Commit graph

142 commits

Author SHA1 Message Date
Benjamin Otte
981f6185c5 debugging session with wim
Original commit message from CVS:
debugging session with wim
2003-03-13 23:59:36 +00:00
Wim Taymans
4e44ab15b5 - fix unlink between decoupled elements
Original commit message from CVS:
- fix unlink between decoupled elements
- clear the entry point of a group when we remove the element
- fix a potential problem in _merge_groups that could be triggered when
someone is holding an extra refcount to the group.
2003-03-13 18:46:21 +00:00
Benjamin Otte
35391c0907 mutexes want to be free, too
Original commit message from CVS:
mutexes want to be free, too
2003-03-12 21:55:09 +00:00
Wim Taymans
678ea8bbbb Fix optimal scheduler with cothreads
Original commit message from CVS:
Fix optimal scheduler with cothreads
2003-03-11 23:26:18 +00:00
Benjamin Otte
0f2a63a169 Fix bug when joining threads.
Original commit message from CVS:
Fix bug when joining threads.
2003-03-11 21:37:36 +00:00
Benjamin Otte
c3ebe159bd added (hopefully) compatible cothreads emulation by using GThreads.
Original commit message from CVS:
added (hopefully) compatible cothreads emulation by using GThreads.
use '--gst-scheduler=basicgthread' to try it out
includes bugfix for opt to call do_cothreads_init when using threads
2003-03-11 21:01:31 +00:00
Wim Taymans
aab288b33e Never take away the group from a loop based element
Original commit message from CVS:
Never take away the group from a loop based element
2003-03-04 20:55:25 +00:00
Wim Taymans
5bda977d93 Fix the get_wrapper to handle the case where the peer group is disabled
Original commit message from CVS:
Fix the get_wrapper to handle the case where the peer group is disabled
2003-02-28 23:14:50 +00:00
Wim Taymans
f006787228 Stupid bug, always use the runqueue to start scheduling things.
Original commit message from CVS:
Stupid bug, always use the runqueue to start scheduling things.
2003-02-28 17:50:04 +00:00
Wim Taymans
777cae1b0b Schedule next chain on interrupt
Original commit message from CVS:
Schedule next chain on interrupt
2003-02-27 23:47:40 +00:00
Wim Taymans
71775c37e3 - More refcounting fixes that now sustain abuse from spider, key is to be careful when iterating over a list when our...
Original commit message from CVS:
- More refcounting fixes that now sustain abuse from spider, key is to be
careful when iterating over a list when our current pointer could be
removed.
2003-02-27 23:07:30 +00:00
Wim Taymans
e44d5c1d4a Removed an assert that will not hold when the chain currently executing was merged into another one.
Original commit message from CVS:
Removed an assert that will not hold when the chain currently executing
was merged into another one.
2003-02-27 21:23:39 +00:00
Wim Taymans
2b3c56c4d1 print refcount in _show
Original commit message from CVS:
print refcount in _show
2003-02-27 18:29:07 +00:00
Wim Taymans
91d33d766a Added refcounting to the optimal scheduler to guard against modifications during iterations
Original commit message from CVS:
Added refcounting to the optimal scheduler to guard against modifications
during iterations
2003-02-27 18:21:34 +00:00
Wim Taymans
a2b7df0911 More sensible name for the scheduler when compiled without cothreads
Original commit message from CVS:
More sensible name for the scheduler when compiled without cothreads
2003-02-10 20:11:14 +00:00
Erik Walthinsen
0da768dacb Fix various inconsistencies discovered while attempting to fix --disable-*.
Original commit message from CVS:
Fix various inconsistencies discovered while attempting to fix --disable-*.
Uraeus: this should fix the bison.simple problem that you've had.
2003-02-10 07:53:58 +00:00
Wim Taymans
50fccd3791 Clear any queued buffers when going to READY.
Original commit message from CVS:
Clear any queued buffers when going to READY.
2003-02-06 20:22:34 +00:00
Ronald S. Bultje
637326bc8d older gccs crack here
Original commit message from CVS:
older gccs crack here
2003-02-05 17:09:50 +00:00
Wim Taymans
d7e33dd2ce - Use the eventhandler instead of the event function to send events.
Original commit message from CVS:
- Use the eventhandler instead of the event function to send events.
- make the scheduler setup the eventhandlers
- intercept flush events on pads links that can potentially queue data
and flush it.
- some more debugging info in spider.

I can think of one case where this flush might fail: unconnected pads where
the scheduler has not set up the eventhandler yet. I'll come up with a
solution for that soon.
2003-02-03 22:50:55 +00:00
Wim Taymans
b87dbe8c98 Some extra debug info
Original commit message from CVS:
Some extra debug info
2003-02-02 19:14:26 +00:00
Erik Walthinsen
bab7be2c91 revert the revert, switch out the new code for the 'working' code with an #if
Original commit message from CVS:
revert the revert, switch out the new code for the 'working' code with an #if
2003-01-21 20:48:23 +00:00
Wim Taymans
69be241342 Revert 1.66 and 1.67 that seemed to break the unlink scheduler function 1.68 is still added
Original commit message from CVS:
Revert 1.66 and 1.67 that seemed to break the unlink scheduler function
1.68 is still added
2003-01-21 18:39:11 +00:00
Wim Taymans
59a224893a Some more debugging
Original commit message from CVS:
Some more debugging
2003-01-20 23:11:51 +00:00
Wim Taymans
826f848618 - We need to pull UNUSABLE pads too because it might be an event
Original commit message from CVS:
- We need to pull UNUSABLE pads too because it might be an event
2003-01-20 23:10:32 +00:00
Erik Walthinsen
de6bcd890a - added 'remove' argument to _chain_recursive_add to remove elements from old chain before adding to new one
Original commit message from CVS:
- added 'remove' argument to _chain_recursive_add to remove elements from
old chain before adding to new one
- reworked _pad_unlink to be simpler and more correct
2003-01-20 22:22:46 +00:00
Erik Walthinsen
cbbb8371fe moved recursive exit to the front of the call, to make it easier to add prevchain removal
Original commit message from CVS:
moved recursive exit to the front of the call, to make it easier to add prevchain removal
2003-01-20 22:02:46 +00:00
Erik Walthinsen
590604e430 - some minor whitespace cleanups
Original commit message from CVS:
- some minor whitespace cleanups
- rework _chain_add_element to put the element in either the main or
disabled lists depending on whether the element is PLAYING or not.
2003-01-20 21:23:19 +00:00
Wim Taymans
59bd88118f Added interrupt back in spider, make opt scheduler schedule next chain on interrupt (like basic did)
Original commit message from CVS:
Added interrupt back in spider, make opt scheduler schedule next chain
on interrupt (like basic did)
2003-01-12 20:28:08 +00:00
Wim Taymans
89bf2733e9 - return boolean in _interrupt to better inform element
Original commit message from CVS:
- return boolean in _interrupt to better inform element
- fix default bufferpool unref
2003-01-11 16:27:45 +00:00
Iain Holmes
7ec295a0cd Don't remove an element from its chain if it doesn't have a chain
Original commit message from CVS:
Don't remove an element from its chain if it doesn't have a chain
2003-01-11 13:44:33 +00:00
David I. Lehn
5a74101933 more connect -> link updates
Original commit message from CVS:
more connect -> link updates
2003-01-09 22:59:37 +00:00
Thomas Vander Stichele
5529bbc7dc first pass of connect->link gst-plugins and other stuff compiles without change at this point
Original commit message from CVS:
first pass of connect->link
gst-plugins and other stuff compiles without change at this point
2003-01-09 14:15:37 +00:00
Wim Taymans
e8b63df877 Provide intrastructure to not have to pass NULL buffers on errors and interrupts, this should fix some issues with th...
Original commit message from CVS:
Provide intrastructure to not have to pass NULL buffers on errors and
interrupts, this should fix some issues with the optimal scheduler.
