xnu-1504.7.4.tar.gz

[apple/xnu.git] / bsd / netinet / ip_divert.c

diff --git a/bsd/netinet/ip_divert.c b/bsd/netinet/ip_divert.c
index 8a02c85b548375ef324322dce1ad9677cef867e9..f5db5922a0434739d60fce23fd33b3e3032a673e 100644 (file)
--- a/bsd/netinet/ip_divert.c
+++ b/bsd/netinet/ip_divert.c
@@ -1,31 +1,29 @@
 /*
 /*

- * Copyright (c) 2006 Apple Computer, Inc. All Rights Reserved.
+ * Copyright (c) 2000-2008 Apple Inc. All rights reserved.
+ *
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@

  * 
  * 

- * @APPLE_LICENSE_OSREFERENCE_HEADER_START@
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.

  * 
  * 

- * This file contains Original Code and/or Modifications of Original Code 
- * as defined in and that are subject to the Apple Public Source License 
- * Version 2.0 (the 'License'). You may not use this file except in 
- * compliance with the License.  The rights granted to you under the 
- * License may not be used to create, or enable the creation or 
- * redistribution of, unlawful or unlicensed copies of an Apple operating 
- * system, or to circumvent, violate, or enable the circumvention or 
- * violation of, any terms of an Apple operating system software license 
- * agreement.
- *
- * Please obtain a copy of the License at 
- * http://www.opensource.apple.com/apsl/ and read it before using this 
- * file.
- *
- * The Original Code and all software distributed under the License are 
- * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER 
- * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES, 
- * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY, 
- * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT. 
- * Please see the License for the specific language governing rights and 
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ * 
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
+ * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and

  * limitations under the License.
  * limitations under the License.

- *
- * @APPLE_LICENSE_OSREFERENCE_HEADER_END@
+ * 
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@

  */
 /*
  * Copyright (c) 1982, 1986, 1988, 1993
  */
 /*
  * Copyright (c) 1982, 1986, 1988, 1993


@@ -62,7 +60,7 @@
  * $FreeBSD: src/sys/netinet/ip_divert.c,v 1.98 2004/08/17 22:05:54 andre Exp $
  */
 
  * $FreeBSD: src/sys/netinet/ip_divert.c,v 1.98 2004/08/17 22:05:54 andre Exp $
  */
 

-#ifndef INET
+#if !INET

 #error "IPDIVERT requires INET."
 #endif
 
 #error "IPDIVERT requires INET."
 #endif
 


@@ -78,9 +76,11 @@
 #include <sys/systm.h>
 #include <sys/proc.h>
 
 #include <sys/systm.h>
 #include <sys/proc.h>
 

+#include <machine/endian.h>

 
 #include <net/if.h>
 #include <net/route.h>
 
 #include <net/if.h>
 #include <net/route.h>

+#include <net/kpi_protocol.h>

 
 #include <netinet/in.h>
 #include <netinet/in_systm.h>
 
 #include <netinet/in.h>
 #include <netinet/in_systm.h>


@@ -92,6 +92,7 @@
 #include <netinet/ip_divert.h>
 
 #include <kern/zalloc.h>
 #include <netinet/ip_divert.h>
 
 #include <kern/zalloc.h>

+#include <libkern/OSAtomic.h>

 
 /*
  * Divert sockets
 
 /*
  * Divert sockets


@@ -131,11 +132,11 @@
 static struct inpcbhead divcb;
 static struct inpcbinfo divcbinfo;
 
 static struct inpcbhead divcb;
 static struct inpcbinfo divcbinfo;
 

-static u_long  div_sendspace = DIVSNDQ;        /* XXX sysctl ? */
-static u_long  div_recvspace = DIVRCVQ;        /* XXX sysctl ? */
+static u_int32_t       div_sendspace = DIVSNDQ;        /* XXX sysctl ? */
+static u_int32_t       div_recvspace = DIVRCVQ;        /* XXX sysctl ? */

