uipc_socket.c

RTEMS (Real-Time Executive for Multiprocessor Systems) is a free open source real-time operating sys

	 *	receive operation at once if we block (resid <= hiwat).
	 *   3. MSG_DONTWAIT is not set
	 * 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.
	 */
	if (m == 0 || (((flags & MSG_DONTWAIT) == 0 &&
	    so->so_rcv.sb_cc < uio->uio_resid) &&
	    (so->so_rcv.sb_cc < so->so_rcv.sb_lowat ||
	    ((flags & MSG_WAITALL) && uio->uio_resid <= so->so_rcv.sb_hiwat)) &&
	    m->m_nextpkt == 0 && (pr->pr_flags & PR_ATOMIC) == 0)) {
#ifdef DIAGNOSTIC
		if (m == 0 && so->so_rcv.sb_cc)
			panic("receive 1");
#endif
		if (so->so_error) {
			if (m)
				goto dontblock;
			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;
			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)) {
			error = EWOULDBLOCK;
			goto release;
		}
		sbunlock(&so->so_rcv);
		error = sbwait(&so->so_rcv);
		splx(s);
		if (error)
			return (error);
		goto restart;
	}
dontblock:
	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;
		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);
				*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;
			controlp = &(*controlp)->m_next;
		}
	}
	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);
			s = splnet();
			if (error)
				goto release;
		} 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 {
				offset += len;
				if (offset == so->so_oobmark)
					break;
			}
		}
		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) {
			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);
			}
			m = so->so_rcv.sb_mb;
			if (m)
				nextrecord = m->m_nextpkt;
		}
	}

	if (m && pr->pr_flags & PR_ATOMIC) {
		flags |= MSG_TRUNC;
		if ((flags & MSG_PEEK) == 0)
			(void) sbdroprecord(&so->so_rcv);
	}
	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_usrreqs->pru_rcvd)(so, flags);
	}
	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;
	}

	if (flagsp)
		*flagsp |= flags;
release:
	sbunlock(&so->so_rcv);
	splx(s);
	return (error);
}

int
soshutdown(so, how)
	register struct socket *so;
	register int how;
{
	register struct protosw *pr = so->so_proto;

	how++;
	if (how & FREAD)
		sorflush(so);
	if (how & FWRITE)
		return ((*pr->pr_usrreqs->pru_shutdown)(so));
	return (0);
}

void
sorflush(so)
	register struct socket *so;
{
	register struct sockbuf *sb = &so->so_rcv;
	register struct protosw *pr = so->so_proto;
	register int s;
	struct sockbuf asb;

	sb->sb_flags |= SB_NOINTR;
	(void) sblock(sb, M_WAITOK);
	s = splimp();
	socantrcvmore(so);
	sbunlock(sb);
	asb = *sb;
	bzero((caddr_t)sb, sizeof (*sb));
	splx(s);
	if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose)
		(*pr->pr_domain->dom_dispose)(asb.sb_mb);
	sbrelease(&asb);
}

int
sosetopt(so, level, optname, m0)
	register struct socket *so;
	int level, optname;
	struct mbuf *m0;
{
	int error = 0;
	register struct mbuf *m = m0;

	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)) {
				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:
		case SO_OOBINLINE:
		case SO_TIMESTAMP:
			if (m == NULL || m->m_len < sizeof (int)) {
				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)) {
				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;
			}

			switch (optname) {

			case SO_SNDBUF:
			case SO_RCVBUF:
				if (sbreserve(optname == SO_SNDBUF ?
				    &so->so_snd : &so->so_rcv,
				    (u_long) optval) == 0) {
					error = ENOBUFS;
					goto bad;
				}
				break;

			/*
			 * Make sure the low-water is never greater than
			 * the high-water.
			 */
			case SO_SNDLOWAT:
				so->so_snd.sb_lowat =
				    (optval > so->so_snd.sb_hiwat) ?
				    so->so_snd.sb_hiwat : optval;
				break;
			case SO_RCVLOWAT:
				so->so_rcv.sb_lowat =
				    (optval > so->so_rcv.sb_hiwat) ?
				    so->so_rcv.sb_hiwat : optval;
				break;
			}
			break;
		    }

		case SO_SNDTIMEO:
		case SO_RCVTIMEO:
		    {
			struct timeval *tv;
			unsigned long val;

			if (m == NULL || m->m_len < sizeof (*tv)) {
				error = EINVAL;
				goto bad;
			}
			tv = mtod(m, struct timeval *);
			if (tv->tv_sec >= (ULONG_MAX - hz) / hz) {
				error = EDOM;
				goto bad;
			}

			val = tv->tv_sec * hz + tv->tv_usec / tick;
			if ((val == 0) && (tv->tv_sec || tv->tv_usec))
				val = 1;

			switch (optname) {

			case SO_SNDTIMEO:
				so->so_snd.sb_timeo = val;
				break;
			case SO_RCVTIMEO:
				so->so_rcv.sb_timeo = val;
				break;
			}
			break;
		    }

		case SO_PRIVSTATE:
			/* we don't care what the parameter is... */
			so->so_state &= ~SS_PRIV;
			break;

		case SO_SNDWAKEUP:
		case SO_RCVWAKEUP:
		    {
			/* RTEMS addition.  */
			struct sockwakeup *sw;
			struct sockbuf *sb;

			if (m == NULL
			    || m->m_len != sizeof (struct sockwakeup)) {
				error = EINVAL;
				goto bad;
			}
			sw = mtod(m, struct sockwakeup *);
			sb = (optname == SO_SNDWAKEUP
			      ? &so->so_snd
			      : &so->so_rcv);
			sb->sb_wakeup = sw->sw_pfn;
			sb->sb_wakeuparg = sw->sw_arg;
			if (sw->sw_pfn)
				sb->sb_flags |= SB_ASYNC;
			else
				sb->sb_flags &=~ SB_ASYNC;
			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 */
		}
	}
bad:
	if (m)
		(void) m_free(m);
	return (error);
}

int
sogetopt(so, level, optname, mp)
	register struct socket *so;
	int level, optname;
	struct mbuf **mp;
{
	register struct mbuf *m;

	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);

		switch (optname) {

		case SO_LINGER:
			m->m_len = sizeof (struct linger);
			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:
		case SO_BROADCAST:
		case SO_OOBINLINE:
		case SO_TIMESTAMP:
			*mtod(m, int *) = so->so_options & optname;
			break;

		case SO_PRIVSTATE:
			*mtod(m, int *) = so->so_state & SS_PRIV;
			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:
		    {
			unsigned long 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;
			break;
		    }

		case SO_SNDWAKEUP:
		case SO_RCVWAKEUP:
		    {
			struct sockbuf *sb;
			struct sockwakeup *sw;

			/* RTEMS additions.  */
			sb = (optname == SO_SNDWAKEUP
			      ? &so->so_snd
			      : &so->so_rcv);
			m->m_len = sizeof (struct sockwakeup);
			sw = mtod(m, struct sockwakeup *);
			sw->sw_pfn = sb->sb_wakeup;
			sw->sw_arg = sb->sb_wakeuparg;
			break;
		    }

		default:
			(void)m_free(m);
			return (ENOPROTOOPT);
		}
		*mp = m;
		return (0);
	}
}

void
sohasoutofband(so)
	register struct socket *so;
{
#if 0	/* FIXME: For now we just ignore out of band data */
	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);
#endif
}