perf性能分析(5) -- linux perf 工具介绍
1. perf 介绍
perf及子命令可以测量/记录系统性能，可以记录的性能数据项繁多。包括CPU/PMU等硬件数据，以及software counter/tracepoint等系统内核采集的数据。可以关注的几类：
CPU/PMU(Performance Monitoring Unit)数据。包括:dTLB,iTLB,cache计数以及miss计数；branch及branch miss计数。memory延时、阻塞；bus延时、阻塞；front end/back end阻塞；virtual memory相关:TLB相关。pipeline相关。
查看perf命令及子命令帮助信息：
1
2
3
4
5
6
7
8
9
man perf
man perf-top
man perf-stat
man perf-record
man perf-report
查看perf所有子命令:
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
$ perf
usage: perf [--version] [--help] [OPTIONS] COMMAND [ARGS]
The most commonly used perf commands are:
annotate Read perf.data (created by perf record) and display annotated code
archive Create archive with object files with build-ids found in perf.data file
bench General framework for benchmark suites
buildid-cache Manage build-id cache.
buildid-list List the buildids in a perf.data file
c2c Shared Data C2C/HITM Analyzer.
config Get and set variables in a configuration file.
daemon Run record sessions on background
data Data file related processing
diff Read perf.data files and display the differential profile
evlist List the event names in a perf.data file
ftrace simple wrapper for kernel's ftrace functionality
inject Filter to augment the events stream with additional information
iostat Show I/O performance metrics
kallsyms Searches running kernel for symbols
kvm Tool to trace/measure kvm guest os
list List all symbolic event types
mem Profile memory accesses
record Run a command and record its profile into perf.data
report Read perf.data (created by perf record) and display the profile
script Read perf.data (created by perf record) and display trace output
stat Run a command and gather performance counter statistics
test Runs sanity tests.
top System profiling tool.
version display the version of perf binary
probe Define new dynamic tracepoints
See 'perf help COMMAND' for more information on a specific command.
查看子命令的帮助信息，如perf stat：
1
2
3
4
5
6
7
8
9
10
$ perf stat -h
Usage: perf stat [<options>] [<command>]
-a, --all-cpus system-wide collection from all CPUs
-A, --no-aggr disable aggregation across CPUs or PMUs
-B, --big-num print large numbers with thousands' separators
-C, --cpu <cpu> list of cpus to monitor in system-wide
-D, --delay <n> ms to wait before starting measurement after program start (-1: start with events disabled)
# ......
2. Events
perf 记录的性能数据项，称为events。主要分为软件 Events 和硬件 Events。
软件 Events 比如有：context-switchs, minor-fault等等。 硬件 Events 主要记录micro-architecture相关性能数据，由CPU/PMU提供。如果硬件没有提供，该对应该event不可用。
使用 perf list 查看 perf 支持的 events：
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
$ perf list
List of pre-defined events (to be used in -e or -M):
branch-instructions OR branches [Hardware event]
branch-misses [Hardware event]
bus-cycles [Hardware event]
cache-misses [Hardware event]
cache-references [Hardware event]
cpu-cycles OR cycles [Hardware event]
instructions [Hardware event]
ref-cycles [Hardware event]
alignment-faults [Software event]
bpf-output [Software event]
cgroup-switches [Software event]
context-switches OR cs [Software event]
cpu-clock [Software event]
cpu-migrations OR migrations [Software event]
dummy [Software event]
emulation-faults [Software event]
major-faults [Software event]
minor-faults [Software event]
page-faults OR faults [Software event]
task-clock [Software event]
duration_time [Tool event]
user_time [Tool event]
system_time [Tool event]
cpu:
L1-dcache-loads OR cpu/L1-dcache-loads/
L1-dcache-load-misses OR cpu/L1-dcache-load-misses/
L1-dcache-stores OR cpu/L1-dcache-stores/
L1-icache-load-misses OR cpu/L1-icache-load-misses/
LLC-loads OR cpu/LLC-loads/
LLC-load-misses OR cpu/LLC-load-misses/
LLC-stores OR cpu/LLC-stores/
LLC-store-misses OR cpu/LLC-store-misses/
dTLB-loads OR cpu/dTLB-loads/
dTLB-load-misses OR cpu/dTLB-load-misses/
dTLB-stores OR cpu/dTLB-stores/
dTLB-store-misses OR cpu/dTLB-store-misses/
iTLB-loads OR cpu/iTLB-loads/
iTLB-load-misses OR cpu/iTLB-load-misses/
branch-loads OR cpu/branch-loads/
branch-load-misses OR cpu/branch-load-misses/
node-loads OR cpu/node-loads/
node-load-misses OR cpu/node-load-misses/
node-stores OR cpu/node-stores/
node-store-misses OR cpu/node-store-misses/
branch-instructions OR cpu/branch-instructions/ [Kernel PMU event]
branch-misses OR cpu/branch-misses/ [Kernel PMU event]
bus-cycles OR cpu/bus-cycles/ [Kernel PMU event]
cache-misses OR cpu/cache-misses/ [Kernel PMU event]
cache-references OR cpu/cache-references/ [Kernel PMU event]
cpu-cycles OR cpu/cpu-cycles/ [Kernel PMU event]
instructions OR cpu/instructions/ [Kernel PMU event]
mem-loads OR cpu/mem-loads/ [Kernel PMU event]
mem-stores OR cpu/mem-stores/ [Kernel PMU event]
# ......
3. 使用 perf stat 记录性能数据，并输出到终端
使用 perf stat 命令，可以运行被测试程序，并在程序结束之后，统计各不同Events的计数，并打印出来。也可以使用-p参数，指定运行程序的进程号。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
$ perf stat ls -l /usr/bin/ls
Performance counter stats for 'ls -lh /home/hxf0223/':
1.65 msec task-clock # 0.757 CPUs utilized
0 context-switches # 0.000 /sec
0 cpu-migrations # 0.000 /sec
112 page-faults # 67.878 K/sec
4,892,473 cycles # 2.965 GHz
3,714,068 instructions # 0.76 insn per cycle
681,289 branches # 412.897 M/sec
20,522 branch-misses # 3.01% of all branches
0.002179855 seconds time elapsed
0.000000000 seconds user
0.002198000 seconds sys
可以指定需要记录的events (-e参数)，以及repeat次数(-r参数)：
1
2
3
4
5
6
7
perf stat -r 6 -e cache-misses ls -lh ~/
Performance counter stats for 'ls -lh /home/hxf0223' (6 runs):
20,763 cache-misses ( +- 14.14% )
0.0015583 +- 0.0000369 seconds time elapsed ( +- 2.37% )
4. 记录性能数据到文件，以及分析：perf record, perf report, perf annotate
4.1 perf record 记录性能数据到文件
使用 perf record 命令，运行被测试程序，并记录测量数据到perf.data数据文件。
1
2
3
4
5
6
7
8
9
10
$ perf record ./test_grain_size
./test_grain_size. Process ID: 6226
Duration: 232.74 seconds
Sum: 1.25e+07. loop num: 10000
[ perf record: Woken up 934 times to write data ]
[ perf record: Captured and wrote 283.898 MB perf.data (7441535 samples) ]
$ ls -lh perf.data
-rw------- 1 <groupname> <username> 284M Oct 31 21:25 perf.data
4.2 perf report 查看性能数据
使用 perf report 命令，查看perf.data数据文件。使用如下命令，直接打开perf.data文件：
1
perf report
4.3 perf annotate 显示源码级别的性能数据
1
perf annotate -i perf.data
如何编译使用-ggdb，则可以显示源码级别的性能数据。
参考资料
本文由作者按照 CC BY 4.0 进行授权