[lttng-ust.git] / liblttng-ust / ltt-events.c

/*
 * ltt-events.c
 *
 * Copyright 2010 (c) - Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
 *
 * Holds LTTng per-session event registry.
 *
 * Dual LGPL v2.1/GPL v2 license.
 */

#define _GNU_SOURCE
#include <stdio.h>
#include <endian.h>
#include <urcu/list.h>
#include <urcu/hlist.h>
#include <pthread.h>
#include <uuid/uuid.h>
#include <errno.h>
#include <sys/shm.h>
#include <sys/ipc.h>
#include <stdint.h>
#include <stddef.h>
#include <inttypes.h>
#include <time.h>
#include "clock.h"

#include <urcu-bp.h>
#include <urcu/compiler.h>
#include <urcu/uatomic.h>
#include <urcu/arch.h>

#include <lttng/tracepoint.h>
#include <lttng/ust-events.h>

#include <usterr-signal-safe.h>
#include <helper.h>
#include "error.h"

#include "ltt-tracer.h"
#include "ltt-tracer-core.h"
#include "wait.h"
#include "../libringbuffer/shm.h"
#include "jhash.h"

/*
 * The sessions mutex is the centralized mutex across UST tracing
 * control and probe registration. All operations within this file are
 * called by the communication thread, under ust_lock protection.
 */
static pthread_mutex_t sessions_mutex = PTHREAD_MUTEX_INITIALIZER;

void ust_lock(void)
{
	pthread_mutex_lock(&sessions_mutex);
}

void ust_unlock(void)
{
	pthread_mutex_unlock(&sessions_mutex);
}

static CDS_LIST_HEAD(sessions);

/*
 * Pending probes hash table, containing the registered ltt events for
 * which tracepoint probes are still missing. Protected by the sessions
 * mutex.
 */
#define PENDING_PROBE_HASH_BITS		6
#define PENDING_PROBE_HASH_SIZE		(1 << PENDING_PROBE_HASH_BITS)
static struct cds_hlist_head pending_probe_table[PENDING_PROBE_HASH_SIZE];

struct ust_pending_probe {
	struct ltt_event *event;
	struct cds_hlist_node node;
	char name[];
};

static void _ltt_event_destroy(struct ltt_event *event);
static void _ltt_wildcard_destroy(struct session_wildcard *sw);
static void _ltt_channel_destroy(struct ltt_channel *chan);
static int _ltt_event_unregister(struct ltt_event *event);
static
int _ltt_event_metadata_statedump(struct ltt_session *session,
				  struct ltt_channel *chan,
				  struct ltt_event *event);
static
int _ltt_session_metadata_statedump(struct ltt_session *session);

/*
 * called at event creation if probe is missing.
 * called with session mutex held.
 */
static
int add_pending_probe(struct ltt_event *event, const char *name)
{
	struct cds_hlist_head *head;
	struct ust_pending_probe *e;
	size_t name_len = strlen(name);
	uint32_t hash;

	if (name_len > LTTNG_UST_SYM_NAME_LEN - 1) {
		WARN("Truncating tracepoint name %s which exceeds size limits of %u chars", name, LTTNG_UST_SYM_NAME_LEN - 1);
		name_len = LTTNG_UST_SYM_NAME_LEN - 1;
	}
	hash = jhash(name, name_len, 0);
	head = &pending_probe_table[hash & (PENDING_PROBE_HASH_SIZE - 1)];
	e = zmalloc(sizeof(struct ust_pending_probe) + name_len);
	if (!e)
		return -ENOMEM;
	memcpy(&e->name[0], name, name_len + 1);
	e->name[name_len] = '\0';
	cds_hlist_add_head(&e->node, head);
	e->event = event;
	event->pending_probe = e;
	return 0;
}

/*
 * remove a pending probe. called when at event teardown and when an
 * event is fixed (probe is loaded).
 * called with session mutex held.
 */
static
void remove_pending_probe(struct ust_pending_probe *e)
{
	if (!e)
		return;
	cds_hlist_del(&e->node);
	free(e);
}

/*
 * Called at library load: connect the probe on the events pending on
 * probe load.
 * called with session mutex held.
 */
int pending_probe_fix_events(const struct lttng_event_desc *desc)
{
	struct cds_hlist_head *head;
	struct cds_hlist_node *node, *p;
	struct ust_pending_probe *e;
	const char *name = desc->name;
	int ret = 0;
	struct lttng_ust_event event_param;
	size_t name_len = strlen(name);
	uint32_t hash;

	/* Wildcard */
	{
		struct wildcard_entry *wildcard;

				/* TODO: get value from loglevel. */

				/* TODO: check if loglevel match */
		wildcard = match_wildcard(desc->name);
		if (strcmp(desc->name, "lttng_ust:metadata") && wildcard) {
			struct session_wildcard *sw;

			cds_list_for_each_entry(sw, &wildcard->session_list,
					session_list) {
				struct ltt_event *ev;
				int ret;

				memcpy(&event_param, &sw->event_param,
						sizeof(event_param));
				memcpy(event_param.name,
					desc->name,
					sizeof(event_param.name));
				/* create event */
				ret = ltt_event_create(sw->chan,
					&event_param, NULL,
					&ev);
				if (ret) {
					DBG("Error creating event");
					continue;
				}
				cds_list_add(&ev->wildcard_list,
					&sw->events);
			}
		}
	}

	if (name_len > LTTNG_UST_SYM_NAME_LEN - 1) {
		WARN("Truncating tracepoint name %s which exceeds size limits of %u chars", name, LTTNG_UST_SYM_NAME_LEN - 1);
		name_len = LTTNG_UST_SYM_NAME_LEN - 1;
	}
	hash = jhash(name, name_len, 0);
	head = &pending_probe_table[hash & (PENDING_PROBE_HASH_SIZE - 1)];
	cds_hlist_for_each_entry_safe(e, node, p, head, node) {
		struct ltt_event *event;
		struct ltt_channel *chan;

				/* TODO: get value from loglevel. */

				/* TODO: check if loglevel match */
		if (strncmp(name, e->name, LTTNG_UST_SYM_NAME_LEN - 1))
			continue;
		event = e->event;
		chan = event->chan;
		assert(!event->desc);
		event->desc = desc;
		event->pending_probe = NULL;
		remove_pending_probe(e);
		ret |= __tracepoint_probe_register(name,
				event->desc->probe_callback,
				event);
		if (ret)
			continue;
		event->id = chan->free_event_id++;
		ret |= _ltt_event_metadata_statedump(chan->session, chan,
				event);
	}
	return ret;
}

void synchronize_trace(void)
{
	synchronize_rcu();
}

struct ltt_session *ltt_session_create(void)
{
	struct ltt_session *session;

	session = zmalloc(sizeof(struct ltt_session));
	if (!session)
		return NULL;
	CDS_INIT_LIST_HEAD(&session->chan);
	CDS_INIT_LIST_HEAD(&session->events);
	CDS_INIT_LIST_HEAD(&session->wildcards);
	uuid_generate(session->uuid);
	cds_list_add(&session->list, &sessions);
	return session;
}

void ltt_session_destroy(struct ltt_session *session)
{
	struct ltt_channel *chan, *tmpchan;
	struct ltt_event *event, *tmpevent;
	struct session_wildcard *wildcard, *tmpwildcard;
	int ret;

	CMM_ACCESS_ONCE(session->active) = 0;
	cds_list_for_each_entry(event, &session->events, list) {
		ret = _ltt_event_unregister(event);
		WARN_ON(ret);
	}
	synchronize_trace();	/* Wait for in-flight events to complete */
	cds_list_for_each_entry_safe(wildcard, tmpwildcard, &session->wildcards, list)
		_ltt_wildcard_destroy(wildcard);
	cds_list_for_each_entry_safe(event, tmpevent, &session->events, list)
		_ltt_event_destroy(event);
	cds_list_for_each_entry_safe(chan, tmpchan, &session->chan, list)
		_ltt_channel_destroy(chan);
	cds_list_del(&session->list);
	free(session);
}

int ltt_session_enable(struct ltt_session *session)
{
	int ret = 0;
	struct ltt_channel *chan;

	if (session->active) {
		ret = -EBUSY;
		goto end;
	}

	/*
	 * Snapshot the number of events per channel to know the type of header
	 * we need to use.
	 */
	cds_list_for_each_entry(chan, &session->chan, list) {
		if (chan->header_type)
			continue;		/* don't change it if session stop/restart */
		if (chan->free_event_id < 31)
			chan->header_type = 1;	/* compact */
		else
			chan->header_type = 2;	/* large */
	}

