NetBSD/sys/kern/uipc_socket.c

1137 lines
26 KiB
C
Raw Normal View History

2000-03-30 13:27:11 +04:00
/* $NetBSD: uipc_socket.c,v 1.50 2000/03/30 09:27:14 augustss Exp $ */
1993-03-21 12:45:37 +03:00
/*
* Copyright (c) 1982, 1986, 1988, 1990, 1993
* The Regents of the University of California. All rights reserved.
1993-03-21 12:45:37 +03:00
*
* 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.
*
1998-03-01 05:20:01 +03:00
* @(#)uipc_socket.c 8.6 (Berkeley) 5/2/95
1993-03-21 12:45:37 +03:00
*/
1998-06-26 03:40:33 +04:00
#include "opt_compat_sunos.h"
1993-12-18 07:21:37 +03:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/file.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/domain.h>
#include <sys/kernel.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
1996-02-04 05:17:43 +03:00
#include <sys/signalvar.h>
1993-12-18 07:21:37 +03:00
#include <sys/resourcevar.h>
1998-08-02 08:53:11 +04:00
#include <sys/pool.h>
struct pool socket_pool;
extern int somaxconn; /* patchable (XXX sysctl) */
int somaxconn = SOMAXCONN;
1998-08-02 08:53:11 +04:00
void
soinit()
{
pool_init(&socket_pool, sizeof(struct socket), 0, 0, 0,
"sockpl", 0, NULL, NULL, M_SOCKET);
}
1993-03-21 12:45:37 +03:00
/*
* Socket operation routines.
* These routines are called by the routines in
* sys_socket.c or from a system process, and
* implement the semantics of socket operations by
* switching out to the protocol specific routines.
*/
/*ARGSUSED*/
int
1993-03-21 12:45:37 +03:00
socreate(dom, aso, type, proto)
1994-04-25 12:22:07 +04:00
int dom;
1993-03-21 12:45:37 +03:00
struct socket **aso;
2000-03-30 13:27:11 +04:00
int type;
1993-03-21 12:45:37 +03:00
int proto;
{
struct proc *p = curproc; /* XXX */
2000-03-30 13:27:11 +04:00
struct protosw *prp;
struct socket *so;
int error;
int s;
1993-03-21 12:45:37 +03:00
if (proto)
prp = pffindproto(dom, proto, type);
else
prp = pffindtype(dom, type);
if (prp == 0 || prp->pr_usrreq == 0)
1993-03-21 12:45:37 +03:00
return (EPROTONOSUPPORT);
if (prp->pr_type != type)
return (EPROTOTYPE);
s = splsoftnet();
1998-08-02 08:53:11 +04:00
so = pool_get(&socket_pool, PR_WAITOK);
memset((caddr_t)so, 0, sizeof(*so));
TAILQ_INIT(&so->so_q0);
TAILQ_INIT(&so->so_q);
1993-03-21 12:45:37 +03:00
so->so_type = type;
so->so_proto = prp;
so->so_send = sosend;
so->so_receive = soreceive;
if (p != 0)
so->so_uid = p->p_ucred->cr_uid;
error = (*prp->pr_usrreq)(so, PRU_ATTACH, (struct mbuf *)0,
(struct mbuf *)(long)proto, (struct mbuf *)0, p);
1993-03-21 12:45:37 +03:00
if (error) {
so->so_state |= SS_NOFDREF;
sofree(so);
splx(s);
1993-03-21 12:45:37 +03:00
return (error);
}
1994-01-23 09:06:21 +03:00
#ifdef COMPAT_SUNOS
{
extern struct emul emul_sunos;
if (p->p_emul == &emul_sunos && type == SOCK_DGRAM)
so->so_options |= SO_BROADCAST;
}
1994-01-23 09:06:21 +03:00
#endif
splx(s);
1993-03-21 12:45:37 +03:00
*aso = so;
return (0);
}
int
1993-03-21 12:45:37 +03:00
sobind(so, nam)
struct socket *so;
struct mbuf *nam;
{
struct proc *p = curproc; /* XXX */
1995-08-13 03:59:09 +04:00
int s = splsoftnet();
1993-03-21 12:45:37 +03:00
int error;
error = (*so->so_proto->pr_usrreq)(so, PRU_BIND, (struct mbuf *)0,
nam, (struct mbuf *)0, p);
1993-03-21 12:45:37 +03:00
splx(s);
return (error);
}
int
1993-03-21 12:45:37 +03:00
solisten(so, backlog)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
int backlog;
{
1995-08-13 03:59:09 +04:00
int s = splsoftnet(), error;
1993-03-21 12:45:37 +03:00
error = (*so->so_proto->pr_usrreq)(so, PRU_LISTEN, (struct mbuf *)0,
(struct mbuf *)0, (struct mbuf *)0, (struct proc *)0);
1993-03-21 12:45:37 +03:00
if (error) {
splx(s);
return (error);
}
if (so->so_q.tqh_first == NULL)
1993-03-21 12:45:37 +03:00
so->so_options |= SO_ACCEPTCONN;
if (backlog < 0)
backlog = 0;
so->so_qlimit = min(backlog, somaxconn);
1993-03-21 12:45:37 +03:00
splx(s);
return (0);
}
1996-02-04 05:17:43 +03:00
void
1993-03-21 12:45:37 +03:00
sofree(so)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
{
if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0)
1993-03-21 12:45:37 +03:00
return;
if (so->so_head) {
/*
* We must not decommission a socket that's on the accept(2)
* queue. If we do, then accept(2) may hang after select(2)
* indicated that the listening socket was ready.
