Merge branch 'for-5.17/core' into for-linus

- support for USI style pens (Tero Kristo, Mika Westerberg)
- quirk for devices that need inverted X/Y axes (Alistair Francis)
- small core code cleanups and deduplication (Benjamin Tissoires)

80 files changed, 3381 insertions, 1437 deletions

diff --git a/tools/perf/util/Build b/tools/perf/util/Build
index f2914d5bed6e..2e5bfbb69960 100644
--- a/tools/perf/util/Build
+++ b/tools/perf/util/Build
@@ -138,6 +138,7 @@ perf-y += expr.o
 perf-y += branch.o
 perf-y += mem2node.o
 perf-y += clockid.o
+perf-y += list_sort.o
 
 perf-$(CONFIG_LIBBPF) += bpf-loader.o
 perf-$(CONFIG_LIBBPF) += bpf_map.o
@@ -201,6 +202,7 @@ endif
 perf-y += perf-hooks.o
 
 perf-$(CONFIG_LIBBPF) += bpf-event.o
+perf-$(CONFIG_LIBBPF) += bpf-utils.o
 
 perf-$(CONFIG_CXX) += c++/
 
@@ -315,3 +317,7 @@ $(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/list_sort.o: ../lib/list_sort.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 0bae061b2d6d..01900689dc00 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -28,6 +28,7 @@
 #include "evsel.h"
 #include "evlist.h"
 #include "bpf-event.h"
+#include "bpf-utils.h"
 #include "block-range.h"
 #include "string2.h"
 #include "util/event.h"
@@ -151,6 +152,7 @@ static int arch__associate_ins_ops(struct arch* arch, const char *name, struct i
 #include "arch/mips/annotate/instructions.c"
 #include "arch/x86/annotate/instructions.c"
 #include "arch/powerpc/annotate/instructions.c"
+#include "arch/riscv64/annotate/instructions.c"
 #include "arch/s390/annotate/instructions.c"
 #include "arch/sparc/annotate/instructions.c"
 
@@ -183,7 +185,6 @@ static struct arch architectures[] = {
 		.init = x86__annotate_init,
 		.instructions = x86__instructions,
 		.nr_instructions = ARRAY_SIZE(x86__instructions),
-		.ins_is_fused = x86__ins_is_fused,
 		.objdump =  {
 			.comment_char = '#',
 		},
@@ -193,6 +194,10 @@ static struct arch architectures[] = {
 		.init = powerpc__annotate_init,
 	},
 	{
+		.name = "riscv64",
+		.init = riscv64__annotate_init,
+	},
+	{
 		.name = "s390",
 		.init = s390__annotate_init,
 		.objdump =  {
@@ -1250,6 +1255,17 @@ int disasm_line__scnprintf(struct disasm_line *dl, char *bf, size_t size, bool r
 	return ins__scnprintf(&dl->ins, bf, size, &dl->ops, max_ins_name);
 }
 
+void annotation__init(struct annotation *notes)
+{
+	pthread_mutex_init(&notes->lock, NULL);
+}
+
+void annotation__exit(struct annotation *notes)
+{
+	annotated_source__delete(notes->src);
+	pthread_mutex_destroy(&notes->lock);
+}
+
 static void annotation_line__add(struct annotation_line *al, struct list_head *head)
 {
 	list_add_tail(&al->node, head);
@@ -1700,12 +1716,12 @@ static int symbol__disassemble_bpf(struct symbol *sym,
 {
 	struct annotation *notes = symbol__annotation(sym);
 	struct annotation_options *opts = args->options;
-	struct bpf_prog_info_linear *info_linear;
 	struct bpf_prog_linfo *prog_linfo = NULL;
 	struct bpf_prog_info_node *info_node;
 	int len = sym->end - sym->start;
 	disassembler_ftype disassemble;
 	struct map *map = args->ms.map;
+	struct perf_bpil *info_linear;
 	struct disassemble_info info;
 	struct dso *dso = map->dso;
 	int pc = 0, count, sub_id;
@@ -3127,7 +3143,7 @@ int symbol__annotate2(struct map_symbol *ms, struct evsel *evsel,
 	notes->nr_events = nr_pcnt;
 
 	annotation__update_column_widths(notes);
-	sym->annotate2 = true;
+	sym->annotate2 = 1;
 
 	return 0;
 
diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h
index 3757416bcf46..986f2bbe4870 100644
--- a/tools/perf/util/annotate.h
+++ b/tools/perf/util/annotate.h
@@ -299,6 +299,9 @@ struct annotation {
 	struct annotated_source *src;
 };
 
+void annotation__init(struct annotation *notes);
+void annotation__exit(struct annotation *notes);
+
 static inline int annotation__cycles_width(struct annotation *notes)
 {
 	if (notes->have_cycles && notes->options->show_minmax_cycle)
diff --git a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c b/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
index 32fe41835fa6..3fc528c9270c 100644
--- a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
+++ b/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
@@ -151,6 +151,7 @@ static int arm_spe_read_record(struct arm_spe_decoder *decoder)
 	u64 payload, ip;
 
 	memset(&decoder->record, 0x0, sizeof(decoder->record));
+	decoder->record.context_id = (u64)-1;
 
 	while (1) {
 		err = arm_spe_get_next_packet(decoder);
@@ -180,6 +181,7 @@ static int arm_spe_read_record(struct arm_spe_decoder *decoder)
 		case ARM_SPE_COUNTER:
 			break;
 		case ARM_SPE_CONTEXT:
+			decoder->record.context_id = payload;
 			break;
 		case ARM_SPE_OP_TYPE:
 			if (idx == SPE_OP_PKT_HDR_CLASS_LD_ST_ATOMIC) {
diff --git a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h b/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
index 59bdb7309674..46a8556a9e95 100644
--- a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
+++ b/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
@@ -38,6 +38,7 @@ struct arm_spe_record {
 	u64 timestamp;
 	u64 virt_addr;
 	u64 phys_addr;
+	u64 context_id;
 };
 
 struct arm_spe_insn;
diff --git a/tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c b/tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c
index 2e5eff4f8f03..2f311189c6e8 100644
--- a/tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c
+++ b/tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c
@@ -13,7 +13,7 @@
 
 #include "arm-spe-pkt-decoder.h"
 
-#if __BYTE_ORDER == __BIG_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
 #define le16_to_cpu bswap_16
 #define le32_to_cpu bswap_32
 #define le64_to_cpu bswap_64
diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c
index 58b7069c5a5f..fccac06b573a 100644
--- a/tools/perf/util/arm-spe.c
+++ b/tools/perf/util/arm-spe.c
@@ -51,6 +51,7 @@ struct arm_spe {
 	u8				timeless_decoding;
 	u8				data_queued;
 
+	u64				sample_type;
 	u8				sample_flc;
 	u8				sample_llc;
 	u8				sample_tlb;
@@ -71,6 +72,7 @@ struct arm_spe {
 	u64				kernel_start;
 
 	unsigned long			num_events;
+	u8				use_ctx_pkt_for_pid;
 };
 
 struct arm_spe_queue {
@@ -100,7 +102,7 @@ static void arm_spe_dump(struct arm_spe *spe __maybe_unused,
 	const char *color = PERF_COLOR_BLUE;
 
 	color_fprintf(stdout, color,
-		      ". ... ARM SPE data: size %zu bytes\n",
+		      ". ... ARM SPE data: size %#zx bytes\n",
 		      len);
 
 	while (len) {
@@ -226,6 +228,44 @@ static inline u8 arm_spe_cpumode(struct arm_spe *spe, u64 ip)
 		PERF_RECORD_MISC_USER;
 }
 
+static void arm_spe_set_pid_tid_cpu(struct arm_spe *spe,
+				    struct auxtrace_queue *queue)
+{
+	struct arm_spe_queue *speq = queue->priv;
+	pid_t tid;
+
+	tid = machine__get_current_tid(spe->machine, speq->cpu);
+	if (tid != -1) {
+		speq->tid = tid;
+		thread__zput(speq->thread);
+	} else
+		speq->tid = queue->tid;
+
+	if ((!speq->thread) && (speq->tid != -1)) {
+		speq->thread = machine__find_thread(spe->machine, -1,
+						    speq->tid);
+	}
+
+	if (speq->thread) {
+		speq->pid = speq->thread->pid_;
+		if (queue->cpu == -1)
+			speq->cpu = speq->thread->cpu;
+	}
+}
+
+static int arm_spe_set_tid(struct arm_spe_queue *speq, pid_t tid)
+{
+	struct arm_spe *spe = speq->spe;
+	int err = machine__set_current_tid(spe->machine, speq->cpu, -1, tid);
+
+	if (err)
+		return err;
+
+	arm_spe_set_pid_tid_cpu(spe, &spe->queues.queue_array[speq->queue_nr]);
+
+	return 0;
+}
+
 static void arm_spe_prep_sample(struct arm_spe *spe,
 				struct arm_spe_queue *speq,
 				union perf_event *event,
@@ -248,6 +288,12 @@ static void arm_spe_prep_sample(struct arm_spe *spe,
 	event->sample.header.size = sizeof(struct perf_event_header);
 }
 
+static int arm_spe__inject_event(union perf_event *event, struct perf_sample *sample, u64 type)
+{
+	event->header.size = perf_event__sample_event_size(sample, type, 0);
+	return perf_event__synthesize_sample(event, type, 0, sample);
+}
+
 static inline int
 arm_spe_deliver_synth_event(struct arm_spe *spe,
 			    struct arm_spe_queue *speq __maybe_unused,
@@ -256,6 +302,12 @@ arm_spe_deliver_synth_event(struct arm_spe *spe,
 {
 	int ret;
 
+	if (spe->synth_opts.inject) {
+		ret = arm_spe__inject_event(event, sample, spe->sample_type);
+		if (ret)
+			return ret;
+	}
+
 	ret = perf_session__deliver_synth_event(spe->session, event, sample);
 	if (ret)
 		pr_err("ARM SPE: failed to deliver event, error %d\n", ret);
@@ -460,6 +512,19 @@ static int arm_spe_run_decoder(struct arm_spe_queue *speq, u64 *timestamp)
 		 * can correlate samples between Arm SPE trace data and other
 		 * perf events with correct time ordering.
 		 */
+
+		/*
+		 * Update pid/tid info.
+		 */
+		record = &speq->decoder->record;
+		if (!spe->timeless_decoding && record->context_id != (u64)-1) {
+			ret = arm_spe_set_tid(speq, record->context_id);
+			if (ret)
+				return ret;
+
+			spe->use_ctx_pkt_for_pid = true;
+		}
+
 		ret = arm_spe_sample(speq);
 		if (ret)
 			return ret;
@@ -586,31 +651,6 @@ static bool arm_spe__is_timeless_decoding(struct arm_spe *spe)
 	return timeless_decoding;
 }
 
-static void arm_spe_set_pid_tid_cpu(struct arm_spe *spe,
-				    struct auxtrace_queue *queue)
-{
-	struct arm_spe_queue *speq = queue->priv;
-	pid_t tid;
-
-	tid = machine__get_current_tid(spe->machine, speq->cpu);
-	if (tid != -1) {
-		speq->tid = tid;
-		thread__zput(speq->thread);
-	} else
-		speq->tid = queue->tid;
-
-	if ((!speq->thread) && (speq->tid != -1)) {
-		speq->thread = machine__find_thread(spe->machine, -1,
-						    speq->tid);
-	}
-
-	if (speq->thread) {
-		speq->pid = speq->thread->pid_;
-		if (queue->cpu == -1)
-			speq->cpu = speq->thread->cpu;
-	}
-}
-
 static int arm_spe_process_queues(struct arm_spe *spe, u64 timestamp)
 {
 	unsigned int queue_nr;
@@ -641,7 +681,12 @@ static int arm_spe_process_queues(struct arm_spe *spe, u64 timestamp)
 			ts = timestamp;
 		}
 
-		arm_spe_set_pid_tid_cpu(spe, queue);
+		/*
+		 * A previous context-switch event has set pid/tid in the machine's context, so
+		 * here we need to update the pid/tid in the thread and SPE queue.
+		 */
+		if (!spe->use_ctx_pkt_for_pid)
+			arm_spe_set_pid_tid_cpu(spe, queue);
 
 		ret = arm_spe_run_decoder(speq, &ts);
 		if (ret < 0) {
@@ -681,6 +726,25 @@ static int arm_spe_process_timeless_queues(struct arm_spe *spe, pid_t tid,
 	return 0;
 }
 
+static int arm_spe_context_switch(struct arm_spe *spe, union perf_event *event,
+				  struct perf_sample *sample)
+{
+	pid_t pid, tid;
+	int cpu;
+
+	if (!(event->header.misc & PERF_RECORD_MISC_SWITCH_OUT))
+		return 0;
+
+	pid = event->context_switch.next_prev_pid;
+	tid = event->context_switch.next_prev_tid;
+	cpu = sample->cpu;
+
+	if (tid == -1)
+		pr_warning("context_switch event has no tid\n");
+
+	return machine__set_current_tid(spe->machine, cpu, pid, tid);
+}
+
 static int arm_spe_process_event(struct perf_session *session,
 				 union perf_event *event,
 				 struct perf_sample *sample,
@@ -718,6 +782,13 @@ static int arm_spe_process_event(struct perf_session *session,
 		}
 	} else if (timestamp) {
 		err = arm_spe_process_queues(spe, timestamp);
+		if (err)
+			return err;
+
+		if (!spe->use_ctx_pkt_for_pid &&
+		    (event->header.type == PERF_RECORD_SWITCH_CPU_WIDE ||
+		    event->header.type == PERF_RECORD_SWITCH))
+			err = arm_spe_context_switch(spe, event, sample);
 	}
 
 	return err;
@@ -783,7 +854,15 @@ static int arm_spe_flush(struct perf_session *session __maybe_unused,
 		return arm_spe_process_timeless_queues(spe, -1,
 				MAX_TIMESTAMP - 1);
 
-	return arm_spe_process_queues(spe, MAX_TIMESTAMP);
+	ret = arm_spe_process_queues(spe, MAX_TIMESTAMP);
+	if (ret)
+		return ret;
+
+	if (!spe->use_ctx_pkt_for_pid)
+		ui__warning("Arm SPE CONTEXT packets not found in the traces.\n"
+			    "Matching of TIDs to SPE events could be inaccurate.\n");
+
+	return 0;
 }
 
 static void arm_spe_free_queue(void *priv)
@@ -920,6 +999,8 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session)
 	else
 		attr.sample_type |= PERF_SAMPLE_TIME;
 
+	spe->sample_type = attr.sample_type;
+
 	attr.exclude_user = evsel->core.attr.exclude_user;
 	attr.exclude_kernel = evsel->core.attr.exclude_kernel;
 	attr.exclude_hv = evsel->core.attr.exclude_hv;
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index 8d2865b9ade2..c679394b898d 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -1564,6 +1564,9 @@ int itrace_do_parse_synth_opts(struct itrace_synth_opts *synth_opts,
 		case 'q':
 			synth_opts->quick += 1;
 			break;
+		case 'A':
+			synth_opts->approx_ipc = true;
+			break;
 		case 'Z':
 			synth_opts->timeless_decoding = true;
 			break;
diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
index 5f383908ca6e..bbf0d78c6401 100644
--- a/tools/perf/util/auxtrace.h
+++ b/tools/perf/util/auxtrace.h
@@ -59,6 +59,7 @@ enum itrace_period_type {
 #define AUXTRACE_ERR_FLG_DATA_LOST	(1 << ('l' - 'a'))
 
 #define AUXTRACE_LOG_FLG_ALL_PERF_EVTS	(1 << ('a' - 'a'))
+#define AUXTRACE_LOG_FLG_USE_STDOUT	(1 << ('o' - 'a'))
 
 /**
  * struct itrace_synth_opts - AUX area tracing synthesis options.
@@ -84,6 +85,7 @@ enum itrace_period_type {
  * @thread_stack: feed branches to the thread_stack
  * @last_branch: add branch context to 'instruction' events
  * @add_last_branch: add branch context to existing event records
+ * @approx_ipc: approximate IPC
  * @flc: whether to synthesize first level cache events
  * @llc: whether to synthesize last level cache events
  * @tlb: whether to synthesize TLB events
@@ -127,6 +129,7 @@ struct itrace_synth_opts {
 	bool			thread_stack;
 	bool			last_branch;
 	bool			add_last_branch;
+	bool			approx_ipc;
 	bool			flc;
 	bool			llc;
 	bool			tlb;
@@ -639,6 +642,7 @@ bool auxtrace__evsel_is_auxtrace(struct perf_session *session,
 "				d[flags]:		create a debug log\n" \
 "							each flag must be preceded by + or -\n" \
 "							log flags are: a (all perf events)\n" \
+"							               o (output to stdout)\n" \
 "				f:	    		synthesize first level cache events\n" \
 "				m:	    		synthesize last level cache events\n" \
 "				t:	    		synthesize TLB events\n" \
@@ -649,6 +653,8 @@ bool auxtrace__evsel_is_auxtrace(struct perf_session *session,
 "				L[len]:			synthesize last branch entries on existing event records\n" \
 "				sNUMBER:    		skip initial number of events\n"		\
 "				q:			quicker (less detailed) decoding\n" \
+"				A:			approximate IPC\n" \
+"				Z:			prefer to ignore timestamps (so-called \"timeless\" decoding)\n" \
 "				PERIOD[ns|us|ms|i|t]:   specify period to sample stream\n" \
 "				concatenate multiple options. Default is ibxwpe or cewp\n"
 
diff --git a/tools/perf/util/bpf-event.c b/tools/perf/util/bpf-event.c
index 388847bab6d9..a517eaa51eb3 100644
--- a/tools/perf/util/bpf-event.c
+++ b/tools/perf/util/bpf-event.c
@@ -10,6 +10,7 @@
 #include <internal/lib.h>
 #include <symbol/kallsyms.h>
 #include "bpf-event.h"
+#include "bpf-utils.h"
 #include "debug.h"
 #include "dso.h"
 #include "symbol.h"
@@ -32,7 +33,32 @@ struct btf * __weak btf__load_from_kernel_by_id(__u32 id)
        return err ? ERR_PTR(err) : btf;
 }
 
-#define ptr_to_u64(ptr)    ((__u64)(unsigned long)(ptr))
+struct bpf_program * __weak
+bpf_object__next_program(const struct bpf_object *obj, struct bpf_program *prev)
+{
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+	return bpf_program__next(prev, obj);
+#pragma GCC diagnostic pop
+}
+
+struct bpf_map * __weak
+bpf_object__next_map(const struct bpf_object *obj, const struct bpf_map *prev)
+{
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+	return bpf_map__next(prev, obj);
+#pragma GCC diagnostic pop
+}
+
+const void * __weak
+btf__raw_data(const struct btf *btf_ro, __u32 *size)
+{
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+	return btf__get_raw_data(btf_ro, size);
+#pragma GCC diagnostic pop
+}
 
 static int snprintf_hex(char *buf, size_t size, unsigned char *data, size_t len)
 {
@@ -48,9 +74,9 @@ static int machine__process_bpf_event_load(struct machine *machine,
 					   union perf_event *event,
 					   struct perf_sample *sample __maybe_unused)
 {
-	struct bpf_prog_info_linear *info_linear;
 	struct bpf_prog_info_node *info_node;
 	struct perf_env *env = machine->env;
+	struct perf_bpil *info_linear;
 	int id = event->bpf.id;
 	unsigned int i;
 
@@ -120,7 +146,11 @@ static int perf_env__fetch_btf(struct perf_env *env,
 	node->data_size = data_size;
 	memcpy(node->data, data, data_size);
 
-	perf_env__insert_btf(env, node);
+	if (!perf_env__insert_btf(env, node)) {
+		/* Insertion failed because of a duplicate. */
+		free(node);
+		return -1;
+	}
 	return 0;
 }
 
@@ -175,9 +205,9 @@ static int perf_event__synthesize_one_bpf_prog(struct perf_session *session,
 {
 	struct perf_record_ksymbol *ksymbol_event = &event->ksymbol;
 	struct perf_record_bpf_event *bpf_event = &event->bpf;
-	struct bpf_prog_info_linear *info_linear;
 	struct perf_tool *tool = session->tool;
 	struct bpf_prog_info_node *info_node;
+	struct perf_bpil *info_linear;
 	struct bpf_prog_info *info;
 	struct btf *btf = NULL;
 	struct perf_env *env;
@@ -191,15 +221,15 @@ static int perf_event__synthesize_one_bpf_prog(struct perf_session *session,
 	 */
 	env = session->data ? &session->header.env : &perf_env;
 
-	arrays = 1UL << BPF_PROG_INFO_JITED_KSYMS;
-	arrays |= 1UL << BPF_PROG_INFO_JITED_FUNC_LENS;
-	arrays |= 1UL << BPF_PROG_INFO_FUNC_INFO;
-	arrays |= 1UL << BPF_PROG_INFO_PROG_TAGS;
-	arrays |= 1UL << BPF_PROG_INFO_JITED_INSNS;
-	arrays |= 1UL << BPF_PROG_INFO_LINE_INFO;
-	arrays |= 1UL << BPF_PROG_INFO_JITED_LINE_INFO;
+	arrays = 1UL << PERF_BPIL_JITED_KSYMS;
+	arrays |= 1UL << PERF_BPIL_JITED_FUNC_LENS;
+	arrays |= 1UL << PERF_BPIL_FUNC_INFO;
+	arrays |= 1UL << PERF_BPIL_PROG_TAGS;
+	arrays |= 1UL << PERF_BPIL_JITED_INSNS;
+	arrays |= 1UL << PERF_BPIL_LINE_INFO;
+	arrays |= 1UL << PERF_BPIL_JITED_LINE_INFO;
 
-	info_linear = bpf_program__get_prog_info_linear(fd, arrays);
+	info_linear = get_bpf_prog_info_linear(fd, arrays);
 	if (IS_ERR_OR_NULL(info_linear)) {
 		info_linear = NULL;
 		pr_debug("%s: failed to get BPF program info. aborting\n", __func__);
@@ -452,8 +482,8 @@ int perf_event__synthesize_bpf_events(struct perf_session *session,
 
 static void perf_env__add_bpf_info(struct perf_env *env, u32 id)
 {
-	struct bpf_prog_info_linear *info_linear;
 	struct bpf_prog_info_node *info_node;
+	struct perf_bpil *info_linear;
 	struct btf *btf = NULL;
 	u64 arrays;
 	u32 btf_id;
@@ -463,15 +493,15 @@ static void perf_env__add_bpf_info(struct perf_env *env, u32 id)
 	if (fd < 0)
 		return;
 
-	arrays = 1UL << BPF_PROG_INFO_JITED_KSYMS;
-	arrays |= 1UL << BPF_PROG_INFO_JITED_FUNC_LENS;
-	arrays |= 1UL << BPF_PROG_INFO_FUNC_INFO;
-	arrays |= 1UL << BPF_PROG_INFO_PROG_TAGS;
-	arrays |= 1UL << BPF_PROG_INFO_JITED_INSNS;
-	arrays |= 1UL << BPF_PROG_INFO_LINE_INFO;
-	arrays |= 1UL << BPF_PROG_INFO_JITED_LINE_INFO;
+	arrays = 1UL << PERF_BPIL_JITED_KSYMS;
+	arrays |= 1UL << PERF_BPIL_JITED_FUNC_LENS;
+	arrays |= 1UL << PERF_BPIL_FUNC_INFO;
+	arrays |= 1UL << PERF_BPIL_PROG_TAGS;
+	arrays |= 1UL << PERF_BPIL_JITED_INSNS;
+	arrays |= 1UL << PERF_BPIL_LINE_INFO;
+	arrays |= 1UL << PERF_BPIL_JITED_LINE_INFO;
 
-	info_linear = bpf_program__get_prog_info_linear(fd, arrays);
+	info_linear = get_bpf_prog_info_linear(fd, arrays);
 	if (IS_ERR_OR_NULL(info_linear)) {
 		pr_debug("%s: failed to get BPF program info. aborting\n", __func__);
 		goto out;
@@ -576,7 +606,7 @@ void bpf_event__print_bpf_prog_info(struct bpf_prog_info *info,
 		synthesize_bpf_prog_name(name, KSYM_NAME_LEN, info, btf, 0);
 		fprintf(fp, "# bpf_prog_info %u: %s addr 0x%llx size %u\n",
 			info->id, name, prog_addrs[0], prog_lens[0]);
-		return;
+		goto out;
 	}
 
 	fprintf(fp, "# bpf_prog_info %u:\n", info->id);
@@ -586,4 +616,6 @@ void bpf_event__print_bpf_prog_info(struct bpf_prog_info *info,
 		fprintf(fp, "# \tsub_prog %u: %s addr 0x%llx size %u\n",
 			i, name, prog_addrs[i], prog_lens[i]);
 	}
+out:
+	btf__free(btf);
 }
diff --git a/tools/perf/util/bpf-event.h b/tools/perf/util/bpf-event.h
index 68f315c3df5b..144a8a24cc69 100644
--- a/tools/perf/util/bpf-event.h
+++ b/tools/perf/util/bpf-event.h
@@ -19,7 +19,7 @@ struct evlist;
 struct target;
 
 struct bpf_prog_info_node {
-	struct bpf_prog_info_linear	*info_linear;
+	struct perf_bpil		*info_linear;
 	struct rb_node			rb_node;
 };
 
diff --git a/tools/perf/util/bpf-utils.c b/tools/perf/util/bpf-utils.c
new file mode 100644
index 000000000000..e271e05e51bc
--- /dev/null
+++ b/tools/perf/util/bpf-utils.c
@@ -0,0 +1,261 @@
+// SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause)
+
+#ifndef _GNU_SOURCE
+#define _GNU_SOURCE
+#endif
+
+#include <errno.h>
+#include <stdlib.h>
+#include <linux/err.h>
+#include <linux/kernel.h>
+#include <bpf/bpf.h>
+#include "bpf-utils.h"
+#include "debug.h"
+
+struct bpil_array_desc {
+	int	array_offset;	/* e.g. offset of jited_prog_insns */
+	int	count_offset;	/* e.g. offset of jited_prog_len */
+	int	size_offset;	/* > 0: offset of rec size,
+				 * < 0: fix size of -size_offset
+				 */
+};
+
+static struct bpil_array_desc bpil_array_desc[] = {
+	[PERF_BPIL_JITED_INSNS] = {
+		offsetof(struct bpf_prog_info, jited_prog_insns),
+		offsetof(struct bpf_prog_info, jited_prog_len),
+		-1,
+	},
+	[PERF_BPIL_XLATED_INSNS] = {
+		offsetof(struct bpf_prog_info, xlated_prog_insns),
+		offsetof(struct bpf_prog_info, xlated_prog_len),
+		-1,
+	},
+	[PERF_BPIL_MAP_IDS] = {
+		offsetof(struct bpf_prog_info, map_ids),
+		offsetof(struct bpf_prog_info, nr_map_ids),
+		-(int)sizeof(__u32),
+	},
+	[PERF_BPIL_JITED_KSYMS] = {
+		offsetof(struct bpf_prog_info, jited_ksyms),
+		offsetof(struct bpf_prog_info, nr_jited_ksyms),
+		-(int)sizeof(__u64),
+	},
+	[PERF_BPIL_JITED_FUNC_LENS] = {
+		offsetof(struct bpf_prog_info, jited_func_lens),
+		offsetof(struct bpf_prog_info, nr_jited_func_lens),
+		-(int)sizeof(__u32),
+	},
+	[PERF_BPIL_FUNC_INFO] = {
+		offsetof(struct bpf_prog_info, func_info),
+		offsetof(struct bpf_prog_info, nr_func_info),
+		offsetof(struct bpf_prog_info, func_info_rec_size),
+	},
+	[PERF_BPIL_LINE_INFO] = {
+		offsetof(struct bpf_prog_info, line_info),
+		offsetof(struct bpf_prog_info, nr_line_info),
+		offsetof(struct bpf_prog_info, line_info_rec_size),
+	},
+	[PERF_BPIL_JITED_LINE_INFO] = {
+		offsetof(struct bpf_prog_info, jited_line_info),
+		offsetof(struct bpf_prog_info, nr_jited_line_info),
+		offsetof(struct bpf_prog_info, jited_line_info_rec_size),
+	},
+	[PERF_BPIL_PROG_TAGS] = {
+		offsetof(struct bpf_prog_info, prog_tags),
+		offsetof(struct bpf_prog_info, nr_prog_tags),
+		-(int)sizeof(__u8) * BPF_TAG_SIZE,
+	},
+
+};
+
+static __u32 bpf_prog_info_read_offset_u32(struct bpf_prog_info *info,
+					   int offset)
+{
+	__u32 *array = (__u32 *)info;
+
+	if (offset >= 0)
+		return array[offset / sizeof(__u32)];
+	return -(int)offset;
+}
+
+static __u64 bpf_prog_info_read_offset_u64(struct bpf_prog_info *info,
+					   int offset)
+{
+	__u64 *array = (__u64 *)info;
+
+	if (offset >= 0)
+		return array[offset / sizeof(__u64)];
+	return -(int)offset;
+}
+
+static void bpf_prog_info_set_offset_u32(struct bpf_prog_info *info, int offset,
+					 __u32 val)
+{
+	__u32 *array = (__u32 *)info;
+
+	if (offset >= 0)
+		array[offset / sizeof(__u32)] = val;
+}
+
+static void bpf_prog_info_set_offset_u64(struct bpf_prog_info *info, int offset,
+					 __u64 val)
+{
+	__u64 *array = (__u64 *)info;
+
+	if (offset >= 0)
+		array[offset / sizeof(__u64)] = val;
+}
+
+struct perf_bpil *
+get_bpf_prog_info_linear(int fd, __u64 arrays)
+{
+	struct bpf_prog_info info = {};
+	struct perf_bpil *info_linear;
+	__u32 info_len = sizeof(info);
+	__u32 data_len = 0;
+	int i, err;
+	void *ptr;
+
+	if (arrays >> PERF_BPIL_LAST_ARRAY)
+		return ERR_PTR(-EINVAL);
+
+	/* step 1: get array dimensions */
+	err = bpf_obj_get_info_by_fd(fd, &info, &info_len);
+	if (err) {
+		pr_debug("can't get prog info: %s", strerror(errno));
+		return ERR_PTR(-EFAULT);
+	}
+
+	/* step 2: calculate total size of all arrays */
+	for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) {
+		bool include_array = (arrays & (1UL << i)) > 0;
+		struct bpil_array_desc *desc;
+		__u32 count, size;
+
+		desc = bpil_array_desc + i;
+
+		/* kernel is too old to support this field */
+		if (info_len < desc->array_offset + sizeof(__u32) ||
+		    info_len < desc->count_offset + sizeof(__u32) ||
+		    (desc->size_offset > 0 && info_len < (__u32)desc->size_offset))
+			include_array = false;
+
+		if (!include_array) {
+			arrays &= ~(1UL << i);	/* clear the bit */
+			continue;
+		}
+
+		count = bpf_prog_info_read_offset_u32(&info, desc->count_offset);
+		size  = bpf_prog_info_read_offset_u32(&info, desc->size_offset);
+
+		data_len += count * size;
+	}
+
+	/* step 3: allocate continuous memory */
+	data_len = roundup(data_len, sizeof(__u64));
+	info_linear = malloc(sizeof(struct perf_bpil) + data_len);
+	if (!info_linear)
+		return ERR_PTR(-ENOMEM);
+
+	/* step 4: fill data to info_linear->info */
+	info_linear->arrays = arrays;
+	memset(&info_linear->info, 0, sizeof(info));
+	ptr = info_linear->data;
+
+	for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) {
+		struct bpil_array_desc *desc;
+		__u32 count, size;
+
+		if ((arrays & (1UL << i)) == 0)
+			continue;
+
+		desc  = bpil_array_desc + i;
+		count = bpf_prog_info_read_offset_u32(&info, desc->count_offset);
+		size  = bpf_prog_info_read_offset_u32(&info, desc->size_offset);
+		bpf_prog_info_set_offset_u32(&info_linear->info,
+					     desc->count_offset, count);
+		bpf_prog_info_set_offset_u32(&info_linear->info,
+					     desc->size_offset, size);
+		bpf_prog_info_set_offset_u64(&info_linear->info,
+					     desc->array_offset,
+					     ptr_to_u64(ptr));
+		ptr += count * size;
+	}
+
+	/* step 5: call syscall again to get required arrays */
+	err = bpf_obj_get_info_by_fd(fd, &info_linear->info, &info_len);
+	if (err) {
+		pr_debug("can't get prog info: %s", strerror(errno));
+		free(info_linear);
+		return ERR_PTR(-EFAULT);
+	}
+
+	/* step 6: verify the data */
+	for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) {
+		struct bpil_array_desc *desc;
+		__u32 v1, v2;
+
+		if ((arrays & (1UL << i)) == 0)
+			continue;
+
+		desc = bpil_array_desc + i;
+		v1 = bpf_prog_info_read_offset_u32(&info, desc->count_offset);
+		v2 = bpf_prog_info_read_offset_u32(&info_linear->info,
+						   desc->count_offset);
+		if (v1 != v2)
+			pr_warning("%s: mismatch in element count\n", __func__);
+
+		v1 = bpf_prog_info_read_offset_u32(&info, desc->size_offset);
+		v2 = bpf_prog_info_read_offset_u32(&info_linear->info,
+						   desc->size_offset);
+		if (v1 != v2)
+			pr_warning("%s: mismatch in rec size\n", __func__);
+	}
+
+	/* step 7: update info_len and data_len */
+	info_linear->info_len = sizeof(struct bpf_prog_info);
+	info_linear->data_len = data_len;
+
+	return info_linear;
+}
+
+void bpil_addr_to_offs(struct perf_bpil *info_linear)
+{
+	int i;
+
+	for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) {
+		struct bpil_array_desc *desc;
+		__u64 addr, offs;
+
+		if ((info_linear->arrays & (1UL << i)) == 0)
+			continue;
+
+		desc = bpil_array_desc + i;
+		addr = bpf_prog_info_read_offset_u64(&info_linear->info,
+						     desc->array_offset);
+		offs = addr - ptr_to_u64(info_linear->data);
+		bpf_prog_info_set_offset_u64(&info_linear->info,
+					     desc->array_offset, offs);
+	}
+}
+
+void bpil_offs_to_addr(struct perf_bpil *info_linear)
+{
+	int i;
+
+	for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) {
+		struct bpil_array_desc *desc;
+		__u64 addr, offs;
+
+		if ((info_linear->arrays & (1UL << i)) == 0)
+			continue;
+
+		desc = bpil_array_desc + i;
+		offs = bpf_prog_info_read_offset_u64(&info_linear->info,
+						     desc->array_offset);
+		addr = offs + ptr_to_u64(info_linear->data);
+		bpf_prog_info_set_offset_u64(&info_linear->info,
+					     desc->array_offset, addr);
+	}
+}
diff --git a/tools/perf/util/bpf-utils.h b/tools/perf/util/bpf-utils.h
new file mode 100644
index 000000000000..86a5055cdfad
--- /dev/null
+++ b/tools/perf/util/bpf-utils.h
@@ -0,0 +1,76 @@
+/* SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) */
+
+#ifndef __PERF_BPF_UTILS_H
+#define __PERF_BPF_UTILS_H
+
+#define ptr_to_u64(ptr)    ((__u64)(unsigned long)(ptr))
+
+#ifdef HAVE_LIBBPF_SUPPORT
+
+#include <bpf/libbpf.h>
+
+/*
+ * Get bpf_prog_info in continuous memory
+ *
+ * struct bpf_prog_info has multiple arrays. The user has option to choose
+ * arrays to fetch from kernel. The following APIs provide an uniform way to
+ * fetch these data. All arrays in bpf_prog_info are stored in a single
+ * continuous memory region. This makes it easy to store the info in a
+ * file.
+ *
+ * Before writing perf_bpil to files, it is necessary to
+ * translate pointers in bpf_prog_info to offsets. Helper functions
+ * bpil_addr_to_offs() and bpil_offs_to_addr()
+ * are introduced to switch between pointers and offsets.
+ *
+ * Examples:
+ *   # To fetch map_ids and prog_tags:
+ *   __u64 arrays = (1UL << PERF_BPIL_MAP_IDS) |
+ *           (1UL << PERF_BPIL_PROG_TAGS);
+ *   struct perf_bpil *info_linear =
+ *           get_bpf_prog_info_linear(fd, arrays);
+ *
+ *   # To save data in file
+ *   bpil_addr_to_offs(info_linear);
+ *   write(f, info_linear, sizeof(*info_linear) + info_linear->data_len);
+ *
+ *   # To read data from file
+ *   read(f, info_linear, <proper_size>);
+ *   bpil_offs_to_addr(info_linear);
+ */
+enum perf_bpil_array_types {
+	PERF_BPIL_FIRST_ARRAY = 0,
+	PERF_BPIL_JITED_INSNS = 0,
+	PERF_BPIL_XLATED_INSNS,
+	PERF_BPIL_MAP_IDS,
+	PERF_BPIL_JITED_KSYMS,
+	PERF_BPIL_JITED_FUNC_LENS,
+	PERF_BPIL_FUNC_INFO,
+	PERF_BPIL_LINE_INFO,
+	PERF_BPIL_JITED_LINE_INFO,
+	PERF_BPIL_PROG_TAGS,
+	PERF_BPIL_LAST_ARRAY,
+};
+
+struct perf_bpil {
+	/* size of struct bpf_prog_info, when the tool is compiled */
+	__u32			info_len;
+	/* total bytes allocated for data, round up to 8 bytes */
+	__u32			data_len;
+	/* which arrays are included in data */
+	__u64			arrays;
+	struct bpf_prog_info	info;
+	__u8			data[];
+};
+
+struct perf_bpil *
+get_bpf_prog_info_linear(int fd, __u64 arrays);
+
+void
+bpil_addr_to_offs(struct perf_bpil *info_linear);
+
+void
+bpil_offs_to_addr(struct perf_bpil *info_linear);
+
+#endif /* HAVE_LIBBPF_SUPPORT */
+#endif /* __PERF_BPF_UTILS_H */
diff --git a/tools/perf/util/bpf_counter.c b/tools/perf/util/bpf_counter.c
index ba0f20853651..c17d4a43ce06 100644
--- a/tools/perf/util/bpf_counter.c
+++ b/tools/perf/util/bpf_counter.c
@@ -13,6 +13,7 @@
 #include <perf/bpf_perf.h>
 
 #include "bpf_counter.h"
+#include "bpf-utils.h"
 #include "counts.h"
 #include "debug.h"
 #include "evsel.h"
@@ -61,14 +62,13 @@ static int bpf_program_profiler__destroy(struct evsel *evsel)
 
 static char *bpf_target_prog_name(int tgt_fd)
 {
-	struct bpf_prog_info_linear *info_linear;
 	struct bpf_func_info *func_info;
+	struct perf_bpil *info_linear;
 	const struct btf_type *t;
 	struct btf *btf = NULL;
 	char *name = NULL;
 
-	info_linear = bpf_program__get_prog_info_linear(
-		tgt_fd, 1UL << BPF_PROG_INFO_FUNC_INFO);
+	info_linear = get_bpf_prog_info_linear(tgt_fd, 1UL << PERF_BPIL_FUNC_INFO);
 	if (IS_ERR_OR_NULL(info_linear)) {
 		pr_debug("failed to get info_linear for prog FD %d\n", tgt_fd);
 		return NULL;
@@ -127,9 +127,9 @@ static int bpf_program_profiler_load_one(struct evsel *evsel, u32 prog_id)
 
 	skel->rodata->num_cpu = evsel__nr_cpus(evsel);
 
-	bpf_map__resize(skel->maps.events, evsel__nr_cpus(evsel));
-	bpf_map__resize(skel->maps.fentry_readings, 1);
-	bpf_map__resize(skel->maps.accum_readings, 1);
+	bpf_map__set_max_entries(skel->maps.events, evsel__nr_cpus(evsel));
+	bpf_map__set_max_entries(skel->maps.fentry_readings, 1);
+	bpf_map__set_max_entries(skel->maps.accum_readings, 1);
 
 	prog_name = bpf_target_prog_name(prog_fd);
 	if (!prog_name) {
@@ -399,7 +399,7 @@ static int bperf_reload_leader_program(struct evsel *evsel, int attr_map_fd,
 		return -1;
 	}
 
