X-Git-Url: http://git.liburcu.org/?a=blobdiff_plain;f=lttng-events.c;h=1ccbffd67fe60a961de18cf6b4c58d7334d1c155;hb=1e543e7c292839b931f808ad341c85f24a859d00;hp=f2a9a8a58498f1f1a833e8ec1b97e76093391982;hpb=64af2437213d40fdb1f386ffcad2c83ed5c11185;p=lttng-modules.git

diff --git a/lttng-events.c b/lttng-events.c
index f2a9a8a5..1ccbffd6 100644
--- a/lttng-events.c
+++ b/lttng-events.c
@@ -1,23 +1,10 @@
-/*
+/* SPDX-License-Identifier: (GPL-2.0 or LGPL-2.1)
+ *
  * lttng-events.c
  *
  * Holds LTTng per-session event registry.
  *
  * Copyright (C) 2010-2012 Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
- *
- * This library is free software; you can redistribute it and/or
- * modify it under the terms of the GNU Lesser General Public
- * License as published by the Free Software Foundation; only
- * version 2.1 of the License.
- *
- * This library is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- * Lesser General Public License for more details.
- *
- * You should have received a copy of the GNU Lesser General Public
- * License along with this library; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
 /*
@@ -40,18 +27,20 @@
 #include <linux/jhash.h>
 #include <linux/uaccess.h>
 #include <linux/vmalloc.h>
+#include <linux/uuid.h>
+#include <linux/dmi.h>
 
-#include <wrapper/uuid.h>
 #include <wrapper/vmalloc.h>	/* for wrapper_vmalloc_sync_all() */
 #include <wrapper/random.h>
 #include <wrapper/tracepoint.h>
 #include <wrapper/list.h>
+#include <wrapper/types.h>
 #include <lttng-kernel-version.h>
 #include <lttng-events.h>
 #include <lttng-tracer.h>
 #include <lttng-abi-old.h>
 #include <lttng-endian.h>
-#include <wrapper/vzalloc.h>
+#include <lttng-string-utils.h>
 #include <wrapper/ringbuffer/backend.h>
 #include <wrapper/ringbuffer/frontend.h>
 
@@ -87,7 +76,12 @@ int _lttng_field_statedump(struct lttng_session *session,
 
 void synchronize_trace(void)
 {
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(5,1,0))
+	synchronize_rcu();
+#else
 	synchronize_sched();
+#endif
+
 #if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,4,0))
 #ifdef CONFIG_PREEMPT_RT_FULL
 	synchronize_rcu();
@@ -130,7 +124,7 @@ struct lttng_session *lttng_session_create(void)
 	int i;
 
 	mutex_lock(&sessions_mutex);
-	session = kzalloc(sizeof(struct lttng_session), GFP_KERNEL);
+	session = lttng_kvzalloc(sizeof(struct lttng_session), GFP_KERNEL);
 	if (!session)
 		goto err;
 	INIT_LIST_HEAD(&session->chan);
@@ -141,7 +135,7 @@ struct lttng_session *lttng_session_create(void)
 			GFP_KERNEL);
 	if (!metadata_cache)
 		goto err_free_session;
-	metadata_cache->data = lttng_vzalloc(METADATA_CACHE_DEFAULT_SIZE);
+	metadata_cache->data = vzalloc(METADATA_CACHE_DEFAULT_SIZE);
 	if (!metadata_cache->data)
 		goto err_free_cache;
 	metadata_cache->cache_alloc = METADATA_CACHE_DEFAULT_SIZE;
@@ -155,13 +149,25 @@ struct lttng_session *lttng_session_create(void)
 	for (i = 0; i < LTTNG_EVENT_HT_SIZE; i++)
 		INIT_HLIST_HEAD(&session->events_ht.table[i]);
 	list_add(&session->list, &sessions);
+	session->pid_tracker.session = session;
+	session->pid_tracker.tracker_type = TRACKER_PID;
+	session->vpid_tracker.session = session;
+	session->vpid_tracker.tracker_type = TRACKER_VPID;
+	session->uid_tracker.session = session;
+	session->uid_tracker.tracker_type = TRACKER_UID;
+	session->vuid_tracker.session = session;
+	session->vuid_tracker.tracker_type = TRACKER_VUID;
+	session->gid_tracker.session = session;
+	session->gid_tracker.tracker_type = TRACKER_GID;
+	session->vgid_tracker.session = session;
+	session->vgid_tracker.tracker_type = TRACKER_VGID;
 	mutex_unlock(&sessions_mutex);
 	return session;
 
 err_free_cache:
 	kfree(metadata_cache);
 err_free_session:
-	kfree(session);
+	lttng_kvfree(session);
 err:
 	mutex_unlock(&sessions_mutex);
 	return NULL;
@@ -184,7 +190,7 @@ void lttng_session_destroy(struct lttng_session *session)
 	int ret;
 
 	mutex_lock(&sessions_mutex);
-	ACCESS_ONCE(session->active) = 0;
+	WRITE_ONCE(session->active, 0);
 	list_for_each_entry(chan, &session->chan, list) {
 		ret = lttng_syscalls_unregister(chan);
 		WARN_ON(ret);
@@ -205,12 +211,26 @@ void lttng_session_destroy(struct lttng_session *session)
 	}
 	list_for_each_entry(metadata_stream, &session->metadata_cache->metadata_stream, list)
 		_lttng_metadata_channel_hangup(metadata_stream);
-	if (session->pid_tracker)
-		lttng_pid_tracker_destroy(session->pid_tracker);
+	lttng_id_tracker_destroy(&session->pid_tracker, false);
+	lttng_id_tracker_destroy(&session->vpid_tracker, false);
+	lttng_id_tracker_destroy(&session->uid_tracker, false);
+	lttng_id_tracker_destroy(&session->vuid_tracker, false);
+	lttng_id_tracker_destroy(&session->gid_tracker, false);
+	lttng_id_tracker_destroy(&session->vgid_tracker, false);
 	kref_put(&session->metadata_cache->refcount, metadata_cache_destroy);
 	list_del(&session->list);
 	mutex_unlock(&sessions_mutex);
-	kfree(session);
+	lttng_kvfree(session);
+}
+
+int lttng_session_statedump(struct lttng_session *session)
+{
+	int ret;
+
+	mutex_lock(&sessions_mutex);
+	ret = lttng_statedump_start(session);
+	mutex_unlock(&sessions_mutex);
+	return ret;
 }
 
 int lttng_session_enable(struct lttng_session *session)
@@ -227,6 +247,9 @@ int lttng_session_enable(struct lttng_session *session)
 	/* Set transient enabler state to "enabled" */
 	session->tstate = 1;
 
