diff options
Diffstat (limited to 'net/sched')
-rw-r--r-- | net/sched/act_api.c | 253 | ||||
-rw-r--r-- | net/sched/act_gact.c | 142 | ||||
-rw-r--r-- | net/sched/act_ipt.c | 179 | ||||
-rw-r--r-- | net/sched/act_mirred.c | 159 | ||||
-rw-r--r-- | net/sched/act_pedit.c | 166 | ||||
-rw-r--r-- | net/sched/act_police.c | 508 | ||||
-rw-r--r-- | net/sched/act_simple.c | 183 | ||||
-rw-r--r-- | net/sched/cls_fw.c | 25 | ||||
-rw-r--r-- | net/sched/sch_htb.c | 1363 | ||||
-rw-r--r-- | net/sched/sch_netem.c | 4 |
10 files changed, 1566 insertions, 1416 deletions
diff --git a/net/sched/act_api.c b/net/sched/act_api.c index a2587b5..835070e 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -33,16 +33,230 @@ #include <net/sch_generic.h> #include <net/act_api.h> -#if 0 /* control */ -#define DPRINTK(format, args...) printk(KERN_DEBUG format, ##args) -#else -#define DPRINTK(format, args...) +void tcf_hash_destroy(struct tcf_common *p, struct tcf_hashinfo *hinfo) +{ + unsigned int h = tcf_hash(p->tcfc_index, hinfo->hmask); + struct tcf_common **p1p; + + for (p1p = &hinfo->htab[h]; *p1p; p1p = &(*p1p)->tcfc_next) { + if (*p1p == p) { + write_lock_bh(hinfo->lock); + *p1p = p->tcfc_next; + write_unlock_bh(hinfo->lock); +#ifdef CONFIG_NET_ESTIMATOR + gen_kill_estimator(&p->tcfc_bstats, + &p->tcfc_rate_est); #endif -#if 0 /* data */ -#define D2PRINTK(format, args...) printk(KERN_DEBUG format, ##args) -#else -#define D2PRINTK(format, args...) + kfree(p); + return; + } + } + BUG_TRAP(0); +} +EXPORT_SYMBOL(tcf_hash_destroy); + +int tcf_hash_release(struct tcf_common *p, int bind, + struct tcf_hashinfo *hinfo) +{ + int ret = 0; + + if (p) { + if (bind) + p->tcfc_bindcnt--; + + p->tcfc_refcnt--; + if (p->tcfc_bindcnt <= 0 && p->tcfc_refcnt <= 0) { + tcf_hash_destroy(p, hinfo); + ret = 1; + } + } + return ret; +} +EXPORT_SYMBOL(tcf_hash_release); + +static int tcf_dump_walker(struct sk_buff *skb, struct netlink_callback *cb, + struct tc_action *a, struct tcf_hashinfo *hinfo) +{ + struct tcf_common *p; + int err = 0, index = -1,i = 0, s_i = 0, n_i = 0; + struct rtattr *r ; + + read_lock(hinfo->lock); + + s_i = cb->args[0]; + + for (i = 0; i < (hinfo->hmask + 1); i++) { + p = hinfo->htab[tcf_hash(i, hinfo->hmask)]; + + for (; p; p = p->tcfc_next) { + index++; + if (index < s_i) + continue; + a->priv = p; + a->order = n_i; + r = (struct rtattr*) skb->tail; + RTA_PUT(skb, a->order, 0, NULL); + err = tcf_action_dump_1(skb, a, 0, 0); + if (err < 0) { + index--; + skb_trim(skb, (u8*)r - skb->data); + goto done; + } + r->rta_len = skb->tail - (u8*)r; + n_i++; + if (n_i >= TCA_ACT_MAX_PRIO) + goto done; + } + } +done: + read_unlock(hinfo->lock); + if (n_i) + cb->args[0] += n_i; + return n_i; + +rtattr_failure: + skb_trim(skb, (u8*)r - skb->data); + goto done; +} + +static int tcf_del_walker(struct sk_buff *skb, struct tc_action *a, + struct tcf_hashinfo *hinfo) +{ + struct tcf_common *p, *s_p; + struct rtattr *r ; + int i= 0, n_i = 0; + + r = (struct rtattr*) skb->tail; + RTA_PUT(skb, a->order, 0, NULL); + RTA_PUT(skb, TCA_KIND, IFNAMSIZ, a->ops->kind); + for (i = 0; i < (hinfo->hmask + 1); i++) { + p = hinfo->htab[tcf_hash(i, hinfo->hmask)]; + + while (p != NULL) { + s_p = p->tcfc_next; + if (ACT_P_DELETED == tcf_hash_release(p, 0, hinfo)) + module_put(a->ops->owner); + n_i++; + p = s_p; + } + } + RTA_PUT(skb, TCA_FCNT, 4, &n_i); + r->rta_len = skb->tail - (u8*)r; + + return n_i; +rtattr_failure: + skb_trim(skb, (u8*)r - skb->data); + return -EINVAL; +} + +int tcf_generic_walker(struct sk_buff *skb, struct netlink_callback *cb, + int type, struct tc_action *a) +{ + struct tcf_hashinfo *hinfo = a->ops->hinfo; + + if (type == RTM_DELACTION) { + return tcf_del_walker(skb, a, hinfo); + } else if (type == RTM_GETACTION) { + return tcf_dump_walker(skb, cb, a, hinfo); + } else { + printk("tcf_generic_walker: unknown action %d\n", type); + return -EINVAL; + } +} +EXPORT_SYMBOL(tcf_generic_walker); + +struct tcf_common *tcf_hash_lookup(u32 index, struct tcf_hashinfo *hinfo) +{ + struct tcf_common *p; + + read_lock(hinfo->lock); + for (p = hinfo->htab[tcf_hash(index, hinfo->hmask)]; p; + p = p->tcfc_next) { + if (p->tcfc_index == index) + break; + } + read_unlock(hinfo->lock); + + return p; +} +EXPORT_SYMBOL(tcf_hash_lookup); + +u32 tcf_hash_new_index(u32 *idx_gen, struct tcf_hashinfo *hinfo) +{ + u32 val = *idx_gen; + + do { + if (++val == 0) + val = 1; + } while (tcf_hash_lookup(val, hinfo)); + + return (*idx_gen = val); +} +EXPORT_SYMBOL(tcf_hash_new_index); + +int tcf_hash_search(struct tc_action *a, u32 index) +{ + struct tcf_hashinfo *hinfo = a->ops->hinfo; + struct tcf_common *p = tcf_hash_lookup(index, hinfo); + + if (p) { + a->priv = p; + return 1; + } + return 0; +} +EXPORT_SYMBOL(tcf_hash_search); + +struct tcf_common *tcf_hash_check(u32 index, struct tc_action *a, int bind, + struct tcf_hashinfo *hinfo) +{ + struct tcf_common *p = NULL; + if (index && (p = tcf_hash_lookup(index, hinfo)) != NULL) { + if (bind) { + p->tcfc_bindcnt++; + p->tcfc_refcnt++; + } + a->priv = p; + } + return p; +} +EXPORT_SYMBOL(tcf_hash_check); + +struct tcf_common *tcf_hash_create(u32 index, struct rtattr *est, struct tc_action *a, int size, int bind, u32 *idx_gen, struct tcf_hashinfo *hinfo) +{ + struct tcf_common *p = kzalloc(size, GFP_KERNEL); + + if (unlikely(!p)) + return p; + p->tcfc_refcnt = 1; + if (bind) + p->tcfc_bindcnt = 1; + + spin_lock_init(&p->tcfc_lock); + p->tcfc_stats_lock = &p->tcfc_lock; + p->tcfc_index = index ? index : tcf_hash_new_index(idx_gen, hinfo); + p->tcfc_tm.install = jiffies; + p->tcfc_tm.lastuse = jiffies; +#ifdef CONFIG_NET_ESTIMATOR + if (est) + gen_new_estimator(&p->tcfc_bstats, &p->tcfc_rate_est, + p->tcfc_stats_lock, est); #endif + a->priv = (void *) p; + return p; +} +EXPORT_SYMBOL(tcf_hash_create); + +void tcf_hash_insert(struct tcf_common *p, struct tcf_hashinfo *hinfo) +{ + unsigned int h = tcf_hash(p->tcfc_index, hinfo->hmask); + + write_lock_bh(hinfo->lock); + p->tcfc_next = hinfo->htab[h]; + hinfo->htab[h] = p; + write_unlock_bh(hinfo->lock); +} +EXPORT_SYMBOL(tcf_hash_insert); static struct tc_action_ops *act_base = NULL; static DEFINE_RWLOCK(act_mod_lock); @@ -155,9 +369,6 @@ int tcf_action_exec(struct sk_buff *skb, struct tc_action *act, if (skb->tc_verd & TC_NCLS) { skb->tc_verd = CLR_TC_NCLS(skb->tc_verd); - D2PRINTK("(%p)tcf_action_exec: cleared TC_NCLS in %s out %s\n", - skb, skb->input_dev ? skb->input_dev->name : "xxx", - skb->dev->name); ret = TC_ACT_OK; goto exec_done; } @@ -187,8 +398,6 @@ void tcf_action_destroy(struct tc_action *act, int bind) for (a = act; a; a = act) { if (a->ops && a->ops->cleanup) { - DPRINTK("tcf_action_destroy destroying %p next %p\n", - a, a->next); if (a->ops->cleanup(a, bind) == ACT_P_DELETED) module_put(a->ops->owner); act = act->next; @@ -331,7 +540,6 @@ struct tc_action *tcf_action_init_1(struct rtattr *rta, struct rtattr *est, if (*err != ACT_P_CREATED) module_put(a_o->owner); a->ops = a_o; - DPRINTK("tcf_action_init_1: successfull %s\n", act_name); *err = 0; return a; @@ -392,12 +600,12 @@ int tcf_action_copy_stats(struct sk_buff *skb, struct tc_action *a, if (compat_mode) { if (a->type == TCA_OLD_COMPAT) err = gnet_stats_start_copy_compat(skb, 0, - TCA_STATS, TCA_XSTATS, h->stats_lock, &d); + TCA_STATS, TCA_XSTATS, h->tcf_stats_lock, &d); else return 0; } else err = gnet_stats_start_copy(skb, TCA_ACT_STATS, - h->stats_lock, &d); + h->tcf_stats_lock, &d); if (err < 0) goto errout; @@ -406,11 +614,11 @@ int tcf_action_copy_stats(struct sk_buff *skb, struct tc_action *a, if (a->ops->get_stats(skb, a) < 0) goto errout; - if (gnet_stats_copy_basic(&d, &h->bstats) < 0 || + if (gnet_stats_copy_basic(&d, &h->tcf_bstats) < 0 || #ifdef CONFIG_NET_ESTIMATOR - gnet_stats_copy_rate_est(&d, &h->rate_est) < 0 || + gnet_stats_copy_rate_est(&d, &h->tcf_rate_est) < 0 || #endif - gnet_stats_copy_queue(&d, &h->qstats) < 0) + gnet_stats_copy_queue(&d, &h->tcf_qstats) < 0) goto errout; if (gnet_stats_finish_copy(&d) < 0) @@ -459,7 +667,6 @@ static int act_get_notify(u32 pid, struct nlmsghdr *n, struct tc_action *a, int event) { struct sk_buff *skb; - int err = 0; skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL); if (!skb) @@ -468,10 +675,8 @@ act_get_notify(u32 pid, struct nlmsghdr *n, struct tc_action *a, int event) kfree_skb(skb); return -EINVAL; } - err = netlink_unicast(rtnl, skb, pid, MSG_DONTWAIT); - if (err > 0) - err = 0; - return err; + + return rtnl_unicast(skb, pid); } static struct tc_action * diff --git a/net/sched/act_gact.c b/net/sched/act_gact.c index e75a147..6cff566 100644 --- a/net/sched/act_gact.c +++ b/net/sched/act_gact.c @@ -34,48 +34,43 @@ #include <linux/tc_act/tc_gact.h> #include <net/tc_act/tc_gact.h> -/* use generic hash table */ -#define MY_TAB_SIZE 16 -#define MY_TAB_MASK 15 - -static u32 idx_gen; -static struct tcf_gact *tcf_gact_ht[MY_TAB_SIZE]; +#define GACT_TAB_MASK 15 +static struct tcf_common *tcf_gact_ht[GACT_TAB_MASK + 1]; +static u32 gact_idx_gen; static DEFINE_RWLOCK(gact_lock); -/* ovewrride the defaults */ -#define tcf_st tcf_gact -#define tc_st tc_gact -#define tcf_t_lock gact_lock -#define tcf_ht tcf_gact_ht - -#define CONFIG_NET_ACT_INIT 1 -#include <net/pkt_act.h> +static struct tcf_hashinfo gact_hash_info = { + .htab = tcf_gact_ht, + .hmask = GACT_TAB_MASK, + .lock = &gact_lock, +}; #ifdef CONFIG_GACT_PROB -static int gact_net_rand(struct tcf_gact *p) +static int gact_net_rand(struct tcf_gact *gact) { - if (net_random()%p->pval) - return p->action; - return p->paction; + if (net_random() % gact->tcfg_pval) + return gact->tcf_action; + return gact->tcfg_paction; } -static int gact_determ(struct tcf_gact *p) +static int gact_determ(struct tcf_gact *gact) { - if (p->bstats.packets%p->pval) - return p->action; - return p->paction; + if (gact->tcf_bstats.packets % gact->tcfg_pval) + return gact->tcf_action; + return gact->tcfg_paction; } -typedef int (*g_rand)(struct tcf_gact *p); +typedef int (*g_rand)(struct tcf_gact *gact); static g_rand gact_rand[MAX_RAND]= { NULL, gact_net_rand, gact_determ }; -#endif +#endif /* CONFIG_GACT_PROB */ static int tcf_gact_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, int ovr, int bind) { struct rtattr *tb[TCA_GACT_MAX]; struct tc_gact *parm; - struct tcf_gact *p; + struct tcf_gact *gact; + struct tcf_common *pc; int ret = 0; if (rta == NULL || rtattr_parse_nested(tb, TCA_GACT_MAX, rta) < 0) @@ -94,105 +89,106 @@ static int tcf_gact_init(struct rtattr *rta, struct rtattr *est, return -EOPNOTSUPP; #endif - p = tcf_hash_check(parm->index, a, ovr, bind); - if (p == NULL) { - p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind); - if (p == NULL) + pc = tcf_hash_check(parm->index, a, bind, &gact_hash_info); + if (!pc) { + pc = tcf_hash_create(parm->index, est, a, sizeof(*gact), + bind, &gact_idx_gen, &gact_hash_info); + if (unlikely(!pc)) return -ENOMEM; ret = ACT_P_CREATED; } else { if (!ovr) { - tcf_hash_release(p, bind); + tcf_hash_release(pc, bind, &gact_hash_info); return -EEXIST; } } - spin_lock_bh(&p->lock); - p->action = parm->action; + gact = to_gact(pc); + + spin_lock_bh(&gact->tcf_lock); + gact->tcf_action = parm->action; #ifdef CONFIG_GACT_PROB if (tb[TCA_GACT_PROB-1] != NULL) { struct tc_gact_p *p_parm = RTA_DATA(tb[TCA_GACT_PROB-1]); - p->paction = p_parm->paction; - p->pval = p_parm->pval; - p->ptype = p_parm->ptype; + gact->tcfg_paction = p_parm->paction; + gact->tcfg_pval = p_parm->pval; + gact->tcfg_ptype = p_parm->ptype; } #endif - spin_unlock_bh(&p->lock); + spin_unlock_bh(&gact->tcf_lock); if (ret == ACT_P_CREATED) - tcf_hash_insert(p); + tcf_hash_insert(pc, &gact_hash_info); return ret; } -static int -tcf_gact_cleanup(struct tc_action *a, int bind) +static int tcf_gact_cleanup(struct tc_action *a, int bind) { - struct tcf_gact *p = PRIV(a, gact); + struct tcf_gact *gact = a->priv; - if (p != NULL) - return tcf_hash_release(p, bind); + if (gact) + return tcf_hash_release(&gact->common, bind, &gact_hash_info); return 0; } -static int -tcf_gact(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) +static int tcf_gact(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) { - struct tcf_gact *p = PRIV(a, gact); + struct tcf_gact *gact = a->priv; int action = TC_ACT_SHOT; - spin_lock(&p->lock); + spin_lock(&gact->tcf_lock); #ifdef CONFIG_GACT_PROB - if (p->ptype && gact_rand[p->ptype] != NULL) - action = gact_rand[p->ptype](p); + if (gact->tcfg_ptype && gact_rand[gact->tcfg_ptype] != NULL) + action = gact_rand[gact->tcfg_ptype](gact); else - action = p->action; + action = gact->tcf_action; #else - action = p->action; + action = gact->tcf_action; #endif - p->bstats.bytes += skb->len; - p->bstats.packets++; + gact->tcf_bstats.bytes += skb->len; + gact->tcf_bstats.packets++; if (action == TC_ACT_SHOT) - p->qstats.drops++; - p->tm.lastuse = jiffies; - spin_unlock(&p->lock); + gact->tcf_qstats.drops++; + gact->tcf_tm.lastuse = jiffies; + spin_unlock(&gact->tcf_lock); return action; } -static int -tcf_gact_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) +static int tcf_gact_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) { unsigned char *b = skb->tail; struct tc_gact opt; - struct tcf_gact *p = PRIV(a, gact); + struct tcf_gact *gact = a->priv; struct tcf_t t; - opt.index = p->index; - opt.refcnt = p->refcnt - ref; - opt.bindcnt = p->bindcnt - bind; - opt.action = p->action; + opt.index = gact->tcf_index; + opt.refcnt = gact->tcf_refcnt - ref; + opt.bindcnt = gact->tcf_bindcnt - bind; + opt.action = gact->tcf_action; RTA_PUT(skb, TCA_GACT_PARMS, sizeof(opt), &opt); #ifdef CONFIG_GACT_PROB - if (p->ptype) { + if (gact->tcfg_ptype) { struct tc_gact_p p_opt; - p_opt.paction = p->paction; - p_opt.pval = p->pval; - p_opt.ptype = p->ptype; + p_opt.paction = gact->tcfg_paction; + p_opt.pval = gact->tcfg_pval; + p_opt.ptype = gact->tcfg_ptype; RTA_PUT(skb, TCA_GACT_PROB, sizeof(p_opt), &p_opt); } #endif - t.install = jiffies_to_clock_t(jiffies - p->tm.install); - t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse); - t.expires = jiffies_to_clock_t(p->tm.expires); + t.install = jiffies_to_clock_t(jiffies - gact->tcf_tm.install); + t.lastuse = jiffies_to_clock_t(jiffies - gact->tcf_tm.lastuse); + t.expires = jiffies_to_clock_t(gact->tcf_tm.expires); RTA_PUT(skb, TCA_GACT_TM, sizeof(t), &t); return skb->len; - rtattr_failure: +rtattr_failure: skb_trim(skb, b - skb->data); return -1; } static struct tc_action_ops act_gact_ops = { .kind = "gact", + .hinfo = &gact_hash_info, .type = TCA_ACT_GACT, .capab = TCA_CAP_NONE, .owner = THIS_MODULE, @@ -208,8 +204,7 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)"); MODULE_DESCRIPTION("Generic Classifier actions"); MODULE_LICENSE("GPL"); -static int __init -gact_init_module(void) +static int __init gact_init_module(void) { #ifdef CONFIG_GACT_PROB printk("GACT probability on\n"); @@ -219,8 +214,7 @@ gact_init_module(void) return tcf_register_action(&act_gact_ops); } -static void __exit -gact_cleanup_module(void) +static void __exit gact_cleanup_module(void) { tcf_unregister_action(&act_gact_ops); } diff --git a/net/sched/act_ipt.c b/net/sched/act_ipt.c index d799e01..d8c9310 100644 --- a/net/sched/act_ipt.c +++ b/net/sched/act_ipt.c @@ -38,25 +38,19 @@ #include <linux/netfilter_ipv4/ip_tables.h> -/* use generic hash table */ -#define MY_TAB_SIZE 16 -#define MY_TAB_MASK 15 -static u32 idx_gen; -static struct tcf_ipt *tcf_ipt_ht[MY_TAB_SIZE]; -/* ipt hash table lock */ +#define IPT_TAB_MASK 15 +static struct tcf_common *tcf_ipt_ht[IPT_TAB_MASK + 1]; +static u32 ipt_idx_gen; static DEFINE_RWLOCK(ipt_lock); -/* ovewrride the defaults */ -#define tcf_st tcf_ipt -#define tcf_t_lock ipt_lock -#define tcf_ht tcf_ipt_ht - -#define CONFIG_NET_ACT_INIT -#include <net/pkt_act.h> +static struct tcf_hashinfo ipt_hash_info = { + .htab = tcf_ipt_ht, + .hmask = IPT_TAB_MASK, + .lock = &ipt_lock, +}; -static int -ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook) +static int ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook) { struct ipt_target *target; int ret = 0; @@ -65,7 +59,6 @@ ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook) if (!target) return -ENOENT; - DPRINTK("ipt_init_target: found %s\n", target->name); t->u.kernel.target = target; ret = xt_check_target(target, AF_INET, t->u.target_size - sizeof(*t), @@ -76,10 +69,7 @@ ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook) if (t->u.kernel.target->checkentry && !t->u.kernel.target->checkentry(table, NULL, t->u.kernel.target, t->data, - t->u.target_size - sizeof(*t), hook)) { - DPRINTK("ipt_init_target: check failed for `%s'.\n", - t->u.kernel.target->name); module_put(t->u.kernel.target->me); ret = -EINVAL; } @@ -87,40 +77,37 @@ ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook) return ret; } -static void -ipt_destroy_target(struct ipt_entry_target *t) +static void ipt_destroy_target(struct ipt_entry_target *t) { if (t->u.kernel.target->destroy) - t->u.kernel.target->destroy(t->u.kernel.target, t->data, - t->u.target_size - sizeof(*t)); + t->u.kernel.target->destroy(t->u.kernel.target, t->data); module_put(t->u.kernel.target->me); } -static int -tcf_ipt_release(struct tcf_ipt *p, int bind) +static int tcf_ipt_release(struct tcf_ipt *ipt, int bind) { int ret = 0; - if (p) { + if (ipt) { if (bind) - p->bindcnt--; - p->refcnt--; - if (p->bindcnt <= 0 && p->refcnt <= 0) { - ipt_destroy_target(p->t); - kfree(p->tname); - kfree(p->t); - tcf_hash_destroy(p); + ipt->tcf_bindcnt--; + ipt->tcf_refcnt--; + if (ipt->tcf_bindcnt <= 0 && ipt->tcf_refcnt <= 0) { + ipt_destroy_target(ipt->tcfi_t); + kfree(ipt->tcfi_tname); + kfree(ipt->tcfi_t); + tcf_hash_destroy(&ipt->common, &ipt_hash_info); ret = ACT_P_DELETED; } } return ret; } -static int -tcf_ipt_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, - int ovr, int bind) +static int tcf_ipt_init(struct rtattr *rta, struct rtattr *est, + struct tc_action *a, int ovr, int bind) { struct rtattr *tb[TCA_IPT_MAX]; - struct tcf_ipt *p; + struct tcf_ipt *ipt; + struct tcf_common *pc; struct ipt_entry_target *td, *t; char *tname; int ret = 0, err; @@ -144,49 +131,51 @@ tcf_ipt_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, RTA_PAYLOAD(tb[TCA_IPT_INDEX-1]) >= sizeof(u32)) index = *(u32 *)RTA_DATA(tb[TCA_IPT_INDEX-1]); - p = tcf_hash_check(index, a, ovr, bind); - if (p == NULL) { - p = tcf_hash_create(index, est, a, sizeof(*p), ovr, bind); - if (p == NULL) + pc = tcf_hash_check(index, a, bind, &ipt_hash_info); + if (!pc) { + pc = tcf_hash_create(index, est, a, sizeof(*ipt), bind, + &ipt_idx_gen, &ipt_hash_info); + if (unlikely(!pc)) return -ENOMEM; ret = ACT_P_CREATED; } else { if (!ovr) { - tcf_ipt_release(p, bind); + tcf_ipt_release(to_ipt(pc), bind); return -EEXIST; } } + ipt = to_ipt(pc); hook = *(u32 *)RTA_DATA(tb[TCA_IPT_HOOK-1]); err = -ENOMEM; tname = kmalloc(IFNAMSIZ, GFP_KERNEL); - if (tname == NULL) + if (unlikely(!tname)) goto err1; if (tb[TCA_IPT_TABLE - 1] == NULL || rtattr_strlcpy(tname, tb[TCA_IPT_TABLE-1], IFNAMSIZ) >= IFNAMSIZ) strcpy(tname, "mangle"); t = kmalloc(td->u.target_size, GFP_KERNEL); - if (t == NULL) + if (unlikely(!t)) goto err2; memcpy(t, td, td->u.target_size); if ((err = ipt_init_target(t, tname, hook)) < 0) goto err3; - spin_lock_bh(&p->lock); + spin_lock_bh(&ipt->tcf_lock); if (ret != ACT_P_CREATED) { - ipt_destroy_target(p->t); - kfree(p->tname); - kfree(p->t); + ipt_destroy_target(ipt->tcfi_t); + kfree(ipt->tcfi_tname); + kfree(ipt->tcfi_t); } - p->tname = tname; - p->t = t; - p->hook = hook; - spin_unlock_bh(&p->lock); + ipt->tcfi_tname = tname; + ipt->tcfi_t = t; + ipt->tcfi_hook = hook; + spin_unlock_bh(&ipt->tcf_lock); if (ret == ACT_P_CREATED) - tcf_hash_insert(p); + tcf_hash_insert(pc, &ipt_hash_info); return ret; err3: @@ -194,33 +183,32 @@ err3: err2: kfree(tname); err1: - kfree(p); + kfree(pc); return err; } -static int -tcf_ipt_cleanup(struct tc_action *a, int bind) +static int tcf_ipt_cleanup(struct tc_action *a, int bind) { - struct tcf_ipt *p = PRIV(a, ipt); - return tcf_ipt_release(p, bind); + struct tcf_ipt *ipt = a->priv; + return tcf_ipt_release(ipt, bind); } -static int -tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) +static int tcf_ipt(struct sk_buff *skb, struct tc_action *a, + struct tcf_result *res) { int ret = 0, result = 0; - struct tcf_ipt *p = PRIV(a, ipt); + struct tcf_ipt *ipt = a->priv; if (skb_cloned(skb)) { if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) return TC_ACT_UNSPEC; } - spin_lock(&p->lock); + spin_lock(&ipt->tcf_lock); - p->tm.lastuse = jiffies; - p->bstats.bytes += skb->len; - p->bstats.packets++; + ipt->tcf_tm.lastuse = jiffies; + ipt->tcf_bstats.bytes += skb->len; + ipt->tcf_bstats.packets++; /* yes, we have to worry about both in and out dev worry later - danger - this API seems to have changed @@ -229,16 +217,17 @@ tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) /* iptables targets take a double skb pointer in case the skb * needs to be replaced. We don't own the skb, so this must not * happen. The pskb_expand_head above should make sure of this */ - ret = p->t->u.kernel.target->target(&skb, skb->dev, NULL, p->hook, - p->t->u.kernel.target, p->t->data, - NULL); + ret = ipt->tcfi_t->u.kernel.target->target(&skb, skb->dev, NULL, + ipt->tcfi_hook, + ipt->tcfi_t->u.kernel.target, + ipt->tcfi_t->data); switch (ret) { case NF_ACCEPT: result = TC_ACT_OK; break; case NF_DROP: result = TC_ACT_SHOT; - p->qstats.drops++; + ipt->tcf_qstats.drops++; break; case IPT_CONTINUE: result = TC_ACT_PIPE; @@ -249,53 +238,46 @@ tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) result = TC_POLICE_OK; break; } - spin_unlock(&p->lock); + spin_unlock(&ipt->tcf_lock); return result; } -static int -tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) +static int tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) { + unsigned char *b = skb->tail; + struct tcf_ipt *ipt = a->priv; struct ipt_entry_target *t; struct tcf_t tm; struct tc_cnt c; - unsigned char *b = skb->tail; - struct tcf_ipt *p = PRIV(a, ipt); /* for simple targets kernel size == user size ** user name = target name ** for foolproof you need to not assume this */ - t = kmalloc(p->t->u.user.target_size, GFP_ATOMIC); - if (t == NULL) + t = kmalloc(ipt->tcfi_t->u.user.target_size, GFP_ATOMIC); + if (unlikely(!t)) goto rtattr_failure; - c.bindcnt = p->bindcnt - bind; - c.refcnt = p->refcnt - ref; - memcpy(t, p->t, p->t->u.user.target_size); - strcpy(t->u.user.name, p->t->u.kernel.target->name); - - DPRINTK("\ttcf_ipt_dump tablename %s length %d\n", p->tname, - strlen(p->tname)); - DPRINTK("\tdump target name %s size %d size user %d " - "data[0] %x data[1] %x\n", p->t->u.kernel.target->name, - p->t->u.target_size, p->t->u.user.target_size, - p->t->data[0], p->t->data[1]); - RTA_PUT(skb, TCA_IPT_TARG, p->t->u.user.target_size, t); - RTA_PUT(skb, TCA_IPT_INDEX, 4, &p->index); - RTA_PUT(skb, TCA_IPT_HOOK, 4, &p->hook); + c.bindcnt = ipt->tcf_bindcnt - bind; + c.refcnt = ipt->tcf_refcnt - ref; + memcpy(t, ipt->tcfi_t, ipt->tcfi_t->u.user.target_size); + strcpy(t->u.user.name, ipt->tcfi_t->u.kernel.target->name); + + RTA_PUT(skb, TCA_IPT_TARG, ipt->tcfi_t->u.user.target_size, t); + RTA_PUT(skb, TCA_IPT_INDEX, 4, &ipt->tcf_index); + RTA_PUT(skb, TCA_IPT_HOOK, 4, &ipt->tcfi_hook); RTA_PUT(skb, TCA_IPT_CNT, sizeof(struct tc_cnt), &c); - RTA_PUT(skb, TCA_IPT_TABLE, IFNAMSIZ, p->tname); - tm.install = jiffies_to_clock_t(jiffies - p->tm.install); - tm.