*/
if (!soqremque(so, 0))
return;
}
1993-03-21 12:45:37 +03:00
sbrelease(&so->so_snd);
sorflush(so);
1998-08-02 08:53:11 +04:00
pool_put(&socket_pool, so);
1993-03-21 12:45:37 +03:00
}
/*
* Close a socket on last file table reference removal.
* Initiate disconnect if connected.
* Free socket when disconnect complete.
*/
int
1993-03-21 12:45:37 +03:00
soclose(so)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
{
struct socket *so2;
1995-08-13 03:59:09 +04:00
int s = splsoftnet(); /* conservative */
1993-03-21 12:45:37 +03:00
int error = 0;
if (so->so_options & SO_ACCEPTCONN) {
while ((so2 = so->so_q0.tqh_first) != 0) {
1999-01-20 23:24:12 +03:00
(void) soqremque(so2, 0);
(void) soabort(so2);
}
while ((so2 = so->so_q.tqh_first) != 0) {
1999-01-20 23:24:12 +03:00
(void) soqremque(so2, 1);
(void) soabort(so2);
}
1993-03-21 12:45:37 +03:00
}
if (so->so_pcb == 0)
goto discard;
if (so->so_state & SS_ISCONNECTED) {
if ((so->so_state & SS_ISDISCONNECTING) == 0) {
error = sodisconnect(so);
if (error)
goto drop;
}
if (so->so_options & SO_LINGER) {
if ((so->so_state & SS_ISDISCONNECTING) &&
(so->so_state & SS_NBIO))
goto drop;
1996-02-04 05:17:43 +03:00
while (so->so_state & SS_ISCONNECTED) {
error = tsleep((caddr_t)&so->so_timeo,
PSOCK | PCATCH, netcls,
so->so_linger * hz);
1996-02-04 05:17:43 +03:00
if (error)
1993-03-21 12:45:37 +03:00
break;
1996-02-04 05:17:43 +03:00
}
1993-03-21 12:45:37 +03:00
}
}
drop:
if (so->so_pcb) {
int error2 = (*so->so_proto->pr_usrreq)(so, PRU_DETACH,
(struct mbuf *)0, (struct mbuf *)0, (struct mbuf *)0,
(struct proc *)0);
1993-03-21 12:45:37 +03:00
if (error == 0)
error = error2;
}
discard:
if (so->so_state & SS_NOFDREF)
panic("soclose: NOFDREF");
so->so_state |= SS_NOFDREF;
sofree(so);
splx(s);
return (error);
}
/*
1995-08-13 03:59:09 +04:00
* Must be called at splsoftnet...
1993-03-21 12:45:37 +03:00
*/
int
1993-03-21 12:45:37 +03:00
soabort(so)
struct socket *so;
{
return (*so->so_proto->pr_usrreq)(so, PRU_ABORT, (struct mbuf *)0,
(struct mbuf *)0, (struct mbuf *)0, (struct proc *)0);
1993-03-21 12:45:37 +03:00
}
int
1993-03-21 12:45:37 +03:00
soaccept(so, nam)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
struct mbuf *nam;
{
1995-08-13 03:59:09 +04:00
int s = splsoftnet();
1993-03-21 12:45:37 +03:00
int error;
if ((so->so_state & SS_NOFDREF) == 0)
panic("soaccept: !NOFDREF");
so->so_state &= ~SS_NOFDREF;
if ((so->so_state & SS_ISDISCONNECTED) == 0)
error = (*so->so_proto->pr_usrreq)(so, PRU_ACCEPT,
(struct mbuf *)0, nam, (struct mbuf *)0, (struct proc *)0);
else
error = 0;
1993-03-21 12:45:37 +03:00
splx(s);
return (error);
}
int
1993-03-21 12:45:37 +03:00
soconnect(so, nam)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
struct mbuf *nam;
{
1996-05-22 23:00:52 +04:00
struct proc *p = curproc; /* XXX */
1993-03-21 12:45:37 +03:00
int s;
int error;
if (so->so_options & SO_ACCEPTCONN)
return (EOPNOTSUPP);
1995-08-13 03:59:09 +04:00
s = splsoftnet();
1993-03-21 12:45:37 +03:00
/*
* If protocol is connection-based, can only connect once.
