On Sun, Mar 07, 2021 at 11:51:20PM +0000, Al Viro wrote:
> On Thu, Mar 04, 2021 at 01:11:33PM +0300, Alexander Mikhalitsyn wrote:
>
> > That problem connected with CRIU (Checkpoint-Restore in Userspace) project.
> > In CRIU we have support of autofs mounts C/R. To acheive that we need to use
> > ioctl's from /dev/autofs to get data about mounts, restore mount as catatonic
> > (if needed), change pipe fd and so on. But the problem is that during CRIU
> > dump we may meet situation when VFS subtree where autofs mount present was
> > overmounted as whole.
> >
> > Simpliest example is /proc/sys/fs/binfmt_misc. This mount present on most
> > GNU/Linux distributions by default. For instance on my Fedora 33:
> >
> > trigger automount of binfmt_misc
> > $ ls /proc/sys/fs/binfmt_misc
> >
> > $ cat /proc/1/mountinfo | grep binfmt
> > 35 24 0:36 / /proc/sys/fs/binfmt_misc rw,relatime shared:16 - autofs systemd-1 rw,...,direct,pipe_ino=223
> > 632 35 0:56 / /proc/sys/fs/binfmt_misc rw,...,relatime shared:315 - binfmt_misc binfmt_misc rw
> >
> > $ sudo unshare -m -p --fork --mount-proc sh
> > # cat /proc/self/mountinfo | grep "/proc"
> > 828 809 0:23 / /proc rw,nosuid,nodev,noexec,relatime - proc proc rw
> > 829 828 0:36 / /proc/sys/fs/binfmt_misc rw,relatime - autofs systemd-1 rw,...,direct,pipe_ino=223
> > 943 829 0:56 / /proc/sys/fs/binfmt_misc rw,...,relatime - binfmt_misc binfmt_misc rw
> > 949 828 0:57 / /proc rw...,relatime - proc proc rw
> >
> > As we can see now autofs mount /proc/sys/fs/binfmt_misc is inaccessible.
> > If we do something like:
> >
> > struct autofs_dev_ioctl *param;
> > param = malloc(...);
> > devfd = open("/dev/autofs", O_RDONLY);
> > init_autofs_dev_ioctl(param);
> > param->size = size;
> > strcpy(param->path, "/proc/sys/fs/binfmt_misc");
> > param->openmount.devid = 36;
> > err = ioctl(devfd, AUTOFS_DEV_IOCTL_OPENMOUNT, param)
> >
> > now we get err = -ENOENT.
>
> Where does that -ENOENT come from? AFAICS, pathwalk ought to succeed and
> return you the root of overmounting binfmt_misc. Why doesn't the loop in
> find_autofs_mount() locate anything it would accept?
>
> I really dislike the patch - the whole "normalize path" thing is fundamentally
> bogus, not to mention the iterator over all mounts, etc., so I would like to
> understand what the hell is going on before even thinking of *not* NAKing
> it on sight.
Wait, so you have /proc overmounted, without anything autofs-related on
/proc/sys/fs/binfmt_misc and still want to have the pathname resolved, just
because it would've resolved with that overmount of /proc removed?
I hope I'm misreading you; in case I'm not, the ABI is extremely
tasteless and until you manage to document the exact semantics you want
for param->path, consider it NAKed.
