net/ceph/messenger_v1.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 #include <linux/ceph/ceph_debug.h>
   3
   4 #include <linux/bvec.h>
   5 #include <linux/crc32c.h>
   6 #include <linux/net.h>
   7 #include <linux/socket.h>
   8 #include <net/sock.h>
   9
  10 #include <linux/ceph/ceph_features.h>
  11 #include <linux/ceph/decode.h>
  12 #include <linux/ceph/libceph.h>
  13 #include <linux/ceph/messenger.h>
  14
  15 /* static tag bytes (protocol control messages) */
  16 static char tag_msg = CEPH_MSGR_TAG_MSG;
  17 static char tag_ack = CEPH_MSGR_TAG_ACK;
  18 static char tag_keepalive = CEPH_MSGR_TAG_KEEPALIVE;
  19 static char tag_keepalive2 = CEPH_MSGR_TAG_KEEPALIVE2;
  20
  21 /*
  22  * If @buf is NULL, discard up to @len bytes.
  23  */
  24 static int ceph_tcp_recvmsg(struct socket *sock, void *buf, size_t len)
  25 {
  26         struct kvec iov = {buf, len};
  27         struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL };
  28         int r;
  29
  30         if (!buf)
  31                 msg.msg_flags |= MSG_TRUNC;
  32
  33         iov_iter_kvec(&msg.msg_iter, ITER_DEST, &iov, 1, len);
  34         r = sock_recvmsg(sock, &msg, msg.msg_flags);
  35         if (r == -EAGAIN)
  36                 r = 0;
  37         return r;
  38 }
  39
  40 static int ceph_tcp_recvpage(struct socket *sock, struct page *page,
  41                      int page_offset, size_t length)
  42 {
  43         struct bio_vec bvec;
  44         struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL };
  45         int r;
  46
  47         BUG_ON(page_offset + length > PAGE_SIZE);
  48         bvec_set_page(&bvec, page, length, page_offset);
  49         iov_iter_bvec(&msg.msg_iter, ITER_DEST, &bvec, 1, length);
  50         r = sock_recvmsg(sock, &msg, msg.msg_flags);
  51         if (r == -EAGAIN)
  52                 r = 0;
  53         return r;
  54 }
  55
  56 /*
  57  * write something.  @more is true if caller will be sending more data
  58  * shortly.
  59  */
  60 static int ceph_tcp_sendmsg(struct socket *sock, struct kvec *iov,
  61                             size_t kvlen, size_t len, bool more)
  62 {
  63         struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL };
  64         int r;
  65
  66         if (more)
  67                 msg.msg_flags |= MSG_MORE;
  68         else
  69                 msg.msg_flags |= MSG_EOR;  /* superfluous, but what the hell */
  70
  71         r = kernel_sendmsg(sock, &msg, iov, kvlen, len);
  72         if (r == -EAGAIN)
  73                 r = 0;
  74         return r;
  75 }
  76
  77 /*
  78  * @more: MSG_MORE or 0.
  79  */
  80 static int ceph_tcp_sendpage(struct socket *sock, struct page *page,
  81                              int offset, size_t size, int more)
  82 {
  83         struct msghdr msg = {
  84                 .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL | more,
  85         };
  86         struct bio_vec bvec;
  87         int ret;
  88
  89         /*
  90          * MSG_SPLICE_PAGES cannot properly handle pages with page_count == 0,
  91          * we need to fall back to sendmsg if that's the case.
  92          *
  93          * Same goes for slab pages: skb_can_coalesce() allows
  94          * coalescing neighboring slab objects into a single frag which
  95          * triggers one of hardened usercopy checks.
  96          */
  97         if (sendpage_ok(page))
  98                 msg.msg_flags |= MSG_SPLICE_PAGES;
  99
 100         bvec_set_page(&bvec, page, size, offset);
 101         iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, &bvec, 1, size);
 102
 103         ret = sock_sendmsg(sock, &msg);
 104         if (ret == -EAGAIN)
 105                 ret = 0;
 106
 107         return ret;
 108 }
 109
 110 static void con_out_kvec_reset(struct ceph_connection *con)
 111 {
 112         BUG_ON(con->v1.out_skip);
 113
 114         con->v1.out_kvec_left = 0;
 115         con->v1.out_kvec_bytes = 0;
 116         con->v1.out_kvec_cur = &con->v1.out_kvec[0];
 117 }
 118
 119 static void con_out_kvec_add(struct ceph_connection *con,
 120                                 size_t size, void *data)
 121 {
 122         int index = con->v1.out_kvec_left;
 123
 124         BUG_ON(con->v1.out_skip);
 125         BUG_ON(index >= ARRAY_SIZE(con->v1.out_kvec));
 126
 127         con->v1.out_kvec[index].iov_len = size;
 128         con->v1.out_kvec[index].iov_base = data;
 129         con->v1.out_kvec_left++;
 130         con->v1.out_kvec_bytes += size;
 131 }
 132
 133 /*
 134  * Chop off a kvec from the end.  Return residual number of bytes for
 135  * that kvec, i.e. how many bytes would have been written if the kvec
 136  * hadn't been nuked.
 137  */
 138 static int con_out_kvec_skip(struct ceph_connection *con)
 139 {
 140         int skip = 0;
 141
 142         if (con->v1.out_kvec_bytes > 0) {
 143                 skip = con->v1.out_kvec_cur[con->v1.out_kvec_left - 1].iov_len;
 144                 BUG_ON(con->v1.out_kvec_bytes < skip);
 145                 BUG_ON(!con->v1.out_kvec_left);
 146                 con->v1.out_kvec_bytes -= skip;
 147                 con->v1.out_kvec_left--;
 148         }
 149
 150         return skip;
 151 }
 152
 153 static size_t sizeof_footer(struct ceph_connection *con)
 154 {
 155         return (con->peer_features & CEPH_FEATURE_MSG_AUTH) ?
 156             sizeof(struct ceph_msg_footer) :
 157             sizeof(struct ceph_msg_footer_old);
 158 }
 159
 160 static void prepare_message_data(struct ceph_msg *msg, u32 data_len)
 161 {
 162         /* Initialize data cursor if it's not a sparse read */
 163         u64 len = msg->sparse_read_total ? : data_len;
 164
 165         ceph_msg_data_cursor_init(&msg->cursor, msg, len);
 166 }
 167
 168 /*
 169  * Prepare footer for currently outgoing message, and finish things
 170  * off.  Assumes out_kvec* are already valid.. we just add on to the end.
 171  */
 172 static void prepare_write_message_footer(struct ceph_connection *con)
 173 {
 174         struct ceph_msg *m = con->out_msg;
 175
 176         m->footer.flags |= CEPH_MSG_FOOTER_COMPLETE;
 177
 178         dout("prepare_write_message_footer %p\n", con);
 179         con_out_kvec_add(con, sizeof_footer(con), &m->footer);
 180         if (con->peer_features & CEPH_FEATURE_MSG_AUTH) {
 181                 if (con->ops->sign_message)
 182                         con->ops->sign_message(m);
 183                 else
 184                         m->footer.sig = 0;
 185         } else {
 186                 m->old_footer.flags = m->footer.flags;
 187         }
 188         con->v1.out_more = m->more_to_follow;
 189         con->v1.out_msg_done = true;
 190 }
 191
 192 /*
 193  * Prepare headers for the next outgoing message.