 
 /* Optimization: have this preinitialized */
 
 /* Optimization: have this preinitialized */

-static struct sockaddr_in divsrc = { sizeof(divsrc), AF_INET,  };
+static struct sockaddr_in divsrc = { sizeof(divsrc), AF_INET, 0, { 0 }, { 0,0,0,0,0,0,0,0 } };

 
 /* Internal functions */
 static int div_output(struct socket *so,
 
 /* Internal functions */
 static int div_output(struct socket *so,


@@ -172,14 +173,15 @@ div_init(void)
         * allocate the lock attribute for divert pcb mutexes
         */
        pcbinfo->mtx_attr = lck_attr_alloc_init();
         * allocate the lock attribute for divert pcb mutexes
         */
        pcbinfo->mtx_attr = lck_attr_alloc_init();

-       lck_attr_setdefault(pcbinfo->mtx_attr);

 
        if ((pcbinfo->mtx = lck_rw_alloc_init(pcbinfo->mtx_grp, pcbinfo->mtx_attr)) == NULL)
                return; /* pretty much dead if this fails... */
 
 
        if ((pcbinfo->mtx = lck_rw_alloc_init(pcbinfo->mtx_grp, pcbinfo->mtx_attr)) == NULL)
                return; /* pretty much dead if this fails... */
 

+#if IPFIREWALL

        if (!IPFW_LOADED) {
                load_ipfw();
        }
        if (!IPFW_LOADED) {
                load_ipfw();
        }

+#endif

 }
 
 /*
 }
 
 /*


@@ -189,7 +191,7 @@ div_init(void)
 void
 div_input(struct mbuf *m, __unused int off)
 {
 void
 div_input(struct mbuf *m, __unused int off)
 {

-       ipstat.ips_noproto++;
+       OSAddAtomic(1, &ipstat.ips_noproto);

        m_freem(m);
 }
 
        m_freem(m);
 }
 


@@ -290,8 +292,8 @@ divert_packet(struct mbuf *m, int incoming, int port, int rule)
                socket_unlock(sa, 1);
        } else {
                m_freem(m);
                socket_unlock(sa, 1);
        } else {
                m_freem(m);

-               ipstat.ips_noproto++;
-               ipstat.ips_delivered--;
+               OSAddAtomic(1, &ipstat.ips_noproto);
+               OSAddAtomic(-1, &ipstat.ips_delivered);

         }
        lck_rw_done(divcbinfo.mtx);     
 }
         }
        lck_rw_done(divcbinfo.mtx);     
 }


@@ -305,14 +307,11 @@ divert_packet(struct mbuf *m, int incoming, int port, int rule)
  * ###LOCK  called in inet_proto mutex when from div_send. 
  */
 static int
  * ###LOCK  called in inet_proto mutex when from div_send. 
  */
 static int

-div_output(so, m, addr, control)
-       struct socket *so;
-       register struct mbuf *m;
-       struct sockaddr *addr;
-       struct mbuf *control;
+div_output(struct socket *so, struct mbuf *m, struct sockaddr *addr,
+          struct mbuf *control)

 {
 {

-       register struct inpcb *const inp = sotoinpcb(so);
-       register struct ip *const ip = mtod(m, struct ip *);
+       struct inpcb *const inp = sotoinpcb(so);
+       struct ip *const ip = mtod(m, struct ip *);

        struct sockaddr_in *sin = (struct sockaddr_in *)addr;
        int error = 0;
 
        struct sockaddr_in *sin = (struct sockaddr_in *)addr;
        int error = 0;
 


@@ -353,6 +352,9 @@ div_output(so, m, addr, control)
 
        /* Reinject packet into the system as incoming or outgoing */
        if (!sin || sin->sin_addr.s_addr == 0) {
 
        /* Reinject packet into the system as incoming or outgoing */
        if (!sin || sin->sin_addr.s_addr == 0) {

+               struct ip_out_args ipoa = { IFSCOPE_NONE };
+               struct route ro;
+

                /*
                 * Don't allow both user specified and setsockopt options,
                 * and don't allow packet length sizes that will crash
                /*
                 * Don't allow both user specified and setsockopt options,
                 * and don't allow packet length sizes that will crash


@@ -364,18 +366,33 @@ div_output(so, m, addr, control)
                }
 
                /* Convert fields to host order for ip_output() */
                }
 
                /* Convert fields to host order for ip_output() */

+#if BYTE_ORDER != BIG_ENDIAN

                NTOHS(ip->ip_len);
                NTOHS(ip->ip_off);
                NTOHS(ip->ip_len);
                NTOHS(ip->ip_off);

+#endif
+
+               OSAddAtomic(1, &ipstat.ips_rawout);
+               /* Copy the cached route and take an extra reference */
+               inp_route_copyout(inp, &ro);
+
+#if PKT_PRIORITY
+               if (soisbackground(so))
+                       m_prio_background(m);
+#endif /* PKT_PRIORITY */

 
 

-               /* Send packet to output processing */
-               ipstat.ips_rawout++;                    /* XXX */

                socket_unlock(so, 0);
                socket_unlock(so, 0);

-               error = ip_output(m,
-                           inp->inp_options, &inp->inp_route,
+#if CONFIG_MACF_NET
+               mac_mbuf_label_associate_inpcb(inp, m);
+#endif
+               /* Send packet to output processing */
+               error = ip_output(m, inp->inp_options, &ro,

                        (so->so_options & SO_DONTROUTE) |
                        (so->so_options & SO_DONTROUTE) |

-                       IP_ALLOWBROADCAST | IP_RAWOUTPUT,
-                       inp->inp_moptions);
+                       IP_ALLOWBROADCAST | IP_RAWOUTPUT | IP_OUTARGS,
+                       inp->inp_moptions, &ipoa);
+

                socket_lock(so, 0);
                socket_lock(so, 0);

+               /* Synchronize cached PCB route */
+               inp_route_copyin(inp, &ro);

        } else {
                struct  ifaddr *ifa;
 
        } else {
                struct  ifaddr *ifa;
 


@@ -396,30 +413,9 @@ div_output(so, m, addr, control)
                        m->m_pkthdr.rcvif = ifa->ifa_ifp;
                        ifafree(ifa);
                }
                        m->m_pkthdr.rcvif = ifa->ifa_ifp;
                        ifafree(ifa);
                }

-               
-               if ((~IF_HWASSIST_CSUM_FLAGS(m->m_pkthdr.rcvif->if_hwassist) & 
-                               m->m_pkthdr.csum_flags) == 0) {
-                       if (m->m_pkthdr.csum_flags & CSUM_DELAY_DATA) {
-                               m->m_pkthdr.csum_flags &= ~CSUM_DELAY_DATA;
-                       }
-                       m->m_pkthdr.csum_flags |=
-                               CSUM_DATA_VALID | CSUM_PSEUDO_HDR |
-                               CSUM_IP_CHECKED | CSUM_IP_VALID;
-                       m->m_pkthdr.csum_data = 0xffff;
-               }
-               else if (m->m_pkthdr.csum_flags & CSUM_DELAY_DATA) {
-                       int     hlen;
-                       
-#ifdef _IP_VHL
-                       hlen = IP_VHL_HL(ip->ip_vhl) << 2;
-#else
-                       hlen = ip->ip_hl << 2;
+#if CONFIG_MACF_NET
+               mac_mbuf_label_associate_socket(so, m);

 #endif
 #endif

-                       in_delayed_cksum(m);
-                       m->m_pkthdr.csum_flags &= ~CSUM_DELAY_DATA;
-                       ip->ip_sum = in_cksum(m, hlen);
-               }
-

                /* Send packet to input processing */
                proto_inject(PF_INET, m);
        }
                /* Send packet to input processing */
                proto_inject(PF_INET, m);
        }


