From 114339ee4d66a328d186264ffa23a766542a9a15 Mon Sep 17 00:00:00 2001
From: Collin Funk <collin.funk1@gmail.com>
Date: Fri, 27 Jun 2025 20:41:25 -0700
Subject: perf build: Specify shellcheck should use bash

When someone has a global shellcheckrc file, for example at
~/.config/shellcheckrc, with the directive 'shell=sh', building perf
will fail with many shellcheck errors like:

    In tests/shell/base_probe/test_adding_kernel.sh line 294:
    (( TEST_RESULT += $? ))
    ^---------------------^ SC3006 (warning): In POSIX sh, standalone ((..)) is undefined.

    For more information:
      https://www.shellcheck.net/wiki/SC3006 -- In POSIX sh, standalone ((..)) is...
    make[5]: *** [tests/Build:91: tests/shell/base_probe/test_adding_kernel.sh.shellcheck_log] Error 1

Passing the '-s bash' option ensures that it runs correctly regardless
of a developers global configuration.

This patch adds '-s bash' and other options to the SHELLCHECK variable
in Makefile.perf and makes use of the variable consistently.

Signed-off-by: Collin Funk <collin.funk1@gmail.com>
Link: https://lore.kernel.org/r/63491dbc8439edf2e949d80e264b9d22332fea61.1751082075.git.collin.funk1@gmail.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/arch/x86/Build       | 2 +-
 tools/perf/arch/x86/tests/Build | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'tools/perf/arch')

diff --git a/tools/perf/arch/x86/Build b/tools/perf/arch/x86/Build
index afae7b8f6bd6..d31a1168757c 100644
--- a/tools/perf/arch/x86/Build
+++ b/tools/perf/arch/x86/Build
@@ -10,6 +10,6 @@ endif
 
 $(OUTPUT)%.shellcheck_log: %
 	$(call rule_mkdir)
-	$(Q)$(call echo-cmd,test)shellcheck -a -S warning "$<" > $@ || (cat $@ && rm $@ && false)
+	$(Q)$(call echo-cmd,test)$(SHELLCHECK) "$<" > $@ || (cat $@ && rm $@ && false)
 
 perf-test-y += $(SHELL_TEST_LOGS)
diff --git a/tools/perf/arch/x86/tests/Build b/tools/perf/arch/x86/tests/Build
index 5e00cbfd2d56..01d5527f38c7 100644
--- a/tools/perf/arch/x86/tests/Build
+++ b/tools/perf/arch/x86/tests/Build
@@ -22,6 +22,6 @@ endif
 
 $(OUTPUT)%.shellcheck_log: %
 	$(call rule_mkdir)
-	$(Q)$(call echo-cmd,test)shellcheck -a -S warning "$<" > $@ || (cat $@ && rm $@ && false)
+	$(Q)$(call echo-cmd,test)$(SHELLCHECK) "$<" > $@ || (cat $@ && rm $@ && false)
 
 perf-test-y += $(SHELL_TEST_LOGS)
-- 
cgit v1.2.3


From 5b546de9cc177936a3ed07d7d46ef072db4fdbab Mon Sep 17 00:00:00 2001
From: Ian Rogers <irogers@google.com>
Date: Fri, 18 Jul 2025 20:05:15 -0700
Subject: perf topdown: Use attribute to see an event is a topdown metic or
 slots

The string comparisons were overly broad and could fire for the
incorrect PMU and events. Switch to using the config in the attribute
then add a perf test to confirm the attribute config values match
those of parsed events of that name and don't match others. This
exposed matches for slots events that shouldn't have matched as the
slots fixed counter event, such as topdown.slots_p.

Fixes: fbc798316bef ("perf x86/topdown: Refine helper arch_is_topdown_metrics()")
Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250719030517.1990983-14-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/arch/x86/include/arch-tests.h |  4 ++
 tools/perf/arch/x86/tests/Build          |  1 +
 tools/perf/arch/x86/tests/arch-tests.c   |  1 +
 tools/perf/arch/x86/tests/topdown.c      | 76 ++++++++++++++++++++++++++++++++
 tools/perf/arch/x86/util/evsel.c         | 46 +++++--------------
 tools/perf/arch/x86/util/topdown.c       | 31 +++++--------
 tools/perf/arch/x86/util/topdown.h       |  4 ++
 7 files changed, 108 insertions(+), 55 deletions(-)
 create mode 100644 tools/perf/arch/x86/tests/topdown.c

(limited to 'tools/perf/arch')

diff --git a/tools/perf/arch/x86/include/arch-tests.h b/tools/perf/arch/x86/include/arch-tests.h
index 4fd425157d7d..8713e9122d4c 100644
--- a/tools/perf/arch/x86/include/arch-tests.h
+++ b/tools/perf/arch/x86/include/arch-tests.h
@@ -2,6 +2,8 @@
 #ifndef ARCH_TESTS_H
 #define ARCH_TESTS_H
 
+#include "tests/tests.h"
+
 struct test_suite;
 
 /* Tests */
@@ -17,6 +19,8 @@ int test__amd_ibs_via_core_pmu(struct test_suite *test, int subtest);
 int test__amd_ibs_period(struct test_suite *test, int subtest);
 int test__hybrid(struct test_suite *test, int subtest);
 
+DECLARE_SUITE(x86_topdown);
+
 extern struct test_suite *arch_tests[];
 
 #endif
diff --git a/tools/perf/arch/x86/tests/Build b/tools/perf/arch/x86/tests/Build
index 01d5527f38c7..311b6b53d3d8 100644
--- a/tools/perf/arch/x86/tests/Build
+++ b/tools/perf/arch/x86/tests/Build
@@ -11,6 +11,7 @@ endif
 perf-test-$(CONFIG_X86_64) += bp-modify.o
 perf-test-y += amd-ibs-via-core-pmu.o
 perf-test-y += amd-ibs-period.o
+perf-test-y += topdown.o
 
 ifdef SHELLCHECK
   SHELL_TESTS := gen-insn-x86-dat.sh
diff --git a/tools/perf/arch/x86/tests/arch-tests.c b/tools/perf/arch/x86/tests/arch-tests.c
index bfee2432515b..29ec1861ccef 100644
--- a/tools/perf/arch/x86/tests/arch-tests.c
+++ b/tools/perf/arch/x86/tests/arch-tests.c
@@ -53,5 +53,6 @@ struct test_suite *arch_tests[] = {
 	&suite__amd_ibs_via_core_pmu,
 	&suite__amd_ibs_period,
 	&suite__hybrid,
+	&suite__x86_topdown,
 	NULL,
 };
diff --git a/tools/perf/arch/x86/tests/topdown.c b/tools/perf/arch/x86/tests/topdown.c
new file mode 100644
index 000000000000..8d0ea7a4bbc1
--- /dev/null
+++ b/tools/perf/arch/x86/tests/topdown.c
@@ -0,0 +1,76 @@
+// SPDX-License-Identifier: GPL-2.0
+#include "arch-tests.h"
+#include "../util/topdown.h"
+#include "evlist.h"
+#include "parse-events.h"
+#include "pmu.h"
+#include "pmus.h"
+
+static int event_cb(void *state, struct pmu_event_info *info)
+{
+	char buf[256];
+	struct parse_events_error parse_err;
+	int *ret = state, err;
+	struct evlist *evlist = evlist__new();
+	struct evsel *evsel;
+
+	if (!evlist)
+		return -ENOMEM;
+
+	parse_events_error__init(&parse_err);
+	snprintf(buf, sizeof(buf), "%s/%s/", info->pmu->name, info->name);
+	err = parse_events(evlist, buf, &parse_err);
+	if (err) {
+		parse_events_error__print(&parse_err, buf);
+		*ret = TEST_FAIL;
+	}
+	parse_events_error__exit(&parse_err);
+	evlist__for_each_entry(evlist, evsel) {
+		bool fail = false;
+		bool p_core_pmu = evsel->pmu->type == PERF_TYPE_RAW;
+		const char *name = evsel__name(evsel);
+
+		if (strcasestr(name, "uops_retired.slots") ||
+		    strcasestr(name, "topdown.backend_bound_slots") ||
+		    strcasestr(name, "topdown.br_mispredict_slots") ||
+		    strcasestr(name, "topdown.memory_bound_slots") ||
+		    strcasestr(name, "topdown.bad_spec_slots") ||
+		    strcasestr(name, "topdown.slots_p")) {
+			if (arch_is_topdown_slots(evsel) || arch_is_topdown_metrics(evsel))
+				fail = true;
+		} else if (strcasestr(name, "slots")) {
+			if (arch_is_topdown_slots(evsel) != p_core_pmu ||
+			    arch_is_topdown_metrics(evsel))
+				fail = true;
+		} else if (strcasestr(name, "topdown")) {
+			if (arch_is_topdown_slots(evsel) ||
+			    arch_is_topdown_metrics(evsel) != p_core_pmu)
+				fail = true;
+		} else if (arch_is_topdown_slots(evsel) || arch_is_topdown_metrics(evsel)) {
+			fail = true;
+		}
+		if (fail) {
+			pr_debug("Broken topdown information for '%s'\n", evsel__name(evsel));
+			*ret = TEST_FAIL;
+		}
+	}
+	evlist__delete(evlist);
+	return 0;
+}
+
+static int test__x86_topdown(struct test_suite *test __maybe_unused, int subtest __maybe_unused)
+{
+	int ret = TEST_OK;
+	struct perf_pmu *pmu = NULL;
+
+	if (!topdown_sys_has_perf_metrics())
+		return TEST_OK;
+
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
+		if (perf_pmu__for_each_event(pmu, /*skip_duplicate_pmus=*/false, &ret, event_cb))
+			break;
+	}
+	return ret;
+}
+
+DEFINE_SUITE("x86 topdown", x86_topdown);
diff --git a/tools/perf/arch/x86/util/evsel.c b/tools/perf/arch/x86/util/evsel.c
index 3dd29ba2c23b..9bc80fff3aa0 100644
--- a/tools/perf/arch/x86/util/evsel.c
+++ b/tools/perf/arch/x86/util/evsel.c
@@ -23,47 +23,25 @@ void arch_evsel__set_sample_weight(struct evsel *evsel)
 bool evsel__sys_has_perf_metrics(const struct evsel *evsel)
 {
 	struct perf_pmu *pmu;
-	u32 type = evsel->core.attr.type;
 
-	/*
-	 * The PERF_TYPE_RAW type is the core PMU type, e.g., "cpu" PMU
-	 * on a non-hybrid machine, "cpu_core" PMU on a hybrid machine.
-	 * The slots event is only available for the core PMU, which
-	 * supports the perf metrics feature.
-	 * Checking both the PERF_TYPE_RAW type and the slots event
-	 * should be good enough to detect the perf metrics feature.
-	 */
-again:
-	switch (type) {
-	case PERF_TYPE_HARDWARE:
-	case PERF_TYPE_HW_CACHE:
-		type = evsel->core.attr.config >> PERF_PMU_TYPE_SHIFT;
-		if (type)
-			goto again;
-		break;
-	case PERF_TYPE_RAW:
-		break;
-	default:
+	if (!topdown_sys_has_perf_metrics())
 		return false;
-	}
-
-	pmu = evsel->pmu;
-	if (pmu && perf_pmu__is_fake(pmu))
-		pmu = NULL;
 
-	if (!pmu) {
-		while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
-			if (pmu->type == PERF_TYPE_RAW)
-				break;
-		}
-	}
-	return pmu && perf_pmu__have_event(pmu, "slots");
+	/*
+	 * The PERF_TYPE_RAW type is the core PMU type, e.g., "cpu" PMU on a
+	 * non-hybrid machine, "cpu_core" PMU on a hybrid machine.  The
+	 * topdown_sys_has_perf_metrics checks the slots event is only available
+	 * for the core PMU, which supports the perf metrics feature. Checking
+	 * both the PERF_TYPE_RAW type and the slots event should be good enough
+	 * to detect the perf metrics feature.
+	 */
+	pmu = evsel__find_pmu(evsel);
+	return pmu && pmu->type == PERF_TYPE_RAW;
 }
 
 bool arch_evsel__must_be_in_group(const struct evsel *evsel)
 {
-	if (!evsel__sys_has_perf_metrics(evsel) || !evsel->name ||
-	    strcasestr(evsel->name, "uops_retired.slots"))
+	if (!evsel__sys_has_perf_metrics(evsel))
 		return false;
 
 	return arch_is_topdown_metrics(evsel) || arch_is_topdown_slots(evsel);
diff --git a/tools/perf/arch/x86/util/topdown.c b/tools/perf/arch/x86/util/topdown.c
index d1c654839049..66b231fbf52e 100644
--- a/tools/perf/arch/x86/util/topdown.c
+++ b/tools/perf/arch/x86/util/topdown.c
@@ -1,6 +1,4 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "api/fs/fs.h"
-#include "util/evsel.h"
 #include "util/evlist.h"
 #include "util/pmu.h"
 #include "util/pmus.h"
@@ -8,6 +6,9 @@
 #include "topdown.h"
 #include "evsel.h"
 
+// cmask=0, inv=0, pc=0, edge=0, umask=4, event=0
+#define TOPDOWN_SLOTS		0x0400
+
 /* Check whether there is a PMU which supports the perf metrics. */
 bool topdown_sys_has_perf_metrics(void)
 {
@@ -32,31 +33,19 @@ bool topdown_sys_has_perf_metrics(void)
 	return has_perf_metrics;
 }
 
-#define TOPDOWN_SLOTS		0x0400
 bool arch_is_topdown_slots(const struct evsel *evsel)
 {
-	if (evsel->core.attr.config == TOPDOWN_SLOTS)
-		return true;
-
-	return false;
+	return evsel->core.attr.type == PERF_TYPE_RAW &&
+	       evsel->core.attr.config == TOPDOWN_SLOTS &&
+	       evsel->core.attr.config1 == 0;
 }
 
 bool arch_is_topdown_metrics(const struct evsel *evsel)
 {
-	int config = evsel->core.attr.config;
-	const char *name_from_config;
-	struct perf_pmu *pmu;
-
-	/* All topdown events have an event code of 0. */
-	if ((config & 0xFF) != 0)
-		return false;
-
-	pmu = evsel__find_pmu(evsel);
-	if (!pmu || !pmu->is_core)
-		return false;
-
-	name_from_config = perf_pmu__name_from_config(pmu, config);
-	return name_from_config && strcasestr(name_from_config, "topdown");
+	// cmask=0, inv=0, pc=0, edge=0, umask=0x80-0x87, event=0
+	return evsel->core.attr.type == PERF_TYPE_RAW &&
+		(evsel->core.attr.config & 0xFFFFF8FF) == 0x8000 &&
+		evsel->core.attr.config1 == 0;
 }
 
 /*
diff --git a/tools/perf/arch/x86/util/topdown.h b/tools/perf/arch/x86/util/topdown.h
index 1bae9b1822d7..2349536cf882 100644
--- a/tools/perf/arch/x86/util/topdown.h
+++ b/tools/perf/arch/x86/util/topdown.h
@@ -2,6 +2,10 @@
 #ifndef _TOPDOWN_H
 #define _TOPDOWN_H 1
 
+#include <stdbool.h>
+
+struct evsel;
+
 bool topdown_sys_has_perf_metrics(void);
 bool arch_is_topdown_slots(const struct evsel *evsel);
 bool arch_is_topdown_metrics(const struct evsel *evsel);
-- 
cgit v1.2.3


From 8dcd27b1b8661f64e220bc26a499865261d5d0f1 Mon Sep 17 00:00:00 2001
From: Ian Rogers <irogers@google.com>
Date: Fri, 18 Jul 2025 20:05:16 -0700
Subject: perf parse-events: Fix missing slots for Intel topdown metric events

Topdown metric events require grouping with a slots event. In perf
metrics this is currently achieved by metrics adding an unnecessary
"0 * tma_info_thread_slots". New TMA metrics trigger optimizations of
the metric expression that removes the event and breaks the metric due
to the missing but required event. Add a pass immediately before
sorting and fixing parsed events, that insert a slots event if one is
missing. Update test expectations to match this.

Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250719030517.1990983-15-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/arch/x86/util/evlist.c  | 24 ++++++++++++++++++++++++
 tools/perf/arch/x86/util/topdown.c | 28 ++++++++++++++++++++++++++++
 tools/perf/arch/x86/util/topdown.h |  2 ++
 tools/perf/tests/parse-events.c    | 24 ++++++++++++------------
 tools/perf/util/evlist.h           |  1 +
 tools/perf/util/parse-events.c     | 10 ++++++++++
 6 files changed, 77 insertions(+), 12 deletions(-)

(limited to 'tools/perf/arch')

diff --git a/tools/perf/arch/x86/util/evlist.c b/tools/perf/arch/x86/util/evlist.c
index 1969758cc8c1..75e9d00a1494 100644
--- a/tools/perf/arch/x86/util/evlist.c
+++ b/tools/perf/arch/x86/util/evlist.c
@@ -81,3 +81,27 @@ int arch_evlist__cmp(const struct evsel *lhs, const struct evsel *rhs)
 	/* Default ordering by insertion index. */
 	return lhs->core.idx - rhs->core.idx;
 }
+
+int arch_evlist__add_required_events(struct list_head *list)
+{
+	struct evsel *pos, *metric_event = NULL;
+	int idx = 0;
+
+	if (!topdown_sys_has_perf_metrics())
+		return 0;
+
+	list_for_each_entry(pos, list, core.node) {
+		if (arch_is_topdown_slots(pos)) {
+			/* Slots event already present, nothing to do. */
+			return 0;
+		}
+		if (metric_event == NULL && arch_is_topdown_metrics(pos))
+			metric_event = pos;
+		idx++;
+	}
+	if (metric_event == NULL) {
+		/* No topdown metric events, nothing to do. */
+		return 0;
+	}
+	return topdown_insert_slots_event(list, idx + 1, metric_event);
+}
diff --git a/tools/perf/arch/x86/util/topdown.c b/tools/perf/arch/x86/util/topdown.c
index 66b231fbf52e..0d01b662627a 100644
--- a/tools/perf/arch/x86/util/topdown.c
+++ b/tools/perf/arch/x86/util/topdown.c
@@ -77,3 +77,31 @@ bool arch_topdown_sample_read(struct evsel *leader)
 
 	return false;
 }
+
+/*
+ * Make a copy of the topdown metric event metric_event with the given index but
+ * change its configuration to be a topdown slots event. Copying from
+ * metric_event ensures modifiers are the same.
+ */
+int topdown_insert_slots_event(struct list_head *list, int idx, struct evsel *metric_event)
+{
+	struct evsel *evsel = evsel__new_idx(&metric_event->core.attr, idx);
+
+	if (!evsel)
+		return -ENOMEM;
+
+	evsel->core.attr.config = TOPDOWN_SLOTS;
+	evsel->core.cpus = perf_cpu_map__get(metric_event->core.cpus);
+	evsel->core.pmu_cpus = perf_cpu_map__get(metric_event->core.pmu_cpus);
+	evsel->core.is_pmu_core = true;
+	evsel->pmu = metric_event->pmu;
+	evsel->name = strdup("slots");
+	evsel->precise_max = metric_event->precise_max;
+	evsel->sample_read = metric_event->sample_read;
+	evsel->weak_group = metric_event->weak_group;
+	evsel->bpf_counter = metric_event->bpf_counter;
+	evsel->retire_lat = metric_event->retire_lat;
+	evsel__set_leader(evsel, evsel__leader(metric_event));
+	list_add_tail(&evsel->core.node, list);
+	return 0;
+}
diff --git a/tools/perf/arch/x86/util/topdown.h b/tools/perf/arch/x86/util/topdown.h
index 2349536cf882..69035565e649 100644
--- a/tools/perf/arch/x86/util/topdown.h
+++ b/tools/perf/arch/x86/util/topdown.h
@@ -5,9 +5,11 @@
 #include <stdbool.h>
 
 struct evsel;
+struct list_head;
 
 bool topdown_sys_has_perf_metrics(void);
 bool arch_is_topdown_slots(const struct evsel *evsel);
 bool arch_is_topdown_metrics(const struct evsel *evsel);
+int topdown_insert_slots_event(struct list_head *list, int idx, struct evsel *metric_event);
 
