summaryrefslogtreecommitdiffstats
path: root/nsprpub/pr/src/md/windows/w95sock.c
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-07 19:33:14 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-07 19:33:14 +0000
commit36d22d82aa202bb199967e9512281e9a53db42c9 (patch)
tree105e8c98ddea1c1e4784a60a5a6410fa416be2de /nsprpub/pr/src/md/windows/w95sock.c
parentInitial commit. (diff)
downloadfirefox-esr-upstream.tar.xz
firefox-esr-upstream.zip
Adding upstream version 115.7.0esr.upstream/115.7.0esrupstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r--nsprpub/pr/src/md/windows/w95sock.c851
1 files changed, 851 insertions, 0 deletions
diff --git a/nsprpub/pr/src/md/windows/w95sock.c b/nsprpub/pr/src/md/windows/w95sock.c
new file mode 100644
index 0000000000..abe462fe02
--- /dev/null
+++ b/nsprpub/pr/src/md/windows/w95sock.c
@@ -0,0 +1,851 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/* This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
+
+/* Win95 Sockets module
+ *
+ */
+
+#if defined(_WIN64)
+#include <winsock2.h>
+#endif
+#include "primpl.h"
+
+#define READ_FD 1
+#define WRITE_FD 2
+#define CONNECT_FD 3
+
+static PRInt32 socket_io_wait(
+ PROsfd osfd,
+ PRInt32 fd_type,
+ PRIntervalTime timeout);
+
+
+/* --- SOCKET IO --------------------------------------------------------- */
+
+static PRBool socketFixInet6RcvBuf = PR_FALSE;
+
+void _PR_MD_InitSockets(void)
+{
+ OSVERSIONINFO osvi;
+
+ memset(&osvi, 0, sizeof(osvi));
+ osvi.dwOSVersionInfoSize = sizeof(osvi);
+ GetVersionEx(&osvi);
+
+ if (osvi.dwMajorVersion == 5 && osvi.dwMinorVersion == 1)
+ {
+ /* if Windows XP (32-bit) */
+ socketFixInet6RcvBuf = PR_TRUE;
+ }
+}
+
+void _PR_MD_CleanupSockets(void)
+{
+ socketFixInet6RcvBuf = PR_FALSE;
+}
+
+PROsfd
+_PR_MD_SOCKET(int af, int type, int flags)
+{
+ SOCKET sock;
+ u_long one = 1;
+
+ sock = socket(af, type, flags);
+
+ if (sock == INVALID_SOCKET )
+ {
+ _PR_MD_MAP_SOCKET_ERROR(WSAGetLastError());
+ return (PROsfd)sock;
+ }
+
+ /*
+ ** Make the socket Non-Blocking
+ */
+ if (ioctlsocket( sock, FIONBIO, &one) != 0)
+ {
+ PR_SetError(PR_UNKNOWN_ERROR, WSAGetLastError());
+ closesocket(sock);
+ return -1;
+ }
+
+ if (af == AF_INET6 && socketFixInet6RcvBuf)
+ {
+ int bufsize;
+ int len = sizeof(bufsize);
+ int rv;
+
+ /* Windows XP 32-bit returns an error on getpeername() for AF_INET6
+ * sockets if the receive buffer size is greater than 65535 before
+ * the connection is initiated. The default receive buffer size may
+ * be 128000 so fix it here to always be <= 65535. See bug 513659
+ * and IBM DB2 support technote "Receive/Send IPv6 Socket Size
+ * Problem in Windows XP SP2 & SP3".
+ */
+ rv = getsockopt(sock, SOL_SOCKET, SO_RCVBUF, (char*)&bufsize, &len);
+ if (rv == 0 && bufsize > 65535)
+ {
+ bufsize = 65535;
+ setsockopt(sock, SOL_SOCKET, SO_RCVBUF, (char*)&bufsize, len);
+ }
+ }
+
+ return (PROsfd)sock;
+}
+
+/*
+** _MD_CloseSocket() -- Close a socket
+**
+*/
+PRInt32
+_MD_CloseSocket(PROsfd osfd)
+{
+ PRInt32 rv;
+
+ rv = closesocket((SOCKET) osfd );
+ if (rv < 0) {
+ _PR_MD_MAP_CLOSE_ERROR(WSAGetLastError());
+ }
+
+ return rv;
+}
+
+PRInt32
+_MD_SocketAvailable(PRFileDesc *fd)
+{
+ PRInt32 result;
+
+ if (ioctlsocket(fd->secret->md.osfd, FIONREAD, &result) < 0) {
+ PR_SetError(PR_BAD_DESCRIPTOR_ERROR, WSAGetLastError());
+ return -1;
+ }
+ return result;
+}
+
+PROsfd _MD_Accept(
+ PRFileDesc *fd,
+ PRNetAddr *raddr,
+ PRUint32 *rlen,
+ PRIntervalTime timeout )
+{
+ PROsfd osfd = fd->secret->md.osfd;
+ SOCKET sock;
+ PRInt32 rv, err;
+
+ while ((sock = accept(osfd, (struct sockaddr *) raddr, rlen)) == -1)
+ {
+ err = WSAGetLastError();
+ if ((err == WSAEWOULDBLOCK) && (!fd->secret->nonblocking))
+ {
+ if ((rv = socket_io_wait(osfd, READ_FD, timeout)) < 0)
+ {
+ break;
+ }
+ }
+ else
+ {
+ _PR_MD_MAP_ACCEPT_ERROR(err);
+ break;
+ }
+ }
+ return(sock);
+} /* end _MD_accept() */
+
+PRInt32
+_PR_MD_CONNECT(PRFileDesc *fd, const PRNetAddr *addr, PRUint32 addrlen,
+ PRIntervalTime timeout)
+{
+ PROsfd osfd = fd->secret->md.osfd;
+ PRInt32 rv;
+ int err;
+
+ if ((rv = connect(osfd, (struct sockaddr *) addr, addrlen)) == -1)
+ {
+ err = WSAGetLastError();
+ if ((!fd->secret->nonblocking) && (err == WSAEWOULDBLOCK))
+ {
+ rv = socket_io_wait(osfd, CONNECT_FD, timeout);
+ if ( rv < 0 )
+ {
+ return(-1);
+ }
+ else
+ {
+ PR_ASSERT(rv > 0);
+ /* it's connected */
+ return(0);
+ }
+ }
+ _PR_MD_MAP_CONNECT_ERROR(err);
+ }
+ return rv;
+}
+
+PRInt32
+_PR_MD_BIND(PRFileDesc *fd, const PRNetAddr *addr, PRUint32 addrlen)
+{
+ PRInt32 rv;
+
+ rv = bind(fd->secret->md.osfd, (const struct sockaddr *)&(addr->inet), addrlen);
+
+ if (rv == SOCKET_ERROR) {
+ _PR_MD_MAP_BIND_ERROR(WSAGetLastError());
+ return -1;
+ }
+
+ return 0;
+}
+
+PRInt32
+_PR_MD_LISTEN(PRFileDesc *fd, PRIntn backlog)
+{
+ PRInt32 rv;
+
+ rv = listen(fd->secret->md.osfd, backlog);
+
+ if (rv == SOCKET_ERROR) {
+ _PR_MD_MAP_DEFAULT_ERROR(WSAGetLastError());
+ return -1;
+ }
+
+ return 0;
+}
+
+PRInt32
+_PR_MD_RECV(PRFileDesc *fd, void *buf, PRInt32 amount, PRIntn flags,
+ PRIntervalTime timeout)
+{
+ PROsfd osfd = fd->secret->md.osfd;
+ PRInt32 rv, err;
+ int osflags;
+
+ if (0 == flags) {
+ osflags = 0;
+ } else {
+ PR_ASSERT(PR_MSG_PEEK == flags);
+ osflags = MSG_PEEK;
+ }
+ while ((rv = recv( osfd, buf, amount, osflags)) == -1)
+ {
+ if (((err = WSAGetLastError()) == WSAEWOULDBLOCK)
+ && (!fd->secret->nonblocking))
+ {
+ rv = socket_io_wait(osfd, READ_FD, timeout);
+ if ( rv < 0 )
+ {
+ return -1;
+ }
+ }
+ else
+ {
+ _PR_MD_MAP_RECV_ERROR(err);
+ break;
+ }
+ } /* end while() */
+ return(rv);
+}
+
+PRInt32
+_PR_MD_SEND(PRFileDesc *fd, const void *buf, PRInt32 amount, PRIntn flags,
+ PRIntervalTime timeout)
+{
+ PROsfd osfd = fd->secret->md.osfd;
+ PRInt32 rv, err;
+ PRInt32 bytesSent = 0;
+
+ while(bytesSent < amount )
+ {
+ while ((rv = send( osfd, buf, amount, 0 )) == -1)
+ {
+ if (((err = WSAGetLastError()) == WSAEWOULDBLOCK)
+ && (!fd->secret->nonblocking))
+ {
+ rv = socket_io_wait(osfd, WRITE_FD, timeout);
+ if ( rv < 0 )
+ {
+ return -1;
+ }
+ }
+ else
+ {
+ _PR_MD_MAP_SEND_ERROR(err);
+ return -1;
+ }
+ }
+ bytesSent += rv;
+ if (fd->secret->nonblocking)
+ {
+ break;
+ }
+ if (bytesSent < amount)
+ {
+ rv = socket_io_wait(osfd, WRITE_FD, timeout);
+ if ( rv < 0 )
+ {
+ return -1;
+ }
+ }
+ }
+ return bytesSent;
+}
+
+PRInt32
+_PR_MD_SENDTO(PRFileDesc *fd, const void *buf, PRInt32 amount, PRIntn flags,
+ const PRNetAddr *addr, PRUint32 addrlen, PRIntervalTime timeout)
+{
+ PROsfd osfd = fd->secret->md.osfd;
+ PRInt32 rv, err;
+ PRInt32 bytesSent = 0;
+
+ do {
+ while ((rv = sendto( osfd, buf, amount, 0, (struct sockaddr *) addr,
+ addrlen)) == -1)
+ {
+ if (((err = WSAGetLastError()) == WSAEWOULDBLOCK)
+ && (!fd->secret->nonblocking))
+ {
+ rv = socket_io_wait(osfd, WRITE_FD, timeout);
+ if ( rv < 0 )
+ {
+ return -1;
+ }
+ }
+ else
+ {
+ _PR_MD_MAP_SENDTO_ERROR(err);
+ return -1;
+ }
+ }
+ bytesSent += rv;
+ if (fd->secret->nonblocking)
+ {
+ break;
+ }
+ if (bytesSent < amount)
+ {
+ rv = socket_io_wait(osfd, WRITE_FD, timeout);
+ if (rv < 0)
+ {
+ return -1;
+ }
+ }
+ } while(bytesSent < amount);
+ return bytesSent;
+}
+
+#if defined(_WIN64)
+
+static PRCallOnceType _pr_has_connectex_once;
+typedef BOOL (PASCAL FAR * _pr_win_connectex_ptr)(_In_ SOCKET s, _In_reads_bytes_(namelen) const struct sockaddr FAR *name, _In_ int namelen, _In_reads_bytes_opt_(dwSendDataLength) PVOID lpSendBuffer, _In_ DWORD dwSendDataLength, _Out_ LPDWORD lpdwBytesSent, _Inout_ LPOVERLAPPED lpOverlapped);
+
+#ifndef WSAID_CONNECTEX
+#define WSAID_CONNECTEX \
+ {0x25a207b9,0xddf3,0x4660,{0x8e,0xe9,0x76,0xe5,0x8c,0x74,0x06,0x3e}}
+#endif
+#ifndef SIO_GET_EXTENSION_FUNCTION_POINTER
+#define SIO_GET_EXTENSION_FUNCTION_POINTER 0xC8000006
+#endif
+#ifndef TCP_FASTOPEN
+#define TCP_FASTOPEN 15
+#endif
+
+#ifndef SO_UPDATE_CONNECT_CONTEXT
+#define SO_UPDATE_CONNECT_CONTEXT 0x7010
+#endif
+
+static _pr_win_connectex_ptr _pr_win_connectex = NULL;
+
+static PRStatus PR_CALLBACK _pr_set_connectex(void)
+{
+ _pr_win_connectex = NULL;
+ SOCKET sock;
+ PRInt32 dwBytes;
+ int rc;
+
+ /* Dummy socket needed for WSAIoctl */
+ sock = socket(AF_INET, SOCK_STREAM, 0);
+ if (sock == INVALID_SOCKET) {
+ return PR_SUCCESS;
+ }
+
+ GUID guid = WSAID_CONNECTEX;
+ rc = WSAIoctl(sock, SIO_GET_EXTENSION_FUNCTION_POINTER,
+ &guid, sizeof(guid),
+ &_pr_win_connectex, sizeof(_pr_win_connectex),
+ &dwBytes, NULL, NULL);
+ if (rc != 0) {
+ _pr_win_connectex = NULL;
+ return PR_SUCCESS;
+ }
+
+ rc = closesocket(sock);
+ return PR_SUCCESS;
+}
+
+PRInt32
+_PR_MD_TCPSENDTO(PRFileDesc *fd, const void *buf, PRInt32 amount, PRIntn flags,
+ const PRNetAddr *addr, PRUint32 addrlen, PRIntervalTime timeout)
+{
+ if (!_fd_waiting_for_overlapped_done_lock) {
+ PR_SetError(PR_NOT_IMPLEMENTED_ERROR, 0);
+ return PR_FAILURE;
+ }
+
+ if (PR_CallOnce(&_pr_has_connectex_once, _pr_set_connectex) != PR_SUCCESS) {
+ PR_SetError(PR_NOT_IMPLEMENTED_ERROR, 0);
+ return PR_FAILURE;
+ }
+
+ if (_pr_win_connectex == NULL) {
+ PR_SetError(PR_NOT_IMPLEMENTED_ERROR, 0);
+ return PR_FAILURE;
+ }
+
+ PROsfd osfd = fd->secret->md.osfd;
+ PRInt32 rv, err;
+ PRInt32 bytesSent = 0;
+ DWORD rvSent;
+
+ BOOL option = 1;
+ rv = setsockopt((SOCKET)osfd, IPPROTO_TCP, TCP_FASTOPEN, (char*)&option, sizeof(option));
+ if (rv != 0) {
+ err = WSAGetLastError();
+ PR_LOG(_pr_io_lm, PR_LOG_MIN,
+ ("_PR_MD_TCPSENDTO error set opt TCP_FASTOPEN failed %d\n", err));
+ if (err == WSAENOPROTOOPT) {
+ PR_SetError(PR_NOT_IMPLEMENTED_ERROR, 0);
+ } else {
+ _PR_MD_MAP_SETSOCKOPT_ERROR(err);
+ }
+ return -1;
+ }
+
+ /* ConnectEx requires the socket to be initially bound. We will use INADDR_ANY. */
+ PRNetAddr bindAddr;
+ memset(&bindAddr, 0, sizeof(bindAddr));
+ bindAddr.raw.family = addr->raw.family;
+
+ rv = bind((SOCKET)osfd, (const struct sockaddr *)&(bindAddr.inet), PR_NETADDR_SIZE(&bindAddr));
+ if (rv != 0) {
+ err = WSAGetLastError();
+ PR_LOG(_pr_io_lm, PR_LOG_MIN,
+ ("_PR_MD_TCPSENDTO error bind failed %d\n", err));
+ _PR_MD_MAP_SETSOCKOPT_ERROR(err);
+ return -1;
+ }
+
+ PR_LOG(_pr_io_lm, PR_LOG_MIN,
+ ("_PR_MD_TCPSENDTO calling _pr_win_connectex %d %p\n", amount, (char*)buf));
+
+ rvSent = 0;
+ memset(&fd->secret->ol, 0, sizeof(fd->secret->ol));
+ /* ConnectEx return TRUE on a success and FALSE on an error. */
+ if (_pr_win_connectex( (SOCKET)osfd, (struct sockaddr *) addr,
+ addrlen, buf, amount,
+ &rvSent, &fd->secret->ol) == TRUE) {
+ /* When ConnectEx is used, all previously set socket options and
+ * property are not enabled and to enable them
+ * SO_UPDATE_CONNECT_CONTEXT option need to be set. */
+ rv = setsockopt((SOCKET)osfd, SOL_SOCKET, SO_UPDATE_CONNECT_CONTEXT, NULL, 0);
+ if (rv != 0) {
+ err = WSAGetLastError();
+ PR_LOG(_pr_io_lm, PR_LOG_MIN,
+ ("_PR_MD_TCPSENDTO setting SO_UPDATE_CONNECT_CONTEXT failed %d\n", err));
+ _PR_MD_MAP_SETSOCKOPT_ERROR(err);
+ return -1;
+ }
+ /* We imitate Linux here. SendTo will return number of bytes send but
+ * it can not return connection success at the same time, so we return
+ * number of bytes send and "connection success" will be return on the
+ * connectcontinue. */
+ fd->secret->alreadyConnected = PR_TRUE;
+ return rvSent;
+ } else {
+ err = WSAGetLastError();
+ PR_LOG(_pr_io_lm, PR_LOG_MIN,
+ ("_PR_MD_TCPSENDTO error _pr_win_connectex failed %d\n", err));
+ if (err != ERROR_IO_PENDING) {
+ _PR_MD_MAP_CONNECT_ERROR(err);
+ return -1;
+ } else if (fd->secret->nonblocking) {
+ /* Remember that overlapped structure is set. We will need to get
+ * the final result of ConnectEx call. */
+ fd->secret->overlappedActive = PR_TRUE;
+
+ /* ConnectEx will copy supplied data to a internal buffer and send
+ * them during Fast Open or after connect. Therefore we can assumed
+ * this data already send. */
+ if (amount > 0) {
+ return amount;
+ }
+
+ _PR_MD_MAP_CONNECT_ERROR(WSAEWOULDBLOCK);
+ return -1;
+ }
+ // err is ERROR_IO_PENDING and socket is blocking, so query
+ // GetOverlappedResult.
+ err = ERROR_IO_INCOMPLETE;
+ while (err == ERROR_IO_INCOMPLETE) {
+ rv = socket_io_wait(osfd, WRITE_FD, timeout);
+ if ( rv < 0 ) {
+ return -1;
+ }
+ rv = GetOverlappedResult((HANDLE)osfd, &fd->secret->ol, &rvSent, FALSE);
+ if ( rv == TRUE ) {
+ return rvSent;
+ } else {
+ err = WSAGetLastError();
+ if (err != ERROR_IO_INCOMPLETE) {
+ _PR_MD_MAP_CONNECT_ERROR(err);
+ return -1;
+ }
+ }
+ }
+ }
+ return -1;
+}
+#endif
+
+PRInt32
+_PR_MD_RECVFROM(PRFileDesc *fd, void *buf, PRInt32 amount, PRIntn flags,
+ PRNetAddr *addr, PRUint32 *addrlen, PRIntervalTime timeout)
+{
+ PROsfd osfd = fd->secret->md.osfd;
+ PRInt32 rv, err;
+
+ while ((rv = recvfrom( osfd, buf, amount, 0, (struct sockaddr *) addr,
+ addrlen)) == -1)
+ {
+ if (((err = WSAGetLastError()) == WSAEWOULDBLOCK)
+ && (!fd->secret->nonblocking))
+ {
+ rv = socket_io_wait(osfd, READ_FD, timeout);
+ if ( rv < 0)
+ {
+ return -1;
+ }
+ }
+ else
+ {
+ _PR_MD_MAP_RECVFROM_ERROR(err);
+ break;
+ }
+ }
+ return(rv);
+}
+
+PRInt32
+_PR_MD_WRITEV(PRFileDesc *fd, const PRIOVec *iov, PRInt32 iov_size, PRIntervalTime timeout)
+{
+ int index;
+ int sent = 0;
+ int rv;
+
+ for (index=0; index < iov_size; index++)
+ {
+ rv = _PR_MD_SEND(fd, iov[index].iov_base, iov[index].iov_len, 0, timeout);
+ if (rv > 0) {
+ sent += rv;
+ }
+ if ( rv != iov[index].iov_len )
+ {
+ if (rv < 0)
+ {
+ if (fd->secret->nonblocking
+ && (PR_GetError() == PR_WOULD_BLOCK_ERROR)
+ && (sent > 0))
+ {
+ return sent;
+ }
+ else
+ {
+ return -1;
+ }
+ }
+ /* Only a nonblocking socket can have partial sends */
+ PR_ASSERT(fd->secret->nonblocking);
+ return sent;
+ }
+ }
+ return sent;
+}
+
+PRInt32
+_PR_MD_SHUTDOWN(PRFileDesc *fd, PRIntn how)
+{
+ PRInt32 rv;
+
+ rv = shutdown(fd->secret->md.osfd, how);
+ if (rv < 0) {
+ _PR_MD_MAP_SHUTDOWN_ERROR(WSAGetLastError());
+ }
+ return rv;
+}
+
+PRStatus
+_PR_MD_GETSOCKNAME(PRFileDesc *fd, PRNetAddr *addr, PRUint32 *len)
+{
+ PRInt32 rv;
+
+ rv = getsockname((SOCKET)fd->secret->md.osfd, (struct sockaddr *)addr, len);
+ if (rv==0) {
+ return PR_SUCCESS;
+ } else {
+ _PR_MD_MAP_GETSOCKNAME_ERROR(WSAGetLastError());
+ return PR_FAILURE;
+ }
+}
+
+PRStatus
+_PR_MD_GETPEERNAME(PRFileDesc *fd, PRNetAddr *addr, PRUint32 *len)
+{
+ PRInt32 rv;
+
+ rv = getpeername((SOCKET)fd->secret->md.osfd, (struct sockaddr *)addr, len);
+ if (rv==0) {
+ return PR_SUCCESS;
+ } else {
+ _PR_MD_MAP_GETPEERNAME_ERROR(WSAGetLastError());
+ return PR_FAILURE;
+ }
+}
+
+PRStatus
+_PR_MD_GETSOCKOPT(PRFileDesc *fd, PRInt32 level, PRInt32 optname, char* optval, PRInt32* optlen)
+{
+ PRInt32 rv;
+
+ rv = getsockopt((SOCKET)fd->secret->md.osfd, level, optname, optval, optlen);
+ if (rv==0) {
+ return PR_SUCCESS;
+ } else {
+ _PR_MD_MAP_GETSOCKOPT_ERROR(WSAGetLastError());
+ return PR_FAILURE;
+ }
+}
+
+PRStatus
+_PR_MD_SETSOCKOPT(PRFileDesc *fd, PRInt32 level, PRInt32 optname, const char* optval, PRInt32 optlen)
+{
+ PRInt32 rv;
+
+ rv = setsockopt((SOCKET)fd->secret->md.osfd, level, optname, optval, optlen);
+ if (rv==0) {
+ return PR_SUCCESS;
+ } else {
+ _PR_MD_MAP_SETSOCKOPT_ERROR(WSAGetLastError());
+ return PR_FAILURE;
+ }
+}
+
+void
+_MD_MakeNonblock(PRFileDesc *f)
+{
+ return; /* do nothing */
+}
+
+
+
+/*
+ * socket_io_wait --
+ *
+ * Wait for socket i/o, periodically checking for interrupt.
+ *
+ * This function returns 1 on success. On failure, it returns
+ * -1 and sets the error codes. It never returns 0.
+ */
+#define _PR_INTERRUPT_CHECK_INTERVAL_SECS 5
+
+static PRInt32 socket_io_wait(
+ PROsfd osfd,
+ PRInt32 fd_type,
+ PRIntervalTime timeout)
+{
+ PRInt32 rv = -1;
+ struct timeval tv;
+ PRThread *me = _PR_MD_CURRENT_THREAD();
+ PRIntervalTime elapsed, remaining;
+ PRBool wait_for_remaining;
+ fd_set rd_wr, ex;
+ int err, len;
+
+ switch (timeout) {
+ case PR_INTERVAL_NO_WAIT:
+ PR_SetError(PR_IO_TIMEOUT_ERROR, 0);
+ break;
+ case PR_INTERVAL_NO_TIMEOUT:
+ /*
+ * This is a special case of the 'default' case below.
+ * Please see the comments there.
+ */
+ tv.tv_sec = _PR_INTERRUPT_CHECK_INTERVAL_SECS;
+ tv.tv_usec = 0;
+ FD_ZERO(&rd_wr);
+ FD_ZERO(&ex);
+ do {
+ FD_SET(osfd, &rd_wr);
+ FD_SET(osfd, &ex);
+ switch( fd_type )
+ {
+ case READ_FD:
+ rv = _MD_SELECT(0, &rd_wr, NULL, NULL, &tv);
+ break;
+ case WRITE_FD:
+ rv = _MD_SELECT(0, NULL, &rd_wr, NULL, &tv);
+ break;
+ case CONNECT_FD:
+ rv = _MD_SELECT(0, NULL, &rd_wr, &ex, &tv);
+ break;
+ default:
+ PR_ASSERT(0);
+ break;
+ } /* end switch() */
+ if (rv == -1 )
+ {
+ _PR_MD_MAP_SELECT_ERROR(WSAGetLastError());
+ break;
+ }
+ if ( rv > 0 && fd_type == CONNECT_FD )
+ {
+ /*
+ * Call Sleep(0) to work around a Winsock timing bug.
+ */
+ Sleep(0);
+ if (FD_ISSET((SOCKET)osfd, &ex))
+ {
+ len = sizeof(err);
+ if (getsockopt(osfd, SOL_SOCKET, SO_ERROR,
+ (char *) &err, &len) == SOCKET_ERROR)
+ {
+ _PR_MD_MAP_GETSOCKOPT_ERROR(WSAGetLastError());
+ return -1;
+ }
+ if (err != 0) {
+ _PR_MD_MAP_CONNECT_ERROR(err);
+ }
+ else {
+ PR_SetError(PR_UNKNOWN_ERROR, 0);
+ }
+ return -1;
+ }
+ if (FD_ISSET((SOCKET)osfd, &rd_wr))
+ {
+ /* it's connected */
+ return 1;
+ }
+ PR_ASSERT(0);
+ }
+ if (_PR_PENDING_INTERRUPT(me)) {
+ me->flags &= ~_PR_INTERRUPT;
+ PR_SetError(PR_PENDING_INTERRUPT_ERROR, 0);
+ rv = -1;
+ break;
+ }
+ } while (rv == 0);
+ break;
+ default:
+ remaining = timeout;
+ FD_ZERO(&rd_wr);
+ FD_ZERO(&ex);
+ do {
+ /*
+ * We block in _MD_SELECT for at most
+ * _PR_INTERRUPT_CHECK_INTERVAL_SECS seconds,
+ * so that there is an upper limit on the delay
+ * before the interrupt bit is checked.
+ */
+ wait_for_remaining = PR_TRUE;
+ tv.tv_sec = PR_IntervalToSeconds(remaining);
+ if (tv.tv_sec > _PR_INTERRUPT_CHECK_INTERVAL_SECS) {
+ wait_for_remaining = PR_FALSE;
+ tv.tv_sec = _PR_INTERRUPT_CHECK_INTERVAL_SECS;
+ tv.tv_usec = 0;
+ } else {
+ tv.tv_usec = PR_IntervalToMicroseconds(
+ remaining -
+ PR_SecondsToInterval(tv.tv_sec));
+ }
+ FD_SET(osfd, &rd_wr);
+ FD_SET(osfd, &ex);
+ switch( fd_type )
+ {
+ case READ_FD:
+ rv = _MD_SELECT(0, &rd_wr, NULL, NULL, &tv);
+ break;
+ case WRITE_FD:
+ rv = _MD_SELECT(0, NULL, &rd_wr, NULL, &tv);
+ break;
+ case CONNECT_FD:
+ rv = _MD_SELECT(0, NULL, &rd_wr, &ex, &tv);
+ break;
+ default:
+ PR_ASSERT(0);
+ break;
+ } /* end switch() */
+ if (rv == -1)
+ {
+ _PR_MD_MAP_SELECT_ERROR(WSAGetLastError());
+ break;
+ }
+ if ( rv > 0 && fd_type == CONNECT_FD )
+ {
+ /*
+ * Call Sleep(0) to work around a Winsock timing bug.
+ */
+ Sleep(0);
+ if (FD_ISSET((SOCKET)osfd, &ex))
+ {
+ len = sizeof(err);
+ if (getsockopt(osfd, SOL_SOCKET, SO_ERROR,
+ (char *) &err, &len) == SOCKET_ERROR)
+ {
+ _PR_MD_MAP_GETSOCKOPT_ERROR(WSAGetLastError());
+ return -1;
+ }
+ if (err != 0) {
+ _PR_MD_MAP_CONNECT_ERROR(err);
+ }
+ else {
+ PR_SetError(PR_UNKNOWN_ERROR, 0);
+ }
+ return -1;
+ }
+ if (FD_ISSET((SOCKET)osfd, &rd_wr))
+ {
+ /* it's connected */
+ return 1;
+ }
+ PR_ASSERT(0);
+ }
+ if (_PR_PENDING_INTERRUPT(me)) {
+ me->flags &= ~_PR_INTERRUPT;
+ PR_SetError(PR_PENDING_INTERRUPT_ERROR, 0);
+ rv = -1;
+ break;
+ }
+ /*
+ * We loop again if _MD_SELECT timed out and the
+ * timeout deadline has not passed yet.
+ */
+ if (rv == 0 )
+ {
+ if (wait_for_remaining) {
+ elapsed = remaining;
+ } else {
+ elapsed = PR_SecondsToInterval(tv.tv_sec)
+ + PR_MicrosecondsToInterval(tv.tv_usec);
+ }
+ if (elapsed >= remaining) {
+ PR_SetError(PR_IO_TIMEOUT_ERROR, 0);
+ rv = -1;
+ break;
+ } else {
+ remaining = remaining - elapsed;
+ }
+ }
+ } while (rv == 0 );
+ break;
+ }
+ return(rv);
+} /* end socket_io_wait() */