17cff5722e
Lock the active plugins map early
...
To prevent a race condition when an exiting plugin wants to terminate
the process when it think there are no plugins left just as another
yabridge instance is connecting to the group socket.
2020-05-21 17:12:41 +02:00
8eb01cb519
Listen on the group socket and handle requests
2020-05-20 18:45:33 +02:00
6d6d928838
Move all plugin group handling boilerplate
2020-05-19 15:29:48 +02:00
8bd1dc8c50
Encapsulate the STDOUT/STDERR capturing
2020-05-18 16:15:07 +02:00
53acb1f78a
Move wine-bridge.h -> bridges/vst2.h
...
This way we can structure the group handling and a potential future VST3
bridge in the same way.
2020-05-18 16:15:07 +02:00
b8028b8e13
Remap STDOUT and STDERR in group process to log
...
This is not very pretty, but there's not really another way and I"m
surprised that it actually works.
2020-05-18 16:15:07 +02:00
2f39650322
Remove last traces of Boost.Filesystem from host
...
Missed this in e728dbe5a2
2020-05-18 16:15:07 +02:00
994f3c9e38
Add a plugin group host application
2020-05-18 16:15:07 +02:00
95e716d229
Rename vst-host.cpp -> individual-host.cpp
2020-05-17 14:43:21 +02:00
37a74c8f98
Get rid of the dedicated AEffect socket
2020-05-13 13:15:52 +02:00
0900dc9e18
Work around race condition in certain plugins
...
Some plugins would either crash or freeze on the next Win32 message loop
when `effEditGetRect` gets called before `effEditOpen` and we run the
message loop between these two event calls.
Fixes the issue with Superior Drummer 3 in Bitwig mentioned in #12 and a
similar issue with the Roland Cloud synths.
2020-05-12 13:17:24 +02:00
33e5d2bd56
Add a workaround for compilation with Wine 5.7+
2020-05-11 18:11:44 +02:00
ba91971829
Simplify object reading
...
No longer needs to read into an existing object after the last change,
and reusing that function here too makes it less error prone.
2020-05-10 13:10:58 +02:00
868b0fd357
Finally implement eff{Set,Get}SpeakerConfiguration
...
As mentioned in #1 . This also indirectly allows yabridge to work under
Renoise.
2020-05-07 18:23:41 +02:00
a3aad51e41
Also add a second payload value to event responses
...
This will only be used for `effGetSpeakerArrangement`.
2020-05-07 17:23:22 +02:00
2d0998047c
💥 Rename PluginBridge to WineBridge
...
I had swapped these names around once before but I think going with
PluginBridge for the plugin and WineBridge for the Wine VST host is the
least ambiguous it can get.
2020-05-07 13:04:00 +02:00
e4cfc8f43b
Fix typo in .dll loading error message
2020-05-06 01:39:14 +02:00
af060054d3
Implement audioMasterProcessEvents, closing #5
...
This allows plugins to output MIDI events.
2020-05-06 01:04:20 +02:00
e71fd433f9
Clarify LoadLibrary() related startup error
...
The old message sounds too much like something that would come from
Linux's dynamic linker.
2020-05-06 00:06:46 +02:00
9c6fc78471
Fix editor window handling in Reaper
...
And other hosts that embed the parent window into another window.
2020-05-02 18:28:43 +02:00
ed8e3ba114
Refactor event receiving to optimize MIDI handlign
...
This keeps compatibility with some weirdly designed plugins (such as
Kontakt) while avoiding some unnecessary data transformations. Before
this we'd convert from a `DynamicVstEvents` object to a `VstEvents`
object, back to a `DynamicVstEvents` and then finally back into another
`VstEvents` object. With this change we can skip the second half of the
conversions.
2020-05-01 14:06:06 +02:00
6266072641
Explicitly list opcodes that should return strings
...
The automatic detection works fine in every case I've tested other than
Fabfilter plugins, but this is probably for the best.
