]>
Commit | Line | Data |
---|---|---|
81a97d9d SH |
1 | = Tracing = |
2 | ||
3 | == Introduction == | |
4 | ||
5 | This document describes the tracing infrastructure in QEMU and how to use it | |
6 | for debugging, profiling, and observing execution. | |
7 | ||
8 | == Quickstart == | |
9 | ||
10 | 1. Build with the 'simple' trace backend: | |
11 | ||
12 | ./configure --trace-backend=simple | |
13 | make | |
14 | ||
dd215f64 | 15 | 2. Run the virtual machine to produce a trace file: |
81a97d9d SH |
16 | |
17 | qemu ... # your normal QEMU invocation | |
18 | ||
dd215f64 | 19 | 3. Pretty-print the binary trace file: |
81a97d9d SH |
20 | |
21 | ./simpletrace.py trace-events trace-* | |
22 | ||
23 | == Trace events == | |
24 | ||
7b92e5bc | 25 | There is a set of static trace events declared in the "trace-events" source |
81a97d9d SH |
26 | file. Each trace event declaration names the event, its arguments, and the |
27 | format string which can be used for pretty-printing: | |
28 | ||
29 | qemu_malloc(size_t size, void *ptr) "size %zu ptr %p" | |
30 | qemu_free(void *ptr) "ptr %p" | |
31 | ||
7b92e5bc | 32 | The "trace-events" file is processed by the "tracetool" script during build to |
81a97d9d SH |
33 | generate code for the trace events. Trace events are invoked directly from |
34 | source code like this: | |
35 | ||
36 | #include "trace.h" /* needed for trace event prototype */ | |
49926043 | 37 | |
81a97d9d SH |
38 | void *qemu_malloc(size_t size) |
39 | { | |
40 | void *ptr; | |
41 | if (!size && !allow_zero_malloc()) { | |
42 | abort(); | |
43 | } | |
44 | ptr = oom_check(malloc(size ? size : 1)); | |
45 | trace_qemu_malloc(size, ptr); /* <-- trace event */ | |
46 | return ptr; | |
47 | } | |
48 | ||
49 | === Declaring trace events === | |
50 | ||
7b92e5bc | 51 | The "tracetool" script produces the trace.h header file which is included by |
81a97d9d | 52 | every source file that uses trace events. Since many source files include |
7b92e5bc LV |
53 | trace.h, it uses a minimum of types and other header files included to keep the |
54 | namespace clean and compile times and dependencies down. | |
81a97d9d SH |
55 | |
56 | Trace events should use types as follows: | |
57 | ||
58 | * Use stdint.h types for fixed-size types. Most offsets and guest memory | |
59 | addresses are best represented with uint32_t or uint64_t. Use fixed-size | |
60 | types over primitive types whose size may change depending on the host | |
61 | (32-bit versus 64-bit) so trace events don't truncate values or break | |
62 | the build. | |
63 | ||
64 | * Use void * for pointers to structs or for arrays. The trace.h header | |
65 | cannot include all user-defined struct declarations and it is therefore | |
66 | necessary to use void * for pointers to structs. | |
67 | ||
e6a750aa SH |
68 | Pointers (including char *) cannot be dereferenced easily (or at all) in |
69 | some trace backends. If pointers are used, ensure they are meaningful by | |
70 | themselves and do not assume the data they point to will be traced. Do | |
71 | not pass in string arguments. | |
72 | ||
81a97d9d SH |
73 | * For everything else, use primitive scalar types (char, int, long) with the |
74 | appropriate signedness. | |
75 | ||
9a85d394 SH |
76 | Format strings should reflect the types defined in the trace event. Take |
77 | special care to use PRId64 and PRIu64 for int64_t and uint64_t types, | |
cf85cf8e SH |
78 | respectively. This ensures portability between 32- and 64-bit platforms. Note |
79 | that format strings must begin and end with double quotes. When using | |
80 | portability macros, ensure they are preceded and followed by double quotes: | |
81 | "value %"PRIx64"". | |
9a85d394 | 82 | |
81a97d9d SH |
83 | === Hints for adding new trace events === |
84 | ||
85 | 1. Trace state changes in the code. Interesting points in the code usually | |
86 | involve a state change like starting, stopping, allocating, freeing. State | |
87 | changes are good trace events because they can be used to understand the | |
88 | execution of the system. | |
89 | ||
90 | 2. Trace guest operations. Guest I/O accesses like reading device registers | |
91 | are good trace events because they can be used to understand guest | |
92 | interactions. | |
93 | ||
94 | 3. Use correlator fields so the context of an individual line of trace output | |
95 | can be understood. For example, trace the pointer returned by malloc and | |
96 | used as an argument to free. This way mallocs and frees can be matched up. | |
97 | Trace events with no context are not very useful. | |
98 | ||
99 | 4. Name trace events after their function. If there are multiple trace events | |
100 | in one function, append a unique distinguisher at the end of the name. | |
101 | ||
dd215f64 LV |
102 | 5. If specific trace events are going to be called a huge number of times, this |
103 | might have a noticeable performance impact even when the trace events are | |
104 | programmatically disabled. In this case you should declare the trace event | |
105 | with the "disable" property, which will effectively disable it at compile | |
106 | time (using the "nop" backend). | |
81a97d9d | 107 | |
31965ae2 LV |
108 | == Generic interface and monitor commands == |
109 | ||
110 | You can programmatically query and control the dynamic state of trace events | |
111 | through a backend-agnostic interface: | |
112 | ||
113 | * trace_print_events | |
114 | ||
115 | * trace_event_set_state | |
116 | Enables or disables trace events at runtime inside QEMU. | |
