labath added a comment. My main question would be whether it is possible to specify all the lldb-mi commands up-front, which is kinda required for this approach. I was against this for the lldb-server tests, because there we often deal with values which are not sufficiently deterministic (for example, because they represent addresses in the inferior address space), so we needed the know the output of one command before we could specify the next one. I felt that trying to capture this statically would end up with a reinvention of a domain-specific scripting language.
*However*, the situation may be better in lldb-mi, as here we are sitting on top of layers which will enable us to reason about things symbolically. In a random sample of lldb-mi tests I did not find any inter-command dependencies, *except* the "-file-exec-and-symbols", where we pass the full path of the built executable. This can probably be worked around somehow. Repository: rL LLVM https://reviews.llvm.org/D46588 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