+	/* We need to sync enablers with session before activation. */
+	lttng_session_sync_enablers(session);
+
 	/*
 	 * Snapshot the number of events per channel to know the type of header
 	 * we need to use.
@@ -240,23 +263,22 @@ int lttng_session_enable(struct lttng_session *session)
 			chan->header_type = 2;	/* large */
 	}
 
-	/* We need to sync enablers with session before activation. */
-	lttng_session_sync_enablers(session);
-
 	/* Clear each stream's quiescent state. */
-	list_for_each_entry(chan, &session->chan, list)
-		lib_ring_buffer_clear_quiescent_channel(chan->chan);
+	list_for_each_entry(chan, &session->chan, list) {
+		if (chan->channel_type != METADATA_CHANNEL)
+			lib_ring_buffer_clear_quiescent_channel(chan->chan);
+	}
 
-	ACCESS_ONCE(session->active) = 1;
-	ACCESS_ONCE(session->been_active) = 1;
+	WRITE_ONCE(session->active, 1);
+	WRITE_ONCE(session->been_active, 1);
 	ret = _lttng_session_metadata_statedump(session);
 	if (ret) {
-		ACCESS_ONCE(session->active) = 0;
+		WRITE_ONCE(session->active, 0);
 		goto end;
 	}
 	ret = lttng_statedump_start(session);
 	if (ret)
-		ACCESS_ONCE(session->active) = 0;
+		WRITE_ONCE(session->active, 0);
 end:
 	mutex_unlock(&sessions_mutex);
 	return ret;
@@ -272,15 +294,17 @@ int lttng_session_disable(struct lttng_session *session)
 		ret = -EBUSY;
 		goto end;
 	}
-	ACCESS_ONCE(session->active) = 0;
+	WRITE_ONCE(session->active, 0);
 
 	/* Set transient enabler state to "disabled" */
 	session->tstate = 0;
 	lttng_session_sync_enablers(session);
 
 	/* Set each stream's quiescent state. */
-	list_for_each_entry(chan, &session->chan, list)
-		lib_ring_buffer_set_quiescent_channel(chan->chan);
+	list_for_each_entry(chan, &session->chan, list) {
+		if (chan->channel_type != METADATA_CHANNEL)
+			lib_ring_buffer_set_quiescent_channel(chan->chan);
+	}
 end:
 	mutex_unlock(&sessions_mutex);
 	return ret;
@@ -326,8 +350,6 @@ end:
 	return ret;
 }
 
-
-
 int lttng_channel_enable(struct lttng_channel *channel)
 {
 	int ret = 0;
@@ -345,7 +367,7 @@ int lttng_channel_enable(struct lttng_channel *channel)
 	channel->tstate = 1;
 	lttng_session_sync_enablers(channel->session);
 	/* Set atomically the state to "enabled" */
-	ACCESS_ONCE(channel->enabled) = 1;
+	WRITE_ONCE(channel->enabled, 1);
 end:
 	mutex_unlock(&sessions_mutex);
 	return ret;
@@ -365,7 +387,7 @@ int lttng_channel_disable(struct lttng_channel *channel)
 		goto end;
 	}
 	/* Set atomically the state to "disabled" */
-	ACCESS_ONCE(channel->enabled) = 0;
+	WRITE_ONCE(channel->enabled, 0);
 	/* Set transient enabler state to "enabled" */
 	channel->tstate = 0;
 	lttng_session_sync_enablers(channel->session);
@@ -394,8 +416,9 @@ int lttng_event_enable(struct lttng_event *event)
 		break;
 	case LTTNG_KERNEL_KPROBE:
 	case LTTNG_KERNEL_FUNCTION:
+	case LTTNG_KERNEL_UPROBE:
 	case LTTNG_KERNEL_NOOP:
-		ACCESS_ONCE(event->enabled) = 1;
+		WRITE_ONCE(event->enabled, 1);
 		break;
 	case LTTNG_KERNEL_KRETPROBE:
 		ret = lttng_kretprobes_event_enable_state(event, 1);
@@ -429,8 +452,9 @@ int lttng_event_disable(struct lttng_event *event)
 		break;
 	case LTTNG_KERNEL_KPROBE:
 	case LTTNG_KERNEL_FUNCTION:
+	case LTTNG_KERNEL_UPROBE:
 	case LTTNG_KERNEL_NOOP:
-		ACCESS_ONCE(event->enabled) = 0;
+		WRITE_ONCE(event->enabled, 0);
 		break;
 	case LTTNG_KERNEL_KRETPROBE:
 		ret = lttng_kretprobes_event_enable_state(event, 0);
@@ -575,6 +599,7 @@ struct lttng_event *_lttng_event_create(struct lttng_channel *chan,
 		event_name = event_desc->name;
 		break;
 	case LTTNG_KERNEL_KPROBE:
+	case LTTNG_KERNEL_UPROBE:
 	case LTTNG_KERNEL_KRETPROBE:
 	case LTTNG_KERNEL_FUNCTION:
 	case LTTNG_KERNEL_NOOP:
@@ -738,6 +763,28 @@ struct lttng_event *_lttng_event_create(struct lttng_channel *chan,
 			goto register_error;
 		}
 		break;
+	case LTTNG_KERNEL_UPROBE:
+		/*
+		 * Needs to be explicitly enabled after creation, since
+		 * we may want to apply filters.
+		 */
+		event->enabled = 0;
+		event->registered = 1;
+
+		/*
+		 * Populate lttng_event structure before event
+		 * registration.
+		 */
+		smp_wmb();
+
+		ret = lttng_uprobes_register(event_param->name,
+				event_param->u.uprobe.fd,
+				event);
+		if (ret)
+			goto register_error;
+		ret = try_module_get(event->desc->owner);
+		WARN_ON_ONCE(!ret);
+		break;
 	default:
 		WARN_ON_ONCE(1);
 		ret = -EINVAL;
@@ -800,6 +847,7 @@ void register_event(struct lttng_event *event)
 			desc->name);
 		break;
 	case LTTNG_KERNEL_KPROBE:
+	case LTTNG_KERNEL_UPROBE:
 	case LTTNG_KERNEL_KRETPROBE:
 	case LTTNG_KERNEL_FUNCTION:
 	case LTTNG_KERNEL_NOOP:
@@ -849,6 +897,10 @@ int _lttng_event_unregister(struct lttng_event *event)
 	case LTTNG_KERNEL_NOOP:
 		ret = 0;
 		break;
+	case LTTNG_KERNEL_UPROBE:
+		lttng_uprobes_unregister(event);
+		ret = 0;
+		break;
 	default:
 		WARN_ON_ONCE(1);
 	}
@@ -882,6 +934,10 @@ void _lttng_event_destroy(struct lttng_event *event)
 	case LTTNG_KERNEL_NOOP:
 	case LTTNG_KERNEL_SYSCALL:
 		break;
+	case LTTNG_KERNEL_UPROBE:
+		module_put(event->desc->owner);
+		lttng_uprobes_destroy_private(event);
+		break;
 	default:
 		WARN_ON_ONCE(1);
 	}
@@ -890,91 +946,85 @@ void _lttng_event_destroy(struct lttng_event *event)
 	kmem_cache_free(event_cache, event);
 }
 