* Otherwise, if connected, try to disconnect first.
* This allows user to disconnect by connecting to, e.g.,
* a null address.
*/
if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) &&
((so->so_proto->pr_flags & PR_CONNREQUIRED) ||
(error = sodisconnect(so))))
error = EISCONN;
else
error = (*so->so_proto->pr_usrreq)(so, PRU_CONNECT,
1996-05-22 23:00:52 +04:00
(struct mbuf *)0, nam, (struct mbuf *)0, p);
1993-03-21 12:45:37 +03:00
splx(s);
return (error);
}
int
1993-03-21 12:45:37 +03:00
soconnect2(so1, so2)
2000-03-30 13:27:11 +04:00
struct socket *so1;
1993-03-21 12:45:37 +03:00
struct socket *so2;
{
1995-08-13 03:59:09 +04:00
int s = splsoftnet();
1993-03-21 12:45:37 +03:00
int error;
error = (*so1->so_proto->pr_usrreq)(so1, PRU_CONNECT2,
(struct mbuf *)0, (struct mbuf *)so2, (struct mbuf *)0,
(struct proc *)0);
1993-03-21 12:45:37 +03:00
splx(s);
return (error);
}
int
1993-03-21 12:45:37 +03:00
sodisconnect(so)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
{
1995-08-13 03:59:09 +04:00
int s = splsoftnet();
1993-03-21 12:45:37 +03:00
int error;
if ((so->so_state & SS_ISCONNECTED) == 0) {
error = ENOTCONN;
goto bad;
}
if (so->so_state & SS_ISDISCONNECTING) {
error = EALREADY;
goto bad;
}
error = (*so->so_proto->pr_usrreq)(so, PRU_DISCONNECT,
(struct mbuf *)0, (struct mbuf *)0, (struct mbuf *)0,
(struct proc *)0);
1993-03-21 12:45:37 +03:00
bad:
splx(s);
return (error);
}
#define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK)
1993-03-21 12:45:37 +03:00
/*
* Send on a socket.
* If send must go all at once and message is larger than
* send buffering, then hard error.
* Lock against other senders.
* If must go all at once and not enough room now, then
* inform user that this would block and do nothing.
* Otherwise, if nonblocking, send as much as possible.
* The data to be sent is described by "uio" if nonzero,
* otherwise by the mbuf chain "top" (which must be null
* if uio is not). Data provided in mbuf chain must be small
* enough to send all at once.
*
* Returns nonzero on error, timeout or signal; callers
* must check for short counts if EINTR/ERESTART are returned.
* Data and control buffers are freed on return.
*/
int
1993-03-21 12:45:37 +03:00
sosend(so, addr, uio, top, control, flags)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
struct mbuf *addr;
struct uio *uio;
struct mbuf *top;
struct mbuf *control;
int flags;
{
struct proc *p = curproc; /* XXX */
1993-03-21 12:45:37 +03:00
struct mbuf **mp;
2000-03-30 13:27:11 +04:00
struct mbuf *m;
long space, len, resid;
1993-03-21 12:45:37 +03:00
int clen = 0, error, s, dontroute, mlen;
int atomic = sosendallatonce(so) || top;
if (uio)
resid = uio->uio_resid;
else
resid = top->m_pkthdr.len;
/*
* In theory resid should be unsigned.
* However, space must be signed, as it might be less than 0
* if we over-committed, and we must use a signed comparison
* of space and resid. On the other hand, a negative resid
* causes us to loop sending 0-length segments to the protocol.
*/
if (resid < 0) {
error = EINVAL;
goto out;
}
1993-03-21 12:45:37 +03:00
dontroute =
(flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 &&
(so->so_proto->pr_flags & PR_ATOMIC);
1994-04-25 12:41:03 +04:00
p->p_stats->p_ru.ru_msgsnd++;
1993-03-21 12:45:37 +03:00
if (control)
clen = control->m_len;
#define snderr(errno) { error = errno; splx(s); goto release; }
restart:
1996-02-04 05:17:43 +03:00
if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0)
1993-03-21 12:45:37 +03:00
goto out;
do {
1995-08-13 03:59:09 +04:00
s = splsoftnet();
1993-03-21 12:45:37 +03:00
if (so->so_state & SS_CANTSENDMORE)
snderr(EPIPE);
if (so->so_error) {
error = so->so_error;
so->so_error = 0;
splx(s);
goto release;
}
1993-03-21 12:45:37 +03:00
if ((so->so_state & SS_ISCONNECTED) == 0) {
if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
if ((so->so_state & SS_ISCONFIRMING) == 0 &&
!(resid == 0 && clen != 0))
snderr(ENOTCONN);
} else if (addr == 0)
snderr(EDESTADDRREQ);
}
space = sbspace(&so->so_snd);
if (flags & MSG_OOB)
space += 1024;
1996-02-04 05:17:43 +03:00
if ((atomic && resid > so->so_snd.sb_hiwat) ||
1994-04-25 12:22:07 +04:00
clen > so->so_snd.sb_hiwat)
snderr(EMSGSIZE);
if (space < resid + clen && uio &&
1993-03-21 12:45:37 +03:00
(atomic || space < so->so_snd.sb_lowat || space < clen)) {
if (so->so_state & SS_NBIO)
snderr(EWOULDBLOCK);
sbunlock(&so->so_snd);
error = sbwait(&so->so_snd);
splx(s);
if (error)
goto out;
goto restart;
}
splx(s);
mp = &top;
space -= clen;
do {
if (uio == NULL) {
/*
* Data is prepackaged in "top".
