On Sun, 12 Mar 2017 12:06:32 -0700, Jun Wu wrote: > Excerpts from Yuya Nishihara's message of 2017-03-12 11:58:40 -0700: > > > That's good to know. I actually had 3 patches to move it to vfs, but was > > > concerned about the future "repostorage" layer holding vfs objects that > > > won't get invalidated across chg workers. > > > > If the filesystem at the cached repository location changed, cache should be > > invalidated and repostorage would be recreated anyway. > > Even if location does not change, the user can run "mount" to change the > "mtab" that affects us.
My point is that the repository should be considered changed if he mounted another filesystem onto that location. vfs caches cansymlink, execflag, etc., which are all invalid. > Querying "mtab" confidently is expensive. And "mtab" can answer all vfs > queries. So I think it's better to cache "mtab" per process, instead of > caching "fstype" per vfs. I won't object to it if it can be implemented cleanly. _______________________________________________ Mercurial-devel mailing list Mercurial-devel@mercurial-scm.org https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel