/*
- * Copyright (c) 1982 Regents of the University of California.
- * All rights reserved. The Berkeley software License Agreement
- * specifies the terms and conditions for redistribution.
*
- * @(#)uipc_usrreq.c 6.16 (Berkeley) %G%
+ * Redistribution and use in source and binary forms are permitted
+ * provided that the above copyright notice and this paragraph are
+ * duplicated in all such forms and that any documentation,
+ * advertising materials, and other materials related to such
+ * distribution and use acknowledge that the software was developed
+ * by the University of California, Berkeley. The name of the
+ * University may not be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
+ * IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * @(#)uipc_usrreq.c 7.19 (Berkeley) %G%
*/
#include "param.h"
-#include "dir.h"
#include "user.h"
-#include "mbuf.h"
#include "domain.h"
#include "protosw.h"
#include "socket.h"
#include "socketvar.h"
#include "unpcb.h"
#include "un.h"
-#include "inode.h"
+#include "vnode.h"
#include "file.h"
#include "stat.h"
+#include "mbuf.h"
/*
* Unix communications domain.
* rethink name space problems
* need a proper out-of-band
*/
-struct sockaddr sun_noname = { AF_UNIX };
-ino_t unp_ino; /* fake inode numbers */
+struct sockaddr sun_noname = { sizeof(sun_noname), AF_UNIX };
+ino_t unp_ino; /* prototype for fake inode numbers */
/*ARGSUSED*/
-uipc_usrreq(so, req, m, nam, rights)
+uipc_usrreq(so, req, m, nam, control)
struct socket *so;
int req;
- struct mbuf *m, *nam, *rights;
+ struct mbuf *m, *nam, *control;
{
struct unpcb *unp = sotounpcb(so);
register struct socket *so2;
- int error = 0;
+ register int error = 0;
- if (req != PRU_SEND && rights && rights->m_len) {
+ if (req == PRU_CONTROL)
+ return (EOPNOTSUPP);
+ if (req != PRU_SEND && control && control->m_len) {
error = EOPNOTSUPP;
goto release;
}
break;
case PRU_LISTEN:
- if (unp->unp_inode == 0)
+ if (unp->unp_vnode == 0)
error = EINVAL;
break;
break;
case PRU_CONNECT2:
- error = unp_connect2(so, (struct mbuf *)0,
- (struct socket *)nam);
+ error = unp_connect2(so, (struct socket *)nam);
break;
case PRU_DISCONNECT:
break;
case PRU_ACCEPT:
- nam->m_len = unp->unp_remaddr->m_len;
- bcopy(mtod(unp->unp_remaddr, caddr_t),
- mtod(nam, caddr_t), (unsigned)nam->m_len);
+ /*
+ * Pass back name of connected socket,
+ * if it was bound and we are still connected
+ * (our peer may have closed already!).
+ */
+ if (unp->unp_conn && unp->unp_conn->unp_addr) {
+ nam->m_len = unp->unp_conn->unp_addr->m_len;
+ bcopy(mtod(unp->unp_conn->unp_addr, caddr_t),
+ mtod(nam, caddr_t), (unsigned)nam->m_len);
+ } else {
+ nam->m_len = sizeof(sun_noname);
+ *(mtod(nam, struct sockaddr *)) = sun_noname;
+ }
break;
case PRU_SHUTDOWN:
break;
so2 = unp->unp_conn->unp_socket;
/*
- * Transfer resources back to send port
+ * Adjust backpressure on sender
* and wakeup any waiting to write.
*/
- snd->sb_mbmax += rcv->sb_mbmax - rcv->sb_mbcnt;
- rcv->sb_mbmax = rcv->sb_mbcnt;
- snd->sb_hiwat += rcv->sb_hiwat - rcv->sb_cc;
- rcv->sb_hiwat = rcv->sb_cc;
+ snd->sb_mbmax += unp->unp_mbcnt - rcv->sb_mbcnt;
+ unp->unp_mbcnt = rcv->sb_mbcnt;
+ snd->sb_hiwat += unp->unp_cc - rcv->sb_cc;
+ unp->unp_cc = rcv->sb_cc;
sowwakeup(so2);
#undef snd
#undef rcv
break;
case PRU_SEND:
+ if (control && (error = unp_internalize(control)))
+ break;
switch (so->so_type) {
- case SOCK_DGRAM:
+ case SOCK_DGRAM: {
+ struct sockaddr *from;
+
if (nam) {
if (unp->unp_conn) {
error = EISCONN;
}
}
so2 = unp->unp_conn->unp_socket;
- /* BEGIN XXX */
- if (rights) {
- error = unp_internalize(rights);
- if (error)
- break;
- }
- if (sbspace(&so2->so_rcv) > 0) {
- /*
- * There's no record of source socket's
- * name, so send null name for the moment.
