A fix was posted on this list which seems to avoid segfaults when using solr+tika. I reposted it not long ago in response to another question. Before exploring the other segfaults it would be worthwhile getting that fix. If other segfaults continue then I suggest posting the logging from Dovecot
https://github.com/dovecot/core/commit/043d111cea78533cf22d3bfe0872c8376d842... (DOV-4090)
Scheduled for 2.3.13.
Meanwhile, I also found the fix from October. I can confirm that the issue occurred on 2.3.11.3 only. I switched to the previous version (which is pretty easy thanks to Docker) and the issue disappeared. All messages were indexed properly afterwards.
I applied the fix and will observe it for the next couple of days. The assertions have been misplaced anyway from my viewpoint because the asserted fields (or properties or whatever, I don't know how they're called in C) are not accessed in every case.
I am happy that this is scheduled for 2.3.13. Although, I hope that the payload_input or payload_output won't be NULL when payload_finished is true. :)
Thanks a lot to you guys for all the effort. I would not be able to resolve this on my own without your help.
Best, Manuel