* [PATCH net 2/2] qlcnic: Fix ethtool statistics length calculation
From: Shahed Shaikh @ 2014-01-09 17:41 UTC (permalink / raw)
To: davem; +Cc: netdev, Dept_NX_Linux_NIC_Driver, Shahed Shaikh
In-Reply-To: <1389289265-9586-1-git-send-email-shahed.shaikh@qlogic.com>
From: Shahed Shaikh <shahed.shaikh@qlogic.com>
o Consider number of Tx queues while calculating the length of
Tx statistics as part of ethtool stats.
o Calculate statistics lenght properly for 82xx and 83xx adapter
Signed-off-by: Shahed Shaikh <shahed.shaikh@qlogic.com>
---
.../net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c | 39 ++++++++++++----------
1 file changed, 21 insertions(+), 18 deletions(-)
diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c
index 7b36340..6b08194 100644
--- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c
+++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c
@@ -167,27 +167,35 @@ static const char qlcnic_gstrings_test[][ETH_GSTRING_LEN] = {
#define QLCNIC_TEST_LEN ARRAY_SIZE(qlcnic_gstrings_test)
-static inline int qlcnic_82xx_statistics(void)
+static inline int qlcnic_82xx_statistics(struct qlcnic_adapter *adapter)
{
- return ARRAY_SIZE(qlcnic_device_gstrings_stats) +
- ARRAY_SIZE(qlcnic_83xx_mac_stats_strings);
+ return ARRAY_SIZE(qlcnic_gstrings_stats) +
+ ARRAY_SIZE(qlcnic_83xx_mac_stats_strings) +
+ QLCNIC_TX_STATS_LEN * adapter->drv_tx_rings;
}
-static inline int qlcnic_83xx_statistics(void)
+static inline int qlcnic_83xx_statistics(struct qlcnic_adapter *adapter)
{
- return ARRAY_SIZE(qlcnic_83xx_tx_stats_strings) +
+ return ARRAY_SIZE(qlcnic_gstrings_stats) +
+ ARRAY_SIZE(qlcnic_83xx_tx_stats_strings) +
ARRAY_SIZE(qlcnic_83xx_mac_stats_strings) +
- ARRAY_SIZE(qlcnic_83xx_rx_stats_strings);
+ ARRAY_SIZE(qlcnic_83xx_rx_stats_strings) +
+ QLCNIC_TX_STATS_LEN * adapter->drv_tx_rings;
}
static int qlcnic_dev_statistics_len(struct qlcnic_adapter *adapter)
{
- if (qlcnic_82xx_check(adapter))
- return qlcnic_82xx_statistics();
- else if (qlcnic_83xx_check(adapter))
- return qlcnic_83xx_statistics();
- else
- return -1;
+ int len = -1;
+
+ if (qlcnic_82xx_check(adapter)) {
+ len = qlcnic_82xx_statistics(adapter);
+ if (adapter->flags & QLCNIC_ESWITCH_ENABLED)
+ len += ARRAY_SIZE(qlcnic_device_gstrings_stats);
+ } else if (qlcnic_83xx_check(adapter)) {
+ len = qlcnic_83xx_statistics(adapter);
+ }
+
+ return len;
}
#define QLCNIC_TX_INTR_NOT_CONFIGURED 0X78563412
@@ -920,18 +928,13 @@ static int qlcnic_eeprom_test(struct net_device *dev)
static int qlcnic_get_sset_count(struct net_device *dev, int sset)
{
- int len;
struct qlcnic_adapter *adapter = netdev_priv(dev);
switch (sset) {
case ETH_SS_TEST:
return QLCNIC_TEST_LEN;
case ETH_SS_STATS:
- len = qlcnic_dev_statistics_len(adapter) + QLCNIC_STATS_LEN;
- if ((adapter->flags & QLCNIC_ESWITCH_ENABLED) ||
- qlcnic_83xx_check(adapter))
- return len;
- return qlcnic_82xx_statistics();
+ return qlcnic_dev_statistics_len(adapter);
default:
return -EOPNOTSUPP;
}
--
1.8.1.4
^ permalink raw reply related
* Re: [net-next 2/7] ixgbe: Indicate removal state explicitly
From: Rustad, Mark D @ 2014-01-09 18:18 UTC (permalink / raw)
To: Scott Feldman
Cc: Brown, Aaron F, David Miller, Netdev, gospo@redhat.com,
sassmann@redhat.com
In-Reply-To: <2537C4E1-5434-4C6B-95A1-2360BA0803FB@intel.com>
[-- Attachment #1: Type: text/plain, Size: 1915 bytes --]
On Jan 9, 2014, at 9:27 AM, Rustad, Mark D <mark.d.rustad@intel.com> wrote:
> On Jan 8, 2014, at 12:35 AM, Scott Feldman <sfeldma@gmail.com> wrote:
>
>> On Jan 7, 2014, at 11:40 PM, Aaron Brown <aaron.f.brown@intel.com> wrote:
>>
>>> From: Mark Rustad <mark.d.rustad@intel.com>
>>>
>>> diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
>>> index 5e157ac..480c5c1 100644
>>> --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
>>> +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
>>> @@ -124,6 +124,11 @@ s32 ixgbe_reset_pipeline_82599(struct ixgbe_hw *hw);
>>> s32 ixgbe_get_thermal_sensor_data_generic(struct ixgbe_hw *hw);
>>> s32 ixgbe_init_thermal_sensor_thresh_generic(struct ixgbe_hw *hw);
>>>
>>> +#define IXGBE_FAILED_READ_REG 0xffffffffU
>>> +
>>> +void ixgbe_check_remove(struct ixgbe_hw *hw, u32 reg);
>>> +#define IXGBE_REMOVED(a) unlikely(!(a))
>>
>> IXGBE_REMOVED seems pretty closely tied to hw->hw_addr, but the macro turns any input into !input. Maybe an inline that takes a *hw?
>
> An earlier version had two different inputs, but not any more. I will change it to take a hw parameter. Good suggestion.
Actually, I responded too quickly. Although my comment on there previously having been two different inputs was true, that isn't the real reason for this form. The reason is that the check needs to be done on the value retrieved by ACCESS_ONCE - the address that will be used for the access. If the parameter were hw, then a separate dereference could occur, which is definitely not wanted.
One could argue that the macro should just be eliminated and the expansion put everywhere, but I feel that the code is much clearer with the macro. Unless there is another suggestion, I think this should stay as it is in this series.
--
Mark Rustad, Networking Division, Intel Corporation
[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 841 bytes --]
^ permalink raw reply
* [RFC Patch net-next 0/4] net_sched: make ingress filters lockless
From: Cong Wang @ 2014-01-09 18:19 UTC (permalink / raw)
To: netdev
Cc: Cong Wang, John Fastabend, Eric Dumazet, David S. Miller,
Jamal Hadi Salim
This patch tries to switch filter list to using struct
list_head, so that on the read side, the list can be traversed
with RCU read lock. Same for actions.
I hope either on egress or ingress classify is already protected by RCU
read lock, but I don't pretend I fully understanding qdisc locking.
Also, I am not sure I use RCU API's correctly at all. At least
I don't see any warning with CONFIG_PROVE_RCU=y.
Without this patch, the spin_lock easily appears on the top of
my perf top with 4 netperf sessions (4 is the number of CPU)
and with 1000 u32 filters on ingress.
Question 1:
Am I still missing something on qdisc locking?
Question 2:
Is it okay to call qdisc_bstats_update() without any lock?
Please comment.
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Cong Wang (4):
net_sched: switch filter list to list_head
net_sched: avoid holding qdisc lock for filters
net_sched: use RCU for tc actions traverse
net_sched: make ingress qdisc lockless
include/net/pkt_cls.h | 25 ++++--------------------
include/net/pkt_sched.h | 4 ++--
include/net/sch_generic.h | 14 +++++++++-----
net/core/dev.c | 2 --
net/sched/act_api.c | 12 +++++++-----
net/sched/cls_api.c | 48 ++++++++++++++++++++++++++++-------------------
net/sched/sch_api.c | 35 +++++++++++++++++++++-------------
net/sched/sch_atm.c | 14 +++++++-------
net/sched/sch_cbq.c | 9 +++++----
net/sched/sch_choke.c | 11 ++++++-----
net/sched/sch_drr.c | 7 ++++---
net/sched/sch_dsmark.c | 7 ++++---
net/sched/sch_fq_codel.c | 9 +++++----
net/sched/sch_hfsc.c | 15 +++++++++------
net/sched/sch_htb.c | 20 ++++++++++++--------
net/sched/sch_ingress.c | 14 +++++++++++---
net/sched/sch_multiq.c | 7 ++++---
net/sched/sch_prio.c | 9 +++++----
net/sched/sch_qfq.c | 7 ++++---
net/sched/sch_sfb.c | 9 +++++----
net/sched/sch_sfq.c | 11 ++++++-----
21 files changed, 160 insertions(+), 129 deletions(-)
--
1.8.3.1
^ permalink raw reply
* [RFC Patch net-next 1/4] net_sched: switch filter list to list_head
From: Cong Wang @ 2014-01-09 18:19 UTC (permalink / raw)
To: netdev
Cc: Cong Wang, John Fastabend, Eric Dumazet, David S. Miller,
Jamal Hadi Salim
In-Reply-To: <1389291593-2494-1-git-send-email-xiyou.wangcong@gmail.com>
So that it could be potentially traversed with RCU.
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
---
include/net/pkt_sched.h | 4 ++--
include/net/sch_generic.h | 9 ++++++---
net/sched/cls_api.c | 36 ++++++++++++++++++++++++------------
net/sched/sch_api.c | 35 ++++++++++++++++++++++-------------
net/sched/sch_atm.c | 14 +++++++-------
net/sched/sch_cbq.c | 9 +++++----
net/sched/sch_choke.c | 11 ++++++-----
net/sched/sch_drr.c | 7 ++++---
net/sched/sch_dsmark.c | 7 ++++---
net/sched/sch_fq_codel.c | 9 +++++----
net/sched/sch_hfsc.c | 15 +++++++++------
net/sched/sch_htb.c | 20 ++++++++++++--------
net/sched/sch_ingress.c | 14 +++++++++++---
net/sched/sch_multiq.c | 7 ++++---
net/sched/sch_prio.c | 9 +++++----
net/sched/sch_qfq.c | 7 ++++---
net/sched/sch_sfb.c | 9 +++++----
net/sched/sch_sfq.c | 11 ++++++-----
18 files changed, 141 insertions(+), 92 deletions(-)
diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h
index 891d80d..27a1efa 100644
--- a/include/net/pkt_sched.h
+++ b/include/net/pkt_sched.h
@@ -109,9 +109,9 @@ static inline void qdisc_run(struct Qdisc *q)
__qdisc_run(q);
}
-int tc_classify_compat(struct sk_buff *skb, const struct tcf_proto *tp,
+int tc_classify_compat(struct sk_buff *skb, const struct list_head *head,
struct tcf_result *res);
-int tc_classify(struct sk_buff *skb, const struct tcf_proto *tp,
+int tc_classify(struct sk_buff *skb, const struct list_head *head,
struct tcf_result *res);
/* Calculate maximal size of packet seen by hard_start_xmit
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h
index 013d96d..97123cc 100644
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -6,6 +6,7 @@
#include <linux/rcupdate.h>
#include <linux/pkt_sched.h>
#include <linux/pkt_cls.h>
+#include <linux/list.h>
#include <net/gen_stats.h>
#include <net/rtnetlink.h>
@@ -143,7 +144,7 @@ struct Qdisc_class_ops {
void (*walk)(struct Qdisc *, struct qdisc_walker * arg);
/* Filter manipulation */
- struct tcf_proto ** (*tcf_chain)(struct Qdisc *, unsigned long);
+ struct list_head * (*tcf_chain)(struct Qdisc *, unsigned long);
unsigned long (*bind_tcf)(struct Qdisc *, unsigned long,
u32 classid);
void (*unbind_tcf)(struct Qdisc *, unsigned long);
@@ -184,6 +185,8 @@ struct tcf_result {
u32 classid;
};
+struct tcf_proto;
+
struct tcf_proto_ops {
struct list_head head;
char kind[IFNAMSIZ];
@@ -212,7 +215,7 @@ struct tcf_proto_ops {
struct tcf_proto {
/* Fast access part */
- struct tcf_proto *next;
+ struct list_head head;
void *root;
int (*classify)(struct sk_buff *,
const struct tcf_proto *,
@@ -376,7 +379,7 @@ struct Qdisc *qdisc_create_dflt(struct netdev_queue *dev_queue,
void __qdisc_calculate_pkt_len(struct sk_buff *skb,
const struct qdisc_size_table *stab);
void tcf_destroy(struct tcf_proto *tp);
-void tcf_destroy_chain(struct tcf_proto **fl);
+void tcf_destroy_chain(struct list_head *fl);
/* Reset all TX qdiscs greater then index of a device. */
static inline void qdisc_reset_all_tx_gt(struct net_device *dev, unsigned int i)
diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c
index d8c42b1..bcc9987 100644
--- a/net/sched/cls_api.c
+++ b/net/sched/cls_api.c
@@ -125,8 +125,9 @@ static int tc_ctl_tfilter(struct sk_buff *skb, struct nlmsghdr *n)
u32 parent;
struct net_device *dev;
struct Qdisc *q;
- struct tcf_proto **back, **chain;
- struct tcf_proto *tp;
+ struct tcf_proto *back;
+ struct list_head *chain;
+ struct tcf_proto *tp, *res = NULL;
const struct tcf_proto_ops *tp_ops;
const struct Qdisc_class_ops *cops;
unsigned long cl;
@@ -196,21 +197,27 @@ replay:
goto errout;
/* Check the chain for existence of proto-tcf with this priority */
- for (back = chain; (tp = *back) != NULL; back = &tp->next) {
+ rcu_read_lock();
+ list_for_each_entry_rcu(tp, chain, head) {
+ back = list_next_entry(tp, head);
if (tp->prio >= prio) {
if (tp->prio == prio) {
if (!nprio ||
- (tp->protocol != protocol && protocol))
+ (tp->protocol != protocol && protocol)) {
+ rcu_read_unlock();
goto errout;
+ }
+ res = tp;
} else
- tp = NULL;
+ res = NULL;
break;
}
}
+ rcu_read_unlock();
root_lock = qdisc_root_sleeping_lock(q);
- if (tp == NULL) {
+ if ((tp = res) == NULL) {
/* Proto-tcf does not exist, create new one */
if (tca[TCA_KIND] == NULL || !protocol)
@@ -228,6 +235,7 @@ replay:
tp = kzalloc(sizeof(*tp), GFP_KERNEL);
if (tp == NULL)
goto errout;
+ INIT_LIST_HEAD(&tp->head);
err = -ENOENT;
tp_ops = tcf_proto_lookup_ops(tca[TCA_KIND]);
if (tp_ops == NULL) {
@@ -258,7 +266,7 @@ replay:
}
tp->ops = tp_ops;
tp->protocol = protocol;
- tp->prio = nprio ? : TC_H_MAJ(tcf_auto_prio(*back));
+ tp->prio = nprio ? : TC_H_MAJ(tcf_auto_prio(back));
tp->q = q;
tp->classify = tp_ops->classify;
tp->classid = parent;
@@ -280,7 +288,7 @@ replay:
if (fh == 0) {
if (n->nlmsg_type == RTM_DELTFILTER && t->tcm_handle == 0) {
spin_lock_bh(root_lock);
- *back = tp->next;
+ list_del_rcu(&tp->head);
spin_unlock_bh(root_lock);
tfilter_notify(net, skb, n, tp, fh, RTM_DELTFILTER);
@@ -321,8 +329,7 @@ replay:
if (err == 0) {
if (tp_created) {
spin_lock_bh(root_lock);
- tp->next = *back;
- *back = tp;
+ list_add_rcu(&tp->head, chain);
spin_unlock_bh(root_lock);
}
tfilter_notify(net, skb, n, tp, fh, RTM_NEWTFILTER);
@@ -417,7 +424,8 @@ static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb)
int s_t;
struct net_device *dev;
struct Qdisc *q;
- struct tcf_proto *tp, **chain;
+ struct tcf_proto *tp;
+ struct list_head *chain;
struct tcmsg *tcm = nlmsg_data(cb->nlh);
unsigned long cl = 0;
const struct Qdisc_class_ops *cops;
@@ -451,7 +459,9 @@ static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb)
s_t = cb->args[0];
- for (tp = *chain, t = 0; tp; tp = tp->next, t++) {
+ t = 0;
+ rcu_read_lock();
+ list_for_each_entry_rcu(tp, chain, head) {
if (t < s_t)
continue;
if (TC_H_MAJ(tcm->tcm_info) &&
@@ -482,7 +492,9 @@ static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb)
cb->args[1] = arg.w.count + 1;
if (arg.w.stop)
break;
+ t++;
}
+ rcu_read_unlock();
cb->args[0] = t;
diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c
index 1313145..df86686 100644
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -29,6 +29,7 @@
#include <linux/hrtimer.h>
#include <linux/lockdep.h>
#include <linux/slab.h>
+#include <linux/rculist.h>
#include <net/net_namespace.h>
#include <net/sock.h>
@@ -1772,13 +1773,15 @@ done:
* to this qdisc, (optionally) tests for protocol and asks
* specific classifiers.
