tcp_connect.c 4.5 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167
  1. /*
  2. * Copyright (c) 2006 Oracle. All rights reserved.
  3. *
  4. * This software is available to you under a choice of one of two
  5. * licenses. You may choose to be licensed under the terms of the GNU
  6. * General Public License (GPL) Version 2, available from the file
  7. * COPYING in the main directory of this source tree, or the
  8. * OpenIB.org BSD license below:
  9. *
  10. * Redistribution and use in source and binary forms, with or
  11. * without modification, are permitted provided that the following
  12. * conditions are met:
  13. *
  14. * - Redistributions of source code must retain the above
  15. * copyright notice, this list of conditions and the following
  16. * disclaimer.
  17. *
  18. * - Redistributions in binary form must reproduce the above
  19. * copyright notice, this list of conditions and the following
  20. * disclaimer in the documentation and/or other materials
  21. * provided with the distribution.
  22. *
  23. * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  24. * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  25. * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
  26. * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
  27. * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
  28. * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
  29. * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  30. * SOFTWARE.
  31. *
  32. */
  33. #include <linux/kernel.h>
  34. #include <linux/in.h>
  35. #include <net/tcp.h>
  36. #include "rds.h"
  37. #include "tcp.h"
  38. void rds_tcp_state_change(struct sock *sk)
  39. {
  40. void (*state_change)(struct sock *sk);
  41. struct rds_connection *conn;
  42. struct rds_tcp_connection *tc;
  43. read_lock_bh(&sk->sk_callback_lock);
  44. conn = sk->sk_user_data;
  45. if (!conn) {
  46. state_change = sk->sk_state_change;
  47. goto out;
  48. }
  49. tc = conn->c_transport_data;
  50. state_change = tc->t_orig_state_change;
  51. rdsdebug("sock %p state_change to %d\n", tc->t_sock, sk->sk_state);
  52. switch(sk->sk_state) {
  53. /* ignore connecting sockets as they make progress */
  54. case TCP_SYN_SENT:
  55. case TCP_SYN_RECV:
  56. break;
  57. case TCP_ESTABLISHED:
  58. rds_connect_complete(conn);
  59. break;
  60. case TCP_CLOSE_WAIT:
  61. case TCP_CLOSE:
  62. rds_conn_drop(conn);
  63. default:
  64. break;
  65. }
  66. out:
  67. read_unlock_bh(&sk->sk_callback_lock);
  68. state_change(sk);
  69. }
  70. int rds_tcp_conn_connect(struct rds_connection *conn)
  71. {
  72. struct socket *sock = NULL;
  73. struct sockaddr_in src, dest;
  74. int ret;
  75. struct rds_tcp_connection *tc = conn->c_transport_data;
  76. mutex_lock(&tc->t_conn_lock);
  77. if (rds_conn_up(conn)) {
  78. mutex_unlock(&tc->t_conn_lock);
  79. return 0;
  80. }
  81. ret = sock_create_kern(rds_conn_net(conn), PF_INET,
  82. SOCK_STREAM, IPPROTO_TCP, &sock);
  83. if (ret < 0)
  84. goto out;
  85. rds_tcp_tune(sock);
  86. src.sin_family = AF_INET;
  87. src.sin_addr.s_addr = (__force u32)conn->c_laddr;
  88. src.sin_port = (__force u16)htons(0);
  89. ret = sock->ops->bind(sock, (struct sockaddr *)&src, sizeof(src));
  90. if (ret) {
  91. rdsdebug("bind failed with %d at address %pI4\n",
  92. ret, &conn->c_laddr);
  93. goto out;
  94. }
  95. dest.sin_family = AF_INET;
  96. dest.sin_addr.s_addr = (__force u32)conn->c_faddr;
  97. dest.sin_port = (__force u16)htons(RDS_TCP_PORT);
  98. /*
  99. * once we call connect() we can start getting callbacks and they
  100. * own the socket
  101. */
  102. rds_tcp_set_callbacks(sock, conn);
  103. ret = sock->ops->connect(sock, (struct sockaddr *)&dest, sizeof(dest),
  104. O_NONBLOCK);
  105. rdsdebug("connect to address %pI4 returned %d\n", &conn->c_faddr, ret);
  106. if (ret == -EINPROGRESS)
  107. ret = 0;
  108. if (ret == 0) {
  109. rds_tcp_keepalive(sock);
  110. sock = NULL;
  111. } else {
  112. rds_tcp_restore_callbacks(sock, conn->c_transport_data);
  113. }
  114. out:
  115. mutex_unlock(&tc->t_conn_lock);
  116. if (sock)
  117. sock_release(sock);
  118. return ret;
  119. }
  120. /*
  121. * Before killing the tcp socket this needs to serialize with callbacks. The
  122. * caller has already grabbed the sending sem so we're serialized with other
  123. * senders.
  124. *
  125. * TCP calls the callbacks with the sock lock so we hold it while we reset the
  126. * callbacks to those set by TCP. Our callbacks won't execute again once we
  127. * hold the sock lock.
  128. */
  129. void rds_tcp_conn_shutdown(struct rds_connection *conn)
  130. {
  131. struct rds_tcp_connection *tc = conn->c_transport_data;
  132. struct socket *sock = tc->t_sock;
  133. rdsdebug("shutting down conn %p tc %p sock %p\n", conn, tc, sock);
  134. if (sock) {
  135. sock->ops->shutdown(sock, RCV_SHUTDOWN | SEND_SHUTDOWN);
  136. lock_sock(sock->sk);
  137. rds_tcp_restore_callbacks(sock, tc); /* tc->tc_sock = NULL */
  138. release_sock(sock->sk);
  139. sock_release(sock);
  140. }
  141. if (tc->t_tinc) {
  142. rds_inc_put(&tc->t_tinc->ti_inc);
  143. tc->t_tinc = NULL;
  144. }
  145. tc->t_tinc_hdr_rem = sizeof(struct rds_header);
  146. tc->t_tinc_data_rem = 0;
  147. }