2011-11-22 06:10:51 +01:00
|
|
|
/*
|
|
|
|
* net/core/netprio_cgroup.c Priority Control Group
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version
|
|
|
|
* 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* Authors: Neil Horman <nhorman@tuxdriver.com>
|
|
|
|
*/
|
|
|
|
|
2012-05-16 21:58:40 +02:00
|
|
|
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
|
|
|
|
|
2011-11-22 06:10:51 +01:00
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/skbuff.h>
|
|
|
|
#include <linux/cgroup.h>
|
|
|
|
#include <linux/rcupdate.h>
|
|
|
|
#include <linux/atomic.h>
|
|
|
|
#include <net/rtnetlink.h>
|
|
|
|
#include <net/pkt_cls.h>
|
|
|
|
#include <net/sock.h>
|
|
|
|
#include <net/netprio_cgroup.h>
|
|
|
|
|
2012-07-20 12:39:25 +02:00
|
|
|
#include <linux/fdtable.h>
|
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
#define PRIOMAP_MIN_SZ 128
|
2011-11-22 06:10:51 +01:00
|
|
|
|
|
|
|
static inline struct cgroup_netprio_state *cgrp_netprio_state(struct cgroup *cgrp)
|
|
|
|
{
|
|
|
|
return container_of(cgroup_subsys_state(cgrp, net_prio_subsys_id),
|
|
|
|
struct cgroup_netprio_state, css);
|
|
|
|
}
|
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
/*
|
|
|
|
* Extend @dev->priomap so that it's large enough to accomodate
|
|
|
|
* @target_idx. @dev->priomap.priomap_len > @target_idx after successful
|
|
|
|
* return. Must be called under rtnl lock.
|
|
|
|
*/
|
|
|
|
static int extend_netdev_table(struct net_device *dev, u32 target_idx)
|
2011-11-22 06:10:51 +01:00
|
|
|
{
|
2012-11-22 16:32:46 +01:00
|
|
|
struct netprio_map *old, *new;
|
|
|
|
size_t new_sz, new_len;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
/* is the existing priomap large enough? */
|
2012-11-22 16:32:46 +01:00
|
|
|
old = rtnl_dereference(dev->priomap);
|
2012-11-22 16:32:46 +01:00
|
|
|
if (old && old->priomap_len > target_idx)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Determine the new size. Let's keep it power-of-two. We start
|
|
|
|
* from PRIOMAP_MIN_SZ and double it until it's large enough to
|
|
|
|
* accommodate @target_idx.
|
|
|
|
*/
|
|
|
|
new_sz = PRIOMAP_MIN_SZ;
|
|
|
|
while (true) {
|
|
|
|
new_len = (new_sz - offsetof(struct netprio_map, priomap)) /
|
|
|
|
sizeof(new->priomap[0]);
|
|
|
|
if (new_len > target_idx)
|
|
|
|
break;
|
|
|
|
new_sz *= 2;
|
|
|
|
/* overflowed? */
|
|
|
|
if (WARN_ON(new_sz < PRIOMAP_MIN_SZ))
|
|
|
|
return -ENOSPC;
|
|
|
|
}
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
/* allocate & copy */
|
|
|
|
new = kzalloc(new_sz, GFP_KERNEL);
|
2013-02-04 17:48:16 +01:00
|
|
|
if (!new)
|
net: cgroup: fix access the unallocated memory in netprio cgroup
there are some out of bound accesses in netprio cgroup.
now before accessing the dev->priomap.priomap array,we only check
if the dev->priomap exist.and because we don't want to see
additional bound checkings in fast path, so we should make sure
that dev->priomap is null or array size of dev->priomap.priomap
is equal to max_prioidx + 1;
so in write_priomap logic,we should call extend_netdev_table when
dev->priomap is null and dev->priomap.priomap_len < max_len.
and in cgrp_create->update_netdev_tables logic,we should call
extend_netdev_table only when dev->priomap exist and
dev->priomap.priomap_len < max_len.
and it's not needed to call update_netdev_tables in write_priomap,
we can only allocate the net device's priomap which we change through
net_prio.ifpriomap.
this patch also add a return value for update_netdev_tables &
extend_netdev_table, so when new_priomap is allocated failed,
write_priomap will stop to access the priomap,and return -ENOMEM
back to the userspace to tell the user what happend.