-int lttng_session_track_pid(struct lttng_session *session, int pid)
+struct lttng_id_tracker *get_tracker(struct lttng_session *session,
+		enum tracker_type tracker_type)
+{
+	switch (tracker_type) {
+	case TRACKER_PID:
+		return &session->pid_tracker;
+	case TRACKER_VPID:
+		return &session->vpid_tracker;
+	case TRACKER_UID:
+		return &session->uid_tracker;
+	case TRACKER_VUID:
+		return &session->vuid_tracker;
+	case TRACKER_GID:
+		return &session->gid_tracker;
+	case TRACKER_VGID:
+		return &session->vgid_tracker;
+	default:
+		WARN_ON_ONCE(1);
+		return NULL;
+	}
+}
+
+int lttng_session_track_id(struct lttng_session *session,
+		enum tracker_type tracker_type, int id)
 {
+	struct lttng_id_tracker *tracker;
 	int ret;
 
-	if (pid < -1)
+	tracker = get_tracker(session, tracker_type);
+	if (!tracker)
+		return -EINVAL;
+	if (id < -1)
 		return -EINVAL;
 	mutex_lock(&sessions_mutex);
-	if (pid == -1) {
-		/* track all pids: destroy tracker. */
-		if (session->pid_tracker) {
-			struct lttng_pid_tracker *lpf;
-
-			lpf = session->pid_tracker;
-			rcu_assign_pointer(session->pid_tracker, NULL);
-			synchronize_trace();
-			lttng_pid_tracker_destroy(lpf);
-		}
+	if (id == -1) {
+		/* track all ids: destroy tracker. */
+		lttng_id_tracker_destroy(tracker, true);
 		ret = 0;
 	} else {
-		if (!session->pid_tracker) {
-			struct lttng_pid_tracker *lpf;
-
-			lpf = lttng_pid_tracker_create();
-			if (!lpf) {
-				ret = -ENOMEM;
-				goto unlock;
-			}
-			ret = lttng_pid_tracker_add(lpf, pid);
-			rcu_assign_pointer(session->pid_tracker, lpf);
-		} else {
-			ret = lttng_pid_tracker_add(session->pid_tracker, pid);
-		}
+		ret = lttng_id_tracker_add(tracker, id);
 	}
-unlock:
 	mutex_unlock(&sessions_mutex);
 	return ret;
 }
 
-int lttng_session_untrack_pid(struct lttng_session *session, int pid)
+int lttng_session_untrack_id(struct lttng_session *session,
+		enum tracker_type tracker_type, int id)
 {
+	struct lttng_id_tracker *tracker;
 	int ret;
 
-	if (pid < -1)
+	tracker = get_tracker(session, tracker_type);
+	if (!tracker)
+		return -EINVAL;
+	if (id < -1)
 		return -EINVAL;
 	mutex_lock(&sessions_mutex);
-	if (pid == -1) {
-		/* untrack all pids: replace by empty tracker. */
-		struct lttng_pid_tracker *old_lpf = session->pid_tracker;
-		struct lttng_pid_tracker *lpf;
-
-		lpf = lttng_pid_tracker_create();
-		if (!lpf) {
-			ret = -ENOMEM;
-			goto unlock;
-		}
-		rcu_assign_pointer(session->pid_tracker, lpf);
-		synchronize_trace();
-		if (old_lpf)
-			lttng_pid_tracker_destroy(old_lpf);
-		ret = 0;
+	if (id == -1) {
+		/* untrack all ids: replace by empty tracker. */
+		ret = lttng_id_tracker_empty_set(tracker);
 	} else {
-		if (!session->pid_tracker) {
-			ret = -ENOENT;
-			goto unlock;
-		}
-		ret = lttng_pid_tracker_del(session->pid_tracker, pid);
+		ret = lttng_id_tracker_del(tracker, id);
 	}
-unlock:
 	mutex_unlock(&sessions_mutex);
 	return ret;
 }
 
