[Ocfs2-tools-devel] [PATCH 0/6] Move to ocfs2_hb_ctl for all stacks

[Joel Becker]

We have determined that ocfs2 can only leave a cluster group safely in
put_super(7).  The presence of bind mounts, rbind mounts, and shared
subtrees make tracking mountpoints impossible in userspace.

To solve this, we move the ocfs2_hb_ctl call out of o2cb and into the
generic stack glue code.  ocfs2_hb_ctl will always be called for all
cluster stacks.

First, we make ocfs2_hb_ctl ignore non-o2cb stacks, while umount.ocfs2
does no cluster work for o2cb stacks.  This enables ocfs2_hb_ctl for the
classic stack just like it used to be.

Next, ocfs2_hb_ctl has to be able to disconnect filesystems in all
cluster stack environments.  It can easily do this by getting the
cluster stack information with ocfs2_fill_cluster_desc.

Next, ocfs2_controld has to treat all mountpoints as the same 'service'.
This allows one call from the kernel to do the final disconnect.

Finally, umount.ocfs2 just plain goes away.

The code is also available on the 'full-hb-ctl' branch of the
ocfs2-tools git repository.

View:
http://oss.oracle.com/git/?p=ocfs2-tools.git;a=shortlog;h=full-hb-ctl
Pull:
git pull git://oss.oracle.com/git/ocfs2-tools.git full-hb-ctl