Change From v3:
1. add rtnl protect when reading max_prioidx in write_priomap.
2. only call extend_netdev_table when map->priomap_len < max_len,
this will make sure array size of dev->map->priomap always
bigger than any prioidx.
3. add a function write_update_netdev_table to make codes clear.
Change From v2:
1. protect extend_netdev_table by RTNL.
2. when extend_netdev_table failed,call dev_put to reduce device's refcount.
Signed-off-by: Gao feng <gaofeng@cn.fujitsu.com>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Eric Dumazet <edumazet@google.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 23:50:15 +02:00
|
|
|
return -ENOMEM;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
if (old)
|
|
|
|
memcpy(new->priomap, old->priomap,
|
|
|
|
old->priomap_len * sizeof(old->priomap[0]));
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
new->priomap_len = new_len;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
/* install the new priomap */
|
2012-11-22 16:32:46 +01:00
|
|
|
rcu_assign_pointer(dev->priomap, new);
|
|
|
|
if (old)
|
|
|
|
kfree_rcu(old, rcu);
|
net: cgroup: fix access the unallocated memory in netprio cgroup
there are some out of bound accesses in netprio cgroup.
now before accessing the dev->priomap.priomap array,we only check
if the dev->priomap exist.and because we don't want to see
additional bound checkings in fast path, so we should make sure
that dev->priomap is null or array size of dev->priomap.priomap
is equal to max_prioidx + 1;
so in write_priomap logic,we should call extend_netdev_table when
dev->priomap is null and dev->priomap.priomap_len < max_len.
and in cgrp_create->update_netdev_tables logic,we should call
extend_netdev_table only when dev->priomap exist and
dev->priomap.priomap_len < max_len.
and it's not needed to call update_netdev_tables in write_priomap,
we can only allocate the net device's priomap which we change through
net_prio.ifpriomap.
this patch also add a return value for update_netdev_tables &
extend_netdev_table, so when new_priomap is allocated failed,
write_priomap will stop to access the priomap,and return -ENOMEM
back to the userspace to tell the user what happend.
Change From v3:
1. add rtnl protect when reading max_prioidx in write_priomap.
2. only call extend_netdev_table when map->priomap_len < max_len,
this will make sure array size of dev->map->priomap always
bigger than any prioidx.
3. add a function write_update_netdev_table to make codes clear.
Change From v2:
1. protect extend_netdev_table by RTNL.
2. when extend_netdev_table failed,call dev_put to reduce device's refcount.
Signed-off-by: Gao feng <gaofeng@cn.fujitsu.com>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Eric Dumazet <edumazet@google.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 23:50:15 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-11-22 16:32:47 +01:00
|
|
|
/**
|
|
|
|
* netprio_prio - return the effective netprio of a cgroup-net_device pair
|
|
|
|
* @cgrp: cgroup part of the target pair
|
|
|
|
* @dev: net_device part of the target pair
|
|
|
|
*
|
|
|
|
* Should be called under RCU read or rtnl lock.
|
|
|
|
*/
|
|
|
|
static u32 netprio_prio(struct cgroup *cgrp, struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct netprio_map *map = rcu_dereference_rtnl(dev->priomap);
|
|
|
|
|
|
|
|
if (map && cgrp->id < map->priomap_len)
|
|
|
|
return map->priomap[cgrp->id];
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* netprio_set_prio - set netprio on a cgroup-net_device pair
|
|
|
|
* @cgrp: cgroup part of the target pair
|
|
|
|
* @dev: net_device part of the target pair
|
|
|
|
* @prio: prio to set
|
|
|
|
*
|
|
|
|
* Set netprio to @prio on @cgrp-@dev pair. Should be called under rtnl
|
|
|
|
* lock and may fail under memory pressure for non-zero @prio.