-	bpf_map__resize(skel->maps.events, libbpf_num_possible_cpus());
+	bpf_map__set_max_entries(skel->maps.events, libbpf_num_possible_cpus());
 	err = bperf_leader_bpf__load(skel);
 	if (err) {
 		pr_err("Failed to load leader skeleton\n");
diff --git a/tools/perf/util/bpf_counter_cgroup.c b/tools/perf/util/bpf_counter_cgroup.c
index 89aa5e71db1a..cbc6c2bca488 100644
--- a/tools/perf/util/bpf_counter_cgroup.c
+++ b/tools/perf/util/bpf_counter_cgroup.c
@@ -65,14 +65,14 @@ static int bperf_load_program(struct evlist *evlist)
 
 	/* we need one copy of events per cpu for reading */
 	map_size = total_cpus * evlist->core.nr_entries / nr_cgroups;
-	bpf_map__resize(skel->maps.events, map_size);
-	bpf_map__resize(skel->maps.cgrp_idx, nr_cgroups);
+	bpf_map__set_max_entries(skel->maps.events, map_size);
+	bpf_map__set_max_entries(skel->maps.cgrp_idx, nr_cgroups);
 	/* previous result is saved in a per-cpu array */
 	map_size = evlist->core.nr_entries / nr_cgroups;
-	bpf_map__resize(skel->maps.prev_readings, map_size);
+	bpf_map__set_max_entries(skel->maps.prev_readings, map_size);
 	/* cgroup result needs all events (per-cpu) */
 	map_size = evlist->core.nr_entries;
-	bpf_map__resize(skel->maps.cgrp_readings, map_size);
+	bpf_map__set_max_entries(skel->maps.cgrp_readings, map_size);
 
 	set_max_rlimit();
 
diff --git a/tools/perf/util/c++/clang-c.h b/tools/perf/util/c++/clang-c.h
index 2df8a45bd088..d3731a876b6c 100644
--- a/tools/perf/util/c++/clang-c.h
+++ b/tools/perf/util/c++/clang-c.h
@@ -12,8 +12,9 @@ extern "C" {
 extern void perf_clang__init(void);
 extern void perf_clang__cleanup(void);
 
-extern int test__clang_to_IR(void);
-extern int test__clang_to_obj(void);
+struct test_suite;
+extern int test__clang_to_IR(struct test_suite *test, int subtest);
+extern int test__clang_to_obj(struct test_suite *test, int subtest);
 
 extern int perf_clang__compile_bpf(const char *filename,
 				   void **p_obj_buf,
@@ -26,9 +27,6 @@ extern int perf_clang__compile_bpf(const char *filename,
 static inline void perf_clang__init(void) { }
 static inline void perf_clang__cleanup(void) { }
 
-static inline int test__clang_to_IR(void) { return -1; }
-static inline int test__clang_to_obj(void) { return -1;}
-
 static inline int
 perf_clang__compile_bpf(const char *filename __maybe_unused,
 			void **p_obj_buf __maybe_unused,
diff --git a/tools/perf/util/c++/clang-test.cpp b/tools/perf/util/c++/clang-test.cpp
index 21b23605f78b..a4683ca53697 100644
--- a/tools/perf/util/c++/clang-test.cpp
+++ b/tools/perf/util/c++/clang-test.cpp
@@ -35,7 +35,8 @@ __test__clang_to_IR(void)
 }
 
 extern "C" {
-int test__clang_to_IR(void)
+int test__clang_to_IR(struct test_suite *test __maybe_unused,
+                      int subtest __maybe_unused)
 {
 	perf_clang_scope _scope;
 
@@ -48,7 +49,8 @@ int test__clang_to_IR(void)
 	return -1;
 }
 
-int test__clang_to_obj(void)
+int test__clang_to_obj(struct test_suite *test __maybe_unused,
+                       int subtest __maybe_unused)
 {
 	perf_clang_scope _scope;
 
diff --git a/tools/perf/util/c++/clang.cpp b/tools/perf/util/c++/clang.cpp
index c8885dfa3667..df7b18fb6b6e 100644
--- a/tools/perf/util/c++/clang.cpp
+++ b/tools/perf/util/c++/clang.cpp
@@ -43,8 +43,6 @@ createCompilerInvocation(llvm::opt::ArgStringList CFlags, StringRef& Path,
 		"-cc1",
 		"-triple", "bpf-pc-linux",
 		"-fsyntax-only",
-		"-ferror-limit", "19",
-		"-fmessage-length", "127",
 		"-O2",
 		"-nostdsysteminc",
 		"-nobuiltininc",
@@ -55,7 +53,11 @@ createCompilerInvocation(llvm::opt::ArgStringList CFlags, StringRef& Path,
 		"-x", "c"};
 
 	CCArgs.append(CFlags.begin(), CFlags.end());
-	CompilerInvocation *CI = tooling::newInvocation(&Diags, CCArgs);
+	CompilerInvocation *CI = tooling::newInvocation(&Diags, CCArgs
+#if CLANG_VERSION_MAJOR >= 11
+                                                        ,/*BinaryName=*/nullptr
+#endif
+                                                        );
 
 	FrontendOptions& Opts = CI->getFrontendOpts();
 	Opts.Inputs.clear();
@@ -151,13 +153,16 @@ getBPFObjectFromModule(llvm::Module *Module)
 
 	legacy::PassManager PM;
 	bool NotAdded;
-#if CLANG_VERSION_MAJOR < 7
-	NotAdded = TargetMachine->addPassesToEmitFile(PM, ostream,
-						      TargetMachine::CGFT_ObjectFile);
+	NotAdded = TargetMachine->addPassesToEmitFile(PM, ostream
+#if CLANG_VERSION_MAJOR >= 7
+                                                      , /*DwoOut=*/nullptr
+#endif
+#if CLANG_VERSION_MAJOR < 10
+                                                      , TargetMachine::CGFT_ObjectFile
 #else
-	NotAdded = TargetMachine->addPassesToEmitFile(PM, ostream, nullptr,
-						      TargetMachine::CGFT_ObjectFile);
+                                                      , llvm::CGFT_ObjectFile
 #endif
+                                                      );
 	if (NotAdded) {
 		llvm::errs() << "TargetMachine can't emit a file of this type\n";
 		return std::unique_ptr<llvm::SmallVectorImpl<char>>(nullptr);
diff --git a/tools/perf/util/cputopo.c b/tools/perf/util/cputopo.c
index ec77e2a7b3ca..51b429c86f98 100644
--- a/tools/perf/util/cputopo.c
+++ b/tools/perf/util/cputopo.c
@@ -14,14 +14,16 @@
 #include "env.h"
 #include "pmu-hybrid.h"
 
-#define CORE_SIB_FMT \
+#define PACKAGE_CPUS_FMT \
+	"%s/devices/system/cpu/cpu%d/topology/package_cpus_list"
+#define PACKAGE_CPUS_FMT_OLD \
 	"%s/devices/system/cpu/cpu%d/topology/core_siblings_list"
-#define DIE_SIB_FMT \
+#define DIE_CPUS_FMT \
 	"%s/devices/system/cpu/cpu%d/topology/die_cpus_list"
-#define THRD_SIB_FMT \
-	"%s/devices/system/cpu/cpu%d/topology/thread_siblings_list"
-#define THRD_SIB_FMT_NEW \
+#define CORE_CPUS_FMT \
 	"%s/devices/system/cpu/cpu%d/topology/core_cpus_list"
+#define CORE_CPUS_FMT_OLD \
+	"%s/devices/system/cpu/cpu%d/topology/thread_siblings_list"
 #define NODE_ONLINE_FMT \
 	"%s/devices/system/node/online"
 #define NODE_MEMINFO_FMT \
@@ -39,8 +41,12 @@ static int build_cpu_topology(struct cpu_topology *tp, int cpu)
 	u32 i = 0;
 	int ret = -1;
 
-	scnprintf(filename, MAXPATHLEN, CORE_SIB_FMT,
+	scnprintf(filename, MAXPATHLEN, PACKAGE_CPUS_FMT,
 		  sysfs__mountpoint(), cpu);
+	if (access(filename, F_OK) == -1) {
+		scnprintf(filename, MAXPATHLEN, PACKAGE_CPUS_FMT_OLD,
+			sysfs__mountpoint(), cpu);
+	}
 	fp = fopen(filename, "r");
 	if (!fp)
 		goto try_dies;
@@ -54,23 +60,23 @@ static int build_cpu_topology(struct cpu_topology *tp, int cpu)
 	if (p)
 		*p = '\0';
 
-	for (i = 0; i < tp->core_sib; i++) {
-		if (!strcmp(buf, tp->core_siblings[i]))
+	for (i = 0; i < tp->package_cpus_lists; i++) {
+		if (!strcmp(buf, tp->package_cpus_list[i]))
 			break;
 	}
-	if (i == tp->core_sib) {
-		tp->core_siblings[i] = buf;
-		tp->core_sib++;
+	if (i == tp->package_cpus_lists) {
+		tp->package_cpus_list[i] = buf;
+		tp->package_cpus_lists++;
 		buf = NULL;
 		len = 0;
 	}
 	ret = 0;
 
 try_dies:
-	if (!tp->die_siblings)
+	if (!tp->die_cpus_list)
 		goto try_threads;
 
-	scnprintf(filename, MAXPATHLEN, DIE_SIB_FMT,
+	scnprintf(filename, MAXPATHLEN, DIE_CPUS_FMT,
 		  sysfs__mountpoint(), cpu);
 	fp = fopen(filename, "r");
 	if (!fp)
@@ -85,23 +91,23 @@ try_dies:
 	if (p)
 		*p = '\0';
 
-	for (i = 0; i < tp->die_sib; i++) {
-		if (!strcmp(buf, tp->die_siblings[i]))
+	for (i = 0; i < tp->die_cpus_lists; i++) {
+		if (!strcmp(buf, tp->die_cpus_list[i]))
 			break;
 	}
-	if (i == tp->die_sib) {
-		tp->die_siblings[i] = buf;
-		tp->die_sib++;
+	if (i == tp->die_cpus_lists) {
+		tp->die_cpus_list[i] = buf;
+		tp->die_cpus_lists++;
 		buf = NULL;
 		len = 0;
 	}
 	ret = 0;
 
 try_threads:
-	scnprintf(filename, MAXPATHLEN, THRD_SIB_FMT_NEW,
+	scnprintf(filename, MAXPATHLEN, CORE_CPUS_FMT,
 		  sysfs__mountpoint(), cpu);
 	if (access(filename, F_OK) == -1) {
-		scnprintf(filename, MAXPATHLEN, THRD_SIB_FMT,
+		scnprintf(filename, MAXPATHLEN, CORE_CPUS_FMT_OLD,
 			  sysfs__mountpoint(), cpu);
 	}
 	fp = fopen(filename, "r");
@@ -115,13 +121,13 @@ try_threads:
 	if (p)
 		*p = '\0';
 
-	for (i = 0; i < tp->thread_sib; i++) {
-		if (!strcmp(buf, tp->thread_siblings[i]))
+	for (i = 0; i < tp->core_cpus_lists; i++) {
+		if (!strcmp(buf, tp->core_cpus_list[i]))
 			break;
 	}
-	if (i == tp->thread_sib) {
-		tp->thread_siblings[i] = buf;
-		tp->thread_sib++;
+	if (i == tp->core_cpus_lists) {
+		tp->core_cpus_list[i] = buf;
+		tp->core_cpus_lists++;
 		buf = NULL;
 	}
 	ret = 0;
@@ -139,16 +145,14 @@ void cpu_topology__delete(struct cpu_topology *tp)
 	if (!tp)
 		return;
 
-	for (i = 0 ; i < tp->core_sib; i++)
-		zfree(&tp->core_siblings[i]);
+	for (i = 0 ; i < tp->package_cpus_lists; i++)
+		zfree(&tp->package_cpus_list[i]);
 
-	if (tp->die_sib) {
-		for (i = 0 ; i < tp->die_sib; i++)
-			zfree(&tp->die_siblings[i]);
-	}
+	for (i = 0 ; i < tp->die_cpus_lists; i++)
+		zfree(&tp->die_cpus_list[i]);
 
-	for (i = 0 ; i < tp->thread_sib; i++)
-		zfree(&tp->thread_siblings[i]);
+	for (i = 0 ; i < tp->core_cpus_lists; i++)
+		zfree(&tp->core_cpus_list[i]);
 
 	free(tp);
 }
@@ -164,7 +168,7 @@ static bool has_die_topology(void)
 	if (strncmp(uts.machine, "x86_64", 6))
 		return false;
 
-	scnprintf(filename, MAXPATHLEN, DIE_SIB_FMT,
+	scnprintf(filename, MAXPATHLEN, DIE_CPUS_FMT,
 		  sysfs__mountpoint(), 0);
 	if (access(filename, F_OK) == -1)
 		return false;
@@ -205,13 +209,13 @@ struct cpu_topology *cpu_topology__new(void)
 
 	tp = addr;
 	addr += sizeof(*tp);
-	tp->core_siblings = addr;
+	tp->package_cpus_list = addr;
 	addr += sz;
 	if (has_die) {
-		tp->die_siblings = addr;
+		tp->die_cpus_list = addr;
 		addr += sz;
 	}
-	tp->thread_siblings = addr;
+	tp->core_cpus_list = addr;
 
 	for (i = 0; i < nr; i++) {
 		if (!cpu_map__has(map, i))
diff --git a/tools/perf/util/cputopo.h b/tools/perf/util/cputopo.h
index d9af97177068..854e18f9041e 100644
--- a/tools/perf/util/cputopo.h
+++ b/tools/perf/util/cputopo.h
@@ -5,12 +5,33 @@
 #include <linux/types.h>
 
 struct cpu_topology {
-	u32	  core_sib;
-	u32	  die_sib;
-	u32	  thread_sib;
-	char	**core_siblings;
-	char	**die_siblings;
-	char	**thread_siblings;
+	/* The number of unique package_cpus_lists below. */
+	u32	  package_cpus_lists;
+	/* The number of unique die_cpu_lists below. */
+	u32	  die_cpus_lists;
+	/* The number of unique core_cpu_lists below. */
+	u32	  core_cpus_lists;
+	/*
+	 * An array of strings where each string is unique and read from
+	 * /sys/devices/system/cpu/cpuX/topology/package_cpus_list. From the ABI
+	 * each of these is a human-readable list of CPUs sharing the same
+	 * physical_package_id. The format is like 0-3, 8-11, 14,17.
+	 */
+	const char **package_cpus_list;
+	/*
+	 * An array of string where each string is unique and from
+	 * /sys/devices/system/cpu/cpuX/topology/die_cpus_list. From the ABI
+	 * each of these is a human-readable list of CPUs within the same die.
+	 * The format is like 0-3, 8-11, 14,17.
+	 */
+	const char **die_cpus_list;
+	/*
+	 * An array of string where each string is unique and from
+	 * /sys/devices/system/cpu/cpuX/topology/core_cpus_list. From the ABI
+	 * each of these is a human-readable list of CPUs within the same
+	 * core. The format is like 0-3, 8-11, 14,17.
+	 */
+	const char **core_cpus_list;
 };
 
 struct numa_topology_node {
diff --git a/tools/perf/util/cs-etm.c b/tools/perf/util/cs-etm.c
index f323adb1af85..4f672f7d008c 100644
--- a/tools/perf/util/cs-etm.c
+++ b/tools/perf/util/cs-etm.c
@@ -537,7 +537,7 @@ static void cs_etm__dump_event(struct cs_etm_queue *etmq,
 
 	fprintf(stdout, "\n");
 	color_fprintf(stdout, color,
-		     ". ... CoreSight %s Trace data: size %zu bytes\n",
+		     ". ... CoreSight %s Trace data: size %#zx bytes\n",
 		     cs_etm_decoder__get_name(etmq->decoder), buffer->size);
 
 	do {
diff --git a/tools/perf/util/data-convert-bt.c b/tools/perf/util/data-convert-bt.c
index aa862a26d95c..8f7705bbc2da 100644
--- a/tools/perf/util/data-convert-bt.c
+++ b/tools/perf/util/data-convert-bt.c
@@ -1437,7 +1437,7 @@ static struct bt_ctf_field_type *create_int_type(int size, bool sign, bool hex)
 	    bt_ctf_field_type_integer_set_base(type, BT_CTF_INTEGER_BASE_HEXADECIMAL))
 		goto err;
 
-#if __BYTE_ORDER == __BIG_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
 	bt_ctf_field_type_set_byte_order(type, BT_CTF_BYTE_ORDER_BIG_ENDIAN);
 #else
 	bt_ctf_field_type_set_byte_order(type, BT_CTF_BYTE_ORDER_LITTLE_ENDIAN);
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index 2c06abf6dcd2..c7a9fa0ffae9 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -24,6 +24,16 @@
 #include "util/parse-sublevel-options.h"
 
 #include <linux/ctype.h>
+#include <traceevent/event-parse.h>
+
+#define MAKE_LIBTRACEEVENT_VERSION(a, b, c) ((a)*255*255+(b)*255+(c))
+#ifndef LIBTRACEEVENT_VERSION
+/*
+ * If LIBTRACEEVENT_VERSION wasn't computed then set to version 1.1.0 that ships
+ * with the Linux kernel tools.
+ */
+#define LIBTRACEEVENT_VERSION MAKE_LIBTRACEEVENT_VERSION(1, 1, 0)
+#endif
 
 int verbose;
 int debug_peo_args;
@@ -228,6 +238,15 @@ int perf_debug_option(const char *str)
 	/* Allow only verbose value in range (0, 10), otherwise set 0. */
 	verbose = (verbose < 0) || (verbose > 10) ? 0 : verbose;
 
+#if MAKE_LIBTRACEEVENT_VERSION(1, 3, 0) <= LIBTRACEEVENT_VERSION
+	if (verbose == 1)
+		tep_set_loglevel(TEP_LOG_INFO);
+	else if (verbose == 2)
+		tep_set_loglevel(TEP_LOG_DEBUG);
+	else if (verbose >= 3)
+		tep_set_loglevel(TEP_LOG_ALL);
+#endif
+
 	return 0;
 }
 
diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 9ed9a5676d35..9cc8a1772b4b 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -14,6 +14,7 @@
 #ifdef HAVE_LIBBPF_SUPPORT
 #include <bpf/libbpf.h>
 #include "bpf-event.h"
+#include "bpf-utils.h"
 #endif
 #include "compress.h"
 #include "env.h"
diff --git a/tools/perf/util/dso.h b/tools/perf/util/dso.h
index 83723ba11dc8..011da3924fc1 100644
--- a/tools/perf/util/dso.h
+++ b/tools/perf/util/dso.h
@@ -193,7 +193,7 @@ struct dso {
 		int		 fd;
 		int		 status;
 		u32		 status_seen;
-		size_t		 file_size;
+		u64		 file_size;
 		struct list_head open_entry;
 		u64		 debug_frame_offset;
 		u64		 eh_frame_hdr_offset;
diff --git a/tools/perf/util/env.c b/tools/perf/util/env.c
index cf773f0dec38..b9904896eb97 100644
--- a/tools/perf/util/env.c
+++ b/tools/perf/util/env.c
@@ -16,6 +16,7 @@ struct perf_env perf_env;
 
 #ifdef HAVE_LIBBPF_SUPPORT
 #include "bpf-event.h"
+#include "bpf-utils.h"
 #include <bpf/libbpf.h>
 
 void perf_env__insert_bpf_prog_info(struct perf_env *env,
@@ -74,12 +75,13 @@ out:
 	return node;
 }
 
-void perf_env__insert_btf(struct perf_env *env, struct btf_node *btf_node)
+bool perf_env__insert_btf(struct perf_env *env, struct btf_node *btf_node)
 {
 	struct rb_node *parent = NULL;
 	__u32 btf_id = btf_node->id;
 	struct btf_node *node;
 	struct rb_node **p;
+	bool ret = true;
 
 	down_write(&env->bpf_progs.lock);
 	p = &env->bpf_progs.btfs.rb_node;
@@ -93,6 +95,7 @@ void perf_env__insert_btf(struct perf_env *env, struct btf_node *btf_node)
 			p = &(*p)->rb_right;
 		} else {
 			pr_debug("duplicated btf %u\n", btf_id);
+			ret = false;
 			goto out;
 		}
 	}
@@ -102,6 +105,7 @@ void perf_env__insert_btf(struct perf_env *env, struct btf_node *btf_node)
 	env->bpf_progs.btfs_cnt++;
 out:
 	up_write(&env->bpf_progs.lock);
+	return ret;
 }
 
 struct btf_node *perf_env__find_btf(struct perf_env *env, __u32 btf_id)
diff --git a/tools/perf/util/env.h b/tools/perf/util/env.h
index 1383876f72b3..163e5ec503a2 100644
--- a/tools/perf/util/env.h
+++ b/tools/perf/util/env.h
@@ -167,7 +167,7 @@ void perf_env__insert_bpf_prog_info(struct perf_env *env,
 				    struct bpf_prog_info_node *info_node);
 struct bpf_prog_info_node *perf_env__find_bpf_prog_info(struct perf_env *env,
 							__u32 prog_id);
-void perf_env__insert_btf(struct perf_env *env, struct btf_node *btf_node);
+bool perf_env__insert_btf(struct perf_env *env, struct btf_node *btf_node);
 struct btf_node *perf_env__find_btf(struct perf_env *env, __u32 btf_id);
 
 int perf_env__numa_node(struct perf_env *env, int cpu);
diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
index ac706304afe9..fe24801f8e9f 100644
--- a/tools/perf/util/event.c
+++ b/tools/perf/util/event.c
@@ -57,6 +57,7 @@ static const char *perf_event__names[] = {
 	[PERF_RECORD_BPF_EVENT]			= "BPF_EVENT",
 	[PERF_RECORD_CGROUP]			= "CGROUP",
 	[PERF_RECORD_TEXT_POKE]			= "TEXT_POKE",
+	[PERF_RECORD_AUX_OUTPUT_HW_ID]		= "AUX_OUTPUT_HW_ID",
 	[PERF_RECORD_HEADER_ATTR]		= "ATTR",
 	[PERF_RECORD_HEADER_EVENT_TYPE]		= "EVENT_TYPE",
 	[PERF_RECORD_HEADER_TRACING_DATA]	= "TRACING_DATA",
@@ -237,6 +238,14 @@ int perf_event__process_itrace_start(struct perf_tool *tool __maybe_unused,
 	return machine__process_itrace_start_event(machine, event);
 }
 
+int perf_event__process_aux_output_hw_id(struct perf_tool *tool __maybe_unused,
+					 union perf_event *event,
+					 struct perf_sample *sample __maybe_unused,
+					 struct machine *machine)
+{
+	return machine__process_aux_output_hw_id_event(machine, event);
+}
+
 int perf_event__process_lost_samples(struct perf_tool *tool __maybe_unused,
 				     union perf_event *event,
 				     struct perf_sample *sample,
@@ -407,6 +416,12 @@ size_t perf_event__fprintf_itrace_start(union perf_event *event, FILE *fp)
 		       event->itrace_start.pid, event->itrace_start.tid);
 }
 
+size_t perf_event__fprintf_aux_output_hw_id(union perf_event *event, FILE *fp)
+{
+	return fprintf(fp, " hw_id: %#"PRI_lx64"\n",
+		       event->aux_output_hw_id.hw_id);
+}
+
 size_t perf_event__fprintf_switch(union perf_event *event, FILE *fp)
 {
 	bool out = event->header.misc & PERF_RECORD_MISC_SWITCH_OUT;
@@ -534,6 +549,9 @@ size_t perf_event__fprintf(union perf_event *event, struct machine *machine, FIL
 	case PERF_RECORD_TEXT_POKE:
 		ret += perf_event__fprintf_text_poke(event, machine, fp);
 		break;
+	case PERF_RECORD_AUX_OUTPUT_HW_ID:
+		ret += perf_event__fprintf_aux_output_hw_id(event, fp);
+		break;
 	default:
 		ret += fprintf(fp, "\n");
 	}
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index 19ad64f2bd83..95ffed66369c 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -330,6 +330,10 @@ int perf_event__process_itrace_start(struct perf_tool *tool,
 				     union perf_event *event,
 				     struct perf_sample *sample,
 				     struct machine *machine);
+int perf_event__process_aux_output_hw_id(struct perf_tool *tool,
+					 union perf_event *event,
+					 struct perf_sample *sample,
+					 struct machine *machine);
 int perf_event__process_switch(struct perf_tool *tool,
 			       union perf_event *event,
 			       struct perf_sample *sample,
@@ -397,6 +401,7 @@ size_t perf_event__fprintf_mmap2(union perf_event *event, FILE *fp);
 size_t perf_event__fprintf_task(union perf_event *event, FILE *fp);
 size_t perf_event__fprintf_aux(union perf_event *event, FILE *fp);
 size_t perf_event__fprintf_itrace_start(union perf_event *event, FILE *fp);
+size_t perf_event__fprintf_aux_output_hw_id(union perf_event *event, FILE *fp);
 size_t perf_event__fprintf_switch(union perf_event *event, FILE *fp);
 size_t perf_event__fprintf_thread_map(union perf_event *event, FILE *fp);
 size_t perf_event__fprintf_cpu_map(union perf_event *event, FILE *fp);
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index dbfeceb2546c..ac0127be0459 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -241,7 +241,7 @@ void evsel__init(struct evsel *evsel,
 {
 	perf_evsel__init(&evsel->core, attr, idx);
 	evsel->tracking	   = !idx;
-	evsel->unit	   = "";
+	evsel->unit	   = strdup("");
 	evsel->scale	   = 1.0;
 	evsel->max_events  = ULONG_MAX;
 	evsel->evlist	   = NULL;
@@ -276,13 +276,8 @@ struct evsel *evsel__new_idx(struct perf_event_attr *attr, int idx)
 	}
 
 	if (evsel__is_clock(evsel)) {
-		/*
-		 * The evsel->unit points to static alias->unit
-		 * so it's ok to use static string in here.
-		 */
-		static const char *unit = "msec";
-
-		evsel->unit = unit;
+		free((char *)evsel->unit);
+		evsel->unit = strdup("msec");
 		evsel->scale = 1e-6;
 	}
 
@@ -294,7 +289,7 @@ static bool perf_event_can_profile_kernel(void)
 	return perf_event_paranoid_check(1);
 }
 
-struct evsel *evsel__new_cycles(bool precise, __u32 type, __u64 config)
+struct evsel *evsel__new_cycles(bool precise __maybe_unused, __u32 type, __u64 config)
 {
 	struct perf_event_attr attr = {
 		.type	= type,
@@ -305,18 +300,16 @@ struct evsel *evsel__new_cycles(bool precise, __u32 type, __u64 config)
 
 	event_attr_init(&attr);
 
-	if (!precise)
-		goto new_event;
-
 	/*
 	 * Now let the usual logic to set up the perf_event_attr defaults
 	 * to kick in when we return and before perf_evsel__open() is called.
 	 */
-new_event:
 	evsel = evsel__new(&attr);
 	if (evsel == NULL)
 		goto out;
 
+	arch_evsel__fixup_new_cycles(&evsel->core.attr);
+
 	evsel->precise_max = true;
 
 	/* use asprintf() because free(evsel) assumes name is allocated */
@@ -410,6 +403,11 @@ struct evsel *evsel__clone(struct evsel *orig)
 		if (evsel->filter == NULL)
 			goto out_err;
 	}
+	if (orig->metric_id) {
+		evsel->metric_id = strdup(orig->metric_id);
+		if (evsel->metric_id == NULL)
+			goto out_err;
+	}
 	evsel->cgrp = cgroup__get(orig->cgrp);
 	evsel->tp_format = orig->tp_format;
 	evsel->handler = orig->handler;
@@ -417,7 +415,11 @@ struct evsel *evsel__clone(struct evsel *orig)
 
 	evsel->max_events = orig->max_events;
 	evsel->tool_event = orig->tool_event;
-	evsel->unit = orig->unit;
+	free((char *)evsel->unit);
+	evsel->unit = strdup(orig->unit);
+	if (evsel->unit == NULL)
+		goto out_err;
+
 	evsel->scale = orig->scale;
 	evsel->snapshot = orig->snapshot;
 	evsel->per_pkg = orig->per_pkg;
@@ -779,6 +781,17 @@ out_unknown:
 	return "unknown";
 }
 
+const char *evsel__metric_id(const struct evsel *evsel)
+{
+	if (evsel->metric_id)
+		return evsel->metric_id;
+
+	if (evsel->core.attr.type == PERF_TYPE_SOFTWARE && evsel->tool_event)
+		return "duration_time";
+
+	return "unknown";
+}
+
 const char *evsel__group_name(struct evsel *evsel)
 {
 	return evsel->group_name ?: "anon group";
@@ -1047,6 +1060,10 @@ void __weak arch_evsel__set_sample_weight(struct evsel *evsel)
 	evsel__set_sample_bit(evsel, WEIGHT);
 }
 
+void __weak arch_evsel__fixup_new_cycles(struct perf_event_attr *attr __maybe_unused)
+{
+}
+
 /*
  * The enable_on_exec/disabled value strategy:
  *
@@ -1423,6 +1440,8 @@ void evsel__exit(struct evsel *evsel)
 	zfree(&evsel->group_name);
 	zfree(&evsel->name);
 	zfree(&evsel->pmu_name);
+	zfree(&evsel->unit);
+	zfree(&evsel->metric_id);
 	evsel__zero_per_pkg(evsel);
 	hashmap__free(evsel->per_pkg_mask);
 	evsel->per_pkg_mask = NULL;
@@ -1807,7 +1826,7 @@ static void evsel__disable_missing_features(struct evsel *evsel)
 		evsel->open_flags &= ~(unsigned long)PERF_FLAG_FD_CLOEXEC;
 	if (perf_missing_features.mmap2)
 		evsel->core.attr.mmap2 = 0;
-	if (perf_missing_features.exclude_guest)
+	if (evsel->pmu && evsel->pmu->missing_features.exclude_guest)
 		evsel->core.attr.exclude_guest = evsel->core.attr.exclude_host = 0;
 	if (perf_missing_features.lbr_flags)
 		evsel->core.attr.branch_sample_type &= ~(PERF_SAMPLE_BRANCH_NO_FLAGS |
@@ -1900,10 +1919,27 @@ bool evsel__detect_missing_features(struct evsel *evsel)
 		perf_missing_features.mmap2 = true;
 		pr_debug2_peo("switching off mmap2\n");
 		return true;
-	} else if (!perf_missing_features.exclude_guest &&
-		   (evsel->core.attr.exclude_guest || evsel->core.attr.exclude_host)) {
-		perf_missing_features.exclude_guest = true;
-		pr_debug2_peo("switching off exclude_guest, exclude_host\n");
+	} else if ((evsel->core.attr.exclude_guest || evsel->core.attr.exclude_host) &&
+		   (evsel->pmu == NULL || evsel->pmu->missing_features.exclude_guest)) {
+		if (evsel->pmu == NULL) {
+			evsel->pmu = evsel__find_pmu(evsel);
+			if (evsel->pmu)
+				evsel->pmu->missing_features.exclude_guest = true;
+			else {
+				/* we cannot find PMU, disable attrs now */
+				evsel->core.attr.exclude_host = false;
+				evsel->core.attr.exclude_guest = false;
+			}
+		}
+
+		if (evsel->exclude_GH) {
+			pr_debug2_peo("PMU has no exclude_host/guest support, bailing out\n");
+			return false;
+		}
+		if (!perf_missing_features.exclude_guest) {
+			perf_missing_features.exclude_guest = true;
+			pr_debug2_peo("switching off exclude_guest, exclude_host\n");
+		}
 		return true;
 	} else if (!perf_missing_features.sample_id_all) {
 		perf_missing_features.sample_id_all = true;
@@ -2221,6 +2257,54 @@ void __weak arch_perf_parse_sample_weight(struct perf_sample *data,
 	data->weight = *array;
 }
 
+u64 evsel__bitfield_swap_branch_flags(u64 value)
+{
+	u64 new_val = 0;
+
+	/*
+	 * branch_flags
+	 * union {
+	 * 	u64 values;
+	 * 	struct {
+	 * 		mispred:1	//target mispredicted
+	 * 		predicted:1	//target predicted
+	 * 		in_tx:1		//in transaction
+	 * 		abort:1		//transaction abort
+	 * 		cycles:16	//cycle count to last branch
+	 * 		type:4		//branch type
+	 * 		reserved:40
+	 * 	}
+	 * }
+	 *
+	 * Avoid bswap64() the entire branch_flag.value,
+	 * as it has variable bit-field sizes. Instead the
+	 * macro takes the bit-field position/size,
+	 * swaps it based on the host endianness.
+	 *
+	 * tep_is_bigendian() is used here instead of
+	 * bigendian() to avoid python test fails.
+	 */
+	if (tep_is_bigendian()) {
+		new_val = bitfield_swap(value, 0, 1);
+		new_val |= bitfield_swap(value, 1, 1);
+		new_val |= bitfield_swap(value, 2, 1);
+		new_val |= bitfield_swap(value, 3, 1);
+		new_val |= bitfield_swap(value, 4, 16);
+		new_val |= bitfield_swap(value, 20, 4);
+		new_val |= bitfield_swap(value, 24, 40);
+	} else {
+		new_val = bitfield_swap(value, 63, 1);
+		new_val |= bitfield_swap(value, 62, 1);
+		new_val |= bitfield_swap(value, 61, 1);
+		new_val |= bitfield_swap(value, 60, 1);
+		new_val |= bitfield_swap(value, 44, 16);
+		new_val |= bitfield_swap(value, 40, 4);
+		new_val |= bitfield_swap(value, 0, 40);
+	}
+
+	return new_val;
+}
+
 int evsel__parse_sample(struct evsel *evsel, union perf_event *event,
 			struct perf_sample *data)
 {
@@ -2408,6 +2492,8 @@ int evsel__parse_sample(struct evsel *evsel, union perf_event *event,
 	if (type & PERF_SAMPLE_BRANCH_STACK) {
 		const u64 max_branch_nr = UINT64_MAX /
 					  sizeof(struct branch_entry);
+		struct branch_entry *e;
+		unsigned int i;
 
 		OVERFLOW_CHECK_u64(array);
 		data->branch_stack = (struct branch_stack *)array++;
@@ -2416,10 +2502,33 @@ int evsel__parse_sample(struct evsel *evsel, union perf_event *event,
 			return -EFAULT;
 
 		sz = data->branch_stack->nr * sizeof(struct branch_entry);
-		if (evsel__has_branch_hw_idx(evsel))
+		if (evsel__has_branch_hw_idx(evsel)) {
 			sz += sizeof(u64);
-		else
+			e = &data->branch_stack->entries[0];
+		} else {
 			data->no_hw_idx = true;
+			/*
+			 * if the PERF_SAMPLE_BRANCH_HW_INDEX is not applied,
+			 * only nr and entries[] will be output by kernel.
+			 */
+			e = (struct branch_entry *)&data->branch_stack->hw_idx;
+		}
+
+		if (swapped) {
+			/*
+			 * struct branch_flag does not have endian
+			 * specific bit field definition. And bswap
+			 * will not resolve the issue, since these
+			 * are bit fields.
+			 *
+			 * evsel__bitfield_swap_branch_flags() uses a
+			 * bitfield_swap macro to swap the bit position
+			 * based on the host endians.
+			 */
+			for (i = 0; i < data->branch_stack->nr; i++, e++)
+				e->flags.value = evsel__bitfield_swap_branch_flags(e->flags.value);
+		}
+
 		OVERFLOW_CHECK(array, sz, max_size);
 		array = (void *)array + sz;
 	}
@@ -2928,3 +3037,15 @@ void evsel__set_leader(struct evsel *evsel, struct evsel *leader)
 {
 	evsel->core.leader = &leader->core;
 }
+
+int evsel__source_count(const struct evsel *evsel)
+{
+	struct evsel *pos;
+	int count = 0;
+
+	evlist__for_each_entry(evsel->evlist, pos) {
+		if (pos->metric_leader == evsel)
+			count++;
+	}
+	return count;
+}
diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
index 1f7edfa8568a..29d49a8c1e92 100644
--- a/tools/perf/util/evsel.h
+++ b/tools/perf/util/evsel.h
@@ -22,6 +22,7 @@ struct target;
 struct hashmap;
 struct bperf_leader_bpf;
 struct bperf_follower_bpf;
+struct perf_pmu;
 
 typedef int (evsel__sb_cb_t)(union perf_event *event, void *data);
 
@@ -68,6 +69,7 @@ struct evsel {
 		double			scale;
 		const char		*unit;
 		struct cgroup		*cgrp;
+		const char		*metric_id;
 		enum perf_tool_event	tool_event;
 		/* parse modifier helper */
 		int			exclude_GH;
@@ -152,6 +154,9 @@ struct evsel {
 	};
 	unsigned long		open_flags;
 	int			precise_ip_original;
+
+	/* for missing_features */
+	struct perf_pmu		*pmu;
 };
 
 struct perf_missing_features {
@@ -261,6 +266,7 @@ bool evsel__match_bpf_counter_events(const char *name);
 
 int __evsel__hw_cache_type_op_res_name(u8 type, u8 op, u8 result, char *bf, size_t size);
 const char *evsel__name(struct evsel *evsel);
+const char *evsel__metric_id(const struct evsel *evsel);
 
 const char *evsel__group_name(struct evsel *evsel);
 int evsel__group_desc(struct evsel *evsel, char *buf, size_t size);
@@ -277,6 +283,7 @@ void __evsel__reset_sample_bit(struct evsel *evsel, enum perf_event_sample_forma
 void evsel__set_sample_id(struct evsel *evsel, bool use_sample_identifier);
 
 void arch_evsel__set_sample_weight(struct evsel *evsel);
+void arch_evsel__fixup_new_cycles(struct perf_event_attr *attr);
 
 int evsel__set_filter(struct evsel *evsel, const char *filter);
 int evsel__append_tp_filter(struct evsel *evsel, const char *filter);
@@ -482,4 +489,18 @@ struct evsel *evsel__leader(struct evsel *evsel);
 bool evsel__has_leader(struct evsel *evsel, struct evsel *leader);
 bool evsel__is_leader(struct evsel *evsel);
 void evsel__set_leader(struct evsel *evsel, struct evsel *leader);
+int evsel__source_count(const struct evsel *evsel);
+
+/*
+ * Macro to swap the bit-field postition and size.
+ * Used when,
+ * - dont need to swap the entire u64 &&
+ * - when u64 has variable bit-field sizes &&
+ * - when presented in a host endian which is different
+ *   than the source endian of the perf.data file
+ */
+#define bitfield_swap(src, pos, size)	\
+	((((src) >> (pos)) & ((1ull << (size)) - 1)) << (63 - ((pos) + (size) - 1)))
+
+u64 evsel__bitfield_swap_branch_flags(u64 value);
 #endif /* __PERF_EVSEL_H */
diff --git a/tools/perf/util/evsel_fprintf.c b/tools/perf/util/evsel_fprintf.c
index bfedd7b23521..8c2ea8001329 100644
--- a/tools/perf/util/evsel_fprintf.c
+++ b/tools/perf/util/evsel_fprintf.c
@@ -11,6 +11,7 @@
 #include "strlist.h"
 #include "symbol.h"
 #include "srcline.h"
+#include "dso.h"
 
 static int comma_fprintf(FILE *fp, bool *first, const char *fmt, ...)
 {
@@ -144,12 +145,17 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 			if (print_arrow && !first)
 				printed += fprintf(fp, " <-");
 
-			if (print_ip)
-				printed += fprintf(fp, "%c%16" PRIx64, s, node->ip);
-
 			if (map)
 				addr = map->map_ip(map, node->ip);
 
