Max Kellermann
ced3f320eb
output/MultipleOutputs: reduce lock/unlock calls in EnableDisable()
...
Use ScopeLock to manage the lock; don't unlock after obtaining the
"really_enabled" flag; keep the same lock during EnableWait() /
DisableWait().
2016-12-14 07:54:05 +01:00
Max Kellermann
a9d7293818
output/Thread: wake up the player thread periodically while playing
...
Without this, the pipe would run empty very often, which may result in
an xrun if the roundtrip to the PlayerThread and back takes too long.
By waking up the PlayerThread before the pipe runs empty, we make MPD
much more latency tolerant, which is a major optimization.
2016-12-13 22:39:49 +01:00
Max Kellermann
265ad4b96a
output/Thread: eliminate redundant nullptr check
2016-12-13 22:20:41 +01:00
Max Kellermann
92eeb4969f
output/Thread: don't wake up the player after receiving command in Play()
...
If we have a pending command, then the player thread is already awake.
Even if not, we'll wake it up as soon as we have finished the command.
2016-12-13 22:17:06 +01:00
Max Kellermann
6fcddaf8fa
output/Thread: use AtScopeExit() to revert the in_playback_loop flag
2016-12-13 22:15:52 +01:00
Max Kellermann
41fd583fbc
output/Thread: eliminate one ScopeLock
2016-12-13 21:57:03 +01:00
Max Kellermann
ceba6816de
output/Thread: split Open()
...
Simplify the code, eliminate duplicate rollback code and eliminate the
"goto" kludge.
2016-12-13 21:46:27 +01:00
Max Kellermann
093cb475bb
output/Thread: simplify the retry_audio_format declaration
2016-12-13 21:34:43 +01:00
Max Kellermann
d44790b35f
output/Thread: reduce locking further in Open()
2016-12-13 21:26:28 +01:00
Max Kellermann
f8164a3dd1
output/Thread: unlock mutex during OpenFilter()
2016-12-13 21:24:14 +01:00
Max Kellermann
c20126598f
output/Thread: remove unused variable
2016-12-13 21:24:05 +01:00
Max Kellermann
6c818bb37a
AudioFormat: add method WithMask(), shortcut for ApplyMask()
2016-12-13 20:57:46 +01:00
Max Kellermann
df91f3738a
output/Thread: remove a useless out_audio_format setting
...
Now that I can see the Reopen() method clearly, it is obvious that the
out_audio_format setting will get overwritten by the following Open()
call.
2016-12-13 20:43:29 +01:00
Max Kellermann
b5b268f606
output/Thread: remove another redundant "open" check
2016-12-13 20:41:54 +01:00
Max Kellermann
ce6b8c94a2
output/Thread: convert redundant runtime check to assertion
2016-12-13 20:29:16 +01:00
Max Kellermann
6e643fe58b
output/Thread: don't deinitialize the pipe in Close()
2016-12-13 20:28:19 +01:00
Max Kellermann
2bf91a0487
output/Internal: merge REOPEN and OPEN
2016-12-13 20:07:00 +01:00
Max Kellermann
2775d747ac
output/Thread: move MusicPipe code to class SharedPipeConsumer
2016-12-13 16:33:23 +01:00
Max Kellermann
99659e4cf9
release v0.19.21
...
-----BEGIN PGP SIGNATURE-----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=qsCR
-----END PGP SIGNATURE-----
Merge tag 'v0.19.21'
release v0.19.21
2016-12-13 11:00:39 +01:00
Max Kellermann
7146f825b2
decoder/ffmpeg: fix double free bug
...
From the avformat_open_input() API documentation:
"Note that a user-supplied AVFormatContext will be freed on failure."
https://bugs.musicpd.org/view.php?id=4607
2016-12-13 08:34:05 +01:00
Max Kellermann
dc05dd7ca1
output/Multiple: make chunk_is_consumed_in() an AudioOutput method
2016-12-11 21:59:28 +01:00
Max Kellermann
57dd344f3b
decoder/wavpack: implement WavpackStreamReader64 if available
2016-12-10 00:24:54 +01:00
Max Kellermann
f7f59df8aa
decoder/wavpack: move code to WavpackInput methods
2016-12-10 00:17:46 +01:00
Max Kellermann
0525a6f90f
decoder/wavpack: use WavpackSeekSample64() if available
2016-12-10 00:12:47 +01:00
Max Kellermann
7b4305d81b
decoder/wavpack: seek errors are fatal
...
The libWavPack documentation says:
"After a FALSE return the file should not be accessed again (other
than to close it); this is a fatal error."
2016-12-10 00:11:29 +01:00
Max Kellermann
94e9e93fa4
ReplayGainGlobal: move replay_gain_config into struct Config
...
Yet another global variable removed.
2016-12-10 00:03:29 +01:00
Max Kellermann
d2a1294e9a
Main: add struct Config, empty currently
2016-12-10 00:00:50 +01:00
Max Kellermann
30bd190b41
decoder/wavpack: implement scan_stream()
2016-12-09 23:23:35 +01:00
Max Kellermann
cf93cd9307
decoder/wavpack: convert WavpackInput::client to pointer
2016-12-09 23:23:02 +01:00
Max Kellermann
f40816e063
decoder/wavpack: add WavpackOpen() wrappers which throw exception
2016-12-09 23:15:40 +01:00
Max Kellermann
3759f29852
decoder/wavpack: move wavpack_scan_file() down
2016-12-09 23:15:06 +01:00
Max Kellermann
bc5555933f
decoder/wavpack: support native DSD
2016-12-09 23:08:54 +01:00
Max Kellermann
ead7ae8504
decoder/wavpack: use template for format_sample_X()
2016-12-09 22:53:01 +01:00
Max Kellermann
ad5abba8c6
decoder/wavpack: move 8 and 16 bit conversion to separate functions
2016-12-09 22:49:07 +01:00
Max Kellermann
61296cedd2
decoder/wavpack: rename format_samples_float() to format_samples_nop()
2016-12-09 22:47:32 +01:00
Max Kellermann
d9c29a2c4f
decoder/wavpack: use std::copy_n()
2016-12-09 22:23:19 +01:00
Max Kellermann
f5172e6b84
decoder/wavpack: use int16_t instead of uint16_t
...
Technically, these samples are signed. There was no practical
difference, only the declaration was formally wrong.
2016-12-09 22:21:31 +01:00
Max Kellermann
544bcbe903
decoder/wavpack: use WavpackGetNumSamples64() if available
2016-12-09 22:09:32 +01:00
Max Kellermann
f35dab1b56
decoder/wavpack: check WavpackGetNumSamples()==-1
2016-12-09 21:46:29 +01:00
Max Kellermann
caee420087
decoder/wavpack: move code to GetDuration()
2016-12-09 21:45:22 +01:00
Max Kellermann
e8c374113e
decoder/wavpack: make is_float const
2016-12-09 21:30:31 +01:00
Max Kellermann
98a241a730
decoder/wavpack: use WavpackGetReducedChannels()
...
Since we don't use OPEN_2CH_MAX, this should be equal to
WavpackGetNumChannels() - but who knows, maybe a future libWavPack
version breaks this assumption.
2016-12-09 21:23:38 +01:00
Max Kellermann
be8297d6f6
decoder/wavpack: basic DSD support (WavPack 5)
...
Enable OPEN_DSD_AS_PCM if available. No OPEN_DSD_NATIVE yet.
https://bugs.musicpd.org/view.php?id=4606
2016-12-09 21:03:47 +01:00
Max Kellermann
09da80e9f4
tag/ApeTag: don't export ape_tags[]
2016-12-09 20:59:07 +01:00
Max Kellermann
df4ecf63a8
decoder/wavpack: disable OPEN_TAGS
...
Use MPD's internal APE tag code, which works just as well. This
removes some duplicate code for reading tags and ReplayGain.
2016-12-09 20:53:54 +01:00
Max Kellermann
b4d1937857
decoder/wavpack: change norm_offset to 0
...
MPD uses a floating point range from -1 to +1 internally - why ask
libwavpack to use -32768..+32768 only to reduce it back to -1..+1
afterwards?
2016-12-09 20:53:45 +01:00
Max Kellermann
0be5fbdc35
decoder/wavpack: use AtScopeExit()
2016-12-09 20:53:04 +01:00
Max Kellermann
ae7e25ea65
release v0.19.20
...
-----BEGIN PGP SIGNATURE-----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=opJg
-----END PGP SIGNATURE-----
Merge tag 'v0.19.20'
release v0.19.20
2016-12-09 20:18:54 +01:00
Max Kellermann
e7353ec7e7
Queue: "setprio" re-enqueues old song if priority has been raised
...
This commit changes a minor queue priority design to something which
makes a little bit more sense.
Previously, a song that had already been played would only be
re-enqueued if its priority had just been raised above the current
song's. This means that if it was already above, it was not
re-enqueued. That is a surprising behavior, because users expect a
song to be played when its priority is raised.
Now the song is always re-enqueued if its priority is raised (and
above the current song's - no matter if it has already been above
before).
https://bugs.musicpd.org/view.php?id=4592
2016-12-09 13:02:26 +01:00
Max Kellermann
5013de6770
LogInit: convert use_stdout flag to out_fd=STDOUT_FILENO
2016-12-04 20:15:45 +01:00