*/
resid = 0;
if (flags & MSG_EOR)
top->m_flags |= M_EOR;
} else do {
if (top == 0) {
MGETHDR(m, M_WAIT, MT_DATA);
mlen = MHLEN;
m->m_pkthdr.len = 0;
m->m_pkthdr.rcvif = (struct ifnet *)0;
} else {
MGET(m, M_WAIT, MT_DATA);
mlen = MLEN;
}
if (resid >= MINCLSIZE && space >= MCLBYTES) {
MCLGET(m, M_WAIT);
if ((m->m_flags & M_EXT) == 0)
goto nopages;
mlen = MCLBYTES;
#ifdef MAPPED_MBUFS
len = min(MCLBYTES, resid);
#else
if (atomic && top == 0) {
len = min(MCLBYTES - max_hdr, resid);
m->m_data += max_hdr;
} else
len = min(MCLBYTES, resid);
#endif
space -= len;
} else {
1993-03-21 12:45:37 +03:00
nopages:
len = min(min(mlen, resid), space);
space -= len;
/*
* For datagram protocols, leave room
* for protocol headers in first mbuf.
*/
if (atomic && top == 0 && len < mlen)
MH_ALIGN(m, len);
}
error = uiomove(mtod(m, caddr_t), (int)len, uio);
resid = uio->uio_resid;
m->m_len = len;
*mp = m;
top->m_pkthdr.len += len;
if (error)
goto release;
mp = &m->m_next;
if (resid <= 0) {
if (flags & MSG_EOR)
top->m_flags |= M_EOR;
break;
}
} while (space > 0 && atomic);
s = splsoftnet();
if (so->so_state & SS_CANTSENDMORE)
snderr(EPIPE);
if (dontroute)
so->so_options |= SO_DONTROUTE;
if (resid > 0)
so->so_state |= SS_MORETOCOME;
error = (*so->so_proto->pr_usrreq)(so,
(flags & MSG_OOB) ? PRU_SENDOOB : PRU_SEND,
top, addr, control, p);
if (dontroute)
so->so_options &= ~SO_DONTROUTE;
if (resid > 0)
so->so_state &= ~SS_MORETOCOME;
splx(s);
clen = 0;
control = 0;
top = 0;
mp = &top;
1993-03-21 12:45:37 +03:00
if (error)
goto release;
} while (resid && space > 0);
} while (resid);
release:
sbunlock(&so->so_snd);
out:
if (top)
m_freem(top);
if (control)
m_freem(control);
return (error);
}
/*
* Implement receive operations on a socket.
* We depend on the way that records are added to the sockbuf
* by sbappend*. In particular, each record (mbufs linked through m_next)
* must begin with an address if the protocol so specifies,
* followed by an optional mbuf or mbufs containing ancillary data,
* and then zero or more mbufs of data.
* In order to avoid blocking network interrupts for the entire time here,
* we splx() while doing the actual copy to user space.
* Although the sockbuf is locked, new data may still be appended,
* and thus we must maintain consistency of the sockbuf during that time.
*
* The caller may receive the data as a single mbuf chain by supplying
* an mbuf **mp0 for use in returning the chain. The uio is then used
* only for the count in uio_resid.
