Messages in this thread | | | Date | Sat, 28 Sep 2013 13:43:49 -0700 | Subject | Re: [rfc][possible solution] RCU vfsmounts | From | Linus Torvalds <> |
| |
On Sat, Sep 28, 2013 at 1:27 PM, Al Viro <viro@zeniv.linux.org.uk> wrote: > FWIW, I think I have a kinda-sorta solution for that and I'd like > to hear your comments on that. I want to replace vfsmount_lock with seqlock > and store additional seq number in nameidata, set to vfsmount_seq in the > beginning and rechecked in unlazy_walk/complete_walk.
Yes, that would be lovely.
> The obvious variant would be to have unlazy_walk/complete_walk to > grab refcount, check vfsmount_seq and mntput on mismatch. The trouble > with that is race with what would've been the final mntput() done by > umount(2); complete_walk() would drop that temporary reference and > fail, all right, but... we would get a umount(2) returning without having > actually shut the filesystem down. Said shutdown would happen in whoever > had been doing pathname resolution that stepped into the race.
Sounds reasonable to to me.
Side note: I really wish there was some way to avoid having to finalize the vfsmount entirely for some common things. For example, "[l]stat[at]()" really doesn't need it for the common cases (network filesystems may need to revalidate), and is a very critical operation, and we *could* just look up the inode under RCU and never finalize the dentry _or_ the vfsmount. However, very annoyingly, the security layer wants the vfsmount, and we don't know if that is RCU-safe...
Linus
| |