+			if (print_ip) {
+				/* Show binary offset for userspace addr */
+				if (map && !map->dso->kernel)
+					printed += fprintf(fp, "%c%16" PRIx64, s, addr);
+				else
+					printed += fprintf(fp, "%c%16" PRIx64, s, node->ip);
+			}
+
 			if (print_sym) {
 				printed += fprintf(fp, " ");
 				node_al.addr = addr;
diff --git a/tools/perf/util/expr.c b/tools/perf/util/expr.c
index a850fd0be3ee..1d532b9fed29 100644
--- a/tools/perf/util/expr.c
+++ b/tools/perf/util/expr.c
@@ -5,13 +5,17 @@
 #include <stdlib.h>
 #include <string.h>
 #include "metricgroup.h"
+#include "cpumap.h"
+#include "cputopo.h"
 #include "debug.h"
 #include "expr.h"
 #include "expr-bison.h"
 #include "expr-flex.h"
+#include "smt.h"
 #include <linux/kernel.h>
 #include <linux/zalloc.h>
 #include <ctype.h>
+#include <math.h>
 
 #ifdef PARSER_DEBUG
 extern int expr_debug;
@@ -19,13 +23,15 @@ extern int expr_debug;
 
 struct expr_id_data {
 	union {
-		double val;
+		struct {
+			double val;
+			int source_count;
+		} val;
 		struct {
 			double val;
 			const char *metric_name;
 			const char *metric_expr;
 		} ref;
-		struct expr_id	*parent;
 	};
 
 	enum {
@@ -35,8 +41,6 @@ struct expr_id_data {
 		EXPR_ID_DATA__REF,
 		/* A reference but the value has been computed. */
 		EXPR_ID_DATA__REF_VALUE,
-		/* A parent is remembered for the recursion check. */
-		EXPR_ID_DATA__PARENT,
 	} kind;
 };
 
@@ -59,21 +63,34 @@ static bool key_equal(const void *key1, const void *key2,
 	return !strcmp((const char *)key1, (const char *)key2);
 }
 
-/* Caller must make sure id is allocated */
-int expr__add_id(struct expr_parse_ctx *ctx, const char *id)
+struct hashmap *ids__new(void)
+{
+	return hashmap__new(key_hash, key_equal, NULL);
+}
+
+void ids__free(struct hashmap *ids)
+{
+	struct hashmap_entry *cur;
+	size_t bkt;
+
+	if (ids == NULL)
+		return;
+
+	hashmap__for_each_entry(ids, cur, bkt) {
+		free((char *)cur->key);
+		free(cur->value);
+	}
+
+	hashmap__free(ids);
+}
+
+int ids__insert(struct hashmap *ids, const char *id)
 {
 	struct expr_id_data *data_ptr = NULL, *old_data = NULL;
 	char *old_key = NULL;
 	int ret;
 
-	data_ptr = malloc(sizeof(*data_ptr));
-	if (!data_ptr)
-		return -ENOMEM;
-
-	data_ptr->parent = ctx->parent;
-	data_ptr->kind = EXPR_ID_DATA__PARENT;
-
-	ret = hashmap__set(&ctx->ids, id, data_ptr,
+	ret = hashmap__set(ids, id, data_ptr,
 			   (const void **)&old_key, (void **)&old_data);
 	if (ret)
 		free(data_ptr);
@@ -82,9 +99,58 @@ int expr__add_id(struct expr_parse_ctx *ctx, const char *id)
 	return ret;
 }
 
+struct hashmap *ids__union(struct hashmap *ids1, struct hashmap *ids2)
+{
+	size_t bkt;
+	struct hashmap_entry *cur;
+	int ret;
+	struct expr_id_data *old_data = NULL;
+	char *old_key = NULL;
+
+	if (!ids1)
+		return ids2;
+
+	if (!ids2)
+		return ids1;
+
+	if (hashmap__size(ids1) <  hashmap__size(ids2)) {
+		struct hashmap *tmp = ids1;
+
+		ids1 = ids2;
+		ids2 = tmp;
+	}
+	hashmap__for_each_entry(ids2, cur, bkt) {
+		ret = hashmap__set(ids1, cur->key, cur->value,
+				(const void **)&old_key, (void **)&old_data);
+		free(old_key);
+		free(old_data);
+
+		if (ret) {
+			hashmap__free(ids1);
+			hashmap__free(ids2);
+			return NULL;
+		}
+	}
+	hashmap__free(ids2);
+	return ids1;
+}
+
+/* Caller must make sure id is allocated */
+int expr__add_id(struct expr_parse_ctx *ctx, const char *id)
+{
+	return ids__insert(ctx->ids, id);
+}
+
 /* Caller must make sure id is allocated */
 int expr__add_id_val(struct expr_parse_ctx *ctx, const char *id, double val)
 {
+	return expr__add_id_val_source_count(ctx, id, val, /*source_count=*/1);
+}
+
+/* Caller must make sure id is allocated */
+int expr__add_id_val_source_count(struct expr_parse_ctx *ctx, const char *id,
+				  double val, int source_count)
+{
 	struct expr_id_data *data_ptr = NULL, *old_data = NULL;
 	char *old_key = NULL;
 	int ret;
@@ -92,10 +158,11 @@ int expr__add_id_val(struct expr_parse_ctx *ctx, const char *id, double val)
 	data_ptr = malloc(sizeof(*data_ptr));
 	if (!data_ptr)
 		return -ENOMEM;
-	data_ptr->val = val;
+	data_ptr->val.val = val;
+	data_ptr->val.source_count = source_count;
 	data_ptr->kind = EXPR_ID_DATA__VALUE;
 
-	ret = hashmap__set(&ctx->ids, id, data_ptr,
+	ret = hashmap__set(ctx->ids, id, data_ptr,
 			   (const void **)&old_key, (void **)&old_data);
 	if (ret)
 		free(data_ptr);
@@ -140,7 +207,7 @@ int expr__add_ref(struct expr_parse_ctx *ctx, struct metric_ref *ref)
 	data_ptr->ref.metric_expr = ref->metric_expr;
 	data_ptr->kind = EXPR_ID_DATA__REF;
 
-	ret = hashmap__set(&ctx->ids, name, data_ptr,
+	ret = hashmap__set(ctx->ids, name, data_ptr,
 			   (const void **)&old_key, (void **)&old_data);
 	if (ret)
 		free(data_ptr);
@@ -156,9 +223,24 @@ int expr__add_ref(struct expr_parse_ctx *ctx, struct metric_ref *ref)
 int expr__get_id(struct expr_parse_ctx *ctx, const char *id,
 		 struct expr_id_data **data)
 {
-	return hashmap__find(&ctx->ids, id, (void **)data) ? 0 : -1;
+	return hashmap__find(ctx->ids, id, (void **)data) ? 0 : -1;
+}
+
+bool expr__subset_of_ids(struct expr_parse_ctx *haystack,
+			 struct expr_parse_ctx *needles)
+{
+	struct hashmap_entry *cur;
+	size_t bkt;
+	struct expr_id_data *data;
+
+	hashmap__for_each_entry(needles->ids, cur, bkt) {
+		if (expr__get_id(haystack, cur->key, &data))
+			return false;
+	}
+	return true;
 }
 
+
 int expr__resolve_id(struct expr_parse_ctx *ctx, const char *id,
 		     struct expr_id_data **datap)
 {
@@ -173,21 +255,18 @@ int expr__resolve_id(struct expr_parse_ctx *ctx, const char *id,
 
 	switch (data->kind) {
 	case EXPR_ID_DATA__VALUE:
-		pr_debug2("lookup(%s): val %f\n", id, data->val);
-		break;
-	case EXPR_ID_DATA__PARENT:
-		pr_debug2("lookup(%s): parent %s\n", id, data->parent->id);
+		pr_debug2("lookup(%s): val %f\n", id, data->val.val);
 		break;
 	case EXPR_ID_DATA__REF:
 		pr_debug2("lookup(%s): ref metric name %s\n", id,
 			data->ref.metric_name);
 		pr_debug("processing metric: %s ENTRY\n", id);
 		data->kind = EXPR_ID_DATA__REF_VALUE;
-		if (expr__parse(&data->ref.val, ctx, data->ref.metric_expr, 1)) {
+		if (expr__parse(&data->ref.val, ctx, data->ref.metric_expr)) {
 			pr_debug("%s failed to count\n", id);
 			return -1;
 		}
-		pr_debug("processing metric: %s EXIT: %f\n", id, data->val);
+		pr_debug("processing metric: %s EXIT: %f\n", id, data->ref.val);
 		break;
 	case EXPR_ID_DATA__REF_VALUE:
 		pr_debug2("lookup(%s): ref val %f metric name %s\n", id,
@@ -205,15 +284,24 @@ void expr__del_id(struct expr_parse_ctx *ctx, const char *id)
 	struct expr_id_data *old_val = NULL;
 	char *old_key = NULL;
 
-	hashmap__delete(&ctx->ids, id,
+	hashmap__delete(ctx->ids, id,
 			(const void **)&old_key, (void **)&old_val);
 	free(old_key);
 	free(old_val);
 }
 
-void expr__ctx_init(struct expr_parse_ctx *ctx)
+struct expr_parse_ctx *expr__ctx_new(void)
 {
-	hashmap__init(&ctx->ids, key_hash, key_equal, NULL);
+	struct expr_parse_ctx *ctx;
+
+	ctx = malloc(sizeof(struct expr_parse_ctx));
+	if (!ctx)
+		return NULL;
+
+	ctx->ids = hashmap__new(key_hash, key_equal, NULL);
+	ctx->runtime = 0;
+
+	return ctx;
 }
 
 void expr__ctx_clear(struct expr_parse_ctx *ctx)
@@ -221,20 +309,32 @@ void expr__ctx_clear(struct expr_parse_ctx *ctx)
 	struct hashmap_entry *cur;
 	size_t bkt;
 
-	hashmap__for_each_entry((&ctx->ids), cur, bkt) {
+	hashmap__for_each_entry(ctx->ids, cur, bkt) {
 		free((char *)cur->key);
 		free(cur->value);
 	}
-	hashmap__clear(&ctx->ids);
+	hashmap__clear(ctx->ids);
+}
+
+void expr__ctx_free(struct expr_parse_ctx *ctx)
+{
+	struct hashmap_entry *cur;
+	size_t bkt;
+
+	hashmap__for_each_entry(ctx->ids, cur, bkt) {
+		free((char *)cur->key);
+		free(cur->value);
+	}
+	hashmap__free(ctx->ids);
+	free(ctx);
 }
 
 static int
 __expr__parse(double *val, struct expr_parse_ctx *ctx, const char *expr,
-	      int start, int runtime)
+	      bool compute_ids)
 {
 	struct expr_scanner_ctx scanner_ctx = {
-		.start_token = start,
-		.runtime = runtime,
+		.runtime = ctx->runtime,
 	};
 	YY_BUFFER_STATE buffer;
 	void *scanner;
@@ -253,7 +353,7 @@ __expr__parse(double *val, struct expr_parse_ctx *ctx, const char *expr,
 	expr_set_debug(1, scanner);
 #endif
 
-	ret = expr_parse(val, ctx, scanner);
+	ret = expr_parse(val, ctx, compute_ids, scanner);
 
 	expr__flush_buffer(buffer, scanner);
 	expr__delete_buffer(buffer, scanner);
@@ -262,15 +362,15 @@ __expr__parse(double *val, struct expr_parse_ctx *ctx, const char *expr,
 }
 
 int expr__parse(double *final_val, struct expr_parse_ctx *ctx,
-		const char *expr, int runtime)
+		const char *expr)
 {
-	return __expr__parse(final_val, ctx, expr, EXPR_PARSE, runtime) ? -1 : 0;
+	return __expr__parse(final_val, ctx, expr, /*compute_ids=*/false) ? -1 : 0;
 }
 
-int expr__find_other(const char *expr, const char *one,
-		     struct expr_parse_ctx *ctx, int runtime)
+int expr__find_ids(const char *expr, const char *one,
+		   struct expr_parse_ctx *ctx)
 {
-	int ret = __expr__parse(NULL, ctx, expr, EXPR_OTHER, runtime);
+	int ret = __expr__parse(NULL, ctx, expr, /*compute_ids=*/true);
 
 	if (one)
 		expr__del_id(ctx, one);
@@ -281,13 +381,47 @@ int expr__find_other(const char *expr, const char *one,
 double expr_id_data__value(const struct expr_id_data *data)
 {
 	if (data->kind == EXPR_ID_DATA__VALUE)
-		return data->val;
+		return data->val.val;
 	assert(data->kind == EXPR_ID_DATA__REF_VALUE);
 	return data->ref.val;
 }
 
-struct expr_id *expr_id_data__parent(struct expr_id_data *data)
+double expr_id_data__source_count(const struct expr_id_data *data)
 {
-	assert(data->kind == EXPR_ID_DATA__PARENT);
-	return data->parent;
+	assert(data->kind == EXPR_ID_DATA__VALUE);
+	return data->val.source_count;
+}
+
+double expr__get_literal(const char *literal)
+{
+	static struct cpu_topology *topology;
+
+	if (!strcmp("#smt_on", literal))
+		return smt_on() > 0 ? 1.0 : 0.0;
+
+	if (!strcmp("#num_cpus", literal))
+		return cpu__max_present_cpu();
+
+	/*
+	 * Assume that topology strings are consistent, such as CPUs "0-1"
+	 * wouldn't be listed as "0,1", and so after deduplication the number of
+	 * these strings gives an indication of the number of packages, dies,
+	 * etc.
+	 */
+	if (!topology) {
+		topology = cpu_topology__new();
+		if (!topology) {
+			pr_err("Error creating CPU topology");
+			return NAN;
+		}
+	}
+	if (!strcmp("#num_packages", literal))
+		return topology->package_cpus_lists;
+	if (!strcmp("#num_dies", literal))
+		return topology->die_cpus_lists;
+	if (!strcmp("#num_cores", literal))
+		return topology->core_cpus_lists;
+
+	pr_err("Unrecognized literal '%s'", literal);
+	return NAN;
 }
diff --git a/tools/perf/util/expr.h b/tools/perf/util/expr.h
index 85df3e4771e4..bd2116983bbb 100644
--- a/tools/perf/util/expr.h
+++ b/tools/perf/util/expr.h
@@ -13,39 +13,51 @@
 
 struct metric_ref;
 
-struct expr_id {
-	char		*id;
-	struct expr_id	*parent;
-};
-
 struct expr_parse_ctx {
-	struct hashmap	 ids;
-	struct expr_id	*parent;
+	struct hashmap	*ids;
+	int runtime;
 };
 
 struct expr_id_data;
 
 struct expr_scanner_ctx {
-	int start_token;
 	int runtime;
 };
 
-void expr__ctx_init(struct expr_parse_ctx *ctx);
+struct hashmap *ids__new(void);
+void ids__free(struct hashmap *ids);
+int ids__insert(struct hashmap *ids, const char *id);
+/*
+ * Union two sets of ids (hashmaps) and construct a third, freeing ids1 and
+ * ids2.
+ */
+struct hashmap *ids__union(struct hashmap *ids1, struct hashmap *ids2);
+
+struct expr_parse_ctx *expr__ctx_new(void);
 void expr__ctx_clear(struct expr_parse_ctx *ctx);
+void expr__ctx_free(struct expr_parse_ctx *ctx);
+
 void expr__del_id(struct expr_parse_ctx *ctx, const char *id);
 int expr__add_id(struct expr_parse_ctx *ctx, const char *id);
 int expr__add_id_val(struct expr_parse_ctx *ctx, const char *id, double val);
+int expr__add_id_val_source_count(struct expr_parse_ctx *ctx, const char *id,
+				double val, int source_count);
 int expr__add_ref(struct expr_parse_ctx *ctx, struct metric_ref *ref);
 int expr__get_id(struct expr_parse_ctx *ctx, const char *id,
 		 struct expr_id_data **data);
+bool expr__subset_of_ids(struct expr_parse_ctx *haystack,
+			 struct expr_parse_ctx *needles);
 int expr__resolve_id(struct expr_parse_ctx *ctx, const char *id,
 		     struct expr_id_data **datap);
+
 int expr__parse(double *final_val, struct expr_parse_ctx *ctx,
-		const char *expr, int runtime);
-int expr__find_other(const char *expr, const char *one,
-		struct expr_parse_ctx *ids, int runtime);
+		const char *expr);
+
+int expr__find_ids(const char *expr, const char *one,
+		   struct expr_parse_ctx *ids);
 
 double expr_id_data__value(const struct expr_id_data *data);
-struct expr_id *expr_id_data__parent(struct expr_id_data *data);
+double expr_id_data__source_count(const struct expr_id_data *data);
+double expr__get_literal(const char *literal);
 
 #endif
diff --git a/tools/perf/util/expr.l b/tools/perf/util/expr.l
index 13e5e3c75f56..0a13eb20c814 100644
--- a/tools/perf/util/expr.l
+++ b/tools/perf/util/expr.l
@@ -6,6 +6,7 @@
 #include <linux/compiler.h>
 #include "expr.h"
 #include "expr-bison.h"
+#include <math.h>
 
 char *expr_get_text(yyscan_t yyscanner);
 YYSTYPE *expr_get_lval(yyscan_t yyscanner);
@@ -41,11 +42,9 @@ static char *normalize(char *str, int runtime)
 	char *dst = str;
 
 	while (*str) {
-		if (*str == '@')
-			*dst++ = '/';
-		else if (*str == '\\')
+		if (*str == '\\')
 			*dst++ = *++str;
-		 else if (*str == '?') {
+		else if (*str == '?') {
 			char *paramval;
 			int i = 0;
 			int size = asprintf(&paramval, "%d", runtime);
@@ -79,6 +78,17 @@ static int str(yyscan_t scanner, int token, int runtime)
 	yylval->str = normalize(yylval->str, runtime);
 	return token;
 }
+
+static int literal(yyscan_t scanner)
+{
+	YYSTYPE *yylval = expr_get_lval(scanner);
+
+	yylval->num = expr__get_literal(expr_get_text(scanner));
+	if (isnan(yylval->num))
+		return EXPR_ERROR;
+
+	return LITERAL;
+}
 %}
 
 number		([0-9]+\.?[0-9]*|[0-9]*\.?[0-9]+)
@@ -87,25 +97,18 @@ sch		[-,=]
 spec		\\{sch}
 sym		[0-9a-zA-Z_\.:@?]+
 symbol		({spec}|{sym})+
+literal		#[0-9a-zA-Z_\.\-]+
 
 %%
 	struct expr_scanner_ctx *sctx = expr_get_extra(yyscanner);
 
-	{
-		int start_token = sctx->start_token;
-
-		if (sctx->start_token) {
-			sctx->start_token = 0;
-			return start_token;
-		}
-	}
-
 d_ratio		{ return D_RATIO; }
 max		{ return MAX; }
 min		{ return MIN; }
 if		{ return IF; }
 else		{ return ELSE; }
-#smt_on		{ return SMT_ON; }
+source_count	{ return SOURCE_COUNT; }
+{literal}	{ return literal(yyscanner); }
 {number}	{ return value(yyscanner); }
 {symbol}	{ return str(yyscanner, ID, sctx->runtime); }
 "|"		{ return '|'; }
diff --git a/tools/perf/util/expr.y b/tools/perf/util/expr.y
index b2ada8f8309a..a30b825adb7b 100644
--- a/tools/perf/util/expr.y
+++ b/tools/perf/util/expr.y
@@ -1,42 +1,43 @@
 /* Simple expression parser */
 %{
 #define YYDEBUG 1
-#include <stdio.h>
-#include "util.h"
+#include <assert.h>
+#include <math.h>
+#include <stdlib.h>
 #include "util/debug.h"
-#include <stdlib.h> // strtod()
 #define IN_EXPR_Y 1
 #include "expr.h"
-#include "smt.h"
-#include <string.h>
-
-static double d_ratio(double val0, double val1)
-{
-	if (val1 == 0) {
-		return 0;
-	}
-	return  val0 / val1;
-}
-
 %}
 
 %define api.pure full
 
 %parse-param { double *final_val }
 %parse-param { struct expr_parse_ctx *ctx }
+%parse-param { bool compute_ids }
 %parse-param {void *scanner}
 %lex-param {void* scanner}
 
 %union {
 	double	 num;
 	char	*str;
+	struct ids {
+		/*
+		 * When creating ids, holds the working set of event ids. NULL
+		 * implies the set is empty.
+		 */
+		struct hashmap *ids;
+		/*
+		 * The metric value. When not creating ids this is the value
+		 * read from a counter, a constant or some computed value. When
+		 * creating ids the value is either a constant or BOTTOM. NAN is
+		 * used as the special BOTTOM value, representing a "set of all
+		 * values" case.
+		 */
+		double val;
+	} ids;
 }
 
-%token EXPR_PARSE EXPR_OTHER EXPR_ERROR
-%token <num> NUMBER
-%token <str> ID
-%destructor { free ($$); } <str>
-%token MIN MAX IF ELSE SMT_ON D_RATIO
+%token ID NUMBER MIN MAX IF ELSE LITERAL D_RATIO SOURCE_COUNT EXPR_ERROR
 %left MIN MAX IF
 %left '|'
 %left '^'
@@ -45,83 +46,256 @@ static double d_ratio(double val0, double val1)
 %left '-' '+'
 %left '*' '/' '%'
 %left NEG NOT
-%type <num> expr if_expr
+%type <num> NUMBER LITERAL
+%type <str> ID
+%destructor { free ($$); } <str>
+%type <ids> expr if_expr
+%destructor { ids__free($$.ids); } <ids>
 
 %{
 static void expr_error(double *final_val __maybe_unused,
 		       struct expr_parse_ctx *ctx __maybe_unused,
+		       bool compute_ids __maybe_unused,
 		       void *scanner,
 		       const char *s)
 {
 	pr_debug("%s\n", s);
 }
 
+/*
+ * During compute ids, the special "bottom" value uses NAN to represent the set
+ * of all values. NAN is selected as it isn't a useful constant value.
+ */
+#define BOTTOM NAN
+
+/* During computing ids, does val represent a constant (non-BOTTOM) value? */
+static bool is_const(double val)
+{
+	return isfinite(val);
+}
+
+static struct ids union_expr(struct ids ids1, struct ids ids2)
+{
+	struct ids result = {
+		.val = BOTTOM,
+		.ids = ids__union(ids1.ids, ids2.ids),
+	};
+	return result;
+}
+
+static struct ids handle_id(struct expr_parse_ctx *ctx, char *id,
+			    bool compute_ids, bool source_count)
+{
+	struct ids result;
+
+	if (!compute_ids) {
+		/*
+		 * Compute the event's value from ID. If the ID isn't known then
+		 * it isn't used to compute the formula so set to NAN.
+		 */
+		struct expr_id_data *data;
+
+		result.val = NAN;
+		if (expr__resolve_id(ctx, id, &data) == 0) {
+			result.val = source_count
+				? expr_id_data__source_count(data)
+				: expr_id_data__value(data);
+		}
+		result.ids = NULL;
+		free(id);
+	} else {
+		/*
+		 * Set the value to BOTTOM to show that any value is possible
+		 * when the event is computed. Create a set of just the ID.
+		 */
+		result.val = BOTTOM;
+		result.ids = ids__new();
+		if (!result.ids || ids__insert(result.ids, id)) {
+			pr_err("Error creating IDs for '%s'", id);
+			free(id);
+		}
+	}
+	return result;
+}
+
+/*
+ * If we're not computing ids or $1 and $3 are constants, compute the new
+ * constant value using OP. Its invariant that there are no ids.  If computing
+ * ids for non-constants union the set of IDs that must be computed.
+ */
+#define BINARY_LONG_OP(RESULT, OP, LHS, RHS)				\
+	if (!compute_ids || (is_const(LHS.val) && is_const(RHS.val))) { \
+		assert(LHS.ids == NULL);				\
+		assert(RHS.ids == NULL);				\
+		RESULT.val = (long)LHS.val OP (long)RHS.val;		\
+		RESULT.ids = NULL;					\
+	} else {							\
+	        RESULT = union_expr(LHS, RHS);				\
+	}
+
+#define BINARY_OP(RESULT, OP, LHS, RHS)					\
+	if (!compute_ids || (is_const(LHS.val) && is_const(RHS.val))) { \
+		assert(LHS.ids == NULL);				\
+		assert(RHS.ids == NULL);				\
+		RESULT.val = LHS.val OP RHS.val;			\
+		RESULT.ids = NULL;					\
+	} else {							\
+	        RESULT = union_expr(LHS, RHS);				\
+	}
+
 %}
 %%
 
-start:
-EXPR_PARSE all_expr
-|
-EXPR_OTHER all_other
+start: if_expr
+{
+	if (compute_ids)
+		ctx->ids = ids__union($1.ids, ctx->ids);
+
+	if (final_val)
+		*final_val = $1.val;
+}
+;
 
-all_other: all_other other
-|
+if_expr: expr IF expr ELSE expr
+{
+	if (fpclassify($3.val) == FP_ZERO) {
+		/*
+		 * The IF expression evaluated to 0 so treat as false, take the
+		 * ELSE and discard everything else.
+		 */
+		$$.val = $5.val;
+		$$.ids = $5.ids;
+		ids__free($1.ids);
+		ids__free($3.ids);
+	} else if (!compute_ids || is_const($3.val)) {
+		/*
+		 * If ids aren't computed then treat the expression as true. If
+		 * ids are being computed and the IF expr is a non-zero
+		 * constant, then also evaluate the true case.
+		 */
+		$$.val = $1.val;
+		$$.ids = $1.ids;
+		ids__free($3.ids);
+		ids__free($5.ids);
+	} else if ($1.val == $5.val) {
+		/*
+		 * LHS == RHS, so both are an identical constant. No need to
+		 * evaluate any events.
+		 */
+		$$.val = $1.val;
+		$$.ids = NULL;
+		ids__free($1.ids);
+		ids__free($3.ids);
+		ids__free($5.ids);
+	} else {
+		/*
+		 * Value is either the LHS or RHS and we need the IF expression
+		 * to compute it.
+		 */
+		$$ = union_expr($1, union_expr($3, $5));
+	}
+}
+| expr
+;
 
-other: ID
+expr: NUMBER
+{
+	$$.val = $1;
+	$$.ids = NULL;
+}
+| ID				{ $$ = handle_id(ctx, $1, compute_ids, /*source_count=*/false); }
+| SOURCE_COUNT '(' ID ')'	{ $$ = handle_id(ctx, $3, compute_ids, /*source_count=*/true); }
+| expr '|' expr { BINARY_LONG_OP($$, |, $1, $3); }
+| expr '&' expr { BINARY_LONG_OP($$, &, $1, $3); }
+| expr '^' expr { BINARY_LONG_OP($$, ^, $1, $3); }
+| expr '<' expr { BINARY_OP($$, <, $1, $3); }
+| expr '>' expr { BINARY_OP($$, >, $1, $3); }
+| expr '+' expr { BINARY_OP($$, +, $1, $3); }
+| expr '-' expr { BINARY_OP($$, -, $1, $3); }
+| expr '*' expr { BINARY_OP($$, *, $1, $3); }
+| expr '/' expr
+{
+	if (fpclassify($3.val) == FP_ZERO) {
+		pr_debug("division by zero\n");
+		YYABORT;
+	} else if (!compute_ids || (is_const($1.val) && is_const($3.val))) {
+		assert($1.ids == NULL);
+		assert($3.ids == NULL);
+		$$.val = $1.val / $3.val;
+		$$.ids = NULL;
+	} else {
+		/* LHS and/or RHS need computing from event IDs so union. */
+		$$ = union_expr($1, $3);
+	}
+}
+| expr '%' expr
+{
+	if (fpclassify($3.val) == FP_ZERO) {
+		pr_debug("division by zero\n");
+		YYABORT;
+	} else if (!compute_ids || (is_const($1.val) && is_const($3.val))) {
+		assert($1.ids == NULL);
+		assert($3.ids == NULL);
+		$$.val = (long)$1.val % (long)$3.val;
+		$$.ids = NULL;
+	} else {
+		/* LHS and/or RHS need computing from event IDs so union. */
+		$$ = union_expr($1, $3);
+	}
+}
+| D_RATIO '(' expr ',' expr ')'
+{
+	if (fpclassify($5.val) == FP_ZERO) {
+		/*
+		 * Division by constant zero always yields zero and no events
+		 * are necessary.
+		 */
+		assert($5.ids == NULL);
+		$$.val = 0.0;
+		$$.ids = NULL;
+		ids__free($3.ids);
+	} else if (!compute_ids || (is_const($3.val) && is_const($5.val))) {
+		assert($3.ids == NULL);
+		assert($5.ids == NULL);
+		$$.val = $3.val / $5.val;
+		$$.ids = NULL;
+	} else {
+		/* LHS and/or RHS need computing from event IDs so union. */
+		$$ = union_expr($3, $5);
+	}
+}
+| '-' expr %prec NEG
+{
+	$$.val = -$2.val;
+	$$.ids = $2.ids;
+}
+| '(' if_expr ')'
 {
-	expr__add_id(ctx, $1);
-}
-|
-MIN | MAX | IF | ELSE | SMT_ON | NUMBER | '|' | '^' | '&' | '-' | '+' | '*' | '/' | '%' | '(' | ')' | ','
-|
-'<' | '>' | D_RATIO
-
-all_expr: if_expr			{ *final_val = $1; }
-	;
-
-if_expr:
-	expr IF expr ELSE expr { $$ = $3 ? $1 : $5; }
-	| expr
-	;
-
-expr:	  NUMBER
-	| ID			{
-					struct expr_id_data *data;
-
-					if (expr__resolve_id(ctx, $1, &data)) {
-						free($1);
-						YYABORT;
-					}
-
-					$$ = expr_id_data__value(data);
-					free($1);
-				}
-	| expr '|' expr		{ $$ = (long)$1 | (long)$3; }
-	| expr '&' expr		{ $$ = (long)$1 & (long)$3; }
-	| expr '^' expr		{ $$ = (long)$1 ^ (long)$3; }
-	| expr '<' expr		{ $$ = $1 < $3; }
-	| expr '>' expr		{ $$ = $1 > $3; }
-	| expr '+' expr		{ $$ = $1 + $3; }
-	| expr '-' expr		{ $$ = $1 - $3; }
-	| expr '*' expr		{ $$ = $1 * $3; }
-	| expr '/' expr		{ if ($3 == 0) {
-					pr_debug("division by zero\n");
-					YYABORT;
-				  }
-				  $$ = $1 / $3;
-	                        }
-	| expr '%' expr		{ if ((long)$3 == 0) {
-					pr_debug("division by zero\n");
-					YYABORT;
-				  }
-				  $$ = (long)$1 % (long)$3;
-	                        }
-	| '-' expr %prec NEG	{ $$ = -$2; }
-	| '(' if_expr ')'	{ $$ = $2; }
-	| MIN '(' expr ',' expr ')' { $$ = $3 < $5 ? $3 : $5; }
-	| MAX '(' expr ',' expr ')' { $$ = $3 > $5 ? $3 : $5; }
-	| SMT_ON		 { $$ = smt_on() > 0; }
-	| D_RATIO '(' expr ',' expr ')' { $$ = d_ratio($3,$5); }
-	;
+	$$ = $2;
+}
+| MIN '(' expr ',' expr ')'
+{
+	if (!compute_ids) {
+		$$.val = $3.val < $5.val ? $3.val : $5.val;
+		$$.ids = NULL;
+	} else {
+		$$ = union_expr($3, $5);
+	}
+}
+| MAX '(' expr ',' expr ')'
+{
+	if (!compute_ids) {
+		$$.val = $3.val > $5.val ? $3.val : $5.val;
+		$$.ids = NULL;
+	} else {
+		$$ = union_expr($3, $5);
+	}
+}
+| LITERAL
+{
+	$$.val = $1;
+	$$.ids = NULL;
+}
+;
 
 %%
diff --git a/tools/perf/util/genelf.h b/tools/perf/util/genelf.h
index d4137559be05..3db3293213a9 100644
--- a/tools/perf/util/genelf.h
+++ b/tools/perf/util/genelf.h
@@ -42,7 +42,7 @@ int jit_add_debug_info(Elf *e, uint64_t code_addr, void *debug, int nr_debug_ent
 #error "unsupported architecture"
 #endif
 
-#if __BYTE_ORDER == __BIG_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
 #define GEN_ELF_ENDIAN	ELFDATA2MSB
 #else
 #define GEN_ELF_ENDIAN	ELFDATA2LSB
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 1c7414f66655..79cce216727e 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -48,6 +48,7 @@
 #include "util/util.h" // perf_exe()
 #include "cputopo.h"
 #include "bpf-event.h"
+#include "bpf-utils.h"
 #include "clockid.h"
 #include "pmu-hybrid.h"
 
@@ -582,21 +583,21 @@ static int write_cpu_topology(struct feat_fd *ff,
 	if (!tp)
 		return -1;
 
-	ret = do_write(ff, &tp->core_sib, sizeof(tp->core_sib));
+	ret = do_write(ff, &tp->package_cpus_lists, sizeof(tp->package_cpus_lists));
 	if (ret < 0)
 		goto done;
 
-	for (i = 0; i < tp->core_sib; i++) {
-		ret = do_write_string(ff, tp->core_siblings[i]);
+	for (i = 0; i < tp->package_cpus_lists; i++) {
+		ret = do_write_string(ff, tp->package_cpus_list[i]);
 		if (ret < 0)
 			goto done;
 	}
-	ret = do_write(ff, &tp->thread_sib, sizeof(tp->thread_sib));
+	ret = do_write(ff, &tp->core_cpus_lists, sizeof(tp->core_cpus_lists));
 	if (ret < 0)
 		goto done;
 
-	for (i = 0; i < tp->thread_sib; i++) {
-		ret = do_write_string(ff, tp->thread_siblings[i]);
+	for (i = 0; i < tp->core_cpus_lists; i++) {
+		ret = do_write_string(ff, tp->core_cpus_list[i]);
 		if (ret < 0)
 			break;
 	}
@@ -616,15 +617,15 @@ static int write_cpu_topology(struct feat_fd *ff,
 			return ret;
 	}
 
-	if (!tp->die_sib)
+	if (!tp->die_cpus_lists)
 		goto done;
 
-	ret = do_write(ff, &tp->die_sib, sizeof(tp->die_sib));
+	ret = do_write(ff, &tp->die_cpus_lists, sizeof(tp->die_cpus_lists));
 	if (ret < 0)
 		goto done;
 
-	for (i = 0; i < tp->die_sib; i++) {
-		ret = do_write_string(ff, tp->die_siblings[i]);
+	for (i = 0; i < tp->die_cpus_lists; i++) {
+		ret = do_write_string(ff, tp->die_cpus_list[i]);
 		if (ret < 0)
 			goto done;
 	}
@@ -1006,17 +1007,17 @@ static int write_bpf_prog_info(struct feat_fd *ff,
 
 		node = rb_entry(next, struct bpf_prog_info_node, rb_node);
 		next = rb_next(&node->rb_node);
-		len = sizeof(struct bpf_prog_info_linear) +
+		len = sizeof(struct perf_bpil) +
 			node->info_linear->data_len;
 
 		/* before writing to file, translate address to offset */
-		bpf_program__bpil_addr_to_offs(node->info_linear);
+		bpil_addr_to_offs(node->info_linear);
 		ret = do_write(ff, node->info_linear, len);
 		/*
 		 * translate back to address even when do_write() fails,
 		 * so that this function never changes the data.
 		 */
-		bpf_program__bpil_offs_to_addr(node->info_linear);
+		bpil_offs_to_addr(node->info_linear);
 		if (ret < 0)
 			goto out;
 	}
@@ -3018,9 +3019,9 @@ static int process_dir_format(struct feat_fd *ff,
 #ifdef HAVE_LIBBPF_SUPPORT
 static int process_bpf_prog_info(struct feat_fd *ff, void *data __maybe_unused)
 {
-	struct bpf_prog_info_linear *info_linear;
 	struct bpf_prog_info_node *info_node;
 	struct perf_env *env = &ff->ph->env;
+	struct perf_bpil *info_linear;
 	u32 count, i;
 	int err = -1;
 
@@ -3049,7 +3050,7 @@ static int process_bpf_prog_info(struct feat_fd *ff, void *data __maybe_unused)
 			goto out;
 		}
 
-		info_linear = malloc(sizeof(struct bpf_prog_info_linear) +
+		info_linear = malloc(sizeof(struct perf_bpil) +
 				     data_len);
 		if (!info_linear)
 			goto out;
@@ -3071,7 +3072,7 @@ static int process_bpf_prog_info(struct feat_fd *ff, void *data __maybe_unused)
 			goto out;
 
 		/* after reading from file, translate offset to address */
-		bpf_program__bpil_offs_to_addr(info_linear);
+		bpil_offs_to_addr(info_linear);
 		info_node->info_linear = info_linear;
 		perf_env__insert_bpf_prog_info(env, info_node);
 	}
@@ -4256,9 +4257,11 @@ int perf_event__process_event_update(struct perf_tool *tool __maybe_unused,
 
 	switch (ev->type) {
 	case PERF_EVENT_UPDATE__UNIT:
+		free((char *)evsel->unit);
 		evsel->unit = strdup(ev->data);
 		break;
 	case PERF_EVENT_UPDATE__NAME:
+		free(evsel->name);
 		evsel->name = strdup(ev->data);
 		break;
 	case PERF_EVENT_UPDATE__SCALE:
@@ -4267,11 +4270,11 @@ int perf_event__process_event_update(struct perf_tool *tool __maybe_unused,
 		break;
 	case PERF_EVENT_UPDATE__CPUS:
 		ev_cpus = (struct perf_record_event_update_cpus *)ev->data;
-
 		map = cpu_map__new_data(&ev_cpus->cpus);
-		if (map)
+		if (map) {
+			perf_cpu_map__put(evsel->core.own_cpus);
 			evsel->core.own_cpus = map;
-		else
+		} else
 			pr_err("failed to get event_update cpus\n");
 	default:
 		break;
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
index 65fe65ba03c2..b776465e04ef 100644
--- a/tools/perf/util/hist.c
+++ b/tools/perf/util/hist.c
@@ -289,15 +289,10 @@ static long hist_time(unsigned long htime)
 	return htime;
 }
 
-static void he_stat__add_period(struct he_stat *he_stat, u64 period,
-				u64 weight, u64 ins_lat, u64 p_stage_cyc)
+static void he_stat__add_period(struct he_stat *he_stat, u64 period)
 {
-
 	he_stat->period		+= period;
-	he_stat->weight		+= weight;
 	he_stat->nr_events	+= 1;
-	he_stat->ins_lat	+= ins_lat;
-	he_stat->p_stage_cyc	+= p_stage_cyc;
 }
 
 static void he_stat__add_stat(struct he_stat *dest, struct he_stat *src)
@@ -308,9 +303,6 @@ static void he_stat__add_stat(struct he_stat *dest, struct he_stat *src)
 	dest->period_guest_sys	+= src->period_guest_sys;
 	dest->period_guest_us	+= src->period_guest_us;
 	dest->nr_events		+= src->nr_events;
-	dest->weight		+= src->weight;
-	dest->ins_lat		+= src->ins_lat;
-	dest->p_stage_cyc		+= src->p_stage_cyc;
 }
 
 static void he_stat__decay(struct he_stat *he_stat)
@@ -598,9 +590,6 @@ static struct hist_entry *hists__findnew_entry(struct hists *hists,
 	struct hist_entry *he;
 	int64_t cmp;
 	u64 period = entry->stat.period;
-	u64 weight = entry->stat.weight;
-	u64 ins_lat = entry->stat.ins_lat;
-	u64 p_stage_cyc = entry->stat.p_stage_cyc;
 	bool leftmost = true;
 
 	p = &hists->entries_in->rb_root.rb_node;
@@ -619,11 +608,11 @@ static struct hist_entry *hists__findnew_entry(struct hists *hists,
 
 		if (!cmp) {
 			if (sample_self) {
-				he_stat__add_period(&he->stat, period, weight, ins_lat, p_stage_cyc);
+				he_stat__add_period(&he->stat, period);
 				hist_entry__add_callchain_period(he, period);
 			}
 			if (symbol_conf.cumulate_callchain)
-				he_stat__add_period(he->stat_acc, period, weight, ins_lat, p_stage_cyc);
+				he_stat__add_period(he->stat_acc, period);
 