	CMM_ACCESS_ONCE(session->active) = 1;
	CMM_ACCESS_ONCE(session->been_active) = 1;
	ret = _ltt_session_metadata_statedump(session);
	if (ret)
		CMM_ACCESS_ONCE(session->active) = 0;
end:
	return ret;
}

int ltt_session_disable(struct ltt_session *session)
{
	int ret = 0;

	if (!session->active) {
		ret = -EBUSY;
		goto end;
	}
	CMM_ACCESS_ONCE(session->active) = 0;
end:
	return ret;
}

int ltt_channel_enable(struct ltt_channel *channel)
{
	int old;

	if (channel == channel->session->metadata)
		return -EPERM;
	old = uatomic_xchg(&channel->enabled, 1);
	if (old)
		return -EEXIST;
	return 0;
}

int ltt_channel_disable(struct ltt_channel *channel)
{
	int old;

	if (channel == channel->session->metadata)
		return -EPERM;
	old = uatomic_xchg(&channel->enabled, 0);
	if (!old)
		return -EEXIST;
	return 0;
}

int ltt_event_enable(struct ltt_event *event)
{
	int old;

	if (event->chan == event->chan->session->metadata)
		return -EPERM;
	old = uatomic_xchg(&event->enabled, 1);
	if (old)
		return -EEXIST;
	return 0;
}

int ltt_event_disable(struct ltt_event *event)
{
	int old;

	if (event->chan == event->chan->session->metadata)
		return -EPERM;
	old = uatomic_xchg(&event->enabled, 0);
	if (!old)
		return -EEXIST;
	return 0;
}

struct ltt_channel *ltt_channel_create(struct ltt_session *session,
				       const char *transport_name,
				       void *buf_addr,
				       size_t subbuf_size, size_t num_subbuf,
				       unsigned int switch_timer_interval,
				       unsigned int read_timer_interval,
				       int **shm_fd, int **wait_fd,
				       uint64_t **memory_map_size,
				       struct ltt_channel *chan_priv_init)
{
	struct ltt_channel *chan = NULL;
	struct ltt_transport *transport;

	if (session->been_active)
		goto active;	/* Refuse to add channel to active session */
	transport = ltt_transport_find(transport_name);
	if (!transport) {
		DBG("LTTng transport %s not found\n",
		       transport_name);
		goto notransport;
	}
	chan_priv_init->id = session->free_chan_id++;
	chan_priv_init->session = session;
	/*
	 * Note: the channel creation op already writes into the packet
	 * headers. Therefore the "chan" information used as input
	 * should be already accessible.
	 */
	chan = transport->ops.channel_create("[lttng]", buf_addr,
			subbuf_size, num_subbuf, switch_timer_interval,
			read_timer_interval, shm_fd, wait_fd,
			memory_map_size, chan_priv_init);
	if (!chan)
		goto create_error;
	chan->enabled = 1;
	chan->ops = &transport->ops;
	cds_list_add(&chan->list, &session->chan);
	return chan;

create_error:
notransport:
active:
	return NULL;
}

/*
 * Only used internally at session destruction.
 */
static
void _ltt_channel_destroy(struct ltt_channel *chan)
{
	cds_list_del(&chan->list);
	lttng_destroy_context(chan->ctx);
	chan->ops->channel_destroy(chan);
}

int ltt_wildcard_create(struct ltt_channel *chan,
	struct lttng_ust_event *event_param,
	struct session_wildcard **_sw)
{
	struct session_wildcard *sw;

	sw = add_wildcard(event_param->name, chan, event_param);
	if (!sw || IS_ERR(sw)) {
		return PTR_ERR(sw);
	}
	*_sw = sw;
	return 0;
}

static
void _ltt_wildcard_destroy(struct session_wildcard *sw)
{
	_remove_wildcard(sw);
}

/*
 * Supports event creation while tracing session is active.
 */
int ltt_event_create(struct ltt_channel *chan,
		struct lttng_ust_event *event_param,
		void *filter,
		struct ltt_event **_event)
{
	struct ltt_event *event;
	int ret = 0;

	if (chan->used_event_id == -1UL) {
		ret = -ENOMEM;
		goto full;
	}
	/*
	 * This is O(n^2) (for each event, the loop is called at event
	 * creation). Might require a hash if we have lots of events.
	 */
	cds_list_for_each_entry(event, &chan->session->events, list) {
		if (event->desc && !strncmp(event->desc->name,
				event_param->name,
				LTTNG_UST_SYM_NAME_LEN - 1)) {
			ret = -EEXIST;
			goto exist;
		}
	}
	event = zmalloc(sizeof(struct ltt_event));
	if (!event) {
		ret = -ENOMEM;
		goto cache_error;
	}
	event->chan = chan;
	event->filter = filter;
	/*
	 * used_event_id counts the maximum number of event IDs that can
	 * register if all probes register.
	 */
	chan->used_event_id++;
	event->enabled = 1;
	event->instrumentation = event_param->instrumentation;
	/* Populate ltt_event structure before tracepoint registration. */
	cmm_smp_wmb();
	switch (event_param->instrumentation) {
	case LTTNG_UST_TRACEPOINT:
		event->desc = ltt_event_get(event_param->name);
				/* TODO: get value from loglevel. */

				/* TODO: check if loglevel match */
		if (event->desc) {
			ret = __tracepoint_probe_register(event_param->name,
					event->desc->probe_callback,
					event);
			if (ret)
				goto register_error;
			event->id = chan->free_event_id++;
		} else {
			/*
			 * If the probe is not present, event->desc stays NULL,
			 * waiting for the probe to register, and the event->id
			 * stays unallocated.
			 */
			ret = add_pending_probe(event, event_param->name);
			if (ret)
				goto add_pending_error;
		}
		break;
	default:
		WARN_ON_ONCE(1);
	}
	if (event->desc) {
		ret = _ltt_event_metadata_statedump(chan->session, chan, event);
		if (ret)
			goto statedump_error;
	}
	cds_list_add(&event->list, &chan->session->events);
	*_event = event;
	return 0;

statedump_error:
	if (event->desc) {
		WARN_ON_ONCE(__tracepoint_probe_unregister(event_param->name,
					event->desc->probe_callback,
					event));
		ltt_event_put(event->desc);
	}
add_pending_error:
register_error:
	free(event);
cache_error:
exist:
full:
	return ret;
}

/*
 * Only used internally at session destruction.
 */
int _ltt_event_unregister(struct ltt_event *event)
{
	int ret = -EINVAL;

	switch (event->instrumentation) {
	case LTTNG_UST_TRACEPOINT:
		if (event->desc) {
			ret = __tracepoint_probe_unregister(event->desc->name,
							  event->desc->probe_callback,
							  event);
			if (ret)
				return ret;
		} else {
			remove_pending_probe(event->pending_probe);
			ret = 0;
		}
		break;
	default:
		WARN_ON_ONCE(1);
	}
	return ret;
}

/*
 * Only used internally at session destruction.
 */
static
void _ltt_event_destroy(struct ltt_event *event)
{
	switch (event->instrumentation) {
	case LTTNG_UST_TRACEPOINT:
		if (event->desc) {
			ltt_event_put(event->desc);
		}
		break;
	default:
		WARN_ON_ONCE(1);
	}
	cds_list_del(&event->list);
	lttng_destroy_context(event->ctx);
	free(event);
}

/*
 * We have exclusive access to our metadata buffer (protected by the
 * ust_lock), so we can do racy operations such as looking for
 * remaining space left in packet and write, since mutual exclusion
 * protects us from concurrent writes.
 */
int lttng_metadata_printf(struct ltt_session *session,
			  const char *fmt, ...)
{
	struct lttng_ust_lib_ring_buffer_ctx ctx;
	struct ltt_channel *chan = session->metadata;
	char *str = NULL;
	int ret = 0, waitret;
	size_t len, reserve_len, pos;
	va_list ap;

	WARN_ON_ONCE(!CMM_ACCESS_ONCE(session->active));

	va_start(ap, fmt);
	ret = vasprintf(&str, fmt, ap);
	va_end(ap);
	if (ret < 0)
		return -ENOMEM;

	len = strlen(str);
	pos = 0;

	for (pos = 0; pos < len; pos += reserve_len) {
		reserve_len = min_t(size_t,
				chan->ops->packet_avail_size(chan->chan, chan->handle),
				len - pos);
		lib_ring_buffer_ctx_init(&ctx, chan->chan, NULL, reserve_len,
					 sizeof(char), -1, chan->handle);
		/*
		 * We don't care about metadata buffer's records lost
		 * count, because we always retry here. Report error if
		 * we need to bail out after timeout or being
		 * interrupted.
		 */
		waitret = wait_cond_interruptible_timeout(
			({
				ret = chan->ops->event_reserve(&ctx, 0);
				ret != -ENOBUFS || !ret;
			}),
			LTTNG_METADATA_TIMEOUT_MSEC);
		if (waitret == -ETIMEDOUT || waitret == -EINTR || ret) {
			DBG("LTTng: Failure to write metadata to buffers (%s)\n",
				waitret == -EINTR ? "interrupted" :
					(ret == -ENOBUFS ? "timeout" : "I/O error"));
			if (waitret == -EINTR)
				ret = waitret;
			goto end;
		}
		chan->ops->event_write(&ctx, &str[pos], reserve_len);
		chan->ops->event_commit(&ctx);
	}
end:
	free(str);
	return ret;
}

static
int _ltt_field_statedump(struct ltt_session *session,
			 const struct lttng_event_field *field)
{
	int ret = 0;

	switch (field->type.atype) {
	case atype_integer:
		ret = lttng_metadata_printf(session,
			"		integer { size = %u; align = %u; signed = %u; encoding = %s; base = %u;%s } _%s;\n",
			field->type.u.basic.integer.size,
			field->type.u.basic.integer.alignment,
			field->type.u.basic.integer.signedness,
			(field->type.u.basic.integer.encoding == lttng_encode_none)
				? "none"
				: (field->type.u.basic.integer.encoding == lttng_encode_UTF8)
					? "UTF8"
					: "ASCII",
			field->type.u.basic.integer.base,
#if (BYTE_ORDER == BIG_ENDIAN)
			field->type.u.basic.integer.reverse_byte_order ? " byte_order = le;" : "",
#else
			field->type.u.basic.integer.reverse_byte_order ? " byte_order = be;" : "",
#endif
			field->name);
		break;
	case atype_float:
		ret = lttng_metadata_printf(session,
			"		floating_point { exp_dig = %u; mant_dig = %u; align = %u;%s } _%s;\n",
			field->type.u.basic._float.exp_dig,
			field->type.u.basic._float.mant_dig,
			field->type.u.basic._float.alignment,
#if (BYTE_ORDER == BIG_ENDIAN)
			field->type.u.basic.integer.reverse_byte_order ? " byte_order = le;" : "",
#else
			field->type.u.basic.integer.reverse_byte_order ? " byte_order = be;" : "",
#endif
			field->name);
		break;
	case atype_enum:
		ret = lttng_metadata_printf(session,
			"		%s %s;\n",
			field->type.u.basic.enumeration.name,
			field->name);
		break;
	case atype_array:
	{
		const struct lttng_basic_type *elem_type;

		elem_type = &field->type.u.array.elem_type;
		ret = lttng_metadata_printf(session,
			"		integer { size = %u; align = %u; signed = %u; encoding = %s; base = %u;%s } _%s[%u];\n",
			elem_type->u.basic.integer.size,
			elem_type->u.basic.integer.alignment,
			elem_type->u.basic.integer.signedness,
			(elem_type->u.basic.integer.encoding == lttng_encode_none)
				? "none"
				: (elem_type->u.basic.integer.encoding == lttng_encode_UTF8)
					? "UTF8"
					: "ASCII",
			elem_type->u.basic.integer.base,
#if (BYTE_ORDER == BIG_ENDIAN)
			elem_type->u.basic.integer.reverse_byte_order ? " byte_order = le;" : "",
#else
			elem_type->u.basic.integer.reverse_byte_order ? " byte_order = be;" : "",
#endif
			field->name, field->type.u.array.length);
		break;
	}
	case atype_sequence:
	{
		const struct lttng_basic_type *elem_type;
		const struct lttng_basic_type *length_type;

		elem_type = &field->type.u.sequence.elem_type;
		length_type = &field->type.u.sequence.length_type;
		ret = lttng_metadata_printf(session,
			"		integer { size = %u; align = %u; signed = %u; encoding = %s; base = %u;%s } __%s_length;\n",
			length_type->u.basic.integer.size,
			(unsigned int) length_type->u.basic.integer.alignment,
			length_type->u.basic.integer.signedness,
			(length_type->u.basic.integer.encoding == lttng_encode_none)
				? "none"
				: ((length_type->u.basic.integer.encoding == lttng_encode_UTF8)
					? "UTF8"
					: "ASCII"),
			length_type->u.basic.integer.base,
#if (BYTE_ORDER == BIG_ENDIAN)
			length_type->u.basic.integer.reverse_byte_order ? " byte_order = le;" : "",
#else
			length_type->u.basic.integer.reverse_byte_order ? " byte_order = be;" : "",
#endif
			field->name);
		if (ret)
			return ret;

		ret = lttng_metadata_printf(session,
			"		integer { size = %u; align = %u; signed = %u; encoding = %s; base = %u;%s } _%s[ __%s_length ];\n",
			elem_type->u.basic.integer.size,
			(unsigned int) elem_type->u.basic.integer.alignment,
			elem_type->u.basic.integer.signedness,
			(elem_type->u.basic.integer.encoding == lttng_encode_none)
				? "none"
				: ((elem_type->u.basic.integer.encoding == lttng_encode_UTF8)
					? "UTF8"
					: "ASCII"),
			elem_type->u.basic.integer.base,
#if (BYTE_ORDER == BIG_ENDIAN)
			elem_type->u.basic.integer.reverse_byte_order ? " byte_order = le;" : "",
#else
			elem_type->u.basic.integer.reverse_byte_order ? " byte_order = be;" : "",
#endif
			field->name,
			field->name);
		break;
	}

	case atype_string:
		/* Default encoding is UTF8 */
		ret = lttng_metadata_printf(session,
			"		string%s _%s;\n",
			field->type.u.basic.string.encoding == lttng_encode_ASCII ?
				" { encoding = ASCII; }" : "",
			field->name);
		break;
	default:
		WARN_ON_ONCE(1);
		return -EINVAL;
	}
	return ret;
}

static
int _ltt_context_metadata_statedump(struct ltt_session *session,
				    struct lttng_ctx *ctx)
{
	int ret = 0;
	int i;

	if (!ctx)
		return 0;
	for (i = 0; i < ctx->nr_fields; i++) {
		const struct lttng_ctx_field *field = &ctx->fields[i];

		ret = _ltt_field_statedump(session, &field->event_field);
		if (ret)
			return ret;
	}
	return ret;
}

static
int _ltt_fields_metadata_statedump(struct ltt_session *session,
				   struct ltt_event *event)
{
	const struct lttng_event_desc *desc = event->desc;
	int ret = 0;
	int i;

	for (i = 0; i < desc->nr_fields; i++) {
		const struct lttng_event_field *field = &desc->fields[i];

		ret = _ltt_field_statedump(session, field);
		if (ret)
			return ret;
	}
	return ret;
}

static
int _ltt_event_metadata_statedump(struct ltt_session *session,
				  struct ltt_channel *chan,
				  struct ltt_event *event)
{
	int ret = 0;

	if (event->metadata_dumped || !CMM_ACCESS_ONCE(session->active))
		return 0;
	if (chan == session->metadata)
		return 0;
	/*
	 * Don't print events for which probe load is pending.
	 */
	if (!event->desc)
		return 0;

	ret = lttng_metadata_printf(session,
		"event {\n"
		"	name = \"%s\";\n"
		"	id = %u;\n"
		"	stream_id = %u;\n",
		event->desc->name,
		event->id,
		event->chan->id);
	if (ret)
		goto end;

	if (event->desc->loglevel) {
		const struct tracepoint_loglevel_entry *ll_entry;

		ll_entry = *event->desc->loglevel;
		ret = lttng_metadata_printf(session,
			"	loglevel.identifier = \"%s\";\n"
			"	loglevel.value = %lld;\n",
			ll_entry->identifier,
			(long long) ll_entry->value);
		if (ret)
			goto end;
	}

	if (event->ctx) {
		ret = lttng_metadata_printf(session,
			"	context := struct {\n");
		if (ret)
			goto end;
	}
	ret = _ltt_context_metadata_statedump(session, event->ctx);
	if (ret)
		goto end;
	if (event->ctx) {
		ret = lttng_metadata_printf(session,
			"	};\n");
		if (ret)
			goto end;
	}

	ret = lttng_metadata_printf(session,
		"	fields := struct {\n"
		);
	if (ret)
		goto end;

	ret = _ltt_fields_metadata_statedump(session, event);
	if (ret)
		goto end;

	/*
	 * LTTng space reservation can only reserve multiples of the
	 * byte size.
	 */
	ret = lttng_metadata_printf(session,
		"	};\n"
		"};\n\n");
	if (ret)
		goto end;

	event->metadata_dumped = 1;
end:
	return ret;

}

static
int _ltt_channel_metadata_statedump(struct ltt_session *session,
				    struct ltt_channel *chan)
{
	int ret = 0;

	if (chan->metadata_dumped || !CMM_ACCESS_ONCE(session->active))
		return 0;
	if (chan == session->metadata)
		return 0;

	WARN_ON_ONCE(!chan->header_type);
	ret = lttng_metadata_printf(session,
		"stream {\n"
		"	id = %u;\n"
		"	event.header := %s;\n"
		"	packet.context := struct packet_context;\n",
		chan->id,
		chan->header_type == 1 ? "struct event_header_compact" :
			"struct event_header_large");
	if (ret)
		goto end;

	if (chan->ctx) {
		ret = lttng_metadata_printf(session,
			"	event.context := struct {\n");
		if (ret)
			goto end;
	}
	ret = _ltt_context_metadata_statedump(session, chan->ctx);
	if (ret)
		goto end;
	if (chan->ctx) {
		ret = lttng_metadata_printf(session,
			"	};\n");
		if (ret)
			goto end;
	}

	ret = lttng_metadata_printf(session,
		"};\n\n");

	chan->metadata_dumped = 1;
end:
	return ret;
}

static
int _ltt_stream_packet_context_declare(struct ltt_session *session)
{
	return lttng_metadata_printf(session,
		"struct packet_context {\n"
		"	uint64_clock_monotonic_t timestamp_begin;\n"
		"	uint64_clock_monotonic_t timestamp_end;\n"
		"	uint32_t events_discarded;\n"
		"	uint32_t content_size;\n"
		"	uint32_t packet_size;\n"
		"	uint32_t cpu_id;\n"
		"};\n\n"
		);
}

/*
 * Compact header:
 * id: range: 0 - 30.
 * id 31 is reserved to indicate an extended header.
 *
 * Large header:
 * id: range: 0 - 65534.
 * id 65535 is reserved to indicate an extended header.
 */
static
int _ltt_event_header_declare(struct ltt_session *session)
{
	return lttng_metadata_printf(session,
	"struct event_header_compact {\n"
	"	enum : uint5_t { compact = 0 ... 30, extended = 31 } id;\n"
	"	variant <id> {\n"
	"		struct {\n"
	"			uint27_clock_monotonic_t timestamp;\n"
	"		} compact;\n"
	"		struct {\n"
	"			uint32_t id;\n"
	"			uint64_clock_monotonic_t timestamp;\n"
	"		} extended;\n"
	"	} v;\n"
	"} align(%u);\n"
	"\n"
	"struct event_header_large {\n"
	"	enum : uint16_t { compact = 0 ... 65534, extended = 65535 } id;\n"
	"	variant <id> {\n"
	"		struct {\n"
	"			uint32_clock_monotonic_t timestamp;\n"
	"		} compact;\n"
	"		struct {\n"
	"			uint32_t id;\n"
	"			uint64_clock_monotonic_t timestamp;\n"
	"		} extended;\n"
	"	} v;\n"
	"} align(%u);\n\n",
	lttng_alignof(uint32_t) * CHAR_BIT,
	lttng_alignof(uint16_t) * CHAR_BIT
	);
}

/*
 * Approximation of NTP time of day to clock monotonic correlation,
 * taken at start of trace.
 * Yes, this is only an approximation. Yes, we can (and will) do better
 * in future versions.
 */
static
uint64_t measure_clock_offset(void)
{
	uint64_t offset, monotonic[2], realtime;
	struct timespec rts = { 0, 0 };
	int ret;

	monotonic[0] = trace_clock_read64();
	ret = clock_gettime(CLOCK_REALTIME, &rts);	
	if (ret < 0)
		return 0;
	monotonic[1] = trace_clock_read64();
	offset = (monotonic[0] + monotonic[1]) >> 1;
	realtime = rts.tv_sec * 1000000000ULL;
	realtime += rts.tv_nsec;
	offset = realtime - offset;
	return offset;
}

