1==========================
2Kprobe-based Event Tracing
3==========================
4
5:Author: Masami Hiramatsu
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 except those with
12__kprobes/nokprobe_inline annotation and those marked NOKPROBE_SYMBOL).
13Unlike the Tracepoint based event, this can be added and removed
14dynamically, on the fly.
15
16To enable this feature, build your kernel with CONFIG_KPROBE_EVENTS=y.
17
18Similar to the events tracer, this doesn't need to be activated via
19current_tracer. Instead of that, add probe points via
20/sys/kernel/debug/tracing/kprobe_events, and enable it via
21/sys/kernel/debug/tracing/events/kprobes/<EVENT>/enable.
22
23
24Synopsis of kprobe_events
25-------------------------
26::
27
28  p[:[GRP/]EVENT] [MOD:]SYM[+offs]|MEMADDR [FETCHARGS]	: Set a probe
29  r[MAXACTIVE][:[GRP/]EVENT] [MOD:]SYM[+0] [FETCHARGS]	: Set a return probe
30  -:[GRP/]EVENT						: Clear a probe
31
32 GRP		: Group name. If omitted, use "kprobes" for it.
33 EVENT		: Event name. If omitted, the event name is generated
34		  based on SYM+offs or MEMADDR.
35 MOD		: Module name which has given SYM.
36 SYM[+offs]	: Symbol+offset where the probe is inserted.
37 MEMADDR	: Address where the probe is inserted.
38 MAXACTIVE	: Maximum number of instances of the specified function that
39		  can be probed simultaneously, or 0 for the default value
40		  as defined in Documentation/kprobes.txt section 1.3.1.
41
42 FETCHARGS	: Arguments. Each probe can have up to 128 args.
43  %REG		: Fetch register REG
44  @ADDR		: Fetch memory at ADDR (ADDR should be in kernel)
45  @SYM[+|-offs]	: Fetch memory at SYM +|- offs (SYM should be a data symbol)
46  $stackN	: Fetch Nth entry of stack (N >= 0)
47  $stack	: Fetch stack address.
48  $argN		: Fetch the Nth function argument. (N >= 1) (\*1)
49  $retval	: Fetch return value.(\*2)
50  $comm		: Fetch current task comm.
51  +|-offs(FETCHARG) : Fetch memory at FETCHARG +|- offs address.(\*3)
52  NAME=FETCHARG : Set NAME as the argument name of FETCHARG.
53  FETCHARG:TYPE : Set TYPE as the type of FETCHARG. Currently, basic types
54		  (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal types
55		  (x8/x16/x32/x64), "string" and bitfield are supported.
56
57  (\*1) only for the probe on function entry (offs == 0).
58  (\*2) only for return probe.
59  (\*3) this is useful for fetching a field of data structures.
60
61Types
62-----
63Several types are supported for fetch-args. Kprobe tracer will access memory
64by given type. Prefix 's' and 'u' means those types are signed and unsigned
65respectively. 'x' prefix implies it is unsigned. Traced arguments are shown
66in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32'
67or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and
68x86-64 uses x64).
69These value types can be an array. To record array data, you can add '[N]'
70(where N is a fixed number, less than 64) to the base type.
71E.g. 'x16[4]' means an array of x16 (2bytes hex) with 4 elements.
72Note that the array can be applied to memory type fetchargs, you can not
73apply it to registers/stack-entries etc. (for example, '$stack1:x8[8]' is
74wrong, but '+8($stack):x8[8]' is OK.)
75String type is a special type, which fetches a "null-terminated" string from
76kernel space. This means it will fail and store NULL if the string container
77has been paged out.
78The string array type is a bit different from other types. For other base
79types, <base-type>[1] is equal to <base-type> (e.g. +0(%di):x32[1] is same
80as +0(%di):x32.) But string[1] is not equal to string. The string type itself
81represents "char array", but string array type represents "char * array".
82So, for example, +0(%di):string[1] is equal to +0(+0(%di)):string.
83Bitfield is another special type, which takes 3 parameters, bit-width, bit-
84offset, and container-size (usually 32). The syntax is::
85
86 b<bit-width>@<bit-offset>/<container-size>
87
88Symbol type('symbol') is an alias of u32 or u64 type (depends on BITS_PER_LONG)
89which shows given pointer in "symbol+offset" style.
90For $comm, the default type is "string"; any other type is invalid.
91
92
93Per-Probe Event Filtering
94-------------------------
95Per-probe event filtering feature allows you to set different filter on each
96probe and gives you what arguments will be shown in trace buffer. If an event
97name is specified right after 'p:' or 'r:' in kprobe_events, it adds an event
98under tracing/events/kprobes/<EVENT>, at the directory you can see 'id',
99'enable', 'format', 'filter' and 'trigger'.
