xref: /linux/Documentation/trace/uprobetracer.rst (revision 1f2367a39f17bd553a75e179a747f9b257bc9478)
1=========================================
2Uprobe-tracer: Uprobe-based Event Tracing
3=========================================
4
5:Author: Srikar Dronamraju
6
7
8Overview
9--------
10Uprobe based trace events are similar to kprobe based trace events.
11To enable this feature, build your kernel with CONFIG_UPROBE_EVENTS=y.
12
13Similar to the kprobe-event tracer, this doesn't need to be activated via
14current_tracer. Instead of that, add probe points via
15/sys/kernel/debug/tracing/uprobe_events, and enable it via
16/sys/kernel/debug/tracing/events/uprobes/<EVENT>/enable.
17
18However unlike kprobe-event tracer, the uprobe event interface expects the
19user to calculate the offset of the probepoint in the object.
20
21You can also use /sys/kernel/debug/tracing/dynamic_events instead of
22uprobe_events. That interface will provide unified access to other
23dynamic events too.
24
25Synopsis of uprobe_tracer
26-------------------------
27::
28
29  p[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a uprobe
30  r[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a return uprobe (uretprobe)
31  -:[GRP/]EVENT                           : Clear uprobe or uretprobe event
32
33  GRP           : Group name. If omitted, "uprobes" is the default value.
34  EVENT         : Event name. If omitted, the event name is generated based
35                  on PATH+OFFSET.
36  PATH          : Path to an executable or a library.
37  OFFSET        : Offset where the probe is inserted.
38
39  FETCHARGS     : Arguments. Each probe can have up to 128 args.
40   %REG         : Fetch register REG
41   @ADDR	: Fetch memory at ADDR (ADDR should be in userspace)
42   @+OFFSET	: Fetch memory at OFFSET (OFFSET from same file as PATH)
43   $stackN	: Fetch Nth entry of stack (N >= 0)
44   $stack	: Fetch stack address.
45   $retval	: Fetch return value.(*)
46   $comm	: Fetch current task comm.
47   +|-offs(FETCHARG) : Fetch memory at FETCHARG +|- offs address.(**)
48   NAME=FETCHARG     : Set NAME as the argument name of FETCHARG.
49   FETCHARG:TYPE     : Set TYPE as the type of FETCHARG. Currently, basic types
50		       (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal types
51		       (x8/x16/x32/x64), "string" and bitfield are supported.
52
53  (*) only for return probe.
54  (**) this is useful for fetching a field of data structures.
55
56Types
57-----
58Several types are supported for fetch-args. Uprobe tracer will access memory
59by given type. Prefix 's' and 'u' means those types are signed and unsigned
60respectively. 'x' prefix implies it is unsigned. Traced arguments are shown
61in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32'
62or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and
63x86-64 uses x64).
64String type is a special type, which fetches a "null-terminated" string from
65user space.
66Bitfield is another special type, which takes 3 parameters, bit-width, bit-
67offset, and container-size (usually 32). The syntax is::
68
69 b<bit-width>@<bit-offset>/<container-size>
70
71For $comm, the default type is "string"; any other type is invalid.
72
73
74Event Profiling
75---------------
76You can check the total number of probe hits per event via
77/sys/kernel/debug/tracing/uprobe_profile. The first column is the filename,
78the second is the event name, the third is the number of probe hits.
79
80Usage examples
81--------------
82 * Add a probe as a new uprobe event, write a new definition to uprobe_events
83   as below (sets a uprobe at an offset of 0x4245c0 in the executable /bin/bash)::
84
85    echo 'p /bin/bash:0x4245c0' > /sys/kernel/debug/tracing/uprobe_events
86
87 * Add a probe as a new uretprobe event::
88
89    echo 'r /bin/bash:0x4245c0' > /sys/kernel/debug/tracing/uprobe_events
90
91 * Unset registered event::
92
93    echo '-:p_bash_0x4245c0' >> /sys/kernel/debug/tracing/uprobe_events
94
95 * Print out the events that are registered::
96
97    cat /sys/kernel/debug/tracing/uprobe_events
98
99 * Clear all events::
100
101    echo > /sys/kernel/debug/tracing/uprobe_events
102
103Following example shows how to dump the instruction pointer and %ax register
104at the probed text address. Probe zfree function in /bin/zsh::
105
106    # cd /sys/kernel/debug/tracing/
107    # cat /proc/`pgrep zsh`/maps | grep /bin/zsh | grep r-xp
108    00400000-0048a000 r-xp 00000000 08:03 130904 /bin/zsh
109    # objdump -T /bin/zsh | grep -w zfree
110    0000000000446420 g    DF .text  0000000000000012  Base        zfree
111
1120x46420 is the offset of zfree in object /bin/zsh that is loaded at
1130x00400000. Hence the command to uprobe would be::
114
115    # echo 'p:zfree_entry /bin/zsh:0x46420 %ip %ax' > uprobe_events
116
117And the same for the uretprobe would be::
118
119    # echo 'r:zfree_exit /bin/zsh:0x46420 %ip %ax' >> uprobe_events
120
121.. note:: User has to explicitly calculate the offset of the probe-point
122	in the object.
123
124We can see the events that are registered by looking at the uprobe_events file.
125::
126
127    # cat uprobe_events
128    p:uprobes/zfree_entry /bin/zsh:0x00046420 arg1=%ip arg2=%ax
129    r:uprobes/zfree_exit /bin/zsh:0x00046420 arg1=%ip arg2=%ax
130
131Format of events can be seen by viewing the file events/uprobes/zfree_entry/format.
132::
133
134    # cat events/uprobes/zfree_entry/format
135    name: zfree_entry
136    ID: 922
137    format:
138         field:unsigned short common_type;         offset:0;  size:2; signed:0;
139         field:unsigned char common_flags;         offset:2;  size:1; signed:0;
140         field:unsigned char common_preempt_count; offset:3;  size:1; signed:0;
141         field:int common_pid;                     offset:4;  size:4; signed:1;
142         field:int common_padding;                 offset:8;  size:4; signed:1;
143
144         field:unsigned long __probe_ip;           offset:12; size:4; signed:0;
145         field:u32 arg1;                           offset:16; size:4; signed:0;
146         field:u32 arg2;                           offset:20; size:4; signed:0;
147
148    print fmt: "(%lx) arg1=%lx arg2=%lx", REC->__probe_ip, REC->arg1, REC->arg2
149
150Right after definition, each event is disabled by default. For tracing these
151events, you need to enable it by::
152
153    # echo 1 > events/uprobes/enable
154
155Lets disable the event after sleeping for some time.
156::
157
158    # sleep 20
159    # echo 0 > events/uprobes/enable
160
161And you can see the traced information via /sys/kernel/debug/tracing/trace.
162::
163
164    # cat trace
165    # tracer: nop
166    #
167    #           TASK-PID    CPU#    TIMESTAMP  FUNCTION
168    #              | |       |          |         |
169                 zsh-24842 [006] 258544.995456: zfree_entry: (0x446420) arg1=446420 arg2=79
170                 zsh-24842 [007] 258545.000270: zfree_exit:  (0x446540 <- 0x446420) arg1=446540 arg2=0
171                 zsh-24842 [002] 258545.043929: zfree_entry: (0x446420) arg1=446420 arg2=79
172                 zsh-24842 [004] 258547.046129: zfree_exit:  (0x446540 <- 0x446420) arg1=446540 arg2=0
173
174Output shows us uprobe was triggered for a pid 24842 with ip being 0x446420
175and contents of ax register being 79. And uretprobe was triggered with ip at
1760x446540 with counterpart function entry at 0x446420.
177