On Friday, July 19, 2019, Kang-Che Sung wrote:
>
> There's side benefit for this patch: In case that overmount fails, we can
have
> a rootfs kept intact (instead of almost destroyed).
>
Correction. It's just a side effect, not a "benefit" worth talking about.
Speaking of, since we are now
On Fri, Jul 19, 2019 at 3:27 PM 阿保 純一 wrote:
>
> As the author said in the comment of util-linux/switch_root.c, current
> implementation leaves the mount point of new root-file-system without rmdir().
> As long as I experimented on a linux kernel, current process of "/" still
> points old
As the author said in the comment of util-linux/switch_root.c, current
implementation leaves the mount point of new root-file-system without rmdir().
As long as I experimented on a linux kernel, current process of "/" still
points old root-file-system even "/" is overmounted. So we can still
