sun4v/etm/etm.c

/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License, Version 1.0 only
 * (the "License").  You may not use this file except in compliance
 * with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */

/*
 * Copyright 2005 Sun Microsystems, Inc.  All rights reserved.
 * Use is subject to license terms.
 */

/*
 * etm.c	FMA Event Transport Module implementation, a plugin of FMD
 *		for sun4v/Ontario
 *
 * plugin for sending/receiving FMA events to/from service processor
 */

#pragma ident	"%Z%%M%	%I%	%E% SMI"

/*
 * --------------------------------- includes --------------------------------
 */

#include <sys/fm/protocol.h>
#include <sys/sysevent/eventdefs.h>
#include <sys/fm/util.h>
#include <netinet/in.h>
#include <fm/fmd_api.h>
#include <libsysevent.h>

#include "etm_xport_api.h"
#include "etm_etm_proto.h"
#include "etm_impl.h"

#include <pthread.h>
#include <signal.h>
#include <stropts.h>
#include <locale.h>
#include <strings.h>
#include <stdlib.h>
#include <unistd.h>
#include <limits.h>
#include <values.h>
#include <alloca.h>
#include <errno.h>
#include <fcntl.h>
#include <time.h>

/*
 * ----------------------------- forward decls -------------------------------
 */

static void
etm_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl, const char *class);

/*
 * ------------------------- data structs for FMD ----------------------------
 */

static const fmd_hdl_ops_t fmd_ops = {
	etm_recv,	/* fmdo_recv */
	NULL,		/* fmdo_timeout */
	NULL,		/* fmdo_close */
	NULL,		/* fmdo_stats */
	NULL,		/* fmdo_gc */
};

static const fmd_prop_t fmd_props[] = {
	{ ETM_PROP_NM_XPORT_ADDRS,	FMD_TYPE_STRING, "" },
	{ ETM_PROP_NM_DEBUG_LVL,	FMD_TYPE_INT32, "0" },
	{ ETM_PROP_NM_DEBUG_MAX_EV_CNT,	FMD_TYPE_INT32, "-1" },
	{ NULL, 0, NULL }
};

static const fmd_hdl_info_t fmd_info = {
	"FMA Event Transport Module", "1.0", &fmd_ops, fmd_props
};

/*
 * ----------------------- private consts and defns --------------------------
 */

/* misc buffer for variable sized protocol header fields */

#define	ETM_MISC_BUF_SZ	(4 * 1024)

/* try limit for IO operations w/ capped exp backoff sleep on retry */

/*
 * Design_Note:	ETM will potentially retry forever IO operations that the
 *		transport fails with EAGAIN (aka EWOULDBLOCK) rather than
 *		giving up after some number of seconds. This avoids
 *		dropping FMA events while the service processor is down,
 *		but at the risk of pending fmdo_recv() forever and
 *		overflowing FMD's event queue for ETM.
 *		A future TBD enhancement would be to always recv
 *		and send each ETM msg in a single read/write() to reduce
 *		the risk of failure between ETM msg hdr and body,
 *		assuming the MTU_SZ is large enough.
 */

#define	ETM_TRY_MAX_CNT		(MAXINT - 1)
#define	ETM_TRY_BACKOFF_RATE	(4)
#define	ETM_TRY_BACKOFF_CAP	(60)

/* protocol transaction id consts for starting id and increment amt */

#define	ETM_XID_BEG	(1)
#define	ETM_XID_INC	(2)

/*
 * ---------------------------- global data ----------------------------------
 */

static int
etm_debug_lvl = 0;	/* debug level: 0 is off, 1 is on, 2 is more, etc */

static int
etm_debug_max_ev_cnt = -1; /* max allowed event count for debugging */

static pthread_t
etm_svr_tid = NULL;	/* thread id of connection acceptance server */

static volatile int
etm_is_dying = 0;	/* bool for dying (killing self) */

static uint32_t
etm_xid_cur = ETM_XID_BEG; /* current transaction id for sends */

static uint32_t
etm_xid_ping = 0;	/* xid of last CONTROL msg sent requesting ping */

static uint32_t
etm_xid_ver_set = 0;	/* xid of last CONTROL msg sent requesting ver set */

static struct stats {

	/* ETM msg counters */

	fmd_stat_t etm_rd_hdr_fmaevent;
	fmd_stat_t etm_rd_hdr_control;
	fmd_stat_t etm_rd_hdr_response;
	fmd_stat_t etm_rd_body_fmaevent;
	fmd_stat_t etm_rd_body_control;
	fmd_stat_t etm_rd_body_response;
	fmd_stat_t etm_wr_hdr_fmaevent;
	fmd_stat_t etm_wr_hdr_control;
	fmd_stat_t etm_wr_hdr_response;
	fmd_stat_t etm_wr_body_fmaevent;
	fmd_stat_t etm_wr_body_control;
	fmd_stat_t etm_wr_body_response;

	/* ETM byte counters */

	fmd_stat_t etm_wr_fmd_bytes;
	fmd_stat_t etm_rd_fmd_bytes;
	fmd_stat_t etm_wr_xport_bytes;
	fmd_stat_t etm_rd_xport_bytes;

	fmd_stat_t etm_magic_drop_bytes;

	/* ETM [dropped] FMA event counters */

	fmd_stat_t etm_rd_fmd_fmaevent;
	fmd_stat_t etm_wr_fmd_fmaevent;

	fmd_stat_t etm_rd_drop_fmaevent;
	fmd_stat_t etm_wr_drop_fmaevent;

	/* ETM protocol failures */

	fmd_stat_t etm_magic_bad;
	fmd_stat_t etm_ver_bad;
	fmd_stat_t etm_msgtype_bad;
	fmd_stat_t etm_subtype_bad;
	fmd_stat_t etm_xid_bad;
	fmd_stat_t etm_fmaeventlen_bad;
	fmd_stat_t etm_respcode_bad;
	fmd_stat_t etm_timeout_bad;
	fmd_stat_t etm_evlens_bad;

	/* IO operation failures */

	fmd_stat_t etm_xport_wr_fail;
	fmd_stat_t etm_xport_rd_fail;
	fmd_stat_t etm_xport_pk_fail;

	/* IO operation retries */

	fmd_stat_t etm_xport_wr_retry;
	fmd_stat_t etm_xport_rd_retry;
	fmd_stat_t etm_xport_pk_retry;

	/* system and library failures */

	fmd_stat_t etm_os_sysevent_publish_fail;
	fmd_stat_t etm_os_sysevent_bind_fail;
	fmd_stat_t etm_os_nvlist_pack_fail;
	fmd_stat_t etm_os_nvlist_unpack_fail;
	fmd_stat_t etm_os_nvlist_size_fail;
	fmd_stat_t etm_os_pthread_create_fail;

	/* xport API failures */

	fmd_stat_t etm_xport_get_ev_addrv_fail;
	fmd_stat_t etm_xport_open_fail;
	fmd_stat_t etm_xport_close_fail;
	fmd_stat_t etm_xport_accept_fail;
	fmd_stat_t etm_xport_open_retry;

	/* FMD entry point bad arguments */

	fmd_stat_t etm_fmd_recv_badargs;
	fmd_stat_t etm_fmd_init_badargs;
	fmd_stat_t etm_fmd_fini_badargs;

} etm_stats = {

	/* ETM msg counters */

	{ "etm_rd_hdr_fmaevent", FMD_TYPE_UINT64,
		"ETM fmaevent msg headers rcvd from xport" },
	{ "etm_rd_hdr_control", FMD_TYPE_UINT64,
		"ETM control msg headers rcvd from xport" },
	{ "etm_rd_hdr_response", FMD_TYPE_UINT64,
		"ETM response msg headers rcvd from xport" },
	{ "etm_rd_body_fmaevent", FMD_TYPE_UINT64,
		"ETM fmaevent msg bodies rcvd from xport" },
	{ "etm_rd_body_control", FMD_TYPE_UINT64,
		"ETM control msg bodies rcvd from xport" },
	{ "etm_rd_body_response", FMD_TYPE_UINT64,
		"ETM response msg bodies rcvd from xport" },
	{ "etm_wr_hdr_fmaevent", FMD_TYPE_UINT64,
		"ETM fmaevent msg headers sent to xport" },
	{ "etm_wr_hdr_control", FMD_TYPE_UINT64,
		"ETM control msg headers sent to xport" },
	{ "etm_wr_hdr_response", FMD_TYPE_UINT64,
		"ETM response msg headers sent to xport" },
	{ "etm_wr_body_fmaevent", FMD_TYPE_UINT64,
		"ETM fmaevent msg bodies sent to xport" },
	{ "etm_wr_body_control", FMD_TYPE_UINT64,
		"ETM control msg bodies sent to xport" },
	{ "etm_wr_body_response", FMD_TYPE_UINT64,
		"ETM response msg bodies sent to xport" },

	/* ETM byte counters */

	{ "etm_wr_fmd_bytes", FMD_TYPE_UINT64,
		"bytes of FMA events sent to FMD" },
	{ "etm_rd_fmd_bytes", FMD_TYPE_UINT64,
		"bytes of FMA events rcvd from FMD" },
	{ "etm_wr_xport_bytes", FMD_TYPE_UINT64,
		"bytes of FMA events sent to xport" },
	{ "etm_rd_xport_bytes", FMD_TYPE_UINT64,
		"bytes of FMA events rcvd from xport" },

	{ "etm_magic_drop_bytes", FMD_TYPE_UINT64,
		"bytes dropped from xport pre magic num" },

	/* ETM [dropped] FMA event counters */

	{ "etm_rd_fmd_fmaevent", FMD_TYPE_UINT64,
		"FMA events rcvd from FMD" },
	{ "etm_wr_fmd_fmaevent", FMD_TYPE_UINT64,
		"FMA events sent to FMD" },

	{ "etm_rd_drop_fmaevent", FMD_TYPE_UINT64,
		"dropped FMA events from xport" },
	{ "etm_wr_drop_fmaevent", FMD_TYPE_UINT64,
		"dropped FMA events to xport" },

	/* ETM protocol failures */

	{ "etm_magic_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid magic num" },
	{ "etm_ver_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid protocol version" },
	{ "etm_msgtype_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid message type" },
	{ "etm_subtype_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid sub type" },
	{ "etm_xid_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ unmatched xid" },
	{ "etm_fmaeventlen_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid FMA event length" },
	{ "etm_respcode_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid response code" },
	{ "etm_timeout_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ invalid timeout value" },
	{ "etm_evlens_bad", FMD_TYPE_UINT64,
		"ETM msgs w/ too many event lengths" },

	/* IO operation failures */

	{ "etm_xport_wr_fail", FMD_TYPE_UINT64,
		"xport write failures" },
	{ "etm_xport_rd_fail", FMD_TYPE_UINT64,
		"xport read failures" },
	{ "etm_xport_pk_fail", FMD_TYPE_UINT64,
		"xport peek failures" },

	/* IO operation retries */

	{ "etm_xport_wr_retry", FMD_TYPE_UINT64,
		"xport write retries" },
	{ "etm_xport_rd_retry", FMD_TYPE_UINT64,
		"xport read retries" },
	{ "etm_xport_pk_retry", FMD_TYPE_UINT64,
		"xport peek retries" },

	/* system and library failures */

	{ "etm_os_sysevent_publish_fail", FMD_TYPE_UINT64,
		"sysevent_evc_publish failures" },
	{ "etm_os_sysevent_bind_fail", FMD_TYPE_UINT64,
		"sysevent_evc_bind failures" },
	{ "etm_os_nvlist_pack_fail", FMD_TYPE_UINT64,
		"nvlist_pack failures" },
	{ "etm_os_nvlist_unpack_fail", FMD_TYPE_UINT64,
		"nvlist_unpack failures" },
	{ "etm_os_nvlist_size_fail", FMD_TYPE_UINT64,
		"nvlist_size failures" },
	{ "etm_os_pthread_create_fail", FMD_TYPE_UINT64,
		"pthread_create failures" },

	/* transport API failures */

	{ "etm_xport_get_ev_addrv_fail", FMD_TYPE_UINT64,
		"xport get event addrv API failures" },
	{ "etm_xport_open_fail", FMD_TYPE_UINT64,
		"xport open API failures" },
	{ "etm_xport_close_fail", FMD_TYPE_UINT64,
		"xport close API failures" },
	{ "etm_xport_accept_fail", FMD_TYPE_UINT64,
		"xport accept API failures" },
	{ "etm_xport_open_retry", FMD_TYPE_UINT64,
		"xport open API retries" },

	/* FMD entry point bad arguments */

	{ "etm_fmd_recv_badargs", FMD_TYPE_UINT64,
		"bad arguments from fmd_recv entry point" },
	{ "etm_fmd_init_badargs", FMD_TYPE_UINT64,
		"bad arguments from fmd_init entry point" },
	{ "etm_fmd_fini_badargs", FMD_TYPE_UINT64,
		"bad arguments from fmd_fini entry point" }
};

/*
 * -------------------------- support functions ------------------------------
 */

/*
 * Design_Note:	Each failure worth reporting to FMD should be done using
 *		a single call to fmd_hdl_error() as it logs an FMA event
 *		for each call. Also be aware that all the fmd_hdl_*()
 *		format strings currently use platform specific *printf()
 *		routines; so "%p" under Solaris does not prepend "0x" to
 *		the outputted hex digits, while Linux and VxWorks do.
 */

/*
 * etm_hexdump - hexdump the given buffer (for debugging) using
 *		the given FMD module handle
 */

static void
etm_hexdump(fmd_hdl_t *hdl, void *buf, size_t byte_cnt)
{
	uint8_t		*bp;		/* byte ptr */
	int		i, j;		/* index */
	char		cb[80];		/* char buf */
	unsigned int	n;		/* a byte of data for sprintf() */

	bp = buf;
	j = 0;

	/*
	 * Design_Note:	fmd_hdl_debug() auto adds a newline if missing;
	 *		hence cb exists to accumulate a longer string.
	 */

	for (i = 1; i <= byte_cnt; i++) {
		n = *bp++;
		(void) sprintf(&cb[j], "%2.2x ", n);
		j += 3;
		/* add a newline every 16 bytes or at the buffer's end */
		if (((i % 16) == 0) || (i >= byte_cnt)) {
			cb[j-1] = '\0';
			fmd_hdl_debug(hdl, "%s\n", cb);
			j = 0;
		}
	} /* for each byte in the buffer */

} /* etm_hexdump() */

/*
 * etm_sleep - sleep the caller for the given number of seconds,
 *		return 0 or -errno value
 *
 * Design_Note:	To avoid interfering with FMD's signal mask (SIGALRM)
 *		do not use [Solaris] sleep(3C) and instead use
 *		pthread_cond_wait() or nanosleep(), both of which
 *		are POSIX spec-ed to leave signal masks alone.
 *		This is needed for Solaris and Linux (domain and SP).
 */

static int
etm_sleep(unsigned sleep_sec)
{
	struct timespec	tms;	/* for nanosleep() */

	tms.tv_sec = sleep_sec;
	tms.tv_nsec = 0;

	if (nanosleep(&tms, NULL) < 0) {
		/* errno assumed set by above call */
		return (-errno);
	}
	return (0);

} /* etm_sleep() */

/*
 * etm_conn_open - open a connection to the given transport address,
 *		return 0 and the opened connection handle
 *		or -errno value
 *
 * caveats:	the err_substr is used in failure cases for calling
 *		fmd_hdl_error()
 */

static int
etm_conn_open(fmd_hdl_t *hdl, char *err_substr,
		etm_xport_addr_t addr, etm_xport_conn_t *connp)
{
	etm_xport_conn_t	conn;	/* connection to return */
	int			nev;	/* -errno value */

	if ((conn = etm_xport_open(hdl, addr)) == NULL) {
		nev = (-errno);
		fmd_hdl_error(hdl, "error: %s: errno %d\n",
					err_substr, errno);
		etm_stats.etm_xport_open_fail.fmds_value.ui64++;
		return (nev);
	} else {
		*connp = conn;
		return (0);
	}
} /* etm_conn_open() */

/*
 * etm_conn_close - close the given connection,
 *		return 0 or -errno value
 *
 * caveats:	the err_substr is used in failure cases for calling
 *		fmd_hdl_error()
 */

static int
etm_conn_close(fmd_hdl_t *hdl, char *err_substr, etm_xport_conn_t conn)
{
	int	nev;	/* -errno value */

	if (etm_xport_close(hdl, conn) == NULL) {
		nev = (-errno);
		fmd_hdl_error(hdl, "warning: %s: errno %d\n",
					err_substr, errno);
		etm_stats.etm_xport_close_fail.fmds_value.ui64++;
		return (nev);
	} else {
		return (0);
	}
} /* etm_conn_close() */

/*
 * etm_io_op - perform an IO operation on the given connection
 *		with the given buffer,
 *		accommodating MTU size and retrying op if needed,
 *		return how many bytes actually done by the op
 *		or -errno value
 *
 * caveats:	the err_substr is used in failure cases for calling
 *		fmd_hdl_error()
 */

static ssize_t
etm_io_op(fmd_hdl_t *hdl, char *err_substr, etm_xport_conn_t conn,
				void *buf, size_t byte_cnt, int io_op)
{
	ssize_t		rv;		/* ret val / byte count */
	ssize_t		n;		/* gen use */
	uint8_t		*datap;		/* ptr to data */
	size_t		mtu_sz;		/* MTU size in bytes */
	int		(*io_func_ptr)(fmd_hdl_t *, etm_xport_conn_t,
							void *, size_t);
	size_t		io_sz;		/* byte count for io_func_ptr */
	int		try_cnt;	/* number of tries done */
	int		sleep_sec;	/* exp backoff sleep period in sec */
	int		sleep_rv;	/* ret val from sleeping */
	fmd_stat_t	io_retry_stat;	/* IO retry stat to update */
	fmd_stat_t	io_fail_stat;	/* IO failure stat to update */

	if ((conn == NULL) || (buf == NULL)) {
		return (-EINVAL);
	}
	switch (io_op) {
		case ETM_IO_OP_RD:
			io_func_ptr = etm_xport_read;
			io_retry_stat = etm_stats.etm_xport_rd_retry;
			io_fail_stat = etm_stats.etm_xport_rd_fail;
			break;
		case ETM_IO_OP_WR:
			io_func_ptr = etm_xport_write;
			io_retry_stat = etm_stats.etm_xport_wr_retry;
			io_fail_stat = etm_stats.etm_xport_wr_fail;
			break;
		case ETM_IO_OP_PK:
			io_func_ptr = etm_xport_peek;
			io_retry_stat = etm_stats.etm_xport_pk_retry;
			io_fail_stat = etm_stats.etm_xport_pk_fail;
			break;
		default:
			return (-EINVAL);
	}
	if (byte_cnt == 0) {
		return (byte_cnt);	/* nop */
	}

	/* obtain [current] MTU size */

	if ((n = etm_xport_get_opt(hdl, conn, ETM_XPORT_OPT_MTU_SZ)) < 0) {
		mtu_sz = ETM_XPORT_MTU_SZ_DEF;
	} else {
		mtu_sz = n;
	}

	/* loop until all IO done, try limit exceeded, or real failure */

	rv = 0;
	datap = buf;
	while (rv < byte_cnt) {
		io_sz = MIN((byte_cnt - rv), mtu_sz);
		try_cnt = 0;
		sleep_sec = 0;

		/* when give up, return -errno value even if partly done */

		while ((n = (*io_func_ptr)(hdl, conn, datap, io_sz)) ==
								(-EAGAIN)) {
			try_cnt++;
			if (try_cnt > ETM_TRY_MAX_CNT) {
				rv = n;
				goto func_ret;
			}
			if (etm_is_dying) {
				rv = (-EINTR);
				goto func_ret;
			}
			if ((sleep_rv = etm_sleep(sleep_sec)) < 0) {
				rv = sleep_rv;
				goto func_ret;
			}
			sleep_sec = ((sleep_sec == 0) ? 1 :
					(sleep_sec * ETM_TRY_BACKOFF_RATE));
			sleep_sec = MIN(sleep_sec, ETM_TRY_BACKOFF_CAP);
			io_retry_stat.fmds_value.ui64++;
			if (etm_debug_lvl >= 1) {
				fmd_hdl_debug(hdl, "info: retrying io op %d "
						"due to EAGAIN\n", io_op);
			}
		} /* while trying the io operation */

		if (etm_is_dying) {
			rv = (-EINTR);
			goto func_ret;
		}
		if (n < 0) {
			rv = n;
			goto func_ret;
		}
		/* avoid spinning CPU when given 0 bytes but no error */
		if (n == 0) {
			if ((sleep_rv = etm_sleep(ETM_SLEEP_QUIK)) < 0) {
				rv = sleep_rv;
				goto func_ret;
			}
		}
		rv += n;
		datap += n;
	} /* while still have more data */

func_ret:

	if (rv < 0) {
		io_fail_stat.fmds_value.ui64++;
		fmd_hdl_error(hdl, "error: %s: errno %d\n",
					err_substr, (int)(-rv));
	}
	if (etm_debug_lvl >= 3) {
		fmd_hdl_debug(hdl, "info: io op %d ret %d of %d\n",
					io_op, (int)rv, (int)byte_cnt);
	}
	return (rv);

} /* etm_io_op() */

/*
 * etm_magic_read - read the magic number of an ETM message header
 *		from the given connection into the given buffer,
 *		return 0 or -errno value
 *
 * Design_Note:	This routine is intended to help protect ETM from protocol
 *		framing errors as might be caused by an SP reset / crash in
 *		the middle of an ETM message send; the connection will be
 *		read from for as many bytes as needed until the magic number
 *		is found using a sliding buffer for comparisons.
 */

static int
etm_magic_read(fmd_hdl_t *hdl, etm_xport_conn_t conn, uint32_t *magic_ptr)
{
	int		rv;		/* ret val */
	uint32_t	magic_num;	/* magic number */
	int		byte_cnt;	/* count of bytes read */
	uint8_t		buf5[4+1];	/* sliding input buffer */
	int		i, j;		/* indices into buf5 */
	ssize_t		n;		/* gen use */
	uint8_t		drop_buf[1024];	/* dropped bytes buffer */

	rv = 0;		/* assume success */
	magic_num = 0;
	byte_cnt = 0;
	j = 0;

