4 **apitrace** consists of a set of tools to:
6 * trace OpenGL, OpenGL ES, Direct3D, and DirectDraw APIs calls to a file;
8 * replay OpenGL and OpenGL ES calls from a file;
10 * inspect OpenGL state at any call while retracing;
12 * visualize and edit trace files.
14 See the [apitrace homepage](http://apitrace.github.io/) for more details.
17 Obtaining **apitrace**
18 ======================
20 To obtain apitrace either [download the latest
21 binaries](http://apitrace.github.io/#download) for your platform if
22 available, or follow the instructions in INSTALL.markdown to build it yourself.
23 On 64bits Linux and Windows platforms you'll need apitrace binaries that match
24 the architecture (32bits or 64bits) of the application being traced.
30 Run the application you want to trace as
32 apitrace trace --api API /path/to/application [args...]
34 and it will generate a trace named `application.trace` in the current
35 directory. You can specify the written trace filename by passing the
36 `--output` command line option.
38 Problems while tracing (e.g, if the application uses calls/parameters
39 unsupported by apitrace) will be reported via stderr output on Unices. On
40 Windows you'll need to run
41 [DebugView](http://technet.microsoft.com/en-us/sysinternals/bb896647) to view
44 Follow the "Tracing manually" instructions below if you cannot obtain a trace.
48 apitrace dump application.trace
50 Replay an OpenGL trace with
52 apitrace replay application.trace
54 Pass the `--sb` option to use a single buffered visual. Pass `--help` to
55 `apitrace replay` for more options.
63 qapitrace application.trace
65 You can also tell the GUI to go directly to a specific call
67 qapitrace application.trace 12345
70 Advanced command line usage
71 ===========================
77 Several tools take `CALLSET` arguments, e.g:
79 apitrace dump --calls=CALLSET foo.trace
80 apitrace dump-images --calls=CALLSET foo.trace
82 The call syntax is very flexible. Here are a few examples:
86 * `0,2,4,5` set of calls
88 * `"0 2 4 5"` set of calls (commas are optional and can be replaced with whitespace)
90 * `0-100/2` calls 1, 3, 5, ..., 99
92 * `0-1000/draw` all draw calls between 0 and 1000
94 * `0-1000/fbo` all fbo changes between calls 0 and 1000
96 * `frame` all calls at end of frames
98 * `@foo.txt` read call numbers from `foo.txt`, using the same syntax as above
107 On 64 bits systems, you'll need to determine ether the application is 64 bits
108 or 32 bits. This can be done by doing
110 file /path/to/application
112 But beware of wrapper shell scripts -- what matters is the architecture of the
115 Run the GLX application you want to trace as
117 LD_PRELOAD=/path/to/apitrace/wrappers/glxtrace.so /path/to/application
119 and it will generate a trace named `application.trace` in the current
120 directory. You can specify the written trace filename by setting the
121 `TRACE_FILE` environment variable before running.
123 For EGL applications you will need to use `egltrace.so` instead of
126 The `LD_PRELOAD` mechanism should work with the majority applications. There
127 are some applications (e.g., Unigine Heaven, Android GPU emulator, etc.), that
128 have global function pointers with the same name as GL entrypoints, living in a
129 shared object that wasn't linked with `-Bsymbolic` flag, so relocations to
130 those globals function pointers get overwritten with the address to our wrapper
131 library, and the application will segfault when trying to write to them. For
132 these applications it is possible to trace by using `glxtrace.so` as an
133 ordinary `libGL.so` and injecting it via `LD_LIBRARY_PATH`:
135 ln -s glxtrace.so wrappers/libGL.so
136 ln -s glxtrace.so wrappers/libGL.so.1
137 ln -s glxtrace.so wrappers/libGL.so.1.2
138 export LD_LIBRARY_PATH=/path/to/apitrace/wrappers:$LD_LIBRARY_PATH
139 export TRACE_LIBGL=/path/to/real/libGL.so.1
142 If you are an application developer, you can avoid this either by linking with
143 `-Bsymbolic` flag, or by using some unique prefix for your function pointers.
