/*
 * Copyright 2000, International Business Machines Corporation and others.
 * All Rights Reserved.
 * 
 * This software has been released under the terms of the IBM Public
 * License.  For details, see the LICENSE file in the top-level source
 * directory or online at http://www.openafs.org/dl/license10.html
 */

#include <afsconfig.h>
#include "afs/param.h"
#include "afs/opr.h"


#include <sys/malloc.h>
#include "rx/rx_kcommon.h"
#include "rx/rx_atomic.h"
#include "rx/rx_internal.h"

#ifdef RXK_LISTENER_ENV
int
osi_NetReceive(osi_socket asocket, struct sockaddr_in *addr,
	       struct iovec *dvec, int nvecs, int *alength)
{
    struct uio u;
    int i;
    struct iovec iov[RX_MAXIOVECS];
    struct sockaddr *sa = NULL;
    int code;

    int haveGlock = ISAFS_GLOCK();

    memset(&u, 0, sizeof(u));
    memset(&iov, 0, sizeof(iov));

    /*AFS_STATCNT(osi_NetReceive); */

    if (nvecs > RX_MAXIOVECS)
	osi_Panic("osi_NetReceive: %d: Too many iovecs.\n", nvecs);

    for (i = 0; i < nvecs; i++)
	iov[i] = dvec[i];

    u.uio_iov = &iov[0];
    u.uio_iovcnt = nvecs;
    u.uio_offset = 0;
    u.uio_resid = *alength;
    u.uio_segflg = UIO_SYSSPACE;
    u.uio_rw = UIO_READ;
    u.uio_td = NULL;

    if (haveGlock)
	AFS_GUNLOCK();
    code = soreceive(asocket, &sa, &u, NULL, NULL, NULL);
    if (haveGlock)
	AFS_GLOCK();

    if (code) {
#if KNET_DEBUG
	if (code == EINVAL)
	    Debugger("afs NetReceive busted");
	else
	    printf("y");
#else
	return code;
#endif
    }
    *alength -= u.uio_resid;
    if (sa) {
	if (sa->sa_family == AF_INET) {
	    if (addr)
		*addr = *(struct sockaddr_in *)sa;
	} else
	    printf("Unknown socket family %d in NetReceive\n", sa->sa_family);
	free(sa, M_SONAME);
    }
    return code;
}

extern int rxk_ListenerPid;
void
osi_StopListener(void)
{
    struct sockaddr_in taddr;
    struct iovec dvec;
    struct proc *p;
    char c;
    c = '\0';

    /*
     * Have to drop global lock to safely do this.
     * soclose() is currently protected by Giant,
     * but pfind and psignal are MPSAFE.
     */
    int haveGlock = ISAFS_GLOCK();
    if (haveGlock)
	AFS_GUNLOCK();
    soshutdown(rx_socket, SHUT_RDWR);
    p = pfind(rxk_ListenerPid);
    if (p) {
	afs_warn("osi_StopListener: rxk_ListenerPid %u\n", rxk_ListenerPid);
	kern_psignal(p, SIGUSR1);
	PROC_UNLOCK(p);
    } else
	afs_warn("osi_StopListener: rxk_Listener not found (pid %u)\n",
	    rxk_ListenerPid);

    /* Avoid destroying socket until osi_NetReceive has
    * had a chance to clean up.  Otherwise we can't restart. */
    bzero(&taddr, sizeof(taddr));
    taddr.sin_len = sizeof(struct sockaddr_in);
    taddr.sin_family = AF_INET;
    rxi_GetLocalAddr(&taddr);
    bzero(&dvec, sizeof(dvec));
    dvec.iov_base = &c;
    dvec.iov_len = 1;
    /* afs_osi_Sleep requires the GLOCK */
    AFS_GLOCK();
    while(rxk_ListenerPid) {
	afs_warn("waiting for rxk_ListenerPid to die\n");
	osi_NetSend(rx_socket, &taddr, &dvec, 1, 1, 0);
	afs_osi_Sleep(&rxk_ListenerPid);
    }
    AFS_GUNLOCK();
    /* in theory, we are now the only people doing anything with rx_socket */
    soclose(rx_socket);

    if (haveGlock)
	AFS_GLOCK();
}

int
osi_NetSend(osi_socket asocket, struct sockaddr_in *addr, struct iovec *dvec,
	    int nvecs, afs_int32 alength, int istack)
{
    afs_int32 code;
    int i;
    struct iovec iov[RX_MAXIOVECS];
    struct uio u;
    int haveGlock = ISAFS_GLOCK();

    memset(&u, 0, sizeof(u));
    memset(&iov, 0, sizeof(iov));

