Extend rate estimator new and replace APIs with additional spinlock
parameter used by lockless actions to protect rate_est pointer from
concurrent modification.
Signed-off-by: Vlad Buslov <vladbu@xxxxxxxxxxxx>
---
include/net/gen_stats.h | 2 ++
net/core/gen_estimator.c | 58 +++++++++++++++++++++++++++++++++++-----------
net/netfilter/xt_RATEEST.c | 2 +-
net/sched/act_api.c | 2 +-
net/sched/act_police.c | 2 +-
net/sched/sch_api.c | 2 ++
net/sched/sch_cbq.c | 4 ++--
net/sched/sch_drr.c | 4 ++--
net/sched/sch_hfsc.c | 4 ++--
net/sched/sch_htb.c | 4 ++--
net/sched/sch_qfq.c | 4 ++--
11 files changed, 61 insertions(+), 27 deletions(-)
diff --git a/include/net/gen_stats.h b/include/net/gen_stats.h
index 0304ba2..d1ef63d 100644
--- a/include/net/gen_stats.h
+++ b/include/net/gen_stats.h
@@ -59,12 +59,14 @@ int gnet_stats_finish_copy(struct gnet_dump *d);
int gen_new_estimator(struct gnet_stats_basic_packed *bstats,
struct gnet_stats_basic_cpu __percpu *cpu_bstats,
struct net_rate_estimator __rcu **rate_est,
+ spinlock_t *rate_est_lock,
spinlock_t *stats_lock,
seqcount_t *running, struct nlattr *opt);
void gen_kill_estimator(struct net_rate_estimator __rcu **ptr);
int gen_replace_estimator(struct gnet_stats_basic_packed *bstats,
struct gnet_stats_basic_cpu __percpu *cpu_bstats,
struct net_rate_estimator __rcu **ptr,
+ spinlock_t *rate_est_lock,
spinlock_t *stats_lock,
seqcount_t *running, struct nlattr *opt);
bool gen_estimator_active(struct net_rate_estimator __rcu **ptr);
diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c
index 98fd127..3512720 100644
--- a/net/core/gen_estimator.c
+++ b/net/core/gen_estimator.c
@@ -107,11 +107,43 @@ static void est_timer(struct timer_list *t)
mod_timer(&est->timer, est->next_jiffies);
}
+static void __replace_estimator(struct net_rate_estimator __rcu **rate_est,
+ struct net_rate_estimator *new)
+{
+ struct net_rate_estimator *old = rcu_dereference_protected(*rate_est,
+ 1);
+
+ if (old) {
+ del_timer_sync(&old->timer);
+ new->avbps = old->avbps;
+ new->avpps = old->avpps;
+ }
+
+ new->next_jiffies = jiffies + ((HZ/4) << new->intvl_log);
+ timer_setup(&new->timer, est_timer, 0);
+ mod_timer(&new->timer, new->next_jiffies);
+
+ rcu_assign_pointer(*rate_est, new);
+
+ if (old)
+ kfree_rcu(old, rcu);
+}
+
+static void replace_estimator(struct net_rate_estimator __rcu **rate_est,
+ struct net_rate_estimator *new,
+ spinlock_t *rate_est_lock)
+{
+ spin_lock(rate_est_lock);
+ __replace_estimator(rate_est, new);
+ spin_unlock(rate_est_lock);
+}
+
/**
* gen_new_estimator - create a new rate estimator
* @bstats: basic statistics
* @cpu_bstats: bstats per cpu
* @rate_est: rate estimator statistics
+ * @rate_est_lock: rate_est lock (might be NULL)
* @stats_lock: statistics lock
* @running: qdisc running seqcount
* @opt: rate estimator configuration TLV
@@ -128,12 +160,13 @@ static void est_timer(struct timer_list *t)
int gen_new_estimator(struct gnet_stats_basic_packed *bstats,
struct gnet_stats_basic_cpu __percpu *cpu_bstats,
struct net_rate_estimator __rcu **rate_est,
+ spinlock_t *rate_est_lock,
spinlock_t *stats_lock,
seqcount_t *running,
struct nlattr *opt)
{
struct gnet_estimator *parm = nla_data(opt);
- struct net_rate_estimator *old, *est;
+ struct net_rate_estimator *est;
struct gnet_stats_basic_packed b;
int intvl_log;
@@ -167,20 +200,15 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats,
local_bh_enable();
est->last_bytes = b.bytes;
est->last_packets = b.packets;
- old = rcu_dereference_protected(*rate_est, 1);
- if (old) {
- del_timer_sync(&old->timer);
- est->avbps = old->avbps;
- est->avpps = old->avpps;
- }
- est->next_jiffies = jiffies + ((HZ/4) << intvl_log);
- timer_setup(&est->timer, est_timer, 0);
- mod_timer(&est->timer, est->next_jiffies);
+ if (rate_est_lock)
+ replace_estimator(rate_est, est, rate_est_lock);
+ else
+ /* If no spinlock argument provided,
+ * then assume that caller is already synchronized.
