These changes are the raw update to linux-4.4.6-rt14. Kernel sources

are taken from kernel.org, and rt patch from the rt wiki download page.

During the rebasing, the following patch collided:

Force tick interrupt and get rid of softirq magic(I70131fb85).

Collisions have been removed because its logic was found on the
source already.

Change-Id: I7f57a4081d9deaa0d9ccfc41a6c8daccdee3b769
Signed-off-by: José Pekkarinen <jose.pekkarinen@nokia.com>

10 files changed, 964 insertions, 444 deletions

diff --git a/kernel/tools/perf/bench/Build b/kernel/tools/perf/bench/Build
index 5ce98023d..60bf11943 100644
--- a/kernel/tools/perf/bench/Build
+++ b/kernel/tools/perf/bench/Build
@@ -1,9 +1,11 @@
 perf-y += sched-messaging.o
 perf-y += sched-pipe.o
-perf-y += mem-memcpy.o
+perf-y += mem-functions.o
 perf-y += futex-hash.o
 perf-y += futex-wake.o
+perf-y += futex-wake-parallel.o
 perf-y += futex-requeue.o
+perf-y += futex-lock-pi.o
 
 perf-$(CONFIG_X86_64) += mem-memcpy-x86-64-asm.o
 perf-$(CONFIG_X86_64) += mem-memset-x86-64-asm.o
diff --git a/kernel/tools/perf/bench/bench.h b/kernel/tools/perf/bench/bench.h
index 3c4dd44d4..a50df86f2 100644
--- a/kernel/tools/perf/bench/bench.h
+++ b/kernel/tools/perf/bench/bench.h
@@ -33,7 +33,11 @@ extern int bench_mem_memcpy(int argc, const char **argv,
 extern int bench_mem_memset(int argc, const char **argv, const char *prefix);
 extern int bench_futex_hash(int argc, const char **argv, const char *prefix);
 extern int bench_futex_wake(int argc, const char **argv, const char *prefix);
+extern int bench_futex_wake_parallel(int argc, const char **argv,
+				     const char *prefix);
 extern int bench_futex_requeue(int argc, const char **argv, const char *prefix);
+/* pi futexes */
+extern int bench_futex_lock_pi(int argc, const char **argv, const char *prefix);
 
