Skip to contentSkip to conversation

perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu

PATCH Applied Ian Rogers 3 participants 9 messages May 5, 2024 – May 13, 2024 perf events

v3 of 6 Applied Latest is v6 →

lore
Share & tools
lore

Ian Rogers

author reviewer
#0 lore raw

v3 cover perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix. Current PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Add support for hexadecimal suffixes and add tests.

v3. Rebase and move tests from pmus.c to the existing pmu.c.

Ian Rogers (2):
perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu perf tests: Add some pmu core functionality tests

 tools/perf/tests/pmu.c | 95 ++++++++++++++++++++++++++++++++++++++++++
 tools/perf/util/pmu.c  |  4 +-
 tools/perf/util/pmus.c | 53 ++++++++++++-----------
 tools/perf/util/pmus.h |  7 +++-
 4 files changed, 131 insertions(+), 28 deletions(-)
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
author reviewer
#1 lore raw

PATCH v3 1/2 perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix while the previous PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Because of the previous assumption it isn't possible to wildcard the mrvl_ddr_pmu.

Modify pmu_name_len_no_suffix but also remove the suffix number out argument, this is because we don't know if a suffix number of say 10 is in hexadecimal or decimal. As the only use of the suffix number is in comparisons, it is safe there to compare the values as hexadecimal.

---
 tools/perf/util/pmu.c  |  4 ++--
 tools/perf/util/pmus.c | 53 ++++++++++++++++++++++--------------------
 tools/perf/util/pmus.h |  7 +++++-
 3 files changed, 36 insertions(+), 28 deletions(-)
tools/perf/util/pmu.c +2 -2
index b3b072feef02..32e5a73273d7 100644
@@ -1761,7 +1761,7 @@ static char *format_alias(char *buf, int len, const struct perf_pmu *pmu,
 {
 	struct parse_events_term *term;
 	int pmu_name_len = skip_duplicate_pmus
-		? pmu_name_len_no_suffix(pmu->name, /*num=*/NULL)
+		? pmu_name_len_no_suffix(pmu->name)
 		: (int)strlen(pmu->name);
 	int used = snprintf(buf, len, "%.*s/%s", pmu_name_len, pmu->name, alias->name);
 
@@ -1805,7 +1805,7 @@ int perf_pmu__for_each_event(struct perf_pmu *pmu, bool skip_duplicate_pmus,
 
 		info.pmu_name = event->pmu_name ?: pmu->name;
 		pmu_name_len = skip_duplicate_pmus
-			? pmu_name_len_no_suffix(info.pmu_name, /*num=*/NULL)
+			? pmu_name_len_no_suffix(info.pmu_name)
 			: (int)strlen(info.pmu_name);
 		info.alias = NULL;
 		if (event->desc) {
tools/perf/util/pmus.c +28 -25

104 lines not on this page · raw message

tools/perf/util/pmus.h +6 -1
index 9d4ded80b8e9..69d64fbd4259 100644
@@ -2,10 +2,15 @@
 #ifndef __PMUS_H
 #define __PMUS_H
 
+#include <stdbool.h>
+#include <linux/list.h>
+
 struct perf_pmu;
 struct print_callbacks;
 
-int pmu_name_len_no_suffix(const char *str, unsigned long *num);
+int pmu_name_len_no_suffix(const char *str);
+/* Exposed for testing only. */
+int pmu_name_cmp(const char *lhs_pmu_name, const char *rhs_pmu_name);
 
 void perf_pmus__destroy(void);
 
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
#2 lore raw
Quoted 9 lines · Ian Rogers

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix while the previous PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Because of the previous assumption it isn't possible to wildcard the mrvl_ddr_pmu.

Modify pmu_name_len_no_suffix but also remove the suffix number out argument, this is because we don't know if a suffix number of say 10 is in hexadecimal or decimal. As the only use of the suffix number is in comparisons, it is safe there to compare the values as hexadecimal.

