int xs_tcp_fin_timeout __read_mostly = XS_TCP_LINGER_TO;
  77
  78 /*
  79  * We can register our own files under /proc/sys/sunrpc by
  80  * calling register_sysctl_table() again.  The files in that
  81  * directory become the union of all files registered there.
  82  *
  83  * We simply need to make sure that we don't collide with
  84  * someone else's file names!
  85  */
  86
  87 static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE;
  88 static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE;
  89 static unsigned int max_tcp_slot_table_limit = RPC_MAX_SLOT_TABLE_LIMIT;
  90 static unsigned int xprt_min_resvport_limit = RPC_MIN_RESVPORT;
  91 static unsigned int xprt_max_resvport_limit = RPC_MAX_RESVPORT;
  92
  93 static struct ctl_table_header *sunrpc_table_header;
  94
  95 static struct xprt_class xs_local_transport;
  96 static struct xprt_class xs_udp_transport;
  97 static struct xprt_class xs_tcp_transport;
  98 static struct xprt_class xs_bc_tcp_transport;
  99
 100 /*
 101  * FIXME: changing the UDP slot table size should also resize the UDP
 102  *        socket buffers for existing UDP transports
 103  */
 104 static struct ctl_table xs_tunables_table[] = {
 105         {
 106                 .procname       = "udp_slot_table_entries",
 107                 .data           = &xprt_udp_slot_table_entries,
 108                 .maxlen         = sizeof(unsigned int),
 109                 .mode           = 0644,
 110                 .proc_handler   = proc_dointvec_minmax,
 111                 .extra1         = &min_slot_table_size,
 112                 .extra2         = &max_slot_table_size
 113         },
 114         {
 115                 .procname       = "tcp_slot_table_entries",
 116                 .data           = &xprt_tcp_slot_table_entries,
 117                 .maxlen         = sizeof(unsigned int),
 118                 .mode           = 0644,
 119                 .proc_handler   = proc_dointvec_minmax,
 120                 .extra1         = &min_slot_table_size,
 121                 .extra2         = &max_slot_table_size
 122         },
 123         {
 124                 .procname       = "tcp_max_slot_table_entries",
 125                 .data           = &xprt_max_tcp_slot_table_entries,
 126                 .maxlen         = sizeof(unsigned int),
 127                 .mode           = 0644,
 128                 .proc_handler   = proc_dointvec_minmax,
 129                 .extra1         = &min_slot_table_size,
 130                 .extra2         = &max_tcp_slot_table_limit
 131         },
 132         {
 133                 .procname       = "min_resvport",
 134                 .data           = &xprt_min_resvport,
 135                 .maxlen         = sizeof(unsigned int),
 136                 .mode           = 0644,
 137                 .proc_handler   = proc_dointvec_minmax,
 138                 .extra1         = &xprt_min_resvport_limit,
 139                 .extra2         = &xprt_max_resvport_limit
 140         },
 141         {
 142                 .procname       = "max_resvport",
 143                 .data           = &xprt_max_resvport,
 144                 .maxlen         = sizeof(unsigned int),
 145                 .mode           = 0644,
 146                 .proc_handler   = proc_dointvec_minmax,
 147                 .extra1         = &xprt_min_resvport_limit,
 148                 .extra2         = &xprt_max_resvport_limit
 149         },
 150         {
 151                 .procname       = "tcp_fin_timeout",
 152                 .data           = &xs_tcp_fin_timeout,
 153                 .maxlen         = sizeof(xs_tcp_fin_timeout),
 154                 .mode           = 0644,
 155                 .proc_handler   = proc_dointvec_jiffies,
 156         },
 157         { },
 158 };
 159
 160 static struct ctl_table sunrpc_table[] = {
 161         {
 162                 .procname       = "sunrpc",
 163                 .mode           = 0555,
 164                 .child          = xs_tunables_table
 165         },
 166         { },
 167 };
 168
 169 /*
 170  * Wait duration for a reply from the RPC portmapper.
 171  */
 172 #define XS_BIND_TO              (60U * HZ)
 173
 174 /*
 175  * Delay if a UDP socket connect error occurs.  This is most likely some
 176  * kind of resource problem on the local host.
 177  */
 178 #define XS_UDP_REEST_TO         (2U * HZ)
 179
 180 /*
 181  * The reestablish timeout allows clients to delay for a bit before attempting
 182  * to reconnect to a server that just dropped our connection.
 183  *
 184  * We implement an exponential backoff when trying to reestablish a TCP
 185  * transport connection with the server.  Some servers like to drop a TCP
 186  * connection when they are overworked, so we start with a short timeout and
 187  * increase over time if the server is down or not responding.
 188  */
 189 #define XS_TCP_INIT_REEST_TO    (3U * HZ)
 190
 191 /*
 192  * TCP idle timeout; client drops the transport socket if it is idle
 193  * for this long.  Note that we also timeout UDP sockets to prevent
 194  * holding port numbers when there is no RPC traffic.
 195  */
 196 #define XS_IDLE_DISC_TO         (5U * 60 * HZ)
 197
 198 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 199 # undef  RPC_DEBUG_DATA
 200 # define RPCDBG_FACILITY        RPCDBG_TRANS
 201 #endif
 202
 203 #ifdef RPC_DEBUG_DATA
 204 static void xs_pktdump(char *msg, u32 *packet, unsigned int count)
 205 {
 206         u8 *buf = (u8 *) packet;
 207         int j;
 208
 209         dprintk("RPC:       %s\n", msg);
 210         for (j = 0; j < count && j < 128; j += 4) {
 211                 if (!(j & 31)) {
 212                         if (j)
 213                                 dprintk("\n");
 214                         dprintk("0x%04x ", j);
 215                 }
 216                 dprintk("%02x%02x%02x%02x ",
 217                         buf[j], buf[j+1], buf[j+2], buf[j+3]);
 218         }
 219         dprintk("\n");
 220 }
 221 #else
 222 static inline void xs_pktdump(char *msg, u32 *packet, unsigned int count)
 223 {
 224         /* NOP */
 225 }
 226 #endif
 227
 228 static inline struct rpc_xprt *xprt_from_sock(struct sock *sk)
 229 {
 230         return (struct rpc_xprt *) sk->sk_user_data;
 231 }
 232
 233 static inline struct sockaddr *xs_addr(struct rpc_xprt *xprt)
 234 {
 235         return (struct sockaddr *) &xprt->addr;
 236 }
 237
 238 static inline struct sockaddr_un *xs_addr_un(struct rpc_xprt *xprt)
 239 {
 240         return (struct sockaddr_un *) &xprt->addr;
 241 }
 242
 243 static inline struct sockaddr_in *xs_addr_in(struct rpc_xprt *xprt)
 244 {
 245         return (struct sockaddr_in *) &xprt->addr;
 246 }
 247
 248 static inline struct sockaddr_in6 *xs_addr_in6(struct rpc_xprt *xprt)
 249 {
 250         return (struct sockaddr_in6 *) &xprt->addr;
 251 }
 252
 253 static void xs_format_common_peer_addresses(struct rpc_xprt *xprt)
 254 {
 255         struct sockaddr *sap = xs_addr(xprt);
 256         struct sockaddr_in6 *sin6;
 257         struct sockaddr_in *sin;
 258         struct sockaddr_un *sun;
 259         char buf[128];
 260
 261         switch (sap->sa_family) {
 262         case AF_LOCAL:
 263                 sun = xs_addr_un(xprt);
 264                 strlcpy(buf, sun->sun_path, sizeof(buf));
 265                 xprt->address_strings[RPC_DISPLAY_ADDR] =
 266                                                 kstrdup(buf, GFP_KERNEL);
 267                 break;
 268         case AF_INET:
 269                 (void)rpc_ntop(sap, buf, sizeof(buf));
 270                 xprt->address_strings[RPC_DISPLAY_ADDR] =
 271                                                 kstrdup(buf, GFP_KERNEL);
 272                 sin = xs_addr_in(xprt);
 273                 snprintf(buf, sizeof(buf), "%08x", ntohl(sin->sin_addr.s_addr));
 274                 break;
 275         case AF_INET6:
 276                 (void)rpc_ntop(sap, buf, sizeof(buf));
 277                 xprt->address_strings[RPC_DISPLAY_ADDR] =
 278                                                 kstrdup(buf, GFP_KERNEL);
 279                 sin6 = xs_addr_in6(xprt);
 280                 snprintf(buf, sizeof(buf), "%pi6", &sin6->sin6_addr);
 281                 break;
 282         default:
 283                 BUG();
 284         }
 285
 286         xprt->address_strings[RPC_DISPLAY_HEX_ADDR] = kstrdup(buf, GFP_KERNEL);
 287 }
 288
 289 static void xs_format_common_peer_ports(struct rpc_xprt *xprt)
 290 {
 291         struct sockaddr *sap = xs_addr(xprt);
 292         char buf[128];
 293
 294         snprintf(buf, sizeof(buf), "%u", rpc_get_port(sap));
 295         xprt->address_strings[RPC_DISPLAY_PORT] = kstrdup(buf, GFP_KERNEL);
 296
 297         snprintf(buf, sizeof(buf), "%4hx", rpc_get_port(sap));
 298         xprt->address_strings[RPC_DISPLAY_HEX_PORT] = kstrdup(buf, GFP_KERNEL);
 299 }
 300
 301 static void xs_format_peer_addresses(struct rpc_xprt *xprt,
 302                                      const char *protocol,
 303                                      const char *netid)
 304 {
 305         xprt->address_strings[RPC_DISPLAY_PROTO] = protocol;
 306         xprt->address_strings[RPC_DISPLAY_NETID] = netid;
 307         xs_format_common_peer_addresses(xprt);
 308         xs_format_common_peer_ports(xprt);
 309 }
 310
 311 static void xs_update_peer_port(struct rpc_xprt *xprt)
 312 {
 313         kfree(xprt->address_strings[RPC_DISPLAY_HEX_PORT]);
 314         kfree(xprt->address_strings[RPC_DISPLAY_PORT]);
 315
 316         xs_format_common_peer_ports(xprt);
 317 }
 318
 319 static void xs_free_peer_addresses(struct rpc_xprt *xprt)
 320 {
 321         unsigned int i;
 322
 323         for (i = 0; i < RPC_DISPLAY_MAX; i++)
 324                 switch (i) {
 325                 case RPC_DISPLAY_PROTO:
 326                 case RPC_DISPLAY_NETID:
 327                         continue;
 328                 default:
 329                         kfree(xprt->address_strings[i]);
 330                 }
 331 }
 332
 333 static size_t
 334 xs_alloc_sparse_pages(struct xdr_buf *buf, size_t want, gfp_t gfp)
 335 {
 336         size_t i,n;
 337
 338         if (!want || !(buf->flags & XDRBUF_SPARSE_PAGES))
 339                 return want;
 340         n = (buf->page_base + want + PAGE_SIZE - 1) >> PAGE_SHIFT;
 341         for (i = 0; i < n; i++) {
 342                 if (buf->pages[i])
 343                         continue;
 344                 buf->bvec[i].bv_page = buf->pages[i] = alloc_page(gfp);
 345                 if (!buf->pages[i]) {
 346                         i *= PAGE_SIZE;
 347                         return i > buf->page_base ? i - buf->page_base : 0;
 348                 }
 349         }
 350         return want;
 351 }
 352
 353 static ssize_t
 354 xs_sock_recvmsg(struct socket *sock, struct msghdr *msg, int flags, size_t seek)
 355 {
 356         ssize_t ret;
 357         if (seek != 0)
 358                 iov_iter_advance(&msg->msg_iter, seek);
 359         ret = sock_recvmsg(sock, msg, flags);
 360         return ret > 0 ? ret + seek : ret;
 361 }
 362
 363 static ssize_t
 364 xs_read_kvec(struct socket *sock, struct msghdr *msg, int flags,
 365                 struct kvec *kvec, size_t count, size_t seek)
 366 {
 367         iov_iter_kvec(&msg->msg_iter, READ, kvec, 1, count);
 368         return xs_sock_recvmsg(sock, msg, flags, seek);
 369 }
 370
 371 static ssize_t
 372 xs_read_bvec(struct socket *sock, struct msghdr *msg, int flags,
 373                 struct bio_vec *bvec, unsigned long nr, size_t count,
 374                 size_t seek)
 375 {
 376         iov_iter_bvec(&msg->msg_iter, READ, bvec, nr, count);
 377         return xs_sock_recvmsg(sock, msg, flags, seek);
 378 }
 379
 380 static ssize_t
 381 xs_read_discard(struct socket *sock, struct msghdr *msg, int flags,
 382                 size_t count)
 383 {
 384         iov_iter_discard(&msg->msg_iter, READ, count);
 385         return sock_recvmsg(sock, msg, flags);
 386 }
 387
 388 #if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE
 389 static void
 390 xs_flush_bvec(const struct bio_vec *bvec, size_t count, size_t seek)
 391 {
 392         struct bvec_iter bi = {
 393                 .bi_size = count,
 394         };
 395         struct bio_vec bv;
 396
 397         bvec_iter_advance(bvec, &bi, seek & PAGE_MASK);
 398         for_each_bvec(bv, bvec, bi, bi)
 399                 flush_dcache_page(bv.bv_page);
 400 }
 401 #else
 402 static inline void
 403 xs_flush_bvec(const struct bio_vec *bvec, size_t count, size_t seek)
 404 {
 405 }
 406 #endif
 407
 408 static ssize_t
 409 xs_read_xdr_buf(struct socket *sock, struct msghdr *msg, int flags,
 410                 struct xdr_buf *buf, size_t count, size_t seek, size_t *read)
 411 {
 412         size_t want, seek_init = seek, offset = 0;
 413         ssize_t ret;
 414
 415         want = min_t(size_t, count, buf->head[0].iov_len);
 416         if (seek < want) {
 417                 ret = xs_read_kvec(sock, msg, flags, &buf->head[0], want, seek);
 418                 if (ret <= 0)
 419                         goto sock_err;
 420                 offset += ret;
 421                 if (offset == count || msg->msg_flags & (MSG_EOR|MSG_TRUNC))
 422                         goto out;
 423                 if (ret != want)
 424                         goto out;
 425                 seek = 0;
 426         } else {
 427                 seek -= want;
 428                 offset += want;
 429         }
 430
 431         want = xs_alloc_sparse_pages(
 432                 buf, min_t(size_t, count - offset, buf->page_len),
 433                 GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN);
 434         if (seek < want) {
 435                 ret = xs_read_bvec(sock, msg, flags, buf->bvec,
 436                                 xdr_buf_pagecount(buf),
 437                                 want + buf->page_base,
 438                                 seek + buf->page_base);
 439                 if (ret <= 0)
 440                         goto sock_err;
 441                 xs_flush_bvec(buf->bvec, ret, seek + buf->page_base);
 442                 ret -= buf->page_base;
 443                 offset += ret;
 444                 if (offset == count || msg->msg_flags & (MSG_EOR|MSG_TRUNC))
