Merge tag 'kvmarm-fixes-for-5.3' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD

KVM/arm fixes for 5.3

- A bunch of switch/case fall-through annotation, fixing one actual bug
- Fix PMU reset bug
- Add missing exception class debug strings

90 files changed, 567 insertions, 330 deletions

diff --git a/tools/perf/util/Build b/tools/perf/util/Build
index d3408a463060..14f812bb07a7 100644
--- a/tools/perf/util/Build
+++ b/tools/perf/util/Build
@@ -20,11 +20,13 @@ perf-y += parse-events.o
 perf-y += perf_regs.o
 perf-y += path.o
 perf-y += print_binary.o
+perf-y += rlimit.o
 perf-y += argv_split.o
 perf-y += rbtree.o
 perf-y += libstring.o
 perf-y += bitmap.o
 perf-y += hweight.o
+perf-y += zalloc.o
 perf-y += smt.o
 perf-y += strbuf.o
 perf-y += string.o
@@ -241,3 +243,7 @@ $(OUTPUT)util/hweight.o: ../lib/hweight.c FORCE
 $(OUTPUT)util/vsprintf.o: ../lib/vsprintf.c FORCE
 	$(call rule_mkdir)
 	$(call if_changed_dep,cc_o_c)
+
+$(OUTPUT)util/zalloc.o: ../lib/zalloc.c FORCE
+	$(call rule_mkdir)
+	$(call if_changed_dep,cc_o_c)
diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index ec7aaf31c2b2..ac9ad2330f93 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -1119,16 +1119,14 @@ static int disasm_line__parse(char *line, const char **namep, char **rawp)
 	*namep = strdup(name);
 
 	if (*namep == NULL)
-		goto out_free_name;
+		goto out;
 
 	(*rawp)[0] = tmp;
 	*rawp = skip_spaces(*rawp);
 
 	return 0;
 
-out_free_name:
-	free((void *)namep);
-	*namep = NULL;
+out:
 	return -1;
 }
 
@@ -1237,8 +1235,7 @@ void disasm_line__free(struct disasm_line *dl)
 		dl->ins.ops->free(&dl->ops);
 	else
 		ins__delete(&dl->ops);
-	free((void *)dl->ins.name);
-	dl->ins.name = NULL;
+	zfree(&dl->ins.name);
 	annotation_line__delete(&dl->al);
 }
 
@@ -1589,7 +1586,7 @@ static void delete_last_nop(struct symbol *sym)
 				return;
 		}
 
-		list_del(&dl->al.node);
+		list_del_init(&dl->al.node);
 		disasm_line__free(dl);
 	}
 }
@@ -2466,7 +2463,7 @@ void annotated_source__purge(struct annotated_source *as)
 	struct annotation_line *al, *n;
 
 	list_for_each_entry_safe(al, n, &as->source, node) {
-		list_del(&al->node);
+		list_del_init(&al->node);
 		disasm_line__free(disasm_line(al));
 	}
 }
diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c
index 6067267cc76c..a314e5b26e9d 100644
--- a/tools/perf/util/arm-spe.c
+++ b/tools/perf/util/arm-spe.c
@@ -12,6 +12,7 @@
 #include <linux/types.h>
 #include <linux/bitops.h>
 #include <linux/log2.h>
+#include <linux/zalloc.h>
 
 #include "cpumap.h"
 #include "color.h"
@@ -19,7 +20,6 @@
 #include "evlist.h"
 #include "machine.h"
 #include "session.h"
-#include "util.h"
 #include "thread.h"
 #include "debug.h"
 #include "auxtrace.h"
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index bc215fe0b4b4..ec0af36697c4 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -24,9 +24,9 @@
 #include <stdlib.h>
 #include <stdio.h>
 #include <linux/list.h>
+#include <linux/zalloc.h>
 
 #include "../perf.h"
-#include "util.h"
 #include "evlist.h"
 #include "dso.h"
 #include "map.h"
@@ -408,7 +408,7 @@ void auxtrace_queues__free(struct auxtrace_queues *queues)
 
 			buffer = list_entry(queues->queue_array[i].head.next,
 					    struct auxtrace_buffer, list);
-			list_del(&buffer->list);
+			list_del_init(&buffer->list);
 			auxtrace_buffer__free(buffer);
 		}
 	}
@@ -612,7 +612,7 @@ void auxtrace_index__free(struct list_head *head)
 	struct auxtrace_index *auxtrace_index, *n;
 
 	list_for_each_entry_safe(auxtrace_index, n, head, list) {
-		list_del(&auxtrace_index->list);
+		list_del_init(&auxtrace_index->list);
 		free(auxtrace_index);
 	}
 }
@@ -1413,7 +1413,7 @@ void auxtrace_cache__free(struct auxtrace_cache *c)
 		return;
 
 	auxtrace_cache__drop(c);
-	free(c->hashtable);
+	zfree(&c->hashtable);
 	free(c);
 }
 
@@ -1459,12 +1459,11 @@ void *auxtrace_cache__lookup(struct auxtrace_cache *c, u32 key)
 
 static void addr_filter__free_str(struct addr_filter *filt)
 {
-	free(filt->str);
+	zfree(&filt->str);
 	filt->action   = NULL;
 	filt->sym_from = NULL;
 	filt->sym_to   = NULL;
 	filt->filename = NULL;
-	filt->str      = NULL;
 }
 
 static struct addr_filter *addr_filter__new(void)
diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index 251d9ea6252f..c61974a50aa5 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -12,6 +12,7 @@
 #include <linux/err.h>
 #include <linux/kernel.h>
 #include <linux/string.h>
+#include <linux/zalloc.h>
 #include <errno.h>
 #include "perf.h"
 #include "debug.h"
@@ -828,7 +829,7 @@ static void
 bpf_map_op__delete(struct bpf_map_op *op)
 {
 	if (!list_empty(&op->list))
-		list_del(&op->list);
+		list_del_init(&op->list);
 	if (op->key_type == BPF_MAP_KEY_RANGES)
 		parse_events__clear_array(&op->k.array);
 	free(op);
diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c
index 89c6913dfc25..f1abfab7aa8c 100644
--- a/tools/perf/util/build-id.c
+++ b/tools/perf/util/build-id.c
@@ -30,6 +30,7 @@
 #include "strlist.h"
 
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 
 static bool no_buildid_cache;
 
diff --git a/tools/perf/util/call-path.c b/tools/perf/util/call-path.c
index c5b90300304d..5c60b8be1cf6 100644
--- a/tools/perf/util/call-path.c
+++ b/tools/perf/util/call-path.c
@@ -6,8 +6,9 @@
 
 #include <linux/rbtree.h>
 #include <linux/list.h>
+#include <linux/zalloc.h>
+#include <stdlib.h>
 
-#include "util.h"
 #include "call-path.h"
 
 static void call_path__init(struct call_path *cp, struct call_path *parent,
@@ -39,7 +40,7 @@ void call_path_root__free(struct call_path_root *cpr)
 	struct call_path_block *pos, *n;
 
 	list_for_each_entry_safe(pos, n, &cpr->blocks, node) {
-		list_del(&pos->node);
+		list_del_init(&pos->node);
 		free(pos);
 	}
 	free(cpr);
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index abb608b09269..8d7d8f62fcca 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -16,11 +16,11 @@
 #include <stdbool.h>
 #include <errno.h>
 #include <math.h>
+#include <linux/zalloc.h>
 
 #include "asm/bug.h"
 
 #include "hist.h"
-#include "util.h"
 #include "sort.h"
 #include "machine.h"
 #include "map.h"
@@ -636,7 +636,7 @@ add_child(struct callchain_node *parent,
 		struct callchain_list *call, *tmp;
 
 		list_for_each_entry_safe(call, tmp, &new->val, list) {
-			list_del(&call->list);
+			list_del_init(&call->list);
 			map__zput(call->ms.map);
 			free(call);
 		}
@@ -1002,7 +1002,7 @@ merge_chain_branch(struct callchain_cursor *cursor,
 		callchain_cursor_append(cursor, list->ip,
 					list->ms.map, list->ms.sym,
 					false, NULL, 0, 0, 0, list->srcline);
-		list_del(&list->list);
+		list_del_init(&list->list);
 		map__zput(list->ms.map);
 		free(list);
 	}
@@ -1453,13 +1453,13 @@ static void free_callchain_node(struct callchain_node *node)
 	struct rb_node *n;
 
 	list_for_each_entry_safe(list, tmp, &node->parent_val, list) {
-		list_del(&list->list);
+		list_del_init(&list->list);
 		map__zput(list->ms.map);
 		free(list);
 	}
 
 	list_for_each_entry_safe(list, tmp, &node->val, list) {
-		list_del(&list->list);
+		list_del_init(&list->list);
 		map__zput(list->ms.map);
 		free(list);
 	}
@@ -1544,7 +1544,7 @@ int callchain_node__make_parent_list(struct callchain_node *node)
 
 out:
 	list_for_each_entry_safe(chain, new, &head, list) {
-		list_del(&chain->list);
+		list_del_init(&chain->list);
 		map__zput(chain->ms.map);
 		free(chain);
 	}
diff --git a/tools/perf/util/cgroup.c b/tools/perf/util/cgroup.c
index ccd02634a616..484c29830a81 100644
--- a/tools/perf/util/cgroup.c
+++ b/tools/perf/util/cgroup.c
@@ -1,11 +1,11 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "util.h"
 #include "../perf.h"
 #include <subcmd/parse-options.h>
 #include "evsel.h"
 #include "cgroup.h"
 #include "evlist.h"
 #include <linux/stringify.h>
+#include <linux/zalloc.h>
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <fcntl.h>
@@ -124,7 +124,7 @@ static struct cgroup *cgroup__new(const char *name)
 	return cgroup;
 
 out_free_name:
-	free(cgroup->name);
+	zfree(&cgroup->name);
 out_err:
 	free(cgroup);
 	return NULL;
diff --git a/tools/perf/util/comm.c b/tools/perf/util/comm.c
index 1066de92af12..afb8d4fd2644 100644
--- a/tools/perf/util/comm.c
+++ b/tools/perf/util/comm.c
@@ -1,12 +1,12 @@
 // SPDX-License-Identifier: GPL-2.0
 #include "comm.h"
-#include "util.h"
 #include <errno.h>
 #include <stdlib.h>
 #include <stdio.h>
 #include <string.h>
 #include <linux/refcount.h>
 #include <linux/rbtree.h>
+#include <linux/zalloc.h>
 #include "rwsem.h"
 
 struct comm_str {
diff --git a/tools/perf/util/config.c b/tools/perf/util/config.c
index 752cce853e51..042ffbc8c53f 100644
--- a/tools/perf/util/config.c
+++ b/tools/perf/util/config.c
@@ -11,7 +11,6 @@
  */
 #include <errno.h>
 #include <sys/param.h>
-#include "util.h"
 #include "cache.h"
 #include "callchain.h"
 #include <subcmd/exec-cmd.h>
@@ -23,7 +22,7 @@
 #include <sys/stat.h>
 #include <unistd.h>
 #include <linux/string.h>
-
+#include <linux/zalloc.h>
 #include <linux/ctype.h>
 
 #define MAXNAME (256)
diff --git a/tools/perf/util/counts.c b/tools/perf/util/counts.c
index 03032b410c29..88be9c4365e0 100644
--- a/tools/perf/util/counts.c
+++ b/tools/perf/util/counts.c
@@ -3,7 +3,7 @@
 #include <stdlib.h>
 #include "evsel.h"
 #include "counts.h"
-#include "util.h"
+#include <linux/zalloc.h>
 
 struct perf_counts *perf_counts__new(int ncpus, int nthreads)
 {
diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
index 0d8fbedf7bd5..3acfbe34ebaf 100644
--- a/tools/perf/util/cpumap.c
+++ b/tools/perf/util/cpumap.c
@@ -1,5 +1,4 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "util.h"
 #include <api/fs/fs.h>
 #include "../perf.h"
 #include "cpumap.h"
@@ -11,6 +10,7 @@
 #include "asm/bug.h"
 
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 
 static int max_cpu_num;
 static int max_present_cpu_num;
diff --git a/tools/perf/util/cputopo.c b/tools/perf/util/cputopo.c
index 26e73a4bd4fe..64336a280967 100644
--- a/tools/perf/util/cputopo.c
+++ b/tools/perf/util/cputopo.c
@@ -2,11 +2,12 @@
 #include <sys/param.h>
 #include <sys/utsname.h>
 #include <inttypes.h>
+#include <stdlib.h>
 #include <api/fs/fs.h>
+#include <linux/zalloc.h>
 
 #include "cputopo.h"
 #include "cpumap.h"
-#include "util.h"
 #include "env.h"
 