Removed these patches due to:

util/pmus.c: In function ‘pmu_name_cmp’:
util/pmus.c:67:19: error: ‘strncmp’ specified bound [18446744071562067968, 18446744073709551615] exceeds maximum object size 9223372036854775807 [-Werror=stringop-overread]

   67 |         int ret = strncmp(lhs_pmu_name, rhs_pmu_name,
      |                   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   68 |                         lhs_pmu_name_len < rhs_pmu_name_len ? lhs_pmu_name_len : rhs_pmu_name_len);
      |                         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

cc1: all warnings being treated as errors
make[4]: *** [/tmp/perf-6.9.0-rc5/tools/build/Makefile.build:105: util/pmus.o] Error 1

Noticed while doing a 'make -C tools/perf build-test', the tarbz test.

acme@x1:~/git/perf-tools-next$ head /etc/os-release 
NAME="Fedora Linux"
VERSION="39 (Workstation Edition)"
ID=fedora
VERSION_ID=39
VERSION_CODENAME=""
PLATFORM_ID="platform:f39"
PRETTY_NAME="Fedora Linux 39 (Workstation Edition)"
ANSI_COLOR="0;38;2;60;110;180"
LOGO=fedora-logo-icon
CPE_NAME="cpe:/o:fedoraproject:fedora:39"
acme@x1:~/git/perf-tools-next$

- Arnaldo

Quoted 165 lines
---
 tools/perf/util/pmu.c  |  4 ++--
 tools/perf/util/pmus.c | 53 ++++++++++++++++++++++--------------------
 tools/perf/util/pmus.h |  7 +++++-
 3 files changed, 36 insertions(+), 28 deletions(-)
tools/perf/util/pmu.c +2 -2
index b3b072feef02..32e5a73273d7 100644
@@ -1761,7 +1761,7 @@ static char *format_alias(char *buf, int len, const struct perf_pmu *pmu,
 {
 	struct parse_events_term *term;
 	int pmu_name_len = skip_duplicate_pmus
-		? pmu_name_len_no_suffix(pmu->name, /*num=*/NULL)
+		? pmu_name_len_no_suffix(pmu->name)
 		: (int)strlen(pmu->name);
 	int used = snprintf(buf, len, "%.*s/%s", pmu_name_len, pmu->name, alias->name);
 
@@ -1805,7 +1805,7 @@ int perf_pmu__for_each_event(struct perf_pmu *pmu, bool skip_duplicate_pmus,
 
 		info.pmu_name = event->pmu_name ?: pmu->name;
 		pmu_name_len = skip_duplicate_pmus
-			? pmu_name_len_no_suffix(info.pmu_name, /*num=*/NULL)
+			? pmu_name_len_no_suffix(info.pmu_name)
 			: (int)strlen(info.pmu_name);
 		info.alias = NULL;
 		if (event->desc) {
tools/perf/util/pmus.c +28 -25
index b9b4c5eb5002..28cfa7adbfd7 100644
@@ -40,31 +40,44 @@ static bool read_sysfs_all_pmus;
 
 static void pmu_read_sysfs(bool core_only);
 
-int pmu_name_len_no_suffix(const char *str, unsigned long *num)
+int pmu_name_len_no_suffix(const char *str)
 {
 	int orig_len, len;
 
 	orig_len = len = strlen(str);
 
-	/* Non-uncore PMUs have their full length, for example, i915. */
-	if (!strstarts(str, "uncore_"))
-		return len;
-
 	/*
 	 * Count trailing digits and '_', if '_{num}' suffix isn't present use
 	 * the full length.
 	 */
-	while (len > 0 && isdigit(str[len - 1]))
+	while (len > 0 && isxdigit(str[len - 1]))
 		len--;
 
-	if (len > 0 && len != orig_len && str[len - 1] == '_') {
-		if (num)
-			*num = strtoul(&str[len], NULL, 10);
+	if (len > 0 && len != orig_len && str[len - 1] == '_')
 		return len - 1;
-	}
+
 	return orig_len;
 }
 
+int pmu_name_cmp(const char *lhs_pmu_name, const char *rhs_pmu_name)
+{
+	unsigned long lhs_num = 0, rhs_num = 0;
+	int lhs_pmu_name_len = pmu_name_len_no_suffix(lhs_pmu_name);
+	int rhs_pmu_name_len = pmu_name_len_no_suffix(rhs_pmu_name);
+	int ret = strncmp(lhs_pmu_name, rhs_pmu_name,
+			lhs_pmu_name_len < rhs_pmu_name_len ? lhs_pmu_name_len : rhs_pmu_name_len);
+
+	if (lhs_pmu_name_len != rhs_pmu_name_len || ret != 0 || lhs_pmu_name_len == 0)
+		return ret;
+
+	if (lhs_pmu_name_len + 1 < (int)strlen(lhs_pmu_name))
+		lhs_num = strtoul(&lhs_pmu_name[lhs_pmu_name_len + 1], NULL, 16);
+	if (rhs_pmu_name_len + 1 < (int)strlen(rhs_pmu_name))
+		rhs_num = strtoul(&rhs_pmu_name[rhs_pmu_name_len + 1], NULL, 16);
+
+	return lhs_num < rhs_num ? -1 : (lhs_num > rhs_num ? 1 : 0);
+}
+
 void perf_pmus__destroy(void)
 {
 	struct perf_pmu *pmu, *tmp;
@@ -167,20 +180,10 @@ static struct perf_pmu *perf_pmu__find2(int dirfd, const char *name)
 static int pmus_cmp(void *priv __maybe_unused,
 		    const struct list_head *lhs, const struct list_head *rhs)
 {
-	unsigned long lhs_num = 0, rhs_num = 0;
 	struct perf_pmu *lhs_pmu = container_of(lhs, struct perf_pmu, list);
 	struct perf_pmu *rhs_pmu = container_of(rhs, struct perf_pmu, list);
-	const char *lhs_pmu_name = lhs_pmu->name ?: "";
-	const char *rhs_pmu_name = rhs_pmu->name ?: "";
-	int lhs_pmu_name_len = pmu_name_len_no_suffix(lhs_pmu_name, &lhs_num);
-	int rhs_pmu_name_len = pmu_name_len_no_suffix(rhs_pmu_name, &rhs_num);
-	int ret = strncmp(lhs_pmu_name, rhs_pmu_name,
-			lhs_pmu_name_len < rhs_pmu_name_len ? lhs_pmu_name_len : rhs_pmu_name_len);
 
-	if (lhs_pmu_name_len != rhs_pmu_name_len || ret != 0 || lhs_pmu_name_len == 0)
-		return ret;
-
-	return lhs_num < rhs_num ? -1 : (lhs_num > rhs_num ? 1 : 0);
+	return pmu_name_cmp(lhs_pmu->name ?: "", rhs_pmu->name ?: "");
 }
 
 /* Add all pmus in sysfs to pmu list: */
@@ -300,11 +303,11 @@ static struct perf_pmu *perf_pmus__scan_skip_duplicates(struct perf_pmu *pmu)
 		pmu_read_sysfs(/*core_only=*/false);
 		pmu = list_prepare_entry(pmu, &core_pmus, list);
 	} else
-		last_pmu_name_len = pmu_name_len_no_suffix(pmu->name ?: "", NULL);
+		last_pmu_name_len = pmu_name_len_no_suffix(pmu->name ?: "");
 
 	if (use_core_pmus) {
 		list_for_each_entry_continue(pmu, &core_pmus, list) {
-			int pmu_name_len = pmu_name_len_no_suffix(pmu->name ?: "", /*num=*/NULL);
+			int pmu_name_len = pmu_name_len_no_suffix(pmu->name ?: "");
 
 			if (last_pmu_name_len == pmu_name_len &&
 			    !strncmp(last_pmu_name, pmu->name ?: "", pmu_name_len))
@@ -316,7 +319,7 @@ static struct perf_pmu *perf_pmus__scan_skip_duplicates(struct perf_pmu *pmu)
 		pmu = list_prepare_entry(pmu, &other_pmus, list);
 	}
 	list_for_each_entry_continue(pmu, &other_pmus, list) {
-		int pmu_name_len = pmu_name_len_no_suffix(pmu->name ?: "", /*num=*/NULL);
+		int pmu_name_len = pmu_name_len_no_suffix(pmu->name ?: "");
 
 		if (last_pmu_name_len == pmu_name_len &&
 		    !strncmp(last_pmu_name, pmu->name ?: "", pmu_name_len))
@@ -566,7 +569,7 @@ void perf_pmus__print_raw_pmu_events(const struct print_callbacks *print_cb, voi
 			.long_string = STRBUF_INIT,
 			.num_formats = 0,
 		};
-		int len = pmu_name_len_no_suffix(pmu->name, /*num=*/NULL);
+		int len = pmu_name_len_no_suffix(pmu->name);
 		const char *desc = "(see 'man perf-list' or 'man perf-record' on how to encode it)";
 
 		if (!pmu->is_core)
tools/perf/util/pmus.h +6 -1
index 9d4ded80b8e9..69d64fbd4259 100644
@@ -2,10 +2,15 @@
 #ifndef __PMUS_H
 #define __PMUS_H
 
+#include <stdbool.h>
+#include <linux/list.h>
+
 struct perf_pmu;
 struct print_callbacks;
 
-int pmu_name_len_no_suffix(const char *str, unsigned long *num);
+int pmu_name_len_no_suffix(const char *str);
+/* Exposed for testing only. */
+int pmu_name_cmp(const char *lhs_pmu_name, const char *rhs_pmu_name);
 
 void perf_pmus__destroy(void);
 
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
author reviewer
#3 lore raw

PATCH v3 2/2 perf tests: Add some pmu core functionality tests

In reply to Ian Rogers

Test behavior of PMU names and comparisons wrt suffixes using Intel uncore_cha and marvell mrvl_ddr_pmu as examples.

---
 tools/perf/tests/pmu.c | 95 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 95 insertions(+)
tools/perf/tests/pmu.c +95 -0
index 06cc0e46cb28..9dc77f29f723 100644
@@ -3,6 +3,7 @@
 #include "evsel.h"
 #include "parse-events.h"
 #include "pmu.h"
+#include "pmus.h"
 #include "tests.h"
 #include "debug.h"
 #include "fncache.h"
@@ -340,10 +341,104 @@ static int test__pmu_event_names(struct test_suite *test __maybe_unused,
 	return ret;
 }
 
+static const char * const uncore_chas[] = {
+	"uncore_cha_0",
+	"uncore_cha_1",
+	"uncore_cha_2",
+	"uncore_cha_3",
+	"uncore_cha_4",
+	"uncore_cha_5",
+	"uncore_cha_6",
+	"uncore_cha_7",
+	"uncore_cha_8",
+	"uncore_cha_9",
+	"uncore_cha_10",
+	"uncore_cha_11",
+	"uncore_cha_12",
+	"uncore_cha_13",
+	"uncore_cha_14",
+	"uncore_cha_15",
+	"uncore_cha_16",
+	"uncore_cha_17",
+	"uncore_cha_18",
+	"uncore_cha_19",
+	"uncore_cha_20",
+	"uncore_cha_21",
+	"uncore_cha_22",
+	"uncore_cha_23",
+	"uncore_cha_24",
+	"uncore_cha_25",
+	"uncore_cha_26",
+	"uncore_cha_27",
+	"uncore_cha_28",
+	"uncore_cha_29",
+	"uncore_cha_30",
+	"uncore_cha_31",
+};
+
+static const char * const mrvl_ddrs[] = {
+	"mrvl_ddr_pmu_87e1b0000000",
+	"mrvl_ddr_pmu_87e1b1000000",
+	"mrvl_ddr_pmu_87e1b2000000",
+	"mrvl_ddr_pmu_87e1b3000000",
+	"mrvl_ddr_pmu_87e1b4000000",
+	"mrvl_ddr_pmu_87e1b5000000",
+	"mrvl_ddr_pmu_87e1b6000000",
+	"mrvl_ddr_pmu_87e1b7000000",
+	"mrvl_ddr_pmu_87e1b8000000",
+	"mrvl_ddr_pmu_87e1b9000000",
+	"mrvl_ddr_pmu_87e1ba000000",
+	"mrvl_ddr_pmu_87e1bb000000",
+	"mrvl_ddr_pmu_87e1bc000000",
+	"mrvl_ddr_pmu_87e1bd000000",
+	"mrvl_ddr_pmu_87e1be000000",
+	"mrvl_ddr_pmu_87e1bf000000",
+};
+
+static int test__name_len(struct test_suite *test __maybe_unused, int subtest __maybe_unused)
+{
+	TEST_ASSERT_EQUAL("cpu", pmu_name_len_no_suffix("cpu"), (int)strlen("cpu"));
+	TEST_ASSERT_EQUAL("i915", pmu_name_len_no_suffix("i915"), (int)strlen("i915"));
+	for (size_t i = 0; i < ARRAY_SIZE(uncore_chas); i++) {
+		TEST_ASSERT_EQUAL("Strips uncore_cha suffix",
+				pmu_name_len_no_suffix(uncore_chas[i]),
+				(int)strlen("uncore_cha"));
+	}
+	for (size_t i = 0; i < ARRAY_SIZE(mrvl_ddrs); i++) {
+		TEST_ASSERT_EQUAL("Strips mrvl_ddr_pmu suffix",
+				pmu_name_len_no_suffix(mrvl_ddrs[i]),
+				(int)strlen("mrvl_ddr_pmu"));
+	}
+	return TEST_OK;
+}
+
+static int test__name_cmp(struct test_suite *test __maybe_unused, int subtest __maybe_unused)
+{
+	TEST_ASSERT_EQUAL("cpu", pmu_name_cmp("cpu", "cpu"), 0);
+	TEST_ASSERT_EQUAL("i915", pmu_name_cmp("i915", "i915"), 0);
+	TEST_ASSERT_VAL("i915", pmu_name_cmp("cpu", "i915") < 0);
+	TEST_ASSERT_VAL("i915", pmu_name_cmp("i915", "cpu") > 0);
+	for (size_t i = 1; i < ARRAY_SIZE(uncore_chas); i++) {
+		TEST_ASSERT_VAL("uncore_cha suffixes ordered lt",
+				pmu_name_cmp(uncore_chas[i-1], uncore_chas[i]) < 0);
+		TEST_ASSERT_VAL("uncore_cha suffixes ordered gt",
+				pmu_name_cmp(uncore_chas[i], uncore_chas[i-1]) > 0);
+	}
+	for (size_t i = 1; i < ARRAY_SIZE(mrvl_ddrs); i++) {
+		TEST_ASSERT_VAL("mrvl_ddr_pmu suffixes ordered lt",
+				pmu_name_cmp(mrvl_ddrs[i-1], mrvl_ddrs[i]) < 0);
+		TEST_ASSERT_VAL("mrvl_ddr_pmu suffixes ordered gt",
+				pmu_name_cmp(mrvl_ddrs[i], mrvl_ddrs[i-1]) > 0);
+	}
+	return TEST_OK;
+}
+
 static struct test_case tests__pmu[] = {
 	TEST_CASE("Parsing with PMU format directory", pmu_format),
 	TEST_CASE("Parsing with PMU event", pmu_events),
 	TEST_CASE("PMU event names", pmu_event_names),
+	TEST_CASE("PMU name combining", name_len),
+	TEST_CASE("PMU name comparison", name_cmp),
 	{	.name = NULL, }
 };
 
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
author reviewer
#4 lore raw
In reply to Ian Rogers
Quoted 11 lines · Ian Rogers

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix. Current PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Add support for hexadecimal suffixes and add tests.

v3. Rebase and move tests from pmus.c to the existing pmu.c.

Ian Rogers (2):
perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu perf tests: Add some pmu core functionality tests

Hi, these patches have been hanging around since March [1], it would be nice to either be landing them or getting feedback on what to fix.

Thanks,
Ian

[1] https://lore.kernel.org/lkml/[email protected]/

Quoted 9 lines
 tools/perf/tests/pmu.c | 95 ++++++++++++++++++++++++++++++++++++++++++
 tools/perf/util/pmu.c  |  4 +-
 tools/perf/util/pmus.c | 53 ++++++++++++-----------
 tools/perf/util/pmus.h |  7 +++-
 4 files changed, 131 insertions(+), 28 deletions(-)

--
2.45.0.rc1.225.g2a3ae87e7f-goog

author reviewer
#5 lore raw
Quoted 14 lines · Ian Rogers
Quoted 11 lines · Ian Rogers

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix. Current PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Add support for hexadecimal suffixes and add tests.

v3. Rebase and move tests from pmus.c to the existing pmu.c.

Ian Rogers (2):
perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu perf tests: Add some pmu core functionality tests

Quoted 7 lines

Hi, these patches have been hanging around since March [1], it would be nice to either be landing them or getting feedback on what to fix.

Thanks,
Ian

[1] https://lore.kernel.org/lkml/[email protected]/

+Tuan Phan, Robin Murphy

Here is another PMU with the same suffix convention/issue:
https://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/drivers/perf/arm_dmc620_pmu.c?h=perf-tools-next#n706

Thanks,
Ian

Quoted 9 lines
Quoted 9 lines
 tools/perf/tests/pmu.c | 95 ++++++++++++++++++++++++++++++++++++++++++
 tools/perf/util/pmu.c  |  4 +-
 tools/perf/util/pmus.c | 53 ++++++++++++-----------
 tools/perf/util/pmus.h |  7 +++-
 4 files changed, 131 insertions(+), 28 deletions(-)

--
2.45.0.rc1.225.g2a3ae87e7f-goog

#6 lore raw
Quoted 21 lines · Ian Rogers
Quoted 20 lines · Ian Rogers
Quoted 11 lines · Ian Rogers

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix. Current PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Add support for hexadecimal suffixes and add tests.

v3. Rebase and move tests from pmus.c to the existing pmu.c.

Ian Rogers (2):
perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu perf tests: Add some pmu core functionality tests

Hi, these patches have been hanging around since March [1], it would be nice to either be landing them or getting feedback on what to fix.

Thanks,
Ian

Quoted 6 lines
Quoted 1 line

[1] https://lore.kernel.org/lkml/[email protected]/

+Tuan Phan, Robin Murphy

Here is another PMU with the same suffix convention/issue:
https://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/drivers/perf/arm_dmc620_pmu.c?h=perf-tools-next#n706

There are at least one or two more as well - certainly arm_smmuv3_pmu which I think may have been where this pattern first started. Now that we've finally done the right thing with the parent mechanism to provide a user-visible relationship of PMU instances to their corresponding Devicetree/ACPI devices, hopefully we can discourage any further use of this rather clunky trick of using the MMIO address as an identifier in the PMU name. However there's then also stuff like dwc_pcie_pmu which encodes a PCI ID as a hex suffix, so understanding hex suffixes in general might still be a reasonable idea for the tool, if the alternative would be maintaining a list of specific prefixes (even if that would be hoped to remain fairly small).

Thanks,
Robin.

author reviewer
#7 lore raw
Quoted 30 lines · Robin Murphy
Quoted 21 lines · Ian Rogers
Quoted 20 lines · Ian Rogers
Quoted 11 lines · Ian Rogers

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix. Current PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Add support for hexadecimal suffixes and add tests.

v3. Rebase and move tests from pmus.c to the existing pmu.c.

Ian Rogers (2):
perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu perf tests: Add some pmu core functionality tests

Hi, these patches have been hanging around since March [1], it would be nice to either be landing them or getting feedback on what to fix.

Thanks,
Ian

Quoted 6 lines
Quoted 1 line

[1] https://lore.kernel.org/lkml/[email protected]/

+Tuan Phan, Robin Murphy

Here is another PMU with the same suffix convention/issue:
https://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/drivers/perf/arm_dmc620_pmu.c?h=perf-tools-next#n706

Quoted 11 lines

There are at least one or two more as well - certainly arm_smmuv3_pmu which I think may have been where this pattern first started. Now that we've finally done the right thing with the parent mechanism to provide a user-visible relationship of PMU instances to their corresponding Devicetree/ACPI devices, hopefully we can discourage any further use of this rather clunky trick of using the MMIO address as an identifier in the PMU name. However there's then also stuff like dwc_pcie_pmu which encodes a PCI ID as a hex suffix, so understanding hex suffixes in general might still be a reasonable idea for the tool, if the alternative would be maintaining a list of specific prefixes (even if that would be hoped to remain fairly small).

Thanks Robin, the problem that Kan identified was that IBM s390 have a PMU called cpum_cf. So the _cf is a valid hex suffix. Maybe there needs to be a minimum hex suffix length to deal with this. We made the 0x optional in perf "raw" events like perf stat -e 'r12abcd'... and now we have the problem that the word read is both quite a desirable perf event name and also a valid hex encoding. We prefer the event name if it exists, but we have to make the code more complex to deal with this. It would be nice if we were capturing the conventions in:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/ABI/testing
Like with the sysfs-bus-event_source-* files. I'll do a v4 of these changes and see if I can do something to document the suffixes, would be great to get an Acked-by/Tested-by.

Thanks,
Ian

Quoted 2 lines

Thanks,
Robin.

maintainer Applied
#8 lore raw
In reply to Ian Rogers
Quoted 7 lines · Ian Rogers
Quoted 6 lines · Ian Rogers

The mrvl_ddr_pmu is uncore and has a hexadecimal address suffix. Current PMU sorting/merging code assumes uncore PMU names start with uncore_ and have a decimal suffix. Add support for hexadecimal suffixes and add tests.

Quoted 8 lines
Quoted 5 lines

v3. Rebase and move tests from pmus.c to the existing pmu.c.

Ian Rogers (2):
perf pmus: Sort/merge/aggregate PMUs like mrvl_ddr_pmu perf tests: Add some pmu core functionality tests

Hi, these patches have been hanging around since March [1], it would be nice to either be landing them or getting feedback on what to fix.

Thanks, applied to perf-tools-next,

- Arnaldo

Quoted 14 lines

Thanks,
Ian

[1] https://lore.kernel.org/lkml/[email protected]/

Quoted 9 lines
 tools/perf/tests/pmu.c | 95 ++++++++++++++++++++++++++++++++++++++++++
 tools/perf/util/pmu.c  |  4 +-
 tools/perf/util/pmus.c | 53 ++++++++++++-----------
 tools/perf/util/pmus.h |  7 +++-
 4 files changed, 131 insertions(+), 28 deletions(-)

--
2.45.0.rc1.225.g2a3ae87e7f-goog