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse); - tm.expires = jiffies_to_clock_t(p->tm.expires); + RTA_PUT(skb, TCA_IPT_TABLE, IFNAMSIZ, ipt->tcfi_tname); + tm.install = jiffies_to_clock_t(jiffies - ipt->tcf_tm.install); + tm.lastuse = jiffies_to_clock_t(jiffies - ipt->tcf_tm.lastuse); + tm.expires = jiffies_to_clock_t(ipt->tcf_tm.expires); RTA_PUT(skb, TCA_IPT_TM, sizeof (tm), &tm); kfree(t); return skb->len; - rtattr_failure: +rtattr_failure: skb_trim(skb, b - skb->data); kfree(t); return -1; @@ -303,6 +285,7 @@ tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) static struct tc_action_ops act_ipt_ops = { .kind = "ipt", + .hinfo = &ipt_hash_info, .type = TCA_ACT_IPT, .capab = TCA_CAP_NONE, .owner = THIS_MODULE, @@ -318,14 +301,12 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)"); MODULE_DESCRIPTION("Iptables target actions"); MODULE_LICENSE("GPL"); -static int __init -ipt_init_module(void) +static int __init ipt_init_module(void) { return tcf_register_action(&act_ipt_ops); } -static void __exit -ipt_cleanup_module(void) +static void __exit ipt_cleanup_module(void) { tcf_unregister_action(&act_ipt_ops); } diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c index fc56204..4838972 100644 --- a/net/sched/act_mirred.c +++ b/net/sched/act_mirred.c @@ -39,46 +39,39 @@ #include <linux/etherdevice.h> #include <linux/if_arp.h> - -/* use generic hash table */ -#define MY_TAB_SIZE 8 -#define MY_TAB_MASK (MY_TAB_SIZE - 1) -static u32 idx_gen; -static struct tcf_mirred *tcf_mirred_ht[MY_TAB_SIZE]; +#define MIRRED_TAB_MASK 7 +static struct tcf_common *tcf_mirred_ht[MIRRED_TAB_MASK + 1]; +static u32 mirred_idx_gen; static DEFINE_RWLOCK(mirred_lock); -/* ovewrride the defaults */ -#define tcf_st tcf_mirred -#define tc_st tc_mirred -#define tcf_t_lock mirred_lock -#define tcf_ht tcf_mirred_ht - -#define CONFIG_NET_ACT_INIT 1 -#include <net/pkt_act.h> +static struct tcf_hashinfo mirred_hash_info = { + .htab = tcf_mirred_ht, + .hmask = MIRRED_TAB_MASK, + .lock = &mirred_lock, +}; -static inline int -tcf_mirred_release(struct tcf_mirred *p, int bind) +static inline int tcf_mirred_release(struct tcf_mirred *m, int bind) { - if (p) { + if (m) { if (bind) - p->bindcnt--; - p->refcnt--; - if(!p->bindcnt && p->refcnt <= 0) { - dev_put(p->dev); - tcf_hash_destroy(p); + m->tcf_bindcnt--; + m->tcf_refcnt--; + if(!m->tcf_bindcnt && m->tcf_refcnt <= 0) { + dev_put(m->tcfm_dev); + tcf_hash_destroy(&m->common, &mirred_hash_info); return 1; } } return 0; } -static int -tcf_mirred_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, - int ovr, int bind) +static int tcf_mirred_init(struct rtattr *rta, struct rtattr *est, + struct tc_action *a, int ovr, int bind) { struct rtattr *tb[TCA_MIRRED_MAX]; struct tc_mirred *parm; - struct tcf_mirred *p; + struct tcf_mirred *m; + struct tcf_common *pc; struct net_device *dev = NULL; int ret = 0; int ok_push = 0; @@ -110,64 +103,62 @@ tcf_mirred_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, } } - p = tcf_hash_check(parm->index, a, ovr, bind); - if (p == NULL) { + pc = tcf_hash_check(parm->index, a, bind, &mirred_hash_info); + if (!pc) { if (!parm->ifindex) return -EINVAL; - p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind); - if (p == NULL) + pc = tcf_hash_create(parm->index, est, a, sizeof(*m), bind, + &mirred_idx_gen, &mirred_hash_info); + if (unlikely(!pc)) return -ENOMEM; ret = ACT_P_CREATED; } else { if (!ovr) { - tcf_mirred_release(p, bind); + tcf_mirred_release(to_mirred(pc), bind); return -EEXIST; } } + m = to_mirred(pc); - spin_lock_bh(&p->lock); - p->action = parm->action; - p->eaction = parm->eaction; + spin_lock_bh(&m->tcf_lock); + m->tcf_action = parm->action; + m->tcfm_eaction = parm->eaction; if (parm->ifindex) { - p->ifindex = parm->ifindex; + m->tcfm_ifindex = parm->ifindex; if (ret != ACT_P_CREATED) - dev_put(p->dev); - p->dev = dev; + dev_put(m->tcfm_dev); + m->tcfm_dev = dev; dev_hold(dev); - p->ok_push = ok_push; + m->tcfm_ok_push = ok_push; } - spin_unlock_bh(&p->lock); + spin_unlock_bh(&m->tcf_lock); if (ret == ACT_P_CREATED) - tcf_hash_insert(p); + tcf_hash_insert(pc, &mirred_hash_info); - DPRINTK("tcf_mirred_init index %d action %d eaction %d device %s " - "ifindex %d\n", parm->index, parm->action, parm->eaction, - dev->name, parm->ifindex); return ret; } -static int -tcf_mirred_cleanup(struct tc_action *a, int bind) +static int tcf_mirred_cleanup(struct tc_action *a, int bind) { - struct tcf_mirred *p = PRIV(a, mirred); + struct tcf_mirred *m = a->priv; - if (p != NULL) - return tcf_mirred_release(p, bind); + if (m) + return tcf_mirred_release(m, bind); return 0; } -static int -tcf_mirred(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) +static int tcf_mirred(struct sk_buff *skb, struct tc_action *a, + struct tcf_result *res) { - struct tcf_mirred *p = PRIV(a, mirred); + struct tcf_mirred *m = a->priv; struct net_device *dev; struct sk_buff *skb2 = NULL; u32 at = G_TC_AT(skb->tc_verd); - spin_lock(&p->lock); + spin_lock(&m->tcf_lock); - dev = p->dev; - p->tm.lastuse = jiffies; + dev = m->tcfm_dev; + m->tcf_tm.lastuse = jiffies; if (!(dev->flags&IFF_UP) ) { if (net_ratelimit()) @@ -176,10 +167,10 @@ tcf_mirred(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) bad_mirred: if (skb2 != NULL) kfree_skb(skb2); - p->qstats.overlimits++; - p->bstats.bytes += skb->len; - p->bstats.packets++; - spin_unlock(&p->lock); + m->tcf_qstats.overlimits++; + m->tcf_bstats.bytes += skb->len; + m->tcf_bstats.packets++; + spin_unlock(&m->tcf_lock); /* should we be asking for packet to be dropped? * may make sense for redirect case only */ @@ -189,59 +180,59 @@ bad_mirred: skb2 = skb_clone(skb, GFP_ATOMIC); if (skb2 == NULL) goto bad_mirred; - if (p->eaction != TCA_EGRESS_MIRROR && p->eaction != TCA_EGRESS_REDIR) { + if (m->tcfm_eaction != TCA_EGRESS_MIRROR && + m->tcfm_eaction != TCA_EGRESS_REDIR) { if (net_ratelimit()) - printk("tcf_mirred unknown action %d\n", p->eaction); + printk("tcf_mirred unknown action %d\n", + m->tcfm_eaction); goto bad_mirred; } - p->bstats.bytes += skb2->len; - p->bstats.packets++; + m->tcf_bstats.bytes += skb2->len; + m->tcf_bstats.packets++; if (!(at & AT_EGRESS)) - if (p->ok_push) + if (m->tcfm_ok_push) skb_push(skb2, skb2->dev->hard_header_len); /* mirror is always swallowed */ - if (p->eaction != TCA_EGRESS_MIRROR) + if (m->tcfm_eaction != TCA_EGRESS_MIRROR) skb2->tc_verd = SET_TC_FROM(skb2->tc_verd, at); skb2->dev = dev; skb2->input_dev = skb->dev; dev_queue_xmit(skb2); - spin_unlock(&p->lock); - return p->action; + spin_unlock(&m->tcf_lock); + return m->tcf_action; } -static int -tcf_mirred_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) +static int tcf_mirred_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) { unsigned char *b = skb->tail; + struct tcf_mirred *m = a->priv; struct tc_mirred opt; - struct tcf_mirred *p = PRIV(a, mirred); struct tcf_t t; - opt.index = p->index; - opt.action = p->action; - opt.refcnt = p->refcnt - ref; - opt.bindcnt = p->bindcnt - bind; - opt.eaction = p->eaction; - opt.ifindex = p->ifindex; - DPRINTK("tcf_mirred_dump index %d action %d eaction %d ifindex %d\n", - p->index, p->action, p->eaction, p->ifindex); + opt.index = m->tcf_index; + opt.action = m->tcf_action; + opt.refcnt = m->tcf_refcnt - ref; + opt.bindcnt = m->tcf_bindcnt - bind; + opt.eaction = m->tcfm_eaction; + opt.ifindex = m->tcfm_ifindex; RTA_PUT(skb, TCA_MIRRED_PARMS, sizeof(opt), &opt); - t.install = jiffies_to_clock_t(jiffies - p->tm.install); - t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse); - t.expires = jiffies_to_clock_t(p->tm.expires); + t.install = jiffies_to_clock_t(jiffies - m->tcf_tm.install); + t.lastuse = jiffies_to_clock_t(jiffies - m->tcf_tm.lastuse); + t.expires = jiffies_to_clock_t(m->tcf_tm.expires); RTA_PUT(skb, TCA_MIRRED_TM, sizeof(t), &t); return skb->len; - rtattr_failure: +rtattr_failure: skb_trim(skb, b - skb->data); return -1; } static struct tc_action_ops act_mirred_ops = { .kind = "mirred", + .hinfo = &mirred_hash_info, .type = TCA_ACT_MIRRED, .capab = TCA_CAP_NONE, .owner = THIS_MODULE, @@ -257,15 +248,13 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002)"); MODULE_DESCRIPTION("Device Mirror/redirect actions"); MODULE_LICENSE("GPL"); -static int __init -mirred_init_module(void) +static int __init mirred_init_module(void) { printk("Mirror/redirect action on\n"); return tcf_register_action(&act_mirred_ops); } -static void __exit -mirred_cleanup_module(void) +static void __exit mirred_cleanup_module(void) { tcf_unregister_action(&act_mirred_ops); } diff --git a/net/sched/act_pedit.c b/net/sched/act_pedit.c index f257475..8ac65c2 100644 --- a/net/sched/act_pedit.c +++ b/net/sched/act_pedit.c @@ -33,32 +33,25 @@ #include <linux/tc_act/tc_pedit.h> #include <net/tc_act/tc_pedit.h> - -#define PEDIT_DEB 1 - -/* use generic hash table */ -#define MY_TAB_SIZE 16 -#define MY_TAB_MASK 15 -static u32 idx_gen; -static struct tcf_pedit *tcf_pedit_ht[MY_TAB_SIZE]; +#define PEDIT_TAB_MASK 15 +static struct tcf_common *tcf_pedit_ht[PEDIT_TAB_MASK + 1]; +static u32 pedit_idx_gen; static DEFINE_RWLOCK(pedit_lock); -#define tcf_st tcf_pedit -#define tc_st tc_pedit -#define tcf_t_lock pedit_lock -#define tcf_ht tcf_pedit_ht - -#define CONFIG_NET_ACT_INIT 1 -#include <net/pkt_act.h> +static struct tcf_hashinfo pedit_hash_info = { + .htab = tcf_pedit_ht, + .hmask = PEDIT_TAB_MASK, + .lock = &pedit_lock, +}; -static int -tcf_pedit_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, - int ovr, int bind) +static int tcf_pedit_init(struct rtattr *rta, struct rtattr *est, + struct tc_action *a, int ovr, int bind) { struct rtattr *tb[TCA_PEDIT_MAX]; struct tc_pedit *parm; int ret = 0; struct tcf_pedit *p; + struct tcf_common *pc; struct tc_pedit_key *keys = NULL; int ksize; @@ -73,54 +66,56 @@ tcf_pedit_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a, if (RTA_PAYLOAD(tb[TCA_PEDIT_PARMS-1]) < sizeof(*parm) + ksize) return -EINVAL; - p = tcf_hash_check(parm->index, a, ovr, bind); - if (p == NULL) { + pc = tcf_hash_check(parm->index, a, bind, &pedit_hash_info); + if (!pc) { if (!parm->nkeys) return -EINVAL; - p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind); - if (p == NULL) + pc = tcf_hash_create(parm->index, est, a, sizeof(*p), bind, + &pedit_idx_gen, &pedit_hash_info); + if (unlikely(!pc)) return -ENOMEM; + p = to_pedit(pc); keys = kmalloc(ksize, GFP_KERNEL); if (keys == NULL) { - kfree(p); + kfree(pc); return -ENOMEM; } ret = ACT_P_CREATED; } else { + p = to_pedit(pc); if (!ovr) { - tcf_hash_release(p, bind); + tcf_hash_release(pc, bind, &pedit_hash_info); return -EEXIST; } - if (p->nkeys && p->nkeys != parm->nkeys) { + if (p->tcfp_nkeys && p->tcfp_nkeys != parm->nkeys) { keys = kmalloc(ksize, GFP_KERNEL); if (keys == NULL) return -ENOMEM; } } - spin_lock_bh(&p->lock); - p->flags = parm->flags; - p->action = parm->action; + spin_lock_bh(&p->tcf_lock); + p->tcfp_flags = parm->flags; + p->tcf_action = parm->action; if (keys) { - kfree(p->keys); - p->keys = keys; - p->nkeys = parm->nkeys; + kfree(p->tcfp_keys); + p->tcfp_keys = keys; + p->tcfp_nkeys = parm->nkeys; } - memcpy(p->keys, parm->keys, ksize); - spin_unlock_bh(&p->lock); + memcpy(p->tcfp_keys, parm->keys, ksize); + spin_unlock_bh(&p->tcf_lock); if (ret == ACT_P_CREATED) - tcf_hash_insert(p); + tcf_hash_insert(pc, &pedit_hash_info); return ret; } -static int -tcf_pedit_cleanup(struct tc_action *a, int bind) +static int tcf_pedit_cleanup(struct tc_action *a, int bind) { - struct tcf_pedit *p = PRIV(a, pedit); + struct tcf_pedit *p = a->priv; - if (p != NULL) { - struct tc_pedit_key *keys = p->keys; - if (tcf_hash_release(p, bind)) { + if (p) { + struct tc_pedit_key *keys = p->tcfp_keys; + if (tcf_hash_release(&p->common, bind, &pedit_hash_info)) { kfree(keys); return 1; } @@ -128,30 +123,30 @@ tcf_pedit_cleanup(struct tc_action *a, int bind) return 0; } -static int -tcf_pedit(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) +static int tcf_pedit(struct sk_buff *skb, struct tc_action *a, + struct tcf_result *res) { - struct tcf_pedit *p = PRIV(a, pedit); + struct tcf_pedit *p = a->priv; int i, munged = 0; u8 *pptr; if (!