 445                         goto out;
 446                 if (ret != want)
 447                         goto out;
 448                 seek = 0;
 449         } else {
 450                 seek -= want;
 451                 offset += want;
 452         }
 453
 454         want = min_t(size_t, count - offset, buf->tail[0].iov_len);
 455         if (seek < want) {
 456                 ret = xs_read_kvec(sock, msg, flags, &buf->tail[0], want, seek);
 457                 if (ret <= 0)
 458                         goto sock_err;
 459                 offset += ret;
 460                 if (offset == count || msg->msg_flags & (MSG_EOR|MSG_TRUNC))
 461                         goto out;
 462                 if (ret != want)
 463                         goto out;
 464         } else if (offset < seek_init)
 465                 offset = seek_init;
 466         ret = -EMSGSIZE;
 467 out:
 468         *read = offset - seek_init;
 469         return ret;
 470 sock_err:
 471         offset += seek;
 472         goto out;
 473 }
 474
 475 static void
 476 xs_read_header(struct sock_xprt *transport, struct xdr_buf *buf)
 477 {
 478         if (!transport->recv.copied) {
 479                 if (buf->head[0].iov_len >= transport->recv.offset)
 480                         memcpy(buf->head[0].iov_base,
 481                                         &transport->recv.xid,
 482                                         transport->recv.offset);
 483                 transport->recv.copied = transport->recv.offset;
 484         }
 485 }
 486
 487 static bool
 488 xs_read_stream_request_done(struct sock_xprt *transport)
 489 {
 490         return transport->recv.fraghdr & cpu_to_be32(RPC_LAST_STREAM_FRAGMENT);
 491 }
 492
 493 static void
 494 xs_read_stream_check_eor(struct sock_xprt *transport,
 495                 struct msghdr *msg)
 496 {
 497         if (xs_read_stream_request_done(transport))
 498                 msg->msg_flags |= MSG_EOR;
 499 }
 500
 501 static ssize_t
 502 xs_read_stream_request(struct sock_xprt *transport, struct msghdr *msg,
 503                 int flags, struct rpc_rqst *req)
 504 {
 505         struct xdr_buf *buf = &req->rq_private_buf;
 506         size_t want, read;
 507         ssize_t ret;
 508
 509         xs_read_header(transport, buf);
 510
 511         want = transport->recv.len - transport->recv.offset;
 512         if (want != 0) {
 513                 ret = xs_read_xdr_buf(transport->sock, msg, flags, buf,
 514                                 transport->recv.copied + want,
 515                                 transport->recv.copied,
 516                                 &read);
 517                 transport->recv.offset += read;
 518                 transport->recv.copied += read;
 519         }
 520
 521         if (transport->recv.offset == transport->recv.len)
 522                 xs_read_stream_check_eor(transport, msg);
 523
 524         if (want == 0)
 525                 return 0;
 526
 527         switch (ret) {
 528         default:
 529                 break;
 530         case -EFAULT:
 531         case -EMSGSIZE:
 532                 msg->msg_flags |= MSG_TRUNC;
 533                 return read;
 534         case 0:
 535                 return -ESHUTDOWN;
 536         }
 537         return ret < 0 ? ret : read;
 538 }
 539
 540 static size_t
 541 xs_read_stream_headersize(bool isfrag)
 542 {
 543         if (isfrag)
 544                 return sizeof(__be32);
 545         return 3 * sizeof(__be32);
 546 }
 547
 548 static ssize_t
 549 xs_read_stream_header(struct sock_xprt *transport, struct msghdr *msg,
 550                 int flags, size_t want, size_t seek)
 551 {
 552         struct kvec kvec = {
 553                 .iov_base = &transport->recv.fraghdr,
 554                 .iov_len = want,
 555         };
 556         return xs_read_kvec(transport->sock, msg, flags, &kvec, want, seek);
 557 }
 558
 559 #if defined(CONFIG_SUNRPC_BACKCHANNEL)
 560 static ssize_t
 561 xs_read_stream_call(struct sock_xprt *transport, struct msghdr *msg, int flags)
 562 {
 563         struct rpc_xprt *xprt = &transport->xprt;
 564         struct rpc_rqst *req;
 565         ssize_t ret;
 566
 567         /* Is this transport associated with the backchannel? */
 568         if (!xprt->bc_serv)
 569                 return -ESHUTDOWN;
 570
 571         /* Look up and lock the request corresponding to the given XID */
 572         req = xprt_lookup_bc_request(xprt, transport->recv.xid);
 573         if (!req) {
 574                 printk(KERN_WARNING "Callback slot table overflowed\n");
 575                 return -ESHUTDOWN;
 576         }
 577         if (transport->recv.copied && !req->rq_private_buf.len)
 578                 return -ESHUTDOWN;
 579
 580         ret = xs_read_stream_request(transport, msg, flags, req);
 581         if (msg->msg_flags & (MSG_EOR|MSG_TRUNC))
 582                 xprt_complete_bc_request(req, transport->recv.copied);
 583         else
 584                 req->rq_private_buf.len = transport->recv.copied;
 585
 586         return ret;
 587 }
 588 #else /* CONFIG_SUNRPC_BACKCHANNEL */
 589 static ssize_t
 590 xs_read_stream_call(struct sock_xprt *transport, struct msghdr *msg, int flags)
 591 {
 592         return -ESHUTDOWN;
 593 }
 594 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
 595
 596 static ssize_t
 597 xs_read_stream_reply(struct sock_xprt *transport, struct msghdr *msg, int flags)
 598 {
 599         struct rpc_xprt *xprt = &transport->xprt;
 600         struct rpc_rqst *req;
 601         ssize_t ret = 0;
 602
 603         /* Look up and lock the request corresponding to the given XID */
 604         spin_lock(&xprt->queue_lock);
 605         req = xprt_lookup_rqst(xprt, transport->recv.xid);
 606         if (!req || (transport->recv.copied && !req->rq_private_buf.len)) {
 607                 msg->msg_flags |= MSG_TRUNC;
 608                 goto out;
 609         }
 610         xprt_pin_rqst(req);
 611         spin_unlock(&xprt->queue_lock);
 612
 613         ret = xs_read_stream_request(transport, msg, flags, req);
 614
 615         spin_lock(&xprt->queue_lock);
 616         if (msg->msg_flags & (MSG_EOR|MSG_TRUNC))
 617                 xprt_complete_rqst(req->rq_task, transport->recv.copied);
 618         else
 619                 req->rq_private_buf.len = transport->recv.copied;
 620         xprt_unpin_rqst(req);
 621 out:
 622         spin_unlock(&xprt->queue_lock);
 623         return ret;
 624 }
 625
 626 static ssize_t
 627 xs_read_stream(struct sock_xprt *transport, int flags)
 628 {
 629         struct msghdr msg = { 0 };
 630         size_t want, read = 0;
 631         ssize_t ret = 0;
 632
 633         if (transport->recv.len == 0) {
 634                 want = xs_read_stream_headersize(transport->recv.copied != 0);
 635                 ret = xs_read_stream_header(transport, &msg, flags, want,
 636                                 transport->recv.offset);
 637                 if (ret <= 0)
 638                         goto out_err;
 639                 transport->recv.offset = ret;
 640                 if (transport->recv.offset != want)
 641                         return transport->recv.offset;
 642                 transport->recv.len = be32_to_cpu(transport->recv.fraghdr) &
 643                         RPC_FRAGMENT_SIZE_MASK;
 644                 transport->recv.offset -= sizeof(transport->recv.fraghdr);
 645                 read = ret;
 646         }
 647
 648         switch (be32_to_cpu(transport->recv.calldir)) {
 649         default:
 650                 msg.msg_flags |= MSG_TRUNC;
 651                 break;
 652         case RPC_CALL:
 653                 ret = xs_read_stream_call(transport, &msg, flags);
 654                 break;
 655         case RPC_REPLY:
 656                 ret = xs_read_stream_reply(transport, &msg, flags);
 657         }
 658         if (msg.msg_flags & MSG_TRUNC) {
 659                 transport->recv.calldir = cpu_to_be32(-1);
 660                 transport->recv.copied = -1;
 661         }
 662         if (ret < 0)
 663                 goto out_err;
 664         read += ret;
 665         if (transport->recv.offset < transport->recv.len) {
 666                 if (!(msg.msg_flags & MSG_TRUNC))
 667                         return read;
 668                 msg.msg_flags = 0;
 669                 ret = xs_read_discard(transport->sock, &msg, flags,
 670                                 transport->recv.len - transport->recv.offset);
 671                 if (ret <= 0)
 672                         goto out_err;
 673                 transport->recv.offset += ret;
 674                 read += ret;
 675                 if (transport->recv.offset != transport->recv.len)
 676                         return read;
 677         }
 678         if (xs_read_stream_request_done(transport)) {
 679                 trace_xs_stream_read_request(transport);
 680                 transport->recv.copied = 0;
 681         }
 682         transport->recv.offset = 0;
 683         transport->recv.len = 0;
 684         return read;
 685 out_err:
 686         return ret != 0 ? ret : -ESHUTDOWN;
 687 }
 688
 689 static __poll_t xs_poll_socket(struct sock_xprt *transport)
 690 {
 691         return transport->sock->ops->poll(transport->file, transport->sock,
 692                         NULL);
 693 }
 694
 695 static bool xs_poll_socket_readable(struct sock_xprt *transport)
 696 {
 697         __poll_t events = xs_poll_socket(transport);
 698
 699         return (events & (EPOLLIN | EPOLLRDNORM)) && !(events & EPOLLRDHUP);
 700 }
 701
 702 static void xs_poll_check_readable(struct sock_xprt *transport)
 703 {
 704
 705         clear_bit(XPRT_SOCK_DATA_READY, &transport->sock_state);
 706         if (!xs_poll_socket_readable(transport))
 707                 return;
 708         if (!test_and_set_bit(XPRT_SOCK_DATA_READY, &transport->sock_state))
 709                 queue_work(xprtiod_workqueue, &transport->recv_worker);
 710 }
 711
 712 static void xs_stream_data_receive(struct sock_xprt *transport)
 713 {
 714         size_t read = 0;
 715         ssize_t ret = 0;
 716
 717         mutex_lock(&transport->recv_mutex);
 718         if (transport->sock == NULL)
 719                 goto out;
 720         for (;;) {
 721                 ret = xs_read_stream(transport, MSG_DONTWAIT);
 722                 if (ret < 0)
 723                         break;
 724                 read += ret;
 725                 cond_resched();
 726         }
 727         if (ret == -ESHUTDOWN)
 728                 kernel_sock_shutdown(transport->sock, SHUT_RDWR);
 729         else
 730                 xs_poll_check_readable(transport);
 731 out:
 732         mutex_unlock(&transport->recv_mutex);
 733         trace_xs_stream_read_data(&transport->xprt, ret, read);
 734 }
 735
 736 static void xs_stream_data_receive_workfn(struct work_struct *work)
 737 {
 738         struct sock_xprt *transport =
 739                 container_of(work, struct sock_xprt, recv_worker);
 740         unsigned int pflags = memalloc_nofs_save();
 741
 742         xs_stream_data_receive(transport);
 743         memalloc_nofs_restore(pflags);
 744 }
 745
 746 static void
 747 xs_stream_reset_connect(struct sock_xprt *transport)
 748 {
 749         transport->recv.offset = 0;
 750         transport->recv.len = 0;
 751         transport->recv.copied = 0;
 752         transport->xmit.offset = 0;
 753 }
 754
 755 static void
 756 xs_stream_start_connect(struct sock_xprt *transport)
 757 {
 758         transport->xprt.stat.connect_count++;
 759         transport->xprt.stat.connect_start = jiffies;
 760 }
 761
 762 #define XS_SENDMSG_FLAGS        (MSG_DONTWAIT | MSG_NOSIGNAL)
 763
 764 /**
 765  * xs_nospace - handle transmit was incomplete
 766  * @req: pointer to RPC request
 767  * @transport: pointer to struct sock_xprt
 768  *
 769  */
 770 static int xs_nospace(struct rpc_rqst *req, struct sock_xprt *transport)
 771 {
 772         struct rpc_xprt *xprt = &transport->xprt;
 773         struct sock *sk = transport->inet;
 774         int ret = -EAGAIN;
 775
 776         trace_rpc_socket_nospace(req, transport);
 777
 778         /* Protect against races with write_space */
 779         spin_lock(&xprt->transport_lock);
 780
 781         /* Don't race with disconnect */
 782         if (xprt_connected(xprt)) {
 783                 /* wait for more buffer space */
 784                 set_bit(XPRT_SOCK_NOSPACE, &transport->sock_state);
 785                 set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
 786                 sk->sk_write_pending++;
 787                 xprt_wait_for_buffer_space(xprt);
 788         } else
 789                 ret = -ENOTCONN;
 790
 791         spin_unlock(&xprt->transport_lock);
 792         return ret;
 793 }
 794
 795 static int xs_sock_nospace(struct rpc_rqst *req)
 796 {
 797         struct sock_xprt *transport =
 798                 container_of(req->rq_xprt, struct sock_xprt, xprt);
 799         struct sock *sk = transport->inet;
 800         int ret = -EAGAIN;
 801
 802         lock_sock(sk);
 803         if (!sock_writeable(sk))
 804                 ret = xs_nospace(req, transport);
 805         release_sock(sk);
 806         return ret;
 807 }
 808
 809 static int xs_stream_nospace(struct rpc_rqst *req, bool vm_wait)
 810 {
 811         struct sock_xprt *transport =
 812                 container_of(req->rq_xprt, struct sock_xprt, xprt);
 813         struct sock *sk = transport->inet;
 814         int ret = -EAGAIN;
 815
 816         if (vm_wait)
 817                 return -ENOBUFS;
 818         lock_sock(sk);
 819         if (!sk_stream_memory_free(sk))
 820                 ret = xs_nospace(req, transport);
 821         release_sock(sk);
 822         return ret;
 823 }
 824
 825 static int
 826 xs_stream_prepare_request(struct rpc_rqst *req)
 827 {
 828         gfp_t gfp = rpc_task_gfp_mask();
 829         int ret;
 830
 831         ret = xdr_alloc_bvec(&req->rq_snd_buf, gfp);
 832         if (ret < 0)
 833                 return ret;
 834         xdr_free_bvec(&req->rq_rcv_buf);
 835         return xdr_alloc_bvec(&req->rq_rcv_buf, gfp);
 836 }
 837
 838 /*
 839  * Determine if the previous message in the stream was aborted before it
 840  * could complete transmission.
