ci files with extensive code blocks result in incorrect line numbering in syntax errors and in gdb
unclear exactly what trips up the translator, but I've found the charmxi emitted line numbers for errors to be garbage (wrong by hundreds of line) on an application with extensive serial blocks.
Same issue results in gdb tracing showing wildly incorrect source lines for parts of the stack that are in the ci file.
Affected code is CharmVSS, which is under NDA. So may need to find a way to replicate this to get some insight into bug reproduction.