 #define BENCH_FORMAT_DEFAULT_STR	"default"
 #define BENCH_FORMAT_DEFAULT		0
diff --git a/kernel/tools/perf/bench/futex-lock-pi.c b/kernel/tools/perf/bench/futex-lock-pi.c
new file mode 100644
index 000000000..bc6a16adb
--- /dev/null
+++ b/kernel/tools/perf/bench/futex-lock-pi.c
@@ -0,0 +1,219 @@
+/*
+ * Copyright (C) 2015 Davidlohr Bueso.
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/stat.h"
+#include "../util/parse-options.h"
+#include "../util/header.h"
+#include "bench.h"
+#include "futex.h"
+
+#include <err.h>
+#include <stdlib.h>
+#include <sys/time.h>
+#include <pthread.h>
+
+struct worker {
+	int tid;
+	u_int32_t *futex;
+	pthread_t thread;
+	unsigned long ops;
+};
+
+static u_int32_t global_futex = 0;
+static struct worker *worker;
+static unsigned int nsecs = 10;
+static bool silent = false, multi = false;
+static bool done = false, fshared = false;
+static unsigned int ncpus, nthreads = 0;
+static int futex_flag = 0;
+struct timeval start, end, runtime;
+static pthread_mutex_t thread_lock;
+static unsigned int threads_starting;
+static struct stats throughput_stats;
+static pthread_cond_t thread_parent, thread_worker;
+
+static const struct option options[] = {
+	OPT_UINTEGER('t', "threads",  &nthreads, "Specify amount of threads"),
+	OPT_UINTEGER('r', "runtime", &nsecs,     "Specify runtime (in seconds)"),
+	OPT_BOOLEAN( 'M', "multi",   &multi,     "Use multiple futexes"),
+	OPT_BOOLEAN( 's', "silent",  &silent,    "Silent mode: do not display data/details"),
+	OPT_BOOLEAN( 'S', "shared",  &fshared,   "Use shared futexes instead of private ones"),
+	OPT_END()
+};
+
+static const char * const bench_futex_lock_pi_usage[] = {
+	"perf bench futex requeue <options>",
+	NULL
+};
+
+static void print_summary(void)
+{
+	unsigned long avg = avg_stats(&throughput_stats);
+	double stddev = stddev_stats(&throughput_stats);
+
+	printf("%sAveraged %ld operations/sec (+- %.2f%%), total secs = %d\n",
+	       !silent ? "\n" : "", avg, rel_stddev_stats(stddev, avg),
+	       (int) runtime.tv_sec);
+}
+
+static void toggle_done(int sig __maybe_unused,
+			siginfo_t *info __maybe_unused,
+			void *uc __maybe_unused)
+{
+	/* inform all threads that we're done for the day */
+	done = true;
+	gettimeofday(&end, NULL);
+	timersub(&end, &start, &runtime);
+}
+
+static void *workerfn(void *arg)
+{
+	struct worker *w = (struct worker *) arg;
+
+	pthread_mutex_lock(&thread_lock);
+	threads_starting--;
+	if (!threads_starting)
+		pthread_cond_signal(&thread_parent);
+	pthread_cond_wait(&thread_worker, &thread_lock);
+	pthread_mutex_unlock(&thread_lock);
+
+	do {
+		int ret;
+	again:
+		ret = futex_lock_pi(w->futex, NULL, 0, futex_flag);
+
+		if (ret) { /* handle lock acquisition */
+			if (!silent)
+				warn("thread %d: Could not lock pi-lock for %p (%d)",
+				     w->tid, w->futex, ret);
+			if (done)
+				break;
+
+			goto again;
+		}
+
+		usleep(1);
+		ret = futex_unlock_pi(w->futex, futex_flag);
+		if (ret && !silent)
+			warn("thread %d: Could not unlock pi-lock for %p (%d)",
+			     w->tid, w->futex, ret);
+		w->ops++; /* account for thread's share of work */
+	}  while (!done);
+
+	return NULL;
+}
+
+static void create_threads(struct worker *w, pthread_attr_t thread_attr)
+{
+	cpu_set_t cpu;
+	unsigned int i;
+
+	threads_starting = nthreads;
+
+	for (i = 0; i < nthreads; i++) {
+		worker[i].tid = i;
+
+		if (multi) {
+			worker[i].futex = calloc(1, sizeof(u_int32_t));
+			if (!worker[i].futex)
+				err(EXIT_FAILURE, "calloc");
+		} else
+			worker[i].futex = &global_futex;
+
+		CPU_ZERO(&cpu);
+		CPU_SET(i % ncpus, &cpu);
+
+		if (pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpu))
+			err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
+
+		if (pthread_create(&w[i].thread, &thread_attr, workerfn, &worker[i]))
+			err(EXIT_FAILURE, "pthread_create");
+	}
+}
+
+int bench_futex_lock_pi(int argc, const char **argv,
+			const char *prefix __maybe_unused)
+{
+	int ret = 0;
+	unsigned int i;
+	struct sigaction act;
+	pthread_attr_t thread_attr;
+
+	argc = parse_options(argc, argv, options, bench_futex_lock_pi_usage, 0);
+	if (argc)
+		goto err;
+
+	ncpus = sysconf(_SC_NPROCESSORS_ONLN);
+
+	sigfillset(&act.sa_mask);
+	act.sa_sigaction = toggle_done;
+	sigaction(SIGINT, &act, NULL);
+
+	if (!nthreads)
+		nthreads = ncpus;
+
+	worker = calloc(nthreads, sizeof(*worker));
+	if (!worker)
+		err(EXIT_FAILURE, "calloc");
+
+	if (!fshared)
+		futex_flag = FUTEX_PRIVATE_FLAG;
+
+	printf("Run summary [PID %d]: %d threads doing pi lock/unlock pairing for %d secs.\n\n",
+	       getpid(), nthreads, nsecs);
+
+	init_stats(&throughput_stats);
+	pthread_mutex_init(&thread_lock, NULL);
+	pthread_cond_init(&thread_parent, NULL);
+	pthread_cond_init(&thread_worker, NULL);
+
+	threads_starting = nthreads;
+	pthread_attr_init(&thread_attr);
+	gettimeofday(&start, NULL);
+
+	create_threads(worker, thread_attr);
+	pthread_attr_destroy(&thread_attr);
+
+	pthread_mutex_lock(&thread_lock);
+	while (threads_starting)
+		pthread_cond_wait(&thread_parent, &thread_lock);
+	pthread_cond_broadcast(&thread_worker);
+	pthread_mutex_unlock(&thread_lock);
+
+	sleep(nsecs);
+	toggle_done(0, NULL, NULL);
+
+	for (i = 0; i < nthreads; i++) {
+		ret = pthread_join(worker[i].thread, NULL);
+		if (ret)
+			err(EXIT_FAILURE, "pthread_join");
+	}
+
+	/* cleanup & report results */
+	pthread_cond_destroy(&thread_parent);
+	pthread_cond_destroy(&thread_worker);
+	pthread_mutex_destroy(&thread_lock);
+
+	for (i = 0; i < nthreads; i++) {
+		unsigned long t = worker[i].ops/runtime.tv_sec;
+
+		update_stats(&throughput_stats, t);
+		if (!silent)
+			printf("[thread %3d] futex: %p [ %ld ops/sec ]\n",
+			       worker[i].tid, worker[i].futex, t);
+
+		if (multi)
+			free(worker[i].futex);
+	}
+
+	print_summary();
+
+	free(worker);
+	return ret;
+err:
+	usage_with_options(bench_futex_lock_pi_usage, options);
+	exit(EXIT_FAILURE);
+}
diff --git a/kernel/tools/perf/bench/futex-wake-parallel.c b/kernel/tools/perf/bench/futex-wake-parallel.c
new file mode 100644
index 000000000..6d8c9fa2a
--- /dev/null
+++ b/kernel/tools/perf/bench/futex-wake-parallel.c
@@ -0,0 +1,294 @@
+/*
+ * Copyright (C) 2015 Davidlohr Bueso.
+ *
+ * Block a bunch of threads and let parallel waker threads wakeup an
+ * equal amount of them. The program output reflects the avg latency
+ * for each individual thread to service its share of work. Ultimately
+ * it can be used to measure futex_wake() changes.
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/stat.h"
+#include "../util/parse-options.h"
+#include "../util/header.h"
+#include "bench.h"
+#include "futex.h"
+
+#include <err.h>
+#include <stdlib.h>
+#include <sys/time.h>
+#include <pthread.h>
+
+struct thread_data {
+	pthread_t worker;
+	unsigned int nwoken;
+	struct timeval runtime;
+};
+
+static unsigned int nwakes = 1;
+
+/* all threads will block on the same futex -- hash bucket chaos ;) */
+static u_int32_t futex = 0;
+
+static pthread_t *blocked_worker;
+static bool done = false, silent = false, fshared = false;
+static unsigned int nblocked_threads = 0, nwaking_threads = 0;
+static pthread_mutex_t thread_lock;
+static pthread_cond_t thread_parent, thread_worker;
+static struct stats waketime_stats, wakeup_stats;
+static unsigned int ncpus, threads_starting;
+static int futex_flag = 0;
+
+static const struct option options[] = {
+	OPT_UINTEGER('t', "threads", &nblocked_threads, "Specify amount of threads"),
