Linux build of HISE

davephillips

Hi Christoph, Am I correct assuming that the Export function still doesn't work for Linux ? IIRC you mentioned that that process is handled by a shell script. If so, where is that script ? I can try to debug it for Linux.

The export process fails with this error :

[dlphilp@The6300 build]$ ./HISE\ Standalone
Creating /home/dlphilp/HISE-MusicBox/Binaries/Source/PresetData.h and /home/dlphilp/HISE-MusicBox/Binaries/Source/PresetData.cpp from files in /home/dlphilp/HISE-MusicBox/Binaries/temp...

Adding preset: 10528 bytes
Adding impulses: 14 bytes
Adding externalFiles: 58 bytes
Adding userPresets: 14 bytes
Adding images: 10 bytes

Total size of binary data: 10624 bytes
Couldn't get a file descriptor referring to the console

What's happening there ? I see a batchCompileOSX file in my project Binaries directory, it's set up for Xcode, not Linux. I presume a similar file is needed for successful export to a native Linux VST/VSTi plugin.

Best,

dp

davephillips

Hi Christoph,

I loaded and played the PatchArena Marimba SFZ on the HISE sampler. I'll try a few more SFZ files today.

Best,

dp

Christoph Hart

What's happening there ? I see a batchCompileOSX file in my project Binaries directory, it's set up for Xcode, not Linux. I presume a similar file is needed for successful export to a native Linux VST/VSTi plugin.

Yes this needs to be customized for every OS. Currently it just checks if it's Windows and if not it creates the stuff for OS X.

It's not totally trivial though. The compilation chain is:

HISE -> Projucer -> Makefile -> Make -> Compiler (eg. clang++)

so I need HISE to create a .jucer file that creates a Makefile that tells make what to do. You can call it a slight overkill of metaproject management :)

diedeno

I tried to use the standalone application on Linux (Ubuntu 16.4 LTS).
The software compiles without any error.
But when i try to start it, it (always) get a segmentaion fault.

startVoiceOpen
renderVoiceOpen
onNoteOnOpen
onNoteOffOpen
onControllerOpen
onControlOpen
externalPopupShown
Segmentation fault

strace:
write(2, "onControllerOpen", 16onControllerOpen) = 16
write(2, "\n", 1
) = 1
write(2, "onControlOpen", 13onControlOpen) = 13
write(2, "\n", 1
) = 1
write(2, "externalPopupShown", 18externalPopupShown) = 18
write(2, "\n", 1
) = 1
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x18} ---
+++ killed by SIGSEGV +++
Segmentation fault

in syslog:
Feb 4 19:12:28 SLEIPNIR-L kernel: [27915.222550] HISE Standalone[7153]: segfault at 18 ip 0000000000488162 sp 00007ffd96acd300 error 4 in HISE Standalone[400000+138e000]

Whaty could be the problem here?
Thank you very much.

Christoph Hart

Hmm, I can compile and run it without the segfaults. I uploaded the recent build for Ubuntu here:

HISE Standalone Linux

Can you check if this build runs on your system?

diedeno

Hi Christoph,

I've tried your build, and i have exactly the same problem as before .
That's weird because i have a fresh install of 16.04 LTS.
Distributor ID: Ubuntu
Description: Ubuntu 16.04.1 LTS
Release: 16.04
Codename: xenial

Analysing the core dump:

(gdb) set args
(gdb) run
Starting program: /home/charles/DEV/HISE Standalone
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ffff4a24700 (LWP 15684)]
[New Thread 0x7ffff0a25700 (LWP 15685)]
[New Thread 0x7ffff7efe700 (LWP 15686)]
[New Thread 0x7ffff7e7d700 (LWP 15687)]
[Thread 0x7ffff7e7d700 (LWP 15687) exited]
[Thread 0x7ffff7efe700 (LWP 15686) exited]
[New Thread 0x7fffd7ffe700 (LWP 15688)]
[Thread 0x7fffd7ffe700 (LWP 15688) exited]
[New Thread 0x7fffd7ffe700 (LWP 15689)]
[New Thread 0x7fffcfffe700 (LWP 15690)]
[New Thread 0x7fffcbfff700 (LWP 15691)]
[Thread 0x7fffcbfff700 (LWP 15691) exited]

Thread 1 "HISE Standalone" received signal SIGSEGV, Segmentation fault.
0x00000000006d7be9 in ScriptingApi::Synth::Synth(ProcessorWithScriptingContent, ModulatorSynth) ()
(gdb) Quit

Thank you very much.

diedeno

Oeps, those big fonts were not meant to be. Sorry

Christoph Hart

This is weird because the constructor is not called until you add a script processor. I'll check again for static initialisers...

Christoph Hart

I found the issue - it's because you don't have a app data folder with the ModuleEnum.xml file which causes HISE to rescan all modules but this method is crashing due to a recent commit (which I didn't noticed because I have this file). Thanks for letting me know!

