zdoom/fluidsynth
0
0
Fork 0
mirror of https://github.com/ZDoom/fluidsynth.git synced 2025-02-07 17:01:18 +00:00
Code Issues Projects Releases Packages Wiki Activity
1835 commits 10 branches 46 tags 9.2 MiB
Commit graph

915 commits

Author SHA1 Message Date
derselbst
a9662bf331 fix gcc8 complaining about stringop-truncation 2018-12-14 19:42:08 +01:00
derselbst
4f1f7349db fix cmake not finding make_tables when cross-compiling 
Fixes #477.
 2018-12-14 17:50:03 +01:00
derselbst
cb485806c3 remove unused io.h include 2018-12-14 17:50:03 +01:00
jjceresa
ba9092ef98 Reducing defsfont loader code (#480) 2018-12-02 17:42:51 +01:00
carlo-bramini
ebb383b47f
Force default device if none is selected 2018-11-25 21:20:18 +01:00
carlo-bramini
26325edb22
Optimize device search 
Search of the MIDI device can be done before allocating the struct for the driver.
 2018-11-25 21:13:25 +01:00
carlo-bramini
fa924516f0
Max length of device name is MAXPNAMELEN 2018-11-25 21:10:16 +01:00
carlo-bramini
9aeac6ca6e
Unprepare MIDI buffers 
This was missing.
 2018-11-25 20:14:18 +01:00
carlo-bramini
c57d7fffe8
move fluid_winmidi_callback() 
Function fluid_winmidi_callback() has been moved on top of the source and made static.
 2018-11-25 20:06:56 +01:00
carlo-bramini
cfa8019a53
Remove old fluid_winmidi_input_error() 2018-11-25 20:05:40 +01:00
carlo-bramini
cea2645530
Makes fluid_winmidi_input_error() thread-safe 
The changes are:
1) made the function static
2) implemented UNICODE support
3) made thread-safe
4) delete unused static buffer
 2018-11-25 20:03:12 +01:00
carlo-bramini
b8b41f781f Improve chunk id search (#471) 
If we move UNKN_ID to the bottom of the enum, in the for() cycle inside chunkid() if the value is not found then the "i" variable will be already UNKN_ID.
 2018-11-25 10:02:33 +01:00
jjceresa
36da011191 Avoid redundant dsp parameter updates upon modulation. (#461) 2018-11-25 08:21:51 +01:00
carlo-bramini
98e4046096 Fix handle leak in winmidi driver (#469) 2018-11-23 19:49:20 +01:00
jjceresa
6e9d84f02a Forbid DATA_ENTRY_LSB to modulate (#465) 
and document illegal CC modulation
 2018-11-18 15:43:03 +01:00
derselbst
c8d1ec3841 Merge branch 'recent-docs' into master 2018-11-18 09:12:22 +01:00
carlo-bramini
fe92a0a655 Introduce enable-threads option to cmake (#463) 2018-11-18 08:55:52 +01:00
carlo-bramini
29c668683f Improve float calculations (#462) 
I tried to check the generated ASM code when WITH_FLOAT macro is defined and when it is undefined. In both cases I noticed that there are some points where the math number are converted from float to double and viceversa. Actually this happens whether the FPU is present or not, perhaps for granting the right precision in digits. I also noticed few points that could be simplified a bit by using integer math, so I also included them.

 * Since fluid_voice_get_actual_key() returns an integer value, the value 60 could be subtracted directly as integer, the code generated looks a bit better.
 * FLUID_NOISE_FLOOR needs to be cast to fluid_real_t to generate a bit more efficient code.
 * DC_OFFSET needs a cast to fluid_real_t to generate a bit more efficient code.
 * "last_fres" and "output_rate" are already fluid_real_t, so they do not need type cast.
 * "chan_add", "par1_add" and "par2_add" are integer types, so they can be added without FPU. "chan_mul", "par1_mul" and "par2_mul" are already fluid_real_t, so they do not need type cast. Instead, the constant 0.5 needs cast to fluid_real_t to be more efficient.
 2018-11-16 16:52:34 +01:00
carlo-bramini
bd41795263 Remove redundant dependency to readline in fluid_cmd.c 
Libreadline is already used into fluid_istream_readline().
In my opinion, there is no need to have this duplicated code, so I would suggest to move the call to add_history() into fluid_sys.h