2003-01-08 21:33:20 +00:00
Wim Taymans
6653167f11 - Remove annoying \n from cothread INFO/DEBUG messages
Original commit message from CVS:
- Remove annoying \n from cothread INFO/DEBUG messages
- only set an element to PAUSED when it gave an error during PLAYING
- ensure internal consistency of opt scheduler structures
- Added mex_recursion to opt scheduler
2003-01-01 04:28:33 +00:00
Wim Taymans
a03425c8b3 - disable pads when going to PAUSED, we want to make sure no data is passing when an element is not PLAYING.
Original commit message from CVS:
- disable pads when going to PAUSED, we want to make sure no data is
passing when an element is not PLAYING.
- changed the clock sync API, element should now get a ClockID first and
sync on that. This makes it possible to cancel clock requests.
2002-12-27 22:57:13 +00:00
Wim Taymans
2694db659d Fix merge_groups
Original commit message from CVS:
Fix merge_groups
2002-12-26 22:11:04 +00:00
Wim Taymans
abfc7a0ed8 implemented yield/interrupt
Original commit message from CVS:
implemented yield/interrupt
2002-12-21 02:25:54 +00:00
Wim Taymans
4a63ac4d02 Destroy cothreads first before destroying context
Original commit message from CVS:
Destroy cothreads first before destroying context
2002-12-20 18:49:17 +00:00
Wim Taymans
68b53ac4ab - Cleanups
Original commit message from CVS:
- Cleanups
- Added padding to structs
- fixed typechecking/casts
- reduced casts
- implemented remove_element in gstbin
- implemented set index on bin
2002-12-19 21:31:03 +00:00
Wim Taymans
7e16eb73de Delete chains only once..
Original commit message from CVS:
Delete chains only once..
2002-12-19 19:45:13 +00:00
Wim Taymans
e7ccc2ec08 - Factor out some error functions
Original commit message from CVS:
- Factor out some error functions
- handle the cases where multipad get to multi-sink loop/chain based
elements are connected.
2002-12-16 21:21:02 +00:00
Wim Taymans
103d810e2c Decoupled elements needs special case when we look for grouped elements.
Original commit message from CVS:
Decoupled elements needs special case when we look for grouped elements.
2002-12-15 22:52:48 +00:00
Wim Taymans
bd9dc3604f Some cleanups, added FIXMEs
Original commit message from CVS:
Some cleanups, added FIXMEs
2002-12-15 22:02:20 +00:00
Wim Taymans
8f06b55c38 - Removed old deprecated fastscheduler
Original commit message from CVS:
- Removed old deprecated fastscheduler
- ifdef out cothread specific code in optimalscheduler
- added more g_asserts to optimalscheduler
- create separate scheduler called "opt", removed property from scheduler
- fixed iterations property
2002-12-15 18:21:43 +00:00
Wim Taymans
11b376c838 Improve the optimal scheduler, mainly element remove and pad disconnect
Original commit message from CVS:
Improve the optimal scheduler, mainly element remove and pad disconnect
2002-12-15 12:18:04 +00:00
Wim Taymans
0d994e41ad turn off cothreads, fix some typos
Original commit message from CVS:
turn off cothreads, fix some typos
2002-12-09 22:31:16 +00:00
Wim Taymans
f07281a771 Destroying the cothread seems like a good idea here
Original commit message from CVS:
Destroying the cothread seems like a good idea here
2002-12-09 20:33:27 +00:00
Thomas Vander Stichele
7eea6a04b7 fix for parallel installability
Original commit message from CVS:
fix for parallel installability
2002-12-08 14:39:38 +00:00
David Schleef
2fed56b4a8 Removing unused code related to stack allocation (cothreads_stackquery(), gst_scheduler_get_preferred_stack())
Original commit message from CVS:
Removing unused code related to stack allocation (cothreads_stackquery(),
gst_scheduler_get_preferred_stack())
2002-12-06 23:05:26 +00:00
Wim Taymans
86376a2402 Make different names for the different cothreads implementations
Original commit message from CVS:
Make different names for the different cothreads implementations
2002-11-25 00:20:52 +00:00