Prev: Help in understanding the tasklets
Next: [PATCH 3/4] procfs: move pid_ns_prepare_proc() from copy_process() to create_pid_namespace()
From: Oleg Nesterov on 19 Jun 2010 15:20
Depending on whether it is called by the init_pid_ns or not, proc_get_sb()
does different things to initialize PROC_I(s_root->d_inode)->pid, and both
look ugly, imho.

- init_pid_ns tries to initialize ->pid every time proc_get_sb() is
called, and we check the global proc_mnt != NULL to ensure proc_get_sb()
was already called in the past and thus it is safe to use sb->s_root.

- sub-namespaces set ->pid at MS_KERNMOUNT stage and thus proc_get_sb()
can'be called before the first alloc_pid() in copy_process(), ugly.

Consolidate this code, and initialize PROC_I(s_root->d_inode)->pid when
the proc fs is actually mounted (MS_KERNMOUNT is not set). This also
allows us to do more cleanups.

Signed-off-by: Oleg Nesterov <oleg(a)redhat.com>
---

fs/proc/root.c | 23 +++++++----------------
1 file changed, 7 insertions(+), 16 deletions(-)

--- 35-rc3/fs/proc/root.c~PNS_1_PROC_GET_SB_ROOT_PID 2010-06-19 19:20:23.000000000 +0200
+++ 35-rc3/fs/proc/root.c 2010-06-19 19:21:38.000000000 +0200
@@ -41,18 +41,6 @@ static int proc_get_sb(struct file_syste
int err;
struct super_block *sb;
struct pid_namespace *ns;
- struct proc_inode *ei;
-
- if (proc_mnt) {
- /* Seed the root directory with a pid so it doesn't need
- * to be special in base.c. I would do this earlier but
- * the only task alive when /proc is mounted the first time
- * is the init_task and it doesn't have any pids.
- */
- ei = PROC_I(proc_mnt->mnt_sb->s_root->d_inode);
- if (!ei->pid)
- ei->pid = find_get_pid(1);
- }

if (flags & MS_KERNMOUNT)
ns = (struct pid_namespace *)data;
@@ -71,15 +59,18 @@ static int proc_get_sb(struct file_syste
return err;
}

- ei = PROC_I(sb->s_root->d_inode);
+ sb->s_flags |= MS_ACTIVE;
+ ns->proc_mnt = mnt;
+ }
+
+ /* MS_KERNMOUNT is called before ns has any pids */
+ if (!(flags & MS_KERNMOUNT)) {
+ struct proc_inode *ei = PROC_I(sb->s_root->d_inode);
if (!ei->pid) {
rcu_read_lock();
ei->pid = get_pid(find_pid_ns(1, ns));
rcu_read_unlock();
}
-
- sb->s_flags |= MS_ACTIVE;
- ns->proc_mnt = mnt;
}

simple_set_mnt(mnt, sb);

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: Help in understanding the tasklets
Next: [PATCH 3/4] procfs: move pid_ns_prepare_proc() from copy_process() to create_pid_namespace()