Commit c049d56e authored by Vlad Buslov's avatar Vlad Buslov Committed by David S. Miller

net: sched: flower: refactor reoffload for concurrent access

Recent changes that introduced unlocked flower did not properly account for
case when reoffload is initiated concurrently with filter updates. To fix
the issue, extend flower with 'hw_filters' list that is used to store
filters that don't have 'skip_hw' flag set. Filter is added to the list
when it is inserted to hardware and only removed from it after being
unoffloaded from all drivers that parent block is attached to. This ensures
that concurrent reoffload can still access filter that is being deleted and
prevents race condition when driver callback can be removed when filter is
no longer accessible trough idr, but is still present in hardware.

Refactor fl_change() to respect new filter reference counter and to release
filter reference with __fl_put() in case of error, instead of directly
deallocating filter memory. This allows for concurrent access to filter
from fl_reoffload() and protects it with reference counting. Refactor
fl_reoffload() to iterate over hw_filters list instead of idr. Implement
fl_get_next_hw_filter() helper function that is used to iterate over
hw_filters list with reference counting and skips filters that are being
concurrently deleted.

Fixes: 92149190 ("net: sched: flower: set unlocked flag for flower proto ops")
Signed-off-by: default avatarVlad Buslov <vladbu@mellanox.com>
Reviewed-by: default avatarSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent a3ddd94f
...@@ -90,6 +90,7 @@ struct cls_fl_head { ...@@ -90,6 +90,7 @@ struct cls_fl_head {
struct rhashtable ht; struct rhashtable ht;
spinlock_t masks_lock; /* Protect masks list */ spinlock_t masks_lock; /* Protect masks list */
struct list_head masks; struct list_head masks;
struct list_head hw_filters;
struct rcu_work rwork; struct rcu_work rwork;
struct idr handle_idr; struct idr handle_idr;
}; };
...@@ -102,6 +103,7 @@ struct cls_fl_filter { ...@@ -102,6 +103,7 @@ struct cls_fl_filter {
struct tcf_result res; struct tcf_result res;
struct fl_flow_key key; struct fl_flow_key key;
struct list_head list; struct list_head list;
struct list_head hw_list;
u32 handle; u32 handle;
u32 flags; u32 flags;
u32 in_hw_count; u32 in_hw_count;
...@@ -315,6 +317,7 @@ static int fl_init(struct tcf_proto *tp) ...@@ -315,6 +317,7 @@ static int fl_init(struct tcf_proto *tp)
spin_lock_init(&head->masks_lock); spin_lock_init(&head->masks_lock);
INIT_LIST_HEAD_RCU(&head->masks); INIT_LIST_HEAD_RCU(&head->masks);
INIT_LIST_HEAD(&head->hw_filters);
rcu_assign_pointer(tp->root, head); rcu_assign_pointer(tp->root, head);
idr_init(&head->handle_idr); idr_init(&head->handle_idr);
...@@ -352,6 +355,16 @@ static bool fl_mask_put(struct cls_fl_head *head, struct fl_flow_mask *mask) ...@@ -352,6 +355,16 @@ static bool fl_mask_put(struct cls_fl_head *head, struct fl_flow_mask *mask)
return true; return true;
} }
static struct cls_fl_head *fl_head_dereference(struct tcf_proto *tp)
{
/* Flower classifier only changes root pointer during init and destroy.
* Users must obtain reference to tcf_proto instance before calling its
* API, so tp->root pointer is protected from concurrent call to
* fl_destroy() by reference counting.