    AFS_STATCNT(osi_NetSend);
    if (nvecs > RX_MAXIOVECS)
	osi_Panic("osi_NetSend: %d: Too many iovecs.\n", nvecs);

    for (i = 0; i < nvecs; i++)
	iov[i] = dvec[i];

    u.uio_iov = &iov[0];
    u.uio_iovcnt = nvecs;
    u.uio_offset = 0;
    u.uio_resid = alength;
    u.uio_segflg = UIO_SYSSPACE;
    u.uio_rw = UIO_WRITE;
    u.uio_td = NULL;

    addr->sin_len = sizeof(struct sockaddr_in);

    if (haveGlock)
	AFS_GUNLOCK();
#if KNET_DEBUG
    printf("+");
#endif
    code =
	sosend(asocket, (struct sockaddr *)addr, &u, NULL, NULL, 0,
	       curthread);
#if KNET_DEBUG
    if (code) {
	if (code == EINVAL)
	    Debugger("afs NetSend busted");
	else
	    printf("z");
    }
#endif
    if (haveGlock)
	AFS_GLOCK();
    return code;
}
#else
/* This code *almost* works :( */
static struct protosw parent_proto;	/* udp proto switch */
static void rxk_input(struct mbuf *am, int iphlen);
static void rxk_fasttimo(void);

/* start intercepting basic calls */
rxk_init()
{
    struct protosw *tpro, *last;
    if (rxk_initDone)
	return 0;

    last = inetdomain.dom_protoswNPROTOSW;
    for (tpro = inetdomain.dom_protosw; tpro < last; tpro++)
	if (tpro->pr_protocol == IPPROTO_UDP) {
	    memcpy(&parent_proto, tpro, sizeof(parent_proto));
	    tpro->pr_input = rxk_input;
	    tpro->pr_fasttimo = rxk_fasttimo;
	    /*
	     * don't bother with pr_drain and pr_ctlinput
	     * until we have something to do
	     */
	    rxk_initDone = 1;
	    return 0;
	}
    osi_Panic("inet:no udp");
}


static void
rxk_input(struct mbuf *am, int iphlen)
{
    void (*tproc) ();
    unsigned short *tsp;
    int hdr;
    struct udphdr *tu;
    struct ip *ti;
    struct udpiphdr *tvu;
    int i;
    char *phandle;
    afs_int32 code;
    struct sockaddr_in taddr;
    int tlen;
    short port;
    int data_len, comp_sum;

    SPLVAR;
    NETPRI;

    /* make sure we have base ip and udp headers in first mbuf */
    if (iphlen > sizeof(struct ip)) {
	ip_stripoptions(am, NULL);
	iphlen = sizeof(struct ip);
    }

    if (am->m_len < sizeof(struct udpiphdr)) {
	am = m_pullup(am, sizeof(struct udpiphdr));
	if (!am) {
	    USERPRI;
	    return;
	}
    }

    ti = mtod(am, struct ip *);
    /* skip basic ip hdr */
    tu = (struct udphdr *)(((char *)ti) + sizeof(struct ip));

    /* now read the port out */
    port = tu->uh_dport;

    if (port) {
	for (tsp = rxk_ports, i = 0; i < MAXRXPORTS; i++) {
	    if (*tsp++ == port) {
		/* checksum the packet */
		/*
		 * Make mbuf data length reflect UDP length.
		 * If not enough data to reflect UDP length, drop.
		 */
		tvu = (struct udpiphdr *)ti;
		tlen = ntohs((u_short) tvu->ui_ulen);
		if ((int)ti->ip_len != tlen) {
		    if (tlen > (int)ti->ip_len) {
			m_free(am);
			USERPRI;
			return;
		    }
		    m_adj(am, tlen - (int)ti->ip_len);
		}
		/* deliver packet to rx */
		taddr.sin_family = AF_INET;	/* compute source address */
		taddr.sin_port = tu->uh_sport;
		taddr.sin_addr.s_addr = ti->ip_src.s_addr;
		taddr.sin_len = sizeof(taddr);
		tvu = (struct udpiphdr *)ti;	/* virtual udp structure, for cksum */
		/* handle the checksum.  Note that this code damages the actual ip
		 * header (replacing it with the virtual one, which is the same size),
		 * so we must ensure we get everything out we need, first */
		if (tu->uh_sum != 0) {
		    /* if the checksum is there, always check it. It's crazy not
		     * to, unless you can really be sure that your
		     * underlying network (and interfaces and drivers and
		     * DMA hardware, etc!) is error-free. First, fill
		     * in entire virtual ip header. */
		    memset(tvu->ui_i.ih_x1, 0, 9);
		    tvu->ui_len = tvu->ui_ulen;
		    tlen = ntohs((unsigned short)(tvu->ui_ulen));
		    if (in_cksum(am, sizeof(struct ip) + tlen)) {
			/* checksum, including cksum field, doesn't come out 0, so
			 * this packet is bad */
			m_freem(am);
			USERPRI;
			return;
		    }
		}