*/
-int tc_classify_compat(struct sk_buff *skb, const struct tcf_proto *tp,
+int tc_classify_compat(struct sk_buff *skb, const struct list_head *head,
struct tcf_result *res)
{
+ struct tcf_proto *tp;
__be16 protocol = skb->protocol;
- int err;
+ int err = -1;
- for (; tp; tp = tp->next) {
+ WARN_ON_ONCE(!rcu_read_lock_held());
+ list_for_each_entry_rcu(tp, head, head) {
if (tp->protocol != protocol &&
tp->protocol != htons(ETH_P_ALL))
continue;
@@ -1789,23 +1792,25 @@ int tc_classify_compat(struct sk_buff *skb, const struct tcf_proto *tp,
if (err != TC_ACT_RECLASSIFY && skb->tc_verd)
skb->tc_verd = SET_TC_VERD(skb->tc_verd, 0);
#endif
- return err;
+ break;
}
}
- return -1;
+
+ return err;
}
EXPORT_SYMBOL(tc_classify_compat);
-int tc_classify(struct sk_buff *skb, const struct tcf_proto *tp,
+int tc_classify(struct sk_buff *skb, const struct list_head *head,
struct tcf_result *res)
{
int err = 0;
#ifdef CONFIG_NET_CLS_ACT
- const struct tcf_proto *otp = tp;
+ const struct tcf_proto *tp;
+ const struct tcf_proto *otp = list_first_entry(head, struct tcf_proto, head);
reclassify:
#endif
- err = tc_classify_compat(skb, tp, res);
+ err = tc_classify_compat(skb, head, res);
#ifdef CONFIG_NET_CLS_ACT
if (err == TC_ACT_RECLASSIFY) {
u32 verd = G_TC_VERD(skb->tc_verd);
@@ -1830,16 +1835,20 @@ void tcf_destroy(struct tcf_proto *tp)
{
tp->ops->destroy(tp);
module_put(tp->ops->owner);
+ synchronize_rcu();
kfree(tp);
}
-void tcf_destroy_chain(struct tcf_proto **fl)
+void tcf_destroy_chain(struct list_head *fl)
{
- struct tcf_proto *tp;
+ struct tcf_proto *tp, *n;
+ LIST_HEAD(list);
+ list_splice_init_rcu(fl, &list, synchronize_rcu);
- while ((tp = *fl) != NULL) {
- *fl = tp->next;
- tcf_destroy(tp);
+ list_for_each_entry_safe(tp, n, &list, head) {
+ tp->ops->destroy(tp);
+ module_put(tp->ops->owner);
+ kfree(tp);
}
}
EXPORT_SYMBOL(tcf_destroy_chain);
diff --git a/net/sched/sch_atm.c b/net/sched/sch_atm.c
index 1f9c314..20a07c2 100644
--- a/net/sched/sch_atm.c
+++ b/net/sched/sch_atm.c
@@ -41,7 +41,7 @@
struct atm_flow_data {
struct Qdisc *q; /* FIFO, TBF, etc. */
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
struct atm_vcc *vcc; /* VCC; NULL if VCC is closed */
void (*old_pop)(struct atm_vcc *vcc,
struct sk_buff *skb); /* chaining */
@@ -273,7 +273,7 @@ static int atm_tc_change(struct Qdisc *sch, u32 classid, u32 parent,
error = -ENOBUFS;
goto err_out;
}
- flow->filter_list = NULL;
+ INIT_LIST_HEAD(&flow->filter_list);
flow->q = qdisc_create_dflt(sch->dev_queue, &pfifo_qdisc_ops, classid);
if (!flow->q)
flow->q = &noop_qdisc;
@@ -311,7 +311,7 @@ static int atm_tc_delete(struct Qdisc *sch, unsigned long arg)
pr_debug("atm_tc_delete(sch %p,[qdisc %p],flow %p)\n", sch, p, flow);
if (list_empty(&flow->list))
return -EINVAL;
- if (flow->filter_list || flow == &p->link)
+ if (!list_empty(&flow->filter_list) || flow == &p->link)
return -EBUSY;
/*
* Reference count must be 2: one for "keepalive" (set at class
@@ -345,7 +345,7 @@ static void atm_tc_walk(struct Qdisc *sch, struct qdisc_walker *walker)
}
}
-static struct tcf_proto **atm_tc_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *atm_tc_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct atm_qdisc_data *p = qdisc_priv(sch);
struct atm_flow_data *flow = (struct atm_flow_data *)cl;
@@ -370,9 +370,9 @@ static int atm_tc_enqueue(struct sk_buff *skb, struct Qdisc *sch)
if (TC_H_MAJ(skb->priority) != sch->handle ||
!(flow = (struct atm_flow_data *)atm_tc_get(sch, skb->priority))) {
list_for_each_entry(flow, &p->flows, list) {
- if (flow->filter_list) {
+ if (!list_empty(&flow->filter_list)) {
result = tc_classify_compat(skb,
- flow->filter_list,
+ &flow->filter_list,
&res);
if (result < 0)
continue;
@@ -544,7 +544,7 @@ static int atm_tc_init(struct Qdisc *sch, struct nlattr *opt)
if (!p->link.q)
p->link.q = &noop_qdisc;
pr_debug("atm_tc_init: link (%p) qdisc %p\n", &p->link, p->link.q);
- p->link.filter_list = NULL;
+ INIT_LIST_HEAD(&p->link.filter_list);
p->link.vcc = NULL;
p->link.sock = NULL;
p->link.classid = sch->handle;
diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c
index 2f80d01..a5914ff 100644
--- a/net/sched/sch_cbq.c
+++ b/net/sched/sch_cbq.c
@@ -133,7 +133,7 @@ struct cbq_class {
struct gnet_stats_rate_est64 rate_est;
struct tc_cbq_xstats xstats;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
int refcnt;
int filters;
@@ -239,8 +239,8 @@ cbq_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
/*
* Step 2+n. Apply classifier.
*/
- if (!head->filter_list ||
- (result = tc_classify_compat(skb, head->filter_list, &res)) < 0)
+ if (list_empty(&head->filter_list) ||
+ (result = tc_classify_compat(skb, &head->filter_list, &res)) < 0)
goto fallback;
cl = (void *)res.class;
@@ -1881,6 +1881,7 @@ cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **t
}
}
+ INIT_LIST_HEAD(&cl->filter_list);
cl->R_tab = rtab;
rtab = NULL;
cl->refcnt = 1;
@@ -1976,7 +1977,7 @@ static int cbq_delete(struct Qdisc *sch, unsigned long arg)
return 0;
}
-static struct tcf_proto **cbq_find_tcf(struct Qdisc *sch, unsigned long arg)
+static struct list_head *cbq_find_tcf(struct Qdisc *sch, unsigned long arg)
{
struct cbq_sched_data *q = qdisc_priv(sch);
struct cbq_class *cl = (struct cbq_class *)arg;
diff --git a/net/sched/sch_choke.c b/net/sched/sch_choke.c
index ddd73cb..9b36830 100644
--- a/net/sched/sch_choke.c
+++ b/net/sched/sch_choke.c
@@ -58,7 +58,7 @@ struct choke_sched_data {
/* Variables */
struct red_vars vars;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
struct {
u32 prob_drop; /* Early probability drops */
u32 prob_mark; /* Early probability marks */
@@ -202,7 +202,7 @@ static bool choke_classify(struct sk_buff *skb,
struct tcf_result res;
int result;
- result = tc_classify(skb, q->filter_list, &res);
+ result = tc_classify(skb, &q->filter_list, &res);
if (result >= 0) {
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
@@ -256,7 +256,7 @@ static bool choke_match_random(const struct choke_sched_data *q,
return false;
oskb = choke_peek_random(q, pidx);
- if (q->filter_list)
+ if (!list_empty(&q->filter_list))
return choke_get_classid(nskb) == choke_get_classid(oskb);
return choke_match_flow(oskb, nskb);
@@ -268,7 +268,7 @@ static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch)
const struct red_parms *p = &q->parms;
int ret = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- if (q->filter_list) {
+ if (!list_empty(&q->filter_list)) {
/* If using external classifiers, get result and record it. */
if (!choke_classify(skb, sch, &ret))
goto other_drop; /* Packet was eaten by filter */
@@ -476,6 +476,7 @@ static int choke_change(struct Qdisc *sch, struct nlattr *opt)
q->flags = ctl->flags;
q->limit = ctl->limit;
+ INIT_LIST_HEAD(&q->filter_list);
red_set_parms(&q->parms, ctl->qth_min, ctl->qth_max, ctl->Wlog,
ctl->Plog, ctl->Scell_log,
@@ -566,7 +567,7 @@ static unsigned long choke_bind(struct Qdisc *sch, unsigned long parent,
return 0;
}
-static struct tcf_proto **choke_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *choke_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct choke_sched_data *q = qdisc_priv(sch);
diff --git a/net/sched/sch_drr.c b/net/sched/sch_drr.c
index 8302717..62f45ac 100644
--- a/net/sched/sch_drr.c
+++ b/net/sched/sch_drr.c
@@ -35,7 +35,7 @@ struct drr_class {
struct drr_sched {
struct list_head active;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
struct Qdisc_class_hash clhash;
};
@@ -184,7 +184,7 @@ static void drr_put_class(struct Qdisc *sch, unsigned long arg)
drr_destroy_class(sch, cl);
}
-static struct tcf_proto **drr_tcf_chain(struct Qdisc *sch, unsigned long cl)
+static struct list_head *drr_tcf_chain(struct Qdisc *sch, unsigned long cl)
{
struct drr_sched *q = qdisc_priv(sch);
@@ -328,7 +328,7 @@ static struct drr_class *drr_classify(struct sk_buff *skb, struct Qdisc *sch,
}
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- result = tc_classify(skb, q->filter_list, &res);
+ result = tc_classify(skb, &q->filter_list, &res);
if (result >= 0) {
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
@@ -443,6 +443,7 @@ static int drr_init_qdisc(struct Qdisc *sch, struct nlattr *opt)
if (err < 0)
return err;
INIT_LIST_HEAD(&q->active);
+ INIT_LIST_HEAD(&q->filter_list);
return 0;
}
diff --git a/net/sched/sch_dsmark.c b/net/sched/sch_dsmark.c
index 49d6ef3..4489ffe 100644
--- a/net/sched/sch_dsmark.c
+++ b/net/sched/sch_dsmark.c
@@ -37,7 +37,7 @@
struct dsmark_qdisc_data {
struct Qdisc *q;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
u8 *mask; /* "owns" the array */
u8 *value;
u16 indices;
@@ -186,7 +186,7 @@ ignore:
}
}
-static inline struct tcf_proto **dsmark_find_tcf(struct Qdisc *sch,
+static inline struct list_head *dsmark_find_tcf(struct Qdisc *sch,
unsigned long cl)
{
struct dsmark_qdisc_data *p = qdisc_priv(sch);
@@ -229,7 +229,7 @@ static int dsmark_enqueue(struct sk_buff *skb, struct Qdisc *sch)
skb->tc_index = TC_H_MIN(skb->priority);
else {
struct tcf_result res;
- int result = tc_classify(skb, p->filter_list, &res);
+ int result = tc_classify(skb, &p->filter_list, &res);
pr_debug("result %d class 0x%04x\n", result, res.classid);
@@ -380,6 +380,7 @@ static int dsmark_init(struct Qdisc *sch, struct nlattr *opt)
p->indices = indices;
p->default_index = default_index;
p->set_tc_index = nla_get_flag(tb[TCA_DSMARK_SET_TC_INDEX]);
+ INIT_LIST_HEAD(&p->filter_list);
p->q = qdisc_create_dflt(sch->dev_queue, &pfifo_qdisc_ops, sch->handle);
if (p->q == NULL)
diff --git a/net/sched/sch_fq_codel.c b/net/sched/sch_fq_codel.c
index 5578628..3c19917 100644
--- a/net/sched/sch_fq_codel.c
+++ b/net/sched/sch_fq_codel.c
@@ -52,7 +52,7 @@ struct fq_codel_flow {
}; /* please try to keep this structure <= 64 bytes */
struct fq_codel_sched_data {
- struct tcf_proto *filter_list; /* optional external classifier */
+ struct list_head filter_list; /* optional external classifier */
struct fq_codel_flow *flows; /* Flows table [flows_cnt] */
u32 *backlogs; /* backlog table [flows_cnt] */
u32 flows_cnt; /* number of flows */
@@ -92,11 +92,11 @@ static unsigned int fq_codel_classify(struct sk_buff *skb, struct Qdisc *sch,
TC_H_MIN(skb->priority) <= q->flows_cnt)
return TC_H_MIN(skb->priority);
- if (!q->filter_list)
+ if (list_emty(&q->filter_list))
return fq_codel_hash(q, skb) + 1;
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- result = tc_classify(skb, q->filter_list, &res);
+ result = tc_classify(skb, &q->filter_list, &res);
if (result >= 0) {
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
@@ -393,6 +393,7 @@ static int fq_codel_init(struct Qdisc *sch, struct nlattr *opt)
q->perturbation = net_random();
INIT_LIST_HEAD(&q->new_flows);
INIT_LIST_HEAD(&q->old_flows);
+ INIT_LIST_HEAD(&q->filter_list);
codel_params_init(&q->cparams);
codel_stats_init(&q->cstats);
q->cparams.ecn = true;
@@ -501,7 +502,7 @@ static void fq_codel_put(struct Qdisc *q, unsigned long cl)
{
}
-static struct tcf_proto **fq_codel_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *fq_codel_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct fq_codel_sched_data *q = qdisc_priv(sch);
diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c
index c407561..8bb4ade 100644
--- a/net/sched/sch_hfsc.c
+++ b/net/sched/sch_hfsc.c
@@ -116,7 +116,7 @@ struct hfsc_class {
struct gnet_stats_queue qstats;
struct gnet_stats_rate_est64 rate_est;
unsigned int level; /* class level in hierarchy */
- struct tcf_proto *filter_list; /* filter list */
+ struct list_head filter_list; /* filter list */
unsigned int filter_cnt; /* filter count */
struct hfsc_sched *sched; /* scheduler data */
@@ -1083,6 +1083,7 @@ hfsc_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
cl->refcnt = 1;
cl->sched = q;
cl->cl_parent = parent;
+ INIT_LIST_HEAD(&cl->filter_list);
cl->qdisc = qdisc_create_dflt(sch->dev_queue,
&pfifo_qdisc_ops, classid);
if (cl->qdisc == NULL)
@@ -1151,7 +1152,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
struct hfsc_sched *q = qdisc_priv(sch);
struct hfsc_class *head, *cl;
struct tcf_result res;
- struct tcf_proto *tcf;
+ struct list_head *list;
int result;
if (TC_H_MAJ(skb->priority ^ sch->handle) == 0 &&
@@ -1161,8 +1162,10 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
head = &q->root;
- tcf = q->root.filter_list;
- while (tcf && (result = tc_classify(skb, tcf, &res)) >= 0) {
+ list = &q->root.filter_list;
+ while (list) {
+ if ((result = tc_classify(skb, list, &res)) < 0)
+ break;
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
case TC_ACT_QUEUED:
@@ -1185,7 +1188,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
return cl; /* hit leaf class */
/* apply inner filter chain */
- tcf = cl->filter_list;
+ list = &cl->filter_list;
head = cl;
}
@@ -1285,7 +1288,7 @@ hfsc_unbind_tcf(struct Qdisc *sch, unsigned long arg)
cl->filter_cnt--;
}
-static struct tcf_proto **
+static struct list_head *
hfsc_tcf_chain(struct Qdisc *sch, unsigned long arg)
{
struct hfsc_sched *q = qdisc_priv(sch);
diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c
index 0db5a6e..c28e405 100644
--- a/net/sched/sch_htb.c
+++ b/net/sched/sch_htb.c
@@ -103,7 +103,7 @@ struct htb_class {
u32 prio; /* these two are used only by leaves... */
int quantum; /* but stored for parent-to-leaf return */
- struct tcf_proto *filter_list; /* class attached filters */
+ struct list_head filter_list; /* class attached filters */
int filter_cnt;
int refcnt; /* usage count of this class */
@@ -153,7 +153,7 @@ struct htb_sched {
int rate2quantum; /* quant = rate / rate2quantum */
/* filters for qdisc itself */
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
#define HTB_WARN_TOOMANYEVENTS 0x1
unsigned int warned; /* only one warning */
@@ -209,7 +209,7 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch,
struct htb_sched *q = qdisc_priv(sch);
struct htb_class *cl;
struct tcf_result res;
- struct tcf_proto *tcf;
+ struct list_head *head;
int result;
/* allow to select class by setting skb->priority to valid classid;
@@ -223,8 +223,10 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch,
return cl;
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- tcf = q->filter_list;
- while (tcf && (result = tc_classify(skb, tcf, &res)) >= 0) {
+ head = &q->filter_list;
+ while (!list_empty(head)) {
+ if ((result = tc_classify(skb, head, &res)) < 0)
+ break;
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
case TC_ACT_QUEUED:
@@ -246,7 +248,7 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch,
return cl; /* we hit leaf; return it */
/* we have got inner class; apply inner filter chain */
- tcf = cl->filter_list;
+ head = &cl->filter_list;
}
/* classification failed; try to use default class */
cl = htb_find(TC_H_MAKE(TC_H_MAJ(sch->handle), q->defcls), sch);
@@ -1040,6 +1042,7 @@ static int htb_init(struct Qdisc *sch, struct nlattr *opt)
qdisc_watchdog_init(&q->watchdog, sch);
INIT_WORK(&q->work, htb_work_func);
skb_queue_head_init(&q->direct_queue);
+ INIT_LIST_HEAD(&q->filter_list);
if (tb[TCA_HTB_DIRECT_QLEN])
q->direct_qlen = nla_get_u32(tb[TCA_HTB_DIRECT_QLEN]);
@@ -1413,6 +1416,7 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
cl->refcnt = 1;
cl->children = 0;
INIT_LIST_HEAD(&cl->un.leaf.drop_list);
+ INIT_LIST_HEAD(&cl->filter_list);
RB_CLEAR_NODE(&cl->pq_node);
for (prio = 0; prio < TC_HTB_NUMPRIO; prio++)
@@ -1518,11 +1522,11 @@ failure:
return err;
}
-static struct tcf_proto **htb_find_tcf(struct Qdisc *sch, unsigned long arg)
+static struct list_head *htb_find_tcf(struct Qdisc *sch, unsigned long arg)
{
struct htb_sched *q = qdisc_priv(sch);
struct htb_class *cl = (struct htb_class *)arg;