*/
int
1993-03-21 12:45:37 +03:00
soreceive(so, paddr, uio, mp0, controlp, flagsp)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
struct mbuf **paddr;
struct uio *uio;
struct mbuf **mp0;
struct mbuf **controlp;
int *flagsp;
{
2000-03-30 13:27:11 +04:00
struct mbuf *m, **mp;
int flags, len, error, s, offset;
1993-03-21 12:45:37 +03:00
struct protosw *pr = so->so_proto;
struct mbuf *nextrecord;
1996-02-04 05:17:43 +03:00
int moff, type = 0;
int orig_resid = uio->uio_resid;
1993-03-21 12:45:37 +03:00
mp = mp0;
if (paddr)
*paddr = 0;
if (controlp)
*controlp = 0;
if (flagsp)
flags = *flagsp &~ MSG_EOR;
else
flags = 0;
if (flags & MSG_OOB) {
m = m_get(M_WAIT, MT_DATA);
error = (*pr->pr_usrreq)(so, PRU_RCVOOB, m,
(struct mbuf *)(long)(flags & MSG_PEEK), (struct mbuf *)0,
(struct proc *)0);
1993-03-21 12:45:37 +03:00
if (error)
goto bad;
do {
error = uiomove(mtod(m, caddr_t),
(int) min(uio->uio_resid, m->m_len), uio);
m = m_free(m);
} while (uio->uio_resid && error == 0 && m);
bad:
if (m)
m_freem(m);
return (error);
}
if (mp)
*mp = (struct mbuf *)0;
if (so->so_state & SS_ISCONFIRMING && uio->uio_resid)
(*pr->pr_usrreq)(so, PRU_RCVD, (struct mbuf *)0,
(struct mbuf *)0, (struct mbuf *)0, (struct proc *)0);
1993-03-21 12:45:37 +03:00
restart:
1996-02-04 05:17:43 +03:00
if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0)
1993-03-21 12:45:37 +03:00
return (error);
1995-08-13 03:59:09 +04:00
s = splsoftnet();
1993-03-21 12:45:37 +03:00
m = so->so_rcv.sb_mb;
/*
* If we have less data than requested, block awaiting more
* (subject to any timeout) if:
* 1. the current count is less than the low water mark,
1993-03-21 12:45:37 +03:00
* 2. MSG_WAITALL is set, and it is possible to do the entire
* receive operation at once if we block (resid <= hiwat), or
* 3. MSG_DONTWAIT is not set.
1993-03-21 12:45:37 +03:00
* If MSG_WAITALL is set but resid is larger than the receive buffer,
* we have to do the receive in sections, and thus risk returning
* a short count if a timeout or signal occurs after we start.
*/
1996-02-04 05:17:43 +03:00
if (m == 0 || (((flags & MSG_DONTWAIT) == 0 &&
so->so_rcv.sb_cc < uio->uio_resid) &&
1993-03-21 12:45:37 +03:00
(so->so_rcv.sb_cc < so->so_rcv.sb_lowat ||
((flags & MSG_WAITALL) && uio->uio_resid <= so->so_rcv.sb_hiwat)) &&
1996-02-04 05:17:43 +03:00
m->m_nextpkt == 0 && (pr->pr_flags & PR_ATOMIC) == 0)) {
1993-03-21 12:45:37 +03:00
#ifdef DIAGNOSTIC
if (m == 0 && so->so_rcv.sb_cc)
panic("receive 1");
#endif
if (so->so_error) {
if (m)
goto dontblock;
1993-03-21 12:45:37 +03:00
error = so->so_error;
if ((flags & MSG_PEEK) == 0)
so->so_error = 0;
goto release;
}
if (so->so_state & SS_CANTRCVMORE) {
if (m)
goto dontblock;
1993-03-21 12:45:37 +03:00
else
goto release;
}
for (; m; m = m->m_next)
if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) {
m = so->so_rcv.sb_mb;
goto dontblock;
}
if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
(so->so_proto->pr_flags & PR_CONNREQUIRED)) {
error = ENOTCONN;
goto release;
}
if (uio->uio_resid == 0)
goto release;
if ((so->so_state & SS_NBIO) || (flags & MSG_DONTWAIT)) {
1993-03-21 12:45:37 +03:00
error = EWOULDBLOCK;
goto release;
}
sbunlock(&so->so_rcv);
error = sbwait(&so->so_rcv);
splx(s);
if (error)
return (error);
goto restart;
}
dontblock:
#ifdef notyet /* XXXX */
if (uio->uio_procp)
uio->uio_procp->p_stats->p_ru.ru_msgrcv++;
#endif
1993-03-21 12:45:37 +03:00
nextrecord = m->m_nextpkt;
if (pr->pr_flags & PR_ADDR) {
#ifdef DIAGNOSTIC
if (m->m_type != MT_SONAME)
panic("receive 1a");
#endif
orig_resid = 0;
1993-03-21 12:45:37 +03:00
if (flags & MSG_PEEK) {
if (paddr)
*paddr = m_copy(m, 0, m->m_len);
m = m->m_next;
} else {
sbfree(&so->so_rcv, m);
if (paddr) {
*paddr = m;
so->so_rcv.sb_mb = m->m_next;
m->m_next = 0;
m = so->so_rcv.sb_mb;
} else {
MFREE(m, so->so_rcv.sb_mb);
m = so->so_rcv.sb_mb;
}
}
}
while (m && m->m_type == MT_CONTROL && error == 0) {
if (flags & MSG_PEEK) {
if (controlp)
*controlp = m_copy(m, 0, m->m_len);
m = m->m_next;
} else {
sbfree(&so->so_rcv, m);
if (controlp) {
if (pr->pr_domain->dom_externalize &&
mtod(m, struct cmsghdr *)->cmsg_type ==
SCM_RIGHTS)
error = (*pr->pr_domain->dom_externalize)(m);
1993-03-21 12:45:37 +03:00
*controlp = m;
so->so_rcv.sb_mb = m->m_next;
m->m_next = 0;
m = so->so_rcv.sb_mb;
} else {
MFREE(m, so->so_rcv.sb_mb);
m = so->so_rcv.sb_mb;
}
}
if (controlp) {
orig_resid = 0;
1993-03-21 12:45:37 +03:00
controlp = &(*controlp)->m_next;
}
1993-03-21 12:45:37 +03:00
}
if (m) {
if ((flags & MSG_PEEK) == 0)
m->m_nextpkt = nextrecord;
type = m->m_type;
if (type == MT_OOBDATA)
flags |= MSG_OOB;
}
moff = 0;
offset = 0;
while (m && uio->uio_resid > 0 && error == 0) {
if (m->m_type == MT_OOBDATA) {
if (type != MT_OOBDATA)
break;
} else if (type == MT_OOBDATA)
break;
#ifdef DIAGNOSTIC
else if (m->m_type != MT_DATA && m->m_type != MT_HEADER)
panic("receive 3");
#endif
so->so_state &= ~SS_RCVATMARK;
len = uio->uio_resid;
if (so->so_oobmark && len > so->so_oobmark - offset)
len = so->so_oobmark - offset;
if (len > m->m_len - moff)
len = m->m_len - moff;
/*
* If mp is set, just pass back the mbufs.
* Otherwise copy them out via the uio, then free.
* Sockbuf must be consistent here (points to current mbuf,
* it points to next record) when we drop priority;
* we must note any additions to the sockbuf when we
* block interrupts again.
*/
if (mp == 0) {
splx(s);
error = uiomove(mtod(m, caddr_t) + moff, (int)len, uio);
1995-08-13 03:59:09 +04:00
s = splsoftnet();
1993-03-21 12:45:37 +03:00
} else
uio->uio_resid -= len;
if (len == m->m_len - moff) {
if (m->m_flags & M_EOR)
flags |= MSG_EOR;
if (flags & MSG_PEEK) {
m = m->m_next;
moff = 0;
} else {
nextrecord = m->m_nextpkt;
sbfree(&so->so_rcv, m);
if (mp) {
*mp = m;
mp = &m->m_next;
so->so_rcv.sb_mb = m = m->m_next;
*mp = (struct mbuf *)0;
} else {
MFREE(m, so->so_rcv.sb_mb);
m = so->so_rcv.sb_mb;
}
if (m)
m->m_nextpkt = nextrecord;
}
} else {
if (flags & MSG_PEEK)
moff += len;
else {
if (mp)
*mp = m_copym(m, 0, len, M_WAIT);
m->m_data += len;
m->m_len -= len;
so->so_rcv.sb_cc -= len;
}
}
if (so->so_oobmark) {
if ((flags & MSG_PEEK) == 0) {
so->so_oobmark -= len;
if (so->so_oobmark == 0) {
so->so_state |= SS_RCVATMARK;
break;
}
} else {
1993-03-21 12:45:37 +03:00
offset += len;
if (offset == so->so_oobmark)
break;
}
1993-03-21 12:45:37 +03:00
}
if (flags & MSG_EOR)
break;
/*
* If the MSG_WAITALL flag is set (for non-atomic socket),
* we must not quit until "uio->uio_resid == 0" or an error
* termination. If a signal/timeout occurs, return
* with a short count but without error.
* Keep sockbuf locked against other readers.
*/
while (flags & MSG_WAITALL && m == 0 && uio->uio_resid > 0 &&
!sosendallatonce(so) && !nextrecord) {
1993-03-21 12:45:37 +03:00
if (so->so_error || so->so_state & SS_CANTRCVMORE)
break;
error = sbwait(&so->so_rcv);
if (error) {
sbunlock(&so->so_rcv);
splx(s);
return (0);
}
1996-02-04 05:17:43 +03:00
if ((m = so->so_rcv.sb_mb) != NULL)
1993-03-21 12:45:37 +03:00
nextrecord = m->m_nextpkt;
}
}
if (m && pr->pr_flags & PR_ATOMIC) {
flags |= MSG_TRUNC;
if ((flags & MSG_PEEK) == 0)
(void) sbdroprecord(&so->so_rcv);
}
1993-03-21 12:45:37 +03:00
if ((flags & MSG_PEEK) == 0) {
if (m == 0)
so->so_rcv.sb_mb = nextrecord;
if (pr->pr_flags & PR_WANTRCVD && so->so_pcb)
(*pr->pr_usrreq)(so, PRU_RCVD, (struct mbuf *)0,
(struct mbuf *)(long)flags, (struct mbuf *)0,
(struct proc *)0);
1993-03-21 12:45:37 +03:00
}
if (orig_resid == uio->uio_resid && orig_resid &&
(flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) {
sbunlock(&so->so_rcv);
splx(s);
goto restart;
}
1993-03-21 12:45:37 +03:00
if (flagsp)
*flagsp |= flags;
release:
sbunlock(&so->so_rcv);
splx(s);
return (error);
}
1994-05-04 15:24:06 +04:00
int
1993-03-21 12:45:37 +03:00
soshutdown(so, how)
struct socket *so;
int how;
1993-03-21 12:45:37 +03:00
{
struct protosw *pr = so->so_proto;
if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR))
return (EINVAL);
1993-03-21 12:45:37 +03:00
if (how == SHUT_RD || how == SHUT_RDWR)
1993-03-21 12:45:37 +03:00
sorflush(so);
if (how == SHUT_WR || how == SHUT_RDWR)
return (*pr->pr_usrreq)(so, PRU_SHUTDOWN, (struct mbuf *)0,
(struct mbuf *)0, (struct mbuf *)0, (struct proc *)0);
1993-03-21 12:45:37 +03:00
return (0);
}
1994-05-04 15:24:06 +04:00
void
1993-03-21 12:45:37 +03:00
sorflush(so)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
{
2000-03-30 13:27:11 +04:00
struct sockbuf *sb = &so->so_rcv;
struct protosw *pr = so->so_proto;
int s;
1993-03-21 12:45:37 +03:00
struct sockbuf asb;
sb->sb_flags |= SB_NOINTR;
(void) sblock(sb, M_WAITOK);
1993-03-21 12:45:37 +03:00
s = splimp();
socantrcvmore(so);
sbunlock(sb);
asb = *sb;
memset((caddr_t)sb, 0, sizeof(*sb));
1993-03-21 12:45:37 +03:00
splx(s);
if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose)
(*pr->pr_domain->dom_dispose)(asb.sb_mb);
sbrelease(&asb);
}
1994-05-04 15:24:06 +04:00
int
1993-03-21 12:45:37 +03:00
sosetopt(so, level, optname, m0)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
int level, optname;
struct mbuf *m0;
{
int error = 0;
2000-03-30 13:27:11 +04:00
struct mbuf *m = m0;
1993-03-21 12:45:37 +03:00
if (level != SOL_SOCKET) {
if (so->so_proto && so->so_proto->pr_ctloutput)
return ((*so->so_proto->pr_ctloutput)
(PRCO_SETOPT, so, level, optname, &m0));
error = ENOPROTOOPT;
} else {
switch (optname) {
case SO_LINGER:
if (m == NULL || m->m_len != sizeof(struct linger)) {
1993-03-21 12:45:37 +03:00
error = EINVAL;
goto bad;
}
so->so_linger = mtod(m, struct linger *)->l_linger;
/* fall thru... */
case SO_DEBUG:
case SO_KEEPALIVE:
case SO_DONTROUTE:
case SO_USELOOPBACK:
case SO_BROADCAST:
case SO_REUSEADDR:
case SO_REUSEPORT:
1993-03-21 12:45:37 +03:00
case SO_OOBINLINE:
case SO_TIMESTAMP:
if (m == NULL || m->m_len < sizeof(int)) {
1993-03-21 12:45:37 +03:00
error = EINVAL;
goto bad;
}
if (*mtod(m, int *))
so->so_options |= optname;
else
so->so_options &= ~optname;
break;
case SO_SNDBUF:
case SO_RCVBUF:
case SO_SNDLOWAT:
case SO_RCVLOWAT:
{
int optval;
if (m == NULL || m->m_len < sizeof(int)) {
1993-03-21 12:45:37 +03:00
error = EINVAL;
goto bad;
}
/*
* Values < 1 make no sense for any of these
* options, so disallow them.
*/
optval = *mtod(m, int *);
if (optval < 1) {
error = EINVAL;
goto bad;
}
1993-03-21 12:45:37 +03:00
switch (optname) {
case SO_SNDBUF:
case SO_RCVBUF:
if (sbreserve(optname == SO_SNDBUF ?
&so->so_snd : &so->so_rcv,
(u_long) optval) == 0) {
1993-03-21 12:45:37 +03:00
error = ENOBUFS;
goto bad;
}
break;
/*
* Make sure the low-water is never greater than
* the high-water.
