RE: [PATCH v14 1/4] lib: add generic support for reading PMU events

[Tomasz Duszynski <tduszynski@marvell.com>]

>> Add support for programming PMU counters and reading their values in
>> runtime bypassing kernel completely.
>>
>> This is especially useful in cases where CPU cores are isolated i.e
>> run dedicated tasks. In such cases one cannot use standard perf
>> utility without sacrificing latency and performance.
>>
>> Signed-off-by: Tomasz Duszynski <tduszynski@marvell.com>
>> ---
>>  MAINTAINERS                            |   5 +
>>  app/test/meson.build                   |   1 +
>>  app/test/test_pmu.c                    |  53 +++
>>  doc/api/doxy-api-index.md              |   3 +-
>>  doc/api/doxy-api.conf.in               |   1 +
>>  doc/guides/prog_guide/profile_app.rst  |  29 ++
>>  doc/guides/rel_notes/release_24_11.rst |   7 +
>>  lib/meson.build                        |   1 +
>>  lib/pmu/meson.build                    |  12 +
>>  lib/pmu/pmu_private.h                  |  32 ++
>>  lib/pmu/rte_pmu.c                      | 463 +++++++++++++++++++++++++
>>  lib/pmu/rte_pmu.h                      | 227 ++++++++++++
>>  lib/pmu/version.map                    |  14 +
>>  13 files changed, 847 insertions(+), 1 deletion(-)  create mode
>> 100644 app/test/test_pmu.c  create mode 100644 lib/pmu/meson.build
>> create mode 100644 lib/pmu/pmu_private.h  create mode 100644
>> lib/pmu/rte_pmu.c  create mode 100644 lib/pmu/rte_pmu.h  create mode
>> 100644 lib/pmu/version.map
>>
>> diff --git a/MAINTAINERS b/MAINTAINERS index 7d171e3d45..fbd46905b6
>> 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -1811,6 +1811,11 @@ M: Nithin Dabilpuram <ndabilpuram@marvell.com>
>>  M: Pavan Nikhilesh <pbhagavatula@marvell.com>
>>  F: lib/node/
>>
>> +PMU - EXPERIMENTAL
>> +M: Tomasz Duszynski <tduszynski@marvell.com>
>> +F: lib/pmu/
>> +F: app/test/test_pmu*
>> +
>>
>>  Test Applications
>>  -----------------
>> diff --git a/app/test/meson.build b/app/test/meson.build index
>> e29258e6ec..45f56d8aae 100644
>> --- a/app/test/meson.build
>> +++ b/app/test/meson.build
>> @@ -139,6 +139,7 @@ source_file_deps = {
>>      'test_pmd_perf.c': ['ethdev', 'net'] + packet_burst_generator_deps,
>>      'test_pmd_ring.c': ['net_ring', 'ethdev', 'bus_vdev'],
>>      'test_pmd_ring_perf.c': ['ethdev', 'net_ring', 'bus_vdev'],
>> +    'test_pmu.c': ['pmu'],
>>      'test_power.c': ['power'],
>>      'test_power_cpufreq.c': ['power'],
>>      'test_power_intel_uncore.c': ['power'], diff --git
>> a/app/test/test_pmu.c b/app/test/test_pmu.c new file mode 100644 index
>> 0000000000..79376ea2e8
>> --- /dev/null
>> +++ b/app/test/test_pmu.c
>> @@ -0,0 +1,53 @@
>> +/* SPDX-License-Identifier: BSD-3-Clause
>> + * Copyright(C) 2024 Marvell International Ltd.
>> + */
>> +
>> +#include <rte_pmu.h>
>> +
>> +#include "test.h"
>> +
>> +static int
>> +test_pmu_read(void)
>> +{
>> +	const char *name = NULL;
>> +	int tries = 10, event;
>> +	uint64_t val = 0;
>> +
>> +	if (name == NULL) {
>> +		printf("PMU not supported on this arch\n");
>> +		return TEST_SKIPPED;
>> +	}
>> +
>> +	if (rte_pmu_init() == -ENOTSUP) {
>> +		printf("pmu_autotest only supported on Linux, skipping test\n");
>> +		return TEST_SKIPPED;
>> +	}
>> +	if (rte_pmu_init() < 0)
>> +		return TEST_SKIPPED;
>> +
>> +	event = rte_pmu_add_event(name);
>> +	while (tries--)
>> +		val += rte_pmu_read(event);
>> +
>> +	rte_pmu_fini();
>> +
>> +	return val ? TEST_SUCCESS : TEST_FAILED; }
>> +
>> +static struct unit_test_suite pmu_tests = {
>> +	.suite_name = "pmu autotest",
>> +	.setup = NULL,
>> +	.teardown = NULL,
>> +	.unit_test_cases = {
>> +		TEST_CASE(test_pmu_read),
>> +		TEST_CASES_END()
>> +	}
>> +};
>> +
>> +static int
>> +test_pmu(void)
>> +{
>> +	return unit_test_suite_runner(&pmu_tests);
>> +}
>> +
>> +REGISTER_FAST_TEST(pmu_autotest, true, true, test_pmu);
>> diff --git a/doc/api/doxy-api-index.md b/doc/api/doxy-api-index.md
>> index f9f0300126..805efc6520 100644
>> --- a/doc/api/doxy-api-index.md
>> +++ b/doc/api/doxy-api-index.md
>> @@ -237,7 +237,8 @@ The public API headers are grouped by topics:
>>    [log](@ref rte_log.h),
>>    [errno](@ref rte_errno.h),
>>    [trace](@ref rte_trace.h),
>> -  [trace_point](@ref rte_trace_point.h)
>> +  [trace_point](@ref rte_trace_point.h),  [pmu](@ref rte_pmu.h)
>>
>>  - **misc**:
>>    [EAL config](@ref rte_eal.h),
>> diff --git a/doc/api/doxy-api.conf.in b/doc/api/doxy-api.conf.in index
>> a8823c046f..658490b6a2 100644
>> --- a/doc/api/doxy-api.conf.in
>> +++ b/doc/api/doxy-api.conf.in
>> @@ -69,6 +69,7 @@ INPUT                   = @TOPDIR@/doc/api/doxy-api-index.md \
>>                            @TOPDIR@/lib/pdcp \
>>                            @TOPDIR@/lib/pdump \
>>                            @TOPDIR@/lib/pipeline \
>> +                          @TOPDIR@/lib/pmu \
>>                            @TOPDIR@/lib/port \
>>                            @TOPDIR@/lib/power \
>>                            @TOPDIR@/lib/ptr_compress \ diff --git
>> a/doc/guides/prog_guide/profile_app.rst
>> b/doc/guides/prog_guide/profile_app.rst
>> index a6b5fb4d5e..854c515a61 100644
>> --- a/doc/guides/prog_guide/profile_app.rst
>> +++ b/doc/guides/prog_guide/profile_app.rst
>> @@ -7,6 +7,35 @@ Profile Your Application  The following sections
>> describe methods of profiling DPDK applications on  different
>> architectures.
>>
>> +Performance counter based profiling
>> +-----------------------------------
>> +
>> +Majority of architectures support some performance monitoring unit (PMU).
>> +Such unit provides programmable counters that monitor specific events.
>> +
>> +Different tools gather that information, like for example perf.
>> +However, in some scenarios when CPU cores are isolated and run
>> +dedicated tasks interrupting those tasks with perf may be undesirable.
>> +
>> +In such cases, an application can use the PMU library to read such events via
>``rte_pmu_read()``.
>> +
>> +By default, userspace applications are not allowed to access PMU
>> +internals. That can be changed by setting
>> +``/sys/kernel/perf_event_paranoid`` to 2 (that should be a default
>> +value anyway) and adding ``CAP_PERFMON`` capability to DPDK
>> +application. Please refer to ``Documentation/admin-guide/perf-security.rst`` under Linux sources
>for more information. Fairly recent kernel, i.e >= 5.9, is advised too.
>> +
>> +As of now implementation imposes certain limitations:
>> +
>> +* Management APIs that normally return a non-negative value will
>> +return error (``-ENOTSUP``) while
>> +  ``rte_pmu_read()`` will return ``UINT64_MAX`` if running under unsupported operating system.
>> +
>> +* Only EAL lcores are supported
>
>Not that I am against that limitation - it is quite common within DPDK...
>but reading through the code, not sure where it comes from?
>