+	OPT_UINTEGER('w', "nwakers", &nwaking_threads, "Specify amount of waking threads"),
+	OPT_BOOLEAN( 's', "silent",  &silent,   "Silent mode: do not display data/details"),
+	OPT_BOOLEAN( 'S', "shared",  &fshared,  "Use shared futexes instead of private ones"),
+	OPT_END()
+};
+
+static const char * const bench_futex_wake_parallel_usage[] = {
+	"perf bench futex wake-parallel <options>",
+	NULL
+};
+
+static void *waking_workerfn(void *arg)
+{
+	struct thread_data *waker = (struct thread_data *) arg;
+	struct timeval start, end;
+
+	gettimeofday(&start, NULL);
+
+	waker->nwoken = futex_wake(&futex, nwakes, futex_flag);
+	if (waker->nwoken != nwakes)
+		warnx("couldn't wakeup all tasks (%d/%d)",
+		      waker->nwoken, nwakes);
+
+	gettimeofday(&end, NULL);
+	timersub(&end, &start, &waker->runtime);
+
+	pthread_exit(NULL);
+	return NULL;
+}
+
+static void wakeup_threads(struct thread_data *td, pthread_attr_t thread_attr)
+{
+	unsigned int i;
+
+	pthread_attr_setdetachstate(&thread_attr, PTHREAD_CREATE_JOINABLE);
+
+	/* create and block all threads */
+	for (i = 0; i < nwaking_threads; i++) {
+		/*
+		 * Thread creation order will impact per-thread latency
+		 * as it will affect the order to acquire the hb spinlock.
+		 * For now let the scheduler decide.
+		 */
+		if (pthread_create(&td[i].worker, &thread_attr,
+				   waking_workerfn, (void *)&td[i]))
+			err(EXIT_FAILURE, "pthread_create");
+	}
+
+	for (i = 0; i < nwaking_threads; i++)
+		if (pthread_join(td[i].worker, NULL))
+			err(EXIT_FAILURE, "pthread_join");
+}
+
+static void *blocked_workerfn(void *arg __maybe_unused)
+{
+	pthread_mutex_lock(&thread_lock);
+	threads_starting--;
+	if (!threads_starting)
+		pthread_cond_signal(&thread_parent);
+	pthread_cond_wait(&thread_worker, &thread_lock);
+	pthread_mutex_unlock(&thread_lock);
+
+	while (1) { /* handle spurious wakeups */
+		if (futex_wait(&futex, 0, NULL, futex_flag) != EINTR)
+			break;
+	}
+
+	pthread_exit(NULL);
+	return NULL;
+}
+
+static void block_threads(pthread_t *w, pthread_attr_t thread_attr)
+{
+	cpu_set_t cpu;
+	unsigned int i;
+
+	threads_starting = nblocked_threads;
+
+	/* create and block all threads */
+	for (i = 0; i < nblocked_threads; i++) {
+		CPU_ZERO(&cpu);
+		CPU_SET(i % ncpus, &cpu);
+
+		if (pthread_attr_setaffinity_np(&thread_attr, sizeof(cpu_set_t), &cpu))
+			err(EXIT_FAILURE, "pthread_attr_setaffinity_np");
+
+		if (pthread_create(&w[i], &thread_attr, blocked_workerfn, NULL))
+			err(EXIT_FAILURE, "pthread_create");
+	}
+}
+
+static void print_run(struct thread_data *waking_worker, unsigned int run_num)
+{
+	unsigned int i, wakeup_avg;
+	double waketime_avg, waketime_stddev;
+	struct stats __waketime_stats, __wakeup_stats;
+
+	init_stats(&__wakeup_stats);
+	init_stats(&__waketime_stats);
+
+	for (i = 0; i < nwaking_threads; i++) {
+		update_stats(&__waketime_stats, waking_worker[i].runtime.tv_usec);
+		update_stats(&__wakeup_stats, waking_worker[i].nwoken);
+	}
+
+	waketime_avg = avg_stats(&__waketime_stats);
+	waketime_stddev = stddev_stats(&__waketime_stats);
+	wakeup_avg = avg_stats(&__wakeup_stats);
+
+	printf("[Run %d]: Avg per-thread latency (waking %d/%d threads) "
+	       "in %.4f ms (+-%.2f%%)\n", run_num + 1, wakeup_avg,
+	       nblocked_threads, waketime_avg/1e3,
+	       rel_stddev_stats(waketime_stddev, waketime_avg));
+}
+
+static void print_summary(void)
+{
+	unsigned int wakeup_avg;
+	double waketime_avg, waketime_stddev;
+
+	waketime_avg = avg_stats(&waketime_stats);
+	waketime_stddev = stddev_stats(&waketime_stats);
+	wakeup_avg = avg_stats(&wakeup_stats);
+
+	printf("Avg per-thread latency (waking %d/%d threads) in %.4f ms (+-%.2f%%)\n",
+	       wakeup_avg,
+	       nblocked_threads,
+	       waketime_avg/1e3,
+	       rel_stddev_stats(waketime_stddev, waketime_avg));
+}
+
+
+static void do_run_stats(struct thread_data *waking_worker)
+{
+	unsigned int i;
+
+	for (i = 0; i < nwaking_threads; i++) {
+		update_stats(&waketime_stats, waking_worker[i].runtime.tv_usec);
+		update_stats(&wakeup_stats, waking_worker[i].nwoken);
+	}
+
+}
+
+static void toggle_done(int sig __maybe_unused,
+			siginfo_t *info __maybe_unused,
+			void *uc __maybe_unused)
+{
+	done = true;
+}
+
+int bench_futex_wake_parallel(int argc, const char **argv,
+			      const char *prefix __maybe_unused)
+{
+	int ret = 0;
+	unsigned int i, j;
+	struct sigaction act;
+	pthread_attr_t thread_attr;
+	struct thread_data *waking_worker;
+
+	argc = parse_options(argc, argv, options,
+			     bench_futex_wake_parallel_usage, 0);
+	if (argc) {
+		usage_with_options(bench_futex_wake_parallel_usage, options);
+		exit(EXIT_FAILURE);
+	}
+
+	sigfillset(&act.sa_mask);
+	act.sa_sigaction = toggle_done;
+	sigaction(SIGINT, &act, NULL);
+
+	ncpus = sysconf(_SC_NPROCESSORS_ONLN);
+	if (!nblocked_threads)
+		nblocked_threads = ncpus;
+
+	/* some sanity checks */
+	if (nwaking_threads > nblocked_threads || !nwaking_threads)
+		nwaking_threads = nblocked_threads;
+
+	if (nblocked_threads % nwaking_threads)
+		errx(EXIT_FAILURE, "Must be perfectly divisible");
+	/*
+	 * Each thread will wakeup nwakes tasks in
+	 * a single futex_wait call.
+	 */
+	nwakes = nblocked_threads/nwaking_threads;
+
+	blocked_worker = calloc(nblocked_threads, sizeof(*blocked_worker));
+	if (!blocked_worker)
+		err(EXIT_FAILURE, "calloc");
+
+	if (!fshared)
+		futex_flag = FUTEX_PRIVATE_FLAG;
+
+	printf("Run summary [PID %d]: blocking on %d threads (at [%s] "
+	       "futex %p), %d threads waking up %d at a time.\n\n",
+	       getpid(), nblocked_threads, fshared ? "shared":"private",
+	       &futex, nwaking_threads, nwakes);
+
+	init_stats(&wakeup_stats);
+	init_stats(&waketime_stats);
+
+	pthread_attr_init(&thread_attr);
+	pthread_mutex_init(&thread_lock, NULL);
+	pthread_cond_init(&thread_parent, NULL);
+	pthread_cond_init(&thread_worker, NULL);
+
+	for (j = 0; j < bench_repeat && !done; j++) {
+		waking_worker = calloc(nwaking_threads, sizeof(*waking_worker));
+		if (!waking_worker)
+			err(EXIT_FAILURE, "calloc");
+
+		/* create, launch & block all threads */
+		block_threads(blocked_worker, thread_attr);
+
+		/* make sure all threads are already blocked */
+		pthread_mutex_lock(&thread_lock);
+		while (threads_starting)
+			pthread_cond_wait(&thread_parent, &thread_lock);
+		pthread_cond_broadcast(&thread_worker);
+		pthread_mutex_unlock(&thread_lock);
+
+		usleep(100000);
+
+		/* Ok, all threads are patiently blocked, start waking folks up */
+		wakeup_threads(waking_worker, thread_attr);
+
+		for (i = 0; i < nblocked_threads; i++) {
+			ret = pthread_join(blocked_worker[i], NULL);
+			if (ret)
+				err(EXIT_FAILURE, "pthread_join");
+		}
+
+		do_run_stats(waking_worker);
+		if (!silent)
+			print_run(waking_worker, j);
+
+		free(waking_worker);
+	}
+
+	/* cleanup & report results */
+	pthread_cond_destroy(&thread_parent);
+	pthread_cond_destroy(&thread_worker);
+	pthread_mutex_destroy(&thread_lock);
+	pthread_attr_destroy(&thread_attr);
+
+	print_summary();
+
+	free(blocked_worker);
+	return ret;
+}
diff --git a/kernel/tools/perf/bench/futex-wake.c b/kernel/tools/perf/bench/futex-wake.c
index 929f762be..e5e41d3bd 100644
--- a/kernel/tools/perf/bench/futex-wake.c
+++ b/kernel/tools/perf/bench/futex-wake.c
@@ -60,7 +60,12 @@ static void *workerfn(void *arg __maybe_unused)
 	pthread_cond_wait(&thread_worker, &thread_lock);
 	pthread_mutex_unlock(&thread_lock);
 