/*
 * Output metadata into this session's metadata buffers.
 */
static
int _ltt_session_metadata_statedump(struct ltt_session *session)
{
	unsigned char *uuid_c = session->uuid;
	char uuid_s[37], clock_uuid_s[CLOCK_UUID_LEN];
	struct ltt_channel *chan;
	struct ltt_event *event;
	int ret = 0;

	if (!CMM_ACCESS_ONCE(session->active))
		return 0;
	if (session->metadata_dumped)
		goto skip_session;
	if (!session->metadata) {
		DBG("LTTng: attempt to start tracing, but metadata channel is not found. Operation abort.\n");
		return -EPERM;
	}

	snprintf(uuid_s, sizeof(uuid_s),
		"%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x",
		uuid_c[0], uuid_c[1], uuid_c[2], uuid_c[3],
		uuid_c[4], uuid_c[5], uuid_c[6], uuid_c[7],
		uuid_c[8], uuid_c[9], uuid_c[10], uuid_c[11],
		uuid_c[12], uuid_c[13], uuid_c[14], uuid_c[15]);

	ret = lttng_metadata_printf(session,
		"typealias integer { size = 8; align = %u; signed = false; } := uint8_t;\n"
		"typealias integer { size = 16; align = %u; signed = false; } := uint16_t;\n"
		"typealias integer { size = 32; align = %u; signed = false; } := uint32_t;\n"
		"typealias integer { size = 64; align = %u; signed = false; } := uint64_t;\n"
		"typealias integer { size = 5; align = 1; signed = false; } := uint5_t;\n"
		"typealias integer { size = 27; align = 1; signed = false; } := uint27_t;\n"
		"\n"
		"trace {\n"
		"	major = %u;\n"
		"	minor = %u;\n"
		"	uuid = \"%s\";\n"
		"	byte_order = %s;\n"
		"	packet.header := struct {\n"
		"		uint32_t magic;\n"
		"		uint8_t  uuid[16];\n"
		"		uint32_t stream_id;\n"
		"	};\n"
		"};\n\n",
		lttng_alignof(uint8_t) * CHAR_BIT,
		lttng_alignof(uint16_t) * CHAR_BIT,
		lttng_alignof(uint32_t) * CHAR_BIT,
		lttng_alignof(uint64_t) * CHAR_BIT,
		CTF_VERSION_MAJOR,
		CTF_VERSION_MINOR,
		uuid_s,
#if (BYTE_ORDER == BIG_ENDIAN)
		"be"
#else
		"le"
#endif
		);
	if (ret)
		goto end;

	ret = lttng_metadata_printf(session,
		"clock {\n"
		"	name = %s;\n",
		"monotonic"
		);
	if (ret)
		goto end;

	if (!trace_clock_uuid(clock_uuid_s)) {
		ret = lttng_metadata_printf(session,
			"	uuid = \"%s\";\n",
			clock_uuid_s
			);
		if (ret)
			goto end;
	}

	ret = lttng_metadata_printf(session,
		"	description = \"Monotonic Clock\";\n"
		"	freq = %" PRIu64 "; /* Frequency, in Hz */\n"
		"	/* clock value offset from Epoch is: offset * (1/freq) */\n"
		"	offset = %" PRIu64 ";\n"
		"};\n\n",
		trace_clock_freq(),
		measure_clock_offset()
		);
	if (ret)
		goto end;

	ret = lttng_metadata_printf(session,
		"typealias integer {\n"
		"	size = 27; align = 1; signed = false;\n"
		"	map = clock.monotonic.value;\n"
		"} := uint27_clock_monotonic_t;\n"
		"\n"
		"typealias integer {\n"
		"	size = 32; align = %u; signed = false;\n"
		"	map = clock.monotonic.value;\n"
		"} := uint32_clock_monotonic_t;\n"
		"\n"
		"typealias integer {\n"
		"	size = 64; align = %u; signed = false;\n"
		"	map = clock.monotonic.value;\n"
		"} := uint64_clock_monotonic_t;\n\n",
		lttng_alignof(uint32_t) * CHAR_BIT,
		lttng_alignof(uint64_t) * CHAR_BIT
		);
	if (ret)
		goto end;

	ret = _ltt_stream_packet_context_declare(session);
	if (ret)
		goto end;

	ret = _ltt_event_header_declare(session);
	if (ret)
		goto end;

skip_session:
	cds_list_for_each_entry(chan, &session->chan, list) {
		ret = _ltt_channel_metadata_statedump(session, chan);
		if (ret)
			goto end;
	}

	cds_list_for_each_entry(event, &session->events, list) {
		ret = _ltt_event_metadata_statedump(session, event->chan, event);
		if (ret)
			goto end;
	}
	session->metadata_dumped = 1;
end:
	return ret;
}

void lttng_ust_events_exit(void)
{
	struct ltt_session *session, *tmpsession;

	cds_list_for_each_entry_safe(session, tmpsession, &sessions, list)
		ltt_session_destroy(session);
}