Closes #460.
 2018-11-10 08:25:55 +01:00
carlo-bramini
0bf5c28275 "ready" is unused if ENABLE_MIXER_THREADS=0 2018-11-10 08:09:08 +01:00
Tom M
06bcf8db9f
Fix an incorrectly aligned result when converting between pointer types (#457) 
fix alignment issue of idlist
 2018-11-10 08:07:30 +01:00
Tom M
edc59ee3b5
Merge pull request #458 from FluidSynth/clang-tidying 
Fix clang-tidy warnings
 2018-11-10 08:01:10 +01:00
carlo-bramini
d20f6cff66
Use integer math. 2018-11-03 21:07:37 +01:00
carlo-bramini
afbd818f8a
Use M_LN10 macro instead of log(10.) 2018-11-03 17:07:20 +01:00
Tom M
0160543cdd
Compile time constant lookup tables with cmake (#438) 
Autogenerate lookup tables with a C helper tool, allowing them to be compile time constant and reduce memory requirements esp. on embedded systems.
 2018-11-03 14:38:54 +01:00
derselbst
28edbbfe83 fluid_settings_option_concat: dont count NULL options 
should never happen though
 2018-11-03 14:17:40 +01:00
derselbst
4d8954ee30 fix mistaken usage of fluid_synth_process in coreaudio driver 2018-11-03 14:11:05 +01:00
derselbst
31ccc6b6a0 fix a NULL deref in coreaudio driver 2018-11-03 14:11:05 +01:00
derselbst
ffdf5b89d7 constify function parameters if possible 2018-11-03 14:11:05 +01:00
derselbst
16d81d50ea remove unused fluid_hz2ct() 
Closes #455.
 2018-11-03 09:13:56 +01:00
carlo-bramini
2205e4a8fc Use integer math 
Since 'data' is an integer value clipped between +/-8192 and 'nrpn_scale' is a char value, this calculation could be done with integers and then return the fluid_real_t value.

Addresses #455.
 2018-11-03 08:06:56 +01:00
carlo-bramini
a948f7c5b2 Use integer math 
Actually, using an integer division produces the same result.