*/
return rcu_dereference_raw(tp->root);
}
static void __fl_destroy_filter(struct cls_fl_filter *f) static void __fl_destroy_filter(struct cls_fl_filter *f)
{ {
tcf_exts_destroy(&f->exts); tcf_exts_destroy(&f->exts);
...@@ -382,6 +395,7 @@ static void fl_hw_destroy_filter(struct tcf_proto *tp, struct cls_fl_filter *f, ...@@ -382,6 +395,7 @@ static void fl_hw_destroy_filter(struct tcf_proto *tp, struct cls_fl_filter *f,
tc_setup_cb_call(block, TC_SETUP_CLSFLOWER, &cls_flower, false); tc_setup_cb_call(block, TC_SETUP_CLSFLOWER, &cls_flower, false);
spin_lock(&tp->lock); spin_lock(&tp->lock);
list_del_init(&f->hw_list);
tcf_block_offload_dec(block, &f->flags); tcf_block_offload_dec(block, &f->flags);
spin_unlock(&tp->lock); spin_unlock(&tp->lock);
...@@ -393,6 +407,7 @@ static int fl_hw_replace_filter(struct tcf_proto *tp, ...@@ -393,6 +407,7 @@ static int fl_hw_replace_filter(struct tcf_proto *tp,
struct cls_fl_filter *f, bool rtnl_held, struct cls_fl_filter *f, bool rtnl_held,
struct netlink_ext_ack *extack) struct netlink_ext_ack *extack)
{ {
struct cls_fl_head *head = fl_head_dereference(tp);
struct tc_cls_flower_offload cls_flower = {}; struct tc_cls_flower_offload cls_flower = {};
struct tcf_block *block = tp->chain->block; struct tcf_block *block = tp->chain->block;
bool skip_sw = tc_skip_sw(f->flags); bool skip_sw = tc_skip_sw(f->flags);
...@@ -444,6 +459,9 @@ static int fl_hw_replace_filter(struct tcf_proto *tp, ...@@ -444,6 +459,9 @@ static int fl_hw_replace_filter(struct tcf_proto *tp,
goto errout; goto errout;
} }
spin_lock(&tp->lock);
list_add(&f->hw_list, &head->hw_filters);
spin_unlock(&tp->lock);
errout: errout:
if (!rtnl_held) if (!rtnl_held)
rtnl_unlock(); rtnl_unlock();
...@@ -475,23 +493,11 @@ static void fl_hw_update_stats(struct tcf_proto *tp, struct cls_fl_filter *f, ...@@ -475,23 +493,11 @@ static void fl_hw_update_stats(struct tcf_proto *tp, struct cls_fl_filter *f,
rtnl_unlock(); rtnl_unlock();
} }
static struct cls_fl_head *fl_head_dereference(struct tcf_proto *tp)
{
/* Flower classifier only changes root pointer during init and destroy.
* Users must obtain reference to tcf_proto instance before calling its
* API, so tp->root pointer is protected from concurrent call to
* fl_destroy() by reference counting.
*/
return rcu_dereference_raw(tp->root);
}
static void __fl_put(struct cls_fl_filter *f) static void __fl_put(struct cls_fl_filter *f)
{ {
if (!refcount_dec_and_test(&f->refcnt)) if (!refcount_dec_and_test(&f->refcnt))
return; return;
WARN_ON(!f->deleted);
if (tcf_exts_get_net(&f->exts)) if (tcf_exts_get_net(&f->exts))
tcf_queue_work(&f->rwork, fl_destroy_filter_work); tcf_queue_work(&f->rwork, fl_destroy_filter_work);
else else
...@@ -1522,6 +1528,7 @@ static int fl_change(struct net *net, struct sk_buff *in_skb, ...@@ -1522,6 +1528,7 @@ static int fl_change(struct net *net, struct sk_buff *in_skb,
err = -ENOBUFS; err = -ENOBUFS;
goto errout_tb; goto errout_tb;
} }
INIT_LIST_HEAD(&fnew->hw_list);
refcount_set(&fnew->refcnt, 1); refcount_set(&fnew->refcnt, 1);
err = tcf_exts_init(&fnew->exts, net, TCA_FLOWER_ACT, 0); err = tcf_exts_init(&fnew->exts, net, TCA_FLOWER_ACT, 0);
...@@ -1569,7 +1576,6 @@ static int fl_change(struct net *net, struct sk_buff *in_skb, ...@@ -1569,7 +1576,6 @@ static int fl_change(struct net *net, struct sk_buff *in_skb,
goto errout_hw; goto errout_hw;
} }
refcount_inc(&fnew->refcnt);
if (fold) { if (fold) {
/* Fold filter was deleted concurrently. Retry lookup. */ /* Fold filter was deleted concurrently. Retry lookup. */
if (fold->deleted) { if (fold->deleted) {
...@@ -1591,6 +1597,7 @@ static int fl_change(struct net *net, struct sk_buff *in_skb, ...@@ -1591,6 +1597,7 @@ static int fl_change(struct net *net, struct sk_buff *in_skb,
in_ht = true; in_ht = true;
} }
