[PATCH] making namespace_sem global

This removes the per-namespace semaphore in favor of a global semaphore.
This can have an effect on namespace scalability.

Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

diff --git a/fs/namespace.c b/fs/namespace.c
index 65f9c0ecc21c9c0a327ac5508602483dcecc8cf3..4abee9ab009f2c2dd3603b982cdca1152c3dbd8e 100644 (file)
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -46,6 +46,7 @@ static int event;
 static struct list_head *mount_hashtable;
 static int hash_mask __read_mostly, hash_bits __read_mostly;
 static kmem_cache_t *mnt_cache;
+static struct rw_semaphore namespace_sem;
 
 static inline unsigned long hash(struct vfsmount *mnt, struct dentry *dentry)
 {
@@ -250,7 +251,7 @@ static void *m_start(struct seq_file *m, loff_t *pos)
        struct list_head *p;
        loff_t l = *pos;
 
-       down_read(&n->sem);
+       down_read(&namespace_sem);
        list_for_each(p, &n->list)
                if (!l--)
                        return list_entry(p, struct vfsmount, mnt_list);
@@ -267,8 +268,7 @@ static void *m_next(struct seq_file *m, void *v, loff_t *pos)
 
 static void m_stop(struct seq_file *m, void *v)
 {
-       struct namespace *n = m->private;
-       up_read(&n->sem);
+       up_read(&namespace_sem);
 }
 
 static inline void mangle(struct seq_file *m, const char *s)
@@ -487,7 +487,7 @@ static int do_umount(struct vfsmount *mnt, int flags)
                return retval;
        }
 
-       down_write(&current->namespace->sem);
+       down_write(&namespace_sem);
        spin_lock(&vfsmount_lock);
        event++;
 
@@ -500,7 +500,7 @@ static int do_umount(struct vfsmount *mnt, int flags)
        spin_unlock(&vfsmount_lock);
        if (retval)
                security_sb_umount_busy(mnt);
-       up_write(&current->namespace->sem);
+       up_write(&namespace_sem);
        release_mounts(&umount_list);
        return retval;
 }
@@ -678,7 +678,7 @@ static int do_loopback(struct nameidata *nd, char *old_name, int recurse)
        if (err)
                return err;
 
-       down_write(&current->namespace->sem);
+       down_write(&namespace_sem);
        err = -EINVAL;
        if (!check_mnt(nd->mnt) || !check_mnt(old_nd.mnt))
                goto out;
@@ -702,7 +702,7 @@ static int do_loopback(struct nameidata *nd, char *old_name, int recurse)
        }
 
 out:
-       up_write(&current->namespace->sem);
+       up_write(&namespace_sem);
        path_release(&old_nd);
        return err;
 }
@@ -750,7 +750,7 @@ static int do_move_mount(struct nameidata *nd, char *old_name)
        if (err)
                return err;
 
-       down_write(&current->namespace->sem);
+       down_write(&namespace_sem);
        while (d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry))
                ;
        err = -EINVAL;
@@ -795,7 +795,7 @@ out2:
 out1:
        up(&nd->dentry->d_inode->i_sem);
 out:
-       up_write(&current->namespace->sem);
+       up_write(&namespace_sem);
        if (!err)
                path_release(&parent_nd);
        path_release(&old_nd);
@@ -834,7 +834,7 @@ int do_add_mount(struct vfsmount *newmnt, struct nameidata *nd,
 {
        int err;
 
-       down_write(&current->namespace->sem);
+       down_write(&namespace_sem);
        /* Something was mounted here while we slept */
        while (d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry))
                ;
@@ -862,11 +862,11 @@ int do_add_mount(struct vfsmount *newmnt, struct nameidata *nd,
                list_add_tail(&newmnt->mnt_expire, fslist);
                spin_unlock(&vfsmount_lock);
        }
-       up_write(&current->namespace->sem);
+       up_write(&namespace_sem);
        return 0;
 
 unlock:
-       up_write(&current->namespace->sem);
+       up_write(&namespace_sem);
        mntput(newmnt);
        return err;
 }
@@ -958,9 +958,9 @@ void mark_mounts_for_expiry(struct list_head *mounts)
                get_namespace(namespace);
 