2020-04-30 20:55:49 +02:00
dcc1a34a41
Use a better check for the 32-bit host
2020-04-30 15:09:13 +02:00
79e8a37c39
Simplify the 32-bit mode detection
2020-04-30 12:59:11 +02:00
6ca7eae16a
Mention 32-bit support in the readme
2020-04-30 12:48:03 +02:00
e33f418206
Add a notice when the host is in 23-bit mode
2020-04-29 19:12:29 +02:00
e414c1a341
Fix screen resolution detection
2020-04-28 12:22:56 +02:00
8adb944445
Use maximum display resolution for the window size
...
Instead of it being hardcoded to 1440p.
2020-04-28 11:47:39 +02:00
4b84f663ab
Always run the event loop, fixing processing issue
...
Also remove any special `effEditIdle` handling.
Apparently plugins rely on the message loop for their internal tasks,
even for things that have nothing to do with GUIs, such as deferring
initialization.
2020-04-27 18:53:06 +02:00
7f0a8c8efd
Extend the midi event lock
...
This should not make any difference, but if this event ever gets called
at the same time as the processing then this at least won't cause any
issues with plugins that are not thread safe themselves.
2020-04-27 16:53:44 +02:00
af6ae2c2ad
Fix capitalization of the word MIDI
2020-04-26 18:19:10 +02:00
85afd4107e
Extend the lifetime of MIDI events, fixing Kontakt
2020-04-26 18:17:30 +02:00
aee890fbf6
Remove the now unused ability to skip opcodes
...
If this is needed again in the future (it shouldn't be), then it should
be re-implemented using a separate function.
2020-04-26 16:30:11 +02:00
5ce20be9d9
Allow storing and loading of larger chunks
2020-04-25 17:35:02 +02:00
631e82f2ad
Print the version and any VCS info at startup
2020-04-25 16:14:22 +02:00
0efdd1a790
Add missing license headers
2020-04-24 15:41:18 +02:00
d0c8d8a2e7
Clean up Win32 thread API usage using RAII
2020-04-24 15:40:44 +02:00
9dfa83df20
Change todo regarding drag-and-drop
2020-04-23 23:16:31 +02:00
060712ee4d
Remote the WS_EX_ACCEPTFILES
...
It doesn't matter, but we shouldn't be the window accepting drag and
drop operations.
2020-04-23 23:13:55 +02:00
259d23ec57
Add support for legacy plugins with only process()
...
By emulating the behavior of `processReplacing()`. Not that there should
still be any plugins around that don't support processReplacing.
2020-04-22 16:57:18 +02:00
27ca18869b
Change variable names for consistency
2020-04-21 17:02:28 +02:00
d2bf5c35a4
Fix hanging when closing the editor
2020-04-21 16:56:36 +02:00
421ed21901
Get rid of now no longer needed synchronisation
...
This was mostly a workaround to get Serum to not crash when audio was
being processed during a specific point of its `WM_PAINT` message
handler. This is no longer needed when using `CreateThread` instead of
`std::thread`.
2020-04-20 23:36:17 +02:00
86974ebe2d
Fix typo
2020-04-20 23:25:01 +02:00
357dec315a
Swap out std::thread for CreateThread
...
Not entirely sure why, but this gets rid of the impossible to debug data
races when Serum's GUI is being repainted while another thread is
calling `processReplacing`. This is possibly because std::thread does
not respect Windows calling conventions and CreateThread does.
2020-04-20 23:24:39 +02:00
1a6a094c2b
Fix reopening closed editor windows
2020-04-19 20:36:39 +02:00
b44d98b4c7
Mention more Wine debugging strategies
2020-04-19 16:38:09 +02:00
9f3ed85208
Replace std::monostate with std::nullptr_t
...
This represents the underlying meaning better since we're mostly dealing
with the `data` void pointer argument.
2020-04-19 16:05:37 +02:00
6025efb0fc
Fix grammar in comments
2020-04-19 15:46:32 +02:00
0cce91d460
Reuse output buffers
2020-04-19 15:38:31 +02:00
Robbert van der Helm