 #define CORE_SIB_FMT \
@@ -343,7 +344,7 @@ void numa_topology__delete(struct numa_topology *tp)
 	u32 i;
 
 	for (i = 0; i < tp->nr; i++)
-		free(tp->nodes[i].cpus);
+		zfree(&tp->nodes[i].cpus);
 
 	free(tp);
 }
diff --git a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
index bb45e23018ee..37d7c492b155 100644
--- a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
+++ b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
@@ -8,6 +8,7 @@
 
 #include <linux/err.h>
 #include <linux/list.h>
+#include <linux/zalloc.h>
 #include <stdlib.h>
 #include <opencsd/c_api/opencsd_c_api.h>
 #include <opencsd/etmv4/trc_pkt_types_etmv4.h>
diff --git a/tools/perf/util/cs-etm.c b/tools/perf/util/cs-etm.c
index 0c7776b51045..3d1c34fc4d68 100644
--- a/tools/perf/util/cs-etm.c
+++ b/tools/perf/util/cs-etm.c
@@ -11,6 +11,7 @@
 #include <linux/kernel.h>
 #include <linux/log2.h>
 #include <linux/types.h>
+#include <linux/zalloc.h>
 
 #include <opencsd/ocsd_if_types.h>
 #include <stdlib.h>
@@ -554,8 +555,7 @@ static void cs_etm__free_traceid_queues(struct cs_etm_queue *etmq)
 	etmq->traceid_queues_list = NULL;
 
 	/* finally free the traceid_queues array */
-	free(etmq->traceid_queues);
-	etmq->traceid_queues = NULL;
+	zfree(&etmq->traceid_queues);
 }
 
 static void cs_etm__free_queue(void *priv)
@@ -2460,7 +2460,7 @@ int cs_etm__process_auxtrace_info(union perf_event *event,
 
 		/* Something went wrong, no need to continue */
 		if (!inode) {
-			err = PTR_ERR(inode);
+			err = -ENOMEM;
 			goto err_free_metadata;
 		}
 
@@ -2517,8 +2517,10 @@ int cs_etm__process_auxtrace_info(union perf_event *event,
 	session->auxtrace = &etm->auxtrace;
 
 	etm->unknown_thread = thread__new(999999999, 999999999);
-	if (!etm->unknown_thread)
+	if (!etm->unknown_thread) {
+		err = -ENOMEM;
 		goto err_free_queues;
+	}
 
 	/*
 	 * Initialize list node so that at thread__zput() we can avoid
@@ -2530,15 +2532,17 @@ int cs_etm__process_auxtrace_info(union perf_event *event,
 	if (err)
 		goto err_delete_thread;
 
-	if (thread__init_map_groups(etm->unknown_thread, etm->machine))
+	if (thread__init_map_groups(etm->unknown_thread, etm->machine)) {
+		err = -ENOMEM;
 		goto err_delete_thread;
+	}
 
 	if (dump_trace) {
 		cs_etm__print_auxtrace_info(auxtrace_info->priv, num_cpu);
 		return 0;
 	}
 
-	if (session->itrace_synth_opts && session->itrace_synth_opts->set) {
+	if (session->itrace_synth_opts->set) {
 		etm->synth_opts = *session->itrace_synth_opts;
 	} else {
 		itrace_synth_opts__set_default(&etm->synth_opts,
@@ -2568,12 +2572,12 @@ err_free_etm:
 err_free_metadata:
 	/* No need to check @metadata[j], free(NULL) is supported */
 	for (j = 0; j < num_cpu; j++)
-		free(metadata[j]);
+		zfree(&metadata[j]);
 	zfree(&metadata);
 err_free_traceid_list:
 	intlist__delete(traceid_list);
 err_free_hdr:
 	zfree(&hdr);
 
-	return -EINVAL;
+	return err;
 }
diff --git a/tools/perf/util/data-convert-bt.c b/tools/perf/util/data-convert-bt.c
index 7b06e7373b9e..ddbcd59f2d9b 100644
--- a/tools/perf/util/data-convert-bt.c
+++ b/tools/perf/util/data-convert-bt.c
@@ -10,6 +10,7 @@
 #include <inttypes.h>
 #include <linux/compiler.h>
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 #include <babeltrace/ctf-writer/writer.h>
 #include <babeltrace/ctf-writer/clock.h>
 #include <babeltrace/ctf-writer/stream.h>
@@ -22,7 +23,6 @@
 #include "asm/bug.h"
 #include "data-convert-bt.h"
 #include "session.h"
-#include "util.h"
 #include "debug.h"
 #include "tool.h"
 #include "evlist.h"
@@ -1353,7 +1353,7 @@ static void free_streams(struct ctf_writer *cw)
 	for (cpu = 0; cpu < cw->stream_cnt; cpu++)
 		ctf_stream__delete(cw->stream[cpu]);
 
-	free(cw->stream);
+	zfree(&cw->stream);
 }
 
 static int ctf_writer__setup_env(struct ctf_writer *cw,
diff --git a/tools/perf/util/data.c b/tools/perf/util/data.c
index 6a64f713710d..1d1b97a92c3f 100644
--- a/tools/perf/util/data.c
+++ b/tools/perf/util/data.c
@@ -1,6 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <linux/compiler.h>
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <errno.h>
@@ -20,7 +21,7 @@ static void close_dir(struct perf_data_file *files, int nr)
 {
 	while (--nr >= 1) {
 		close(files[nr].fd);
-		free(files[nr].path);
+		zfree(&files[nr].path);
 	}
 	free(files);
 }
diff --git a/tools/perf/util/db-export.c b/tools/perf/util/db-export.c
index 2182f552aac6..ffbb3e7d3288 100644
--- a/tools/perf/util/db-export.c
+++ b/tools/perf/util/db-export.c
@@ -5,6 +5,7 @@
  */
 
 #include <errno.h>
+#include <stdlib.h>
 
 #include "evsel.h"
 #include "machine.h"
@@ -13,76 +14,20 @@
 #include "symbol.h"
 #include "map.h"
 #include "event.h"
-#include "util.h"
 #include "thread-stack.h"
 #include "callchain.h"
 #include "call-path.h"
 #include "db-export.h"
-
-struct deferred_export {
-	struct list_head node;
-	struct comm *comm;
-};
-
-static int db_export__deferred(struct db_export *dbe)
-{
-	struct deferred_export *de;
-	int err;
-
-	while (!list_empty(&dbe->deferred)) {
-		de = list_entry(dbe->deferred.next, struct deferred_export,
-				node);
-		err = dbe->export_comm(dbe, de->comm);
-		list_del(&de->node);
-		free(de);
-		if (err)
-			return err;
-	}
-
-	return 0;
-}
-
-static void db_export__free_deferred(struct db_export *dbe)
-{
-	struct deferred_export *de;
-
-	while (!list_empty(&dbe->deferred)) {
-		de = list_entry(dbe->deferred.next, struct deferred_export,
-				node);
-		list_del(&de->node);
-		free(de);
-	}
-}
-
-static int db_export__defer_comm(struct db_export *dbe, struct comm *comm)
-{
-	struct deferred_export *de;
-
-	de = zalloc(sizeof(struct deferred_export));
-	if (!de)
-		return -ENOMEM;
-
-	de->comm = comm;
-	list_add_tail(&de->node, &dbe->deferred);
-
-	return 0;
-}
+#include <linux/zalloc.h>
 
 int db_export__init(struct db_export *dbe)
 {
 	memset(dbe, 0, sizeof(struct db_export));
-	INIT_LIST_HEAD(&dbe->deferred);
 	return 0;
 }
 
-int db_export__flush(struct db_export *dbe)
-{
-	return db_export__deferred(dbe);
-}
-
 void db_export__exit(struct db_export *dbe)
 {
-	db_export__free_deferred(dbe);
 	call_return_processor__free(dbe->crp);
 	dbe->crp = NULL;
 }
@@ -114,71 +59,73 @@ int db_export__machine(struct db_export *dbe, struct machine *machine)
 }
 
 int db_export__thread(struct db_export *dbe, struct thread *thread,
-		      struct machine *machine, struct comm *comm)
+		      struct machine *machine, struct thread *main_thread)
 {
-	struct thread *main_thread;
 	u64 main_thread_db_id = 0;
-	int err;
 
 	if (thread->db_id)
 		return 0;
 
 	thread->db_id = ++dbe->thread_last_db_id;
 
-	if (thread->pid_ != -1) {
-		if (thread->pid_ == thread->tid) {
-			main_thread = thread;
-		} else {
-			main_thread = machine__findnew_thread(machine,
-							      thread->pid_,
-							      thread->pid_);
-			if (!main_thread)
-				return -ENOMEM;
-			err = db_export__thread(dbe, main_thread, machine,
-						comm);
-			if (err)
-				goto out_put;
-			if (comm) {
-				err = db_export__comm_thread(dbe, comm, thread);
-				if (err)
-					goto out_put;
-			}
-		}
+	if (main_thread)
 		main_thread_db_id = main_thread->db_id;
-		if (main_thread != thread)
-			thread__put(main_thread);
-	}
 
 	if (dbe->export_thread)
 		return dbe->export_thread(dbe, thread, main_thread_db_id,
 					  machine);
 
 	return 0;
+}
 
-out_put:
-	thread__put(main_thread);
-	return err;
+static int __db_export__comm(struct db_export *dbe, struct comm *comm,
+			     struct thread *thread)
+{
+	comm->db_id = ++dbe->comm_last_db_id;
+
+	if (dbe->export_comm)
+		return dbe->export_comm(dbe, comm, thread);
+
+	return 0;
 }
 
 int db_export__comm(struct db_export *dbe, struct comm *comm,
-		    struct thread *main_thread)
+		    struct thread *thread)
+{
+	if (comm->db_id)
+		return 0;
+
+	return __db_export__comm(dbe, comm, thread);
+}
+
+/*
+ * Export the "exec" comm. The "exec" comm is the program / application command
+ * name at the time it first executes. It is used to group threads for the same
+ * program. Note that the main thread pid (or thread group id tgid) cannot be
+ * used because it does not change when a new program is exec'ed.
+ */
+int db_export__exec_comm(struct db_export *dbe, struct comm *comm,
+			 struct thread *main_thread)
 {
 	int err;
 
 	if (comm->db_id)
 		return 0;
 
-	comm->db_id = ++dbe->comm_last_db_id;
-
-	if (dbe->export_comm) {
-		if (main_thread->comm_set)
-			err = dbe->export_comm(dbe, comm);
-		else
-			err = db_export__defer_comm(dbe, comm);
-		if (err)
-			return err;
-	}
+	err = __db_export__comm(dbe, comm, main_thread);
+	if (err)
+		return err;
 
+	/*
+	 * Record the main thread for this comm. Note that the main thread can
+	 * have many "exec" comms because there will be a new one every time it
+	 * exec's. An "exec" comm however will only ever have 1 main thread.
+	 * That is different to any other threads for that same program because
+	 * exec() will effectively kill them, so the relationship between the
+	 * "exec" comm and non-main threads is 1-to-1. That is why
+	 * db_export__comm_thread() is called here for the main thread, but it
+	 * is called for non-main threads when they are exported.
+	 */
 	return db_export__comm_thread(dbe, comm, main_thread);
 }
 
@@ -339,11 +286,65 @@ int db_export__branch_type(struct db_export *dbe, u32 branch_type,
 	return 0;
 }
 
+static int db_export__threads(struct db_export *dbe, struct thread *thread,
+			      struct thread *main_thread,
+			      struct machine *machine, struct comm **comm_ptr)
+{
+	struct comm *comm = NULL;
+	struct comm *curr_comm;
+	int err;
+
+	if (main_thread) {
+		/*
+		 * A thread has a reference to the main thread, so export the
+		 * main thread first.
+		 */
+		err = db_export__thread(dbe, main_thread, machine, main_thread);
+		if (err)
+			return err;
+		/*
+		 * Export comm before exporting the non-main thread because
+		 * db_export__comm_thread() can be called further below.
+		 */
+		comm = machine__thread_exec_comm(machine, main_thread);
+		if (comm) {
+			err = db_export__exec_comm(dbe, comm, main_thread);
+			if (err)
+				return err;
+			*comm_ptr = comm;
+		}
+	}
+
+	if (thread != main_thread) {
+		/*
+		 * For a non-main thread, db_export__comm_thread() must be
+		 * called only if thread has not previously been exported.
+		 */
+		bool export_comm_thread = comm && !thread->db_id;
+
+		err = db_export__thread(dbe, thread, machine, main_thread);
+		if (err)
+			return err;
+
+		if (export_comm_thread) {
+			err = db_export__comm_thread(dbe, comm, thread);
+			if (err)
+				return err;
+		}
+	}
+
+	curr_comm = thread__comm(thread);
+	if (curr_comm)
+		return db_export__comm(dbe, curr_comm, thread);
+
+	return 0;
+}
+
 int db_export__sample(struct db_export *dbe, union perf_event *event,
 		      struct perf_sample *sample, struct perf_evsel *evsel,
 		      struct addr_location *al)
 {
-	struct thread* thread = al->thread;
+	struct thread *thread = al->thread;
 	struct export_sample es = {
 		.event = event,
 		.sample = sample,
@@ -363,19 +364,13 @@ int db_export__sample(struct db_export *dbe, union perf_event *event,
 		return err;
 
 	main_thread = thread__main_thread(al->machine, thread);
-	if (main_thread)
-		comm = machine__thread_exec_comm(al->machine, main_thread);
 
