jimingham wrote: > @jimingham > > > Have you checked to make sure this also works when there is a source-path > > mapping in place for the files that also differ by realpath vrs. symlink in > > the part past the part the source-path mapping substitutes? That might > > "just work" but it also seems like it might not, so it would be good to > > have a test for that. > > I checked one scenario and it worked: The user set a breakpoint in file A, > which is source-mapped to file B. The line table has file C, which is > realpath'ed into file B. Hence a match. See [this > example](https://github.com/royitaqi/lldb_demos/blob/main/realpath/README.md#run-with-realpathing-and-source-map). > It works because the source-map is done [here](https://fburl.com/afz6bxlk) > on the input file A, while the realpath'ing is done on the support file C, so > the two are really orthogonal. > > I didn't fully understand your question. Did you have a different scenario in > mind? > > Also, I have updated the patch according to most of your comments. Only two > still open, for which I hope for your clarification.
No, I wanted to make sure that scenario worked and that it had a test (since we want it to keep working). Looks like you did show that it works, but I didn't see a test for that scenario so that it will keep working. Did I miss that? https://github.com/llvm/llvm-project/pull/102223 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
