labath wrote: > > With this in mind, deleting the file is representative of an editor wanting > > to save the file using the deletion strategy,... > > Yes. I had meant it is not representative of what Android Studio (or _some_ > > editors) does. > > > think it's not our goal here to ensure the editor cannot save the file if > > the setting is true. > > I agree that having a test that sets `use_source_cache=true` and expects > failure behavior is not a goal. But, it is how we _observe_ that the test > input `cpp` source file we use satisfies the condition to use mmap. If the > condition to use mmap changes (e.g., from filesize=16K to filesize=32K), then > the test input source file will no longer trigger mmap, yet the > `use_source_cache=false` test will keep passing. So, I think having that > negative test case, despite being weird, still serves some purpose.
Makes sense (and now that I think about it, I think I used some line of reasoning like this on the original patch that introduced this). https://github.com/llvm/llvm-project/pull/111237 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