*/
1993-03-21 12:45:37 +03:00
case SO_SNDLOWAT:
so->so_snd.sb_lowat =
(optval > so->so_snd.sb_hiwat) ?
so->so_snd.sb_hiwat : optval;
1993-03-21 12:45:37 +03:00
break;
case SO_RCVLOWAT:
so->so_rcv.sb_lowat =
(optval > so->so_rcv.sb_hiwat) ?
so->so_rcv.sb_hiwat : optval;
1993-03-21 12:45:37 +03:00
break;
}
break;
}
1993-03-21 12:45:37 +03:00
case SO_SNDTIMEO:
case SO_RCVTIMEO:
{
struct timeval *tv;
short val;
if (m == NULL || m->m_len < sizeof(*tv)) {
1993-03-21 12:45:37 +03:00
error = EINVAL;
goto bad;
}
tv = mtod(m, struct timeval *);
if (tv->tv_sec * hz + tv->tv_usec / tick > SHRT_MAX) {
1993-03-21 12:45:37 +03:00
error = EDOM;
goto bad;
}
val = tv->tv_sec * hz + tv->tv_usec / tick;
switch (optname) {
case SO_SNDTIMEO:
so->so_snd.sb_timeo = val;
break;
case SO_RCVTIMEO:
so->so_rcv.sb_timeo = val;
break;
}
break;
}
default:
error = ENOPROTOOPT;
break;
}
if (error == 0 && so->so_proto && so->so_proto->pr_ctloutput) {
(void) ((*so->so_proto->pr_ctloutput)
(PRCO_SETOPT, so, level, optname, &m0));
m = NULL; /* freed by protocol */
}
1993-03-21 12:45:37 +03:00
}
bad:
if (m)
(void) m_free(m);
return (error);
}
1994-05-04 15:24:06 +04:00
int
1993-03-21 12:45:37 +03:00
sogetopt(so, level, optname, mp)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
int level, optname;
struct mbuf **mp;
{
2000-03-30 13:27:11 +04:00
struct mbuf *m;
1993-03-21 12:45:37 +03:00
if (level != SOL_SOCKET) {
if (so->so_proto && so->so_proto->pr_ctloutput) {
return ((*so->so_proto->pr_ctloutput)
(PRCO_GETOPT, so, level, optname, mp));
} else
return (ENOPROTOOPT);
} else {
m = m_get(M_WAIT, MT_SOOPTS);
m->m_len = sizeof(int);
1993-03-21 12:45:37 +03:00
switch (optname) {
case SO_LINGER:
m->m_len = sizeof(struct linger);
1993-03-21 12:45:37 +03:00
mtod(m, struct linger *)->l_onoff =
so->so_options & SO_LINGER;
mtod(m, struct linger *)->l_linger = so->so_linger;
break;
case SO_USELOOPBACK:
case SO_DONTROUTE:
case SO_DEBUG:
case SO_KEEPALIVE:
case SO_REUSEADDR:
case SO_REUSEPORT:
1993-03-21 12:45:37 +03:00
case SO_BROADCAST:
case SO_OOBINLINE:
case SO_TIMESTAMP:
1993-03-21 12:45:37 +03:00
*mtod(m, int *) = so->so_options & optname;
break;
case SO_TYPE:
*mtod(m, int *) = so->so_type;
break;
case SO_ERROR:
*mtod(m, int *) = so->so_error;
so->so_error = 0;
break;
case SO_SNDBUF:
*mtod(m, int *) = so->so_snd.sb_hiwat;
break;
case SO_RCVBUF:
*mtod(m, int *) = so->so_rcv.sb_hiwat;
break;
case SO_SNDLOWAT:
*mtod(m, int *) = so->so_snd.sb_lowat;
break;
case SO_RCVLOWAT:
*mtod(m, int *) = so->so_rcv.sb_lowat;
break;
case SO_SNDTIMEO:
case SO_RCVTIMEO:
{
int val = (optname == SO_SNDTIMEO ?
so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
m->m_len = sizeof(struct timeval);
mtod(m, struct timeval *)->tv_sec = val / hz;
mtod(m, struct timeval *)->tv_usec =
(val % hz) * tick;
1993-03-21 12:45:37 +03:00
break;
}
default:
(void)m_free(m);
return (ENOPROTOOPT);
}
*mp = m;
return (0);
}
}
1994-05-04 15:24:06 +04:00
void
1993-03-21 12:45:37 +03:00
sohasoutofband(so)
2000-03-30 13:27:11 +04:00
struct socket *so;
1993-03-21 12:45:37 +03:00
{
struct proc *p;
if (so->so_pgid < 0)
gsignal(-so->so_pgid, SIGURG);
else if (so->so_pgid > 0 && (p = pfind(so->so_pgid)) != 0)
psignal(p, SIGURG);
selwakeup(&so->so_rcv.sb_sel);
1993-03-21 12:45:37 +03:00
}