+ */
+ __replace_estimator(rate_est, est);
- rcu_assign_pointer(*rate_est, est);
- if (old)
- kfree_rcu(old, rcu);
return 0;
}
EXPORT_SYMBOL(gen_new_estimator);
@@ -209,6 +237,7 @@ EXPORT_SYMBOL(gen_kill_estimator);
* @bstats: basic statistics
* @cpu_bstats: bstats per cpu
* @rate_est: rate estimator statistics
+ * @rate_est_lock: rate_est lock (might be NULL)
* @stats_lock: statistics lock
* @running: qdisc running seqcount (might be NULL)
* @opt: rate estimator configuration TLV
@@ -221,10 +250,11 @@ EXPORT_SYMBOL(gen_kill_estimator);
int gen_replace_estimator(struct gnet_stats_basic_packed *bstats,
struct gnet_stats_basic_cpu __percpu *cpu_bstats,
struct net_rate_estimator __rcu **rate_est,
+ spinlock_t *rate_est_lock,
spinlock_t *stats_lock,
seqcount_t *running, struct nlattr *opt)
{
- return gen_new_estimator(bstats, cpu_bstats, rate_est,
+ return gen_new_estimator(bstats, cpu_bstats, rate_est, rate_est_lock,
stats_lock, running, opt);
}
EXPORT_SYMBOL(gen_replace_estimator);
diff --git a/net/netfilter/xt_RATEEST.c b/net/netfilter/xt_RATEEST.c
index dec843c..8e79bd5 100644
--- a/net/netfilter/xt_RATEEST.c
+++ b/net/netfilter/xt_RATEEST.c
@@ -154,7 +154,7 @@ static int xt_rateest_tg_checkentry(const struct xt_tgchk_param *par)
cfg.est.interval = info->interval;
cfg.est.ewma_log = info->ewma_log;
- ret = gen_new_estimator(&est->bstats, NULL, &est->rate_est,
+ ret = gen_new_estimator(&est->bstats, NULL, &est->rate_est, NULL,
&est->lock, NULL, &cfg.opt);
if (ret < 0)
goto err2;
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index a5193dc..1dc092e 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -409,7 +409,7 @@ int tcf_idr_create(struct tc_action_net *tn, u32 index, struct nlattr *est,
p->tcfa_tm.firstuse = 0;
if (est) {
err = gen_new_estimator(&p->tcfa_bstats, p->cpu_bstats,
- &p->tcfa_rate_est,
+ &p->tcfa_rate_est, NULL,
&p->tcfa_lock, NULL, est);
if (err)
goto err4;
diff --git a/net/sched/act_police.c b/net/sched/act_police.c
index 86d9417..c480d68 100644
--- a/net/sched/act_police.c
+++ b/net/sched/act_police.c
@@ -133,7 +133,7 @@ static int tcf_act_police_init(struct net *net, struct nlattr *nla,
if (est) {
err = gen_replace_estimator(&police->tcf_bstats, NULL,
- &police->tcf_rate_est,
+ &police->tcf_rate_est, NULL,
&police->tcf_lock,
NULL, est);
if (err)
diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c
index 106dae7e..de6a297 100644
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -1178,6 +1178,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev,
sch->cpu_bstats,
&sch->rate_est,
NULL,
+ NULL,
running,
tca[TCA_RATE]);
if (err) {
@@ -1253,6 +1254,7 @@ static int qdisc_change(struct Qdisc *sch, struct nlattr **tca,
sch->cpu_bstats,
&sch->rate_est,
NULL,
+ NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
}
diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c
index f42025d..2a7ff53 100644
--- a/net/sched/sch_cbq.c
+++ b/net/sched/sch_cbq.c
@@ -1503,7 +1503,7 @@ cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **t
if (tca[TCA_RATE]) {
err = gen_replace_estimator(&cl->bstats, NULL,
- &cl->rate_est,
+ &cl->rate_est, NULL,
NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
@@ -1605,7 +1605,7 @@ cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **t
if (tca[TCA_RATE]) {
err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est,
- NULL,
+ NULL, NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
if (err) {
diff --git a/net/sched/sch_drr.c b/net/sched/sch_drr.c
index e0b0cf8..0896e23 100644
--- a/net/sched/sch_drr.c
+++ b/net/sched/sch_drr.c
@@ -95,7 +95,7 @@ static int drr_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
if (cl != NULL) {
if (tca[TCA_RATE]) {
err = gen_replace_estimator(&cl->bstats, NULL,
- &cl->rate_est,
+ &cl->rate_est, NULL,
NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
@@ -129,7 +129,7 @@ static int drr_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
if (tca[TCA_RATE]) {
err = gen_replace_estimator(&cl->bstats, NULL, &cl->rate_est,
- NULL,
+ NULL, NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
if (err) {
diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c
index 3ae9877..f341324 100644
--- a/net/sched/sch_hfsc.c
+++ b/net/sched/sch_hfsc.c
@@ -972,7 +972,7 @@ hfsc_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
if (tca[TCA_RATE]) {
err = gen_replace_estimator(&cl->bstats, NULL,
- &cl->rate_est,
+ &cl->rate_est, NULL,
NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
@@ -1042,7 +1042,7 @@ hfsc_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
if (tca[TCA_RATE]) {
err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est,
- NULL,
+ NULL, NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
if (err) {
diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c
index 2a4ab7c..acc0355 100644
--- a/net/sched/sch_htb.c
+++ b/net/sched/sch_htb.c
@@ -1407,7 +1407,7 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
if (htb_rate_est || tca[TCA_RATE]) {
err = gen_new_estimator(&cl->bstats, NULL,
&cl->rate_est,
- NULL,
+ NULL, NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE] ? : &est.nla);
if (err) {
@@ -1473,7 +1473,7 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
} else {
if (tca[TCA_RATE]) {
err = gen_replace_estimator(&cl->bstats, NULL,
- &cl->rate_est,
+ &cl->rate_est, NULL,
NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c
index bb1a9c1..8026c1e 100644
--- a/net/sched/sch_qfq.c
+++ b/net/sched/sch_qfq.c
@@ -461,7 +461,7 @@ static int qfq_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
if (cl != NULL) { /* modify existing class */
if (tca[TCA_RATE]) {
err = gen_replace_estimator(&cl->bstats, NULL,
- &cl->rate_est,
+ &cl->rate_est, NULL,
NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
@@ -488,7 +488,7 @@ static int qfq_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
if (tca[TCA_RATE]) {
err = gen_new_estimator(&cl->bstats, NULL,
&cl->rate_est,
- NULL,
+ NULL, NULL,
qdisc_root_sleeping_running(sch),
tca[TCA_RATE]);
if (err)
--
2.7.5
--
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