 			/*
 			 * This mem info was allocated from sample__resolve_mem
@@ -733,9 +722,6 @@ __hists__add_entry(struct hists *hists,
 		.stat = {
 			.nr_events = 1,
 			.period	= sample->period,
-			.weight = sample->weight,
-			.ins_lat = sample->ins_lat,
-			.p_stage_cyc = sample->p_stage_cyc,
 		},
 		.parent = sym_parent,
 		.filtered = symbol__parent_filter(sym_parent) | al->filtered,
@@ -748,6 +734,9 @@ __hists__add_entry(struct hists *hists,
 		.raw_size = sample->raw_size,
 		.ops = ops,
 		.time = hist_time(sample->time),
+		.weight = sample->weight,
+		.ins_lat = sample->ins_lat,
+		.p_stage_cyc = sample->p_stage_cyc,
 	}, *he = hists__findnew_entry(hists, &entry, al, sample_self);
 
 	if (!hists->has_callchains && he && he->callchain_size != 0)
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
index 5343b62476e6..621f35ae1efa 100644
--- a/tools/perf/util/hist.h
+++ b/tools/perf/util/hist.h
@@ -369,7 +369,6 @@ enum {
 };
 
 void perf_hpp__init(void);
-void perf_hpp__column_unregister(struct perf_hpp_fmt *format);
 void perf_hpp__cancel_cumulate(void);
 void perf_hpp__setup_output_field(struct perf_hpp_list *list);
 void perf_hpp__reset_output_field(struct perf_hpp_list *list);
diff --git a/tools/perf/util/intel-bts.c b/tools/perf/util/intel-bts.c
index af1e78d76228..2c8147a62203 100644
--- a/tools/perf/util/intel-bts.c
+++ b/tools/perf/util/intel-bts.c
@@ -35,7 +35,7 @@
 #define INTEL_BTS_ERR_NOINSN  5
 #define INTEL_BTS_ERR_LOST    9
 
-#if __BYTE_ORDER == __BIG_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
 #define le64_to_cpu bswap_64
 #else
 #define le64_to_cpu
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 5ab631702769..5f83937bf8f3 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -608,6 +608,7 @@ static inline void intel_pt_update_sample_time(struct intel_pt_decoder *decoder)
 {
 	decoder->sample_timestamp = decoder->timestamp;
 	decoder->sample_insn_cnt = decoder->timestamp_insn_cnt;
+	decoder->state.cycles = decoder->tot_cyc_cnt;
 }
 
 static void intel_pt_reposition(struct intel_pt_decoder *decoder)
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
index 4b5e79fcf557..8fd68f7a0963 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
@@ -218,6 +218,7 @@ struct intel_pt_state {
 	uint64_t to_ip;
 	uint64_t tot_insn_cnt;
 	uint64_t tot_cyc_cnt;
+	uint64_t cycles;
 	uint64_t timestamp;
 	uint64_t est_timestamp;
 	uint64_t trace_nr;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c
index 593f20e9774c..9d5e65cec89b 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c
@@ -143,7 +143,7 @@ static void intel_pt_insn_decoder(struct insn *insn,
 
 	if (branch == INTEL_PT_BR_CONDITIONAL ||
 	    branch == INTEL_PT_BR_UNCONDITIONAL) {
-#if __BYTE_ORDER == __BIG_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
 		switch (insn->immediate.nbytes) {
 		case 1:
 			intel_pt_insn->rel = insn->immediate.value;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-log.c b/tools/perf/util/intel-pt-decoder/intel-pt-log.c
index 09feb5b07d32..5f5dfc8753f3 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-log.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-log.c
@@ -82,10 +82,10 @@ static int intel_pt_log_open(void)
 	if (f)
 		return 0;
 
-	if (!log_name[0])
-		return -1;
-
-	f = fopen(log_name, "w+");
+	if (log_name[0])
+		f = fopen(log_name, "w+");
+	else
+		f = stdout;
 	if (!f) {
 		intel_pt_enable_logging = false;
 		return -1;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
index 02a3395d6ce3..4bd154848cad 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
@@ -16,7 +16,7 @@
 
 #define BIT63		((uint64_t)1 << 63)
 
-#if __BYTE_ORDER == __BIG_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
 #define le16_to_cpu bswap_16
 #define le32_to_cpu bswap_32
 #define le64_to_cpu bswap_64
diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 6f852b305e92..556a893508da 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -111,6 +111,7 @@ struct intel_pt {
 	u64 cbr_id;
 	u64 psb_id;
 
+	bool single_pebs;
 	bool sample_pebs;
 	struct evsel *pebs_evsel;
 
@@ -148,6 +149,14 @@ enum switch_state {
 	INTEL_PT_SS_EXPECTING_SWITCH_IP,
 };
 
+/* applicable_counters is 64-bits */
+#define INTEL_PT_MAX_PEBS 64
+
+struct intel_pt_pebs_event {
+	struct evsel *evsel;
+	u64 id;
+};
+
 struct intel_pt_queue {
 	struct intel_pt *pt;
 	unsigned int queue_nr;
@@ -163,6 +172,7 @@ struct intel_pt_queue {
 	bool step_through_buffers;
 	bool use_buffer_pid_tid;
 	bool sync_switch;
+	bool sample_ipc;
 	pid_t pid, tid;
 	int cpu;
 	int switch_state;
@@ -189,6 +199,7 @@ struct intel_pt_queue {
 	u64 last_br_cyc_cnt;
 	unsigned int cbr_seen;
 	char insn[INTEL_PT_INSN_BUF_SZ];
+	struct intel_pt_pebs_event pebs[INTEL_PT_MAX_PEBS];
 };
 
 static void intel_pt_dump(struct intel_pt *pt __maybe_unused,
@@ -1571,7 +1582,7 @@ static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq)
 		sample.branch_stack = (struct branch_stack *)&dummy_bs;
 	}
 
-	if (ptq->state->flags & INTEL_PT_SAMPLE_IPC)
+	if (ptq->sample_ipc)
 		sample.cyc_cnt = ptq->ipc_cyc_cnt - ptq->last_br_cyc_cnt;
 	if (sample.cyc_cnt) {
 		sample.insn_cnt = ptq->ipc_insn_cnt - ptq->last_br_insn_cnt;
@@ -1622,7 +1633,7 @@ static int intel_pt_synth_instruction_sample(struct intel_pt_queue *ptq)
 	else
 		sample.period = ptq->state->tot_insn_cnt - ptq->last_insn_cnt;
 
-	if (ptq->state->flags & INTEL_PT_SAMPLE_IPC)
+	if (ptq->sample_ipc)
 		sample.cyc_cnt = ptq->ipc_cyc_cnt - ptq->last_in_cyc_cnt;
 	if (sample.cyc_cnt) {
 		sample.insn_cnt = ptq->ipc_insn_cnt - ptq->last_in_insn_cnt;
@@ -1978,15 +1989,13 @@ static void intel_pt_add_lbrs(struct branch_stack *br_stack,
 	}
 }
 
-static int intel_pt_synth_pebs_sample(struct intel_pt_queue *ptq)
+static int intel_pt_do_synth_pebs_sample(struct intel_pt_queue *ptq, struct evsel *evsel, u64 id)
 {
 	const struct intel_pt_blk_items *items = &ptq->state->items;
 	struct perf_sample sample = { .ip = 0, };
 	union perf_event *event = ptq->event_buf;
 	struct intel_pt *pt = ptq->pt;
-	struct evsel *evsel = pt->pebs_evsel;
 	u64 sample_type = evsel->core.attr.sample_type;
-	u64 id = evsel->core.id[0];
 	u8 cpumode;
 	u64 regs[8 * sizeof(sample.intr_regs.mask)];
 
@@ -2112,6 +2121,45 @@ static int intel_pt_synth_pebs_sample(struct intel_pt_queue *ptq)
 	return intel_pt_deliver_synth_event(pt, event, &sample, sample_type);
 }
 
+static int intel_pt_synth_single_pebs_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	struct evsel *evsel = pt->pebs_evsel;
+	u64 id = evsel->core.id[0];
+
+	return intel_pt_do_synth_pebs_sample(ptq, evsel, id);
+}
+
+static int intel_pt_synth_pebs_sample(struct intel_pt_queue *ptq)
+{
+	const struct intel_pt_blk_items *items = &ptq->state->items;
+	struct intel_pt_pebs_event *pe;
+	struct intel_pt *pt = ptq->pt;
+	int err = -EINVAL;
+	int hw_id;
+
+	if (!items->has_applicable_counters || !items->applicable_counters) {
+		if (!pt->single_pebs)
+			pr_err("PEBS-via-PT record with no applicable_counters\n");
+		return intel_pt_synth_single_pebs_sample(ptq);
+	}
+
+	for_each_set_bit(hw_id, (unsigned long *)&items->applicable_counters, INTEL_PT_MAX_PEBS) {
+		pe = &ptq->pebs[hw_id];
+		if (!pe->evsel) {
+			if (!pt->single_pebs)
+				pr_err("PEBS-via-PT record with no matching event, hw_id %d\n",
+				       hw_id);
+			return intel_pt_synth_single_pebs_sample(ptq);
+		}
+		err = intel_pt_do_synth_pebs_sample(ptq, pe->evsel, pe->id);
+		if (err)
+			return err;
+	}
+
+	return err;
+}
+
 static int intel_pt_synth_error(struct intel_pt *pt, int code, int cpu,
 				pid_t pid, pid_t tid, u64 ip, u64 timestamp)
 {
@@ -2198,8 +2246,15 @@ static int intel_pt_sample(struct intel_pt_queue *ptq)
 
 	ptq->have_sample = false;
 
-	ptq->ipc_insn_cnt = ptq->state->tot_insn_cnt;
-	ptq->ipc_cyc_cnt = ptq->state->tot_cyc_cnt;
+	if (pt->synth_opts.approx_ipc) {
+		ptq->ipc_insn_cnt = ptq->state->tot_insn_cnt;
+		ptq->ipc_cyc_cnt = ptq->state->cycles;
+		ptq->sample_ipc = true;
+	} else {
+		ptq->ipc_insn_cnt = ptq->state->tot_insn_cnt;
+		ptq->ipc_cyc_cnt = ptq->state->tot_cyc_cnt;
+		ptq->sample_ipc = ptq->state->flags & INTEL_PT_SAMPLE_IPC;
+	}
 
 	/*
 	 * Do PEBS first to allow for the possibility that the PEBS timestamp
@@ -2882,6 +2937,30 @@ static int intel_pt_process_itrace_start(struct intel_pt *pt,
 					event->itrace_start.tid);
 }
 
+static int intel_pt_process_aux_output_hw_id(struct intel_pt *pt,
+					     union perf_event *event,
+					     struct perf_sample *sample)
+{
+	u64 hw_id = event->aux_output_hw_id.hw_id;
+	struct auxtrace_queue *queue;
+	struct intel_pt_queue *ptq;
+	struct evsel *evsel;
+
+	queue = auxtrace_queues__sample_queue(&pt->queues, sample, pt->session);
+	evsel = evlist__id2evsel_strict(pt->session->evlist, sample->id);
+	if (!queue || !queue->priv || !evsel || hw_id > INTEL_PT_MAX_PEBS) {
+		pr_err("Bad AUX output hardware ID\n");
+		return -EINVAL;
+	}
+
+	ptq = queue->priv;
+
+	ptq->pebs[hw_id].evsel = evsel;
+	ptq->pebs[hw_id].id = sample->id;
+
+	return 0;
+}
+
 static int intel_pt_find_map(struct thread *thread, u8 cpumode, u64 addr,
 			     struct addr_location *al)
 {
@@ -3009,6 +3088,8 @@ static int intel_pt_process_event(struct perf_session *session,
 		err = intel_pt_process_switch(pt, sample);
 	else if (event->header.type == PERF_RECORD_ITRACE_START)
 		err = intel_pt_process_itrace_start(pt, event, sample);
+	else if (event->header.type == PERF_RECORD_AUX_OUTPUT_HW_ID)
+		err = intel_pt_process_aux_output_hw_id(pt, event, sample);
 	else if (event->header.type == PERF_RECORD_SWITCH ||
 		 event->header.type == PERF_RECORD_SWITCH_CPU_WIDE)
 		err = intel_pt_context_switch(pt, event, sample);
@@ -3393,9 +3474,13 @@ static void intel_pt_setup_pebs_events(struct intel_pt *pt)
 
 	evlist__for_each_entry(pt->session->evlist, evsel) {
 		if (evsel->core.attr.aux_output && evsel->core.id) {
+			if (pt->single_pebs) {
+				pt->single_pebs = false;
+				return;
+			}
+			pt->single_pebs = true;
 			pt->sample_pebs = true;
 			pt->pebs_evsel = evsel;
-			return;
 		}
 	}
 }
@@ -3651,8 +3736,6 @@ int intel_pt_process_auxtrace_info(union perf_event *event,
 	if (err)
 		goto err_free;
 
-	intel_pt_log_set_name(INTEL_PT_PMU_NAME);
-
 	if (session->itrace_synth_opts->set) {
 		pt->synth_opts = *session->itrace_synth_opts;
 	} else {
@@ -3667,6 +3750,9 @@ int intel_pt_process_auxtrace_info(union perf_event *event,
 		pt->synth_opts.thread_stack = opts->thread_stack;
 	}
 
+	if (!(pt->synth_opts.log_plus_flags & AUXTRACE_LOG_FLG_USE_STDOUT))
+		intel_pt_log_set_name(INTEL_PT_PMU_NAME);
+
 	pt->session = session;
 	pt->machine = &session->machines.host; /* No kvm support */
 	pt->auxtrace_type = auxtrace_info->type;
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index 44e40bad0e33..fb8496df8432 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -755,6 +755,14 @@ int machine__process_itrace_start_event(struct machine *machine __maybe_unused,
 	return 0;
 }
 
+int machine__process_aux_output_hw_id_event(struct machine *machine __maybe_unused,
+					    union perf_event *event)
+{
+	if (dump_trace)
+		perf_event__fprintf_aux_output_hw_id(event, stdout);
+	return 0;
+}
+
 int machine__process_switch_event(struct machine *machine __maybe_unused,
 				  union perf_event *event)
 {
@@ -2028,6 +2036,8 @@ int machine__process_event(struct machine *machine, union perf_event *event,
 		ret = machine__process_bpf(machine, event, sample); break;
 	case PERF_RECORD_TEXT_POKE:
 		ret = machine__process_text_poke(machine, event, sample); break;
+	case PERF_RECORD_AUX_OUTPUT_HW_ID:
+		ret = machine__process_aux_output_hw_id_event(machine, event); break;
 	default:
 		ret = -1;
 		break;
diff --git a/tools/perf/util/machine.h b/tools/perf/util/machine.h
index 7377ed6efdf1..a143087eeb47 100644
--- a/tools/perf/util/machine.h
+++ b/tools/perf/util/machine.h
@@ -124,6 +124,8 @@ int machine__process_aux_event(struct machine *machine,
 			       union perf_event *event);
 int machine__process_itrace_start_event(struct machine *machine,
 					union perf_event *event);
+int machine__process_aux_output_hw_id_event(struct machine *machine,
+					    union perf_event *event);
 int machine__process_switch_event(struct machine *machine,
 				  union perf_event *event);
 int machine__process_namespaces_event(struct machine *machine,
diff --git a/tools/perf/util/metricgroup.c b/tools/perf/util/metricgroup.c
index 29b747ac31c1..fffe02aae3ed 100644
--- a/tools/perf/util/metricgroup.c
+++ b/tools/perf/util/metricgroup.c
@@ -11,6 +11,7 @@
 #include "evsel.h"
 #include "strbuf.h"
 #include "pmu.h"
+#include "pmu-hybrid.h"
 #include "expr.h"
 #include "rblist.h"
 #include <string.h>
@@ -18,6 +19,7 @@
 #include "strlist.h"
 #include <assert.h>
 #include <linux/ctype.h>
+#include <linux/list_sort.h>
 #include <linux/string.h>
 #include <linux/zalloc.h>
 #include <subcmd/parse-options.h>
@@ -84,6 +86,7 @@ static void metric_event_delete(struct rblist *rblist __maybe_unused,
 	struct metric_expr *expr, *tmp;
 
 	list_for_each_entry_safe(expr, tmp, &me->head, nd) {
+		free((char *)expr->metric_name);
 		free(expr->metric_refs);
 		free(expr->metric_events);
 		free(expr);
@@ -116,289 +119,207 @@ struct metric_ref_node {
 	struct list_head list;
 };
 
+/**
+ * The metric under construction. The data held here will be placed in a
+ * metric_expr.
+ */
 struct metric {
 	struct list_head nd;
-	struct expr_parse_ctx pctx;
+	/**
+	 * The expression parse context importantly holding the IDs contained
+	 * within the expression.
+	 */
+	struct expr_parse_ctx *pctx;
+	/** The name of the metric such as "IPC". */
 	const char *metric_name;
+	/** Modifier on the metric such as "u" or NULL for none. */
+	const char *modifier;
+	/** The expression to parse, for example, "instructions/cycles". */
 	const char *metric_expr;
+	/**
+	 * The "ScaleUnit" that scales and adds a unit to the metric during
+	 * output.
+	 */
 	const char *metric_unit;
-	struct list_head metric_refs;
-	int metric_refs_cnt;
-	int runtime;
+	/** Optional null terminated array of referenced metrics. */
+	struct metric_ref *metric_refs;
+	/**
+	 * Is there a constraint on the group of events? In which case the
+	 * events won't be grouped.
+	 */
 	bool has_constraint;
+	/**
+	 * Parsed events for the metric. Optional as events may be taken from a
+	 * different metric whose group contains all the IDs necessary for this
+	 * one.
+	 */
+	struct evlist *evlist;
 };
 
-#define RECURSION_ID_MAX 1000
+static void metricgroup___watchdog_constraint_hint(const char *name, bool foot)
+{
+	static bool violate_nmi_constraint;
 
-struct expr_ids {
-	struct expr_id	id[RECURSION_ID_MAX];
-	int		cnt;
-};
+	if (!foot) {
+		pr_warning("Splitting metric group %s into standalone metrics.\n", name);
+		violate_nmi_constraint = true;
+		return;
+	}
 
-static struct expr_id *expr_ids__alloc(struct expr_ids *ids)
+	if (!violate_nmi_constraint)
+		return;
+
+	pr_warning("Try disabling the NMI watchdog to comply NO_NMI_WATCHDOG metric constraint:\n"
+		   "    echo 0 > /proc/sys/kernel/nmi_watchdog\n"
+		   "    perf stat ...\n"
+		   "    echo 1 > /proc/sys/kernel/nmi_watchdog\n");
+}
+
+static bool metricgroup__has_constraint(const struct pmu_event *pe)
 {
-	if (ids->cnt >= RECURSION_ID_MAX)
-		return NULL;
-	return &ids->id[ids->cnt++];
+	if (!pe->metric_constraint)
+		return false;
+
+	if (!strcmp(pe->metric_constraint, "NO_NMI_WATCHDOG") &&
+	    sysctl__nmi_watchdog_enabled()) {
+		metricgroup___watchdog_constraint_hint(pe->metric_name, false);
+		return true;
+	}
+
+	return false;
 }
 
-static void expr_ids__exit(struct expr_ids *ids)
+static struct metric *metric__new(const struct pmu_event *pe,
+				  const char *modifier,
+				  bool metric_no_group,
+				  int runtime)
 {
-	int i;
+	struct metric *m;
+
+	m = zalloc(sizeof(*m));
+	if (!m)
+		return NULL;
+
+	m->pctx = expr__ctx_new();
+	if (!m->pctx) {
+		free(m);
+		return NULL;
+	}
 
-	for (i = 0; i < ids->cnt; i++)
-		free(ids->id[i].id);
+	m->metric_name = pe->metric_name;
+	m->modifier = modifier ? strdup(modifier) : NULL;
+	if (modifier && !m->modifier) {
+		free(m);
+		expr__ctx_free(m->pctx);
+		return NULL;
+	}
+	m->metric_expr = pe->metric_expr;
+	m->metric_unit = pe->unit;
+	m->pctx->runtime = runtime;
+	m->has_constraint = metric_no_group || metricgroup__has_constraint(pe);
+	m->metric_refs = NULL;
+	m->evlist = NULL;
+
+	return m;
 }
 
-static bool contains_event(struct evsel **metric_events, int num_events,
-			const char *event_name)
+static void metric__free(struct metric *m)
+{
+	free(m->metric_refs);
+	expr__ctx_free(m->pctx);
+	free((char *)m->modifier);
+	evlist__delete(m->evlist);
+	free(m);
+}
+
+static bool contains_metric_id(struct evsel **metric_events, int num_events,
+			       const char *metric_id)
 {
 	int i;
 
 	for (i = 0; i < num_events; i++) {
-		if (!strcmp(metric_events[i]->name, event_name))
+		if (!strcmp(evsel__metric_id(metric_events[i]), metric_id))
 			return true;
 	}
 	return false;
 }
 
-static bool evsel_same_pmu_or_none(struct evsel *ev1, struct evsel *ev2)
-{
-	if (!ev1->pmu_name || !ev2->pmu_name)
-		return true;
-
-	return !strcmp(ev1->pmu_name, ev2->pmu_name);
-}
-
 /**
- * Find a group of events in perf_evlist that correspond to those from a parsed
- * metric expression. Note, as find_evsel_group is called in the same order as
- * perf_evlist was constructed, metric_no_merge doesn't need to test for
- * underfilling a group.
- * @perf_evlist: a list of events something like: {metric1 leader, metric1
- * sibling, metric1 sibling}:W,duration_time,{metric2 leader, metric2 sibling,
- * metric2 sibling}:W,duration_time
- * @pctx: the parse context for the metric expression.
- * @metric_no_merge: don't attempt to share events for the metric with other
- * metrics.
- * @has_constraint: is there a constraint on the group of events? In which case
- * the events won't be grouped.
- * @metric_events: out argument, null terminated array of evsel's associated
- * with the metric.
- * @evlist_used: in/out argument, bitmap tracking which evlist events are used.
- * @return the first metric event or NULL on failure.
+ * setup_metric_events - Find a group of events in metric_evlist that correspond
+ *                       to the IDs from a parsed metric expression.
+ * @ids: the metric IDs to match.
+ * @metric_evlist: the list of perf events.
+ * @out_metric_events: holds the created metric events array.
  */
-static struct evsel *find_evsel_group(struct evlist *perf_evlist,
-				      struct expr_parse_ctx *pctx,
-				      bool metric_no_merge,
-				      bool has_constraint,
-				      struct evsel **metric_events,
-				      unsigned long *evlist_used)
+static int setup_metric_events(struct hashmap *ids,
+			       struct evlist *metric_evlist,
+			       struct evsel ***out_metric_events)
 {
-	struct evsel *ev, *current_leader = NULL;
-	struct expr_id_data *val_ptr;
-	int i = 0, matched_events = 0, events_to_match;
-	const int idnum = (int)hashmap__size(&pctx->ids);
+	struct evsel **metric_events;
+	const char *metric_id;
+	struct evsel *ev;
+	size_t ids_size, matched_events, i;
 
-	/*
-	 * duration_time is always grouped separately, when events are grouped
-	 * (ie has_constraint is false) then ignore it in the matching loop and
-	 * add it to metric_events at the end.
-	 */
-	if (!has_constraint &&
-	    hashmap__find(&pctx->ids, "duration_time", (void **)&val_ptr))
-		events_to_match = idnum - 1;
-	else
-		events_to_match = idnum;
+	*out_metric_events = NULL;
+	ids_size = hashmap__size(ids);
+
+	metric_events = calloc(sizeof(void *), ids_size + 1);
+	if (!metric_events)
+		return -ENOMEM;
+
+	matched_events = 0;
+	evlist__for_each_entry(metric_evlist, ev) {
+		struct expr_id_data *val_ptr;
 
-	evlist__for_each_entry (perf_evlist, ev) {
 		/*
-		 * Events with a constraint aren't grouped and match the first
-		 * events available.
+		 * Check for duplicate events with the same name. For
+		 * example, uncore_imc/cas_count_read/ will turn into 6
+		 * events per socket on skylakex. Only the first such
+		 * event is placed in metric_events.
 		 */
-		if (has_constraint && ev->weak_group)
-			continue;
-		/* Ignore event if already used and merging is disabled. */
-		if (metric_no_merge && test_bit(ev->core.idx, evlist_used))
+		metric_id = evsel__metric_id(ev);
+		if (contains_metric_id(metric_events, matched_events, metric_id))
 			continue;
-		if (!has_constraint && !evsel__has_leader(ev, current_leader)) {
-			/*
-			 * Start of a new group, discard the whole match and
-			 * start again.
-			 */
-			matched_events = 0;
-			memset(metric_events, 0,
-				sizeof(struct evsel *) * idnum);
-			current_leader = evsel__leader(ev);
-		}
 		/*
-		 * Check for duplicate events with the same name. For example,
-		 * uncore_imc/cas_count_read/ will turn into 6 events per socket
-		 * on skylakex. Only the first such event is placed in
-		 * metric_events. If events aren't grouped then this also
-		 * ensures that the same event in different sibling groups
-		 * aren't both added to metric_events.
+		 * Does this event belong to the parse context? For
+		 * combined or shared groups, this metric may not care
+		 * about this event.
 		 */
-		if (contains_event(metric_events, matched_events, ev->name))
-			continue;
-		/* Does this event belong to the parse context? */
-		if (hashmap__find(&pctx->ids, ev->name, (void **)&val_ptr))
+		if (hashmap__find(ids, metric_id, (void **)&val_ptr)) {
 			metric_events[matched_events++] = ev;
 
-		if (matched_events == events_to_match)
-			break;
-	}
-
-	if (events_to_match != idnum) {
-		/* Add the first duration_time. */
-		evlist__for_each_entry(perf_evlist, ev) {
-			if (!strcmp(ev->name, "duration_time")) {
-				metric_events[matched_events++] = ev;
+			if (matched_events >= ids_size)
 				break;
-			}
 		}
 	}
-
-	if (matched_events != idnum) {
-		/* Not a whole match */
-		return NULL;
+	if (matched_events < ids_size) {
+		free(metric_events);
+		return -EINVAL;
 	}
-
-	metric_events[idnum] = NULL;
-
-	for (i = 0; i < idnum; i++) {
+	for (i = 0; i < ids_size; i++) {
 		ev = metric_events[i];
-		/* Don't free the used events. */
-		set_bit(ev->core.idx, evlist_used);
+		ev->collect_stat = true;
+
 		/*
-		 * The metric leader points to the identically named event in
-		 * metric_events.
+		 * The metric leader points to the identically named
+		 * event in metric_events.
 		 */
 		ev->metric_leader = ev;
 		/*
-		 * Mark two events with identical names in the same group (or
-		 * globally) as being in use as uncore events may be duplicated
-		 * for each pmu. Set the metric leader of such events to be the
-		 * event that appears in metric_events.
+		 * Mark two events with identical names in the same
+		 * group (or globally) as being in use as uncore events
+		 * may be duplicated for each pmu. Set the metric leader
+		 * of such events to be the event that appears in
+		 * metric_events.
 		 */
-		evlist__for_each_entry_continue(perf_evlist, ev) {
-			/*
-			 * If events are grouped then the search can terminate
-			 * when then group is left.
-			 */
-			if (!has_constraint &&
-			    ev->core.leader != metric_events[i]->core.leader &&
-			    evsel_same_pmu_or_none(evsel__leader(ev), evsel__leader(metric_events[i])))
-				break;
-			if (!strcmp(metric_events[i]->name, ev->name)) {
-				set_bit(ev->core.idx, evlist_used);
+		metric_id = evsel__metric_id(ev);
+		evlist__for_each_entry_continue(metric_evlist, ev) {
+			if (!strcmp(evsel__metric_id(metric_events[i]), metric_id))
 				ev->metric_leader = metric_events[i];
-			}
 		}
 	}
-
-	return metric_events[0];
-}
-
-static int metricgroup__setup_events(struct list_head *groups,
-				     bool metric_no_merge,
-				     struct evlist *perf_evlist,
-				     struct rblist *metric_events_list)
-{
-	struct metric_event *me;
-	struct metric_expr *expr;
-	int i = 0;
-	int ret = 0;
-	struct metric *m;
-	struct evsel *evsel, *tmp;
-	unsigned long *evlist_used;
-
-	evlist_used = bitmap_zalloc(perf_evlist->core.nr_entries);
-	if (!evlist_used)
-		return -ENOMEM;
-
-	list_for_each_entry (m, groups, nd) {
-		struct evsel **metric_events;
-		struct metric_ref *metric_refs = NULL;
-
-		metric_events = calloc(sizeof(void *),
-				hashmap__size(&m->pctx.ids) + 1);
-		if (!metric_events) {
-			ret = -ENOMEM;
-			break;
-		}
-		evsel = find_evsel_group(perf_evlist, &m->pctx,
-					 metric_no_merge,
-					 m->has_constraint, metric_events,
-					 evlist_used);
-		if (!evsel) {
-			pr_debug("Cannot resolve %s: %s\n",
-					m->metric_name, m->metric_expr);
-			free(metric_events);
-			continue;
-		}
-		for (i = 0; metric_events[i]; i++)
-			metric_events[i]->collect_stat = true;
-		me = metricgroup__lookup(metric_events_list, evsel, true);
-		if (!me) {
-			ret = -ENOMEM;
-			free(metric_events);
-			break;
-		}
-		expr = malloc(sizeof(struct metric_expr));
-		if (!expr) {
-			ret = -ENOMEM;
-			free(metric_events);
-			break;
-		}
-
-		/*
-		 * Collect and store collected nested expressions
-		 * for metric processing.
-		 */
-		if (m->metric_refs_cnt) {
-			struct metric_ref_node *ref;
-
-			metric_refs = zalloc(sizeof(struct metric_ref) * (m->metric_refs_cnt + 1));
-			if (!metric_refs) {
-				ret = -ENOMEM;
-				free(metric_events);
-				free(expr);
-				break;
-			}
-
-			i = 0;
-			list_for_each_entry(ref, &m->metric_refs, list) {
-				/*
-				 * Intentionally passing just const char pointers,
-				 * originally from 'struct pmu_event' object.
-				 * We don't need to change them, so there's no
-				 * need to create our own copy.
-				 */
-				metric_refs[i].metric_name = ref->metric_name;
-				metric_refs[i].metric_expr = ref->metric_expr;
-				i++;
-			}
-		}
-
-		expr->metric_refs = metric_refs;
-		expr->metric_expr = m->metric_expr;
-		expr->metric_name = m->metric_name;
-		expr->metric_unit = m->metric_unit;
-		expr->metric_events = metric_events;
-		expr->runtime = m->runtime;
-		list_add(&expr->nd, &me->head);
-	}
-
-	evlist__for_each_entry_safe(perf_evlist, tmp, evsel) {
-		if (!test_bit(evsel->core.idx, evlist_used)) {
-			evlist__remove(perf_evlist, evsel);
-			evsel__delete(evsel);
-		}
-	}
-	bitmap_free(evlist_used);
-
-	return ret;
+	*out_metric_events = metric_events;
+	return 0;
 }
 
 static bool match_metric(const char *n, const char *list)
@@ -422,7 +343,7 @@ static bool match_metric(const char *n, const char *list)
 	return false;
 }
 
-static bool match_pe_metric(struct pmu_event *pe, const char *metric)
+static bool match_pe_metric(const struct pmu_event *pe, const char *metric)
 {
 	return match_metric(pe->metric_group, metric) ||
 	       match_metric(pe->metric_name, metric);
@@ -506,7 +427,7 @@ static void metricgroup__print_strlist(struct strlist *metrics, bool raw)
 		putchar('\n');
 }
 
-static int metricgroup__print_pmu_event(struct pmu_event *pe,
+static int metricgroup__print_pmu_event(const struct pmu_event *pe,
 					bool metricgroups, char *filter,
 					bool raw, bool details,
 					struct rblist *groups,
@@ -581,14 +502,14 @@ struct metricgroup_print_sys_idata {
 	bool details;
 };
 
-typedef int (*metricgroup_sys_event_iter_fn)(struct pmu_event *pe, void *);
+typedef int (*metricgroup_sys_event_iter_fn)(const struct pmu_event *pe, void *);
 
 struct metricgroup_iter_data {
 	metricgroup_sys_event_iter_fn fn;
 	void *data;
 };
 
-static int metricgroup__sys_event_iter(struct pmu_event *pe, void *data)
+static int metricgroup__sys_event_iter(const struct pmu_event *pe, void *data)
 {
 	struct metricgroup_iter_data *d = data;
 	struct perf_pmu *pmu = NULL;
@@ -607,7 +528,7 @@ static int metricgroup__sys_event_iter(struct pmu_event *pe, void *data)
 	return 0;
 }
 
-static int metricgroup__print_sys_event_iter(struct pmu_event *pe, void *data)
+static int metricgroup__print_sys_event_iter(const struct pmu_event *pe, void *data)
 {
 	struct metricgroup_print_sys_idata *d = data;
 
@@ -616,10 +537,10 @@ static int metricgroup__print_sys_event_iter(struct pmu_event *pe, void *data)
 }
 
 void metricgroup__print(bool metrics, bool metricgroups, char *filter,
-			bool raw, bool details)
+			bool raw, bool details, const char *pmu_name)
 {
-	struct pmu_events_map *map = pmu_events_map__find();
-	struct pmu_event *pe;
+	const struct pmu_events_map *map = pmu_events_map__find();
+	const struct pmu_event *pe;
 	int i;
 	struct rblist groups;
 	struct rb_node *node, *next;
@@ -642,6 +563,10 @@ void metricgroup__print(bool metrics, bool metricgroups, char *filter,
 			break;
 		if (!pe->metric_expr)
 			continue;
+		if (pmu_name && perf_pmu__is_hybrid(pe->pmu) &&
+		    strcmp(pmu_name, pe->pmu)) {
+			continue;
+		}
 		if (metricgroup__print_pmu_event(pe, metricgroups, filter,
 						 raw, details, &groups,
 						 metriclist) < 0)
@@ -686,150 +611,391 @@ void metricgroup__print(bool metrics, bool metricgroups, char *filter,
 	strlist__delete(metriclist);
 }
 
-static void metricgroup__add_metric_weak_group(struct strbuf *events,
-					       struct expr_parse_ctx *ctx)
+static const char *code_characters = ",-=@";
+
+static int encode_metric_id(struct strbuf *sb, const char *x)
 {
-	struct hashmap_entry *cur;
-	size_t bkt;
-	bool no_group = true, has_duration = false;
+	char *c;
+	int ret = 0;
 
-	hashmap__for_each_entry((&ctx->ids), cur, bkt) {
-		pr_debug("found event %s\n", (const char *)cur->key);
-		/*
-		 * Duration time maps to a software event and can make
-		 * groups not count. Always use it outside a
-		 * group.
-		 */
-		if (!strcmp(cur->key, "duration_time")) {
-			has_duration = true;
-			continue;
+	for (; *x; x++) {
+		c = strchr(code_characters, *x);
+		if (c) {
+			ret = strbuf_addch(sb, '!');
+			if (ret)
+				break;
+
+			ret = strbuf_addch(sb, '0' + (c - code_characters));
+			if (ret)
+				break;
+		} else {
+			ret = strbuf_addch(sb, *x);
+			if (ret)
+				break;
 		}
-		strbuf_addf(events, "%s%s",
-			no_group ? "{" : ",",
-			(const char *)cur->key);
-		no_group = false;
 	}
-	if (!no_group) {
-		strbuf_addf(events, "}:W");
-		if (has_duration)
-			strbuf_addf(events, ",duration_time");
-	} else if (has_duration)
-		strbuf_addf(events, "duration_time");
+	return ret;
 }
 
-static void metricgroup__add_metric_non_group(struct strbuf *events,
-					      struct expr_parse_ctx *ctx)
+static int decode_metric_id(struct strbuf *sb, const char *x)
 {
-	struct hashmap_entry *cur;
-	size_t bkt;
-	bool first = true;
+	const char *orig = x;
+	size_t i;
+	char c;
+	int ret;
 
-	hashmap__for_each_entry((&ctx->ids), cur, bkt) {
-		if (!first)
-			strbuf_addf(events, ",");
-		strbuf_addf(events, "%s", (const char *)cur->key);
-		first = false;
+	for (; *x; x++) {
+		c = *x;
+		if (*x == '!') {
+			x++;
+			i = *x - '0';
+			if (i > strlen(code_characters)) {
+				pr_err("Bad metric-id encoding in: '%s'", orig);
+				return -1;
+			}
+			c = code_characters[i];
+		}
+		ret = strbuf_addch(sb, c);
+		if (ret)
+			return ret;
 	}
+	return 0;
 }
 
-static void metricgroup___watchdog_constraint_hint(const char *name, bool foot)
+static int decode_all_metric_ids(struct evlist *perf_evlist, const char *modifier)
 {
-	static bool violate_nmi_constraint;
+	struct evsel *ev;
+	struct strbuf sb = STRBUF_INIT;
+	char *cur;
+	int ret = 0;
 
-	if (!foot) {
-		pr_warning("Splitting metric group %s into standalone metrics.\n", name);
-		violate_nmi_constraint = true;
-		return;
-	}
+	evlist__for_each_entry(perf_evlist, ev) {
+		if (!ev->metric_id)
+			continue;
 
-	if (!violate_nmi_constraint)
-		return;
+		ret = strbuf_setlen(&sb, 0);
+		if (ret)
+			break;
 
-	pr_warning("Try disabling the NMI watchdog to comply NO_NMI_WATCHDOG metric constraint:\n"
-		   "    echo 0 > /proc/sys/kernel/nmi_watchdog\n"
-		   "    perf stat ...\n"
-		   "    echo 1 > /proc/sys/kernel/nmi_watchdog\n");
+		ret = decode_metric_id(&sb, ev->metric_id);
+		if (ret)
+			break;
+
+		free((char *)ev->metric_id);
+		ev->metric_id = strdup(sb.buf);
+		if (!ev->metric_id) {
+			ret = -ENOMEM;
+			break;
+		}
+		/*
+		 * If the name is just the parsed event, use the metric-id to
+		 * give a more friendly display version.
+		 */
+		if (strstr(ev->name, "metric-id=")) {
+			bool has_slash = false;
+
+			free(ev->name);
+			for (cur = strchr(sb.buf, '@') ; cur; cur = strchr(++cur, '@')) {
+				*cur = '/';
+				has_slash = true;
+			}
+
+			if (modifier) {
+				if (!has_slash && !strchr(sb.buf, ':')) {
+					ret = strbuf_addch(&sb, ':');
+					if (ret)
+						break;
+				}
+				ret = strbuf_addstr(&sb, modifier);
+				if (ret)
+					break;
+			}
+			ev->name = strdup(sb.buf);
+			if (!ev->name) {
+				ret = -ENOMEM;
+				break;
+			}
+		}
+	}
+	strbuf_release(&sb);
+	return ret;
 }
 
-static bool metricgroup__has_constraint(struct pmu_event *pe)
+static int metricgroup__build_event_string(struct strbuf *events,
+					   const struct expr_parse_ctx *ctx,
+					   const char *modifier,
+					   bool has_constraint)
 {
-	if (!pe->metric_constraint)
-		return false;
+	struct hashmap_entry *cur;
+	size_t bkt;
+	bool no_group = true, has_duration = false;
+	int ret = 0;
 
