6 files changed, 390 insertions, 5 deletions

diff --git a/tools/perf/util/bpf_skel/lock_contention.bpf.c b/tools/perf/util/bpf_skel/lock_contention.bpf.c
index 69be7a4234e0..96e7d853b9ed 100644
--- a/tools/perf/util/bpf_skel/lock_contention.bpf.c
+++ b/tools/perf/util/bpf_skel/lock_contention.bpf.c
@@ -11,6 +11,12 @@
 /* for collect_lock_syms().  4096 was rejected by the verifier */
 #define MAX_CPUS  1024
 
+/* for collect_zone_lock().  It should be more than the actual zones. */
+#define MAX_ZONES  10
+
+/* for do_lock_delay().  Arbitrarily set to 1 million. */
+#define MAX_LOOP  (1U << 20)
+
 /* lock contention flags from include/trace/events/lock.h */
 #define LCB_F_SPIN	(1U << 0)
 #define LCB_F_READ	(1U << 1)
@@ -146,6 +152,13 @@ struct {
 	__uint(max_entries, 1);
 } slab_caches SEC(".maps");
 
+struct {
+	__uint(type, BPF_MAP_TYPE_HASH);
+	__uint(key_size, sizeof(__u64));
+	__uint(value_size, sizeof(__u64));
+	__uint(max_entries, 1);
+} lock_delays SEC(".maps");
+
 struct rw_semaphore___old {
 	struct task_struct *owner;
 } __attribute__((preserve_access_index));
@@ -176,6 +189,7 @@ const volatile int stack_skip;
 const volatile int lock_owner;
 const volatile int use_cgroup_v2;
 const volatile int max_stack;
+const volatile int lock_delay;
 
 /* determine the key of lock stat */
 const volatile int aggr_mode;
@@ -384,6 +398,35 @@ static inline __u32 check_lock_type(__u64 lock, __u32 flags)
 	return 0;
 }
 
+static inline long delay_callback(__u64 idx, void *arg)
+{
+	__u64 target = *(__u64 *)arg;
+
+	if (target <= bpf_ktime_get_ns())
+		return 1;
+
+	/* just to kill time */
+	(void)bpf_get_prandom_u32();
+
+	return 0;
+}
+
+static inline void do_lock_delay(__u64 duration)
+{
+	__u64 target = bpf_ktime_get_ns() + duration;
+
+	bpf_loop(MAX_LOOP, delay_callback, &target, /*flags=*/0);
+}
+
+static inline void check_lock_delay(__u64 lock)
+{
+	__u64 *delay;
+
+	delay = bpf_map_lookup_elem(&lock_delays, &lock);
+	if (delay)
+		do_lock_delay(*delay);
+}
+
 static inline struct tstamp_data *get_tstamp_elem(__u32 flags)
 {
 	__u32 pid;
@@ -793,6 +836,9 @@ found:
 	update_contention_data(data, duration, 1);
 
 out:
+	if (lock_delay)
+		check_lock_delay(pelem->lock);
+
 	pelem->lock = 0;
 	if (need_delete)
 		bpf_map_delete_elem(&tstamp, &pid);
@@ -801,6 +847,11 @@ out:
 
 extern struct rq runqueues __ksym;
 
+const volatile __u64 contig_page_data_addr;
+const volatile __u64 node_data_addr;
+const volatile int nr_nodes;
+const volatile int sizeof_zone;
+
 struct rq___old {
 	raw_spinlock_t lock;
 } __attribute__((preserve_access_index));
@@ -809,6 +860,59 @@ struct rq___new {
 	raw_spinlock_t __lock;
 } __attribute__((preserve_access_index));
 