 static
-void *pid_list_start(struct seq_file *m, loff_t *pos)
+void *id_list_start(struct seq_file *m, loff_t *pos)
 {
-	struct lttng_session *session = m->private;
-	struct lttng_pid_tracker *lpf;
-	struct lttng_pid_hash_node *e;
+	struct lttng_id_tracker *id_tracker = m->private;
+	struct lttng_id_tracker_rcu *id_tracker_p = id_tracker->p;
+	struct lttng_id_hash_node *e;
 	int iter = 0, i;
 
 	mutex_lock(&sessions_mutex);
-	lpf = session->pid_tracker;
-	if (lpf) {
-		for (i = 0; i < LTTNG_PID_TABLE_SIZE; i++) {
-			struct hlist_head *head = &lpf->pid_hash[i];
+	if (id_tracker_p) {
+		for (i = 0; i < LTTNG_ID_TABLE_SIZE; i++) {
+			struct hlist_head *head = &id_tracker_p->id_hash[i];
 
 			lttng_hlist_for_each_entry(e, head, hlist) {
 				if (iter++ >= *pos)
@@ -982,9 +1032,9 @@ void *pid_list_start(struct seq_file *m, loff_t *pos)
 			}
 		}
 	} else {
-		/* PID tracker disabled. */
+		/* ID tracker disabled. */
 		if (iter >= *pos && iter == 0) {
-			return session;	/* empty tracker */
+			return id_tracker_p;	/* empty tracker */
 		}
 		iter++;
 	}
@@ -994,18 +1044,17 @@ void *pid_list_start(struct seq_file *m, loff_t *pos)
 
 /* Called with sessions_mutex held. */
 static
-void *pid_list_next(struct seq_file *m, void *p, loff_t *ppos)
+void *id_list_next(struct seq_file *m, void *p, loff_t *ppos)
 {
-	struct lttng_session *session = m->private;
-	struct lttng_pid_tracker *lpf;
-	struct lttng_pid_hash_node *e;
+	struct lttng_id_tracker *id_tracker = m->private;
+	struct lttng_id_tracker_rcu *id_tracker_p = id_tracker->p;
+	struct lttng_id_hash_node *e;
 	int iter = 0, i;
 
 	(*ppos)++;
-	lpf = session->pid_tracker;
-	if (lpf) {
-		for (i = 0; i < LTTNG_PID_TABLE_SIZE; i++) {
-			struct hlist_head *head = &lpf->pid_hash[i];
+	if (id_tracker_p) {
+		for (i = 0; i < LTTNG_ID_TABLE_SIZE; i++) {
+			struct hlist_head *head = &id_tracker_p->id_hash[i];
 
 			lttng_hlist_for_each_entry(e, head, hlist) {
 				if (iter++ >= *ppos)
@@ -1013,9 +1062,9 @@ void *pid_list_next(struct seq_file *m, void *p, loff_t *ppos)
 			}
 		}
 	} else {
-		/* PID tracker disabled. */
+		/* ID tracker disabled. */
 		if (iter >= *ppos && iter == 0)
-			return session;	/* empty tracker */
+			return p;	/* empty tracker */
 		iter++;
 	}
 
@@ -1024,67 +1073,91 @@ void *pid_list_next(struct seq_file *m, void *p, loff_t *ppos)
 }
 
 static
-void pid_list_stop(struct seq_file *m, void *p)
+void id_list_stop(struct seq_file *m, void *p)
 {
 	mutex_unlock(&sessions_mutex);
 }
 
 static
-int pid_list_show(struct seq_file *m, void *p)
+int id_list_show(struct seq_file *m, void *p)
 {
-	int pid;
+	struct lttng_id_tracker *id_tracker = m->private;
+	struct lttng_id_tracker_rcu *id_tracker_p = id_tracker->p;
+	int id;
 
-	if (p == m->private) {
+	if (p == id_tracker_p) {
 		/* Tracker disabled. */
-		pid = -1;
+		id = -1;
 	} else {
-		const struct lttng_pid_hash_node *e = p;
+		const struct lttng_id_hash_node *e = p;
 
-		pid = lttng_pid_tracker_get_node_pid(e);
+		id = lttng_id_tracker_get_node_id(e);
+	}
+	switch (id_tracker->tracker_type) {
+	case TRACKER_PID:
+		seq_printf(m,	"process { pid = %d; };\n", id);
+		break;
+	case TRACKER_VPID:
+		seq_printf(m,	"process { vpid = %d; };\n", id);
+		break;
+	case TRACKER_UID:
+		seq_printf(m,	"user { uid = %d; };\n", id);
+		break;
+	case TRACKER_VUID:
+		seq_printf(m,	"user { vuid = %d; };\n", id);
+		break;
+	case TRACKER_GID:
+		seq_printf(m,	"group { gid = %d; };\n", id);
+		break;
+	case TRACKER_VGID:
+		seq_printf(m,	"group { vgid = %d; };\n", id);
+		break;
+	default:
+		seq_printf(m,	"UNKNOWN { field = %d };\n", id);
 	}
-	seq_printf(m,	"process { pid = %d; };\n", pid);
 	return 0;
 }
 
 static
-const struct seq_operations lttng_tracker_pids_list_seq_ops = {
-	.start = pid_list_start,
-	.next = pid_list_next,
-	.stop = pid_list_stop,
-	.show = pid_list_show,
+const struct seq_operations lttng_tracker_ids_list_seq_ops = {
+	.start = id_list_start,
+	.next = id_list_next,
+	.stop = id_list_stop,
+	.show = id_list_show,
 };
 
 static
-int lttng_tracker_pids_list_open(struct inode *inode, struct file *file)
+int lttng_tracker_ids_list_open(struct inode *inode, struct file *file)
 {
-	return seq_open(file, &lttng_tracker_pids_list_seq_ops);
+	return seq_open(file, &lttng_tracker_ids_list_seq_ops);
 }
 
 static
-int lttng_tracker_pids_list_release(struct inode *inode, struct file *file)
+int lttng_tracker_ids_list_release(struct inode *inode, struct file *file)
 {
 	struct seq_file *m = file->private_data;
-	struct lttng_session *session = m->private;
+	struct lttng_id_tracker *id_tracker = m->private;
 	int ret;
 