		/*
		 * 28 is IP (20) + UDP (8) header.  ulen includes
		 * udp header, and we *don't* tell RX about udp
		 * header either.  So, we remove those 8 as well.
		 */
		data_len = ntohs(tu->uh_ulen);
		data_len -= 8;
		if (!(*rxk_GetPacketProc) (&phandle, data_len)) {
		    if (rx_mb_to_packet(am, m_freem, 28, data_len, phandle)) {
			/* XXX should just increment counter here.. */
			printf("rx: truncated UDP packet\n");
			rxi_FreePacket(phandle);
		    } else
			(*rxk_PacketArrivalProc) (phandle, &taddr,
						  rxk_portRocks[i], data_len);
		} else
		    m_freem(am);
		USERPRI;
		return;
	    }
	}
    }

    /* if we get here, try to deliver packet to udp */
    if (tproc = parent_proto.pr_input)
	(*tproc) (am, iphlen);
    USERPRI;
    return;
}


/* 
 * UDP fast timer to raise events for all but Solaris and NCR. 
 * Called about 5 times per second (at unknown priority?).  Must go to
 * splnet or obtain global lock before touching anything significant.
 */
static void
rxk_fasttimo(void)
{
    void (*tproc) ();
    struct clock temp;

    /* do rx fasttimo processing here */
    rxevent_RaiseEvents(&temp);
    if (tproc = parent_proto.pr_fasttimo)
	(*tproc) ();
}

/* rx_NetSend - send asize bytes at adata from asocket to host at addr.
 *
 * Now, why do we allocate a new buffer when we could theoretically use the one
 * pointed to by adata?  Because PRU_SEND returns after queueing the message,
 * not after sending it.  If the sender changes the data after queueing it,
 * we'd see the already-queued data change.  One attempt to fix this without
 * adding a copy would be to have this function wait until the datagram is
 * sent; however this doesn't work well.  In particular, if a host is down, and
 * an ARP fails to that host, this packet will be queued until the ARP request
 * comes back, which could be hours later.  We can't block in this routine that
 * long, since it prevents RPC timeouts from happening.
 */
/* XXX In the brave new world, steal the data bufs out of the rx_packet iovec,
 * and just queue those.  XXX
 */

/* set lock on sockbuf sb; can't call sblock since we're at interrupt level
 * sometimes */
static
trysblock(sb)
     struct sockbuf *sb;
{
    AFS_STATCNT(trysblock);
    if (sb->sb_flags & SB_LOCK) {
	return -1;		/* can't lock socket */
    }
    sb->sb_flags |= SB_LOCK;
    return 0;
}

/* We only have to do all the mbuf management ourselves if we can be called at
   interrupt time. in RXK_LISTENER_ENV, we can just call sosend() */
int
osi_NetSend(osi_socket asocket, struct sockaddr_in *addr, struct iovec *dvec,
	    int nvec, afs_int32 asize, int istack)
{
    struct mbuf *tm, *um;
    afs_int32 code;
    int s;
    struct mbuf *top = 0;
    struct mbuf *m, **mp;
    int len;
    char *tdata;
    caddr_t tpa;
    int i, tl, rlen;
    int mlen;
    int haveGlock;
#if KNET_DEBUG
    static int before = 0;
#endif