	/* magic number bytes are sent in network (big endian) order */

	while (magic_num != ETM_PROTO_MAGIC_NUM) {
		if ((n = etm_io_op(hdl, "bad io read on magic",
				conn, &buf5[j], 1, ETM_IO_OP_RD)) < 0) {
			rv = n;
			goto func_ret;
		}
		byte_cnt++;
		j = MIN((j + 1), sizeof (magic_num));
		if (byte_cnt < sizeof (magic_num)) {
			continue;
		}

		if (byte_cnt > sizeof (magic_num)) {
			etm_stats.etm_magic_drop_bytes.fmds_value.ui64++;
			i = MIN(byte_cnt - j - 1, sizeof (drop_buf) - 1);
			drop_buf[i] = buf5[0];
			for (i = 0; i < j; i++) {
				buf5[i] = buf5[i+1];
			} /* for sliding the buffer contents */
		}
		(void) memcpy(&magic_num, &buf5[0], sizeof (magic_num));
		magic_num = ntohl(magic_num);
	} /* for reading bytes until find magic number */

func_ret:

	if (byte_cnt != sizeof (magic_num)) {
		fmd_hdl_error(hdl, "warning: bad proto frame "
				"implies corrupt/lost msg(s)\n");
	}
	if ((byte_cnt > sizeof (magic_num)) && (etm_debug_lvl >= 2)) {
		i = MIN(byte_cnt - sizeof (magic_num), sizeof (drop_buf));
		fmd_hdl_debug(hdl, "info: magic drop hexdump "
				"first %d of %d bytes:\n",
				i, byte_cnt - sizeof (magic_num));
		etm_hexdump(hdl, drop_buf, i);
	}

	if (rv == 0) {
		*magic_ptr = magic_num;
	}
	return (rv);

} /* etm_magic_read() */

/*
 * etm_hdr_read - allocate, read, and validate a [variable sized]
 *		ETM message header from the given connection,
 *		return the allocated ETM message header
 *		(which is guaranteed to be large enough to reuse as a
 *		RESPONSE msg hdr) and its size
 *		or NULL and set errno on failure
 */

static void *
etm_hdr_read(fmd_hdl_t *hdl, etm_xport_conn_t conn, size_t *szp)
{
	uint8_t			*hdrp;		/* ptr to header to return */
	size_t			hdr_sz;		/* sizeof *hdrp */
	etm_proto_v1_pp_t	pp; 		/* protocol preamble */
	etm_proto_v1_ev_hdr_t	*ev_hdrp;	/* for FMA_EVENT msg */
	etm_proto_v1_ctl_hdr_t	*ctl_hdrp;	/* for CONTROL msg */
	etm_proto_v1_resp_hdr_t *resp_hdrp;	/* for RESPONSE msg */
	uint32_t		*lenp;		/* ptr to FMA event length */
	ssize_t			i, n;		/* gen use */
	uint8_t	misc_buf[ETM_MISC_BUF_SZ];	/* for var sized hdrs */
	int			dummy_int;	/* dummy var to appease lint */

	hdrp = NULL; hdr_sz = 0;

	/* read the magic number which starts the protocol preamble */

	if ((n = etm_magic_read(hdl, conn, &pp.pp_magic_num)) < 0) {
		errno = (-n);
		etm_stats.etm_magic_bad.fmds_value.ui64++;
		return (NULL);
	}

	/* read the rest of the protocol preamble all at once */

	if ((n = etm_io_op(hdl, "bad io read on preamble",
				conn, &pp.pp_proto_ver,
				sizeof (pp) - sizeof (pp.pp_magic_num),
				ETM_IO_OP_RD)) < 0) {
		errno = (-n);
		return (NULL);
	}

	/*
	 * Design_Note:	The magic number was already network decoded; but
	 *		some other preamble fields also need to be decoded,
	 *		specifically pp_xid and pp_timeout. The rest of the
	 *		preamble fields are byte sized and hence need no
	 *		decoding.
	 */

	pp.pp_xid = ntohl(pp.pp_xid);
	pp.pp_timeout = ntohl(pp.pp_timeout);

	/* sanity check the header as best we can */

	if (pp.pp_proto_ver != ETM_PROTO_V1) {
		fmd_hdl_error(hdl, "error: bad proto ver %d\n",
					(int)pp.pp_proto_ver);
		errno = EPROTO;
		etm_stats.etm_ver_bad.fmds_value.ui64++;
		return (NULL);
	}

	dummy_int = pp.pp_msg_type;
	if ((dummy_int <= ETM_MSG_TYPE_TOO_LOW) ||
	    (dummy_int >= ETM_MSG_TYPE_TOO_BIG)) {
		fmd_hdl_error(hdl, "error: bad msg type %d", dummy_int);
		errno = EBADMSG;
		etm_stats.etm_msgtype_bad.fmds_value.ui64++;
		return (NULL);
	}

	/* handle [var sized] hdrs for FMA_EVENT, CONTROL, RESPONSE msgs */

	if (pp.pp_msg_type == ETM_MSG_TYPE_FMA_EVENT) {

		ev_hdrp = (void*)&misc_buf[0];
		hdr_sz = sizeof (*ev_hdrp);
		(void) memcpy(&ev_hdrp->ev_pp, &pp, sizeof (pp));

		/* sanity check the header's timeout */

		if (ev_hdrp->ev_pp.pp_timeout != ETM_PROTO_V1_TIMEOUT_NONE) {
			errno = ETIME;
			etm_stats.etm_timeout_bad.fmds_value.ui64++;
			return (NULL);
		}

		/* get all FMA event lengths from the header */

		lenp = (uint32_t *)&ev_hdrp->ev_lens[0]; lenp--;
		i = -1;	/* cnt of length entries preceding 0 */
		do {
			i++; lenp++;
			if ((sizeof (*ev_hdrp) + (i * sizeof (*lenp))) >=
							ETM_MISC_BUF_SZ) {
				errno = E2BIG;	/* ridiculous size */
				etm_stats.etm_evlens_bad.fmds_value.ui64++;
				return (NULL);
			}
			if ((n = etm_io_op(hdl, "bad io read on event len",
						conn, lenp, sizeof (*lenp),
						ETM_IO_OP_RD)) < 0) {
				errno = (-n);
				return (NULL);
			}
			*lenp = ntohl(*lenp);

		} while (*lenp != 0);
		i += 0; /* first len already counted by sizeof(ev_hdr) */
		hdr_sz += (i * sizeof (*lenp));

		etm_stats.etm_rd_hdr_fmaevent.fmds_value.ui64++;

	} else if (pp.pp_msg_type == ETM_MSG_TYPE_CONTROL) {

		ctl_hdrp = (void*)&misc_buf[0];
		hdr_sz = sizeof (*ctl_hdrp);
		(void) memcpy(&ctl_hdrp->ctl_pp, &pp, sizeof (pp));

		/* sanity check the header's sub type (control selector) */

		if ((ctl_hdrp->ctl_pp.pp_sub_type <= ETM_CTL_SEL_TOO_LOW) ||
		    (ctl_hdrp->ctl_pp.pp_sub_type >= ETM_CTL_SEL_TOO_BIG)) {
			fmd_hdl_error(hdl, "error: bad ctl sub type %d\n",
					(int)ctl_hdrp->ctl_pp.pp_sub_type);
			errno = EBADMSG;
			etm_stats.etm_subtype_bad.fmds_value.ui64++;
			return (NULL);
		}

		/* get the control length */

		if ((n = etm_io_op(hdl, "bad io read on ctl len",
					conn, &ctl_hdrp->ctl_len,
					sizeof (ctl_hdrp->ctl_len),
					ETM_IO_OP_RD)) < 0) {
			errno = (-n);
			return (NULL);
		}

		ctl_hdrp->ctl_len = ntohl(ctl_hdrp->ctl_len);

		etm_stats.etm_rd_hdr_control.fmds_value.ui64++;

	} else if (pp.pp_msg_type == ETM_MSG_TYPE_RESPONSE) {

		resp_hdrp = (void*)&misc_buf[0];
		hdr_sz = sizeof (*resp_hdrp);
		(void) memcpy(&resp_hdrp->resp_pp, &pp, sizeof (pp));