@@ -441,7 +437,7 @@ div_attach(struct socket *so, int proto, struct proc *p)
        inp  = sotoinpcb(so);
        if (inp)
                panic("div_attach");
        inp  = sotoinpcb(so);
        if (inp)
                panic("div_attach");

-       if (p && (error = proc_suser(p)) != 0)
+       if ((error = proc_suser(p)) != 0)

                return error;
 
        error = soreserve(so, div_sendspace, div_recvspace);
                return error;
 
        error = soreserve(so, div_sendspace, div_recvspace);


@@ -459,7 +455,7 @@ div_attach(struct socket *so, int proto, struct proc *p)
        so->so_state |= SS_ISCONNECTED;
 
 #ifdef MORE_DICVLOCK_DEBUG
        so->so_state |= SS_ISCONNECTED;
 
 #ifdef MORE_DICVLOCK_DEBUG

-       printf("div_attach: so=%x sopcb=%x lock=%x ref=%x\n",
+       printf("div_attach: so=%p sopcb=%p lock=%x ref=%x\n",

                        so, so->so_pcb, ((struct inpcb *)so->so_pcb)->inpcb_mtx, so->so_usecount);
 #endif
        return 0;
                        so, so->so_pcb, ((struct inpcb *)so->so_pcb)->inpcb_mtx, so->so_usecount);
 #endif
        return 0;


@@ -471,12 +467,12 @@ div_detach(struct socket *so)
        struct inpcb *inp;
 
 #ifdef MORE_DICVLOCK_DEBUG
        struct inpcb *inp;
 
 #ifdef MORE_DICVLOCK_DEBUG

-       printf("div_detach: so=%x sopcb=%x lock=%x ref=%x\n",
+       printf("div_detach: so=%p sopcb=%p lock=%x ref=%x\n",

                        so, so->so_pcb, ((struct inpcb *)so->so_pcb)->inpcb_mtx, so->so_usecount);
 #endif
        inp = sotoinpcb(so);
        if (inp == 0)
                        so, so->so_pcb, ((struct inpcb *)so->so_pcb)->inpcb_mtx, so->so_usecount);
 #endif
        inp = sotoinpcb(so);
        if (inp == 0)

-               panic("div_detach: so=%x null inp\n", so);
+               panic("div_detach: so=%p null inp\n", so);

        in_pcbdetach(inp);
        inp->inp_state = INPCB_STATE_DEAD;
        return 0;
        in_pcbdetach(inp);
        inp->inp_state = INPCB_STATE_DEAD;
        return 0;


@@ -534,7 +530,7 @@ div_send(struct socket *so, __unused int flags, struct mbuf *m, struct sockaddr
        /* Packet must have a header (but that's about it) */
        if (m->m_len < sizeof (struct ip) &&
            (m = m_pullup(m, sizeof (struct ip))) == 0) {
        /* Packet must have a header (but that's about it) */
        if (m->m_len < sizeof (struct ip) &&
            (m = m_pullup(m, sizeof (struct ip))) == 0) {

-               ipstat.ips_toosmall++;
+               OSAddAtomic(1, &ipstat.ips_toosmall);

                m_freem(m);
                return EINVAL;
        }
                m_freem(m);
                return EINVAL;
        }


@@ -543,9 +539,11 @@ div_send(struct socket *so, __unused int flags, struct mbuf *m, struct sockaddr
        return div_output(so, m, nam, control);
 }
 
        return div_output(so, m, nam, control);
 }
 

+#if 0

 static int
 div_pcblist SYSCTL_HANDLER_ARGS
 {
 static int
 div_pcblist SYSCTL_HANDLER_ARGS
 {

+#pragma unused(oidp, arg1, arg2)

        int error, i, n;
        struct inpcb *inp, **inp_list;
        inp_gen_t gencnt;
        int error, i, n;
        struct inpcb *inp, **inp_list;
        inp_gen_t gencnt;


@@ -637,86 +635,92 @@ div_pcblist SYSCTL_HANDLER_ARGS
        lck_rw_done(divcbinfo.mtx);
        return error;
 }
        lck_rw_done(divcbinfo.mtx);
        return error;
 }

