JACK saves those flags as-is, which is great for us.

Fixes #1045

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Found via `codepell` (v1.17.0.dev0)  
```
codespell -q 3 -L iff,iif,inport,sord,sinc,te,uint -S ./source/includes,./source/modules,./source/native-plugins,./data/windows
```

Signed-off-by: falkTX <falktx@falktx.com>

The type of JackBridge::client_open_ptr is:

    jack_client_t* (*)(const char*, jack_options_t, jack_status_t*)

client_open_ptr is set to the address of jack_client_open (loaded via
dlopen/dlsym), but the signature of jack_client_open is:

    jack_client_t* jack_client_open(
        const char*, jack_options_t, jack_status_t*, ...
    )

jack_client_open is variadic, while client_open_ptr is not. PowerPC
segfaults when calling a variadic function through a non-variadic
pointer, as PowerPC reserves different amounts of stack space when
calling a variadic vs. non-variadic function. In this case,
jack_client_open assumes the caller allocated more space than it
actually did, so it overwrites parts of the caller's stack frame that
were not intended to be modified.

This commit changes the type of JackBridge::client_open_ptr to match
the signature of jack_client_open.

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@falktx.com>

Expose linux call over bridge, so we can set the flag

Signed-off-by: falkTX <falktx@falktx.com>

Expose linux call over bridge, so we can set the flag

Signed-off-by: falkTX <falktx@falktx.com>

Signed-off-by: falkTX <falktx@gmail.com>

Fixes #581

(needs a full rebuild of carla and bridges)

Also some needed cleanup