Hello! For a while now we have been compiling our fuzz targets with llvm-cov coverage collection instrumentation like that, with -fcoverage-mapping and -fprofile-instr-generate :

And everything seems to be working fine. We are collecting our coverage in fuzz-time, it works fine even with parallel fuzzing because we are just merging profdata. But we know that often different instrumentations don't work good with one another and shouldn't be combined. Are we missing something or afl-lto instrumentation was built with that in mind? Is there any well-known conflicts between these two?