 194  */
 195 static void prepare_write_message(struct ceph_connection *con)
 196 {
 197         struct ceph_msg *m;
 198         u32 crc;
 199
 200         con_out_kvec_reset(con);
 201         con->v1.out_msg_done = false;
 202
 203         /* Sneak an ack in there first?  If we can get it into the same
 204          * TCP packet that's a good thing. */
 205         if (con->in_seq > con->in_seq_acked) {
 206                 con->in_seq_acked = con->in_seq;
 207                 con_out_kvec_add(con, sizeof (tag_ack), &tag_ack);
 208                 con->v1.out_temp_ack = cpu_to_le64(con->in_seq_acked);
 209                 con_out_kvec_add(con, sizeof(con->v1.out_temp_ack),
 210                         &con->v1.out_temp_ack);
 211         }
 212
 213         ceph_con_get_out_msg(con);
 214         m = con->out_msg;
 215
 216         dout("prepare_write_message %p seq %lld type %d len %d+%d+%zd\n",
 217              m, con->out_seq, le16_to_cpu(m->hdr.type),
 218              le32_to_cpu(m->hdr.front_len), le32_to_cpu(m->hdr.middle_len),
 219              m->data_length);
 220         WARN_ON(m->front.iov_len != le32_to_cpu(m->hdr.front_len));
 221         WARN_ON(m->data_length != le32_to_cpu(m->hdr.data_len));
 222
 223         /* tag + hdr + front + middle */
 224         con_out_kvec_add(con, sizeof (tag_msg), &tag_msg);
 225         con_out_kvec_add(con, sizeof(con->v1.out_hdr), &con->v1.out_hdr);
 226         con_out_kvec_add(con, m->front.iov_len, m->front.iov_base);
 227
 228         if (m->middle)
 229                 con_out_kvec_add(con, m->middle->vec.iov_len,
 230                         m->middle->vec.iov_base);
 231
 232         /* fill in hdr crc and finalize hdr */
 233         crc = crc32c(0, &m->hdr, offsetof(struct ceph_msg_header, crc));
 234         con->out_msg->hdr.crc = cpu_to_le32(crc);
 235         memcpy(&con->v1.out_hdr, &con->out_msg->hdr, sizeof(con->v1.out_hdr));
 236
 237         /* fill in front and middle crc, footer */
 238         crc = crc32c(0, m->front.iov_base, m->front.iov_len);
 239         con->out_msg->footer.front_crc = cpu_to_le32(crc);
 240         if (m->middle) {
 241                 crc = crc32c(0, m->middle->vec.iov_base,
 242                                 m->middle->vec.iov_len);
 243                 con->out_msg->footer.middle_crc = cpu_to_le32(crc);
 244         } else
 245                 con->out_msg->footer.middle_crc = 0;
 246         dout("%s front_crc %u middle_crc %u\n", __func__,
 247              le32_to_cpu(con->out_msg->footer.front_crc),
 248              le32_to_cpu(con->out_msg->footer.middle_crc));
 249         con->out_msg->footer.flags = 0;
 250
 251         /* is there a data payload? */
 252         con->out_msg->footer.data_crc = 0;
 253         if (m->data_length) {
 254                 prepare_message_data(con->out_msg, m->data_length);
 255                 con->v1.out_more = 1;  /* data + footer will follow */
 256         } else {
 257                 /* no, queue up footer too and be done */
 258                 prepare_write_message_footer(con);
 259         }
 260
 261         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
 262 }
 263
 264 /*
 265  * Prepare an ack.
 266  */
 267 static void prepare_write_ack(struct ceph_connection *con)
 268 {
 269         dout("prepare_write_ack %p %llu -> %llu\n", con,
 270              con->in_seq_acked, con->in_seq);
 271         con->in_seq_acked = con->in_seq;
 272
 273         con_out_kvec_reset(con);
 274
 275         con_out_kvec_add(con, sizeof (tag_ack), &tag_ack);
 276
 277         con->v1.out_temp_ack = cpu_to_le64(con->in_seq_acked);
 278         con_out_kvec_add(con, sizeof(con->v1.out_temp_ack),
 279                          &con->v1.out_temp_ack);
 280
 281         con->v1.out_more = 1;  /* more will follow.. eventually.. */
 282         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
 283 }
 284
 285 /*
 286  * Prepare to share the seq during handshake
 287  */
 288 static void prepare_write_seq(struct ceph_connection *con)
 289 {
 290         dout("prepare_write_seq %p %llu -> %llu\n", con,
 291              con->in_seq_acked, con->in_seq);
 292         con->in_seq_acked = con->in_seq;
 293
 294         con_out_kvec_reset(con);
 295
 296         con->v1.out_temp_ack = cpu_to_le64(con->in_seq_acked);
 297         con_out_kvec_add(con, sizeof(con->v1.out_temp_ack),
 298                          &con->v1.out_temp_ack);
 299
 300         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
 301 }
 302
 303 /*
 304  * Prepare to write keepalive byte.
 305  */
 306 static void prepare_write_keepalive(struct ceph_connection *con)
 307 {
 308         dout("prepare_write_keepalive %p\n", con);
 309         con_out_kvec_reset(con);
 310         if (con->peer_features & CEPH_FEATURE_MSGR_KEEPALIVE2) {
 311                 struct timespec64 now;
 312
 313                 ktime_get_real_ts64(&now);
 314                 con_out_kvec_add(con, sizeof(tag_keepalive2), &tag_keepalive2);
 315                 ceph_encode_timespec64(&con->v1.out_temp_keepalive2, &now);
 316                 con_out_kvec_add(con, sizeof(con->v1.out_temp_keepalive2),
 317                                  &con->v1.out_temp_keepalive2);
 318         } else {
 319                 con_out_kvec_add(con, sizeof(tag_keepalive), &tag_keepalive);
 320         }
 321         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
 322 }
 323
 324 /*
 325  * Connection negotiation.
 326  */
 327
 328 static int get_connect_authorizer(struct ceph_connection *con)
 329 {
 330         struct ceph_auth_handshake *auth;
 331         int auth_proto;
 332
 333         if (!con->ops->get_authorizer) {
 334                 con->v1.auth = NULL;
 335                 con->v1.out_connect.authorizer_protocol = CEPH_AUTH_UNKNOWN;
 336                 con->v1.out_connect.authorizer_len = 0;
 337                 return 0;
 338         }
 339
 340         auth = con->ops->get_authorizer(con, &auth_proto, con->v1.auth_retry);
 341         if (IS_ERR(auth))
 342                 return PTR_ERR(auth);
 343
 344         con->v1.auth = auth;
 345         con->v1.out_connect.authorizer_protocol = cpu_to_le32(auth_proto);
 346         con->v1.out_connect.authorizer_len =
 347                 cpu_to_le32(auth->authorizer_buf_len);
 348         return 0;
 349 }
 350
 351 /*
 352  * We connected to a peer and are saying hello.
 353  */
 354 static void prepare_write_banner(struct ceph_connection *con)
 355 {
 356         con_out_kvec_add(con, strlen(CEPH_BANNER), CEPH_BANNER);
 357         con_out_kvec_add(con, sizeof (con->msgr->my_enc_addr),
 358                                         &con->msgr->my_enc_addr);
 359
 360         con->v1.out_more = 0;
 361         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
 362 }
 363
 364 static void __prepare_write_connect(struct ceph_connection *con)
 365 {
 366         con_out_kvec_add(con, sizeof(con->v1.out_connect),
 367                          &con->v1.out_connect);
 368         if (con->v1.auth)
 369                 con_out_kvec_add(con, con->v1.auth->authorizer_buf_len,
 370                                  con->v1.auth->authorizer_buf);
 371
 372         con->v1.out_more = 0;
 373         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
 374 }
 375
 376 static int prepare_write_connect(struct ceph_connection *con)
 377 {
 378         unsigned int global_seq = ceph_get_global_seq(con->msgr, 0);
 379         int proto;
 380         int ret;
 381
 382         switch (con->peer_name.type) {
 383         case CEPH_ENTITY_TYPE_MON:
 384                 proto = CEPH_MONC_PROTOCOL;
 385                 break;
 386         case CEPH_ENTITY_TYPE_OSD:
 387                 proto = CEPH_OSDC_PROTOCOL;
 388                 break;
 389         case CEPH_ENTITY_TYPE_MDS:
 390                 proto = CEPH_MDSC_PROTOCOL;
 391                 break;
 392         default:
 393                 BUG();
 394         }
 395
 396         dout("prepare_write_connect %p cseq=%d gseq=%d proto=%d\n", con,
 397              con->v1.connect_seq, global_seq, proto);
 398
 399         con->v1.out_connect.features =
 400                 cpu_to_le64(from_msgr(con->msgr)->supported_features);
 401         con->v1.out_connect.host_type = cpu_to_le32(CEPH_ENTITY_TYPE_CLIENT);
 402         con->v1.out_connect.connect_seq = cpu_to_le32(con->v1.connect_seq);
 403         con->v1.out_connect.global_seq = cpu_to_le32(global_seq);
 404         con->v1.out_connect.protocol_version = cpu_to_le32(proto);
 405         con->v1.out_connect.flags = 0;
 406
 407         ret = get_connect_authorizer(con);
 408         if (ret)
 409                 return ret;
 410
 411         __prepare_write_connect(con);
 412         return 0;
 413 }
 414
 415 /*
 416  * write as much of pending kvecs to the socket as we can.
