Commit graph

857 commits

Author SHA1 Message Date
Edward Hervey
55274bc89f validate-pad-monitor: Use GST_SEQNUM_INVALID
Instead of 0 (which is valid)
2018-02-10 14:20:44 +01:00
Edward Hervey
92285ef261 validate: Use a single TCPServer for subprocess communication
Instead of creating a separate TCPServer for each test, just create
one which handles all connections in a threaded fashion.

Shaves off ~500ms per test

https://bugzilla.gnome.org/show_bug.cgi?id=791159
2017-12-03 12:23:29 +01:00
Edward Hervey
9acfa7fe4b validate: Don't leak strings
We only use them in the error/debug case anyway
2017-11-25 13:10:41 +01:00
Edward Hervey
c33b2e240c validate-report: Plug leaks
The trace was never freed, nor were the output of g_str_split
2017-11-25 12:46:05 +01:00
Edward Hervey
b5bb7d7016 validate-scenario: Don't assume element have factories
Some elements might not originate from factories (like custom/internal
elements).

Avoids dereferencing a NULL pointer
2017-11-23 12:27:11 +01:00
Edward Hervey
ff6055b195 validate-scenario: Handle non-relative switch
Make sure we stay within the number of present streams (and avoid
out-of-bound read).

CID #1415470
2017-11-22 16:37:44 +01:00
Edward Hervey
81c3ab0b12 validate-scenario: Handle switching stream of type not present
Unlikely to happen, but at least don't end up doing unsafe calculation
with n == 0 afterwards

CID #1415453
2017-11-22 16:37:44 +01:00
Edward Hervey
9658e12086 validate: Call g-ir-scanner with the same toolchain as the rest 2017-11-08 17:22:47 +01:00
Tim-Philipp Müller
8b91fef492 validate: hide some private symbols 2017-08-12 12:04:42 +01:00
Thibault Saunier
8393a3d5ca validate: Plug a potential leak when retrieving peer pad 2017-08-10 21:44:14 -04:00
Thibault Saunier
7be8ecd628 validate: Fix going over ghostpads/proxypads 2017-08-10 19:29:07 -04:00
Thibault Saunier
74a560163a validate: Mark symbols explicitly for export with GST_EXPORT
With an exception:
  * gst_validate_monitor_setup

which was never declared in headers and should always have been static.
2017-08-07 15:56:21 -04:00
Thibault Saunier
823c9ca834 validate: Fix building the tracer
It fails on some platforms, I guess this is the reason
2017-07-26 17:25:25 -04:00
Thibault Saunier
8ef1050d1e validate: Also mkenums with autotools
And fix the build with stricter gcc arguments.
2017-07-26 16:18:02 -04:00
Thibault Saunier
b9d6f9df9e validate: Add a way to print information about pipeline status
Similare to what is done with gst-launch.

And finally generate GTypes for our flags and enums.
2017-07-26 15:45:37 -04:00
Edward Hervey
40eb48d21f validate-scenario: Fix NULL pointer usage
for good this time ...

CID #1415570
2017-07-26 15:22:49 +02:00
Edward Hervey
1b48ffdff5 validate-report: Fix a leak in error cases
CID #1415494
2017-07-26 15:18:57 +02:00
Edward Hervey
6ad8bd3e7c validate-scenario: Protect against priv NULL usage
CID #1415570
2017-07-20 14:21:59 +02:00
Thibault Saunier
1a28e7b043 validate: Factor out a method to set properties on elements in utils
Make sure to use it where appropriate and add some logging when
setting an object property from an action.

And use the valgrind.conf to set all the properties instead of having
a mixture of a config scenario and the config file (making sure the
max-lateness is set on any sink)
2017-07-19 11:34:19 -04:00
Thibault Saunier
1d568ff11f validate:scenario: Allow not config action to be executed from config files
When those are special cased to support that, such as the `set-property`
action.

This special handling was added in

  4927c65710
  validate: disable QOS features when running with valgrind

before we started to support executing arbitrary config action from
configuration files.
2017-07-19 11:34:12 -04:00
Thibault Saunier
b3134e89d9 validate: scenario: Fix running config action from the config file 2017-07-19 10:23:34 -04:00
Edward Hervey
8b9b6ead3f validate: Cast GList data content before usage
Apart from code readability, it allows compilers to detect wrong usages,
such as the call to gst_validate_action_new() which was using the wrong
argument
2017-07-19 11:02:44 +02:00
Thibault Saunier
f7252cbd48 validate: Remove some dead code
CID 1415457
2017-07-18 10:47:00 -04:00
Thibault Saunier
e70557dc3e validate: Plug a minor string leak
CID 1415459
2017-07-18 10:45:29 -04:00
Thibault Saunier
9ee7b4483c validate: Don't create scenario on a monitor which has no target
Not very probable but avoids a potential NULL pointer dereferencing.

CID 1415460
2017-07-18 10:42:00 -04:00
Thibault Saunier
6572af916b validate: Always only use the first description in scenarios
Also pluging a leak of the descrption copied structure

CID 1415463
2017-07-18 10:38:24 -04:00
Thibault Saunier
c7483a90f0 validate: Plug a string leak
CID 1415475
2017-07-18 10:23:31 -04:00
Thibault Saunier
28f8787c17 validate: Add missing break statement
CID 1415485
2017-07-18 10:16:07 -04:00
Thibault Saunier
7eb9746adc Check g_file_set_contents() return value
CID 1415486
2017-07-18 10:14:59 -04:00
Thibault Saunier
44260d70bd validate: Plug leak of copy of a va_list
CID 1415490
2017-07-18 10:12:07 -04:00
Thibault Saunier
11ef28f3dc validate: Do not check NULL pointer uselessly
CID 141593
2017-07-18 10:10:06 -04:00
Thibault Saunier
fc49d18425 validate: Plug minor leak in issue creation error path
CID 1415494
2017-07-18 10:10:06 -04:00
Thibault Saunier
54e35bb082 validate: Remove some dead code
Next will never be NULL as `done` is always set to TRUE when next is
set.

CID 1415503
2017-07-18 10:00:03 -04:00
Thibault Saunier
5c537bd627 validate: monitor: Add missing break; statement
Fixes CID 1415500
2017-07-18 09:59:42 -04:00
Edward Hervey
4f1242968e validate-scenario: Fix wrong return value
We were always returning ok ...

CID #1415484
2017-07-18 15:50:35 +02:00
Edward Hervey
3904f2523b validate-scenario: Check g_file_set_contents() return value
CID #1415487
2017-07-18 15:48:01 +02:00
Edward Hervey
e6073a411c validate-scenario: Fix copy/paste error
CID #1415502
2017-07-18 15:45:13 +02:00
Edward Hervey
e5903f1d05 validate-scenario: Fix string usage
Use the string representation of the index if it *IS* present (and
not the opposite).

CID #1415506
2017-07-18 15:43:26 +02:00
Edward Hervey
5b54127c5a validate-utils: Add NULL check
Unlikely to be triggered. And fix typo at the same time

CID #1415464
2017-07-18 15:38:04 +02:00
Thibault Saunier
c1f613d8e7 validate:scenario: Allow setting properties by element factory name 2017-06-30 12:30:40 -04:00
Thibault Saunier
f2fc6a4550 validate:scenario: Enhance the set_property action to handle enum props
User needs to specify the enum value as a string, to be used
as with gst_util_set_object_arg.

Also enhance reporting and verify that the set value has actually
been taken into account.
2017-06-30 09:46:57 -04:00
Thibault Saunier
dbc598b378 validate: Use Gst printing utils in our reporting system
Allowing us to use GST_PTR_FORMAT and friends!
2017-06-30 09:45:02 -04:00
Thibault Saunier
6a8b0111cd validate:scenario: Enhance playbin3 stream selection error message 2017-06-22 11:52:26 -04:00
Thibault Saunier
eac403f87c validate: Add missing space in segment mismatch issue description 2017-06-20 15:51:27 -04:00
Thibault Saunier
a728a7df49 validate: Fix the way we check if _CONFIG is a list of structures 2017-06-16 17:36:34 -04:00
Thibault Saunier
6439494f40 validate: overrides: Add a hook about newly added elements in a bin 2017-06-16 17:36:34 -04:00
Thibault Saunier
e0484a7cdd validate: Attach overrides before calling monitor.setup() 2017-06-16 17:36:34 -04:00
Thibault Saunier
237a976d07 validate: Add a way to use config actions in GST_VALIDATE_CONFIG files
The synthax is:

core, action=action-name, param1=1, param2=param2
2017-06-07 11:17:53 -04:00
Thibault Saunier
8d251f12d9 scenario: Fallback to using media_info to determine stream duration
In the case the query duration return CLOCK_TIME_NONE.
2017-06-07 11:17:53 -04:00
Thibault Saunier
b9774b14b5 validate: Do not print \r in actual files 2017-06-07 11:17:53 -04:00
Thibault Saunier
3c62c315a9 validate: Implement RTSP support 2017-06-07 11:17:53 -04:00
Reynaldo H. Verdejo Pinochet
a18cef9c3f validate: reporter: add doc entry for gst_validate_report()'s varargs
Avoids GTK-Doc parser warnings
2017-06-07 00:04:34 -07:00
Reynaldo H. Verdejo Pinochet
4349e11cb0 validate: scenario: fix typo/grammar issues in function documentation 2017-06-07 00:04:34 -07:00
Reynaldo H. Verdejo Pinochet
018ca98fd9 validate: actions: drop needless newline before feature-rank and wait
Stale new-lines messed the help output of gst-validate -t
2017-06-07 00:02:54 -07:00
Reynaldo H. Verdejo Pinochet
7f396ec660 validate: actions: add info on mandatory fields for set-property
Additionally, drop a comment that becomes redundant after adding this
info to the action description
2017-06-06 23:04:52 -07:00
Reynaldo H. Verdejo Pinochet
89fff02bb3 validate: scenario: fix description for 'seek' and 'stop' 2017-06-06 23:04:28 -07:00
Reynaldo H. Verdejo Pinochet
4b57c0bbb4 validate: fix documentation for gst_validate_init() 2017-06-06 18:50:47 -07:00
Thibault Saunier
9d3b760cf3 validate: Make Reporter.runner a MT safe weak reference
It can be used in any thread!
2017-06-01 16:57:50 -04:00
Thibault Saunier
e27367f040 validate: Fix json serialized object leaks 2017-06-01 16:57:50 -04:00
Thibault Saunier
86e9135b56 validate: Use GWeakRefs on monitor target and pipeline
Making it thread safe and more future proof (though having them point
to NULL might not be handled all around).

https://bugzilla.gnome.org/show_bug.cgi?id=782784
2017-06-01 16:57:50 -04:00
Thibault Saunier
8cc1b39919 validate: monitor: Unref our weak reference to the pipeline 2017-05-31 14:53:05 -04:00
Thibault Saunier
8f123cbc6f validate: Make accessing Scenario.pipeline thread safe
The fact that Scenario.pipeline was not accessible in a thread way lead
to the fact that all users had to take the unref the last pipeline ref
in the main thread, otherwise we were crying. This was an ugly
restriction which lead to issue when using scenario on gst-rtsp-server.

This break the API as this commit remove the GstValidateScenario.pipeline
field but it is worth it.
2017-05-31 14:53:05 -04:00
Víctor Manuel Jáquez Leal
29480d006c validate: g_object_newv() is deprecated
Since glib version 2.54, g_object_newv() is deprecated.

This patch changes that function with a simpler g_object_new(),
since no properties are set.

https://bugzilla.gnome.org/show_bug.cgi?id=782860
2017-05-20 12:40:26 +02:00
Edward Hervey
a4b16d54a0 validate-report: Don't leak GError 2017-05-05 15:27:37 +02:00
Edward Hervey
506e537f82 validate-scenario: Don't end up with invalid action name
When replacing an action structure, also update the action name with
the (new) name from the new structure. Otherwise we end up with
a bogus name from the previous (deleted) structure.
2017-05-05 15:27:37 +02:00
Edward Hervey
790e78c4f8 validate-scenario: Avoid invalid memory access
The name of the action comes directly (i.e. not copied) from the
contained GstStructure field. Therefore make sure to take that
name from the proper structure field (copied just before) and
not from an outside one.
2017-05-05 14:43:25 +02:00
Reynaldo H. Verdejo Pinochet
9d33d0ec9c validate: drop unneeded break+concat in some string literals
Perform other related improvements while at it.
2017-04-12 10:31:21 -07:00
Thibault Saunier
26b388c349 validate: Do not expect track switch to be synchronous for playbin3
And let following actions to be executed (setting the action as
INTERLACED) which will make sure the track switch happened at some
point. It means the user has to set the pipeline to PLAYING so we can
make it works but we do not have choice here I think

https://bugzilla.gnome.org/show_bug.cgi?id=781213
2017-04-12 11:28:50 -03:00
Edward Hervey
1f5a900bce validate-pad-monitor: Fix previous commit
Post-send handling was only meant for seek events
2017-04-10 11:43:28 +02:00
Edward Hervey
b156616b5c pad-monitor: Safely store expected seek values
Protect the expected seek values with the same lock as the one
that will be used to read/validate the resulting segments and flush
values.

Avoids races with duplicated seeks (i.e. a seek that was already
sent and handled via another pad, such as in demuxers).

https://bugzilla.gnome.org/show_bug.cgi?id=781112
2017-04-10 09:15:50 +02:00
Reynaldo H. Verdejo Pinochet
22d99ac89c validate: scenario: fix error message on scenario parsing failure 2017-04-05 15:26:31 -07:00
Reynaldo H. Verdejo Pinochet
4ec6bda043 validate: report: avoid pointless var & associated dup/free
Just pass the replacement string literal to _regex_replace().
2017-04-05 14:40:32 -07:00
Reynaldo H. Verdejo Pinochet
38440a5931 validate: report: avoid _strdup_printf() if not needed
g_strdup() is enough for this string literal.
2017-04-05 14:36:10 -07:00
Reynaldo H. Verdejo Pinochet
c8478b87f8 validate: report: fix description for timestamp out of range
Additionally: Fix nits in the descriptions for
SCENARIO_ACTION_EXECUTION_ISSUE and CONFIG_ACTION_TYPE
2017-04-05 12:24:06 -07:00
Reynaldo H. Verdejo Pinochet
d5da8682b2 validate: scenario: improve description blurb for execute-on-idle
Additionally: Fix issues in _register_action_type() documentation.
2017-03-28 15:03:54 -07:00
Reynaldo H. Verdejo Pinochet
eb2353f037 validate: scenario: fix msg on _lookup_feature() failure 2017-03-28 15:03:54 -07:00
Reynaldo H. Verdejo Pinochet
d6169ccb9f validate: scenario: fix documentation for GstValidateAction
Additionally:
 Improve GstValidatePrepareAction documentation
 Correct one-off use of 'eos' instead of EOS.
2017-03-28 15:03:54 -07:00
Reynaldo H. Verdejo Pinochet
fd796d2976 validate: scenario: fix type field in description parameter
Nothing wrong with Sting though :)

Additionally: Fix typo in need-clock-sync's description
2017-03-28 15:03:47 -07:00
Reynaldo H. Verdejo Pinochet
b2ec10ace3 validate: do not check for NULL before g_free()
g_free() is NULL-safe.
2017-03-23 15:21:58 -07:00
Reynaldo H. Verdejo Pinochet
51a921b34b validate: fix issues with a couple of string constants
Fix missing and/or dupplicated separators, bogus breaks, typos, etc.
2017-03-23 15:20:05 -07:00
Nicolas Dufresne
bdd152484c Rename plugin filenames to match plugin names
- libgstvalidateplugin.so -> libgstvalidatetracer.so
- faultinjection -> validatefaultinjection
- gstvalidategtk -> validategtk
- ssim -> validatessim

https://bugzilla.gnome.org/show_bug.cgi?id=779344
2017-03-08 20:04:29 -05:00
Reynaldo H. Verdejo Pinochet
7f5d22f616 validate: report: avoid unnecessary calls to _append_printf
Additionally: simplify %format magic used for padding
2017-03-02 17:15:20 -08:00
Reynaldo H. Verdejo Pinochet
9588ccd02b validate: report: clarify misleading message on skipped actions
Intention is to notify that not all actions were
executed but previous message suggested none were.

Additionally: fix problems in two others
2017-03-02 17:14:37 -08:00
Reynaldo H. Verdejo Pinochet
c62e33a7b8 validate: report: fix _issue_new()'s parameter description 2017-03-02 17:04:16 -08:00
Thibault Saunier
95a5a3a62c validate: Ignore more parser related fields when comparing media caps 2017-02-27 12:12:38 -03:00
Reynaldo H. Verdejo Pinochet
158507585c validate: fix multiple occurrences of 'parametter' 2017-02-17 14:05:00 -08:00
Thibault Saunier
40016b9956 validate:scenario: Report action done with the execution duration 2017-02-17 15:29:41 -03:00
Reynaldo H. Verdejo Pinochet
f92569724a validate: runner: fix error msg for duplicated runner
- Add trailing whitespace to avoid accidental concatenation
- Fix bunch of typos and some grammar while at it
2017-02-16 14:34:33 -08:00
Thibault Saunier
aa006f78c4 meson: Fix Gir *_prefix value to be the sames as with autotools
We want to be able to do GstValidate.Monitor and not
GstValidate.ValidateMonitor.

And do not pass header to the list of sources to build libraries as
it is not needed.
2017-02-15 18:36:55 -03:00
Thibault Saunier
53e878ea8b meson: Do not forget to install headers 2017-02-15 18:03:36 -03:00
Thibault Saunier
cd78d29c56 validate:scenario: Do not switch relative track when no track of type avalaible
This case was not handled and leaded to a division by zero.
2017-02-15 13:52:35 -03:00
Guillaume Desmottes
1cebf3a24f valdate: fix element leak in overide-registry
The elements were never released once created.

https://bugzilla.gnome.org/show_bug.cgi?id=778279
2017-02-08 17:00:08 -03:00
Guillaume Desmottes
4c5ba39bfe validate: pad-monitor: fix caps leaks
These caps were not released when disposing the pad monitor.

https://bugzilla.gnome.org/show_bug.cgi?id=778279
2017-02-08 17:00:08 -03:00
Thibault Saunier
addd76fc85 validate: Mark tests as SKIPPED when installation is missing a GStreamer plugin 2017-02-07 13:12:09 -03:00
Thibault Saunier
9790823d2a validate: Do not fail media check when fields are related to (decoding) elements
In the case of h264 the stream might very well be in `nal` format but the decoder
might not accept it thus the parser converts to `byte-stream`, leading
to a correct stream detection but a failure in the validate-media-check
tool.
2017-02-06 16:02:55 -03:00
Thibault Saunier
bfc24038ba validate: Avoid assertion when trying to pass the config as a caps 2017-01-31 08:25:05 -03:00
Wonchul Lee
74085a4460 validate: fix linker flags for validate plugin
https://bugzilla.gnome.org/show_bug.cgi?id=777938
2017-01-30 16:53:21 +02:00
Thibault Saunier
7daf21d2e7 validate: Handle the case where QUERY has an EMPTY caps filter
In our algorithm describing caps negotiation issues.
2017-01-27 16:44:46 -03:00
Thibault Saunier
2d83384263 validate: Stop comparing buffers offset values in media descriptor
This value can vary for some external reasons and should not matter
for now.
2017-01-24 18:43:52 -03:00
Sebastian Dröge
93fa16f1ba validate-runner: Use correct enum in return value
gst-validate-runner.c:856:7: error: implicit conversion from enumeration type 'GstValidateReportLevel' to different enumeration type 'GstValidateReportingDetails' [-Werror,-Wenum-conversion]
      GST_VALIDATE_REPORT_LEVEL_UNKNOWN);
      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2016-12-22 19:24:22 +02:00