- */
- if (sbappendaddr(&so2->so_rcv,
- &sun_noname, m, rights)) {
- sorwakeup(so2);
- m = 0;
- } else
- error = ENOBUFS;
- }
- /* END XXX */
+ if (unp->unp_addr)
+ from = mtod(unp->unp_addr, struct sockaddr *);
+ else
+ from = &sun_noname;
+ if (sbappendaddr(&so2->so_rcv, from, m, control)) {
+ sorwakeup(so2);
+ m = 0;
+ control = 0;
+ } else
+ error = ENOBUFS;
if (nam)
unp_disconnect(unp);
break;
+ }
case SOCK_STREAM:
#define rcv (&so2->so_rcv)
#define snd (&so->so_snd)
- if (rights && rights->m_len) {
- error = EOPNOTSUPP;
- break;
- }
if (so->so_state & SS_CANTSENDMORE) {
error = EPIPE;
break;
panic("uipc 3");
so2 = unp->unp_conn->unp_socket;
/*
- * Send to paired receive port, and then
- * give it enough resources to hold what it already has.
+ * Send to paired receive port, and then reduce
+ * send buffer hiwater marks to maintain backpressure.
* Wake up readers.
*/
- sbappend(rcv, m);
- snd->sb_mbmax -= rcv->sb_mbcnt - rcv->sb_mbmax;
- rcv->sb_mbmax = rcv->sb_mbcnt;
- snd->sb_hiwat -= rcv->sb_cc - rcv->sb_hiwat;
- rcv->sb_hiwat = rcv->sb_cc;
+ if (control) {
+ (void)sbappendcontrol(rcv, m, control);
+ control = 0;
+ } else
+ sbappend(rcv, m);
+ snd->sb_mbmax -=
+ rcv->sb_mbcnt - unp->unp_conn->unp_mbcnt;
+ unp->unp_conn->unp_mbcnt = rcv->sb_mbcnt;
+ snd->sb_hiwat -= rcv->sb_cc - unp->unp_conn->unp_cc;
+ unp->unp_conn->unp_cc = rcv->sb_cc;
sorwakeup(so2);
m = 0;
#undef snd
unp_drop(unp, ECONNABORTED);
break;
-/* SOME AS YET UNIMPLEMENTED HOOKS */
- case PRU_CONTROL:
- return (EOPNOTSUPP);
-
-/* END UNIMPLEMENTED HOOKS */
case PRU_SENSE:
((struct stat *) m)->st_blksize = so->so_snd.sb_hiwat;
if (so->so_type == SOCK_STREAM && unp->unp_conn != 0) {
((struct stat *) m)->st_blksize += so2->so_rcv.sb_cc;
}
((struct stat *) m)->st_dev = NODEV;
- ((struct stat *) m)->st_ino = unp_ino++;
+ if (unp->unp_ino == 0)
+ unp->unp_ino = unp_ino++;
+ ((struct stat *) m)->st_ino = unp->unp_ino;
return (0);
case PRU_RCVOOB:
break;
case PRU_SOCKADDR:
+ if (unp->unp_addr) {
+ nam->m_len = unp->unp_addr->m_len;
+ bcopy(mtod(unp->unp_addr, caddr_t),
+ mtod(nam, caddr_t), (unsigned)nam->m_len);
+ } else
+ nam->m_len = 0;
break;
case PRU_PEERADDR:
+ if (unp->unp_conn && unp->unp_conn->unp_addr) {
+ nam->m_len = unp->unp_conn->unp_addr->m_len;
+ bcopy(mtod(unp->unp_conn->unp_addr, caddr_t),
+ mtod(nam, caddr_t), (unsigned)nam->m_len);
+ } else
+ nam->m_len = 0;
break;
case PRU_SLOWTIMO:
panic("piusrreq");
}
release:
+ if (control)
+ m_freem(control);
if (m)
m_freem(m);
return (error);
}
/*
- * We assign all buffering for stream sockets to the source,
- * as that is where the flow control is implemented.
