Fatal signal 22 / need some help interpreting crash log.txt

hi all!

At a recent performance my vcv 2 patch file decided to keep crashing upon launch. Luckily I had manual incremental saves from the night before, but it was a little odd and frustrating since when I tested the patch that morning before the show, everything seemed to be fine.

The crush only happened once I set up my PC at the venue a few hours later. The only variables were that I had only one monitor and one audio interface attached to my PC rather than two and two, respectively, as I normally do at home. I didn’t see how that would cause a crash though, and have since tested the crashing patch files at home with two audio interface and two monitors connected. They still crash.

I suspect it’s a module causing the problem, but I’m having difficulties reading and understanding the stack trace in the log.txt. It goes as follows:

[3.114 fatal adapters/standalone.cpp:49 fatalSignalHandler] Fatal signal 22. Stack trace:
17:  0x0
16: raise 0x7ffcdbebabe0
15: abort 0x7ffcdbebf1e0
14: wassert 0x7ffcdbebd210
13: ZN9HostedVST17addBusPortMappingEiiRKSt6vectorIiSaIiEE 0x7ffc676f9290
12: ZN5THostILi0ELi2ELi2ELi16EE25reshapeBusChannelMappingsEv 0x7ffc6776ea80
11: ZN5THostILi0ELi2ELi2ELi16EE12dataFromJsonEP6json_t 0x7ffc6776da90
10: ZN4rack6engine6Engine8fromJsonEP6json_t 0x7ffc851135d0
9: ZN4rack5patch7Manager8fromJsonEP6json_t 0x7ffc850c3350
8: ZN4rack5patch7Manager12loadAutosaveEv 0x7ffc850c3690
7: ZN4rack5patch7Manager10loadActionENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE 0x7ffc850c4a70
6: ZN4rack5patch7Manager6launchENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE 0x7ffc850c4cc0
5: ZN4rack5patch7Manager6launchENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE 0x7ffc850c4cc0
4: ZN4rack5patch7Manager6launchENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE 0x7ffc850c4cc0
3: ZN4rack5patch7Manager6launchENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE 0x7ffc850c4cc0
2: ZN4rack5patch7Manager6launchENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE 0x7ffc850c4cc0
1: BaseThreadInitThunk 0x7ffcdc377330
0: RtlUserThreadStart 0x7ffcdd682690

Since I can’t upload files yet, I put up the entire log on pastebin here if anyone’s interested. Any help interpreting the crash would be greatly appreciated!! Luckily I was able to perform still but would like to avoid crashes in the future if possible, whether that means stopping using a particular module or else…

Thanks for reading!! xxdd

Maybe someone else will see something more specific, but it appears to me that the problem is a VST that is loaded into one of your Host-FX modules.

1 Like

This is an assert firing in the Host module in its dataFromJson override. Since the VCV Host module is paid and provided by VCV, open a support ticket at VCV - Support (vcvrack.com).

The issue probably isn’t in the hosted VST: the crashing code is in Host.

VCV will probably want the settings.json, patch json, and the log.txt in your support request. At the least, they’ll need to know what VST you’re using, and the OS you’re on.

1 Like

Awesome, I submitted a ticket so hopefully I’ll hear something back.

My suspicion was also the ‘addBusPortMapping’ part. One of the VST Hosts has a CV connecting to it for some modulation. But somehow the crashes only happens with 2 of the later files. The previous saves were all fine. Baffling…

1 Like

the salty to be downtalked user again:

Having here Fatal signal 22 or 27 crashes also with the host modules when using it with some modulations going on. Found that while the modulation is steady, means coming from controller knobs or faders the file does not crash after saving and reloading. But if the modulation is moving , from a LFO or some random values the Host module crashes after saving and reloading.

Means the project works over hours without any problems while modulation is going wild, the reload results in a crash at start up.

There are few tricks to avoid this startup crashes, like move your modulation cable to an unconneted Modulation input before saving and reconnect after loading your file. But if you forgot one connection —— crash

There are even ways to rescue your crashing files by some manual work, just to let all know!

All true for osx 10.13 , 12.6.7 Intel and 14 on M1

Think all this gurus here still need NO tester for your Apple versions? Please feel free to downtalk me again

If you have a patch that only uses freely-available VSTs and crashes reliably, VCV Support would very much appreciate a copy!

Oooh interesting!! That makes sense & was my suspicion too! Ok really good to know that someone else also had the same problem hehe. Losing work really sucks!

What have you tried to rescue the file? I’ll avoid having modulation to Hosts altogether from now on but would love to have a work at the back of my mind just in case:))

the salty to be downtalked user again:

As expected, no reaction at all! Thank You so much