100
101enable:
102  You can enable/disable the probe by writing 1 or 0 on it.
103
104format:
105  This shows the format of this probe event.
106
107filter:
108  You can write filtering rules of this event.
109
110id:
111  This shows the id of this probe event.
112
113trigger:
114  This allows to install trigger commands which are executed when the event is
115  hit (for details, see Documentation/trace/events.rst, section 6).
116
117Event Profiling
118---------------
119You can check the total number of probe hits and probe miss-hits via
120/sys/kernel/debug/tracing/kprobe_profile.
121The first column is event name, the second is the number of probe hits,
122the third is the number of probe miss-hits.
123
124
125Usage examples
126--------------
127To add a probe as a new event, write a new definition to kprobe_events
128as below::
129
130  echo 'p:myprobe do_sys_open dfd=%ax filename=%dx flags=%cx mode=+4($stack)' > /sys/kernel/debug/tracing/kprobe_events
131
132This sets a kprobe on the top of do_sys_open() function with recording
1331st to 4th arguments as "myprobe" event. Note, which register/stack entry is
134assigned to each function argument depends on arch-specific ABI. If you unsure
135the ABI, please try to use probe subcommand of perf-tools (you can find it
136under tools/perf/).
137As this example shows, users can choose more familiar names for each arguments.
138::
139
140  echo 'r:myretprobe do_sys_open $retval' >> /sys/kernel/debug/tracing/kprobe_events
141
142This sets a kretprobe on the return point of do_sys_open() function with
143recording return value as "myretprobe" event.
144You can see the format of these events via
145/sys/kernel/debug/tracing/events/kprobes/<EVENT>/format.
146::
147
148  cat /sys/kernel/debug/tracing/events/kprobes/myprobe/format
149  name: myprobe
150  ID: 780
151  format:
152          field:unsigned short common_type;       offset:0;       size:2; signed:0;
153          field:unsigned char common_flags;       offset:2;       size:1; signed:0;
154          field:unsigned char common_preempt_count;       offset:3; size:1;signed:0;
155          field:int common_pid;   offset:4;       size:4; signed:1;
156
157          field:unsigned long __probe_ip; offset:12;      size:4; signed:0;
158          field:int __probe_nargs;        offset:16;      size:4; signed:1;
159          field:unsigned long dfd;        offset:20;      size:4; signed:0;
160          field:unsigned long filename;   offset:24;      size:4; signed:0;
161          field:unsigned long flags;      offset:28;      size:4; signed:0;
162          field:unsigned long mode;       offset:32;      size:4; signed:0;
163
164
165  print fmt: "(%lx) dfd=%lx filename=%lx flags=%lx mode=%lx", REC->__probe_ip,
166  REC->dfd, REC->filename, REC->flags, REC->mode
167
168You can see that the event has 4 arguments as in the expressions you specified.
169::
170
171  echo > /sys/kernel/debug/tracing/kprobe_events
172
173This clears all probe points.
174
175Or,
176::
177
178  echo -:myprobe >> kprobe_events
179
180This clears probe points selectively.
181
182Right after definition, each event is disabled by default. For tracing these
183events, you need to enable it.
184::
185
186  echo 1 > /sys/kernel/debug/tracing/events/kprobes/myprobe/enable
187  echo 1 > /sys/kernel/debug/tracing/events/kprobes/myretprobe/enable
188
189And you can see the traced information via /sys/kernel/debug/tracing/trace.
190::
191
192  cat /sys/kernel/debug/tracing/trace
193  # tracer: nop
194  #
195  #           TASK-PID    CPU#    TIMESTAMP  FUNCTION
196  #              | |       |          |         |
197             <...>-1447  [001] 1038282.286875: myprobe: (do_sys_open+0x0/0xd6) dfd=3 filename=7fffd1ec4440 flags=8000 mode=0
198             <...>-1447  [001] 1038282.286878: myretprobe: (sys_openat+0xc/0xe <- do_sys_open) $retval=fffffffffffffffe
199             <...>-1447  [001] 1038282.286885: myprobe: (do_sys_open+0x0/0xd6) dfd=ffffff9c filename=40413c flags=8000 mode=1b6
200             <...>-1447  [001] 1038282.286915: myretprobe: (sys_open+0x1b/0x1d <- do_sys_open) $retval=3
201             <...>-1447  [001] 1038282.286969: myprobe: (do_sys_open+0x0/0xd6) dfd=ffffff9c filename=4041c6 flags=98800 mode=10
202             <...>-1447  [001] 1038282.286976: myretprobe: (sys_open+0x1b/0x1d <- do_sys_open) $retval=3
203
204
205Each line shows when the kernel hits an event, and <- SYMBOL means kernel
206returns from SYMBOL(e.g. "sys_open+0x1b/0x1d <- do_sys_open" means kernel
207returns from do_sys_open to sys_open+0x1b).
208
209