Addresses #455.
 2018-11-03 08:04:16 +01:00
Tom M
3f4c08e34b
Merge pull request #442 from carlo-bramini/float-to-s16-1 
Rounded samples do not need to be float.
 2018-10-31 13:18:33 +01:00
Colin Kinloch
cf09b654ab jack midi autoconnect (#450) 2018-10-31 13:10:29 +01:00
derselbst
b628115092 optimize rounding and clipping to int16 samples 2018-10-31 10:40:32 +01:00
derselbst
ee5b0ea1c7 remove redundant declarations 2018-10-28 16:42:20 +01:00
derselbst
a4ddc9396c avoid leaking memory in pulse driver 2018-10-28 16:37:44 +01:00
Tom M
804cbf91c9
Merge pull request #449 from carlo-bramini/win-driver-selection 
let cmake find WinMidi and DSound
 2018-10-27 20:26:18 +02:00
derselbst
e4ab5067f1 fix linker error on win32 without network support 
by adding missing #ifdefs
 2018-10-27 16:32:54 +02:00
carlo-bramini
e298fefd7f cleanup audio and midi driver instantiation (#448) 
Instead of saving the name of the driver, it would be worth to save to pointer to the selected definition instead.
In this way, the function for deleting the driver does not need to search its name by parsing all the list, but it just needs to call the pointer into the saved definition (less code). This fix can be applied to MIDI drivers too. 
I also moved the FLUID_FREE(allnames) inside the "if(allnames != NULL)" block, nothing bad should happen even by keeping that instruction outside, but actually there is no need to call the free if allnames is NULL.
 2018-10-27 16:22:44 +02:00
carlo-bramini
30886a3d78
Remove handwritten macros 2018-10-26 19:44:27 +02:00
carlo-bramini
413bcbeb84
Add Windows driver support macro. 2018-10-26 19:43:27 +02:00
carlo-bramini
8ef6ac7dcc
Enable VersionInfo also for MINGW 2018-10-26 19:42:15 +02:00
carlo-bramini
7b6ebbea5b
Add Windows driver selection 2018-10-26 19:41:34 +02:00
Tom M
121ed7124e
Merge pull request #447 from FluidSynth/zero-size-arr 
Fix build when no audio drivers are supported
 2018-10-23 17:41:35 +02:00
derselbst
e03e2edf8d register settings before adding options 2018-10-23 17:07:34 +02:00
derselbst
f6e70b38c7 clarify log message when no MIDI drivers are available 2018-10-21 20:26:46 +02:00
derselbst
327b3fb45b clarify log message when no audio drivers are available 2018-10-21 20:21:21 +02:00
carlo-bramini
a7c5b66872 Simplify MIDI driver installation 2018-10-21 13:35:41 +02:00
carlo-bramini
70f4551e90 Simplify audio driver installation 
In my opionion, it should be possible to simplify the code by installing the drivers with a for() cycle instead of doing tons of #ifdef...#endif.
The size of the binary code is basically the same as before, but the source lines are much less. I think that it could be done also for MIDI input drivers.
 2018-10-21 13:35:32 +02:00
carlo-bramini
326c20c0e3 Remove duplicated init_dither() declaration (#446) 2018-10-20 20:06:38 +02:00
derselbst
670cdf1e8f conditionally compile fluid_aufile.c 2018-10-20 15:15:43 +02:00
derselbst
08b5333398 remove macro workaround when no midi drivers are supported 2018-10-20 15:05:18 +02:00
derselbst
5bd07c120c fix build when no audio drivers are supported 2018-10-20 15:04:55 +02:00
carlo-bramini
ebc177f48f Consistently use FLUID_LOG macro (#443) 2018-10-16 18:02:27 +02:00
carlo-bramini
d8890038b6
Rounded samples do not need to be float. 
Actually, the function roundi() already returns an "int" type value, so in my opinion there is no need to use a floating point value for saturating the values in the range -32768/+32767. The generated assembly code looks more efficient after that.
 2018-10-13 14:37:41 +02:00
derselbst
7517c17524 Revert "make fluid_log() thread safe" 
This reverts commit d25cdae17c.
Mistakenly committed too early on the wrong branch.
 2018-10-12 09:24:07 +02:00
derselbst
2b563071f6 initialize logging at compile time 2018-10-09 18:11:37 +02:00
derselbst
d25cdae17c make fluid_log() thread safe 
by using a local buffer rather than global one, intentionally breaks
fluid_synth_error()
 2018-10-09 18:00:19 +02:00
derselbst
f0312c13d3 deprecate fluid_synth_error() 2018-10-09 17:52:33 +02:00
derselbst
c93fdd4b10 avoid chorus depth from spamming the console 
on any chorus setting, if the previous depth was out of range
 2018-10-07 13:15:05 +02:00
derselbst
0920db6d1f abort 'set' shell command if param not found 2018-10-07 12:56:18 +02:00
derselbst
ce57220d65 dont cast const qualifier away 2018-10-06 14:27:10 +02:00
carlo-bramini
f74f415a69 Avoid duplication of fluid_chorus_sine() call. 
Closes #439.
 2018-10-06 12:05:10 +02:00
Tom M
b768ad6d14
Merge pull request #437 from FluidSynth/null-fixes 
NULL derefs and mem leaks in soundfont loader
 2018-10-06 11:55:11 +02:00
derselbst
3ef6af5ade Merge branch 'recent-docs' 2018-10-06 11:45:18 +02:00
derselbst
fd7db023dd set upper chorus depth to 256 ms 
upper depth limit (ms) = (MAX_SAMPLES * 1000) / lower sample rate limit

with MAX_SAMPLES==2048 && lowersrate==8000
 2018-10-05 18:39:54 +02:00
derselbst
f972fbf1ec remove experimental hint from fluid_synth_set_sample_rate() 2018-10-05 18:12:31 +02:00
derselbst
57ef2dfed8 fix memory leaks in load_phdr() and load_ihdr() 2018-10-02 20:20:22 +02:00
derselbst
7f2d655b9c fix various NULL derefs in fluid_sffile 2018-10-02 20:14:56 +02:00
derselbst
b31afd65a5 fix NULL deref in new_fluid_hashtable_full() 2018-10-02 19:33:50 +02:00
carlo-bramini
6f8a574e36 Various fixes to dsound driver (#435) 
- fluid_dsound_audio_run() and fluid_win32_error() are now static functions.

- WAVEFORMATEX is not required by DirectSound during the playback. We do not need it anymore after calling ::SetFormat(), so allocating this structure dynamically inside driver structure is useless.

- Implemented support for float sample type.

- Uses an event object for handling the end of thread, it allows to combine the quit event with the later wait in milliseconds in a single block.

- Calculates the amount of milliseconds to sleep rather than sleeping always for one millisecond.

- Fix an error into a FLUID_LOG() call.

- Fix handle leak of the thread, now it is correctly closed with CloseHandle().

- ExitThread() is a nonsense in that position, since the thread is already exiting.

- Fix error when compiling with WSDK 8.1, by defining NOBITMAP, in case NOGDI macro added somewhere
 2018-09-29 10:08:34 +02:00
derselbst
923f5f3544 update API doc of delete_fluid_sequencer() 2018-09-21 19:00:28 +02:00
Tom M
fb684aa5d8
Merge pull request #434 from rncbc/jack_driver-last_client-reuse-fix 
jack_driver: last_client reuse fix.
 2018-09-20 20:20:55 +02:00
jjceresa
75b77f0186 Remove redundant invalid preset generators checking in fluid_defpreset_noteon() (#430) 2018-09-20 20:11:11 +02:00
rncbc
99e235e279 jack_driver: last_client reuse fix. 2018-09-20 09:04:21 +01:00
Colin Kinloch
f4b88fc747 coremidi input port and autoconnect (#427) 2018-09-15 14:08:55 +02:00
jjceresa
f566f0ef33 Simplify fluid_mod_get_xxx_value() signature. (#429) 2018-09-15 14:01:15 +02:00
Tom M
e4b8e2b44c
Implement fluid_get_userconf() on windows (#425) 2018-09-09 22:34:35 +02:00
derselbst
eb132d8196 correctly typedef fluid_stat_buf_t for glib < 2.26 2018-08-30 22:14:27 +02:00
derselbst
b9dd426046 correctly typedef fluid_stat_buf_t 
fixing copy'n paste mistake... excuse me
 2018-08-30 19:53:40 +02:00
derselbst
9052bad56c silence MSVC type narrowing warnings in fluid_cmd.c 2018-08-30 18:45:29 +02:00
derselbst
7d9a02d836 replace strlen with FLUID_STRLEN 2018-08-30 18:40:49 +02:00
derselbst
cdc8d38f4e silence MSVC type narrowing warnings in fluid_settings.c 2018-08-30 18:33:05 +02:00
derselbst
70fbabf97c fix incompatible types warning: 'fluid_stat_buf_t *' to '_stat64i32 *' 
only occurs on Win64 with glib < 2.26
 2018-08-30 17:38:19 +02:00
derselbst
4d8574d38a change fluid_log_* prototypes to const char* 2018-08-23 16:56:22 +02:00
Tom M
55a4bb5597
Merge pull request #421 from FluidSynth/test-flag 
Always build unit tests, remove `enable-tests` cmake flag
 2018-08-10 16:39:17 +02:00
derselbst
500959701e move fluid_file_renderer_settings() declaration to fluid_adriver.h 2018-08-08 20:55:32 +02:00
derselbst
908aaeb5cc cmake: setup linker directories before creating target 2018-08-07 15:19:05 +02:00
derselbst
39f574942e cmake: remove redundant include and link dir variables 
cf. docs of pkg_check_modules()
 2018-08-07 15:07:20 +02:00
derselbst
ff1482ec06 fix dll visibility on win32 for unittests 2018-08-06 20:29:50 +02:00
derselbst
521e5451fc remove cmake option enable-tests 
Always build unit tests instead and use cmakes OBJECT library to bypass visibility control and the value of BUILD_SHARED_LIBS.
 2018-08-05 20:50:41 +02:00
derselbst
063ab5d7d2 remove unused midi driver status functions 2018-07-21 10:43:38 +02:00
derselbst
47c4a5b1c7 remove various redundant audio / midi driver declarations 
by moving already existing function declarations to fluid_adriver.h and
fluid_mdriver.h resp.
 2018-07-21 10:37:43 +02:00
derselbst
f6f66d53fd make fluid_synth_write_float() safe when called with NULL buffers 2018-07-20 19:30:18 +02:00
derselbst
8afb46515f fix comment for sinc interpolation table 
a Hanning window is applied, rather than a Hamming  window
 2018-07-19 17:22:44 +02:00
derselbst
bb413011f7 update deprecated note of fluid_synth_nwrite_float() 2018-07-16 12:21:55 +02:00
derselbst
2185194d43 cleanup fluid_synth_process() 
remove code duplications
 2018-07-15 19:52:48 +02:00
derselbst
af1fc75717 add fluid_synth_count_effects_groups() 2018-07-12 10:34:02 +02:00
derselbst
7e627c575c fix NULL deref in jack driver 2018-07-11 21:07:37 +02:00