Root/Documentation/trace/kprobetrace.txt

1                        Kprobe-based Event Tracing
2                        ==========================
3
4                 Documentation is written by Masami Hiramatsu
5
6
7Overview
8--------
9These events are similar to tracepoint based events. Instead of Tracepoint,
10this is based on kprobes (kprobe and kretprobe). So it can probe wherever
11kprobes can probe (this means, all functions body except for __kprobes
12functions). Unlike the Tracepoint based event, this can be added and removed
13dynamically, on the fly.
14
15To enable this feature, build your kernel with CONFIG_KPROBE_TRACING=y.
16
17Similar to the events tracer, this doesn't need to be activated via
18current_tracer. Instead of that, add probe points via
19/sys/kernel/debug/tracing/kprobe_events, and enable it via
20/sys/kernel/debug/tracing/events/kprobes/<EVENT>/enabled.
21
22
23Synopsis of kprobe_events
24-------------------------
25  p[:[GRP/]EVENT] SYMBOL[+offs]|MEMADDR [FETCHARGS] : Set a probe
26  r[:[GRP/]EVENT] SYMBOL[+0] [FETCHARGS] : Set a return probe
27  -:[GRP/]EVENT : Clear a probe
28
29 GRP : Group name. If omitted, use "kprobes" for it.
30 EVENT : Event name. If omitted, the event name is generated
31          based on SYMBOL+offs or MEMADDR.
32 SYMBOL[+offs] : Symbol+offset where the probe is inserted.
33 MEMADDR : Address where the probe is inserted.
34
35 FETCHARGS : Arguments. Each probe can have up to 128 args.
36  %REG : Fetch register REG
37  @ADDR : Fetch memory at ADDR (ADDR should be in kernel)
38  @SYM[+|-offs] : Fetch memory at SYM +|- offs (SYM should be a data symbol)
39  $stackN : Fetch Nth entry of stack (N >= 0)
40  $stack : Fetch stack address.
41  $retval : Fetch return value.(*)
42  +|-offs(FETCHARG) : Fetch memory at FETCHARG +|- offs address.(**)
43  NAME=FETCHARG: Set NAME as the argument name of FETCHARG.
44
45  (*) only for return probe.
46  (**) this is useful for fetching a field of data structures.
47
48
49Per-Probe Event Filtering
50-------------------------
51 Per-probe event filtering feature allows you to set different filter on each
52probe and gives you what arguments will be shown in trace buffer. If an event
53name is specified right after 'p:' or 'r:' in kprobe_events, it adds an event
54under tracing/events/kprobes/<EVENT>, at the directory you can see 'id',
55'enabled', 'format' and 'filter'.
56
57enabled:
58  You can enable/disable the probe by writing 1 or 0 on it.
59
60format:
61  This shows the format of this probe event.
62
63filter:
64  You can write filtering rules of this event.
65
66id:
67  This shows the id of this probe event.
68
69
70Event Profiling
71---------------
72 You can check the total number of probe hits and probe miss-hits via
73/sys/kernel/debug/tracing/kprobe_profile.
74 The first column is event name, the second is the number of probe hits,
75the third is the number of probe miss-hits.
76
77
78Usage examples
79--------------
80To add a probe as a new event, write a new definition to kprobe_events
81as below.
82
83  echo 'p:myprobe do_sys_open dfd=%ax filename=%dx flags=%cx mode=+4($stack)' > /sys/kernel/debug/tracing/kprobe_events
84
85 This sets a kprobe on the top of do_sys_open() function with recording
861st to 4th arguments as "myprobe" event. Note, which register/stack entry is
87assigned to each function argument depends on arch-specific ABI. If you unsure
88the ABI, please try to use probe subcommand of perf-tools (you can find it
89under tools/perf/).
90As this example shows, users can choose more familiar names for each arguments.
91
92  echo 'r:myretprobe do_sys_open $retval' >> /sys/kernel/debug/tracing/kprobe_events
93
94 This sets a kretprobe on the return point of do_sys_open() function with
95recording return value as "myretprobe" event.
96 You can see the format of these events via
97/sys/kernel/debug/tracing/events/kprobes/<EVENT>/format.
98
99  cat /sys/kernel/debug/tracing/events/kprobes/myprobe/format
100name: myprobe
101ID: 780
102format:
103        field:unsigned short common_type; offset:0; size:2; signed:0;
104        field:unsigned char common_flags; offset:2; size:1; signed:0;
105        field:unsigned char common_preempt_count; offset:3; size:1;signed:0;
106        field:int common_pid; offset:4; size:4; signed:1;
107        field:int common_lock_depth; offset:8; size:4; signed:1;
108
109        field:unsigned long __probe_ip; offset:12; size:4; signed:0;
110        field:int __probe_nargs; offset:16; size:4; signed:1;
111        field:unsigned long dfd; offset:20; size:4; signed:0;
112        field:unsigned long filename; offset:24; size:4; signed:0;
113        field:unsigned long flags; offset:28; size:4; signed:0;
114        field:unsigned long mode; offset:32; size:4; signed:0;
115
116
117print fmt: "(%lx) dfd=%lx filename=%lx flags=%lx mode=%lx", REC->__probe_ip,
118REC->dfd, REC->filename, REC->flags, REC->mode
119
120 You can see that the event has 4 arguments as in the expressions you specified.
121
122  echo > /sys/kernel/debug/tracing/kprobe_events
123
124 This clears all probe points.
125
126 Or,
127
128  echo -:myprobe >> kprobe_events
129
130 This clears probe points selectively.
131
132 Right after definition, each event is disabled by default. For tracing these
133events, you need to enable it.
134
135  echo 1 > /sys/kernel/debug/tracing/events/kprobes/myprobe/enable
136  echo 1 > /sys/kernel/debug/tracing/events/kprobes/myretprobe/enable
137
138 And you can see the traced information via /sys/kernel/debug/tracing/trace.
139
140  cat /sys/kernel/debug/tracing/trace
141# tracer: nop
142#
143# TASK-PID CPU# TIMESTAMP FUNCTION
144# | | | | |
145           <...>-1447 [001] 1038282.286875: myprobe: (do_sys_open+0x0/0xd6) dfd=3 filename=7fffd1ec4440 flags=8000 mode=0
146           <...>-1447 [001] 1038282.286878: myretprobe: (sys_openat+0xc/0xe <- do_sys_open) $retval=fffffffffffffffe
147           <...>-1447 [001] 1038282.286885: myprobe: (do_sys_open+0x0/0xd6) dfd=ffffff9c filename=40413c flags=8000 mode=1b6
148           <...>-1447 [001] 1038282.286915: myretprobe: (sys_open+0x1b/0x1d <- do_sys_open) $retval=3
149           <...>-1447 [001] 1038282.286969: myprobe: (do_sys_open+0x0/0xd6) dfd=ffffff9c filename=4041c6 flags=98800 mode=10
150           <...>-1447 [001] 1038282.286976: myretprobe: (sys_open+0x1b/0x1d <- do_sys_open) $retval=3
151
152
153 Each line shows when the kernel hits an event, and <- SYMBOL means kernel
154returns from SYMBOL(e.g. "sys_open+0x1b/0x1d <- do_sys_open" means kernel
155returns from do_sys_open to sys_open+0x1b).
156
157

Archive Download this file



interactive