Commit graph

15677 commits

Author SHA1 Message Date
Sebastian Dröge
db1e4422b6 gst: Store more basic type GTypes in variables
Micro optimization to change a function call to a variable access
for all our basic types.
2014-06-19 08:06:55 +02:00
Sebastian Dröge
f47a95f4cb value: Store our fundamental type GTypes in variables
Micro optimization to change a function call to a variable access
for all our basic types.
2014-06-19 08:06:55 +02:00
Edward Hervey
b0b20b44e5 gstvalue: Speed up gst_value_intersect/_subtract
Both gst_value_intersect and gst_value_subtract will call
gst_value_compare if one of their arguments isn't a list.

gst_value_compare will then re-do a check to see if one of
the arguments is a list (for the special case of comparing a unitary
value with a list of length 1).

The problem is that the various G_VALUE_HOLDS represent an expensive
amount of calling gst_value_compare (almost half of it) to see if
the provided arguments are list. These checks can be done without
when we know that the arguments aren't lists.

* Create a new "nolist" gst_value_compare which avoids that special
  case comparision

Benchmarks:
 valgrind/callgrind: average speedup in instruction calls for
 gst_value_intersect and gst_value_subtract is around 56% (Makes 63%
 of the calls it used to take previously)

 tests/benchmarks/capsnego: With default settings (depth 4, children 3
 607 elements), time taken for transition from READY to PAUSED:
   Before : 00.391519153
   After  : 00.220397492
    56% of the time previously used, +77% speedup

https://bugzilla.gnome.org/show_bug.cgi?id=731756
2014-06-17 16:27:07 +02:00
Tim-Philipp Müller
5c62bce85c tests: remove some cruft from the bufferlist test
Buffers no longer carry caps, and bufferlists don't have
groups where buffers may need to be merged into one any more.
2014-06-17 14:39:00 +01:00
Tim-Philipp Müller
b314ac8291 tests: add test for gst_buffer_list_remove() 2014-06-16 20:30:32 +01:00
Tim-Philipp Müller
87fcf1e078 bufferlist: fix buffer leak in _remove() 2014-06-16 20:29:56 +01:00
Tim-Philipp Müller
f6e2dee6f9 flowcombiner: fix g-i transfer annotations 2014-06-16 09:18:45 +01:00
Edward Hervey
ecd8fd3511 flowcombiner: Fixed GBoxedCopyFunc
I'll just quote the most interesting man in the world:

"I don't usually push commits, but when I do I don't compile it
first"
2014-06-16 08:43:57 +02:00
Tim-Philipp Müller
bd26fa7ccb devicemonitor: some docs additions and fixes 2014-06-14 16:30:49 +01:00
Tim-Philipp Müller
97fe853e42 win32: add exports for new get_type() function 2014-06-14 16:28:48 +01:00
Tim-Philipp Müller
b05eeb9df4 flowcombiner: keep a ref to the pads we're using
Needed for use via the boxed type.

https://bugzilla.gnome.org/show_bug.cgi?id=731355
2014-06-14 11:32:42 +01:00
Tim-Philipp Müller
49fedb521f flowcombiner: add boxed type for bindings
https://bugzilla.gnome.org/show_bug.cgi?id=731355
2014-06-14 10:54:41 +01:00
Evan Nemerson
2a14482c62 introspection: minor annotation additions
https://bugzilla.gnome.org/show_bug.cgi?id=731541
2014-06-12 09:26:38 +01:00
Evan Nemerson
ba1212f63b introspection: include gstversion.h in GIR generation
https://bugzilla.gnome.org/show_bug.cgi?id=703021
2014-06-12 08:22:20 +01:00
Nicolas Dufresne
d99e621533 bytereader: Use concistant derefence method
This is minor style fix to not mix *var and var[N].
2014-06-10 10:23:13 -04:00
Sungho Bae
d3b2f6e4b8 bytereader: Use pointer instead of index access
Currently the scan uses Boyer-moore method and its performance is good.
but, it can be optimized from an implementation of view.