-	if (!strcmp(pe->metric_constraint, "NO_NMI_WATCHDOG") &&
-	    sysctl__nmi_watchdog_enabled()) {
-		metricgroup___watchdog_constraint_hint(pe->metric_name, false);
-		return true;
+#define RETURN_IF_NON_ZERO(x) do { if (x) return x; } while (0)
+
+	hashmap__for_each_entry(ctx->ids, cur, bkt) {
+		const char *sep, *rsep, *id = cur->key;
+
+		pr_debug("found event %s\n", id);
+		/*
+		 * Duration time maps to a software event and can make
+		 * groups not count. Always use it outside a
+		 * group.
+		 */
+		if (!strcmp(id, "duration_time")) {
+			has_duration = true;
+			continue;
+		}
+		/* Separate events with commas and open the group if necessary. */
+		if (no_group) {
+			if (!has_constraint) {
+				ret = strbuf_addch(events, '{');
+				RETURN_IF_NON_ZERO(ret);
+			}
+
+			no_group = false;
+		} else {
+			ret = strbuf_addch(events, ',');
+			RETURN_IF_NON_ZERO(ret);
+		}
+		/*
+		 * Encode the ID as an event string. Add a qualifier for
+		 * metric_id that is the original name except with characters
+		 * that parse-events can't parse replaced. For example,
+		 * 'msr@tsc@' gets added as msr/tsc,metric-id=msr!3tsc!3/
+		 */
+		sep = strchr(id, '@');
+		if (sep != NULL) {
+			ret = strbuf_add(events, id, sep - id);
+			RETURN_IF_NON_ZERO(ret);
+			ret = strbuf_addch(events, '/');
+			RETURN_IF_NON_ZERO(ret);
+			rsep = strrchr(sep, '@');
+			ret = strbuf_add(events, sep + 1, rsep - sep - 1);
+			RETURN_IF_NON_ZERO(ret);
+			ret = strbuf_addstr(events, ",metric-id=");
+			RETURN_IF_NON_ZERO(ret);
+			sep = rsep;
+		} else {
+			sep = strchr(id, ':');
+			if (sep != NULL) {
+				ret = strbuf_add(events, id, sep - id);
+				RETURN_IF_NON_ZERO(ret);
+			} else {
+				ret = strbuf_addstr(events, id);
+				RETURN_IF_NON_ZERO(ret);
+			}
+			ret = strbuf_addstr(events, "/metric-id=");
+			RETURN_IF_NON_ZERO(ret);
+		}
+		ret = encode_metric_id(events, id);
+		RETURN_IF_NON_ZERO(ret);
+		ret = strbuf_addstr(events, "/");
+		RETURN_IF_NON_ZERO(ret);
+
+		if (sep != NULL) {
+			ret = strbuf_addstr(events, sep + 1);
+			RETURN_IF_NON_ZERO(ret);
+		}
+		if (modifier) {
+			ret = strbuf_addstr(events, modifier);
+			RETURN_IF_NON_ZERO(ret);
+		}
 	}
+	if (has_duration) {
+		if (no_group) {
+			/* Strange case of a metric of just duration_time. */
+			ret = strbuf_addf(events, "duration_time");
+		} else if (!has_constraint)
+			ret = strbuf_addf(events, "}:W,duration_time");
+		else
+			ret = strbuf_addf(events, ",duration_time");
+	} else if (!no_group && !has_constraint)
+		ret = strbuf_addf(events, "}:W");
 
-	return false;
+	return ret;
+#undef RETURN_IF_NON_ZERO
 }
 
-int __weak arch_get_runtimeparam(struct pmu_event *pe __maybe_unused)
+int __weak arch_get_runtimeparam(const struct pmu_event *pe __maybe_unused)
 {
 	return 1;
 }
 
+/*
+ * A singly linked list on the stack of the names of metrics being
+ * processed. Used to identify recursion.
+ */
+struct visited_metric {
+	const char *name;
+	const struct visited_metric *parent;
+};
+
 struct metricgroup_add_iter_data {
 	struct list_head *metric_list;
-	const char *metric;
-	struct expr_ids *ids;
+	const char *metric_name;
+	const char *modifier;
 	int *ret;
 	bool *has_match;
 	bool metric_no_group;
+	struct metric *root_metric;
+	const struct visited_metric *visited;
+	const struct pmu_events_map *map;
 };
 
+static int add_metric(struct list_head *metric_list,
+		      const struct pmu_event *pe,
+		      const char *modifier,
+		      bool metric_no_group,
+		      struct metric *root_metric,
+		      const struct visited_metric *visited,
+		      const struct pmu_events_map *map);
+
+/**
+ * resolve_metric - Locate metrics within the root metric and recursively add
+ *                    references to them.
+ * @metric_list: The list the metric is added to.
+ * @modifier: if non-null event modifiers like "u".
+ * @metric_no_group: Should events written to events be grouped "{}" or
+ *                   global. Grouping is the default but due to multiplexing the
+ *                   user may override.
+ * @root_metric: Metrics may reference other metrics to form a tree. In this
+ *               case the root_metric holds all the IDs and a list of referenced
+ *               metrics. When adding a root this argument is NULL.
+ * @visited: A singly linked list of metric names being added that is used to
+ *           detect recursion.
+ * @map: The map that is searched for metrics, most commonly the table for the
+ *       architecture perf is running upon.
+ */
+static int resolve_metric(struct list_head *metric_list,
+			  const char *modifier,
+			  bool metric_no_group,
+			  struct metric *root_metric,
+			  const struct visited_metric *visited,
+			  const struct pmu_events_map *map)
+{
+	struct hashmap_entry *cur;
+	size_t bkt;
+	struct to_resolve {
+		/* The metric to resolve. */
+		const struct pmu_event *pe;
+		/*
+		 * The key in the IDs map, this may differ from in case,
+		 * etc. from pe->metric_name.
+		 */
+		const char *key;
+	} *pending = NULL;
+	int i, ret = 0, pending_cnt = 0;
+
+	/*
+	 * Iterate all the parsed IDs and if there's a matching metric and it to
+	 * the pending array.
+	 */
+	hashmap__for_each_entry(root_metric->pctx->ids, cur, bkt) {
+		const struct pmu_event *pe;
+
+		pe = metricgroup__find_metric(cur->key, map);
+		if (pe) {
+			pending = realloc(pending,
+					(pending_cnt + 1) * sizeof(struct to_resolve));
+			if (!pending)
+				return -ENOMEM;
+
+			pending[pending_cnt].pe = pe;
+			pending[pending_cnt].key = cur->key;
+			pending_cnt++;
+		}
+	}
+
+	/* Remove the metric IDs from the context. */
+	for (i = 0; i < pending_cnt; i++)
+		expr__del_id(root_metric->pctx, pending[i].key);
+
+	/*
+	 * Recursively add all the metrics, IDs are added to the root metric's
+	 * context.
+	 */
+	for (i = 0; i < pending_cnt; i++) {
+		ret = add_metric(metric_list, pending[i].pe, modifier, metric_no_group,
+				root_metric, visited, map);
+		if (ret)
+			break;
+	}
+
+	free(pending);
+	return ret;
+}
+
+/**
+ * __add_metric - Add a metric to metric_list.
+ * @metric_list: The list the metric is added to.
+ * @pe: The pmu_event containing the metric to be added.
+ * @modifier: if non-null event modifiers like "u".
+ * @metric_no_group: Should events written to events be grouped "{}" or
+ *                   global. Grouping is the default but due to multiplexing the
+ *                   user may override.
+ * @runtime: A special argument for the parser only known at runtime.
+ * @root_metric: Metrics may reference other metrics to form a tree. In this
+ *               case the root_metric holds all the IDs and a list of referenced
+ *               metrics. When adding a root this argument is NULL.
+ * @visited: A singly linked list of metric names being added that is used to
+ *           detect recursion.
+ * @map: The map that is searched for metrics, most commonly the table for the
+ *       architecture perf is running upon.
+ */
 static int __add_metric(struct list_head *metric_list,
-			struct pmu_event *pe,
+			const struct pmu_event *pe,
+			const char *modifier,
 			bool metric_no_group,
 			int runtime,
-			struct metric **mp,
-			struct expr_id *parent,
-			struct expr_ids *ids)
+			struct metric *root_metric,
+			const struct visited_metric *visited,
+			const struct pmu_events_map *map)
 {
-	struct metric_ref_node *ref;
-	struct metric *m;
+	const struct visited_metric *vm;
+	int ret;
+	bool is_root = !root_metric;
+	struct visited_metric visited_node = {
+		.name = pe->metric_name,
+		.parent = visited,
+	};
 
-	if (*mp == NULL) {
+	for (vm = visited; vm; vm = vm->parent) {
+		if (!strcmp(pe->metric_name, vm->name)) {
+			pr_err("failed: recursion detected for %s\n", pe->metric_name);
+			return -1;
+		}
+	}
+
+	if (is_root) {
 		/*
-		 * We got in here for the parent group,
-		 * allocate it and put it on the list.
+		 * This metric is the root of a tree and may reference other
+		 * metrics that are added recursively.
 		 */
-		m = zalloc(sizeof(*m));
-		if (!m)
+		root_metric = metric__new(pe, modifier, metric_no_group, runtime);
+		if (!root_metric)
 			return -ENOMEM;
 
-		expr__ctx_init(&m->pctx);
-		m->metric_name = pe->metric_name;
-		m->metric_expr = pe->metric_expr;
-		m->metric_unit = pe->unit;
-		m->runtime = runtime;
-		m->has_constraint = metric_no_group || metricgroup__has_constraint(pe);
-		INIT_LIST_HEAD(&m->metric_refs);
-		m->metric_refs_cnt = 0;
-
-		parent = expr_ids__alloc(ids);
-		if (!parent) {
-			free(m);
-			return -EINVAL;
-		}
-
-		parent->id = strdup(pe->metric_name);
-		if (!parent->id) {
-			free(m);
-			return -ENOMEM;
-		}
-		*mp = m;
 	} else {
+		int cnt = 0;
+
 		/*
-		 * We got here for the referenced metric, via the
-		 * recursive metricgroup__add_metric call, add
-		 * it to the parent group.
+		 * This metric was referenced in a metric higher in the
+		 * tree. Check if the same metric is already resolved in the
+		 * metric_refs list.
 		 */
-		m = *mp;
+		if (root_metric->metric_refs) {
+			for (; root_metric->metric_refs[cnt].metric_name; cnt++) {
+				if (!strcmp(pe->metric_name,
+					    root_metric->metric_refs[cnt].metric_name))
+					return 0;
+			}
+		}
 
-		ref = malloc(sizeof(*ref));
-		if (!ref)
+		/* Create reference. Need space for the entry and the terminator. */
+		root_metric->metric_refs = realloc(root_metric->metric_refs,
+						(cnt + 2) * sizeof(struct metric_ref));
+		if (!root_metric->metric_refs)
 			return -ENOMEM;
 
 		/*
@@ -838,54 +1004,35 @@ static int __add_metric(struct list_head *metric_list,
 		 * need to change them, so there's no need to create
 		 * our own copy.
 		 */
-		ref->metric_name = pe->metric_name;
-		ref->metric_expr = pe->metric_expr;
+		root_metric->metric_refs[cnt].metric_name = pe->metric_name;
+		root_metric->metric_refs[cnt].metric_expr = pe->metric_expr;
 
-		list_add(&ref->list, &m->metric_refs);
-		m->metric_refs_cnt++;
+		/* Null terminate array. */
+		root_metric->metric_refs[cnt+1].metric_name = NULL;
+		root_metric->metric_refs[cnt+1].metric_expr = NULL;
 	}
 
-	/* Force all found IDs in metric to have us as parent ID. */
-	WARN_ON_ONCE(!parent);
-	m->pctx.parent = parent;
-
 	/*
 	 * For both the parent and referenced metrics, we parse
-	 * all the metric's IDs and add it to the parent context.
+	 * all the metric's IDs and add it to the root context.
 	 */
-	if (expr__find_other(pe->metric_expr, NULL, &m->pctx, runtime) < 0) {
-		if (m->metric_refs_cnt == 0) {
-			expr__ctx_clear(&m->pctx);
-			free(m);
-			*mp = NULL;
-		}
-		return -EINVAL;
+	if (expr__find_ids(pe->metric_expr, NULL, root_metric->pctx) < 0) {
+		/* Broken metric. */
+		ret = -EINVAL;
+	} else {
+		/* Resolve referenced metrics. */
+		ret = resolve_metric(metric_list, modifier, metric_no_group, root_metric,
+				     &visited_node, map);
 	}
 
-	/*
-	 * We add new group only in the 'parent' call,
-	 * so bail out for referenced metric case.
-	 */
-	if (m->metric_refs_cnt)
-		return 0;
-
-	if (list_empty(metric_list))
-		list_add(&m->nd, metric_list);
-	else {
-		struct list_head *pos;
-
-		/* Place the largest groups at the front. */
-		list_for_each_prev(pos, metric_list) {
-			struct metric *old = list_entry(pos, struct metric, nd);
+	if (ret) {
+		if (is_root)
+			metric__free(root_metric);
 
-			if (hashmap__size(&m->pctx.ids) <=
-			    hashmap__size(&old->pctx.ids))
-				break;
-		}
-		list_add(&m->nd, pos);
-	}
+	} else if (is_root)
+		list_add(&root_metric->nd, metric_list);
 
-	return 0;
+	return ret;
 }
 
 #define map_for_each_event(__pe, __idx, __map)					\
@@ -900,10 +1047,10 @@ static int __add_metric(struct list_head *metric_list,
 		    (match_metric(__pe->metric_group, __metric) ||	\
 		     match_metric(__pe->metric_name, __metric)))
 
-struct pmu_event *metricgroup__find_metric(const char *metric,
-					   struct pmu_events_map *map)
+const struct pmu_event *metricgroup__find_metric(const char *metric,
+						 const struct pmu_events_map *map)
 {
-	struct pmu_event *pe;
+	const struct pmu_event *pe;
 	int i;
 
 	map_for_each_event(pe, i, map) {
@@ -914,136 +1061,21 @@ struct pmu_event *metricgroup__find_metric(const char *metric,
 	return NULL;
 }
 
-static int recursion_check(struct metric *m, const char *id, struct expr_id **parent,
-			   struct expr_ids *ids)
-{
-	struct expr_id_data *data;
-	struct expr_id *p;
-	int ret;
-
-	/*
-	 * We get the parent referenced by 'id' argument and
-	 * traverse through all the parent object IDs to check
-	 * if we already processed 'id', if we did, it's recursion
-	 * and we fail.
-	 */
-	ret = expr__get_id(&m->pctx, id, &data);
-	if (ret)
-		return ret;
-
-	p = expr_id_data__parent(data);
-
-	while (p->parent) {
-		if (!strcmp(p->id, id)) {
-			pr_err("failed: recursion detected for %s\n", id);
-			return -1;
-		}
-		p = p->parent;
-	}
-
-	/*
-	 * If we are over the limit of static entris, the metric
-	 * is too difficult/nested to process, fail as well.
-	 */
-	p = expr_ids__alloc(ids);
-	if (!p) {
-		pr_err("failed: too many nested metrics\n");
-		return -EINVAL;
-	}
-
-	p->id     = strdup(id);
-	p->parent = expr_id_data__parent(data);
-	*parent   = p;
-
-	return p->id ? 0 : -ENOMEM;
-}
-
 static int add_metric(struct list_head *metric_list,
-		      struct pmu_event *pe,
+		      const struct pmu_event *pe,
+		      const char *modifier,
 		      bool metric_no_group,
-		      struct metric **mp,
-		      struct expr_id *parent,
-		      struct expr_ids *ids);
-
-static int __resolve_metric(struct metric *m,
-			    bool metric_no_group,
-			    struct list_head *metric_list,
-			    struct pmu_events_map *map,
-			    struct expr_ids *ids)
+		      struct metric *root_metric,
+		      const struct visited_metric *visited,
+		      const struct pmu_events_map *map)
 {
-	struct hashmap_entry *cur;
-	size_t bkt;
-	bool all;
-	int ret;
-
-	/*
-	 * Iterate all the parsed IDs and if there's metric,
-	 * add it to the context.
-	 */
-	do {
-		all = true;
-		hashmap__for_each_entry((&m->pctx.ids), cur, bkt) {
-			struct expr_id *parent;
-			struct pmu_event *pe;
-
-			pe = metricgroup__find_metric(cur->key, map);
-			if (!pe)
-				continue;
-
-			ret = recursion_check(m, cur->key, &parent, ids);
-			if (ret)
-				return ret;
-
-			all = false;
-			/* The metric key itself needs to go out.. */
-			expr__del_id(&m->pctx, cur->key);
-
-			/* ... and it gets resolved to the parent context. */
-			ret = add_metric(metric_list, pe, metric_no_group, &m, parent, ids);
-			if (ret)
-				return ret;
-
-			/*
-			 * We added new metric to hashmap, so we need
-			 * to break the iteration and start over.
-			 */
-			break;
-		}
-	} while (!all);
-
-	return 0;
-}
-
-static int resolve_metric(bool metric_no_group,
-			  struct list_head *metric_list,
-			  struct pmu_events_map *map,
-			  struct expr_ids *ids)
-{
-	struct metric *m;
-	int err;
-
-	list_for_each_entry(m, metric_list, nd) {
-		err = __resolve_metric(m, metric_no_group, metric_list, map, ids);
-		if (err)
-			return err;
-	}
-	return 0;
-}
-
-static int add_metric(struct list_head *metric_list,
-		      struct pmu_event *pe,
-		      bool metric_no_group,
-		      struct metric **m,
-		      struct expr_id *parent,
-		      struct expr_ids *ids)
-{
-	struct metric *orig = *m;
 	int ret = 0;
 
 	pr_debug("metric expr %s for %s\n", pe->metric_expr, pe->metric_name);
 
 	if (!strstr(pe->metric_expr, "?")) {
-		ret = __add_metric(metric_list, pe, metric_no_group, 1, m, parent, ids);
+		ret = __add_metric(metric_list, pe, modifier, metric_no_group, 0,
+				   root_metric, visited, map);
 	} else {
 		int j, count;
 
@@ -1054,29 +1086,25 @@ static int add_metric(struct list_head *metric_list,
 		 * those events to metric_list.
 		 */
 
-		for (j = 0; j < count && !ret; j++, *m = orig)
-			ret = __add_metric(metric_list, pe, metric_no_group, j, m, parent, ids);
+		for (j = 0; j < count && !ret; j++)
+			ret = __add_metric(metric_list, pe, modifier, metric_no_group, j,
+					root_metric, visited, map);
 	}
 
 	return ret;
 }
 
-static int metricgroup__add_metric_sys_event_iter(struct pmu_event *pe,
+static int metricgroup__add_metric_sys_event_iter(const struct pmu_event *pe,
 						  void *data)
 {
 	struct metricgroup_add_iter_data *d = data;
-	struct metric *m = NULL;
 	int ret;
 
-	if (!match_pe_metric(pe, d->metric))
+	if (!match_pe_metric(pe, d->metric_name))
 		return 0;
 
-	ret = add_metric(d->metric_list, pe, d->metric_no_group, &m, NULL, d->ids);
-	if (ret)
-		goto out;
-
-	ret = resolve_metric(d->metric_no_group,
-				     d->metric_list, NULL, d->ids);
+	ret = add_metric(d->metric_list, pe, d->modifier, d->metric_no_group,
+			 d->root_metric, d->visited, d->map);
 	if (ret)
 		goto out;
 
@@ -1087,32 +1115,47 @@ out:
 	return ret;
 }
 
-static int metricgroup__add_metric(const char *metric, bool metric_no_group,
-				   struct strbuf *events,
+static int metric_list_cmp(void *priv __maybe_unused, const struct list_head *l,
+			   const struct list_head *r)
+{
+	const struct metric *left = container_of(l, struct metric, nd);
+	const struct metric *right = container_of(r, struct metric, nd);
+
+	return hashmap__size(right->pctx->ids) - hashmap__size(left->pctx->ids);
+}
+
+/**
+ * metricgroup__add_metric - Find and add a metric, or a metric group.
+ * @metric_name: The name of the metric or metric group. For example, "IPC"
+ *               could be the name of a metric and "TopDownL1" the name of a
+ *               metric group.
+ * @modifier: if non-null event modifiers like "u".
+ * @metric_no_group: Should events written to events be grouped "{}" or
+ *                   global. Grouping is the default but due to multiplexing the
+ *                   user may override.
+ * @metric_list: The list that the metric or metric group are added to.
+ * @map: The map that is searched for metrics, most commonly the table for the
+ *       architecture perf is running upon.
+ */
+static int metricgroup__add_metric(const char *metric_name, const char *modifier,
+				   bool metric_no_group,
 				   struct list_head *metric_list,
-				   struct pmu_events_map *map)
+				   const struct pmu_events_map *map)
 {
-	struct expr_ids ids = { .cnt = 0, };
-	struct pmu_event *pe;
-	struct metric *m;
+	const struct pmu_event *pe;
 	LIST_HEAD(list);
 	int i, ret;
 	bool has_match = false;
 
-	map_for_each_metric(pe, i, map, metric) {
+	/*
+	 * Iterate over all metrics seeing if metric matches either the name or
+	 * group. When it does add the metric to the list.
+	 */
+	map_for_each_metric(pe, i, map, metric_name) {
 		has_match = true;
-		m = NULL;
-
-		ret = add_metric(&list, pe, metric_no_group, &m, NULL, &ids);
-		if (ret)
-			goto out;
-
-		/*
-		 * Process any possible referenced metrics
-		 * included in the expression.
-		 */
-		ret = resolve_metric(metric_no_group,
-				     &list, map, &ids);
+		ret = add_metric(&list, pe, modifier, metric_no_group,
+				 /*root_metric=*/NULL,
+				 /*visited_metrics=*/NULL, map);
 		if (ret)
 			goto out;
 	}
@@ -1122,34 +1165,20 @@ static int metricgroup__add_metric(const char *metric, bool metric_no_group,
 			.fn = metricgroup__add_metric_sys_event_iter,
 			.data = (void *) &(struct metricgroup_add_iter_data) {
 				.metric_list = &list,
-				.metric = metric,
+				.metric_name = metric_name,
+				.modifier = modifier,
 				.metric_no_group = metric_no_group,
-				.ids = &ids,
 				.has_match = &has_match,
 				.ret = &ret,
+				.map = map,
 			},
 		};
 
 		pmu_for_each_sys_event(metricgroup__sys_event_iter, &data);
 	}
 	/* End of pmu events. */
-	if (!has_match) {
+	if (!has_match)
 		ret = -EINVAL;
-		goto out;
-	}
-
-	list_for_each_entry(m, &list, nd) {
-		if (events->len > 0)
-			strbuf_addf(events, ",");
-
-		if (m->has_constraint) {
-			metricgroup__add_metric_non_group(events,
-							  &m->pctx);
-		} else {
-			metricgroup__add_metric_weak_group(events,
-							   &m->pctx);
-		}
-	}
 
 out:
 	/*
@@ -1157,95 +1186,315 @@ out:
 	 * even if it's failed
 	 */
 	list_splice(&list, metric_list);
-	expr_ids__exit(&ids);
 	return ret;
 }
 
+/**
+ * metricgroup__add_metric_list - Find and add metrics, or metric groups,
+ *                                specified in a list.
+ * @list: the list of metrics or metric groups. For example, "IPC,CPI,TopDownL1"
+ *        would match the IPC and CPI metrics, and TopDownL1 would match all
+ *        the metrics in the TopDownL1 group.
+ * @metric_no_group: Should events written to events be grouped "{}" or
+ *                   global. Grouping is the default but due to multiplexing the
+ *                   user may override.
+ * @metric_list: The list that metrics are added to.
+ * @map: The map that is searched for metrics, most commonly the table for the
+ *       architecture perf is running upon.
+ */
 static int metricgroup__add_metric_list(const char *list, bool metric_no_group,
-					struct strbuf *events,
 					struct list_head *metric_list,
-					struct pmu_events_map *map)
+					const struct pmu_events_map *map)
 {
-	char *llist, *nlist, *p;
-	int ret = -EINVAL;
+	char *list_itr, *list_copy, *metric_name, *modifier;
+	int ret, count = 0;
 
-	nlist = strdup(list);
-	if (!nlist)
+	list_copy = strdup(list);
+	if (!list_copy)
 		return -ENOMEM;
-	llist = nlist;
+	list_itr = list_copy;
+
+	while ((metric_name = strsep(&list_itr, ",")) != NULL) {
+		modifier = strchr(metric_name, ':');
+		if (modifier)
+			*modifier++ = '\0';
 
-	strbuf_init(events, 100);
-	strbuf_addf(events, "%s", "");
+		ret = metricgroup__add_metric(metric_name, modifier,
+					      metric_no_group, metric_list,
+					      map);
+		if (ret == -EINVAL)
+			pr_err("Cannot find metric or group `%s'\n", metric_name);
 
-	while ((p = strsep(&llist, ",")) != NULL) {
-		ret = metricgroup__add_metric(p, metric_no_group, events,
-					      metric_list, map);
-		if (ret == -EINVAL) {
-			fprintf(stderr, "Cannot find metric or group `%s'\n",
-					p);
+		if (ret)
 			break;
-		}
+
+		count++;
 	}
-	free(nlist);
+	free(list_copy);
 
-	if (!ret)
+	if (!ret) {
+		/*
+		 * Warn about nmi_watchdog if any parsed metrics had the
+		 * NO_NMI_WATCHDOG constraint.
+		 */
 		metricgroup___watchdog_constraint_hint(NULL, true);
-
+		/* No metrics. */
+		if (count == 0)
+			return -EINVAL;
+	}
 	return ret;
 }
 
-static void metric__free_refs(struct metric *metric)
+static void metricgroup__free_metrics(struct list_head *metric_list)
 {
-	struct metric_ref_node *ref, *tmp;
+	struct metric *m, *tmp;
 
-	list_for_each_entry_safe(ref, tmp, &metric->metric_refs, list) {
-		list_del(&ref->list);
-		free(ref);
+	list_for_each_entry_safe (m, tmp, metric_list, nd) {
+		list_del_init(&m->nd);
+		metric__free(m);
 	}
 }
 
-static void metricgroup__free_metrics(struct list_head *metric_list)
+/**
+ * build_combined_expr_ctx - Make an expr_parse_ctx with all has_constraint
+ *                           metric IDs, as the IDs are held in a set,
+ *                           duplicates will be removed.
+ * @metric_list: List to take metrics from.
+ * @combined: Out argument for result.
+ */
+static int build_combined_expr_ctx(const struct list_head *metric_list,
+				   struct expr_parse_ctx **combined)
 {
-	struct metric *m, *tmp;
+	struct hashmap_entry *cur;
+	size_t bkt;
+	struct metric *m;
+	char *dup;
+	int ret;
 
-	list_for_each_entry_safe (m, tmp, metric_list, nd) {
-		metric__free_refs(m);
-		expr__ctx_clear(&m->pctx);
-		list_del_init(&m->nd);
-		free(m);
+	*combined = expr__ctx_new();
+	if (!*combined)
+		return -ENOMEM;
+
+	list_for_each_entry(m, metric_list, nd) {
+		if (m->has_constraint && !m->modifier) {
+			hashmap__for_each_entry(m->pctx->ids, cur, bkt) {
+				dup = strdup(cur->key);
+				if (!dup) {
+					ret = -ENOMEM;
+					goto err_out;
+				}
+				ret = expr__add_id(*combined, dup);
+				if (ret)
+					goto err_out;
+			}
+		}
 	}
+	return 0;
+err_out:
+	expr__ctx_free(*combined);
+	*combined = NULL;
+	return ret;
+}
+
+/**
+ * parse_ids - Build the event string for the ids and parse them creating an
+ *             evlist. The encoded metric_ids are decoded.
+ * @fake_pmu: used when testing metrics not supported by the current CPU.
+ * @ids: the event identifiers parsed from a metric.
+ * @modifier: any modifiers added to the events.
+ * @has_constraint: false if events should be placed in a weak group.
+ * @out_evlist: the created list of events.
+ */
+static int parse_ids(struct perf_pmu *fake_pmu, struct expr_parse_ctx *ids,
+		     const char *modifier, bool has_constraint, struct evlist **out_evlist)
+{
+	struct parse_events_error parse_error;
+	struct evlist *parsed_evlist;
+	struct strbuf events = STRBUF_INIT;
+	int ret;
+
+	*out_evlist = NULL;
+	if (hashmap__size(ids->ids) == 0) {
+		char *tmp;
+		/*
+		 * No ids/events in the expression parsing context. Events may
+		 * have been removed because of constant evaluation, e.g.:
+		 *  event1 if #smt_on else 0
+		 * Add a duration_time event to avoid a parse error on an empty
+		 * string.
+		 */
+		tmp = strdup("duration_time");
+		if (!tmp)
+			return -ENOMEM;
+
+		ids__insert(ids->ids, tmp);
+	}
+	ret = metricgroup__build_event_string(&events, ids, modifier,
+					      has_constraint);
+	if (ret)
+		return ret;
+
+	parsed_evlist = evlist__new();
+	if (!parsed_evlist) {
+		ret = -ENOMEM;
+		goto err_out;
+	}
+	pr_debug("Parsing metric events '%s'\n", events.buf);
+	parse_events_error__init(&parse_error);
+	ret = __parse_events(parsed_evlist, events.buf, &parse_error, fake_pmu);
+	if (ret) {
+		parse_events_error__print(&parse_error, events.buf);
+		goto err_out;
+	}
+	ret = decode_all_metric_ids(parsed_evlist, modifier);
+	if (ret)
+		goto err_out;
+
+	*out_evlist = parsed_evlist;
+	parsed_evlist = NULL;
+err_out:
+	parse_events_error__exit(&parse_error);
+	evlist__delete(parsed_evlist);
+	strbuf_release(&events);
+	return ret;
 }
 
 static int parse_groups(struct evlist *perf_evlist, const char *str,
 			bool metric_no_group,
 			bool metric_no_merge,
 			struct perf_pmu *fake_pmu,
-			struct rblist *metric_events,
-			struct pmu_events_map *map)
+			struct rblist *metric_events_list,
+			const struct pmu_events_map *map)
 {
-	struct parse_events_error parse_error;
-	struct strbuf extra_events;
+	struct evlist *combined_evlist = NULL;
 	LIST_HEAD(metric_list);
+	struct metric *m;
 	int ret;
 
-	if (metric_events->nr_entries == 0)
-		metricgroup__rblist_init(metric_events);
+	if (metric_events_list->nr_entries == 0)
+		metricgroup__rblist_init(metric_events_list);
 	ret = metricgroup__add_metric_list(str, metric_no_group,
-					   &extra_events, &metric_list, map);
+					   &metric_list, map);
 	if (ret)
 		goto out;
-	pr_debug("adding %s\n", extra_events.buf);
-	bzero(&parse_error, sizeof(parse_error));
-	ret = __parse_events(perf_evlist, extra_events.buf, &parse_error, fake_pmu);
-	if (ret) {
-		parse_events_print_error(&parse_error, extra_events.buf);
-		goto out;
+
+	/* Sort metrics from largest to smallest. */
+	list_sort(NULL, &metric_list, metric_list_cmp);
+
+	if (!metric_no_merge) {
+		struct expr_parse_ctx *combined = NULL;
+
+		ret = build_combined_expr_ctx(&metric_list, &combined);
+
+		if (!ret && combined && hashmap__size(combined->ids)) {
+			ret = parse_ids(fake_pmu, combined, /*modifier=*/NULL,
+					/*has_constraint=*/true,
+					&combined_evlist);
+		}
+		if (combined)
+			expr__ctx_free(combined);
+
+		if (ret)
+			goto out;
 	}
-	ret = metricgroup__setup_events(&metric_list, metric_no_merge,
-					perf_evlist, metric_events);
+
+	list_for_each_entry(m, &metric_list, nd) {
+		struct metric_event *me;
+		struct evsel **metric_events;
+		struct evlist *metric_evlist = NULL;
+		struct metric *n;
+		struct metric_expr *expr;
+
+		if (combined_evlist && m->has_constraint) {
+			metric_evlist = combined_evlist;
+		} else if (!metric_no_merge) {
+			/*
+			 * See if the IDs for this metric are a subset of an
+			 * earlier metric.
+			 */
+			list_for_each_entry(n, &metric_list, nd) {
+				if (m == n)
+					break;
+
+				if (n->evlist == NULL)
+					continue;
+
+				if ((!m->modifier && n->modifier) ||
+				    (m->modifier && !n->modifier) ||
+				    (m->modifier && n->modifier &&
+					    strcmp(m->modifier, n->modifier)))
+					continue;
+
+				if (expr__subset_of_ids(n->pctx, m->pctx)) {
+					pr_debug("Events in '%s' fully contained within '%s'\n",
+						 m->metric_name, n->metric_name);
+					metric_evlist = n->evlist;
+					break;
+				}
+
+			}
+		}
+		if (!metric_evlist) {
+			ret = parse_ids(fake_pmu, m->pctx, m->modifier,
+					m->has_constraint, &m->evlist);
+			if (ret)
+				goto out;
+
+			metric_evlist = m->evlist;
+		}
+		ret = setup_metric_events(m->pctx->ids, metric_evlist, &metric_events);
+		if (ret) {
+			pr_debug("Cannot resolve IDs for %s: %s\n",
+				m->metric_name, m->metric_expr);
+			goto out;
+		}
+
+		me = metricgroup__lookup(metric_events_list, metric_events[0], true);
+
+		expr = malloc(sizeof(struct metric_expr));
+		if (!expr) {
+			ret = -ENOMEM;
+			free(metric_events);
+			goto out;
+		}
+
+		expr->metric_refs = m->metric_refs;
+		m->metric_refs = NULL;
+		expr->metric_expr = m->metric_expr;
+		if (m->modifier) {
+			char *tmp;
+
+			if (asprintf(&tmp, "%s:%s", m->metric_name, m->modifier) < 0)
+				expr->metric_name = NULL;
+			else
+				expr->metric_name = tmp;
+		} else
+			expr->metric_name = strdup(m->metric_name);
+
+		if (!expr->metric_name) {
+			ret = -ENOMEM;
+			free(metric_events);
+			goto out;
+		}
+		expr->metric_unit = m->metric_unit;
+		expr->metric_events = metric_events;
+		expr->runtime = m->pctx->runtime;
+		list_add(&expr->nd, &me->head);
+	}
+
+
+	if (combined_evlist) {
+		evlist__splice_list_tail(perf_evlist, &combined_evlist->core.entries);
+		evlist__delete(combined_evlist);
+	}
+
+	list_for_each_entry(m, &metric_list, nd) {
+		if (m->evlist)
+			evlist__splice_list_tail(perf_evlist, &m->evlist->core.entries);
+	}
+
 out:
 	metricgroup__free_metrics(&metric_list);
-	strbuf_release(&extra_events);
 	return ret;
 }
 
@@ -1256,14 +1505,14 @@ int metricgroup__parse_groups(const struct option *opt,
 			      struct rblist *metric_events)
 {
 	struct evlist *perf_evlist = *(struct evlist **)opt->value;
-	struct pmu_events_map *map = pmu_events_map__find();
+	const struct pmu_events_map *map = pmu_events_map__find();
 
 	return parse_groups(perf_evlist, str, metric_no_group,
 			    metric_no_merge, NULL, metric_events, map);
 }
 
 int metricgroup__parse_groups_test(struct evlist *evlist,
-				   struct pmu_events_map *map,
+				   const struct pmu_events_map *map,
 				   const char *str,
 				   bool metric_no_group,
 				   bool metric_no_merge,
@@ -1275,8 +1524,8 @@ int metricgroup__parse_groups_test(struct evlist *evlist,
 
 bool metricgroup__has_metric(const char *metric)
 {
-	struct pmu_events_map *map = pmu_events_map__find();
-	struct pmu_event *pe;
+	const struct pmu_events_map *map = pmu_events_map__find();
+	const struct pmu_event *pe;
 	int i;
 
 	if (!map)
@@ -1328,7 +1577,10 @@ int metricgroup__copy_metric_events(struct evlist *evlist, struct cgroup *cgrp,
 				return -ENOMEM;
 
 			new_expr->metric_expr = old_expr->metric_expr;
-			new_expr->metric_name = old_expr->metric_name;
+			new_expr->metric_name = strdup(old_expr->metric_name);
+			if (!new_expr->metric_name)
+				return -ENOMEM;
+
 			new_expr->metric_unit = old_expr->metric_unit;
 			new_expr->runtime = old_expr->runtime;
 
diff --git a/tools/perf/util/metricgroup.h b/tools/perf/util/metricgroup.h
index cc4a92492a61..2b42b778d1bf 100644
--- a/tools/perf/util/metricgroup.h
+++ b/tools/perf/util/metricgroup.h
@@ -14,24 +14,51 @@ struct rblist;
 struct pmu_events_map;
 struct cgroup;
 
+/**
+ * A node in a rblist keyed by the evsel. The global rblist of metric events
+ * generally exists in perf_stat_config. The evsel is looked up in the rblist
+ * yielding a list of metric_expr.
+ */
 struct metric_event {
 	struct rb_node nd;
 	struct evsel *evsel;
 	struct list_head head; /* list of metric_expr */
 };
 
+/**
+ * A metric referenced by a metric_expr. When parsing a metric expression IDs
+ * will be looked up, matching either a value (from metric_events) or a
+ * metric_ref. A metric_ref will then be parsed recursively. The metric_refs and
+ * metric_events need to be known before parsing so that their values may be
+ * placed in the parse context for lookup.
+ */
 struct metric_ref {
 	const char *metric_name;
 	const char *metric_expr;
 };
 
+/**
+ * One in a list of metric_expr associated with an evsel. The data is used to
+ * generate a metric value during stat output.
+ */
 struct metric_expr {
 	struct list_head nd;
+	/** The expression to parse, for example, "instructions/cycles". */
 	const char *metric_expr;
+	/** The name of the meric such as "IPC". */
 	const char *metric_name;
+	/**
+	 * The "ScaleUnit" that scales and adds a unit to the metric during
+	 * output. For example, "6.4e-05MiB" means to scale the resulting metric
+	 * by 6.4e-05 (typically converting a unit like cache lines to something
+	 * more human intelligible) and then add "MiB" afterward when displayed.
+	 */
 	const char *metric_unit;
+	/** Null terminated array of events used by the metric. */
 	struct evsel **metric_events;
+	/** Null terminated array of referenced metrics. */
 	struct metric_ref *metric_refs;
+	/** A value substituted for '?' during parsing. */
 	int runtime;
 };
 
@@ -43,19 +70,19 @@ int metricgroup__parse_groups(const struct option *opt,
 			      bool metric_no_group,
 			      bool metric_no_merge,
 			      struct rblist *metric_events);
-struct pmu_event *metricgroup__find_metric(const char *metric,
-					   struct pmu_events_map *map);
+const struct pmu_event *metricgroup__find_metric(const char *metric,
+						 const struct pmu_events_map *map);
 int metricgroup__parse_groups_test(struct evlist *evlist,
-				   struct pmu_events_map *map,
+				   const struct pmu_events_map *map,
 				   const char *str,
 				   bool metric_no_group,
 				   bool metric_no_merge,
 				   struct rblist *metric_events);
 
 void metricgroup__print(bool metrics, bool groups, char *filter,
-			bool raw, bool details);
+			bool raw, bool details, const char *pmu_name);
 bool metricgroup__has_metric(const char *metric);
-int arch_get_runtimeparam(struct pmu_event *pe __maybe_unused);
+int arch_get_runtimeparam(const struct pmu_event *pe __maybe_unused);
 void metricgroup__rblist_exit(struct rblist *metric_events);
 
 int metricgroup__copy_metric_events(struct evlist *evlist, struct cgroup *cgrp,
diff --git a/tools/perf/util/mmap.c b/tools/perf/util/mmap.c
index 512dc8b9c168..23ecdba9e670 100644
--- a/tools/perf/util/mmap.c
+++ b/tools/perf/util/mmap.c
@@ -350,3 +350,14 @@ int perf_mmap__push(struct mmap *md, void *to,
 out:
 	return rc;
 }
+
+int mmap_cpu_mask__duplicate(struct mmap_cpu_mask *original, struct mmap_cpu_mask *clone)
+{
+	clone->nbits = original->nbits;
+	clone->bits  = bitmap_zalloc(original->nbits);
+	if (!clone->bits)
+		return -ENOMEM;
+
+	memcpy(clone->bits, original->bits, MMAP_CPU_MASK_BYTES(original));
+	return 0;
+}
diff --git a/tools/perf/util/mmap.h b/tools/perf/util/mmap.h
index af33118354dd..8e259b9610f8 100644
--- a/tools/perf/util/mmap.h
+++ b/tools/perf/util/mmap.h
@@ -64,4 +64,7 @@ size_t mmap__mmap_len(struct mmap *map);
 
 void mmap_cpu_mask__scnprintf(struct mmap_cpu_mask *mask, const char *tag);
 