-	err = db_export__thread(dbe, thread, al->machine, comm);
+	err = db_export__threads(dbe, thread, main_thread, al->machine, &comm);
 	if (err)
 		goto out_put;
 
-	if (comm) {
-		err = db_export__comm(dbe, comm, main_thread);
-		if (err)
-			goto out_put;
+	if (comm)
 		es.comm_db_id = comm->db_id;
-	}
 
 	es.db_id = ++dbe->sample_last_db_id;
 
@@ -524,3 +519,92 @@ int db_export__call_return(struct db_export *dbe, struct call_return *cr,
 
 	return 0;
 }
+
+static int db_export__pid_tid(struct db_export *dbe, struct machine *machine,
+			      pid_t pid, pid_t tid, u64 *db_id,
+			      struct comm **comm_ptr, bool *is_idle)
+{
+	struct thread *thread = machine__find_thread(machine, pid, tid);
+	struct thread *main_thread;
+	int err = 0;
+
+	if (!thread || !thread->comm_set)
+		goto out_put;
+
+	*is_idle = !thread->pid_ && !thread->tid;
+
+	main_thread = thread__main_thread(machine, thread);
+
+	err = db_export__threads(dbe, thread, main_thread, machine, comm_ptr);
+
+	*db_id = thread->db_id;
+
+	thread__put(main_thread);
+out_put:
+	thread__put(thread);
+
+	return err;
+}
+
+int db_export__switch(struct db_export *dbe, union perf_event *event,
+		      struct perf_sample *sample, struct machine *machine)
+{
+	bool out = event->header.misc & PERF_RECORD_MISC_SWITCH_OUT;
+	bool out_preempt = out &&
+		(event->header.misc & PERF_RECORD_MISC_SWITCH_OUT_PREEMPT);
+	int flags = out | (out_preempt << 1);
+	bool is_idle_a = false, is_idle_b = false;
+	u64 th_a_id = 0, th_b_id = 0;
+	u64 comm_out_id, comm_in_id;
+	struct comm *comm_a = NULL;
+	struct comm *comm_b = NULL;
+	u64 th_out_id, th_in_id;
+	u64 db_id;
+	int err;
+
+	err = db_export__machine(dbe, machine);
+	if (err)
+		return err;
+
+	err = db_export__pid_tid(dbe, machine, sample->pid, sample->tid,
+				 &th_a_id, &comm_a, &is_idle_a);
+	if (err)
+		return err;
+
+	if (event->header.type == PERF_RECORD_SWITCH_CPU_WIDE) {
+		pid_t pid = event->context_switch.next_prev_pid;
+		pid_t tid = event->context_switch.next_prev_tid;
+
+		err = db_export__pid_tid(dbe, machine, pid, tid, &th_b_id,
+					 &comm_b, &is_idle_b);
+		if (err)
+			return err;
+	}
+
+	/*
+	 * Do not export if both threads are unknown (i.e. not being traced),
+	 * or one is unknown and the other is the idle task.
+	 */
+	if ((!th_a_id || is_idle_a) && (!th_b_id || is_idle_b))
+		return 0;
+
+	db_id = ++dbe->context_switch_last_db_id;
+
+	if (out) {
+		th_out_id   = th_a_id;
+		th_in_id    = th_b_id;
+		comm_out_id = comm_a ? comm_a->db_id : 0;
+		comm_in_id  = comm_b ? comm_b->db_id : 0;
+	} else {
+		th_out_id   = th_b_id;
+		th_in_id    = th_a_id;
+		comm_out_id = comm_b ? comm_b->db_id : 0;
+		comm_in_id  = comm_a ? comm_a->db_id : 0;
+	}
+
+	if (dbe->export_context_switch)
+		return dbe->export_context_switch(dbe, db_id, machine, sample,
+						  th_out_id, comm_out_id,
+						  th_in_id, comm_in_id, flags);
+	return 0;
+}
diff --git a/tools/perf/util/db-export.h b/tools/perf/util/db-export.h
index e8a64028a386..ba1f62a5fe10 100644
--- a/tools/perf/util/db-export.h
+++ b/tools/perf/util/db-export.h
@@ -43,7 +43,8 @@ struct db_export {
 	int (*export_machine)(struct db_export *dbe, struct machine *machine);
 	int (*export_thread)(struct db_export *dbe, struct thread *thread,
 			     u64 main_thread_db_id, struct machine *machine);
-	int (*export_comm)(struct db_export *dbe, struct comm *comm);
+	int (*export_comm)(struct db_export *dbe, struct comm *comm,
+			   struct thread *thread);
 	int (*export_comm_thread)(struct db_export *dbe, u64 db_id,
 				  struct comm *comm, struct thread *thread);
 	int (*export_dso)(struct db_export *dbe, struct dso *dso,
@@ -56,6 +57,11 @@ struct db_export {
 	int (*export_call_path)(struct db_export *dbe, struct call_path *cp);
 	int (*export_call_return)(struct db_export *dbe,
 				  struct call_return *cr);
+	int (*export_context_switch)(struct db_export *dbe, u64 db_id,
+				     struct machine *machine,
+				     struct perf_sample *sample,
+				     u64 th_out_id, u64 comm_out_id,
+				     u64 th_in_id, u64 comm_in_id, int flags);
 	struct call_return_processor *crp;
 	struct call_path_root *cpr;
 	u64 evsel_last_db_id;
@@ -68,18 +74,19 @@ struct db_export {
 	u64 sample_last_db_id;
 	u64 call_path_last_db_id;
 	u64 call_return_last_db_id;
-	struct list_head deferred;
+	u64 context_switch_last_db_id;
 };
 
 int db_export__init(struct db_export *dbe);
-int db_export__flush(struct db_export *dbe);
 void db_export__exit(struct db_export *dbe);
 int db_export__evsel(struct db_export *dbe, struct perf_evsel *evsel);
 int db_export__machine(struct db_export *dbe, struct machine *machine);
 int db_export__thread(struct db_export *dbe, struct thread *thread,
-		      struct machine *machine, struct comm *comm);
+		      struct machine *machine, struct thread *main_thread);
 int db_export__comm(struct db_export *dbe, struct comm *comm,
-		    struct thread *main_thread);
+		    struct thread *thread);
+int db_export__exec_comm(struct db_export *dbe, struct comm *comm,
+			 struct thread *main_thread);
 int db_export__comm_thread(struct db_export *dbe, struct comm *comm,
 			   struct thread *thread);
 int db_export__dso(struct db_export *dbe, struct dso *dso,
@@ -97,5 +104,7 @@ int db_export__branch_types(struct db_export *dbe);
 int db_export__call_path(struct db_export *dbe, struct call_path *cp);
 int db_export__call_return(struct db_export *dbe, struct call_return *cr,
 			   u64 *parent_db_id);
+int db_export__switch(struct db_export *dbe, union perf_event *event,
+		      struct perf_sample *sample, struct machine *machine);
 
 #endif
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index 3cc578343f48..3780fe42453b 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -7,6 +7,7 @@
 #include <string.h>
 #include <stdarg.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <sys/wait.h>
 #include <api/debug.h>
 #include <linux/time64.h>
diff --git a/tools/perf/util/demangle-java.c b/tools/perf/util/demangle-java.c
index 5b4900d67c80..763328c151e9 100644
--- a/tools/perf/util/demangle-java.c
+++ b/tools/perf/util/demangle-java.c
@@ -1,14 +1,15 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <sys/types.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <string.h>
-#include "util.h"
 #include "debug.h"
 #include "symbol.h"
 
 #include "demangle-java.h"
 
 #include <linux/ctype.h>
+#include <linux/kernel.h>
 
 enum {
 	MODE_PREFIX = 0,
diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index c7fde04400f7..ebc9d46c15a7 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -2,6 +2,7 @@
 #include <asm/bug.h>
 #include <linux/kernel.h>
 #include <linux/string.h>
+#include <linux/zalloc.h>
 #include <sys/time.h>
 #include <sys/resource.h>
 #include <sys/types.h>
@@ -21,7 +22,7 @@
 #include "dso.h"
 #include "machine.h"
 #include "auxtrace.h"
-#include "util.h"
+#include "util.h" /* O_CLOEXEC for older systems */
 #include "debug.h"
 #include "string2.h"
 #include "vdso.h"
@@ -433,7 +434,7 @@ static void dso__list_add(struct dso *dso)
 
 static void dso__list_del(struct dso *dso)
 {
-	list_del(&dso->data.open_entry);
+	list_del_init(&dso->data.open_entry);
 	WARN_ONCE(dso__data_open_cnt <= 0,
 		  "DSO data fd counter out of bounds.");
 	dso__data_open_cnt--;
diff --git a/tools/perf/util/dwarf-aux.c b/tools/perf/util/dwarf-aux.c
index 218bfea8f8a8..03b2de1f5a35 100644
--- a/tools/perf/util/dwarf-aux.c
+++ b/tools/perf/util/dwarf-aux.c
@@ -6,7 +6,7 @@
 #include <errno.h>
 #include <inttypes.h>
 #include <stdbool.h>
-#include "util.h"
+#include <stdlib.h>
 #include "debug.h"
 #include "dwarf-aux.h"
 #include "string2.h"
diff --git a/tools/perf/util/env.c b/tools/perf/util/env.c
index 22eee8942527..9909ec40c6d2 100644
--- a/tools/perf/util/env.c
+++ b/tools/perf/util/env.c
@@ -2,11 +2,12 @@
 #include "cpumap.h"
 #include "env.h"
 #include <linux/ctype.h>
-#include "util.h"
+#include <linux/zalloc.h>
 #include "bpf-event.h"
 #include <errno.h>
 #include <sys/utsname.h>
 #include <bpf/libbpf.h>
+#include <stdlib.h>
 
 struct perf_env perf_env;
 
@@ -186,7 +187,7 @@ void perf_env__exit(struct perf_env *env)
 	zfree(&env->caches);
 
 	for (i = 0; i < env->nr_memory_nodes; i++)
-		free(env->memory_nodes[i].set);
+		zfree(&env->memory_nodes[i].set);
 	zfree(&env->memory_nodes);
 }
 
@@ -286,9 +287,9 @@ int perf_env__nr_cpus_avail(struct perf_env *env)
 
 void cpu_cache_level__free(struct cpu_cache_level *cache)
 {
-	free(cache->type);
-	free(cache->map);
-	free(cache->size);
+	zfree(&cache->type);
+	zfree(&cache->map);
+	zfree(&cache->size);
 }
 
 /*
diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
index e1d0c5ba1f92..f1f4848947ce 100644
--- a/tools/perf/util/event.c
+++ b/tools/perf/util/event.c
@@ -11,6 +11,7 @@
 #include <uapi/linux/mman.h> /* To get things like MAP_HUGETLB even on older libc headers */
 #include <api/fs/fs.h>
 #include <linux/perf_event.h>
+#include <linux/zalloc.h>
 #include "event.h"
 #include "debug.h"
 #include "hist.h"
@@ -855,7 +856,7 @@ free_threads:
 	free(synthesize_threads);
 free_dirent:
 	for (i = 0; i < n; i++)
-		free(dirent[i]);
+		zfree(&dirent[i]);
 	free(dirent);
 
 	return err;
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index a474ede17cd6..b0364d923f76 100644
--- a/tools/perf/util/evlist.c
+++ b/tools/perf/util/evlist.c
@@ -5,7 +5,6 @@
  * Parts came from builtin-{top,stat,record}.c, see those files for further
  * copyright notes.
  */
-#include "util.h"
 #include <api/fs/fs.h>
 #include <errno.h>
 #include <inttypes.h>
@@ -33,6 +32,7 @@
 #include <linux/hash.h>
 #include <linux/log2.h>
 #include <linux/err.h>
+#include <linux/zalloc.h>
 
 #ifdef LACKS_SIGQUEUE_PROTOTYPE
 int sigqueue(pid_t pid, int sig, const union sigval value);
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index 7fb4ae82f34c..ebb46da4dfe5 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -17,6 +17,7 @@
 #include <linux/perf_event.h>
 #include <linux/compiler.h>
 #include <linux/err.h>
+#include <linux/zalloc.h>
 #include <sys/ioctl.h>
 #include <sys/resource.h>
 #include <sys/types.h>
@@ -27,7 +28,6 @@
 #include "event.h"
 #include "evsel.h"
 #include "evlist.h"
-#include "util.h"
 #include "cpumap.h"
 #include "thread_map.h"
 #include "target.h"
@@ -1298,7 +1298,7 @@ static void perf_evsel__free_config_terms(struct perf_evsel *evsel)
 	struct perf_evsel_config_term *term, *h;
 
 	list_for_each_entry_safe(term, h, &evsel->config_terms, list) {
-		list_del(&term->list);
+		list_del_init(&term->list);
 		free(term);
 	}
 }
diff --git a/tools/perf/util/get_current_dir_name.c b/tools/perf/util/get_current_dir_name.c
index 267aa609a582..01f32f26552d 100644
--- a/tools/perf/util/get_current_dir_name.c
+++ b/tools/perf/util/get_current_dir_name.c
@@ -1,8 +1,8 @@
-// SPDX-License-Identifier: GPL-2.0
-// Copyright (C) 2018, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+// SPDX-License-Identifier: LGPL-2.1
+// Copyright (C) 2018, 2019 Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
 //
 #ifndef HAVE_GET_CURRENT_DIR_NAME
-#include "util.h"
+#include "get_current_dir_name.h"
 #include <unistd.h>
 #include <stdlib.h>
 #include <stdlib.h>
diff --git a/tools/perf/util/get_current_dir_name.h b/tools/perf/util/get_current_dir_name.h
new file mode 100644
index 000000000000..69f7d5537d32
--- /dev/null
+++ b/tools/perf/util/get_current_dir_name.h
@@ -0,0 +1,8 @@
+// SPDX-License-Identifier: LGPL-2.1
+// Copyright (C) 2018, 2019 Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+//
+#ifndef __PERF_GET_CURRENT_DIR_NAME_H
+#ifndef HAVE_GET_CURRENT_DIR_NAME
+char *get_current_dir_name(void);
+#endif // HAVE_GET_CURRENT_DIR_NAME
+#endif // __PERF_GET_CURRENT_DIR_NAME_H
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 6a93ff5d8db5..c24db7f4909c 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -1,7 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <errno.h>
 #include <inttypes.h>
-#include "util.h"
 #include "string2.h"
 #include <sys/param.h>
 #include <sys/types.h>
@@ -15,6 +14,7 @@
 #include <linux/bitops.h>
 #include <linux/string.h>
 #include <linux/stringify.h>
+#include <linux/zalloc.h>
 #include <sys/stat.h>
 #include <sys/utsname.h>
 #include <linux/time64.h>
@@ -1052,7 +1052,7 @@ static int cpu_cache_level__read(struct cpu_cache_level *cache, u32 cpu, u16 lev
 
 	scnprintf(file, PATH_MAX, "%s/size", path);
 	if (sysfs__read_str(file, &cache->size, &len)) {
-		free(cache->type);
+		zfree(&cache->type);
 		return -1;
 	}
 
@@ -1061,8 +1061,8 @@ static int cpu_cache_level__read(struct cpu_cache_level *cache, u32 cpu, u16 lev
 
 	scnprintf(file, PATH_MAX, "%s/shared_cpu_list", path);
 	if (sysfs__read_str(file, &cache->map, &len)) {
-		free(cache->map);
-		free(cache->type);
+		zfree(&cache->map);
+		zfree(&cache->type);
 		return -1;
 	}
 
diff --git a/tools/perf/util/help-unknown-cmd.c b/tools/perf/util/help-unknown-cmd.c
index 4f07a5ba5030..ab9e16123626 100644
--- a/tools/perf/util/help-unknown-cmd.c
+++ b/tools/perf/util/help-unknown-cmd.c
@@ -3,9 +3,11 @@
 #include "config.h"
 #include <poll.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <subcmd/help.h>
 #include "../builtin.h"
 #include "levenshtein.h"
+#include <linux/zalloc.h>
 
 static int autocorrect;
 
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
index 27cecb59f866..f24fd1954f6c 100644
--- a/tools/perf/util/hist.c
+++ b/tools/perf/util/hist.c
@@ -1,6 +1,5 @@
 // SPDX-License-Identifier: GPL-2.0
 #include "callchain.h"
-#include "util.h"
 #include "build-id.h"
 #include "hist.h"
 #include "map.h"
@@ -20,6 +19,7 @@
 #include <inttypes.h>
 #include <sys/param.h>
 #include <linux/time64.h>
+#include <linux/zalloc.h>
 
 static bool hists__filter_entry_by_dso(struct hists *hists,
 				       struct hist_entry *he);
@@ -472,16 +472,16 @@ static int hist_entry__init(struct hist_entry *he,
 	return 0;
 
 err_srcline:
-	free(he->srcline);
+	zfree(&he->srcline);
 
 err_rawdata:
-	free(he->raw_data);
+	zfree(&he->raw_data);
 
 err_infos:
 	if (he->branch_info) {
 		map__put(he->branch_info->from.map);
 		map__put(he->branch_info->to.map);
-		free(he->branch_info);
+		zfree(&he->branch_info);
 	}
 	if (he->mem_info) {
 		map__put(he->mem_info->iaddr.map);
@@ -489,7 +489,7 @@ err_infos:
 	}
 err:
 	map__zput(he->ms.map);
-	free(he->stat_acc);
+	zfree(&he->stat_acc);
 	return -ENOMEM;
 }
 
@@ -1254,10 +1254,10 @@ void hist_entry__delete(struct hist_entry *he)
 	zfree(&he->stat_acc);
 	free_srcline(he->srcline);
 	if (he->srcfile && he->srcfile[0])
-		free(he->srcfile);
+		zfree(&he->srcfile);
 	free_callchain(he->callchain);
-	free(he->trace_output);
-	free(he->raw_data);
+	zfree(&he->trace_output);
+	zfree(&he->raw_data);
 	ops->free(he);
 }
 
@@ -2741,10 +2741,10 @@ static void hists_evsel__exit(struct perf_evsel *evsel)
 
 	list_for_each_entry_safe(node, tmp, &hists->hpp_formats, list) {
 		perf_hpp_list__for_each_format_safe(&node->hpp, fmt, pos) {
-			list_del(&fmt->list);
+			list_del_init(&fmt->list);
 			free(fmt);
 		}
-		list_del(&node->list);
+		list_del_init(&node->list);
 		free(node);
 	}
 }
diff --git a/tools/perf/util/intel-bts.c b/tools/perf/util/intel-bts.c
index e32dbffebb2f..5560e95afdda 100644
--- a/tools/perf/util/intel-bts.c
+++ b/tools/perf/util/intel-bts.c
@@ -12,6 +12,7 @@
 #include <linux/types.h>
 #include <linux/bitops.h>
 #include <linux/log2.h>
+#include <linux/zalloc.h>
 
 #include "cpumap.h"
 #include "color.h"
@@ -21,7 +22,6 @@
 #include "map.h"
 #include "symbol.h"
 #include "session.h"
-#include "util.h"
 #include "thread.h"
 #include "thread-stack.h"
 #include "debug.h"
@@ -891,13 +891,12 @@ int intel_bts_process_auxtrace_info(union perf_event *event,
 	if (dump_trace)
 		return 0;
 
-	if (session->itrace_synth_opts && session->itrace_synth_opts->set) {
+	if (session->itrace_synth_opts->set) {
 		bts->synth_opts = *session->itrace_synth_opts;
 	} else {
 		itrace_synth_opts__set_default(&bts->synth_opts,
 				session->itrace_synth_opts->default_no_sample);
-		if (session->itrace_synth_opts)
-			bts->synth_opts.thread_stack =
+		bts->synth_opts.thread_stack =
 				session->itrace_synth_opts->thread_stack;
 	}
 
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 4d14e78c5927..3bfdf2b7a96a 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -14,9 +14,9 @@
 #include <stdint.h>
 #include <inttypes.h>
 #include <linux/compiler.h>
+#include <linux/zalloc.h>
 
 #include "../cache.h"
-#include "../util.h"
 #include "../auxtrace.h"
 
 #include "intel-pt-insn-decoder.h"
diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 470aaae9d930..df061599fef4 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -10,6 +10,7 @@
 #include <errno.h>
 #include <linux/kernel.h>
 #include <linux/types.h>
+#include <linux/zalloc.h>
 
 #include "../perf.h"
 #include "session.h"
@@ -22,7 +23,6 @@
 #include "evsel.h"
 #include "map.h"
 #include "color.h"
-#include "util.h"
 #include "thread.h"
 #include "thread-stack.h"
 #include "symbol.h"
@@ -3210,7 +3210,7 @@ int intel_pt_process_auxtrace_info(union perf_event *event,
 		goto err_delete_thread;
 	}
 
-	if (session->itrace_synth_opts && session->itrace_synth_opts->set) {
+	if (session->itrace_synth_opts->set) {
 		pt->synth_opts = *session->itrace_synth_opts;
 	} else {
 		itrace_synth_opts__set_default(&pt->synth_opts,
@@ -3220,8 +3220,7 @@ int intel_pt_process_auxtrace_info(union perf_event *event,
 			pt->synth_opts.branches = false;
 			pt->synth_opts.callchain = true;
 		}
-		if (session->itrace_synth_opts)
-			pt->synth_opts.thread_stack =
+		pt->synth_opts.thread_stack =
 				session->itrace_synth_opts->thread_stack;
 	}
 
@@ -3241,11 +3240,9 @@ int intel_pt_process_auxtrace_info(union perf_event *event,
 		pt->cbr2khz = tsc_freq / pt->max_non_turbo_ratio / 1000;
 	}
 
-	if (session->itrace_synth_opts) {
-		err = intel_pt_setup_time_ranges(pt, session->itrace_synth_opts);
-		if (err)
-			goto err_delete_thread;
-	}
+	err = intel_pt_setup_time_ranges(pt, session->itrace_synth_opts);
+	if (err)
+		goto err_delete_thread;
 
 	if (pt->synth_opts.calls)
 		pt->branches_filter |= PERF_IP_FLAG_CALL | PERF_IP_FLAG_ASYNC |
diff --git a/tools/perf/util/jitdump.c b/tools/perf/util/jitdump.c
index 28908afedec4..18c34f0c1966 100644
--- a/tools/perf/util/jitdump.c
+++ b/tools/perf/util/jitdump.c
@@ -29,6 +29,7 @@
 #include "../builtin.h"
 
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 
 struct jit_buf_desc {
 	struct perf_data *output;
@@ -431,14 +432,12 @@ static int jit_repipe_code_load(struct jit_buf_desc *jd, union jr_entry *jr)
 			   jd->unwinding_data, jd->eh_frame_hdr_size, jd->unwinding_size);
 
 	if (jd->debug_data && jd->nr_debug_entries) {
-		free(jd->debug_data);
-		jd->debug_data = NULL;
+		zfree(&jd->debug_data);
 		jd->nr_debug_entries = 0;
 	}
 
 	if (jd->unwinding_data && jd->eh_frame_hdr_size) {
-		free(jd->unwinding_data);
-		jd->unwinding_data = NULL;
+		zfree(&jd->unwinding_data);
 		jd->eh_frame_hdr_size = 0;
 		jd->unwinding_mapped_size = 0;
 		jd->unwinding_size = 0;
diff --git a/tools/perf/util/llvm-utils.c b/tools/perf/util/llvm-utils.c
index 5b0b60f00275..9f0470ecbca9 100644
--- a/tools/perf/util/llvm-utils.c
+++ b/tools/perf/util/llvm-utils.c
@@ -9,6 +9,7 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <linux/err.h>
+#include <linux/zalloc.h>
 #include "debug.h"
 #include "llvm-utils.h"
 #include "config.h"
@@ -352,8 +353,7 @@ void llvm__get_kbuild_opts(char **kbuild_dir, char **kbuild_include_opts)
 "     \toption in [llvm] to \"\" to suppress this detection.\n\n",
 			*kbuild_dir);
 
-		free(*kbuild_dir);
-		*kbuild_dir = NULL;
+		zfree(kbuild_dir);
 		goto errout;
 	}
 
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index 147ed85ea2bc..cf826eca3aaf 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -15,7 +15,6 @@
 #include "strlist.h"
 #include "thread.h"
 #include "vdso.h"
-#include "util.h"
 #include <stdbool.h>
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -28,6 +27,7 @@
 #include <linux/ctype.h>
 #include <symbol/kallsyms.h>
 #include <linux/mman.h>
+#include <linux/zalloc.h>
 
 static void __machine__remove_thread(struct machine *machine, struct thread *th, bool lock);
 
@@ -810,7 +810,7 @@ struct map *machine__findnew_module_map(struct machine *machine, u64 start,
 out:
 	/* put the dso here, corresponding to  machine__findnew_module_dso */
 	dso__put(dso);
-	free(m.name);
+	zfree(&m.name);
 	return map;
 }
 