-	WARN_ON_ONCE(!session);
+	WARN_ON_ONCE(!id_tracker);
 	ret = seq_release(inode, file);
-	if (!ret && session)
-		fput(session->file);
+	if (!ret)
+		fput(id_tracker->session->file);
 	return ret;
 }
 
-const struct file_operations lttng_tracker_pids_list_fops = {
+const struct file_operations lttng_tracker_ids_list_fops = {
 	.owner = THIS_MODULE,
-	.open = lttng_tracker_pids_list_open,
+	.open = lttng_tracker_ids_list_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
-	.release = lttng_tracker_pids_list_release,
+	.release = lttng_tracker_ids_list_release,
 };
 
-int lttng_session_list_tracker_pids(struct lttng_session *session)
+int lttng_session_list_tracker_ids(struct lttng_session *session,
+		enum tracker_type tracker_type)
 {
-	struct file *tracker_pids_list_file;
+	struct file *tracker_ids_list_file;
 	struct seq_file *m;
 	int file_fd, ret;
 
@@ -1094,30 +1167,32 @@ int lttng_session_list_tracker_pids(struct lttng_session *session)
 		goto fd_error;
 	}
 
-	tracker_pids_list_file = anon_inode_getfile("[lttng_tracker_pids_list]",
-					  &lttng_tracker_pids_list_fops,
+	tracker_ids_list_file = anon_inode_getfile("[lttng_tracker_ids_list]",
+					  &lttng_tracker_ids_list_fops,
 					  NULL, O_RDWR);
-	if (IS_ERR(tracker_pids_list_file)) {
-		ret = PTR_ERR(tracker_pids_list_file);
+	if (IS_ERR(tracker_ids_list_file)) {
+		ret = PTR_ERR(tracker_ids_list_file);
 		goto file_error;
 	}
-	if (atomic_long_add_unless(&session->file->f_count,
-		1, INT_MAX) == INT_MAX) {
+	if (!atomic_long_add_unless(&session->file->f_count, 1, LONG_MAX)) {
+		ret = -EOVERFLOW;
 		goto refcount_error;
 	}
-	ret = lttng_tracker_pids_list_fops.open(NULL, tracker_pids_list_file);
+	ret = lttng_tracker_ids_list_fops.open(NULL, tracker_ids_list_file);
 	if (ret < 0)
 		goto open_error;
-	m = tracker_pids_list_file->private_data;
-	m->private = session;
-	fd_install(file_fd, tracker_pids_list_file);
+	m = tracker_ids_list_file->private_data;
+
+	m->private = get_tracker(session, tracker_type);
+	BUG_ON(!m->private);
+	fd_install(file_fd, tracker_ids_list_file);
 
 	return file_fd;
 
 open_error:
 	atomic_long_dec(&session->file->f_count);
 refcount_error:
-	fput(tracker_pids_list_file);
+	fput(tracker_ids_list_file);
 file_error:
 	put_unused_fd(file_fd);
 fd_error:
@@ -1128,11 +1203,11 @@ fd_error:
  * Enabler management.
  */
 static
-int lttng_match_enabler_wildcard(const char *desc_name,
-		const char *name)
+int lttng_match_enabler_star_glob(const char *desc_name,
+		const char *pattern)
 {
-	/* Compare excluding final '*' */
-	if (strncmp(desc_name, name, strlen(name) - 1))
+	if (!strutils_star_glob_match(pattern, LTTNG_SIZE_MAX,
+			desc_name, LTTNG_SIZE_MAX))
 		return 0;
 	return 1;
 }
@@ -1177,8 +1252,8 @@ int lttng_desc_match_enabler(const struct lttng_event_desc *desc,
 		return -EINVAL;
 	}
 	switch (enabler->type) {
-	case LTTNG_ENABLER_WILDCARD:
-		return lttng_match_enabler_wildcard(desc_name, enabler_name);
+	case LTTNG_ENABLER_STAR_GLOB:
+		return lttng_match_enabler_star_glob(desc_name, enabler_name);
 	case LTTNG_ENABLER_NAME:
 		return lttng_match_enabler_name(desc_name, enabler_name);
 	default:
@@ -1431,6 +1506,18 @@ error_free:
 	return ret;
 }
 
+int lttng_event_add_callsite(struct lttng_event *event,
+		struct lttng_kernel_event_callsite __user *callsite)
+{
+
+	switch (event->instrumentation) {
+	case LTTNG_KERNEL_UPROBE:
+		return lttng_uprobes_add_callsite(event, callsite);
+	default:
+		return -EINVAL;
+	}
+}
+
 int lttng_enabler_attach_context(struct lttng_enabler *enabler,
 		struct lttng_kernel_context *context_param)
 {
@@ -1501,7 +1588,7 @@ void lttng_session_sync_enablers(struct lttng_session *session)
 		 */
 		enabled = enabled && session->tstate && event->chan->tstate;
 
-		ACCESS_ONCE(event->enabled) = enabled;
+		WRITE_ONCE(event->enabled, enabled);
 		/*
 		 * Sync tracepoint registration with event enabled
 		 * state.
@@ -1627,7 +1714,7 @@ int lttng_metadata_printf(struct lttng_session *session,
 	va_list ap;
 	struct lttng_metadata_stream *stream;
 
-	WARN_ON_ONCE(!ACCESS_ONCE(session->active));
+	WARN_ON_ONCE(!READ_ONCE(session->active));
 
 	va_start(ap, fmt);
 	str = kvasprintf(GFP_KERNEL, fmt, ap);
@@ -1645,7 +1732,7 @@ int lttng_metadata_printf(struct lttng_session *session,
 		tmp_cache_alloc_size = max_t(unsigned int,
 				session->metadata_cache->cache_alloc + len,
 				session->metadata_cache->cache_alloc << 1);
-		tmp_cache_realloc = lttng_vzalloc(tmp_cache_alloc_size);
+		tmp_cache_realloc = vzalloc(tmp_cache_alloc_size);
 		if (!tmp_cache_realloc)
 			goto err;
 		if (session->metadata_cache->data) {
@@ -1895,7 +1982,7 @@ int _lttng_enum_statedump(struct lttng_session *session,
 	if (ret)
 		goto end;
 	ret = lttng_metadata_printf(session,
-		"enum : integer { size = %u; align = %u; signed = %u; encoding = %s; base = %u; } {\n",
+		"enum : integer { size = %u; align = %u; signed = %u; encoding = %s; base = %u;%s } {\n",
 		container_type->size,
 		container_type->alignment,
 		container_type->signedness,
@@ -1904,7 +1991,13 @@ int _lttng_enum_statedump(struct lttng_session *session,
 			: (container_type->encoding == lttng_encode_UTF8)
 				? "UTF8"
 				: "ASCII",
-		container_type->base);
+		container_type->base,
+#if __BYTE_ORDER == __BIG_ENDIAN
+		container_type->reverse_byte_order ? " byte_order = le;" : ""
+#else
+		container_type->reverse_byte_order ? " byte_order = be;" : ""
+#endif
+		);
 	if (ret)
 	        goto end;
 	/* Dump all entries */
@@ -1941,33 +2034,46 @@ int _lttng_enum_statedump(struct lttng_session *session,
 			if (ret)
 				goto end;
 		}
-		ret = lttng_metadata_printf(session,
-				"\" = ");
+		ret = lttng_metadata_printf(session, "\"");
 		if (ret)
 			goto end;
-		if (entry->start.signedness)
-			ret = lttng_metadata_printf(session,
-				"%lld", (long long) entry->start.value);
-		else
-			ret = lttng_metadata_printf(session,
-				"%llu", entry->start.value);
-		if (ret)
-			goto end;
-		if (entry->start.signedness == entry->end.signedness &&
-				entry->start.value == entry->end.value) {
-			ret = lttng_metadata_printf(session,
-				",\n");
+
+		if (entry->options.is_auto) {
+			ret = lttng_metadata_printf(session, ",\n");
+			if (ret)
+				goto end;
 		} else {
-			if (entry->end.signedness) {
+			ret = lttng_metadata_printf(session,
+					" = ");
+			if (ret)
+				goto end;
+			if (entry->start.signedness)
 				ret = lttng_metadata_printf(session,
-					" ... %lld,\n", (long long) entry->end.value);
-			} else {
+					"%lld", (long long) entry->start.value);
+			else
+				ret = lttng_metadata_printf(session,
+					"%llu", entry->start.value);
+			if (ret)
+				goto end;
+			if (entry->start.signedness == entry->end.signedness &&
+					entry->start.value
+						== entry->end.value) {
 				ret = lttng_metadata_printf(session,
-					" ... %llu,\n", entry->end.value);
+					",\n");
+			} else {
+				if (entry->end.signedness) {
+					ret = lttng_metadata_printf(session,
+						" ... %lld,\n",
+						(long long) entry->end.value);
+				} else {
+					ret = lttng_metadata_printf(session,
+						" ... %llu,\n",
+						entry->end.value);
+				}
 			}
+			if (ret)
+				goto end;
 		}
-		if (ret)
-			goto end;
 	}
 	ret = print_tabs(session, nesting);
 	if (ret)
@@ -2015,6 +2121,7 @@ int _lttng_field_statedump(struct lttng_session *session,
 		ret = _lttng_enum_statedump(session, field, nesting);
 		break;
 	case atype_array:
+	case atype_array_bitfield:
 	{
 		const struct lttng_basic_type *elem_type;
 
@@ -2053,6 +2160,7 @@ int _lttng_field_statedump(struct lttng_session *session,
 		break;
 	}
 	case atype_sequence:
+	case atype_sequence_bitfield:
 	{
 		const struct lttng_basic_type *elem_type;
 		const struct lttng_basic_type *length_type;
@@ -2195,7 +2303,7 @@ int _lttng_event_metadata_statedump(struct lttng_session *session,
 {
 	int ret = 0;
 
-	if (event->metadata_dumped || !ACCESS_ONCE(session->active))
+	if (event->metadata_dumped || !READ_ONCE(session->active))
 		return 0;
 	if (chan->channel_type == METADATA_CHANNEL)
 		return 0;
@@ -2262,7 +2370,7 @@ int _lttng_channel_metadata_statedump(struct lttng_session *session,
 {
 	int ret = 0;
 
-	if (chan->metadata_dumped || !ACCESS_ONCE(session->active))
+	if (chan->metadata_dumped || !READ_ONCE(session->active))
 		return 0;
 
 	if (chan->channel_type == METADATA_CHANNEL)
@@ -2388,7 +2496,7 @@ int64_t measure_clock_offset(void)
 	/* Disable interrupts to increase correlation precision. */
 	local_irq_save(flags);
 	monotonic[0] = trace_clock_read64();
-	getnstimeofday(&rts);      
+	getnstimeofday(&rts);
 	monotonic[1] = trace_clock_read64();
 	local_irq_restore(flags);
 
@@ -2406,6 +2514,61 @@ int64_t measure_clock_offset(void)
 	return offset;
 }
 
+static
+int print_escaped_ctf_string(struct lttng_session *session, const char *string)
+{
+	int ret;
+	size_t i;
+	char cur;
+
+	i = 0;
+	cur = string[i];
+	while (cur != '\0') {
+		switch (cur) {
+		case '\n':
+			ret = lttng_metadata_printf(session, "%s", "\\n");
+			break;
+		case '\\':
+		case '"':
+			ret = lttng_metadata_printf(session, "%c", '\\');
+			if (ret)
+				goto error;
+			/* We still print the current char */
+			/* Fallthrough */
+		default:
+			ret = lttng_metadata_printf(session, "%c", cur);
+			break;
+		}
+
+		if (ret)
+			goto error;
+
+		cur = string[++i];
+	}
+error:
+	return ret;
+}
+
+static
+int print_metadata_escaped_field(struct lttng_session *session, const char *field,
+		const char *field_value)
+{
+	int ret;
+
+	ret = lttng_metadata_printf(session, "	%s = \"", field);
+	if (ret)
+		goto error;
+
+	ret = print_escaped_ctf_string(session, field_value);
+	if (ret)
+		goto error;
+
+	ret = lttng_metadata_printf(session, "\";\n");
+
+error:
+	return ret;
+}
+
 /*
  * Output metadata into this session's metadata buffers.
  * Must be called with sessions_mutex held.
@@ -2415,11 +2578,12 @@ int _lttng_session_metadata_statedump(struct lttng_session *session)
 {
 	unsigned char *uuid_c = session->uuid.b;
 	unsigned char uuid_s[37], clock_uuid_s[BOOT_ID_LEN];
+	const char *product_uuid;
 	struct lttng_channel *chan;
 	struct lttng_event *event;
 	int ret = 0;
 
