1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
|
// SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
// Copyright (c) 2021 Google
#include "vmlinux.h"
#include <bpf/bpf_helpers.h>
#include <bpf/bpf_tracing.h>
// This should be in sync with "util/ftrace.h"
#define NUM_BUCKET 22
struct {
__uint(type, BPF_MAP_TYPE_HASH);
__uint(key_size, sizeof(__u64));
__uint(value_size, sizeof(__u64));
__uint(max_entries, 10000);
} functime SEC(".maps");
struct {
__uint(type, BPF_MAP_TYPE_HASH);
__uint(key_size, sizeof(__u32));
__uint(value_size, sizeof(__u8));
__uint(max_entries, 1);
} cpu_filter SEC(".maps");
struct {
__uint(type, BPF_MAP_TYPE_HASH);
__uint(key_size, sizeof(__u32));
__uint(value_size, sizeof(__u8));
__uint(max_entries, 1);
} task_filter SEC(".maps");
struct {
__uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
__uint(key_size, sizeof(__u32));
__uint(value_size, sizeof(__u64));
__uint(max_entries, NUM_BUCKET);
} latency SEC(".maps");
int enabled = 0;
// stats
__s64 total;
__s64 count;
__s64 max;
__s64 min;
const volatile int has_cpu = 0;
const volatile int has_task = 0;
const volatile int use_nsec = 0;
const volatile unsigned int bucket_range;
const volatile unsigned int min_latency;
const volatile unsigned int max_latency;
const volatile unsigned int bucket_num = NUM_BUCKET;
static bool can_record(void)
{
if (has_cpu) {
__u32 cpu = bpf_get_smp_processor_id();
__u8 *ok;
ok = bpf_map_lookup_elem(&cpu_filter, &cpu);
if (!ok)
return false;
}
if (has_task) {
__u32 pid = bpf_get_current_pid_tgid();
__u8 *ok;
ok = bpf_map_lookup_elem(&task_filter, &pid);
if (!ok)
return false;
}
return true;
}
static void update_latency(__s64 delta)
{
__u64 val = delta;
__u32 key = 0;
__u64 *hist;
__u64 cmp_base = use_nsec ? 1 : 1000;
if (delta < 0)
return;
if (bucket_range != 0) {
val = delta / cmp_base;
if (min_latency > 0) {
if (val > min_latency)
val -= min_latency;
else
goto do_lookup;
}
// Less than 1 unit (ms or ns), or, in the future,
// than the min latency desired.
if (val > 0) { // 1st entry: [ 1 unit .. bucket_range units )
key = val / bucket_range + 1;
if (key >= bucket_num)
key = bucket_num - 1;
}
goto do_lookup;
}
// calculate index using delta
for (key = 0; key < (bucket_num - 1); key++) {
if (delta < (cmp_base << key))
break;
}
do_lookup:
hist = bpf_map_lookup_elem(&latency, &key);
if (!hist)
return;
__sync_fetch_and_add(hist, 1);
__sync_fetch_and_add(&total, delta); // always in nsec
__sync_fetch_and_add(&count, 1);
if (delta > max)
max = delta;
if (delta < min)
min = delta;
}
SEC("kprobe/func")
int BPF_PROG(func_begin)
{
__u64 key, now;
if (!enabled || !can_record())
return 0;
key = bpf_get_current_pid_tgid();
now = bpf_ktime_get_ns();
// overwrite timestamp for nested functions
bpf_map_update_elem(&functime, &key, &now, BPF_ANY);
return 0;
}
SEC("kretprobe/func")
int BPF_PROG(func_end)
{
__u64 tid;
__u64 *start;
if (!enabled)
return 0;
tid = bpf_get_current_pid_tgid();
start = bpf_map_lookup_elem(&functime, &tid);
if (start) {
update_latency(bpf_ktime_get_ns() - *start);
bpf_map_delete_elem(&functime, &tid);
}
return 0;
}
SEC("raw_tp")
int BPF_PROG(event_begin)
{
__u64 key, now;
if (!enabled || !can_record())
return 0;
key = bpf_get_current_pid_tgid();
now = bpf_ktime_get_ns();
// overwrite timestamp for nested events
bpf_map_update_elem(&functime, &key, &now, BPF_ANY);
return 0;
}
SEC("raw_tp")
int BPF_PROG(event_end)
{
__u64 tid;
__u64 *start;
if (!enabled)
return 0;
tid = bpf_get_current_pid_tgid();
start = bpf_map_lookup_elem(&functime, &tid);
if (start) {
update_latency(bpf_ktime_get_ns() - *start);
bpf_map_delete_elem(&functime, &tid);
}
return 0;
}
|
