minimal is just that – a minimal practical BPF application example. It
doesn't use or require BPF CO-RE, so should run on quite old kernels. It
installs a tracepoint handler which is triggered once every second. It uses
bpf_printk() BPF helper to communicate with the world. To see it's output,
read /sys/kernel/debug/tracing/trace_pipe file as a root:
$ make minimal
$ sudo ./minimal
$ sudo cat /sys/kernel/debug/tracing/trace_pipe
<...>-3840345 [010] d... 3220701.101143: bpf_trace_printk: BPF triggered from PID 3840345.
<...>-3840345 [010] d... 3220702.101265: bpf_trace_printk: BPF triggered from PID 3840345.minimal is great as a bare-bones experimental playground to quickly try out
new ideas or BPF features.
bootstrap is an example of a simple (but realistic) BPF application. It
tracks process starts (exec() family of syscalls, to be precise) and exits
and emits data about filename, PID and parent PID, as well as exit status and
duration of the process life. With -d <min-duration-ms> you can specify
minimum duration of the process to log. In such mode process start
(technically, exec()) events are not output (see example output below).
bootstrap was created in the similar spirit as
libbpf-tools from
BCC package, but is designed to be more stand-alone and with simpler Makefile
to simplify adoption to user's particular needs. It demonstrates the use of
typical BPF features:
- cooperating BPF programs (tracepoint handlers for process
execandexitevents, in this particular case); - BPF map for maintaining the state;
- BPF ring buffer for sending data to user-space;
- global variables for application behavior parameterization.
- it utilizes BPF CO-RE and vmlinux.h to read extra process information from
kernel's
struct task_struct.
bootstrap is intended to be the starting point for your own BPF application,
with things like BPF CO-RE and vmlinux.h, consuming BPF ring buffer data,
command line arguments parsing, graceful Ctrl-C handling, etc. all taken care
of for you, which are crucial but mundane tasks that are no fun, but necessary
to be able to do anything useful. Just copy/paste and do simple renaming to get
yourself started.
Here's an example output in minimum process duration mode:
$ sudo ./bootstrap -d 50
TIME EVENT COMM PID PPID FILENAME/EXIT CODE
19:18:32 EXIT timeout 3817109 402466 [0] (126ms)
19:18:32 EXIT sudo 3817117 3817111 [0] (259ms)
19:18:32 EXIT timeout 3817110 402466 [0] (264ms)
19:18:33 EXIT python3.7 3817083 1 [0] (1026ms)
19:18:38 EXIT python3 3817429 3817424 [1] (60ms)
19:18:38 EXIT sh 3817424 3817420 [0] (79ms)
19:18:38 EXIT timeout 3817420 402466 [0] (80ms)
19:18:43 EXIT timeout 3817610 402466 [0] (70ms)
19:18:43 EXIT grep 3817619 3817617 [1] (271ms)
19:18:43 EXIT timeout 3817609 402466 [0] (321ms)
19:18:44 EXIT iostat 3817585 3817531 [0] (3006ms)
19:18:44 EXIT tee 3817587 3817531 [0] (3005ms)
...uprobe is an example of dealing with user-space entry and exit (return) probes,
uprobe and uretprobe in libbpf lingo. It attached uprobe and uretprobe
BPF programs to its own function (uprobe_trigger()) and logs input arguments
and return result, respectively, using bpf_printk() macro. The user-space
function is triggered once every second:
$ sudo ./uprobe
libbpf: loading object 'uprobe_bpf' from buffer
...
Successfully started!
...........You can see uprobe demo output in /sys/kernel/debug/tracing/trace_pipe:
$ sudo cat /sys/kernel/debug/tracing/trace_pipe
<...>-461101 [018] d... 505432.345032: bpf_trace_printk: UPROBE ENTRY: a = 0, b = 1
<...>-461101 [018] d... 505432.345042: bpf_trace_printk: UPROBE EXIT: return = 1
<...>-461101 [018] d... 505433.345186: bpf_trace_printk: UPROBE ENTRY: a = 1, b = 2
<...>-461101 [018] d... 505433.345202: bpf_trace_printk: UPROBE EXIT: return = 3
<...>-461101 [018] d... 505434.345342: bpf_trace_printk: UPROBE ENTRY: a = 2, b = 3
<...>-461101 [018] d... 505434.345367: bpf_trace_printk: UPROBE EXIT: return = 5fentry is an example that uses fentry and fexit BPF programs for tracing. It
attaches fentry and fexit traces to do_unlinkat() which is called when a
file is deleted and logs the return value, PID, and filename to the
trace pipe.
Important differences, compared to kprobes, are improved performance and usability. In this example, better usability is shown with the ability to directly dereference pointer arguments, like in normal C, instead of using various read helpers. The big distinction between fexit and kretprobe programs is that fexit one has access to both input arguments and returned result, while kretprobe can only access the result.
fentry and fexit programs are available starting from 5.5 kernels.
$ sudo ./fentry
libbpf: loading object 'fentry_bpf' from buffer
...
Successfully started!
..........The fentry output in /sys/kernel/debug/tracing/trace_pipe should look
something like this:
$ sudo cat /sys/kernel/debug/tracing/trace_pipe
rm-9290 [004] d..2 4637.798698: bpf_trace_printk: fentry: pid = 9290, filename = test_file
rm-9290 [004] d..2 4637.798843: bpf_trace_printk: fexit: pid = 9290, filename = test_file, ret = 0
rm-9290 [004] d..2 4637.798698: bpf_trace_printk: fentry: pid = 9290, filename = test_file2
rm-9290 [004] d..2 4637.798843: bpf_trace_printk: fexit: pid = 9290, filename = test_file2, ret = 0kprobe is an example of dealing with kernel-space entry and exit (return)
probes, kprobe and kretprobe in libbpf lingo. It attaches kprobe and
kretprobe BPF programs to the do_unlinkat() function and logs the PID,
filename, and return result, respectively, using bpf_printk() macro.
$ sudo ./kprobe
libbpf: loading object 'kprobe_bpf' from buffer
...
Successfully started!
...........The kprobe demo output in /sys/kernel/debug/tracing/trace_pipe should look
something like this:
$ sudo cat /sys/kernel/debug/tracing/trace_pipe
rm-9346 [005] d..3 4710.951696: bpf_trace_printk: KPROBE ENTRY pid = 9346, filename = test1
rm-9346 [005] d..4 4710.951819: bpf_trace_printk: KPROBE EXIT: ret = 0
rm-9346 [005] d..3 4710.951852: bpf_trace_printk: KPROBE ENTRY pid = 9346, filename = test2
rm-9346 [005] d..4 4710.951895: bpf_trace_printk: KPROBE EXIT: ret = 0$ git submodule update --init --recursive # check out libbpf
$ cd src
$ make
$ sudo ./bootstrap
TIME EVENT COMM PID PPID FILENAME/EXIT CODE
00:21:22 EXIT python3.8 4032353 4032352 [0] (123ms)
00:21:22 EXEC mkdir 4032379 4032337 /usr/bin/mkdir
00:21:22 EXIT mkdir 4032379 4032337 [0] (1ms)
00:21:22 EXEC basename 4032382 4032381 /usr/bin/basename
00:21:22 EXIT basename 4032382 4032381 [0] (0ms)
00:21:22 EXEC sh 4032381 4032380 /bin/sh
00:21:22 EXEC dirname 4032384 4032381 /usr/bin/dirname
00:21:22 EXIT dirname 4032384 4032381 [0] (1ms)
00:21:22 EXEC readlink 4032387 4032386 /usr/bin/readlink
^CLibbpf debug logs are quire helpful to pinpoint the exact source of problems, so it's usually a good idea to look at them before starting to debug or posting question online.
./minimal is always running with libbpf debug logs turned on.
For ./bootstrap, run it in verbose mode (-v) to see libbpf debug logs:
$ sudo ./bootstrap -v
libbpf: loading object 'bootstrap_bpf' from buffer
libbpf: elf: section(2) tp/sched/sched_process_exec, size 384, link 0, flags 6, type=1
libbpf: sec 'tp/sched/sched_process_exec': found program 'handle_exec' at insn offset 0 (0 bytes), code size 48 insns (384 bytes)
libbpf: elf: section(3) tp/sched/sched_process_exit, size 432, link 0, flags 6, type=1
libbpf: sec 'tp/sched/sched_process_exit': found program 'handle_exit' at insn offset 0 (0 bytes), code size 54 insns (432 bytes)
libbpf: elf: section(4) license, size 13, link 0, flags 3, type=1
libbpf: license of bootstrap_bpf is Dual BSD/GPL
...