 #endif
diff --git a/tools/perf/tests/parse-events.c b/tools/perf/tests/parse-events.c
index 5ec2e5607987..bb8004397650 100644
--- a/tools/perf/tests/parse-events.c
+++ b/tools/perf/tests/parse-events.c
@@ -719,20 +719,20 @@ static int test__checkevent_pmu_partial_time_callgraph(struct evlist *evlist)
 
 static int test__checkevent_pmu_events(struct evlist *evlist)
 {
-	struct evsel *evsel = evlist__first(evlist);
+	struct evsel *evsel;
 
-	TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->core.nr_entries);
-	TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->core.attr.type ||
-				      strcmp(evsel->pmu->name, "cpu"));
-	TEST_ASSERT_VAL("wrong exclude_user",
-			!evsel->core.attr.exclude_user);
-	TEST_ASSERT_VAL("wrong exclude_kernel",
-			evsel->core.attr.exclude_kernel);
-	TEST_ASSERT_VAL("wrong exclude_hv", evsel->core.attr.exclude_hv);
-	TEST_ASSERT_VAL("wrong precise_ip", !evsel->core.attr.precise_ip);
-	TEST_ASSERT_VAL("wrong pinned", !evsel->core.attr.pinned);
-	TEST_ASSERT_VAL("wrong exclusive", !evsel->core.attr.exclusive);
+	TEST_ASSERT_VAL("wrong number of entries", 1 <= evlist->core.nr_entries);
 
+	evlist__for_each_entry(evlist, evsel) {
+		TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->core.attr.type ||
+					      strcmp(evsel->pmu->name, "cpu"));
+		TEST_ASSERT_VAL("wrong exclude_user", !evsel->core.attr.exclude_user);
+		TEST_ASSERT_VAL("wrong exclude_kernel", evsel->core.attr.exclude_kernel);
+		TEST_ASSERT_VAL("wrong exclude_hv", evsel->core.attr.exclude_hv);
+		TEST_ASSERT_VAL("wrong precise_ip", !evsel->core.attr.precise_ip);
+		TEST_ASSERT_VAL("wrong pinned", !evsel->core.attr.pinned);
+		TEST_ASSERT_VAL("wrong exclusive", !evsel->core.attr.exclusive);
+	}
 	return TEST_OK;
 }
 
diff --git a/tools/perf/util/evlist.h b/tools/perf/util/evlist.h
index fac1a01ba13f..1472d2179be1 100644
--- a/tools/perf/util/evlist.h
+++ b/tools/perf/util/evlist.h
@@ -111,6 +111,7 @@ void evlist__add(struct evlist *evlist, struct evsel *entry);
 void evlist__remove(struct evlist *evlist, struct evsel *evsel);
 
 int arch_evlist__cmp(const struct evsel *lhs, const struct evsel *rhs);
+int arch_evlist__add_required_events(struct list_head *list);
 
 int evlist__add_dummy(struct evlist *evlist);
 struct evsel *evlist__add_aux_dummy(struct evlist *evlist, bool system_wide);
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index fe2073c6b549..01fa8c80998b 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -2190,6 +2190,11 @@ static int evlist__cmp(void *_fg_idx, const struct list_head *l, const struct li
 	return arch_evlist__cmp(lhs, rhs);
 }
 
+int __weak arch_evlist__add_required_events(struct list_head *list __always_unused)
+{
+	return 0;
+}
+
 static int parse_events__sort_events_and_fix_groups(struct list_head *list)
 {
 	int idx = 0, force_grouped_idx = -1;
@@ -2201,6 +2206,11 @@ static int parse_events__sort_events_and_fix_groups(struct list_head *list)
 	struct evsel *force_grouped_leader = NULL;
 	bool last_event_was_forced_leader = false;
 
+	/* On x86 topdown metrics events require a slots event. */
+	ret = arch_evlist__add_required_events(list);
+	if (ret)
+		return ret;
+
 	/*
 	 * Compute index to insert ungrouped events at. Place them where the
 	 * first ungrouped event appears.
-- 
cgit v1.2.3


From 8882095b1d4d785524a7a4df8e04e35cfd039142 Mon Sep 17 00:00:00 2001
From: Ian Rogers <irogers@google.com>
Date: Thu, 24 Jul 2025 09:33:00 -0700
Subject: perf sample: Remove arch notion of sample parsing

By definition arch sample parsing and synthesis will inhibit certain
kinds of cross-platform record then analysis (report, script,
etc.). Remove arch_perf_parse_sample_weight and
arch_perf_synthesize_sample_weight replacing with a common
implementation. Combine perf_sample p_stage_cyc and retire_lat as
weight3 to capture the differing uses regardless of compiled for
architecture.

Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250724163302.596743-21-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/arch/powerpc/util/event.c       | 26 --------------------------
 tools/perf/arch/x86/tests/sample-parsing.c |  4 ++--
 tools/perf/arch/x86/util/event.c           | 27 ---------------------------
 tools/perf/builtin-script.c                |  2 +-
 tools/perf/util/dlfilter.c                 |  2 +-
 tools/perf/util/event.h                    |  2 --
 tools/perf/util/evsel.c                    | 17 ++++++++++++-----
 tools/perf/util/hist.c                     |  4 ++--
 tools/perf/util/hist.h                     |  3 ++-
 tools/perf/util/intel-tpebs.c              |  4 ++--
 tools/perf/util/sample.h                   |  6 ++----
 tools/perf/util/session.c                  |  2 +-
 tools/perf/util/sort.c                     |  7 +++----
 tools/perf/util/synthetic-events.c         | 10 ++++++++--
 14 files changed, 36 insertions(+), 80 deletions(-)

(limited to 'tools/perf/arch')

diff --git a/tools/perf/arch/powerpc/util/event.c b/tools/perf/arch/powerpc/util/event.c
index 77d8cc2b5691..024ac8b54c33 100644
--- a/tools/perf/arch/powerpc/util/event.c
+++ b/tools/perf/arch/powerpc/util/event.c
@@ -11,32 +11,6 @@
 #include "../../../util/debug.h"
 #include "../../../util/sample.h"
 
-void arch_perf_parse_sample_weight(struct perf_sample *data,
-				   const __u64 *array, u64 type)
-{
-	union perf_sample_weight weight;
-
-	weight.full = *array;
-	if (type & PERF_SAMPLE_WEIGHT)
-		data->weight = weight.full;
-	else {
-		data->weight = weight.var1_dw;
-		data->ins_lat = weight.var2_w;
-		data->p_stage_cyc = weight.var3_w;
-	}
-}
-
-void arch_perf_synthesize_sample_weight(const struct perf_sample *data,
-					__u64 *array, u64 type)
-{
-	*array = data->weight;
-
-	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
-		*array &= 0xffffffff;
-		*array |= ((u64)data->ins_lat << 32);
-	}
-}
-
 const char *arch_perf_header_entry(const char *se_header)
 {
 	if (!strcmp(se_header, "Local INSTR Latency"))
diff --git a/tools/perf/arch/x86/tests/sample-parsing.c b/tools/perf/arch/x86/tests/sample-parsing.c
index a061e8619267..22feec23e53d 100644
--- a/tools/perf/arch/x86/tests/sample-parsing.c
+++ b/tools/perf/arch/x86/tests/sample-parsing.c
@@ -29,7 +29,7 @@ static bool samples_same(const struct perf_sample *s1,
 {
 	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
 		COMP(ins_lat);
-		COMP(retire_lat);
+		COMP(weight3);
 	}
 
 	return true;
@@ -50,7 +50,7 @@ static int do_test(u64 sample_type)
 	struct perf_sample sample = {
 		.weight		= 101,
 		.ins_lat        = 102,
-		.retire_lat     = 103,
+		.weight3	= 103,
 	};
 	struct perf_sample sample_out;
 	size_t i, sz, bufsz;
diff --git a/tools/perf/arch/x86/util/event.c b/tools/perf/arch/x86/util/event.c
index a0400707180c..576c1c36046c 100644
--- a/tools/perf/arch/x86/util/event.c
+++ b/tools/perf/arch/x86/util/event.c
@@ -92,33 +92,6 @@ int perf_event__synthesize_extra_kmaps(const struct perf_tool *tool,
 