-	futex_wait(&futex1, 0, NULL, futex_flag);
+	while (1) {
+		if (futex_wait(&futex1, 0, NULL, futex_flag) != EINTR)
+			break;
+	}
+
+	pthread_exit(NULL);
 	return NULL;
 }
 
diff --git a/kernel/tools/perf/bench/futex.h b/kernel/tools/perf/bench/futex.h
index 7ed22ff1e..d44de9f44 100644
--- a/kernel/tools/perf/bench/futex.h
+++ b/kernel/tools/perf/bench/futex.h
@@ -56,6 +56,26 @@ futex_wake(u_int32_t *uaddr, int nr_wake, int opflags)
 }
 
 /**
+ * futex_lock_pi() - block on uaddr as a PI mutex
+ * @detect:	whether (1) or not (0) to perform deadlock detection
+ */
+static inline int
+futex_lock_pi(u_int32_t *uaddr, struct timespec *timeout, int detect,
+	      int opflags)
+{
+	return futex(uaddr, FUTEX_LOCK_PI, detect, timeout, NULL, 0, opflags);
+}
+
+/**
+ * futex_unlock_pi() - release uaddr as a PI mutex, waking the top waiter
+ */
+static inline int
+futex_unlock_pi(u_int32_t *uaddr, int opflags)
+{
+	return futex(uaddr, FUTEX_UNLOCK_PI, 0, NULL, NULL, 0, opflags);
+}
+
+/**
 * futex_cmp_requeue() - requeue tasks from uaddr to uaddr2
 * @nr_wake:        wake up to this many tasks
 * @nr_requeue:        requeue up to this many tasks
diff --git a/kernel/tools/perf/bench/mem-functions.c b/kernel/tools/perf/bench/mem-functions.c
new file mode 100644
index 000000000..9419b9442
--- /dev/null
+++ b/kernel/tools/perf/bench/mem-functions.c
@@ -0,0 +1,379 @@
+/*
+ * mem-memcpy.c
+ *
+ * Simple memcpy() and memset() benchmarks
+ *
+ * Written by Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/parse-options.h"
+#include "../util/header.h"
+#include "../util/cloexec.h"
+#include "bench.h"
+#include "mem-memcpy-arch.h"
+#include "mem-memset-arch.h"
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/time.h>
+#include <errno.h>
+
+#define K 1024
+
+static const char	*size_str	= "1MB";
+static const char	*function_str	= "all";
+static int		nr_loops	= 1;
+static bool		use_cycles;
+static int		cycles_fd;
+
+static const struct option options[] = {
+	OPT_STRING('s', "size", &size_str, "1MB",
+		    "Specify the size of the memory buffers. "
+		    "Available units: B, KB, MB, GB and TB (case insensitive)"),
+
+	OPT_STRING('f', "function", &function_str, "all",
+		    "Specify the function to run, \"all\" runs all available functions, \"help\" lists them"),
+
+	OPT_INTEGER('l', "nr_loops", &nr_loops,
+		    "Specify the number of loops to run. (default: 1)"),
+
+	OPT_BOOLEAN('c', "cycles", &use_cycles,
+		    "Use a cycles event instead of gettimeofday() to measure performance"),
+
+	OPT_END()
+};
+
+typedef void *(*memcpy_t)(void *, const void *, size_t);
+typedef void *(*memset_t)(void *, int, size_t);
+
+struct function {
+	const char *name;
+	const char *desc;
+	union {
+		memcpy_t memcpy;
+		memset_t memset;
+	} fn;
+};
+
+static struct perf_event_attr cycle_attr = {
+	.type		= PERF_TYPE_HARDWARE,
+	.config		= PERF_COUNT_HW_CPU_CYCLES
+};
+
+static void init_cycles(void)
+{
+	cycles_fd = sys_perf_event_open(&cycle_attr, getpid(), -1, -1, perf_event_open_cloexec_flag());
+
+	if (cycles_fd < 0 && errno == ENOSYS)
+		die("No CONFIG_PERF_EVENTS=y kernel support configured?\n");
+	else
+		BUG_ON(cycles_fd < 0);
+}
+
+static u64 get_cycles(void)
+{
+	int ret;
+	u64 clk;
+
+	ret = read(cycles_fd, &clk, sizeof(u64));
+	BUG_ON(ret != sizeof(u64));
+
+	return clk;
+}
+
+static double timeval2double(struct timeval *ts)
+{
+	return (double)ts->tv_sec + (double)ts->tv_usec / (double)1000000;
+}
+
+#define print_bps(x) do {						\
+		if (x < K)						\
+			printf(" %14lf bytes/sec\n", x);		\
+		else if (x < K * K)					\
+			printf(" %14lfd KB/sec\n", x / K);		\
+		else if (x < K * K * K)					\
+			printf(" %14lf MB/sec\n", x / K / K);		\
+		else							\
+			printf(" %14lf GB/sec\n", x / K / K / K);	\
+	} while (0)
+
+struct bench_mem_info {
+	const struct function *functions;
+	u64 (*do_cycles)(const struct function *r, size_t size);
+	double (*do_gettimeofday)(const struct function *r, size_t size);
+	const char *const *usage;
+};
+
+static void __bench_mem_function(struct bench_mem_info *info, int r_idx, size_t size, double size_total)
+{
+	const struct function *r = &info->functions[r_idx];
+	double result_bps = 0.0;
+	u64 result_cycles = 0;
+
+	printf("# function '%s' (%s)\n", r->name, r->desc);
+
+	if (bench_format == BENCH_FORMAT_DEFAULT)
+		printf("# Copying %s bytes ...\n\n", size_str);
+
+	if (use_cycles) {
+		result_cycles = info->do_cycles(r, size);
+	} else {
+		result_bps = info->do_gettimeofday(r, size);
+	}
+
+	switch (bench_format) {
+	case BENCH_FORMAT_DEFAULT:
+		if (use_cycles) {
+			printf(" %14lf cycles/byte\n", (double)result_cycles/size_total);
+		} else {
+			print_bps(result_bps);
+		}
+		break;
+
+	case BENCH_FORMAT_SIMPLE:
+		if (use_cycles) {
+			printf("%lf\n", (double)result_cycles/size_total);
+		} else {
+			printf("%lf\n", result_bps);
+		}
+		break;
+
+	default:
+		BUG_ON(1);
+		break;
+	}
+}
+
+static int bench_mem_common(int argc, const char **argv, struct bench_mem_info *info)
+{
+	int i;
+	size_t size;
+	double size_total;
+
+	argc = parse_options(argc, argv, options, info->usage, 0);
+
+	if (use_cycles)
+		init_cycles();
+
+	size = (size_t)perf_atoll((char *)size_str);
+	size_total = (double)size * nr_loops;
+
+	if ((s64)size <= 0) {
+		fprintf(stderr, "Invalid size:%s\n", size_str);
+		return 1;
+	}
+
+	if (!strncmp(function_str, "all", 3)) {
+		for (i = 0; info->functions[i].name; i++)
+			__bench_mem_function(info, i, size, size_total);