+static void collect_zone_lock(void)
+{
+	__u64 nr_zones, zone_off;
+	__u64 lock_addr, lock_off;
+	__u32 lock_flag = LOCK_CLASS_ZONE_LOCK;
+
+	zone_off = offsetof(struct pglist_data, node_zones);
+	lock_off = offsetof(struct zone, lock);
+
+	if (contig_page_data_addr) {
+		struct pglist_data *contig_page_data;
+
+		contig_page_data = (void *)(long)contig_page_data_addr;
+		nr_zones = BPF_CORE_READ(contig_page_data, nr_zones);
+
+		for (int i = 0; i < MAX_ZONES; i++) {
+			__u64 zone_addr;
+
+			if (i >= nr_zones)
+				break;
+
+			zone_addr = contig_page_data_addr + (sizeof_zone * i) + zone_off;
+			lock_addr = zone_addr + lock_off;
+
+			bpf_map_update_elem(&lock_syms, &lock_addr, &lock_flag, BPF_ANY);
+		}
+	} else if (nr_nodes > 0) {
+		struct pglist_data **node_data = (void *)(long)node_data_addr;
+
+		for (int i = 0; i < nr_nodes; i++) {
+			struct pglist_data *pgdat = NULL;
+			int err;
+
+			err = bpf_core_read(&pgdat, sizeof(pgdat), &node_data[i]);
+			if (err < 0 || pgdat == NULL)
+				break;
+
+			nr_zones = BPF_CORE_READ(pgdat, nr_zones);
+			for (int k = 0; k < MAX_ZONES; k++) {
+				__u64 zone_addr;
+
+				if (k >= nr_zones)
+					break;
+
+				zone_addr = (__u64)(void *)pgdat + (sizeof_zone * k) + zone_off;
+				lock_addr = zone_addr + lock_off;
+
+				bpf_map_update_elem(&lock_syms, &lock_addr, &lock_flag, BPF_ANY);
+			}
+		}
+	}
+}
+
 SEC("raw_tp/bpf_test_finish")
 int BPF_PROG(collect_lock_syms)
 {
@@ -830,6 +934,9 @@ int BPF_PROG(collect_lock_syms)
 		lock_flag = LOCK_CLASS_RQLOCK;
 		bpf_map_update_elem(&lock_syms, &lock_addr, &lock_flag, BPF_ANY);
 	}
+
+	collect_zone_lock();
+
 	return 0;
 }
 
diff --git a/tools/perf/util/bpf_skel/lock_data.h b/tools/perf/util/bpf_skel/lock_data.h
index 15f5743bd409..28c5e5aced7f 100644
--- a/tools/perf/util/bpf_skel/lock_data.h
+++ b/tools/perf/util/bpf_skel/lock_data.h
@@ -67,6 +67,7 @@ enum lock_aggr_mode {
 enum lock_class_sym {
 	LOCK_CLASS_NONE,
 	LOCK_CLASS_RQLOCK,
+	LOCK_CLASS_ZONE_LOCK,
 };
 
 struct slab_cache_data {
diff --git a/tools/perf/util/bpf_skel/off_cpu.bpf.c b/tools/perf/util/bpf_skel/off_cpu.bpf.c
index c152116df72f..72763bb8d1de 100644
--- a/tools/perf/util/bpf_skel/off_cpu.bpf.c
+++ b/tools/perf/util/bpf_skel/off_cpu.bpf.c
@@ -18,10 +18,19 @@
 #define MAX_STACKS   32
 #define MAX_ENTRIES  102400
 
+#define MAX_CPUS  4096
+#define MAX_OFFCPU_LEN 37
+
+// We have a 'struct stack' in vmlinux.h when building with GEN_VMLINUX_H=1
+struct __stack {
+	u64 array[MAX_STACKS];
+};
+
 struct tstamp_data {
 	__u32 stack_id;
 	__u32 state;
 	__u64 timestamp;
+	struct __stack stack;
 };
 
 struct offcpu_key {
@@ -39,6 +48,24 @@ struct {
 	__uint(max_entries, MAX_ENTRIES);
 } stacks SEC(".maps");
 
+struct offcpu_data {
+	u64 array[MAX_OFFCPU_LEN];
+};
+
+struct {
+	__uint(type, BPF_MAP_TYPE_PERF_EVENT_ARRAY);
+	__uint(key_size, sizeof(int));
+	__uint(value_size, sizeof(int));
+	__uint(max_entries, MAX_CPUS);
+} offcpu_output SEC(".maps");
+
+struct {
+	__uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
+	__uint(key_size, sizeof(__u32));
+	__uint(value_size, sizeof(struct offcpu_data));
+	__uint(max_entries, 1);
+} offcpu_payload SEC(".maps");
+
 struct {
 	__uint(type, BPF_MAP_TYPE_TASK_STORAGE);
 	__uint(map_flags, BPF_F_NO_PREALLOC);
@@ -97,6 +124,8 @@ const volatile bool uses_cgroup_v1 = false;
 
 int perf_subsys_id = -1;
 
+__u64 offcpu_thresh_ns;
+
 /*
  * Old kernel used to call it task_struct->state and now it's '__state'.
  * Use BPF CO-RE "ignored suffix rule" to deal with it like below:
@@ -183,6 +212,47 @@ static inline int can_record(struct task_struct *t, int state)
 	return 1;
 }
 
+static inline int copy_stack(struct __stack *from, struct offcpu_data *to, int n)
+{
+	int len = 0;
+
+	for (int i = 0; i < MAX_STACKS && from->array[i]; ++i, ++len)
+		to->array[n + 2 + i] = from->array[i];
+
+	return len;
+}
+
+/**
+ * off_cpu_dump - dump off-cpu samples to ring buffer
+ * @data: payload for dumping off-cpu samples
+ * @key: off-cpu data
+ * @stack: stack trace of the task before being scheduled out
+ *
+ * If the threshold of off-cpu time is reached, acquire tid, period, callchain, and cgroup id
+ * information of the task, and dump it as a raw sample to perf ring buffer
+ */
+static int off_cpu_dump(void *ctx, struct offcpu_data *data, struct offcpu_key *key,
+			struct __stack *stack, __u64 delta)
+{
+	int n = 0, len = 0;
+
+	data->array[n++] = (u64)key->tgid << 32 | key->pid;
+	data->array[n++] = delta;
+
+	/* data->array[n] is callchain->nr (updated later) */
+	data->array[n + 1] = PERF_CONTEXT_USER;
+	data->array[n + 2] = 0;
+	len = copy_stack(stack, data, n);
+
+	/* update length of callchain */
+	data->array[n] = len + 1;
+	n += len + 2;
+
+	data->array[n++] = key->cgroup_id;
+
+	return bpf_perf_event_output(ctx, &offcpu_output, BPF_F_CURRENT_CPU, data, n * sizeof(u64));
+}
+
 static int off_cpu_stat(u64 *ctx, struct task_struct *prev,
 			struct task_struct *next, int state)
 {
@@ -207,6 +277,16 @@ static int off_cpu_stat(u64 *ctx, struct task_struct *prev,
 	pelem->state = state;
 	pelem->stack_id = stack_id;
 
+	/*
+	 * If stacks are successfully collected by bpf_get_stackid(), collect them once more
+	 * in task_storage for direct off-cpu sample dumping
+	 */
+	if (stack_id > 0 && bpf_get_stack(ctx, &pelem->stack, MAX_STACKS * sizeof(u64), BPF_F_USER_STACK)) {
+		/*
+		 * This empty if block is used to avoid 'result unused warning' from bpf_get_stack().
+		 * If the collection fails, continue with the logic for the next task.
+		 */
+	}
 next:
 	pelem = bpf_task_storage_get(&tstamp, next, NULL, 0);
 
@@ -221,11 +301,19 @@ next:
 		__u64 delta = ts - pelem->timestamp;
 		__u64 *total;
 
-		total = bpf_map_lookup_elem(&off_cpu, &key);
-		if (total)
-			*total += delta;
-		else
-			bpf_map_update_elem(&off_cpu, &key, &delta, BPF_ANY);
+		if (delta >= offcpu_thresh_ns) {
+			int zero = 0;
+			struct offcpu_data *data = bpf_map_lookup_elem(&offcpu_payload, &zero);
+
+			if (data)
+				off_cpu_dump(ctx, data, &key, &pelem->stack, delta);
+		} else {
+			total = bpf_map_lookup_elem(&off_cpu, &key);
+			if (total)
+				*total += delta;
+			else
+				bpf_map_update_elem(&off_cpu, &key, &delta, BPF_ANY);
+		}
 
 		/* prevent to reuse the timestamp later */
 		pelem->timestamp = 0;
diff --git a/tools/perf/util/bpf_skel/syscall_summary.bpf.c b/tools/perf/util/bpf_skel/syscall_summary.bpf.c
new file mode 100644
index 000000000000..1bcd066a5199
--- /dev/null
+++ b/tools/perf/util/bpf_skel/syscall_summary.bpf.c
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Trace raw_syscalls tracepoints to collect system call statistics.
+ */
+
+#include "vmlinux.h"
+#include "syscall_summary.h"
+
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+#include <bpf/bpf_core_read.h>
+
+/* This is to calculate a delta between sys-enter and sys-exit for each thread */
+struct syscall_trace {
+	int nr; /* syscall number is only available at sys-enter */
+	int unused;
+	u64 timestamp;
+};
+
+#define MAX_ENTRIES	(128 * 1024)
+
+struct syscall_trace_map {
+	__uint(type, BPF_MAP_TYPE_HASH);
+	__type(key, int); /* tid */
+	__type(value, struct syscall_trace);
+	__uint(max_entries, MAX_ENTRIES);
+} syscall_trace_map SEC(".maps");
+
+struct syscall_stats_map {
+	__uint(type, BPF_MAP_TYPE_HASH);
+	__type(key, struct syscall_key);
+	__type(value, struct syscall_stats);
+	__uint(max_entries, MAX_ENTRIES);
+} syscall_stats_map SEC(".maps");
+
+int enabled; /* controlled from userspace */
+
+const volatile enum syscall_aggr_mode aggr_mode;
+const volatile int use_cgroup_v2;
+
+int perf_subsys_id = -1;
+
+static inline __u64 get_current_cgroup_id(void)
+{
+	struct task_struct *task;
+	struct cgroup *cgrp;
+
+	if (use_cgroup_v2)
+		return bpf_get_current_cgroup_id();
+
+	task = bpf_get_current_task_btf();
+
+	if (perf_subsys_id == -1) {
+#if __has_builtin(__builtin_preserve_enum_value)
+		perf_subsys_id = bpf_core_enum_value(enum cgroup_subsys_id,
+						     perf_event_cgrp_id);
+#else
+		perf_subsys_id = perf_event_cgrp_id;
+#endif
+	}
+
+	cgrp = BPF_CORE_READ(task, cgroups, subsys[perf_subsys_id], cgroup);
+	return BPF_CORE_READ(cgrp, kn, id);
+}
+
+static void update_stats(int cpu_or_tid, u64 cgroup_id, int nr, s64 duration,
+			 long ret)
+{
+	struct syscall_key key = {
+		.cpu_or_tid = cpu_or_tid,
+		.cgroup = cgroup_id,
+		.nr = nr,
+	};
+	struct syscall_stats *stats;
+
+	stats = bpf_map_lookup_elem(&syscall_stats_map, &key);
+	if (stats == NULL) {
+		struct syscall_stats zero = {};
+
+		bpf_map_update_elem(&syscall_stats_map, &key, &zero, BPF_NOEXIST);
+		stats = bpf_map_lookup_elem(&syscall_stats_map, &key);
+		if (stats == NULL)
+			return;
+	}
+
+	__sync_fetch_and_add(&stats->count, 1);
+	if (ret < 0)
+		__sync_fetch_and_add(&stats->error, 1);
+
+	if (duration > 0) {
+		__sync_fetch_and_add(&stats->total_time, duration);
+		__sync_fetch_and_add(&stats->squared_sum, duration * duration);
+		if (stats->max_time < duration)
+			stats->max_time = duration;
+		if (stats->min_time > duration || stats->min_time == 0)
+			stats->min_time = duration;
+	}
+
+	return;
+}
+
+SEC("tp_btf/sys_enter")
+int sys_enter(u64 *ctx)
+{
+	int tid;
+	struct syscall_trace st;
+
+	if (!enabled)
+		return 0;
+
+	st.nr = ctx[1]; /* syscall number */
+	st.unused = 0;
+	st.timestamp = bpf_ktime_get_ns();
+
+	tid = bpf_get_current_pid_tgid();
+	bpf_map_update_elem(&syscall_trace_map, &tid, &st, BPF_ANY);
+
+	return 0;
+}
+
+SEC("tp_btf/sys_exit")
+int sys_exit(u64 *ctx)
+{
+	int tid;
+	int key = 0;
+	u64 cgroup = 0;
+	long ret = ctx[1]; /* return value of the syscall */
+	struct syscall_trace *st;
+	s64 delta;
+
+	if (!enabled)
+		return 0;
+
+	tid = bpf_get_current_pid_tgid();
+	st = bpf_map_lookup_elem(&syscall_trace_map, &tid);
+	if (st == NULL)
+		return 0;
+
+	if (aggr_mode == SYSCALL_AGGR_THREAD)
+		key = tid;
+	else if (aggr_mode == SYSCALL_AGGR_CGROUP)
+		cgroup = get_current_cgroup_id();
+	else
+		key = bpf_get_smp_processor_id();
+
+	delta = bpf_ktime_get_ns() - st->timestamp;
+	update_stats(key, cgroup, st->nr, delta, ret);
+
+	bpf_map_delete_elem(&syscall_trace_map, &tid);
+	return 0;
+}
+
+char _license[] SEC("license") = "GPL";
diff --git a/tools/perf/util/bpf_skel/syscall_summary.h b/tools/perf/util/bpf_skel/syscall_summary.h
new file mode 100644
index 000000000000..72ccccb45925
--- /dev/null
+++ b/tools/perf/util/bpf_skel/syscall_summary.h
@@ -0,0 +1,27 @@
+// SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+/* Data structures shared between BPF and tools. */
+#ifndef UTIL_BPF_SKEL_SYSCALL_SUMMARY_H
+#define UTIL_BPF_SKEL_SYSCALL_SUMMARY_H
+
+enum syscall_aggr_mode {
+	SYSCALL_AGGR_THREAD,
+	SYSCALL_AGGR_CPU,
+	SYSCALL_AGGR_CGROUP,
+};
+
+struct syscall_key {
+	u64 cgroup;
+	int cpu_or_tid;
+	int nr;
+};
+
+struct syscall_stats {
+	u64 total_time;
+	u64 squared_sum;
+	u64 max_time;
+	u64 min_time;
+	u32 count;
+	u32 error;
+};
+
+#endif /* UTIL_BPF_SKEL_SYSCALL_SUMMARY_H */
diff --git a/tools/perf/util/bpf_skel/vmlinux/vmlinux.h b/tools/perf/util/bpf_skel/vmlinux/vmlinux.h
index 7b81d3173917..a59ce912be18 100644
--- a/tools/perf/util/bpf_skel/vmlinux/vmlinux.h
+++ b/tools/perf/util/bpf_skel/vmlinux/vmlinux.h
@@ -203,4 +203,13 @@ struct bpf_iter__kmem_cache {
 	struct kmem_cache *s;
 } __attribute__((preserve_access_index));
 
+struct zone {
+	spinlock_t lock;
+} __attribute__((preserve_access_index));
+
+struct pglist_data {
+	struct zone node_zones[6]; /* value for all possible config */
+	int nr_zones;
+} __attribute__((preserve_access_index));
+
 #endif // __VMLINUX_H