It was tested only on lcores (look TLS for example). I don't know how other constructs behave and what are their specific requirements. 

This library has some room for improvements so I think future changes will address that limitations. 

>> +
>> +* EAL lcores must not share a cpu
>> +
>> +* Each EAL lcore measures same group of events
>> +
>>
>>  Profiling on x86
>>  ----------------
>> diff --git a/doc/guides/rel_notes/release_24_11.rst
>> b/doc/guides/rel_notes/release_24_11.rst
>> index 53d8661365..06fa294787 100644
>> --- a/doc/guides/rel_notes/release_24_11.rst
>> +++ b/doc/guides/rel_notes/release_24_11.rst
>> @@ -150,6 +150,13 @@ New Features
>>
>>    * Added independent enqueue feature.
>>
>> +* **Added PMU library.**
>> +
>> +  Added a new performance monitoring unit (PMU) library which allows
>> + applications  to perform self monitoring activities without depending on external utilities
>like perf.
>> +  After integration with :doc:`../prog_guide/trace_lib` data gathered
>> + from hardware counters  can be stored in CTF format for further analysis.
>> +
>>
>>  Removed Items
>>  -------------
>> diff --git a/lib/meson.build b/lib/meson.build index
>> 162287753f..cc7a4bd535 100644
>> --- a/lib/meson.build
>> +++ b/lib/meson.build
>> @@ -13,6 +13,7 @@ libraries = [
>>          'kvargs', # eal depends on kvargs
>>          'argparse',
>>          'telemetry', # basic info querying
>> +        'pmu',
>>          'eal', # everything depends on eal
>>          'ptr_compress',
>>          'ring',
>> diff --git a/lib/pmu/meson.build b/lib/pmu/meson.build new file mode
>> 100644 index 0000000000..386232e5c7
>> --- /dev/null
>> +++ b/lib/pmu/meson.build
>> @@ -0,0 +1,12 @@
>> +# SPDX-License-Identifier: BSD-3-Clause # Copyright(C) 2024 Marvell
>> +International Ltd.
>> +
>> +headers = files('rte_pmu.h')
>> +
>> +if not is_linux
>> +    subdir_done()
>> +endif
>> +
>> +sources = files('rte_pmu.c')
>> +
>> +deps += ['log']
>> diff --git a/lib/pmu/pmu_private.h b/lib/pmu/pmu_private.h new file
>> mode 100644 index 0000000000..d2b15615bf
>> --- /dev/null
>> +++ b/lib/pmu/pmu_private.h
>> @@ -0,0 +1,32 @@
>> +/* SPDX-License-Identifier: BSD-3-Clause
>> + * Copyright(c) 2024 Marvell
>> + */
>> +
>> +#ifndef _PMU_PRIVATE_H_
>> +#define _PMU_PRIVATE_H_
>> +
>> +/**
>> + * Architecture specific PMU init callback.
>> + *
>> + * @return
>> + *   0 in case of success, negative value otherwise.
>> + */
>> +int
>> +pmu_arch_init(void);
>> +
>> +/**
>> + * Architecture specific PMU cleanup callback.
>> + */
>> +void
>> +pmu_arch_fini(void);
>> +
>> +/**
>> + * Apply architecture specific settings to config before passing it to syscall.
>> + *
>> + * @param config
>> + *   Architecture specific event configuration. Consult kernel sources for available options.
>> + */
>> +void
>> +pmu_arch_fixup_config(uint64_t config[3]);
>> +
>> +#endif /* _PMU_PRIVATE_H_ */
>> diff --git a/lib/pmu/rte_pmu.c b/lib/pmu/rte_pmu.c new file mode
>> 100644 index 0000000000..3a861bca71
>> --- /dev/null
>> +++ b/lib/pmu/rte_pmu.c
>> @@ -0,0 +1,463 @@
>> +/* SPDX-License-Identifier: BSD-3-Clause
>> + * Copyright(C) 2024 Marvell International Ltd.
>> + */
>> +
>> +#include <ctype.h>
>> +#include <dirent.h>
>> +#include <errno.h>
>> +#include <regex.h>
>> +#include <stdlib.h>
>> +#include <string.h>
>> +#include <sys/ioctl.h>
>> +#include <sys/mman.h>
>> +#include <sys/queue.h>
>> +#include <sys/syscall.h>
>> +#include <unistd.h>
>> +
>> +#include <rte_atomic.h>
>> +#include <rte_lcore.h>
>> +#include <rte_per_lcore.h>
>> +#include <rte_pmu.h>
>> +#include <rte_spinlock.h>
>> +#include <rte_tailq.h>
>> +
>> +#include "pmu_private.h"
>> +
>> +#define EVENT_SOURCE_DEVICES_PATH "/sys/bus/event_source/devices"
>> +
>> +#define GENMASK_ULL(h, l) ((~0ULL - (1ULL << (l)) + 1) & (~0ULL >>
>> +((64 - 1 - (h)))))
>
>Would RTE_GENMASK64(high, low) work for you here?

If that RTE_GENMASK64 generates continuous mask of 1s then it would work. I didn't realize DPDK has it. 

>
>> +#define FIELD_PREP(m, v) (((uint64_t)(v) << (__builtin_ffsll(m) - 1))
>> +& (m))
>> +
>> +/* A structure describing an event */ struct rte_pmu_event {
>> +	char *name;
>> +	unsigned int index;
>> +	TAILQ_ENTRY(rte_pmu_event) next;
>> +};
>> +
>> +RTE_DEFINE_PER_LCORE(struct rte_pmu_event_group, _event_group);
>> +struct rte_pmu rte_pmu;
>> +
>> +/*
>> + * Following __rte_weak functions provide default no-op.
>> +Architectures should override them if
>> + * necessary.
>> + */
>> +
>> +int
>> +__rte_weak pmu_arch_init(void)
>> +{
>> +	return 0;
>> +}
>> +
>> +void
>> +__rte_weak pmu_arch_fini(void)
>> +{
>> +}
>> +
>> +void
>> +__rte_weak pmu_arch_fixup_config(uint64_t __rte_unused config[3]) { }
>> +
>> +static int
>> +get_term_format(const char *name, int *num, uint64_t *mask) {
>> +	char path[PATH_MAX];
>> +	char *config = NULL;
>> +	int high, low, ret;
>> +	FILE *fp;
>> +
>> +	*num = *mask = 0;
>> +	snprintf(path, sizeof(path), EVENT_SOURCE_DEVICES_PATH "/%s/format/%s", rte_pmu.name, name);
>> +	fp = fopen(path, "r");
>> +	if (fp == NULL)
>> +		return -errno;
>> +
>> +	errno = 0;
>> +	ret = fscanf(fp, "%m[^:]:%d-%d", &config, &low, &high);
>> +	if (ret < 2) {
>> +		ret = -ENODATA;
>> +		goto out;
>> +	}
>> +	if (errno) {
>> +		ret = -errno;
>> +		goto out;
>> +	}
>> +
>> +	if (ret == 2)
>> +		high = low;
>> +
>> +	*mask = GENMASK_ULL(high, low);
>> +	/* Last digit should be [012]. If last digit is missing 0 is implied. */
>> +	*num = config[strlen(config) - 1];
>> +	*num = isdigit(*num) ? *num - '0' : 0;
>> +
>> +	ret = 0;
>> +out:
>> +	free(config);
>> +	fclose(fp);
>> +
>> +	return ret;
>> +}
>> +
>> +static int
>> +parse_event(char *buf, uint64_t config[3]) {
>> +	char *token, *term;
>> +	int num, ret, val;
>> +	uint64_t mask;
>> +
>> +	config[0] = config[1] = config[2] = 0;
>> +
>> +	token = strtok(buf, ",");
>> +	while (token) {
>> +		errno = 0;
>> +		/* <term>=<value> */
>> +		ret = sscanf(token, "%m[^=]=%i", &term, &val);
>> +		if (ret < 1)
>> +			return -ENODATA;
>> +		if (errno)
>> +			return -errno;
>> +		if (ret == 1)
>> +			val = 1;
>> +
>> +		ret = get_term_format(term, &num, &mask);
>> +		free(term);
>> +		if (ret)
>> +			return ret;
>> +
>> +		config[num] |= FIELD_PREP(mask, val);
>> +		token = strtok(NULL, ",");
>> +	}
>> +
>> +	return 0;
>> +}
>> +
>> +static int
>> +get_event_config(const char *name, uint64_t config[3]) {
>> +	char path[PATH_MAX], buf[BUFSIZ];
>> +	FILE *fp;
>> +	int ret;
>> +
>> +	snprintf(path, sizeof(path), EVENT_SOURCE_DEVICES_PATH "/%s/events/%s", rte_pmu.name, name);
>> +	fp = fopen(path, "r");
>> +	if (fp == NULL)
>> +		return -errno;
>> +
>> +	ret = fread(buf, 1, sizeof(buf), fp);
>> +	if (ret == 0) {
>> +		fclose(fp);
>> +
>> +		return -EINVAL;
>> +	}
>> +	fclose(fp);
>> +	buf[ret] = '\0';
>> +
>> +	return parse_event(buf, config);
>> +}
>> +
>> +static int
>> +do_perf_event_open(uint64_t config[3], int group_fd) {
>> +	struct perf_event_attr attr = {
>> +		.size = sizeof(struct perf_event_attr),
>> +		.type = PERF_TYPE_RAW,
>> +		.exclude_kernel = 1,
>> +		.exclude_hv = 1,
>> +		.disabled = 1,
>> +		.pinned = group_fd == -1,
>> +	};
>> +
>> +	pmu_arch_fixup_config(config);
>> +
>> +	attr.config = config[0];
>> +	attr.config1 = config[1];
>> +	attr.config2 = config[2];
>> +
>> +	return syscall(SYS_perf_event_open, &attr, 0, -1, group_fd, 0); }
>> +
>> +static int
>> +open_events(struct rte_pmu_event_group *group) {
>> +	struct rte_pmu_event *event;
>> +	uint64_t config[3];
>> +	int num = 0, ret;
>> +
>> +	/* group leader gets created first, with fd = -1 */
>> +	group->fds[0] = -1;
>> +
>> +	TAILQ_FOREACH(event, &rte_pmu.event_list, next) {
>> +		ret = get_event_config(event->name, config);
>> +		if (ret)
>> +			continue;
>> +
>> +		ret = do_perf_event_open(config, group->fds[0]);
>> +		if (ret == -1) {
>> +			ret = -errno;
>> +			goto out;
>> +		}
>> +
>> +		group->fds[event->index] = ret;
>> +		num++;
>> +	}
>> +
>> +	return 0;
>> +out:
>> +	for (--num; num >= 0; num--) {
>> +		close(group->fds[num]);
>> +		group->fds[num] = -1;
>> +	}
>> +
>> +
>> +	return ret;
>> +}
>> +
>> +static int
>> +mmap_events(struct rte_pmu_event_group *group) {
>> +	long page_size = sysconf(_SC_PAGE_SIZE);
>> +	unsigned int i;
>> +	void *addr;
>> +	int ret;
>> +
>> +	for (i = 0; i < rte_pmu.num_group_events; i++) {
>> +		addr = mmap(0, page_size, PROT_READ, MAP_SHARED, group->fds[i], 0);
>> +		if (addr == MAP_FAILED) {
>> +			ret = -errno;
>> +			goto out;
>> +		}
>> +
>> +		group->mmap_pages[i] = addr;
>> +	}
>> +
>> +	return 0;
>> +out:
>> +	for (; i; i--) {
>> +		munmap(group->mmap_pages[i - 1], page_size);
>> +		group->mmap_pages[i - 1] = NULL;
>> +	}
>> +
>> +	return ret;
>> +}
>> +
>> +static void
>> +cleanup_events(struct rte_pmu_event_group *group) {
>> +	unsigned int i;
>> +
>> +	if (group->fds[0] != -1)
>> +		ioctl(group->fds[0], PERF_EVENT_IOC_DISABLE, PERF_IOC_FLAG_GROUP);
>> +
>> +	for (i = 0; i < rte_pmu.num_group_events; i++) {
>> +		if (group->mmap_pages[i]) {
>> +			munmap(group->mmap_pages[i], sysconf(_SC_PAGE_SIZE));
>> +			group->mmap_pages[i] = NULL;
>> +		}
>> +
>> +		if (group->fds[i] != -1) {
>> +			close(group->fds[i]);
>> +			group->fds[i] = -1;
>> +		}
>> +	}
>> +
>> +	group->enabled = false;
>> +}
>> +
>> +int
>> +__rte_pmu_enable_group(void)
>> +{
>> +	struct rte_pmu_event_group *group = &RTE_PER_LCORE(_event_group);
>> +	int ret;
>> +
>> +	if (rte_pmu.num_group_events == 0)
>> +		return -ENODEV;
>> +
>> +	ret = open_events(group);
>> +	if (ret)
>> +		goto out;
>> +
>> +	ret = mmap_events(group);
>> +	if (ret)
>> +		goto out;
>> +
>> +	if (ioctl(group->fds[0], PERF_EVENT_IOC_RESET, PERF_IOC_FLAG_GROUP) == -1) {
>> +		ret = -errno;
>> +		goto out;
>> +	}
>> +
>> +	if (ioctl(group->fds[0], PERF_EVENT_IOC_ENABLE, PERF_IOC_FLAG_GROUP) == -1) {
>> +		ret = -errno;
>> +		goto out;
>> +	}
>> +
>> +	rte_spinlock_lock(&rte_pmu.lock);
>> +	TAILQ_INSERT_TAIL(&rte_pmu.event_group_list, group, next);
>> +	rte_spinlock_unlock(&rte_pmu.lock);
>
>I thought that after previous round of reviews, we got a consensus that it is a bad idea to insert
>pointer of TLS variable into the global list:
>https://urldefense.proofpoint.com/v2/url?u=https-
>3A__patchwork.dpdk.org_project_dpdk_patch_20230216175502.3164820-2D2-2Dtduszynski-
>40marvell.com_&d=DwIFAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=PZNXgrbjdlXxVEEGYkxIxRndyEUwWU_ad5ce22YI6Is&m=oJ
>-
>eSnmJoK0r1zVFhKrkWMnfelOkxqpjtX2fCrXaG2RdWagOqAQ7vcFCJ0dOWrTt&s=TvGxqQmUz_U3xLOroMxOmsCiaxdqbNLi6GZ
>pHIefniw&e=

I don't think there was any consensus. It was rather your point of view solely. So I still believe
that given currently that only runs on lcores and lcores do not terminate before the main 
one hence it's safe to access TLS from a main. 

Once support for running on anything other than lcore gets added current mechanics is likely to change.

And as you pointed out, there are plenty of options available. 

>As I said before - I believe it is error prone, and needs to be addressed.
>There are plenty of other ways to have one object per lcore (if that's really necessary):
>array of RTE_MAX_LCORE indexed by lcore_id, or new rte_lcore_var.h API introduced by Mattias
>recently, or even simple TLS pointer with malloc() behind would be better.
>
>> +	group->enabled = true;
>> +
>> +	return 0;
>> +
>> +out:
>> +	cleanup_events(group);
>> +
>> +	return ret;
>> +}
>> +
>> +static int
>> +scan_pmus(void)
>> +{
>> +	char path[PATH_MAX];
>> +	struct dirent *dent;
>> +	const char *name;
>> +	DIR *dirp;
>> +
>> +	dirp = opendir(EVENT_SOURCE_DEVICES_PATH);
>> +	if (dirp == NULL)
>> +		return -errno;
>> +
>> +	while ((dent = readdir(dirp))) {
>> +		name = dent->d_name;
>> +		if (name[0] == '.')
>> +			continue;
>> +
>> +		/* sysfs entry should either contain cpus or be a cpu */
>> +		if (!strcmp(name, "cpu"))
>> +			break;
>> +
>> +		snprintf(path, sizeof(path), EVENT_SOURCE_DEVICES_PATH "/%s/cpus", name);
>> +		if (access(path, F_OK) == 0)
>> +			break;
>> +	}
>> +
>> +	if (dent) {
>> +		rte_pmu.name = strdup(name);
>> +		if (rte_pmu.name == NULL) {
>> +			closedir(dirp);
>> +
>> +			return -ENOMEM;
>> +		}
>> +	}
>> +
>> +	closedir(dirp);
>> +
>> +	return rte_pmu.name ? 0 : -ENODEV;
>> +}
>> +
>> +static struct rte_pmu_event *
>> +new_event(const char *name)
>> +{
>> +	struct rte_pmu_event *event;
>> +
>> +	event = calloc(1, sizeof(*event));
>> +	if (event == NULL)
>> +		goto out;
>> +
>> +	event->name = strdup(name);
>> +	if (event->name == NULL) {
>> +		free(event);
>> +		event = NULL;
>> +	}
>> +
>> +out:
>> +	return event;
>> +}
>> +
>> +static void
>> +free_event(struct rte_pmu_event *event) {
>> +	free(event->name);
>> +	free(event);
>> +}
>> +
>> +int
>> +rte_pmu_add_event(const char *name)
>> +{
>> +	struct rte_pmu_event *event;
>> +	char path[PATH_MAX];
>> +
>> +	if (rte_pmu.name == NULL)
>> +		return -ENODEV;
>> +
>> +	if (rte_pmu.num_group_events + 1 >= RTE_MAX_NUM_GROUP_EVENTS)
>> +		return -ENOSPC;
>> +
>> +	snprintf(path, sizeof(path), EVENT_SOURCE_DEVICES_PATH "/%s/events/%s", rte_pmu.name, name);
>> +	if (access(path, R_OK))
>> +		return -ENODEV;
>> +
>> +	TAILQ_FOREACH(event, &rte_pmu.event_list, next) {
>> +		if (!strcmp(event->name, name))
>> +			return event->index;
>> +		continue;
>> +	}
>> +
>> +	event = new_event(name);
>> +	if (event == NULL)
>> +		return -ENOMEM;
>> +
>> +	event->index = rte_pmu.num_group_events++;
>> +	TAILQ_INSERT_TAIL(&rte_pmu.event_list, event, next);
>> +
>> +	return event->index;
>> +}
>> +
>> +int
>> +rte_pmu_init(void)
>> +{
>> +	int ret;
>> +
>> +	/* Allow calling init from multiple contexts within a single thread. This simplifies
>> +	 * resource management a bit e.g in case fast-path tracepoint has already been enabled
>> +	 * via command line but application doesn't care enough and performs init/fini again.
>> +	 */
>> +	if (rte_atomic_fetch_add_explicit(&rte_pmu.initialized, 1, rte_memory_order_seq_cst) != 0)
>> +		return 0;
>
>Not sure I understand the comment above: rte_pmu_init() supposed to be MT safe or not?
>If yes, then that function probably need to wait till initialization finishes.
>

It's supposed to be MT safe. Generally there are two options here.

# one you pointed out i.e spinning in a loop until everything is up and running 
# checking if tracing is initialized during event read-out

Initially second option was chosen and rte_pmu_read() performs check prior to doing anything.
Still, I see a problem there so I'm glad you raised that. 