    AFS_STATCNT(osi_NetSend);
/* Actually, the Ultrix way is as good as any for us, so we don't bother with
 * special mbufs any more.  Used to think we could get away with not copying
 * the data to the interface, but there's no way to tell the caller not to
 * reuse the buffers after sending, so we lost out on that trick anyway */
    s = splnet();
    if (trysblock(&asocket->so_snd)) {
	splx(s);
	return 1;
    }
    mp = &top;
    i = 0;
    tdata = dvec[i].iov_base;
    tl = dvec[i].iov_len;
    while (1) {
	mlen = MLEN;
	if (top == 0) {
	    MGETHDR(m, M_DONTWAIT, MT_DATA);
	    if (!m) {
		sbunlock(&asocket->so_snd);
		splx(s);
		return 1;
	    }
	    mlen = MHLEN;
	    m->m_pkthdr.len = 0;
	    m->m_pkthdr.rcvif = NULL;
	} else
	    MGET(m, M_DONTWAIT, MT_DATA);
	if (!m) {
	    /* can't get an mbuf, give up */
	    if (top)
		m_freem(top);	/* free mbuf list we're building */
	    sbunlock(&asocket->so_snd);
	    splx(s);
	    return 1;
	}
	/*
	 * WARNING: the `4 * MLEN' is somewhat dubious.  It is better than
	 * `NBPG', which may have no relation to `CLBYTES'.  Also, `CLBYTES'
	 * may be so large that we never use clusters, resulting in far
	 * too many mbufs being used.  It is often better to briefly use
	 * a cluster, even if we are only using a portion of it.  Since
	 * we are on the xmit side, it shouldn't end up sitting on a queue
	 * for a potentially unbounded time (except perhaps if we are talking
	 * to ourself).
	 */
	if (asize >= 4 * MLEN) {	/* try to get cluster mbuf */
	    /* different algorithms for getting cluster mbuf */
	    MCLGET(m, M_DONTWAIT);
	    if ((m->m_flags & M_EXT) == 0)
		goto nopages;
	    mlen = MCLBYTES;

	    /* now compute usable size */
	    len = opr_min(mlen, asize);
/* Should I look at MAPPED_MBUFS??? */
	} else {
	  nopages:
	    len = opr_min(mlen, asize);
	}
	m->m_len = 0;
	*mp = m;		/* XXXX */
	top->m_pkthdr.len += len;
	tpa = mtod(m, caddr_t);
	while (len) {
	    rlen = opr_min(len, tl);
	    memcpy(tpa, tdata, rlen);
	    asize -= rlen;
	    len -= rlen;
	    tpa += rlen;
	    m->m_len += rlen;
	    tdata += rlen;
	    tl -= rlen;
	    if (tl <= 0) {
		i++;
		if (i > nvec) {
		    /* shouldn't come here! */
		    asize = 0;	/* so we make progress toward completion */
		    break;
		}
		tdata = dvec[i].iov_base;
		tl = dvec[i].iov_len;
	    }
	}
	*mp = m;
	mp = &m->m_next;
	if (asize <= 0)
	    break;
    }
    tm = top;

    tm->m_act = NULL;

    /* setup mbuf corresponding to destination address */
    um = m_get(M_DONTWAIT, MT_SONAME);
    if (!um) {
	if (top)
	    m_freem(top);	/* free mbuf chain */
	sbunlock(&asocket->so_snd);
	splx(s);
	return 1;
    }
    memcpy(mtod(um, caddr_t), addr, sizeof(*addr));
    addr->sin_len = um->m_len = sizeof(*addr);
    /* note that udp_usrreq frees funny mbuf.  We hold onto data, but mbuf
     * around it is gone. */
    /*    haveGlock = ISAFS_GLOCK();
     * if (haveGlock) {
     * AFS_GUNLOCK();
     * }  */
    /* SOCKET_LOCK(asocket); */
    /* code = (*asocket->so_proto->pr_usrreq)(asocket, PRU_SEND, tm, um, 0); */
#if KNET_DEBUG
    if (before)
	Debugger("afs NetSend before");
#endif
    code =
	(*asocket->so_proto->pr_usrreqs->pru_send) (asocket, 0, tm,
						    (struct sockaddr *)
						    addr, um, &proc0);
    /* SOCKET_UNLOCK(asocket); */
    /* if (haveGlock) {
     * AFS_GLOCK();
     * } */
    sbunlock(&asocket->so_snd);
    splx(s);
#if KNET_DEBUG
    if (code) {
	if (code == EINVAL)
	    Debugger("afs NetSend busted");
	else
	    printf("z");
    }
#endif
    return code;
}
#endif