		/* sanity check the header's timeout */

		if (resp_hdrp->resp_pp.pp_timeout !=
						ETM_PROTO_V1_TIMEOUT_NONE) {
			errno = ETIME;
			etm_stats.etm_timeout_bad.fmds_value.ui64++;
			return (NULL);
		}

		/* get the response code and length */

		if ((n = etm_io_op(hdl, "bad io read on resp code+len",
					conn, &resp_hdrp->resp_code,
					sizeof (resp_hdrp->resp_code) +
					sizeof (resp_hdrp->resp_len),
					ETM_IO_OP_RD)) < 0) {
			errno = (-n);
			return (NULL);
		}

		resp_hdrp->resp_code = ntohl(resp_hdrp->resp_code);
		resp_hdrp->resp_len = ntohl(resp_hdrp->resp_len);

		etm_stats.etm_rd_hdr_response.fmds_value.ui64++;

	} /* whether we have FMA_EVENT, CONTROL, RESPONSE msg */

	/*
	 * choose a header size that allows hdr reuse for RESPONSE msgs,
	 * allocate and populate the message header, and
	 * return alloc size to caller for later free of hdrp
	 */

	hdr_sz = MAX(hdr_sz, sizeof (*resp_hdrp));
	hdrp = fmd_hdl_zalloc(hdl, hdr_sz, FMD_SLEEP);
	(void) memcpy(hdrp, misc_buf, hdr_sz);

	if (etm_debug_lvl >= 3) {
		fmd_hdl_debug(hdl, "info: msg hdr hexdump %d bytes:\n",
								hdr_sz);
		etm_hexdump(hdl, hdrp, hdr_sz);
	}
	*szp = hdr_sz;
	return (hdrp);

} /* etm_hdr_read() */

/*
 * etm_hdr_write - create and write a [variable sized] ETM message header
 *		to the given connection appropriate for the given FMA event
 *		and type of nvlist encoding,
 *		return the allocated ETM message header and its size
 *		or NULL and set errno on failure
 */

static void*
etm_hdr_write(fmd_hdl_t *hdl, etm_xport_conn_t conn, nvlist_t *evp,
						int encoding, size_t *szp)
{
	etm_proto_v1_ev_hdr_t	*hdrp;		/* for FMA_EVENT msg */
	size_t			hdr_sz;		/* sizeof *hdrp */
	uint32_t		*lenp;		/* ptr to FMA event length */
	size_t			evsz;		/* packed FMA event size */
	ssize_t			n;		/* gen use */

	/* allocate and populate the message header for 1 FMA event */

	hdr_sz = sizeof (*hdrp) + (1 * sizeof (hdrp->ev_lens[0]));

	hdrp = fmd_hdl_zalloc(hdl, hdr_sz, FMD_SLEEP);

	/*
	 * Design_Note: Although the ETM protocol supports it, sun4v/Ontario
	 *		does not wait for responses/ACKs on FMA events. All
	 *		such msgs are sent with ETM_PROTO_V1_TIMEOUT_NONE.
	 */

	hdrp->ev_pp.pp_magic_num = ETM_PROTO_MAGIC_NUM;
	hdrp->ev_pp.pp_magic_num = htonl(hdrp->ev_pp.pp_magic_num);
	hdrp->ev_pp.pp_proto_ver = ETM_PROTO_V1;
	hdrp->ev_pp.pp_msg_type = ETM_MSG_TYPE_FMA_EVENT;
	hdrp->ev_pp.pp_sub_type = 0;
	hdrp->ev_pp.pp_rsvd_pad = 0;
	hdrp->ev_pp.pp_xid = etm_xid_cur;
	hdrp->ev_pp.pp_xid = htonl(hdrp->ev_pp.pp_xid);
	etm_xid_cur += ETM_XID_INC;
	hdrp->ev_pp.pp_timeout = ETM_PROTO_V1_TIMEOUT_NONE;
	hdrp->ev_pp.pp_timeout = htonl(hdrp->ev_pp.pp_timeout);

	lenp = &hdrp->ev_lens[0];

	if ((n = nvlist_size(evp, &evsz, encoding)) != 0) {
		errno = n;
		fmd_hdl_free(hdl, hdrp, hdr_sz);
		etm_stats.etm_os_nvlist_size_fail.fmds_value.ui64++;
		return (NULL);
	}

	/* indicate 1 FMA event, network encode its length, and 0-terminate */

	*lenp = evsz; *lenp = htonl(*lenp); lenp++;
	*lenp = 0; *lenp = htonl(*lenp); lenp++;

	/*
	 * write the network encoded header to the transport, and
	 * return alloc size to caller for later free
	 */

	if ((n = etm_io_op(hdl, "bad io write on event hdr",
				conn, hdrp, hdr_sz, ETM_IO_OP_WR)) < 0) {
		errno = (-n);
		fmd_hdl_free(hdl, hdrp, hdr_sz);
		return (NULL);
	}

	*szp = hdr_sz;
	return (hdrp);

} /* etm_hdr_write() */

/*
 * etm_post_to_fmd - post the given FMA event to FMD
 *			[via sysevent or via a FMD transport API call ],
 *			return 0 or -errno value
 *
 * Design_Note:	This routine exists to ease future porting to both
 *		FMA Phase 2 FMD as well as porting to Linux which lacks
 *		a native sysevent.
 */

static int
etm_post_to_fmd(fmd_hdl_t *hdl, nvlist_t *evp)
{
	int			rv;		/* ret val */
	evchan_t		*scp;		/* sysevent channel ptr */
	ssize_t			n;		/* gen use */

	rv = 0; /* default success */

	scp = NULL;

	if ((n = sysevent_evc_bind(FM_ERROR_CHAN, &scp,
				EVCH_CREAT | EVCH_HOLD_PEND)) != 0) {
		rv = (-n);
		fmd_hdl_error(hdl, "error: FMA event dropped: "
				"sysevent bind errno %d\n", n);
		etm_stats.etm_os_sysevent_bind_fail.fmds_value.ui64++;
		etm_stats.etm_rd_drop_fmaevent.fmds_value.ui64++;
		goto func_ret;
	}

	if ((n = sysevent_evc_publish(scp, EC_FM, ESC_FM_ERROR, "com.sun",
				getexecname(), evp, EVCH_SLEEP)) != 0) {
		rv = (-n);
		fmd_hdl_error(hdl, "error: FMA event dropped: "
				"sysevent publish errno %d\n", n);
		etm_stats.etm_os_sysevent_publish_fail.fmds_value.ui64++;
		etm_stats.etm_rd_drop_fmaevent.fmds_value.ui64++;
		goto func_ret;
	}

func_ret:

	if (scp != NULL) {
		sysevent_evc_unbind(scp);
	}
	if (rv == 0) {
		etm_stats.etm_wr_fmd_fmaevent.fmds_value.ui64++;
		(void) nvlist_size(evp, (size_t *)&n, NV_ENCODE_XDR);
		etm_stats.etm_wr_fmd_bytes.fmds_value.ui64 += n;
		if (etm_debug_lvl >= 1) {
			fmd_hdl_debug(hdl, "info: event %p post ok to FMD\n",
								evp);
		}
	}
	return (rv);

} /* etm_post_to_fmd() */

/*
 * etm_req_ver_negot - send an ETM control message to the other end requesting
 *			that the ETM protocol version be negotiated/set
 */

static void
etm_req_ver_negot(fmd_hdl_t *hdl)
{
	etm_xport_addr_t	*addrv;		/* default dst addr(s) */
	etm_xport_conn_t	conn;		/* connection to other end */
	etm_proto_v1_ctl_hdr_t	*ctl_hdrp;	/* for CONTROL msg */
	size_t			hdr_sz;		/* sizeof header */
	uint8_t			*body_buf;	/* msg body buffer */
	uint32_t		body_sz;	/* sizeof *body_buf */
	ssize_t			i;		/* gen use */