+ * Both send and receive buffers are allocated PIPSIZ bytes of buffering
+ * for stream sockets, although the total for sender and receiver is
+ * actually only PIPSIZ.
* Datagram sockets really use the sendspace as the maximum datagram size,
* and don't really want to reserve the sendspace. Their recvspace should
* be large enough for at least one max-size datagram plus address.
*/
#define PIPSIZ 4096
-int unpst_sendspace = PIPSIZ;
-int unpst_recvspace = 0;
-int unpdg_sendspace = 2*1024; /* really max datagram size */
-int unpdg_recvspace = 4*1024;
+u_long unpst_sendspace = PIPSIZ;
+u_long unpst_recvspace = PIPSIZ;
+u_long unpdg_sendspace = 2*1024; /* really max datagram size */
+u_long unpdg_recvspace = 4*1024;
+
+int unp_rights; /* file descriptors in flight */
unp_attach(so)
struct socket *so;
register struct unpcb *unp;
int error;
- switch (so->so_type) {
+ if (so->so_snd.sb_hiwat == 0 || so->so_rcv.sb_hiwat == 0) {
+ switch (so->so_type) {
- case SOCK_STREAM:
- error = soreserve(so, unpst_sendspace, unpst_recvspace);
- break;
+ case SOCK_STREAM:
+ error = soreserve(so, unpst_sendspace, unpst_recvspace);
+ break;
- case SOCK_DGRAM:
- error = soreserve(so, unpdg_sendspace, unpdg_recvspace);
- break;
+ case SOCK_DGRAM:
+ error = soreserve(so, unpdg_sendspace, unpdg_recvspace);
+ break;
+ }
+ if (error)
+ return (error);
}
- if (error)
- return (error);
m = m_getclr(M_DONTWAIT, MT_PCB);
if (m == NULL)
return (ENOBUFS);
register struct unpcb *unp;
{
- if (unp->unp_inode) {
- unp->unp_inode->i_socket = 0;
- irele(unp->unp_inode);
- unp->unp_inode = 0;
+ if (unp->unp_vnode) {
+ unp->unp_vnode->v_socket = 0;
+ vrele(unp->unp_vnode);
+ unp->unp_vnode = 0;
}
if (unp->unp_conn)
unp_disconnect(unp);
unp_drop(unp->unp_refs, ECONNRESET);
soisdisconnected(unp->unp_socket);
unp->unp_socket->so_pcb = 0;
- m_freem(unp->unp_remaddr);
+ m_freem(unp->unp_addr);
(void) m_free(dtom(unp));
+ if (unp_rights)
+ unp_gc();
}
unp_bind(unp, nam)
struct mbuf *nam;
{
struct sockaddr_un *soun = mtod(nam, struct sockaddr_un *);
- register struct inode *ip;
+ register struct vnode *vp;
register struct nameidata *ndp = &u.u_nd;
+ struct vattr vattr;
int error;
ndp->ni_dirp = soun->sun_path;
- if (unp->unp_inode != NULL || nam->m_len == MLEN)
+ if (unp->unp_vnode != NULL)
return (EINVAL);
- *(mtod(nam, caddr_t) + nam->m_len) = 0;
+ if (nam->m_len == MLEN) {
+ if (*(mtod(nam, caddr_t) + nam->m_len - 1) != 0)
+ return (EINVAL);
+ } else
+ *(mtod(nam, caddr_t) + nam->m_len) = 0;
/* SHOULD BE ABLE TO ADOPT EXISTING AND wakeup() ALA FIFO's */
- ndp->ni_nameiop = CREATE | FOLLOW;
+ ndp->ni_nameiop = CREATE | FOLLOW | LOCKPARENT;
ndp->ni_segflg = UIO_SYSSPACE;
- ip = namei(ndp);
- if (ip) {
- iput(ip);
- return (EADDRINUSE);
- }
- if (error = u.u_error) {
- u.u_error = 0; /* XXX */
+ if (error = namei(ndp))
return (error);
+ vp = ndp->ni_vp;
+ if (vp != NULL) {
+ VOP_ABORTOP(ndp);
+ if (ndp->ni_dvp == vp)
+ vrele(ndp->ni_dvp);
+ else
+ vput(ndp->ni_dvp);
+ vrele(vp);