@@ -1350,7 +1350,7 @@ static int map_groups__set_modules_path_dir(struct map_groups *mg,
 			if (m.kmod)
 				ret = map_groups__set_module_path(mg, path, &m);
 
-			free(m.name);
+			zfree(&m.name);
 
 			if (ret)
 				goto out;
diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index 6fce983c6115..668410b1d426 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -12,10 +12,10 @@
 #include "thread.h"
 #include "vdso.h"
 #include "build-id.h"
-#include "util.h"
 #include "debug.h"
 #include "machine.h"
 #include <linux/string.h>
+#include <linux/zalloc.h>
 #include "srcline.h"
 #include "namespaces.h"
 #include "unwind.h"
@@ -476,8 +476,11 @@ int map__fprintf_srccode(struct map *map, u64 addr,
 		goto out_free_line;
 
 	ret = fprintf(fp, "|%-8d %.*s", line, len, srccode);
-	state->srcfile = srcfile;
-	state->line = line;
+
+	if (state) {
+		state->srcfile = srcfile;
+		state->line = line;
+	}
 	return ret;
 
 out_free_line:
diff --git a/tools/perf/util/mem2node.c b/tools/perf/util/mem2node.c
index c6fd81c02586..cacc2fc4dcbd 100644
--- a/tools/perf/util/mem2node.c
+++ b/tools/perf/util/mem2node.c
@@ -1,8 +1,8 @@
 #include <errno.h>
 #include <inttypes.h>
 #include <linux/bitmap.h>
+#include <linux/zalloc.h>
 #include "mem2node.h"
-#include "util.h"
 
 struct phys_entry {
 	struct rb_node	rb_node;
diff --git a/tools/perf/util/metricgroup.c b/tools/perf/util/metricgroup.c
index d8164574cb16..416a9015405e 100644
--- a/tools/perf/util/metricgroup.c
+++ b/tools/perf/util/metricgroup.c
@@ -18,6 +18,7 @@
 #include "strlist.h"
 #include <assert.h>
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 
 struct metric_event *metricgroup__lookup(struct rblist *metric_events,
 					 struct perf_evsel *evsel,
@@ -235,7 +236,7 @@ static struct rb_node *mep_new(struct rblist *rl __maybe_unused,
 		goto out_name;
 	return &me->nd;
 out_name:
-	free((char *)me->name);
+	zfree(&me->name);
 out_me:
 	free(me);
 	return NULL;
@@ -263,7 +264,7 @@ static void mep_delete(struct rblist *rl __maybe_unused,
 	struct mep *me = container_of(nd, struct mep, nd);
 
 	strlist__delete(me->metrics);
-	free((void *)me->name);
+	zfree(&me->name);
 	free(me);
 }
 
@@ -489,8 +490,9 @@ static void metricgroup__free_egroups(struct list_head *group_list)
 
 	list_for_each_entry_safe (eg, egtmp, group_list, nd) {
 		for (i = 0; i < eg->idnum; i++)
-			free((char *)eg->ids[i]);
-		free(eg->ids);
+			zfree(&eg->ids[i]);
+		zfree(&eg->ids);
+		list_del_init(&eg->nd);
 		free(eg);
 	}
 }
diff --git a/tools/perf/util/mmap.c b/tools/perf/util/mmap.c
index 768c632b0d82..9f0b6391af33 100644
--- a/tools/perf/util/mmap.c
+++ b/tools/perf/util/mmap.c
@@ -9,6 +9,7 @@
 #include <sys/mman.h>
 #include <inttypes.h>
 #include <asm/bug.h>
+#include <linux/zalloc.h>
 #ifdef HAVE_LIBNUMA_SUPPORT
 #include <numaif.h>
 #endif
diff --git a/tools/perf/util/namespaces.c b/tools/perf/util/namespaces.c
index 023c4efd788d..46d3a7754897 100644
--- a/tools/perf/util/namespaces.c
+++ b/tools/perf/util/namespaces.c
@@ -5,8 +5,8 @@
  */
 
 #include "namespaces.h"
-#include "util.h"
 #include "event.h"
+#include "get_current_dir_name.h"
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <fcntl.h>
@@ -17,6 +17,7 @@
 #include <string.h>
 #include <unistd.h>
 #include <asm/bug.h>
+#include <linux/zalloc.h>
 
 struct namespaces *namespaces__new(struct namespaces_event *event)
 {
diff --git a/tools/perf/util/namespaces.h b/tools/perf/util/namespaces.h
index 15a5a276c478..004430c0de93 100644
--- a/tools/perf/util/namespaces.h
+++ b/tools/perf/util/namespaces.h
@@ -13,6 +13,10 @@
 #include <linux/refcount.h>
 #include <linux/types.h>
 
+#ifndef HAVE_SETNS_SUPPORT
+int setns(int fd, int nstype);
+#endif
+
 struct namespaces_event;
 
 struct namespaces {
diff --git a/tools/perf/util/ordered-events.c b/tools/perf/util/ordered-events.c
index 989fed6f43b5..bb5f34b7ab44 100644
--- a/tools/perf/util/ordered-events.c
+++ b/tools/perf/util/ordered-events.c
@@ -138,7 +138,7 @@ static struct ordered_event *alloc_event(struct ordered_events *oe,
 
 	if (!list_empty(cache)) {
 		new = list_entry(cache->next, struct ordered_event, list);
-		list_del(&new->list);
+		list_del_init(&new->list);
 	} else if (oe->buffer) {
 		new = &oe->buffer->event[oe->buffer_idx];
 		if (++oe->buffer_idx == MAX_SAMPLE_BUFFER)
@@ -394,13 +394,13 @@ void ordered_events__free(struct ordered_events *oe)
 	 * yet, we need to free only allocated ones ...
 	 */
 	if (oe->buffer) {
-		list_del(&oe->buffer->list);
+		list_del_init(&oe->buffer->list);
 		ordered_events_buffer__free(oe->buffer, oe->buffer_idx, oe);
 	}
 
 	/* ... and continue with the rest */
 	list_for_each_entry_safe(buffer, tmp, &oe->to_free, list) {
-		list_del(&buffer->list);
+		list_del_init(&buffer->list);
 		ordered_events_buffer__free(buffer, MAX_SAMPLE_BUFFER, oe);
 	}
 }
diff --git a/tools/perf/util/parse-branch-options.c b/tools/perf/util/parse-branch-options.c
index bd779d9f4d1e..726e8d9e8c54 100644
--- a/tools/perf/util/parse-branch-options.c
+++ b/tools/perf/util/parse-branch-options.c
@@ -1,9 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0
 #include "perf.h"
-#include "util/util.h"
 #include "util/debug.h"
 #include <subcmd/parse-options.h>
 #include "util/parse-branch-options.h"
+#include <stdlib.h>
 
 #define BRANCH_OPT(n, m) \
 	{ .name = n, .mode = (m) }
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index cf0b9b81c5aa..371ff3aee769 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -1,6 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <linux/hw_breakpoint.h>
 #include <linux/err.h>
+#include <linux/zalloc.h>
 #include <dirent.h>
 #include <errno.h>
 #include <sys/ioctl.h>
@@ -651,7 +652,7 @@ static int add_bpf_event(const char *group, const char *event, int fd,
 		pr_debug("Failed to add BPF event %s:%s\n",
 			 group, event);
 		list_for_each_entry_safe(evsel, tmp, &new_evsels, node) {
-			list_del(&evsel->node);
+			list_del_init(&evsel->node);
 			perf_evsel__delete(evsel);
 		}
 		return err;
diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
index 6ad8d4914969..f1c36ed1cf36 100644
--- a/tools/perf/util/parse-events.y
+++ b/tools/perf/util/parse-events.y
@@ -480,7 +480,6 @@ event_bpf_file:
 PE_BPF_OBJECT opt_event_config
 {
 	struct parse_events_state *parse_state = _parse_state;
-	struct parse_events_error *error = parse_state->error;
 	struct list_head *list;
 
 	ALLOC_LIST(list);
@@ -626,7 +625,6 @@ PE_TERM
 PE_NAME array '=' PE_NAME
 {
 	struct parse_events_term *term;
-	int i;
 
 	ABORT_ON(parse_events_term__str(&term, PARSE_EVENTS__TERM_TYPE_USER,
 					$1, $4, &@1, &@4));
diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c
index 08581e276225..ef46c2848808 100644
--- a/tools/perf/util/parse-regs-options.c
+++ b/tools/perf/util/parse-regs-options.c
@@ -1,8 +1,12 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "perf.h"
-#include "util/util.h"
+#include <stdbool.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <string.h>
+#include <stdio.h>
 #include "util/debug.h"
 #include <subcmd/parse-options.h>
+#include "util/perf_regs.h"
 #include "util/parse-regs-options.h"
 
 static int
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 55f4de6442e3..f32b710347db 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -2,6 +2,7 @@
 #include <linux/list.h>
 #include <linux/compiler.h>
 #include <linux/string.h>
+#include <linux/zalloc.h>
 #include <sys/types.h>
 #include <errno.h>
 #include <fcntl.h>
@@ -14,7 +15,6 @@
 #include <api/fs/fs.h>
 #include <locale.h>
 #include <regex.h>
-#include "util.h"
 #include "pmu.h"
 #include "parse-events.h"
 #include "cpumap.h"
@@ -1245,7 +1245,7 @@ int perf_pmu__check_alias(struct perf_pmu *pmu, struct list_head *head_terms,
 		info->metric_expr = alias->metric_expr;
 		info->metric_name = alias->metric_name;
 
-		list_del(&term->list);
+		list_del_init(&term->list);
 		free(term);
 	}
 
diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
index 6f24eaf6e504..cd1eb73cfe83 100644
--- a/tools/perf/util/probe-event.c
+++ b/tools/perf/util/probe-event.c
@@ -19,7 +19,6 @@
 #include <limits.h>
 #include <elf.h>
 
-#include "util.h"
 #include "event.h"
 #include "namespaces.h"
 #include "strlist.h"
@@ -40,6 +39,7 @@
 #include "string2.h"
 
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 
 #define PERFPROBE_GROUP "probe"
 
@@ -214,9 +214,9 @@ out:
 
 static void clear_perf_probe_point(struct perf_probe_point *pp)
 {
-	free(pp->file);
-	free(pp->function);
-	free(pp->lazy_line);
+	zfree(&pp->file);
+	zfree(&pp->function);
+	zfree(&pp->lazy_line);
 }
 
 static void clear_probe_trace_events(struct probe_trace_event *tevs, int ntevs)
@@ -1175,12 +1175,11 @@ int show_available_vars(struct perf_probe_event *pevs __maybe_unused,
 
 void line_range__clear(struct line_range *lr)
 {
-	free(lr->function);
-	free(lr->file);
-	free(lr->path);
-	free(lr->comp_dir);
+	zfree(&lr->function);
+	zfree(&lr->file);
+	zfree(&lr->path);
+	zfree(&lr->comp_dir);
 	intlist__delete(lr->line_list);
-	memset(lr, 0, sizeof(*lr));
 }
 
 int line_range__init(struct line_range *lr)
@@ -1563,6 +1562,17 @@ static int parse_perf_probe_arg(char *str, struct perf_probe_arg *arg)
 		str = tmp + 1;
 	}
 
+	tmp = strchr(str, '@');
+	if (tmp && tmp != str && strcmp(tmp + 1, "user")) { /* user attr */
+		if (!user_access_is_supported()) {
+			semantic_error("ftrace does not support user access\n");
+			return -EINVAL;
+		}
+		*tmp = '\0';
+		arg->user_access = true;
+		pr_debug("user_access ");
+	}
+
 	tmp = strchr(str, ':');
 	if (tmp) {	/* Type setting */
 		*tmp = '\0';
@@ -2203,15 +2213,15 @@ void clear_perf_probe_event(struct perf_probe_event *pev)
 	struct perf_probe_arg_field *field, *next;
 	int i;
 
-	free(pev->event);
-	free(pev->group);
-	free(pev->target);
+	zfree(&pev->event);
+	zfree(&pev->group);
+	zfree(&pev->target);
 	clear_perf_probe_point(&pev->point);
 
 	for (i = 0; i < pev->nargs; i++) {
-		free(pev->args[i].name);
-		free(pev->args[i].var);
-		free(pev->args[i].type);
+		zfree(&pev->args[i].name);
+		zfree(&pev->args[i].var);
+		zfree(&pev->args[i].type);
 		field = pev->args[i].field;
 		while (field) {
 			next = field->next;
@@ -2220,8 +2230,7 @@ void clear_perf_probe_event(struct perf_probe_event *pev)
 			field = next;
 		}
 	}
-	free(pev->args);
-	memset(pev, 0, sizeof(*pev));
+	zfree(&pev->args);
 }
 