|
|
|
|
*/
|
|
|
|
static int netprio_set_prio(struct cgroup *cgrp, struct net_device *dev,
|
|
|
|
u32 prio)
|
|
|
|
{
|
|
|
|
struct netprio_map *map;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
/* avoid extending priomap for zero writes */
|
|
|
|
map = rtnl_dereference(dev->priomap);
|
|
|
|
if (!prio && (!map || map->priomap_len <= cgrp->id))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
ret = extend_netdev_table(dev, cgrp->id);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
map = rtnl_dereference(dev->priomap);
|
|
|
|
map->priomap[cgrp->id] = prio;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-11-19 17:13:38 +01:00
|
|
|
static struct cgroup_subsys_state *cgrp_css_alloc(struct cgroup *cgrp)
|
2011-11-22 06:10:51 +01:00
|
|
|
{
|
|
|
|
struct cgroup_netprio_state *cs;
|
2012-11-22 16:32:47 +01:00
|
|
|
|
2011-11-22 06:10:51 +01:00
|
|
|
cs = kzalloc(sizeof(*cs), GFP_KERNEL);
|
|
|
|
if (!cs)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
|
|
|
return &cs->css;
|
|
|
|
}
|
|
|
|
|
2012-11-22 16:32:47 +01:00
|
|
|
static int cgrp_css_online(struct cgroup *cgrp)
|
2011-11-22 06:10:51 +01:00
|
|
|
{
|
2012-11-22 16:32:47 +01:00
|
|
|
struct cgroup *parent = cgrp->parent;
|
2011-11-22 06:10:51 +01:00
|
|
|
struct net_device *dev;
|
2012-11-22 16:32:47 +01:00
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
if (!parent)
|
|
|
|
return 0;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
|
|
|
rtnl_lock();
|
2012-11-22 16:32:47 +01:00
|
|
|
/*
|
|
|
|
* Inherit prios from the parent. As all prios are set during
|
|
|
|
* onlining, there is no need to clear them on offline.
|
|
|
|
*/
|
|
|
|
for_each_netdev(&init_net, dev) {
|
|
|
|
u32 prio = netprio_prio(parent, dev);
|
|
|
|
|
|
|
|
ret = netprio_set_prio(cgrp, dev, prio);
|
|
|
|
if (ret)
|
|
|
|
break;
|
|
|
|
}
|
2011-11-22 06:10:51 +01:00
|
|
|
rtnl_unlock();
|
2012-11-22 16:32:47 +01:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cgrp_css_free(struct cgroup *cgrp)
|
|
|
|
{
|
|
|
|
kfree(cgrp_netprio_state(cgrp));
|
2011-11-22 06:10:51 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static u64 read_prioidx(struct cgroup *cgrp, struct cftype *cft)
|
|
|
|
{
|
2012-11-22 16:32:47 +01:00
|
|
|
return cgrp->id;
|
2011-11-22 06:10:51 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static int read_priomap(struct cgroup *cont, struct cftype *cft,
|
|
|
|
struct cgroup_map_cb *cb)
|
|
|
|
{
|
|
|
|
struct net_device *dev;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
2012-11-22 16:32:47 +01:00
|
|
|
for_each_netdev_rcu(&init_net, dev)
|
|
|
|
cb->fill(cb, dev->name, netprio_prio(cont, dev));
|
2011-11-22 06:10:51 +01:00
|
|
|
rcu_read_unlock();
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int write_priomap(struct cgroup *cgrp, struct cftype *cft,
|
|
|
|
const char *buffer)
|
|
|
|
{
|
2012-11-22 16:32:46 +01:00
|
|
|
char devname[IFNAMSIZ + 1];
|
2011-11-22 06:10:51 +01:00
|
|
|
struct net_device *dev;
|
2012-11-22 16:32:46 +01:00
|
|
|
u32 prio;
|
|
|
|
int ret;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-11-22 16:32:46 +01:00
|
|
|
if (sscanf(buffer, "%"__stringify(IFNAMSIZ)"s %u", devname, &prio) != 2)
|
|
|
|
return -EINVAL;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
|
|
|
dev = dev_get_by_name(&init_net, devname);
|
|
|
|
if (!dev)
|
2012-11-22 16:32:46 +01:00
|
|
|
return -ENODEV;
|
2011-11-22 06:10:51 +01:00
|
|
|
|
2012-08-14 14:34:35 +02:00
|
|
|
rtnl_lock();
|
2012-11-22 16:32:46 +01:00
|
|
|
|
2012-11-22 16:32:47 +01:00
|
|
|
ret = netprio_set_prio(cgrp, dev, prio);
|
net: cgroup: fix access the unallocated memory in netprio cgroup
there are some out of bound accesses in netprio cgroup.
now before accessing the dev->priomap.priomap array,we only check
if the dev->priomap exist.and because we don't want to see
additional bound checkings in fast path, so we should make sure
that dev->priomap is null or array size of dev->priomap.priomap
is equal to max_prioidx + 1;
so in write_priomap logic,we should call extend_netdev_table when
dev->priomap is null and dev->priomap.priomap_len < max_len.
and in cgrp_create->update_netdev_tables logic,we should call
extend_netdev_table only when dev->priomap exist and
dev->priomap.priomap_len < max_len.
and it's not needed to call update_netdev_tables in write_priomap,
we can only allocate the net device's priomap which we change through
net_prio.ifpriomap.
this patch also add a return value for update_netdev_tables &
extend_netdev_table, so when new_priomap is allocated failed,
write_priomap will stop to access the priomap,and return -ENOMEM
back to the userspace to tell the user what happend.
Change From v3:
1. add rtnl protect when reading max_prioidx in write_priomap.
2. only call extend_netdev_table when map->priomap_len < max_len,
this will make sure array size of dev->map->priomap always
bigger than any prioidx.
3. add a function write_update_netdev_table to make codes clear.
Change From v2:
1. protect extend_netdev_table by RTNL.
2. when extend_netdev_table failed,call dev_put to reduce device's refcount.
Signed-off-by: Gao feng <gaofeng@cn.fujitsu.com>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Eric Dumazet <edumazet@google.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 23:50:15 +02:00
|
|
|
|
2012-08-14 14:34:35 +02:00
|
|
|
rtnl_unlock();
|
2011-11-22 06:10:51 +01:00
|
|
|
dev_put(dev);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2012-08-22 04:32:06 +02:00
|
|
|
static int update_netprio(const void *v, struct file *file, unsigned n)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
struct socket *sock = sock_from_file(file, &err);
|
|
|
|
if (sock)
|
|
|
|
sock->sk->sk_cgrp_prioidx = (u32)(unsigned long)v;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-10-25 06:16:55 +02:00
|
|
|
static void net_prio_attach(struct cgroup *cgrp, struct cgroup_taskset *tset)
|
2012-07-20 12:39:25 +02:00
|
|
|
{
|
|
|
|
struct task_struct *p;
|
2012-08-22 04:32:06 +02:00
|
|
|
void *v;
|
2012-07-20 12:39:25 +02:00
|
|
|
|
|
|
|
cgroup_taskset_for_each(p, cgrp, tset) {
|
|
|
|
task_lock(p);
|
2012-08-22 04:32:06 +02:00
|
|
|
v = (void *)(unsigned long)task_netprioidx(p);
|
|
|
|
iterate_fd(p->files, 0, update_netprio, v);
|
2012-07-20 12:39:25 +02:00
|
|
|
task_unlock(p);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-11-22 06:10:51 +01:00
|
|
|
static struct cftype ss_files[] = {
|
|
|
|
{
|
|
|
|
.name = "prioidx",
|
|
|
|
.read_u64 = read_prioidx,
|
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "ifpriomap",
|
|
|
|
.read_map = read_priomap,
|
|
|
|
.write_string = write_priomap,
|
|
|
|
},
|
2012-04-01 21:09:55 +02:00
|
|
|
{ } /* terminate */
|
2011-11-22 06:10:51 +01:00
|
|
|
};
|
|
|
|
|
2012-04-01 21:09:55 +02:00
|
|
|
struct cgroup_subsys net_prio_subsys = {
|
|
|
|
.name = "net_prio",
|
2012-11-19 17:13:38 +01:00
|
|
|
.css_alloc = cgrp_css_alloc,
|
2012-11-22 16:32:47 +01:00
|
|
|
.css_online = cgrp_css_online,
|
2012-11-19 17:13:38 +01:00
|
|
|
.css_free = cgrp_css_free,
|
2012-07-20 12:39:25 +02:00
|
|
|
.attach = net_prio_attach,
|
2012-04-01 21:09:55 +02:00
|
|
|
.subsys_id = net_prio_subsys_id,
|
2012-04-01 21:09:55 +02:00
|
|
|
.base_cftypes = ss_files,
|
2012-09-13 21:20:58 +02:00
|
|
|
.module = THIS_MODULE,
|
2012-04-01 21:09:55 +02:00
|
|
|
};
|
2011-11-22 06:10:51 +01:00
|
|
|
|
|
|
|
static int netprio_device_event(struct notifier_block *unused,
|
|
|
|
unsigned long event, void *ptr)
|
|
|
|
{
|
2013-05-28 03:30:21 +02:00
|
|
|
struct net_device *dev = netdev_notifier_info_to_dev(ptr);
|
2011-11-22 06:10:51 +01:00
|
|
|
struct netprio_map *old;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Note this is called with rtnl_lock held so we have update side
|
|
|
|
* protection on our rcu assignments
|
|
|
|
*/
|
|
|
|
|
|
|
|
switch (event) {
|
|
|
|
case NETDEV_UNREGISTER:
|
|
|
|
old = rtnl_dereference(dev->priomap);
|
2011-11-23 08:09:32 +01:00
|
|
|
RCU_INIT_POINTER(dev->priomap, NULL);
|
2011-11-22 06:10:51 +01:00
|
|
|
if (old)
|
|
|
|
kfree_rcu(old, rcu);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return NOTIFY_DONE;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct notifier_block netprio_device_notifier = {
|
|
|
|
.notifier_call = netprio_device_event
|
|
|
|
};
|
|
|
|
|
|
|
|
static int __init init_cgroup_netprio(void)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = cgroup_load_subsys(&net_prio_subsys);
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
register_netdevice_notifier(&netprio_device_notifier);
|
|
|
|
|
|
|
|
out:
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __exit exit_cgroup_netprio(void)
|
|
|
|
{
|
|
|
|
struct netprio_map *old;
|
|
|
|
struct net_device *dev;
|
|
|
|
|
|
|
|
unregister_netdevice_notifier(&netprio_device_notifier);
|
|
|
|
|
|
|
|
cgroup_unload_subsys(&net_prio_subsys);
|
|
|
|
|
|
|
|
rtnl_lock();
|
|
|
|
for_each_netdev(&init_net, dev) {
|
|
|
|
old = rtnl_dereference(dev->priomap);
|
2011-11-23 08:09:32 +01:00
|
|
|
RCU_INIT_POINTER(dev->priomap, NULL);
|
2011-11-22 06:10:51 +01:00
|
|
|
if (old)
|
|
|
|
kfree_rcu(old, rcu);
|
|
|
|
}
|
|
|
|
rtnl_unlock();
|
|
|
|
}
|
|
|
|
|
|
|
|
module_init(init_cgroup_netprio);
|
|
|
|
module_exit(exit_cgroup_netprio);
|
|
|
|
MODULE_LICENSE("GPL v2");
|