I am busy fixing another issue right now, but I'll commit a fix for this tomorrow.

diedeno

OK, thank you very much.
I"m looking forward to try it.

Christoph Hart

Should be fixed with the latest commit. I probably have to extend the Travis CI test coverage to include running HISE and not just compiling it to avoid this kind of issues in the future :)

diedeno

OK, this works. Thx

lacipor

Hello Everybody!

I've tried to build HISE on Ubuntu Studio 16.04.3 64-bit,
but it stops at the point where the last line says "Linking HISE Standalone".
This happens right after the "Compiling juce_tracktion_marketplace.cpp" line.
All this happens in the terminal, but everything else looks the same as in Dave's Windows HISE build video on YouTube.

I also made a build with Debug and Release in the Makefile, but of course it didn't change the thing, so it still doesn't work.

I also disabled the IPP in Projucer, but it didn't help either.

As I opened and saved the HISE Standalone project in Projucer,
Projucer tried to update itself,
but I clicked the "Cancel" button,
so I did the saving with the Projucer version that comes with the GitHub Clone/Download,
which is 4.3.0.

I also have a few warnings while building,
but I don't know what exactly to do about them.

I have the "build" folder,
and also a sub-folder in the "build" folder named "Intermediate" with a lot of files in it,
and I have a "HISE Standalone" executable in it,
but for some reason it doesn't execute,
it just won't start.
Am I missing something here?

What should I do, any idea?

Thank you for your help!

THIS IS FROM THE TERMINAL (from START to (not) FINISH):
Compiling DspUnitTests.cpp
Compiling HiseEventBufferUnitTests.cpp
Compiling MainComponent.cpp
Compiling Main.cpp
Compiling BinaryData.cpp
Compiling hi_backend.cpp
In file included from ../../../../../HISE-master/hi_backend/hi_backend.cpp:62:0,
from ../../JuceLibraryCode/hi_backend.cpp:9:
../../../../../HISE-master/hi_backend/backend/CompileExporter.cpp: In member function ‘CompileExporter::ErrorCodes CompileExporter::compileSolution(CompileExporter::BuildOption, CompileExporter::TargetTypes)’:
../../../../../HISE-master/hi_backend/backend/CompileExporter.cpp:934:47: warning: ignoring return value of ‘int system(const char*)’, declared with attribute warn_unused_result [-Wunused-result]
system(permissionCommand.getCharPointer());
^
../../../../../HISE-master/hi_backend/backend/CompileExporter.cpp: In static member function ‘static void CompileExporter::BatchFileCreator::createBatchFile(CompileExporter*, CompileExporter::BuildOption, CompileExporter::TargetTypes)’:
../../../../../HISE-master/hi_backend/backend/CompileExporter.cpp:1927:43: warning: ignoring return value of ‘int system(const char*)’, declared with attribute warn_unused_result [-Wunused-result]
system(lineEndChange.getCharPointer());
^
At global scope:
cc1plus: warning: unrecognized command line option ‘-Wno-inconsistent-missing-override’
Compiling hi_components.cpp
Compiling hi_core.cpp
Compiling hi_dsp.cpp
Compiling hi_dsp_library.cpp
Compiling hi_lac.cpp
Compiling hi_modules.cpp
Compiling hi_sampler.cpp
Compiling hi_scripting.cpp
Compiling juce_audio_basics.cpp
Compiling juce_audio_devices.cpp
Compiling juce_audio_formats.cpp
Compiling juce_audio_processors.cpp
In file included from ../../../../JUCE/modules/juce_audio_processors/format_types/juce_VSTPluginFormat.cpp:55:0,
from ../../../../JUCE/modules/juce_audio_processors/juce_audio_processors.cpp:153,
from ../../JuceLibraryCode/juce_audio_processors.cpp:9:
../../../../JUCE/modules/juce_audio_processors/format_types/juce_VSTMidiEventList.h:37:7: warning: ‘juce::VSTMidiEventList’ has a field ‘juce::VSTMidiEventList::events’ whose type uses the anonymous namespace
class VSTMidiEventList
^
In file included from ../../../../JUCE/modules/juce_audio_processors/juce_audio_processors.cpp:153:0,
from ../../JuceLibraryCode/juce_audio_processors.cpp:9:
../../../../JUCE/modules/juce_audio_processors/format_types/juce_VSTPluginFormat.cpp:325:7: warning: ‘juce::ModuleHandle’ has a field ‘juce::ModuleHandle::moduleMain’ whose type uses the anonymous namespace
class ModuleHandle : public ReferenceCountedObject
^
../../../../JUCE/modules/juce_audio_processors/format_types/juce_VSTPluginFormat.cpp:325:7: warning: ‘juce::ModuleHandle’ has a field ‘juce::ModuleHandle::customMain’ whose type uses the anonymous namespace
../../../../JUCE/modules/juce_audio_processors/format_types/juce_VSTPluginFormat.cpp:589:7: warning: ‘juce::VSTPluginInstance’ has a field ‘juce::VSTPluginInstance::vstEffect’ whose type uses the anonymous namespace
class VSTPluginInstance : public AudioPluginInstance,
^
../../../../JUCE/modules/juce_audio_processors/format_types/juce_VSTPluginFormat.cpp:589:7: warning: ‘juce::VSTPluginInstance’ has a field ‘juce::VSTPluginInstance::vstHostTime’ whose type uses the anonymous namespace
cc1plus: warning: unrecognized command line option ‘-Wno-inconsistent-missing-override’
Compiling juce_audio_utils.cpp
Compiling juce_core.cpp
Compiling juce_cryptography.cpp
Compiling juce_data_structures.cpp
Compiling juce_events.cpp
Compiling juce_graphics.cpp
Compiling juce_gui_basics.cpp
Compiling juce_gui_extra.cpp
Compiling juce_opengl.cpp
Compiling juce_tracktion_marketplace.cpp
Linking HISE Standalone
lacipor@laciporpc:~/laciporpcc/downloads/HISE-master/projects/standalone/Builds/LinuxMakefile$ sudo make
[sudo] password for lacipor:
Linking HISE Standalone

d.healey

While we're digging up this old thread I too am having an issue with HISE on Linux Mint. The build completes just fine, but when I try and run the binary via the terminal I get this.

JUCE v4.3.0
JUCE Assertion failure in UtilityClasses.h:489
terminate called after throwing an instance of 'juce::String'
Aborted

I tried enabling JACK in projucer (I'd previously disabled it for some reason I can't recall) now I get this build error:

In file included from ../../JuceLibraryCode/juce_audio_devices.cpp:9:0:
../../../../JUCE/modules/juce_audio_devices/juce_audio_devices.cpp:119:25: fatal error: jack/jack.h: No such file or directory
   #include <jack/jack.h>

Christoph Hart

Not sure about the JACK thing, but I'll take a look about the other issue. One thing however is that you are running a Debug build on Linux with vastly decreased performance (otherwise you wouldn't hit an assertion failure). That error looks like there is going something wrong and instead of the error being handled by the script engine it slips through and calls abort.

d.healey

I tried it set to release too. I have an older version on my laptop which builds and runs fine so I'm going to play spot the difference tomorrow and see if I can find anything obvious.

d.healey

I think there is a dependency issue. The latest github version builds and runs on my laptop perfectly but I'm having trouble on my main system. I changed to Release mode, and installed sudo apt-get install libjack-jackd2-dev and that got rid of the jack error but when I run it I get the juce::String error again.

Cannot lock down 82274202 byte memory area (Cannot allocate memory)
terminate called after throwing an instance of 'juce::String'
Aborted

Update: Actually I'm starting to think the issue is with my jack configuration. On my laptop I'm using Cadence but on my PC I'm using qjackctl because I get less latency with it.
If I close jack I get this mess when trying to run HISE.

Cannot connect to server socket err = No such file or directory
Cannot connect to server request channel
jack server is not running or cannot be started
JackShmReadWritePtr::~JackShmReadWritePtr - Init not done for 4294967295, skipping unlock
JackShmReadWritePtr::~JackShmReadWritePtr - Init not done for 4294967295, skipping unlock
terminate called after throwing an instance of 'juce::String'
Aborted

d.healey

I managed to get rid of the memory lock down thing by increasing the limit in etc/security/limit.config to 1000000. Now there's just the juce::String assertion issue to solve.

d.healey

I'm getting the same error on my laptop now too :(

After looking at this line in the source I'm thinking that this is still some memory issue... any ideas @Christoph-Hart?
JUCE Assertion failure in UtilityClasses.h:489

Christoph Hart

Hmm, everything is running fine for me, but I am not doing JACK or anything serious because I am just running it inside VirtualBox (the graphics are messed up but I think the culprit is the VM since Visual Code behaves similar).

The problem is that the exception handling on Linux is a bit more crude than on OSX or Windows and calls terminate whenever an error is raised. Can you hook up a Debugger and see where it throws the String?

Linux build of HISE

Thread 1 "HISE Standalone" received signal SIGSEGV, Segmentation fault.
0x00000000006d7be9 in ScriptingApi::Synth::Synth(ProcessorWithScriptingContent, ModulatorSynth) ()
(gdb) Quit

35

1.8k

10.3k

89.7k

Linux build of HISE

Thread 1 "HISE Standalone" received signal SIGSEGV, Segmentation fault. 0x00000000006d7be9 in ScriptingApi::Synth::Synth(ProcessorWithScriptingContent*, ModulatorSynth*) () (gdb) Quit

35

1.8k

10.3k

89.7k

Thread 1 "HISE Standalone" received signal SIGSEGV, Segmentation fault.
0x00000000006d7be9 in ScriptingApi::Synth::Synth(ProcessorWithScriptingContent, ModulatorSynth) ()
(gdb) Quit