	/* populate an ETM control msg to send */

	hdr_sz = sizeof (*ctl_hdrp);
	body_sz = (1 + 1);		/* V1 byte plus null byte */

	ctl_hdrp = fmd_hdl_zalloc(hdl, hdr_sz + body_sz, FMD_SLEEP);

	ctl_hdrp->ctl_pp.pp_magic_num = htonl(ETM_PROTO_MAGIC_NUM);
	ctl_hdrp->ctl_pp.pp_proto_ver = ETM_PROTO_V1;
	ctl_hdrp->ctl_pp.pp_msg_type = ETM_MSG_TYPE_CONTROL;
	ctl_hdrp->ctl_pp.pp_sub_type = ETM_CTL_SEL_VER_SET_REQ;
	ctl_hdrp->ctl_pp.pp_rsvd_pad = 0;
	etm_xid_ver_set = etm_xid_cur;
	etm_xid_cur += ETM_XID_INC;
	ctl_hdrp->ctl_pp.pp_xid = htonl(etm_xid_ver_set);
	ctl_hdrp->ctl_pp.pp_timeout = htonl(ETM_PROTO_V1_TIMEOUT_FOREVER);
	ctl_hdrp->ctl_len = htonl(body_sz);

	body_buf = (void*)&ctl_hdrp->ctl_len;
	body_buf += sizeof (ctl_hdrp->ctl_len);
	*body_buf++ = ETM_PROTO_V1;
	*body_buf++ = '\0';

	/*
	 * open and close a connection to send the ETM control msg
	 * to any/all of the default dst addrs
	 */

	if ((addrv = etm_xport_get_ev_addrv(hdl, NULL)) == NULL) {
		fmd_hdl_error(hdl,
			"error: bad ctl dst addrs errno %d\n", errno);
		etm_stats.etm_xport_get_ev_addrv_fail.fmds_value.ui64++;
		goto func_ret;
	}

	for (i = 0; addrv[i] != NULL; i++) {

		if (etm_conn_open(hdl, "bad conn open during ver negot",
					addrv[i], &conn) < 0) {
			continue;
		}
		if (etm_io_op(hdl, "bad io write on ctl hdr+body",
					conn, ctl_hdrp, hdr_sz + body_sz,
					ETM_IO_OP_WR) >= 0) {
			etm_stats.etm_wr_hdr_control.fmds_value.ui64++;
			etm_stats.etm_wr_body_control.fmds_value.ui64++;
		}
		(void) etm_conn_close(hdl, "bad conn close during ver negot",
									conn);

	} /* foreach dst addr */

func_ret:

	if (addrv != NULL) {
		etm_xport_free_addrv(hdl, addrv);
	}
	fmd_hdl_free(hdl, ctl_hdrp, hdr_sz + body_sz);

} /* etm_req_ver_negot() */

/*
 * etm_handle_new_conn - receive an ETM message sent from the other end via
 *			the given open connection, pull out any FMA events
 *			and post them to the local FMD (or handle any ETM
 *			control or response msg); when done, close the
 *			connection
 */

static void
etm_handle_new_conn(fmd_hdl_t *hdl, etm_xport_conn_t conn)
{
	etm_proto_v1_ev_hdr_t	*ev_hdrp;	/* for FMA_EVENT msg */
	etm_proto_v1_ctl_hdr_t	*ctl_hdrp;	/* for CONTROL msg */
	etm_proto_v1_resp_hdr_t *resp_hdrp;	/* for RESPONSE msg */
	size_t			hdr_sz;		/* sizeof header */
	uint8_t			*body_buf;	/* msg body buffer */
	uint32_t		body_sz;	/* sizeof body_buf */
	uint8_t			*bp;		/* byte ptr within body_buf */
	nvlist_t		*evp;		/* ptr to unpacked FMA event */
	char			*class;		/* FMA event class */
	ssize_t			i, n;		/* gen use */

	fmd_hdl_debug(hdl, "info: handling new conn %p\n", conn);

	ev_hdrp = NULL;
	ctl_hdrp = NULL;
	resp_hdrp = NULL;
	body_buf = NULL;
	class = NULL;
	evp = NULL;

	/* read a network decoded message header from the connection */

	/*
	 * Design_Note:	We rely on the fact that all message types have
	 *		a common protocol preamble; if this fact should
	 *		ever change it may break the code below. We also
	 *		rely on the fact that FMA_EVENT and CONTROL headers
	 *		returned will be sized large enough to reuse them
	 *		as RESPONSE headers if the remote endpt asked
	 *		for a response via the pp_timeout field.
	 */

	if ((ev_hdrp = etm_hdr_read(hdl, conn, &hdr_sz)) == NULL) {
		/* errno assumed set by above call */
		fmd_hdl_error(hdl, "error: FMA event dropped: "
					"bad hdr read errno %d\n", errno);
		etm_stats.etm_rd_drop_fmaevent.fmds_value.ui64++;
		goto func_ret;
	}

	/*
	 * handle the message based on its preamble pp_msg_type
	 * which is known to be valid from etm_hdr_read() checks
	 */

	if (ev_hdrp->ev_pp.pp_msg_type == ETM_MSG_TYPE_FMA_EVENT) {

		fmd_hdl_debug(hdl, "info: rcvd FMA_EVENT msg from xport\n");

		/* allocate buf large enough for whole body / all FMA events */

		body_sz = 0;
		for (i = 0; ev_hdrp->ev_lens[i] != 0; i++) {
			body_sz += ev_hdrp->ev_lens[i];
		} /* for summing sizes of all FMA events */

		if (etm_debug_lvl >= 1) {
			fmd_hdl_debug(hdl, "info: event lengths %d sum %d\n",
								i, body_sz);
		}

		body_buf = fmd_hdl_zalloc(hdl, body_sz, FMD_SLEEP);

		/* read all the FMA events at once */

		if ((n = etm_io_op(hdl, "FMA event dropped: "
					"bad io read on event bodies",
					conn, body_buf, body_sz,
					ETM_IO_OP_RD)) < 0) {
			etm_stats.etm_rd_drop_fmaevent.fmds_value.ui64++;
			goto func_ret;
		}

		etm_stats.etm_rd_xport_bytes.fmds_value.ui64 += body_sz;

		/* immediately close the connection to improve xport thruput */

		(void) etm_conn_close(hdl, "bad conn close "
					"after event body read", conn);
		conn = NULL;

		/* unpack each FMA event and post it to FMD */

		bp = body_buf;
		for (i = 0; ev_hdrp->ev_lens[i] != 0; i++) {
			if ((n = nvlist_unpack((char *)bp,
					ev_hdrp->ev_lens[i], &evp, 0)) != 0) {
				fmd_hdl_error(hdl, "error: FMA event dropped: "
						"bad event body unpack "
						"errno %d\n", n);
				if (etm_debug_lvl >= 2) {
					fmd_hdl_debug(hdl, "info: FMA event "
						"hexdump %d bytes:\n",
						ev_hdrp->ev_lens[i]);
					etm_hexdump(hdl, bp,
						ev_hdrp->ev_lens[i]);
				}
				etm_stats.etm_os_nvlist_unpack_fail.fmds_value.
					ui64++;
				etm_stats.etm_rd_drop_fmaevent.fmds_value.
					ui64++;
				bp += ev_hdrp->ev_lens[i];
				continue;
			}
			etm_stats.etm_rd_body_fmaevent.fmds_value.ui64++;
			if (etm_debug_lvl >= 1) {
				(void) nvlist_lookup_string(evp, FM_CLASS,
								&class);
				if (class == NULL) {
					class = "NULL";
				}
				fmd_hdl_debug(hdl, "info: FMA event %p "
						"class %s\n", evp, class);
			}
			(void) etm_post_to_fmd(hdl, evp);
			nvlist_free(evp);
			bp += ev_hdrp->ev_lens[i];
		} /* foreach FMA event in the body buffer */