145 See the `ld.so` man page for more information about `LD_PRELOAD` and
146 `LD_LIBRARY_PATH` environment flags.
150 To trace standalone native OpenGL ES applications, use
151 `LD_PRELOAD=/path/to/egltrace.so /path/to/application` like described in the
152 previous section. To trace Java applications, refer to Dalvik.markdown.
156 Run the application you want to trace as
158 DYLD_LIBRARY_PATH=/path/to/apitrace/wrappers /path/to/application
160 Note that although Mac OS X has an `LD_PRELOAD` equivalent,
161 `DYLD_INSERT_LIBRARIES`, it is mostly useless because it only works with
162 `DYLD_FORCE_FLAT_NAMESPACE=1` which breaks most applications. See the `dyld` man
163 page for more details about these environment flags.
167 When tracing third-party applications, you can identify the target
168 application's main executable, either by:
170 * right clicking on the application's icon in the _Start Menu_, choose
171 _Properties_, and see the _Target_ field;
173 * or by starting the application, run Windows Task Manager (taskmgr.exe), right
174 click on the application name in the _Applications_ tab, choose _Go To Process_,
175 note the highlighted _Image Name_, and search it on `C:\Program Files` or
176 `C:\Program Files (x86)`.
178 On 64 bits Windows, you'll need to determine ether the application is a 64 bits
179 or 32 bits. 32 bits applications will have a `*32` suffix in the _Image Name_
180 column of the _Processes_ tab of _Windows Task Manager_ window.
182 Copy the appropriate `opengl32.dll`, `d3d8.dll`, or `d3d9.dll` from the
183 wrappers directory to the directory with the application you want to trace.
184 Then run the application as usual.
186 You can specify the written trace filename by setting the `TRACE_FILE`
187 environment variable before running.
189 For D3D10 and higher you really must use `apitrace trace -a DXGI ...`. This is
190 because D3D10-11 API span many DLLs which depend on each other, and once a DLL
191 with a given name is loaded Windows will reuse it for LoadLibrary calls of the
192 same name, causing internal calls to be traced erroneously. `apitrace trace`
193 solves this issue by injecting a DLL `dxgitrace.dll` and patching all modules
194 to hook only the APIs of interest.
197 Emitting annotations to the trace
198 ---------------------------------
200 From OpenGL applications you can embed annotations in the trace file through the
201 [`GL_GREMEDY_string_marker`](http://www.opengl.org/registry/specs/GREMEDY/string_marker.txt)
203 [`GL_GREMEDY_frame_terminator`](http://www.opengl.org/registry/specs/GREMEDY/frame_terminator.txt)
206 **apitrace** will advertise and intercept these GL extensions independently of
207 the GL implementation. So all you have to do is to use these extensions when
210 For example, if you use [GLEW](http://glew.sourceforge.net/) to dynamically
211 detect and use GL extensions, you could easily accomplish this by doing:
215 if (GLEW_GREMEDY_string_marker) {
216 glStringMarkerGREMEDY(0, __FUNCTION__ ": enter");
221 if (GLEW_GREMEDY_string_marker) {
222 glStringMarkerGREMEDY(0, __FUNCTION__ ": leave");
227 This has the added advantage of working equally well with gDEBugger.
230 From OpenGL ES applications you can embed annotations in the trace file through the
231 [`GL_EXT_debug_marker`](http://www.khronos.org/registry/gles/extensions/EXT/EXT_debug_marker.txt)
235 For Direct3D applications you can follow the standard procedure for
236 [adding user defined events to Visual Studio Graphics Debugger / PIX](http://msdn.microsoft.com/en-us/library/vstudio/hh873200.aspx):
238 - `D3DPERF_BeginEvent`, `D3DPERF_EndEvent`, and `D3DPERF_SetMarker` for D3D9 applications.