-	if (!ACCESS_ONCE(session->active))
+	if (!READ_ONCE(session->active))
 		return 0;
 	if (session->metadata_dumped)
 		goto skip_session;
@@ -2481,7 +2645,7 @@ int _lttng_session_metadata_statedump(struct lttng_session *session)
 		"	tracer_major = %d;\n"
 		"	tracer_minor = %d;\n"
 		"	tracer_patchlevel = %d;\n"
-		"};\n\n",
+		"	trace_buffering_scheme = \"global\";\n",
 		current->nsproxy->uts_ns->name.nodename,
 		utsname()->sysname,
 		utsname()->release,
@@ -2493,6 +2657,30 @@ int _lttng_session_metadata_statedump(struct lttng_session *session)
 	if (ret)
 		goto end;
 
+	ret = print_metadata_escaped_field(session, "trace_name", session->name);
+	if (ret)
+		goto end;
+	ret = print_metadata_escaped_field(session, "trace_creation_datetime",
+			session->creation_time);
+	if (ret)
+		goto end;
+
+	/* Add the product UUID to the 'env' section */
+	product_uuid = dmi_get_system_info(DMI_PRODUCT_UUID);
+	if (product_uuid) {
+		ret = lttng_metadata_printf(session,
+				"	product_uuid = \"%s\";\n",
+				product_uuid
+				);
+		if (ret)
+			goto end;
+	}
+
+	/* Close the 'env' section */
+	ret = lttng_metadata_printf(session, "};\n\n");
+	if (ret)
+		goto end;
+
 	ret = lttng_metadata_printf(session,
 		"clock {\n"
 		"	name = \"%s\";\n",
@@ -2611,6 +2799,133 @@ void lttng_transport_unregister(struct lttng_transport *transport)
 }
 EXPORT_SYMBOL_GPL(lttng_transport_unregister);
 
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0))
+
+enum cpuhp_state lttng_hp_prepare;
+enum cpuhp_state lttng_hp_online;
+
+static int lttng_hotplug_prepare(unsigned int cpu, struct hlist_node *node)
+{
+	struct lttng_cpuhp_node *lttng_node;
+
+	lttng_node = container_of(node, struct lttng_cpuhp_node, node);
+	switch (lttng_node->component) {
+	case LTTNG_RING_BUFFER_FRONTEND:
+		return 0;
+	case LTTNG_RING_BUFFER_BACKEND:
+		return lttng_cpuhp_rb_backend_prepare(cpu, lttng_node);
+	case LTTNG_RING_BUFFER_ITER:
+		return 0;
+	case LTTNG_CONTEXT_PERF_COUNTERS:
+		return 0;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int lttng_hotplug_dead(unsigned int cpu, struct hlist_node *node)
+{
+	struct lttng_cpuhp_node *lttng_node;
+
+	lttng_node = container_of(node, struct lttng_cpuhp_node, node);
+	switch (lttng_node->component) {
+	case LTTNG_RING_BUFFER_FRONTEND:
+		return lttng_cpuhp_rb_frontend_dead(cpu, lttng_node);
+	case LTTNG_RING_BUFFER_BACKEND:
+		return 0;
+	case LTTNG_RING_BUFFER_ITER:
+		return 0;
+	case LTTNG_CONTEXT_PERF_COUNTERS:
+		return lttng_cpuhp_perf_counter_dead(cpu, lttng_node);
+	default:
+		return -EINVAL;
+	}
+}
+
+static int lttng_hotplug_online(unsigned int cpu, struct hlist_node *node)
+{
+	struct lttng_cpuhp_node *lttng_node;
+
+	lttng_node = container_of(node, struct lttng_cpuhp_node, node);
+	switch (lttng_node->component) {
+	case LTTNG_RING_BUFFER_FRONTEND:
+		return lttng_cpuhp_rb_frontend_online(cpu, lttng_node);
+	case LTTNG_RING_BUFFER_BACKEND:
+		return 0;
+	case LTTNG_RING_BUFFER_ITER:
+		return lttng_cpuhp_rb_iter_online(cpu, lttng_node);
+	case LTTNG_CONTEXT_PERF_COUNTERS:
+		return lttng_cpuhp_perf_counter_online(cpu, lttng_node);
+	default:
+		return -EINVAL;
+	}
+}
+
+static int lttng_hotplug_offline(unsigned int cpu, struct hlist_node *node)
+{
+	struct lttng_cpuhp_node *lttng_node;
+
+	lttng_node = container_of(node, struct lttng_cpuhp_node, node);
+	switch (lttng_node->component) {
+	case LTTNG_RING_BUFFER_FRONTEND:
+		return lttng_cpuhp_rb_frontend_offline(cpu, lttng_node);
+	case LTTNG_RING_BUFFER_BACKEND:
+		return 0;
+	case LTTNG_RING_BUFFER_ITER:
+		return 0;
+	case LTTNG_CONTEXT_PERF_COUNTERS:
+		return 0;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int __init lttng_init_cpu_hotplug(void)
+{
+	int ret;
+
+	ret = cpuhp_setup_state_multi(CPUHP_BP_PREPARE_DYN, "lttng:prepare",
+			lttng_hotplug_prepare,
+			lttng_hotplug_dead);
+	if (ret < 0) {
+		return ret;
+	}
+	lttng_hp_prepare = ret;
+	lttng_rb_set_hp_prepare(ret);
+
+	ret = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, "lttng:online",
+			lttng_hotplug_online,
+			lttng_hotplug_offline);
+	if (ret < 0) {
+		cpuhp_remove_multi_state(lttng_hp_prepare);
+		lttng_hp_prepare = 0;
+		return ret;
+	}
+	lttng_hp_online = ret;
+	lttng_rb_set_hp_online(ret);
+
+	return 0;
+}
+
+static void __exit lttng_exit_cpu_hotplug(void)
+{
+	lttng_rb_set_hp_online(0);
+	cpuhp_remove_multi_state(lttng_hp_online);
+	lttng_rb_set_hp_prepare(0);
+	cpuhp_remove_multi_state(lttng_hp_prepare);
+}
+
+#else /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)) */
+static int lttng_init_cpu_hotplug(void)
+{
+	return 0;
+}
+static void lttng_exit_cpu_hotplug(void)
+{
+}
+#endif /* #else #if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)) */
+
+
 static int __init lttng_events_init(void)
 {
 	int ret;
@@ -2644,8 +2959,29 @@ static int __init lttng_events_init(void)
 	ret = lttng_logger_init();
 	if (ret)
 		goto error_logger;
+	ret = lttng_init_cpu_hotplug();
+	if (ret)
+		goto error_hotplug;
+	printk(KERN_NOTICE "LTTng: Loaded modules v%s.%s.%s%s (%s)%s%s\n",
+		__stringify(LTTNG_MODULES_MAJOR_VERSION),
+		__stringify(LTTNG_MODULES_MINOR_VERSION),
+		__stringify(LTTNG_MODULES_PATCHLEVEL_VERSION),
+		LTTNG_MODULES_EXTRAVERSION,
+		LTTNG_VERSION_NAME,
+#ifdef LTTNG_EXTRA_VERSION_GIT
+		LTTNG_EXTRA_VERSION_GIT[0] == '\0' ? "" : " - " LTTNG_EXTRA_VERSION_GIT,
+#else
+		"",
+#endif
+#ifdef LTTNG_EXTRA_VERSION_NAME
+		LTTNG_EXTRA_VERSION_NAME[0] == '\0' ? "" : " - " LTTNG_EXTRA_VERSION_NAME);
+#else
+		"");
+#endif
 	return 0;
 