 417  *  1 -> done
 418  *  0 -> socket full, but more to do
 419  * <0 -> error
 420  */
 421 static int write_partial_kvec(struct ceph_connection *con)
 422 {
 423         int ret;
 424
 425         dout("write_partial_kvec %p %d left\n", con, con->v1.out_kvec_bytes);
 426         while (con->v1.out_kvec_bytes > 0) {
 427                 ret = ceph_tcp_sendmsg(con->sock, con->v1.out_kvec_cur,
 428                                        con->v1.out_kvec_left,
 429                                        con->v1.out_kvec_bytes,
 430                                        con->v1.out_more);
 431                 if (ret <= 0)
 432                         goto out;
 433                 con->v1.out_kvec_bytes -= ret;
 434                 if (!con->v1.out_kvec_bytes)
 435                         break;            /* done */
 436
 437                 /* account for full iov entries consumed */
 438                 while (ret >= con->v1.out_kvec_cur->iov_len) {
 439                         BUG_ON(!con->v1.out_kvec_left);
 440                         ret -= con->v1.out_kvec_cur->iov_len;
 441                         con->v1.out_kvec_cur++;
 442                         con->v1.out_kvec_left--;
 443                 }
 444                 /* and for a partially-consumed entry */
 445                 if (ret) {
 446                         con->v1.out_kvec_cur->iov_len -= ret;
 447                         con->v1.out_kvec_cur->iov_base += ret;
 448                 }
 449         }
 450         con->v1.out_kvec_left = 0;
 451         ret = 1;
 452 out:
 453         dout("write_partial_kvec %p %d left in %d kvecs ret = %d\n", con,
 454              con->v1.out_kvec_bytes, con->v1.out_kvec_left, ret);
 455         return ret;  /* done! */
 456 }
 457
 458 /*
 459  * Write as much message data payload as we can.  If we finish, queue
 460  * up the footer.
 461  *  1 -> done, footer is now queued in out_kvec[].
 462  *  0 -> socket full, but more to do
 463  * <0 -> error
 464  */
 465 static int write_partial_message_data(struct ceph_connection *con)
 466 {
 467         struct ceph_msg *msg = con->out_msg;
 468         struct ceph_msg_data_cursor *cursor = &msg->cursor;
 469         bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
 470         u32 crc;
 471
 472         dout("%s %p msg %p\n", __func__, con, msg);
 473
 474         if (!msg->num_data_items)
 475                 return -EINVAL;
 476
 477         /*
 478          * Iterate through each page that contains data to be
 479          * written, and send as much as possible for each.
 480          *
 481          * If we are calculating the data crc (the default), we will
 482          * need to map the page.  If we have no pages, they have
 483          * been revoked, so use the zero page.
 484          */
 485         crc = do_datacrc ? le32_to_cpu(msg->footer.data_crc) : 0;
 486         while (cursor->total_resid) {
 487                 struct page *page;
 488                 size_t page_offset;
 489                 size_t length;
 490                 int ret;
 491
 492                 if (!cursor->resid) {
 493                         ceph_msg_data_advance(cursor, 0);
 494                         continue;
 495                 }
 496
 497                 page = ceph_msg_data_next(cursor, &page_offset, &length);
 498                 ret = ceph_tcp_sendpage(con->sock, page, page_offset, length,
 499                                         MSG_MORE);
 500                 if (ret <= 0) {
 501                         if (do_datacrc)
 502                                 msg->footer.data_crc = cpu_to_le32(crc);
 503
 504                         return ret;
 505                 }
 506                 if (do_datacrc && cursor->need_crc)
 507                         crc = ceph_crc32c_page(crc, page, page_offset, length);
 508                 ceph_msg_data_advance(cursor, (size_t)ret);
 509         }
 510
 511         dout("%s %p msg %p done\n", __func__, con, msg);
 512
 513         /* prepare and queue up footer, too */
 514         if (do_datacrc)
 515                 msg->footer.data_crc = cpu_to_le32(crc);
 516         else
 517                 msg->footer.flags |= CEPH_MSG_FOOTER_NOCRC;
 518         con_out_kvec_reset(con);
 519         prepare_write_message_footer(con);
 520
 521         return 1;       /* must return > 0 to indicate success */
 522 }
 523
 524 /*
 525  * write some zeros
 526  */
 527 static int write_partial_skip(struct ceph_connection *con)
 528 {
 529         int ret;
 530
 531         dout("%s %p %d left\n", __func__, con, con->v1.out_skip);
 532         while (con->v1.out_skip > 0) {
 533                 size_t size = min(con->v1.out_skip, (int)PAGE_SIZE);
 534
 535                 ret = ceph_tcp_sendpage(con->sock, ceph_zero_page, 0, size,
 536                                         MSG_MORE);
 537                 if (ret <= 0)
 538                         goto out;
 539                 con->v1.out_skip -= ret;
 540         }
 541         ret = 1;
 542 out:
 543         return ret;
 544 }
 545
 546 /*
 547  * Prepare to read connection handshake, or an ack.
 548  */
 549 static void prepare_read_banner(struct ceph_connection *con)
 550 {
 551         dout("prepare_read_banner %p\n", con);
 552         con->v1.in_base_pos = 0;
 553 }
 554
 555 static void prepare_read_connect(struct ceph_connection *con)
 556 {
 557         dout("prepare_read_connect %p\n", con);
 558         con->v1.in_base_pos = 0;
 559 }
 560
 561 static void prepare_read_ack(struct ceph_connection *con)
 562 {
 563         dout("prepare_read_ack %p\n", con);
 564         con->v1.in_base_pos = 0;
 565 }
 566
 567 static void prepare_read_seq(struct ceph_connection *con)
 568 {
 569         dout("prepare_read_seq %p\n", con);
 570         con->v1.in_base_pos = 0;
 571         con->v1.in_tag = CEPH_MSGR_TAG_SEQ;
 572 }
 573
 574 static void prepare_read_tag(struct ceph_connection *con)
 575 {
 576         dout("prepare_read_tag %p\n", con);
 577         con->v1.in_base_pos = 0;
 578         con->v1.in_tag = CEPH_MSGR_TAG_READY;
 579 }
 580
 581 static void prepare_read_keepalive_ack(struct ceph_connection *con)
 582 {
 583         dout("prepare_read_keepalive_ack %p\n", con);
 584         con->v1.in_base_pos = 0;
 585 }
 586
 587 /*
 588  * Prepare to read a message.