+ return (EADDRINUSE);
}
- ip = maknode(IFSOCK | 0777, ndp);
- if (ip == NULL) {
- error = u.u_error; /* XXX */
- u.u_error = 0; /* XXX */
+ VATTR_NULL(&vattr);
+ vattr.va_type = VSOCK;
+ vattr.va_mode = 0777;
+ if (error = VOP_CREATE(ndp, &vattr))
return (error);
- }
- ip->i_socket = unp->unp_socket;
- unp->unp_inode = ip;
- iunlock(ip); /* but keep reference */
+ vp = ndp->ni_vp;
+ vp->v_socket = unp->unp_socket;
+ unp->unp_vnode = vp;
+ unp->unp_addr = m_copy(nam, 0, (int)M_COPYALL);
+ VOP_UNLOCK(vp);
return (0);
}
struct mbuf *nam;
{
register struct sockaddr_un *soun = mtod(nam, struct sockaddr_un *);
- register struct inode *ip;
- int error;
- register struct socket *so2;
+ register struct vnode *vp;
+ register struct socket *so2, *so3;
register struct nameidata *ndp = &u.u_nd;
+ struct unpcb *unp2, *unp3;
+ int error;
ndp->ni_dirp = soun->sun_path;
- if (nam->m_len + (nam->m_off - MMINOFF) == MLEN)
- return (EMSGSIZE);
- *(mtod(nam, caddr_t) + nam->m_len) = 0;
- ndp->ni_nameiop = LOOKUP | FOLLOW;
+ if (nam->m_data + nam->m_len == &nam->m_dat[MLEN]) { /* XXX */
+ if (*(mtod(nam, caddr_t) + nam->m_len - 1) != 0)
+ return (EMSGSIZE);
+ } else
+ *(mtod(nam, caddr_t) + nam->m_len) = 0;
+ ndp->ni_nameiop = LOOKUP | FOLLOW | LOCKLEAF;
ndp->ni_segflg = UIO_SYSSPACE;
- ip = namei(ndp);
- if (ip == 0) {
- error = u.u_error;
- u.u_error = 0;
- return (error); /* XXX */
- }
- if (access(ip, IWRITE)) {
- error = u.u_error;
- u.u_error = 0; /* XXX */
- goto bad;
- }
- if ((ip->i_mode&IFMT) != IFSOCK) {
+ if (error = namei(ndp))
+ return (error);
+ vp = ndp->ni_vp;
+ if (vp->v_type != VSOCK) {
error = ENOTSOCK;
goto bad;
}
- so2 = ip->i_socket;
+ if (error = VOP_ACCESS(vp, VWRITE, ndp->ni_cred))
+ goto bad;
+ so2 = vp->v_socket;
if (so2 == 0) {
error = ECONNREFUSED;
goto bad;
error = EPROTOTYPE;
goto bad;
}
- if (so->so_proto->pr_flags & PR_CONNREQUIRED &&
- ((so2->so_options&SO_ACCEPTCONN) == 0 ||
- (so2 = sonewconn(so2)) == 0)) {
- error = ECONNREFUSED;
- goto bad;
+ if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
+ if ((so2->so_options & SO_ACCEPTCONN) == 0 ||
+ (so3 = sonewconn(so2, 0)) == 0) {
+ error = ECONNREFUSED;
+ goto bad;
+ }
+ unp2 = sotounpcb(so2);
+ unp3 = sotounpcb(so3);
+ if (unp2->unp_addr)
+ unp3->unp_addr =
+ m_copy(unp2->unp_addr, 0, (int)M_COPYALL);
+ so2 = so3;
}
- error = unp_connect2(so, nam, so2);
+ error = unp_connect2(so, so2);
bad:
- iput(ip);
+ vput(vp);
return (error);
}
-unp_connect2(so, sonam, so2)
+unp_connect2(so, so2)
register struct socket *so;
- struct mbuf *sonam;
register struct socket *so2;
{
register struct unpcb *unp = sotounpcb(so);
case SOCK_STREAM:
unp2->unp_conn = unp;
- if (sonam)
- unp2->unp_remaddr = m_copy(sonam, 0, (int)M_COPYALL);
- soisconnected(so2);
soisconnected(so);
+ soisconnected(so2);
break;
default:
unp_disconnect(unp);
if (so->so_head) {
so->so_pcb = (caddr_t) 0;
- m_freem(unp->unp_remaddr);
+ m_freem(unp->unp_addr);
(void) m_free(dtom(unp));
sofree(so);
}