 #define strdup_or_goto(str, label)	\
@@ -2302,15 +2311,15 @@ void clear_probe_trace_event(struct probe_trace_event *tev)
 	struct probe_trace_arg_ref *ref, *next;
 	int i;
 
-	free(tev->event);
-	free(tev->group);
-	free(tev->point.symbol);
-	free(tev->point.realname);
-	free(tev->point.module);
+	zfree(&tev->event);
+	zfree(&tev->group);
+	zfree(&tev->point.symbol);
+	zfree(&tev->point.realname);
+	zfree(&tev->point.module);
 	for (i = 0; i < tev->nargs; i++) {
-		free(tev->args[i].name);
-		free(tev->args[i].value);
-		free(tev->args[i].type);
+		zfree(&tev->args[i].name);
+		zfree(&tev->args[i].value);
+		zfree(&tev->args[i].type);
 		ref = tev->args[i].ref;
 		while (ref) {
 			next = ref->next;
@@ -2318,8 +2327,7 @@ void clear_probe_trace_event(struct probe_trace_event *tev)
 			ref = next;
 		}
 	}
-	free(tev->args);
-	memset(tev, 0, sizeof(*tev));
+	zfree(&tev->args);
 }
 
 struct kprobe_blacklist_node {
@@ -2336,8 +2344,8 @@ static void kprobe_blacklist__delete(struct list_head *blacklist)
 	while (!list_empty(blacklist)) {
 		node = list_first_entry(blacklist,
 					struct kprobe_blacklist_node, list);
-		list_del(&node->list);
-		free(node->symbol);
+		list_del_init(&node->list);
+		zfree(&node->symbol);
 		free(node);
 	}
 }
diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h
index 05c8d571a901..96a319cd2378 100644
--- a/tools/perf/util/probe-event.h
+++ b/tools/perf/util/probe-event.h
@@ -37,6 +37,7 @@ struct probe_trace_point {
 struct probe_trace_arg_ref {
 	struct probe_trace_arg_ref	*next;	/* Next reference */
 	long				offset;	/* Offset value */
+	bool				user_access;	/* User-memory access */
 };
 
 /* kprobe-tracer and uprobe-tracer tracing argument */
@@ -82,6 +83,7 @@ struct perf_probe_arg {
 	char				*var;	/* Variable name */
 	char				*type;	/* Type name */
 	struct perf_probe_arg_field	*field;	/* Structure fields */
+	bool				user_access;	/* User-memory access */
 };
 
 /* Perf probe probing event (point + arg) */
diff --git a/tools/perf/util/probe-file.c b/tools/perf/util/probe-file.c
index 0ed1900454eb..5b4d49382932 100644
--- a/tools/perf/util/probe-file.c
+++ b/tools/perf/util/probe-file.c
@@ -10,8 +10,8 @@
 #include <sys/types.h>
 #include <sys/uio.h>
 #include <unistd.h>
+#include <linux/zalloc.h>
 #include "namespaces.h"
-#include "util.h"
 #include "event.h"
 #include "strlist.h"
 #include "strfilter.h"
@@ -1005,6 +1005,7 @@ enum ftrace_readme {
 	FTRACE_README_PROBE_TYPE_X = 0,
 	FTRACE_README_KRETPROBE_OFFSET,
 	FTRACE_README_UPROBE_REF_CTR,
+	FTRACE_README_USER_ACCESS,
 	FTRACE_README_END,
 };
 
@@ -1017,6 +1018,7 @@ static struct {
 	DEFINE_TYPE(FTRACE_README_PROBE_TYPE_X, "*type: * x8/16/32/64,*"),
 	DEFINE_TYPE(FTRACE_README_KRETPROBE_OFFSET, "*place (kretprobe): *"),
 	DEFINE_TYPE(FTRACE_README_UPROBE_REF_CTR, "*ref_ctr_offset*"),
+	DEFINE_TYPE(FTRACE_README_USER_ACCESS, "*[u]<offset>*"),
 };
 
 static bool scan_ftrace_readme(enum ftrace_readme type)
@@ -1077,3 +1079,8 @@ bool uprobe_ref_ctr_is_supported(void)
 {
 	return scan_ftrace_readme(FTRACE_README_UPROBE_REF_CTR);
 }
+
+bool user_access_is_supported(void)
+{
+	return scan_ftrace_readme(FTRACE_README_USER_ACCESS);
+}
diff --git a/tools/perf/util/probe-file.h b/tools/perf/util/probe-file.h
index 2a249182f2a6..986c1c94f64f 100644
--- a/tools/perf/util/probe-file.h
+++ b/tools/perf/util/probe-file.h
@@ -70,6 +70,7 @@ int probe_cache__show_all_caches(struct strfilter *filter);
 bool probe_type_is_available(enum probe_type type);
 bool kretprobe_offset_is_supported(void);
 bool uprobe_ref_ctr_is_supported(void);
+bool user_access_is_supported(void);
 #else	/* ! HAVE_LIBELF_SUPPORT */
 static inline struct probe_cache *probe_cache__new(const char *tgt __maybe_unused, struct nsinfo *nsi __maybe_unused)
 {
diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
index 6b40cc691a2d..025fc4491993 100644
--- a/tools/perf/util/probe-finder.c
+++ b/tools/perf/util/probe-finder.c
@@ -19,11 +19,11 @@
 #include <dwarf-regs.h>
 
 #include <linux/bitops.h>
+#include <linux/zalloc.h>
 #include "event.h"
 #include "dso.h"
 #include "debug.h"
 #include "intlist.h"
-#include "util.h"
 #include "strlist.h"
 #include "symbol.h"
 #include "probe-finder.h"
@@ -280,7 +280,7 @@ static_var:
 
 static int convert_variable_type(Dwarf_Die *vr_die,
 				 struct probe_trace_arg *tvar,
-				 const char *cast)
+				 const char *cast, bool user_access)
 {
 	struct probe_trace_arg_ref **ref_ptr = &tvar->ref;
 	Dwarf_Die type;
@@ -320,7 +320,8 @@ static int convert_variable_type(Dwarf_Die *vr_die,
 	pr_debug("%s type is %s.\n",
 		 dwarf_diename(vr_die), dwarf_diename(&type));
 
-	if (cast && strcmp(cast, "string") == 0) {	/* String type */
+	if (cast && (!strcmp(cast, "string") || !strcmp(cast, "ustring"))) {
+		/* String type */
 		ret = dwarf_tag(&type);
 		if (ret != DW_TAG_pointer_type &&
 		    ret != DW_TAG_array_type) {
@@ -343,6 +344,7 @@ static int convert_variable_type(Dwarf_Die *vr_die,
 				pr_warning("Out of memory error\n");
 				return -ENOMEM;
 			}
+			(*ref_ptr)->user_access = user_access;
 		}
 		if (!die_compare_name(&type, "char") &&
 		    !die_compare_name(&type, "unsigned char")) {
@@ -397,7 +399,7 @@ formatted:
 static int convert_variable_fields(Dwarf_Die *vr_die, const char *varname,
 				    struct perf_probe_arg_field *field,
 				    struct probe_trace_arg_ref **ref_ptr,
-				    Dwarf_Die *die_mem)
+				    Dwarf_Die *die_mem, bool user_access)
 {
 	struct probe_trace_arg_ref *ref = *ref_ptr;
 	Dwarf_Die type;
@@ -434,6 +436,7 @@ static int convert_variable_fields(Dwarf_Die *vr_die, const char *varname,
 				*ref_ptr = ref;
 		}
 		ref->offset += dwarf_bytesize(&type) * field->index;
+		ref->user_access = user_access;
 		goto next;
 	} else if (tag == DW_TAG_pointer_type) {
 		/* Check the pointer and dereference */
@@ -505,17 +508,18 @@ static int convert_variable_fields(Dwarf_Die *vr_die, const char *varname,
 		}
 	}
 	ref->offset += (long)offs;
+	ref->user_access = user_access;
 
 	/* If this member is unnamed, we need to reuse this field */
 	if (!dwarf_diename(die_mem))
 		return convert_variable_fields(die_mem, varname, field,
-						&ref, die_mem);
+						&ref, die_mem, user_access);
 
 next:
 	/* Converting next field */
 	if (field->next)
 		return convert_variable_fields(die_mem, field->name,
-					field->next, &ref, die_mem);
+				field->next, &ref, die_mem, user_access);
 	else
 		return 0;
 }
@@ -541,11 +545,12 @@ static int convert_variable(Dwarf_Die *vr_die, struct probe_finder *pf)
 	else if (ret == 0 && pf->pvar->field) {
 		ret = convert_variable_fields(vr_die, pf->pvar->var,
 					      pf->pvar->field, &pf->tvar->ref,
-					      &die_mem);
+					      &die_mem, pf->pvar->user_access);
 		vr_die = &die_mem;
 	}
 	if (ret == 0)
-		ret = convert_variable_type(vr_die, pf->tvar, pf->pvar->type);
+		ret = convert_variable_type(vr_die, pf->tvar, pf->pvar->type,
+					    pf->pvar->user_access);
 	/* *expr will be cached in libdw. Don't free it. */
 	return ret;
 }
diff --git a/tools/perf/util/pstack.c b/tools/perf/util/pstack.c
index 797fe1ae2d2e..28de8a4c2ce8 100644
--- a/tools/perf/util/pstack.c
+++ b/tools/perf/util/pstack.c
@@ -5,10 +5,10 @@
  * (c) 2010 Arnaldo Carvalho de Melo <acme@redhat.com>
  */
 
-#include "util.h"
 #include "pstack.h"
 #include "debug.h"
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 #include <stdlib.h>
 
 struct pstack {
diff --git a/tools/perf/util/python-ext-sources b/tools/perf/util/python-ext-sources
index 2237bac9fadb..ceb8afdf9a89 100644
--- a/tools/perf/util/python-ext-sources
+++ b/tools/perf/util/python-ext-sources
@@ -18,6 +18,7 @@ util/namespaces.c
 ../lib/hweight.c
 ../lib/string.c
 ../lib/vsprintf.c
+../lib/zalloc.c
 util/thread_map.c
 util/util.c
 util/xyarray.c
diff --git a/tools/perf/util/rlimit.c b/tools/perf/util/rlimit.c
new file mode 100644
index 000000000000..13521d392a22
--- /dev/null
+++ b/tools/perf/util/rlimit.c
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: LGPL-2.1 */
+
+#include "util/debug.h"
+#include "util/rlimit.h"
+#include <sys/time.h>
+#include <sys/resource.h>
+
+/*
+ * Bump the memlock so that we can get bpf maps of a reasonable size,
+ * like the ones used with 'perf trace' and with 'perf test bpf',
+ * improve this to some specific request if needed.
+ */
+void rlimit__bump_memlock(void)
+{
+	struct rlimit rlim;
+
+	if (getrlimit(RLIMIT_MEMLOCK, &rlim) == 0) {
+		rlim.rlim_cur *= 4;
+		rlim.rlim_max *= 4;
+
+		if (setrlimit(RLIMIT_MEMLOCK, &rlim) < 0) {
+			rlim.rlim_cur /= 2;
+			rlim.rlim_max /= 2;
+
+			if (setrlimit(RLIMIT_MEMLOCK, &rlim) < 0)
+				pr_debug("Couldn't bump rlimit(MEMLOCK), failures may take place when creating BPF maps, etc\n");
+		}
+	}
+}
diff --git a/tools/perf/util/rlimit.h b/tools/perf/util/rlimit.h
new file mode 100644
index 000000000000..9f59d8e710a3
--- /dev/null
+++ b/tools/perf/util/rlimit.h
@@ -0,0 +1,6 @@
+#ifndef __PERF_RLIMIT_H_
+#define __PERF_RLIMIT_H_
+/* SPDX-License-Identifier: LGPL-2.1 */
+
+void rlimit__bump_memlock(void);
+#endif // __PERF_RLIMIT_H_
diff --git a/tools/perf/util/s390-cpumsf.c b/tools/perf/util/s390-cpumsf.c
index 10d36d9b7909..83d2e149ef19 100644
--- a/tools/perf/util/s390-cpumsf.c
+++ b/tools/perf/util/s390-cpumsf.c
@@ -146,6 +146,7 @@
 #include <linux/types.h>
 #include <linux/bitops.h>
 #include <linux/log2.h>
+#include <linux/zalloc.h>
 