240 - `ID3DUserDefinedAnnotation::BeginEvent`,
241 `ID3DUserDefinedAnnotation::EndEvent`, and
242 `ID3DUserDefinedAnnotation::SetMarker` for D3D11.1 applications.
245 Dump GL state at a particular call
246 ----------------------------------
248 You can get a dump of the bound GL state at call 12345 by doing:
250 apitrace replay -D 12345 application.trace > 12345.json
252 This is precisely the mechanism the GUI obtains its own state.
254 You can compare two state dumps by doing:
256 apitrace diff-state 12345.json 67890.json
259 Comparing two traces side by side
260 ---------------------------------
262 apitrace diff trace1.trace trace2.trace
264 This works only on Unices, and it will truncate the traces due to performance
268 Recording a video with FFmpeg/Libav
269 -----------------------------------
271 You can make a video of the output with FFmpeg by doing
273 apitrace dump-images -o - application.trace \
274 | ffmpeg -r 30 -f image2pipe -vcodec ppm -i pipe: -vcodec mpeg4 -y output.mp4
276 or Libav (which replaces FFmpeg on recent Debian/Ubuntu distros) doing
278 apitrace dump-images -o - application.trace \
279 | avconv -r 30 -f image2pipe -vcodec ppm -i - -vcodec mpeg4 -y output.mp4
281 Recording a video with gstreamer
282 --------------------------------------
284 You can make a video of the output with gstreamer by doing
286 glretrace --snapshot-format=RGB -s - smokinguns.trace | gst-launch-0.10 fdsrc blocksize=409600 ! queue \
287 ! videoparse format=rgb width=1920 height=1080 ! queue ! ffmpegcolorspace ! queue \
288 ! vaapiupload direct-rendering=0 ! queue ! vaapiencodeh264 ! filesink location=xxx.264
293 You can truncate a trace by doing:
295 apitrace trim --exact --calls 0-12345 -o trimed.trace application.trace
297 If you need precise control over which calls to trim you can specify the
298 individual call numbers a plaintext file, as described in the 'Call sets'
301 There is also experimental support for automatically trimming the calls
302 necessary for a given frame or call:
304 apitrace trim --auto --calls=12345 -o trimed.trace application.trace
305 apitrace trim --auto --frames=12345 -o trimed.trace application.trace
311 You can perform gpu and cpu profiling with the command line options:
313 * `--pgpu` record gpu times for frames and draw calls.
315 * `--pcpu` record cpu times for frames and draw calls.
317 * `--ppd` record pixels drawn for each draw call.
319 The results from this can then be read by hand or analysed with a script.
321 `scripts/profileshader.py` will read the profile results and format them into a
322 table which displays profiling results per shader.
324 For example, to record all profiling data and utilise the per shader script:
326 apitrace replay --pgpu --pcpu --ppd foo.trace | ./scripts/profileshader.py
329 Advanced usage for OpenGL implementors
330 ======================================
332 There are several advanced usage examples meant for OpenGL implementors.
338 These are the steps to create a regression test-suite around **apitrace**:
342 * obtain reference snapshots, by doing on a reference system:
344 mkdir /path/to/reference/snapshots/
345 apitrace dump-images -o /path/to/reference/snapshots/ application.trace
347 * prune the snapshots which are not interesting
349 * to do a regression test, use `apitrace diff-images`:
351 apitrace dump-images -o /path/to/test/snapshots/ application.trace
352 apitrace diff-images --output summary.html /path/to/reference/snapshots/ /path/to/test/snapshots/
355 Automated git-bisection
356 -----------------------
358 With tracecheck.py it is possible to automate git bisect and pinpoint the
359 commit responsible for a regression.
361 Below is an example of using tracecheck.py to bisect a regression in the
362 Mesa-based Intel 965 driver. But the procedure could be applied to any GL
363 driver hosted on a git repository.