 841  */
 842 static bool
 843 xs_send_request_was_aborted(struct sock_xprt *transport, struct rpc_rqst *req)
 844 {
 845         return transport->xmit.offset != 0 && req->rq_bytes_sent == 0;
 846 }
 847
 848 /*
 849  * Return the stream record marker field for a record of length < 2^31-1
 850  */
 851 static rpc_fraghdr
 852 xs_stream_record_marker(struct xdr_buf *xdr)
 853 {
 854         if (!xdr->len)
 855                 return 0;
 856         return cpu_to_be32(RPC_LAST_STREAM_FRAGMENT | (u32)xdr->len);
 857 }
 858
 859 /**
 860  * xs_local_send_request - write an RPC request to an AF_LOCAL socket
 861  * @req: pointer to RPC request
 862  *
 863  * Return values:
 864  *        0:    The request has been sent
 865  *   EAGAIN:    The socket was blocked, please call again later to
 866  *              complete the request
 867  * ENOTCONN:    Caller needs to invoke connect logic then call again
 868  *    other:    Some other error occurred, the request was not sent
 869  */
 870 static int xs_local_send_request(struct rpc_rqst *req)
 871 {
 872         struct rpc_xprt *xprt = req->rq_xprt;
 873         struct sock_xprt *transport =
 874                                 container_of(xprt, struct sock_xprt, xprt);
 875         struct xdr_buf *xdr = &req->rq_snd_buf;
 876         rpc_fraghdr rm = xs_stream_record_marker(xdr);
 877         unsigned int msglen = rm ? req->rq_slen + sizeof(rm) : req->rq_slen;
 878         struct msghdr msg = {
 879                 .msg_flags      = XS_SENDMSG_FLAGS,
 880         };
 881         bool vm_wait;
 882         unsigned int sent;
 883         int status;
 884
 885         /* Close the stream if the previous transmission was incomplete */
 886         if (xs_send_request_was_aborted(transport, req)) {
 887                 xprt_force_disconnect(xprt);
 888                 return -ENOTCONN;
 889         }
 890
 891         xs_pktdump("packet data:",
 892                         req->rq_svec->iov_base, req->rq_svec->iov_len);
 893
 894         vm_wait = sk_stream_is_writeable(transport->inet) ? true : false;
 895
 896         req->rq_xtime = ktime_get();
 897         status = xprt_sock_sendmsg(transport->sock, &msg, xdr,
 898                                    transport->xmit.offset, rm, &sent);
 899         dprintk("RPC:       %s(%u) = %d\n",
 900                         __func__, xdr->len - transport->xmit.offset, status);
 901
 902         if (likely(sent > 0) || status == 0) {
 903                 transport->xmit.offset += sent;
 904                 req->rq_bytes_sent = transport->xmit.offset;
 905                 if (likely(req->rq_bytes_sent >= msglen)) {
 906                         req->rq_xmit_bytes_sent += transport->xmit.offset;
 907                         transport->xmit.offset = 0;
 908                         return 0;
 909                 }
 910                 status = -EAGAIN;
 911                 vm_wait = false;
 912         }
 913
 914         switch (status) {
 915         case -EAGAIN:
 916                 status = xs_stream_nospace(req, vm_wait);
 917                 break;
 918         default:
 919                 dprintk("RPC:       sendmsg returned unrecognized error %d\n",
 920                         -status);
 921                 fallthrough;
 922         case -EPIPE:
 923                 xprt_force_disconnect(xprt);
 924                 status = -ENOTCONN;
 925         }
 926
 927         return status;
 928 }
 929
 930 /**
 931  * xs_udp_send_request - write an RPC request to a UDP socket
 932  * @req: pointer to RPC request
 933  *
 934  * Return values:
 935  *        0:    The request has been sent
 936  *   EAGAIN:    The socket was blocked, please call again later to
 937  *              complete the request
 938  * ENOTCONN:    Caller needs to invoke connect logic then call again
 939  *    other:    Some other error occurred, the request was not sent
 940  */
 941 static int xs_udp_send_request(struct rpc_rqst *req)
 942 {
 943         struct rpc_xprt *xprt = req->rq_xprt;
 944         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
 945         struct xdr_buf *xdr = &req->rq_snd_buf;
 946         struct msghdr msg = {
 947                 .msg_name       = xs_addr(xprt),
 948                 .msg_namelen    = xprt->addrlen,
 949                 .msg_flags      = XS_SENDMSG_FLAGS,
 950         };
 951         unsigned int sent;
 952         int status;
 953
 954         xs_pktdump("packet data:",
 955                                 req->rq_svec->iov_base,
 956                                 req->rq_svec->iov_len);
 957
 958         if (!xprt_bound(xprt))
 959                 return -ENOTCONN;
 960
 961         if (!xprt_request_get_cong(xprt, req))
 962                 return -EBADSLT;
 963
 964         status = xdr_alloc_bvec(xdr, rpc_task_gfp_mask());
 965         if (status < 0)
 966                 return status;
 967         req->rq_xtime = ktime_get();
 968         status = xprt_sock_sendmsg(transport->sock, &msg, xdr, 0, 0, &sent);
 969
 970         dprintk("RPC:       xs_udp_send_request(%u) = %d\n",
 971                         xdr->len, status);
 972
 973         /* firewall is blocking us, don't return -EAGAIN or we end up looping */
 974         if (status == -EPERM)
 975                 goto process_status;
 976
 977         if (status == -EAGAIN && sock_writeable(transport->inet))
 978                 status = -ENOBUFS;
 979
 980         if (sent > 0 || status == 0) {
 981                 req->rq_xmit_bytes_sent += sent;
 982                 if (sent >= req->rq_slen)
 983                         return 0;
 984                 /* Still some bytes left; set up for a retry later. */
 985                 status = -EAGAIN;
 986         }
 987
 988 process_status:
 989         switch (status) {
 990         case -ENOTSOCK:
 991                 status = -ENOTCONN;
 992                 /* Should we call xs_close() here? */
 993                 break;
 994         case -EAGAIN:
 995                 status = xs_sock_nospace(req);
 996                 break;
 997         case -ENETUNREACH:
 998         case -ENOBUFS:
 999         case -EPIPE:
1000         case -ECONNREFUSED:
1001         case -EPERM:
1002                 /* When the server has died, an ICMP port unreachable message
1003                  * prompts ECONNREFUSED. */
1004                 break;
1005         default:
1006                 dprintk("RPC:       sendmsg returned unrecognized error %d\n",
1007                         -status);
1008         }
1009
1010         return status;
1011 }
1012
1013 /**
1014  * xs_tcp_send_request - write an RPC request to a TCP socket
1015  * @req: pointer to RPC request
1016  *
1017  * Return values:
1018  *        0:    The request has been sent
1019  *   EAGAIN:    The socket was blocked, please call again later to
1020  *              complete the request
1021  * ENOTCONN:    Caller needs to invoke connect logic then call again
1022  *    other:    Some other error occurred, the request was not sent
1023  *
1024  * XXX: In the case of soft timeouts, should we eventually give up
1025  *      if sendmsg is not able to make progress?
1026  */
1027 static int xs_tcp_send_request(struct rpc_rqst *req)
1028 {
1029         struct rpc_xprt *xprt = req->rq_xprt;
1030         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1031         struct xdr_buf *xdr = &req->rq_snd_buf;
1032         rpc_fraghdr rm = xs_stream_record_marker(xdr);
1033         unsigned int msglen = rm ? req->rq_slen + sizeof(rm) : req->rq_slen;
1034         struct msghdr msg = {
1035                 .msg_flags      = XS_SENDMSG_FLAGS,
1036         };
1037         bool vm_wait;
1038         unsigned int sent;
1039         int status;
1040
1041         /* Close the stream if the previous transmission was incomplete */
1042         if (xs_send_request_was_aborted(transport, req)) {
1043                 if (transport->sock != NULL)
1044                         kernel_sock_shutdown(transport->sock, SHUT_RDWR);
1045                 return -ENOTCONN;
1046         }
1047         if (!transport->inet)
1048                 return -ENOTCONN;
1049
1050         xs_pktdump("packet data:",
1051                                 req->rq_svec->iov_base,
1052                                 req->rq_svec->iov_len);
1053
1054         if (test_bit(XPRT_SOCK_UPD_TIMEOUT, &transport->sock_state))
1055                 xs_tcp_set_socket_timeouts(xprt, transport->sock);
1056
1057         xs_set_srcport(transport, transport->sock);
1058
1059         /* Continue transmitting the packet/record. We must be careful
1060          * to cope with writespace callbacks arriving _after_ we have
1061          * called sendmsg(). */
1062         req->rq_xtime = ktime_get();
1063         tcp_sock_set_cork(transport->inet, true);
1064
1065         vm_wait = sk_stream_is_writeable(transport->inet) ? true : false;
1066
1067         do {
1068                 status = xprt_sock_sendmsg(transport->sock, &msg, xdr,
1069                                            transport->xmit.offset, rm, &sent);
1070
1071                 dprintk("RPC:       xs_tcp_send_request(%u) = %d\n",
1072                                 xdr->len - transport->xmit.offset, status);
1073
1074                 /* If we've sent the entire packet, immediately
1075                  * reset the count of bytes sent. */
1076                 transport->xmit.offset += sent;
1077                 req->rq_bytes_sent = transport->xmit.offset;
1078                 if (likely(req->rq_bytes_sent >= msglen)) {
1079                         req->rq_xmit_bytes_sent += transport->xmit.offset;
1080                         transport->xmit.offset = 0;
1081                         if (atomic_long_read(&xprt->xmit_queuelen) == 1)
1082                                 tcp_sock_set_cork(transport->inet, false);
1083                         return 0;
1084                 }
1085
1086                 WARN_ON_ONCE(sent == 0 && status == 0);
1087
1088                 if (sent > 0)
1089                         vm_wait = false;
1090
1091         } while (status == 0);
1092
1093         switch (status) {
1094         case -ENOTSOCK:
1095                 status = -ENOTCONN;
1096                 /* Should we call xs_close() here? */
1097                 break;
1098         case -EAGAIN:
1099                 status = xs_stream_nospace(req, vm_wait);
1100                 break;
1101         case -ECONNRESET:
1102         case -ECONNREFUSED:
1103         case -ENOTCONN:
1104         case -EADDRINUSE:
1105         case -ENOBUFS:
1106         case -EPIPE:
1107                 break;
1108         default:
1109                 dprintk("RPC:       sendmsg returned unrecognized error %d\n",
1110                         -status);
1111         }
1112
1113         return status;
1114 }
1115
1116 static void xs_save_old_callbacks(struct sock_xprt *transport, struct sock *sk)
1117 {
1118         transport->old_data_ready = sk->sk_data_ready;
1119         transport->old_state_change = sk->sk_state_change;
1120         transport->old_write_space = sk->sk_write_space;
1121         transport->old_error_report = sk->sk_error_report;
1122 }
1123
1124 static void xs_restore_old_callbacks(struct sock_xprt *transport, struct sock *sk)
1125 {
1126         sk->sk_data_ready = transport->old_data_ready;
1127         sk->sk_state_change = transport->old_state_change;
1128         sk->sk_write_space = transport->old_write_space;
1129         sk->sk_error_report = transport->old_error_report;
1130 }
1131
1132 static void xs_sock_reset_state_flags(struct rpc_xprt *xprt)
1133 {
1134         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1135
1136         clear_bit(XPRT_SOCK_DATA_READY, &transport->sock_state);
1137         clear_bit(XPRT_SOCK_WAKE_ERROR, &transport->sock_state);
1138         clear_bit(XPRT_SOCK_WAKE_WRITE, &transport->sock_state);
1139         clear_bit(XPRT_SOCK_WAKE_DISCONNECT, &transport->sock_state);
1140         clear_bit(XPRT_SOCK_NOSPACE, &transport->sock_state);
1141 }
1142
1143 static void xs_run_error_worker(struct sock_xprt *transport, unsigned int nr)
1144 {
1145         set_bit(nr, &transport->sock_state);
1146         queue_work(xprtiod_workqueue, &transport->error_worker);
1147 }
1148
1149 static void xs_sock_reset_connection_flags(struct rpc_xprt *xprt)
1150 {
1151         xprt->connect_cookie++;
1152         smp_mb__before_atomic();
1153         clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
1154         clear_bit(XPRT_CLOSING, &xprt->state);
1155         xs_sock_reset_state_flags(xprt);
1156         smp_mb__after_atomic();
1157 }
1158
1159 /**
1160  * xs_error_report - callback to handle TCP socket state errors
1161  * @sk: socket
1162  *
1163  * Note: we don't call sock_error() since there may be a rpc_task
1164  * using the socket, and so we don't want to clear sk->sk_err.
1165  */
1166 static void xs_error_report(struct sock *sk)
1167 {
1168         struct sock_xprt *transport;
1169         struct rpc_xprt *xprt;
1170
1171         if (!(xprt = xprt_from_sock(sk)))
1172                 return;
1173
1174         transport = container_of(xprt, struct sock_xprt, xprt);
1175         transport->xprt_err = -sk->sk_err;
1176         if (transport->xprt_err == 0)
1177                 return;
1178         dprintk("RPC:       xs_error_report client %p, error=%d...\n",
1179                         xprt, -transport->xprt_err);
1180         trace_rpc_socket_error(xprt, sk->sk_socket, transport->xprt_err);
1181
1182         /* barrier ensures xprt_err is set before XPRT_SOCK_WAKE_ERROR */
1183         smp_mb__before_atomic();
1184         xs_run_error_worker(transport, XPRT_SOCK_WAKE_ERROR);
1185 }
1186
1187 static void xs_reset_transport(struct sock_xprt *transport)
1188 {
1189         struct socket *sock = transport->sock;
1190         struct sock *sk = transport->inet;
1191         struct rpc_xprt *xprt = &transport->xprt;
1192         struct file *filp = transport->file;
1193
1194         if (sk == NULL)
1195                 return;
1196         /*
1197          * Make sure we're calling this in a context from which it is safe
1198          * to call __fput_sync(). In practice that means rpciod and the
1199          * system workqueue.
1200          */
1201         if (!(current->flags & PF_WQ_WORKER)) {
1202                 WARN_ON_ONCE(1);
1203                 set_bit(XPRT_CLOSE_WAIT, &xprt->state);
1204                 return;
1205         }
1206
1207         if (atomic_read(&transport->xprt.swapper))
1208                 sk_clear_memalloc(sk);
1209
1210         kernel_sock_shutdown(sock, SHUT_RDWR);
1211
1212         mutex_lock(&transport->recv_mutex);
1213         lock_sock(sk);
1214         transport->inet = NULL;
1215         transport->sock = NULL;
1216         transport->file = NULL;
1217
1218         sk->sk_user_data = NULL;
1219
1220         xs_restore_old_callbacks(transport, sk);
1221         xprt_clear_connected(xprt);
1222         xs_sock_reset_connection_flags(xprt);
1223         /* Reset stream record info */
1224         xs_stream_reset_connect(transport);
1225         release_sock(sk);
1226         mutex_unlock(&transport->recv_mutex);
1227
1228         trace_rpc_socket_close(xprt, sock);
1229         __fput_sync(filp);
1230
1231         xprt_disconnect_done(xprt);
1232 }
1233
1234 /**
1235  * xs_close - close a socket
1236  * @xprt: transport
1237  *
1238  * This is used when all requests are complete; ie, no DRC state remains
1239  * on the server we want to save.
1240  *
1241  * The caller _must_ be holding XPRT_LOCKED in order to avoid issues with
1242  * xs_reset_transport() zeroing the socket from underneath a writer.