+#endif

 
 __private_extern__ int
 
 __private_extern__ int

-div_lock(struct socket *so, int refcount, int lr)
- {
-       int lr_saved;
-#ifdef __ppc__
-       if (lr == 0) {
-                       __asm__ volatile("mflr %0" : "=r" (lr_saved));
-       }
-       else lr_saved = lr;
-#endif
-       
+div_lock(struct socket *so, int refcount, void *lr)
+{
+       void *lr_saved;
+
+       if (lr == NULL)
+               lr_saved = __builtin_return_address(0);
+       else
+               lr_saved = lr;
+

 #ifdef MORE_DICVLOCK_DEBUG
 #ifdef MORE_DICVLOCK_DEBUG

-       printf("div_lock: so=%x sopcb=%x lock=%x ref=%x lr=%x\n",
-                       so, 
-                       so->so_pcb, 
-                       so->so_pcb ? ((struct inpcb *)so->so_pcb)->inpcb_mtx : 0, 
-                       so->so_usecount, 
-                       lr_saved);
+       printf("div_lock: so=%p sopcb=%p lock=%p ref=%x lr=%p\n",
+           so, so->so_pcb, so->so_pcb ?
+           ((struct inpcb *)so->so_pcb)->inpcb_mtx : NULL,
+           so->so_usecount, lr_saved);

 #endif
        if (so->so_pcb) {
                lck_mtx_lock(((struct inpcb *)so->so_pcb)->inpcb_mtx);
        } else  {
 #endif
        if (so->so_pcb) {
                lck_mtx_lock(((struct inpcb *)so->so_pcb)->inpcb_mtx);
        } else  {

-               panic("div_lock: so=%x NO PCB! lr=%x\n", so, lr_saved);
-               lck_mtx_lock(so->so_proto->pr_domain->dom_mtx);
+               panic("div_lock: so=%p NO PCB! lr=%p lrh= lrh= %s\n", 
+                   so, lr_saved, solockhistory_nr(so));
+               /* NOTREACHED */

        }
        }

-       
-       if (so->so_usecount < 0)
-               panic("div_lock: so=%x so_pcb=%x lr=%x ref=%x\n",
-               so, so->so_pcb, lr_saved, so->so_usecount);
-       
+
+       if (so->so_usecount < 0) {
+               panic("div_lock: so=%p so_pcb=%p lr=%p ref=%x lrh= %s\n",
+                   so, so->so_pcb, lr_saved, so->so_usecount,
+                   solockhistory_nr(so));
+               /* NOTREACHED */
+       }
+

        if (refcount)
                so->so_usecount++;
        if (refcount)
                so->so_usecount++;

-       so->reserved3 = (void *)lr_saved;
+       so->lock_lr[so->next_lock_lr] = lr_saved;
+       so->next_lock_lr = (so->next_lock_lr+1) % SO_LCKDBG_MAX;

 
        return (0);
 }
 
 __private_extern__ int
 
        return (0);
 }
 
 __private_extern__ int

-div_unlock(struct socket *so, int refcount, int lr)
+div_unlock(struct socket *so, int refcount, void *lr)

 {
 {

-       int lr_saved;
+       void *lr_saved;

        lck_mtx_t * mutex_held;
        lck_mtx_t * mutex_held;

-       struct inpcb *inp = sotoinpcb(so);      
-#ifdef __ppc__
-       if (lr == 0) {
-               __asm__ volatile("mflr %0" : "=r" (lr_saved));
-       }
-       else lr_saved = lr;
-#endif
-       
+       struct inpcb *inp = sotoinpcb(so);
+
+       if (lr == NULL)
+               lr_saved = __builtin_return_address(0);
+       else
+               lr_saved = lr;
+