 #include <sys/stat.h>
 #include <sys/types.h>
@@ -156,7 +157,6 @@
 #include "evlist.h"
 #include "machine.h"
 #include "session.h"
-#include "util.h"
 #include "thread.h"
 #include "debug.h"
 #include "auxtrace.h"
@@ -756,7 +756,7 @@ static int s390_cpumsf_run_decoder(struct s390_cpumsf_queue *sfq,
 	 */
 	if (err) {
 		sfq->buffer = NULL;
-		list_del(&buffer->list);
+		list_del_init(&buffer->list);
 		auxtrace_buffer__free(buffer);
 		if (err > 0)		/* Buffer done, no error */
 			err = 0;
@@ -1044,7 +1044,7 @@ static void s390_cpumsf_free(struct perf_session *session)
 	auxtrace_heap__free(&sf->heap);
 	s390_cpumsf_free_queues(session);
 	session->auxtrace = NULL;
-	free(sf->logdir);
+	zfree(&sf->logdir);
 	free(sf);
 }
 
@@ -1101,8 +1101,7 @@ static int s390_cpumsf__config(const char *var, const char *value, void *cb)
 	if (rc == -1 || !S_ISDIR(stbuf.st_mode)) {
 		pr_err("Missing auxtrace log directory %s,"
 		       " continue with current directory...\n", value);
-		free(sf->logdir);
-		sf->logdir = NULL;
+		zfree(&sf->logdir);
 	}
 	return 1;
 }
@@ -1162,7 +1161,7 @@ err_free_queues:
 	auxtrace_queues__free(&sf->queues);
 	session->auxtrace = NULL;
 err_free:
-	free(sf->logdir);
+	zfree(&sf->logdir);
 	free(sf);
 	return err;
 }
diff --git a/tools/perf/util/scripting-engines/trace-event-python.c b/tools/perf/util/scripting-engines/trace-event-python.c
index 112bed65232f..25dc1d765553 100644
--- a/tools/perf/util/scripting-engines/trace-event-python.c
+++ b/tools/perf/util/scripting-engines/trace-event-python.c
@@ -113,6 +113,7 @@ struct tables {
 	PyObject		*call_path_handler;
 	PyObject		*call_return_handler;
 	PyObject		*synth_handler;
+	PyObject		*context_switch_handler;
 	bool			db_export_mode;
 };
 
@@ -1011,15 +1012,19 @@ static int python_export_thread(struct db_export *dbe, struct thread *thread,
 	return 0;
 }
 
-static int python_export_comm(struct db_export *dbe, struct comm *comm)
+static int python_export_comm(struct db_export *dbe, struct comm *comm,
+			      struct thread *thread)
 {
 	struct tables *tables = container_of(dbe, struct tables, dbe);
 	PyObject *t;
 
-	t = tuple_new(2);
+	t = tuple_new(5);
 
 	tuple_set_u64(t, 0, comm->db_id);
 	tuple_set_string(t, 1, comm__str(comm));
+	tuple_set_u64(t, 2, thread->db_id);
+	tuple_set_u64(t, 3, comm->start);
+	tuple_set_s32(t, 4, comm->exec);
 
 	call_object(tables->comm_handler, t, "comm_table");
 
@@ -1233,6 +1238,34 @@ static int python_export_call_return(struct db_export *dbe,
 	return 0;
 }
 
+static int python_export_context_switch(struct db_export *dbe, u64 db_id,
+					struct machine *machine,
+					struct perf_sample *sample,
+					u64 th_out_id, u64 comm_out_id,
+					u64 th_in_id, u64 comm_in_id, int flags)
+{
+	struct tables *tables = container_of(dbe, struct tables, dbe);
+	PyObject *t;
+
+	t = tuple_new(9);
+
+	tuple_set_u64(t, 0, db_id);
+	tuple_set_u64(t, 1, machine->db_id);
+	tuple_set_u64(t, 2, sample->time);
+	tuple_set_s32(t, 3, sample->cpu);
+	tuple_set_u64(t, 4, th_out_id);
+	tuple_set_u64(t, 5, comm_out_id);
+	tuple_set_u64(t, 6, th_in_id);
+	tuple_set_u64(t, 7, comm_in_id);
+	tuple_set_s32(t, 8, flags);
+
+	call_object(tables->context_switch_handler, t, "context_switch");
+
+	Py_DECREF(t);
+
+	return 0;
+}
+
 static int python_process_call_return(struct call_return *cr, u64 *parent_db_id,
 				      void *data)
 {
@@ -1296,6 +1329,16 @@ static void python_process_event(union perf_event *event,
 	}
 }
 
+static void python_process_switch(union perf_event *event,
+				  struct perf_sample *sample,
+				  struct machine *machine)
+{
+	struct tables *tables = &tables_global;
+
+	if (tables->db_export_mode)
+		db_export__switch(&tables->dbe, event, sample, machine);
+}
+
 static void get_handler_name(char *str, size_t size,
 			     struct perf_evsel *evsel)
 {
@@ -1511,6 +1554,7 @@ static void set_table_handlers(struct tables *tables)
 	SET_TABLE_HANDLER(sample);
 	SET_TABLE_HANDLER(call_path);
 	SET_TABLE_HANDLER(call_return);
+	SET_TABLE_HANDLER(context_switch);
 
 	/*
 	 * Synthesized events are samples but with architecture-specific data
@@ -1620,9 +1664,7 @@ error:
 
 static int python_flush_script(void)
 {
-	struct tables *tables = &tables_global;
-
-	return db_export__flush(&tables->dbe);
+	return 0;
 }
 
 /*
@@ -1831,6 +1873,7 @@ struct scripting_ops python_scripting_ops = {
 	.flush_script		= python_flush_script,
 	.stop_script		= python_stop_script,
 	.process_event		= python_process_event,
+	.process_switch		= python_process_switch,
 	.process_stat		= python_process_stat,
 	.process_stat_interval	= python_process_stat_interval,
 	.generate_script	= python_generate_script,
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 54cf163347f7..d0fd6c614e68 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -2,6 +2,7 @@
 #include <errno.h>
 #include <inttypes.h>
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 #include <traceevent/event-parse.h>
 #include <api/fs/fs.h>
 
@@ -18,7 +19,6 @@
 #include "session.h"
 #include "tool.h"
 #include "sort.h"
-#include "util.h"
 #include "cpumap.h"
 #include "perf_regs.h"
 #include "asm/bug.h"
@@ -1246,9 +1246,12 @@ static void dump_read(struct perf_evsel *evsel, union perf_event *event)
 		return;
 
 	printf(": %d %d %s %" PRIu64 "\n", event->read.pid, event->read.tid,
-	       evsel ? perf_evsel__name(evsel) : "FAIL",
+	       perf_evsel__name(evsel),
 	       event->read.value);
 
+	if (!evsel)
+		return;
+
 	read_format = evsel->attr.read_format;
 
 	if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED)
diff --git a/tools/perf/util/setns.c b/tools/perf/util/setns.c
index ce8fc290fce8..48f9c0af63b2 100644
--- a/tools/perf/util/setns.c
+++ b/tools/perf/util/setns.c
@@ -1,4 +1,6 @@
-#include "util.h"
+// SPDX-License-Identifier: LGPL-2.1
+
+#include "namespaces.h"
 #include <unistd.h>
 #include <sys/syscall.h>
 
diff --git a/tools/perf/util/srccode.c b/tools/perf/util/srccode.c
index 684b155c222a..adfcf1ff464c 100644
--- a/tools/perf/util/srccode.c
+++ b/tools/perf/util/srccode.c
@@ -4,7 +4,8 @@
  * Copyright (c) 2017, Intel Corporation.
  * Author: Andi Kleen
  */
-#include "linux/list.h"
+#include <linux/list.h>
+#include <linux/zalloc.h>
 #include <stdlib.h>
 #include <sys/mman.h>
 #include <sys/stat.h>
@@ -82,12 +83,12 @@ static void fill_lines(char **lines, int maxline, char *map, int maplen)
 
 static void free_srcfile(struct srcfile *sf)
 {
-	list_del(&sf->nd);
+	list_del_init(&sf->nd);
 	hlist_del(&sf->hash_nd);
 	map_total_sz -= sf->maplen;
 	munmap(sf->map, sf->maplen);
-	free(sf->lines);
-	free(sf->fn);
+	zfree(&sf->lines);
+	zfree(&sf->fn);
 	free(sf);
 	num_srcfiles--;
 }
@@ -153,7 +154,7 @@ static struct srcfile *find_srcfile(char *fn)
 out_map:
 	munmap(h->map, sz);
 out_fn:
-	free(h->fn);
+	zfree(&h->fn);
 out_h:
 	free(h);
 	return NULL;
diff --git a/tools/perf/util/srcline.c b/tools/perf/util/srcline.c
index dcad75daf5e4..6ccf6f6d09df 100644
--- a/tools/perf/util/srcline.c
+++ b/tools/perf/util/srcline.c
@@ -6,9 +6,9 @@
 
 #include <linux/kernel.h>
 #include <linux/string.h>
+#include <linux/zalloc.h>
 
 #include "util/dso.h"
-#include "util/util.h"
 #include "util/debug.h"
 #include "util/callchain.h"
 #include "util/symbol_conf.h"
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
index cb891e5c2969..656065af4971 100644
--- a/tools/perf/util/stat-shadow.c
+++ b/tools/perf/util/stat-shadow.c
@@ -8,6 +8,7 @@
 #include "evlist.h"
 #include "expr.h"
 #include "metricgroup.h"
+#include <linux/zalloc.h>
 
 /*
  * AGGR_GLOBAL: Use CPU 0
@@ -775,7 +776,7 @@ static void generic_metric(struct perf_stat_config *config,
 		print_metric(config, ctxp, NULL, NULL, "", 0);
 
 	for (i = 1; i < pctx.num_ids; i++)
-		free((void *)pctx.ids[i].name);
+		zfree(&pctx.ids[i].name);
 }
 
 void perf_stat__print_shadow_stats(struct perf_stat_config *config,
diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
index d91fe754b6d2..db8a6cf336be 100644
--- a/tools/perf/util/stat.c
+++ b/tools/perf/util/stat.c
@@ -6,6 +6,7 @@
 #include "evlist.h"
 #include "evsel.h"
 #include "thread_map.h"
+#include <linux/zalloc.h>
 
 void update_stats(struct stats *stats, u64 val)
 {
@@ -132,7 +133,7 @@ static void perf_evsel__free_stat_priv(struct perf_evsel *evsel)
 	struct perf_stat_evsel *ps = evsel->stats;
 
 	if (ps)
-		free(ps->group_data);
+		zfree(&ps->group_data);
 	zfree(&evsel->stats);
 }
 
diff --git a/tools/perf/util/strbuf.c b/tools/perf/util/strbuf.c
index 23092fd6451d..2ce0dc887364 100644
--- a/tools/perf/util/strbuf.c
+++ b/tools/perf/util/strbuf.c
@@ -1,8 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0
 #include "debug.h"
-#include "util.h"
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 #include <errno.h>
+#include <stdlib.h>
 
 /*
  * Used as the default ->buf value, so that people can always assume
diff --git a/tools/perf/util/strfilter.c b/tools/perf/util/strfilter.c
index 90ea2b209cbb..78aa4c3b990d 100644
--- a/tools/perf/util/strfilter.c
+++ b/tools/perf/util/strfilter.c
@@ -1,11 +1,12 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "util.h"
 #include "string2.h"
 #include "strfilter.h"
 
 #include <errno.h>
+#include <stdlib.h>
 #include <linux/ctype.h>
 #include <linux/string.h>
+#include <linux/zalloc.h>
 
 /* Operators */
 static const char *OP_and	= "&";	/* Logical AND */
diff --git a/tools/perf/util/strlist.c b/tools/perf/util/strlist.c
index af45c6fd97db..8a868cbeffae 100644
--- a/tools/perf/util/strlist.c
+++ b/tools/perf/util/strlist.c
@@ -4,12 +4,12 @@
  */
 
 #include "strlist.h"
-#include "util.h"
 #include <errno.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <unistd.h>
+#include <linux/zalloc.h>
 
 static
 struct rb_node *strlist__node_new(struct rblist *rblist, const void *entry)
diff --git a/tools/perf/util/svghelper.c b/tools/perf/util/svghelper.c
index fab8a048d31b..76cc54000483 100644
--- a/tools/perf/util/svghelper.c
+++ b/tools/perf/util/svghelper.c
@@ -15,10 +15,10 @@
 #include <string.h>
 #include <linux/bitmap.h>
 #include <linux/time64.h>
+#include <linux/zalloc.h>
 
 #include "perf.h"
 #include "svghelper.h"
-#include "util.h"
 #include "cpumap.h"
 
 static u64 first_time, last_time;
diff --git a/tools/perf/util/symbol-elf.c b/tools/perf/util/symbol-elf.c
index 62008756d8cc..7d504dc22108 100644
--- a/tools/perf/util/symbol-elf.c
+++ b/tools/perf/util/symbol-elf.c
@@ -2,6 +2,7 @@
 #include <fcntl.h>
 #include <stdio.h>
 #include <errno.h>
+#include <stdlib.h>
 #include <string.h>
 #include <unistd.h>
 #include <inttypes.h>
@@ -16,6 +17,7 @@
 #include "debug.h"
 #include "util.h"
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 #include <symbol/kallsyms.h>
 
