Mike Brady
539c501174
Fix the volume-changed command caller. A space went missing.
2018-09-28 09:33:10 +01:00
Mike Brady
1b18f45fac
Clang Format Cleanup
2018-09-09 20:43:01 +01:00
Mike Brady
c9b3d2a2f4
Many changes to compilation and linking flags. Stop using HAVE_* flags except where necessary, use CONFIG_* for optional stuff, use HAS_* for immediate definitions, use USE_* for AM definitions. Probably introduced bgs, sigh.
2018-09-09 20:41:31 +01:00
Mike Brady
69642bb7de
Partly clean up use of automake, autoconfig and compiler flags. Works but still not consistent.
2018-09-09 14:12:20 +01:00
Mike Brady
6c6dda9cce
Add ability for the jack back end to automaticlly connect a client when the jack server comes up and to drop the client when a play session is finished.
2018-09-02 17:25:29 +01:00
Mike Brady
76c5dd9285
Stop using SIGUSR1 for cancelling threads, use pthread_cancel and friends instead, fix some memory leaks, add accurate ouput rate calculation to statistics, add "quit" verb to MPRIS and native d-bus interfaces. Probably still buggy.
2018-07-30 10:07:11 +01:00
Mike Brady
ea20840d6c
Make the play function return a status (ignored for the present). Clean up audio_alsa by removing all the die calls and placing the alsa_mutex calls in pthread cleanup routines.
2018-07-29 13:45:42 +01:00
Mike Brady
f153a1f570
Remove some unwanted debug messages
2018-07-22 16:26:47 +01:00
Mike Brady
5f750ba9b7
Fix a few memory leaks.
2018-07-22 16:12:52 +01:00
Mike Brady
f1d4503431
Try to add proper cancellation code to all threads -- not complete, buggy.
2018-07-20 13:50:38 +01:00
Mike Brady
01f2a374b9
Use pthread_cancel to stop a player thread rather than SIGUSR1 and pthread_kill.
2018-07-15 13:54:33 +01:00
Mike Brady
768dc7759a
Tidy up log output and add timers to inform statements. Move more stuff to the player thread pop function. Warn about cancellation points in debug versions of mutex lock and unlock. Ensure they are not cancellation points when debug not in use. Warn about Mac OS X cancellation points.
2018-07-15 11:54:48 +01:00
Mike Brady
d3b79b965d
Make more of the code amenable to pthread cancel
2018-07-14 19:34:31 +01:00
Mike Brady
0d48a3d96a
improve the timed mutex lock code and make it compatible with os x, we hope.
2018-06-21 16:15:53 +01:00
Till Zimmermann
3d1086ea2e
Fixed not connected bug, added version string
2018-06-20 18:58:34 +02:00
Mike Brady
eee822b932
add some more debugging messages.
2018-06-01 16:00:20 +01:00
Mike Brady
d224f47e37
Remove unneeded volume control mutex, adjust some timings, add wait time to slow mutex locks.
2018-06-01 11:21:16 +01:00
Mike Brady
93f410599f
Add debug_mutex_unlock code in place of a regular pthread_mutex_unlock in some places.
2018-06-01 09:28:28 +01:00
Mike Brady
18b37ea254
Replace all pthread_mutex_locks with debug_mutex_lock with a message and timer. Not really tested.
2018-06-01 08:29:48 +01:00
Mike Brady
d392f75c1c
Add warnings to tell us if the time info might overflow, quieten some debug messages in rtp.
2018-04-11 20:56:38 +01:00
Mike Brady
89a6a7d29e
Reformulate the conversion of seconds and nanoseconds into fixed point 64-bit time.
2018-04-11 20:13:53 +01:00
Mike Brady
08a7d4ffb5
Use snprintf in place of sprintf.
2018-04-08 15:36:05 +01:00
Mike Brady
1c437f1aab
Rememeber the debug message of the previous request to lock the mutex.
2018-04-08 12:51:27 +01:00
Mike Brady
7fb8913532
Use snprintf in place of sprintf and initialise mutexes for debugging.
2018-04-08 12:36:00 +01:00
Mike Brady
8bf83c149f
Modify latency detection to use the second word of a sync packet to decide whether to add an extra delay or not. Use all sync packets, not just one with the exception flag. Use 1024 buffers -- a bit profligate, but hey.
2018-04-02 16:42:46 +01:00
Mike Brady
69b3b5d776
Send rtsp events to the metadata thteread for handling. Begin to clean up signals.
2018-03-31 12:39:26 +01:00
Mike Brady
340e365a62
Format after adding a few new elements ot the native dbus interface
2018-03-23 21:29:38 +00:00
Mike Brady
6dc30c6ca9
Add a few new elements ot the native dbus interface
2018-03-23 21:28:28 +00:00
Mike Brady
d692fab9e4
Clean up native dbus interface files and associated changes to existing files.
2018-03-20 16:31:59 +00:00
Mike Brady
fba40733f3
Add more basic remote control stuff, but SPS gets confused about who it can control.
2018-03-18 19:10:45 +00:00
Mike Brady
13931664ed
Check if locks are already acquired.
2018-03-16 15:57:35 +00:00
Mike Brady
9d9b0a1824
Add a Shairport Sync Diagnostics DBus Onject, only part working
2018-03-10 17:01:42 +00:00
Mike Brady
c36a7822a3
Tidy up using clang-format
2018-03-07 14:35:25 +00:00
Mike Brady
07e46565ca
Eliminate almost all warnings from -Wall
2018-03-07 14:29:14 +00:00
Mike Brady
a7d82ae823
Begin to avoid strict aliasing violations -- not finished yet.
2018-03-07 08:58:23 +00:00
Mike Brady
71a9a0cb81
Add timings to warnings and die messages if debugging is enabled.
2018-02-20 09:20:21 +00:00
Mike Brady
283e4930e8
Add support for OpenBSD compilation
2018-02-09 21:11:39 +00:00
Mike Brady
a89742e3a0
Quieten some debug messages.
2018-02-05 18:10:49 +00:00
Mike Brady
a0bb2993df
Add timing to debug messages, make timing sender wait interruptable eveyr 20 ms, thus making TEARDOWN faster.
2018-02-03 20:24:58 +00:00
Mike Brady
f83b86ec23
Add the option to use a flat volume control profile.
2018-01-29 22:28:22 +00:00
Mike Brady
86b72a981a
Fix small compilation bug evident in FreeBSD
2018-01-23 10:18:02 +00:00
Mike Brady
7a73e8fd2a
Brown bag update. Fix compilation errors when neither mpris nor dbus is selected.
2018-01-17 10:32:31 +00:00
Mike Brady
4aaa8d19fe
Implement a rudimentary cover art file writing facility.
2018-01-16 22:22:03 +00:00
Belbo
b98e6a9ac0
Fixed race condition when stopping other threads with SIGUSR1
...
Many threads have a main loop that checks some exit variable in the loop condition. To stop a thread, another thread sets that exit variable, sends a SIGUSR1 to that thread to get it out of blocking calls like read(), write() or select(), and finally joins it.
There was a race condition though: if the signal arrived after the loop condition was checked but before the blocking system function was entered, the target thread essentially ignored the signal, made the blocking call anyway, and could get stuck forever there.
The only way around this race condition is to block SIGUSR1 in the destination thread during the critical time interval. This requires that the blocking system call and the call to unblock SIGUSR1 be an atomic unit. There is only one system call that makes this possible: pselect(), which takes a signal mask to replace the current signal mask with, but only while pselect() runs. Other functions like read(), recv(), write() and sendto() may only be called if we can be sure that they don't block. For this purpose, the file descriptors must be set to non-blocking mode and we must prepend the call to them with a call to pselect() to block until the socket becomes readable/writable.
2017-11-01 20:46:05 +01:00
Belbo
3e54765193
Fixed loop condition in non_blocking_write()
...
The while loop should be repeated, if the last write attempt was a partial write only. In this case, however, rc is greater than 0 (should be 1), not equal to 0.
2017-10-31 23:19:22 +01:00
Belbo
19ff6042cf
Prevent buffer overflow in log functions
...
die(), warn(), debug() and inform() write the string into a limited internal buffer first. They now use vsnprintf() instead of vsprintf() to prevent an overflow of that buffer.
2017-10-31 23:19:21 +01:00
Mike Brady
3001f39bc5
3.1.2 pretty-print using clang-format
2017-09-07 12:02:00 +01:00
Mike Brady
48c1d3eb20
Add fixes suggested by using "infer"
2017-09-03 22:46:57 +01:00
Mike Brady
31a09bb368
Fix some silent issues that were highlighted by "infer"
2017-09-03 15:08:13 +01:00
Mike Brady
d3148eb464
Improve handling of player lock and simplify handling of loss of connection and improve some debug messages
2017-08-27 12:43:42 +01:00
Mike Brady
Till Zimmermann
Mike Brady
Belbo