My understanding is the new mount API will eventually replace the current mount API, so it needs to be feature for feature compatible before that happens. I was looking to use the read only bind mount as a template for reimplementing shiftfs when I discovered that you can't actually create a read only bind mount with the new API. The problem is that fspick() will only reconfigure the underlying superblock, which you don't want because you only want the bound subtree to become read only and open_tree()/move_mount() doesn't give you any facility to add or change options on the bind. The way it works with the current API is that you create a bind mount with MS_BIND and then remount it with MS_BIND|MS_REMOUNT|MS_RDONLY which makes the subtree go read only. It looks like the nearest way of getting this to work with the new API would be to add an additional fspick() flag, say FSPICK_BIND, that tells it to operate on the subtree not the superblock and then pass this down to fsconfig() so it will accept the "ro" and "rw" flags and the reconfigure command and nothing else (although I'd like it to accept additional flags on down the road for the shiftfs rewrite). There are some caveats, in that fspick/fsconfig seem to be fairly adamantly coded to work with the superblock, so the subtree handling will have to be special cased, but if everyone's happy with that, I can code up something. Regards, James