refcount_inc(&fnew->refcnt);
rhashtable_remove_fast(&fold->mask->ht, rhashtable_remove_fast(&fold->mask->ht,
&fold->ht_node, &fold->ht_node,
fold->mask->filter_ht_params); fold->mask->filter_ht_params);
...@@ -1631,6 +1638,7 @@ static int fl_change(struct net *net, struct sk_buff *in_skb, ...@@ -1631,6 +1638,7 @@ static int fl_change(struct net *net, struct sk_buff *in_skb,
if (err) if (err)
goto errout_hw; goto errout_hw;
refcount_inc(&fnew->refcnt);
fnew->handle = handle; fnew->handle = handle;
list_add_tail_rcu(&fnew->list, &fnew->mask->filters); list_add_tail_rcu(&fnew->list, &fnew->mask->filters);
spin_unlock(&tp->lock); spin_unlock(&tp->lock);
...@@ -1642,19 +1650,20 @@ static int fl_change(struct net *net, struct sk_buff *in_skb, ...@@ -1642,19 +1650,20 @@ static int fl_change(struct net *net, struct sk_buff *in_skb,
kfree(mask); kfree(mask);
return 0; return 0;
errout_ht:
spin_lock(&tp->lock);
errout_hw: errout_hw:
fnew->deleted = true;
spin_unlock(&tp->lock); spin_unlock(&tp->lock);
if (!tc_skip_hw(fnew->flags)) if (!tc_skip_hw(fnew->flags))
fl_hw_destroy_filter(tp, fnew, rtnl_held, NULL); fl_hw_destroy_filter(tp, fnew, rtnl_held, NULL);
errout_ht:
if (in_ht) if (in_ht)
rhashtable_remove_fast(&fnew->mask->ht, &fnew->ht_node, rhashtable_remove_fast(&fnew->mask->ht, &fnew->ht_node,
fnew->mask->filter_ht_params); fnew->mask->filter_ht_params);
errout_mask: errout_mask:
fl_mask_put(head, fnew->mask); fl_mask_put(head, fnew->mask);
errout: errout:
tcf_exts_get_net(&fnew->exts); __fl_put(fnew);
tcf_queue_work(&fnew->rwork, fl_destroy_filter_work);
errout_tb: errout_tb:
kfree(tb); kfree(tb);
errout_mask_alloc: errout_mask_alloc:
...@@ -1699,19 +1708,46 @@ static void fl_walk(struct tcf_proto *tp, struct tcf_walker *arg, ...@@ -1699,19 +1708,46 @@ static void fl_walk(struct tcf_proto *tp, struct tcf_walker *arg,
} }
} }
static struct cls_fl_filter *
fl_get_next_hw_filter(struct tcf_proto *tp, struct cls_fl_filter *f, bool add)
{
struct cls_fl_head *head = fl_head_dereference(tp);
spin_lock(&tp->lock);
if (list_empty(&head->hw_filters)) {
spin_unlock(&tp->lock);
return NULL;
}
if (!f)
f = list_entry(&head->hw_filters, struct cls_fl_filter,
hw_list);
list_for_each_entry_continue(f, &head->hw_filters, hw_list) {
if (!(add && f->deleted) && refcount_inc_not_zero(&f->refcnt)) {
spin_unlock(&tp->lock);
return f;
}
}
spin_unlock(&tp->lock);
return NULL;
}
static int fl_reoffload(struct tcf_proto *tp, bool add, tc_setup_cb_t *cb, static int fl_reoffload(struct tcf_proto *tp, bool add, tc_setup_cb_t *cb,
void *cb_priv, struct netlink_ext_ack *extack) void *cb_priv, struct netlink_ext_ack *extack)
{ {
struct tc_cls_flower_offload cls_flower = {}; struct tc_cls_flower_offload cls_flower = {};
struct tcf_block *block = tp->chain->block; struct tcf_block *block = tp->chain->block;
unsigned long handle = 0; struct cls_fl_filter *f = NULL;
struct cls_fl_filter *f;
int err; int err;
while ((f = fl_get_next_filter(tp, &handle))) { /* hw_filters list can only be changed by hw offload functions after
if (tc_skip_hw(f->flags)) * obtaining rtnl lock. Make sure it is not changed while reoffload is
goto next_flow; * iterating it.
*/
ASSERT_RTNL();
while ((f = fl_get_next_hw_filter(tp, f, add))) {
cls_flower.rule = cls_flower.rule =
flow_rule_alloc(tcf_exts_num_actions(&f->exts)); flow_rule_alloc(tcf_exts_num_actions(&f->exts));
if (!cls_flower.rule) { if (!cls_flower.rule) {
...@@ -1757,7 +1793,6 @@ static int fl_reoffload(struct tcf_proto *tp, bool add, tc_setup_cb_t *cb, ...@@ -1757,7 +1793,6 @@ static int fl_reoffload(struct tcf_proto *tp, bool add, tc_setup_cb_t *cb,
add); add);
spin_unlock(&tp->lock); spin_unlock(&tp->lock);
next_flow: next_flow:
handle++;
__fl_put(f); __fl_put(f);
} }
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment