fix(container): preserve submount propagation across pivot_root

In do_pivot_root, the post-pivot call
  do_propagation_mount(old_root, MS_REC | MS_PRIVATE)
used old_root.current_path() (a readlink of /proc/self/fd/<old_root>) as
the mount target. old_root is opened on the host root mount *before*
pivot_root(".", "."); after the pivot that readlink resolves to "/"
(i.e. the *new* root X), not to the old root. The recursive MS_PRIVATE
therefore landed on X and its entire subtree, clobbering the propagation
type that /run/media, /sys, etc. had just inherited from their host
sources during do_mounts.

Root cause: pivot_root(".", ".") makes the new rootfs X the namespace
root and parks the old root at the cwd. A subsequent readlink of the
pre-pivot old_root fd returns "/" because the detached old root mount is
reported through the new root, so the MS_REC|MS_PRIVATE recursively
re-privatized X's subtree instead of the old root's.

Fix: drop the fd-based target and use the cwd literal "." —
fchdir(old_root) has just set cwd to the old root, so "." unambiguously
refers to it. The recursive private now only touches the old root's
subtree (host-side mounts), leaving the new root
and its inherited-slave submounts untouched.

Also open new_root with O_PATH instead of O_RDONLY: fchdir accepts
O_PATH descriptors since Linux 3.5 and pivot_root(".", ".") operates on
the cwd, so a read-write descriptor is unnecessary.

Signed-off-by: Yuming He <ComixHe1895@outlook.com>

Author: ComixHe

src/linyaps_box/container.cpp

@@ -1467,7 +1467,7 @@ void do_pivot_root(const linyaps_box::container &container, const std::filesyste
     LINYAPS_BOX_DEBUG() << "start pivot root";
     LINYAPS_BOX_DEBUG() << linyaps_box::utils::inspect_fds();
     auto old_root = linyaps_box::utils::open("/", O_DIRECTORY | O_PATH | O_CLOEXEC);
-    auto new_root = linyaps_box::utils::open(rootfs.c_str(), O_DIRECTORY | O_RDONLY | O_CLOEXEC);
+    auto new_root = linyaps_box::utils::open(rootfs.c_str(), O_DIRECTORY | O_PATH | O_CLOEXEC);
 
     auto old_root_stat = linyaps_box::utils::statfs(old_root);
     LINYAPS_BOX_DEBUG() << "Pivot root old root: " << dump_mount_flags(old_root_stat.f_flags);
@@ -1492,8 +1492,12 @@ void do_pivot_root(const linyaps_box::container &container, const std::filesyste
         throw std::system_error(errno, std::system_category(), "fchdir");
     }
 
-    // make sure that umount event couldn't propagate to host
-    do_propagation_mount(old_root, MS_REC | MS_PRIVATE);
+    // make sure that umount event couldn't propagate to host.
+    // target the old root via cwd "." (set by fchdir(old_root) above); do NOT use
+    // old_root.current_path(); after pivot_root(".", ".") the old_root fd's readlink
+    // resolves to "/" (= new root X), which would recursively privatize X and all
+    // submounts (e.g. /run/media, /sys), overriding their inherited propagation.
+    syscall_mount(nullptr, ".", nullptr, MS_REC | MS_PRIVATE, nullptr);
 
     // umount old root
     ret = umount2(".", MNT_DETACH);