+		return 0;
+	}
+
+	for (i = 0; info->functions[i].name; i++) {
+		if (!strcmp(info->functions[i].name, function_str))
+			break;
+	}
+	if (!info->functions[i].name) {
+		if (strcmp(function_str, "help") && strcmp(function_str, "h"))
+			printf("Unknown function: %s\n", function_str);
+		printf("Available functions:\n");
+		for (i = 0; info->functions[i].name; i++) {
+			printf("\t%s ... %s\n",
+			       info->functions[i].name, info->functions[i].desc);
+		}
+		return 1;
+	}
+
+	__bench_mem_function(info, i, size, size_total);
+
+	return 0;
+}
+
+static void memcpy_alloc_mem(void **dst, void **src, size_t size)
+{
+	*dst = zalloc(size);
+	if (!*dst)
+		die("memory allocation failed - maybe size is too large?\n");
+
+	*src = zalloc(size);
+	if (!*src)
+		die("memory allocation failed - maybe size is too large?\n");
+
+	/* Make sure to always prefault zero pages even if MMAP_THRESH is crossed: */
+	memset(*src, 0, size);
+}
+
+static u64 do_memcpy_cycles(const struct function *r, size_t size)
+{
+	u64 cycle_start = 0ULL, cycle_end = 0ULL;
+	void *src = NULL, *dst = NULL;
+	memcpy_t fn = r->fn.memcpy;
+	int i;
+
+	memcpy_alloc_mem(&dst, &src, size);
+
+	/*
+	 * We prefault the freshly allocated memory range here,
+	 * to not measure page fault overhead:
+	 */
+	fn(dst, src, size);
+
+	cycle_start = get_cycles();
+	for (i = 0; i < nr_loops; ++i)
+		fn(dst, src, size);
+	cycle_end = get_cycles();
+
+	free(src);
+	free(dst);
+	return cycle_end - cycle_start;
+}
+
+static double do_memcpy_gettimeofday(const struct function *r, size_t size)
+{
+	struct timeval tv_start, tv_end, tv_diff;
+	memcpy_t fn = r->fn.memcpy;
+	void *src = NULL, *dst = NULL;
+	int i;
+
+	memcpy_alloc_mem(&dst, &src, size);
+
+	/*
+	 * We prefault the freshly allocated memory range here,
+	 * to not measure page fault overhead:
+	 */
+	fn(dst, src, size);
+
+	BUG_ON(gettimeofday(&tv_start, NULL));
+	for (i = 0; i < nr_loops; ++i)
+		fn(dst, src, size);
+	BUG_ON(gettimeofday(&tv_end, NULL));
+
+	timersub(&tv_end, &tv_start, &tv_diff);
+
+	free(src);
+	free(dst);
+
+	return (double)(((double)size * nr_loops) / timeval2double(&tv_diff));
+}
+
+struct function memcpy_functions[] = {
+	{ .name		= "default",
+	  .desc		= "Default memcpy() provided by glibc",
+	  .fn.memcpy	= memcpy },
+
+#ifdef HAVE_ARCH_X86_64_SUPPORT
+# define MEMCPY_FN(_fn, _name, _desc) {.name = _name, .desc = _desc, .fn.memcpy = _fn},
+# include "mem-memcpy-x86-64-asm-def.h"
+# undef MEMCPY_FN
+#endif
+
+	{ .name = NULL, }
+};
+
+static const char * const bench_mem_memcpy_usage[] = {
+	"perf bench mem memcpy <options>",
+	NULL
+};
+
+int bench_mem_memcpy(int argc, const char **argv, const char *prefix __maybe_unused)
+{
+	struct bench_mem_info info = {
+		.functions		= memcpy_functions,
+		.do_cycles		= do_memcpy_cycles,
+		.do_gettimeofday	= do_memcpy_gettimeofday,
+		.usage			= bench_mem_memcpy_usage,
+	};
+
+	return bench_mem_common(argc, argv, &info);
+}
+
+static void memset_alloc_mem(void **dst, size_t size)
+{
+	*dst = zalloc(size);
+	if (!*dst)
+		die("memory allocation failed - maybe size is too large?\n");
+}
+
+static u64 do_memset_cycles(const struct function *r, size_t size)
+{
+	u64 cycle_start = 0ULL, cycle_end = 0ULL;
+	memset_t fn = r->fn.memset;
+	void *dst = NULL;
+	int i;
+
+	memset_alloc_mem(&dst, size);
+
+	/*
+	 * We prefault the freshly allocated memory range here,
+	 * to not measure page fault overhead:
+	 */
+	fn(dst, -1, size);
+
+	cycle_start = get_cycles();
+	for (i = 0; i < nr_loops; ++i)
+		fn(dst, i, size);
+	cycle_end = get_cycles();
+
+	free(dst);
+	return cycle_end - cycle_start;
+}
+
+static double do_memset_gettimeofday(const struct function *r, size_t size)
+{
+	struct timeval tv_start, tv_end, tv_diff;
+	memset_t fn = r->fn.memset;
+	void *dst = NULL;
+	int i;
+
+	memset_alloc_mem(&dst, size);
+
+	/*
+	 * We prefault the freshly allocated memory range here,
+	 * to not measure page fault overhead:
+	 */
+	fn(dst, -1, size);
+
+	BUG_ON(gettimeofday(&tv_start, NULL));
+	for (i = 0; i < nr_loops; ++i)
+		fn(dst, i, size);
+	BUG_ON(gettimeofday(&tv_end, NULL));
+
+	timersub(&tv_end, &tv_start, &tv_diff);
+
+	free(dst);
+	return (double)(((double)size * nr_loops) / timeval2double(&tv_diff));
+}
+
+static const char * const bench_mem_memset_usage[] = {
+	"perf bench mem memset <options>",
+	NULL
+};
+
+static const struct function memset_functions[] = {
+	{ .name		= "default",
+	  .desc		= "Default memset() provided by glibc",
+	  .fn.memset	= memset },
+
+#ifdef HAVE_ARCH_X86_64_SUPPORT
+# define MEMSET_FN(_fn, _name, _desc) { .name = _name, .desc = _desc, .fn.memset = _fn },
+# include "mem-memset-x86-64-asm-def.h"
+# undef MEMSET_FN
+#endif
+
+	{ .name = NULL, }
+};
+
+int bench_mem_memset(int argc, const char **argv, const char *prefix __maybe_unused)
+{
+	struct bench_mem_info info = {
+		.functions		= memset_functions,
+		.do_cycles		= do_memset_cycles,
+		.do_gettimeofday	= do_memset_gettimeofday,
+		.usage			= bench_mem_memset_usage,
+	};
+
+	return bench_mem_common(argc, argv, &info);
+}
diff --git a/kernel/tools/perf/bench/mem-memcpy.c b/kernel/tools/perf/bench/mem-memcpy.c
deleted file mode 100644
index d3dfb7936..000000000
--- a/kernel/tools/perf/bench/mem-memcpy.c
+++ /dev/null
@@ -1,434 +0,0 @@
-/*
- * mem-memcpy.c
- *
- * memcpy: Simple memory copy in various ways
- *
- * Written by Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
- */
-
-#include "../perf.h"
-#include "../util/util.h"