 #endif
 
-void arch_perf_parse_sample_weight(struct perf_sample *data,
-				   const __u64 *array, u64 type)
-{
-	union perf_sample_weight weight;
-
-	weight.full = *array;
-	if (type & PERF_SAMPLE_WEIGHT)
-		data->weight = weight.full;
-	else {
-		data->weight = weight.var1_dw;
-		data->ins_lat = weight.var2_w;
-		data->retire_lat = weight.var3_w;
-	}
-}
-
-void arch_perf_synthesize_sample_weight(const struct perf_sample *data,
-					__u64 *array, u64 type)
-{
-	*array = data->weight;
-
-	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
-		*array &= 0xffffffff;
-		*array |= ((u64)data->ins_lat << 32);
-		*array |= ((u64)data->retire_lat << 48);
-	}
-}
-
 const char *arch_perf_header_entry(const char *se_header)
 {
 	if (!strcmp(se_header, "Local Pipeline Stage Cycle"))
diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index f2b5620165b4..d9fbdcf72f25 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -2252,7 +2252,7 @@ static void process_event(struct perf_script *script,
 		fprintf(fp, "%16" PRIu16, sample->ins_lat);
 
 	if (PRINT_FIELD(RETIRE_LAT))
-		fprintf(fp, "%16" PRIu16, sample->retire_lat);
+		fprintf(fp, "%16" PRIu16, sample->weight3);
 
 	if (PRINT_FIELD(CGROUP)) {
 		const char *cgrp_name;
diff --git a/tools/perf/util/dlfilter.c b/tools/perf/util/dlfilter.c
index ddacef881af2..c0afcbd954f8 100644
--- a/tools/perf/util/dlfilter.c
+++ b/tools/perf/util/dlfilter.c
@@ -513,6 +513,7 @@ int dlfilter__do_filter_event(struct dlfilter *d,
 	d->d_addr_al   = &d_addr_al;
 
 	d_sample.size  = sizeof(d_sample);
+	d_sample.p_stage_cyc = sample->weight3;
 	d_ip_al.size   = 0; /* To indicate d_ip_al is not initialized */
 	d_addr_al.size = 0; /* To indicate d_addr_al is not initialized */
 
@@ -526,7 +527,6 @@ int dlfilter__do_filter_event(struct dlfilter *d,
 	ASSIGN(period);
 	ASSIGN(weight);
 	ASSIGN(ins_lat);
-	ASSIGN(p_stage_cyc);
 	ASSIGN(transaction);
 	ASSIGN(insn_cnt);
 	ASSIGN(cyc_cnt);
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index 67ad4a2014bc..b13385a6068b 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -391,8 +391,6 @@ extern unsigned int proc_map_timeout;
 #define PAGE_SIZE_NAME_LEN	32
 char *get_page_size_name(u64 size, char *str);
 
-void arch_perf_parse_sample_weight(struct perf_sample *data, const __u64 *array, u64 type);
-void arch_perf_synthesize_sample_weight(const struct perf_sample *data, __u64 *array, u64 type);
 const char *arch_perf_header_entry(const char *se_header);
 int arch_support_sort_key(const char *sort_key);
 
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index aa6efcc4404c..3d27e9bdd66b 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -2880,11 +2880,18 @@ perf_event__check_size(union perf_event *event, unsigned int sample_size)
 	return 0;
 }
 
-void __weak arch_perf_parse_sample_weight(struct perf_sample *data,
-					  const __u64 *array,
-					  u64 type __maybe_unused)
+static void perf_parse_sample_weight(struct perf_sample *data, const __u64 *array, u64 type)
 {
-	data->weight = *array;
+	union perf_sample_weight weight;
+
+	weight.full = *array;
+	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
+		data->weight = weight.var1_dw;
+		data->ins_lat = weight.var2_w;
+		data->weight3 = weight.var3_w;
+	} else {
+		data->weight = weight.full;
+	}
 }
 
 u64 evsel__bitfield_swap_branch_flags(u64 value)
@@ -3270,7 +3277,7 @@ int evsel__parse_sample(struct evsel *evsel, union perf_event *event,
 
 	if (type & PERF_SAMPLE_WEIGHT_TYPE) {
 		OVERFLOW_CHECK_u64(array);
-		arch_perf_parse_sample_weight(data, array, type);
+		perf_parse_sample_weight(data, array, type);
 		array++;
 	}
 
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
index afc6855327ab..64ff427040c3 100644
--- a/tools/perf/util/hist.c
+++ b/tools/perf/util/hist.c
@@ -829,7 +829,7 @@ __hists__add_entry(struct hists *hists,
 			.period	= sample->period,
 			.weight1 = sample->weight,
 			.weight2 = sample->ins_lat,
-			.weight3 = sample->p_stage_cyc,
+			.weight3 = sample->weight3,
 			.latency = al->latency,
 		},
 		.parent = sym_parent,
@@ -846,7 +846,7 @@ __hists__add_entry(struct hists *hists,
 		.time = hist_time(sample->time),
 		.weight = sample->weight,
 		.ins_lat = sample->ins_lat,
-		.p_stage_cyc = sample->p_stage_cyc,
+		.weight3 = sample->weight3,
 		.simd_flags = sample->simd_flags,
 	}, *he = hists__findnew_entry(hists, &entry, al, sample_self);
 
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
index c64254088fc7..70438d03ca9c 100644
--- a/tools/perf/util/hist.h
+++ b/tools/perf/util/hist.h
@@ -255,7 +255,8 @@ struct hist_entry {
 	u64			code_page_size;
 	u64			weight;
 	u64			ins_lat;
-	u64			p_stage_cyc;
+	/** @weight3: On x86 holds retire_lat, on powerpc holds p_stage_cyc. */
+	u64			weight3;
 	s32			socket;
 	s32			cpu;
 	int			parallelism;
diff --git a/tools/perf/util/intel-tpebs.c b/tools/perf/util/intel-tpebs.c
index 3b92ebf5c112..8c9aee157ec4 100644
--- a/tools/perf/util/intel-tpebs.c
+++ b/tools/perf/util/intel-tpebs.c
@@ -210,8 +210,8 @@ static int process_sample_event(const struct perf_tool *tool __maybe_unused,
 	 * latency value will be used. Save the number of samples and the sum of
 	 * retire latency value for each event.
 	 */
-	t->last = sample->retire_lat;
-	update_stats(&t->stats, sample->retire_lat);
+	t->last = sample->weight3;
+	update_stats(&t->stats, sample->weight3);
 	mutex_unlock(tpebs_mtx_get());
 	return 0;
 }
diff --git a/tools/perf/util/sample.h b/tools/perf/util/sample.h
index 0e96240052e9..fae834144ef4 100644
--- a/tools/perf/util/sample.h
+++ b/tools/perf/util/sample.h
@@ -104,10 +104,8 @@ struct perf_sample {
 	u8  cpumode;
 	u16 misc;
 	u16 ins_lat;
-	union {
-		u16 p_stage_cyc;
-		u16 retire_lat;
-	};
+	/** @weight3: On x86 holds retire_lat, on powerpc holds p_stage_cyc. */
+	u16 weight3;
 	bool no_hw_idx;		/* No hw_idx collected in branch_stack */
 	char insn[MAX_INSN];
 	void *raw_data;
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 2a79e6844f36..26ae078278cd 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -1099,7 +1099,7 @@ static void dump_sample(struct evsel *evsel, union perf_event *event,
 		printf("... weight: %" PRIu64 "", sample->weight);
 			if (sample_type & PERF_SAMPLE_WEIGHT_STRUCT) {
 				printf(",0x%"PRIx16"", sample->ins_lat);
-				printf(",0x%"PRIx16"", sample->p_stage_cyc);
+				printf(",0x%"PRIx16"", sample->weight3);
 			}
 		printf("\n");
 	}
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 7969d64a47bf..0ba2ce1b1c07 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -1884,21 +1884,20 @@ struct sort_entry sort_global_ins_lat = {
 static int64_t
 sort__p_stage_cyc_cmp(struct hist_entry *left, struct hist_entry *right)
 {
-	return left->p_stage_cyc - right->p_stage_cyc;
+	return left->weight3 - right->weight3;
 }
 
 static int hist_entry__global_p_stage_cyc_snprintf(struct hist_entry *he, char *bf,
 					size_t size, unsigned int width)
 {
-	return repsep_snprintf(bf, size, "%-*u", width,
-			he->p_stage_cyc * he->stat.nr_events);
+	return repsep_snprintf(bf, size, "%-*u", width, he->weight3 * he->stat.nr_events);
 }
 
 
 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->p_stage_cyc);
+	return repsep_snprintf(bf, size, "%-*u", width, he->weight3);
 }
 
 struct sort_entry sort_local_p_stage_cyc = {
diff --git a/tools/perf/util/synthetic-events.c b/tools/perf/util/synthetic-events.c
index e7ca3f5eb493..cb2c1ace304a 100644
--- a/tools/perf/util/synthetic-events.c
+++ b/tools/perf/util/synthetic-events.c
@@ -1573,10 +1573,16 @@ size_t perf_event__sample_event_size(const struct perf_sample *sample, u64 type,
 	return result;
 }
 
-void __weak arch_perf_synthesize_sample_weight(const struct perf_sample *data,
+static void perf_synthesize_sample_weight(const struct perf_sample *data,
 					       __u64 *array, u64 type __maybe_unused)
 {
 	*array = data->weight;
+
+	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
+		*array &= 0xffffffff;
+		*array |= ((u64)data->ins_lat << 32);
+		*array |= ((u64)data->weight3 << 48);
+	}
 }
 
 static __u64 *copy_read_group_values(__u64 *array, __u64 read_format,
@@ -1736,7 +1742,7 @@ int perf_event__synthesize_sample(union perf_event *event, u64 type, u64 read_fo
 	}
 
 	if (type & PERF_SAMPLE_WEIGHT_TYPE) {
-		arch_perf_synthesize_sample_weight(sample, array, type);
+		perf_synthesize_sample_weight(sample, array, type);
 		array++;
 	}
 
-- 
cgit v1.2.3


From a563c9f3bb8c23416f3e72edfbc75d1a4937f7e0 Mon Sep 17 00:00:00 2001
From: Ian Rogers <irogers@google.com>
Date: Thu, 24 Jul 2025 09:33:01 -0700
Subject: perf test: Move PERF_SAMPLE_WEIGHT_STRUCT parsing to common test

test__x86_sample_parsing is identical to test__sample_parsing except
it explicitly tested PERF_SAMPLE_WEIGHT_STRUCT. Now the parsing code
is common move the PERF_SAMPLE_WEIGHT_STRUCT to the common sample
parsing test and remove the x86 version.

Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250724163302.596743-22-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/arch/x86/include/arch-tests.h   |   1 -
 tools/perf/arch/x86/tests/Build            |   1 -
 tools/perf/arch/x86/tests/arch-tests.c     |   2 -
 tools/perf/arch/x86/tests/sample-parsing.c | 125 -----------------------------
 tools/perf/tests/sample-parsing.c          |  14 ++++
 5 files changed, 14 insertions(+), 129 deletions(-)
 delete mode 100644 tools/perf/arch/x86/tests/sample-parsing.c

(limited to 'tools/perf/arch')

diff --git a/tools/perf/arch/x86/include/arch-tests.h b/tools/perf/arch/x86/include/arch-tests.h
index 8713e9122d4c..7d65b9e51840 100644
--- a/tools/perf/arch/x86/include/arch-tests.h
+++ b/tools/perf/arch/x86/include/arch-tests.h
@@ -14,7 +14,6 @@ int test__insn_x86(struct test_suite *test, int subtest);
 int test__intel_pt_pkt_decoder(struct test_suite *test, int subtest);
 int test__intel_pt_hybrid_compat(struct test_suite *test, int subtest);
 int test__bp_modify(struct test_suite *test, int subtest);
-int test__x86_sample_parsing(struct test_suite *test, int subtest);
 int test__amd_ibs_via_core_pmu(struct test_suite *test, int subtest);
 int test__amd_ibs_period(struct test_suite *test, int subtest);
 int test__hybrid(struct test_suite *test, int subtest);
diff --git a/tools/perf/arch/x86/tests/Build b/tools/perf/arch/x86/tests/Build
index 311b6b53d3d8..7790b3e20f4e 100644
--- a/tools/perf/arch/x86/tests/Build
+++ b/tools/perf/arch/x86/tests/Build
@@ -2,7 +2,6 @@ perf-test-$(CONFIG_DWARF_UNWIND) += regs_load.o
 perf-test-$(CONFIG_DWARF_UNWIND) += dwarf-unwind.o
 
 perf-test-y += arch-tests.o
-perf-test-y += sample-parsing.o
 perf-test-y += hybrid.o
 perf-test-$(CONFIG_AUXTRACE) += intel-pt-test.o
 ifeq ($(CONFIG_EXTRA_TESTS),y)
diff --git a/tools/perf/arch/x86/tests/arch-tests.c b/tools/perf/arch/x86/tests/arch-tests.c
index 29ec1861ccef..8f9cfeaa170f 100644
--- a/tools/perf/arch/x86/tests/arch-tests.c
+++ b/tools/perf/arch/x86/tests/arch-tests.c
@@ -23,7 +23,6 @@ struct test_suite suite__intel_pt = {
 #if defined(__x86_64__)
 DEFINE_SUITE("x86 bp modify", bp_modify);
 #endif
-DEFINE_SUITE("x86 Sample parsing", x86_sample_parsing);
 DEFINE_SUITE("AMD IBS via core pmu", amd_ibs_via_core_pmu);
 DEFINE_SUITE_EXCLUSIVE("AMD IBS sample period", amd_ibs_period);
 static struct test_case hybrid_tests[] = {
@@ -49,7 +48,6 @@ struct test_suite *arch_tests[] = {
 #if defined(__x86_64__)
 	&suite__bp_modify,
 #endif
-	&suite__x86_sample_parsing,
 	&suite__amd_ibs_via_core_pmu,
 	&suite__amd_ibs_period,
 	&suite__hybrid,
diff --git a/tools/perf/arch/x86/tests/sample-parsing.c b/tools/perf/arch/x86/tests/sample-parsing.c
deleted file mode 100644
index 22feec23e53d..000000000000
--- a/tools/perf/arch/x86/tests/sample-parsing.c
+++ /dev/null
@@ -1,125 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-#include <stdbool.h>
-#include <inttypes.h>
-#include <stdlib.h>
-#include <string.h>
-#include <linux/bitops.h>
-#include <linux/kernel.h>
-#include <linux/types.h>
-
-#include "event.h"
-#include "evsel.h"
-#include "debug.h"
-#include "util/sample.h"
-#include "util/synthetic-events.h"
-
-#include "tests/tests.h"
-#include "arch-tests.h"
-
-#define COMP(m) do {					\
-	if (s1->m != s2->m) {				\
-		pr_debug("Samples differ at '"#m"'\n");	\
-		return false;				\
-	}						\
-} while (0)
-
-static bool samples_same(const struct perf_sample *s1,
-			 const struct perf_sample *s2,
-			 u64 type)
-{
-	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
-		COMP(ins_lat);
-		COMP(weight3);
-	}
-
-	return true;
-}
-
-static int do_test(u64 sample_type)
-{
-	struct evsel evsel = {
-		.needs_swap = false,
-		.core = {
-			. attr = {
-				.sample_type = sample_type,
-				.read_format = 0,
-			},
-		},
-	};
-	union perf_event *event;
-	struct perf_sample sample = {
-		.weight		= 101,
-		.ins_lat        = 102,
-		.weight3	= 103,
-	};
-	struct perf_sample sample_out;
-	size_t i, sz, bufsz;
-	int err, ret = -1;
-
-	sz = perf_event__sample_event_size(&sample, sample_type, 0);
-	bufsz = sz + 4096; /* Add a bit for overrun checking */
-	event = malloc(bufsz);
-	if (!event) {
-		pr_debug("malloc failed\n");
-		return -1;
-	}
-
-	memset(event, 0xff, bufsz);
-	event->header.type = PERF_RECORD_SAMPLE;
-	event->header.misc = 0;
-	event->header.size = sz;
-
-	err = perf_event__synthesize_sample(event, sample_type, 0, &sample);
-	if (err) {
-		pr_debug("%s failed for sample_type %#"PRIx64", error %d\n",
-			 "perf_event__synthesize_sample", sample_type, err);
-		goto out_free;
-	}
-
-	/* The data does not contain 0xff so we use that to check the size */
-	for (i = bufsz; i > 0; i--) {
-		if (*(i - 1 + (u8 *)event) != 0xff)
-			break;
-	}
-	if (i != sz) {
-		pr_debug("Event size mismatch: actual %zu vs expected %zu\n",
-			 i, sz);
-		goto out_free;
-	}
-
-	evsel.sample_size = __evsel__sample_size(sample_type);
-
-	err = evsel__parse_sample(&evsel, event, &sample_out);
-	if (err) {
-		pr_debug("%s failed for sample_type %#"PRIx64", error %d\n",
-			 "evsel__parse_sample", sample_type, err);
-		goto out_free;
-	}
-
-	if (!samples_same(&sample, &sample_out, sample_type)) {
-		pr_debug("parsing failed for sample_type %#"PRIx64"\n",
-			 sample_type);
-		goto out_free;
-	}
-
-	ret = 0;
-out_free:
-	free(event);
-
-	return ret;
-}
-
-/**
- * test__x86_sample_parsing - test X86 specific sample parsing
- *
- * This function implements a test that synthesizes a sample event, parses it
- * and then checks that the parsed sample matches the original sample. If the
- * test passes %0 is returned, otherwise %-1 is returned.
- *
- * For now, the PERF_SAMPLE_WEIGHT_STRUCT is the only X86 specific sample type.
- * The test only checks the PERF_SAMPLE_WEIGHT_STRUCT type.
- */
-int test__x86_sample_parsing(struct test_suite *test __maybe_unused, int subtest __maybe_unused)
-{
-	return do_test(PERF_SAMPLE_WEIGHT_STRUCT);
-}
diff --git a/tools/perf/tests/sample-parsing.c b/tools/perf/tests/sample-parsing.c
index 72411580f869..a7327c942ca2 100644
--- a/tools/perf/tests/sample-parsing.c
+++ b/tools/perf/tests/sample-parsing.c
@@ -152,6 +152,12 @@ static bool samples_same(struct perf_sample *s1,
 	if (type & PERF_SAMPLE_WEIGHT)
 		COMP(weight);
 
+	if (type & PERF_SAMPLE_WEIGHT_STRUCT) {
+		COMP(weight);
+		COMP(ins_lat);
+		COMP(weight3);
+	}
+
 	if (type & PERF_SAMPLE_DATA_SRC)
 		COMP(data_src);
 
@@ -269,6 +275,8 @@ static int do_test(u64 sample_type, u64 sample_regs, u64 read_format)
 		.cgroup		= 114,
 		.data_page_size = 115,
 		.code_page_size = 116,
+		.ins_lat	= 117,
+		.weight3	= 118,
 		.aux_sample	= {
 			.size	= sizeof(aux_data),
 			.data	= (void *)aux_data,
@@ -439,6 +447,12 @@ static int test__sample_parsing(struct test_suite *test __maybe_unused, int subt
 		if (err)
 			return err;
 	}
+	sample_type = (PERF_SAMPLE_MAX - 1) & ~PERF_SAMPLE_WEIGHT_STRUCT;
+	for (i = 0; i < ARRAY_SIZE(rf); i++) {
+		err = do_test(sample_type, sample_regs, rf[i]);
+		if (err)
+			return err;
+	}
 
 	return 0;
 }
-- 
cgit v1.2.3


From 6e19839a80b8713b836722ba9d99a3ab12cfb651 Mon Sep 17 00:00:00 2001
From: Ian Rogers <irogers@google.com>
Date: Thu, 24 Jul 2025 09:33:02 -0700
Subject: perf sort: Use perf_env to set arch sort keys and header

Previously arch_support_sort_key and arch_perf_header_entry used a
weak symbol to compile as appropriate for x86 and powerpc. A
limitation to this is that the handling of a data file could vary in
cross-platform development. Change to using the perf_env of the
current session to determine the architecture kind and set the sort
key and header entries as appropriate.

Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250724163302.596743-23-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/arch/powerpc/util/Build   |  1 -
 tools/perf/arch/powerpc/util/event.c | 34 --------------------
 tools/perf/arch/x86/util/event.c     | 19 -----------
 tools/perf/builtin-annotate.c        |  2 +-
 tools/perf/builtin-c2c.c             | 53 +++++++++++++++++--------------
 tools/perf/builtin-diff.c            |  2 +-
 tools/perf/builtin-report.c          |  2 +-
 tools/perf/builtin-top.c             | 22 ++++++-------
 tools/perf/tests/hists_cumulate.c    |  8 ++---
 tools/perf/tests/hists_filter.c      |  8 ++---
 tools/perf/tests/hists_link.c        |  8 ++---
 tools/perf/tests/hists_output.c      | 10 +++---
 tools/perf/util/event.h              |  3 --
 tools/perf/util/sort.c               | 61 +++++++++++++++++++++++++-----------
 tools/perf/util/sort.h               |  5 +--
 15 files changed, 107 insertions(+), 131 deletions(-)
 delete mode 100644 tools/perf/arch/powerpc/util/event.c

(limited to 'tools/perf/arch')

diff --git a/tools/perf/arch/powerpc/util/Build b/tools/perf/arch/powerpc/util/Build
index ed82715080f9..fdd6a77a3432 100644
--- a/tools/perf/arch/powerpc/util/Build
+++ b/tools/perf/arch/powerpc/util/Build
@@ -5,7 +5,6 @@ perf-util-y += mem-events.o
 perf-util-y += pmu.o
 perf-util-y += sym-handling.o
 perf-util-y += evsel.o
-perf-util-y += event.o
 
 perf-util-$(CONFIG_LIBDW) += skip-callchain-idx.o
 
diff --git a/tools/perf/arch/powerpc/util/event.c b/tools/perf/arch/powerpc/util/event.c
deleted file mode 100644
index 024ac8b54c33..000000000000
--- a/tools/perf/arch/powerpc/util/event.c
+++ /dev/null
@@ -1,34 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0
-#include <linux/types.h>
-#include <linux/string.h>
-#include <linux/zalloc.h>
-
-#include "../../../util/event.h"
-#include "../../../util/synthetic-events.h"
-#include "../../../util/machine.h"
-#include "../../../util/tool.h"
-#include "../../../util/map.h"
-#include "../../../util/debug.h"
-#include "../../../util/sample.h"
-
-const char *arch_perf_header_entry(const char *se_header)
-{
-	if (!strcmp(se_header, "Local INSTR Latency"))
-		return "Finish Cyc";
-	else if (!strcmp(se_header, "INSTR Latency"))
-		return "Global Finish_cyc";
-	else if (!strcmp(se_header, "Local Pipeline Stage Cycle"))
-		return "Dispatch Cyc";
-	else if (!strcmp(se_header, "Pipeline Stage Cycle"))
-		return "Global Dispatch_cyc";
-	return se_header;
-}
-
-int arch_support_sort_key(const char *sort_key)
-{
-	if (!strcmp(sort_key, "p_stage_cyc"))
-		return 1;
-	if (!strcmp(sort_key, "local_p_stage_cyc"))
-		return 1;
-	return 0;
-}
diff --git a/tools/perf/arch/x86/util/event.c b/tools/perf/arch/x86/util/event.c
index 576c1c36046c..3cd384317739 100644
--- a/tools/perf/arch/x86/util/event.c
+++ b/tools/perf/arch/x86/util/event.c
@@ -91,22 +91,3 @@ int perf_event__synthesize_extra_kmaps(const struct perf_tool *tool,
 }
 
 #endif
-
-const char *arch_perf_header_entry(const char *se_header)
-{
-	if (!strcmp(se_header, "Local Pipeline Stage Cycle"))
-		return "Local Retire Latency";
-	else if (!strcmp(se_header, "Pipeline Stage Cycle"))
-		return "Retire Latency";
-
-	return se_header;
-}
-
-int arch_support_sort_key(const char *sort_key)
-{
-	if (!strcmp(sort_key, "p_stage_cyc"))
-		return 1;
-	if (!strcmp(sort_key, "local_p_stage_cyc"))
-		return 1;
-	return 0;
-}
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 326593862998..5d57d2913f3d 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -947,7 +947,7 @@ int cmd_annotate(int argc, const char **argv)
 			annotate_opts.show_br_cntr = true;
 	}
 
-	if (setup_sorting(NULL) < 0)
+	if (setup_sorting(/*evlist=*/NULL, perf_session__env(annotate.session)) < 0)
 		usage_with_options(annotate_usage, options);
 
 	ret = __cmd_annotate(&annotate);
diff --git a/tools/perf/builtin-c2c.c b/tools/perf/builtin-c2c.c
index 8cb36d9433f8..9e9ff471ddd1 100644
--- a/tools/perf/builtin-c2c.c
+++ b/tools/perf/builtin-c2c.c
@@ -195,12 +195,14 @@ static struct hist_entry_ops c2c_entry_ops = {
 
 static int c2c_hists__init(struct c2c_hists *hists,
 			   const char *sort,
-			   int nr_header_lines);
+			   int nr_header_lines,
+			   struct perf_env *env);
 
 static struct c2c_hists*
 he__get_c2c_hists(struct hist_entry *he,
 		  const char *sort,
-		  int nr_header_lines)
+		  int nr_header_lines,
+		  struct perf_env *env)
 {
 	struct c2c_hist_entry *c2c_he;
 	struct c2c_hists *hists;
@@ -214,7 +216,7 @@ he__get_c2c_hists(struct hist_entry *he,
 	if (!hists)
 		return NULL;
 
-	ret = c2c_hists__init(hists, sort, nr_header_lines);
+	ret = c2c_hists__init(hists, sort, nr_header_lines, env);
 	if (ret) {
 		free(hists);
 		return NULL;
@@ -350,7 +352,7 @@ static int process_sample_event(const struct perf_tool *tool __maybe_unused,
 
 		mi = mi_dup;
 
-		c2c_hists = he__get_c2c_hists(he, c2c.cl_sort, 2);
+		c2c_hists = he__get_c2c_hists(he, c2c.cl_sort, 2, machine->env);
 		if (!c2c_hists)
 			goto free_mi;
 
@@ -1966,7 +1968,8 @@ static struct c2c_fmt *get_format(const char *name)
 	return c2c_fmt;
 }
 
-static int c2c_hists__init_output(struct perf_hpp_list *hpp_list, char *name)
+static int c2c_hists__init_output(struct perf_hpp_list *hpp_list, char *name,
+				  struct perf_env *env __maybe_unused)
 {
 	struct c2c_fmt *c2c_fmt = get_format(name);
 	int level = 0;
@@ -1980,14 +1983,14 @@ static int c2c_hists__init_output(struct perf_hpp_list *hpp_list, char *name)
 	return 0;
 }
 
-static int c2c_hists__init_sort(struct perf_hpp_list *hpp_list, char *name)
+static int c2c_hists__init_sort(struct perf_hpp_list *hpp_list, char *name, struct perf_env *env)
 {
 	struct c2c_fmt *c2c_fmt = get_format(name);
 	struct c2c_dimension *dim;
 
 	if (!c2c_fmt) {
 		reset_dimensions();
-		return sort_dimension__add(hpp_list, name, NULL, 0);
+		return sort_dimension__add(hpp_list, name, /*evlist=*/NULL, env, /*level=*/0);
 	}
 
 	dim = c2c_fmt->dim;
@@ -2008,7 +2011,7 @@ static int c2c_hists__init_sort(struct perf_hpp_list *hpp_list, char *name)
 										\
 		for (tok = strtok_r((char *)_list, ", ", &tmp);			\
 				tok; tok = strtok_r(NULL, ", ", &tmp)) {	\
-			ret = _fn(hpp_list, tok);				\
+			ret = _fn(hpp_list, tok, env);				\
 			if (ret == -EINVAL) {					\
 				pr_err("Invalid --fields key: `%s'", tok);	\
 				break;						\
@@ -2021,7 +2024,8 @@ static int c2c_hists__init_sort(struct perf_hpp_list *hpp_list, char *name)
 
 static int hpp_list__parse(struct perf_hpp_list *hpp_list,
 			   const char *output_,
-			   const char *sort_)
+			   const char *sort_,
+			   struct perf_env *env)
 {
 	char *output = output_ ? strdup(output_) : NULL;
 	char *sort   = sort_   ? strdup(sort_) : NULL;
@@ -2052,7 +2056,8 @@ static int hpp_list__parse(struct perf_hpp_list *hpp_list,
 
 static int c2c_hists__init(struct c2c_hists *hists,
 			   const char *sort,
-			   int nr_header_lines)
+			   int nr_header_lines,
+			   struct perf_env *env)
 {
 	__hists__init(&hists->hists, &hists->list);
 
@@ -2066,15 +2071,16 @@ static int c2c_hists__init(struct c2c_hists *hists,
 	/* Overload number of header lines.*/
 	hists->list.nr_header_lines = nr_header_lines;
 
-	return hpp_list__parse(&hists->list, NULL, sort);
+	return hpp_list__parse(&hists->list, /*output=*/NULL, sort, env);
 }
 
 static int c2c_hists__reinit(struct c2c_hists *c2c_hists,
 			     const char *output,
-			     const char *sort)
+			     const char *sort,
+			     struct perf_env *env)
 {
 	perf_hpp__reset_output_field(&c2c_hists->list);
-	return hpp_list__parse(&c2c_hists->list, output, sort);
+	return hpp_list__parse(&c2c_hists->list, output, sort, env);
 }
 
 #define DISPLAY_LINE_LIMIT  0.001
@@ -2207,8 +2213,9 @@ static int filter_cb(struct hist_entry *he, void *arg __maybe_unused)
 	return 0;
 }
 
-static int resort_cl_cb(struct hist_entry *he, void *arg __maybe_unused)
+static int resort_cl_cb(struct hist_entry *he, void *arg)
 {
+	struct perf_env *env = arg;
 	struct c2c_hist_entry *c2c_he;
 	struct c2c_hists *c2c_hists;
 	bool display = he__display(he, &c2c.shared_clines_stats);
@@ -2222,7 +2229,7 @@ static int resort_cl_cb(struct hist_entry *he, void *arg __maybe_unused)
 		c2c_he->cacheline_idx = idx++;
 		calc_width(c2c_he);
 
-		c2c_hists__reinit(c2c_hists, c2c.cl_output, c2c.cl_resort);
+		c2c_hists__reinit(c2c_hists, c2c.cl_output, c2c.cl_resort, env);
 
 		hists__collapse_resort(&c2c_hists->hists, NULL);
 		hists__output_resort_cb(&c2c_hists->hists, NULL, filter_cb);
@@ -2334,7 +2341,7 @@ static int resort_shared_cl_cb(struct hist_entry *he, void *arg __maybe_unused)
 	return 0;
 }
 
-static int hists__iterate_cb(struct hists *hists, hists__resort_cb_t cb)
+static int hists__iterate_cb(struct hists *hists, hists__resort_cb_t cb, void *arg)
 {
 	struct rb_node *next = rb_first_cached(&hists->entries);
 	int ret = 0;
@@ -2343,7 +2350,7 @@ static int hists__iterate_cb(struct hists *hists, hists__resort_cb_t cb)
 		struct hist_entry *he;
 
 		he = rb_entry(next, struct hist_entry, rb_node);
-		ret = cb(he, NULL);
+		ret = cb(he, arg);
 		if (ret)
 			break;
 		next = rb_next(&he->rb_node);
@@ -2449,7 +2456,7 @@ static void print_cacheline(struct c2c_hists *c2c_hists,
 	hists__fprintf(&c2c_hists->hists, false, 0, 0, 0, out, false);
 }
 
-static void print_pareto(FILE *out)
+static void print_pareto(FILE *out, struct perf_env *env)
 {
 	struct perf_hpp_list hpp_list;
 	struct rb_node *nd;
@@ -2474,7 +2481,7 @@ static void print_pareto(FILE *out)
 			    "dcacheline";
 
 	perf_hpp_list__init(&hpp_list);
-	ret = hpp_list__parse(&hpp_list, cl_output, NULL);
+	ret = hpp_list__parse(&hpp_list, cl_output, /*evlist=*/NULL, env);
 
 	if (WARN_ONCE(ret, "failed to setup sort entries\n"))
 		return;
@@ -2539,7 +2546,7 @@ static void perf_c2c__hists_fprintf(FILE *out, struct perf_session *session)
 	fprintf(out, "=================================================\n");
 	fprintf(out, "#\n");
 
-	print_pareto(out);
+	print_pareto(out, perf_session__env(session));
 }
 
 #ifdef HAVE_SLANG_SUPPORT
@@ -3097,7 +3104,7 @@ static int perf_c2c__report(int argc, const char **argv)
 		goto out_session;
 	}
 
-	err = c2c_hists__init(&c2c.hists, "dcacheline", 2);
+	err = c2c_hists__init(&c2c.hists, "dcacheline", 2, perf_session__env(session));
 	if (err) {
 		pr_debug("Failed to initialize hists\n");
 		goto out_session;
@@ -3181,13 +3188,13 @@ static int perf_c2c__report(int argc, const char **argv)
 	else if (c2c.display == DISPLAY_SNP_PEER)
 		sort_str = "tot_peer";
 
-	c2c_hists__reinit(&c2c.hists, output_str, sort_str);
+	c2c_hists__reinit(&c2c.hists, output_str, sort_str, perf_session__env(session));
 
 	ui_progress__init(&prog, c2c.hists.hists.nr_entries, "Sorting...");
 
 	hists__collapse_resort(&c2c.hists.hists, NULL);
 	hists__output_resort_cb(&c2c.hists.hists, &prog, resort_shared_cl_cb);
-	hists__iterate_cb(&c2c.hists.hists, resort_cl_cb);
+	hists__iterate_cb(&c2c.hists.hists, resort_cl_cb, perf_session__env(session));
 
 	ui_progress__finish();
 
diff --git a/tools/perf/builtin-diff.c b/tools/perf/builtin-diff.c
index ae490d58af92..53d5ea4a6a4f 100644
--- a/tools/perf/builtin-diff.c
+++ b/tools/perf/builtin-diff.c
@@ -2003,7 +2003,7 @@ int cmd_diff(int argc, const char **argv)
 		sort__mode = SORT_MODE__DIFF;
 	}
 
-	if (setup_sorting(NULL) < 0)
+	if (setup_sorting(/*evlist=*/NULL, perf_session__env(data__files[0].session)) < 0)
 		usage_with_options(diff_usage, options);
 
 	setup_pager();
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index ada8e0166c78..35df04dad2fd 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -1790,7 +1790,7 @@ repeat:
 	}
 
 	if ((last_key != K_SWITCH_INPUT_DATA && last_key != K_RELOAD) &&
-	    (setup_sorting(session->evlist) < 0)) {
+	    (setup_sorting(session->evlist, perf_session__env(session)) < 0)) {
 		if (sort_order)
 			parse_options_usage(report_usage, options, "s", 1);
 		if (field_order)
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index e9743f17bd0c..a11f629c7d76 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -1767,7 +1767,17 @@ int cmd_top(int argc, const char **argv)
 
 	setup_browser(false);
 
-	if (setup_sorting(top.evlist) < 0) {
+	top.session = __perf_session__new(/*data=*/NULL, /*tool=*/NULL,
+					  /*trace_event_repipe=*/false,
+					  &host_env);
+	if (IS_ERR(top.session)) {
+		status = PTR_ERR(top.session);
+		top.session = NULL;
+		goto out_delete_evlist;
+	}
+	top.evlist->session = top.session;
+
+	if (setup_sorting(top.evlist, perf_session__env(top.session)) < 0) {
 		if (sort_order)
 			parse_options_usage(top_usage, options, "s", 1);
 		if (field_order)
@@ -1842,16 +1852,6 @@ int cmd_top(int argc, const char **argv)
 		signal(SIGWINCH, winch_sig);
 	}
 
-	top.session = __perf_session__new(/*data=*/NULL, /*tool=*/NULL,
-					  /*trace_event_repipe=*/false,
-					  &host_env);
-	if (IS_ERR(top.session)) {
-		status = PTR_ERR(top.session);
-		top.session = NULL;
-		goto out_delete_evlist;
-	}
-	top.evlist->session = top.session;
-
 	if (!evlist__needs_bpf_sb_event(top.evlist))
 		top.record_opts.no_bpf_event = true;
 
diff --git a/tools/perf/tests/hists_cumulate.c b/tools/perf/tests/hists_cumulate.c
index 1e0f5a310fd5..3eb9ef8d7ec6 100644
--- a/tools/perf/tests/hists_cumulate.c
+++ b/tools/perf/tests/hists_cumulate.c
@@ -295,7 +295,7 @@ static int test1(struct evsel *evsel, struct machine *machine)
 	symbol_conf.cumulate_callchain = false;
 	evsel__reset_sample_bit(evsel, CALLCHAIN);
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 	callchain_register_param(&callchain_param);
 
 	err = add_hist_entries(hists, machine);
@@ -442,7 +442,7 @@ static int test2(struct evsel *evsel, struct machine *machine)
 	symbol_conf.cumulate_callchain = false;
 	evsel__set_sample_bit(evsel, CALLCHAIN);
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 	callchain_register_param(&callchain_param);
 
 	err = add_hist_entries(hists, machine);
@@ -500,7 +500,7 @@ static int test3(struct evsel *evsel, struct machine *machine)
 	symbol_conf.cumulate_callchain = true;
 	evsel__reset_sample_bit(evsel, CALLCHAIN);
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 	callchain_register_param(&callchain_param);
 
 	err = add_hist_entries(hists, machine);
@@ -684,7 +684,7 @@ static int test4(struct evsel *evsel, struct machine *machine)
 	symbol_conf.cumulate_callchain = true;
 	evsel__set_sample_bit(evsel, CALLCHAIN);
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 
 	callchain_param = callchain_param_default;
 	callchain_register_param(&callchain_param);
diff --git a/tools/perf/tests/hists_filter.c b/tools/perf/tests/hists_filter.c
index 4b2e4f2fbe48..1cebd20cc91c 100644
--- a/tools/perf/tests/hists_filter.c
+++ b/tools/perf/tests/hists_filter.c
@@ -131,10 +131,6 @@ static int test__hists_filter(struct test_suite *test __maybe_unused, int subtes
 		goto out;
 	err = TEST_FAIL;
 
-	/* default sort order (comm,dso,sym) will be used */
-	if (setup_sorting(NULL) < 0)
-		goto out;
-
 	machines__init(&machines);
 
 	/* setup threads/dso/map/symbols also */
@@ -145,6 +141,10 @@ static int test__hists_filter(struct test_suite *test __maybe_unused, int subtes
 	if (verbose > 1)
 		machine__fprintf(machine, stderr);
 
+	/* default sort order (comm,dso,sym) will be used */
+	if (setup_sorting(evlist, machine->env) < 0)
+		goto out;
+
 	/* process sample events */
 	err = add_hist_entries(evlist, machine);
 	if (err < 0)
diff --git a/tools/perf/tests/hists_link.c b/tools/perf/tests/hists_link.c
index 5b6f1e883466..996f5f0b3bd1 100644
--- a/tools/perf/tests/hists_link.c
+++ b/tools/perf/tests/hists_link.c
@@ -303,10 +303,6 @@ static int test__hists_link(struct test_suite *test __maybe_unused, int subtest
 		goto out;
 
 	err = TEST_FAIL;
-	/* default sort order (comm,dso,sym) will be used */
-	if (setup_sorting(NULL) < 0)
-		goto out;
-
 	machines__init(&machines);
 
 	/* setup threads/dso/map/symbols also */
@@ -317,6 +313,10 @@ static int test__hists_link(struct test_suite *test __maybe_unused, int subtest
 	if (verbose > 1)
 		machine__fprintf(machine, stderr);
 
+	/* default sort order (comm,dso,sym) will be used */
+	if (setup_sorting(evlist, machine->env) < 0)
+		goto out;
+
 	/* process sample events */
 	err = add_hist_entries(evlist, machine);
 	if (err < 0)
diff --git a/tools/perf/tests/hists_output.c b/tools/perf/tests/hists_output.c
index 33b5cc8352a7..ee5ec8bda60e 100644
--- a/tools/perf/tests/hists_output.c
+++ b/tools/perf/tests/hists_output.c
@@ -146,7 +146,7 @@ static int test1(struct evsel *evsel, struct machine *machine)
 	field_order = NULL;
 	sort_order = NULL; /* equivalent to sort_order = "comm,dso,sym" */
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 
 	/*
 	 * expected output:
@@ -248,7 +248,7 @@ static int test2(struct evsel *evsel, struct machine *machine)
 	field_order = "overhead,cpu";
 	sort_order = "pid";
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 
 	/*
 	 * expected output:
@@ -304,7 +304,7 @@ static int test3(struct evsel *evsel, struct machine *machine)
 	field_order = "comm,overhead,dso";
 	sort_order = NULL;
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 
 	/*
 	 * expected output:
@@ -378,7 +378,7 @@ static int test4(struct evsel *evsel, struct machine *machine)
 	field_order = "dso,sym,comm,overhead,dso";
 	sort_order = "sym";
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 
 	/*
 	 * expected output:
@@ -480,7 +480,7 @@ static int test5(struct evsel *evsel, struct machine *machine)
 	field_order = "cpu,pid,comm,dso,sym";
 	sort_order = "dso,pid";
 
-	setup_sorting(NULL);
+	setup_sorting(/*evlist=*/NULL, machine->env);
 
 	/*
 	 * expected output:
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index b13385a6068b..e40d16d3246c 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -391,9 +391,6 @@ extern unsigned int proc_map_timeout;
 #define PAGE_SIZE_NAME_LEN	32
 char *get_page_size_name(u64 size, char *str);
 
-const char *arch_perf_header_entry(const char *se_header);
-int arch_support_sort_key(const char *sort_key);
-
 static inline bool perf_event_header__cpumode_is_guest(u8 cpumode)
 {
 	return cpumode == PERF_RECORD_MISC_GUEST_KERNEL ||
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 0ba2ce1b1c07..f3a565b0e230 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -2530,19 +2530,44 @@ struct sort_dimension {
 	int			taken;
 };
 
-int __weak arch_support_sort_key(const char *sort_key __maybe_unused)
+static int arch_support_sort_key(const char *sort_key, struct perf_env *env)
 {
+	const char *arch = perf_env__arch(env);
+
+	if (!strcmp("x86", arch) || !strcmp("powerpc", arch)) {
+		if (!strcmp(sort_key, "p_stage_cyc"))
+			return 1;
+		if (!strcmp(sort_key, "local_p_stage_cyc"))
+			return 1;
+	}
 	return 0;
 }
 
-const char * __weak arch_perf_header_entry(const char *se_header)
-{
+static const char *arch_perf_header_entry(const char *se_header, struct perf_env *env)
+{
+	const char *arch = perf_env__arch(env);
+
+	if (!strcmp("x86", arch)) {
+		if (!strcmp(se_header, "Local Pipeline Stage Cycle"))
+			return "Local Retire Latency";
+		else if (!strcmp(se_header, "Pipeline Stage Cycle"))
+			return "Retire Latency";
+	} else if (!strcmp("powerpc", arch)) {
+		if (!strcmp(se_header, "Local INSTR Latency"))
+			return "Finish Cyc";
+		else if (!strcmp(se_header, "INSTR Latency"))
+			return "Global Finish_cyc";
+		else if (!strcmp(se_header, "Local Pipeline Stage Cycle"))
+			return "Dispatch Cyc";
+		else if (!strcmp(se_header, "Pipeline Stage Cycle"))
+			return "Global Dispatch_cyc";
+	}
 	return se_header;
 }
 
-static void sort_dimension_add_dynamic_header(struct sort_dimension *sd)
+static void sort_dimension_add_dynamic_header(struct sort_dimension *sd, struct perf_env *env)
 {
-	sd->entry->se_header = arch_perf_header_entry(sd->entry->se_header);
+	sd->entry->se_header = arch_perf_header_entry(sd->entry->se_header, env);
 }
 
 #define DIM(d, n, func) [d] = { .name = n, .entry = &(func) }
@@ -3594,7 +3619,7 @@ int hpp_dimension__add_output(unsigned col, bool implicit)
 }
 
 int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
-			struct evlist *evlist,
+			struct evlist *evlist, struct perf_env *env,
 			int level)
 {
 	unsigned int i, j;
@@ -3607,7 +3632,7 @@ int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
 	 */
 	for (j = 0; j < ARRAY_SIZE(arch_specific_sort_keys); j++) {
 		if (!strcmp(arch_specific_sort_keys[j], tok) &&
-				!arch_support_sort_key(tok)) {
+		    !arch_support_sort_key(tok, env)) {
 			return 0;
 		}
 	}
@@ -3620,7 +3645,7 @@ int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
 
 		for (j = 0; j < ARRAY_SIZE(dynamic_headers); j++) {
 			if (sd->name && !strcmp(dynamic_headers[j], sd->name))
-				sort_dimension_add_dynamic_header(sd);
+				sort_dimension_add_dynamic_header(sd, env);
 		}
 
 		if (sd->entry == &sort_parent && parent_pattern) {
@@ -3716,13 +3741,13 @@ int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
 }
 
 /* This should match with sort_dimension__add() above */
-static bool is_hpp_sort_key(const char *key)
+static bool is_hpp_sort_key(const char *key, struct perf_env *env)
 {
 	unsigned i;
 
 	for (i = 0; i < ARRAY_SIZE(arch_specific_sort_keys); i++) {
 		if (!strcmp(arch_specific_sort_keys[i], key) &&
-		    !arch_support_sort_key(key)) {
+		    !arch_support_sort_key(key, env)) {
 			return false;
 		}
 	}
@@ -3744,7 +3769,7 @@ static bool is_hpp_sort_key(const char *key)
 }
 
 static int setup_sort_list(struct perf_hpp_list *list, char *str,
-			   struct evlist *evlist)
+			   struct evlist *evlist, struct perf_env *env)
 {
 	char *tmp, *tok;
 	int ret = 0;
@@ -3773,7 +3798,7 @@ static int setup_sort_list(struct perf_hpp_list *list, char *str,
 		}
 
 		if (*tok) {
-			if (is_hpp_sort_key(tok)) {
+			if (is_hpp_sort_key(tok, env)) {
 				/* keep output (hpp) sort keys in the same level */
 				if (prev_was_hpp) {
 					bool next_same = (level == next_level);
@@ -3786,7 +3811,7 @@ static int setup_sort_list(struct perf_hpp_list *list, char *str,
 				prev_was_hpp = false;
 			}
 
-			ret = sort_dimension__add(list, tok, evlist, level);
+			ret = sort_dimension__add(list, tok, evlist, env, level);
 			if (ret == -EINVAL) {
 				if (!cacheline_size() && !strncasecmp(tok, "dcacheline", strlen(tok)))
 					ui__error("The \"dcacheline\" --sort key needs to know the cacheline size and it couldn't be determined on this system");
@@ -3915,7 +3940,7 @@ static char *setup_overhead(char *keys)
 	return keys;
 }
 
-static int __setup_sorting(struct evlist *evlist)
+static int __setup_sorting(struct evlist *evlist, struct perf_env *env)
 {
 	char *str;
 	const char *sort_keys;
@@ -3955,7 +3980,7 @@ static int __setup_sorting(struct evlist *evlist)
 		}
 	}
 
-	ret = setup_sort_list(&perf_hpp_list, str, evlist);
+	ret = setup_sort_list(&perf_hpp_list, str, evlist, env);
 
 	free(str);
 	return ret;
@@ -4191,16 +4216,16 @@ out:
 	return ret;
 }
 
-int setup_sorting(struct evlist *evlist)
+int setup_sorting(struct evlist *evlist, struct perf_env *env)
 {
 	int err;
 
-	err = __setup_sorting(evlist);
+	err = __setup_sorting(evlist, env);
 	if (err < 0)
 		return err;
 
 	if (parent_pattern != default_parent_pattern) {
-		err = sort_dimension__add(&perf_hpp_list, "parent", evlist, -1);
+		err = sort_dimension__add(&perf_hpp_list, "parent", evlist, env, -1);
 		if (err < 0)
 			return err;
 	}
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
index a742ab7f3c67..d7787958e06b 100644
--- a/tools/perf/util/sort.h
+++ b/tools/perf/util/sort.h
@@ -6,6 +6,7 @@
 #include "hist.h"
 
 struct option;
+struct perf_env;
 
 extern regex_t parent_regex;
 extern const char *sort_order;
@@ -130,7 +131,7 @@ extern struct sort_entry sort_thread;
 
 struct evlist;
 struct tep_handle;
-int setup_sorting(struct evlist *evlist);
+int setup_sorting(struct evlist *evlist, struct perf_env *env);
 int setup_output_field(void);
 void reset_output_field(void);
 void sort__setup_elide(FILE *fp);
@@ -145,7 +146,7 @@ bool is_strict_order(const char *order);
 int hpp_dimension__add_output(unsigned col, bool implicit);
 void reset_dimensions(void);
 int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
-			struct evlist *evlist,
+			struct evlist *evlist, struct perf_env *env,
 			int level);
 int output_field_add(struct perf_hpp_list *list, const char *tok, int *level);
 int64_t
-- 
cgit v1.2.3