	} else if (ev_hdrp->ev_pp.pp_msg_type == ETM_MSG_TYPE_CONTROL) {

		ctl_hdrp = (void*)ev_hdrp;

		fmd_hdl_debug(hdl, "info: rcvd CONTROL msg from xport\n");
		if (etm_debug_lvl >= 1) {
			fmd_hdl_debug(hdl, "info: ctl sel %d xid 0x%x\n",
					(int)ctl_hdrp->ctl_pp.pp_sub_type,
					ctl_hdrp->ctl_pp.pp_xid);
		}

		/*
		 * if we have a VER_SET_REQ read the body and validate
		 * the protocol version set contained therein,
		 * otherwise we have a PING_REQ (which has no body)
		 * and we [also] fall thru to the code which sends a
		 * response msg if the pp_timeout field requested one
		 */

		if (ctl_hdrp->ctl_pp.pp_sub_type == ETM_CTL_SEL_VER_SET_REQ) {

			body_sz = ctl_hdrp->ctl_len;
			body_buf = fmd_hdl_zalloc(hdl, body_sz, FMD_SLEEP);

			if ((n = etm_io_op(hdl, "bad io read on ctl body",
						conn, body_buf, body_sz,
						ETM_IO_OP_RD)) < 0) {
				goto func_ret;
			}

			/* complain if version set lacks our version */

			n = 0;
			for (i = 0; i < body_sz; i++) {
				if (body_buf[i] == ETM_PROTO_V1) {
					n = 1;
					break;
				}
			}
			if (n == 0) {
				etm_stats.etm_ver_bad.fmds_value.ui64++;
			}

		} /* if got version set request */

		etm_stats.etm_rd_body_control.fmds_value.ui64++;

		/* if a response is requested send one (reuse received hdr) */

		if (ctl_hdrp->ctl_pp.pp_timeout != ETM_PROTO_V1_TIMEOUT_NONE) {
			resp_hdrp = (void*)ctl_hdrp;
			resp_hdrp->resp_len = 0;
			if (ctl_hdrp->ctl_pp.pp_sub_type ==
						ETM_CTL_SEL_VER_SET_REQ) {
				resp_hdrp->resp_len = 1;
			}
			resp_hdrp->resp_code = 0;
			resp_hdrp->resp_pp.pp_timeout =
						ETM_PROTO_V1_TIMEOUT_NONE;
			resp_hdrp->resp_pp.pp_msg_type = ETM_MSG_TYPE_RESPONSE;
			if ((n = etm_io_op(hdl, "bad io write on resp hdr",
						conn, resp_hdrp,
						sizeof (*resp_hdrp),
						ETM_IO_OP_WR)) < 0) {
				goto func_ret;
			}
			etm_stats.etm_wr_hdr_response.fmds_value.ui64++;
			if (resp_hdrp->resp_pp.pp_sub_type ==
						ETM_CTL_SEL_VER_SET_REQ) {
				/* send our default proto ver in resp body */
				bp = (void*)&i;
				*bp = ETM_PROTO_V1;
				if ((n = etm_io_op(hdl,
						"bad io write on resp body",
						conn, bp, 1,
						ETM_IO_OP_WR)) < 0) {
					goto func_ret;
				}
			} /* if need to send proto ver in [tmp] msg body */
			etm_stats.etm_wr_body_response.fmds_value.ui64++;
			fmd_hdl_debug(hdl, "info: response sent "
					"xid 0x%x code %d body len %d\n",
					resp_hdrp->resp_pp.pp_xid,
					resp_hdrp->resp_code,
					resp_hdrp->resp_len);
		} /* if a response was requested */

	} else if (ev_hdrp->ev_pp.pp_msg_type == ETM_MSG_TYPE_RESPONSE) {

		resp_hdrp = (void*)ev_hdrp;

		fmd_hdl_debug(hdl, "info: rcvd RESPONSE msg from xport\n");
		if (etm_debug_lvl >= 1) {
			fmd_hdl_debug(hdl, "info: resp xid 0x%x\n",
					(int)resp_hdrp->resp_pp.pp_xid);
		}

		body_sz = resp_hdrp->resp_len;
		body_buf = fmd_hdl_zalloc(hdl, body_sz, FMD_SLEEP);

		if ((n = etm_io_op(hdl, "bad io read on resp len",
				conn, body_buf, body_sz, ETM_IO_OP_RD)) < 0) {
			goto func_ret;
		}

		etm_stats.etm_rd_body_response.fmds_value.ui64++;

		/*
		 * look up the xid to interpret response body
		 *
		 * ping is a nop; for ver set just confirm ETM_PROTO_V1
		 * was negotiated
		 */

		if ((resp_hdrp->resp_pp.pp_xid != etm_xid_ping) &&
			(resp_hdrp->resp_pp.pp_xid != etm_xid_ver_set)) {
			etm_stats.etm_xid_bad.fmds_value.ui64++;
			goto func_ret;
		}

		if (resp_hdrp->resp_pp.pp_xid == etm_xid_ver_set) {
			if (body_buf[0] != ETM_PROTO_V1) {
				etm_stats.etm_ver_bad.fmds_value.ui64++;
				goto func_ret;
			}
		} /* if have resp to last req to set proto ver */

	} /* whether we have a FMA_EVENT, CONTROL, or RESPONSE msg */

func_ret:

	if (conn != NULL) {
		(void) etm_conn_close(hdl, "bad conn close after event recv",
									conn);
	}
	if (ev_hdrp != NULL) {
		fmd_hdl_free(hdl, ev_hdrp, hdr_sz);
	}
	if (body_buf != NULL) {
		fmd_hdl_free(hdl, body_buf, body_sz);
	}
} /* etm_handle_new_conn() */

/*
 * etm_server - loop forever accepting new connections
 *		using the given FMD handle,
 *		handling any ETM msgs sent from the other side
 *		via each such connection
 */

static void
etm_server(void *arg)
{
	etm_xport_conn_t	conn;		/* connection handle */
	ssize_t			n;		/* gen use */
	fmd_hdl_t		*hdl;		/* FMD handle */

	hdl = arg;

	fmd_hdl_debug(hdl, "info: connection server starting\n");

	while (!etm_is_dying) {
		if ((conn = etm_xport_accept(hdl, NULL)) == NULL) {
			/* errno assumed set by above call */
			n = errno;
			if (etm_is_dying) {
				break;
			}
			fmd_hdl_debug(hdl,
				"error: bad conn accept errno %d\n", n);
			etm_stats.etm_xport_accept_fail.fmds_value.ui64++;
			/* avoid spinning CPU */
			(void) etm_sleep(ETM_SLEEP_SLOW);
			continue;
		}

		/*
		 * Design_Note: etm_handle_new_conn() will close the
		 *		accepted connection when done. In early designs
		 *		etm_handle_new_conn() was spawned as a
		 *		separate thread via pthread_create();
		 *		however fmd_thr_create() constrains thread
		 *		creation to prevent spawned threads from
		 *		spawning others (ie, no grandchildren).
		 *		Hence etm_handle_new_conn() is now called
		 *		as a simple function [w/ multiple args].
		 */

		etm_handle_new_conn(hdl, conn);

	} /* while accepting new connections until ETM dies */

	/* ETM is dying (probably due to "fmadm unload etm") */

	if (etm_debug_lvl >= 1) {
		fmd_hdl_debug(hdl, "info: connection server is dying\n");
	}
} /* etm_server() */

/*
 * -------------------------- FMD entry points -------------------------------
 */

/*
 * _fmd_init - initialize the transport for use by ETM and start the
 *		server daemon to accept new connections to us
 *
 *		FMD will read our *.conf and subscribe us to FMA events
 */

void
_fmd_init(fmd_hdl_t *hdl)
{
	ssize_t			n;		/* gen use */

	if (fmd_hdl_register(hdl, FMD_API_VERSION, &fmd_info) != 0) {
		return; /* invalid data in configuration file */
	}

	fmd_hdl_debug(hdl, "info: module initializing\n");

	/* setup statistics and properties from FMD */

	(void) fmd_stat_create(hdl, FMD_STAT_NOALLOC,
				sizeof (etm_stats) / sizeof (fmd_stat_t),
				(fmd_stat_t *)&etm_stats);

	etm_debug_lvl = fmd_prop_get_int32(hdl, ETM_PROP_NM_DEBUG_LVL);
	etm_debug_max_ev_cnt = fmd_prop_get_int32(hdl,
						ETM_PROP_NM_DEBUG_MAX_EV_CNT);
	fmd_hdl_debug(hdl, "info: etm_debug_lvl %d "
			"etm_debug_max_ev_cnt %d\n",
			etm_debug_lvl, etm_debug_max_ev_cnt);

	/*
	 * init the transport,
	 * start the connection acceptance server, and
	 * request protocol version be negotiated
	 */

	if ((n = etm_xport_init(hdl)) != 0) {
		fmd_hdl_error(hdl, "error: bad xport init errno %d\n", (-n));
		fmd_hdl_unregister(hdl);
		return;
	}

	etm_svr_tid = fmd_thr_create(hdl, etm_server, hdl);
	etm_req_ver_negot(hdl);

	fmd_hdl_debug(hdl, "info: module initialized ok\n");

} /* _fmd_init() */

/*
 * etm_recv - receive an FMA event from FMD and transport it
 *		to the remote endpoint
 */

/*ARGSUSED*/
void
etm_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *evp, const char *class)
{
	etm_xport_addr_t	*addrv;	/* vector of transport addresses */
	etm_xport_conn_t	conn;	/* connection handle */
	etm_proto_v1_ev_hdr_t	*hdrp;	/* for FMA_EVENT msg */
	ssize_t			i, n;	/* gen use */
	size_t			sz;	/* header size */
	size_t			buflen;	/* size of packed FMA event */
	uint8_t			*buf;	/* tmp buffer for packed FMA event */

	buflen = 0;
	(void) nvlist_size(evp, &buflen, NV_ENCODE_XDR);
	etm_stats.etm_rd_fmd_bytes.fmds_value.ui64 += buflen;
	etm_stats.etm_rd_fmd_fmaevent.fmds_value.ui64++;

	fmd_hdl_debug(hdl, "info: rcvd event %p from FMD\n", evp);
	fmd_hdl_debug(hdl, "info: cnt %llu class %s\n",
		etm_stats.etm_rd_fmd_fmaevent.fmds_value.ui64, class);

	/*
	 * if the debug limit has been set, avoid excessive traffic,
	 * for example, an infinite cycle using loopback nodes
	 */

	if ((etm_debug_max_ev_cnt >= 0) &&
		(etm_stats.etm_rd_fmd_fmaevent.fmds_value.ui64 >
						etm_debug_max_ev_cnt)) {
		fmd_hdl_debug(hdl, "warning: FMA event dropped: "
			"event %p cnt %llu > debug max %d\n", evp,
			etm_stats.etm_rd_fmd_fmaevent.fmds_value.ui64,
			etm_debug_max_ev_cnt);
		etm_stats.etm_wr_drop_fmaevent.fmds_value.ui64++;
		return;
	}

	/* allocate a buffer for the FMA event and nvlist pack it */

	buf = fmd_hdl_zalloc(hdl, buflen, FMD_SLEEP);

	if ((n = nvlist_pack(evp, (char **)&buf, &buflen,
					NV_ENCODE_XDR, 0)) != 0) {
		fmd_hdl_error(hdl, "error: FMA event dropped: "
				"event pack errno %d\n", n);
		etm_stats.etm_os_nvlist_pack_fail.fmds_value.ui64++;
		etm_stats.etm_wr_drop_fmaevent.fmds_value.ui64++;
		fmd_hdl_free(hdl, buf, buflen);
		return;
	}

	/* get vector of dst addrs and send the FMA event to each one */

	if ((addrv = etm_xport_get_ev_addrv(hdl, evp)) == NULL) {
		fmd_hdl_error(hdl, "error: FMA event dropped: "
				"bad event dst addrs errno %d\n", errno);
		etm_stats.etm_xport_get_ev_addrv_fail.fmds_value.ui64++;
		etm_stats.etm_wr_drop_fmaevent.fmds_value.ui64++;
		fmd_hdl_free(hdl, buf, buflen);
		return;
	}

	for (i = 0; addrv[i] != NULL; i++) {

		/* open a new connection to this dst addr */

		if ((n = etm_conn_open(hdl, "FMA event dropped: "
				"bad conn open on new ev",
				addrv[i], &conn)) < 0) {
			etm_stats.etm_wr_drop_fmaevent.fmds_value.ui64++;
			continue;
		}

		/* write the ETM message header */

		if ((hdrp = etm_hdr_write(hdl, conn, evp, NV_ENCODE_XDR,
							&sz)) == NULL) {
			fmd_hdl_error(hdl, "error: FMA event dropped: "
					"bad hdr write errno %d\n", errno);
			(void) etm_conn_close(hdl,
				"bad conn close per bad hdr wr", conn);
			etm_stats.etm_wr_drop_fmaevent.fmds_value.ui64++;
			continue;
		}

		fmd_hdl_free(hdl, hdrp, sz);	/* header not needed */
		etm_stats.etm_wr_hdr_fmaevent.fmds_value.ui64++;
		fmd_hdl_debug(hdl, "info: hdr xport write ok for event %p\n",
								evp);

		/* write the ETM message body, ie, the packed nvlist */

		if ((n = etm_io_op(hdl, "FMA event dropped: "
					"bad io write on event", conn,
					buf, buflen, ETM_IO_OP_WR)) < 0) {
			(void) etm_conn_close(hdl,
				"bad conn close per bad body wr", conn);
			etm_stats.etm_wr_drop_fmaevent.fmds_value.ui64++;
			continue;
		}

		etm_stats.etm_wr_body_fmaevent.fmds_value.ui64++;
		etm_stats.etm_wr_xport_bytes.fmds_value.ui64 += buflen;
		fmd_hdl_debug(hdl, "info: body xport write ok for event %p\n",
								evp);

		/* close the connection */

		(void) etm_conn_close(hdl, "bad conn close after event send",
									conn);
	} /* foreach dst addr in the vector */

	etm_xport_free_addrv(hdl, addrv);
	fmd_hdl_free(hdl, buf, buflen);

} /* etm_recv() */

/*
 * _fmd_fini - stop the server daemon and teardown the transport
 */

void
_fmd_fini(fmd_hdl_t *hdl)
{
	ssize_t	n;	/* gen use */

	fmd_hdl_debug(hdl, "info: module finializing\n");

	/* kill the connection server ; wait for it to die */

	etm_is_dying = 1;

	if (etm_svr_tid != NULL) {
		fmd_thr_signal(hdl, etm_svr_tid);
		fmd_thr_destroy(hdl, etm_svr_tid);
		etm_svr_tid = NULL;
	} /* if server thread was successfully created */

	/* teardown the transport */

	if ((n = etm_xport_fini(hdl)) != 0) {
		fmd_hdl_error(hdl, "warning: xport fini errno %d\n", (-n));
	}

	fmd_hdl_debug(hdl, "info: module finalized ok\n");

} /* _fmd_fini() */