+error_hotplug:
+	lttng_logger_exit();
 error_logger:
 	lttng_abi_exit();
 error_abi:
@@ -2654,6 +2990,22 @@ error_kmem:
 	lttng_tracepoint_exit();
 error_tp:
 	lttng_context_exit();
+	printk(KERN_NOTICE "LTTng: Failed to load modules v%s.%s.%s%s (%s)%s%s\n",
+		__stringify(LTTNG_MODULES_MAJOR_VERSION),
+		__stringify(LTTNG_MODULES_MINOR_VERSION),
+		__stringify(LTTNG_MODULES_PATCHLEVEL_VERSION),
+		LTTNG_MODULES_EXTRAVERSION,
+		LTTNG_VERSION_NAME,
+#ifdef LTTNG_EXTRA_VERSION_GIT
+		LTTNG_EXTRA_VERSION_GIT[0] == '\0' ? "" : " - " LTTNG_EXTRA_VERSION_GIT,
+#else
+		"",
+#endif
+#ifdef LTTNG_EXTRA_VERSION_NAME
+		LTTNG_EXTRA_VERSION_NAME[0] == '\0' ? "" : " - " LTTNG_EXTRA_VERSION_NAME);
+#else
+		"");
+#endif
 	return ret;
 }
 
@@ -2663,6 +3015,7 @@ static void __exit lttng_events_exit(void)
 {
 	struct lttng_session *session, *tmpsession;
 
+	lttng_exit_cpu_hotplug();
 	lttng_logger_exit();
 	lttng_abi_exit();
 	list_for_each_entry_safe(session, tmpsession, &sessions, list)
@@ -2670,13 +3023,36 @@ static void __exit lttng_events_exit(void)
 	kmem_cache_destroy(event_cache);
 	lttng_tracepoint_exit();
 	lttng_context_exit();
+	printk(KERN_NOTICE "LTTng: Unloaded modules v%s.%s.%s%s (%s)%s%s\n",
+		__stringify(LTTNG_MODULES_MAJOR_VERSION),
+		__stringify(LTTNG_MODULES_MINOR_VERSION),
+		__stringify(LTTNG_MODULES_PATCHLEVEL_VERSION),
+		LTTNG_MODULES_EXTRAVERSION,
+		LTTNG_VERSION_NAME,
+#ifdef LTTNG_EXTRA_VERSION_GIT
+		LTTNG_EXTRA_VERSION_GIT[0] == '\0' ? "" : " - " LTTNG_EXTRA_VERSION_GIT,
+#else
+		"",
+#endif
+#ifdef LTTNG_EXTRA_VERSION_NAME
+		LTTNG_EXTRA_VERSION_NAME[0] == '\0' ? "" : " - " LTTNG_EXTRA_VERSION_NAME);
+#else
+		"");
+#endif
 }
 
 module_exit(lttng_events_exit);
 
+#include "extra_version/patches.i"
+#ifdef LTTNG_EXTRA_VERSION_GIT
+MODULE_INFO(extra_version_git, LTTNG_EXTRA_VERSION_GIT);
+#endif
+#ifdef LTTNG_EXTRA_VERSION_NAME
+MODULE_INFO(extra_version_name, LTTNG_EXTRA_VERSION_NAME);
+#endif
 MODULE_LICENSE("GPL and additional rights");
 MODULE_AUTHOR("Mathieu Desnoyers <mathieu.desnoyers@efficios.com>");
-MODULE_DESCRIPTION("LTTng Events");
+MODULE_DESCRIPTION("LTTng tracer");
 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
 	__stringify(LTTNG_MODULES_MINOR_VERSION) "."
 	__stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)