JDevlieghere added a comment. In D55827#1334671 <https://reviews.llvm.org/D55827#1334671>, @labath wrote:
> Hm... this seems like an important issue in the `RealFileSystem` (or our > usage of it), and I'm not sure it should be papered over like that. I mean, > lldb is a library, and requiring every call to `chdir` in the whole process > go through "our" implementation is not very friendly towards everyone else we > happen to be sharing a process with. On the other hand (playing the devil's advocate) it could be considered nice that lldb as a library isn't affected by that if we explicitly set the cwd? > It sounds like we want a VFS that is slightly more "real" than the > RealFileSystem, and which shares the notion of the CWD with the OS. Did you have a look at the comment in `VirtualFileSystem.cpp` that explains why they went this route? // FIXME: chdir is thread hostile; on the other hand, creating the same // behavior as chdir is complex: chdir resolves the path once, thus // guaranteeing that all subsequent relative path operations work // on the same path the original chdir resulted in. This makes a // difference for example on network filesystems, where symlinks might be // switched during runtime of the tool. Fixing this depends on having a // file system abstraction that allows openat() style interactions. Wouldn't we encounter the same problem? Repository: rLLDB LLDB CHANGES SINCE LAST ACTION https://reviews.llvm.org/D55827/new/ https://reviews.llvm.org/D55827 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