The original scan code is implemented by byte array and index-based access.
In _scan_for_start_code(), the index is increasing from start to end and the
base address of the byte array is referred to as return value.

In the case, index-based access can be replaced by pointer access, which
improve the performance by removing index-related operations.

Its performace is enhanced by approximately 8% on arm-based embedded devices.
Although it seems trivial, it can affect the overall performance because the
_scan_for_start_code() function is very often called when H.264/H.265 video is
played.

In addition, the technique can apply for all architectures and it is good in
view of readability and maintainability.

https://bugzilla.gnome.org/show_bug.cgi?id=731442
2014-06-10 09:35:38 -04:00
Tim-Philipp Müller
12088d87bb globaldevicemonitor: prettify header 2014-06-07 10:13:56 +01:00
Tim-Philipp Müller
9c4e1d3689 tests: add unit test for queuearray expansion from 1
https://bugzilla.gnome.org/show_bug.cgi?id=731349
2014-06-07 09:46:42 +01:00
Evan Nemerson
1cd4bd64b0 queuearray: fix expanding size of queue from 1
Without we would not actually expand and access
memory beyond the allocated region for the array.

https://bugzilla.gnome.org/show_bug.cgi?id=731349
2014-06-07 09:45:39 +01:00
Evan Nemerson
75e0a3c676 dataqueue: clear up documentation of gst_data_queue_new
The gpointer argument is passed to all three callbacks, not just one.

https://bugzilla.gnome.org/show_bug.cgi?id=731302
2014-06-07 09:30:14 +01:00
Evan Nemerson
1fe6d3ad08 introspection: fix some minor annotation bugs
https://bugzilla.gnome.org/show_bug.cgi?id=730982
2014-06-06 15:15:12 -04:00
Evan Nemerson
25af4ad652 base: use correct syntax in documentation more consistently
Previously, many constants were prefixed with # or unprefixed,
some functions and macros were prefixed with # instead of suffixed
with (), etc.

https://bugzilla.gnome.org/show_bug.cgi?id=731293
2014-06-06 15:09:21 -04:00
zhouming
2793f808ee baseparse: Pass rate of input segment to output segment
https://bugzilla.gnome.org/show_bug.cgi?id=729701
2014-06-06 13:53:35 +03:00
Vincent Penquerc'h
a8ca56b42d gstbuffer: factor three flags-to-string loops 2014-06-04 12:30:36 +01:00
Tim-Philipp Müller
b65a7a7034 info: make printing datetimes work with GST_PTR_FORMAT 2014-06-03 23:49:26 +01:00
Tim-Philipp Müller
8998cf840d datetime: change internal implementation to mini object
And move type stuff from GstValue to GstDateTime.
2014-06-03 23:49:21 +01:00
Wim Taymans
f446280740 downloadbuffer: fix uninitialized variable 2014-06-03 22:19:33 +02:00
Wim Taymans
45c61db152 downloadbuffer: improve start/stop in buffering query
The start and stop should represent the currently downloading region.
The estimated-total should represent the remaining time to download
the currently downloading region. This makes it a lot more useful
for applications because they can then use those values to update
the fill region and use the estimated time to delay playback.
Update the docs with this clarification.
2014-06-03 22:17:17 +02:00
Vincent Penquerc'h
557bb0de16 identity: add static and const where appropriate 2014-06-03 15:36:15 +01:00
Vincent Penquerc'h
1e063bc6e6 identity: fix potential buffer overflow
Coverity 1037155
2014-06-03 15:36:11 +01:00
Wim Taymans
8cbc35a93d downloadbuffer: reset read and write positions
Reset the read and write positions right after we open the file or flush
it. We are also in the buffering state with 0 percent buffered when we
start.
2014-06-03 14:49:44 +02:00
Wim Taymans
36d4c4c218 info: first handle all miniobjects, then GObjects
First handle all miniobjects before we attempt to dereference the first
field pointer and look at the GType. With the recent glib change to
speed up G_IS_OBJECT, this causes crashes on miniobjects otherwise.
2014-06-03 14:47:17 +02:00
Wim Taymans
460be3e124 info: GstDateTime does not have a GType as first field
GstDateTime does not have the GType as the first field so we can't use
it to detect its type.
2014-06-03 14:46:11 +02:00
Wim Taymans
e457739ac9 info: use macros to check types
Use the macros to check the type of objects instead of directly poking
at the first field.
2014-06-03 14:45:22 +02:00
Tim-Philipp Müller
64c71323a6 globaldevicemonitor: connect sync-message signal on the right object
Fixes criticals at runtime and makes stuff actually work.
2014-06-01 23:51:20 +01:00
Sebastian Dröge
a91b51c9b7 typefind: Keep still meaningfull pending events on FLUSH_STOP
Only EOS and segment should be deleted in that case.

https://bugzilla.gnome.org/show_bug.cgi?id=709868
2014-05-31 17:35:52 +02:00
Sebastian Dröge
066fdcd432 Revert "miniobject: Add missing (nullable) annotations"
This reverts commit 96361e9b5c.

This was not supposed to be pushed yet!
2014-05-30 09:13:16 +02:00
Sebastian Dröge
67138f4ee8 bufferpool: It's pool, not poo... even when talking about flushing 2014-05-30 09:12:50 +02:00
Philip Withnall
96361e9b5c miniobject: Add missing (nullable) annotations
gst_mini_object_replace() can take NULL mini-objects.

https://bugzilla.gnome.org/show_bug.cgi?id=730873
2014-05-30 09:12:49 +02:00
Thiago Santos
baf1688474 tests: multiqueue: fix leaks 2014-05-30 01:42:17 -03:00
Evan Nemerson
e10266e3f3 docs: convert NULL, TRUE, and FALSE to %NULL, %TRUE, and %FALSE
This should help improve documentation generated for
languages other than C.

https://bugzilla.gnome.org/show_bug.cgi?id=730961
2014-05-30 00:20:27 +01:00
Tim-Philipp Müller
88191aa880 docs: fix type in GstObject docs 2014-05-30 00:14:01 +01:00
Evan Nemerson
4397150d54 bufferpool: fix gst_buffer_pool_has_option() documentation
https://bugzilla.gnome.org/show_bug.cgi?id=730962
2014-05-29 23:18:18 +01:00
Thiago Santos
1911554cff element: set pads need-parent flag to false when removing
When a pad is added the need-parent flag is set to true, so when
they are removed the flag should be set back to false

This was preventing GstPads to be reused in elements (removed and
later re-added). A unit tests was added to verify that this is
working now.

The use case is tsdemux that has a program-number property and
allows the user to switch programs. In order to do that tsdemux
will remove the pads of the current program and add from the new
ones. The removed pads are kept in the demuxer for later if the
user selects the old program again.
2014-05-29 17:01:20 -03:00
Thiago Santos
01c9ae2630 multiqueue: post buffering message when queues flush
The buffering status goes back to 0, so inform the application about it

https://bugzilla.gnome.org/show_bug.cgi?id=726423
2014-05-29 16:51:41 -03:00
Nicolas Dufresne
9abc55491d gitignore: Ignore VIM swap files 2014-05-29 14:39:36 -04:00
Tim-Philipp Müller
ba3073ffce pad: two minor docs fixes 2014-05-27 13:37:05 +01:00
Tim-Philipp Müller
2f265b7571 flowcombiner: beautify headers a little 2014-05-27 10:09:02 +01:00
Tim-Philipp Müller
a6d9c1a3c3 docs: add GstFlowCombiner 2014-05-27 10:05:51 +01:00
Tim-Philipp Müller
0b194cf664 base: include flowcombiner header from base.h 2014-05-27 09:55:27 +01:00