>> +
>> +	ret = scan_pmus();
>> +	if (ret)
>> +		goto out;
>> +
>> +	ret = pmu_arch_init();
>> +	if (ret)
>> +		goto out;
>> +
>> +	TAILQ_INIT(&rte_pmu.event_list);
>> +	TAILQ_INIT(&rte_pmu.event_group_list);
>> +	rte_spinlock_init(&rte_pmu.lock);
>> +
>> +	return 0;
>> +out:
>> +	free(rte_pmu.name);
>> +	rte_pmu.name = NULL;
>> +
>> +	return ret;
>> +}
>> +
>> +void
>> +rte_pmu_fini(void)
>> +{
>> +	struct rte_pmu_event_group *group, *tmp_group;
>> +	struct rte_pmu_event *event, *tmp_event;
>> +
>> +	/* cleanup once init count drops to zero */
>> +	if (rte_atomic_fetch_sub_explicit(&rte_pmu.initialized, 1,
>> +					  rte_memory_order_seq_cst) - 1 != 0)
>> +		return;
>> +
>> +	RTE_TAILQ_FOREACH_SAFE(event, &rte_pmu.event_list, next, tmp_event) {
>> +		TAILQ_REMOVE(&rte_pmu.event_list, event, next);
>> +		free_event(event);
>> +	}
>> +
>> +	RTE_TAILQ_FOREACH_SAFE(group, &rte_pmu.event_group_list, next, tmp_group) {
>> +		TAILQ_REMOVE(&rte_pmu.event_group_list, group, next);
>> +		cleanup_events(group);
>> +	}
>> +
>> +	pmu_arch_fini();
>> +	free(rte_pmu.name);
>> +	rte_pmu.name = NULL;
>> +	rte_pmu.num_group_events = 0;
>> +}
>> diff --git a/lib/pmu/rte_pmu.h b/lib/pmu/rte_pmu.h new file mode
>> 100644 index 0000000000..09238ee33d
>> --- /dev/null
>> +++ b/lib/pmu/rte_pmu.h
>> @@ -0,0 +1,227 @@
>> +/* SPDX-License-Identifier: BSD-3-Clause
>> + * Copyright(c) 2024 Marvell
>> + */
>> +
>> +#ifndef _RTE_PMU_H_
>> +#define _RTE_PMU_H_
>> +
>> +/**
>> + * @file
>> + *
>> + * PMU event tracing operations
>> + *
>> + * This file defines generic API and types necessary to setup PMU and
>> + * read selected counters in runtime.
>> + */
>> +
>> +#ifdef __cplusplus
>> +extern "C" {
>> +#endif
>> +
>> +#include <errno.h>
>> +
>> +#include <rte_common.h>
>> +#include <rte_compat.h>
>> +
>> +#ifdef RTE_EXEC_ENV_LINUX
>> +
>> +#include <linux/perf_event.h>
>> +
>> +#include <rte_atomic.h>
>> +#include <rte_branch_prediction.h>
>> +#include <rte_spinlock.h>
>> +
>> +/** Maximum number of events in a group */ #define
>> +RTE_MAX_NUM_GROUP_EVENTS 8
>> +
>> +/**
>> + * A structure describing a group of events.
>> + */
>> +struct __rte_cache_aligned rte_pmu_event_group {
>> +	/** array of user pages */
>> +	struct perf_event_mmap_page *mmap_pages[RTE_MAX_NUM_GROUP_EVENTS];
>> +	int fds[RTE_MAX_NUM_GROUP_EVENTS]; /**< array of event descriptors */
>> +	bool enabled; /**< true if group was enabled on particular lcore */
>> +	TAILQ_ENTRY(rte_pmu_event_group) next; /**< list entry */ };
>> +
>> +/**
>> + * A PMU state container.
>> + */
>> +struct rte_pmu {
>> +	char *name; /**< name of core PMU listed under /sys/bus/event_source/devices */
>> +	rte_spinlock_t lock; /**< serialize access to event group list */
>> +	TAILQ_HEAD(, rte_pmu_event_group) event_group_list; /**< list of event groups */
>> +	unsigned int num_group_events; /**< number of events in a group */
>> +	TAILQ_HEAD(, rte_pmu_event) event_list; /**< list of matching events */
>> +	unsigned int initialized; /**< initialization counter */ };
>> +
>> +/** lcore event group */
>> +RTE_DECLARE_PER_LCORE(struct rte_pmu_event_group, _event_group);
>> +
>> +/** PMU state container */
>> +extern struct rte_pmu rte_pmu;
>> +
>> +/** Each architecture supporting PMU needs to provide its own version
>> +*/ #ifndef rte_pmu_pmc_read #define rte_pmu_pmc_read(index) ({
>> +(void)(index); 0; }) #endif
>> +
>> +/**
>> + * @warning
>> + * @b EXPERIMENTAL: this API may change without prior notice
>> + *
>> + * Read PMU counter.
>> + *
>> + * @warning This should be not called directly.
>> + *
>> + * @param pc
>> + *   Pointer to the mmapped user page.
>> + * @return
>> + *   Counter value read from hardware.
>> + */
>
>As I already asked you several times in previous review rounds:
>https://urldefense.proofpoint.com/v2/url?u=https-
>3A__patchwork.dpdk.org_project_dpdk_patch_20230216175502.3164820-2D2-2Dtduszynski-
>40marvell.com_&d=DwIFAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=PZNXgrbjdlXxVEEGYkxIxRndyEUwWU_ad5ce22YI6Is&m=oJ
>-
>eSnmJoK0r1zVFhKrkWMnfelOkxqpjtX2fCrXaG2RdWagOqAQ7vcFCJ0dOWrTt&s=TvGxqQmUz_U3xLOroMxOmsCiaxdqbNLi6GZ
>pHIefniw&e=
>please add a detailed comment here:
>i.e. that function (and rte_pmu_read()) to work properly supposed to be used only on
>a thread that *always* binded to just one cpu core.
>

Okay I will add comment under limitations that we want one lcore per cpu core. 

>> +__rte_experimental
>> +static __rte_always_inline uint64_t
>> +__rte_pmu_read_userpage(struct perf_event_mmap_page *pc) { #define
>> +__RTE_PMU_READ_ONCE(x) (*(const volatile typeof(x) *)&(x))
>> +	uint64_t width, offset;
>> +	uint32_t seq, index;
>> +	int64_t pmc;
>> +
>> +	for (;;) {
>> +		seq = __RTE_PMU_READ_ONCE(pc->lock);
>> +		rte_compiler_barrier();
>> +		index = __RTE_PMU_READ_ONCE(pc->index);
>> +		offset = __RTE_PMU_READ_ONCE(pc->offset);
>> +		width = __RTE_PMU_READ_ONCE(pc->pmc_width);
>> +
>> +		/* index set to 0 means that particular counter cannot be used */
>> +		if (likely(pc->cap_user_rdpmc && index)) {
>> +			pmc = rte_pmu_pmc_read(index - 1);
>> +			pmc <<= 64 - width;
>> +			pmc >>= 64 - width;
>> +			offset += pmc;
>> +		}
>> +
>> +		rte_compiler_barrier();
>> +
>> +		if (likely(__RTE_PMU_READ_ONCE(pc->lock) == seq))
>> +			return offset;
>> +	}
>> +
>> +	return 0;
>> +}
>> +
>> +/**
>> + * @warning
>> + * @b EXPERIMENTAL: this API may change without prior notice
>> + *
>> + * Enable group of events on the calling lcore.
>> + *
>> + * @warning This should be not called directly.
>> + *
>> + * @return
>> + *   0 in case of success, negative value otherwise.
>> + */
>> +__rte_experimental
>> +int
>> +__rte_pmu_enable_group(void);
>> +
>> +/**
>> + * @warning
>> + * @b EXPERIMENTAL: this API may change without prior notice
>> + *
>> + * Initialize PMU library.
>> + *
>> + * @warning This should be not called directly.
>
>Same questions as before: if it is internal, then why:
>a) it is located in public API .h file
>b) why it is called directly by user-code in app/test/test_pmu.c ?
>

Some functions are located in public API file out of convenience. They are either macros or inline functions. 
Documentation points out which are safe to be called directly which are not so user-space knows the rules. 

>> + *
>> + * @return
>> + *   0 in case of success, negative value otherwise.
>> + */
>> +__rte_experimental
>> +int
>> +rte_pmu_init(void);
>> +
>> +/**
>> + * @warning
>> + * @b EXPERIMENTAL: this API may change without prior notice
>> + *
>> + * Finalize PMU library. This should be called after PMU counters are no longer being read.
>> + */
>> +__rte_experimental
>> +void
>> +rte_pmu_fini(void);
>
>Hmm..., why _fini_() is allowed to be called directly while _init_() doesn't?
>

Typo. Functions that should not be called directly are prefixed with double underscore. 
I'll scan through header file once again and remove invalid comments. Thanks for 
catching this. 

>> +/**
>> + * @warning
>> + * @b EXPERIMENTAL: this API may change without prior notice
>> + *
>> + * Add event to the group of enabled events.
>> + *
>> + * @param name
>> + *   Name of an event listed under /sys/bus/event_source/devices/pmu/events.
>> + * @return
>> + *   Event index in case of success, negative value otherwise.
>> + */
>> +__rte_experimental
>> +int
>> +rte_pmu_add_event(const char *name);
>> +
>> +/**
>> + * @warning
>> + * @b EXPERIMENTAL: this API may change without prior notice
>> + *
>> + * Read hardware counter configured to count occurrences of an event.
>> + *
>> + * @param index
>> + *   Index of an event to be read.
>> + * @return
>> + *   Event value read from register. In case of errors or lack of support
>> + *   0 is returned. In other words, stream of zeros in a trace file
>> + *   indicates problem with reading particular PMU event register.
>> + */
>> +__rte_experimental
>> +static __rte_always_inline uint64_t
>> +rte_pmu_read(unsigned int index)
>> +{
>> +	struct rte_pmu_event_group *group = &RTE_PER_LCORE(_event_group);
>> +	int ret;
>> +
>> +	if (unlikely(!rte_pmu.initialized))
>> +		return 0;
>> +
>> +	if (unlikely(!group->enabled)) {
>> +		ret = __rte_pmu_enable_group();
>> +		if (ret)
>> +			return 0;
>> +	}
>> +
>> +	if (unlikely(index >= rte_pmu.num_group_events))
>> +		return 0;
>> +
>> +	return __rte_pmu_read_userpage(group->mmap_pages[index]);
>> +}
>> +
>> +#else /* !RTE_EXEC_ENV_LINUX */
>> +
>> +__rte_experimental
>> +static inline int rte_pmu_init(void) { return -ENOTSUP; }
>> +
>> +__rte_experimental
>> +static inline void rte_pmu_fini(void) { }
>> +
>> +__rte_experimental
>> +static inline int rte_pmu_add_event(const char *name __rte_unused) {
>> +return -ENOTSUP; }
>> +
>> +__rte_experimental
>> +static inline uint64_t rte_pmu_read(unsigned int index __rte_unused)
>> +{ return UINT64_MAX; }
>> +
>> +#endif /* RTE_EXEC_ENV_LINUX */
>> +
>> +#ifdef __cplusplus
>> +}
>> +#endif
>> +
>> +#endif /* _RTE_PMU_H_ */
>> diff --git a/lib/pmu/version.map b/lib/pmu/version.map new file mode
>> 100644 index 0000000000..d7c80ce4ce
>> --- /dev/null
>> +++ b/lib/pmu/version.map
>> @@ -0,0 +1,14 @@
>> +EXPERIMENTAL {
>> +	global:
>> +
>> +	# added in 24.11
>> +	__rte_pmu_enable_group;
>> +	per_lcore__event_group;
>> +	rte_pmu;
>> +	rte_pmu_add_event;
>> +	rte_pmu_fini;
>> +	rte_pmu_init;
>> +	rte_pmu_read;
>> +
>> +	local: *;
>> +};
>> --
>> 2.34.1