 #ifdef MORE_DICVLOCK_DEBUG
 #ifdef MORE_DICVLOCK_DEBUG

-       printf("div_unlock: so=%x sopcb=%x lock=%x ref=%x lr=%x\n",
-                       so, 
-                       so->so_pcb, 
-                       so->so_pcb ? ((struct inpcb *)so->so_pcb)->inpcb_mtx : 0, 
-                       so->so_usecount, 
-                       lr_saved);
+       printf("div_unlock: so=%p sopcb=%p lock=%p ref=%x lr=%p\n",
+           so, so->so_pcb, so->so_pcb ?
+           ((struct inpcb *)so->so_pcb)->inpcb_mtx : NULL,
+           so->so_usecount, lr_saved);

 #endif
        if (refcount)
                so->so_usecount--;
 #endif
        if (refcount)
                so->so_usecount--;

-       
-       if (so->so_usecount < 0)
-               panic("div_unlock: so=%x usecount=%x\n", so, so->so_usecount);
+
+       if (so->so_usecount < 0) {
+               panic("div_unlock: so=%p usecount=%x lrh= %s\n", 
+                   so, so->so_usecount, solockhistory_nr(so));
+               /* NOTREACHED */
+       }

        if (so->so_pcb == NULL) {
        if (so->so_pcb == NULL) {

-               panic("div_unlock: so=%x NO PCB usecount=%x lr=%x\n", so, so->so_usecount, lr_saved);
-               mutex_held = so->so_proto->pr_domain->dom_mtx;
-       } else {
-               mutex_held = ((struct inpcb *)so->so_pcb)->inpcb_mtx;
+               panic("div_unlock: so=%p NO PCB usecount=%x lr=%p lrh= %s\n",
+                   so, so->so_usecount, lr_saved, solockhistory_nr(so));
+               /* NOTREACHED */

        }
        }

+       mutex_held = ((struct inpcb *)so->so_pcb)->inpcb_mtx;

 
        if (so->so_usecount == 0 && (inp->inp_wantcnt == WNT_STOPUSING)) {
                lck_rw_lock_exclusive(divcbinfo.mtx);
 
        if (so->so_usecount == 0 && (inp->inp_wantcnt == WNT_STOPUSING)) {
                lck_rw_lock_exclusive(divcbinfo.mtx);

+               if (inp->inp_state != INPCB_STATE_DEAD)
+                       in_pcbdetach(inp);

                in_pcbdispose(inp);
                lck_rw_done(divcbinfo.mtx);
                return (0);
        }
        lck_mtx_assert(mutex_held, LCK_MTX_ASSERT_OWNED);
                in_pcbdispose(inp);
                lck_rw_done(divcbinfo.mtx);
                return (0);
        }
        lck_mtx_assert(mutex_held, LCK_MTX_ASSERT_OWNED);

+       so->unlock_lr[so->next_unlock_lr] = lr_saved;
+       so->next_unlock_lr = (so->next_unlock_lr+1) % SO_LCKDBG_MAX;

        lck_mtx_unlock(mutex_held);
        lck_mtx_unlock(mutex_held);

-       so->reserved4 = (void *)lr_saved;

        return (0);
 }
 
        return (0);
 }
 


@@ -727,10 +731,12 @@ div_getlock(struct socket *so, __unused int locktype)
        
        if (so->so_pcb)  {
                if (so->so_usecount < 0)
        
        if (so->so_pcb)  {
                if (so->so_usecount < 0)

-                       panic("div_getlock: so=%x usecount=%x\n", so, so->so_usecount);
+                       panic("div_getlock: so=%p usecount=%x lrh= %s\n", 
+                           so, so->so_usecount, solockhistory_nr(so));

                return(inpcb->inpcb_mtx);
        } else {
                return(inpcb->inpcb_mtx);
        } else {

-               panic("div_getlock: so=%x NULL so_pcb\n", so);
+               panic("div_getlock: so=%p NULL NO PCB lrh= %s\n", 
+                   so, solockhistory_nr(so));

                return (so->so_proto->pr_domain->dom_mtx);
        }
 }
                return (so->so_proto->pr_domain->dom_mtx);
        }
 }