> No, there is nothing natural (and this can even be wrong). The statements > must have the source location corresponding to the source construct they are > generated for, which may be totally different from that of the insertion > point. Yes, that can generate jumpiness in GDB, but this is far better that > breaking the coverage info by giving the same source location to instructions > that have different coverage status.
For the unittest I provided, setting the inserted stmt with UNKNOWN_LOCATION will: * break the coverage * increase jumpiness in gdb Setting location to UNKNOWN_LOCATION is like setting it to random because compiler may put this stmt as the entry point of a BB (as in the unittest). Thus setting deterministic locations for inserted stmt will improve debugability and reduce jumpiness. Thanks, Dehao > > -- > Eric Botcazou