-#include "../util/parse-options.h"
-#include "../util/header.h"
-#include "../util/cloexec.h"
-#include "bench.h"
-#include "mem-memcpy-arch.h"
-#include "mem-memset-arch.h"
-
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include <sys/time.h>
-#include <errno.h>
-
-#define K 1024
-
-static const char	*length_str	= "1MB";
-static const char	*routine	= "default";
-static int		iterations	= 1;
-static bool		use_cycle;
-static int		cycle_fd;
-static bool		only_prefault;
-static bool		no_prefault;
-
-static const struct option options[] = {
-	OPT_STRING('l', "length", &length_str, "1MB",
-		    "Specify length of memory to copy. "
-		    "Available units: B, KB, MB, GB and TB (upper and lower)"),
-	OPT_STRING('r', "routine", &routine, "default",
-		    "Specify routine to copy, \"all\" runs all available routines"),
-	OPT_INTEGER('i', "iterations", &iterations,
-		    "repeat memcpy() invocation this number of times"),
-	OPT_BOOLEAN('c', "cycle", &use_cycle,
-		    "Use cycles event instead of gettimeofday() for measuring"),
-	OPT_BOOLEAN('o', "only-prefault", &only_prefault,
-		    "Show only the result with page faults before memcpy()"),
-	OPT_BOOLEAN('n', "no-prefault", &no_prefault,
-		    "Show only the result without page faults before memcpy()"),
-	OPT_END()
-};
-
-typedef void *(*memcpy_t)(void *, const void *, size_t);
-typedef void *(*memset_t)(void *, int, size_t);
-
-struct routine {
-	const char *name;
-	const char *desc;
-	union {
-		memcpy_t memcpy;
-		memset_t memset;
-	} fn;
-};
-
-struct routine memcpy_routines[] = {
-	{ .name = "default",
-	  .desc = "Default memcpy() provided by glibc",
-	  .fn.memcpy = memcpy },
-#ifdef HAVE_ARCH_X86_64_SUPPORT
-
-#define MEMCPY_FN(_fn, _name, _desc) {.name = _name, .desc = _desc, .fn.memcpy = _fn},
-#include "mem-memcpy-x86-64-asm-def.h"
-#undef MEMCPY_FN
-
-#endif
-
-	{ NULL,
-	  NULL,
-	  {NULL}   }
-};
-
-static const char * const bench_mem_memcpy_usage[] = {
-	"perf bench mem memcpy <options>",
-	NULL
-};
-
-static struct perf_event_attr cycle_attr = {
-	.type		= PERF_TYPE_HARDWARE,
-	.config		= PERF_COUNT_HW_CPU_CYCLES
-};
-
-static void init_cycle(void)
-{
-	cycle_fd = sys_perf_event_open(&cycle_attr, getpid(), -1, -1,
-				       perf_event_open_cloexec_flag());
-
-	if (cycle_fd < 0 && errno == ENOSYS)
-		die("No CONFIG_PERF_EVENTS=y kernel support configured?\n");
-	else
-		BUG_ON(cycle_fd < 0);
-}
-
-static u64 get_cycle(void)
-{
-	int ret;
-	u64 clk;
-
-	ret = read(cycle_fd, &clk, sizeof(u64));
-	BUG_ON(ret != sizeof(u64));
-
-	return clk;
-}
-
-static double timeval2double(struct timeval *ts)
-{
-	return (double)ts->tv_sec +
-		(double)ts->tv_usec / (double)1000000;
-}
-
-#define pf (no_prefault ? 0 : 1)
-
-#define print_bps(x) do {					\
-		if (x < K)					\
-			printf(" %14lf B/Sec", x);		\
-		else if (x < K * K)				\
-			printf(" %14lfd KB/Sec", x / K);	\
-		else if (x < K * K * K)				\
-			printf(" %14lf MB/Sec", x / K / K);	\
-		else						\
-			printf(" %14lf GB/Sec", x / K / K / K); \
-	} while (0)
-
-struct bench_mem_info {
-	const struct routine *routines;
-	u64 (*do_cycle)(const struct routine *r, size_t len, bool prefault);
-	double (*do_gettimeofday)(const struct routine *r, size_t len, bool prefault);
-	const char *const *usage;
-};
-
-static void __bench_mem_routine(struct bench_mem_info *info, int r_idx, size_t len, double totallen)
-{
-	const struct routine *r = &info->routines[r_idx];
-	double result_bps[2];
-	u64 result_cycle[2];
-
-	result_cycle[0] = result_cycle[1] = 0ULL;
-	result_bps[0] = result_bps[1] = 0.0;
-
-	printf("Routine %s (%s)\n", r->name, r->desc);
-
-	if (bench_format == BENCH_FORMAT_DEFAULT)
-		printf("# Copying %s Bytes ...\n\n", length_str);
-
-	if (!only_prefault && !no_prefault) {
-		/* show both of results */
-		if (use_cycle) {
-			result_cycle[0] = info->do_cycle(r, len, false);
-			result_cycle[1] = info->do_cycle(r, len, true);
-		} else {
-			result_bps[0]   = info->do_gettimeofday(r, len, false);
-			result_bps[1]   = info->do_gettimeofday(r, len, true);
-		}
-	} else {
-		if (use_cycle)
-			result_cycle[pf] = info->do_cycle(r, len, only_prefault);
-		else
-			result_bps[pf] = info->do_gettimeofday(r, len, only_prefault);
-	}
-
-	switch (bench_format) {
-	case BENCH_FORMAT_DEFAULT:
-		if (!only_prefault && !no_prefault) {
-			if (use_cycle) {
-				printf(" %14lf Cycle/Byte\n",
-					(double)result_cycle[0]
-					/ totallen);
-				printf(" %14lf Cycle/Byte (with prefault)\n",
-					(double)result_cycle[1]
-					/ totallen);
-			} else {
-				print_bps(result_bps[0]);
-				printf("\n");
-				print_bps(result_bps[1]);
-				printf(" (with prefault)\n");
-			}
-		} else {
-			if (use_cycle) {
-				printf(" %14lf Cycle/Byte",
-					(double)result_cycle[pf]
-					/ totallen);
-			} else
-				print_bps(result_bps[pf]);
-
-			printf("%s\n", only_prefault ? " (with prefault)" : "");
-		}
-		break;
-	case BENCH_FORMAT_SIMPLE:
-		if (!only_prefault && !no_prefault) {
-			if (use_cycle) {
-				printf("%lf %lf\n",
-					(double)result_cycle[0] / totallen,
-					(double)result_cycle[1] / totallen);
-			} else {
-				printf("%lf %lf\n",
-					result_bps[0], result_bps[1]);
-			}
-		} else {
-			if (use_cycle) {
-				printf("%lf\n", (double)result_cycle[pf]
-					/ totallen);
-			} else
-				printf("%lf\n", result_bps[pf]);
-		}
-		break;
-	default:
-		/* reaching this means there's some disaster: */
-		die("unknown format: %d\n", bench_format);
-		break;
-	}
-}
-
-static int bench_mem_common(int argc, const char **argv,
-		     const char *prefix __maybe_unused,
-		     struct bench_mem_info *info)
-{
-	int i;
-	size_t len;
-	double totallen;
-
-	argc = parse_options(argc, argv, options,
-			     info->usage, 0);
-
-	if (no_prefault && only_prefault) {
-		fprintf(stderr, "Invalid options: -o and -n are mutually exclusive\n");
-		return 1;
-	}
-
-	if (use_cycle)
-		init_cycle();
-
-	len = (size_t)perf_atoll((char *)length_str);
-	totallen = (double)len * iterations;
-
-	if ((s64)len <= 0) {
-		fprintf(stderr, "Invalid length:%s\n", length_str);
-		return 1;
-	}
-
-	/* same to without specifying either of prefault and no-prefault */
-	if (only_prefault && no_prefault)
-		only_prefault = no_prefault = false;
-
-	if (!strncmp(routine, "all", 3)) {
-		for (i = 0; info->routines[i].name; i++)
-			__bench_mem_routine(info, i, len, totallen);
-		return 0;
-	}
-
-	for (i = 0; info->routines[i].name; i++) {
-		if (!strcmp(info->routines[i].name, routine))
-			break;
-	}
-	if (!info->routines[i].name) {
-		printf("Unknown routine:%s\n", routine);
-		printf("Available routines...\n");
-		for (i = 0; info->routines[i].name; i++) {
-			printf("\t%s ... %s\n",
-			       info->routines[i].name, info->routines[i].desc);
-		}
-		return 1;
-	}
-
-	__bench_mem_routine(info, i, len, totallen);
-
-	return 0;
-}
-
-static void memcpy_alloc_mem(void **dst, void **src, size_t length)
-{
-	*dst = zalloc(length);
-	if (!*dst)
-		die("memory allocation failed - maybe length is too large?\n");
-
-	*src = zalloc(length);
-	if (!*src)
-		die("memory allocation failed - maybe length is too large?\n");
-	/* Make sure to always replace the zero pages even if MMAP_THRESH is crossed */
-	memset(*src, 0, length);
-}
-
-static u64 do_memcpy_cycle(const struct routine *r, size_t len, bool prefault)
-{
-	u64 cycle_start = 0ULL, cycle_end = 0ULL;
-	void *src = NULL, *dst = NULL;
-	memcpy_t fn = r->fn.memcpy;
-	int i;
-
-	memcpy_alloc_mem(&dst, &src, len);
-
-	if (prefault)
-		fn(dst, src, len);
-
-	cycle_start = get_cycle();
-	for (i = 0; i < iterations; ++i)
-		fn(dst, src, len);
-	cycle_end = get_cycle();
-
-	free(src);
-	free(dst);
-	return cycle_end - cycle_start;
-}
-
-static double do_memcpy_gettimeofday(const struct routine *r, size_t len,
-				     bool prefault)
-{
-	struct timeval tv_start, tv_end, tv_diff;
-	memcpy_t fn = r->fn.memcpy;
-	void *src = NULL, *dst = NULL;
-	int i;
-
-	memcpy_alloc_mem(&dst, &src, len);
-
-	if (prefault)
-		fn(dst, src, len);
-
-	BUG_ON(gettimeofday(&tv_start, NULL));
-	for (i = 0; i < iterations; ++i)
-		fn(dst, src, len);
-	BUG_ON(gettimeofday(&tv_end, NULL));
-
-	timersub(&tv_end, &tv_start, &tv_diff);
-
-	free(src);
-	free(dst);
-	return (double)(((double)len * iterations) / timeval2double(&tv_diff));
-}
-
-int bench_mem_memcpy(int argc, const char **argv,
-		     const char *prefix __maybe_unused)
-{
-	struct bench_mem_info info = {
-		.routines = memcpy_routines,
-		.do_cycle = do_memcpy_cycle,
-		.do_gettimeofday = do_memcpy_gettimeofday,
-		.usage = bench_mem_memcpy_usage,
-	};
-
-	return bench_mem_common(argc, argv, prefix, &info);
-}
-
-static void memset_alloc_mem(void **dst, size_t length)
-{
-	*dst = zalloc(length);
-	if (!*dst)
-		die("memory allocation failed - maybe length is too large?\n");
-}
-
-static u64 do_memset_cycle(const struct routine *r, size_t len, bool prefault)
-{
-	u64 cycle_start = 0ULL, cycle_end = 0ULL;
-	memset_t fn = r->fn.memset;
-	void *dst = NULL;
-	int i;
-
-	memset_alloc_mem(&dst, len);
-
-	if (prefault)
-		fn(dst, -1, len);
-
-	cycle_start = get_cycle();
-	for (i = 0; i < iterations; ++i)
-		fn(dst, i, len);
-	cycle_end = get_cycle();
-
-	free(dst);
-	return cycle_end - cycle_start;
-}
-
-static double do_memset_gettimeofday(const struct routine *r, size_t len,
-				     bool prefault)
-{
-	struct timeval tv_start, tv_end, tv_diff;
-	memset_t fn = r->fn.memset;
-	void *dst = NULL;
-	int i;
-
-	memset_alloc_mem(&dst, len);
-
-	if (prefault)
-		fn(dst, -1, len);
-
-	BUG_ON(gettimeofday(&tv_start, NULL));
-	for (i = 0; i < iterations; ++i)
-		fn(dst, i, len);
-	BUG_ON(gettimeofday(&tv_end, NULL));
-
-	timersub(&tv_end, &tv_start, &tv_diff);
-
-	free(dst);
-	return (double)(((double)len * iterations) / timeval2double(&tv_diff));
-}
-
-static const char * const bench_mem_memset_usage[] = {
-	"perf bench mem memset <options>",
-	NULL
-};
-
-static const struct routine memset_routines[] = {
-	{ .name ="default",
-	  .desc = "Default memset() provided by glibc",
-	  .fn.memset = memset },
-#ifdef HAVE_ARCH_X86_64_SUPPORT
-
-#define MEMSET_FN(_fn, _name, _desc) { .name = _name, .desc = _desc, .fn.memset = _fn },
-#include "mem-memset-x86-64-asm-def.h"
-#undef MEMSET_FN
-
-#endif
-
-	{ .name = NULL,
-	  .desc = NULL,
-	  .fn.memset = NULL   }
-};
-
-int bench_mem_memset(int argc, const char **argv,
-		     const char *prefix __maybe_unused)
-{
-	struct bench_mem_info info = {
-		.routines = memset_routines,
-		.do_cycle = do_memset_cycle,
-		.do_gettimeofday = do_memset_gettimeofday,
-		.usage = bench_mem_memset_usage,
-	};
-
-	return bench_mem_common(argc, argv, prefix, &info);
-}
diff --git a/kernel/tools/perf/bench/numa.c b/kernel/tools/perf/bench/numa.c
index ba5efa471..492df2752 100644
--- a/kernel/tools/perf/bench/numa.c
+++ b/kernel/tools/perf/bench/numa.c
@@ -8,6 +8,7 @@
 #include "../builtin.h"
 #include "../util/util.h"
 #include "../util/parse-options.h"
+#include "../util/cloexec.h"
 
