| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|\ \ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
- handle setVideoSurface in deferred action, and return
to client immediately
- handle GenericSource's prepareAsync on its own looper,
instead of sharing looper with NuPlayer
- let HTTPLiveSource share looper with LiveSession, instead
of NuPlayer
- remove reflector in RTSPSource
Bug: 16892748
Change-Id: I1aed557320052012065f5a90adbcb03c238da988
|
| |_|/
|/| |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Offload audio playback is not guaranteed even if AudioSystem says it
is allowed. Create AudioSink early to verify offload is really possible.
Move AudioSink open / close into functions.
Bug: 16732303
Bug: 16978805
Change-Id: Ie1c73a96656863c1281bed3280a84b86d3cbadf5
|
|/ /
| |
| |
| |
| | |
Bug: 16892521
Change-Id: I0e2da4134a37e632f9f2c29d1d27d8fded7a7863
|
|\ \
| |/
|/| |
|
| |
| |
| |
| |
| | |
Bug: 16885604
Change-Id: Ib6c60f0ef8cf72aa1aac67a75121142bff304018
|
|/
|
|
| |
Change-Id: If5251f89b881e7f268e11a06cd3685d794c958b6
|
|
|
|
|
|
|
|
| |
need to consider rotation when notifying video size change after
initial prepare, in addition to after an output format change.
Bug: 16870964
Change-Id: Icd1c9e901ac932e720ce95b8a983c0edcf251c35
|
|
|
|
|
|
| |
Bug: 16708180
Change-Id: I9d578ef5e2edaed50279d28d3831c68556468f39
|
|
|
|
|
|
|
|
|
|
|
|
| |
The previous patch results in MediaBuffer leakage.
This reverts commit 69a85b792c31033a99c8858e3b1a3c2ea68b6278.
Bug: 14955925
Bug: 16303659
Bug: 16467066
Bug: 16849601
Change-Id: Ib1892b7603a97e12b7ee228fd5a4009700cdc988
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
Bug: 15153976
Bug: 15763638
Bug: 16351654
Change-Id: I4462276d4b7342647286a0ca4be11692ce52ff6d
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- prefetch data for sniffing
- notify error instead of crashing if extractor is NULL
Bug: 16818302
Change-Id: I56ff4996d99ac2811d19d141f7ff7acdd7c1da17
|
|\ \ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Bug: 14955925
Bug: 16303659
Bug: 16467066
Change-Id: I7c258b986ae711357bd7e636b79853f8f3959d32
|
|\ \ \ |
|
| |/ /
| | |
| | |
| | |
| | | |
Bug: 16786456
Change-Id: I06f18b0e9c3f65e3e127369baa1f9ea8c9307762
|
|/ /
| |
| |
| |
| |
| |
| |
| | |
Pass audio aac sub formats in offloadinfo according
to aac profile. Audio HAL can take decision about offload
using DSP capabilities
Change-Id: Ice277e8b6561b7a7db92c474f23f8041ebb35e8c
|
|\ \ |
|
| |/
| |
| |
| |
| | |
Bug: 16385674
Change-Id: Ib322cc13488dcadf2409921272cd469304b0af37
|
| |
| |
| |
| |
| | |
Bug: 16641557
Change-Id: I433158e6e585b4714cfd6d89562ecd0587a07ae2
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Assumption:
NuPlayer must see both (a/v) discontinuities before resuming.
Example issue:
Suppose we're changing the video track, so the video source queues a
DISCONTINUITY_FORMATCHANGE, and the audio source queues a DISCONTINUITY_NONE.
Consider this sequence of events without this change:
1. audio discontinuity seen by player
2. mFlushingAudio = FLUSHED
3. video discontinuity seen by player
4. mFlushingAudio != NONE so no performScanSources is queued
5. video decoding shuts down without being restarted
Bug: 15153976
Change-Id: I659550d9c0f883828103142abead857f576f25b1
|
|
|
|
|
|
| |
Bug: 15153976
Change-Id: Idfce856b4ca5fa7853da857757490798bc3ffe1e
|
|
|
|
|
| |
Bug: 16641557
Change-Id: I77aad6e710117a16093726bf3b36360904fbe148
|
|\ |
|
| |
| |
| |
| |
| | |
Bug: 16653284
Change-Id: I54165041da5a13498d627eee1b3ec59ef3c923b0
|
|/
|
|
|
|
|
|
| |
This reverts commit bace25b0fbcd5101d2db288b0c70c4a239f0e815.