365 First, create a build script, named build-script.sh, containing:
369 export PATH=/usr/lib/ccache:$PATH
372 ./autogen.sh --disable-egl --disable-gallium --disable-glut --disable-glu --disable-glw --with-dri-drivers=i965
376 It is important that builds are both robust, and efficient. Due to broken
377 dependency discovery in Mesa's makefile system, it was necessary invoke `make
378 clean` in every iteration step. `ccache` should be installed to avoid
379 recompiling unchanged source files.
384 export LIBGL_DEBUG=verbose
385 export LD_LIBRARY_PATH=$PWD/lib
386 export LIBGL_DRIVERS_DIR=$PWD/lib
388 6491e9593d5cbc5644eb02593a2f562447efdcbb 71acbb54f49089b03d3498b6f88c1681d3f649ac \
389 -- src/mesa/drivers/dri/intel src/mesa/drivers/dri/i965/
390 git bisect run /path/to/tracecheck.py \
391 --precision-threshold 8.0 \
392 --build /path/to/build-script.sh \
393 --gl-renderer '.*Mesa.*Intel.*' \
394 --retrace=/path/to/glretrace \
395 -c /path/to/reference/snapshots/ \
396 topogun-1.06-orc-84k.trace
398 The trace-check.py script will skip automatically when there are build
401 The `--gl-renderer` option will also cause a commit to be skipped if the
402 `GL_RENDERER` is unexpected (e.g., when a software renderer or another GL
403 driver is unintentionally loaded due to missing symbol in the DRI driver, or
404 another runtime fault).
407 Side by side retracing
408 ----------------------
410 In order to determine which draw call a regression first manifests one could
411 generate snapshots for every draw call, using the `-S` option. That is, however,
412 very inefficient for big traces with many draw calls.
414 A faster approach is to run both the bad and a good GL driver side-by-side.
415 The latter can be either a previously known good build of the GL driver, or a
416 reference software renderer.
418 This can be achieved with retracediff.py script, which invokes glretrace with
419 different environments, allowing to choose the desired GL driver by
420 manipulating variables such as `LD_LIBRARY_PATH`, `LIBGL_DRIVERS_DIR`, or
423 For example, on Linux:
425 ./scripts/retracediff.py \
426 --ref-env LD_LIBRARY_PATH=/path/to/reference/GL/implementation \
427 --retrace /path/to/glretrace \
428 --diff-prefix=/path/to/output/diffs \
433 python scripts\retracediff.py --retrace \path\to\glretrace.exe --ref-env TRACE_LIBGL=\path\to\reference\opengl32.dll application.trace
439 qapitrace has rudimentary support for replaying traces on a remote
440 target device. This can be useful, for example, when developing for an
441 embedded system. The primary GUI will run on the local host, while any
442 replays will be performed on the target device.
444 In order to target a remote device, use the command-line:
446 qapitrace --remote-target <HOST> <trace-file>
448 In order for this to work, the following must be available in the
449 system configuration:
451 1. It must be possible for the current user to initiate an ssh session
452 that has access to the target's window system. The command to be
453 exectuted by qapitrace will be:
457 For example, if the target device is using the X window system, one
458 can test whether an ssh session has access to the target X server
463 If this command fails with something like "cannot open display"
464 then the user will have to configure the target to set the DISPLAY
465 environment variable, (for example, setting DISPLAY=:0 in the
466 .bashrc file on the target or similar).
468 Also, note that if the ssh session requires a custom username, then
469 this must be configured on the host side so that ssh can be
470 initiated without a username.
472 For example, if you normally connect with `ssh user@192.168.0.2`
473 you could configure ~/.ssh/config on the host with a block such as:
479 And after this you should be able to connect with `ssh target` so
480 that you can also use `qapitrace --remote-target target`.
482 2. The target host must have a functional glretrace binary available
484 3. The target host must have access to <trace-file> at the same path
485 in the filesystem as the <trace-file> path on the host system being
486 passed to the qapitrace command line.
489 [![githalytics.com alpha](https://cruel-carlota.pagodabox.com/c1062ad633aa7a458e9d7520021307e4 "githalytics.com")](http://githalytics.com/apitrace/apitrace)