                spin_unlock(&vfsmount_lock);
-               down_write(&namespace->sem);
+               down_write(&namespace_sem);
                expire_mount(mnt, mounts, &umounts);
-               up_write(&namespace->sem);
+               up_write(&namespace_sem);
                release_mounts(&umounts);
                mntput(mnt);
                put_namespace(namespace);
@@ -1127,17 +1127,16 @@ int copy_namespace(int flags, struct task_struct *tsk)
                goto out;
 
        atomic_set(&new_ns->count, 1);
-       init_rwsem(&new_ns->sem);
        INIT_LIST_HEAD(&new_ns->list);
        init_waitqueue_head(&new_ns->poll);
        new_ns->event = 0;
 
-       down_write(&tsk->namespace->sem);
+       down_write(&namespace_sem);
        /* First pass: copy the tree topology */
        new_ns->root = copy_tree(namespace->root, namespace->root->mnt_root,
                                        CL_EXPIRE);
        if (!new_ns->root) {
-               up_write(&tsk->namespace->sem);
+               up_write(&namespace_sem);
                kfree(new_ns);
                goto out;
        }
@@ -1171,7 +1170,7 @@ int copy_namespace(int flags, struct task_struct *tsk)
                p = next_mnt(p, namespace->root);
                q = next_mnt(q, new_ns->root);
        }
-       up_write(&tsk->namespace->sem);
+       up_write(&namespace_sem);
 
        tsk->namespace = new_ns;
 
@@ -1356,7 +1355,7 @@ asmlinkage long sys_pivot_root(const char __user * new_root,
        user_nd.mnt = mntget(current->fs->rootmnt);
        user_nd.dentry = dget(current->fs->root);
        read_unlock(&current->fs->lock);
-       down_write(&current->namespace->sem);
+       down_write(&namespace_sem);
        down(&old_nd.dentry->d_inode->i_sem);
        error = -EINVAL;
        if (!check_mnt(user_nd.mnt))
@@ -1407,7 +1406,7 @@ asmlinkage long sys_pivot_root(const char __user * new_root,
        path_release(&parent_nd);
 out2:
        up(&old_nd.dentry->d_inode->i_sem);
-       up_write(&current->namespace->sem);
+       up_write(&namespace_sem);
        path_release(&user_nd);
        path_release(&old_nd);
 out1:
@@ -1434,7 +1433,6 @@ static void __init init_mount_tree(void)
                panic("Can't allocate initial namespace");
        atomic_set(&namespace->count, 1);
        INIT_LIST_HEAD(&namespace->list);
-       init_rwsem(&namespace->sem);
        init_waitqueue_head(&namespace->poll);
        namespace->event = 0;
        list_add(&mnt->mnt_list, &namespace->list);
@@ -1459,6 +1457,8 @@ void __init mnt_init(unsigned long mempages)
        unsigned int nr_hash;
        int i;
 
+       init_rwsem(&namespace_sem);
+
        mnt_cache = kmem_cache_create("mnt_cache", sizeof(struct vfsmount),
                        0, SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL, NULL);
 
@@ -1507,11 +1507,11 @@ void __put_namespace(struct namespace *namespace)
        LIST_HEAD(umount_list);
        namespace->root = NULL;
        spin_unlock(&vfsmount_lock);
-       down_write(&namespace->sem);
+       down_write(&namespace_sem);
        spin_lock(&vfsmount_lock);
        umount_tree(root, &umount_list);
        spin_unlock(&vfsmount_lock);
-       up_write(&namespace->sem);
+       up_write(&namespace_sem);
        release_mounts(&umount_list);
        kfree(namespace);
 }

diff --git a/include/linux/namespace.h b/include/linux/namespace.h
index 6f0f25d64c383dbb556e7ef7fdf058d979b804f5..6731977c4c13a00e39a251a12d7be31311960062 100644 (file)
--- a/include/linux/namespace.h
+++ b/include/linux/namespace.h
@@ -9,7 +9,6 @@ struct namespace {
        atomic_t                count;
        struct vfsmount *       root;
        struct list_head        list;
-       struct rw_semaphore     sem;
        wait_queue_head_t poll;
        int event;
 };