+int mmap_cpu_mask__duplicate(struct mmap_cpu_mask *original,
+				struct mmap_cpu_mask *clone);
+
 #endif /*__PERF_MMAP_H */
diff --git a/tools/perf/util/parse-events-hybrid.c b/tools/perf/util/parse-events-hybrid.c
index b234d95fb10a..9fc86971027b 100644
--- a/tools/perf/util/parse-events-hybrid.c
+++ b/tools/perf/util/parse-events-hybrid.c
@@ -38,7 +38,8 @@ static void config_hybrid_attr(struct perf_event_attr *attr,
 
 static int create_event_hybrid(__u32 config_type, int *idx,
 			       struct list_head *list,
-			       struct perf_event_attr *attr, char *name,
+			       struct perf_event_attr *attr, const char *name,
+			       const char *metric_id,
 			       struct list_head *config_terms,
 			       struct perf_pmu *pmu)
 {
@@ -47,7 +48,7 @@ static int create_event_hybrid(__u32 config_type, int *idx,
 	__u64 config = attr->config;
 
 	config_hybrid_attr(attr, config_type, pmu->type);
-	evsel = parse_events__add_event_hybrid(list, idx, attr, name,
+	evsel = parse_events__add_event_hybrid(list, idx, attr, name, metric_id,
 					       pmu, config_terms);
 	if (evsel)
 		evsel->pmu_name = strdup(pmu->name);
@@ -70,7 +71,8 @@ static int pmu_cmp(struct parse_events_state *parse_state,
 
 static int add_hw_hybrid(struct parse_events_state *parse_state,
 			 struct list_head *list, struct perf_event_attr *attr,
-			 char *name, struct list_head *config_terms)
+			 const char *name, const char *metric_id,
+			 struct list_head *config_terms)
 {
 	struct perf_pmu *pmu;
 	int ret;
@@ -84,7 +86,7 @@ static int add_hw_hybrid(struct parse_events_state *parse_state,
 		copy_config_terms(&terms, config_terms);
 		ret = create_event_hybrid(PERF_TYPE_HARDWARE,
 					  &parse_state->idx, list, attr, name,
-					  &terms, pmu);
+					  metric_id, &terms, pmu);
 		free_config_terms(&terms);
 		if (ret)
 			return ret;
@@ -94,14 +96,16 @@ static int add_hw_hybrid(struct parse_events_state *parse_state,
 }
 
 static int create_raw_event_hybrid(int *idx, struct list_head *list,
-				   struct perf_event_attr *attr, char *name,
+				   struct perf_event_attr *attr,
+				   const char *name,
+				   const char *metric_id,
 				   struct list_head *config_terms,
 				   struct perf_pmu *pmu)
 {
 	struct evsel *evsel;
 
 	attr->type = pmu->type;
-	evsel = parse_events__add_event_hybrid(list, idx, attr, name,
+	evsel = parse_events__add_event_hybrid(list, idx, attr, name, metric_id,
 					       pmu, config_terms);
 	if (evsel)
 		evsel->pmu_name = strdup(pmu->name);
@@ -113,7 +117,8 @@ static int create_raw_event_hybrid(int *idx, struct list_head *list,
 
 static int add_raw_hybrid(struct parse_events_state *parse_state,
 			  struct list_head *list, struct perf_event_attr *attr,
-			  char *name, struct list_head *config_terms)
+			  const char *name, const char *metric_id,
+			  struct list_head *config_terms)
 {
 	struct perf_pmu *pmu;
 	int ret;
@@ -126,7 +131,7 @@ static int add_raw_hybrid(struct parse_events_state *parse_state,
 
 		copy_config_terms(&terms, config_terms);
 		ret = create_raw_event_hybrid(&parse_state->idx, list, attr,
-					      name, &terms, pmu);
+					      name, metric_id, &terms, pmu);
 		free_config_terms(&terms);
 		if (ret)
 			return ret;
@@ -138,7 +143,8 @@ static int add_raw_hybrid(struct parse_events_state *parse_state,
 int parse_events__add_numeric_hybrid(struct parse_events_state *parse_state,
 				     struct list_head *list,
 				     struct perf_event_attr *attr,
-				     char *name, struct list_head *config_terms,
+				     const char *name, const char *metric_id,
+				     struct list_head *config_terms,
 				     bool *hybrid)
 {
 	*hybrid = false;
@@ -150,16 +156,18 @@ int parse_events__add_numeric_hybrid(struct parse_events_state *parse_state,
 
 	*hybrid = true;
 	if (attr->type != PERF_TYPE_RAW) {
-		return add_hw_hybrid(parse_state, list, attr, name,
+		return add_hw_hybrid(parse_state, list, attr, name, metric_id,
 				     config_terms);
 	}
 
-	return add_raw_hybrid(parse_state, list, attr, name,
+	return add_raw_hybrid(parse_state, list, attr, name, metric_id,
 			      config_terms);
 }
 
 int parse_events__add_cache_hybrid(struct list_head *list, int *idx,
-				   struct perf_event_attr *attr, char *name,
+				   struct perf_event_attr *attr,
+				   const char *name,
+				   const char *metric_id,
 				   struct list_head *config_terms,
 				   bool *hybrid,
 				   struct parse_events_state *parse_state)
@@ -180,7 +188,7 @@ int parse_events__add_cache_hybrid(struct list_head *list, int *idx,
 
 		copy_config_terms(&terms, config_terms);
 		ret = create_event_hybrid(PERF_TYPE_HW_CACHE, idx, list,
-					  attr, name, &terms, pmu);
+					  attr, name, metric_id, &terms, pmu);
 		free_config_terms(&terms);
 		if (ret)
 			return ret;
diff --git a/tools/perf/util/parse-events-hybrid.h b/tools/perf/util/parse-events-hybrid.h
index f33bd67aa851..cbc05fec02a2 100644
--- a/tools/perf/util/parse-events-hybrid.h
+++ b/tools/perf/util/parse-events-hybrid.h
@@ -11,11 +11,13 @@
 int parse_events__add_numeric_hybrid(struct parse_events_state *parse_state,
 				     struct list_head *list,
 				     struct perf_event_attr *attr,
-				     char *name, struct list_head *config_terms,
+				     const char *name, const char *metric_id,
+				     struct list_head *config_terms,
 				     bool *hybrid);
 
 int parse_events__add_cache_hybrid(struct list_head *list, int *idx,
-				   struct perf_event_attr *attr, char *name,
+				   struct perf_event_attr *attr,
+				   const char *name, const char *metric_id,
 				   struct list_head *config_terms,
 				   bool *hybrid,
 				   struct parse_events_state *parse_state);
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 51a2219df601..ba74fdf74af9 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -19,8 +19,6 @@
 #include <subcmd/exec-cmd.h>
 #include "string2.h"
 #include "strlist.h"
-#include "symbol.h"
-#include "header.h"
 #include "bpf-loader.h"
 #include "debug.h"
 #include <api/fs/tracing_path.h>
@@ -193,39 +191,6 @@ static int tp_event_has_id(const char *dir_path, struct dirent *evt_dir)
 
 #define MAX_EVENT_LENGTH 512
 
-void parse_events__handle_error(struct parse_events_error *err, int idx,
-				char *str, char *help)
-{
-	if (WARN(!str, "WARNING: failed to provide error string\n")) {
-		free(help);
-		return;
-	}
-	switch (err->num_errors) {
-	case 0:
-		err->idx = idx;
-		err->str = str;
-		err->help = help;
-		break;
-	case 1:
-		err->first_idx = err->idx;
-		err->idx = idx;
-		err->first_str = err->str;
-		err->str = str;
-		err->first_help = err->help;
-		err->help = help;
-		break;
-	default:
-		pr_debug("Multiple errors dropping message: %s (%s)\n",
-			err->str, err->help);
-		free(err->str);
-		err->str = str;
-		free(err->help);
-		err->help = help;
-		break;
-	}
-	err->num_errors++;
-}
-
 struct tracepoint_path *tracepoint_id_to_path(u64 config)
 {
 	struct tracepoint_path *path = NULL;
@@ -334,12 +299,7 @@ const char *event_type(int type)
 	return "unknown";
 }
 
-static int parse_events__is_name_term(struct parse_events_term *term)
-{
-	return term->type_term == PARSE_EVENTS__TERM_TYPE_NAME;
-}
-
-static char *get_config_name(struct list_head *head_terms)
+static char *get_config_str(struct list_head *head_terms, int type_term)
 {
 	struct parse_events_term *term;
 
@@ -347,17 +307,27 @@ static char *get_config_name(struct list_head *head_terms)
 		return NULL;
 
 	list_for_each_entry(term, head_terms, list)
-		if (parse_events__is_name_term(term))
+		if (term->type_term == type_term)
 			return term->val.str;
 
 	return NULL;
 }
 
+static char *get_config_metric_id(struct list_head *head_terms)
+{
+	return get_config_str(head_terms, PARSE_EVENTS__TERM_TYPE_METRIC_ID);
+}
+
+static char *get_config_name(struct list_head *head_terms)
+{
+	return get_config_str(head_terms, PARSE_EVENTS__TERM_TYPE_NAME);
+}
+
 static struct evsel *
 __add_event(struct list_head *list, int *idx,
 	    struct perf_event_attr *attr,
 	    bool init_attr,
-	    char *name, struct perf_pmu *pmu,
+	    const char *name, const char *metric_id, struct perf_pmu *pmu,
 	    struct list_head *config_terms, bool auto_merge_stats,
 	    const char *cpu_list)
 {
@@ -386,6 +356,9 @@ __add_event(struct list_head *list, int *idx,
 	if (name)
 		evsel->name = strdup(name);
 
+	if (metric_id)
+		evsel->metric_id = strdup(metric_id);
+
 	if (config_terms)
 		list_splice_init(config_terms, &evsel->config_terms);
 
@@ -396,18 +369,21 @@ __add_event(struct list_head *list, int *idx,
 }
 
 struct evsel *parse_events__add_event(int idx, struct perf_event_attr *attr,
-					char *name, struct perf_pmu *pmu)
+				      const char *name, const char *metric_id,
+				      struct perf_pmu *pmu)
 {
-	return __add_event(NULL, &idx, attr, false, name, pmu, NULL, false,
-			   NULL);
+	return __add_event(/*list=*/NULL, &idx, attr, /*init_attr=*/false, name,
+			   metric_id, pmu, /*config_terms=*/NULL,
+			   /*auto_merge_stats=*/false, /*cpu_list=*/NULL);
 }
 
 static int add_event(struct list_head *list, int *idx,
-		     struct perf_event_attr *attr, char *name,
-		     struct list_head *config_terms)
+		     struct perf_event_attr *attr, const char *name,
+		     const char *metric_id, struct list_head *config_terms)
 {
-	return __add_event(list, idx, attr, true, name, NULL, config_terms,
-			   false, NULL) ? 0 : -ENOMEM;
+	return __add_event(list, idx, attr, /*init_attr*/true, name, metric_id,
+			   /*pmu=*/NULL, config_terms,
+			   /*auto_merge_stats=*/false, /*cpu_list=*/NULL) ? 0 : -ENOMEM;
 }
 
 static int add_event_tool(struct list_head *list, int *idx,
@@ -419,13 +395,17 @@ static int add_event_tool(struct list_head *list, int *idx,
 		.config = PERF_COUNT_SW_DUMMY,
 	};
 
-	evsel = __add_event(list, idx, &attr, true, NULL, NULL, NULL, false,
-			    "0");
+	evsel = __add_event(list, idx, &attr, /*init_attr=*/true, /*name=*/NULL,
+			    /*metric_id=*/NULL, /*pmu=*/NULL,
+			    /*config_terms=*/NULL, /*auto_merge_stats=*/false,
+			    /*cpu_list=*/"0");
 	if (!evsel)
 		return -ENOMEM;
 	evsel->tool_event = tool_event;
-	if (tool_event == PERF_TOOL_DURATION_TIME)
-		evsel->unit = "ns";
+	if (tool_event == PERF_TOOL_DURATION_TIME) {
+		free((char *)evsel->unit);
+		evsel->unit = strdup("ns");
+	}
 	return 0;
 }
 
@@ -466,7 +446,8 @@ int parse_events_add_cache(struct list_head *list, int *idx,
 {
 	struct perf_event_attr attr;
 	LIST_HEAD(config_terms);
-	char name[MAX_NAME_LEN], *config_name;
+	char name[MAX_NAME_LEN];
+	const char *config_name, *metric_id;
 	int cache_type = -1, cache_op = -1, cache_result = -1;
 	char *op_result[2] = { op_result1, op_result2 };
 	int i, n, ret;
@@ -531,13 +512,17 @@ int parse_events_add_cache(struct list_head *list, int *idx,
 			return -ENOMEM;
 	}
 
+	metric_id = get_config_metric_id(head_config);
 	ret = parse_events__add_cache_hybrid(list, idx, &attr,
-					     config_name ? : name, &config_terms,
+					     config_name ? : name,
+					     metric_id,
+					     &config_terms,
 					     &hybrid, parse_state);
 	if (hybrid)
 		goto out_free_terms;
 
-	ret = add_event(list, idx, &attr, config_name ? : name, &config_terms);
+	ret = add_event(list, idx, &attr, config_name ? : name, metric_id,
+			&config_terms);
 out_free_terms:
 	free_config_terms(&config_terms);
 	return ret;
@@ -571,7 +556,7 @@ static void tracepoint_error(struct parse_events_error *e, int err,
 	}
 
 	tracing_path__strerror_open_tp(err, help, sizeof(help), sys, name);
-	parse_events__handle_error(e, 0, strdup(str), strdup(help));
+	parse_events_error__handle(e, 0, strdup(str), strdup(help));
 }
 
 static int add_tracepoint(struct list_head *list, int *idx,
@@ -795,7 +780,7 @@ int parse_events_load_bpf_obj(struct parse_events_state *parse_state,
 
 	return 0;
 errout:
-	parse_events__handle_error(parse_state->error, 0,
+	parse_events_error__handle(parse_state->error, 0,
 				strdup(errbuf), strdup("(add -v to see detail)"));
 	return err;
 }
@@ -815,7 +800,7 @@ parse_events_config_bpf(struct parse_events_state *parse_state,
 		int err;
 
 		if (term->type_term != PARSE_EVENTS__TERM_TYPE_USER) {
-			parse_events__handle_error(parse_state->error, term->err_term,
+			parse_events_error__handle(parse_state->error, term->err_term,
 						strdup("Invalid config term for BPF object"),
 						NULL);
 			return -EINVAL;
@@ -835,7 +820,7 @@ parse_events_config_bpf(struct parse_events_state *parse_state,
 			else
 				idx = term->err_term + error_pos;
 
-			parse_events__handle_error(parse_state->error, idx,
+			parse_events_error__handle(parse_state->error, idx,
 						strdup(errbuf),
 						strdup(
 "Hint:\tValid config terms:\n"
@@ -907,7 +892,7 @@ int parse_events_load_bpf(struct parse_events_state *parse_state,
 						   -err, errbuf,
 						   sizeof(errbuf));
 
-		parse_events__handle_error(parse_state->error, 0,
+		parse_events_error__handle(parse_state->error, 0,
 					strdup(errbuf), strdup("(add -v to see detail)"));
 		return err;
 	}
@@ -931,7 +916,7 @@ int parse_events_load_bpf_obj(struct parse_events_state *parse_state,
 			      struct bpf_object *obj __maybe_unused,
 			      struct list_head *head_config __maybe_unused)
 {
-	parse_events__handle_error(parse_state->error, 0,
+	parse_events_error__handle(parse_state->error, 0,
 				   strdup("BPF support is not compiled"),
 				   strdup("Make sure libbpf-devel is available at build time."));
 	return -ENOTSUP;
@@ -943,7 +928,7 @@ int parse_events_load_bpf(struct parse_events_state *parse_state,
 			  bool source __maybe_unused,
 			  struct list_head *head_config __maybe_unused)
 {
-	parse_events__handle_error(parse_state->error, 0,
+	parse_events_error__handle(parse_state->error, 0,
 				   strdup("BPF support is not compiled"),
 				   strdup("Make sure libbpf-devel is available at build time."));
 	return -ENOTSUP;
@@ -1014,7 +999,8 @@ int parse_events_add_breakpoint(struct list_head *list, int *idx,
 	attr.type = PERF_TYPE_BREAKPOINT;
 	attr.sample_period = 1;
 
-	return add_event(list, idx, &attr, NULL, NULL);
+	return add_event(list, idx, &attr, /*name=*/NULL, /*mertic_id=*/NULL,
+			 /*config_terms=*/NULL);
 }
 
 static int check_type_val(struct parse_events_term *term,
@@ -1025,7 +1011,7 @@ static int check_type_val(struct parse_events_term *term,
 		return 0;
 
 	if (err) {
-		parse_events__handle_error(err, term->err_val,
+		parse_events_error__handle(err, term->err_val,
 					type == PARSE_EVENTS__TERM_TYPE_NUM
 					? strdup("expected numeric value")
 					: strdup("expected string value"),
@@ -1059,6 +1045,7 @@ static const char *config_term_names[__PARSE_EVENTS__TERM_TYPE_NR] = {
 	[PARSE_EVENTS__TERM_TYPE_PERCORE]		= "percore",
 	[PARSE_EVENTS__TERM_TYPE_AUX_OUTPUT]		= "aux-output",
 	[PARSE_EVENTS__TERM_TYPE_AUX_SAMPLE_SIZE]	= "aux-sample-size",
+	[PARSE_EVENTS__TERM_TYPE_METRIC_ID]		= "metric-id",
 };
 
 static bool config_term_shrinked;
@@ -1069,7 +1056,7 @@ config_term_avail(int term_type, struct parse_events_error *err)
 	char *err_str;
 
 	if (term_type < 0 || term_type >= __PARSE_EVENTS__TERM_TYPE_NR) {
-		parse_events__handle_error(err, -1,
+		parse_events_error__handle(err, -1,
 					strdup("Invalid term_type"), NULL);
 		return false;
 	}
@@ -1081,6 +1068,7 @@ config_term_avail(int term_type, struct parse_events_error *err)
 	case PARSE_EVENTS__TERM_TYPE_CONFIG1:
 	case PARSE_EVENTS__TERM_TYPE_CONFIG2:
 	case PARSE_EVENTS__TERM_TYPE_NAME:
+	case PARSE_EVENTS__TERM_TYPE_METRIC_ID:
 	case PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD:
 	case PARSE_EVENTS__TERM_TYPE_PERCORE:
 		return true;
@@ -1091,7 +1079,7 @@ config_term_avail(int term_type, struct parse_events_error *err)
 		/* term_type is validated so indexing is safe */
 		if (asprintf(&err_str, "'%s' is not usable in 'perf stat'",
 				config_term_names[term_type]) >= 0)
-			parse_events__handle_error(err, -1, err_str, NULL);
+			parse_events_error__handle(err, -1, err_str, NULL);
 		return false;
 	}
 }
@@ -1135,7 +1123,7 @@ do {									   \
 		if (strcmp(term->val.str, "no") &&
 		    parse_branch_str(term->val.str,
 				    &attr->branch_sample_type)) {
-			parse_events__handle_error(err, term->err_val,
+			parse_events_error__handle(err, term->err_val,
 					strdup("invalid branch sample type"),
 					NULL);
 			return -EINVAL;
@@ -1144,7 +1132,7 @@ do {									   \
 	case PARSE_EVENTS__TERM_TYPE_TIME:
 		CHECK_TYPE_VAL(NUM);
 		if (term->val.num > 1) {
-			parse_events__handle_error(err, term->err_val,
+			parse_events_error__handle(err, term->err_val,
 						strdup("expected 0 or 1"),
 						NULL);
 			return -EINVAL;
@@ -1171,6 +1159,9 @@ do {									   \
 	case PARSE_EVENTS__TERM_TYPE_NAME:
 		CHECK_TYPE_VAL(STR);
 		break;
+	case PARSE_EVENTS__TERM_TYPE_METRIC_ID:
+		CHECK_TYPE_VAL(STR);
+		break;
 	case PARSE_EVENTS__TERM_TYPE_MAX_STACK:
 		CHECK_TYPE_VAL(NUM);
 		break;
@@ -1180,7 +1171,7 @@ do {									   \
 	case PARSE_EVENTS__TERM_TYPE_PERCORE:
 		CHECK_TYPE_VAL(NUM);
 		if ((unsigned int)term->val.num > 1) {
-			parse_events__handle_error(err, term->err_val,
+			parse_events_error__handle(err, term->err_val,
 						strdup("expected 0 or 1"),
 						NULL);
 			return -EINVAL;
@@ -1192,14 +1183,14 @@ do {									   \
 	case PARSE_EVENTS__TERM_TYPE_AUX_SAMPLE_SIZE:
 		CHECK_TYPE_VAL(NUM);
 		if (term->val.num > UINT_MAX) {
-			parse_events__handle_error(err, term->err_val,
+			parse_events_error__handle(err, term->err_val,
 						strdup("too big"),
 						NULL);
 			return -EINVAL;
 		}
 		break;
 	default:
-		parse_events__handle_error(err, term->err_term,
+		parse_events_error__handle(err, term->err_term,
 				strdup("unknown term"),
 				parse_events_formats_error_string(NULL));
 		return -EINVAL;
@@ -1253,7 +1244,7 @@ static int config_term_tracepoint(struct perf_event_attr *attr,
 		return config_term_common(attr, term, err);
 	default:
 		if (err) {
-			parse_events__handle_error(err, term->err_term,
+			parse_events_error__handle(err, term->err_term,
 				strdup("unknown term"),
 				strdup("valid terms: call-graph,stack-size\n"));
 		}
@@ -1440,6 +1431,7 @@ int parse_events_add_numeric(struct parse_events_state *parse_state,
 {
 	struct perf_event_attr attr;
 	LIST_HEAD(config_terms);
+	const char *name, *metric_id;
 	bool hybrid;
 	int ret;
 
@@ -1456,14 +1448,16 @@ int parse_events_add_numeric(struct parse_events_state *parse_state,
 			return -ENOMEM;
 	}
 
+	name = get_config_name(head_config);
+	metric_id = get_config_metric_id(head_config);
 	ret = parse_events__add_numeric_hybrid(parse_state, list, &attr,
-					       get_config_name(head_config),
+					       name, metric_id,
 					       &config_terms, &hybrid);
 	if (hybrid)
 		goto out_free_terms;
 
-	ret = add_event(list, &parse_state->idx, &attr,
-			get_config_name(head_config), &config_terms);
+	ret = add_event(list, &parse_state->idx, &attr, name, metric_id,
+			&config_terms);
 out_free_terms:
 	free_config_terms(&config_terms);
 	return ret;
@@ -1471,7 +1465,7 @@ out_free_terms:
 
 int parse_events_add_tool(struct parse_events_state *parse_state,
 			  struct list_head *list,
-			  enum perf_tool_event tool_event)
+			  int tool_event)
 {
 	return add_event_tool(list, &parse_state->idx, tool_event);
 }
@@ -1549,7 +1543,7 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
 		if (asprintf(&err_str,
 				"Cannot find PMU `%s'. Missing kernel support?",
 				name) >= 0)
-			parse_events__handle_error(err, 0, err_str, NULL);
+			parse_events_error__handle(err, 0, err_str, NULL);
 		return -EINVAL;
 	}
 
@@ -1564,8 +1558,11 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
 
 	if (!head_config) {
 		attr.type = pmu->type;
-		evsel = __add_event(list, &parse_state->idx, &attr, true, NULL,
-				    pmu, NULL, auto_merge_stats, NULL);
+		evsel = __add_event(list, &parse_state->idx, &attr,
+				    /*init_attr=*/true, /*name=*/NULL,
+				    /*metric_id=*/NULL, pmu,
+				    /*config_terms=*/NULL, auto_merge_stats,
+				    /*cpu_list=*/NULL);
 		if (evsel) {
 			evsel->pmu_name = name ? strdup(name) : NULL;
 			evsel->use_uncore_alias = use_uncore_alias;
@@ -1618,9 +1615,10 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
 		return -EINVAL;
 	}
 
-	evsel = __add_event(list, &parse_state->idx, &attr, true,
-			    get_config_name(head_config), pmu,
-			    &config_terms, auto_merge_stats, NULL);
+	evsel = __add_event(list, &parse_state->idx, &attr, /*init_attr=*/true,
+			    get_config_name(head_config),
+			    get_config_metric_id(head_config), pmu,
+			    &config_terms, auto_merge_stats, /*cpu_list=*/NULL);
 	if (!evsel)
 		return -ENOMEM;
 
@@ -1634,7 +1632,8 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
 	if (parse_state->fake_pmu)
 		return 0;
 
-	evsel->unit = info.unit;
+	free((char *)evsel->unit);
+	evsel->unit = strdup(info.unit);
 	evsel->scale = info.scale;
 	evsel->per_pkg = info.per_pkg;
 	evsel->snapshot = info.snapshot;
@@ -1644,44 +1643,50 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
 }
 
 int parse_events_multi_pmu_add(struct parse_events_state *parse_state,
-			       char *str, struct list_head **listp)
+			       char *str, struct list_head *head,
+			       struct list_head **listp)
 {
 	struct parse_events_term *term;
-	struct list_head *list;
+	struct list_head *list = NULL;
 	struct perf_pmu *pmu = NULL;
 	int ok = 0;
+	char *config;
 
 	*listp = NULL;
+
+	if (!head) {
+		head = malloc(sizeof(struct list_head));
+		if (!head)
+			goto out_err;
+
+		INIT_LIST_HEAD(head);
+	}
+	config = strdup(str);
+	if (!config)
+		goto out_err;
+
+	if (parse_events_term__num(&term,
+				   PARSE_EVENTS__TERM_TYPE_USER,
+				   config, 1, false, &config,
+					NULL) < 0) {
+		free(config);
+		goto out_err;
+	}
+	list_add_tail(&term->list, head);
+
+
 	/* Add it for all PMUs that support the alias */
 	list = malloc(sizeof(struct list_head));
 	if (!list)
-		return -1;
+		goto out_err;
+
 	INIT_LIST_HEAD(list);
+
 	while ((pmu = perf_pmu__scan(pmu)) != NULL) {
 		struct perf_pmu_alias *alias;
 
 		list_for_each_entry(alias, &pmu->aliases, list) {
 			if (!strcasecmp(alias->name, str)) {
-				struct list_head *head;
-				char *config;
-
-				head = malloc(sizeof(struct list_head));
-				if (!head)
-					return -1;
-				INIT_LIST_HEAD(head);
-				config = strdup(str);
-				if (!config)
-					return -1;
-				if (parse_events_term__num(&term,
-						   PARSE_EVENTS__TERM_TYPE_USER,
-						   config, 1, false, &config,
-						   NULL) < 0) {
-					free(list);
-					free(config);
-					return -1;
-				}
-				list_add_tail(&term->list, head);
-
 				if (!parse_events_add_pmu(parse_state, list,
 							  pmu->name, head,
 							  true, true)) {
@@ -1689,17 +1694,17 @@ int parse_events_multi_pmu_add(struct parse_events_state *parse_state,
 						 pmu->name, alias->str);
 					ok++;
 				}
-
-				parse_events_terms__delete(head);
 			}
 		}
 	}
-	if (!ok) {
+out_err:
+	if (ok)
+		*listp = list;
+	else
 		free(list);
-		return -1;
-	}
-	*listp = list;
-	return 0;
+
+	parse_events_terms__delete(head);
+	return ok ? 0 : -1;
 }
 
 int parse_events__modifier_group(struct list_head *list,
@@ -2029,7 +2034,7 @@ int parse_events__modifier_event(struct list_head *list, char *str, bool add)
 	return 0;
 }
 
-int parse_events_name(struct list_head *list, char *name)
+int parse_events_name(struct list_head *list, const char *name)
 {
 	struct evsel *evsel;
 
@@ -2299,6 +2304,52 @@ int __parse_events(struct evlist *evlist, const char *str,
 	return ret;
 }
 
+void parse_events_error__init(struct parse_events_error *err)
+{
+	bzero(err, sizeof(*err));
+}
+
+void parse_events_error__exit(struct parse_events_error *err)
+{
+	zfree(&err->str);
+	zfree(&err->help);
+	zfree(&err->first_str);
+	zfree(&err->first_help);
+}
+
+void parse_events_error__handle(struct parse_events_error *err, int idx,
+				char *str, char *help)
+{
+	if (WARN(!str, "WARNING: failed to provide error string\n")) {
+		free(help);
+		return;
+	}
+	switch (err->num_errors) {
+	case 0:
+		err->idx = idx;
+		err->str = str;
+		err->help = help;
+		break;
+	case 1:
+		err->first_idx = err->idx;
+		err->idx = idx;
+		err->first_str = err->str;
+		err->str = str;
+		err->first_help = err->help;
+		err->help = help;
+		break;
+	default:
+		pr_debug("Multiple errors dropping message: %s (%s)\n",
+			err->str, err->help);
+		free(err->str);
+		err->str = str;
+		free(err->help);
+		err->help = help;
+		break;
+	}
+	err->num_errors++;
+}
+
 #define MAX_WIDTH 1000
 static int get_term_width(void)
 {
@@ -2308,8 +2359,8 @@ static int get_term_width(void)
 	return ws.ws_col > MAX_WIDTH ? MAX_WIDTH : ws.ws_col;
 }
 
-static void __parse_events_print_error(int err_idx, const char *err_str,
-				const char *err_help, const char *event)
+static void __parse_events_error__print(int err_idx, const char *err_str,
+					const char *err_help, const char *event)
 {
 	const char *str = "invalid or unsupported event: ";
 	char _buf[MAX_WIDTH];
@@ -2363,22 +2414,18 @@ static void __parse_events_print_error(int err_idx, const char *err_str,
 	}
 }
 
-void parse_events_print_error(struct parse_events_error *err,
-			      const char *event)
+void parse_events_error__print(struct parse_events_error *err,
+			       const char *event)
 {
 	if (!err->num_errors)
 		return;
 
-	__parse_events_print_error(err->idx, err->str, err->help, event);
-	zfree(&err->str);
-	zfree(&err->help);
+	__parse_events_error__print(err->idx, err->str, err->help, event);
 
 	if (err->num_errors > 1) {
 		fputs("\nInitial error:\n", stderr);
-		__parse_events_print_error(err->first_idx, err->first_str,
+		__parse_events_error__print(err->first_idx, err->first_str,
 					err->first_help, event);
-		zfree(&err->first_str);
-		zfree(&err->first_help);
 	}
 }
 
@@ -2391,13 +2438,14 @@ int parse_events_option(const struct option *opt, const char *str,
 	struct parse_events_error err;
 	int ret;
 
-	bzero(&err, sizeof(err));
+	parse_events_error__init(&err);
 	ret = parse_events(evlist, str, &err);
 
 	if (ret) {
-		parse_events_print_error(&err, str);
+		parse_events_error__print(&err, str);
 		fprintf(stderr, "Run 'perf list' for a list of valid events\n");
 	}
+	parse_events_error__exit(&err);
 
 	return ret;
 }
@@ -2703,7 +2751,7 @@ next:
 	return 0;
 }
 
-static bool is_event_supported(u8 type, unsigned config)
+static bool is_event_supported(u8 type, u64 config)
 {
 	bool ret = true;
 	int open_return;
@@ -2823,10 +2871,18 @@ void print_sdt_events(const char *subsys_glob, const char *event_glob,
 
 int print_hwcache_events(const char *event_glob, bool name_only)
 {
-	unsigned int type, op, i, evt_i = 0, evt_num = 0;
-	char name[64];
-	char **evt_list = NULL;
+	unsigned int type, op, i, evt_i = 0, evt_num = 0, npmus = 0;
+	char name[64], new_name[128];
+	char **evt_list = NULL, **evt_pmus = NULL;
 	bool evt_num_known = false;
+	struct perf_pmu *pmu = NULL;
+
+	if (perf_pmu__has_hybrid()) {
+		npmus = perf_pmu__hybrid_pmu_num();
+		evt_pmus = zalloc(sizeof(char *) * npmus);
+		if (!evt_pmus)
+			goto out_enomem;
+	}
 
 restart:
 	if (evt_num_known) {
@@ -2842,20 +2898,61 @@ restart:
 				continue;
 
 			for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
+				unsigned int hybrid_supported = 0, j;
+				bool supported;
+
 				__evsel__hw_cache_type_op_res_name(type, op, i, name, sizeof(name));
 				if (event_glob != NULL && !strglobmatch(name, event_glob))
 					continue;
 
-				if (!is_event_supported(PERF_TYPE_HW_CACHE,
-							type | (op << 8) | (i << 16)))
-					continue;
+				if (!perf_pmu__has_hybrid()) {
+					if (!is_event_supported(PERF_TYPE_HW_CACHE,
+								type | (op << 8) | (i << 16))) {
+						continue;
+					}
+				} else {
+					perf_pmu__for_each_hybrid_pmu(pmu) {
+						if (!evt_num_known) {
+							evt_num++;
+							continue;
+						}
+
+						supported = is_event_supported(
+									PERF_TYPE_HW_CACHE,
+									type | (op << 8) | (i << 16) |
+									((__u64)pmu->type << PERF_PMU_TYPE_SHIFT));
+						if (supported) {
+							snprintf(new_name, sizeof(new_name), "%s/%s/",
+								 pmu->name, name);
+							evt_pmus[hybrid_supported] = strdup(new_name);
+							hybrid_supported++;
+						}
+					}
+
+					if (hybrid_supported == 0)
+						continue;
+				}
 
 				if (!evt_num_known) {
 					evt_num++;
 					continue;
 				}
 
-				evt_list[evt_i] = strdup(name);
+				if ((hybrid_supported == 0) ||
+				    (hybrid_supported == npmus)) {
+					evt_list[evt_i] = strdup(name);
+					if (npmus > 0) {
+						for (j = 0; j < npmus; j++)
+							zfree(&evt_pmus[j]);
+					}
+				} else {
+					for (j = 0; j < hybrid_supported; j++) {
+						evt_list[evt_i++] = evt_pmus[j];
+						evt_pmus[j] = NULL;
+					}
+					continue;
+				}
+
 				if (evt_list[evt_i] == NULL)
 					goto out_enomem;
 				evt_i++;
@@ -2867,6 +2964,13 @@ restart:
 		evt_num_known = true;
 		goto restart;
 	}
+
+	for (evt_i = 0; evt_i < evt_num; evt_i++) {
+		if (!evt_list[evt_i])
+			break;
+	}
+
+	evt_num = evt_i;
 	qsort(evt_list, evt_num, sizeof(char *), cmp_string);
 	evt_i = 0;
 	while (evt_i < evt_num) {
@@ -2885,6 +2989,10 @@ out_free:
 	for (evt_i = 0; evt_i < evt_num; evt_i++)
 		zfree(&evt_list[evt_i]);
 	zfree(&evt_list);
+
+	for (evt_i = 0; evt_i < npmus; evt_i++)
+		zfree(&evt_pmus[evt_i]);
+	zfree(&evt_pmus);
 	return evt_num;
 
 out_enomem:
@@ -2994,7 +3102,8 @@ out_enomem:
  * Print the help text for the event symbols:
  */
 void print_events(const char *event_glob, bool name_only, bool quiet_flag,
-			bool long_desc, bool details_flag, bool deprecated)
+			bool long_desc, bool details_flag, bool deprecated,
+			const char *pmu_name)
 {
 	print_symbol_events(event_glob, PERF_TYPE_HARDWARE,
 			    event_symbols_hw, PERF_COUNT_HW_MAX, name_only);
@@ -3006,7 +3115,7 @@ void print_events(const char *event_glob, bool name_only, bool quiet_flag,
 	print_hwcache_events(event_glob, name_only);
 
 	print_pmu_events(event_glob, name_only, quiet_flag, long_desc,
-			details_flag, deprecated);
+			details_flag, deprecated, pmu_name);
 
 	if (event_glob != NULL)
 		return;
@@ -3032,7 +3141,8 @@ void print_events(const char *event_glob, bool name_only, bool quiet_flag,
 
 	print_sdt_events(NULL, NULL, name_only);
 
-	metricgroup__print(true, true, NULL, name_only, details_flag);
+	metricgroup__print(true, true, NULL, name_only, details_flag,
+			   pmu_name);
 
 	print_libpfm_events(name_only, long_desc);
 }
@@ -3083,7 +3193,7 @@ int parse_events_term__num(struct parse_events_term **term,
 	struct parse_events_term temp = {
 		.type_val  = PARSE_EVENTS__TERM_TYPE_NUM,
 		.type_term = type_term,
-		.config    = config,
+		.config    = config ? : strdup(config_term_names[type_term]),
 		.no_value  = no_value,
 		.err_term  = loc_term ? loc_term->first_column : 0,
 		.err_val   = loc_val  ? loc_val->first_column  : 0,
@@ -3227,7 +3337,7 @@ void parse_events_evlist_error(struct parse_events_state *parse_state,
 	if (!parse_state->error)
 		return;
 
-	parse_events__handle_error(parse_state->error, idx, strdup(str), NULL);
+	parse_events_error__handle(parse_state->error, idx, strdup(str), NULL);
 }
 
 static void config_terms_list(char *buf, size_t buf_sz)
@@ -3286,9 +3396,12 @@ fail:
 
 struct evsel *parse_events__add_event_hybrid(struct list_head *list, int *idx,
 					     struct perf_event_attr *attr,
-					     char *name, struct perf_pmu *pmu,
+					     const char *name,
+					     const char *metric_id,
+					     struct perf_pmu *pmu,
 					     struct list_head *config_terms)
 {
-	return __add_event(list, idx, attr, true, name, pmu,
-			   config_terms, false, NULL);
+	return __add_event(list, idx, attr, /*init_attr=*/true, name, metric_id,
+			   pmu, config_terms, /*auto_merge_stats=*/false,
+			   /*cpu_list=*/NULL);
 }
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index bf6e41aa9b6a..c7fc93f54577 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -87,6 +87,7 @@ enum {
 	PARSE_EVENTS__TERM_TYPE_PERCORE,
 	PARSE_EVENTS__TERM_TYPE_AUX_OUTPUT,
 	PARSE_EVENTS__TERM_TYPE_AUX_SAMPLE_SIZE,
+	PARSE_EVENTS__TERM_TYPE_METRIC_ID,
 	__PARSE_EVENTS__TERM_TYPE_NR,
 };
 
@@ -141,8 +142,6 @@ struct parse_events_state {
 	char			  *hybrid_pmu_name;
 };
 
-void parse_events__handle_error(struct parse_events_error *err, int idx,
-				char *str, char *help);
 void parse_events__shrink_config_terms(void);
 int parse_events__is_hardcoded_term(struct parse_events_term *term);
 int parse_events_term__num(struct parse_events_term **term,
@@ -162,7 +161,7 @@ void parse_events_terms__purge(struct list_head *terms);
 void parse_events__clear_array(struct parse_events_array *a);
 int parse_events__modifier_event(struct list_head *list, char *str, bool add);
 int parse_events__modifier_group(struct list_head *list, char *event_mod);
-int parse_events_name(struct list_head *list, char *name);
+int parse_events_name(struct list_head *list, const char *name);
 int parse_events_add_tracepoint(struct list_head *list, int *idx,
 				const char *sys, const char *event,
 				struct parse_events_error *error,
@@ -182,10 +181,9 @@ int parse_events_add_numeric(struct parse_events_state *parse_state,
 			     struct list_head *list,
 			     u32 type, u64 config,
 			     struct list_head *head_config);
-enum perf_tool_event;
 int parse_events_add_tool(struct parse_events_state *parse_state,
 			  struct list_head *list,
-			  enum perf_tool_event tool_event);
+			  int tool_event);
 int parse_events_add_cache(struct list_head *list, int *idx,
 			   char *type, char *op_result1, char *op_result2,
 			   struct parse_events_error *error,
@@ -200,10 +198,12 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
 			 bool use_alias);
 
 struct evsel *parse_events__add_event(int idx, struct perf_event_attr *attr,
-					char *name, struct perf_pmu *pmu);
+				      const char *name, const char *metric_id,
+				      struct perf_pmu *pmu);
 
 int parse_events_multi_pmu_add(struct parse_events_state *parse_state,
 			       char *str,
+			       struct list_head *head_config,
 			       struct list_head **listp);
 
 int parse_events_copy_term_list(struct list_head *old,
@@ -219,7 +219,8 @@ void parse_events_evlist_error(struct parse_events_state *parse_state,
 			       int idx, const char *str);
 
 void print_events(const char *event_glob, bool name_only, bool quiet,
-		  bool long_desc, bool details_flag, bool deprecated);
+		  bool long_desc, bool details_flag, bool deprecated,
+		  const char *pmu_name);
 
 struct event_symbol {
 	const char	*symbol;
@@ -241,8 +242,12 @@ int is_valid_tracepoint(const char *event_string);
 int valid_event_mount(const char *eventfs);
 char *parse_events_formats_error_string(char *additional_terms);
 
-void parse_events_print_error(struct parse_events_error *err,
-			      const char *event);
+void parse_events_error__init(struct parse_events_error *err);
+void parse_events_error__exit(struct parse_events_error *err);
+void parse_events_error__handle(struct parse_events_error *err, int idx,
+				char *str, char *help);
+void parse_events_error__print(struct parse_events_error *err,
+			       const char *event);
 
 #ifdef HAVE_LIBELF_SUPPORT
 /*
@@ -267,7 +272,9 @@ int perf_pmu__test_parse_init(void);
 
 struct evsel *parse_events__add_event_hybrid(struct list_head *list, int *idx,
 					     struct perf_event_attr *attr,
-					     char *name, struct perf_pmu *pmu,
+					     const char *name,
+					     const char *metric_id,
+					     struct perf_pmu *pmu,
 					     struct list_head *config_terms);
 
 #endif /* __PERF_PARSE_EVENTS_H */
diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l
index 923849024b15..4efe9872c667 100644
--- a/tools/perf/util/parse-events.l
+++ b/tools/perf/util/parse-events.l
@@ -12,7 +12,6 @@
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <unistd.h>
-#include "../perf.h"
 #include "parse-events.h"
 #include "parse-events-bison.h"
 #include "evsel.h"
@@ -139,18 +138,23 @@ static int pmu_str_check(yyscan_t scanner, struct parse_events_state *parse_stat
 
 	yylval->str = strdup(text);
 
-	if (parse_state->fake_pmu)
-		return PE_PMU_EVENT_FAKE;
-
+	/*
+	 * If we're not testing then parse check determines the PMU event type
+	 * which if it isn't a PMU returns PE_NAME. When testing the result of
+	 * parse check can't be trusted so we return PE_PMU_EVENT_FAKE unless
+	 * an '!' is present in which case the text can't be a PMU name.
+	 */
 	switch (perf_pmu__parse_check(text)) {
 		case PMU_EVENT_SYMBOL_PREFIX:
 			return PE_PMU_EVENT_PRE;
 		case PMU_EVENT_SYMBOL_SUFFIX:
 			return PE_PMU_EVENT_SUF;
 		case PMU_EVENT_SYMBOL:
-			return PE_KERNEL_PMU_EVENT;
+			return parse_state->fake_pmu
+				? PE_PMU_EVENT_FAKE : PE_KERNEL_PMU_EVENT;
 		default:
-			return PE_NAME;
+			return parse_state->fake_pmu && !strchr(text,'!')
+				? PE_PMU_EVENT_FAKE : PE_NAME;
 	}
 }
 
@@ -205,7 +209,7 @@ bpf_source	[^,{}]+\.c[a-zA-Z0-9._]*
 num_dec		[0-9]+
 num_hex		0x[a-fA-F0-9]+
 num_raw_hex	[a-fA-F0-9]+
-name		[a-zA-Z_*?\[\]][a-zA-Z0-9_*?.\[\]]*
+name		[a-zA-Z_*?\[\]][a-zA-Z0-9_*?.\[\]!]*
 name_tag	[\'][a-zA-Z_*?\[\]][a-zA-Z0-9_*?\-,\.\[\]:=]*[\']
 name_minus	[a-zA-Z_*?][a-zA-Z0-9\-_*?.:]*
 drv_cfg_term	[a-zA-Z0-9_\.]+(=[a-zA-Z0-9_*?\.:]+)?
@@ -295,6 +299,7 @@ no-overwrite		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_NOOVERWRITE); }
 percore			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_PERCORE); }
 aux-output		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_AUX_OUTPUT); }
 aux-sample-size		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_AUX_SAMPLE_SIZE); }
+metric-id		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_METRIC_ID); }
 r{num_raw_hex}		{ return raw(yyscanner); }
 r0x{num_raw_hex}	{ return raw(yyscanner); }
 ,			{ return ','; }
diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
index d94e48e1ff9b..174158982fae 100644
--- a/tools/perf/util/parse-events.y
+++ b/tools/perf/util/parse-events.y
@@ -183,6 +183,11 @@ group_def ':' PE_MODIFIER_EVENT
 	err = parse_events__modifier_group(list, $3);
 	free($3);
 	if (err) {
+		struct parse_events_state *parse_state = _parse_state;
+		struct parse_events_error *error = parse_state->error;
+
+		parse_events_error__handle(error, @3.first_column,
+					   strdup("Bad modifier"), NULL);
 		free_list_evsel(list);
 		YYABORT;
 	}
@@ -240,6 +245,11 @@ event_name PE_MODIFIER_EVENT
 	err = parse_events__modifier_event(list, $2, false);
 	free($2);
 	if (err) {
+		struct parse_events_state *parse_state = _parse_state;
+		struct parse_events_error *error = parse_state->error;
+
+		parse_events_error__handle(error, @2.first_column,
+					   strdup("Bad modifier"), NULL);
 		free_list_evsel(list);
 		YYABORT;
 	}
@@ -342,7 +352,20 @@ PE_KERNEL_PMU_EVENT sep_dc
 	struct list_head *list;
 	int err;
 
-	err = parse_events_multi_pmu_add(_parse_state, $1, &list);
+	err = parse_events_multi_pmu_add(_parse_state, $1, NULL, &list);
+	free($1);
+	if (err < 0)
+		YYABORT;
+	$$ = list;
+}
+|
+PE_KERNEL_PMU_EVENT opt_pmu_config
+{
+	struct list_head *list;
+	int err;
+
+	/* frees $2 */
+	err = parse_events_multi_pmu_add(_parse_state, $1, $2, &list);
 	free($1);
 	if (err < 0)
 		YYABORT;
@@ -357,7 +380,7 @@ PE_PMU_EVENT_PRE '-' PE_PMU_EVENT_SUF sep_dc
 	snprintf(pmu_name, sizeof(pmu_name), "%s-%s", $1, $3);
 	free($1);
 	free($3);
-	if (parse_events_multi_pmu_add(_parse_state, pmu_name, &list) < 0)
+	if (parse_events_multi_pmu_add(_parse_state, pmu_name, NULL, &list) < 0)
 		YYABORT;
 	$$ = list;
 }
diff --git a/tools/perf/util/pfm.c b/tools/perf/util/pfm.c
index 756295dedccc..f0bcfcab1a93 100644
--- a/tools/perf/util/pfm.c
+++ b/tools/perf/util/pfm.c
@@ -87,7 +87,8 @@ int parse_libpfm_events_option(const struct option *opt, const char *str,
 
 		pmu = perf_pmu__find_by_type((unsigned int)attr.type);
 		evsel = parse_events__add_event(evlist->core.nr_entries,
-						&attr, q, pmu);
+						&attr, q, /*metric_id=*/NULL,
+						pmu);
 		if (evsel == NULL)
 			goto error;
 
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index bdabd62170d2..6ae58406f4fc 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -315,7 +315,7 @@ static bool perf_pmu_merge_alias(struct perf_pmu_alias *newalias,
 }
 
 static int __perf_pmu__new_alias(struct list_head *list, char *dir, char *name,
-				 char *desc, char *val, struct pmu_event *pe)
+				 char *desc, char *val, const struct pmu_event *pe)
 {
 	struct parse_events_term *term;
 	struct perf_pmu_alias *alias;
@@ -710,9 +710,9 @@ static char *perf_pmu__getcpuid(struct perf_pmu *pmu)
 	return cpuid;
 }
 
-struct pmu_events_map *perf_pmu__find_map(struct perf_pmu *pmu)
+const struct pmu_events_map *perf_pmu__find_map(struct perf_pmu *pmu)
 {
-	struct pmu_events_map *map;
+	const struct pmu_events_map *map;
 	char *cpuid = perf_pmu__getcpuid(pmu);
 	int i;
 
@@ -737,7 +737,7 @@ struct pmu_events_map *perf_pmu__find_map(struct perf_pmu *pmu)
 	return map;
 }
 
-struct pmu_events_map *__weak pmu_events_map__find(void)
+const struct pmu_events_map *__weak pmu_events_map__find(void)
 {
 	return perf_pmu__find_map(NULL);
 }
@@ -824,7 +824,7 @@ out:
  * as aliases.
  */
 void pmu_add_cpu_aliases_map(struct list_head *head, struct perf_pmu *pmu,
-			     struct pmu_events_map *map)
+			     const struct pmu_events_map *map)
 {
 	int i;
 	const char *name = pmu->name;
@@ -834,7 +834,7 @@ void pmu_add_cpu_aliases_map(struct list_head *head, struct perf_pmu *pmu,
 	i = 0;
 	while (1) {
 		const char *cpu_name = is_arm_pmu_core(name) ? name : "cpu";
-		struct pmu_event *pe = &map->table[i++];
+		const struct pmu_event *pe = &map->table[i++];
 		const char *pname = pe->pmu ? pe->pmu : cpu_name;
 
 		if (!pe->name) {
@@ -859,7 +859,7 @@ new_alias:
 
 static void pmu_add_cpu_aliases(struct list_head *head, struct perf_pmu *pmu)
 {
-	struct pmu_events_map *map;
+	const struct pmu_events_map *map;
 
 	map = perf_pmu__find_map(pmu);
 	if (!map)
@@ -873,7 +873,7 @@ void pmu_for_each_sys_event(pmu_sys_event_iter_fn fn, void *data)
 	int i = 0;
 
 	while (1) {
-		struct pmu_sys_events *event_table;
+		const struct pmu_sys_events *event_table;
 		int j = 0;
 
 		event_table = &pmu_sys_event_tables[i++];
@@ -882,7 +882,7 @@ void pmu_for_each_sys_event(pmu_sys_event_iter_fn fn, void *data)
 			break;
 
 		while (1) {
-			struct pmu_event *pe = &event_table->table[j++];
+			const struct pmu_event *pe = &event_table->table[j++];
 			int ret;
 
 			if (!pe->name && !pe->metric_group && !pe->metric_name)
@@ -900,7 +900,7 @@ struct pmu_sys_event_iter_data {
 	struct perf_pmu *pmu;
 };
 
-static int pmu_add_sys_aliases_iter_fn(struct pmu_event *pe, void *data)
+static int pmu_add_sys_aliases_iter_fn(const struct pmu_event *pe, void *data)
 {
 	struct pmu_sys_event_iter_data *idata = data;
 	struct perf_pmu *pmu = idata->pmu;
@@ -1283,7 +1283,7 @@ static int pmu_config_term(const char *pmu_name,
 			unknown_term = NULL;
 		help_msg = parse_events_formats_error_string(pmu_term);
 		if (err) {
-			parse_events__handle_error(err, term->err_term,
+			parse_events_error__handle(err, term->err_term,
 						   unknown_term,
 						   help_msg);
 		} else {
@@ -1316,7 +1316,7 @@ static int pmu_config_term(const char *pmu_name,
 		if (term->no_value &&
 		    bitmap_weight(format->bits, PERF_PMU_FORMAT_BITS) > 1) {
 			if (err) {
-				parse_events__handle_error(err, term->err_val,
+				parse_events_error__handle(err, term->err_val,
 					   strdup("no value assigned for term"),
 					   NULL);
 			}
@@ -1331,7 +1331,7 @@ static int pmu_config_term(const char *pmu_name,
 						term->config, term->val.str);
 			}
 			if (err) {
-				parse_events__handle_error(err, term->err_val,
+				parse_events_error__handle(err, term->err_val,
 					strdup("expected numeric value"),
 					NULL);
 			}
@@ -1348,7 +1348,7 @@ static int pmu_config_term(const char *pmu_name,
 		if (err) {
 			char *err_str;
 
-			parse_events__handle_error(err, term->err_val,
+			parse_events_error__handle(err, term->err_val,
 				asprintf(&err_str,
 				    "value too big for format, maximum is %llu",
 				    (unsigned long long)max_val) < 0
@@ -1608,6 +1608,7 @@ static int cmp_sevent(const void *a, const void *b)
 {
 	const struct sevent *as = a;
 	const struct sevent *bs = b;
+	int ret;
 
 	/* Put extra events last */
 	if (!!as->desc != !!bs->desc)
@@ -1623,7 +1624,13 @@ static int cmp_sevent(const void *a, const void *b)
 	if (as->is_cpu != bs->is_cpu)
 		return bs->is_cpu - as->is_cpu;
 
-	return strcmp(as->name, bs->name);
+	ret = strcmp(as->name, bs->name);
+	if (!ret) {
+		if (as->pmu && bs->pmu)
+			return strcmp(as->pmu, bs->pmu);
+	}
+
+	return ret;
 }
 
 static void wordwrap(char *s, int start, int max, int corr)
@@ -1653,7 +1660,8 @@ bool is_pmu_core(const char *name)
 }
 
 void print_pmu_events(const char *event_glob, bool name_only, bool quiet_flag,
-			bool long_desc, bool details_flag, bool deprecated)
+			bool long_desc, bool details_flag, bool deprecated,
+			const char *pmu_name)
 {
 	struct perf_pmu *pmu;
 	struct perf_pmu_alias *alias;
@@ -1679,10 +1687,16 @@ void print_pmu_events(const char *event_glob, bool name_only, bool quiet_flag,
 	pmu = NULL;
 	j = 0;
 	while ((pmu = perf_pmu__scan(pmu)) != NULL) {
+		if (pmu_name && perf_pmu__is_hybrid(pmu->name) &&
+		    strcmp(pmu_name, pmu->name)) {
+			continue;
+		}
+
 		list_for_each_entry(alias, &pmu->aliases, list) {
 			char *name = alias->desc ? alias->name :
 				format_alias(buf, sizeof(buf), pmu, alias);
-			bool is_cpu = is_pmu_core(pmu->name);
+			bool is_cpu = is_pmu_core(pmu->name) ||
+				      perf_pmu__is_hybrid(pmu->name);
 
 			if (alias->deprecated && !deprecated)
 				continue;
@@ -1730,8 +1744,13 @@ void print_pmu_events(const char *event_glob, bool name_only, bool quiet_flag,
 	qsort(aliases, len, sizeof(struct sevent), cmp_sevent);
 	for (j = 0; j < len; j++) {
 		/* Skip duplicates */
-		if (j > 0 && !strcmp(aliases[j].name, aliases[j - 1].name))
-			continue;
+		if (j > 0 && !strcmp(aliases[j].name, aliases[j - 1].name)) {
+			if (!aliases[j].pmu || !aliases[j - 1].pmu ||
+			    !strcmp(aliases[j].pmu, aliases[j - 1].pmu)) {
+				continue;
+			}
+		}
+
 		if (name_only) {
 			printf("%s ", aliases[j].name);
 			continue;
@@ -1906,7 +1925,7 @@ int perf_pmu__caps_parse(struct perf_pmu *pmu)
 }
 
 void perf_pmu__warn_invalid_config(struct perf_pmu *pmu, __u64 config,
-				   char *name)
+				   const char *name)
 {
 	struct perf_pmu_format *format;
 	__u64 masks = 0, bits;
diff --git a/tools/perf/util/pmu.h b/tools/perf/util/pmu.h
index 394898b07fd9..541889fa9f9c 100644
--- a/tools/perf/util/pmu.h
+++ b/tools/perf/util/pmu.h
@@ -49,6 +49,10 @@ struct perf_pmu {
 	struct list_head caps;    /* HEAD struct perf_pmu_caps -> list */
 	struct list_head list;    /* ELEM */
 	struct list_head hybrid_list;
+
+	struct {
+		bool exclude_guest;
+	} missing_features;
 };
 
 extern struct perf_pmu perf_pmu__fake;
@@ -111,7 +115,7 @@ struct perf_pmu *perf_pmu__scan(struct perf_pmu *pmu);
 bool is_pmu_core(const char *name);
 void print_pmu_events(const char *event_glob, bool name_only, bool quiet,
 		      bool long_desc, bool details_flag,
-		      bool deprecated);
+		      bool deprecated, const char *pmu_name);
 bool pmu_have_event(const char *pname, const char *name);
 
 int perf_pmu__scan_file(struct perf_pmu *pmu, const char *name, const char *fmt, ...) __scanf(3, 4);
@@ -120,21 +124,21 @@ int perf_pmu__test(void);
 
 struct perf_event_attr *perf_pmu__get_default_config(struct perf_pmu *pmu);
 void pmu_add_cpu_aliases_map(struct list_head *head, struct perf_pmu *pmu,
-			     struct pmu_events_map *map);
+			     const struct pmu_events_map *map);
 
-struct pmu_events_map *perf_pmu__find_map(struct perf_pmu *pmu);
-struct pmu_events_map *pmu_events_map__find(void);
+const struct pmu_events_map *perf_pmu__find_map(struct perf_pmu *pmu);
+const struct pmu_events_map *pmu_events_map__find(void);
 bool pmu_uncore_alias_match(const char *pmu_name, const char *name);
 void perf_pmu_free_alias(struct perf_pmu_alias *alias);
 
-typedef int (*pmu_sys_event_iter_fn)(struct pmu_event *pe, void *data);
+typedef int (*pmu_sys_event_iter_fn)(const struct pmu_event *pe, void *data);
 void pmu_for_each_sys_event(pmu_sys_event_iter_fn fn, void *data);
 int perf_pmu__convert_scale(const char *scale, char **end, double *sval);
 
 int perf_pmu__caps_parse(struct perf_pmu *pmu);
 
 void perf_pmu__warn_invalid_config(struct perf_pmu *pmu, __u64 config,
-				   char *name);
+				   const char *name);
 
 bool perf_pmu__has_hybrid(void);
 int perf_pmu__match(char *pattern, char *name, char *tok);
diff --git a/tools/perf/util/python-ext-sources b/tools/perf/util/python-ext-sources
index d7c976671e3a..a685d20165f7 100644
--- a/tools/perf/util/python-ext-sources
+++ b/tools/perf/util/python-ext-sources
@@ -18,6 +18,7 @@ util/mmap.c
 util/namespaces.c
 ../lib/bitmap.c
 ../lib/find_bit.c
+../lib/list_sort.c
 ../lib/hweight.c
 ../lib/string.c
 ../lib/vsprintf.c
diff --git a/tools/perf/util/python.c b/tools/perf/util/python.c
index 8feef3a05af7..563a9ba8954f 100644
--- a/tools/perf/util/python.c
+++ b/tools/perf/util/python.c
@@ -70,6 +70,18 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 }
 
 /*
+ * This one is needed not to drag the PMU bandwagon, jevents generated
+ * pmu_sys_event_tables, etc and evsel__find_pmu() is used so far just for
+ * doing per PMU perf_event_attr.exclude_guest handling, not really needed, so
+ * far, for the perf python binding known usecases, revisit if this become
+ * necessary.
+ */
+struct perf_pmu *evsel__find_pmu(struct evsel *evsel __maybe_unused)
+{
+	return NULL;
+}
+
+/*
  * Add this one here not to drag util/metricgroup.c
  */
 int metricgroup__copy_metric_events(struct evlist *evlist, struct cgroup *cgrp,
diff --git a/tools/perf/util/record.h b/tools/perf/util/record.h
index 68f471d9a88b..ef6c2715fdd9 100644
--- a/tools/perf/util/record.h
+++ b/tools/perf/util/record.h
@@ -77,6 +77,7 @@ struct record_opts {
 	int	      ctl_fd;
 	int	      ctl_fd_ack;
 	bool	      ctl_fd_close;
+	int	      synth;
 };
 
 extern const char * const *record_usage;
diff --git a/tools/perf/util/s390-cpumsf.c b/tools/perf/util/s390-cpumsf.c
index 8130b56aa04b..f3fdad28a852 100644
--- a/tools/perf/util/s390-cpumsf.c
+++ b/tools/perf/util/s390-cpumsf.c
@@ -244,7 +244,7 @@ static bool s390_cpumsf_basic_show(const char *color, size_t pos,
 				   struct hws_basic_entry *basicp)
 {
 	struct hws_basic_entry *basic = basicp;
-#if __BYTE_ORDER == __LITTLE_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
 	struct hws_basic_entry local;
 	unsigned long long word = be64toh(*(unsigned long long *)basicp);
 
@@ -288,7 +288,7 @@ static bool s390_cpumsf_diag_show(const char *color, size_t pos,
 				  struct hws_diag_entry *diagp)
 {
 	struct hws_diag_entry *diag = diagp;
-#if __BYTE_ORDER == __LITTLE_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
 	struct hws_diag_entry local;
 	unsigned long long word = be64toh(*(unsigned long long *)diagp);
 
@@ -322,7 +322,7 @@ static unsigned long long trailer_timestamp(struct hws_trailer_entry *te,
 static bool s390_cpumsf_trailer_show(const char *color, size_t pos,
 				     struct hws_trailer_entry *te)
 {
-#if __BYTE_ORDER == __LITTLE_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
 	struct hws_trailer_entry local;
 	const unsigned long long flags = be64toh(te->flags);
 
@@ -552,7 +552,7 @@ static unsigned long long get_trailer_time(const unsigned char *buf)
 	te = (struct hws_trailer_entry *)(buf + S390_CPUMSF_PAGESZ
 					      - sizeof(*te));
 
-#if __BYTE_ORDER == __LITTLE_ENDIAN
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
 	clock_base = be64toh(te->progusage[0]) >> 63 & 0x1;
 	progusage2 = be64toh(te->progusage[1]);
 #else
diff --git a/tools/perf/util/s390-sample-raw.c b/tools/perf/util/s390-sample-raw.c
index 08ec3c3ae0ee..cd3a34840389 100644
--- a/tools/perf/util/s390-sample-raw.c
+++ b/tools/perf/util/s390-sample-raw.c
@@ -135,12 +135,12 @@ static int get_counterset_start(int setnr)
  * the name of this counter.
  * If no match is found a NULL pointer is returned.
  */
-static const char *get_counter_name(int set, int nr, struct pmu_events_map *map)
+static const char *get_counter_name(int set, int nr, const struct pmu_events_map *map)
 {
 	int rc, event_nr, wanted = get_counterset_start(set) + nr;
 
 	if (map) {
-		struct pmu_event *evp = map->table;
+		const struct pmu_event *evp = map->table;
 
 		for (; evp->name || evp->event || evp->desc; ++evp) {
 			if (evp->name == NULL || evp->event == NULL)
@@ -159,7 +159,7 @@ static void s390_cpumcfdg_dump(struct perf_sample *sample)
 	unsigned char *buf = sample->raw_data;
 	const char *color = PERF_COLOR_BLUE;
 	struct cf_ctrset_entry *cep, ce;
-	struct pmu_events_map *map;
+	const struct pmu_events_map *map;
 	u64 *p;
 
 	map = pmu_events_map__find();
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 352f16076e01..d8857d1b6d7c 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -44,7 +44,7 @@ static int perf_session__process_compressed_event(struct perf_session *session,
 	size_t decomp_size, src_size;
 	u64 decomp_last_rem = 0;
 	size_t mmap_len, decomp_len = session->header.env.comp_mmap_len;
-	struct decomp *decomp, *decomp_last = session->decomp_last;
+	struct decomp *decomp, *decomp_last = session->active_decomp->decomp_last;
 
 	if (decomp_last) {
 		decomp_last_rem = decomp_last->size - decomp_last->head;
@@ -71,7 +71,7 @@ static int perf_session__process_compressed_event(struct perf_session *session,
 	src = (void *)event + sizeof(struct perf_record_compressed);
 	src_size = event->pack.header.size - sizeof(struct perf_record_compressed);
 
-	decomp_size = zstd_decompress_stream(&(session->zstd_data), src, src_size,
+	decomp_size = zstd_decompress_stream(session->active_decomp->zstd_decomp, src, src_size,
 				&(decomp->data[decomp_last_rem]), decomp_len - decomp_last_rem);
 	if (!decomp_size) {
 		munmap(decomp, mmap_len);
@@ -81,13 +81,12 @@ static int perf_session__process_compressed_event(struct perf_session *session,
 
 	decomp->size += decomp_size;
 
-	if (session->decomp == NULL) {
-		session->decomp = decomp;
-		session->decomp_last = decomp;
-	} else {
-		session->decomp_last->next = decomp;
-		session->decomp_last = decomp;
-	}
+	if (session->active_decomp->decomp == NULL)
+		session->active_decomp->decomp = decomp;
+	else
+		session->active_decomp->decomp_last->next = decomp;
+
+	session->active_decomp->decomp_last = decomp;
 
 	pr_debug("decomp (B): %zd to %zd\n", src_size, decomp_size);
 
@@ -197,6 +196,8 @@ struct perf_session *__perf_session__new(struct perf_data *data,
 
 	session->repipe = repipe;
 	session->tool   = tool;
+	session->decomp_data.zstd_decomp = &session->zstd_data;
+	session->active_decomp = &session->decomp_data;
 	INIT_LIST_HEAD(&session->auxtrace_index);
 	machines__init(&session->machines);
 	ordered_events__init(&session->ordered_events,
@@ -276,11 +277,11 @@ static void perf_session__delete_threads(struct perf_session *session)
 	machine__delete_threads(&session->machines.host);
 }
 
-static void perf_session__release_decomp_events(struct perf_session *session)
+static void perf_decomp__release_events(struct decomp *next)
 {
-	struct decomp *next, *decomp;
+	struct decomp *decomp;
 	size_t mmap_len;
-	next = session->decomp;
+
 	do {
 		decomp = next;
 		if (decomp == NULL)
@@ -299,7 +300,7 @@ void perf_session__delete(struct perf_session *session)
 	auxtrace_index__free(&session->auxtrace_index);
 	perf_session__destroy_kernel_maps(session);
 	perf_session__delete_threads(session);
-	perf_session__release_decomp_events(session);
+	perf_decomp__release_events(session->decomp_data.decomp);
 	perf_env__exit(&session->header.env);
 	machines__exit(&session->machines);
 	if (session->data) {
@@ -509,6 +510,8 @@ void perf_tool__fill_defaults(struct perf_tool *tool)
 		tool->bpf = perf_event__process_bpf;
 	if (tool->text_poke == NULL)
 		tool->text_poke = perf_event__process_text_poke;
+	if (tool->aux_output_hw_id == NULL)
+		tool->aux_output_hw_id = perf_event__process_aux_output_hw_id;
 	if (tool->read == NULL)
 		tool->read = process_event_sample_stub;
 	if (tool->throttle == NULL)
@@ -1000,6 +1003,7 @@ static perf_event__swap_op perf_event__swap_ops[] = {
 	[PERF_RECORD_NAMESPACES]	  = perf_event__namespaces_swap,
 	[PERF_RECORD_CGROUP]		  = perf_event__cgroup_swap,
 	[PERF_RECORD_TEXT_POKE]		  = perf_event__text_poke_swap,
+	[PERF_RECORD_AUX_OUTPUT_HW_ID]	  = perf_event__all64_swap,
 	[PERF_RECORD_HEADER_ATTR]	  = perf_event__hdr_attr_swap,
 	[PERF_RECORD_HEADER_EVENT_TYPE]	  = perf_event__event_type_swap,
 	[PERF_RECORD_HEADER_TRACING_DATA] = perf_event__tracing_data_swap,
@@ -1556,6 +1560,8 @@ static int machines__deliver_event(struct machines *machines,
 		return tool->bpf(tool, event, sample, machine);
 	case PERF_RECORD_TEXT_POKE:
 		return tool->text_poke(tool, event, sample, machine);
+	case PERF_RECORD_AUX_OUTPUT_HW_ID:
+		return tool->aux_output_hw_id(tool, event, sample, machine);
 	default:
 		++evlist->stats.nr_unknown_events;
 		return -1;
@@ -2117,7 +2123,7 @@ static int __perf_session__process_decomp_events(struct perf_session *session)
 {
 	s64 skip;
 	u64 size;
-	struct decomp *decomp = session->decomp_last;
+	struct decomp *decomp = session->active_decomp->decomp_last;
 
 	if (!decomp)
 		return 0;
@@ -2171,35 +2177,55 @@ struct reader {
 	u64		 data_offset;
 	reader_cb_t	 process;
 	bool		 in_place_update;
+	char		 *mmaps[NUM_MMAPS];
+	size_t		 mmap_size;
+	int		 mmap_idx;
+	char		 *mmap_cur;
+	u64		 file_pos;
+	u64		 file_offset;
+	u64		 head;
+	struct zstd_data   zstd_data;
+	struct decomp_data decomp_data;
 };
 
 static int
-reader__process_events(struct reader *rd, struct perf_session *session,
-		       struct ui_progress *prog)
+reader__init(struct reader *rd, bool *one_mmap)
 {
 	u64 data_size = rd->data_size;
-	u64 head, page_offset, file_offset, file_pos, size;
-	int err = 0, mmap_prot, mmap_flags, map_idx = 0;
-	size_t	mmap_size;
-	char *buf, *mmaps[NUM_MMAPS];
-	union perf_event *event;
-	s64 skip;
-
-	page_offset = page_size * (rd->data_offset / page_size);
-	file_offset = page_offset;
-	head = rd->data_offset - page_offset;
-
-	ui_progress__init_size(prog, data_size, "Processing events...");
+	char **mmaps = rd->mmaps;
 
+	rd->head = rd->data_offset;
 	data_size += rd->data_offset;
 
-	mmap_size = MMAP_SIZE;
-	if (mmap_size > data_size) {
-		mmap_size = data_size;
-		session->one_mmap = true;
+	rd->mmap_size = MMAP_SIZE;
+	if (rd->mmap_size > data_size) {
+		rd->mmap_size = data_size;
+		if (one_mmap)
+			*one_mmap = true;
 	}
 
-	memset(mmaps, 0, sizeof(mmaps));
+	memset(mmaps, 0, sizeof(rd->mmaps));
+
+	if (zstd_init(&rd->zstd_data, 0))
+		return -1;
+	rd->decomp_data.zstd_decomp = &rd->zstd_data;
+
+	return 0;
+}
+
+static void
+reader__release_decomp(struct reader *rd)
+{
+	perf_decomp__release_events(rd->decomp_data.decomp);
+	zstd_fini(&rd->zstd_data);
+}
+
+static int
+reader__mmap(struct reader *rd, struct perf_session *session)
+{
+	int mmap_prot, mmap_flags;
+	char *buf, **mmaps = rd->mmaps;
+	u64 page_offset;
 
 	mmap_prot  = PROT_READ;
 	mmap_flags = MAP_SHARED;
@@ -2210,47 +2236,63 @@ reader__process_events(struct reader *rd, struct perf_session *session,
 		mmap_prot  |= PROT_WRITE;
 		mmap_flags = MAP_PRIVATE;
 	}
-remap:
-	buf = mmap(NULL, mmap_size, mmap_prot, mmap_flags, rd->fd,
-		   file_offset);
+
+	if (mmaps[rd->mmap_idx]) {
+		munmap(mmaps[rd->mmap_idx], rd->mmap_size);
+		mmaps[rd->mmap_idx] = NULL;
+	}
+
+	page_offset = page_size * (rd->head / page_size);
+	rd->file_offset += page_offset;
+	rd->head -= page_offset;
+
+	buf = mmap(NULL, rd->mmap_size, mmap_prot, mmap_flags, rd->fd,
+		   rd->file_offset);
 	if (buf == MAP_FAILED) {
 		pr_err("failed to mmap file\n");
-		err = -errno;
-		goto out;
+		return -errno;
 	}
-	mmaps[map_idx] = buf;
-	map_idx = (map_idx + 1) & (ARRAY_SIZE(mmaps) - 1);
-	file_pos = file_offset + head;
+	mmaps[rd->mmap_idx] = rd->mmap_cur = buf;
+	rd->mmap_idx = (rd->mmap_idx + 1) & (ARRAY_SIZE(rd->mmaps) - 1);
+	rd->file_pos = rd->file_offset + rd->head;
 	if (session->one_mmap) {
 		session->one_mmap_addr = buf;
-		session->one_mmap_offset = file_offset;
+		session->one_mmap_offset = rd->file_offset;
 	}
 
-more:
-	event = fetch_mmaped_event(head, mmap_size, buf, session->header.needs_swap);
+	return 0;
+}
+
+enum {
+	READER_OK,
+	READER_NODATA,
+};
+
+static int
+reader__read_event(struct reader *rd, struct perf_session *session,
+		   struct ui_progress *prog)
+{
+	u64 size;
+	int err = READER_OK;
+	union perf_event *event;
+	s64 skip;
+
+	event = fetch_mmaped_event(rd->head, rd->mmap_size, rd->mmap_cur,
+				   session->header.needs_swap);
 	if (IS_ERR(event))
 		return PTR_ERR(event);
 
-	if (!event) {
-		if (mmaps[map_idx]) {
-			munmap(mmaps[map_idx], mmap_size);
-			mmaps[map_idx] = NULL;
-		}
-
-		page_offset = page_size * (head / page_size);
-		file_offset += page_offset;
-		head -= page_offset;
-		goto remap;
-	}
+	if (!event)
+		return READER_NODATA;
 
 	size = event->header.size;
 
 	skip = -EINVAL;
 
 	if (size < sizeof(struct perf_event_header) ||
-	    (skip = rd->process(session, event, file_pos)) < 0) {
+	    (skip = rd->process(session, event, rd->file_pos)) < 0) {
 		pr_err("%#" PRIx64 " [%#x]: failed to process type: %d [%s]\n",
-		       file_offset + head, event->header.size,
+		       rd->file_offset + rd->head, event->header.size,
 		       event->header.type, strerror(-skip));
 		err = skip;
 		goto out;
@@ -2259,8 +2301,8 @@ more:
 	if (skip)
 		size += skip;
 
-	head += size;
-	file_pos += size;
+	rd->head += size;
+	rd->file_pos += size;
 
 	err = __perf_session__process_decomp_events(session);
 	if (err)
@@ -2268,13 +2310,48 @@ more:
 
 	ui_progress__update(prog, size);
 
+out:
+	return err;
+}
+
+static inline bool
+reader__eof(struct reader *rd)
+{
+	return (rd->file_pos >= rd->data_size + rd->data_offset);
+}
+
+static int
+reader__process_events(struct reader *rd, struct perf_session *session,
+		       struct ui_progress *prog)
+{
+	int err;
+
+	err = reader__init(rd, &session->one_mmap);
+	if (err)
+		goto out;
+
+	session->active_decomp = &rd->decomp_data;
+
+remap:
+	err = reader__mmap(rd, session);
+	if (err)
+		goto out;
+
+more:
+	err = reader__read_event(rd, session, prog);
+	if (err < 0)
+		goto out;
+	else if (err == READER_NODATA)
+		goto remap;
+
 	if (session_done())
 		goto out;
 
-	if (file_pos < data_size)
+	if (!reader__eof(rd))
 		goto more;
 
 out:
+	session->active_decomp = &session->decomp_data;
 	return err;
 }
 
@@ -2327,6 +2404,7 @@ out_err:
 	 */
 	ordered_events__reinit(&session->ordered_events);
 	auxtrace__free_events(session);
+	reader__release_decomp(&rd);
 	session->one_mmap = false;
 	return err;
 }
diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h
index 5d8bd14a0a39..46c854292ad6 100644
--- a/tools/perf/util/session.h
+++ b/tools/perf/util/session.h
@@ -20,6 +20,12 @@ struct thread;
 struct auxtrace;
 struct itrace_synth_opts;
 
+struct decomp_data {
+	struct decomp	 *decomp;
+	struct decomp	 *decomp_last;
+	struct zstd_data *zstd_decomp;
+};
+
 struct perf_session {
 	struct perf_header	header;
 	struct machines		machines;
@@ -39,8 +45,8 @@ struct perf_session {
 	u64			bytes_transferred;
 	u64			bytes_compressed;
 	struct zstd_data	zstd_data;
-	struct decomp		*decomp;
-	struct decomp		*decomp_last;
+	struct decomp_data	decomp_data;
+	struct decomp_data	*active_decomp;
 };
 
 struct decomp {
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 568a88c001c6..a111065b484e 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -1325,88 +1325,68 @@ struct sort_entry sort_mispredict = {
 	.se_width_idx	= HISTC_MISPREDICT,
 };
 
-static u64 he_weight(struct hist_entry *he)
-{
-	return he->stat.nr_events ? he->stat.weight / he->stat.nr_events : 0;
-}
-
 static int64_t
-sort__local_weight_cmp(struct hist_entry *left, struct hist_entry *right)
+sort__weight_cmp(struct hist_entry *left, struct hist_entry *right)
 {
-	return he_weight(left) - he_weight(right);
+	return left->weight - right->weight;
 }
 
 static int hist_entry__local_weight_snprintf(struct hist_entry *he, char *bf,
 				    size_t size, unsigned int width)
 {
-	return repsep_snprintf(bf, size, "%-*llu", width, he_weight(he));
+	return repsep_snprintf(bf, size, "%-*llu", width, he->weight);
 }
 
 struct sort_entry sort_local_weight = {
 	.se_header	= "Local Weight",
-	.se_cmp		= sort__local_weight_cmp,
+	.se_cmp		= sort__weight_cmp,
 	.se_snprintf	= hist_entry__local_weight_snprintf,
 	.se_width_idx	= HISTC_LOCAL_WEIGHT,
 };
 
-static int64_t
-sort__global_weight_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	return left->stat.weight - right->stat.weight;
-}
-
 static int hist_entry__global_weight_snprintf(struct hist_entry *he, char *bf,
 					      size_t size, unsigned int width)
 {
-	return repsep_snprintf(bf, size, "%-*llu", width, he->stat.weight);
+	return repsep_snprintf(bf, size, "%-*llu", width,
+			       he->weight * he->stat.nr_events);
 }
 
 struct sort_entry sort_global_weight = {
 	.se_header	= "Weight",
-	.se_cmp		= sort__global_weight_cmp,
+	.se_cmp		= sort__weight_cmp,
 	.se_snprintf	= hist_entry__global_weight_snprintf,
 	.se_width_idx	= HISTC_GLOBAL_WEIGHT,
 };
 
-static u64 he_ins_lat(struct hist_entry *he)
-{
-		return he->stat.nr_events ? he->stat.ins_lat / he->stat.nr_events : 0;
-}
-
 static int64_t
-sort__local_ins_lat_cmp(struct hist_entry *left, struct hist_entry *right)
+sort__ins_lat_cmp(struct hist_entry *left, struct hist_entry *right)
 {
-		return he_ins_lat(left) - he_ins_lat(right);
+	return left->ins_lat - right->ins_lat;
 }
 
 static int hist_entry__local_ins_lat_snprintf(struct hist_entry *he, char *bf,
 					      size_t size, unsigned int width)
 {
-		return repsep_snprintf(bf, size, "%-*u", width, he_ins_lat(he));
+	return repsep_snprintf(bf, size, "%-*u", width, he->ins_lat);
 }
 
 struct sort_entry sort_local_ins_lat = {
 	.se_header	= "Local INSTR Latency",
-	.se_cmp		= sort__local_ins_lat_cmp,
+	.se_cmp		= sort__ins_lat_cmp,
 	.se_snprintf	= hist_entry__local_ins_lat_snprintf,
 	.se_width_idx	= HISTC_LOCAL_INS_LAT,
 };
 
