JackLockedEngine.h: internal compiler error: in objc_eh_runtime_type, at objc/objc-next-runtime-abi-01.cc

[barracuda156]

Build is broken on PowerPC. Tested on 10.5.8 and 10.6, with gcc7 and gcc12.

during GIMPLE pass: eh
In file included from ../macosx/coreaudio/JackCoreAudioDriver.mm:29:
../common/JackLockedEngine.h: In member function 'int Jack::JackLockedEngine::PortRegister(int, const char*, const char*, unsigned int, unsigned int, jack_port_id_t*)':
../common/JackLockedEngine.h:196:13: internal compiler error: in objc_eh_runtime_type, at objc/objc-next-runtime-abi-01.cc:2795
  196 |         int PortRegister(int refnum, const char* name, const char *type, unsigned int flags, unsigned int buffer_size, jack_port_id_t* port)
      |             ^~~~~~~~~~~~
Please submit a full bug report, with preprocessed source (by using -freport-bug).
See <https://trac.macports.org/newticket> for instructions.

../macosx/coremidi/JackCoreMidiInputPort.mm: In member function 'void Jack::JackCoreMidiInputPort::ProcessCoreMidi(const MIDIPacketList*)':
../macosx/coremidi/JackCoreMidiInputPort.mm:163:31: warning: comparison of integer expressions of different signedness: 'int' and 'long unsigned int' [-Wsign-compare]
  163 |             if (chunkSize + 1 <= sizeof(running_status_buf)) {
      |                 ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../macosx/coremidi/JackCoreMidiInputPort.mm:208:15: warning: comparison of integer expressions of different signedness: 'int' and 'size_t' {aka 'long unsigned int'} [-Wsign-compare]
  208 |         if (k < size) goto parse_event;
      |             ~~^~~~~~
At global scope:
cc1objplus: note: unrecognized command-line option '-Wno-deprecated-register' may have been intended to silence earlier diagnostics

Waf: Leaving directory `/opt/local/var/macports/build/_opt_PPCSnowLeopardPorts_audio_jack/jack/work/jack2-1.9.22/build'
Build failed
 -> task in 'coreaudio' failed with exit status 1: 
	{task 23367152: cxx JackCoreAudioDriver.mm -> JackCoreAudioDriver.mm.7.o}
['/opt/local/bin/g++-mp-12', '-pipe', '-Os', '-arch', 'ppc', '-Wall', '-Wno-invalid-offsetof', '-std=gnu++11', '-Wno-deprecated-register', '-fPIC', '-I.', '-I..', '-Icommon', '-I../common', '-Icommon/jack', '-I../common/jack', '-Imacosx', '-I../macosx', '-Iposix', '-I../posix', '-DAFTEN=1', '-DHAVE_DOXYGEN=0', '-DHAVE_ALSA=0', '-DHAVE_FIREWIRE=0', '-DHAVE_IIO=0', '-DHAVE_PORTAUDIO=0', '-DHAVE_WINMME=0', '-DHAVE_CELT=0', '-DHAVE_TESTS=0', '-DHAVE_OPUS=0', '-DHAVE_SAMPLERATE=1', '-DHAVE_SYSTEMD=0', '-DHAVE_DB=0', '-DHAVE_ALLOCA_H=1', '-DHAVE_EXECINFO_H=1', '-DUSE_CLASSIC_AUTOLAUNCH=1', '-DCLIENT_NUM=256', '-DPORT_NUM_FOR_CLIENT=2048', '-DADDON_DIR="/opt/local/lib/jack"', '-DJACK_LOCATION="/opt/local/bin"', '-DUSE_POSIX_SHM=1', '-DJACKMP=1', '-DHAVE_CONFIG_H', '-DSERVER_SIDE', '../macosx/coreaudio/JackCoreAudioDriver.mm', '-c', '-o/opt/local/var/macports/build/_opt_PPCSnowLeopardPorts_audio_jack/jack/work/jack2-1.9.22/build/macosx/coreaudio/JackCoreAudioDriver.mm.7.o', '-I/opt/local/include']
Command failed:  cd "/opt/local/var/macports/build/_opt_PPCSnowLeopardPorts_audio_jack/jack/work/jack2-1.9.22" && /opt/local/bin/python2.7 ./waf build --verbose 
Exit code: 1

https://trac.macports.org/ticket/67794
The bug is not new: https://trac.macports.org/ticket/60420

[falkTX]

an internal compiler error is not an issue in jack2 but in the compiler. not much we can do to fix this

[barracuda156]

@falkTX Thank you for responding. Yes, I understand ICE is not your bug, but maybe you could suggest a possible workaround? It it something not generally acceptable, we can have a local patch in Macports to use only with GCC.
Otherwise jack is broken, at least on PowerPC where no Clang works.

P. S. I need to check, but I recall GCC upstream is aware of the problem, but ObjC no one cares.

[falkTX]

You can always try to disable optimizations, either globally or in that specific part of the code by use of __attribute__((optnone))

[barracuda156]

You can always try to disable optimizations, either globally or in that specific part of the code by use of __attribute__((optnone))

I will try that, thank you. For GCC, apparently __attribute__((optimize(0))) is correct one: argtable/argtable3@7704006

[barracuda156]

@falkTX Unfortunately, building the whole thing with -O0 (was easier to try and has the same effect as setting respective __attribute__, I assume) did not help, same ICE.

In the meanwhile, a couple of headers are missing in macosx files: #951

[barracuda156]

I have found a ticket on GCC Bugzilla, it is a bug in GCC for 32-bit targets:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61759
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107530

[barracuda156]

@alarixnia Could you say if jack2 builds via pkgsrc for Darwin ppc? If you know or could check.

[alarixnia]

@alarixnia Could you say if jack2 builds via pkgsrc for Darwin ppc? If you know or could check.

A dependency of it requires CMake and I don't have a compiler capable of building CMake for Darwin/ppc.

[barracuda156]

@alarixnia Could you say if jack2 builds via pkgsrc for Darwin ppc? If you know or could check.

A dependency of it requires CMake and I don't have a compiler capable of building CMake for Darwin/ppc.

Unless that dependency is extremely picky, CMake 3.9.6 builds without C++11.
Also, gcc7 and gcc10 build on 10.4.11. I think even gcc13 (though gcc13 requires C++11, so one must build gcc7 or gcc10 first).

[cooljeanius]

I have found a ticket on GCC Bugzilla, it is a bug in GCC for 32-bit targets:

gcc.gnu.org/bugzilla/show_bug.cgi?id=61759 gcc.gnu.org/bugzilla/show_bug.cgi?id=107530

If anyone wants to help debug these, see the following GCC wiki page: https://gcc.gnu.org/wiki/DebuggingGCC