1243  */
1244 static void xs_close(struct rpc_xprt *xprt)
1245 {
1246         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1247
1248         dprintk("RPC:       xs_close xprt %p\n", xprt);
1249
1250         xs_reset_transport(transport);
1251         xprt->reestablish_timeout = 0;
1252 }
1253
1254 static void xs_inject_disconnect(struct rpc_xprt *xprt)
1255 {
1256         dprintk("RPC:       injecting transport disconnect on xprt=%p\n",
1257                 xprt);
1258         xprt_disconnect_done(xprt);
1259 }
1260
1261 static void xs_xprt_free(struct rpc_xprt *xprt)
1262 {
1263         xs_free_peer_addresses(xprt);
1264         xprt_free(xprt);
1265 }
1266
1267 /**
1268  * xs_destroy - prepare to shutdown a transport
1269  * @xprt: doomed transport
1270  *
1271  */
1272 static void xs_destroy(struct rpc_xprt *xprt)
1273 {
1274         struct sock_xprt *transport = container_of(xprt,
1275                         struct sock_xprt, xprt);
1276         dprintk("RPC:       xs_destroy xprt %p\n", xprt);
1277
1278         cancel_delayed_work_sync(&transport->connect_worker);
1279         xs_close(xprt);
1280         cancel_work_sync(&transport->recv_worker);
1281         cancel_work_sync(&transport->error_worker);
1282         xs_xprt_free(xprt);
1283         module_put(THIS_MODULE);
1284 }
1285
1286 /**
1287  * xs_udp_data_read_skb - receive callback for UDP sockets
1288  * @xprt: transport
1289  * @sk: socket
1290  * @skb: skbuff
1291  *
1292  */
1293 static void xs_udp_data_read_skb(struct rpc_xprt *xprt,
1294                 struct sock *sk,
1295                 struct sk_buff *skb)
1296 {
1297         struct rpc_task *task;
1298         struct rpc_rqst *rovr;
1299         int repsize, copied;
1300         u32 _xid;
1301         __be32 *xp;
1302
1303         repsize = skb->len;
1304         if (repsize < 4) {
1305                 dprintk("RPC:       impossible RPC reply size %d!\n", repsize);
1306                 return;
1307         }
1308
1309         /* Copy the XID from the skb... */
1310         xp = skb_header_pointer(skb, 0, sizeof(_xid), &_xid);
1311         if (xp == NULL)
1312                 return;
1313
1314         /* Look up and lock the request corresponding to the given XID */
1315         spin_lock(&xprt->queue_lock);
1316         rovr = xprt_lookup_rqst(xprt, *xp);
1317         if (!rovr)
1318                 goto out_unlock;
1319         xprt_pin_rqst(rovr);
1320         xprt_update_rtt(rovr->rq_task);
1321         spin_unlock(&xprt->queue_lock);
1322         task = rovr->rq_task;
1323
1324         if ((copied = rovr->rq_private_buf.buflen) > repsize)
1325                 copied = repsize;
1326
1327         /* Suck it into the iovec, verify checksum if not done by hw. */
1328         if (csum_partial_copy_to_xdr(&rovr->rq_private_buf, skb)) {
1329                 spin_lock(&xprt->queue_lock);
1330                 __UDPX_INC_STATS(sk, UDP_MIB_INERRORS);
1331                 goto out_unpin;
1332         }
1333
1334
1335         spin_lock(&xprt->transport_lock);
1336         xprt_adjust_cwnd(xprt, task, copied);
1337         spin_unlock(&xprt->transport_lock);
1338         spin_lock(&xprt->queue_lock);
1339         xprt_complete_rqst(task, copied);
1340         __UDPX_INC_STATS(sk, UDP_MIB_INDATAGRAMS);
1341 out_unpin:
1342         xprt_unpin_rqst(rovr);
1343  out_unlock:
1344         spin_unlock(&xprt->queue_lock);
1345 }
1346
1347 static void xs_udp_data_receive(struct sock_xprt *transport)
1348 {
1349         struct sk_buff *skb;
1350         struct sock *sk;
1351         int err;
1352
1353         mutex_lock(&transport->recv_mutex);
1354         sk = transport->inet;
1355         if (sk == NULL)
1356                 goto out;
1357         for (;;) {
1358                 skb = skb_recv_udp(sk, MSG_DONTWAIT, &err);
1359                 if (skb == NULL)
1360                         break;
1361                 xs_udp_data_read_skb(&transport->xprt, sk, skb);
1362                 consume_skb(skb);
1363                 cond_resched();
1364         }
1365         xs_poll_check_readable(transport);
1366 out:
1367         mutex_unlock(&transport->recv_mutex);
1368 }
1369
1370 static void xs_udp_data_receive_workfn(struct work_struct *work)
1371 {
1372         struct sock_xprt *transport =
1373                 container_of(work, struct sock_xprt, recv_worker);
1374         unsigned int pflags = memalloc_nofs_save();
1375
1376         xs_udp_data_receive(transport);
1377         memalloc_nofs_restore(pflags);
1378 }
1379
1380 /**
1381  * xs_data_ready - "data ready" callback for UDP sockets
1382  * @sk: socket with data to read
1383  *
1384  */
1385 static void xs_data_ready(struct sock *sk)
1386 {
1387         struct rpc_xprt *xprt;
1388
1389         dprintk("RPC:       xs_data_ready...\n");
1390         xprt = xprt_from_sock(sk);
1391         if (xprt != NULL) {
1392                 struct sock_xprt *transport = container_of(xprt,
1393                                 struct sock_xprt, xprt);
1394                 transport->old_data_ready(sk);
1395                 /* Any data means we had a useful conversation, so
1396                  * then we don't need to delay the next reconnect
1397                  */
1398                 if (xprt->reestablish_timeout)
1399                         xprt->reestablish_timeout = 0;
1400                 if (!test_and_set_bit(XPRT_SOCK_DATA_READY, &transport->sock_state))
1401                         queue_work(xprtiod_workqueue, &transport->recv_worker);
1402         }
1403 }
1404
1405 /*
1406  * Helper function to force a TCP close if the server is sending
1407  * junk and/or it has put us in CLOSE_WAIT
1408  */
1409 static void xs_tcp_force_close(struct rpc_xprt *xprt)
1410 {
1411         xprt_force_disconnect(xprt);
1412 }
1413
1414 #if defined(CONFIG_SUNRPC_BACKCHANNEL)
1415 static size_t xs_tcp_bc_maxpayload(struct rpc_xprt *xprt)
1416 {
1417         return PAGE_SIZE;
1418 }
1419 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
1420
1421 /**
1422  * xs_local_state_change - callback to handle AF_LOCAL socket state changes
1423  * @sk: socket whose state has changed
1424  *
1425  */
1426 static void xs_local_state_change(struct sock *sk)
1427 {
1428         struct rpc_xprt *xprt;
1429         struct sock_xprt *transport;
1430
1431         if (!(xprt = xprt_from_sock(sk)))
1432                 return;
1433         transport = container_of(xprt, struct sock_xprt, xprt);
1434         if (sk->sk_shutdown & SHUTDOWN_MASK) {
1435                 clear_bit(XPRT_CONNECTED, &xprt->state);
1436                 /* Trigger the socket release */
1437                 xs_run_error_worker(transport, XPRT_SOCK_WAKE_DISCONNECT);
1438         }
1439 }
1440
1441 /**
1442  * xs_tcp_state_change - callback to handle TCP socket state changes
1443  * @sk: socket whose state has changed
1444  *
1445  */
1446 static void xs_tcp_state_change(struct sock *sk)
1447 {
1448         struct rpc_xprt *xprt;
1449         struct sock_xprt *transport;
1450
1451         if (!(xprt = xprt_from_sock(sk)))
1452                 return;
1453         dprintk("RPC:       xs_tcp_state_change client %p...\n", xprt);
1454         dprintk("RPC:       state %x conn %d dead %d zapped %d sk_shutdown %d\n",
1455                         sk->sk_state, xprt_connected(xprt),
1456                         sock_flag(sk, SOCK_DEAD),
1457                         sock_flag(sk, SOCK_ZAPPED),
1458                         sk->sk_shutdown);
1459
1460         transport = container_of(xprt, struct sock_xprt, xprt);
1461         trace_rpc_socket_state_change(xprt, sk->sk_socket);
1462         switch (sk->sk_state) {
1463         case TCP_ESTABLISHED:
1464                 if (!xprt_test_and_set_connected(xprt)) {
1465                         xprt->connect_cookie++;
1466                         clear_bit(XPRT_SOCK_CONNECTING, &transport->sock_state);
1467                         xprt_clear_connecting(xprt);
1468
1469                         xprt->stat.connect_count++;
1470                         xprt->stat.connect_time += (long)jiffies -
1471                                                    xprt->stat.connect_start;
1472                         xs_run_error_worker(transport, XPRT_SOCK_WAKE_PENDING);
1473                 }
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_run_error_worker(transport, XPRT_SOCK_WAKE_DISCONNECT);
1490                 fallthrough;
1491         case TCP_CLOSING:
1492                 /*
1493                  * If the server closed down the connection, make sure that
1494                  * we back off before reconnecting
1495                  */
1496                 if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
1497                         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
1498                 break;
1499         case TCP_LAST_ACK:
1500                 set_bit(XPRT_CLOSING, &xprt->state);
1501                 smp_mb__before_atomic();
1502                 clear_bit(XPRT_CONNECTED, &xprt->state);
1503                 smp_mb__after_atomic();
1504                 break;
1505         case TCP_CLOSE:
1506                 if (test_and_clear_bit(XPRT_SOCK_CONNECTING,
1507                                         &transport->sock_state))
1508                         xprt_clear_connecting(xprt);
1509                 clear_bit(XPRT_CLOSING, &xprt->state);
1510                 /* Trigger the socket release */
1511                 xs_run_error_worker(transport, XPRT_SOCK_WAKE_DISCONNECT);
1512         }
1513 }
1514
1515 static void xs_write_space(struct sock *sk)
1516 {
1517         struct sock_xprt *transport;
1518         struct rpc_xprt *xprt;
1519
1520         if (!sk->sk_socket)
1521                 return;
1522         clear_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
1523
1524         if (unlikely(!(xprt = xprt_from_sock(sk))))
1525                 return;
1526         transport = container_of(xprt, struct sock_xprt, xprt);
1527         if (!test_and_clear_bit(XPRT_SOCK_NOSPACE, &transport->sock_state))
1528                 return;
1529         xs_run_error_worker(transport, XPRT_SOCK_WAKE_WRITE);
1530         sk->sk_write_pending--;
1531 }
1532
1533 /**
1534  * xs_udp_write_space - callback invoked when socket buffer space
1535  *                             becomes available
1536  * @sk: socket whose state has changed
1537  *
1538  * Called when more output buffer space is available for this socket.
1539  * We try not to wake our writers until they can make "significant"
1540  * progress, otherwise we'll waste resources thrashing kernel_sendmsg
1541  * with a bunch of small requests.
1542  */
1543 static void xs_udp_write_space(struct sock *sk)
1544 {
1545         /* from net/core/sock.c:sock_def_write_space */
1546         if (sock_writeable(sk))
1547                 xs_write_space(sk);
1548 }
1549
1550 /**
1551  * xs_tcp_write_space - callback invoked when socket buffer space
1552  *                             becomes available
1553  * @sk: socket whose state has changed
1554  *
1555  * Called when more output buffer space is available for this socket.
1556  * We try not to wake our writers until they can make "significant"
1557  * progress, otherwise we'll waste resources thrashing kernel_sendmsg
1558  * with a bunch of small requests.
1559  */
1560 static void xs_tcp_write_space(struct sock *sk)
1561 {
1562         /* from net/core/stream.c:sk_stream_write_space */
1563         if (sk_stream_is_writeable(sk))
1564                 xs_write_space(sk);
1565 }
1566
1567 static void xs_udp_do_set_buffer_size(struct rpc_xprt *xprt)
1568 {
1569         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1570         struct sock *sk = transport->inet;
1571
1572         if (transport->rcvsize) {
1573                 sk->sk_userlocks |= SOCK_RCVBUF_LOCK;
1574                 sk->sk_rcvbuf = transport->rcvsize * xprt->max_reqs * 2;
1575         }
1576         if (transport->sndsize) {
1577                 sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
1578                 sk->sk_sndbuf = transport->sndsize * xprt->max_reqs * 2;
1579                 sk->sk_write_space(sk);
1580         }
1581 }
1582
1583 /**
1584  * xs_udp_set_buffer_size - set send and receive limits
1585  * @xprt: generic transport
1586  * @sndsize: requested size of send buffer, in bytes
1587  * @rcvsize: requested size of receive buffer, in bytes
1588  *
1589  * Set socket send and receive buffer size limits.
1590  */
1591 static void xs_udp_set_buffer_size(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize)
1592 {
1593         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1594
1595         transport->sndsize = 0;
1596         if (sndsize)
1597                 transport->sndsize = sndsize + 1024;
1598         transport->rcvsize = 0;
1599         if (rcvsize)
1600                 transport->rcvsize = rcvsize + 1024;
1601
1602         xs_udp_do_set_buffer_size(xprt);
1603 }
1604
1605 /**
1606  * xs_udp_timer - called when a retransmit timeout occurs on a UDP transport
1607  * @xprt: controlling transport
1608  * @task: task that timed out
1609  *
1610  * Adjust the congestion window after a retransmit timeout has occurred.
1611  */
1612 static void xs_udp_timer(struct rpc_xprt *xprt, struct rpc_task *task)
1613 {
1614         spin_lock(&xprt->transport_lock);
1615         xprt_adjust_cwnd(xprt, task, -ETIMEDOUT);
1616         spin_unlock(&xprt->transport_lock);
1617 }
1618
1619 static int xs_get_random_port(void)
1620 {
1621         unsigned short min = xprt_min_resvport, max = xprt_max_resvport;
1622         unsigned short range;
1623         unsigned short rand;
1624
1625         if (max < min)
1626                 return -EADDRINUSE;
1627         range = max - min + 1;
1628         rand = (unsigned short) prandom_u32() % range;
1629         return rand + min;
1630 }
1631
1632 static unsigned short xs_sock_getport(struct socket *sock)
1633 {
1634         struct sockaddr_storage buf;
1635         unsigned short port = 0;
1636
1637         if (kernel_getsockname(sock, (struct sockaddr *)&buf) < 0)
1638                 goto out;
1639         switch (buf.ss_family) {
1640         case AF_INET6:
1641                 port = ntohs(((struct sockaddr_in6 *)&buf)->sin6_port);
1642                 break;
1643         case AF_INET:
1644                 port = ntohs(((struct sockaddr_in *)&buf)->sin_port);
1645         }
1646 out:
1647         return port;
1648 }
1649
1650 /**
1651  * xs_set_port - reset the port number in the remote endpoint address
1652  * @xprt: generic transport
1653  * @port: new port number
1654  *
1655  */
1656 static void xs_set_port(struct rpc_xprt *xprt, unsigned short port)
1657 {
1658         dprintk("RPC:       setting port for xprt %p to %u\n", xprt, port);
1659
1660         rpc_set_port(xs_addr(xprt), port);
1661         xs_update_peer_port(xprt);
1662 }
1663
1664 static void xs_set_srcport(struct sock_xprt *transport, struct socket *sock)
1665 {
1666         if (transport->srcport == 0 && transport->xprt.reuseport)
1667                 transport->srcport = xs_sock_getport(sock);
1668 }
1669
1670 static int xs_get_srcport(struct sock_xprt *transport)
1671 {
1672         int port = transport->srcport;
1673
1674         if (port == 0 && transport->xprt.resvport)
1675                 port = xs_get_random_port();
1676         return port;
1677 }
1678
1679 static unsigned short xs_sock_srcport(struct rpc_xprt *xprt)
1680 {
1681         struct sock_xprt *sock = container_of(xprt, struct sock_xprt, xprt);
1682         unsigned short ret = 0;
1683         mutex_lock(&sock->recv_mutex);
1684         if (sock->sock)
1685                 ret = xs_sock_getport(sock->sock);
1686         mutex_unlock(&sock->recv_mutex);
1687         return ret;
1688 }
1689
1690 static int xs_sock_srcaddr(struct rpc_xprt *xprt, char *buf, size_t buflen)
1691 {
1692         struct sock_xprt *sock = container_of(xprt, struct sock_xprt, xprt);
1693         union {
1694                 struct sockaddr sa;
1695                 struct sockaddr_storage st;
1696         } saddr;
1697         int ret = -ENOTCONN;
1698
1699         mutex_lock(&sock->recv_mutex);
1700         if (sock->sock) {
1701                 ret = kernel_getsockname(sock->sock, &saddr.sa);
1702                 if (ret >= 0)
1703                         ret = snprintf(buf, buflen, "%pISc", &saddr.sa);
1704         }
1705         mutex_unlock(&sock->recv_mutex);
1706         return ret;
1707 }
1708
1709 static unsigned short xs_next_srcport(struct sock_xprt *transport, unsigned short port)
1710 {
1711         if (transport->srcport != 0)
1712                 transport->srcport = 0;
1713         if (!transport->xprt.resvport)
1714                 return 0;
1715         if (port <= xprt_min_resvport || port > xprt_max_resvport)
1716                 return xprt_max_resvport;
1717         return --port;
1718 }
1719 static int xs_bind(struct sock_xprt *transport, struct socket *sock)
1720 {
1721         struct sockaddr_storage myaddr;
1722         int err, nloop = 0;
1723         int port = xs_get_srcport(transport);
1724         unsigned short last;
1725
1726         /*
1727          * If we are asking for any ephemeral port (i.e. port == 0 &&
1728          * transport->xprt.resvport == 0), don't bind.  Let the local
1729          * port selection happen implicitly when the socket is used
1730          * (for example at connect time).
1731          *
1732          * This ensures that we can continue to establish TCP
1733          * connections even when all local ephemeral ports are already
1734          * a part of some TCP connection.  This makes no difference
1735          * for UDP sockets, but also doesn't harm them.
1736          *
1737          * If we're asking for any reserved port (i.e. port == 0 &&
1738          * transport->xprt.resvport == 1) xs_get_srcport above will
1739          * ensure that port is non-zero and we will bind as needed.
1740          */
1741         if (port <= 0)
1742                 return port;
1743
1744         memcpy(&myaddr, &transport->srcaddr, transport->xprt.addrlen);
1745         do {
1746                 rpc_set_port((struct sockaddr *)&myaddr, port);
1747                 err = kernel_bind(sock, (struct sockaddr *)&myaddr,
1748                                 transport->xprt.addrlen);
1749                 if (err == 0) {
1750                         if (transport->xprt.reuseport)
1751                                 transport->srcport = port;
1752                         break;
1753                 }
1754                 last = port;
1755                 port = xs_next_srcport(transport, port);
1756                 if (port > last)
1757                         nloop++;
1758         } while (err == -EADDRINUSE && nloop != 2);
1759
1760         if (myaddr.ss_family == AF_INET)
1761                 dprintk("RPC:       %s %pI4:%u: %s (%d)\n", __func__,
1762                                 &((struct sockaddr_in *)&myaddr)->sin_addr,
1763                                 port, err ? "failed" : "ok", err);
1764         else
1765                 dprintk("RPC:       %s %pI6:%u: %s (%d)\n", __func__,
1766                                 &((struct sockaddr_in6 *)&myaddr)->sin6_addr,
1767                                 port, err ? "failed" : "ok", err);
1768         return err;
1769 }
1770
1771 /*
1772  * We don't support autobind on AF_LOCAL sockets
1773  */
1774 static void xs_local_rpcbind(struct rpc_task *task)
1775 {
1776         xprt_set_bound(task->tk_xprt);
1777 }
1778
1779 static void xs_local_set_port(struct rpc_xprt *xprt, unsigned short port)
1780 {
1781 }
1782
1783 #ifdef CONFIG_DEBUG_LOCK_ALLOC
1784 static struct lock_class_key xs_key[3];
1785 static struct lock_class_key xs_slock_key[3];
1786
1787 static inline void xs_reclassify_socketu(struct socket *sock)
1788 {
1789         struct sock *sk = sock->sk;
1790
1791         sock_lock_init_class_and_name(sk, "slock-AF_LOCAL-RPC",
1792                 &xs_slock_key[0], "sk_lock-AF_LOCAL-RPC", &xs_key[0]);
1793 }
1794
1795 static inline void xs_reclassify_socket4(struct socket *sock)
1796 {
1797         struct sock *sk = sock->sk;
1798
1799         sock_lock_init_class_and_name(sk, "slock-AF_INET-RPC",
1800                 &xs_slock_key[1], "sk_lock-AF_INET-RPC", &xs_key[1]);
1801 }
1802
1803 static inline void xs_reclassify_socket6(struct socket *sock)
1804 {
1805         struct sock *sk = sock->sk;
1806
1807         sock_lock_init_class_and_name(sk, "slock-AF_INET6-RPC",
1808                 &xs_slock_key[2], "sk_lock-AF_INET6-RPC", &xs_key[2]);
1809 }
1810
1811 static inline void xs_reclassify_socket(int family, struct socket *sock)
1812 {
1813         if (WARN_ON_ONCE(!sock_allow_reclassification(sock->sk)))
1814                 return;
1815
1816         switch (family) {
1817         case AF_LOCAL:
1818                 xs_reclassify_socketu(sock);
1819                 break;
1820         case AF_INET:
1821                 xs_reclassify_socket4(sock);
1822                 break;
1823         case AF_INET6:
1824                 xs_reclassify_socket6(sock);
1825                 break;
1826         }
1827 }
1828 #else
1829 static inline void xs_reclassify_socket(int family, struct socket *sock)
1830 {
1831 }
1832 #endif
1833
1834 static void xs_dummy_setup_socket(struct work_struct *work)
1835 {
1836 }
1837
1838 static struct socket *xs_create_sock(struct rpc_xprt *xprt,
1839                 struct sock_xprt *transport, int family, int type,
1840                 int protocol, bool reuseport)
1841 {
1842         struct file *filp;
1843         struct socket *sock;
1844         int err;
1845
1846         err = __sock_create(xprt->xprt_net, family, type, protocol, &sock, 1);
1847         if (err < 0) {
1848                 dprintk("RPC:       can't create %d transport socket (%d).\n",
1849                                 protocol, -err);
1850                 goto out;
1851         }
1852         xs_reclassify_socket(family, sock);
1853
1854         if (reuseport)
1855                 sock_set_reuseport(sock->sk);
1856
1857         err = xs_bind(transport, sock);
1858         if (err) {
1859                 sock_release(sock);
1860                 goto out;
1861         }
1862
1863         filp = sock_alloc_file(sock, O_NONBLOCK, NULL);
1864         if (IS_ERR(filp))
1865                 return ERR_CAST(filp);
1866         transport->file = filp;
1867
1868         return sock;
1869 out:
1870         return ERR_PTR(err);
1871 }
1872
1873 static int xs_local_finish_connecting(struct rpc_xprt *xprt,
1874                                       struct socket *sock)
1875 {
1876         struct sock_xprt *transport = container_of(xprt, struct sock_xprt,
1877                                                                         xprt);
1878
1879         if (!transport->inet) {
1880                 struct sock *sk = sock->sk;
1881
1882                 lock_sock(sk);
1883
1884                 xs_save_old_callbacks(transport, sk);
1885
1886                 sk->sk_user_data = xprt;
1887                 sk->sk_data_ready = xs_data_ready;
1888                 sk->sk_write_space = xs_udp_write_space;
1889                 sk->sk_state_change = xs_local_state_change;
1890                 sk->sk_error_report = xs_error_report;
1891
1892                 xprt_clear_connected(xprt);
1893
1894                 /* Reset to new socket */
1895                 transport->sock = sock;
1896                 transport->inet = sk;
1897
1898                 release_sock(sk);
1899         }
1900
1901         xs_stream_start_connect(transport);
1902
1903         return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, 0);
1904 }
1905
1906 /**
1907  * xs_local_setup_socket - create AF_LOCAL socket, connect to a local endpoint
1908  * @transport: socket transport to connect
1909  */
1910 static int xs_local_setup_socket(struct sock_xprt *transport)
1911 {
1912         struct rpc_xprt *xprt = &transport->xprt;
1913         struct file *filp;
1914         struct socket *sock;
1915         int status;
1916
1917         status = __sock_create(xprt->xprt_net, AF_LOCAL,
1918                                         SOCK_STREAM, 0, &sock, 1);
1919         if (status < 0) {
1920                 dprintk("RPC:       can't create AF_LOCAL "
1921                         "transport socket (%d).\n", -status);
1922                 goto out;
1923         }
1924         xs_reclassify_socket(AF_LOCAL, sock);
1925
1926         filp = sock_alloc_file(sock, O_NONBLOCK, NULL);
1927         if (IS_ERR(filp)) {
1928                 status = PTR_ERR(filp);
1929                 goto out;
1930         }
1931         transport->file = filp;
1932
1933         dprintk("RPC:       worker connecting xprt %p via AF_LOCAL to %s\n",
1934                         xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1935
1936         status = xs_local_finish_connecting(xprt, sock);
1937         trace_rpc_socket_connect(xprt, sock, status);
1938         switch (status) {
1939         case 0:
1940                 dprintk("RPC:       xprt %p connected to %s\n",
1941                                 xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1942                 xprt->stat.connect_count++;
1943                 xprt->stat.connect_time += (long)jiffies -
1944                                            xprt->stat.connect_start;
1945                 xprt_set_connected(xprt);
1946                 break;
1947         case -ENOBUFS:
1948                 break;
1949         case -ENOENT:
1950                 dprintk("RPC:       xprt %p: socket %s does not exist\n",
1951                                 xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1952                 break;
1953         case -ECONNREFUSED:
1954                 dprintk("RPC:       xprt %p: connection refused for %s\n",
1955                                 xprt, xprt->address_strings[RPC_DISPLAY_ADDR]);
1956                 break;
1957         default:
1958                 printk(KERN_ERR "%s: unhandled error (%d) connecting to %s\n",
1959                                 __func__, -status,
1960                                 xprt->address_strings[RPC_DISPLAY_ADDR]);
1961         }
1962
1963 out:
1964         xprt_clear_connecting(xprt);
1965         xprt_wake_pending_tasks(xprt, status);
1966         return status;
1967 }
1968
1969 static void xs_local_connect(struct rpc_xprt *xprt, struct rpc_task *task)
1970 {
1971         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
1972         int ret;
1973
1974         if (transport->file)
1975                 goto force_disconnect;
1976
1977         if (RPC_IS_ASYNC(task)) {
1978                 /*
1979                  * We want the AF_LOCAL connect to be resolved in the
1980                  * filesystem namespace of the process making the rpc
1981                  * call.  Thus we connect synchronously.
1982                  *
1983                  * If we want to support asynchronous AF_LOCAL calls,
1984                  * we'll need to figure out how to pass a namespace to
1985                  * connect.
1986                  */
1987                 task->tk_rpc_status = -ENOTCONN;
1988                 rpc_exit(task, -ENOTCONN);
1989                 goto out_wake;
1990         }
1991         ret = xs_local_setup_socket(transport);
1992         if (ret && !RPC_IS_SOFTCONN(task))
1993                 msleep_interruptible(15000);
1994         return;
1995 force_disconnect:
1996         xprt_force_disconnect(xprt);
1997 out_wake:
1998         xprt_clear_connecting(xprt);
1999         xprt_wake_pending_tasks(xprt, -ENOTCONN);
2000 }
2001
2002 #if IS_ENABLED(CONFIG_SUNRPC_SWAP)
2003 /*
2004  * Note that this should be called with XPRT_LOCKED held, or recv_mutex
2005  * held, or when we otherwise know that we have exclusive access to the
2006  * socket, to guard against races with xs_reset_transport.
2007  */
2008 static void xs_set_memalloc(struct rpc_xprt *xprt)
2009 {
2010         struct sock_xprt *transport = container_of(xprt, struct sock_xprt,
2011                         xprt);
2012
2013         /*
2014          * If there's no sock, then we have nothing to set. The
2015          * reconnecting process will get it for us.
2016          */
2017         if (!transport->inet)
2018                 return;
2019         if (atomic_read(&xprt->swapper))
2020                 sk_set_memalloc(transport->inet);
2021 }
2022
2023 /**
2024  * xs_enable_swap - Tag this transport as being used for swap.
2025  * @xprt: transport to tag
2026  *
2027  * Take a reference to this transport on behalf of the rpc_clnt, and
2028  * optionally mark it for swapping if it wasn't already.
2029  */
2030 static int
2031 xs_enable_swap(struct rpc_xprt *xprt)
2032 {
2033         struct sock_xprt *xs = container_of(xprt, struct sock_xprt, xprt);
2034
2035         mutex_lock(&xs->recv_mutex);
2036         if (atomic_inc_return(&xprt->swapper) == 1 &&
2037             xs->inet)
2038                 sk_set_memalloc(xs->inet);
2039         mutex_unlock(&xs->recv_mutex);
2040         return 0;
2041 }
2042
2043 /**
2044  * xs_disable_swap - Untag this transport as being used for swap.
2045  * @xprt: transport to tag
2046  *
2047  * Drop a "swapper" reference to this xprt on behalf of the rpc_clnt. If the
2048  * swapper refcount goes to 0, untag the socket as a memalloc socket.
2049  */
2050 static void
2051 xs_disable_swap(struct rpc_xprt *xprt)
2052 {
2053         struct sock_xprt *xs = container_of(xprt, struct sock_xprt, xprt);
2054
2055         mutex_lock(&xs->recv_mutex);
2056         if (atomic_dec_and_test(&xprt->swapper) &&
2057             xs->inet)
2058                 sk_clear_memalloc(xs->inet);
2059         mutex_unlock(&xs->recv_mutex);
2060 }
2061 #else
2062 static void xs_set_memalloc(struct rpc_xprt *xprt)
2063 {
2064 }
2065
2066 static int
2067 xs_enable_swap(struct rpc_xprt *xprt)
2068 {
2069         return -EINVAL;
2070 }
2071
2072 static void
2073 xs_disable_swap(struct rpc_xprt *xprt)
2074 {
2075 }
2076 #endif
2077
2078 static void xs_udp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
2079 {
2080         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2081
2082         if (!transport->inet) {
2083                 struct sock *sk = sock->sk;
2084
2085                 lock_sock(sk);
2086
2087                 xs_save_old_callbacks(transport, sk);
2088
2089                 sk->sk_user_data = xprt;
2090                 sk->sk_data_ready = xs_data_ready;
2091                 sk->sk_write_space = xs_udp_write_space;
2092
2093                 xprt_set_connected(xprt);
2094
2095                 /* Reset to new socket */
2096                 transport->sock = sock;
2097                 transport->inet = sk;
2098
2099                 xs_set_memalloc(xprt);
2100
2101                 release_sock(sk);
2102         }
2103         xs_udp_do_set_buffer_size(xprt);
2104
2105         xprt->stat.connect_start = jiffies;
2106 }
2107
2108 static void xs_udp_setup_socket(struct work_struct *work)
2109 {
2110         struct sock_xprt *transport =
2111                 container_of(work, struct sock_xprt, connect_worker.work);
2112         struct rpc_xprt *xprt = &transport->xprt;
2113         struct socket *sock;
2114         int status = -EIO;
2115         unsigned int pflags = current->flags;
2116
2117         if (atomic_read(&xprt->swapper))
2118                 current->flags |= PF_MEMALLOC;
2119         sock = xs_create_sock(xprt, transport,
2120                         xs_addr(xprt)->sa_family, SOCK_DGRAM,
2121                         IPPROTO_UDP, false);
2122         if (IS_ERR(sock))
2123                 goto out;
2124
2125         dprintk("RPC:       worker connecting xprt %p via %s to "
2126                                 "%s (port %s)\n", xprt,
2127                         xprt->address_strings[RPC_DISPLAY_PROTO],
2128                         xprt->address_strings[RPC_DISPLAY_ADDR],
2129                         xprt->address_strings[RPC_DISPLAY_PORT]);
2130
2131         xs_udp_finish_connecting(xprt, sock);
2132         trace_rpc_socket_connect(xprt, sock, 0);
2133         status = 0;
2134 out:
2135         xprt_clear_connecting(xprt);
2136         xprt_unlock_connect(xprt, transport);
2137         xprt_wake_pending_tasks(xprt, status);
2138         current_restore_flags(pflags, PF_MEMALLOC);
2139 }
2140
2141 /**
2142  * xs_tcp_shutdown - gracefully shut down a TCP socket
2143  * @xprt: transport
2144  *
2145  * Initiates a graceful shutdown of the TCP socket by calling the
2146  * equivalent of shutdown(SHUT_RDWR);
2147  */
2148 static void xs_tcp_shutdown(struct rpc_xprt *xprt)
2149 {
2150         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2151         struct socket *sock = transport->sock;
2152         int skst = transport->inet ? transport->inet->sk_state : TCP_CLOSE;
2153
2154         if (sock == NULL)
2155                 return;
2156         if (!xprt->reuseport) {
2157                 xs_close(xprt);
2158                 return;
2159         }
2160         switch (skst) {
2161         case TCP_FIN_WAIT1:
2162         case TCP_FIN_WAIT2:
2163                 break;
2164         case TCP_ESTABLISHED:
2165         case TCP_CLOSE_WAIT:
2166                 kernel_sock_shutdown(sock, SHUT_RDWR);
2167                 trace_rpc_socket_shutdown(xprt, sock);
2168                 break;
2169         default:
2170                 xs_reset_transport(transport);
2171         }
2172 }
2173
2174 static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt,
2175                 struct socket *sock)
2176 {
2177         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2178         unsigned int keepidle;
2179         unsigned int keepcnt;
2180         unsigned int timeo;
2181
2182         spin_lock(&xprt->transport_lock);
2183         keepidle = DIV_ROUND_UP(xprt->timeout->to_initval, HZ);
2184         keepcnt = xprt->timeout->to_retries + 1;
2185         timeo = jiffies_to_msecs(xprt->timeout->to_initval) *
2186                 (xprt->timeout->to_retries + 1);
2187         clear_bit(XPRT_SOCK_UPD_TIMEOUT, &transport->sock_state);
2188         spin_unlock(&xprt->transport_lock);
2189
2190         /* TCP Keepalive options */
2191         sock_set_keepalive(sock->sk);
2192         tcp_sock_set_keepidle(sock->sk, keepidle);
2193         tcp_sock_set_keepintvl(sock->sk, keepidle);
2194         tcp_sock_set_keepcnt(sock->sk, keepcnt);
2195
2196         /* TCP user timeout (see RFC5482) */
2197         tcp_sock_set_user_timeout(sock->sk, timeo);
2198 }
2199
2200 static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt,
2201                 unsigned long connect_timeout,
2202                 unsigned long reconnect_timeout)
2203 {
2204         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2205         struct rpc_timeout to;
2206         unsigned long initval;
2207
2208         spin_lock(&xprt->transport_lock);
2209         if (reconnect_timeout < xprt->max_reconnect_timeout)
2210                 xprt->max_reconnect_timeout = reconnect_timeout;
2211         if (connect_timeout < xprt->connect_timeout) {
2212                 memcpy(&to, xprt->timeout, sizeof(to));
2213                 initval = DIV_ROUND_UP(connect_timeout, to.to_retries + 1);
2214                 /* Arbitrary lower limit */
2215                 if (initval <  XS_TCP_INIT_REEST_TO << 1)
2216                         initval = XS_TCP_INIT_REEST_TO << 1;
2217                 to.to_initval = initval;
2218                 to.to_maxval = initval;
2219                 memcpy(&transport->tcp_timeout, &to,
2220                                 sizeof(transport->tcp_timeout));
2221                 xprt->timeout = &transport->tcp_timeout;
2222                 xprt->connect_timeout = connect_timeout;
2223         }
2224         set_bit(XPRT_SOCK_UPD_TIMEOUT, &transport->sock_state);
2225         spin_unlock(&xprt->transport_lock);
2226 }
2227
2228 static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
2229 {
2230         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2231
2232         if (!transport->inet) {
2233                 struct sock *sk = sock->sk;
2234
2235                 /* Avoid temporary address, they are bad for long-lived
2236                  * connections such as NFS mounts.
2237                  * RFC4941, section 3.6 suggests that:
2238                  *    Individual applications, which have specific
2239                  *    knowledge about the normal duration of connections,
2240                  *    MAY override this as appropriate.
2241                  */
2242                 if (xs_addr(xprt)->sa_family == PF_INET6) {
2243                         ip6_sock_set_addr_preferences(sk,
2244                                 IPV6_PREFER_SRC_PUBLIC);
2245                 }
2246
2247                 xs_tcp_set_socket_timeouts(xprt, sock);
2248                 tcp_sock_set_nodelay(sk);
2249
2250                 lock_sock(sk);
2251
2252                 xs_save_old_callbacks(transport, sk);
2253
2254                 sk->sk_user_data = xprt;
2255                 sk->sk_data_ready = xs_data_ready;
2256                 sk->sk_state_change = xs_tcp_state_change;
2257                 sk->sk_write_space = xs_tcp_write_space;
2258                 sk->sk_error_report = xs_error_report;
2259
2260                 /* socket options */
2261                 sock_reset_flag(sk, SOCK_LINGER);
2262
2263                 xprt_clear_connected(xprt);
2264
2265                 /* Reset to new socket */
2266                 transport->sock = sock;
2267                 transport->inet = sk;
2268
2269                 release_sock(sk);
2270         }
2271
2272         if (!xprt_bound(xprt))
2273                 return -ENOTCONN;
2274
2275         xs_set_memalloc(xprt);
2276
2277         xs_stream_start_connect(transport);
2278
2279         /* Tell the socket layer to start connecting... */
2280         set_bit(XPRT_SOCK_CONNECTING, &transport->sock_state);
2281         return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, O_NONBLOCK);
2282 }
2283
2284 /**
2285  * xs_tcp_setup_socket - create a TCP socket and connect to a remote endpoint
2286  * @work: queued work item
2287  *
2288  * Invoked by a work queue tasklet.
2289  */
2290 static void xs_tcp_setup_socket(struct work_struct *work)
2291 {
2292         struct sock_xprt *transport =
2293                 container_of(work, struct sock_xprt, connect_worker.work);
2294         struct socket *sock = transport->sock;
2295         struct rpc_xprt *xprt = &transport->xprt;
2296         int status;
2297         unsigned int pflags = current->flags;
2298
2299         if (atomic_read(&xprt->swapper))
2300                 current->flags |= PF_MEMALLOC;
2301
2302         if (xprt_connected(xprt))
2303                 goto out;
2304         if (test_and_clear_bit(XPRT_SOCK_CONNECT_SENT,
2305                                &transport->sock_state) ||
2306             !sock) {
2307                 xs_reset_transport(transport);
2308                 sock = xs_create_sock(xprt, transport, xs_addr(xprt)->sa_family,
2309                                       SOCK_STREAM, IPPROTO_TCP, true);
2310                 if (IS_ERR(sock)) {
2311                         xprt_wake_pending_tasks(xprt, PTR_ERR(sock));
2312                         goto out;
2313                 }
2314         }
2315
2316         dprintk("RPC:       worker connecting xprt %p via %s to "
2317                                 "%s (port %s)\n", xprt,
2318                         xprt->address_strings[RPC_DISPLAY_PROTO],
2319                         xprt->address_strings[RPC_DISPLAY_ADDR],
2320                         xprt->address_strings[RPC_DISPLAY_PORT]);
2321
2322         status = xs_tcp_finish_connecting(xprt, sock);
2323         trace_rpc_socket_connect(xprt, sock, status);
2324         dprintk("RPC:       %p connect status %d connected %d sock state %d\n",
2325                         xprt, -status, xprt_connected(xprt),
2326                         sock->sk->sk_state);
2327         switch (status) {
2328         case 0:
2329         case -EINPROGRESS:
2330                 /* SYN_SENT! */
2331                 set_bit(XPRT_SOCK_CONNECT_SENT, &transport->sock_state);
2332                 if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
2333                         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
2334                 fallthrough;
2335         case -EALREADY:
2336                 goto out_unlock;
2337         case -EADDRNOTAVAIL:
2338                 /* Source port number is unavailable. Try a new one! */
2339                 transport->srcport = 0;
2340                 status = -EAGAIN;
2341                 break;
2342         case -EINVAL:
2343                 /* Happens, for instance, if the user specified a link
2344                  * local IPv6 address without a scope-id.
2345                  */
2346         case -ECONNREFUSED:
2347         case -ECONNRESET:
2348         case -ENETDOWN:
2349         case -ENETUNREACH:
2350         case -EHOSTUNREACH:
2351         case -EADDRINUSE:
2352         case -ENOBUFS:
2353                 break;
2354         default:
2355                 printk("%s: connect returned unhandled error %d\n",
2356                         __func__, status);
2357                 status = -EAGAIN;
2358         }
2359
2360         /* xs_tcp_force_close() wakes tasks with a fixed error code.
2361          * We need to wake them first to ensure the correct error code.
2362          */
2363         xprt_wake_pending_tasks(xprt, status);
2364         xs_tcp_force_close(xprt);
2365 out:
2366         xprt_clear_connecting(xprt);
2367 out_unlock:
2368         xprt_unlock_connect(xprt, transport);
2369         current_restore_flags(pflags, PF_MEMALLOC);
2370 }
2371
2372 /**
2373  * xs_connect - connect a socket to a remote endpoint
2374  * @xprt: pointer to transport structure
2375  * @task: address of RPC task that manages state of connect request
2376  *
2377  * TCP: If the remote end dropped the connection, delay reconnecting.
2378  *
2379  * UDP socket connects are synchronous, but we use a work queue anyway
2380  * to guarantee that even unprivileged user processes can set up a
2381  * socket on a privileged port.
2382  *
2383  * If a UDP socket connect fails, the delay behavior here prevents
2384  * retry floods (hard mounts).
2385  */
2386 static void xs_connect(struct rpc_xprt *xprt, struct rpc_task *task)
2387 {
2388         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2389         unsigned long delay = 0;
2390
2391         WARN_ON_ONCE(!xprt_lock_connect(xprt, task, transport));
2392
2393         if (transport->sock != NULL) {
2394                 dprintk("RPC:       xs_connect delayed xprt %p for %lu "
2395                         "seconds\n", xprt, xprt->reestablish_timeout / HZ);
2396
2397                 delay = xprt_reconnect_delay(xprt);
2398                 xprt_reconnect_backoff(xprt, XS_TCP_INIT_REEST_TO);
2399
2400         } else
2401                 dprintk("RPC:       xs_connect scheduled xprt %p\n", xprt);
2402
2403         queue_delayed_work(xprtiod_workqueue,
2404                         &transport->connect_worker,
2405                         delay);
2406 }
2407
2408 static void xs_wake_disconnect(struct sock_xprt *transport)
2409 {
2410         if (test_and_clear_bit(XPRT_SOCK_WAKE_DISCONNECT, &transport->sock_state))
2411                 xs_tcp_force_close(&transport->xprt);
2412 }
2413
2414 static void xs_wake_write(struct sock_xprt *transport)
2415 {
2416         if (test_and_clear_bit(XPRT_SOCK_WAKE_WRITE, &transport->sock_state))
2417                 xprt_write_space(&transport->xprt);
2418 }
2419
2420 static void xs_wake_error(struct sock_xprt *transport)
2421 {
2422         int sockerr;
2423
2424         if (!test_bit(XPRT_SOCK_WAKE_ERROR, &transport->sock_state))
2425                 return;
2426         mutex_lock(&transport->recv_mutex);
2427         if (transport->sock == NULL)
2428                 goto out;
2429         if (!test_and_clear_bit(XPRT_SOCK_WAKE_ERROR, &transport->sock_state))
2430                 goto out;
2431         sockerr = xchg(&transport->xprt_err, 0);
2432         if (sockerr < 0)
2433                 xprt_wake_pending_tasks(&transport->xprt, sockerr);
2434 out:
2435         mutex_unlock(&transport->recv_mutex);
2436 }
2437
2438 static void xs_wake_pending(struct sock_xprt *transport)
2439 {
2440         if (test_and_clear_bit(XPRT_SOCK_WAKE_PENDING, &transport->sock_state))
2441                 xprt_wake_pending_tasks(&transport->xprt, -EAGAIN);
2442 }
2443
2444 static void xs_error_handle(struct work_struct *work)
2445 {
2446         struct sock_xprt *transport = container_of(work,
2447                         struct sock_xprt, error_worker);
2448
2449         xs_wake_disconnect(transport);
2450         xs_wake_write(transport);
2451         xs_wake_error(transport);
2452         xs_wake_pending(transport);
2453 }
2454
2455 /**
2456  * xs_local_print_stats - display AF_LOCAL socket-specific stats
2457  * @xprt: rpc_xprt struct containing statistics
2458  * @seq: output file
2459  *
2460  */
2461 static void xs_local_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
2462 {
2463         long idle_time = 0;
2464
2465         if (xprt_connected(xprt))
2466                 idle_time = (long)(jiffies - xprt->last_used) / HZ;
2467
2468         seq_printf(seq, "\txprt:\tlocal %lu %lu %lu %ld %lu %lu %lu "
2469                         "%llu %llu %lu %llu %llu\n",
2470                         xprt->stat.bind_count,
2471                         xprt->stat.connect_count,
2472                         xprt->stat.connect_time / HZ,
2473                         idle_time,
2474                         xprt->stat.sends,
2475                         xprt->stat.recvs,
2476                         xprt->stat.bad_xids,
2477                         xprt->stat.req_u,
2478                         xprt->stat.bklog_u,
2479                         xprt->stat.max_slots,
2480                         xprt->stat.sending_u,
2481                         xprt->stat.pending_u);
2482 }
2483
2484 /**
2485  * xs_udp_print_stats - display UDP socket-specific stats
2486  * @xprt: rpc_xprt struct containing statistics
2487  * @seq: output file
2488  *
2489  */
2490 static void xs_udp_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
2491 {
2492         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2493
2494         seq_printf(seq, "\txprt:\tudp %u %lu %lu %lu %lu %llu %llu "
2495                         "%lu %llu %llu\n",
2496                         transport->srcport,
2497                         xprt->stat.bind_count,
2498                         xprt->stat.sends,
2499                         xprt->stat.recvs,
2500                         xprt->stat.bad_xids,
2501                         xprt->stat.req_u,
2502                         xprt->stat.bklog_u,
2503                         xprt->stat.max_slots,
2504                         xprt->stat.sending_u,
2505                         xprt->stat.pending_u);
2506 }
2507
2508 /**
2509  * xs_tcp_print_stats - display TCP socket-specific stats
2510  * @xprt: rpc_xprt struct containing statistics
2511  * @seq: output file
2512  *
2513  */
2514 static void xs_tcp_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
2515 {
2516         struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
2517         long idle_time = 0;
2518
2519         if (xprt_connected(xprt))
2520                 idle_time = (long)(jiffies - xprt->last_used) / HZ;
2521
2522         seq_printf(seq, "\txprt:\ttcp %u %lu %lu %lu %ld %lu %lu %lu "
2523                         "%llu %llu %lu %llu %llu\n",
2524                         transport->srcport,
2525                         xprt->stat.bind_count,
2526                         xprt->stat.connect_count,
2527                         xprt->stat.connect_time / HZ,
2528                         idle_time,
2529                         xprt->stat.sends,
2530                         xprt->stat.recvs,
2531                         xprt->stat.bad_xids,
2532                         xprt->stat.req_u,
2533                         xprt->stat.bklog_u,
2534                         xprt->stat.max_slots,
2535                         xprt->stat.sending_u,
2536                         xprt->stat.pending_u);
2537 }
2538
2539 /*
2540  * Allocate a bunch of pages for a scratch buffer for the rpc code. The reason
2541  * we allocate pages instead doing a kmalloc like rpc_malloc is because we want
2542  * to use the server side send routines.
2543  */
2544 static int bc_malloc(struct rpc_task *task)
2545 {
2546         struct rpc_rqst *rqst = task->tk_rqstp;
2547         size_t size = rqst->rq_callsize;
2548         struct page *page;
2549         struct rpc_buffer *buf;
2550
2551         if (size > PAGE_SIZE - sizeof(struct rpc_buffer)) {
2552                 WARN_ONCE(1, "xprtsock: large bc buffer request (size %zu)\n",
2553                           size);
2554                 return -EINVAL;
2555         }
2556
2557         page = alloc_page(GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN);
2558         if (!page)
2559                 return -ENOMEM;
2560
2561         buf = page_address(page);
2562         buf->len = PAGE_SIZE;
2563
2564         rqst->rq_buffer = buf->data;
2565         rqst->rq_rbuffer = (char *)rqst->rq_buffer + rqst->rq_callsize;
2566         return 0;
2567 }
2568
2569 /*
2570  * Free the space allocated in the bc_alloc routine
2571  */
2572 static void bc_free(struct rpc_task *task)
2573 {
2574         void *buffer = task->tk_rqstp->rq_buffer;
2575         struct rpc_buffer *buf;
2576
2577         buf = container_of(buffer, struct rpc_buffer, data);
2578         free_page((unsigned long)buf);
2579 }
2580
2581 static int bc_sendto(struct rpc_rqst *req)
2582 {
2583         struct xdr_buf *xdr = &req->rq_snd_buf;
2584         struct sock_xprt *transport =
2585                         container_of(req->rq_xprt, struct sock_xprt, xprt);
2586         struct msghdr msg = {
2587                 .msg_flags      = 0,
2588         };
2589         rpc_fraghdr marker = cpu_to_be32(RPC_LAST_STREAM_FRAGMENT |
2590                                          (u32)xdr->len);
2591         unsigned int sent = 0;
2592         int err;
2593
2594         req->rq_xtime = ktime_get();
2595         err = xdr_alloc_bvec(xdr, rpc_task_gfp_mask());
2596         if (err < 0)
2597                 return err;
2598         err = xprt_sock_sendmsg(transport->sock, &msg, xdr, 0, marker, &sent);
2599         xdr_free_bvec(xdr);
2600         if (err < 0 || sent != (xdr->len + sizeof(marker)))
2601                 return -EAGAIN;
2602         return sent;
2603 }
2604
2605 /**
2606  * bc_send_request - Send a backchannel Call on a TCP socket
2607  * @req: rpc_rqst containing Call message to be sent
2608  *
2609  * xpt_mutex ensures @rqstp's whole message is written to the socket
2610  * without interruption.
2611  *
2612  * Return values:
2613  *   %0 if the message was sent successfully
2614  *   %ENOTCONN if the message was not sent
2615  */
2616 static int bc_send_request(struct rpc_rqst *req)
2617 {
2618         struct svc_xprt *xprt;
2619         int len;
2620
2621         /*
2622          * Get the server socket associated with this callback xprt
2623          */
2624         xprt = req->rq_xprt->bc_xprt;
2625
2626         /*
2627          * Grab the mutex to serialize data as the connection is shared
2628          * with the fore channel
2629          */
2630         mutex_lock(&xprt->xpt_mutex);
2631         if (test_bit(XPT_DEAD, &xprt->xpt_flags))
2632                 len = -ENOTCONN;
2633         else
2634                 len = bc_sendto(req);
2635         mutex_unlock(&xprt->xpt_mutex);
2636
2637         if (len > 0)
2638                 len = 0;
2639
2640         return len;
2641 }
2642
2643 /*
2644  * The close routine. Since this is client initiated, we do nothing
2645  */
2646
2647 static void bc_close(struct rpc_xprt *xprt)
2648 {
2649         xprt_disconnect_done(xprt);
2650 }
2651
2652 /*
2653  * The xprt destroy routine. Again, because this connection is client
2654  * initiated, we do nothing
2655  */
2656
2657 static void bc_destroy(struct rpc_xprt *xprt)
2658 {
2659         dprintk("RPC:       bc_destroy xprt %p\n", xprt);
2660
2661         xs_xprt_free(xprt);
2662         module_put(THIS_MODULE);
2663 }
2664
2665 static const struct rpc_xprt_ops xs_local_ops = {
2666         .reserve_xprt           = xprt_reserve_xprt,
2667         .release_xprt           = xprt_release_xprt,
2668         .alloc_slot             = xprt_alloc_slot,
2669         .free_slot              = xprt_free_slot,
2670         .rpcbind                = xs_local_rpcbind,
2671         .set_port               = xs_local_set_port,
2672         .connect                = xs_local_connect,
2673         .buf_alloc              = rpc_malloc,
2674         .buf_free               = rpc_free,
2675         .prepare_request        = xs_stream_prepare_request,
2676         .send_request           = xs_local_send_request,
2677         .wait_for_reply_request = xprt_wait_for_reply_request_def,
2678         .close                  = xs_close,
2679         .destroy                = xs_destroy,
2680         .print_stats            = xs_local_print_stats,
2681         .enable_swap            = xs_enable_swap,
2682         .disable_swap           = xs_disable_swap,
2683 };
2684
2685 static const struct rpc_xprt_ops xs_udp_ops = {
2686         .set_buffer_size        = xs_udp_set_buffer_size,
2687         .reserve_xprt           = xprt_reserve_xprt_cong,
2688         .release_xprt           = xprt_release_xprt_cong,
2689         .alloc_slot             = xprt_alloc_slot,
2690         .free_slot              = xprt_free_slot,
2691         .rpcbind                = rpcb_getport_async,
2692         .set_port               = xs_set_port,
2693         .connect                = xs_connect,
2694         .get_srcaddr            = xs_sock_srcaddr,
2695         .get_srcport            = xs_sock_srcport,
2696         .buf_alloc              = rpc_malloc,
2697         .buf_free               = rpc_free,
2698         .send_request           = xs_udp_send_request,
2699         .wait_for_reply_request = xprt_wait_for_reply_request_rtt,
2700         .timer                  = xs_udp_timer,
2701         .release_request        = xprt_release_rqst_cong,
2702         .close                  = xs_close,
2703         .destroy                = xs_destroy,
2704         .print_stats            = xs_udp_print_stats,
2705         .enable_swap            = xs_enable_swap,
2706         .disable_swap           = xs_disable_swap,
2707         .inject_disconnect      = xs_inject_disconnect,
2708 };
2709
2710 static const struct rpc_xprt_ops xs_tcp_ops = {
2711         .reserve_xprt           = xprt_reserve_xprt,
2712         .release_xprt           = xprt_release_xprt,
2713         .alloc_slot             = xprt_alloc_slot,
2714         .free_slot              = xprt_free_slot,
2715         .rpcbind                = rpcb_getport_async,
2716         .set_port               = xs_set_port,
2717         .connect                = xs_connect,
2718         .get_srcaddr            = xs_sock_srcaddr,
2719         .get_srcport            = xs_sock_srcport,
2720         .buf_alloc              = rpc_malloc,
2721         .buf_free               = rpc_free,
2722         .prepare_request        = xs_stream_prepare_request,
2723         .send_request           = xs_tcp_send_request,
2724         .wait_for_reply_request = xprt_wait_for_reply_request_def,
2725         .close                  = xs_tcp_shutdown,
2726         .destroy                = xs_destroy,
2727         .set_connect_timeout    = xs_tcp_set_connect_timeout,
2728         .print_stats            = xs_tcp_print_stats,
2729         .enable_swap            = xs_enable_swap,
2730         .disable_swap           = xs_disable_swap,
2731         .inject_disconnect      = xs_inject_disconnect,
2732 #ifdef CONFIG_SUNRPC_BACKCHANNEL
2733         .bc_setup               = xprt_setup_bc,
2734         .bc_maxpayload          = xs_tcp_bc_maxpayload,
2735         .bc_num_slots           = xprt_bc_max_slots,
2736         .bc_free_rqst           = xprt_free_bc_rqst,
2737         .bc_destroy             = xprt_destroy_bc,
2738 #endif
2739 };
2740
2741 /*
2742  * The rpc_xprt_ops for the server backchannel
2743  */
2744
2745 static const struct rpc_xprt_ops bc_tcp_ops = {
2746         .reserve_xprt           = xprt_reserve_xprt,
2747         .release_xprt           = xprt_release_xprt,
2748         .alloc_slot             = xprt_alloc_slot,
2749         .free_slot              = xprt_free_slot,
2750         .buf_alloc              = bc_malloc,
2751         .buf_free               = bc_free,
2752         .send_request           = bc_send_request,
2753         .wait_for_reply_request = xprt_wait_for_reply_request_def,
2754         .close                  = bc_close,
2755         .destroy                = bc_destroy,
2756         .print_stats            = xs_tcp_print_stats,
2757         .enable_swap            = xs_enable_swap,
2758         .disable_swap           = xs_disable_swap,
2759         .inject_disconnect      = xs_inject_disconnect,
2760 };
2761
2762 static int xs_init_anyaddr(const int family, struct sockaddr *sap)
2763 {
2764         static const struct sockaddr_in sin = {
2765                 .sin_family             = AF_INET,
2766                 .sin_addr.s_addr        = htonl(INADDR_ANY),
2767         };
2768         static const struct sockaddr_in6 sin6 = {
2769                 .sin6_family            = AF_INET6,
2770                 .sin6_addr              = IN6ADDR_ANY_INIT,
2771         };
2772
2773         switch (family) {
2774         case AF_LOCAL:
2775                 break;
2776         case AF_INET:
2777                 memcpy(sap, &sin, sizeof(sin));
2778                 break;
2779         case AF_INET6:
2780                 memcpy(sap, &sin6, sizeof(sin6));
2781                 break;
2782         default:
2783                 dprintk("RPC:       %s: Bad address family\n", __func__);
2784                 return -EAFNOSUPPORT;
2785         }
2786         return 0;
2787 }
2788
2789 static struct rpc_xprt *xs_setup_xprt(struct xprt_create *args,
2790                                       unsigned int slot_table_size,
2791                                       unsigned int max_slot_table_size)
2792 {
2793         struct rpc_xprt *xprt;
2794         struct sock_xprt *new;
2795
2796         if (args->addrlen > sizeof(xprt->addr)) {
2797                 dprintk("RPC:       xs_setup_xprt: address too large\n");
2798                 return ERR_PTR(-EBADF);
2799         }
2800
2801         xprt = xprt_alloc(args->net, sizeof(*new), slot_table_size,
2802                         max_slot_table_size);
2803         if (xprt == NULL) {
2804                 dprintk("RPC:       xs_setup_xprt: couldn't allocate "
2805                                 "rpc_xprt\n");
2806                 return ERR_PTR(-ENOMEM);
2807         }
2808
2809         new = container_of(xprt, struct sock_xprt, xprt);
2810         mutex_init(&new->recv_mutex);
2811         memcpy(&xprt->addr, args->dstaddr, args->addrlen);
2812         xprt->addrlen = args->addrlen;
2813         if (args->srcaddr)
2814                 memcpy(&new->srcaddr, args->srcaddr, args->addrlen);
2815         else {
2816                 int err;
2817                 err = xs_init_anyaddr(args->dstaddr->sa_family,
2818                                         (struct sockaddr *)&new->srcaddr);
2819                 if (err != 0) {
2820                         xprt_free(xprt);
2821                         return ERR_PTR(err);
2822                 }
2823         }
2824
2825         return xprt;
2826 }
2827
2828 static const struct rpc_timeout xs_local_default_timeout = {
2829         .to_initval = 10 * HZ,
2830         .to_maxval = 10 * HZ,
2831         .to_retries = 2,
2832 };
2833
2834 /**
2835  * xs_setup_local - Set up transport to use an AF_LOCAL socket
2836  * @args: rpc transport creation arguments
2837  *
2838  * AF_LOCAL is a "tpi_cots_ord" transport, just like TCP
2839  */
2840 static struct rpc_xprt *xs_setup_local(struct xprt_create *args)
2841 {
2842         struct sockaddr_un *sun = (struct sockaddr_un *)args->dstaddr;
2843         struct sock_xprt *transport;
2844         struct rpc_xprt *xprt;
2845         struct rpc_xprt *ret;
2846
2847         xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
2848                         xprt_max_tcp_slot_table_entries);
2849         if (IS_ERR(xprt))
2850                 return xprt;
2851         transport = container_of(xprt, struct sock_xprt, xprt);
2852
2853         xprt->prot = 0;
2854         xprt->xprt_class = &xs_local_transport;
2855         xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
2856
2857         xprt->bind_timeout = XS_BIND_TO;
2858         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
2859         xprt->idle_timeout = XS_IDLE_DISC_TO;
2860
2861         xprt->ops = &xs_local_ops;
2862         xprt->timeout = &xs_local_default_timeout;
2863
2864         INIT_WORK(&transport->recv_worker, xs_stream_data_receive_workfn);
2865         INIT_WORK(&transport->error_worker, xs_error_handle);
2866         INIT_DELAYED_WORK(&transport->connect_worker, xs_dummy_setup_socket);
2867
2868         switch (sun->sun_family) {
2869         case AF_LOCAL:
2870                 if (sun->sun_path[0] != '/') {
2871                         dprintk("RPC:       bad AF_LOCAL address: %s\n",
2872                                         sun->sun_path);
2873                         ret = ERR_PTR(-EINVAL);
2874                         goto out_err;
2875                 }
2876                 xprt_set_bound(xprt);
2877                 xs_format_peer_addresses(xprt, "local", RPCBIND_NETID_LOCAL);
2878                 break;
2879         default:
2880                 ret = ERR_PTR(-EAFNOSUPPORT);
2881                 goto out_err;
2882         }
2883
2884         dprintk("RPC:       set up xprt to %s via AF_LOCAL\n",
2885                         xprt->address_strings[RPC_DISPLAY_ADDR]);
2886
2887         if (try_module_get(THIS_MODULE))
2888                 return xprt;
2889         ret = ERR_PTR(-EINVAL);
2890 out_err:
2891         xs_xprt_free(xprt);
2892         return ret;
2893 }
2894
2895 static const struct rpc_timeout xs_udp_default_timeout = {
2896         .to_initval = 5 * HZ,
2897         .to_maxval = 30 * HZ,
2898         .to_increment = 5 * HZ,
2899         .to_retries = 5,
2900 };
2901
2902 /**
2903  * xs_setup_udp - Set up transport to use a UDP socket
2904  * @args: rpc transport creation arguments
2905  *
2906  */
2907 static struct rpc_xprt *xs_setup_udp(struct xprt_create *args)
2908 {
2909         struct sockaddr *addr = args->dstaddr;
2910         struct rpc_xprt *xprt;
2911         struct sock_xprt *transport;
2912         struct rpc_xprt *ret;
2913
2914         xprt = xs_setup_xprt(args, xprt_udp_slot_table_entries,
2915                         xprt_udp_slot_table_entries);
2916         if (IS_ERR(xprt))
2917                 return xprt;
2918         transport = container_of(xprt, struct sock_xprt, xprt);
2919
2920         xprt->prot = IPPROTO_UDP;
2921         xprt->xprt_class = &xs_udp_transport;
2922         /* XXX: header size can vary due to auth type, IPv6, etc. */
2923         xprt->max_payload = (1U << 16) - (MAX_HEADER << 3);
2924
2925         xprt->bind_timeout = XS_BIND_TO;
2926         xprt->reestablish_timeout = XS_UDP_REEST_TO;
2927         xprt->idle_timeout = XS_IDLE_DISC_TO;
2928
2929         xprt->ops = &xs_udp_ops;
2930
2931         xprt->timeout = &xs_udp_default_timeout;
2932
2933         INIT_WORK(&transport->recv_worker, xs_udp_data_receive_workfn);
2934         INIT_WORK(&transport->error_worker, xs_error_handle);
2935         INIT_DELAYED_WORK(&transport->connect_worker, xs_udp_setup_socket);
2936
2937         switch (addr->sa_family) {
2938         case AF_INET:
2939                 if (((struct sockaddr_in *)addr)->sin_port != htons(0))
2940                         xprt_set_bound(xprt);
2941
2942                 xs_format_peer_addresses(xprt, "udp", RPCBIND_NETID_UDP);
2943                 break;
2944         case AF_INET6:
2945                 if (((struct sockaddr_in6 *)addr)->sin6_port != htons(0))
2946                         xprt_set_bound(xprt);
2947
2948                 xs_format_peer_addresses(xprt, "udp", RPCBIND_NETID_UDP6);
2949                 break;
2950         default:
2951                 ret = ERR_PTR(-EAFNOSUPPORT);
2952                 goto out_err;
2953         }
2954
2955         if (xprt_bound(xprt))
2956                 dprintk("RPC:       set up xprt to %s (port %s) via %s\n",
2957                                 xprt->address_strings[RPC_DISPLAY_ADDR],
2958                                 xprt->address_strings[RPC_DISPLAY_PORT],
2959                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
2960         else
2961                 dprintk("RPC:       set up xprt to %s (autobind) via %s\n",
2962                                 xprt->address_strings[RPC_DISPLAY_ADDR],
2963                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
2964
2965         if (try_module_get(THIS_MODULE))
2966                 return xprt;
2967         ret = ERR_PTR(-EINVAL);
2968 out_err:
2969         xs_xprt_free(xprt);
2970         return ret;
2971 }
2972
2973 static const struct rpc_timeout xs_tcp_default_timeout = {
2974         .to_initval = 60 * HZ,
2975         .to_maxval = 60 * HZ,
2976         .to_retries = 2,
2977 };
2978
2979 /**
2980  * xs_setup_tcp - Set up transport to use a TCP socket
2981  * @args: rpc transport creation arguments
2982  *
2983  */
2984 static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args)
2985 {
2986         struct sockaddr *addr = args->dstaddr;
2987         struct rpc_xprt *xprt;
2988         struct sock_xprt *transport;
2989         struct rpc_xprt *ret;
2990         unsigned int max_slot_table_size = xprt_max_tcp_slot_table_entries;
2991
2992         if (args->flags & XPRT_CREATE_INFINITE_SLOTS)
2993                 max_slot_table_size = RPC_MAX_SLOT_TABLE_LIMIT;
2994
2995         xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
2996                         max_slot_table_size);
2997         if (IS_ERR(xprt))
2998                 return xprt;
2999         transport = container_of(xprt, struct sock_xprt, xprt);
3000
3001         xprt->prot = IPPROTO_TCP;
3002         xprt->xprt_class = &xs_tcp_transport;
3003         xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
3004
3005         xprt->bind_timeout = XS_BIND_TO;
3006         xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
3007         xprt->idle_timeout = XS_IDLE_DISC_TO;
3008
3009         xprt->ops = &xs_tcp_ops;
3010         xprt->timeout = &xs_tcp_default_timeout;
3011
3012         xprt->max_reconnect_timeout = xprt->timeout->to_maxval;
3013         xprt->connect_timeout = xprt->timeout->to_initval *
3014                 (xprt->timeout->to_retries + 1);
3015
3016         INIT_WORK(&transport->recv_worker, xs_stream_data_receive_workfn);
3017         INIT_WORK(&transport->error_worker, xs_error_handle);
3018         INIT_DELAYED_WORK(&transport->connect_worker, xs_tcp_setup_socket);
3019
3020         switch (addr->sa_family) {
3021         case AF_INET:
3022                 if (((struct sockaddr_in *)addr)->sin_port != htons(0))
3023                         xprt_set_bound(xprt);
3024
3025                 xs_format_peer_addresses(xprt, "tcp", RPCBIND_NETID_TCP);
3026                 break;
3027         case AF_INET6:
3028                 if (((struct sockaddr_in6 *)addr)->sin6_port != htons(0))
3029                         xprt_set_bound(xprt);
3030
3031                 xs_format_peer_addresses(xprt, "tcp", RPCBIND_NETID_TCP6);
3032                 break;
3033         default:
3034                 ret = ERR_PTR(-EAFNOSUPPORT);
3035                 goto out_err;
3036         }
3037
3038         if (xprt_bound(xprt))
3039                 dprintk("RPC:       set up xprt to %s (port %s) via %s\n",
3040                                 xprt->address_strings[RPC_DISPLAY_ADDR],
3041                                 xprt->address_strings[RPC_DISPLAY_PORT],
3042                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
3043         else
3044                 dprintk("RPC:       set up xprt to %s (autobind) via %s\n",
3045                                 xprt->address_strings[RPC_DISPLAY_ADDR],
3046                                 xprt->address_strings[RPC_DISPLAY_PROTO]);
3047
3048         if (try_module_get(THIS_MODULE))
3049                 return xprt;
3050         ret = ERR_PTR(-EINVAL);
3051 out_err:
3052         xs_xprt_free(xprt);
3053         return ret;
3054 }
3055
3056 /**
3057  * xs_setup_bc_tcp - Set up transport to use a TCP backchannel socket
3058  * @args: rpc transport creation arguments
3059  *
3060  */
3061 static struct rpc_xprt *xs_setup_bc_tcp(struct xprt_create *args)
3062 {
3063         struct sockaddr *addr = args->dstaddr;
3064         struct rpc_xprt *xprt;
3065         struct sock_xprt *transport;
3066         struct svc_sock *bc_sock;
3067         struct rpc_xprt *ret;
3068
3069         xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
3070                         xprt_tcp_slot_table_entries);
3071         if (IS_ERR(xprt))
3072                 return xprt;
3073         transport = container_of(xprt, struct sock_xprt, xprt);
3074
3075         xprt->prot = IPPROTO_TCP;
3076         xprt->xprt_class = &xs_bc_tcp_transport;
3077         xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
3078         xprt->timeout = &xs_tcp_default_timeout;
3079
3080         /* backchannel */
3081         xprt_set_bound(xprt);
3082         xprt->bind_timeout = 0;
3083         xprt->reestablish_timeout = 0;
3084         xprt->idle_timeout = 0;
3085
3086         xprt->ops = &bc_tcp_ops;
3087
3088         switch (addr->sa_family) {
3089         case AF_INET:
3090                 xs_format_peer_addresses(xprt, "tcp",
3091                                          RPCBIND_NETID_TCP);
3092                 break;
3093         case AF_INET6:
3094                 xs_format_peer_addresses(xprt, "tcp",
3095                                    RPCBIND_NETID_TCP6);
3096                 break;
3097         default:
3098                 ret = ERR_PTR(-EAFNOSUPPORT);
3099                 goto out_err;
3100         }
3101
3102         dprintk("RPC:       set up xprt to %s (port %s) via %s\n",
3103                         xprt->address_strings[RPC_DISPLAY_ADDR],
3104                         xprt->address_strings[RPC_DISPLAY_PORT],
3105                         xprt->address_strings[RPC_DISPLAY_PROTO]);
3106
3107         /*
3108          * Once we've associated a backchannel xprt with a connection,
3109          * we want to keep it around as long as the connection lasts,
3110          * in case we need to start using it for a backchannel again;
3111          * this reference won't be dropped until bc_xprt is destroyed.
3112          */
3113         xprt_get(xprt);
3114         args->bc_xprt->xpt_bc_xprt = xprt;
3115         xprt->bc_xprt = args->bc_xprt;
3116         bc_sock = container_of(args->bc_xprt, struct svc_sock, sk_xprt);
3117         transport->sock = bc_sock->sk_sock;
3118         transport->inet = bc_sock->sk_sk;
3119
3120         /*
3121          * Since we don't want connections for the backchannel, we set
3122          * the xprt status to connected
3123          */
3124         xprt_set_connected(xprt);
3125
3126         if (try_module_get(THIS_MODULE))
3127                 return xprt;
3128
3129         args->bc_xprt->xpt_bc_xprt = NULL;
3130         args->bc_xprt->xpt_bc_xps = NULL;
3131         xprt_put(xprt);
3132         ret = ERR_PTR(-EINVAL);
3133 out_err:
3134         xs_xprt_free(xprt);
3135         return ret;
3136 }
3137
3138 static struct xprt_class        xs_local_transport = {
3139         .list           = LIST_HEAD_INIT(xs_local_transport.list),
3140         .name           = "named UNIX socket",
3141         .owner          = THIS_MODULE,
3142         .ident          = XPRT_TRANSPORT_LOCAL,
3143         .setup          = xs_setup_local,
3144         .netid          = { "" },
3145 };
3146
3147 static struct xprt_class        xs_udp_transport = {
3148         .list           = LIST_HEAD_INIT(xs_udp_transport.list),
3149         .name           = "udp",
3150         .owner          = THIS_MODULE,
3151         .ident          = XPRT_TRANSPORT_UDP,
3152         .setup          = xs_setup_udp,
3153         .netid          = { "udp", "udp6", "" },
3154 };
3155
3156 static struct xprt_class        xs_tcp_transport = {
3157         .list           = LIST_HEAD_INIT(xs_tcp_transport.list),
3158         .name           = "tcp",
3159         .owner          = THIS_MODULE,
3160         .ident          = XPRT_TRANSPORT_TCP,
3161         .setup          = xs_setup_tcp,
3162         .netid          = { "tcp", "tcp6", "" },
3163 };
3164
3165 static struct xprt_class        xs_bc_tcp_transport = {
3166         .list           = LIST_HEAD_INIT(xs_bc_tcp_transport.list),
3167         .name           = "tcp NFSv4.1 backchannel",
3168         .owner          = THIS_MODULE,
3169         .ident          = XPRT_TRANSPORT_BC_TCP,
3170         .setup          = xs_setup_bc_tcp,
3171         .netid          = { "" },
3172 };
3173
3174 /**
3175  * init_socket_xprt - set up xprtsock's sysctls, register with RPC client
3176  *
3177  */
3178 int init_socket_xprt(void)
3179 {
3180         if (!sunrpc_table_header)
3181                 sunrpc_table_header = register_sysctl_table(sunrpc_table);
3182
3183         xprt_register_transport(&xs_local_transport);
3184         xprt_register_transport(&xs_udp_transport);
3185         xprt_register_transport(&xs_tcp_transport);
3186         xprt_register_transport(&xs_bc_tcp_transport);
3187
3188         return 0;
3189 }
3190
3191 /**
3192  * cleanup_socket_xprt - remove xprtsock's sysctls, unregister
3193  *
3194  */
3195 void cleanup_socket_xprt(void)
3196 {
3197         if (sunrpc_table_header) {
3198                 unregister_sysctl_table(sunrpc_table_header);
3199                 sunrpc_table_header = NULL;
3200         }
3201
3202         xprt_unregister_transport(&xs_local_transport);
3203         xprt_unregister_transport(&xs_udp_transport);
3204         xprt_unregister_transport(&xs_tcp_transport);
3205         xprt_unregister_transport(&xs_bc_tcp_transport);
3206 }
3207
3208 static int param_set_portnr(const char *val, const struct kernel_param *kp)
3209 {
3210         return param_set_uint_minmax(val, kp,
3211                         RPC_MIN_RESVPORT,
3212                         RPC_MAX_RESVPORT);
3213 }
3214
3215 static const struct kernel_param_ops param_ops_portnr = {
3216         .set = param_set_portnr,
3217         .get = param_get_uint,
3218 };
3219
3220 #define param_check_portnr(name, p) \
3221         __param_check(name, p, unsigned int);
3222
3223 module_param_named(min_resvport, xprt_min_resvport, portnr, 0644);
3224 module_param_named(max_resvport, xprt_max_resvport, portnr, 0644);
3225
3226 static int param_set_slot_table_size(const char *val,
3227                                      const struct kernel_param *kp)
3228 {
3229         return param_set_uint_minmax(val, kp,
3230                         RPC_MIN_SLOT_TABLE,
3231                         RPC_MAX_SLOT_TABLE);
3232 }
3233
3234 static const struct kernel_param_ops param_ops_slot_table_size = {
3235         .set = param_set_slot_table_size,
3236         .get = param_get_uint,
3237 };
3238
3239 #define param_check_slot_table_size(name, p) \
3240         __param_check(name, p, unsigned int);
3241
3242 static int param_set_max_slot_table_size(const char *val,
3243                                      const struct kernel_param *kp)
3244 {
3245         return param_set_uint_minmax(val, kp,
3246                         RPC_MIN_SLOT_TABLE,
3247                         RPC_MAX_SLOT_TABLE_LIMIT);
3248 }
3249
3250 static const struct kernel_param_ops param_ops_max_slot_table_size = {
3251         .set = param_set_max_slot_table_size,
3252         .get = param_get_uint,
3253 };
3254
3255 #define param_check_max_slot_table_size(name, p) \
3256         __param_check(name, p, unsigned int);
3257
3258 module_param_named(tcp_slot_table_entries, xprt_tcp_slot_table_entries,
3259                    slot_table_size, 0644);
3260 module_param_named(tcp_max_slot_table_entries, xprt_max_tcp_slot_table_entries,
3261                    max_slot_table_size, 0644);
3262 module_param_named(udp_slot_table_entries, xprt_udp_slot_table_entries,
3263                    slot_table_size, 0644);