On Wed, Sep 28 2022, John Cai via GitGitGadget wrote: > From: John Cai <johncai86@xxxxxxxxx> This looks good! Just a clarifying question/comment: > Since we can't do the cleanup in a portable and signal-safe way, skip > the cleanup when we're handling a signal. > > This means that when signal handling, the temporary directory may not > get cleaned up properly. This is mitigated by b3cecf49ea (tmp-objdir: new > API for creating temporary writable databases, 2021-12-06) which changed > the default name and allows gc to clean up these temporary directories. I think this still doesn't cover the common case of the "atexit" handler saving the day, as Jeff King pointed out in v1: https://lore.kernel.org/git/YzLiI1HZeBszsIJq@xxxxxxxxxxxxxxxxxxxxxxx/ I think it's fine to have this proceed as-is (although if you're doing a v4 anyway, maybe we want to update tho commit message). I.e. is it correct that we'll now only skip this if we get a signal, *and* it's a fatal signal, or we otherwise die/exit/abort/whatever before we can get to our atexit() handler?