This reverts commit eb1735e343a93830df259ae8882160bb0d79dcb5.
This reverts commit 05312bc7478feec11d9ae88e951c0857a7a3f28d.
Bug: 16660839
|
|
|
|
| |
Change-Id: I10343151f930a4ab778a6d2574ab5165af2013ba
|
|
|
|
|
| |
Bug: 16385674
Change-Id: I954cc463b55af4b93c6a6e714115db4fda729dc8
|
|\ |
|
| |
| |
| |
| |
| | |
Bug: 16385674
Change-Id: Ifa9fbc17a1276aa774d57b82cdce500876a0b543
|
| |
| |
| |
| |
| | |
Bug: 16467066
Change-Id: I3674e07779a59f4d4fa639d066ed1988df89b3f5
|
|\ \
| |/ |
|
| |
| |
| |
| |
| | |
Bug: 15153976
Change-Id: I522b1f9f0ffedf4edbea03a6654a6dbc0262860a
|
|/
|
|
| |
Change-Id: Icde3d65c964b2a13fb1c6636adcce52ae048a3fb
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- handle widevine:// scheme
- add separate looper for renderer (as it can block initial buffer
handling if all buffers are used)
- initiate secure codecs before source is started
- don't read secure buffers
- share ACodec's input buffers with Widevine source
on the decoder side
- keep track of mediabuffers released by widevine source
- keep track of dequeued input buffers (for safety)
- release mediabuffer when buffer is subsequently dequeued. (This
was hardcoded into OMXCodec to do this when buffer-empties message
was handled, but MediaCodec does not support such functionality.)
Bug: 15699665
Change-Id: I4a369443294e45c644be8b0257010e52db1d7c9b
|
|
|
|
|
| |
Bug: 15699665
Change-Id: Ided823bd0b1118bbabb288cf62d6389518f820a9
|
|
|
|
|
| |
Bug: 15153976
Change-Id: I0204c4188d485cda026497469c7cde24f7bd5c95
|
|
|
|
| |
Change-Id: Ic83973339fb46a83b48382e6097925f45d200867
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
in case of explicit discontinuity (#EXT-X-DISCONTINUITY tag in the
playlist), do not discard packets that's already queued.
(when seeking, the old fetcher will be discarded so this won't affect
seeking case.)
Bug: 15020526
Change-Id: I5d3fc489df35e3bc44acb1ee7851571cb99cb56b
|
|/
|
|
|
| |
bug: 15917805
Change-Id: I824fe7eea807f8faba6b149c31890b7a5df87825
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also get duration from mvhd if available, use that in addition to track
durations to determine total length of file, clear track header
struct so we don't read uninitialized garbage from it, and take
composition time offset into account for proper ordering when B frames
are used.
b/15669839
Change-Id: Ibf3e35b5c7299bac11d0e78a391545fc325dd2d0
|
|
|
|
|
|
| |
Bug: 15470448
Change-Id: Ic6a527f5c35a8ee0a08a5b043336e4d193216083
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(this is in preparation for supporting other cc source)
- split into two methods: getTrackCount() and getTrackInfo()
- move track info parcelling to NuPlayer
- parcel in the mime type of the subtitle format
Bug: 15470448
Change-Id: If00724d8c3a2b2319cb9c5f29d3fe76347bfe947
|
|
|
|
|
|
|
|
|
|
| |
This prevents decoder from requesting new buffer until the decoder
is resumed, and prevents processing a potential DISCONTINUITY while
the player is still flushing.
Bug: 13133027
Change-Id: I2f9fa9f00c8583aa6908809cb7c31ddde07cfaf0
|
|
|
|
|
|
|
| |
Fixes OnSeekCompleted callback and calling getDuration immediately
after prepare().
Change-Id: Ie58c509005cded278a0e50c87240b0a2d920b7d7
|
|
|
|
|
| |
Bug: 11785204
Change-Id: I1455bfc683469c7a69e565b179aceacbc5c459f5
|
|
|
|
|
|
|
| |
Remove software renderer support from NuPlayer because it is included
in MediaCodec.
Change-Id: I06e001df76df2e18ea8169620ff33aeec57e5ca2
|
|
|
|
|
|
| |
Avoid reinstantiating decoder if seamless format switch is supported.
Change-Id: I2c2be08d6da90cc835ec747d04a76db2313dfc7c
|