-static int64_t
-sort__global_ins_lat_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-		return left->stat.ins_lat - right->stat.ins_lat;
-}
-
 static int hist_entry__global_ins_lat_snprintf(struct hist_entry *he, char *bf,
 					       size_t size, unsigned int width)
 {
-		return repsep_snprintf(bf, size, "%-*u", width, he->stat.ins_lat);
+	return repsep_snprintf(bf, size, "%-*u", width,
+			       he->ins_lat * he->stat.nr_events);
 }
 
 struct sort_entry sort_global_ins_lat = {
 	.se_header	= "INSTR Latency",
-	.se_cmp		= sort__global_ins_lat_cmp,
+	.se_cmp		= sort__ins_lat_cmp,
 	.se_snprintf	= hist_entry__global_ins_lat_snprintf,
 	.se_width_idx	= HISTC_GLOBAL_INS_LAT,
 };
@@ -1414,13 +1394,13 @@ struct sort_entry sort_global_ins_lat = {
 static int64_t
 sort__global_p_stage_cyc_cmp(struct hist_entry *left, struct hist_entry *right)
 {
-	return left->stat.p_stage_cyc - right->stat.p_stage_cyc;
+	return left->p_stage_cyc - right->p_stage_cyc;
 }
 
 static int hist_entry__p_stage_cyc_snprintf(struct hist_entry *he, char *bf,
 					size_t size, unsigned int width)
 {
-	return repsep_snprintf(bf, size, "%-*u", width, he->stat.p_stage_cyc);
+	return repsep_snprintf(bf, size, "%-*u", width, he->p_stage_cyc);
 }
 
 struct sort_entry sort_p_stage_cyc = {
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
index b67c469aba79..7b7145501933 100644
--- a/tools/perf/util/sort.h
+++ b/tools/perf/util/sort.h
@@ -49,9 +49,6 @@ struct he_stat {
 	u64			period_us;
 	u64			period_guest_sys;
 	u64			period_guest_us;
-	u64			weight;
-	u64			ins_lat;
-	u64			p_stage_cyc;
 	u32			nr_events;
 };
 
@@ -109,6 +106,9 @@ struct hist_entry {
 	s32			socket;
 	s32			cpu;
 	u64			code_page_size;
+	u64			weight;
+	u64			ins_lat;
+	u64			p_stage_cyc;
 	u8			cpumode;
 	u8			depth;
 
diff --git a/tools/perf/util/srcline.c b/tools/perf/util/srcline.c
index 5b7d6c16d33f..af468e3bb6fa 100644
--- a/tools/perf/util/srcline.c
+++ b/tools/perf/util/srcline.c
@@ -1,8 +1,10 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <inttypes.h>
+#include <signal.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
+#include <sys/types.h>
 
 #include <linux/kernel.h>
 #include <linux/string.h>
@@ -15,6 +17,7 @@
 #include "srcline.h"
 #include "string2.h"
 #include "symbol.h"
+#include "subcmd/run-command.h"
 
 bool srcline_full_filename;
 
@@ -119,6 +122,8 @@ static struct symbol *new_inline_sym(struct dso *dso,
 	return inline_sym;
 }
 
+#define MAX_INLINE_NEST 1024
+
 #ifdef HAVE_LIBBFD_SUPPORT
 
 /*
@@ -273,8 +278,6 @@ static void addr2line_cleanup(struct a2l_data *a2l)
 	free(a2l);
 }
 
-#define MAX_INLINE_NEST 1024
-
 static int inline_list__append_dso_a2l(struct dso *dso,
 				       struct inline_node *node,
 				       struct symbol *sym)
@@ -361,26 +364,14 @@ void dso__free_a2l(struct dso *dso)
 	dso->a2l = NULL;
 }
 
-static struct inline_node *addr2inlines(const char *dso_name, u64 addr,
-					struct dso *dso, struct symbol *sym)
-{
-	struct inline_node *node;
-
-	node = zalloc(sizeof(*node));
-	if (node == NULL) {
-		perror("not enough memory for the inline node");
-		return NULL;
-	}
-
-	INIT_LIST_HEAD(&node->val);
-	node->addr = addr;
-
-	addr2line(dso_name, addr, NULL, NULL, dso, true, node, sym);
-	return node;
-}
-
 #else /* HAVE_LIBBFD_SUPPORT */
 
+struct a2l_subprocess {
+	struct child_process addr2line;
+	FILE *to_child;
+	FILE *from_child;
+};
+
 static int filename_split(char *filename, unsigned int *line_nr)
 {
 	char *sep;
@@ -402,114 +393,285 @@ static int filename_split(char *filename, unsigned int *line_nr)
 	return 0;
 }
 
-static int addr2line(const char *dso_name, u64 addr,
-		     char **file, unsigned int *line_nr,
-		     struct dso *dso __maybe_unused,
-		     bool unwind_inlines __maybe_unused,
-		     struct inline_node *node __maybe_unused,
-		     struct symbol *sym __maybe_unused)
+static void addr2line_subprocess_cleanup(struct a2l_subprocess *a2l)
 {
-	FILE *fp;
-	char cmd[PATH_MAX];
-	char *filename = NULL;
-	size_t len;
-	int ret = 0;
+	if (a2l->addr2line.pid != -1) {
+		kill(a2l->addr2line.pid, SIGKILL);
+		finish_command(&a2l->addr2line); /* ignore result, we don't care */
+		a2l->addr2line.pid = -1;
+	}
 
-	scnprintf(cmd, sizeof(cmd), "addr2line -e %s %016"PRIx64,
-		  dso_name, addr);
+	if (a2l->to_child != NULL) {
+		fclose(a2l->to_child);
+		a2l->to_child = NULL;
+	}
 
-	fp = popen(cmd, "r");
-	if (fp == NULL) {
-		pr_warning("popen failed for %s\n", dso_name);
-		return 0;
+	if (a2l->from_child != NULL) {
+		fclose(a2l->from_child);
+		a2l->from_child = NULL;
+	}
+
+	free(a2l);
+}
+
+static struct a2l_subprocess *addr2line_subprocess_init(const char *path)
+{
+	const char *argv[] = { "addr2line", "-e", path, "-i", "-f", NULL };
+	struct a2l_subprocess *a2l = zalloc(sizeof(*a2l));
+	int start_command_status = 0;
+
+	if (a2l == NULL)
+		goto out;
+
+	a2l->to_child = NULL;
+	a2l->from_child = NULL;
+
+	a2l->addr2line.pid = -1;
+	a2l->addr2line.in = -1;
+	a2l->addr2line.out = -1;
+	a2l->addr2line.no_stderr = 1;
+
+	a2l->addr2line.argv = argv;
+	start_command_status = start_command(&a2l->addr2line);
+	a2l->addr2line.argv = NULL; /* it's not used after start_command; avoid dangling pointers */
+
+	if (start_command_status != 0) {
+		pr_warning("could not start addr2line for %s: start_command return code %d\n",
+			   path,
+			   start_command_status);
+		goto out;
 	}
 
-	if (getline(&filename, &len, fp) < 0 || !len) {
-		pr_warning("addr2line has no output for %s\n", dso_name);
+	a2l->to_child = fdopen(a2l->addr2line.in, "w");
+	if (a2l->to_child == NULL) {
+		pr_warning("could not open write-stream to addr2line of %s\n", path);
 		goto out;
 	}
 
-	ret = filename_split(filename, line_nr);
-	if (ret != 1) {
-		free(filename);
+	a2l->from_child = fdopen(a2l->addr2line.out, "r");
+	if (a2l->from_child == NULL) {
+		pr_warning("could not open read-stream from addr2line of %s\n", path);
 		goto out;
 	}
 
-	*file = filename;
+	return a2l;
 
 out:
-	pclose(fp);
-	return ret;
+	if (a2l)
+		addr2line_subprocess_cleanup(a2l);
+
+	return NULL;
 }
 
-void dso__free_a2l(struct dso *dso __maybe_unused)
+static int read_addr2line_record(struct a2l_subprocess *a2l,
+				 char **function,
+				 char **filename,
+				 unsigned int *line_nr)
 {
+	/*
+	 * Returns:
+	 * -1 ==> error
+	 * 0 ==> sentinel (or other ill-formed) record read
+	 * 1 ==> a genuine record read
+	 */
+	char *line = NULL;
+	size_t line_len = 0;
+	unsigned int dummy_line_nr = 0;
+	int ret = -1;
+
+	if (function != NULL)
+		zfree(function);
+
+	if (filename != NULL)
+		zfree(filename);
+
+	if (line_nr != NULL)
+		*line_nr = 0;
+
+	if (getline(&line, &line_len, a2l->from_child) < 0 || !line_len)
+		goto error;
+
+	if (function != NULL)
+		*function = strdup(strim(line));
+
+	zfree(&line);
+	line_len = 0;
+
+	if (getline(&line, &line_len, a2l->from_child) < 0 || !line_len)
+		goto error;
+
+	if (filename_split(line, line_nr == NULL ? &dummy_line_nr : line_nr) == 0) {
+		ret = 0;
+		goto error;
+	}
+
+	if (filename != NULL)
+		*filename = strdup(line);
+
+	zfree(&line);
+	line_len = 0;
+
+	return 1;
+
+error:
+	free(line);
+	if (function != NULL)
+		zfree(function);
+	if (filename != NULL)
+		zfree(filename);
+	return ret;
 }
 
-static struct inline_node *addr2inlines(const char *dso_name, u64 addr,
-					struct dso *dso __maybe_unused,
-					struct symbol *sym)
+static int inline_list__append_record(struct dso *dso,
+				      struct inline_node *node,
+				      struct symbol *sym,
+				      const char *function,
+				      const char *filename,
+				      unsigned int line_nr)
 {
-	FILE *fp;
-	char cmd[PATH_MAX];
-	struct inline_node *node;
-	char *filename = NULL;
-	char *funcname = NULL;
-	size_t filelen, funclen;
-	unsigned int line_nr = 0;
+	struct symbol *inline_sym = new_inline_sym(dso, sym, function);
 
-	scnprintf(cmd, sizeof(cmd), "addr2line -e %s -i -f %016"PRIx64,
-		  dso_name, addr);
+	return inline_list__append(inline_sym, srcline_from_fileline(filename, line_nr), node);
+}
 
-	fp = popen(cmd, "r");
-	if (fp == NULL) {
-		pr_err("popen failed for %s\n", dso_name);
-		return NULL;
+static int addr2line(const char *dso_name, u64 addr,
+		     char **file, unsigned int *line_nr,
+		     struct dso *dso,
+		     bool unwind_inlines,
+		     struct inline_node *node,
+		     struct symbol *sym __maybe_unused)
+{
+	struct a2l_subprocess *a2l = dso->a2l;
+	char *record_function = NULL;
+	char *record_filename = NULL;
+	unsigned int record_line_nr = 0;
+	int record_status = -1;
+	int ret = 0;
+	size_t inline_count = 0;
+
+	if (!a2l) {
+		dso->a2l = addr2line_subprocess_init(dso_name);
+		a2l = dso->a2l;
 	}
 
-	node = zalloc(sizeof(*node));
-	if (node == NULL) {
-		perror("not enough memory for the inline node");
+	if (a2l == NULL) {
+		if (!symbol_conf.disable_add2line_warn)
+			pr_warning("%s %s: addr2line_subprocess_init failed\n", __func__, dso_name);
 		goto out;
 	}
 
-	INIT_LIST_HEAD(&node->val);
-	node->addr = addr;
-
-	/* addr2line -f generates two lines for each inlined functions */
-	while (getline(&funcname, &funclen, fp) != -1) {
-		char *srcline;
-		struct symbol *inline_sym;
+	/*
+	 * Send our request and then *deliberately* send something that can't be interpreted as
+	 * a valid address to ask addr2line about (namely, ","). This causes addr2line to first
+	 * write out the answer to our request, in an unbounded/unknown number of records, and
+	 * then to write out the lines "??" and "??:0", so that we can detect when it has
+	 * finished giving us anything useful. We have to be careful about the first record,
+	 * though, because it may be genuinely unknown, in which case we'll get two sets of
+	 * "??"/"??:0" lines.
+	 */
+	if (fprintf(a2l->to_child, "%016"PRIx64"\n,\n", addr) < 0 || fflush(a2l->to_child) != 0) {
+		pr_warning("%s %s: could not send request\n", __func__, dso_name);
+		goto out;
+	}
 
-		strim(funcname);
+	switch (read_addr2line_record(a2l, &record_function, &record_filename, &record_line_nr)) {
+	case -1:
+		pr_warning("%s %s: could not read first record\n", __func__, dso_name);
+		goto out;
+	case 0:
+		/*
+		 * The first record was invalid, so return failure, but first read another
+		 * record, since we asked a junk question and have to clear the answer out.
+		 */
+		switch (read_addr2line_record(a2l, NULL, NULL, NULL)) {
+		case -1:
+			pr_warning("%s %s: could not read delimiter record\n", __func__, dso_name);
+			break;
+		case 0:
+			/* As expected. */
+			break;
+		default:
+			pr_warning("%s %s: unexpected record instead of sentinel",
+				   __func__, dso_name);
+			break;
+		}
+		goto out;
+	default:
+		break;
+	}
 
-		if (getline(&filename, &filelen, fp) == -1)
-			goto out;
+	if (file) {
+		*file = strdup(record_filename);
+		ret = 1;
+	}
+	if (line_nr)
+		*line_nr = record_line_nr;
 
-		if (filename_split(filename, &line_nr) != 1)
+	if (unwind_inlines) {
+		if (node && inline_list__append_record(dso, node, sym,
+						       record_function,
+						       record_filename,
+						       record_line_nr)) {
+			ret = 0;
 			goto out;
+		}
+	}
 
-		srcline = srcline_from_fileline(filename, line_nr);
-		inline_sym = new_inline_sym(dso, sym, funcname);
-
-		if (inline_list__append(inline_sym, srcline, node) != 0) {
-			free(srcline);
-			if (inline_sym && inline_sym->inlined)
-				symbol__delete(inline_sym);
-			goto out;
+	/* We have to read the records even if we don't care about the inline info. */
+	while ((record_status = read_addr2line_record(a2l,
+						      &record_function,
+						      &record_filename,
+						      &record_line_nr)) == 1) {
+		if (unwind_inlines && node && inline_count++ < MAX_INLINE_NEST) {
+			if (inline_list__append_record(dso, node, sym,
+						       record_function,
+						       record_filename,
+						       record_line_nr)) {
+				ret = 0;
+				goto out;
+			}
+			ret = 1; /* found at least one inline frame */
 		}
 	}
 
 out:
-	pclose(fp);
-	free(filename);
-	free(funcname);
+	free(record_function);
+	free(record_filename);
+	return ret;
+}
 
-	return node;
+void dso__free_a2l(struct dso *dso)
+{
+	struct a2l_subprocess *a2l = dso->a2l;
+
+	if (!a2l)
+		return;
+
+	addr2line_subprocess_cleanup(a2l);
+
+	dso->a2l = NULL;
 }
 
 #endif /* HAVE_LIBBFD_SUPPORT */
 
+static struct inline_node *addr2inlines(const char *dso_name, u64 addr,
+					struct dso *dso, struct symbol *sym)
+{
+	struct inline_node *node;
+
+	node = zalloc(sizeof(*node));
+	if (node == NULL) {
+		perror("not enough memory for the inline node");
+		return NULL;
+	}
+
+	INIT_LIST_HEAD(&node->val);
+	node->addr = addr;
+
+	addr2line(dso_name, addr, NULL, NULL, dso, true, node, sym);
+	return node;
+}
+
 /*
  * Number of addr2line failures (without success) before disabling it for that
  * dso.
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
index 34a7f5c1fff7..5c7308efa768 100644
--- a/tools/perf/util/stat-shadow.c
+++ b/tools/perf/util/stat-shadow.c
@@ -1,8 +1,10 @@
 // SPDX-License-Identifier: GPL-2.0
+#include <math.h>
 #include <stdio.h>
 #include "evsel.h"
 #include "stat.h"
 #include "color.h"
+#include "debug.h"
 #include "pmu.h"
 #include "rblist.h"
 #include "evlist.h"
@@ -370,12 +372,16 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 {
 	struct evsel *counter, *leader, **metric_events, *oc;
 	bool found;
-	struct expr_parse_ctx ctx;
+	struct expr_parse_ctx *ctx;
 	struct hashmap_entry *cur;
 	size_t bkt;
 	int i;
 
-	expr__ctx_init(&ctx);
+	ctx = expr__ctx_new();
+	if (!ctx) {
+		pr_debug("expr__ctx_new failed");
+		return;
+	}
 	evlist__for_each_entry(evsel_list, counter) {
 		bool invalid = false;
 
@@ -383,25 +389,25 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 		if (!counter->metric_expr)
 			continue;
 
-		expr__ctx_clear(&ctx);
+		expr__ctx_clear(ctx);
 		metric_events = counter->metric_events;
 		if (!metric_events) {
-			if (expr__find_other(counter->metric_expr,
-					     counter->name,
-					     &ctx, 1) < 0)
+			if (expr__find_ids(counter->metric_expr,
+					   counter->name,
+					   ctx) < 0)
 				continue;
 
 			metric_events = calloc(sizeof(struct evsel *),
-					       hashmap__size(&ctx.ids) + 1);
+					       hashmap__size(ctx->ids) + 1);
 			if (!metric_events) {
-				expr__ctx_clear(&ctx);
+				expr__ctx_free(ctx);
 				return;
 			}
 			counter->metric_events = metric_events;
 		}
 
 		i = 0;
-		hashmap__for_each_entry((&ctx.ids), cur, bkt) {
+		hashmap__for_each_entry(ctx->ids, cur, bkt) {
 			const char *metric_name = (const char *)cur->key;
 
 			found = false;
@@ -438,6 +444,7 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 						"Add %s event to groups to get metric expression for %s\n",
 						metric_name,
 						counter->name);
+					free(printed);
 					printed = strdup(metric_name);
 				}
 				invalid = true;
@@ -453,7 +460,7 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 			counter->metric_expr = NULL;
 		}
 	}
-	expr__ctx_clear(&ctx);
+	expr__ctx_free(ctx);
 }
 
 static double runtime_stat_avg(struct runtime_stat *st,
@@ -815,18 +822,19 @@ static int prepare_metric(struct evsel **metric_events,
 			  struct runtime_stat *st)
 {
 	double scale;
-	char *n, *pn;
+	char *n;
 	int i, j, ret;
 
-	expr__ctx_init(pctx);
 	for (i = 0; metric_events[i]; i++) {
 		struct saved_value *v;
 		struct stats *stats;
 		u64 metric_total = 0;
+		int source_count;
 
 		if (!strcmp(metric_events[i]->name, "duration_time")) {
 			stats = &walltime_nsecs_stats;
 			scale = 1e-9;
+			source_count = 1;
 		} else {
 			v = saved_value_lookup(metric_events[i], cpu, false,
 					       STAT_NONE, 0, st,
@@ -835,27 +843,18 @@ static int prepare_metric(struct evsel **metric_events,
 				break;
 			stats = &v->stats;
 			scale = 1.0;
+			source_count = evsel__source_count(metric_events[i]);
 
 			if (v->metric_other)
 				metric_total = v->metric_total;
 		}
-
-		n = strdup(metric_events[i]->name);
+		n = strdup(evsel__metric_id(metric_events[i]));
 		if (!n)
 			return -ENOMEM;
-		/*
-		 * This display code with --no-merge adds [cpu] postfixes.
-		 * These are not supported by the parser. Remove everything
-		 * after the space.
-		 */
-		pn = strchr(n, ' ');
-		if (pn)
-			*pn = 0;
-
-		if (metric_total)
-			expr__add_id_val(pctx, n, metric_total);
-		else
-			expr__add_id_val(pctx, n, avg_stats(stats)*scale);
+
+		expr__add_id_val_source_count(pctx, n,
+					metric_total ? : avg_stats(stats) * scale,
+					source_count);
 	}
 
 	for (j = 0; metric_refs && metric_refs[j].metric_name; j++) {
@@ -880,17 +879,23 @@ static void generic_metric(struct perf_stat_config *config,
 			   struct runtime_stat *st)
 {
 	print_metric_t print_metric = out->print_metric;
-	struct expr_parse_ctx pctx;
+	struct expr_parse_ctx *pctx;
 	double ratio, scale;
 	int i;
 	void *ctxp = out->ctx;
 
-	i = prepare_metric(metric_events, metric_refs, &pctx, cpu, st);
-	if (i < 0)
+	pctx = expr__ctx_new();
+	if (!pctx)
 		return;
 
+	pctx->runtime = runtime;
+	i = prepare_metric(metric_events, metric_refs, pctx, cpu, st);
+	if (i < 0) {
+		expr__ctx_free(pctx);
+		return;
+	}
 	if (!metric_events[i]) {
-		if (expr__parse(&ratio, &pctx, metric_expr, runtime) == 0) {
+		if (expr__parse(&ratio, pctx, metric_expr) == 0) {
 			char *unit;
 			char metric_bf[64];
 
@@ -926,22 +931,26 @@ static void generic_metric(struct perf_stat_config *config,
 			     (metric_name ? metric_name : name) : "", 0);
 	}
 
-	expr__ctx_clear(&pctx);
+	expr__ctx_free(pctx);
 }
 
 double test_generic_metric(struct metric_expr *mexp, int cpu, struct runtime_stat *st)
 {
-	struct expr_parse_ctx pctx;
+	struct expr_parse_ctx *pctx;
 	double ratio = 0.0;
 
-	if (prepare_metric(mexp->metric_events, mexp->metric_refs, &pctx, cpu, st) < 0)
+	pctx = expr__ctx_new();
+	if (!pctx)
+		return NAN;
+
+	if (prepare_metric(mexp->metric_events, mexp->metric_refs, pctx, cpu, st) < 0)
 		goto out;
 
-	if (expr__parse(&ratio, &pctx, mexp->metric_expr, 1))
+	if (expr__parse(&ratio, pctx, mexp->metric_expr))
 		ratio = 0.0;
 
 out:
-	expr__ctx_clear(&pctx);
+	expr__ctx_free(pctx);
 	return ratio;
 }
 
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 0fc9a5410739..b2ed3140a1fa 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -274,7 +274,7 @@ struct symbol *symbol__new(u64 start, u64 len, u8 binding, u8 type, const char *
 	if (symbol_conf.priv_size) {
 		if (symbol_conf.init_annotation) {
 			struct annotation *notes = (void *)sym;
-			pthread_mutex_init(&notes->lock, NULL);
+			annotation__init(notes);
 		}
 		sym = ((void *)sym) + symbol_conf.priv_size;
 	}
@@ -294,6 +294,13 @@ struct symbol *symbol__new(u64 start, u64 len, u8 binding, u8 type, const char *
 
 void symbol__delete(struct symbol *sym)
 {
+	if (symbol_conf.priv_size) {
+		if (symbol_conf.init_annotation) {
+			struct annotation *notes = symbol__annotation(sym);
+
+			annotation__exit(notes);
+		}
+	}
 	free(((void *)sym) - symbol_conf.priv_size);
 }
 
@@ -702,6 +709,10 @@ static int map__process_kallsym_symbol(void *arg, const char *name,
 	if (!symbol_type__filter(type))
 		return 0;
 
+	/* Ignore local symbols for ARM modules */
+	if (name[0] == '$')
+		return 0;
+
 	/*
 	 * module symbols are not sorted so we add all
 	 * symbols, setting length to 0, and rely on
@@ -2630,3 +2641,25 @@ struct mem_info *mem_info__new(void)
 		refcount_set(&mi->refcnt, 1);
 	return mi;
 }
+
+/*
+ * Checks that user supplied symbol kernel files are accessible because
+ * the default mechanism for accessing elf files fails silently. i.e. if
+ * debug syms for a build ID aren't found perf carries on normally. When
+ * they are user supplied we should assume that the user doesn't want to
+ * silently fail.
+ */
+int symbol__validate_sym_arguments(void)
+{
+	if (symbol_conf.vmlinux_name &&
+	    access(symbol_conf.vmlinux_name, R_OK)) {
+		pr_err("Invalid file: %s\n", symbol_conf.vmlinux_name);
+		return -EINVAL;
+	}
+	if (symbol_conf.kallsyms_name &&
+	    access(symbol_conf.kallsyms_name, R_OK)) {
+		pr_err("Invalid file: %s\n", symbol_conf.kallsyms_name);
+		return -EINVAL;
+	}
+	return 0;
+}
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 954d6a049ee2..fbf866d82dcc 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -40,22 +40,33 @@ Elf_Scn *elf_section_by_name(Elf *elf, GElf_Ehdr *ep,
 			     GElf_Shdr *shp, const char *name, size_t *idx);
 #endif
 
-/** struct symbol - symtab entry
- *
- * @ignore - resolvable but tools ignore it (e.g. idle routines)
+/**
+ * A symtab entry. When allocated this may be preceded by an annotation (see
+ * symbol__annotation), a browser_index (see symbol__browser_index) and rb_node
+ * to sort by name (see struct symbol_name_rb_node).
  */
 struct symbol {
 	struct rb_node	rb_node;
+	/** Range of symbol [start, end). */
 	u64		start;
 	u64		end;
+	/** Length of the string name. */
 	u16		namelen;
+	/** ELF symbol type as defined for st_info. E.g STT_OBJECT or STT_FUNC. */
 	u8		type:4;
+	/** ELF binding type as defined for st_info. E.g. STB_WEAK or STB_GLOBAL. */
 	u8		binding:4;
+	/** Set true for kernel symbols of idle routines. */
 	u8		idle:1;
+	/** Resolvable but tools ignore it (e.g. idle routines). */
 	u8		ignore:1;
+	/** Symbol for an inlined function. */
 	u8		inlined:1;
+	/** Has symbol__annotate2 been performed. */
+	u8		annotate2:1;
+	/** Architecture specific. Unused except on PPC where it holds st_other. */
 	u8		arch_sym;
-	bool		annotate2;
+	/** The name of length namelen associated with the symbol. */
 	char		name[];
 };
 
@@ -286,4 +297,6 @@ static inline void __mem_info__zput(struct mem_info **mi)
 
 #define mem_info__zput(mi) __mem_info__zput(&mi)
 
+int symbol__validate_sym_arguments(void);
+
 #endif /* __PERF_SYMBOL */
diff --git a/tools/perf/util/synthetic-events.c b/tools/perf/util/synthetic-events.c
index a7e981b2d7de..198982109f0f 100644
--- a/tools/perf/util/synthetic-events.c
+++ b/tools/perf/util/synthetic-events.c
@@ -715,7 +715,8 @@ static int __event__synthesize_thread(union perf_event *comm_event,
 				      union perf_event *fork_event,
 				      union perf_event *namespaces_event,
 				      pid_t pid, int full, perf_event__handler_t process,
-				      struct perf_tool *tool, struct machine *machine, bool mmap_data)
+				      struct perf_tool *tool, struct machine *machine,
+				      bool needs_mmap, bool mmap_data)
 {
 	char filename[PATH_MAX];
 	struct dirent **dirent;
@@ -739,7 +740,7 @@ static int __event__synthesize_thread(union perf_event *comm_event,
 		 * send mmap only for thread group leader
 		 * see thread__init_maps()
 		 */
-		if (pid == tgid &&
+		if (pid == tgid && needs_mmap &&
 		    perf_event__synthesize_mmap_events(tool, mmap_event, pid, tgid,
 						       process, machine, mmap_data))
 			return -1;
@@ -786,7 +787,7 @@ static int __event__synthesize_thread(union perf_event *comm_event,
 			break;
 
 		rc = 0;
-		if (_pid == pid && !kernel_thread) {
+		if (_pid == pid && !kernel_thread && needs_mmap) {
 			/* process the parent's maps too */
 			rc = perf_event__synthesize_mmap_events(tool, mmap_event, pid, tgid,
 						process, machine, mmap_data);
@@ -806,7 +807,7 @@ int perf_event__synthesize_thread_map(struct perf_tool *tool,
 				      struct perf_thread_map *threads,
 				      perf_event__handler_t process,
 				      struct machine *machine,
-				      bool mmap_data)
+				      bool needs_mmap, bool mmap_data)
 {
 	union perf_event *comm_event, *mmap_event, *fork_event;
 	union perf_event *namespaces_event;
@@ -836,7 +837,7 @@ int perf_event__synthesize_thread_map(struct perf_tool *tool,
 					       fork_event, namespaces_event,
 					       perf_thread_map__pid(threads, thread), 0,
 					       process, tool, machine,
-					       mmap_data)) {
+					       needs_mmap, mmap_data)) {
 			err = -1;
 			break;
 		}
@@ -862,7 +863,7 @@ int perf_event__synthesize_thread_map(struct perf_tool *tool,
 						       fork_event, namespaces_event,
 						       comm_event->comm.pid, 0,
 						       process, tool, machine,
-						       mmap_data)) {
+						       needs_mmap, mmap_data)) {
 				err = -1;
 				break;
 			}
@@ -882,6 +883,7 @@ out:
 static int __perf_event__synthesize_threads(struct perf_tool *tool,
 					    perf_event__handler_t process,
 					    struct machine *machine,
+					    bool needs_mmap,
 					    bool mmap_data,
 					    struct dirent **dirent,
 					    int start,
@@ -926,7 +928,7 @@ static int __perf_event__synthesize_threads(struct perf_tool *tool,
 		 */
 		__event__synthesize_thread(comm_event, mmap_event, fork_event,
 					   namespaces_event, pid, 1, process,
-					   tool, machine, mmap_data);
+					   tool, machine, needs_mmap, mmap_data);
 	}
 	err = 0;
 
@@ -945,6 +947,7 @@ struct synthesize_threads_arg {
 	struct perf_tool *tool;
 	perf_event__handler_t process;
 	struct machine *machine;
+	bool needs_mmap;
 	bool mmap_data;
 	struct dirent **dirent;
 	int num;
@@ -956,7 +959,8 @@ static void *synthesize_threads_worker(void *arg)
 	struct synthesize_threads_arg *args = arg;
 
 	__perf_event__synthesize_threads(args->tool, args->process,
-					 args->machine, args->mmap_data,
+					 args->machine,
+					 args->needs_mmap, args->mmap_data,
 					 args->dirent,
 					 args->start, args->num);
 	return NULL;
@@ -965,7 +969,7 @@ static void *synthesize_threads_worker(void *arg)
 int perf_event__synthesize_threads(struct perf_tool *tool,
 				   perf_event__handler_t process,
 				   struct machine *machine,
-				   bool mmap_data,
+				   bool needs_mmap, bool mmap_data,
 				   unsigned int nr_threads_synthesize)
 {
 	struct synthesize_threads_arg *args = NULL;
@@ -994,7 +998,8 @@ int perf_event__synthesize_threads(struct perf_tool *tool,
 
 	if (thread_nr <= 1) {
 		err = __perf_event__synthesize_threads(tool, process,
-						       machine, mmap_data,
+						       machine,
+						       needs_mmap, mmap_data,
 						       dirent, base, n);
 		goto free_dirent;
 	}
@@ -1015,6 +1020,7 @@ int perf_event__synthesize_threads(struct perf_tool *tool,
 		args[i].tool = tool;
 		args[i].process = process;
 		args[i].machine = machine;
+		args[i].needs_mmap = needs_mmap;
 		args[i].mmap_data = mmap_data;
 		args[i].dirent = dirent;
 	}
@@ -1775,26 +1781,27 @@ out_err:
 
 int __machine__synthesize_threads(struct machine *machine, struct perf_tool *tool,
 				  struct target *target, struct perf_thread_map *threads,
-				  perf_event__handler_t process, bool data_mmap,
-				  unsigned int nr_threads_synthesize)
+				  perf_event__handler_t process, bool needs_mmap,
+				  bool data_mmap, unsigned int nr_threads_synthesize)
 {
 	if (target__has_task(target))
-		return perf_event__synthesize_thread_map(tool, threads, process, machine, data_mmap);
+		return perf_event__synthesize_thread_map(tool, threads, process, machine,
+							 needs_mmap, data_mmap);
 	else if (target__has_cpu(target))
-		return perf_event__synthesize_threads(tool, process,
-						      machine, data_mmap,
+		return perf_event__synthesize_threads(tool, process, machine,
+						      needs_mmap, data_mmap,
 						      nr_threads_synthesize);
 	/* command specified */
 	return 0;
 }
 
 int machine__synthesize_threads(struct machine *machine, struct target *target,
-				struct perf_thread_map *threads, bool data_mmap,
-				unsigned int nr_threads_synthesize)
+				struct perf_thread_map *threads, bool needs_mmap,
+				bool data_mmap, unsigned int nr_threads_synthesize)
 {
 	return __machine__synthesize_threads(machine, NULL, target, threads,
-					     perf_event__process, data_mmap,
-					     nr_threads_synthesize);
+					     perf_event__process, needs_mmap,
+					     data_mmap, nr_threads_synthesize);
 }
 
 static struct perf_record_event_update *event_update_event__new(size_t size, u64 type, u64 id)
@@ -2230,3 +2237,31 @@ int perf_event__synthesize_for_pipe(struct perf_tool *tool,
 
 	return ret;
 }
+
+int parse_synth_opt(char *synth)
+{
+	char *p, *q;
+	int ret = 0;
+
+	if (synth == NULL)
+		return -1;
+
+	for (q = synth; (p = strsep(&q, ",")); p = q) {
+		if (!strcasecmp(p, "no") || !strcasecmp(p, "none"))
+			return 0;
+
+		if (!strcasecmp(p, "all"))
+			return PERF_SYNTH_ALL;
+
+		if (!strcasecmp(p, "task"))
+			ret |= PERF_SYNTH_TASK;
+		else if (!strcasecmp(p, "mmap"))
+			ret |= PERF_SYNTH_TASK | PERF_SYNTH_MMAP;
+		else if (!strcasecmp(p, "cgroup"))
+			ret |= PERF_SYNTH_CGROUP;
+		else
+			return -1;
+	}
+
+	return ret;
+}
diff --git a/tools/perf/util/synthetic-events.h b/tools/perf/util/synthetic-events.h
index c845e2b9b444..c931433bacbf 100644
--- a/tools/perf/util/synthetic-events.h
+++ b/tools/perf/util/synthetic-events.h
@@ -27,6 +27,18 @@ struct target;
 
 union perf_event;
 
+enum perf_record_synth {
+	PERF_SYNTH_TASK		= 1 << 0,
+	PERF_SYNTH_MMAP		= 1 << 1,
+	PERF_SYNTH_CGROUP	= 1 << 2,
+
+	/* last element */
+	PERF_SYNTH_MAX		= 1 << 3,
+};
+#define PERF_SYNTH_ALL  (PERF_SYNTH_MAX - 1)
+
+int parse_synth_opt(char *str);
+
 typedef int (*perf_event__handler_t)(struct perf_tool *tool, union perf_event *event,
 				     struct perf_sample *sample, struct machine *machine);
 
@@ -53,8 +65,8 @@ int perf_event__synthesize_stat_events(struct perf_stat_config *config, struct p
 int perf_event__synthesize_stat_round(struct perf_tool *tool, u64 time, u64 type, perf_event__handler_t process, struct machine *machine);
 int perf_event__synthesize_stat(struct perf_tool *tool, u32 cpu, u32 thread, u64 id, struct perf_counts_values *count, perf_event__handler_t process, struct machine *machine);
 int perf_event__synthesize_thread_map2(struct perf_tool *tool, struct perf_thread_map *threads, perf_event__handler_t process, struct machine *machine);
-int perf_event__synthesize_thread_map(struct perf_tool *tool, struct perf_thread_map *threads, perf_event__handler_t process, struct machine *machine, bool mmap_data);
-int perf_event__synthesize_threads(struct perf_tool *tool, perf_event__handler_t process, struct machine *machine, bool mmap_data, unsigned int nr_threads_synthesize);
+int perf_event__synthesize_thread_map(struct perf_tool *tool, struct perf_thread_map *threads, perf_event__handler_t process, struct machine *machine, bool needs_mmap, bool mmap_data);
+int perf_event__synthesize_threads(struct perf_tool *tool, perf_event__handler_t process, struct machine *machine, bool needs_mmap, bool mmap_data, unsigned int nr_threads_synthesize);
 int perf_event__synthesize_tracing_data(struct perf_tool *tool, int fd, struct evlist *evlist, perf_event__handler_t process);
 int perf_event__synth_time_conv(const struct perf_event_mmap_page *pc, struct perf_tool *tool, perf_event__handler_t process, struct machine *machine);
 pid_t perf_event__synthesize_comm(struct perf_tool *tool, union perf_event *event, pid_t pid, perf_event__handler_t process, struct machine *machine);
@@ -65,10 +77,10 @@ size_t perf_event__sample_event_size(const struct perf_sample *sample, u64 type,
 
 int __machine__synthesize_threads(struct machine *machine, struct perf_tool *tool,
 				  struct target *target, struct perf_thread_map *threads,
-				  perf_event__handler_t process, bool data_mmap,
+				  perf_event__handler_t process, bool needs_mmap, bool data_mmap,
 				  unsigned int nr_threads_synthesize);
 int machine__synthesize_threads(struct machine *machine, struct target *target,
-				struct perf_thread_map *threads, bool data_mmap,
+				struct perf_thread_map *threads, bool needs_mmap, bool data_mmap,
 				unsigned int nr_threads_synthesize);
 
 #ifdef HAVE_AUXTRACE_SUPPORT
diff --git a/tools/perf/util/tool.h b/tools/perf/util/tool.h
index bbbc0dcd461f..ef873f2cc38f 100644
--- a/tools/perf/util/tool.h
+++ b/tools/perf/util/tool.h
@@ -53,6 +53,7 @@ struct perf_tool {
 			lost_samples,
 			aux,
 			itrace_start,
+			aux_output_hw_id,
 			context_switch,
 			throttle,
 			unthrottle,
diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c
index 37a9492edb3e..df3c4671be72 100644
--- a/tools/perf/util/util.c
+++ b/tools/perf/util/util.c
@@ -379,32 +379,32 @@ fetch_kernel_version(unsigned int *puint, char *str,
 	return 0;
 }
 
-const char *perf_tip(const char *dirpath)
+int perf_tip(char **strp, const char *dirpath)
 {
 	struct strlist *tips;
 	struct str_node *node;
-	char *tip = NULL;
 	struct strlist_config conf = {
 		.dirname = dirpath,
 		.file_only = true,
 	};
+	int ret = 0;
 
+	*strp = NULL;
 	tips = strlist__new("tips.txt", &conf);
 	if (tips == NULL)
-		return errno == ENOENT ? NULL :
-			"Tip: check path of tips.txt or get more memory! ;-p";
+		return -errno;
 
 	if (strlist__nr_entries(tips) == 0)
 		goto out;
 
 	node = strlist__entry(tips, random() % strlist__nr_entries(tips));
-	if (asprintf(&tip, "Tip: %s", node->s) < 0)
-		tip = (char *)"Tip: get more memory! ;-)";
+	if (asprintf(strp, "Tip: %s", node->s) < 0)
+		ret = -ENOMEM;
 
 out:
 	strlist__delete(tips);
 
-	return tip;
+	return ret;
 }
 
 char *perf_exe(char *buf, int len)
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index ad737052e597..9f0d36ba77f2 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -39,7 +39,7 @@ int fetch_kernel_version(unsigned int *puint,
 #define KVER_FMT	"%d.%d.%d"
 #define KVER_PARAM(x)	KVER_VERSION(x), KVER_PATCHLEVEL(x), KVER_SUBLEVEL(x)
 
-const char *perf_tip(const char *dirpath);
+int perf_tip(char **strp, const char *dirpath);
 
 #ifndef HAVE_SCHED_GETCPU_SUPPORT
 int sched_getcpu(void);