- struct tcf_proto **fl = cl ? &cl->filter_list : &q->filter_list;
+ struct list_head *fl = cl ? &cl->filter_list : &q->filter_list;
return fl;
}
diff --git a/net/sched/sch_ingress.c b/net/sched/sch_ingress.c
index bce1665..b182e0c 100644
--- a/net/sched/sch_ingress.c
+++ b/net/sched/sch_ingress.c
@@ -17,7 +17,7 @@
struct ingress_qdisc_data {
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
};
/* ------------------------- Class/flow operations ------------------------- */
@@ -46,7 +46,7 @@ static void ingress_walk(struct Qdisc *sch, struct qdisc_walker *walker)
{
}
-static struct tcf_proto **ingress_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *ingress_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct ingress_qdisc_data *p = qdisc_priv(sch);
@@ -61,7 +61,7 @@ static int ingress_enqueue(struct sk_buff *skb, struct Qdisc *sch)
struct tcf_result res;
int result;
- result = tc_classify(skb, p->filter_list, &res);
+ result = tc_classify(skb, &p->filter_list, &res);
qdisc_bstats_update(sch, skb);
switch (result) {
@@ -108,6 +108,13 @@ nla_put_failure:
return -1;
}
+static int ingress_init(struct Qdisc *sch, struct nlattr *opt)
+{
+ struct ingress_qdisc_data *q = qdisc_priv(sch);
+ INIT_LIST_HEAD(&q->filter_list);
+ return 0;
+}
+
static const struct Qdisc_class_ops ingress_class_ops = {
.leaf = ingress_leaf,
.get = ingress_get,
@@ -119,6 +126,7 @@ static const struct Qdisc_class_ops ingress_class_ops = {
};
static struct Qdisc_ops ingress_qdisc_ops __read_mostly = {
+ .init = ingress_init,
.cl_ops = &ingress_class_ops,
.id = "ingress",
.priv_size = sizeof(struct ingress_qdisc_data),
diff --git a/net/sched/sch_multiq.c b/net/sched/sch_multiq.c
index afb050a..e4be093 100644
--- a/net/sched/sch_multiq.c
+++ b/net/sched/sch_multiq.c
@@ -31,7 +31,7 @@ struct multiq_sched_data {
u16 bands;
u16 max_bands;
u16 curband;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
struct Qdisc **queues;
};
@@ -45,7 +45,7 @@ multiq_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
int err;
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- err = tc_classify(skb, q->filter_list, &res);
+ err = tc_classify(skb, &q->filter_list, &res);
#ifdef CONFIG_NET_CLS_ACT
switch (err) {
case TC_ACT_STOLEN:
@@ -258,6 +258,7 @@ static int multiq_init(struct Qdisc *sch, struct nlattr *opt)
if (opt == NULL)
return -EINVAL;
+ INIT_LIST_HEAD(&q->filter_list);
q->max_bands = qdisc_dev(sch)->num_tx_queues;
q->queues = kcalloc(q->max_bands, sizeof(struct Qdisc *), GFP_KERNEL);
@@ -388,7 +389,7 @@ static void multiq_walk(struct Qdisc *sch, struct qdisc_walker *arg)
}
}
-static struct tcf_proto **multiq_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *multiq_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct multiq_sched_data *q = qdisc_priv(sch);
diff --git a/net/sched/sch_prio.c b/net/sched/sch_prio.c
index 79359b6..f4ee09f 100644
--- a/net/sched/sch_prio.c
+++ b/net/sched/sch_prio.c
@@ -24,7 +24,7 @@
struct prio_sched_data {
int bands;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
u8 prio2band[TC_PRIO_MAX+1];
struct Qdisc *queues[TCQ_PRIO_BANDS];
};
@@ -40,7 +40,7 @@ prio_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
if (TC_H_MAJ(skb->priority) != sch->handle) {
- err = tc_classify(skb, q->filter_list, &res);
+ err = tc_classify(skb, &q->filter_list, &res);
#ifdef CONFIG_NET_CLS_ACT
switch (err) {
case TC_ACT_STOLEN:
@@ -50,7 +50,7 @@ prio_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
return NULL;
}
#endif
- if (!q->filter_list || err < 0) {
+ if (list_empty(&q->filter_list) || err < 0) {
if (TC_H_MAJ(band))
band = 0;
return q->queues[q->prio2band[band & TC_PRIO_MAX]];
@@ -235,6 +235,7 @@ static int prio_init(struct Qdisc *sch, struct nlattr *opt)
if ((err = prio_tune(sch, opt)) != 0)
return err;
}
+ INIT_LIST_HEAD(&q->filter_list);
return 0;
}
@@ -351,7 +352,7 @@ static void prio_walk(struct Qdisc *sch, struct qdisc_walker *arg)
}
}
-static struct tcf_proto **prio_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *prio_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct prio_sched_data *q = qdisc_priv(sch);
diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c
index 8056fb4..c43f85a 100644
--- a/net/sched/sch_qfq.c
+++ b/net/sched/sch_qfq.c
@@ -181,7 +181,7 @@ struct qfq_group {
};
struct qfq_sched {
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
struct Qdisc_class_hash clhash;
u64 oldV, V; /* Precise virtual times. */
@@ -576,7 +576,7 @@ static void qfq_put_class(struct Qdisc *sch, unsigned long arg)
qfq_destroy_class(sch, cl);
}
-static struct tcf_proto **qfq_tcf_chain(struct Qdisc *sch, unsigned long cl)
+static struct list_head *qfq_tcf_chain(struct Qdisc *sch, unsigned long cl)
{
struct qfq_sched *q = qdisc_priv(sch);
@@ -714,7 +714,7 @@ static struct qfq_class *qfq_classify(struct sk_buff *skb, struct Qdisc *sch,
}
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- result = tc_classify(skb, q->filter_list, &res);
+ result = tc_classify(skb, &q->filter_list, &res);
if (result >= 0) {
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
@@ -1498,6 +1498,7 @@ static int qfq_init_qdisc(struct Qdisc *sch, struct nlattr *opt)
}
INIT_HLIST_HEAD(&q->nonfull_aggs);
+ INIT_LIST_HEAD(&q->filter_list);
return 0;
}
diff --git a/net/sched/sch_sfb.c b/net/sched/sch_sfb.c
index 30ea467..20fcc6d 100644
--- a/net/sched/sch_sfb.c
+++ b/net/sched/sch_sfb.c
@@ -55,7 +55,7 @@ struct sfb_bins {
struct sfb_sched_data {
struct Qdisc *qdisc;
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
unsigned long rehash_interval;
unsigned long warmup_time; /* double buffering warmup time in jiffies */
u32 max;
@@ -259,7 +259,7 @@ static bool sfb_classify(struct sk_buff *skb, struct sfb_sched_data *q,
struct tcf_result res;
int result;
- result = tc_classify(skb, q->filter_list, &res);
+ result = tc_classify(skb, &q->filter_list, &res);
if (result >= 0) {
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
@@ -306,7 +306,7 @@ static int sfb_enqueue(struct sk_buff *skb, struct Qdisc *sch)
}
}
- if (q->filter_list) {
+ if (!list_empty(&q->filter_list)) {
/* If using external classifiers, get result and record it. */
if (!sfb_classify(skb, q, &ret, &salt))
goto other_drop;
@@ -533,6 +533,7 @@ static int sfb_change(struct Qdisc *sch, struct nlattr *opt)
q->tokens_avail = ctl->penalty_burst;
q->token_time = jiffies;
+ INIT_LIST_HEAD(&q->filter_list);
q->slot = 0;
q->double_buffering = false;
sfb_zero_all_buckets(q);
@@ -660,7 +661,7 @@ static void sfb_walk(struct Qdisc *sch, struct qdisc_walker *walker)
}
}
-static struct tcf_proto **sfb_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *sfb_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct sfb_sched_data *q = qdisc_priv(sch);
diff --git a/net/sched/sch_sfq.c b/net/sched/sch_sfq.c
index 76f01e0..2a5b2a4 100644
--- a/net/sched/sch_sfq.c
+++ b/net/sched/sch_sfq.c
@@ -125,7 +125,7 @@ struct sfq_sched_data {
u8 cur_depth; /* depth of longest slot */
u8 flags;
unsigned short scaled_quantum; /* SFQ_ALLOT_SIZE(quantum) */
- struct tcf_proto *filter_list;
+ struct list_head filter_list;
sfq_index *ht; /* Hash table ('divisor' slots) */
struct sfq_slot *slots; /* Flows table ('maxflows' entries) */
@@ -194,13 +194,13 @@ static unsigned int sfq_classify(struct sk_buff *skb, struct Qdisc *sch,
TC_H_MIN(skb->priority) <= q->divisor)
return TC_H_MIN(skb->priority);
- if (!q->filter_list) {
+ if (list_empty(&q->filter_list)) {
skb_flow_dissect(skb, &sfq_skb_cb(skb)->keys);
return sfq_hash(q, skb) + 1;
}
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
- result = tc_classify(skb, q->filter_list, &res);
+ result = tc_classify(skb, &q->filter_list, &res);
if (result >= 0) {
#ifdef CONFIG_NET_CLS_ACT
switch (result) {
@@ -630,7 +630,7 @@ static void sfq_perturbation(unsigned long arg)
spin_lock(root_lock);
q->perturbation = net_random();
- if (!q->filter_list && q->tail)
+ if (list_empty(&q->filter_list) && q->tail)
sfq_rehash(sch);
spin_unlock(root_lock);
@@ -760,6 +760,7 @@ static int sfq_init(struct Qdisc *sch, struct nlattr *opt)
q->scaled_quantum = SFQ_ALLOT_SIZE(q->quantum);
q->perturb_period = 0;
q->perturbation = net_random();
+ INIT_LIST_HEAD(&q->filter_list);
if (opt) {
int err = sfq_change(sch, opt);
@@ -846,7 +847,7 @@ static void sfq_put(struct Qdisc *q, unsigned long cl)
{
}
-static struct tcf_proto **sfq_find_tcf(struct Qdisc *sch, unsigned long cl)
+static struct list_head *sfq_find_tcf(struct Qdisc *sch, unsigned long cl)
{
struct sfq_sched_data *q = qdisc_priv(sch);
--
1.8.3.1
^ permalink raw reply related
* [RFC Patch net-next 2/4] net_sched: avoid holding qdisc lock for filters
From: Cong Wang @ 2014-01-09 18:19 UTC (permalink / raw)
To: netdev
Cc: Cong Wang, John Fastabend, Eric Dumazet, David S. Miller,
Jamal Hadi Salim
In-Reply-To: <1389291593-2494-1-git-send-email-xiyou.wangcong@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
---
include/net/pkt_cls.h | 25 ++++---------------------
include/net/sch_generic.h | 5 +++--
net/sched/cls_api.c | 12 +++++-------
3 files changed, 12 insertions(+), 30 deletions(-)
diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h
index 50ea079..9afb15b 100644
--- a/include/net/pkt_cls.h
+++ b/include/net/pkt_cls.h
@@ -18,26 +18,9 @@ int register_tcf_proto_ops(struct tcf_proto_ops *ops);
int unregister_tcf_proto_ops(struct tcf_proto_ops *ops);
static inline unsigned long
-__cls_set_class(unsigned long *clp, unsigned long cl)
+cls_set_class(unsigned long *clp, unsigned long cl)
{
- unsigned long old_cl;
-
- old_cl = *clp;
- *clp = cl;
- return old_cl;
-}
-
-static inline unsigned long
-cls_set_class(struct tcf_proto *tp, unsigned long *clp,
- unsigned long cl)
-{
- unsigned long old_cl;
-
- tcf_tree_lock(tp);
- old_cl = __cls_set_class(clp, cl);
- tcf_tree_unlock(tp);
-
- return old_cl;
+ return xchg(clp, cl);
}
static inline void
@@ -46,7 +29,7 @@ tcf_bind_filter(struct tcf_proto *tp, struct tcf_result *r, unsigned long base)
unsigned long cl;
cl = tp->q->ops->cl_ops->bind_tcf(tp->q, base, r->classid);
- cl = cls_set_class(tp, &r->class, cl);
+ cl = cls_set_class(&r->class, cl);
if (cl)
tp->q->ops->cl_ops->unbind_tcf(tp->q, cl);
}
@@ -56,7 +39,7 @@ tcf_unbind_filter(struct tcf_proto *tp, struct tcf_result *r)
{
unsigned long cl;
- if ((cl = __cls_set_class(&r->class, 0)) != 0)
+ if ((cl = cls_set_class(&r->class, 0)) != 0)
tp->q->ops->cl_ops->unbind_tcf(tp->q, cl);
}
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h
index 97123cc..000ce54 100644
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -215,6 +215,7 @@ struct tcf_proto_ops {
struct tcf_proto {
/* Fast access part */
+ spinlock_t lock;
struct list_head head;
void *root;
int (*classify)(struct sk_buff *,
@@ -312,8 +313,8 @@ static inline void sch_tree_unlock(const struct Qdisc *q)
spin_unlock_bh(qdisc_root_sleeping_lock(q));
}
-#define tcf_tree_lock(tp) sch_tree_lock((tp)->q)
-#define tcf_tree_unlock(tp) sch_tree_unlock((tp)->q)
+#define tcf_tree_lock(tp) spin_lock_bh(&(tp)->lock);
+#define tcf_tree_unlock(tp) spin_unlock_bh(&(tp)->lock);
extern struct Qdisc noop_qdisc;
extern struct Qdisc_ops noop_qdisc_ops;
diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c
index bcc9987..9ba2804 100644
--- a/net/sched/cls_api.c
+++ b/net/sched/cls_api.c
@@ -117,7 +117,6 @@ static int tc_ctl_tfilter(struct sk_buff *skb, struct nlmsghdr *n)
{
struct net *net = sock_net(skb->sk);
struct nlattr *tca[TCA_MAX + 1];
- spinlock_t *root_lock;
struct tcmsg *t;
u32 protocol;
u32 prio;
@@ -215,8 +214,6 @@ replay:
}
rcu_read_unlock();
- root_lock = qdisc_root_sleeping_lock(q);
-
if ((tp = res) == NULL) {
/* Proto-tcf does not exist, create new one */
@@ -236,6 +233,7 @@ replay:
if (tp == NULL)
goto errout;
INIT_LIST_HEAD(&tp->head);
+ spin_lock_init(&tp->lock);
err = -ENOENT;
tp_ops = tcf_proto_lookup_ops(tca[TCA_KIND]);
if (tp_ops == NULL) {
@@ -287,9 +285,9 @@ replay:
if (fh == 0) {
if (n->nlmsg_type == RTM_DELTFILTER && t->tcm_handle == 0) {
- spin_lock_bh(root_lock);
+ tcf_tree_lock(tp);
list_del_rcu(&tp->head);
- spin_unlock_bh(root_lock);
+ tcf_tree_unlock(tp);
tfilter_notify(net, skb, n, tp, fh, RTM_DELTFILTER);
tcf_destroy(tp);
@@ -328,9 +326,9 @@ replay:
err = tp->ops->change(net, skb, tp, cl, t->tcm_handle, tca, &fh);
if (err == 0) {
if (tp_created) {
- spin_lock_bh(root_lock);
+ tcf_tree_lock(tp);
list_add_rcu(&tp->head, chain);
- spin_unlock_bh(root_lock);
+ tcf_tree_unlock(tp);
}
tfilter_notify(net, skb, n, tp, fh, RTM_NEWTFILTER);
} else {
--
1.8.3.1
^ permalink raw reply related
* [RFC Patch net-next 3/4] net_sched: use RCU for tc actions traverse
From: Cong Wang @ 2014-01-09 18:19 UTC (permalink / raw)
To: netdev
Cc: Cong Wang, John Fastabend, Eric Dumazet, David S. Miller,
Jamal Hadi Salim
In-Reply-To: <1389291593-2494-1-git-send-email-xiyou.wangcong@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
---
net/sched/act_api.c | 12 +++++++-----
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index f63e146..e3c655e 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -351,7 +351,7 @@ int tcf_action_exec(struct sk_buff *skb, const struct list_head *actions,
ret = TC_ACT_OK;
goto exec_done;
}
- list_for_each_entry(a, actions, list) {
+ list_for_each_entry_rcu(a, actions, list) {
repeat:
ret = a->ops->act(skb, a, res);
if (TC_MUNGED & skb->tc_verd) {
@@ -372,11 +372,13 @@ EXPORT_SYMBOL(tcf_action_exec);
void tcf_action_destroy(struct list_head *actions, int bind)
{
struct tc_action *a, *tmp;
+ LIST_HEAD(list);
+ list_splice_init_rcu(actions, &list, synchronize_rcu);
- list_for_each_entry_safe(a, tmp, actions, list) {
+ list_for_each_entry_safe(a, tmp, &list, list) {
if (a->ops->cleanup(a, bind) == ACT_P_DELETED)
module_put(a->ops->owner);
- list_del(&a->list);
+ list_del_rcu(&a->list);
kfree(a);
}
}
@@ -420,7 +422,7 @@ tcf_action_dump(struct sk_buff *skb, struct list_head *actions, int bind, int re
int err = -EINVAL;
struct nlattr *nest;
- list_for_each_entry(a, actions, list) {
+ list_for_each_entry_rcu(a, actions, list) {
nest = nla_nest_start(skb, a->order);
if (nest == NULL)
goto nla_put_failure;
@@ -542,7 +544,7 @@ int tcf_action_init(struct net *net, struct nlattr *nla,
goto err;
}
act->order = i;
- list_add_tail(&act->list, actions);
+ list_add_tail_rcu(&act->list, actions);
}
return 0;
--
1.8.3.1
^ permalink raw reply related
* [RFC Patch net-next 4/4] net_sched: make ingress qdisc lockless
From: Cong Wang @ 2014-01-09 18:19 UTC (permalink / raw)
To: netdev
Cc: Cong Wang, John Fastabend, Eric Dumazet, David S. Miller,
Jamal Hadi Salim
In-Reply-To: <1389291593-2494-1-git-send-email-xiyou.wangcong@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
---
net/core/dev.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/net/core/dev.c b/net/core/dev.c
index ce01847..e357d05 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3376,10 +3376,8 @@ static int ing_filter(struct sk_buff *skb, struct netdev_queue *rxq)
q = rxq->qdisc;
if (q != &noop_qdisc) {
- spin_lock(qdisc_lock(q));
if (likely(!test_bit(__QDISC_STATE_DEACTIVATED, &q->state)))
result = qdisc_enqueue_root(skb, q);
- spin_unlock(qdisc_lock(q));
}
return result;
--
1.8.3.1
^ permalink raw reply related
* Re: [Xen-devel] [PATCH net-next v3 1/9] xen-netback: Introduce TX grant map definitions
From: David Vrabel @ 2014-01-09 18:23 UTC (permalink / raw)
To: Stefano Stabellini
Cc: Wei Liu, jonathan.davies, Ian Campbell, netdev, linux-kernel,
Zoltan Kiss, xen-devel, Roger Pau Monné,
Konrad Rzeszutek Wilk
In-Reply-To: <alpine.DEB.2.02.1401091759100.21510@kaball.uk.xensource.com>
On 09/01/14 18:09, Stefano Stabellini wrote:
>
> I agree, but I think that we have different views on the use case.
> To me the m2p_override use case is "everywhere an mfn_to_pfn translation
> is required", that unfortunately is potentially everywhere at this time.
mfn_to_pfn() cannot be made to work correctly with foreign MFNs. It's a
fundamentally unsolvable problem.
IMO, the only sensible use of the m2p_override is to cause mfn_to_pfn()
to BUG() if a foreign MFN is used.
David
^ permalink raw reply
* [PATCH net-next] net/mlx4_en: call gro handler for encapsulated frames
From: Eric Dumazet @ 2014-01-09 18:30 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Amir Vadai, Or Gerlitz, Jerry Chu
From: Eric Dumazet <edumazet@google.com>
In order to use the native GRO handling of encapsulated protocols on
mlx4, we need to call napi_gro_receive() instead of netif_receive_skb()
unless busy polling is in action.
While we are at it, rename mlx4_en_cq_ll_polling() to
mlx4_en_cq_busy_polling()
Tested with GRE tunnel : GRO aggregation is now performed on the
ethernet device instead of being done later on gre device.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Amir Vadai <amirv@mellanox.com>
Cc: Jerry Chu <hkchu@google.com>
Cc: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 8 +++++---
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 4 ++--
2 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index 3b66f26ba049..890922c1c8ee 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -724,7 +724,7 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
* - not an IP fragment
* - no LLS polling in progress
*/
- if (!mlx4_en_cq_ll_polling(cq) &&
+ if (!mlx4_en_cq_busy_polling(cq) &&
(dev->features & NETIF_F_GRO)) {
struct sk_buff *gro_skb = napi_get_frags(&cq->napi);
if (!gro_skb)
@@ -816,8 +816,10 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
skb_mark_napi_id(skb, &cq->napi);
- /* Push it up the stack */
- netif_receive_skb(skb);
+ if (!mlx4_en_cq_busy_polling(cq))
+ napi_gro_receive(&cq->napi, skb);
+ else
+ netif_receive_skb(skb);
next:
for (nr = 0; nr < priv->num_frags; nr++)
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index 2f1e200f2e4c..fe7bdfebf353 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -661,7 +661,7 @@ static inline bool mlx4_en_cq_unlock_poll(struct mlx4_en_cq *cq)
}
/* true if a socket is polling, even if it did not get the lock */
-static inline bool mlx4_en_cq_ll_polling(struct mlx4_en_cq *cq)
+static inline bool mlx4_en_cq_busy_polling(struct mlx4_en_cq *cq)
{
WARN_ON(!(cq->state & MLX4_CQ_LOCKED));
return cq->state & CQ_USER_PEND;
@@ -691,7 +691,7 @@ static inline bool mlx4_en_cq_unlock_poll(struct mlx4_en_cq *cq)
return false;
}
-static inline bool mlx4_en_cq_ll_polling(struct mlx4_en_cq *cq)
+static inline bool mlx4_en_cq_busy_polling(struct mlx4_en_cq *cq)
{
return false;
}
^ permalink raw reply related
* Re: [PATCH v4 4/5] alx: add alx_get_stats64 operation
From: Ben Hutchings @ 2014-01-09 18:47 UTC (permalink / raw)
To: Sabrina Dubroca; +Cc: davem, johannes, stephen, netdev
In-Reply-To: <1389258571-10083-5-git-send-email-sd@queasysnail.net>
On Thu, 2014-01-09 at 10:09 +0100, Sabrina Dubroca wrote:
> Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Reviewed-by: Ben Hutchings <bhutchings@solarflare.com>
> ---
>
> I've modified the assignements of hw stats to netstats here as Ben asked in v3.
>
> The other atheros drivers use the same code as v3. Should I modify
> them as well in another patch?
That seems like a good thing to do.
Ben.
>
> drivers/net/ethernet/atheros/alx/alx.h | 3 ++
> drivers/net/ethernet/atheros/alx/main.c | 50 +++++++++++++++++++++++++++++++++
> 2 files changed, 53 insertions(+)
>
> diff --git a/drivers/net/ethernet/atheros/alx/alx.h b/drivers/net/ethernet/atheros/alx/alx.h
> index d71103d..8fc93c5 100644
> --- a/drivers/net/ethernet/atheros/alx/alx.h
> +++ b/drivers/net/ethernet/atheros/alx/alx.h
> @@ -106,6 +106,9 @@ struct alx_priv {
> u16 msg_enable;
>
> bool msi;
> +
> + /* protects hw.stats */
> + spinlock_t stats_lock;
> };
>
> extern const struct ethtool_ops alx_ethtool_ops;
> diff --git a/drivers/net/ethernet/atheros/alx/main.c b/drivers/net/ethernet/atheros/alx/main.c
> index c3c4c26..e92ffd6 100644
> --- a/drivers/net/ethernet/atheros/alx/main.c
> +++ b/drivers/net/ethernet/atheros/alx/main.c
> @@ -1166,10 +1166,60 @@ static void alx_poll_controller(struct net_device *netdev)
> }
> #endif
>
> +static struct rtnl_link_stats64 *alx_get_stats64(struct net_device *dev,
> + struct rtnl_link_stats64 *net_stats)
> +{
> + struct alx_priv *alx = netdev_priv(dev);
> + struct alx_hw_stats *hw_stats = &alx->hw.stats;
> +
> + spin_lock(&alx->stats_lock);
> +
> + alx_update_hw_stats(&alx->hw);
> +
> + net_stats->tx_bytes = hw_stats->tx_byte_cnt;
> + net_stats->rx_bytes = hw_stats->rx_byte_cnt;
> + net_stats->multicast = hw_stats->rx_mcast;
> + net_stats->collisions = hw_stats->tx_single_col +
> + hw_stats->tx_multi_col +
> + hw_stats->tx_late_col +
> + hw_stats->tx_abort_col;
> +
> + net_stats->rx_errors = hw_stats->rx_frag +
> + hw_stats->rx_fcs_err +
> + hw_stats->rx_len_err +
> + hw_stats->rx_ov_sz +
> + hw_stats->rx_ov_rrd +
> + hw_stats->rx_align_err +
> + hw_stats->rx_ov_rxf;
> +
> + net_stats->rx_fifo_errors = hw_stats->rx_ov_rxf;
> + net_stats->rx_length_errors = hw_stats->rx_len_err;
> + net_stats->rx_crc_errors = hw_stats->rx_fcs_err;
> + net_stats->rx_frame_errors = hw_stats->rx_align_err;
> + net_stats->rx_dropped = hw_stats->rx_ov_rrd;
> +
> + net_stats->tx_errors = hw_stats->tx_late_col +
> + hw_stats->tx_abort_col +
> + hw_stats->tx_underrun +
> + hw_stats->tx_trunc;
> +
> + net_stats->tx_aborted_errors = hw_stats->tx_abort_col;
> + net_stats->tx_fifo_errors = hw_stats->tx_underrun;
> + net_stats->tx_window_errors = hw_stats->tx_late_col;
> +
> + net_stats->tx_packets = hw_stats->tx_ok + net_stats->tx_errors;
> + net_stats->rx_packets = hw_stats->rx_ok + net_stats->rx_errors;
> +
> + spin_unlock(&alx->stats_lock);
> +
> + return net_stats;
> +}
> +
> static const struct net_device_ops alx_netdev_ops = {
> .ndo_open = alx_open,
> .ndo_stop = alx_stop,
> .ndo_start_xmit = alx_start_xmit,
> + .ndo_get_stats64 = alx_get_stats64,
> .ndo_set_rx_mode = alx_set_rx_mode,
> .ndo_validate_addr = eth_validate_addr,
> .ndo_set_mac_address = alx_set_mac_address,
--
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: [PATCH net-next 2/2] include/uapi/linux/xfrm.h: Pack struct xfrm_usersa_info
From: Ben Hutchings @ 2014-01-09 18:58 UTC (permalink / raw)
To: Fan Du; +Cc: steffen.klassert, davem, stephen, dev, netdev
In-Reply-To: <52CE5CBA.9000200@windriver.com>
On Thu, 2014-01-09 at 16:24 +0800, Fan Du wrote:
>
> On 2014年01月09日 04:33, Ben Hutchings wrote:
> > On Tue, 2014-01-07 at 14:48 +0800, Fan Du wrote:
> >> Otherwise 64bits kernel has sizeof(struct xfrm_usersa_info) 224 bytes,
> >> while 32bits compiled iproute2 see the same structure as 220 bytes, which
> >> leading deficit xfrm sa, in turn broken IPsec connectivity.
> >>
> >> Fix this by packing the structure.
> >>
> >> Signed-off-by: Fan Du<fan.du@windriver.com>
> >> ---
> >> include/uapi/linux/xfrm.h | 2 +-
> >> 1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/include/uapi/linux/xfrm.h b/include/uapi/linux/xfrm.h
> >> index 470bfae..61460c4 100644
> >> --- a/include/uapi/linux/xfrm.h
> >> +++ b/include/uapi/linux/xfrm.h
> >> @@ -366,7 +366,7 @@ struct xfrm_usersa_info {
> >> #define XFRM_STATE_AF_UNSPEC 32
> >> #define XFRM_STATE_ALIGN4 64
> >> #define XFRM_STATE_ESN 128
> >> -};
> >> +} __attribute__((packed));
> >>
> >> #define XFRM_SA_XFLAG_DONT_ENCAP_DSCP 1
> >>
> >
> > That change will make access to the structure very slow on some
> > architectures, and I suspect it will cause other compatibility problems.
> >
> > I think the right thing to do is to reduce the minimum length of the
> > structure in the netlink policy so that padding at the end is not
> > required.
>
> Could you please be more specific about this? Thanks.
[...]
Each netlink message and attribute has a specified minimum size, and
messages are rejected if the size provided by the sender is too small.
The minimum size is normally the same as the structure size, but doesn't
have to be.
Something like this might work:
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -2266,8 +2266,15 @@ static int xfrm_send_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
#define XMSGSIZE(type) sizeof(struct type)
+/* Padding at the end of struct xfrm_usersa_info differs between
+ * architectures so for 32/64-bit compat we don't require padding
+ */
+#define XMSGSIZE_XFRM_USERSA_INFO \
+ (offsetof(struct xfrm_usersa_info, flags) + \
+ sizeof(((struct xfrm_usersa_info *)NULL)->flags))
+
static const int xfrm_msg_min[XFRM_NR_MSGTYPES] = {
- [XFRM_MSG_NEWSA - XFRM_MSG_BASE] = XMSGSIZE(xfrm_usersa_info),
+ [XFRM_MSG_NEWSA - XFRM_MSG_BASE] = XMSGSIZE_XFRM_USERSA_INFO,
[XFRM_MSG_DELSA - XFRM_MSG_BASE] = XMSGSIZE(xfrm_usersa_id),
[XFRM_MSG_GETSA - XFRM_MSG_BASE] = XMSGSIZE(xfrm_usersa_id),
[XFRM_MSG_NEWPOLICY - XFRM_MSG_BASE] = XMSGSIZE(xfrm_userpolicy_info),
@@ -2277,7 +2284,7 @@ static const int xfrm_msg_min[XFRM_NR_MSGTYPES] = {
[XFRM_MSG_ACQUIRE - XFRM_MSG_BASE] = XMSGSIZE(xfrm_user_acquire),
[XFRM_MSG_EXPIRE - XFRM_MSG_BASE] = XMSGSIZE(xfrm_user_expire),
[XFRM_MSG_UPDPOLICY - XFRM_MSG_BASE] = XMSGSIZE(xfrm_userpolicy_info),
- [XFRM_MSG_UPDSA - XFRM_MSG_BASE] = XMSGSIZE(xfrm_usersa_info),
+ [XFRM_MSG_UPDSA - XFRM_MSG_BASE] = XMSGSIZE_XFRM_USERSA_INFO,
[XFRM_MSG_POLEXPIRE - XFRM_MSG_BASE] = XMSGSIZE(xfrm_user_polexpire),
[XFRM_MSG_FLUSHSA - XFRM_MSG_BASE] = XMSGSIZE(xfrm_usersa_flush),
[XFRM_MSG_FLUSHPOLICY - XFRM_MSG_BASE] = 0,
@@ -2292,7 +2299,7 @@ static const int xfrm_msg_min[XFRM_NR_MSGTYPES] = {
#undef XMSGSIZE
static const struct nla_policy xfrma_policy[XFRMA_MAX+1] = {
- [XFRMA_SA] = { .len = sizeof(struct xfrm_usersa_info)},
+ [XFRMA_SA] = { .len = XMSGSIZE_XFRM_USERSA_INFO },
[XFRMA_POLICY] = { .len = sizeof(struct xfrm_userpolicy_info)},
[XFRMA_LASTUSED] = { .type = NLA_U64},
[XFRMA_ALG_AUTH_TRUNC] = { .len = sizeof(struct xfrm_algo_auth)},
---
But beware that if the structure is copied in and then copied back out
to userland then the padding will need to be explicitly cleared.
Ben.
--
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* RE: [PATCH net] bnx2x: fix DMA unmapping of TSO split BDs
From: Dmitry Kravkov @ 2014-01-09 19:23 UTC (permalink / raw)
To: Eric Dumazet, Michal Schmidt
Cc: netdev@vger.kernel.org, Ariel Elior, Yuval Mintz, David Miller
In-Reply-To: <1389289923.31367.39.camel@edumazet-glaptop2.roam.corp.google.com>
> -----Original Message-----
> From: Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Thursday, January 09, 2014 7:52 PM
> To: Michal Schmidt
> Cc: netdev@vger.kernel.org; Ariel Elior; Yuval Mintz; Dmitry Kravkov; David
> Miller
> Subject: Re: [PATCH net] bnx2x: fix DMA unmapping of TSO split BDs
>
> On Thu, 2014-01-09 at 14:36 +0100, Michal Schmidt wrote:
> > bnx2x triggers warnings with CONFIG_DMA_API_DEBUG=y:
> >
> > WARNING: CPU: 0 PID: 2253 at lib/dma-debug.c:887
> check_unmap+0xf8/0x920()
> > bnx2x 0000:28:00.0: DMA-API: device driver frees DMA memory with
> > different size [device address=0x00000000da2b389e] [map size=1490
> bytes]
> > [unmap size=66 bytes]
> >
> > The reason is that bnx2x splits a TSO BD into two BDs (headers + data)
> > using one DMA mapping for both, but it uses only the length of the
> > first BD when unmapping.
> >
> > This patch fixes the bug by unmapping the whole length of the two BDs.
> >
> > Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
> > ---
>
> Reviewed-by: Eric Dumazet <edumazet@google.com>
>
Thanks you, Guys.
Acked-by: Dmitry Kravkov <dmitry@broadcom.com>
^ permalink raw reply
* Re: [PATCH -next] openvswitch: Use kmem_cache_free() instead of kfree()
From: David Miller @ 2014-01-09 19:27 UTC (permalink / raw)
To: jesse; +Cc: weiyj.lk, pshelar, yongjun_wei, dev, netdev
In-Reply-To: <CAEP_g=9HVB0UkajKesyC8Dkj=0ts3kExbhhHKj2wrjx9RymEog@mail.gmail.com>
From: Jesse Gross <jesse@nicira.com>
Date: Wed, 8 Jan 2014 09:17:11 -0500
> On Wed, Jan 8, 2014 at 5:13 AM, Wei Yongjun <weiyj.lk@gmail.com> wrote:
>> From: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
>>
>> memory allocated by kmem_cache_alloc() should be freed using
>> kmem_cache_free(), not kfree().
>>
>> Fixes: e298e5057006 ('openvswitch: Per cpu flow stats.')
>> Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
>
> Good catch, thanks.
>
> I'll let David pick this up directly since I just flushed my patch queue.
>
> Acked-by: Jesse Gross <jesse@nicira.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH -next] ip_tunnel: fix sparse non static symbol warning
From: David Miller @ 2014-01-09 19:33 UTC (permalink / raw)
To: weiyj.lk; +Cc: kuznet, jmorris, yoshfuji, kaber, therbert, yongjun_wei, netdev
In-Reply-To: <CAPgLHd__qSufRW4YE63SocWV1gcgzBCBX1bdy+WcJ7v8zhE8qw@mail.gmail.com>
From: Wei Yongjun <weiyj.lk@gmail.com>
Date: Wed, 8 Jan 2014 21:59:30 +0800
> From: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
>
> Fixes the following sparse warning:
>
> net/ipv4/ip_tunnel.c:116:18: warning:
> symbol 'tunnel_dst_check' was not declared. Should it be static?
>
> Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Applied.
