mirrors/gstreamer
mirrors/gstreamer
1
1
Fork 0
mirror of https://gitlab.freedesktop.org/gstreamer/gstreamer.git synced 2024-11-26 19:51:11 +00:00
Code Issues 1 Projects Releases Wiki Activity

docs/design/part-events.txt: Update some docs.

Browse source 
Original commit message from CVS:
* docs/design/part-events.txt:
Update some docs.
* docs/design/part-block.txt:
* gst/gstpad.c: (gst_pad_is_blocking), (handle_pad_block),
(gst_pad_push_event):
Revert BLOCKING patch, it tries to be smart without really having a
clear idea what or how. So, now we discard all FLUSHING events again on
a blocking pad. Should fix gnonlin again.
This commit is contained in:
Wim Taymans 2006-10-23 15:21:12 +00:00
parent 5861382c2f
commit 340a8fee72
4 changed files with 28 additions and 35 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
ChangeLog
View file

@ -1,3 +1,15 @@
2006-10-23 Wim Taymans <wim@fluendo.com>
* docs/design/part-events.txt:
Update some docs.
* docs/design/part-block.txt:
* gst/gstpad.c: (gst_pad_is_blocking), (handle_pad_block),
(gst_pad_push_event):
Revert BLOCKING patch, it tries to be smart without really having a
clear idea what or how. So, now we discard all FLUSHING events again on
a blocking pad. Should fix gnonlin again.
2006-10-23 Wim Taymans <wim@fluendo.com>
Patch by: Sergey Scobich <sergey dot scobich at gmail dot com>

16
docs/design/part-block.txt
View file

@ -69,19 +69,9 @@ Flushing
is blocking somewhere downstream, the callback is not called or the sync block
does not return.
In order for the block to happen, a FLUSH_START needs to be:
_ either sent directly on the downstream blocking element/pad so that it
release the stream lock, and it gives a chance for the elem1.src pad to
block.
_ A FLUSH_START is sent downstream from an upstream element, causing all the
pads down to the blocking element/pad (including elem1.src) to go to the
FLUSHING state. A FLUSH_STOP can now be sent downstream, causing all the
pads down to the previously blocking element to unset their FLUSHING state.
The next push will then properly block on elem1.src.
In this case, the pads have to be careful when handling the FLUSH events
forwarding. Those events should only be forwarded downstream is the BLOCKED
pad was previously BLOCKING.
In order for the block to happen, a FLUSH_START needs to be sent directly on
the downstream blocking element/pad so that it releases the stream lock, and it
gives a chance for the elem1.src pad to block.
Use cases:

8
docs/design/part-events.txt
View file

@ -122,12 +122,12 @@ A NEWSEGMENT event should be generated as soon as possible in the pipeline and
is usually generated by a demuxer or source. The event is generated before
pushing the first buffer and after a seek, right before pushing the new buffer.
The NEWSEGMENT event can be sent from both the application and the streaming
thread and should be serialized with the buffers.
The NEWSEGMENT event should be sent from the streaming thread and should be
serialized with the buffers.
Buffers should be clipped within the range indicated by the newsegment event
start and stop values. Sinks must to drop buffers with timestamps out
of the indicated newsegment range.
start and stop values. Sinks must drop buffers with timestamps out of the
indicated newsegment range.
If a newsegment arrives at an element not preceeded by a flush event, the
streamtime of the pipeline will not be reset to 0 so any element that syncs

27
gst/gstpad.c
View file

@ -1068,9 +1068,7 @@ gst_pad_is_blocking (GstPad * pad)
g_return_val_if_fail (GST_IS_PAD (pad), result);
GST_OBJECT_LOCK (pad);
/* the blocking flag is only valid if the pad is not flushing */
result = GST_OBJECT_FLAG_IS_SET (pad, GST_PAD_BLOCKING) &&
!GST_OBJECT_FLAG_IS_SET (pad, GST_PAD_FLUSHING);
GST_OBJECT_UNLOCK (pad);
@ -3210,9 +3208,7 @@ gst_ghost_pad_save_thyself (GstPad * pad, xmlNodePtr parent)
* the pad block happens.
*
* During the actual blocking state, the GST_PAD_BLOCKING flag is set.
* The GST_PAD_BLOCKING flag is unset when the pad was unblocked without a
* flush. This is to know whether the pad was blocking when GST_PAD_FLUSHING
* was set.
* The GST_PAD_BLOCKING flag is unset when the pad was unblocked.
*
* MT safe.
*/
@ -3271,14 +3267,11 @@ handle_pad_block (GstPad * pad)
"Waiting to be unblocked or set flushing");
GST_OBJECT_FLAG_SET (pad, GST_PAD_BLOCKING);
GST_PAD_BLOCK_WAIT (pad);
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_BLOCKING);
/* see if we got unlocked by a flush or not */
/* see if we got unblocked by a flush or not */
if (GST_PAD_IS_FLUSHING (pad))
goto flushing;
/* If we made it here we were unblocked and not flushing, remove the
* blocking flag now. */
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_BLOCKING);
}
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "got unblocked");
@ -3880,10 +3873,12 @@ gst_pad_push_event (GstPad * pad, GstEvent * event)
case GST_EVENT_FLUSH_START:
GST_PAD_SET_FLUSHING (pad);
if (G_UNLIKELY (GST_PAD_IS_BLOCKING (pad))) {
if (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad))) {
/* flush start will have set the FLUSHING flag and will then
* unlock all threads doing a GCond wait on the blocking pad. This
* will typically unblock the STREAMING thread blocked on a pad. */
GST_LOG_OBJECT (pad, "Pad is blocked, not forwarding flush-start, "
"doing block signal.");
GST_PAD_BLOCK_SIGNAL (pad);
goto flushed;
}
@ -3891,13 +3886,9 @@ gst_pad_push_event (GstPad * pad, GstEvent * event)
case GST_EVENT_FLUSH_STOP:
GST_PAD_UNSET_FLUSHING (pad);
/* If pad was blocking on something when the pad received flush-start, the
* BLOCKING flag was never cleared. we don't forward the flush-stop event
* either then but unset the blocking flag. */
if (G_UNLIKELY (GST_PAD_IS_BLOCKING (pad))) {
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_BLOCKING);
GST_LOG_OBJECT (pad,
"Pad was previously blocking, not forwarding flush-stop");
/* if we are blocked, flush away the FLUSH_STOP event */
if (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad))) {
GST_LOG_OBJECT (pad, "Pad is blocked, not forwarding flush-stop");
goto flushed;
}
break;