#404 ctest does not catch test failures

[Zbigniew Jędrzejewski-Szmek]

This is most probably a problem with ctest, not with libsbml, just putting this here as a heads-up. I have to say I'm rather fed up with cmake... Error upon error upon error.

23: [Thread debugging using libthread_db enabled]
23: Using host libthread_db library "/lib/libthread_db.so.1".
23: 0x001b198c in mono_trace ()
23:   Id   Target Id         Frame 
23:   2    Thread 0xb68a9450 (LWP 22799) "mono" 0xb6e356a4 in read () from /lib/libpthread.so.0
23: * 1    Thread 0xb6f19210 (LWP 22795) "mono" 0x001b198c in mono_trace ()
23: 
23: Thread 2 (Thread 0xb68a9450 (LWP 22799)):
23: #0  0xb6e356a4 in read () from /lib/libpthread.so.0
23: #1  0xb68a9910 in ?? ()
23: Backtrace stopped: previous frame identical to this frame (corrupt stack?)
23: 
23: Thread 1 (Thread 0xb6f19210 (LWP 22795)):
23: #0  0x001b198c in mono_trace ()
23: #1  0x00171b00 in mono_lookup_pinvoke_call ()
23: #2  0x001474a0 in mono_marshal_get_native_wrapper ()
23: #3  0x000488e4 in mono_method_to_ir ()
23: #4  0x00060ba8 in inline_method.isra ()
23: #5  0x000472e0 in mono_method_to_ir ()
23: #6  0x00024ec4 in mini_method_compile ()
23: #7  0x00026a08 in mono_jit_compile_method_with_opt.constprop ()
23: #8  0x000276f8 in mono_jit_runtime_invoke ()
23: #9  0x00155bc0 in mono_runtime_invoke ()
23: #10 0x00158e8c in mono_runtime_invoke_array ()
23: #11 0x00120a30 in ves_icall_InternalInvoke ()
23: #12 0xb65cd96c in ?? ()
23: Backtrace stopped: previous frame identical to this frame (corrupt stack?)
23: 
23: =================================================================
23: Got a SIGSEGV while executing native code. This usually indicates
23: a fatal error in the mono runtime or one of the native libraries 
23: used by your application.
23: =================================================================
23: 
23: ...................................E.E........................EE...............E.........EEE.E................................E.EEE.EE.....EE..EE..EE................................................E.E......EE........E.E.E............EEEE............E...EEE.....E....................................................................................................EE....E..............................................................E...................................E...............
23/29 Test #23: test_csharp_bindings_full .........   Passed    8.27 sec

Full build log: http://in.waw.pl/~zbyszek/fedora/libsbml-csharp-tests-failure.log

[Frank Bergmann]

Sorry that you are having such a bad experience. This issue is fixed in trunk. There we now have a regular expression that only counts the tests as passed if the testrunner prints "failures 0" at the end, which it wouldn't have done in your case.

Sorry you are having such a bad time at this, I don't really think it is cmake that is to blame. Until now we really did not encounter so many underlying failures as you seem to get.

It would be great if you could also attach the java failure:
/builddir/build/BUILD/libSBML-5.11.0-Source/src/bindings/java/hs_err_pid22976.log

I realize that these tests were again from an arm machine. Can you let us know about the specs of it? We looked into buying them, but finding one that would run so many processes in parallel was quite hard.

thanks
Frank

[Sarah Keating]

Closing this as there has been no further response