^ permalink raw reply
* Re: [PATCH net-next v2 3/4] virtio-net: auto-tune mergeable rx buffer size for improved performance
From: Michael Dalton @ 2014-01-09 19:33 UTC (permalink / raw)
To: Michael S. Tsirkin; +Cc: netdev, lf-virt, Eric Dumazet, David S. Miller
In-Reply-To: <20140109132529.GB2712@redhat.com>
Hi Michael,
Your improvements (code changes, more consistent naming, and use of
256-byte alignment only) all sound good to me. I will get started on a v3
patchset in conformance with your recommendations after sorting out
what we want to do with the debugfs/sysfs issues. I will followup soon on
the thread for patch 4/4 so we can close on what changes are needed for
debugfs/sysfs. Thanks!
Best,
Mike
^ permalink raw reply
* Re: [net-next 3/7] ixgbe: Use static inlines instead of macros
From: David Miller @ 2014-01-09 19:39 UTC (permalink / raw)
To: mark.d.rustad; +Cc: sfeldma, aaron.f.brown, netdev, gospo, sassmann
In-Reply-To: <2C8FA848-3185-43BA-BDD7-955B0C7DD835@intel.com>
From: "Rustad, Mark D" <mark.d.rustad@intel.com>
Date: Thu, 9 Jan 2014 17:34:18 +0000
> On Jan 8, 2014, at 12:47 AM, Scott Feldman <sfeldma@cumulusnetworks.com> wrote:
>
>>
>> On Jan 7, 2014, at 11:40 PM, Aaron Brown <aaron.f.brown@intel.com> wrote:
>>
>>> From: Mark Rustad <mark.d.rustad@intel.com>
>>>
>>> -#define IXGBE_WRITE_REG(a, reg, value) writel((value), ((a)->hw_addr + (reg)))
>>> +static inline void IXGBE_WRITE_REG(struct ixgbe_hw *hw, u32 reg, u32 value)
>>
>> Bummer, now you have a all-caps func name.
>
> Agreed, but this is actually a fairly common condition among drivers that used to use macros. It isn't perfect, but at least it is moving in the right direction. I'd rather leave the case change for a later patch series that does only that or has some reason to touch all of the register access sites.
>
> At least the new accessor I introduced is lower case. :-)
Please address this feedback, all caps function names are really not
appropriate.
^ permalink raw reply
* Re: [PATCH net-next v3 1/9] xen-netback: Introduce TX grant map definitions
From: Zoltan Kiss @ 2014-01-09 19:53 UTC (permalink / raw)
To: Wei Liu
Cc: ian.campbell, xen-devel, netdev, linux-kernel, jonathan.davies,
Roger Pau Monné, David Vrabel
In-Reply-To: <20140109153010.GE12164@zion.uk.xensource.com>
On 09/01/14 15:30, Wei Liu wrote:
> On Wed, Jan 08, 2014 at 12:10:10AM +0000, Zoltan Kiss wrote:
>> v3:
>> - fix comment in xenvif_tx_dealloc_action()
>> - call unmap hypercall directly instead of gnttab_unmap_refs(), which does
>> unnecessary m2p_override. Also remove pages_to_[un]map members
>
> Is it worthy to have another function call
> gnttab_unmap_refs_no_m2p_override in Xen core driver, or just add a
> parameter to control wether we need to touch m2p_override? I *think* it
> will benefit block driver as well?
>
> (CC Roger and David for input)
Yep, it worth, but let's make it a different patch
>> --- a/drivers/net/xen-netback/netback.c
>> +++ b/drivers/net/xen-netback/netback.c
>> @@ -771,6 +771,19 @@ static struct page *xenvif_alloc_page(struct xenvif *vif,
>> return page;
>> }
>>
>> +static inline void xenvif_tx_create_gop(struct xenvif *vif, u16 pending_idx,
>> + struct xen_netif_tx_request *txp,
>> + struct gnttab_map_grant_ref *gop)
>
> Indentation.
I fixed it and the later ones up, hopefully I haven't missed anything.
>
>> +
>> + spin_lock_irqsave(&vif->dealloc_lock, flags);
>> + do {
>> + pending_idx = ubuf->desc;
>> + ubuf = (struct ubuf_info *) ubuf->ctx;
>> + index = pending_index(vif->dealloc_prod);
>> + vif->dealloc_ring[index] = pending_idx;
>> + /* Sync with xenvif_tx_action_dealloc:
>
> xenvif_tx_dealloc_action I suppose.
Yes.
>> + /* Already unmapped? */
>> + if (vif->grant_tx_handle[pending_idx] ==
>> + NETBACK_INVALID_HANDLE) {
>> + netdev_err(vif->dev,
>> + "Trying to unmap invalid handle! "
>> + "pending_idx: %x\n", pending_idx);
>> + continue;
>
> You seemed to miss the BUG_ON we discussed?
>
> See thread starting <52AF1A84.3090304@citrix.com>.
Indeed, despite I wrote it in the version history :)
^ permalink raw reply
* Re: [net-next 05/15] i40e: add a comment on barrier and fix panic on reset
From: David Miller @ 2014-01-09 20:12 UTC (permalink / raw)
To: gregory.v.rose
Cc: sergei.shtylyov, jeffrey.t.kirsher, netdev, gospo, sassmann,
mitch.a.williams, jesse.brandeburg
In-Reply-To: <20140109091735.00002cf0@unknown>
From: Greg Rose <gregory.v.rose@intel.com>
Date: Thu, 9 Jan 2014 09:17:35 -0800
> Yes, it's apparent that two different internal patches were compressed
> together. If its unacceptable to do this then I'll speak to Jeff about
> splitting them.
I won't reject the entire pull request on account of this, but please
don't do this in the future.
^ permalink raw reply
* [PATCH v2 0/4] net_dma removal, and dma debug extension
From: Dan Williams @ 2014-01-09 20:12 UTC (permalink / raw)
To: dmaengine; +Cc: netdev, linux-kernel
Follow up patches to 77873803363c "net_dma: mark broken" to remove
net_dma bits and provide debug infrastructure to flag other
get_user_pages() vs dma instances that might violate the dma api.
Will takes this through the dmaengine tree once acked.
Changes since v1 [1]:
1/ net_dma removal patch has been expanded to revert other
net_dma induced changes.
2/ updated the debug_dma_assert_idle() api to be gated on
CONFIG_DMA_VS_CPU_DEBUG
---
Dan Williams (4):
net_dma: simple removal
net_dma: revert 'copied_early'
net: make tcp_cleanup_rbuf private
dma debug: introduce debug_dma_assert_idle()
Documentation/ABI/removed/net_dma | 8 +
Documentation/networking/ip-sysctl.txt | 6 -
drivers/dma/Kconfig | 12 -
drivers/dma/Makefile | 1
drivers/dma/dmaengine.c | 104 ------------
drivers/dma/ioat/dma.c | 1
drivers/dma/ioat/dma.h | 7 -
drivers/dma/ioat/dma_v2.c | 1
drivers/dma/ioat/dma_v3.c | 1
drivers/dma/iovlock.c | 280 --------------------------------
include/linux/dma-debug.h | 6 +
include/linux/dmaengine.h | 22 ---
include/linux/skbuff.h | 8 -
include/linux/tcp.h | 8 -
include/net/netdma.h | 32 ----
include/net/sock.h | 19 --
include/net/tcp.h | 9 -
kernel/sysctl_binary.c | 1
lib/Kconfig.debug | 13 +
lib/dma-debug.c | 130 +++++++++++++--
mm/memory.c | 3
net/core/Makefile | 1
net/core/dev.c | 10 -
net/core/sock.c | 6 -
net/core/user_dma.c | 131 ---------------
net/dccp/proto.c | 4
net/ipv4/sysctl_net_ipv4.c | 9 -
net/ipv4/tcp.c | 157 ++----------------
net/ipv4/tcp_input.c | 83 +--------
net/ipv4/tcp_ipv4.c | 18 --
net/ipv6/tcp_ipv6.c | 13 -
net/llc/af_llc.c | 10 +
32 files changed, 186 insertions(+), 928 deletions(-)
create mode 100644 Documentation/ABI/removed/net_dma
delete mode 100644 drivers/dma/iovlock.c
delete mode 100644 include/net/netdma.h
delete mode 100644 net/core/user_dma.c
^ permalink raw reply
* Re: [net-next 00/15][pull request] Intel Wired LAN Driver Updates
From: David Miller @ 2014-01-09 20:13 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, sassmann
In-Reply-To: <1389271944-26227-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Thu, 9 Jan 2014 04:52:09 -0800
> This series contains updates to i40e only.
...
> git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master
Pulled, thanks Jeff.
^ permalink raw reply
* Re: Multicast routing stops functioning after 4G multicast packets recived.
From: Bob Falken @ 2014-01-09 20:14 UTC (permalink / raw)
To: Hannes Frederic Sowa, Julian Anastasov, netdev
Hello,
Testing this patch as im typing this. will check status in about 12hours.
Unfortuantly, I dont have any receivers avaialble for requesting the multicast stream on the edge point anymore.
So there is not TX traffic a.t.m..
I will have a better test-lab available next week. (hopefully).
Thanks again.
----- Original Message -----
From: Hannes Frederic Sowa
Sent: 01/07/14 09:20 PM
To: Bob Falken, Julian Anastasov, netdev@vger.kernel.org
Subject: Re: Multicast routing stops functioning after 4G multicast packets recived.
On Tue, Jan 07, 2014 at 09:11:47PM +0100, Hannes Frederic Sowa wrote:
> On Tue, Jan 07, 2014 at 06:43:22PM +0100, Hannes Frederic Sowa wrote:
> > On Tue, Jan 07, 2014 at 06:01:44PM +0100, Bob Falken wrote:
> > > Hello,
> > >
> > > I patched, kernel 3.2.53 yesterday,
> > > 8774002632packets, and ~9.1TB later, the multicast routing seems to function properly. :)
> > >
> > > Kudos for fixing this.
> > >
> > > I will keep checking the next days.
> > >
> > > As for IPv6 MR, my current setup i.e: the Multicast source, does not support IPv6, so cannot do a check for that natively.
> > >
> > > Unless I can translate IPv4 multicast into IPv6 multicast easily, using some iptable prerouting rules(?).
> >
> > Thank you for testing!
> >
> > I'll review the RCU regions again and will prepare the patches (before
> > introduction of ebc0ffae5dfb44 ("fib: RCU conversion of fib_lookup()")
> > and after.
>
> It seems ip(6)mr_fib_lookup is not always called from rcu section
> (ndo_start_xmit), so I had to restructure a bit. Could you retest this
> patch as preparation for a submission to stable? Thanks!
>
> RCU conversion can be done later then.
Broken patch, sorry. Please try this one:
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 421a249..e5e9071 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -157,9 +157,12 @@ static struct mr_table *ipmr_get_table(struct net *net, u32 id)
static int ipmr_fib_lookup(struct net *net, struct flowi4 *flp4,
struct mr_table **mrt)
{
- struct ipmr_result res;
- struct fib_lookup_arg arg = { .result = &res};
int err;
+ struct ipmr_result res;
+ struct fib_lookup_arg arg = {
+ .result = &res,
+ .flags = FIB_LOOKUP_NOREF,
+ };
err = fib_rules_lookup(net->ipv4.mr_rules_ops,
flowi4_to_flowi(flp4), 0, &arg);
@@ -448,16 +451,22 @@ failure:
static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, struct net_device *dev)
{
+ int err;
+ struct ipmr_result res;
struct net *net = dev_net(dev);
- struct mr_table *mrt;
+
+ struct fib_lookup_arg arg = {
+ .result = &res,
+ };
+
struct flowi4 fl4 = {
.flowi4_oif = dev->ifindex,
.flowi4_iif = skb->skb_iif,
.flowi4_mark = skb->mark,
};
- int err;
- err = ipmr_fib_lookup(net, &fl4, &mrt);
+ err = fib_rules_lookup(net->ipv4.mr_rules_ops,
+ flowi4_to_flowi(&fl4), 0, &arg);
if (err < 0) {
kfree_skb(skb);
return err;
@@ -466,9 +475,12 @@ static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, struct net_device *dev)
read_lock(&mrt_lock);
dev->stats.tx_bytes += skb->len;
dev->stats.tx_packets++;
- ipmr_cache_report(mrt, skb, mrt->mroute_reg_vif_num, IGMPMSG_WHOLEPKT);
+ ipmr_cache_report(res.mrt, skb, res.mrt->mroute_reg_vif_num,
+ IGMPMSG_WHOLEPKT);
read_unlock(&mrt_lock);
kfree_skb(skb);
+ if (arg.rule)
+ fib_rule_put(arg.rule);
return NETDEV_TX_OK;
}
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c
index f365310..45ec621 100644
--- a/net/ipv6/ip6mr.c
+++ b/net/ipv6/ip6mr.c
@@ -141,9 +141,12 @@ static struct mr6_table *ip6mr_get_table(struct net *net, u32 id)
static int ip6mr_fib_lookup(struct net *net, struct flowi6 *flp6,
struct mr6_table **mrt)
{
- struct ip6mr_result res;
- struct fib_lookup_arg arg = { .result = &res};
int err;
+ struct ip6mr_result res;
+ struct fib_lookup_arg arg = {
+ .result = &res,
+ .flags = FIB_LOOKUP_NOREF,
+ };
err = fib_rules_lookup(net->ipv6.mr6_rules_ops,
flowi6_to_flowi(flp6), 0, &arg);
@@ -693,16 +696,20 @@ static const struct inet6_protocol pim6_protocol = {
static netdev_tx_t reg_vif_xmit(struct sk_buff *skb,
struct net_device *dev)
{
+ int err;
+ struct ip6mr_result res;
struct net *net = dev_net(dev);
- struct mr6_table *mrt;
struct flowi6 fl6 = {
.flowi6_oif = dev->ifindex,
.flowi6_iif = skb->skb_iif,
.flowi6_mark = skb->mark,
};
- int err;
+ struct fib_lookup_arg arg = {
+ .result = &res,
+ };
- err = ip6mr_fib_lookup(net, &fl6, &mrt);
+ err = fib_rules_lookup(net->ipv6.mr6_rules_ops,
+ flowi6_to_flowi(&fl6), 0, &arg);
if (err < 0) {
kfree_skb(skb);
return err;
@@ -711,9 +718,12 @@ static netdev_tx_t reg_vif_xmit(struct sk_buff *skb,
read_lock(&mrt_lock);
dev->stats.tx_bytes += skb->len;
dev->stats.tx_packets++;
- ip6mr_cache_report(mrt, skb, mrt->mroute_reg_vif_num, MRT6MSG_WHOLEPKT);
+ ip6mr_cache_report(res.mrt, skb, res.mrt->mroute_reg_vif_num,
+ MRT6MSG_WHOLEPKT);
read_unlock(&mrt_lock);
kfree_skb(skb);
+ if (arg.rule)
+ fib_rule_put(arg.rule);
return NETDEV_TX_OK;
}
^ permalink raw reply related
* RE: [net-next 05/15] i40e: add a comment on barrier and fix panic on reset
From: Rose, Gregory V @ 2014-01-09 20:14 UTC (permalink / raw)
To: David Miller
Cc: sergei.shtylyov@cogentembedded.com, Kirsher, Jeffrey T,
netdev@vger.kernel.org, gospo@redhat.com, sassmann@redhat.com,
Williams, Mitch A, Brandeburg, Jesse
In-Reply-To: <20140109.151224.1539252523028219271.davem@davemloft.net>
> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net]
> Sent: Thursday, January 09, 2014 12:12 PM
> To: Rose, Gregory V
> Cc: sergei.shtylyov@cogentembedded.com; Kirsher, Jeffrey T;
> netdev@vger.kernel.org; gospo@redhat.com; sassmann@redhat.com; Williams,
> Mitch A; Brandeburg, Jesse
> Subject: Re: [net-next 05/15] i40e: add a comment on barrier and fix panic
> on reset
>
> From: Greg Rose <gregory.v.rose@intel.com>
> Date: Thu, 9 Jan 2014 09:17:35 -0800
>
> > Yes, it's apparent that two different internal patches were compressed
> > together. If its unacceptable to do this then I'll speak to Jeff
> > about splitting them.
>
> I won't reject the entire pull request on account of this, but please
> don't do this in the future.
Thanks Dave,
We've discussed this internally and ahem... mistakes were made. We'll look out for it in the future.
- Greg
^ permalink raw reply
* Re: [net-next 3/7] ixgbe: Use static inlines instead of macros
From: Rustad, Mark D @ 2014-01-09 20:14 UTC (permalink / raw)
To: David Miller
Cc: Scott Feldman, Brown, Aaron F, Netdev, gospo@redhat.com,
sassmann@redhat.com
In-Reply-To: <20140109.143923.437905458430188006.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 1984 bytes --]
On Jan 9, 2014, at 11:39 AM, David Miller <davem@davemloft.net> wrote:
> From: "Rustad, Mark D" <mark.d.rustad@intel.com>
> Date: Thu, 9 Jan 2014 17:34:18 +0000
>
>> On Jan 8, 2014, at 12:47 AM, Scott Feldman <sfeldma@cumulusnetworks.com> wrote:
>>
>>>
>>> On Jan 7, 2014, at 11:40 PM, Aaron Brown <aaron.f.brown@intel.com> wrote:
>>>
>>>> From: Mark Rustad <mark.d.rustad@intel.com>
>>>>
>>>> -#define IXGBE_WRITE_REG(a, reg, value) writel((value), ((a)->hw_addr + (reg)))
>>>> +static inline void IXGBE_WRITE_REG(struct ixgbe_hw *hw, u32 reg, u32 value)
>>>
>>> Bummer, now you have a all-caps func name.
>>
>> Agreed, but this is actually a fairly common condition among drivers that used to use macros. It isn't perfect, but at least it is moving in the right direction. I'd rather leave the case change for a later patch series that does only that or has some reason to touch all of the register access sites.
>>
>> At least the new accessor I introduced is lower case. :-)
>
> Please address this feedback, all caps function names are really not
> appropriate.
I really don't think it is a good idea to do that as part of this patch series. It makes this patch series a pretty solid barrier to any other patches going into this driver because it would change the name of all the register accessors.
This makes me want to deal with that as a separate issue, since there would be no functional reason to drop such a patch and it can be planned into a workflow.
Obviously I could do it here, but I *really* think it is procedurally a really bad idea to change the case as part of a functional change. I thought I was doing a favor my at least making them inlines, but prehaps not.
Anyone want to take on changing the upper case static inlines in mcf8390, 7990, benet, ns83820, s2io, vxge, iwlwifi, ath9k, wil6210, mwifiex, and rtlwifi? And those are just under drivers/net.
--
Mark Rustad, Networking Division, Intel Corporation
[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 841 bytes --]
^ permalink raw reply
* [PATCH v2 2/4] net_dma: revert 'copied_early'
From: Dan Williams @ 2014-01-09 20:16 UTC (permalink / raw)
To: dmaengine
Cc: Hideaki YOSHIFUJI, netdev, Ali Saidi, James Morris,
David S. Miller, Eric Dumazet, Alexey Kuznetsov, Dave Jones,
Patrick McHardy, linux-kernel
In-Reply-To: <20140109201157.28381.94057.stgit@viggo.jf.intel.com>
Now that tcp_dma_try_early_copy() is gone nothing ever sets
copied_early.
Also reverts "53240c208776 tcp: Fix possible double-ack w/ user dma"
since it is no longer necessary.
Cc: Ali Saidi <saidi@engin.umich.edu>
Cc: James Morris <jmorris@namei.org>
Cc: Patrick McHardy <kaber@trash.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
New in v2.
net/ipv4/tcp_input.c | 22 ++++++++--------------
1 files changed, 8 insertions(+), 14 deletions(-)
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 33ef18e550c5..15911a280485 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -5145,19 +5145,15 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
}
} else {
int eaten = 0;
- int copied_early = 0;
bool fragstolen = false;
- if (tp->copied_seq == tp->rcv_nxt &&
- len - tcp_header_len <= tp->ucopy.len) {
- if (tp->ucopy.task == current &&
- sock_owned_by_user(sk) && !copied_early) {
- __set_current_state(TASK_RUNNING);
+ if (tp->ucopy.task == current &&
+ tp->copied_seq == tp->rcv_nxt &&
+ len - tcp_header_len <= tp->ucopy.len &&
+ sock_owned_by_user(sk)) {
+ __set_current_state(TASK_RUNNING);
- if (!tcp_copy_to_iovec(sk, skb, tcp_header_len))
- eaten = 1;
- }
- if (eaten) {
+ if (!tcp_copy_to_iovec(sk, skb, tcp_header_len)) {
/* Predicted packet is in window by definition.
* seq == rcv_nxt and rcv_wup <= rcv_nxt.
* Hence, check seq<=rcv_wup reduces to:
@@ -5173,9 +5169,8 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
__skb_pull(skb, tcp_header_len);
tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq;
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPHPHITSTOUSER);
+ eaten = 1;
}
- if (copied_early)
- tcp_cleanup_rbuf(sk, skb->len);
}
if (!eaten) {
if (tcp_checksum_complete_user(sk, skb))
@@ -5212,8 +5207,7 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
goto no_ack;
}
- if (!copied_early || tp->rcv_nxt != tp->rcv_wup)
- __tcp_ack_snd_check(sk, 0);
+ __tcp_ack_snd_check(sk, 0);
no_ack:
if (eaten)
kfree_skb_partial(skb, fragstolen);
^ permalink raw reply related
* [PATCH v2 1/4] net_dma: simple removal
From: Dan Williams @ 2014-01-09 20:16 UTC (permalink / raw)
To: dmaengine
Cc: Alexander Duyck, Dave Jiang, Vinod Koul, netdev, David Whipple,
linux-kernel, David S. Miller
In-Reply-To: <20140109201157.28381.94057.stgit@viggo.jf.intel.com>
Per commit "77873803363c net_dma: mark broken" net_dma is no longer used
and there is no plan to fix it.
This is the mechanical removal of bits in CONFIG_NET_DMA ifdef guards.
Reverting the remainder of the net_dma induced changes is deferred to
subsequent patches.
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: David Whipple <whipple@securedatainnovations.ch>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
Updated the changelog to foreshadow the other reverts, otherwise the
same straightforward removal as before.
Documentation/ABI/removed/net_dma | 8 +
Documentation/networking/ip-sysctl.txt | 6 -
drivers/dma/Kconfig | 12 -
drivers/dma/Makefile | 1
drivers/dma/dmaengine.c | 104 ------------
drivers/dma/ioat/dma.c | 1
drivers/dma/ioat/dma.h | 7 -
drivers/dma/ioat/dma_v2.c | 1
drivers/dma/ioat/dma_v3.c | 1
drivers/dma/iovlock.c | 280 --------------------------------
include/linux/dmaengine.h | 22 ---
include/linux/skbuff.h | 8 -
include/linux/tcp.h | 8 -
include/net/netdma.h | 32 ----
include/net/sock.h | 19 --
include/net/tcp.h | 8 -
kernel/sysctl_binary.c | 1
net/core/Makefile | 1
net/core/dev.c | 10 -
net/core/sock.c | 6 -
net/core/user_dma.c | 131 ---------------
net/dccp/proto.c | 4
net/ipv4/sysctl_net_ipv4.c | 9 -
net/ipv4/tcp.c | 147 ++---------------
net/ipv4/tcp_input.c | 61 -------
net/ipv4/tcp_ipv4.c | 18 --
net/ipv6/tcp_ipv6.c | 13 -
net/llc/af_llc.c | 10 +
28 files changed, 35 insertions(+), 894 deletions(-)
create mode 100644 Documentation/ABI/removed/net_dma
delete mode 100644 drivers/dma/iovlock.c
delete mode 100644 include/net/netdma.h
delete mode 100644 net/core/user_dma.c
diff --git a/Documentation/ABI/removed/net_dma b/Documentation/ABI/removed/net_dma
new file mode 100644
index 000000000000..a173aecc2f18
--- /dev/null
+++ b/Documentation/ABI/removed/net_dma
@@ -0,0 +1,8 @@
+What: tcp_dma_copybreak sysctl
+Date: Removed in kernel v3.13
+Contact: Dan Williams <dan.j.williams@intel.com>
+Description:
+ Formerly the lower limit, in bytes, of the size of socket reads
+ that will be offloaded to a DMA copy engine. Removed due to
+ coherency issues of the cpu potentially touching the buffers
+ while dma is in flight.
diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index 3c12d9a7ed00..bdd8a67f0be2 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -538,12 +538,6 @@ tcp_workaround_signed_windows - BOOLEAN
not receive a window scaling option from them.
Default: 0
-tcp_dma_copybreak - INTEGER
- Lower limit, in bytes, of the size of socket reads that will be
- offloaded to a DMA copy engine, if one is present in the system
- and CONFIG_NET_DMA is enabled.
- Default: 4096
-
tcp_thin_linear_timeouts - BOOLEAN
Enable dynamic triggering of linear timeouts for thin streams.
If set, a check is performed upon retransmission by timeout to
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index c823daaf9043..b24f13195272 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -351,18 +351,6 @@ config DMA_OF
comment "DMA Clients"
depends on DMA_ENGINE
-config NET_DMA
- bool "Network: TCP receive copy offload"
- depends on DMA_ENGINE && NET
- default (INTEL_IOATDMA || FSL_DMA)
- depends on BROKEN
- help
- This enables the use of DMA engines in the network stack to
- offload receive copy-to-user operations, freeing CPU cycles.
-
- Say Y here if you enabled INTEL_IOATDMA or FSL_DMA, otherwise
- say N.
-
config ASYNC_TX_DMA
bool "Async_tx: Offload support for the async_tx api"
depends on DMA_ENGINE
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 0ce2da97e429..024b008a25de 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -6,7 +6,6 @@ obj-$(CONFIG_DMA_VIRTUAL_CHANNELS) += virt-dma.o
obj-$(CONFIG_DMA_ACPI) += acpi-dma.o
obj-$(CONFIG_DMA_OF) += of-dma.o
-obj-$(CONFIG_NET_DMA) += iovlock.o
obj-$(CONFIG_INTEL_MID_DMAC) += intel_mid_dma.o
obj-$(CONFIG_DMATEST) += dmatest.o
obj-$(CONFIG_INTEL_IOATDMA) += ioat/
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index ef63b9058f3c..d7f4f4e0d71f 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -1029,110 +1029,6 @@ dmaengine_get_unmap_data(struct device *dev, int nr, gfp_t flags)
}
EXPORT_SYMBOL(dmaengine_get_unmap_data);
-/**
- * dma_async_memcpy_pg_to_pg - offloaded copy from page to page
- * @chan: DMA channel to offload copy to
- * @dest_pg: destination page
- * @dest_off: offset in page to copy to
- * @src_pg: source page
- * @src_off: offset in page to copy from
- * @len: length
- *
- * Both @dest_page/@dest_off and @src_page/@src_off must be mappable to a bus
- * address according to the DMA mapping API rules for streaming mappings.
- * Both @dest_page/@dest_off and @src_page/@src_off must stay memory resident
- * (kernel memory or locked user space pages).
- */
-dma_cookie_t
-dma_async_memcpy_pg_to_pg(struct dma_chan *chan, struct page *dest_pg,
- unsigned int dest_off, struct page *src_pg, unsigned int src_off,
- size_t len)
-{
- struct dma_device *dev = chan->device;
- struct dma_async_tx_descriptor *tx;
- struct dmaengine_unmap_data *unmap;
- dma_cookie_t cookie;
- unsigned long flags;
-
- unmap = dmaengine_get_unmap_data(dev->dev, 2, GFP_NOWAIT);
- if (!unmap)
- return -ENOMEM;
-
- unmap->to_cnt = 1;
- unmap->from_cnt = 1;
- unmap->addr[0] = dma_map_page(dev->dev, src_pg, src_off, len,
- DMA_TO_DEVICE);
- unmap->addr[1] = dma_map_page(dev->dev, dest_pg, dest_off, len,
- DMA_FROM_DEVICE);
- unmap->len = len;
- flags = DMA_CTRL_ACK;
- tx = dev->device_prep_dma_memcpy(chan, unmap->addr[1], unmap->addr[0],
- len, flags);
-
- if (!tx) {
- dmaengine_unmap_put(unmap);
- return -ENOMEM;
- }
-
- dma_set_unmap(tx, unmap);
- cookie = tx->tx_submit(tx);
- dmaengine_unmap_put(unmap);
-
- preempt_disable();
- __this_cpu_add(chan->local->bytes_transferred, len);
- __this_cpu_inc(chan->local->memcpy_count);
- preempt_enable();
-
- return cookie;
-}
-EXPORT_SYMBOL(dma_async_memcpy_pg_to_pg);
-
-/**
- * dma_async_memcpy_buf_to_buf - offloaded copy between virtual addresses
- * @chan: DMA channel to offload copy to
- * @dest: destination address (virtual)
- * @src: source address (virtual)
- * @len: length
- *
- * Both @dest and @src must be mappable to a bus address according to the
- * DMA mapping API rules for streaming mappings.
- * Both @dest and @src must stay memory resident (kernel memory or locked
- * user space pages).
- */
-dma_cookie_t
-dma_async_memcpy_buf_to_buf(struct dma_chan *chan, void *dest,
- void *src, size_t len)
-{
- return dma_async_memcpy_pg_to_pg(chan, virt_to_page(dest),
- (unsigned long) dest & ~PAGE_MASK,
- virt_to_page(src),
- (unsigned long) src & ~PAGE_MASK, len);
-}
-EXPORT_SYMBOL(dma_async_memcpy_buf_to_buf);
-
-/**
- * dma_async_memcpy_buf_to_pg - offloaded copy from address to page
- * @chan: DMA channel to offload copy to
- * @page: destination page
- * @offset: offset in page to copy to
- * @kdata: source address (virtual)
- * @len: length
- *
- * Both @page/@offset and @kdata must be mappable to a bus address according
- * to the DMA mapping API rules for streaming mappings.
- * Both @page/@offset and @kdata must stay memory resident (kernel memory or
- * locked user space pages)
- */
-dma_cookie_t
-dma_async_memcpy_buf_to_pg(struct dma_chan *chan, struct page *page,
- unsigned int offset, void *kdata, size_t len)
-{
- return dma_async_memcpy_pg_to_pg(chan, page, offset,
- virt_to_page(kdata),
- (unsigned long) kdata & ~PAGE_MASK, len);
-}
-EXPORT_SYMBOL(dma_async_memcpy_buf_to_pg);
-
void dma_async_tx_descriptor_init(struct dma_async_tx_descriptor *tx,
struct dma_chan *chan)
{
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index 1a49c777607c..97fa394ca855 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -1175,7 +1175,6 @@ int ioat1_dma_probe(struct ioatdma_device *device, int dca)
err = ioat_probe(device);
if (err)
return err;
- ioat_set_tcp_copy_break(4096);
err = ioat_register(device);
if (err)
return err;
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index 11fb877ddca9..664ec9cbd651 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -214,13 +214,6 @@ __dump_desc_dbg(struct ioat_chan_common *chan, struct ioat_dma_descriptor *hw,
#define dump_desc_dbg(c, d) \
({ if (d) __dump_desc_dbg(&c->base, d->hw, &d->txd, desc_id(d)); 0; })
-static inline void ioat_set_tcp_copy_break(unsigned long copybreak)
-{
- #ifdef CONFIG_NET_DMA
- sysctl_tcp_dma_copybreak = copybreak;
- #endif
-}
-
static inline struct ioat_chan_common *
ioat_chan_by_index(struct ioatdma_device *device, int index)
{
diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c
index 5d3affe7e976..31e8098e444f 100644
--- a/drivers/dma/ioat/dma_v2.c
+++ b/drivers/dma/ioat/dma_v2.c
@@ -900,7 +900,6 @@ int ioat2_dma_probe(struct ioatdma_device *device, int dca)
err = ioat_probe(device);
if (err)
return err;
- ioat_set_tcp_copy_break(2048);
list_for_each_entry(c, &dma->channels, device_node) {
chan = to_chan_common(c);
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index 820817e97e62..4bb81346bee2 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -1652,7 +1652,6 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
err = ioat_probe(device);
if (err)
return err;
- ioat_set_tcp_copy_break(262144);
list_for_each_entry(c, &dma->channels, device_node) {
chan = to_chan_common(c);
diff --git a/drivers/dma/iovlock.c b/drivers/dma/iovlock.c
deleted file mode 100644
index bb48a57c2fc1..000000000000
--- a/drivers/dma/iovlock.c
+++ /dev/null
@@ -1,280 +0,0 @@
-/*
- * Copyright(c) 2004 - 2006 Intel Corporation. All rights reserved.
- * Portions based on net/core/datagram.c and copyrighted by their authors.
- *
- * 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.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA 02111-1307, USA.
- *
- * The full GNU General Public License is included in this distribution in the
- * file called COPYING.
- */
-
-/*
- * This code allows the net stack to make use of a DMA engine for
- * skb to iovec copies.
- */
-
-#include <linux/dmaengine.h>
-#include <linux/pagemap.h>
-#include <linux/slab.h>
-#include <net/tcp.h> /* for memcpy_toiovec */
-#include <asm/io.h>
-#include <asm/uaccess.h>
-
-static int num_pages_spanned(struct iovec *iov)
-{
- return
- ((PAGE_ALIGN((unsigned long)iov->iov_base + iov->iov_len) -
- ((unsigned long)iov->iov_base & PAGE_MASK)) >> PAGE_SHIFT);
-}
-
-/*
- * Pin down all the iovec pages needed for len bytes.
- * Return a struct dma_pinned_list to keep track of pages pinned down.
- *
- * We are allocating a single chunk of memory, and then carving it up into
- * 3 sections, the latter 2 whose size depends on the number of iovecs and the
- * total number of pages, respectively.
- */
-struct dma_pinned_list *dma_pin_iovec_pages(struct iovec *iov, size_t len)
-{
- struct dma_pinned_list *local_list;
- struct page **pages;
- int i;
- int ret;
- int nr_iovecs = 0;
- int iovec_len_used = 0;
- int iovec_pages_used = 0;
-
- /* don't pin down non-user-based iovecs */
- if (segment_eq(get_fs(), KERNEL_DS))
- return NULL;
-
- /* determine how many iovecs/pages there are, up front */
- do {
- iovec_len_used += iov[nr_iovecs].iov_len;
- iovec_pages_used += num_pages_spanned(&iov[nr_iovecs]);
- nr_iovecs++;
- } while (iovec_len_used < len);
-
- /* single kmalloc for pinned list, page_list[], and the page arrays */
- local_list = kmalloc(sizeof(*local_list)
- + (nr_iovecs * sizeof (struct dma_page_list))
- + (iovec_pages_used * sizeof (struct page*)), GFP_KERNEL);
- if (!local_list)
- goto out;
-
- /* list of pages starts right after the page list array */
- pages = (struct page **) &local_list->page_list[nr_iovecs];
-
- local_list->nr_iovecs = 0;
-
- for (i = 0; i < nr_iovecs; i++) {
- struct dma_page_list *page_list = &local_list->page_list[i];
-
- len -= iov[i].iov_len;
-
- if (!access_ok(VERIFY_WRITE, iov[i].iov_base, iov[i].iov_len))
- goto unpin;
-
- page_list->nr_pages = num_pages_spanned(&iov[i]);
- page_list->base_address = iov[i].iov_base;
-
- page_list->pages = pages;
- pages += page_list->nr_pages;
-
- /* pin pages down */
- down_read(¤t->mm->mmap_sem);
- ret = get_user_pages(
- current,
- current->mm,
- (unsigned long) iov[i].iov_base,
- page_list->nr_pages,
- 1, /* write */
- 0, /* force */
- page_list->pages,
- NULL);
- up_read(¤t->mm->mmap_sem);
-
- if (ret != page_list->nr_pages)
- goto unpin;
-
- local_list->nr_iovecs = i + 1;
- }
-
- return local_list;
-
-unpin:
- dma_unpin_iovec_pages(local_list);
-out:
- return NULL;
-}
-
-void dma_unpin_iovec_pages(struct dma_pinned_list *pinned_list)
-{
- int i, j;
-
- if (!pinned_list)
- return;
-
- for (i = 0; i < pinned_list->nr_iovecs; i++) {
- struct dma_page_list *page_list = &pinned_list->page_list[i];
- for (j = 0; j < page_list->nr_pages; j++) {
- set_page_dirty_lock(page_list->pages[j]);
- page_cache_release(page_list->pages[j]);
- }
- }
-
- kfree(pinned_list);
-}
-
-
-/*
- * We have already pinned down the pages we will be using in the iovecs.
- * Each entry in iov array has corresponding entry in pinned_list->page_list.
- * Using array indexing to keep iov[] and page_list[] in sync.
- * Initial elements in iov array's iov->iov_len will be 0 if already copied into
- * by another call.
- * iov array length remaining guaranteed to be bigger than len.
- */
-dma_cookie_t dma_memcpy_to_iovec(struct dma_chan *chan, struct iovec *iov,
- struct dma_pinned_list *pinned_list, unsigned char *kdata, size_t len)
-{
- int iov_byte_offset;
- int copy;
- dma_cookie_t dma_cookie = 0;
- int iovec_idx;
- int page_idx;
-
- if (!chan)
- return memcpy_toiovec(iov, kdata, len);
-
- iovec_idx = 0;
- while (iovec_idx < pinned_list->nr_iovecs) {
- struct dma_page_list *page_list;
-
- /* skip already used-up iovecs */
- while (!iov[iovec_idx].iov_len)
- iovec_idx++;
-
- page_list = &pinned_list->page_list[iovec_idx];
-
- iov_byte_offset = ((unsigned long)iov[iovec_idx].iov_base & ~PAGE_MASK);
- page_idx = (((unsigned long)iov[iovec_idx].iov_base & PAGE_MASK)
- - ((unsigned long)page_list->base_address & PAGE_MASK)) >> PAGE_SHIFT;
-
- /* break up copies to not cross page boundary */
- while (iov[iovec_idx].iov_len) {
- copy = min_t(int, PAGE_SIZE - iov_byte_offset, len);
- copy = min_t(int, copy, iov[iovec_idx].iov_len);
-
- dma_cookie = dma_async_memcpy_buf_to_pg(chan,
- page_list->pages[page_idx],
- iov_byte_offset,
- kdata,
- copy);
- /* poll for a descriptor slot */
- if (unlikely(dma_cookie < 0)) {
- dma_async_issue_pending(chan);
- continue;
- }
-
- len -= copy;
- iov[iovec_idx].iov_len -= copy;
- iov[iovec_idx].iov_base += copy;
-
- if (!len)
- return dma_cookie;
-
- kdata += copy;
- iov_byte_offset = 0;
- page_idx++;
- }
- iovec_idx++;
- }
-
- /* really bad if we ever run out of iovecs */
- BUG();
- return -EFAULT;
-}
-
-dma_cookie_t dma_memcpy_pg_to_iovec(struct dma_chan *chan, struct iovec *iov,
- struct dma_pinned_list *pinned_list, struct page *page,
- unsigned int offset, size_t len)
-{
- int iov_byte_offset;
- int copy;
- dma_cookie_t dma_cookie = 0;
- int iovec_idx;
- int page_idx;
- int err;
-
- /* this needs as-yet-unimplemented buf-to-buff, so punt. */
- /* TODO: use dma for this */
- if (!chan || !pinned_list) {
- u8 *vaddr = kmap(page);
- err = memcpy_toiovec(iov, vaddr + offset, len);
- kunmap(page);
- return err;
- }
-
- iovec_idx = 0;
- while (iovec_idx < pinned_list->nr_iovecs) {
- struct dma_page_list *page_list;
-
- /* skip already used-up iovecs */
- while (!iov[iovec_idx].iov_len)
- iovec_idx++;
-
- page_list = &pinned_list->page_list[iovec_idx];
-
- iov_byte_offset = ((unsigned long)iov[iovec_idx].iov_base & ~PAGE_MASK);
- page_idx = (((unsigned long)iov[iovec_idx].iov_base & PAGE_MASK)
- - ((unsigned long)page_list->base_address & PAGE_MASK)) >> PAGE_SHIFT;
-
- /* break up copies to not cross page boundary */
- while (iov[iovec_idx].iov_len) {
- copy = min_t(int, PAGE_SIZE - iov_byte_offset, len);
- copy = min_t(int, copy, iov[iovec_idx].iov_len);
-
- dma_cookie = dma_async_memcpy_pg_to_pg(chan,
- page_list->pages[page_idx],
- iov_byte_offset,
- page,
- offset,
- copy);
- /* poll for a descriptor slot */
- if (unlikely(dma_cookie < 0)) {
- dma_async_issue_pending(chan);
- continue;
- }
-
- len -= copy;
- iov[iovec_idx].iov_len -= copy;
- iov[iovec_idx].iov_base += copy;
-
- if (!len)
- return dma_cookie;
-
- offset += copy;
- iov_byte_offset = 0;
- page_idx++;
- }
- iovec_idx++;
- }
-
- /* really bad if we ever run out of iovecs */
- BUG();
- return -EFAULT;
-}
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index 41cf0c399288..890545871af0 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -875,18 +875,6 @@ static inline void dmaengine_put(void)
}
#endif
-#ifdef CONFIG_NET_DMA
-#define net_dmaengine_get() dmaengine_get()
-#define net_dmaengine_put() dmaengine_put()
-#else
-static inline void net_dmaengine_get(void)
-{
-}
-static inline void net_dmaengine_put(void)
-{
-}
-#endif
-
#ifdef CONFIG_ASYNC_TX_DMA
#define async_dmaengine_get() dmaengine_get()
#define async_dmaengine_put() dmaengine_put()
@@ -908,16 +896,8 @@ async_dma_find_channel(enum dma_transaction_type type)
return NULL;
}
#endif /* CONFIG_ASYNC_TX_DMA */
-
-dma_cookie_t dma_async_memcpy_buf_to_buf(struct dma_chan *chan,
- void *dest, void *src, size_t len);
-dma_cookie_t dma_async_memcpy_buf_to_pg(struct dma_chan *chan,
- struct page *page, unsigned int offset, void *kdata, size_t len);
-dma_cookie_t dma_async_memcpy_pg_to_pg(struct dma_chan *chan,
- struct page *dest_pg, unsigned int dest_off, struct page *src_pg,
- unsigned int src_off, size_t len);
void dma_async_tx_descriptor_init(struct dma_async_tx_descriptor *tx,
- struct dma_chan *chan);
+ struct dma_chan *chan);
static inline void async_tx_ack(struct dma_async_tx_descriptor *tx)
{
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index bec1cc7d5e3c..ac4f84dfa84b 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -28,7 +28,6 @@
#include <linux/textsearch.h>
#include <net/checksum.h>
#include <linux/rcupdate.h>
-#include <linux/dmaengine.h>
#include <linux/hrtimer.h>
#include <linux/dma-mapping.h>
#include <linux/netdev_features.h>
@@ -496,11 +495,8 @@ struct sk_buff {
/* 6/8 bit hole (depending on ndisc_nodetype presence) */
kmemcheck_bitfield_end(flags2);
-#if defined CONFIG_NET_DMA || defined CONFIG_NET_RX_BUSY_POLL
- union {
- unsigned int napi_id;
- dma_cookie_t dma_cookie;
- };
+#ifdef CONFIG_NET_RX_BUSY_POLL
+ unsigned int napi_id;
#endif
#ifdef CONFIG_NETWORK_SECMARK
__u32 secmark;
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index d68633452d9b..26f16021ce1d 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -19,7 +19,6 @@
#include <linux/skbuff.h>
-#include <linux/dmaengine.h>
#include <net/sock.h>
#include <net/inet_connection_sock.h>
#include <net/inet_timewait_sock.h>
@@ -169,13 +168,6 @@ struct tcp_sock {
struct iovec *iov;
int memory;
int len;
-#ifdef CONFIG_NET_DMA
- /* members for async copy */
- struct dma_chan *dma_chan;
- int wakeup;
- struct dma_pinned_list *pinned_list;
- dma_cookie_t dma_cookie;
-#endif
} ucopy;
u32 snd_wl1; /* Sequence for window update */
diff --git a/include/net/netdma.h b/include/net/netdma.h
deleted file mode 100644
index 8ba8ce284eeb..000000000000
--- a/include/net/netdma.h
+++ /dev/null
@@ -1,32 +0,0 @@
-/*
- * Copyright(c) 2004 - 2006 Intel Corporation. All rights reserved.
- *
- * 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.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA 02111-1307, USA.
- *
- * The full GNU General Public License is included in this distribution in the
- * file called COPYING.
- */
-#ifndef NETDMA_H
-#define NETDMA_H
-#ifdef CONFIG_NET_DMA
-#include <linux/dmaengine.h>
-#include <linux/skbuff.h>
-
-int dma_skb_copy_datagram_iovec(struct dma_chan* chan,
- struct sk_buff *skb, int offset, struct iovec *to,
- size_t len, struct dma_pinned_list *pinned_list);
-
-#endif /* CONFIG_NET_DMA */
-#endif /* NETDMA_H */
diff --git a/include/net/sock.h b/include/net/sock.h
index e3a18ff0c38b..9d5f716e921e 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -231,7 +231,6 @@ struct cg_proto;
* @sk_receive_queue: incoming packets
* @sk_wmem_alloc: transmit queue bytes committed
* @sk_write_queue: Packet sending queue
- * @sk_async_wait_queue: DMA copied packets
* @sk_omem_alloc: "o" is "option" or "other"
* @sk_wmem_queued: persistent queue size
* @sk_forward_alloc: space allocated forward
@@ -354,10 +353,6 @@ struct sock {
struct sk_filter __rcu *sk_filter;
struct socket_wq __rcu *sk_wq;
-#ifdef CONFIG_NET_DMA
- struct sk_buff_head sk_async_wait_queue;
-#endif
-
#ifdef CONFIG_XFRM
struct xfrm_policy *sk_policy[2];
#endif
@@ -2200,27 +2195,15 @@ void sock_tx_timestamp(struct sock *sk, __u8 *tx_flags);
* sk_eat_skb - Release a skb if it is no longer needed
* @sk: socket to eat this skb from
* @skb: socket buffer to eat
- * @copied_early: flag indicating whether DMA operations copied this data early
*
* This routine must be called with interrupts disabled or with the socket
* locked so that the sk_buff queue operation is ok.
*/
-#ifdef CONFIG_NET_DMA
-static inline void sk_eat_skb(struct sock *sk, struct sk_buff *skb, bool copied_early)
-{
- __skb_unlink(skb, &sk->sk_receive_queue);
- if (!copied_early)
- __kfree_skb(skb);
- else
- __skb_queue_tail(&sk->sk_async_wait_queue, skb);
-}
-#else
-static inline void sk_eat_skb(struct sock *sk, struct sk_buff *skb, bool copied_early)
+static inline void sk_eat_skb(struct sock *sk, struct sk_buff *skb)
{
__skb_unlink(skb, &sk->sk_receive_queue);
__kfree_skb(skb);
}
-#endif
static inline
struct net *sock_net(const struct sock *sk)
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 70e55d200610..084c163e9d40 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -27,7 +27,6 @@
#include <linux/cache.h>
#include <linux/percpu.h>
#include <linux/skbuff.h>
-#include <linux/dmaengine.h>
#include <linux/crypto.h>
#include <linux/cryptohash.h>
#include <linux/kref.h>
@@ -267,7 +266,6 @@ extern int sysctl_tcp_adv_win_scale;
extern int sysctl_tcp_tw_reuse;
extern int sysctl_tcp_frto;
extern int sysctl_tcp_low_latency;
-extern int sysctl_tcp_dma_copybreak;
extern int sysctl_tcp_nometrics_save;
extern int sysctl_tcp_moderate_rcvbuf;
extern int sysctl_tcp_tso_win_divisor;
@@ -1032,12 +1030,6 @@ static inline void tcp_prequeue_init(struct tcp_sock *tp)
tp->ucopy.len = 0;
tp->ucopy.memory = 0;
skb_queue_head_init(&tp->ucopy.prequeue);
-#ifdef CONFIG_NET_DMA
- tp->ucopy.dma_chan = NULL;
- tp->ucopy.wakeup = 0;
- tp->ucopy.pinned_list = NULL;
- tp->ucopy.dma_cookie = 0;
-#endif
}
bool tcp_prequeue(struct sock *sk, struct sk_buff *skb);
diff --git a/kernel/sysctl_binary.c b/kernel/sysctl_binary.c
index 653cbbd9e7ad..d457005acedf 100644
--- a/kernel/sysctl_binary.c
+++ b/kernel/sysctl_binary.c
@@ -390,7 +390,6 @@ static const struct bin_table bin_net_ipv4_table[] = {
{ CTL_INT, NET_TCP_MTU_PROBING, "tcp_mtu_probing" },
{ CTL_INT, NET_TCP_BASE_MSS, "tcp_base_mss" },
{ CTL_INT, NET_IPV4_TCP_WORKAROUND_SIGNED_WINDOWS, "tcp_workaround_signed_windows" },
- { CTL_INT, NET_TCP_DMA_COPYBREAK, "tcp_dma_copybreak" },
{ CTL_INT, NET_TCP_SLOW_START_AFTER_IDLE, "tcp_slow_start_after_idle" },
{ CTL_INT, NET_CIPSOV4_CACHE_ENABLE, "cipso_cache_enable" },
{ CTL_INT, NET_CIPSOV4_CACHE_BUCKET_SIZE, "cipso_cache_bucket_size" },
diff --git a/net/core/Makefile b/net/core/Makefile
index b33b996f5dd6..5f98e5983bd3 100644
--- a/net/core/Makefile
+++ b/net/core/Makefile
@@ -16,7 +16,6 @@ obj-y += net-sysfs.o
obj-$(CONFIG_PROC_FS) += net-procfs.o
obj-$(CONFIG_NET_PKTGEN) += pktgen.o
obj-$(CONFIG_NETPOLL) += netpoll.o
-obj-$(CONFIG_NET_DMA) += user_dma.o
obj-$(CONFIG_FIB_RULES) += fib_rules.o
obj-$(CONFIG_TRACEPOINTS) += net-traces.o
obj-$(CONFIG_NET_DROP_MONITOR) += drop_monitor.o
diff --git a/net/core/dev.c b/net/core/dev.c
index ba3b7ea5ebb3..677a5a4dcca7 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1262,7 +1262,6 @@ static int __dev_open(struct net_device *dev)
clear_bit(__LINK_STATE_START, &dev->state);
else {
dev->flags |= IFF_UP;
- net_dmaengine_get();
dev_set_rx_mode(dev);
dev_activate(dev);
add_device_randomness(dev->dev_addr, dev->addr_len);
@@ -1338,7 +1337,6 @@ static int __dev_close_many(struct list_head *head)
ops->ndo_stop(dev);
dev->flags &= ~IFF_UP;
- net_dmaengine_put();
}
return 0;
@@ -4362,14 +4360,6 @@ static void net_rx_action(struct softirq_action *h)
out:
net_rps_action_and_irq_enable(sd);
-#ifdef CONFIG_NET_DMA
- /*
- * There may not be any more sk_buffs coming right now, so push
- * any pending DMA copies to hardware
- */
- dma_issue_pending_all();
-#endif
-
return;
softnet_break:
diff --git a/net/core/sock.c b/net/core/sock.c
index ab20ed9b0f31..411dab3a5726 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -1461,9 +1461,6 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority)
atomic_set(&newsk->sk_omem_alloc, 0);
skb_queue_head_init(&newsk->sk_receive_queue);
skb_queue_head_init(&newsk->sk_write_queue);
-#ifdef CONFIG_NET_DMA
- skb_queue_head_init(&newsk->sk_async_wait_queue);
-#endif
spin_lock_init(&newsk->sk_dst_lock);
rwlock_init(&newsk->sk_callback_lock);
@@ -2290,9 +2287,6 @@ void sock_init_data(struct socket *sock, struct sock *sk)
skb_queue_head_init(&sk->sk_receive_queue);
skb_queue_head_init(&sk->sk_write_queue);
skb_queue_head_init(&sk->sk_error_queue);
-#ifdef CONFIG_NET_DMA
- skb_queue_head_init(&sk->sk_async_wait_queue);
-#endif
sk->sk_send_head = NULL;
diff --git a/net/core/user_dma.c b/net/core/user_dma.c
deleted file mode 100644
index 1b5fefdb8198..000000000000
--- a/net/core/user_dma.c
+++ /dev/null
@@ -1,131 +0,0 @@
-/*
- * Copyright(c) 2004 - 2006 Intel Corporation. All rights reserved.
- * Portions based on net/core/datagram.c and copyrighted by their authors.
- *
- * 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.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA 02111-1307, USA.
- *
- * The full GNU General Public License is included in this distribution in the
- * file called COPYING.
- */
-
-/*
- * This code allows the net stack to make use of a DMA engine for
- * skb to iovec copies.
- */
-
-#include <linux/dmaengine.h>
-#include <linux/socket.h>
-#include <linux/export.h>
-#include <net/tcp.h>
-#include <net/netdma.h>
-
-#define NET_DMA_DEFAULT_COPYBREAK 4096
-
-int sysctl_tcp_dma_copybreak = NET_DMA_DEFAULT_COPYBREAK;
-EXPORT_SYMBOL(sysctl_tcp_dma_copybreak);
-
-/**
- * dma_skb_copy_datagram_iovec - Copy a datagram to an iovec.