 589  */
 590 static int prepare_read_message(struct ceph_connection *con)
 591 {
 592         dout("prepare_read_message %p\n", con);
 593         BUG_ON(con->in_msg != NULL);
 594         con->v1.in_base_pos = 0;
 595         con->in_front_crc = con->in_middle_crc = con->in_data_crc = 0;
 596         return 0;
 597 }
 598
 599 static int read_partial(struct ceph_connection *con,
 600                         int end, int size, void *object)
 601 {
 602         while (con->v1.in_base_pos < end) {
 603                 int left = end - con->v1.in_base_pos;
 604                 int have = size - left;
 605                 int ret = ceph_tcp_recvmsg(con->sock, object + have, left);
 606                 if (ret <= 0)
 607                         return ret;
 608                 con->v1.in_base_pos += ret;
 609         }
 610         return 1;
 611 }
 612
 613 /*
 614  * Read all or part of the connect-side handshake on a new connection
 615  */
 616 static int read_partial_banner(struct ceph_connection *con)
 617 {
 618         int size;
 619         int end;
 620         int ret;
 621
 622         dout("read_partial_banner %p at %d\n", con, con->v1.in_base_pos);
 623
 624         /* peer's banner */
 625         size = strlen(CEPH_BANNER);
 626         end = size;
 627         ret = read_partial(con, end, size, con->v1.in_banner);
 628         if (ret <= 0)
 629                 goto out;
 630
 631         size = sizeof(con->v1.actual_peer_addr);
 632         end += size;
 633         ret = read_partial(con, end, size, &con->v1.actual_peer_addr);
 634         if (ret <= 0)
 635                 goto out;
 636         ceph_decode_banner_addr(&con->v1.actual_peer_addr);
 637
 638         size = sizeof(con->v1.peer_addr_for_me);
 639         end += size;
 640         ret = read_partial(con, end, size, &con->v1.peer_addr_for_me);
 641         if (ret <= 0)
 642                 goto out;
 643         ceph_decode_banner_addr(&con->v1.peer_addr_for_me);
 644
 645 out:
 646         return ret;
 647 }
 648
 649 static int read_partial_connect(struct ceph_connection *con)
 650 {
 651         int size;
 652         int end;
 653         int ret;
 654
 655         dout("read_partial_connect %p at %d\n", con, con->v1.in_base_pos);
 656
 657         size = sizeof(con->v1.in_reply);
 658         end = size;
 659         ret = read_partial(con, end, size, &con->v1.in_reply);
 660         if (ret <= 0)
 661                 goto out;
 662
 663         if (con->v1.auth) {
 664                 size = le32_to_cpu(con->v1.in_reply.authorizer_len);
 665                 if (size > con->v1.auth->authorizer_reply_buf_len) {
 666                         pr_err("authorizer reply too big: %d > %zu\n", size,
 667                                con->v1.auth->authorizer_reply_buf_len);
 668                         ret = -EINVAL;
 669                         goto out;
 670                 }
 671
 672                 end += size;
 673                 ret = read_partial(con, end, size,
 674                                    con->v1.auth->authorizer_reply_buf);
 675                 if (ret <= 0)
 676                         goto out;
 677         }
 678
 679         dout("read_partial_connect %p tag %d, con_seq = %u, g_seq = %u\n",
 680              con, con->v1.in_reply.tag,
 681              le32_to_cpu(con->v1.in_reply.connect_seq),
 682              le32_to_cpu(con->v1.in_reply.global_seq));
 683 out:
 684         return ret;
 685 }
 686
 687 /*
 688  * Verify the hello banner looks okay.
 689  */
 690 static int verify_hello(struct ceph_connection *con)
 691 {
 692         if (memcmp(con->v1.in_banner, CEPH_BANNER, strlen(CEPH_BANNER))) {
 693                 pr_err("connect to %s got bad banner\n",
 694                        ceph_pr_addr(&con->peer_addr));
 695                 con->error_msg = "protocol error, bad banner";
 696                 return -1;
 697         }
 698         return 0;
 699 }
 700
 701 static int process_banner(struct ceph_connection *con)
 702 {
 703         struct ceph_entity_addr *my_addr = &con->msgr->inst.addr;
 704
 705         dout("process_banner on %p\n", con);
 706
 707         if (verify_hello(con) < 0)
 708                 return -1;
 709
 710         /*
 711          * Make sure the other end is who we wanted.  note that the other
 712          * end may not yet know their ip address, so if it's 0.0.0.0, give
 713          * them the benefit of the doubt.
 714          */
 715         if (memcmp(&con->peer_addr, &con->v1.actual_peer_addr,
 716                    sizeof(con->peer_addr)) != 0 &&
 717             !(ceph_addr_is_blank(&con->v1.actual_peer_addr) &&
 718               con->v1.actual_peer_addr.nonce == con->peer_addr.nonce)) {
 719                 pr_warn("wrong peer, want %s/%u, got %s/%u\n",
 720                         ceph_pr_addr(&con->peer_addr),
 721                         le32_to_cpu(con->peer_addr.nonce),
 722                         ceph_pr_addr(&con->v1.actual_peer_addr),
 723                         le32_to_cpu(con->v1.actual_peer_addr.nonce));
 724                 con->error_msg = "wrong peer at address";
 725                 return -1;
 726         }
 727
 728         /*
 729          * did we learn our address?
 730          */
 731         if (ceph_addr_is_blank(my_addr)) {
 732                 memcpy(&my_addr->in_addr,
 733                        &con->v1.peer_addr_for_me.in_addr,
 734                        sizeof(con->v1.peer_addr_for_me.in_addr));
 735                 ceph_addr_set_port(my_addr, 0);
 736                 ceph_encode_my_addr(con->msgr);
 737                 dout("process_banner learned my addr is %s\n",
 738                      ceph_pr_addr(my_addr));
 739         }
 740
 741         return 0;
 742 }
 743
 744 static int process_connect(struct ceph_connection *con)
 745 {
 746         u64 sup_feat = from_msgr(con->msgr)->supported_features;
 747         u64 req_feat = from_msgr(con->msgr)->required_features;
 748         u64 server_feat = le64_to_cpu(con->v1.in_reply.features);
 749         int ret;
 750
 751         dout("process_connect on %p tag %d\n", con, con->v1.in_tag);
 752
 753         if (con->v1.auth) {
 754                 int len = le32_to_cpu(con->v1.in_reply.authorizer_len);
 755
 756                 /*
 757                  * Any connection that defines ->get_authorizer()
 758                  * should also define ->add_authorizer_challenge() and
 759                  * ->verify_authorizer_reply().
 760                  *
 761                  * See get_connect_authorizer().
 762                  */
 763                 if (con->v1.in_reply.tag ==
 764                                 CEPH_MSGR_TAG_CHALLENGE_AUTHORIZER) {
 765                         ret = con->ops->add_authorizer_challenge(
 766                                 con, con->v1.auth->authorizer_reply_buf, len);
 767                         if (ret < 0)
 768                                 return ret;
 769
 770                         con_out_kvec_reset(con);
 771                         __prepare_write_connect(con);
 772                         prepare_read_connect(con);
 773                         return 0;
 774                 }
 775
 776                 if (len) {
 777                         ret = con->ops->verify_authorizer_reply(con);
 778                         if (ret < 0) {
 779                                 con->error_msg = "bad authorize reply";
 780                                 return ret;
 781                         }
 782                 }
 783         }
 784
 785         switch (con->v1.in_reply.tag) {
 786         case CEPH_MSGR_TAG_FEATURES:
 787                 pr_err("%s%lld %s feature set mismatch,"
 788                        " my %llx < server's %llx, missing %llx\n",
 789                        ENTITY_NAME(con->peer_name),
 790                        ceph_pr_addr(&con->peer_addr),
 791                        sup_feat, server_feat, server_feat & ~sup_feat);
 792                 con->error_msg = "missing required protocol features";
 793                 return -1;
 794
 795         case CEPH_MSGR_TAG_BADPROTOVER:
 796                 pr_err("%s%lld %s protocol version mismatch,"
 797                        " my %d != server's %d\n",
 798                        ENTITY_NAME(con->peer_name),
 799                        ceph_pr_addr(&con->peer_addr),
 800                        le32_to_cpu(con->v1.out_connect.protocol_version),
 801                        le32_to_cpu(con->v1.in_reply.protocol_version));
 802                 con->error_msg = "protocol version mismatch";
 803                 return -1;
 804
 805         case CEPH_MSGR_TAG_BADAUTHORIZER:
 806                 con->v1.auth_retry++;
 807                 dout("process_connect %p got BADAUTHORIZER attempt %d\n", con,
 808                      con->v1.auth_retry);
 809                 if (con->v1.auth_retry == 2) {
 810                         con->error_msg = "connect authorization failure";
 811                         return -1;
 812                 }
 813                 con_out_kvec_reset(con);
 814                 ret = prepare_write_connect(con);
 815                 if (ret < 0)
 816                         return ret;
 817                 prepare_read_connect(con);
 818                 break;
 819
 820         case CEPH_MSGR_TAG_RESETSESSION:
 821                 /*
 822                  * If we connected with a large connect_seq but the peer
 823                  * has no record of a session with us (no connection, or
 824                  * connect_seq == 0), they will send RESETSESION to indicate
 825                  * that they must have reset their session, and may have
 826                  * dropped messages.
 827                  */
 828                 dout("process_connect got RESET peer seq %u\n",
 829                      le32_to_cpu(con->v1.in_reply.connect_seq));
 830                 pr_info("%s%lld %s session reset\n",
 831                         ENTITY_NAME(con->peer_name),
 832                         ceph_pr_addr(&con->peer_addr));
 833                 ceph_con_reset_session(con);
 834                 con_out_kvec_reset(con);
 835                 ret = prepare_write_connect(con);
 836                 if (ret < 0)
 837                         return ret;
 838                 prepare_read_connect(con);
 839
 840                 /* Tell ceph about it. */
 841                 mutex_unlock(&con->mutex);
 842                 if (con->ops->peer_reset)
 843                         con->ops->peer_reset(con);
 844                 mutex_lock(&con->mutex);
 845                 if (con->state != CEPH_CON_S_V1_CONNECT_MSG)
 846                         return -EAGAIN;
 847                 break;
 848
 849         case CEPH_MSGR_TAG_RETRY_SESSION:
 850                 /*
 851                  * If we sent a smaller connect_seq than the peer has, try
 852                  * again with a larger value.
 853                  */
 854                 dout("process_connect got RETRY_SESSION my seq %u, peer %u\n",
 855                      le32_to_cpu(con->v1.out_connect.connect_seq),
 856                      le32_to_cpu(con->v1.in_reply.connect_seq));
 857                 con->v1.connect_seq = le32_to_cpu(con->v1.in_reply.connect_seq);
 858                 con_out_kvec_reset(con);
 859                 ret = prepare_write_connect(con);
 860                 if (ret < 0)
 861                         return ret;
 862                 prepare_read_connect(con);
 863                 break;
 864
 865         case CEPH_MSGR_TAG_RETRY_GLOBAL:
 866                 /*
 867                  * If we sent a smaller global_seq than the peer has, try
 868                  * again with a larger value.
 869                  */
 870                 dout("process_connect got RETRY_GLOBAL my %u peer_gseq %u\n",
 871                      con->v1.peer_global_seq,
 872                      le32_to_cpu(con->v1.in_reply.global_seq));
 873                 ceph_get_global_seq(con->msgr,
 874                                     le32_to_cpu(con->v1.in_reply.global_seq));
 875                 con_out_kvec_reset(con);
 876                 ret = prepare_write_connect(con);
 877                 if (ret < 0)
 878                         return ret;
 879                 prepare_read_connect(con);
 880                 break;
 881
 882         case CEPH_MSGR_TAG_SEQ:
 883         case CEPH_MSGR_TAG_READY:
 884                 if (req_feat & ~server_feat) {
 885                         pr_err("%s%lld %s protocol feature mismatch,"
 886                                " my required %llx > server's %llx, need %llx\n",
 887                                ENTITY_NAME(con->peer_name),
 888                                ceph_pr_addr(&con->peer_addr),
 889                                req_feat, server_feat, req_feat & ~server_feat);
 890                         con->error_msg = "missing required protocol features";
 891                         return -1;
 892                 }
 893
 894                 WARN_ON(con->state != CEPH_CON_S_V1_CONNECT_MSG);
 895                 con->state = CEPH_CON_S_OPEN;
 896                 con->v1.auth_retry = 0;    /* we authenticated; clear flag */
 897                 con->v1.peer_global_seq =
 898                         le32_to_cpu(con->v1.in_reply.global_seq);
 899                 con->v1.connect_seq++;
 900                 con->peer_features = server_feat;
 901                 dout("process_connect got READY gseq %d cseq %d (%d)\n",
 902                      con->v1.peer_global_seq,
 903                      le32_to_cpu(con->v1.in_reply.connect_seq),
 904                      con->v1.connect_seq);
 905                 WARN_ON(con->v1.connect_seq !=
 906                         le32_to_cpu(con->v1.in_reply.connect_seq));
 907
 908                 if (con->v1.in_reply.flags & CEPH_MSG_CONNECT_LOSSY)
 909                         ceph_con_flag_set(con, CEPH_CON_F_LOSSYTX);
 910
 911                 con->delay = 0;      /* reset backoff memory */
 912
 913                 if (con->v1.in_reply.tag == CEPH_MSGR_TAG_SEQ) {
 914                         prepare_write_seq(con);
 915                         prepare_read_seq(con);
 916                 } else {
 917                         prepare_read_tag(con);
 918                 }
 919                 break;
 920
 921         case CEPH_MSGR_TAG_WAIT:
 922                 /*
 923                  * If there is a connection race (we are opening
 924                  * connections to each other), one of us may just have
 925                  * to WAIT.  This shouldn't happen if we are the
 926                  * client.
 927                  */
 928                 con->error_msg = "protocol error, got WAIT as client";
 929                 return -1;
 930
 931         default:
 932                 con->error_msg = "protocol error, garbage tag during connect";
 933                 return -1;
 934         }
 935         return 0;
 936 }
 937
 938 /*
 939  * read (part of) an ack
 940  */
 941 static int read_partial_ack(struct ceph_connection *con)
 942 {
 943         int size = sizeof(con->v1.in_temp_ack);
 944         int end = size;
 945
 946         return read_partial(con, end, size, &con->v1.in_temp_ack);
 947 }
 948
 949 /*
 950  * We can finally discard anything that's been acked.
 951  */
 952 static void process_ack(struct ceph_connection *con)
 953 {
 954         u64 ack = le64_to_cpu(con->v1.in_temp_ack);
 955
 956         if (con->v1.in_tag == CEPH_MSGR_TAG_ACK)
 957                 ceph_con_discard_sent(con, ack);
 958         else
 959                 ceph_con_discard_requeued(con, ack);
 960
 961         prepare_read_tag(con);
 962 }
 963
 964 static int read_partial_message_chunk(struct ceph_connection *con,
 965                                       struct kvec *section,
 966                                       unsigned int sec_len, u32 *crc)
 967 {
 968         int ret, left;
 969
 970         BUG_ON(!section);
 971
 972         while (section->iov_len < sec_len) {
 973                 BUG_ON(section->iov_base == NULL);
 974                 left = sec_len - section->iov_len;
 975                 ret = ceph_tcp_recvmsg(con->sock, (char *)section->iov_base +
 976                                        section->iov_len, left);
 977                 if (ret <= 0)
 978                         return ret;
 979                 section->iov_len += ret;
 980         }
 981         if (section->iov_len == sec_len)
 982                 *crc = crc32c(*crc, section->iov_base, section->iov_len);
 983
 984         return 1;
 985 }
 986
 987 static inline int read_partial_message_section(struct ceph_connection *con,
 988                                                struct kvec *section,
 989                                                unsigned int sec_len, u32 *crc)
 990 {
 991         *crc = 0;
 992         return read_partial_message_chunk(con, section, sec_len, crc);
 993 }
 994
 995 static int read_partial_sparse_msg_extent(struct ceph_connection *con, u32 *crc)
 996 {
 997         struct ceph_msg_data_cursor *cursor = &con->in_msg->cursor;
 998         bool do_bounce = ceph_test_opt(from_msgr(con->msgr), RXBOUNCE);
 999
1000         if (do_bounce && unlikely(!con->bounce_page)) {
1001                 con->bounce_page = alloc_page(GFP_NOIO);
1002                 if (!con->bounce_page) {
1003                         pr_err("failed to allocate bounce page\n");
1004                         return -ENOMEM;
1005                 }
1006         }
1007
1008         while (cursor->sr_resid > 0) {
1009                 struct page *page, *rpage;
1010                 size_t off, len;
1011                 int ret;
1012
1013                 page = ceph_msg_data_next(cursor, &off, &len);
1014                 rpage = do_bounce ? con->bounce_page : page;
1015
1016                 /* clamp to what remains in extent */
1017                 len = min_t(int, len, cursor->sr_resid);
1018                 ret = ceph_tcp_recvpage(con->sock, rpage, (int)off, len);
1019                 if (ret <= 0)
1020                         return ret;
1021                 *crc = ceph_crc32c_page(*crc, rpage, off, ret);
1022                 ceph_msg_data_advance(cursor, (size_t)ret);
1023                 cursor->sr_resid -= ret;
1024                 if (do_bounce)
1025                         memcpy_page(page, off, rpage, off, ret);
1026         }
1027         return 1;
1028 }
1029
1030 static int read_partial_sparse_msg_data(struct ceph_connection *con)
1031 {
1032         struct ceph_msg_data_cursor *cursor = &con->in_msg->cursor;
1033         bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
1034         u32 crc = 0;
1035         int ret = 1;
1036
1037         if (do_datacrc)
1038                 crc = con->in_data_crc;
1039
1040         while (cursor->total_resid) {
1041                 if (con->v1.in_sr_kvec.iov_base)
1042                         ret = read_partial_message_chunk(con,
1043                                                          &con->v1.in_sr_kvec,
1044                                                          con->v1.in_sr_len,
1045                                                          &crc);
1046                 else if (cursor->sr_resid > 0)
1047                         ret = read_partial_sparse_msg_extent(con, &crc);
1048                 if (ret <= 0)
1049                         break;
1050
1051                 memset(&con->v1.in_sr_kvec, 0, sizeof(con->v1.in_sr_kvec));
1052                 ret = con->ops->sparse_read(con, cursor,
1053                                 (char **)&con->v1.in_sr_kvec.iov_base);
1054                 if (ret <= 0) {
1055                         ret = ret ? ret : 1;  /* must return > 0 to indicate success */
1056                         break;
1057                 }
1058                 con->v1.in_sr_len = ret;
1059         }
1060
1061         if (do_datacrc)
1062                 con->in_data_crc = crc;
1063
1064         return ret;
1065 }
1066
1067 static int read_partial_msg_data(struct ceph_connection *con)
1068 {
1069         struct ceph_msg_data_cursor *cursor = &con->in_msg->cursor;
1070         bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
1071         struct page *page;
1072         size_t page_offset;
1073         size_t length;
1074         u32 crc = 0;
1075         int ret;
1076
1077         if (do_datacrc)
1078                 crc = con->in_data_crc;
1079         while (cursor->total_resid) {
1080                 if (!cursor->resid) {
1081                         ceph_msg_data_advance(cursor, 0);
1082                         continue;
1083                 }
1084
1085                 page = ceph_msg_data_next(cursor, &page_offset, &length);
1086                 ret = ceph_tcp_recvpage(con->sock, page, page_offset, length);
1087                 if (ret <= 0) {
1088                         if (do_datacrc)
1089                                 con->in_data_crc = crc;
1090
1091                         return ret;
1092                 }
1093
1094                 if (do_datacrc)
1095                         crc = ceph_crc32c_page(crc, page, page_offset, ret);
1096                 ceph_msg_data_advance(cursor, (size_t)ret);
1097         }
1098         if (do_datacrc)
1099                 con->in_data_crc = crc;
1100
1101         return 1;       /* must return > 0 to indicate success */
1102 }
1103
1104 static int read_partial_msg_data_bounce(struct ceph_connection *con)
1105 {
1106         struct ceph_msg_data_cursor *cursor = &con->in_msg->cursor;
1107         struct page *page;
1108         size_t off, len;
1109         u32 crc;
1110         int ret;
1111
1112         if (unlikely(!con->bounce_page)) {
1113                 con->bounce_page = alloc_page(GFP_NOIO);
1114                 if (!con->bounce_page) {
1115                         pr_err("failed to allocate bounce page\n");
1116                         return -ENOMEM;
1117                 }
1118         }
1119
1120         crc = con->in_data_crc;
1121         while (cursor->total_resid) {
1122                 if (!cursor->resid) {
1123                         ceph_msg_data_advance(cursor, 0);
1124                         continue;
1125                 }
1126
1127                 page = ceph_msg_data_next(cursor, &off, &len);
1128                 ret = ceph_tcp_recvpage(con->sock, con->bounce_page, 0, len);
1129                 if (ret <= 0) {
1130                         con->in_data_crc = crc;
1131                         return ret;
1132                 }
1133
1134                 crc = crc32c(crc, page_address(con->bounce_page), ret);
1135                 memcpy_to_page(page, off, page_address(con->bounce_page), ret);
1136
1137                 ceph_msg_data_advance(cursor, ret);
1138         }
1139         con->in_data_crc = crc;
1140
1141         return 1;       /* must return > 0 to indicate success */
1142 }
1143
1144 /*
1145  * read (part of) a message.
1146  */
1147 static int read_partial_message(struct ceph_connection *con)
1148 {
1149         struct ceph_msg *m = con->in_msg;
1150         int size;
1151         int end;
1152         int ret;
1153         unsigned int front_len, middle_len, data_len;
1154         bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
1155         bool need_sign = (con->peer_features & CEPH_FEATURE_MSG_AUTH);
1156         u64 seq;
1157         u32 crc;
1158
1159         dout("read_partial_message con %p msg %p\n", con, m);
1160
1161         /* header */
1162         size = sizeof(con->v1.in_hdr);
1163         end = size;
1164         ret = read_partial(con, end, size, &con->v1.in_hdr);
1165         if (ret <= 0)
1166                 return ret;
1167
1168         crc = crc32c(0, &con->v1.in_hdr, offsetof(struct ceph_msg_header, crc));
1169         if (cpu_to_le32(crc) != con->v1.in_hdr.crc) {
1170                 pr_err("read_partial_message bad hdr crc %u != expected %u\n",
1171                        crc, con->v1.in_hdr.crc);
1172                 return -EBADMSG;
1173         }
1174
1175         front_len = le32_to_cpu(con->v1.in_hdr.front_len);
1176         if (front_len > CEPH_MSG_MAX_FRONT_LEN)
1177                 return -EIO;
1178         middle_len = le32_to_cpu(con->v1.in_hdr.middle_len);
1179         if (middle_len > CEPH_MSG_MAX_MIDDLE_LEN)
1180                 return -EIO;
1181         data_len = le32_to_cpu(con->v1.in_hdr.data_len);
1182         if (data_len > CEPH_MSG_MAX_DATA_LEN)
1183                 return -EIO;
1184
1185         /* verify seq# */
1186         seq = le64_to_cpu(con->v1.in_hdr.seq);
1187         if ((s64)seq - (s64)con->in_seq < 1) {
1188                 pr_info("skipping %s%lld %s seq %lld expected %lld\n",
1189                         ENTITY_NAME(con->peer_name),
1190                         ceph_pr_addr(&con->peer_addr),
1191                         seq, con->in_seq + 1);
1192                 con->v1.in_base_pos = -front_len - middle_len - data_len -
1193                                       sizeof_footer(con);
1194                 con->v1.in_tag = CEPH_MSGR_TAG_READY;
1195                 return 1;
1196         } else if ((s64)seq - (s64)con->in_seq > 1) {
1197                 pr_err("read_partial_message bad seq %lld expected %lld\n",
1198                        seq, con->in_seq + 1);
1199                 con->error_msg = "bad message sequence # for incoming message";
1200                 return -EBADE;
1201         }
1202
1203         /* allocate message? */
1204         if (!con->in_msg) {
1205                 int skip = 0;
1206
1207                 dout("got hdr type %d front %d data %d\n", con->v1.in_hdr.type,
1208                      front_len, data_len);
1209                 ret = ceph_con_in_msg_alloc(con, &con->v1.in_hdr, &skip);
1210                 if (ret < 0)
1211                         return ret;
1212
1213                 BUG_ON((!con->in_msg) ^ skip);
1214                 if (skip) {
1215                         /* skip this message */
1216                         dout("alloc_msg said skip message\n");
1217                         con->v1.in_base_pos = -front_len - middle_len -
1218                                               data_len - sizeof_footer(con);
1219                         con->v1.in_tag = CEPH_MSGR_TAG_READY;
1220                         con->in_seq++;
1221                         return 1;
1222                 }
1223
1224                 BUG_ON(!con->in_msg);
1225                 BUG_ON(con->in_msg->con != con);
1226                 m = con->in_msg;
1227                 m->front.iov_len = 0;    /* haven't read it yet */
1228                 if (m->middle)
1229                         m->middle->vec.iov_len = 0;
1230
1231                 /* prepare for data payload, if any */
1232
1233                 if (data_len)
1234                         prepare_message_data(con->in_msg, data_len);
1235         }
1236
1237         /* front */
1238         ret = read_partial_message_section(con, &m->front, front_len,
1239                                            &con->in_front_crc);
1240         if (ret <= 0)
1241                 return ret;
1242
1243         /* middle */
1244         if (m->middle) {
1245                 ret = read_partial_message_section(con, &m->middle->vec,
1246                                                    middle_len,
1247                                                    &con->in_middle_crc);
1248                 if (ret <= 0)
1249                         return ret;
1250         }
1251
1252         /* (page) data */
1253         if (data_len) {
1254                 if (!m->num_data_items)
1255                         return -EIO;
1256
1257                 if (m->sparse_read_total)
1258                         ret = read_partial_sparse_msg_data(con);
1259                 else if (ceph_test_opt(from_msgr(con->msgr), RXBOUNCE))
1260                         ret = read_partial_msg_data_bounce(con);
1261                 else
1262                         ret = read_partial_msg_data(con);
1263                 if (ret <= 0)
1264                         return ret;
1265         }
1266
1267         /* footer */
1268         size = sizeof_footer(con);
1269         end += size;
1270         ret = read_partial(con, end, size, &m->footer);
1271         if (ret <= 0)
1272                 return ret;
1273
1274         if (!need_sign) {
1275                 m->footer.flags = m->old_footer.flags;
1276                 m->footer.sig = 0;
1277         }
1278
1279         dout("read_partial_message got msg %p %d (%u) + %d (%u) + %d (%u)\n",
1280              m, front_len, m->footer.front_crc, middle_len,
1281              m->footer.middle_crc, data_len, m->footer.data_crc);
1282
1283         /* crc ok? */
1284         if (con->in_front_crc != le32_to_cpu(m->footer.front_crc)) {
1285                 pr_err("read_partial_message %p front crc %u != exp. %u\n",
1286                        m, con->in_front_crc, m->footer.front_crc);
1287                 return -EBADMSG;
1288         }
1289         if (con->in_middle_crc != le32_to_cpu(m->footer.middle_crc)) {
1290                 pr_err("read_partial_message %p middle crc %u != exp %u\n",
1291                        m, con->in_middle_crc, m->footer.middle_crc);
1292                 return -EBADMSG;
1293         }
1294         if (do_datacrc &&
1295             (m->footer.flags & CEPH_MSG_FOOTER_NOCRC) == 0 &&
1296             con->in_data_crc != le32_to_cpu(m->footer.data_crc)) {
1297                 pr_err("read_partial_message %p data crc %u != exp. %u\n", m,
1298                        con->in_data_crc, le32_to_cpu(m->footer.data_crc));
1299                 return -EBADMSG;
1300         }
1301
1302         if (need_sign && con->ops->check_message_signature &&
1303             con->ops->check_message_signature(m)) {
1304                 pr_err("read_partial_message %p signature check failed\n", m);
1305                 return -EBADMSG;
1306         }
1307
1308         return 1; /* done! */
1309 }
1310
1311 static int read_keepalive_ack(struct ceph_connection *con)
1312 {
1313         struct ceph_timespec ceph_ts;
1314         size_t size = sizeof(ceph_ts);
1315         int ret = read_partial(con, size, size, &ceph_ts);
1316         if (ret <= 0)
1317                 return ret;
1318         ceph_decode_timespec64(&con->last_keepalive_ack, &ceph_ts);
1319         prepare_read_tag(con);
1320         return 1;
1321 }
1322
1323 /*
1324  * Read what we can from the socket.
1325  */
1326 int ceph_con_v1_try_read(struct ceph_connection *con)
1327 {
1328         int ret = -1;
1329
1330 more:
1331         dout("try_read start %p state %d\n", con, con->state);
1332         if (con->state != CEPH_CON_S_V1_BANNER &&
1333             con->state != CEPH_CON_S_V1_CONNECT_MSG &&
1334             con->state != CEPH_CON_S_OPEN)
1335                 return 0;
1336
1337         BUG_ON(!con->sock);
1338
1339         dout("try_read tag %d in_base_pos %d\n", con->v1.in_tag,
1340              con->v1.in_base_pos);
1341
1342         if (con->state == CEPH_CON_S_V1_BANNER) {
1343                 ret = read_partial_banner(con);
1344                 if (ret <= 0)
1345                         goto out;
1346                 ret = process_banner(con);
1347                 if (ret < 0)
1348                         goto out;
1349
1350                 con->state = CEPH_CON_S_V1_CONNECT_MSG;
1351
1352                 /*
1353                  * Received banner is good, exchange connection info.
1354                  * Do not reset out_kvec, as sending our banner raced
1355                  * with receiving peer banner after connect completed.
1356                  */
1357                 ret = prepare_write_connect(con);
1358                 if (ret < 0)
1359                         goto out;
1360                 prepare_read_connect(con);
1361
1362                 /* Send connection info before awaiting response */
1363                 goto out;
1364         }
1365
1366         if (con->state == CEPH_CON_S_V1_CONNECT_MSG) {
1367                 ret = read_partial_connect(con);
1368                 if (ret <= 0)
1369                         goto out;
1370                 ret = process_connect(con);
1371                 if (ret < 0)
1372                         goto out;
1373                 goto more;
1374         }
1375
1376         WARN_ON(con->state != CEPH_CON_S_OPEN);
1377
1378         if (con->v1.in_base_pos < 0) {
1379                 /*
1380                  * skipping + discarding content.
1381                  */
1382                 ret = ceph_tcp_recvmsg(con->sock, NULL, -con->v1.in_base_pos);
1383                 if (ret <= 0)
1384                         goto out;
1385                 dout("skipped %d / %d bytes\n", ret, -con->v1.in_base_pos);
1386                 con->v1.in_base_pos += ret;
1387                 if (con->v1.in_base_pos)
1388                         goto more;
1389         }
1390         if (con->v1.in_tag == CEPH_MSGR_TAG_READY) {
1391                 /*
1392                  * what's next?
1393                  */
1394                 ret = ceph_tcp_recvmsg(con->sock, &con->v1.in_tag, 1);
1395                 if (ret <= 0)
1396                         goto out;
1397                 dout("try_read got tag %d\n", con->v1.in_tag);
1398                 switch (con->v1.in_tag) {
1399                 case CEPH_MSGR_TAG_MSG:
1400                         prepare_read_message(con);
1401                         break;
1402                 case CEPH_MSGR_TAG_ACK:
1403                         prepare_read_ack(con);
1404                         break;
1405                 case CEPH_MSGR_TAG_KEEPALIVE2_ACK:
1406                         prepare_read_keepalive_ack(con);
1407                         break;
1408                 case CEPH_MSGR_TAG_CLOSE:
1409                         ceph_con_close_socket(con);
1410                         con->state = CEPH_CON_S_CLOSED;
1411                         goto out;
1412                 default:
1413                         goto bad_tag;
1414                 }
1415         }
1416         if (con->v1.in_tag == CEPH_MSGR_TAG_MSG) {
1417                 ret = read_partial_message(con);
1418                 if (ret <= 0) {
1419                         switch (ret) {
1420                         case -EBADMSG:
1421                                 con->error_msg = "bad crc/signature";
1422                                 fallthrough;
1423                         case -EBADE:
1424                                 ret = -EIO;
1425                                 break;
1426                         case -EIO:
1427                                 con->error_msg = "io error";
1428                                 break;
1429                         }
1430                         goto out;
1431                 }
1432                 if (con->v1.in_tag == CEPH_MSGR_TAG_READY)
1433                         goto more;
1434                 ceph_con_process_message(con);
1435                 if (con->state == CEPH_CON_S_OPEN)
1436                         prepare_read_tag(con);
1437                 goto more;
1438         }
1439         if (con->v1.in_tag == CEPH_MSGR_TAG_ACK ||
1440             con->v1.in_tag == CEPH_MSGR_TAG_SEQ) {
1441                 /*
1442                  * the final handshake seq exchange is semantically
1443                  * equivalent to an ACK
1444                  */
1445                 ret = read_partial_ack(con);
1446                 if (ret <= 0)
1447                         goto out;
1448                 process_ack(con);
1449                 goto more;
1450         }
1451         if (con->v1.in_tag == CEPH_MSGR_TAG_KEEPALIVE2_ACK) {
1452                 ret = read_keepalive_ack(con);
1453                 if (ret <= 0)
1454                         goto out;
1455                 goto more;
1456         }
1457
1458 out:
1459         dout("try_read done on %p ret %d\n", con, ret);
1460         return ret;
1461
1462 bad_tag:
1463         pr_err("try_read bad tag %d\n", con->v1.in_tag);
1464         con->error_msg = "protocol error, garbage tag";
1465         ret = -1;
1466         goto out;
1467 }
1468
1469 /*
1470  * Write something to the socket.  Called in a worker thread when the
1471  * socket appears to be writeable and we have something ready to send.
1472  */
1473 int ceph_con_v1_try_write(struct ceph_connection *con)
1474 {
1475         int ret = 1;
1476
1477         dout("try_write start %p state %d\n", con, con->state);
1478         if (con->state != CEPH_CON_S_PREOPEN &&
1479             con->state != CEPH_CON_S_V1_BANNER &&
1480             con->state != CEPH_CON_S_V1_CONNECT_MSG &&
1481             con->state != CEPH_CON_S_OPEN)
1482                 return 0;
1483
1484         /* open the socket first? */
1485         if (con->state == CEPH_CON_S_PREOPEN) {
1486                 BUG_ON(con->sock);
1487                 con->state = CEPH_CON_S_V1_BANNER;
1488
1489                 con_out_kvec_reset(con);
1490                 prepare_write_banner(con);
1491                 prepare_read_banner(con);
1492
1493                 BUG_ON(con->in_msg);
1494                 con->v1.in_tag = CEPH_MSGR_TAG_READY;
1495                 dout("try_write initiating connect on %p new state %d\n",
1496                      con, con->state);
1497                 ret = ceph_tcp_connect(con);
1498                 if (ret < 0) {
1499                         con->error_msg = "connect error";
1500                         goto out;
1501                 }
1502         }
1503
1504 more:
1505         dout("try_write out_kvec_bytes %d\n", con->v1.out_kvec_bytes);
1506         BUG_ON(!con->sock);
1507
1508         /* kvec data queued? */
1509         if (con->v1.out_kvec_left) {
1510                 ret = write_partial_kvec(con);
1511                 if (ret <= 0)
1512                         goto out;
1513         }
1514         if (con->v1.out_skip) {
1515                 ret = write_partial_skip(con);
1516                 if (ret <= 0)
1517                         goto out;
1518         }
1519
1520         /* msg pages? */
1521         if (con->out_msg) {
1522                 if (con->v1.out_msg_done) {
1523                         ceph_msg_put(con->out_msg);
1524                         con->out_msg = NULL;   /* we're done with this one */
1525                         goto do_next;
1526                 }
1527
1528                 ret = write_partial_message_data(con);
1529                 if (ret == 1)
1530                         goto more;  /* we need to send the footer, too! */
1531                 if (ret == 0)
1532                         goto out;
1533                 if (ret < 0) {
1534                         dout("try_write write_partial_message_data err %d\n",
1535                              ret);
1536                         goto out;
1537                 }
1538         }
1539
1540 do_next:
1541         if (con->state == CEPH_CON_S_OPEN) {
1542                 if (ceph_con_flag_test_and_clear(con,
1543                                 CEPH_CON_F_KEEPALIVE_PENDING)) {
1544                         prepare_write_keepalive(con);
1545                         goto more;
1546                 }
1547                 /* is anything else pending? */
1548                 if (!list_empty(&con->out_queue)) {
1549                         prepare_write_message(con);
1550                         goto more;
1551                 }
1552                 if (con->in_seq > con->in_seq_acked) {
1553                         prepare_write_ack(con);
1554                         goto more;
1555                 }
1556         }
1557
1558         /* Nothing to do! */
1559         ceph_con_flag_clear(con, CEPH_CON_F_WRITE_PENDING);
1560         dout("try_write nothing else to write.\n");
1561         ret = 0;
1562 out:
1563         dout("try_write done on %p ret %d\n", con, ret);
1564         return ret;
1565 }
1566
1567 void ceph_con_v1_revoke(struct ceph_connection *con)
1568 {
1569         struct ceph_msg *msg = con->out_msg;
1570
1571         WARN_ON(con->v1.out_skip);
1572         /* footer */
1573         if (con->v1.out_msg_done) {
1574                 con->v1.out_skip += con_out_kvec_skip(con);
1575         } else {
1576                 WARN_ON(!msg->data_length);
1577                 con->v1.out_skip += sizeof_footer(con);
1578         }
1579         /* data, middle, front */
1580         if (msg->data_length)
1581                 con->v1.out_skip += msg->cursor.total_resid;
1582         if (msg->middle)
1583                 con->v1.out_skip += con_out_kvec_skip(con);
1584         con->v1.out_skip += con_out_kvec_skip(con);
1585
1586         dout("%s con %p out_kvec_bytes %d out_skip %d\n", __func__, con,
1587              con->v1.out_kvec_bytes, con->v1.out_skip);
1588 }
1589
1590 void ceph_con_v1_revoke_incoming(struct ceph_connection *con)
1591 {
1592         unsigned int front_len = le32_to_cpu(con->v1.in_hdr.front_len);
1593         unsigned int middle_len = le32_to_cpu(con->v1.in_hdr.middle_len);
1594         unsigned int data_len = le32_to_cpu(con->v1.in_hdr.data_len);
1595
1596         /* skip rest of message */
1597         con->v1.in_base_pos = con->v1.in_base_pos -
1598                         sizeof(struct ceph_msg_header) -
1599                         front_len -
1600                         middle_len -
1601                         data_len -
1602                         sizeof(struct ceph_msg_footer);
1603
1604         con->v1.in_tag = CEPH_MSGR_TAG_READY;
1605         con->in_seq++;
1606
1607         dout("%s con %p in_base_pos %d\n", __func__, con, con->v1.in_base_pos);
1608 }
1609
1610 bool ceph_con_v1_opened(struct ceph_connection *con)
1611 {
1612         return con->v1.connect_seq;
1613 }
1614
1615 void ceph_con_v1_reset_session(struct ceph_connection *con)
1616 {
1617         con->v1.connect_seq = 0;
1618         con->v1.peer_global_seq = 0;
1619 }
1620
1621 void ceph_con_v1_reset_protocol(struct ceph_connection *con)
1622 {
1623         con->v1.out_skip = 0;
1624 }