unp_externalize(rights)
struct mbuf *rights;
{
- int newfds = rights->m_len / sizeof (int);
register int i;
- register struct file **rp = mtod(rights, struct file **);
+ register struct cmsghdr *cm = mtod(rights, struct cmsghdr *);
+ register struct file **rp = (struct file **)(cm + 1);
register struct file *fp;
+ int newfds = (cm->cmsg_len - sizeof(*cm)) / sizeof (int);
int f;
if (newfds > ufavail()) {
return (EMSGSIZE);
}
for (i = 0; i < newfds; i++) {
- f = ufalloc(0);
- if (f < 0)
+ if (ufalloc(0, &f))
panic("unp_externalize");
fp = *rp;
u.u_ofile[f] = fp;
fp->f_msgcount--;
+ unp_rights--;
*(int *)rp++ = f;
}
return (0);
}
-unp_internalize(rights)
- struct mbuf *rights;
+unp_internalize(control)
+ struct mbuf *control;
{
+ register struct cmsghdr *cm = mtod(control, struct cmsghdr *);
register struct file **rp;
- int oldfds = rights->m_len / sizeof (int);
- register int i;
register struct file *fp;
+ register int i, fd;
+ int oldfds;
- rp = mtod(rights, struct file **);
- for (i = 0; i < oldfds; i++)
- if (getf(*(int *)rp++) == 0)
+ if (cm->cmsg_type != SCM_RIGHTS || cm->cmsg_level != SOL_SOCKET ||
+ cm->cmsg_len != control->m_len)
+ return (EINVAL);
+ oldfds = (cm->cmsg_len - sizeof (*cm)) / sizeof (int);
+ rp = (struct file **)(cm + 1);
+ for (i = 0; i < oldfds; i++) {
+ fd = *(int *)rp++;
+ if ((unsigned)fd >= NOFILE || u.u_ofile[fd] == NULL)
return (EBADF);
- rp = mtod(rights, struct file **);
+ }
+ rp = (struct file **)(cm + 1);
for (i = 0; i < oldfds; i++) {
- fp = getf(*(int *)rp);
+ fp = u.u_ofile[*(int *)rp];
*rp++ = fp;
fp->f_count++;
fp->f_msgcount++;
+ unp_rights++;
}
return (0);
}
continue;
fp->f_flag |= FMARK;
}
- if (fp->f_type != DTYPE_SOCKET)
+ if (fp->f_type != DTYPE_SOCKET ||
+ (so = (struct socket *)fp->f_data) == 0)
continue;
- so = (struct socket *)fp->f_data;
if (so->so_proto->pr_domain != &unixdomain ||
(so->so_proto->pr_flags&PR_RIGHTS) == 0)
continue;
for (fp = file; fp < fileNFILE; fp++) {
if (fp->f_count == 0)
continue;
- if (fp->f_count == fp->f_msgcount && (fp->f_flag&FMARK)==0) {
- if (fp->f_type != DTYPE_SOCKET)
- panic("unp_gc");
- (void) soshutdown((struct socket *)fp->f_data, 0);
- }
+ if (fp->f_count == fp->f_msgcount && (fp->f_flag & FMARK) == 0)
+ while (fp->f_msgcount)
+ unp_discard(fp);
}
unp_gcing = 0;
}
{
register struct mbuf *m;
register struct file **rp;
+ register struct cmsghdr *cm;
register int i;
int qfds;
while (m0) {
for (m = m0; m; m = m->m_next)
- if (m->m_type == MT_RIGHTS && m->m_len) {
- qfds = m->m_len / sizeof (struct file *);
- rp = mtod(m, struct file **);
+ if (m->m_type == MT_CONTROL &&
+ m->m_len >= sizeof(*cm)) {
+ cm = mtod(m, struct cmsghdr *);
+ if (cm->cmsg_level != SOL_SOCKET ||
+ cm->cmsg_type != SCM_RIGHTS)
+ continue;
+ qfds = (cm->cmsg_len - sizeof *cm)
+ / sizeof (struct file *);
+ rp = (struct file **)(cm + 1);
for (i = 0; i < qfds; i++)
(*op)(*rp++);
break; /* XXX, but saves time */
{
fp->f_msgcount--;
- closef(fp);
+ unp_rights--;
+ (void) closef(fp);
}