- * @skb - buffer to copy
- * @offset - offset in the buffer to start copying from
- * @iovec - io vector to copy to
- * @len - amount of data to copy from buffer to iovec
- * @pinned_list - locked iovec buffer data
- *
- * Note: the iovec is modified during the copy.
- */
-int dma_skb_copy_datagram_iovec(struct dma_chan *chan,
- struct sk_buff *skb, int offset, struct iovec *to,
- size_t len, struct dma_pinned_list *pinned_list)
-{
- int start = skb_headlen(skb);
- int i, copy = start - offset;
- struct sk_buff *frag_iter;
- dma_cookie_t cookie = 0;
-
- /* Copy header. */
- if (copy > 0) {
- if (copy > len)
- copy = len;
- cookie = dma_memcpy_to_iovec(chan, to, pinned_list,
- skb->data + offset, copy);
- if (cookie < 0)
- goto fault;
- len -= copy;
- if (len == 0)
- goto end;
- offset += copy;
- }
-
- /* Copy paged appendix. Hmm... why does this look so complicated? */
- for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
- int end;
- const skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
-
- WARN_ON(start > offset + len);
-
- end = start + skb_frag_size(frag);
- copy = end - offset;
- if (copy > 0) {
- struct page *page = skb_frag_page(frag);
-
- if (copy > len)
- copy = len;
-
- cookie = dma_memcpy_pg_to_iovec(chan, to, pinned_list, page,
- frag->page_offset + offset - start, copy);
- if (cookie < 0)
- goto fault;
- len -= copy;
- if (len == 0)
- goto end;
- offset += copy;
- }
- start = end;
- }
-
- skb_walk_frags(skb, frag_iter) {
- int end;
-
- WARN_ON(start > offset + len);
-
- end = start + frag_iter->len;
- copy = end - offset;
- if (copy > 0) {
- if (copy > len)
- copy = len;
- cookie = dma_skb_copy_datagram_iovec(chan, frag_iter,
- offset - start,
- to, copy,
- pinned_list);
- if (cookie < 0)
- goto fault;
- len -= copy;
- if (len == 0)
- goto end;
- offset += copy;
- }
- start = end;
- }
-
-end:
- if (!len) {
- skb->dma_cookie = cookie;
- return cookie;
- }
-
-fault:
- return -EFAULT;
-}
diff --git a/net/dccp/proto.c b/net/dccp/proto.c
index eb892b4f4814..f9076f295b13 100644
--- a/net/dccp/proto.c
+++ b/net/dccp/proto.c
@@ -848,7 +848,7 @@ int dccp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
default:
dccp_pr_debug("packet_type=%s\n",
dccp_packet_name(dh->dccph_type));
- sk_eat_skb(sk, skb, false);
+ sk_eat_skb(sk, skb);
}
verify_sock_status:
if (sock_flag(sk, SOCK_DONE)) {
@@ -905,7 +905,7 @@ verify_sock_status:
len = skb->len;
found_fin_ok:
if (!(flags & MSG_PEEK))
- sk_eat_skb(sk, skb, false);
+ sk_eat_skb(sk, skb);
break;
} while (1);
out:
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index 3d69ec8dac57..79a90b92e12d 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -642,15 +642,6 @@ static struct ctl_table ipv4_table[] = {
.mode = 0644,
.proc_handler = proc_dointvec
},
-#ifdef CONFIG_NET_DMA
- {
- .procname = "tcp_dma_copybreak",
- .data = &sysctl_tcp_dma_copybreak,
- .maxlen = sizeof(int),
- .mode = 0644,
- .proc_handler = proc_dointvec
- },
-#endif
{
.procname = "tcp_slow_start_after_idle",
.data = &sysctl_tcp_slow_start_after_idle,
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index c4638e6f0238..8dc913dfbaef 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -274,7 +274,6 @@
#include <net/tcp.h>
#include <net/xfrm.h>
#include <net/ip.h>
-#include <net/netdma.h>
#include <net/sock.h>
#include <asm/uaccess.h>
@@ -1409,39 +1408,6 @@ static void tcp_prequeue_process(struct sock *sk)
tp->ucopy.memory = 0;
}
-#ifdef CONFIG_NET_DMA
-static void tcp_service_net_dma(struct sock *sk, bool wait)
-{
- dma_cookie_t done, used;
- dma_cookie_t last_issued;
- struct tcp_sock *tp = tcp_sk(sk);
-
- if (!tp->ucopy.dma_chan)
- return;
-
- last_issued = tp->ucopy.dma_cookie;
- dma_async_issue_pending(tp->ucopy.dma_chan);
-
- do {
- if (dma_async_is_tx_complete(tp->ucopy.dma_chan,
- last_issued, &done,
- &used) == DMA_COMPLETE) {
- /* Safe to free early-copied skbs now */
- __skb_queue_purge(&sk->sk_async_wait_queue);
- break;
- } else {
- struct sk_buff *skb;
- while ((skb = skb_peek(&sk->sk_async_wait_queue)) &&
- (dma_async_is_complete(skb->dma_cookie, done,
- used) == DMA_COMPLETE)) {
- __skb_dequeue(&sk->sk_async_wait_queue);
- kfree_skb(skb);
- }
- }
- } while (wait);
-}
-#endif
-
static struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off)
{
struct sk_buff *skb;
@@ -1459,7 +1425,7 @@ static struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off)
* splitted a fat GRO packet, while we released socket lock
* in skb_splice_bits()
*/
- sk_eat_skb(sk, skb, false);
+ sk_eat_skb(sk, skb);
}
return NULL;
}
@@ -1525,11 +1491,11 @@ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc,
continue;
}
if (tcp_hdr(skb)->fin) {
- sk_eat_skb(sk, skb, false);
+ sk_eat_skb(sk, skb);
++seq;
break;
}
- sk_eat_skb(sk, skb, false);
+ sk_eat_skb(sk, skb);
if (!desc->count)
break;
tp->copied_seq = seq;
@@ -1567,7 +1533,6 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
int target; /* Read at least this many bytes */
long timeo;
struct task_struct *user_recv = NULL;
- bool copied_early = false;
struct sk_buff *skb;
u32 urg_hole = 0;
@@ -1610,28 +1575,6 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);
-#ifdef CONFIG_NET_DMA
- tp->ucopy.dma_chan = NULL;
- preempt_disable();
- skb = skb_peek_tail(&sk->sk_receive_queue);
- {
- int available = 0;
-
- if (skb)
- available = TCP_SKB_CB(skb)->seq + skb->len - (*seq);
- if ((available < target) &&
- (len > sysctl_tcp_dma_copybreak) && !(flags & MSG_PEEK) &&
- !sysctl_tcp_low_latency &&
- net_dma_find_channel()) {
- preempt_enable_no_resched();
- tp->ucopy.pinned_list =
- dma_pin_iovec_pages(msg->msg_iov, len);
- } else {
- preempt_enable_no_resched();
- }
- }
-#endif
-
do {
u32 offset;
@@ -1762,16 +1705,6 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
/* __ Set realtime policy in scheduler __ */
}
-#ifdef CONFIG_NET_DMA
- if (tp->ucopy.dma_chan) {
- if (tp->rcv_wnd == 0 &&
- !skb_queue_empty(&sk->sk_async_wait_queue)) {
- tcp_service_net_dma(sk, true);
- tcp_cleanup_rbuf(sk, copied);
- } else
- dma_async_issue_pending(tp->ucopy.dma_chan);
- }
-#endif
if (copied >= target) {
/* Do not sleep, just process backlog. */
release_sock(sk);
@@ -1779,11 +1712,6 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
} else
sk_wait_data(sk, &timeo);
-#ifdef CONFIG_NET_DMA
- tcp_service_net_dma(sk, false); /* Don't block */
- tp->ucopy.wakeup = 0;
-#endif
-
if (user_recv) {
int chunk;
@@ -1841,43 +1769,13 @@ do_prequeue:
}
if (!(flags & MSG_TRUNC)) {
-#ifdef CONFIG_NET_DMA
- if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
- tp->ucopy.dma_chan = net_dma_find_channel();
-
- if (tp->ucopy.dma_chan) {
- tp->ucopy.dma_cookie = dma_skb_copy_datagram_iovec(
- tp->ucopy.dma_chan, skb, offset,
- msg->msg_iov, used,
- tp->ucopy.pinned_list);
-
- if (tp->ucopy.dma_cookie < 0) {
-
- pr_alert("%s: dma_cookie < 0\n",
- __func__);
-
- /* Exception. Bailout! */
- if (!copied)
- copied = -EFAULT;
- break;
- }
-
- dma_async_issue_pending(tp->ucopy.dma_chan);
-
- if ((offset + used) == skb->len)
- copied_early = true;
-
- } else
-#endif
- {
- err = skb_copy_datagram_iovec(skb, offset,
- msg->msg_iov, used);
- if (err) {
- /* Exception. Bailout! */
- if (!copied)
- copied = -EFAULT;
- break;
- }
+ err = skb_copy_datagram_iovec(skb, offset,
+ msg->msg_iov, used);
+ if (err) {
+ /* Exception. Bailout! */
+ if (!copied)
+ copied = -EFAULT;
+ break;
}
}
@@ -1897,19 +1795,15 @@ skip_copy:
if (tcp_hdr(skb)->fin)
goto found_fin_ok;
- if (!(flags & MSG_PEEK)) {
- sk_eat_skb(sk, skb, copied_early);
- copied_early = false;
- }
+ if (!(flags & MSG_PEEK))
+ sk_eat_skb(sk, skb);
continue;
found_fin_ok:
/* Process the FIN. */
++*seq;
- if (!(flags & MSG_PEEK)) {
- sk_eat_skb(sk, skb, copied_early);
- copied_early = false;
- }
+ if (!(flags & MSG_PEEK))
+ sk_eat_skb(sk, skb);
break;
} while (len > 0);
@@ -1932,16 +1826,6 @@ skip_copy:
tp->ucopy.len = 0;
}
-#ifdef CONFIG_NET_DMA
- tcp_service_net_dma(sk, true); /* Wait for queue to drain */
- tp->ucopy.dma_chan = NULL;
-
- if (tp->ucopy.pinned_list) {
- dma_unpin_iovec_pages(tp->ucopy.pinned_list);
- tp->ucopy.pinned_list = NULL;
- }
-#endif
-
/* According to UNIX98, msg_name/msg_namelen are ignored
* on connected socket. I was just happy when found this 8) --ANK
*/
@@ -2285,9 +2169,6 @@ int tcp_disconnect(struct sock *sk, int flags)
__skb_queue_purge(&sk->sk_receive_queue);
tcp_write_queue_purge(sk);
__skb_queue_purge(&tp->out_of_order_queue);
-#ifdef CONFIG_NET_DMA
- __skb_queue_purge(&sk->sk_async_wait_queue);
-#endif
inet->inet_dport = 0;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index c53b7f35c51d..33ef18e550c5 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -73,7 +73,6 @@
#include <net/inet_common.h>
#include <linux/ipsec.h>
#include <asm/unaligned.h>
-#include <net/netdma.h>
int sysctl_tcp_timestamps __read_mostly = 1;
int sysctl_tcp_window_scaling __read_mostly = 1;
@@ -4967,53 +4966,6 @@ static inline bool tcp_checksum_complete_user(struct sock *sk,
__tcp_checksum_complete_user(sk, skb);
}
-#ifdef CONFIG_NET_DMA
-static bool tcp_dma_try_early_copy(struct sock *sk, struct sk_buff *skb,
- int hlen)
-{
- struct tcp_sock *tp = tcp_sk(sk);
- int chunk = skb->len - hlen;
- int dma_cookie;
- bool copied_early = false;
-
- if (tp->ucopy.wakeup)
- return false;
-
- if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
- tp->ucopy.dma_chan = net_dma_find_channel();
-
- if (tp->ucopy.dma_chan && skb_csum_unnecessary(skb)) {
-
- dma_cookie = dma_skb_copy_datagram_iovec(tp->ucopy.dma_chan,
- skb, hlen,
- tp->ucopy.iov, chunk,
- tp->ucopy.pinned_list);
-
- if (dma_cookie < 0)
- goto out;
-
- tp->ucopy.dma_cookie = dma_cookie;
- copied_early = true;
-
- tp->ucopy.len -= chunk;
- tp->copied_seq += chunk;
- tcp_rcv_space_adjust(sk);
-
- if ((tp->ucopy.len == 0) ||
- (tcp_flag_word(tcp_hdr(skb)) & TCP_FLAG_PSH) ||
- (atomic_read(&sk->sk_rmem_alloc) > (sk->sk_rcvbuf >> 1))) {
- tp->ucopy.wakeup = 1;
- sk->sk_data_ready(sk, 0);
- }
- } else if (chunk > 0) {
- tp->ucopy.wakeup = 1;
- sk->sk_data_ready(sk, 0);
- }
-out:
- return copied_early;
-}
-#endif /* CONFIG_NET_DMA */
-
/* Does PAWS and seqno based validation of an incoming segment, flags will
* play significant role here.
*/
@@ -5198,14 +5150,6 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
if (tp->copied_seq == tp->rcv_nxt &&
len - tcp_header_len <= tp->ucopy.len) {
-#ifdef CONFIG_NET_DMA
- if (tp->ucopy.task == current &&
- sock_owned_by_user(sk) &&
- tcp_dma_try_early_copy(sk, skb, tcp_header_len)) {
- copied_early = 1;
- eaten = 1;
- }
-#endif
if (tp->ucopy.task == current &&
sock_owned_by_user(sk) && !copied_early) {
__set_current_state(TASK_RUNNING);
@@ -5271,11 +5215,6 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
if (!copied_early || tp->rcv_nxt != tp->rcv_wup)
__tcp_ack_snd_check(sk, 0);
no_ack:
-#ifdef CONFIG_NET_DMA
- if (copied_early)
- __skb_queue_tail(&sk->sk_async_wait_queue, skb);
- else
-#endif
if (eaten)
kfree_skb_partial(skb, fragstolen);
sk->sk_data_ready(sk, 0);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 59a6f8b90cd9..dc92ba9d0350 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -72,7 +72,6 @@
#include <net/inet_common.h>
#include <net/timewait_sock.h>
#include <net/xfrm.h>
-#include <net/netdma.h>
#include <net/secure_seq.h>
#include <net/tcp_memcontrol.h>
#include <net/busy_poll.h>
@@ -2000,18 +1999,8 @@ process:
bh_lock_sock_nested(sk);
ret = 0;
if (!sock_owned_by_user(sk)) {
-#ifdef CONFIG_NET_DMA
- struct tcp_sock *tp = tcp_sk(sk);
- if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
- tp->ucopy.dma_chan = net_dma_find_channel();
- if (tp->ucopy.dma_chan)
+ if (!tcp_prequeue(sk, skb))
ret = tcp_v4_do_rcv(sk, skb);
- else
-#endif
- {
- if (!tcp_prequeue(sk, skb))
- ret = tcp_v4_do_rcv(sk, skb);
- }
} else if (unlikely(sk_add_backlog(sk, skb,
sk->sk_rcvbuf + sk->sk_sndbuf))) {
bh_unlock_sock(sk);
@@ -2170,11 +2159,6 @@ void tcp_v4_destroy_sock(struct sock *sk)
}
#endif
-#ifdef CONFIG_NET_DMA
- /* Cleans up our sk_async_wait_queue */
- __skb_queue_purge(&sk->sk_async_wait_queue);
-#endif
-
/* Clean prequeue, it must be empty really */
__skb_queue_purge(&tp->ucopy.prequeue);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 0740f93a114a..e27972590379 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -59,7 +59,6 @@
#include <net/snmp.h>
#include <net/dsfield.h>
#include <net/timewait_sock.h>
-#include <net/netdma.h>
#include <net/inet_common.h>
#include <net/secure_seq.h>
#include <net/tcp_memcontrol.h>
@@ -1504,18 +1503,8 @@ process:
bh_lock_sock_nested(sk);
ret = 0;
if (!sock_owned_by_user(sk)) {
-#ifdef CONFIG_NET_DMA
- struct tcp_sock *tp = tcp_sk(sk);
- if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
- tp->ucopy.dma_chan = net_dma_find_channel();
- if (tp->ucopy.dma_chan)
+ if (!tcp_prequeue(sk, skb))
ret = tcp_v6_do_rcv(sk, skb);
- else
-#endif
- {
- if (!tcp_prequeue(sk, skb))
- ret = tcp_v6_do_rcv(sk, skb);
- }
} else if (unlikely(sk_add_backlog(sk, skb,
sk->sk_rcvbuf + sk->sk_sndbuf))) {
bh_unlock_sock(sk);
diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c
index 7b01b9f5846c..e1b46709f8d6 100644
--- a/net/llc/af_llc.c
+++ b/net/llc/af_llc.c
@@ -838,7 +838,7 @@ static int llc_ui_recvmsg(struct kiocb *iocb, struct socket *sock,
if (!(flags & MSG_PEEK)) {
spin_lock_irqsave(&sk->sk_receive_queue.lock, cpu_flags);
- sk_eat_skb(sk, skb, false);
+ sk_eat_skb(sk, skb);
spin_unlock_irqrestore(&sk->sk_receive_queue.lock, cpu_flags);
*seq = 0;
}
@@ -860,10 +860,10 @@ copy_uaddr:
llc_cmsg_rcv(msg, skb);
if (!(flags & MSG_PEEK)) {
- spin_lock_irqsave(&sk->sk_receive_queue.lock, cpu_flags);
- sk_eat_skb(sk, skb, false);
- spin_unlock_irqrestore(&sk->sk_receive_queue.lock, cpu_flags);
- *seq = 0;
+ spin_lock_irqsave(&sk->sk_receive_queue.lock, cpu_flags);
+ sk_eat_skb(sk, skb);
+ spin_unlock_irqrestore(&sk->sk_receive_queue.lock, cpu_flags);
+ *seq = 0;
}
goto out;
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox