Revision b91e1302ad9b80c174a4855533f7e3aa2873355e authored by Linus Torvalds on 27 December 2016, 19:40:38 UTC, committed by Linus Torvalds on 29 December 2016, 19:03:15 UTC
In commit 62906027091f ("mm: add PageWaiters indicating tasks are
waiting for a page bit") Nick Piggin made our page locking no longer
unconditionally touch the hashed page waitqueue, which not only helps
performance in general, but is particularly helpful on NUMA machines
where the hashed wait queues can bounce around a lot.
However, the "clear lock bit atomically and then test the waiters bit"
sequence turns out to be much more expensive than it needs to be,
because you get a nasty stall when trying to access the same word that
just got updated atomically.
On architectures where locking is done with LL/SC, this would be trivial
to fix with a new primitive that clears one bit and tests another
atomically, but that ends up not working on x86, where the only atomic
operations that return the result end up being cmpxchg and xadd. The
atomic bit operations return the old value of the same bit we changed,
not the value of an unrelated bit.
On x86, we could put the lock bit in the high bit of the byte, and use
"xadd" with that bit (where the overflow ends up not touching other
bits), and look at the other bits of the result. However, an even
simpler model is to just use a regular atomic "and" to clear the lock
bit, and then the sign bit in eflags will indicate the resulting state
of the unrelated bit #7.
So by moving the PageWaiters bit up to bit #7, we can atomically clear
the lock bit and test the waiters bit on x86 too. And architectures
with LL/SC (which is all the usual RISC suspects), the particular bit
doesn't matter, so they are fine with this approach too.
This avoids the extra access to the same atomic word, and thus avoids
the costly stall at page unlock time.
The only downside is that the interface ends up being a bit odd and
specialized: clear a bit in a byte, and test the sign bit. Nick doesn't
love the resulting name of the new primitive, but I'd rather make the
name be descriptive and very clear about the limitation imposed by
trying to work across all relevant architectures than make it be some
generic thing that doesn't make the odd semantics explicit.
So this introduces the new architecture primitive
clear_bit_unlock_is_negative_byte();
and adds the trivial implementation for x86. We have a generic
non-optimized fallback (that just does a "clear_bit()"+"test_bit(7)"
combination) which can be overridden by any architecture that can do
better. According to Nick, Power has the same hickup x86 has, for
example, but some other architectures may not even care.
All these optimizations mean that my page locking stress-test (which is
just executing a lot of small short-lived shell scripts: "make test" in
the git source tree) no longer makes our page locking look horribly bad.
Before all these optimizations, just the unlock_page() costs were just
over 3% of all CPU overhead on "make test". After this, it's down to
0.66%, so just a quarter of the cost it used to be.
(The difference on NUMA is bigger, but there this micro-optimization is
likely less noticeable, since the big issue on NUMA was not the accesses
to 'struct page', but the waitqueue accesses that were already removed
by Nick's earlier commit).
Acked-by: Nick Piggin <npiggin@gmail.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Andrew Lutomirski <luto@kernel.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent 2d706e7
pnode.c
/*
* linux/fs/pnode.c
*
* (C) Copyright IBM Corporation 2005.
* Released under GPL v2.
* Author : Ram Pai (linuxram@us.ibm.com)
*
*/
#include <linux/mnt_namespace.h>
#include <linux/mount.h>
#include <linux/fs.h>
#include <linux/nsproxy.h>
#include "internal.h"
#include "pnode.h"
/* return the next shared peer mount of @p */
static inline struct mount *next_peer(struct mount *p)
{
return list_entry(p->mnt_share.next, struct mount, mnt_share);
}
static inline struct mount *first_slave(struct mount *p)
{
return list_entry(p->mnt_slave_list.next, struct mount, mnt_slave);
}
static inline struct mount *next_slave(struct mount *p)
{
return list_entry(p->mnt_slave.next, struct mount, mnt_slave);
}
static struct mount *get_peer_under_root(struct mount *mnt,
struct mnt_namespace *ns,
const struct path *root)
{
struct mount *m = mnt;
do {
/* Check the namespace first for optimization */
if (m->mnt_ns == ns && is_path_reachable(m, m->mnt.mnt_root, root))
return m;
m = next_peer(m);
} while (m != mnt);
return NULL;
}
/*
* Get ID of closest dominating peer group having a representative
* under the given root.
*
* Caller must hold namespace_sem
*/
int get_dominating_id(struct mount *mnt, const struct path *root)
{
struct mount *m;
for (m = mnt->mnt_master; m != NULL; m = m->mnt_master) {
struct mount *d = get_peer_under_root(m, mnt->mnt_ns, root);
if (d)
return d->mnt_group_id;
}
return 0;
}
static int do_make_slave(struct mount *mnt)
{
struct mount *master, *slave_mnt;
if (list_empty(&mnt->mnt_share)) {
if (IS_MNT_SHARED(mnt)) {
mnt_release_group_id(mnt);
CLEAR_MNT_SHARED(mnt);
}
master = mnt->mnt_master;
if (!master) {
struct list_head *p = &mnt->mnt_slave_list;
while (!list_empty(p)) {
slave_mnt = list_first_entry(p,
struct mount, mnt_slave);
list_del_init(&slave_mnt->mnt_slave);
slave_mnt->mnt_master = NULL;
}
return 0;
}
} else {
struct mount *m;
/*
* slave 'mnt' to a peer mount that has the
* same root dentry. If none is available then
* slave it to anything that is available.
*/
for (m = master = next_peer(mnt); m != mnt; m = next_peer(m)) {
if (m->mnt.mnt_root == mnt->mnt.mnt_root) {
master = m;
break;
}
}
list_del_init(&mnt->mnt_share);
mnt->mnt_group_id = 0;
CLEAR_MNT_SHARED(mnt);
}
list_for_each_entry(slave_mnt, &mnt->mnt_slave_list, mnt_slave)
slave_mnt->mnt_master = master;
list_move(&mnt->mnt_slave, &master->mnt_slave_list);
list_splice(&mnt->mnt_slave_list, master->mnt_slave_list.prev);
INIT_LIST_HEAD(&mnt->mnt_slave_list);
mnt->mnt_master = master;
return 0;
}
/*
* vfsmount lock must be held for write
*/
void change_mnt_propagation(struct mount *mnt, int type)
{
if (type == MS_SHARED) {
set_mnt_shared(mnt);
return;
}
do_make_slave(mnt);
if (type != MS_SLAVE) {
list_del_init(&mnt->mnt_slave);
mnt->mnt_master = NULL;
if (type == MS_UNBINDABLE)
mnt->mnt.mnt_flags |= MNT_UNBINDABLE;
else
mnt->mnt.mnt_flags &= ~MNT_UNBINDABLE;
}
}
/*
* get the next mount in the propagation tree.
* @m: the mount seen last
* @origin: the original mount from where the tree walk initiated
*
* Note that peer groups form contiguous segments of slave lists.
* We rely on that in get_source() to be able to find out if
* vfsmount found while iterating with propagation_next() is
* a peer of one we'd found earlier.
*/
static struct mount *propagation_next(struct mount *m,
struct mount *origin)
{
/* are there any slaves of this mount? */
if (!IS_MNT_NEW(m) && !list_empty(&m->mnt_slave_list))
return first_slave(m);
while (1) {
struct mount *master = m->mnt_master;
if (master == origin->mnt_master) {
struct mount *next = next_peer(m);
return (next == origin) ? NULL : next;
} else if (m->mnt_slave.next != &master->mnt_slave_list)
return next_slave(m);
/* back at master */
m = master;
}
}
static struct mount *next_group(struct mount *m, struct mount *origin)
{
while (1) {
while (1) {
struct mount *next;
if (!IS_MNT_NEW(m) && !list_empty(&m->mnt_slave_list))
return first_slave(m);
next = next_peer(m);
if (m->mnt_group_id == origin->mnt_group_id) {
if (next == origin)
return NULL;
} else if (m->mnt_slave.next != &next->mnt_slave)
break;
m = next;
}
/* m is the last peer */
while (1) {
struct mount *master = m->mnt_master;
if (m->mnt_slave.next != &master->mnt_slave_list)
return next_slave(m);
m = next_peer(master);
if (master->mnt_group_id == origin->mnt_group_id)
break;
if (master->mnt_slave.next == &m->mnt_slave)
break;
m = master;
}
if (m == origin)
return NULL;
}
}
/* all accesses are serialized by namespace_sem */
static struct user_namespace *user_ns;
static struct mount *last_dest, *first_source, *last_source, *dest_master;
static struct mountpoint *mp;
static struct hlist_head *list;
static inline bool peers(struct mount *m1, struct mount *m2)
{
return m1->mnt_group_id == m2->mnt_group_id && m1->mnt_group_id;
}
static int propagate_one(struct mount *m)
{
struct mount *child;
int type;
/* skip ones added by this propagate_mnt() */
if (IS_MNT_NEW(m))
return 0;
/* skip if mountpoint isn't covered by it */
if (!is_subdir(mp->m_dentry, m->mnt.mnt_root))
return 0;
if (peers(m, last_dest)) {
type = CL_MAKE_SHARED;
} else {
struct mount *n, *p;
bool done;
for (n = m; ; n = p) {
p = n->mnt_master;
if (p == dest_master || IS_MNT_MARKED(p))
break;
}
do {
struct mount *parent = last_source->mnt_parent;
if (last_source == first_source)
break;
done = parent->mnt_master == p;
if (done && peers(n, parent))
break;
last_source = last_source->mnt_master;
} while (!done);
type = CL_SLAVE;
/* beginning of peer group among the slaves? */
if (IS_MNT_SHARED(m))
type |= CL_MAKE_SHARED;
}
/* Notice when we are propagating across user namespaces */
if (m->mnt_ns->user_ns != user_ns)
type |= CL_UNPRIVILEGED;
child = copy_tree(last_source, last_source->mnt.mnt_root, type);
if (IS_ERR(child))
return PTR_ERR(child);
child->mnt.mnt_flags &= ~MNT_LOCKED;
mnt_set_mountpoint(m, mp, child);
last_dest = m;
last_source = child;
if (m->mnt_master != dest_master) {
read_seqlock_excl(&mount_lock);
SET_MNT_MARK(m->mnt_master);
read_sequnlock_excl(&mount_lock);
}
hlist_add_head(&child->mnt_hash, list);
return count_mounts(m->mnt_ns, child);
}
/*
* mount 'source_mnt' under the destination 'dest_mnt' at
* dentry 'dest_dentry'. And propagate that mount to
* all the peer and slave mounts of 'dest_mnt'.
* Link all the new mounts into a propagation tree headed at
* source_mnt. Also link all the new mounts using ->mnt_list
* headed at source_mnt's ->mnt_list
*
* @dest_mnt: destination mount.
* @dest_dentry: destination dentry.
* @source_mnt: source mount.
* @tree_list : list of heads of trees to be attached.
*/
int propagate_mnt(struct mount *dest_mnt, struct mountpoint *dest_mp,
struct mount *source_mnt, struct hlist_head *tree_list)
{
struct mount *m, *n;
int ret = 0;
/*
* we don't want to bother passing tons of arguments to
* propagate_one(); everything is serialized by namespace_sem,
* so globals will do just fine.
*/
user_ns = current->nsproxy->mnt_ns->user_ns;
last_dest = dest_mnt;
first_source = source_mnt;
last_source = source_mnt;
mp = dest_mp;
list = tree_list;
dest_master = dest_mnt->mnt_master;
/* all peers of dest_mnt, except dest_mnt itself */
for (n = next_peer(dest_mnt); n != dest_mnt; n = next_peer(n)) {
ret = propagate_one(n);
if (ret)
goto out;
}
/* all slave groups */
for (m = next_group(dest_mnt, dest_mnt); m;
m = next_group(m, dest_mnt)) {
/* everything in that slave group */
n = m;
do {
ret = propagate_one(n);
if (ret)
goto out;
n = next_peer(n);
} while (n != m);
}
out:
read_seqlock_excl(&mount_lock);
hlist_for_each_entry(n, tree_list, mnt_hash) {
m = n->mnt_parent;
if (m->mnt_master != dest_mnt->mnt_master)
CLEAR_MNT_MARK(m->mnt_master);
}
read_sequnlock_excl(&mount_lock);
return ret;
}
/*
* return true if the refcount is greater than count
*/
static inline int do_refcount_check(struct mount *mnt, int count)
{
return mnt_get_count(mnt) > count;
}
/*
* check if the mount 'mnt' can be unmounted successfully.
* @mnt: the mount to be checked for unmount
* NOTE: unmounting 'mnt' would naturally propagate to all
* other mounts its parent propagates to.
* Check if any of these mounts that **do not have submounts**
* have more references than 'refcnt'. If so return busy.
*
* vfsmount lock must be held for write
*/
int propagate_mount_busy(struct mount *mnt, int refcnt)
{
struct mount *m, *child;
struct mount *parent = mnt->mnt_parent;
int ret = 0;
if (mnt == parent)
return do_refcount_check(mnt, refcnt);
/*
* quickly check if the current mount can be unmounted.
* If not, we don't have to go checking for all other
* mounts
*/
if (!list_empty(&mnt->mnt_mounts) || do_refcount_check(mnt, refcnt))
return 1;
for (m = propagation_next(parent, parent); m;
m = propagation_next(m, parent)) {
child = __lookup_mnt_last(&m->mnt, mnt->mnt_mountpoint);
if (child && list_empty(&child->mnt_mounts) &&
(ret = do_refcount_check(child, 1)))
break;
}
return ret;
}
/*
* Clear MNT_LOCKED when it can be shown to be safe.
*
* mount_lock lock must be held for write
*/
void propagate_mount_unlock(struct mount *mnt)
{
struct mount *parent = mnt->mnt_parent;
struct mount *m, *child;
BUG_ON(parent == mnt);
for (m = propagation_next(parent, parent); m;
m = propagation_next(m, parent)) {
child = __lookup_mnt_last(&m->mnt, mnt->mnt_mountpoint);
if (child)
child->mnt.mnt_flags &= ~MNT_LOCKED;
}
}
/*
* Mark all mounts that the MNT_LOCKED logic will allow to be unmounted.
*/
static void mark_umount_candidates(struct mount *mnt)
{
struct mount *parent = mnt->mnt_parent;
struct mount *m;
BUG_ON(parent == mnt);
for (m = propagation_next(parent, parent); m;
m = propagation_next(m, parent)) {
struct mount *child = __lookup_mnt_last(&m->mnt,
mnt->mnt_mountpoint);
if (child && (!IS_MNT_LOCKED(child) || IS_MNT_MARKED(m))) {
SET_MNT_MARK(child);
}
}
}
/*
* NOTE: unmounting 'mnt' naturally propagates to all other mounts its
* parent propagates to.
*/
static void __propagate_umount(struct mount *mnt)
{
struct mount *parent = mnt->mnt_parent;
struct mount *m;
BUG_ON(parent == mnt);
for (m = propagation_next(parent, parent); m;
m = propagation_next(m, parent)) {
struct mount *child = __lookup_mnt_last(&m->mnt,
mnt->mnt_mountpoint);
/*
* umount the child only if the child has no children
* and the child is marked safe to unmount.
*/
if (!child || !IS_MNT_MARKED(child))
continue;
CLEAR_MNT_MARK(child);
if (list_empty(&child->mnt_mounts)) {
list_del_init(&child->mnt_child);
child->mnt.mnt_flags |= MNT_UMOUNT;
list_move_tail(&child->mnt_list, &mnt->mnt_list);
}
}
}
/*
* collect all mounts that receive propagation from the mount in @list,
* and return these additional mounts in the same list.
* @list: the list of mounts to be unmounted.
*
* vfsmount lock must be held for write
*/
int propagate_umount(struct list_head *list)
{
struct mount *mnt;
list_for_each_entry_reverse(mnt, list, mnt_list)
mark_umount_candidates(mnt);
list_for_each_entry(mnt, list, mnt_list)
__propagate_umount(mnt);
return 0;
}
Computing file changes ...