 #ifndef EM_AARCH64
@@ -1476,7 +1478,7 @@ static void kcore_copy__free_phdrs(struct kcore_copy_info *kci)
 	struct phdr_data *p, *tmp;
 
 	list_for_each_entry_safe(p, tmp, &kci->phdrs, node) {
-		list_del(&p->node);
+		list_del_init(&p->node);
 		free(p);
 	}
 }
@@ -1499,7 +1501,7 @@ static void kcore_copy__free_syms(struct kcore_copy_info *kci)
 	struct sym_data *s, *tmp;
 
 	list_for_each_entry_safe(s, tmp, &kci->syms, node) {
-		list_del(&s->node);
+		list_del_init(&s->node);
 		free(s);
 	}
 }
@@ -2131,11 +2133,11 @@ static int populate_sdt_note(Elf **elf, const char *data, size_t len,
 	return 0;
 
 out_free_args:
-	free(tmp->args);
+	zfree(&tmp->args);
 out_free_name:
-	free(tmp->name);
+	zfree(&tmp->name);
 out_free_prov:
-	free(tmp->provider);
+	zfree(&tmp->provider);
 out_free_note:
 	free(tmp);
 out_err:
@@ -2250,9 +2252,9 @@ int cleanup_sdt_note_list(struct list_head *sdt_notes)
 	int nr_free = 0;
 
 	list_for_each_entry_safe(pos, tmp, sdt_notes, note_list) {
-		list_del(&pos->note_list);
-		free(pos->name);
-		free(pos->provider);
+		list_del_init(&pos->note_list);
+		zfree(&pos->name);
+		zfree(&pos->provider);
 		free(pos);
 		nr_free++;
 	}
diff --git a/tools/perf/util/symbol-minimal.c b/tools/perf/util/symbol-minimal.c
index 17edbd4f6f85..3bc8b7e3300e 100644
--- a/tools/perf/util/symbol-minimal.c
+++ b/tools/perf/util/symbol-minimal.c
@@ -7,9 +7,10 @@
 #include <stdio.h>
 #include <fcntl.h>
 #include <string.h>
+#include <stdlib.h>
 #include <byteswap.h>
 #include <sys/stat.h>
-
+#include <linux/zalloc.h>
 
 static bool check_need_swap(int file_endian)
 {
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index ae2ce255e848..173f3378aaa0 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -26,6 +26,7 @@
 #include "header.h"
 #include "path.h"
 #include <linux/ctype.h>
+#include <linux/zalloc.h>
 
 #include <elf.h>
 #include <limits.h>
diff --git a/tools/perf/util/syscalltbl.c b/tools/perf/util/syscalltbl.c
index c2037ac533f3..022a9c670338 100644
--- a/tools/perf/util/syscalltbl.c
+++ b/tools/perf/util/syscalltbl.c
@@ -10,9 +10,9 @@
 #include <linux/compiler.h>
 
 #ifdef HAVE_SYSCALL_TABLE_SUPPORT
+#include <linux/zalloc.h>
 #include <string.h>
 #include "string2.h"
-#include "util.h"
 
 #if defined(__x86_64__)
 #include <asm/syscalls_64.c>
diff --git a/tools/perf/util/target.c b/tools/perf/util/target.c
index 3852d07c49bd..3adc65480349 100644
--- a/tools/perf/util/target.c
+++ b/tools/perf/util/target.c
@@ -10,9 +10,9 @@
 #include "debug.h"
 
 #include <pwd.h>
+#include <stdlib.h>
 #include <string.h>
 
-
 enum target_errno target__validate(struct target *target)
 {
 	enum target_errno ret = TARGET_ERRNO__SUCCESS;
diff --git a/tools/perf/util/thread-stack.c b/tools/perf/util/thread-stack.c
index 6ff1ff4d4ce7..15134ac9b8f1 100644
--- a/tools/perf/util/thread-stack.c
+++ b/tools/perf/util/thread-stack.c
@@ -7,12 +7,13 @@
 #include <linux/rbtree.h>
 #include <linux/list.h>
 #include <linux/log2.h>
+#include <linux/zalloc.h>
 #include <errno.h>
+#include <stdlib.h>
 #include "thread.h"
 #include "event.h"
 #include "machine.h"
 #include "env.h"
-#include "util.h"
 #include "debug.h"
 #include "symbol.h"
 #include "comm.h"
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
index 3e29a4e8b5e6..873ab505ca80 100644
--- a/tools/perf/util/thread.c
+++ b/tools/perf/util/thread.c
@@ -5,10 +5,10 @@
 #include <stdio.h>
 #include <string.h>
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 #include "session.h"
 #include "thread.h"
 #include "thread-stack.h"
-#include "util.h"
 #include "debug.h"
 #include "namespaces.h"
 #include "comm.h"
@@ -93,14 +93,14 @@ void thread__delete(struct thread *thread)
 	down_write(&thread->namespaces_lock);
 	list_for_each_entry_safe(namespaces, tmp_namespaces,
 				 &thread->namespaces_list, list) {
-		list_del(&namespaces->list);
+		list_del_init(&namespaces->list);
 		namespaces__free(namespaces);
 	}
 	up_write(&thread->namespaces_lock);
 
 	down_write(&thread->comm_lock);
 	list_for_each_entry_safe(comm, tmp_comm, &thread->comm_list, list) {
-		list_del(&comm->list);
+		list_del_init(&comm->list);
 		comm__free(comm);
 	}
 	up_write(&thread->comm_lock);
diff --git a/tools/perf/util/thread_map.c b/tools/perf/util/thread_map.c
index 281bf06f10f2..5b3511f2b6b1 100644
--- a/tools/perf/util/thread_map.c
+++ b/tools/perf/util/thread_map.c
@@ -13,9 +13,9 @@
 #include <string.h>
 #include <api/fs/fs.h>
 #include <linux/string.h>
+#include <linux/zalloc.h>
 #include "asm/bug.h"
 #include "thread_map.h"
-#include "util.h"
 #include "debug.h"
 #include "event.h"
 
@@ -480,7 +480,7 @@ int thread_map__remove(struct thread_map *threads, int idx)
 	/*
 	 * Free the 'idx' item and shift the rest up.
 	 */
-	free(threads->map[idx].comm);
+	zfree(&threads->map[idx].comm);
 
 	for (i = idx; i < threads->nr - 1; i++)
 		threads->map[i] = threads->map[i + 1];
diff --git a/tools/perf/util/trace-event-info.c b/tools/perf/util/trace-event-info.c
index 806a11b334d3..4550015b9d5d 100644
--- a/tools/perf/util/trace-event-info.c
+++ b/tools/perf/util/trace-event-info.c
@@ -18,6 +18,7 @@
 #include <stdbool.h>
 #include <linux/list.h>
 #include <linux/kernel.h>
+#include <linux/zalloc.h>
 
 #include "../perf.h"
 #include "trace-event.h"
diff --git a/tools/perf/util/trace-event-scripting.c b/tools/perf/util/trace-event-scripting.c
index b023db136ef3..ba58f69777a1 100644
--- a/tools/perf/util/trace-event-scripting.c
+++ b/tools/perf/util/trace-event-scripting.c
@@ -12,8 +12,8 @@
 
 #include "../perf.h"
 #include "debug.h"
-#include "util.h"
 #include "trace-event.h"
+#include <linux/zalloc.h>
 
 struct scripting_context *scripting_context;
 
diff --git a/tools/perf/util/trace-event.h b/tools/perf/util/trace-event.h
index d9b0a942090a..c7002fe11673 100644
--- a/tools/perf/util/trace-event.h
+++ b/tools/perf/util/trace-event.h
@@ -81,6 +81,9 @@ struct scripting_ops {
 			       struct perf_sample *sample,
 			       struct perf_evsel *evsel,
 			       struct addr_location *al);
+	void (*process_switch)(union perf_event *event,
+			       struct perf_sample *sample,
+			       struct machine *machine);
 	void (*process_stat)(struct perf_stat_config *config,
 			     struct perf_evsel *evsel, u64 tstamp);
 	void (*process_stat_interval)(u64 tstamp);
diff --git a/tools/perf/util/unwind-libdw.c b/tools/perf/util/unwind-libdw.c
index 407d0167b942..28f71ca6ce1c 100644
--- a/tools/perf/util/unwind-libdw.c
+++ b/tools/perf/util/unwind-libdw.c
@@ -12,6 +12,7 @@
 #include "symbol.h"
 #include "thread.h"
 #include <linux/types.h>
+#include <linux/zalloc.h>
 #include "event.h"
 #include "perf_regs.h"
 #include "callchain.h"
diff --git a/tools/perf/util/unwind-libunwind-local.c b/tools/perf/util/unwind-libunwind-local.c
index 25e1406b1f8b..71a788921b62 100644
--- a/tools/perf/util/unwind-libunwind-local.c
+++ b/tools/perf/util/unwind-libunwind-local.c
@@ -25,6 +25,7 @@
 #include <unistd.h>
 #include <sys/mman.h>
 #include <linux/list.h>
+#include <linux/zalloc.h>
 #ifndef REMOTE_UNWIND_LIBUNWIND
 #include <libunwind.h>
 #include <libunwind-ptrace.h>
@@ -345,7 +346,7 @@ static int read_unwind_spec_debug_frame(struct dso *dso,
 							__func__,
 							dso->symsrc_filename,
 							debuglink);
-					free(dso->symsrc_filename);
+					zfree(&dso->symsrc_filename);
 				}
 				dso->symsrc_filename = debuglink;
 			} else {
diff --git a/tools/perf/util/usage.c b/tools/perf/util/usage.c
index 070d25ceea6a..3949a60b00ae 100644
--- a/tools/perf/util/usage.c
+++ b/tools/perf/util/usage.c
@@ -9,6 +9,9 @@
  */
 #include "util.h"
 #include "debug.h"
+#include <stdio.h>
+#include <stdlib.h>
+#include <linux/compiler.h>
 
 static __noreturn void usage_builtin(const char *err)
 {
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 125e215dd3d8..dc7a469921e9 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -9,8 +9,6 @@
 #include <fcntl.h>
 #include <stdbool.h>
 #include <stddef.h>
-#include <stdlib.h>
-#include <stdarg.h>
 #include <linux/compiler.h>
 #include <sys/types.h>
 
@@ -18,13 +16,6 @@
 void usage(const char *err) __noreturn;
 void die(const char *err, ...) __noreturn __printf(1, 2);
 
-static inline void *zalloc(size_t size)
-{
-	return calloc(1, size);
-}
-
-#define zfree(ptr) ({ free(*ptr); *ptr = NULL; })
-
 struct dirent;
 struct nsinfo;
 struct strlist;
@@ -59,18 +50,10 @@ int fetch_kernel_version(unsigned int *puint,
 
 const char *perf_tip(const char *dirpath);
 
-#ifndef HAVE_GET_CURRENT_DIR_NAME
-char *get_current_dir_name(void);
-#endif
-
 #ifndef HAVE_SCHED_GETCPU_SUPPORT
 int sched_getcpu(void);
 #endif
 
-#ifndef HAVE_SETNS_SUPPORT
-int setns(int fd, int nstype);
-#endif
-
 extern bool perf_singlethreaded;
 
 void perf_set_singlethreaded(void);
diff --git a/tools/perf/util/values.c b/tools/perf/util/values.c
index 4b7a303e4ba8..c59154e2d124 100644
--- a/tools/perf/util/values.c
+++ b/tools/perf/util/values.c
@@ -3,8 +3,8 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <errno.h>
+#include <linux/zalloc.h>
 
-#include "util.h"
 #include "values.h"
 #include "debug.h"
 
diff --git a/tools/perf/util/vdso.c b/tools/perf/util/vdso.c
index 5031b7b22bbd..7f427bab6c12 100644
--- a/tools/perf/util/vdso.c
+++ b/tools/perf/util/vdso.c
@@ -16,6 +16,7 @@
 #include "machine.h"
 #include "thread.h"
 #include "linux/string.h"
+#include <linux/zalloc.h>
 #include "debug.h"
 
 /*
diff --git a/tools/perf/util/xyarray.c b/tools/perf/util/xyarray.c
index dc95154f5646..86889ebc3514 100644
--- a/tools/perf/util/xyarray.c
+++ b/tools/perf/util/xyarray.c
@@ -1,8 +1,8 @@
 // SPDX-License-Identifier: GPL-2.0
 #include "xyarray.h"
-#include "util.h"
 #include <stdlib.h>
 #include <string.h>
+#include <linux/zalloc.h>
 
 struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size)
 {