 #include "bench.h"
 
@@ -23,6 +24,7 @@
 #include <pthread.h>
 #include <sys/mman.h>
 #include <sys/time.h>
+#include <sys/resource.h>
 #include <sys/wait.h>
 #include <sys/prctl.h>
 #include <sys/types.h>
@@ -51,6 +53,9 @@ struct thread_data {
 	unsigned int		loops_done;
 	u64			val;
 	u64			runtime_ns;
+	u64			system_time_ns;
+	u64			user_time_ns;
+	double			speed_gbs;
 	pthread_mutex_t		*process_lock;
 };
 
@@ -159,8 +164,8 @@ static const struct option options[] = {
 	OPT_STRING('L', "mb_proc_locked", &p0.mb_proc_locked_str,"MB", "process serialized/locked memory access (MBs), <= process_memory"),
 	OPT_STRING('T', "mb_thread"	, &p0.mb_thread_str,	"MB", "thread  memory (MBs)"),
 
-	OPT_UINTEGER('l', "nr_loops"	, &p0.nr_loops,		"max number of loops to run"),
-	OPT_UINTEGER('s', "nr_secs"	, &p0.nr_secs,		"max number of seconds to run"),
+	OPT_UINTEGER('l', "nr_loops"	, &p0.nr_loops,		"max number of loops to run (default: unlimited)"),
+	OPT_UINTEGER('s', "nr_secs"	, &p0.nr_secs,		"max number of seconds to run (default: 5 secs)"),
 	OPT_UINTEGER('u', "usleep"	, &p0.sleep_usecs,	"usecs to sleep per loop iteration"),
 
 	OPT_BOOLEAN('R', "data_reads"	, &p0.data_reads,	"access the data via writes (can be mixed with -W)"),
@@ -1042,6 +1047,7 @@ static void *worker_thread(void *__tdata)
 	u64 bytes_done;
 	long work_done;
 	u32 l;
+	struct rusage rusage;
 
 	bind_to_cpumask(td->bind_cpumask);
 	bind_to_memnode(td->bind_node);
@@ -1194,6 +1200,13 @@ static void *worker_thread(void *__tdata)
 	timersub(&stop, &start0, &diff);
 	td->runtime_ns = diff.tv_sec * 1000000000ULL;
 	td->runtime_ns += diff.tv_usec * 1000ULL;
+	td->speed_gbs = bytes_done / (td->runtime_ns / 1e9) / 1e9;
+
+	getrusage(RUSAGE_THREAD, &rusage);
+	td->system_time_ns = rusage.ru_stime.tv_sec * 1000000000ULL;
+	td->system_time_ns += rusage.ru_stime.tv_usec * 1000ULL;
+	td->user_time_ns = rusage.ru_utime.tv_sec * 1000000000ULL;
+	td->user_time_ns += rusage.ru_utime.tv_usec * 1000ULL;
 
 	free_data(thread_data, g->p.bytes_thread);
 
@@ -1420,7 +1433,7 @@ static int __bench_numa(const char *name)
 	double runtime_sec_min;
 	int wait_stat;
 	double bytes;
-	int i, t;
+	int i, t, p;
 
 	if (init())
 		return -1;
@@ -1556,6 +1569,24 @@ static int __bench_numa(const char *name)
 	print_res(name, bytes / runtime_sec_max / 1e9,
 		"GB/sec,", "total-speed",	"GB/sec total speed");
 
+	if (g->p.show_details >= 2) {
+		char tname[32];
+		struct thread_data *td;
+		for (p = 0; p < g->p.nr_proc; p++) {
+			for (t = 0; t < g->p.nr_threads; t++) {
+				memset(tname, 0, 32);
+				td = g->threads + p*g->p.nr_threads + t;
+				snprintf(tname, 32, "process%d:thread%d", p, t);
+				print_res(tname, td->speed_gbs,
+					"GB/sec",	"thread-speed", "GB/sec/thread speed");
+				print_res(tname, td->system_time_ns / 1e9,
+					"secs",	"thread-system-time", "system CPU time/thread");
+				print_res(tname, td->user_time_ns / 1e9,
+					"secs",	"thread-user-time", "user CPU time/thread");
+			}
+		}
+	}
+
 	free(pids);
 
 	deinit();
diff --git a/kernel/tools/perf/bench/sched-messaging.c b/kernel/tools/perf/bench/sched-messaging.c
index d7f281c28..d4ff1b539 100644
--- a/kernel/tools/perf/bench/sched-messaging.c
+++ b/kernel/tools/perf/bench/sched-messaging.c
@@ -33,7 +33,7 @@
 #define DATASIZE 100
 
 static bool use_pipes = false;
-static unsigned int loops = 100;
+static unsigned int nr_loops = 100;
 static bool thread_mode = false;
 static unsigned int num_groups = 10;
 
@@ -79,7 +79,7 @@ static void ready(int ready_out, int wakefd)
 		err(EXIT_FAILURE, "poll");
 }
 
-/* Sender sprays loops messages down each file descriptor */
+/* Sender sprays nr_loops messages down each file descriptor */
 static void *sender(struct sender_context *ctx)
 {
 	char data[DATASIZE];
@@ -88,7 +88,7 @@ static void *sender(struct sender_context *ctx)
 	ready(ctx->ready_out, ctx->wakefd);
 
 	/* Now pump to every receiver. */
-	for (i = 0; i < loops; i++) {
+	for (i = 0; i < nr_loops; i++) {
 		for (j = 0; j < ctx->num_fds; j++) {
 			int ret, done = 0;
 
@@ -213,7 +213,7 @@ static unsigned int group(pthread_t *pth,
 		/* Create the pipe between client and server */
 		fdpair(fds);
 
-		ctx->num_packets = num_fds * loops;
+		ctx->num_packets = num_fds * nr_loops;
 		ctx->in_fds[0] = fds[0];
 		ctx->in_fds[1] = fds[1];
 		ctx->ready_out = ready_out;
@@ -250,7 +250,7 @@ static const struct option options[] = {
 	OPT_BOOLEAN('t', "thread", &thread_mode,
 		    "Be multi thread instead of multi process"),
 	OPT_UINTEGER('g', "group", &num_groups, "Specify number of groups"),
-	OPT_UINTEGER('l', "loop", &loops, "Specify number of loops"),
+	OPT_UINTEGER('l', "nr_loops", &nr_loops, "Specify the number of loops to run (default: 100)"),
 	OPT_END()
 };