uipc_socket2.c

ReactOS是一些高手根据Windows XP的内核编写出的类XP。内核实现机理和API函数调用几乎相同。甚至可以兼容XP的程序。喜欢研究系统内核的人可以看一看。

/*
 * Copyright (c) 1982, 1986, 1988, 1990, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	@(#)uipc_socket2.c	8.1 (Berkeley) 6/10/93
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/file.h>
#include <sys/buf.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/protosw.h>
#include <sys/stat.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/signalvar.h>
#include <oskittcp.h>

/*
 * Primitive routines for operating on sockets and socket buffers
 */

/* strings for sleep message: */
char	netio[] = "netio";
char	netcon[] = "netcon";
char	netcls[] = "netcls";

/* these two are sysctl visible */
u_long	sb_max = SB_MAX;		/* patchable */

u_long	sb_efficiency = 8;		/* parameter for sbreserve() */

int	sominqueue = 0;			/* minimum queue override */

/*
 * Procedures to manipulate state flags of socket
 * and do appropriate wakeups.  Normal sequence from the
 * active (originating) side is that soisconnecting() is
 * called during processing of connect() call,
 * resulting in an eventual call to soisconnected() if/when the
 * connection is established.  When the connection is torn down
 * soisdisconnecting() is called during processing of disconnect() call,
 * and soisdisconnected() is called when the connection to the peer
 * is totally severed.  The semantics of these routines are such that
 * connectionless protocols can call soisconnected() and soisdisconnected()
 * only, bypassing the in-progress calls when setting up a ``connection''
 * takes no time.
 *
 * From the passive side, a socket is created with
 * two queues of sockets: so_q0 for connections in progress
 * and so_q for connections already made and awaiting user acceptance.
 * As a protocol is preparing incoming connections, it creates a socket
 * structure queued on so_q0 by calling sonewconn().  When the connection
 * is established, soisconnected() is called, and transfers the
 * socket structure to so_q, making it available to accept().
 *
 * If a socket is closed with sockets on either
 * so_q0 or so_q, these sockets are dropped.
 *
 * If higher level protocols are implemented in
 * the kernel, the wakeups done here will sometimes
 * cause software-interrupt process scheduling.
 */

void
soisconnecting(so)
	register struct socket *so;
{
    OS_DbgPrint(OSK_MID_TRACE,("Called %x\n", so));
    so->so_state &= ~(SS_ISCONNECTED|SS_ISDISCONNECTING);
    so->so_state |= SS_ISCONNECTING;
}

void
soisconnected(so)
	register struct socket *so;
{
	register struct socket *head = so->so_head;

	OS_DbgPrint(OSK_MID_TRACE,("Called %x\n", so));

	so->so_state &= ~(SS_ISCONNECTING|SS_ISDISCONNECTING|SS_ISCONFIRMING);
	so->so_state |= SS_ISCONNECTED;
	if (head && soqremque(so, 0)) {
		soqinsque(head, so, 1);
		sorwakeup(head);
		wakeup(so, (caddr_t)&head->so_timeo);
	} else {
		wakeup(so, (caddr_t)&so->so_timeo);
		sorwakeup(so);
		sowwakeup(so);
	}
}

void
soisdisconnecting(so)
	register struct socket *so;
{
    OS_DbgPrint(OSK_MID_TRACE,("Called %x\n", so));
    so->so_state &= ~SS_ISCONNECTING;
    so->so_state |= (SS_ISDISCONNECTING|SS_CANTRCVMORE|SS_CANTSENDMORE);
    wakeup(so, (caddr_t)&so->so_timeo);
    sowwakeup(so);
    sorwakeup(so);
}

void
soisdisconnected(so)
	register struct socket *so;
{
    OS_DbgPrint(OSK_MID_TRACE,("Called %x\n", so));

    so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
    so->so_state |= (SS_CANTRCVMORE|SS_CANTSENDMORE);
    wakeup(so, (caddr_t)&so->so_timeo);
    sowwakeup(so);
    sorwakeup(so);
}

/*
 * When an attempt at a new connection is noted on a socket
 * which accepts connections, sonewconn is called.  If the
 * connection is possible (subject to space constraints, etc.)
 * then we allocate a new structure, propoerly linked into the
 * data structure of the original socket, and return this.
 * Connstatus may be 0, or SO_ISCONFIRMING, or SO_ISCONNECTED.
 *
 * Currently, sonewconn() is defined as sonewconn1() in socketvar.h
 * to catch calls that are missing the (new) second parameter.
 */
struct socket *
sonewconn1(head, connstatus)
	register struct socket *head;
	int connstatus;
{
	register struct socket *so;
	int soqueue = connstatus ? 1 : 0;

	OS_DbgPrint(OSK_MID_TRACE,("Called %x\n", head));

	if ((head->so_qlen + head->so_q0len > 3 * head->so_qlimit / 2) &&
	    (head->so_qlen + head->so_q0len > sominqueue))
		return ((struct socket *)0);
	MALLOC(so, struct socket *, sizeof(*so), M_SOCKET, M_DONTWAIT);
	if (so == NULL)
		return ((struct socket *)0);
	bzero((caddr_t)so, sizeof(*so));
	so->so_type = head->so_type;
	so->so_options = head->so_options &~ SO_ACCEPTCONN;
	so->so_linger = head->so_linger;
	so->so_state = head->so_state | SS_NOFDREF;
	so->so_proto = head->so_proto;
	so->so_timeo = head->so_timeo;
	so->so_pgid = head->so_pgid;
	(void) soreserve(so, head->so_snd.sb_hiwat, head->so_rcv.sb_hiwat);
	soqinsque(head, so, soqueue);
	if ((*so->so_proto->pr_usrreq)(so, PRU_ATTACH,
	    (struct mbuf *)0, (struct mbuf *)0, (struct mbuf *)0)) {
		(void) soqremque(so, soqueue);
		(void) free((caddr_t)so, M_SOCKET);
		return ((struct socket *)0);
	}
	if (connstatus) {
		sorwakeup(head);
		wakeup(head, (caddr_t)&head->so_timeo);
		so->so_state |= connstatus;
	}
	return (so);
}

void
soqinsque(head, so, q)
	register struct socket *head, *so;
	int q;
{

	register struct socket **prev;
	so->so_head = head;
	if (q == 0) {
		head->so_q0len++;
		so->so_q0 = 0;
		for (prev = &(head->so_q0); *prev; )
			prev = &((*prev)->so_q0);
	} else {
		head->so_qlen++;
		so->so_q = 0;
		for (prev = &(head->so_q); *prev; )
			prev = &((*prev)->so_q);
	}
	*prev = so;
}

int
soqremque(so, q)
	register struct socket *so;
	int q;
{
	register struct socket *head, *prev, *next;

	head = so->so_head;
	prev = head;
	for (;;) {
		next = q ? prev->so_q : prev->so_q0;
		if (next == so)
			break;
		if (next == 0)
			return (0);
		prev = next;
	}
	if (q == 0) {
		prev->so_q0 = next->so_q0;
		head->so_q0len--;
	} else {
		prev->so_q = next->so_q;
		head->so_qlen--;
	}
	next->so_q0 = next->so_q = 0;
	next->so_head = 0;
	return (1);
}

/*
 * Socantsendmore indicates that no more data will be sent on the
 * socket; it would normally be applied to a socket when the user
 * informs the system that no more data is to be sent, by the protocol
 * code (in case PRU_SHUTDOWN).  Socantrcvmore indicates that no more data
 * will be received, and will normally be applied to the socket by a
 * protocol when it detects that the peer will send no more data.
 * Data queued for reading in the socket may yet be read.
 */

void
socantsendmore(so)
	struct socket *so;
{

	so->so_state |= SS_CANTSENDMORE;
	sowwakeup(so);
}

void
socantrcvmore(so)
	struct socket *so;
{

	so->so_state |= SS_CANTRCVMORE;
	sorwakeup(so);
}

/*
 * Wait for data to arrive at/drain from a socket buffer.
 */
int
sbwait(sb)
	struct sockbuf *sb;
{

	sb->sb_flags |= SB_WAIT;
	return (tsleep((caddr_t)&sb->sb_cc,
	    (sb->sb_flags & SB_NOINTR) ? PSOCK : PSOCK | PCATCH, netio,
	    sb->sb_timeo));
}

/*
 * Lock a sockbuf already known to be locked;
 * return any error returned from sleep (EINTR).
 */
int
sb_lock(sb)
	register struct sockbuf *sb;
{
	int error;

	while (sb->sb_flags & SB_LOCK) {
		sb->sb_flags |= SB_WANT;
		error = tsleep((caddr_t)&sb->sb_flags,
		    (sb->sb_flags & SB_NOINTR) ? PSOCK : PSOCK|PCATCH,
		    netio, 0);
		if (error)
			return (error);
	}
	sb->sb_flags |= SB_LOCK;
	return (0);
}

/*
 * Wakeup processes waiting on a socket buffer.
 * Do asynchronous notification via SIGIO
 * if the socket has the SS_ASYNC flag set.
 */
void
sowakeup(so, sb)
	register struct socket *so;
	register struct sockbuf *sb;
{
	struct proc *p;

	wakeup(so, &sb->sb_sel);
#ifndef OSKIT
	/*
	 * in the OS Kit, we do not want notifications to stop
	 */
	sb->sb_flags &= ~SB_SEL;
#endif
	if (sb->sb_flags & SB_WAIT) {
		sb->sb_flags &= ~SB_WAIT;
		wakeup(so, (caddr_t)&sb->sb_cc);
	}
	if (so->so_state & SS_ASYNC) {
		if (so->so_pgid < 0)
			gsignal(-so->so_pgid, SIGIO);
		else if (so->so_pgid > 0 && (p = pfind(so->so_pgid)) != 0)
			psignal(p, SIGIO);
	}
}

/*
 * Socket buffer (struct sockbuf) utility routines.
 *
 * Each socket contains two socket buffers: one for sending data and
 * one for receiving data.  Each buffer contains a queue of mbufs,
 * information about the number of mbufs and amount of data in the
 * queue, and other fields allowing select() statements and notification
 * on data availability to be implemented.
 *
 * Data stored in a socket buffer is maintained as a list of records.
 * Each record is a list of mbufs chained together with the m_next
 * field.  Records are chained together with the m_nextpkt field. The upper
 * level routine soreceive() expects the following conventions to be
 * observed when placing information in the receive buffer:
 *
 * 1. If the protocol requires each message be preceded by the sender's
 *    name, then a record containing that name must be present before
 *    any associated data (mbuf's must be of type MT_SONAME).
 * 2. If the protocol supports the exchange of ``access rights'' (really
 *    just additional data associated with the message), and there are
 *    ``rights'' to be received, then a record containing this data
 *    should be present (mbuf's must be of type MT_RIGHTS).
 * 3. If a name or rights record exists, then it must be followed by
 *    a data record, perhaps of zero length.
 *
 * Before using a new socket structure it is first necessary to reserve
 * buffer space to the socket, by calling sbreserve().  This should commit
 * some of the available buffer space in the system buffer pool for the
 * socket (currently, it does nothing but enforce limits).  The space
 * should be released by calling sbrelease() when the socket is destroyed.
 */

int
soreserve(so, sndcc, rcvcc)
	register struct socket *so;
	u_long sndcc, rcvcc;
{

	if (sbreserve(&so->so_snd, sndcc) == 0)
		goto bad;
	if (sbreserve(&so->so_rcv, rcvcc) == 0)
		goto bad2;
	if (so->so_rcv.sb_lowat == 0)
		so->so_rcv.sb_lowat = 1;
	if (so->so_snd.sb_lowat == 0)
		so->so_snd.sb_lowat = MCLBYTES;
	if (so->so_snd.sb_lowat > so->so_snd.sb_hiwat)
		so->so_snd.sb_lowat = so->so_snd.sb_hiwat;
	return (0);
bad2:
	sbrelease(&so->so_snd);
bad:
	return (ENOBUFS);
}

/*
 * Allot mbufs to a sockbuf.
 * Attempt to scale mbmax so that mbcnt doesn't become limiting
 * if buffering efficiency is near the normal case.