diff options
author | Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> | 2008-07-18 12:16:16 -0400 |
---|---|---|
committer | Ingo Molnar <mingo@elte.hu> | 2008-10-14 04:28:47 -0400 |
commit | 24b8d831d56aac7907752d22d2aba5d8127db6f6 (patch) | |
tree | 7a92d53f0c5e237099c0d39e7f11e4fef79fe951 /Documentation | |
parent | 97e1c18e8d17bd87e1e383b2e9d9fc740332c8e2 (diff) |
tracing: tracepoints, documentation
Documentation of tracepoint usage.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Acked-by: 'Peter Zijlstra' <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Diffstat (limited to 'Documentation')
-rw-r--r-- | Documentation/tracepoints.txt | 101 |
1 files changed, 101 insertions, 0 deletions
diff --git a/Documentation/tracepoints.txt b/Documentation/tracepoints.txt new file mode 100644 index 000000000000..5d354e167494 --- /dev/null +++ b/Documentation/tracepoints.txt | |||
@@ -0,0 +1,101 @@ | |||
1 | Using the Linux Kernel Tracepoints | ||
2 | |||
3 | Mathieu Desnoyers | ||
4 | |||
5 | |||
6 | This document introduces Linux Kernel Tracepoints and their use. It provides | ||
7 | examples of how to insert tracepoints in the kernel and connect probe functions | ||
8 | to them and provides some examples of probe functions. | ||
9 | |||
10 | |||
11 | * Purpose of tracepoints | ||
12 | |||
13 | A tracepoint placed in code provides a hook to call a function (probe) that you | ||
14 | can provide at runtime. A tracepoint can be "on" (a probe is connected to it) or | ||
15 | "off" (no probe is attached). When a tracepoint is "off" it has no effect, | ||
16 | except for adding a tiny time penalty (checking a condition for a branch) and | ||
17 | space penalty (adding a few bytes for the function call at the end of the | ||
18 | instrumented function and adds a data structure in a separate section). When a | ||
19 | tracepoint is "on", the function you provide is called each time the tracepoint | ||
20 | is executed, in the execution context of the caller. When the function provided | ||
21 | ends its execution, it returns to the caller (continuing from the tracepoint | ||
22 | site). | ||
23 | |||
24 | You can put tracepoints at important locations in the code. They are | ||
25 | lightweight hooks that can pass an arbitrary number of parameters, | ||
26 | which prototypes are described in a tracepoint declaration placed in a header | ||
27 | file. | ||
28 | |||
29 | They can be used for tracing and performance accounting. | ||
30 | |||
31 | |||
32 | * Usage | ||
33 | |||
34 | Two elements are required for tracepoints : | ||
35 | |||
36 | - A tracepoint definition, placed in a header file. | ||
37 | - The tracepoint statement, in C code. | ||
38 | |||
39 | In order to use tracepoints, you should include linux/tracepoint.h. | ||
40 | |||
41 | In include/trace/subsys.h : | ||
42 | |||
43 | #include <linux/tracepoint.h> | ||
44 | |||
45 | DEFINE_TRACE(subsys_eventname, | ||
46 | TPPTOTO(int firstarg, struct task_struct *p), | ||
47 | TPARGS(firstarg, p)); | ||
48 | |||
49 | In subsys/file.c (where the tracing statement must be added) : | ||
50 | |||
51 | #include <trace/subsys.h> | ||
52 | |||
53 | void somefct(void) | ||
54 | { | ||
55 | ... | ||
56 | trace_subsys_eventname(arg, task); | ||
57 | ... | ||
58 | } | ||
59 | |||
60 | Where : | ||
61 | - subsys_eventname is an identifier unique to your event | ||
62 | - subsys is the name of your subsystem. | ||
63 | - eventname is the name of the event to trace. | ||
64 | - TPPTOTO(int firstarg, struct task_struct *p) is the prototype of the function | ||
65 | called by this tracepoint. | ||
66 | - TPARGS(firstarg, p) are the parameters names, same as found in the prototype. | ||
67 | |||
68 | Connecting a function (probe) to a tracepoint is done by providing a probe | ||
69 | (function to call) for the specific tracepoint through | ||
70 | register_trace_subsys_eventname(). Removing a probe is done through | ||
71 | unregister_trace_subsys_eventname(); it will remove the probe sure there is no | ||
72 | caller left using the probe when it returns. Probe removal is preempt-safe | ||
73 | because preemption is disabled around the probe call. See the "Probe example" | ||
74 | section below for a sample probe module. | ||
75 | |||
76 | The tracepoint mechanism supports inserting multiple instances of the same | ||
77 | tracepoint, but a single definition must be made of a given tracepoint name over | ||
78 | all the kernel to make sure no type conflict will occur. Name mangling of the | ||
79 | tracepoints is done using the prototypes to make sure typing is correct. | ||
80 | Verification of probe type correctness is done at the registration site by the | ||
81 | compiler. Tracepoints can be put in inline functions, inlined static functions, | ||
82 | and unrolled loops as well as regular functions. | ||
83 | |||
84 | The naming scheme "subsys_event" is suggested here as a convention intended | ||
85 | to limit collisions. Tracepoint names are global to the kernel: they are | ||
86 | considered as being the same whether they are in the core kernel image or in | ||
87 | modules. | ||
88 | |||
89 | |||
90 | * Probe / tracepoint example | ||
91 | |||
92 | See the example provided in samples/tracepoints/src | ||
93 | |||
94 | Compile them with your kernel. | ||
95 | |||
96 | Run, as root : | ||
97 | modprobe tracepoint-example (insmod order is not important) | ||
98 | modprobe tracepoint-probe-example | ||
99 | cat /proc/tracepoint-example (returns an expected error) | ||
100 | rmmod tracepoint-example tracepoint-probe-example | ||
101 | dmesg | ||