117 | The function returns "true" if the state of the event has been successfully | |
118 | changed, or "false" otherwise: | |
119 | ||
120 | #include "trace/control.h" | |
121 | ||
122 | trace_event_set_state("virtio_irq", true); /* enable */ | |
123 | [...] | |
124 | trace_event_set_state("virtio_irq", false); /* disable */ | |
125 | ||
126 | Note that some of the backends do not provide an implementation for this | |
127 | interface, in which case QEMU will just print a warning. | |
128 | ||
129 | This functionality is also provided through monitor commands: | |
130 | ||
131 | * info trace-events | |
132 | View available trace events and their state. State 1 means enabled, state 0 | |
133 | means disabled. | |
134 | ||
135 | * trace-event NAME on|off | |
136 | Enable/disable a given trace event. | |
137 | ||
23d15e86 LV |
138 | The "-trace events=<file>" command line argument can be used to enable the |
139 | events listed in <file> from the very beginning of the program. This file must | |
140 | contain one event name per line. | |
141 | ||
81a97d9d SH |
142 | == Trace backends == |
143 | ||
7b92e5bc | 144 | The "tracetool" script automates tedious trace event code generation and also |
81a97d9d SH |
145 | keeps the trace event declarations independent of the trace backend. The trace |
146 | events are not tightly coupled to a specific trace backend, such as LTTng or | |
7b92e5bc | 147 | SystemTap. Support for trace backends can be added by extending the "tracetool" |
81a97d9d SH |
148 | script. |
149 | ||
150 | The trace backend is chosen at configure time and only one trace backend can | |
151 | be built into the binary: | |
152 | ||
153 | ./configure --trace-backend=simple | |
154 | ||
155 | For a list of supported trace backends, try ./configure --help or see below. | |
156 | ||
157 | The following subsections describe the supported trace backends. | |
158 | ||
159 | === Nop === | |
160 | ||
161 | The "nop" backend generates empty trace event functions so that the compiler | |
162 | can optimize out trace events completely. This is the default and imposes no | |
163 | performance penalty. | |
164 | ||
dd215f64 LV |
165 | Note that regardless of the selected trace backend, events with the "disable" |
166 | property will be generated with the "nop" backend. | |
167 | ||
b48c20f7 SH |
168 | === Stderr === |
169 | ||
170 | The "stderr" backend sends trace events directly to standard error. This | |
171 | effectively turns trace events into debug printfs. | |
172 | ||
173 | This is the simplest backend and can be used together with existing code that | |
174 | uses DPRINTF(). | |
175 | ||
dd215f64 LV |
176 | Note that with this backend trace events cannot be programmatically |
177 | enabled/disabled. Thus, in order to trim down the amount of output and the | |
178 | performance impact of tracing, you might want to add the "disable" property in | |
179 | the "trace-events" file for those events you are not interested in. | |
180 | ||
81a97d9d SH |
181 | === Simpletrace === |
182 | ||
183 | The "simple" backend supports common use cases and comes as part of the QEMU | |
184 | source tree. It may not be as powerful as platform-specific or third-party | |
185 | trace backends but it is portable. This is the recommended trace backend | |
186 | unless you have specific needs for more advanced backends. | |
187 | ||
188 | ==== Monitor commands ==== | |
189 | ||
190 | * info trace | |
191 | Display the contents of trace buffer. This command dumps the trace buffer | |
192 | with simple formatting. For full pretty-printing, use the simpletrace.py | |
193 | script on a binary trace file. | |
194 | ||
195 | The trace buffer is written into until full. The full trace buffer is | |
196 | flushed and emptied. This means the 'info trace' will display few or no | |
197 | entries if the buffer has just been flushed. | |
198 | ||
81a97d9d SH |
199 | * trace-file on|off|flush|set <path> |
200 | Enable/disable/flush the trace file or set the trace file name. | |
201 | ||
81a97d9d SH |
202 | ==== Analyzing trace files ==== |
203 | ||
204 | The "simple" backend produces binary trace files that can be formatted with the | |
7b92e5bc | 205 | simpletrace.py script. The script takes the "trace-events" file and the binary |
81a97d9d SH |
206 | trace: |
207 | ||
208 | ./simpletrace.py trace-events trace-12345 | |
209 | ||
7b92e5bc | 210 | You must ensure that the same "trace-events" file was used to build QEMU, |
81a97d9d SH |
211 | otherwise trace event declarations may have changed and output will not be |
212 | consistent. | |
213 | ||
214 | === LTTng Userspace Tracer === | |
215 | ||
216 | The "ust" backend uses the LTTng Userspace Tracer library. There are no | |
217 | monitor commands built into QEMU, instead UST utilities should be used to list, | |
218 | enable/disable, and dump traces. | |
b48c20f7 SH |
219 | |
220 | === SystemTap === | |
221 | ||
222 | The "dtrace" backend uses DTrace sdt probes but has only been tested with | |
223 | SystemTap. When SystemTap support is detected a .stp file with wrapper probes | |
224 | is generated to make use in scripts more convenient. This step can also be | |
225 | performed manually after a build in order to change the binary name in the .stp | |
226 | probes: | |
227 | ||
228 | scripts/tracetool --dtrace --stap \ | |
229 | --binary path/to/qemu-binary \ | |
230 | --target-type system \ | |
231 | --target-arch x86_64 \ | |
232 | <trace-events >qemu.stp |