A TestClock will be used automatically when a scenario has a
`crank-clock` action.
And make `validate` and `debug-viewer` options features in meson,
no reason they weren't and now we require gst-check to build validate
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-devtools/-/merge_requests/182>
This way we can have a single file that wraps scenarios,
`gst-validate-1.0` arguments, as well as a configuration.
It changes the name of `description` of scenarios to use `meta`
The goal is to replace tests describes in python with dictionary
to fully self contained `.validatetest` files which look like:
```
meta,
handles-states=true,
ignore-eos=true,
gst-validate-args = {
"videotestsrc pattern=blue ! video/x-raw,format=I420,framerate=1/1 ! timeoverlay ! $(videosink) name=videosink allocation-meta-flags=0",
},
configs = {
"$(validateflow), pad=videosink:sink, buffers-checksum=true, ignored-fields={\"buffers=meta\", }",
}
play
seek, start=0.0, stop=5.0, flags=accurate+flush, rate=1.0
crank-clock, expected-elapsed-time=0.0
crank-clock, repeat=4, expected-elapsed-time=1.0
crank-clock, expected-elapsed-time=1.0
stop, on-message=eos
```
Get a sense of files and line numbers in the parsed GstStructure
and take that information when reporting GstValidateAction errors
by letting the user know where the action comes from in the messages.
And accept non-literal string in printing formats.
Since `,` is the separator between fields of GstStructure we can
safely consider that if a line ends with it, the following line
is the logical continuity of the serialized GstStructure.
This makes writing those files more convenient and reading them
more pleasant as we do not need to add extra `\` at end of lines
anymore
Added two properties to the plugin:
* ignored-event-types: A list of event types to be ignored when logging events
* logged-event-types: A list of event types to be logged when logging events
This commits also moves the "ignored-event-fields" property to using a proper
GstValueList for the list of event fields to be taken into account, instead
of the home grown separated by comas list of string, making the API more
uniform.
This also adds a simple helper method: `gst_validate_utils_get_strv`
`gst_pad_pull_range` should always be called from the streaming thread,
we now check that when pull_range is called, and if the sinkpad calling
the function has a GstTask with a running thread, the function is called
from that thread.
In file included from ../../../../dist/linux_x86_64/include/gstreamer-1.0/gst/gst.h:55,
from ../validate/gst/validate/gst-validate-scenario.c:45:
../validate/gst/validate/gst-validate-scenario.c: In function ‘gst_validate_scenario_load’:
../../../../dist/linux_x86_64/include/gstreamer-1.0/gst/gstinfo.h:645:5: error: ‘%s’ directive argument is null [-Werror=format-overflow=]
645 | gst_debug_log ((cat), (level), __FILE__, GST_FUNCTION, __LINE__, \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
646 | (GObject *) (object), __VA_ARGS__); \
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../../../../dist/linux_x86_64/include/gstreamer-1.0/gst/gstinfo.h:1067:26: note: in expansion of macro ‘GST_CAT_LEVEL_LOG’
1067 | #define GST_ERROR(...) GST_CAT_LEVEL_LOG (GST_CAT_DEFAULT, GST_LEVEL_ERROR, NULL, __VA_ARGS__)
| ^~~~~~~~~~~~~~~~~
../validate/gst/validate/gst-validate-scenario.c:3615:5: note: in expansion of macro ‘GST_ERROR’
3615 | GST_ERROR ("Invalid name for scenario '%s'", scenario_name);
| ^~~~~~~~~
../validate/gst/validate/gst-validate-scenario.c:3615:44: note: format string is defined here
3615 | GST_ERROR ("Invalid name for scenario '%s'", scenario_name);
| ^~
[3470/4053] Compiling C object 'subprojects/gst-devtools/validate/gst/validate/28db7b6@@gstvalidatetracer@sha/gst-validate-reporter.c.o'.
../subprojects/gst-devtools/validate/gst/validate/gst-validate-reporter.c:186:31: warning: format string is not a string literal [-Wformat-nonliteral]
message = g_strdup_vprintf (format, vacopy);
^~~~~~
[3487/4053] Compiling C object 'subprojects/gst-devtools/validate/gst/validate/28db7b6@@gstvalidatetracer@sha/gst-validate-report.c.o'.
../subprojects/gst-devtools/validate/gst/validate/gst-validate-report.c:1007:34: warning: format string is not a string literal [-Wformat-nonliteral]
tmp = gst_info_strdup_vprintf (format, args);
^~~~~~
[76/151] Compiling C object 'subprojects/gst-devtools/validate/plugins/flow/697521d@@gstvalidateflow@sha/gstvalidateflow.c.o'.
../subprojects/gst-devtools/validate/plugins/flow/gstvalidateflow.c:125:65: warning: format string is not a string literal [-Wformat-nonliteral]
if (!flow->error_writing_file && vfprintf (flow->output_file, format, ap) < 0) {
^~~~~~
The way this was implemented was simply wrong, first of all checking
the position against the segment after a seek in PAUSED by a query
of the pipeline position is of limited interest, and can only work
in forward playback.
Furthermore the check was a huge blob of code that didn't even look
like it was pretending to do a good job at checking the position in
reverse playback.
This also adds the notion of global variables which will be useable
in config files too.
And add some documentation about default variables in scenarios
Meaning that instead of getting 1 "Detected on" line per monitor,
there will be one per "branch" like:
Detected on <audioconvert1:sink, audioconvert1:src, audioresample1:sink, audioresample1:src, smart-adder-adder:sink_0, smart-adder-adder:src, smart-adder-capsfilter:sink, smart-adder-capsfilter:src, capsfilter2:sink, capsfilter2:src, tee1:sink, tee1:src_0>
Making it simpler to read and a bit less verbose.
The action is generally useful but was implemented in a way that
was restricting its usage for no good reason. Refactor the
implementation adding more argument so it can be used in a wider
context, such as uvch264src.
Something like:
``` bash
echo "video-request-key-unit, direction=upstream, all-header=true, count=1, target-element-factory-name=h264parse, srcpad=src, playback-time=1.0" > tmp.scenario && \
echo "stop,playback-time=2.0" >> tmp.scenario && \
gst-validate-1.0 --set-scenario=tmp.scenario uvch264src \
device=/dev/video0 name=src iframe-period=33 auto-start=true src.vfsrc ! queue ! fakesink \
src.vidsrc ! queue ! video/x-h264,width=1280,height=720,framerate=30/1 ! h264parse ! fakesink
```
works now.
So it can be used directly in the documentation Also add a special "all"
argument to `gst-validate-1.0 --inspect-action-type` so we can generate
the documentation for all action types easily.
_Q_VALIDATE_MONITOR was defined twice because it wasn't declared
as extern in the header, so it would be defined as variable in all
included files. This doesn't seem to cause problems on Linux, but
seems to cause build failures on macOS.
Fixes#42
Basically nothing guarantees that the set of pads we aggregated the flow
for is the same as the one that was aggregated during the actual data
flow as some pads could have been removed meanwhile.
Commit 394242c224 ("validate:scenario: Enhance variable
implementation") caused the duration parameter to be stored
as a double instead of GstClockTime, which the _execute_pause
implementation expects. Fix the parameter type and use
gst_validate_action_get_clocktime to handle duration correctly.
https://gitlab.freedesktop.org/gstreamer/gst-devtools/merge_requests/73
There is a mockdecryptor that has been added into validate-sources and
this element is base on GstBaseTransform. This added a deps against
gstbase which was leading to linking errors when building with meson.
Especially the init section and the --quiet.
Remove the whole manual build/source dir include addition
to the g-ir-scanner args seeing that things worked fine
without the args being passed to the scanner at all.
Previously validateflow tests did not fail when the pad was not
attached.
This was a limitation caused by how the Validate API worked. Before, the
`notify::validate-runner` signal was not emitted until a monitor was
attached to the override. This made impossible to listen for the
runner's `stopping` signal.
This patch fixes the problem by setting `validate-runner` for all
existing overrides when the runner is initialized and adding checks in
validateflow to error in the case no pad was attached.
While in many cases it's desirable to wait for a buffer to be pushed
downstream when using appsrc-push, in some cases this is not possible as
such pushing action is dependent on following actions that would not be
executed if we wait.
An example for this is prerolling:
appsrc ! qtdemux ! video/x-h264 ! decodebin name=dec ! %(videosink)s
description, seek=false, handles-states=true
appsrc-push, target-element-name=appsrc0, file-name="raw_h264.0.mp4"
set-state, state=playing
appsrc-eos, target-element-name=appsrc0
In order for the preroll to occur, both the appsrc needs to push the
buffer and the state needs to reach PLAYING. But `set-state` cannot
finish if the buffer has not been pushed (the state transition does not
finish) and conversely pushing the buffer will not finish until the
state has reached.
Making appsrc-push not wait for the buffer solves this problem. This
patch makes appsrc-push aware of this issue by only waiting for the
buffer to be pushed if the pipeline is in a state that allows buffers to
flow.
Since gst_validate_action_set_done() is asynchronous, the bus EOS
handler may already be running before the action is actually finished.
This patch ensures that is not a problem.
This change allow tests to check performance of elements by checking the
frequency at which buffers are pushed on src pads.
I re-used most of the logic from fpsdisplaysink to compute the
frequency.
We can now uses something like:
GST_VALIDATE_CONFIG='core,min-buffer-frequency=60,target-element-factory-name=v4l2src'
The 'buffer-frequency-start' optional field can be used to ignore the
frequency during the start of the pipeline. This is useful when testing live
pipelines where configuring and setting up elements can take some time slowing
down the first buffers.
- Stop arbitrarily consider params as ClockTime based on their names
but add a convetion that the `.type` field of the ActionType should
end by `(GstClockTime)` when it is a clock time.