(skb->tc_verd & TC_OK2MUNGE)) { /* should we set skb->cloned? */ if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) { - return p->action; + return p->tcf_action; } } pptr = skb->nh.raw; - spin_lock(&p->lock); + spin_lock(&p->tcf_lock); - p->tm.lastuse = jiffies; + p->tcf_tm.lastuse = jiffies; - if (p->nkeys > 0) { - struct tc_pedit_key *tkey = p->keys; + if (p->tcfp_nkeys > 0) { + struct tc_pedit_key *tkey = p->tcfp_keys; - for (i = p->nkeys; i > 0; i--, tkey++) { + for (i = p->tcfp_nkeys; i > 0; i--, tkey++) { u32 *ptr; int offset = tkey->off; @@ -169,7 +164,8 @@ tcf_pedit(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) printk("offset must be on 32 bit boundaries\n"); goto bad; } - if (skb->len < 0 || (offset > 0 && offset > skb->len)) { + if (skb->len < 0 || + (offset > 0 && offset > skb->len)) { printk("offset %d cant exceed pkt length %d\n", offset, skb->len); goto bad; @@ -185,63 +181,47 @@ tcf_pedit(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) skb->tc_verd = SET_TC_MUNGED(skb->tc_verd); goto done; } else { - printk("pedit BUG: index %d\n",p->index); + printk("pedit BUG: index %d\n", p->tcf_index); } bad: - p->qstats.overlimits++; + p->tcf_qstats.overlimits++; done: - p->bstats.bytes += skb->len; - p->bstats.packets++; - spin_unlock(&p->lock); - return p->action; + p->tcf_bstats.bytes += skb->len; + p->tcf_bstats.packets++; + spin_unlock(&p->tcf_lock); + return p->tcf_action; } -static int -tcf_pedit_dump(struct sk_buff *skb, struct tc_action *a,int bind, int ref) +static int tcf_pedit_dump(struct sk_buff *skb, struct tc_action *a, + int bind, int ref) { unsigned char *b = skb->tail; + struct tcf_pedit *p = a->priv; struct tc_pedit *opt; - struct tcf_pedit *p = PRIV(a, pedit); struct tcf_t t; int s; - s = sizeof(*opt) + p->nkeys * sizeof(struct tc_pedit_key); + s = sizeof(*opt) + p->tcfp_nkeys * sizeof(struct tc_pedit_key); /* netlink spinlocks held above us - must use ATOMIC */ opt = kzalloc(s, GFP_ATOMIC); - if (opt == NULL) + if (unlikely(!opt)) return -ENOBUFS; - memcpy(opt->keys, p->keys, p->nkeys * sizeof(struct tc_pedit_key)); - opt->index = p->index; - opt->nkeys = p->nkeys; - opt->flags = p->flags; - opt->action = p->action; - opt->refcnt = p->refcnt - ref; - opt->bindcnt = p->bindcnt - bind; - - -#ifdef PEDIT_DEB - { - /* Debug - get rid of later */ - int i; - struct tc_pedit_key *key = opt->keys; - - for (i=0; i<opt->nkeys; i++, key++) { - printk( "\n key #%d",i); - printk( " at %d: val %08x mask %08x", - (unsigned int)key->off, - (unsigned int)key->val, - (unsigned int)key->mask); - } - } -#endif + memcpy(opt->keys, p->tcfp_keys, + p->tcfp_nkeys * sizeof(struct tc_pedit_key)); + opt->index = p->tcf_index; + opt->nkeys = p->tcfp_nkeys; + opt->flags = p->tcfp_flags; + opt->action = p->tcf_action; + opt->refcnt = p->tcf_refcnt - ref; + opt->bindcnt = p->tcf_bindcnt - bind; RTA_PUT(skb, TCA_PEDIT_PARMS, s, opt); - t.install = jiffies_to_clock_t(jiffies - p->tm.install); - t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse); - t.expires = jiffies_to_clock_t(p->tm.expires); + t.install = jiffies_to_clock_t(jiffies - p->tcf_tm.install); + t.lastuse = jiffies_to_clock_t(jiffies - p->tcf_tm.lastuse); + t.expires = jiffies_to_clock_t(p->tcf_tm.expires); RTA_PUT(skb, TCA_PEDIT_TM, sizeof(t), &t); kfree(opt); return skb->len; @@ -252,9 +232,9 @@ rtattr_failure: return -1; } -static -struct tc_action_ops act_pedit_ops = { +static struct tc_action_ops act_pedit_ops = { .kind = "pedit", + .hinfo = &pedit_hash_info, .type = TCA_ACT_PEDIT, .capab = TCA_CAP_NONE, .owner = THIS_MODULE, @@ -270,14 +250,12 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)"); MODULE_DESCRIPTION("Generic Packet Editor actions"); MODULE_LICENSE("GPL"); -static int __init -pedit_init_module(void) +static int __init pedit_init_module(void) { return tcf_register_action(&act_pedit_ops); } -static void __exit -pedit_cleanup_module(void) +static void __exit pedit_cleanup_module(void) { tcf_unregister_action(&act_pedit_ops); } diff --git a/net/sched/act_police.c b/net/sched/act_police.c index da905d7..fed47b6 100644 --- a/net/sched/act_police.c +++ b/net/sched/act_police.c @@ -32,43 +32,27 @@ #include <net/sock.h> #include <net/act_api.h> -#define L2T(p,L) ((p)->R_tab->data[(L)>>(p)->R_tab->rate.cell_log]) -#define L2T_P(p,L) ((p)->P_tab->data[(L)>>(p)->P_tab->rate.cell_log]) -#define PRIV(a) ((struct tcf_police *) (a)->priv) - -/* use generic hash table */ -#define MY_TAB_SIZE 16 -#define MY_TAB_MASK 15 -static u32 idx_gen; -static struct tcf_police *tcf_police_ht[MY_TAB_SIZE]; -/* Policer hash table lock */ -static DEFINE_RWLOCK(police_lock); - -/* Each policer is serialized by its individual spinlock */ +#define L2T(p,L) ((p)->tcfp_R_tab->data[(L)>>(p)->tcfp_R_tab->rate.cell_log]) +#define L2T_P(p,L) ((p)->tcfp_P_tab->data[(L)>>(p)->tcfp_P_tab->rate.cell_log]) -static __inline__ unsigned tcf_police_hash(u32 index) -{ - return index&0xF; -} +#define POL_TAB_MASK 15 +static struct tcf_common *tcf_police_ht[POL_TAB_MASK + 1]; +static u32 police_idx_gen; +static DEFINE_RWLOCK(police_lock); -static __inline__ struct tcf_police * tcf_police_lookup(u32 index) -{ - struct tcf_police *p; +static struct tcf_hashinfo police_hash_info = { + .htab = tcf_police_ht, + .hmask = POL_TAB_MASK, + .lock = &police_lock, +}; - read_lock(&police_lock); - for (p = tcf_police_ht[tcf_police_hash(index)]; p; p = p->next) { - if (p->index == index) - break; - } - read_unlock(&police_lock); - return p; -} +/* Each policer is serialized by its individual spinlock */ #ifdef CONFIG_NET_CLS_ACT static int tcf_act_police_walker(struct sk_buff *skb, struct netlink_callback *cb, int type, struct tc_action *a) { - struct tcf_police *p; + struct tcf_common *p; int err = 0, index = -1, i = 0, s_i = 0, n_i = 0; struct rtattr *r; @@ -76,10 +60,10 @@ static int tcf_act_police_walker(struct sk_buff *skb, struct netlink_callback *c s_i = cb->args[0]; - for (i = 0; i < MY_TAB_SIZE; i++) { - p = tcf_police_ht[tcf_police_hash(i)]; + for (i = 0; i < (POL_TAB_MASK + 1); i++) { + p = tcf_police_ht[tcf_hash(i, POL_TAB_MASK)]; - for (; p; p = p->next) { + for (; p; p = p->tcfc_next) { index++; if (index < s_i) continue; @@ -110,48 +94,26 @@ rtattr_failure: skb_trim(skb, (u8*)r - skb->data); goto done; } - -static inline int -tcf_act_police_hash_search(struct tc_action *a, u32 index) -{ - struct tcf_police *p = tcf_police_lookup(index); - - if (p != NULL) { - a->priv = p; - return 1; - } else { - return 0; - } -} #endif -static inline u32 tcf_police_new_index(void) -{ - do { - if (++idx_gen == 0) - idx_gen = 1; - } while (tcf_police_lookup(idx_gen)); - - return idx_gen; -} - void tcf_police_destroy(struct tcf_police *p) { - unsigned h = tcf_police_hash(p->index); - struct tcf_police **p1p; + unsigned int h = tcf_hash(p->tcf_index, POL_TAB_MASK); + struct tcf_common **p1p; - for (p1p = &tcf_police_ht[h]; *p1p; p1p = &(*p1p)->next) { - if (*p1p == p) { + for (p1p = &tcf_police_ht[h]; *p1p; p1p = &(*p1p)->tcfc_next) { + if (*p1p == &p->common) { write_lock_bh(&police_lock); - *p1p = p->next; + *p1p = p->tcf_next; write_unlock_bh(&police_lock); #ifdef CONFIG_NET_ESTIMATOR - gen_kill_estimator(&p->bstats, &p->rate_est); + gen_kill_estimator(&p->tcf_bstats, + &p->tcf_rate_est); #endif - if (p->R_tab) - qdisc_put_rtab(p->R_tab); - if (p->P_tab) - qdisc_put_rtab(p->P_tab); + if (p->tcfp_R_tab) + qdisc_put_rtab(p->tcfp_R_tab); + if (p->tcfp_P_tab) + qdisc_put_rtab(p->tcfp_P_tab); kfree(p); return; } @@ -167,7 +129,7 @@ static int tcf_act_police_locate(struct rtattr *rta, struct rtattr *est, int ret = 0, err; struct rtattr *tb[TCA_POLICE_MAX]; struct tc_police *parm; - struct tcf_police *p; + struct tcf_police *police; struct qdisc_rate_table *R_tab = NULL, *P_tab = NULL; if (rta == NULL || rtattr_parse_nested(tb, TCA_POLICE_MAX, rta) < 0) @@ -185,27 +147,32 @@ static int tcf_act_police_locate(struct rtattr *rta, struct rtattr *est, RTA_PAYLOAD(tb[TCA_POLICE_RESULT-1]) != sizeof(u32)) return -EINVAL; - if (parm->index && (p = tcf_police_lookup(parm->index)) != NULL) { - a->priv = p; - if (bind) { - p->bindcnt += 1; - p->refcnt += 1; + if (parm->index) { + struct tcf_common *pc; + + pc = tcf_hash_lookup(parm->index, &police_hash_info); + if (pc != NULL) { + a->priv = pc; + police = to_police(pc); + if (bind) { + police->tcf_bindcnt += 1; + police->tcf_refcnt += 1; + } + if (ovr) + goto override; + return ret; } - if (ovr) - goto override; - return ret; } - p = kzalloc(sizeof(*p), GFP_KERNEL); - if (p == NULL) + police = kzalloc(sizeof(*police), GFP_KERNEL); + if (police == NULL) return -ENOMEM; - ret = ACT_P_CREATED; - p->refcnt = 1; - spin_lock_init(&p->lock); - p->stats_lock = &p->lock; + police->tcf_refcnt = 1; + spin_lock_init(&police->tcf_lock); + police->tcf_stats_lock = &police->tcf_lock; if (bind) - p->bindcnt = 1; + police->tcf_bindcnt = 1; override: if (parm->rate.rate) { err = -ENOMEM; @@ -215,67 +182,71 @@ override: if (parm->peakrate.rate) { P_tab = qdisc_get_rtab(&parm->peakrate, tb[TCA_POLICE_PEAKRATE-1]); - if (p->P_tab == NULL) { + if (P_tab == NULL) { qdisc_put_rtab(R_tab); goto failure; } } } /* No failure allowed after this point */ - spin_lock_bh(&p->lock); + spin_lock_bh(&police->tcf_lock); if (R_tab != NULL) { - qdisc_put_rtab(p->R_tab); - p->R_tab = R_tab; + qdisc_put_rtab(police->tcfp_R_tab); + police->tcfp_R_tab = R_tab; } if (P_tab != NULL) { - qdisc_put_rtab(p->P_tab); - p->P_tab = P_tab; + qdisc_put_rtab(police->tcfp_P_tab); + police->tcfp_P_tab = P_tab; } if (tb[TCA_POLICE_RESULT-1]) - p->result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]); - p->toks = p->burst = parm->burst; - p->mtu = parm->mtu; - if (p->mtu == 0) { - p->mtu = ~0; - if (p->R_tab) - p->mtu = 255<<p->R_tab->rate.cell_log; + police->tcfp_result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]); + police->tcfp_toks = police->tcfp_burst = parm->burst; + police->tcfp_mtu = parm->mtu; + if (police->tcfp_mtu == 0) { + police->tcfp_mtu = ~0; + if (police->tcfp_R_tab) + police->tcfp_mtu = 255<<police->tcfp_R_tab->rate.cell_log; } - if (p->P_tab) - p->ptoks = L2T_P(p, p->mtu); - p->action = parm->action; + if (police->tcfp_P_tab) + police->tcfp_ptoks = L2T_P(police, police->tcfp_mtu); + police->tcf_action = parm->action; #ifdef CONFIG_NET_ESTIMATOR if (tb[TCA_POLICE_AVRATE-1]) - p->ewma_rate = *(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]); + police->tcfp_ewma_rate = + *(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]); if (est) - gen_replace_estimator(&p->bstats, &p->rate_est, p->stats_lock, est); + gen_replace_estimator(&police->tcf_bstats, + &police->tcf_rate_est, + police->tcf_stats_lock, est); #endif - spin_unlock_bh(&p->lock); + spin_unlock_bh(&police->tcf_lock); if (ret != ACT_P_CREATED) return ret; - PSCHED_GET_TIME(p->t_c); - p->index = parm->index ? : tcf_police_new_index(); - h = tcf_police_hash(p->index); + PSCHED_GET_TIME(police->tcfp_t_c); + police->tcf_index = parm->index ? parm->index : + tcf_hash_new_index(&police_idx_gen, &police_hash_info); + h = tcf_hash(police->tcf_index, POL_TAB_MASK); write_lock_bh(&police_lock); - p->next = tcf_police_ht[h]; - tcf_police_ht[h] = p; + police->tcf_next = tcf_police_ht[h]; + tcf_police_ht[h] = &police->common; write_unlock_bh(&police_lock); - a->priv = p; + a->priv = police; return ret; failure: if (ret == ACT_P_CREATED) - kfree(p); + kfree(police); return err; } static int tcf_act_police_cleanup(struct tc_action *a, int bind) { - struct tcf_police *p = PRIV(a); + struct tcf_police *p = a->priv; if (p != NULL) return tcf_police_release(p, bind); @@ -285,86 +256,87 @@ static int tcf_act_police_cleanup(struct tc_action *a, int bind) static int tcf_act_police(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) { + struct tcf_police *police = a->priv; psched_time_t now; - struct tcf_police *p = PRIV(a); long toks; long ptoks = 0; - spin_lock(&p->lock); + spin_lock(&police->tcf_lock); - p->bstats.bytes += skb->len; - p->bstats.packets++; + police->tcf_bstats.bytes += skb->len; + police->tcf_bstats.packets++; #ifdef CONFIG_NET_ESTIMATOR - if (p->ewma_rate && p->rate_est.bps >= p->ewma_rate) { - p->qstats.overlimits++; - spin_unlock(&p->lock); - return p->action; + if (police->tcfp_ewma_rate && + police->tcf_rate_est.bps >= police->tcfp_ewma_rate) { + police->tcf_qstats.overlimits++; + spin_unlock(&police->tcf_lock); + return police->tcf_action; } #endif - if (skb->len <= p->mtu) { - if (p->R_tab == NULL) { - spin_unlock(&p->lock); - return p->result; + if (skb->len <= police->tcfp_mtu) { + if (police->tcfp_R_tab == NULL) { + spin_unlock(&police->tcf_lock); + return police->tcfp_result; } PSCHED_GET_TIME(now); - toks = PSCHED_TDIFF_SAFE(now, p->t_c, p->burst); - - if (p->P_tab) { - ptoks = toks + p->ptoks; - if (ptoks > (long)L2T_P(p, p->mtu)) - ptoks = (long)L2T_P(p, p->mtu); - ptoks -= L2T_P(p, skb->len); + toks = PSCHED_TDIFF_SAFE(now, police->tcfp_t_c, + police->tcfp_burst); + if (police->tcfp_P_tab) { + ptoks = toks + police->tcfp_ptoks; + if (ptoks > (long)L2T_P(police, police->tcfp_mtu)) + ptoks = (long)L2T_P(police, police->tcfp_mtu); + ptoks -= L2T_P(police, skb->len); } - toks += p->toks; - if (toks > (long)p->burst) - toks = p->burst; - toks -= L2T(p, skb->len); - + toks += police->tcfp_toks; + if (toks > (long)police->tcfp_burst) + toks = police->tcfp_burst; + toks -= L2T(police, skb->len); if ((toks|ptoks) >= 0) { - p->t_c = now; - p->toks = toks; - p->ptoks = ptoks; - spin_unlock(&p->lock); - return p->result; + police->tcfp_t_c = now; + police->tcfp_toks = toks; + police->tcfp_ptoks = ptoks; + spin_unlock(&police->tcf_lock); + return police->tcfp_result; } } - p->qstats.overlimits++; - spin_unlock(&p->lock); - return p->action; + police->tcf_qstats.overlimits++; + spin_unlock(&police->tcf_lock); + return police->tcf_action; } static int tcf_act_police_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref) { unsigned char *b = skb->tail; + struct tcf_police *police = a->priv; struct tc_police opt; - struct tcf_police *p = PRIV(a); - - opt.index = p->index; - opt.action = p->action; - opt.mtu = p->mtu; - opt.burst = p->burst; - opt.refcnt = p->refcnt - ref; - opt.bindcnt = p->bindcnt - bind; - if (p->R_tab) - opt.rate = p->R_tab->rate; + + opt.index = police->tcf_index; + opt.action = police->tcf_action; + opt.mtu = police->tcfp_mtu; + opt.burst = police->tcfp_burst; + opt.refcnt = police->tcf_refcnt - ref; + opt.bindcnt = police->tcf_bindcnt - bind; + if (police->tcfp_R_tab) + opt.rate = police->tcfp_R_tab->rate; else memset(&opt.rate, 0, sizeof(opt.rate)); - if (p->P_tab) - opt.peakrate = p->P_tab->rate; + if (police->tcfp_P_tab) + opt.peakrate = police->tcfp_P_tab->rate; else memset(&opt.peakrate, 0, sizeof(opt.peakrate)); RTA_PUT(skb, TCA_POLICE_TBF, sizeof(opt), &opt); - if (p->result) - RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int), &p->result); + if (police->tcfp_result) + RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int), + &police->tcfp_result); #ifdef CONFIG_NET_ESTIMATOR - if (p->ewma_rate) - RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &p->ewma_rate); + if (police->tcfp_ewma_rate) + RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &police->tcfp_ewma_rate); #endif return skb->len; @@ -379,13 +351,14 @@ MODULE_LICENSE("GPL"); static struct tc_action_ops act_police_ops = { .kind = "police", + .hinfo = &police_hash_info, .type = TCA_ID_POLICE, .capab = TCA_CAP_NONE, .owner = THIS_MODULE, .act = tcf_act_police, .dump = tcf_act_police_dump, .cleanup = tcf_act_police_cleanup, - .lookup = tcf_act_police_hash_search, + .lookup = tcf_hash_search, .init = tcf_act_police_locate, .walk = tcf_act_police_walker }; @@ -407,10 +380,39 @@ module_exit(police_cleanup_module); #else /* CONFIG_NET_CLS_ACT */ -struct tcf_police * tcf_police_locate(struct rtattr *rta, struct rtattr *est) +static struct tcf_common *tcf_police_lookup(u32 index) { - unsigned h; - struct tcf_police *p; + struct tcf_hashinfo *hinfo = &police_hash_info; + struct tcf_common *p; + + read_lock(hinfo->lock); + for (p = hinfo->htab[tcf_hash(index, hinfo->hmask)]; p; + p = p->tcfc_next) { + if (p->tcfc_index == index) + break; + } + read_unlock(hinfo->lock); + + return p; +} + +static u32 tcf_police_new_index(void) +{ + u32 *idx_gen = &police_idx_gen; + u32 val = *idx_gen; + + do { + if (++val == 0) + val = 1; + } while (tcf_police_lookup(val)); + + return (*idx_gen = val); +} + +struct tcf_police *tcf_police_locate(struct rtattr *rta, struct rtattr *est) +{ + unsigned int h; + struct tcf_police *police; struct rtattr *tb[TCA_POLICE_MAX]; struct tc_police *parm; @@ -423,149 +425,158 @@ struct tcf_police * tcf_police_locate(struct rtattr *rta, struct rtattr *est) parm = RTA_DATA(tb[TCA_POLICE_TBF-1]); - if (parm->index && (p = tcf_police_lookup(parm->index)) != NULL) { - p->refcnt++; - return p; - } + if (parm->index) { + struct tcf_common *pc; - p = kzalloc(sizeof(*p), GFP_KERNEL); - if (p == NULL) + pc = tcf_police_lookup(parm->index); + if (pc) { + police = to_police(pc); + police->tcf_refcnt++; + return police; + } + } + police = kzalloc(sizeof(*police), GFP_KERNEL); + if (unlikely(!police)) return NULL; - p->refcnt = 1; - spin_lock_init(&p->lock); - p->stats_lock = &p->lock; + police->tcf_refcnt = 1; + spin_lock_init(&police->tcf_lock); + police->tcf_stats_lock = &police->tcf_lock; if (parm->rate.rate) { - p->R_tab = qdisc_get_rtab(&parm->rate, tb[TCA_POLICE_RATE-1]); - if (p->R_tab == NULL) + police->tcfp_R_tab = + qdisc_get_rtab(&parm->rate, tb[TCA_POLICE_RATE-1]); + if (police->tcfp_R_tab == NULL) goto failure; if (parm->peakrate.rate) { - p->P_tab = qdisc_get_rtab(&parm->peakrate, - tb[TCA_POLICE_PEAKRATE-1]); - if (p->P_tab == NULL) + police->tcfp_P_tab = + qdisc_get_rtab(&parm->peakrate, + tb[TCA_POLICE_PEAKRATE-1]); + if (police->tcfp_P_tab == NULL) goto failure; } } if (tb[TCA_POLICE_RESULT-1]) { if (RTA_PAYLOAD(tb[TCA_POLICE_RESULT-1]) != sizeof(u32)) goto failure; - p->result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]); + police->tcfp_result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]); } #ifdef CONFIG_NET_ESTIMATOR if (tb[TCA_POLICE_AVRATE-1]) { if (RTA_PAYLOAD(tb[TCA_POLICE_AVRATE-1]) != sizeof(u32)) goto failure; - p->ewma_rate = *(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]); + police->tcfp_ewma_rate = + *(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]); } #endif - p->toks = p->burst = parm->burst; - p->mtu = parm->mtu; - if (p->mtu == 0) { - p->mtu = ~0; - if (p->R_tab) - p->mtu = 255<<p->R_tab->rate.cell_log; + police->tcfp_toks = police->tcfp_burst = parm->burst; + police->tcfp_mtu = parm->mtu; + if (police->tcfp_mtu == 0) { + police->tcfp_mtu = ~0; + if (police->tcfp_R_tab) + police->tcfp_mtu = 255<<police->tcfp_R_tab->rate.cell_log; } - if (p->P_tab) - p->ptoks = L2T_P(p, p->mtu); - PSCHED_GET_TIME(p->t_c); - p->index = parm->index ? : tcf_police_new_index(); - p->action = parm->action; + if (police->tcfp_P_tab) + police->tcfp_ptoks = L2T_P(police, police->tcfp_mtu); + PSCHED_GET_TIME(police->tcfp_t_c); + police->tcf_index = parm->index ? parm->index : + tcf_police_new_index(); + police->tcf_action = parm->action; #ifdef CONFIG_NET_ESTIMATOR if (est) - gen_new_estimator(&p->bstats, &p->rate_est, p->stats_lock, est); + gen_new_estimator(&police->tcf_bstats, &police->tcf_rate_est, + police->tcf_stats_lock, est); #endif - h = tcf_police_hash(p->index); + h = tcf_hash(police->tcf_index, POL_TAB_MASK); write_lock_bh(&police_lock); - p->next = tcf_police_ht[h]; - tcf_police_ht[h] = p; + police->tcf_next = tcf_police_ht[h]; + tcf_police_ht[h] = &police->common; write_unlock_bh(&police_lock); - return p; + return police; failure: - if (p->R_tab) - qdisc_put_rtab(p->R_tab); - kfree(p); + if (police->tcfp_R_tab) + qdisc_put_rtab(police->tcfp_R_tab); + kfree(police); return NULL; } -int tcf_police(struct sk_buff *skb, struct tcf_police *p) +int tcf_police(struct sk_buff *skb, struct tcf_police *police) { psched_time_t now; long toks; long ptoks = 0; - spin_lock(&p->lock); + spin_lock(&police->tcf_lock); - p->bstats.bytes += skb->len; - p->bstats.packets++; + police->tcf_bstats.bytes += skb->len; + police->tcf_bstats.packets++; #ifdef CONFIG_NET_ESTIMATOR - if (p->ewma_rate && p->rate_est.bps >= p->ewma_rate) { - p->qstats.overlimits++; - spin_unlock(&p->lock); - return p->action; + if (police->tcfp_ewma_rate && + police->tcf_rate_est.bps >= police->tcfp_ewma_rate) { + police->tcf_qstats.overlimits++; + spin_unlock(&police->tcf_lock); + return police->tcf_action; } #endif - - if (skb->len <= p->mtu) { - if (p->R_tab == NULL) { - spin_unlock(&p->lock); - return p->result; + if (skb->len <= police->tcfp_mtu) { + if (police->tcfp_R_tab == NULL) { + spin_unlock(&police->tcf_lock); + return police->tcfp_result; } PSCHED_GET_TIME(now); - - toks = PSCHED_TDIFF_SAFE(now, p->t_c, p->burst); - - if (p->P_tab) { - ptoks = toks + p->ptoks; - if (ptoks > (long)L2T_P(p, p->mtu)) - ptoks = (long)L2T_P(p, p->mtu); - ptoks -= L2T_P(p, skb->len); + toks = PSCHED_TDIFF_SAFE(now, police->tcfp_t_c, + police->tcfp_burst); + if (police->tcfp_P_tab) { + ptoks = toks + police->tcfp_ptoks; + if (ptoks > (long)L2T_P(police, police->tcfp_mtu)) + ptoks = (long)L2T_P(police, police->tcfp_mtu); + ptoks -= L2T_P(police, skb->len); } - toks += p->toks; - if (toks > (long)p->burst) - toks = p->burst; - toks -= L2T(p, skb->len); - + toks += police->tcfp_toks; + if (toks > (long)police->tcfp_burst) + toks = police->tcfp_burst; + toks -= L2T(police, skb->len); if ((toks|ptoks) >= 0) { - p->t_c = now; - p->toks = toks; - p->ptoks = ptoks; - spin_unlock(&p->lock); - return p->result; + police->tcfp_t_c = now; + police->tcfp_toks = toks; + police->tcfp_ptoks = ptoks; + spin_unlock(&police->tcf_lock); + return police->tcfp_result; } } - p->qstats.overlimits++; - spin_unlock(&p->lock); - return p->action; + police->tcf_qstats.overlimits++; + spin_unlock(&police->tcf_lock); + return police->tcf_action; } EXPORT_SYMBOL(tcf_police); -int tcf_police_dump(struct sk_buff *skb, struct tcf_police *p) +int tcf_police_dump(struct sk_buff *skb, struct tcf_police *police) { - unsigned char *b = skb->tail; + unsigned char *b = skb->tail; struct tc_police opt; - opt.index = p->index; - opt.action = p->action; - opt.mtu = p->mtu; - opt.burst = p->burst; - if (p->R_tab) - opt.rate = p->R_tab->rate; + opt.index = police->tcf_index; + opt.action = police->tcf_action; + opt.mtu = police->tcfp_mtu; + opt.burst = police->tcfp_burst; + if (police->tcfp_R_tab) + opt.rate = police->tcfp_R_tab->rate; else memset(&opt.rate, 0, sizeof(opt.rate)); - if (p->P_tab) - opt.peakrate = p->P_tab->rate; + if (police->tcfp_P_tab) + opt.peakrate = police->tcfp_P_tab->rate; else memset(&opt.peakrate, 0, sizeof(opt.peakrate)); RTA_PUT(skb, TCA_POLICE_TBF, sizeof(opt), &opt); - if (p->result) - RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int), &p->result); + if (police->tcfp_result) + RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int), + &police->tcfp_result); #ifdef CONFIG_NET_ESTIMATOR - if (p->ewma_rate) - RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &p->ewma_rate); + if (police->tcfp_ewma_rate) + RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &police->tcfp_ewma_rate); #endif return skb->len; @@ -574,19 +585,20 @@ rtattr_failure: return -1; } -int tcf_police_dump_stats(struct sk_buff *skb, struct tcf_police *p) +int tcf_police_dump_stats(struct sk_buff *skb, struct tcf_police *police) { struct gnet_dump d; if (gnet_stats_start_copy_compat(skb, TCA_STATS2, TCA_STATS, - TCA_XSTATS, p->stats_lock, &d) < 0) + TCA_XSTATS, police->tcf_stats_lock, + &d) < 0) goto errout; - if (gnet_stats_copy_basic(&d, &p->bstats) < 0 || + if (gnet_stats_copy_basic(&d, &police->tcf_bstats) < 0 || #ifdef CONFIG_NET_ESTIMATOR - gnet_stats_copy_rate_est(&d, &p->rate_est) < 0 || + gnet_stats_copy_rate_est(&d, &police->tcf_rate_est) < 0 || #endif - gnet_stats_copy_queue(&d, &p->qstats) < 0) + gnet_stats_copy_queue(&d, &police->tcf_qstats) < 0) goto errout; if (gnet_stats_finish_copy(&d) < 0) diff --git a/net/sched/act_simple.c b/net/sched/act_simple.c index 17105c8..901571a 100644 --- a/net/sched/act_simple.c +++ b/net/sched/act_simple.c @@ -20,54 +20,175 @@ #define TCA_ACT_SIMP 22 -/* XXX: Hide all these common elements under some macro - * probably -*/ #include <linux/tc_act/tc_defact.h> #include <net/tc_act/tc_defact.h> -/* use generic hash table with 8 buckets */ -#define MY_TAB_SIZE 8 -#define MY_TAB_MASK (MY_TAB_SIZE - 1) -static u32 idx_gen; -static struct tcf_defact *tcf_simp_ht[MY_TAB_SIZE]; +#define SIMP_TAB_MASK 7 +static struct tcf_common *tcf_simp_ht[SIMP_TAB_MASK + 1]; +static u32 simp_idx_gen; static DEFINE_RWLOCK(simp_lock); -/* override the defaults */ -#define tcf_st tcf_defact -#define tc_st tc_defact -#define tcf_t_lock simp_lock -#define tcf_ht tcf_simp_ht - -#define CONFIG_NET_ACT_INIT 1 -#include <net/pkt_act.h> -#include <net/act_generic.h> +static struct tcf_hashinfo simp_hash_info = { + .htab = tcf_simp_ht, + .hmask = SIMP_TAB_MASK, + .lock = &simp_lock, +}; static int tcf_simp(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) { - struct tcf_defact *p = PRIV(a, defact); + struct tcf_defact *d = a->priv; - spin_lock(&p->lock); - p->tm.lastuse = jiffies; - p->bstats.bytes += skb->len; - p->bstats.packets++; + spin_lock(&d->tcf_lock); + d->tcf_tm.lastuse = jiffies; + d->tcf_bstats.bytes += skb->len; + d->tcf_bstats.packets++; /* print policy string followed by _ then packet count * Example if this was the 3rd packet and the string was "hello" * then it would look like "hello_3" (without quotes) **/ - printk("simple: %s_%d\n", (char *)p->defdata, p->bstats.packets); - spin_unlock(&p->lock); - return p->action; + printk("simple: %s_%d\n", + (char *)d->tcfd_defdata, d->tcf_bstats.packets); + spin_unlock(&d->tcf_lock); + return d->tcf_action; +} + +static int tcf_simp_release(struct tcf_defact *d, int bind) +{ + int ret = 0; + if (d) { + if (bind) + d->tcf_bindcnt--; + d->tcf_refcnt--; + if (d->tcf_bindcnt <= 0 && d->tcf_refcnt <= 0) { + kfree(d->tcfd_defdata); + tcf_hash_destroy(&d->common, &simp_hash_info); + ret = 1; + } + } + return ret; +} + +static int alloc_defdata(struct tcf_defact *d, u32 datalen, void *defdata) +{ + d->tcfd_defdata = kmalloc(datalen, GFP_KERNEL); + if (unlikely(!d->tcfd_defdata)) + return -ENOMEM; + d->tcfd_datalen = datalen; + memcpy(d->tcfd_defdata, defdata, datalen); + return 0; +} + +static int realloc_defdata(struct tcf_defact *d, u32 datalen, void *defdata) +{ + kfree(d->tcfd_defdata); + return alloc_defdata(d, datalen, defdata); +} + +static int tcf_simp_init(struct rtattr *rta, struct rtattr *est, + struct tc_action *a, int ovr, int bind) +{ + struct rtattr *tb[TCA_DEF_MAX]; + struct tc_defact *parm; + struct tcf_defact *d; + struct tcf_common *pc; + void *defdata; + u32 datalen = 0; + int ret = 0; + + if (rta == NULL || rtattr_parse_nested(tb, TCA_DEF_MAX, rta) < 0) + return -EINVAL; + + if (tb[TCA_DEF_PARMS - 1] == NULL || + RTA_PAYLOAD(tb[TCA_DEF_PARMS - 1]) < sizeof(*parm)) + return -EINVAL; + + parm = RTA_DATA(tb[TCA_DEF_PARMS - 1]); + defdata = RTA_DATA(tb[TCA_DEF_DATA - 1]); + if (defdata == NULL) + return -EINVAL; + + datalen = RTA_PAYLOAD(tb[TCA_DEF_DATA - 1]); + if (datalen <= 0) + return -EINVAL; + + pc = tcf_hash_check(parm->index, a, bind, &simp_hash_info); + if (!pc) { + pc = tcf_hash_create(parm->index, est, a, sizeof(*d), bind, + &simp_idx_gen, &simp_hash_info); + if (unlikely(!pc)) + return -ENOMEM; + + d = to_defact(pc); + ret = alloc_defdata(d, datalen, defdata); + if (ret < 0) { + kfree(pc); + return ret; + } + ret = ACT_P_CREATED; + } else { + d = to_defact(pc); + if (!ovr) { + tcf_simp_release(d, bind); + return -EEXIST; + } + realloc_defdata(d, datalen, defdata); + } + + spin_lock_bh(&d->tcf_lock); + d->tcf_action = parm->action; + spin_unlock_bh(&d->tcf_lock); + + if (ret == ACT_P_CREATED) + tcf_hash_insert(pc, &simp_hash_info); + return ret; +} + +static inline int tcf_simp_cleanup(struct tc_action *a, int bind) +{ + struct tcf_defact *d = a->priv; + + if (d) + return tcf_simp_release(d, bind); + return 0; +} + +static inline int tcf_simp_dump(struct sk_buff *skb, struct tc_action *a, + int bind, int ref) +{ + unsigned char *b = skb->tail; + struct tcf_defact *d = a->priv; + struct tc_defact opt; + struct tcf_t t; + + opt.index = d->tcf_index; + opt.refcnt = d->tcf_refcnt - ref; + opt.bindcnt = d->tcf_bindcnt - bind; + opt.action = d->tcf_action; + RTA_PUT(skb, TCA_DEF_PARMS, sizeof(opt), &opt); + RTA_PUT(skb, TCA_DEF_DATA, d->tcfd_datalen, d->tcfd_defdata); + t.install = jiffies_to_clock_t(jiffies - d->tcf_tm.install); + t.lastuse = jiffies_to_clock_t(jiffies - d->tcf_tm.lastuse); + t.expires = jiffies_to_clock_t(d->tcf_tm.expires); + RTA_PUT(skb, TCA_DEF_TM, sizeof(t), &t); + return skb->len; + +rtattr_failure: + skb_trim(skb, b - skb->data); + return -1; } static struct tc_action_ops act_simp_ops = { - .kind = "simple", - .type = TCA_ACT_SIMP, - .capab = TCA_CAP_NONE, - .owner = THIS_MODULE, - .act = tcf_simp, - tca_use_default_ops + .kind = "simple", + .hinfo = &simp_hash_info, + .type = TCA_ACT_SIMP, + .capab = TCA_CAP_NONE, + .owner = THIS_MODULE, + .act = tcf_simp, + .dump = tcf_simp_dump, + .cleanup = tcf_simp_cleanup, + .init = tcf_simp_init, + .walk = tcf_generic_walker, }; MODULE_AUTHOR("Jamal Hadi Salim(2005)"); diff --git a/net/sched/cls_fw.c b/net/sched/cls_fw.c index e6973d9..e54acc6 100644 --- a/net/sched/cls_fw.c +++ b/net/sched/cls_fw.c @@ -50,6 +50,7 @@ struct fw_head { struct fw_filter *ht[HTSIZE]; + u32 mask; }; struct fw_filter @@ -101,7 +102,7 @@ static int fw_classify(struct sk_buff *skb, struct tcf_proto *tp, struct fw_filter *f; int r; #ifdef CONFIG_NETFILTER - u32 id = skb->nfmark; + u32 id = skb->nfmark & head->mask; #else u32 id = 0; #endif @@ -209,7 +210,9 @@ static int fw_change_attrs(struct tcf_proto *tp, struct fw_filter *f, struct rtattr **tb, struct rtattr **tca, unsigned long base) { + struct fw_head *head = (struct fw_head *)tp->root; struct tcf_exts e; + u32 mask; int err; err = tcf_exts_validate(tp, tb, tca[TCA_RATE-1], &e, &fw_ext_map); @@ -232,6 +235,15 @@ fw_change_attrs(struct tcf_proto *tp, struct fw_filter *f, } #endif /* CONFIG_NET_CLS_IND */ + if (tb[TCA_FW_MASK-1]) { + if (RTA_PAYLOAD(tb[TCA_FW_MASK-1]) != sizeof(u32)) + goto errout; + mask = *(u32*)RTA_DATA(tb[TCA_FW_MASK-1]); + if (mask != head->mask) + goto errout; + } else if (head->mask != 0xFFFFFFFF) + goto errout; + tcf_exts_change(tp, &f->exts, &e); return 0; @@ -267,9 +279,17 @@ static int fw_change(struct tcf_proto *tp, unsigned long base, return -EINVAL; if (head == NULL) { + u32 mask = 0xFFFFFFFF; + if (tb[TCA_FW_MASK-1]) { + if (RTA_PAYLOAD(tb[TCA_FW_MASK-1]) != sizeof(u32)) + return -EINVAL; + mask = *(u32*)RTA_DATA(tb[TCA_FW_MASK-1]); + } + head = kzalloc(sizeof(struct fw_head), GFP_KERNEL); if (head == NULL) return -ENOBUFS; + head->mask = mask; tcf_tree_lock(tp); tp->root = head; @@ -330,6 +350,7 @@ static void fw_walk(struct tcf_proto *tp, struct tcf_walker *arg) static int fw_dump(struct tcf_proto *tp, unsigned long fh, struct sk_buff *skb, struct tcmsg *t) { + struct fw_head *head = (struct fw_head *)tp->root; struct fw_filter *f = (struct fw_filter*)fh; unsigned char *b = skb->tail; struct rtattr *rta; @@ -351,6 +372,8 @@ static int fw_dump(struct tcf_proto *tp, unsigned long fh, if (strlen(f->indev)) RTA_PUT(skb, TCA_FW_INDEV, IFNAMSIZ, f->indev); #endif /* CONFIG_NET_CLS_IND */ + if (head->mask != 0xFFFFFFFF) + RTA_PUT(skb, TCA_FW_MASK, 4, &head->mask); if (tcf_exts_dump(skb, &f->exts, &fw_ext_map) < 0) goto rtattr_failure; diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 880a339..bb3ddd4 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -1,4 +1,4 @@ -/* vim: ts=8 sw=8 +/* * net/sched/sch_htb.c Hierarchical token bucket, feed tree version * * This program is free software; you can redistribute it and/or @@ -68,218 +68,165 @@ one less than their parent. */ -#define HTB_HSIZE 16 /* classid hash size */ -#define HTB_EWMAC 2 /* rate average over HTB_EWMAC*HTB_HSIZE sec */ -#undef HTB_DEBUG /* compile debugging support (activated by tc tool) */ -#define HTB_RATECM 1 /* whether to use rate computer */ -#define HTB_HYSTERESIS 1/* whether to use mode hysteresis for speedup */ -#define HTB_QLOCK(S) spin_lock_bh(&(S)->dev->queue_lock) -#define HTB_QUNLOCK(S) spin_unlock_bh(&(S)->dev->queue_lock) -#define HTB_VER 0x30011 /* major must be matched with number suplied by TC as version */ +#define HTB_HSIZE 16 /* classid hash size */ +#define HTB_EWMAC 2 /* rate average over HTB_EWMAC*HTB_HSIZE sec */ +#define HTB_RATECM 1 /* whether to use rate computer */ +#define HTB_HYSTERESIS 1 /* whether to use mode hysteresis for speedup */ +#define HTB_VER 0x30011 /* major must be matched with number suplied by TC as version */ #if HTB_VER >> 16 != TC_HTB_PROTOVER #error "Mismatched sch_htb.c and pkt_sch.h" #endif -/* debugging support; S is subsystem, these are defined: - 0 - netlink messages - 1 - enqueue - 2 - drop & requeue - 3 - dequeue main - 4 - dequeue one prio DRR part - 5 - dequeue class accounting - 6 - class overlimit status computation - 7 - hint tree - 8 - event queue - 10 - rate estimator - 11 - classifier - 12 - fast dequeue cache - - L is level; 0 = none, 1 = basic info, 2 = detailed, 3 = full - q->debug uint32 contains 16 2-bit fields one for subsystem starting - from LSB - */ -#ifdef HTB_DEBUG -#define HTB_DBG_COND(S,L) (((q->debug>>(2*S))&3) >= L) -#define HTB_DBG(S,L,FMT,ARG...) if (HTB_DBG_COND(S,L)) \ - printk(KERN_DEBUG FMT,##ARG) -#define HTB_CHCL(cl) BUG_TRAP((cl)->magic == HTB_CMAGIC) -#define HTB_PASSQ q, -#define HTB_ARGQ struct htb_sched *q, -#define static -#undef __inline__ -#define __inline__ -#undef inline -#define inline -#define HTB_CMAGIC 0xFEFAFEF1 -#define htb_safe_rb_erase(N,R) do { BUG_TRAP((N)->rb_color != -1); \ - if ((N)->rb_color == -1) break; \ - rb_erase(N,R); \ - (N)->rb_color = -1; } while (0) -#else -#define HTB_DBG_COND(S,L) (0) -#define HTB_DBG(S,L,FMT,ARG...) -#define HTB_PASSQ -#define HTB_ARGQ -#define HTB_CHCL(cl) -#define htb_safe_rb_erase(N,R) rb_erase(N,R) -#endif - - /* used internaly to keep status of single class */ enum htb_cmode { - HTB_CANT_SEND, /* class can't send and can't borrow */ - HTB_MAY_BORROW, /* class can't send but may borrow */ - HTB_CAN_SEND /* class can send */ + HTB_CANT_SEND, /* class can't send and can't borrow */ + HTB_MAY_BORROW, /* class can't send but may borrow */ + HTB_CAN_SEND /* class can send */ }; /* interior & leaf nodes; props specific to leaves are marked L: */ -struct htb_class -{ -#ifdef HTB_DEBUG - unsigned magic; -#endif - /* general class parameters */ - u32 classid; - struct gnet_stats_basic bstats; - struct gnet_stats_queue qstats; - struct gnet_stats_rate_est rate_est; - struct tc_htb_xstats xstats;/* our special stats */ - int refcnt; /* usage count of this class */ +struct htb_class { + /* general class parameters */ + u32 classid; + struct gnet_stats_basic bstats; + struct gnet_stats_queue qstats; + struct gnet_stats_rate_est rate_est; + struct tc_htb_xstats xstats; /* our special stats */ + int refcnt; /* usage count of this class */ #ifdef HTB_RATECM - /* rate measurement counters */ - unsigned long rate_bytes,sum_bytes; - unsigned long rate_packets,sum_packets; + /* rate measurement counters */ + unsigned long rate_bytes, sum_bytes; + unsigned long rate_packets, sum_packets; #endif - /* topology */ - int level; /* our level (see above) */ - struct htb_class *parent; /* parent class */ - struct list_head hlist; /* classid hash list item */ - struct list_head sibling; /* sibling list item */ - struct list_head children; /* children list */ - - union { - struct htb_class_leaf { - struct Qdisc *q; - int prio; - int aprio; - int quantum; - int deficit[TC_HTB_MAXDEPTH]; - struct list_head drop_list; - } leaf; - struct htb_class_inner { - struct rb_root feed[TC_HTB_NUMPRIO]; /* feed trees */ - struct rb_node *ptr[TC_HTB_NUMPRIO]; /* current class ptr */ - /* When class changes from state 1->2 and disconnects from - parent's feed then we lost ptr value and start from the - first child again. Here we store classid of the - last valid ptr (used when ptr is NULL). */ - u32 last_ptr_id[TC_HTB_NUMPRIO]; - } inner; - } un; - struct rb_node node[TC_HTB_NUMPRIO]; /* node for self or feed tree */ - struct rb_node pq_node; /* node for event queue */ - unsigned long pq_key; /* the same type as jiffies global */ - - int prio_activity; /* for which prios are we active */ - enum htb_cmode cmode; /* current mode of the class */ - - /* class attached filters */ - struct tcf_proto *filter_list; - int filter_cnt; - - int warned; /* only one warning about non work conserving .. */ - - /* token bucket parameters */ - struct qdisc_rate_table *rate; /* rate table of the class itself */ - struct qdisc_rate_table *ceil; /* ceiling rate (limits borrows too) */ - long buffer,cbuffer; /* token bucket depth/rate */ - psched_tdiff_t mbuffer; /* max wait time */ - long tokens,ctokens; /* current number of tokens */ - psched_time_t t_c; /* checkpoint time */ + /* topology */ + int level; /* our level (see above) */ + struct htb_class *parent; /* parent class */ + struct hlist_node hlist; /* classid hash list item */ + struct list_head sibling; /* sibling list item */ + struct list_head children; /* children list */ + + union { + struct htb_class_leaf { + struct Qdisc *q; + int prio; + int aprio; + int quantum; + int deficit[TC_HTB_MAXDEPTH]; + struct list_head drop_list; + } leaf; + struct htb_class_inner { + struct rb_root feed[TC_HTB_NUMPRIO]; /* feed trees */ + struct rb_node *ptr[TC_HTB_NUMPRIO]; /* current class ptr */ + /* When class changes from state 1->2 and disconnects from + parent's feed then we lost ptr value and start from the + first child again. Here we store classid of the + last valid ptr (used when ptr is NULL). */ + u32 last_ptr_id[TC_HTB_NUMPRIO]; + } inner; + } un; + struct rb_node node[TC_HTB_NUMPRIO]; /* node for self or feed tree */ + struct rb_node pq_node; /* node for event queue */ + unsigned long pq_key; /* the same type as jiffies global */ + + int prio_activity; /* for which prios are we active */ + enum htb_cmode cmode; /* current mode of the class */ + + /* class attached filters */ + struct tcf_proto *filter_list; + int filter_cnt; + + int warned; /* only one warning about non work conserving .. */ + + /* token bucket parameters */ + struct qdisc_rate_table *rate; /* rate table of the class itself */ + struct qdisc_rate_table *ceil; /* ceiling rate (limits borrows too) */ + long buffer, cbuffer; /* token bucket depth/rate */ + psched_tdiff_t mbuffer; /* max wait time */ + long tokens, ctokens; /* current number of tokens */ + psched_time_t t_c; /* checkpoint time */ }; /* TODO: maybe compute rate when size is too large .. or drop ? */ -static __inline__ long L2T(struct htb_class *cl,struct qdisc_rate_table *rate, - int size) -{ - int slot = size >> rate->rate.cell_log; - if (slot > 255) { - cl->xstats.giants++; - slot = 255; - } - return rate->data[slot]; +static inline long L2T(struct htb_class *cl, struct qdisc_rate_table *rate, + int size) +{ + int slot = size >> rate->rate.cell_log; + if (slot > 255) { + cl->xstats.giants++; + slot = 255; + } + return rate->data[slot]; } -struct htb_sched -{ - struct list_head root; /* root classes list */ - struct list_head hash[HTB_HSIZE]; /* hashed by classid */ - struct list_head drops[TC_HTB_NUMPRIO]; /* active leaves (for drops) */ - - /* self list - roots of self generating tree */ - struct rb_root row[TC_HTB_MAXDEPTH][TC_HTB_NUMPRIO]; - int row_mask[TC_HTB_MAXDEPTH]; - struct rb_node *ptr[TC_HTB_MAXDEPTH][TC_HTB_NUMPRIO]; - u32 last_ptr_id[TC_HTB_MAXDEPTH][TC_HTB_NUMPRIO]; - - /* self wait list - roots of wait PQs per row */ - struct rb_root wait_pq[TC_HTB_MAXDEPTH]; - - /* time of nearest event per level (row) */ - unsigned long near_ev_cache[TC_HTB_MAXDEPTH]; - - /* cached value of jiffies in dequeue */ - unsigned long jiffies; - - /* whether we hit non-work conserving class during this dequeue; we use */ - int nwc_hit; /* this to disable mindelay complaint in dequeue */ - - int defcls; /* class where unclassified flows go to */ - u32 debug; /* subsystem debug levels */ - - /* filters for qdisc itself */ - struct tcf_proto *filter_list; - int filter_cnt; - - int rate2quantum; /* quant = rate / rate2quantum */ - psched_time_t now; /* cached dequeue time */ - struct timer_list timer; /* send delay timer */ +struct htb_sched { + struct list_head root; /* root classes list */ + struct hlist_head hash[HTB_HSIZE]; /* hashed by classid */ + struct list_head drops[TC_HTB_NUMPRIO];/* active leaves (for drops) */ + + /* self list - roots of self generating tree */ + struct rb_root row[TC_HTB_MAXDEPTH][TC_HTB_NUMPRIO]; + int row_mask[TC_HTB_MAXDEPTH]; + struct rb_node *ptr[TC_HTB_MAXDEPTH][TC_HTB_NUMPRIO]; + u32 last_ptr_id[TC_HTB_MAXDEPTH][TC_HTB_NUMPRIO]; + + /* self wait list - roots of wait PQs per row */ + struct rb_root wait_pq[TC_HTB_MAXDEPTH]; + + /* time of nearest event per level (row) */ + unsigned long near_ev_cache[TC_HTB_MAXDEPTH]; + + /* cached value of jiffies in dequeue */ + unsigned long jiffies; + + /* whether we hit non-work conserving class during this dequeue; we use */ + int nwc_hit; /* this to disable mindelay complaint in dequeue */ + + int defcls; /* class where unclassified flows go to */ + + /* filters for qdisc itself */ + struct tcf_proto *filter_list; + int filter_cnt; + + int rate2quantum; /* quant = rate / rate2quantum */ + psched_time_t now; /* cached dequeue time */ + struct timer_list timer; /* send delay timer */ #ifdef HTB_RATECM - struct timer_list rttim; /* rate computer timer */ - int recmp_bucket; /* which hash bucket to recompute next */ + struct timer_list rttim; /* rate computer timer */ + int recmp_bucket; /* which hash bucket to recompute next */ #endif - - /* non shaped skbs; let them go directly thru */ - struct sk_buff_head direct_queue; - int direct_qlen; /* max qlen of above */ - long direct_pkts; + /* non shaped skbs; let them go directly thru */ + struct sk_buff_head direct_queue; + int direct_qlen; /* max qlen of above */ + + long direct_pkts; }; /* compute hash of size HTB_HSIZE for given handle */ -static __inline__ int htb_hash(u32 h) +static inline int htb_hash(u32 h) { #if HTB_HSIZE != 16 - #error "Declare new hash for your HTB_HSIZE" +#error "Declare new hash for your HTB_HSIZE" #endif - h ^= h>>8; /* stolen from cbq_hash */ - h ^= h>>4; - return h & 0xf; + h ^= h >> 8; /* stolen from cbq_hash */ + h ^= h >> 4; + return h & 0xf; } /* find class in global hash table using given handle */ -static __inline__ struct htb_class *htb_find(u32 handle, struct Qdisc *sch) +static inline struct htb_class *htb_find(u32 handle, struct Qdisc *sch) { struct htb_sched *q = qdisc_priv(sch); - struct list_head *p; - if (TC_H_MAJ(handle) != sch->handle) + struct hlist_node *p; + struct htb_class *cl; + + if (TC_H_MAJ(handle) != sch->handle) return NULL; - - list_for_each (p,q->hash+htb_hash(handle)) { - struct htb_class *cl = list_entry(p,struct htb_class,hlist); + + hlist_for_each_entry(cl, p, q->hash + htb_hash(handle), hlist) { if (cl->classid == handle) return cl; } @@ -304,7 +251,8 @@ static inline u32 htb_classid(struct htb_class *cl) return (cl && cl != HTB_DIRECT) ? cl->classid : TC_H_UNSPEC; } -static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) +static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, + int *qerr) { struct htb_sched *q = qdisc_priv(sch); struct htb_class *cl; @@ -316,8 +264,8 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, in note that nfmark can be used too by attaching filter fw with no rules in it */ if (skb->priority == sch->handle) - return HTB_DIRECT; /* X:0 (direct flow) selected */ - if ((cl = htb_find(skb->priority,sch)) != NULL && cl->level == 0) + return HTB_DIRECT; /* X:0 (direct flow) selected */ + if ((cl = htb_find(skb->priority, sch)) != NULL && cl->level == 0) return cl; *qerr = NET_XMIT_BYPASS; @@ -326,7 +274,7 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, in #ifdef CONFIG_NET_CLS_ACT switch (result) { case TC_ACT_QUEUED: - case TC_ACT_STOLEN: + case TC_ACT_STOLEN: *qerr = NET_XMIT_SUCCESS; case TC_ACT_SHOT: return NULL; @@ -335,97 +283,44 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, in if (result == TC_POLICE_SHOT) return HTB_DIRECT; #endif - if ((cl = (void*)res.class) == NULL) { + if ((cl = (void *)res.class) == NULL) { if (res.classid == sch->handle) - return HTB_DIRECT; /* X:0 (direct flow) */ - if ((cl = htb_find(res.classid,sch)) == NULL) - break; /* filter selected invalid classid */ + return HTB_DIRECT; /* X:0 (direct flow) */ + if ((cl = htb_find(res.classid, sch)) == NULL) + break; /* filter selected invalid classid */ } if (!cl->level) - return cl; /* we hit leaf; return it */ + return cl; /* we hit leaf; return it */ /* we have got inner class; apply inner filter chain */ tcf = cl->filter_list; } /* classification failed; try to use default class */ - cl = htb_find(TC_H_MAKE(TC_H_MAJ(sch->handle),q->defcls),sch); + cl = htb_find(TC_H_MAKE(TC_H_MAJ(sch->handle), q->defcls), sch); if (!cl || cl->level) - return HTB_DIRECT; /* bad default .. this is safe bet */ + return HTB_DIRECT; /* bad default .. this is safe bet */ return cl; } -#ifdef HTB_DEBUG -static void htb_next_rb_node(struct rb_node **n); -#define HTB_DUMTREE(root,memb) if(root) { \ - struct rb_node *n = (root)->rb_node; \ - while (n->rb_left) n = n->rb_left; \ - while (n) { \ - struct htb_class *cl = rb_entry(n, struct htb_class, memb); \ - printk(" %x",cl->classid); htb_next_rb_node (&n); \ - } } - -static void htb_debug_dump (struct htb_sched *q) -{ - int i,p; - printk(KERN_DEBUG "htb*g j=%lu lj=%lu\n",jiffies,q->jiffies); - /* rows */ - for (i=TC_HTB_MAXDEPTH-1;i>=0;i--) { - printk(KERN_DEBUG "htb*r%d m=%x",i,q->row_mask[i]); - for (p=0;p<TC_HTB_NUMPRIO;p++) { - if (!q->row[i][p].rb_node) continue; - printk(" p%d:",p); - HTB_DUMTREE(q->row[i]+p,node[p]); - } - printk("\n"); - } - /* classes */ - for (i = 0; i < HTB_HSIZE; i++) { - struct list_head *l; - list_for_each (l,q->hash+i) { - struct htb_class *cl = list_entry(l,struct htb_class,hlist); - long diff = PSCHED_TDIFF_SAFE(q->now, cl->t_c, (u32)cl->mbuffer); - printk(KERN_DEBUG "htb*c%x m=%d t=%ld c=%ld pq=%lu df=%ld ql=%d " - "pa=%x f:", - cl->classid,cl->cmode,cl->tokens,cl->ctokens, - cl->pq_node.rb_color==-1?0:cl->pq_key,diff, - cl->level?0:cl->un.leaf.q->q.qlen,cl->prio_activity); - if (cl->level) - for (p=0;p<TC_HTB_NUMPRIO;p++) { - if (!cl->un.inner.feed[p].rb_node) continue; - printk(" p%d a=%x:",p,cl->un.inner.ptr[p]?rb_entry(cl->un.inner.ptr[p], struct htb_class,node[p])->classid:0); - HTB_DUMTREE(cl->un.inner.feed+p,node[p]); - } - printk("\n"); - } - } -} -#endif /** * htb_add_to_id_tree - adds class to the round robin list * * Routine adds class to the list (actually tree) sorted by classid. * Make sure that class is not already on such list for given prio. */ -static void htb_add_to_id_tree (HTB_ARGQ struct rb_root *root, - struct htb_class *cl,int prio) +static void htb_add_to_id_tree(struct rb_root *root, + struct htb_class *cl, int prio) { struct rb_node **p = &root->rb_node, *parent = NULL; - HTB_DBG(7,3,"htb_add_id_tree cl=%X prio=%d\n",cl->classid,prio); -#ifdef HTB_DEBUG - if (cl->node[prio].rb_color != -1) { BUG_TRAP(0); return; } - HTB_CHCL(cl); - if (*p) { - struct htb_class *x = rb_entry(*p,struct htb_class,node[prio]); - HTB_CHCL(x); - } -#endif + while (*p) { - struct htb_class *c; parent = *p; + struct htb_class *c; + parent = *p; c = rb_entry(parent, struct htb_class, node[prio]); - HTB_CHCL(c); + if (cl->classid > c->classid) p = &parent->rb_right; - else + else p = &parent->rb_left; } rb_link_node(&cl->node[prio], parent, p); @@ -439,17 +334,11 @@ static void htb_add_to_id_tree (HTB_ARGQ struct rb_root *root, * change its mode in cl->pq_key microseconds. Make sure that class is not * already in the queue. */ -static void htb_add_to_wait_tree (struct htb_sched *q, - struct htb_class *cl,long delay,int debug_hint) +static void htb_add_to_wait_tree(struct htb_sched *q, + struct htb_class *cl, long delay) { struct rb_node **p = &q->wait_pq[cl->level].rb_node, *parent = NULL; - HTB_DBG(7,3,"htb_add_wt cl=%X key=%lu\n",cl->classid,cl->pq_key); -#ifdef HTB_DEBUG - if (cl->pq_node.rb_color != -1) { BUG_TRAP(0); return; } - HTB_CHCL(cl); - if ((delay <= 0 || delay > cl->mbuffer) && net_ratelimit()) - printk(KERN_ERR "HTB: suspicious delay in wait_tree d=%ld cl=%X h=%d\n",delay,cl->classid,debug_hint); -#endif + cl->pq_key = q->jiffies + PSCHED_US2JIFFIE(delay); if (cl->pq_key == q->jiffies) cl->pq_key++; @@ -457,13 +346,14 @@ static void htb_add_to_wait_tree (struct htb_sched *q, /* update the nearest event cache */ if (time_after(q->near_ev_cache[cl->level], cl->pq_key)) q->near_ev_cache[cl->level] = cl->pq_key; - + while (*p) { - struct htb_class *c; parent = *p; + struct htb_class *c; + parent = *p; c = rb_entry(parent, struct htb_class, pq_node); if (time_after_eq(cl->pq_key, c->pq_key)) p = &parent->rb_right; - else + else p = &parent->rb_left; } rb_link_node(&cl->pq_node, parent, p); @@ -476,7 +366,7 @@ static void htb_add_to_wait_tree (struct htb_sched *q, * When we are past last key we return NULL. * Average complexity is 2 steps per call. */ -static void htb_next_rb_node(struct rb_node **n) +static inline void htb_next_rb_node(struct rb_node **n) { *n = rb_next(*n); } @@ -487,42 +377,51 @@ static void htb_next_rb_node(struct rb_node **n) * The class is added to row at priorities marked in mask. * It does nothing if mask == 0. */ -static inline void htb_add_class_to_row(struct htb_sched *q, - struct htb_class *cl,int mask) +static inline void htb_add_class_to_row(struct htb_sched *q, + struct htb_class *cl, int mask) { - HTB_DBG(7,2,"htb_addrow cl=%X mask=%X rmask=%X\n", - cl->classid,mask,q->row_mask[cl->level]); - HTB_CHCL(cl); q->row_mask[cl->level] |= mask; while (mask) { int prio = ffz(~mask); mask &= ~(1 << prio); - htb_add_to_id_tree(HTB_PASSQ q->row[cl->level]+prio,cl,prio); + htb_add_to_id_tree(q->row[cl->level] + prio, cl, prio); + } +} + +/* If this triggers, it is a bug in this code, but it need not be fatal */ +static void htb_safe_rb_erase(struct rb_node *rb, struct rb_root *root) +{ + if (RB_EMPTY_NODE(rb)) { + WARN_ON(1); + } else { + rb_erase(rb, root); + RB_CLEAR_NODE(rb); } } + /** * htb_remove_class_from_row - removes class from its row * * The class is removed from row at priorities marked in mask. * It does nothing if mask == 0. */ -static __inline__ void htb_remove_class_from_row(struct htb_sched *q, - struct htb_class *cl,int mask) +static inline void htb_remove_class_from_row(struct htb_sched *q, + struct htb_class *cl, int mask) { int m = 0; - HTB_CHCL(cl); + while (mask) { int prio = ffz(~mask); + mask &= ~(1 << prio); - if (q->ptr[cl->level][prio] == cl->node+prio) - htb_next_rb_node(q->ptr[cl->level]+prio); - htb_safe_rb_erase(cl->node + prio,q->row[cl->level]+prio); - if (!q->row[cl->level][prio].rb_node) + if (q->ptr[cl->level][prio] == cl->node + prio) + htb_next_rb_node(q->ptr[cl->level] + prio); + + htb_safe_rb_erase(cl->node + prio, q->row[cl->level] + prio); + if (!q->row[cl->level][prio].rb_node) m |= 1 << prio; } - HTB_DBG(7,2,"htb_delrow cl=%X mask=%X rmask=%X maskdel=%X\n", - cl->classid,mask,q->row_mask[cl->level],m); q->row_mask[cl->level] &= ~m; } @@ -533,34 +432,31 @@ static __inline__ void htb_remove_class_from_row(struct htb_sched *q, * for priorities it is participating on. cl->cmode must be new * (activated) mode. It does nothing if cl->prio_activity == 0. */ -static void htb_activate_prios(struct htb_sched *q,struct htb_class *cl) +static void htb_activate_prios(struct htb_sched *q, struct htb_class *cl) { struct htb_class *p = cl->parent; - long m,mask = cl->prio_activity; - HTB_DBG(7,2,"htb_act_prios cl=%X mask=%lX cmode=%d\n",cl->classid,mask,cl->cmode); - HTB_CHCL(cl); + long m, mask = cl->prio_activity; while (cl->cmode == HTB_MAY_BORROW && p && mask) { - HTB_CHCL(p); - m = mask; while (m) { + m = mask; + while (m) { int prio = ffz(~m); m &= ~(1 << prio); - + if (p->un.inner.feed[prio].rb_node) /* parent already has its feed in use so that reset bit in mask as parent is already ok */ mask &= ~(1 << prio); - - htb_add_to_id_tree(HTB_PASSQ p->un.inner.feed+prio,cl,prio); + + htb_add_to_id_tree(p->un.inner.feed + prio, cl, prio); } - HTB_DBG(7,3,"htb_act_pr_aft p=%X pact=%X mask=%lX pmode=%d\n", - p->classid,p->prio_activity,mask,p->cmode); p->prio_activity |= mask; - cl = p; p = cl->parent; - HTB_CHCL(cl); + cl = p; + p = cl->parent; + } if (cl->cmode == HTB_CAN_SEND && mask) - htb_add_class_to_row(q,cl,mask); + htb_add_class_to_row(q, cl, mask); } /** @@ -573,39 +469,52 @@ static void htb_activate_prios(struct htb_sched *q,struct htb_class *cl) static void htb_deactivate_prios(struct htb_sched *q, struct htb_class *cl) { struct htb_class *p = cl->parent; - long m,mask = cl->prio_activity; - HTB_DBG(7,2,"htb_deact_prios cl=%X mask=%lX cmode=%d\n",cl->classid,mask,cl->cmode); - HTB_CHCL(cl); + long m, mask = cl->prio_activity; while (cl->cmode == HTB_MAY_BORROW && p && mask) { - m = mask; mask = 0; + m = mask; + mask = 0; while (m) { int prio = ffz(~m); m &= ~(1 << prio); - - if (p->un.inner.ptr[prio] == cl->node+prio) { + + if (p->un.inner.ptr[prio] == cl->node + prio) { /* we are removing child which is pointed to from parent feed - forget the pointer but remember classid */ p->un.inner.last_ptr_id[prio] = cl->classid; p->un.inner.ptr[prio] = NULL; } - - htb_safe_rb_erase(cl->node + prio,p->un.inner.feed + prio); - - if (!p->un.inner.feed[prio].rb_node) + + htb_safe_rb_erase(cl->node + prio, p->un.inner.feed + prio); + + if (!p->un.inner.feed[prio].rb_node) mask |= 1 << prio; } - HTB_DBG(7,3,"htb_deact_pr_aft p=%X pact=%X mask=%lX pmode=%d\n", - p->classid,p->prio_activity,mask,p->cmode); + p->prio_activity &= ~mask; - cl = p; p = cl->parent; - HTB_CHCL(cl); + cl = p; + p = cl->parent; + } - if (cl->cmode == HTB_CAN_SEND && mask) - htb_remove_class_from_row(q,cl,mask); + if (cl->cmode == HTB_CAN_SEND && mask) + htb_remove_class_from_row(q, cl, mask); } +#if HTB_HYSTERESIS +static inline long htb_lowater(const struct htb_class *cl) +{ + return cl->cmode != HTB_CANT_SEND ? -cl->cbuffer : 0; +} +static inline long htb_hiwater(const struct htb_class *cl) +{ + return cl->cmode == HTB_CAN_SEND ? -cl->buffer : 0; +} +#else +#define htb_lowater(cl) (0) +#define htb_hiwater(cl) (0) +#endif + /** * htb_class_mode - computes and returns current class mode * @@ -617,28 +526,21 @@ static void htb_deactivate_prios(struct htb_sched *q, struct htb_class *cl) * 0 .. -cl->{c,}buffer range. It is meant to limit number of * mode transitions per time unit. The speed gain is about 1/6. */ -static __inline__ enum htb_cmode -htb_class_mode(struct htb_class *cl,long *diff) +static inline enum htb_cmode +htb_class_mode(struct htb_class *cl, long *diff) { - long toks; + long toks; - if ((toks = (cl->ctokens + *diff)) < ( -#if HTB_HYSTERESIS - cl->cmode != HTB_CANT_SEND ? -cl->cbuffer : -#endif - 0)) { - *diff = -toks; - return HTB_CANT_SEND; - } - if ((toks = (cl->tokens + *diff)) >= ( -#if HTB_HYSTERESIS - cl->cmode == HTB_CAN_SEND ? -cl->buffer : -#endif - 0)) - return HTB_CAN_SEND; + if ((toks = (cl->ctokens + *diff)) < htb_lowater(cl)) { + *diff = -toks; + return HTB_CANT_SEND; + } + + if ((toks = (cl->tokens + *diff)) >= htb_hiwater(cl)) + return HTB_CAN_SEND; - *diff = -toks; - return HTB_MAY_BORROW; + *diff = -toks; + return HTB_MAY_BORROW; } /** @@ -650,24 +552,21 @@ htb_class_mode(struct htb_class *cl,long *diff) * be different from old one and cl->pq_key has to be valid if changing * to mode other than HTB_CAN_SEND (see htb_add_to_wait_tree). */ -static void +static void htb_change_class_mode(struct htb_sched *q, struct htb_class *cl, long *diff) -{ - enum htb_cmode new_mode = htb_class_mode(cl,diff); - - HTB_CHCL(cl); - HTB_DBG(7,1,"htb_chging_clmode %d->%d cl=%X\n",cl->cmode,new_mode,cl->classid); +{ + enum htb_cmode new_mode = htb_class_mode(cl, diff); if (new_mode == cl->cmode) - return; - - if (cl->prio_activity) { /* not necessary: speed optimization */ - if (cl->cmode != HTB_CANT_SEND) - htb_deactivate_prios(q,cl); + return; + + if (cl->prio_activity) { /* not necessary: speed optimization */ + if (cl->cmode != HTB_CANT_SEND) + htb_deactivate_prios(q, cl); cl->cmode = new_mode; - if (new_mode != HTB_CANT_SEND) - htb_activate_prios(q,cl); - } else + if (new_mode != HTB_CANT_SEND) + htb_activate_prios(q, cl); + } else cl->cmode = new_mode; } @@ -678,14 +577,15 @@ htb_change_class_mode(struct htb_sched *q, struct htb_class *cl, long *diff) * for the prio. It can be called on already active leaf safely. * It also adds leaf into droplist. */ -static __inline__ void htb_activate(struct htb_sched *q,struct htb_class *cl) +static inline void htb_activate(struct htb_sched *q, struct htb_class *cl) { BUG_TRAP(!cl->level && cl->un.leaf.q && cl->un.leaf.q->q.qlen); - HTB_CHCL(cl); + if (!cl->prio_activity) { cl->prio_activity = 1 << (cl->un.leaf.aprio = cl->un.leaf.prio); - htb_activate_prios(q,cl); - list_add_tail(&cl->un.leaf.drop_list,q->drops+cl->un.leaf.aprio); + htb_activate_prios(q, cl); + list_add_tail(&cl->un.leaf.drop_list, + q->drops + cl->un.leaf.aprio); } } @@ -695,120 +595,120 @@ static __inline__ void htb_activate(struct htb_sched *q,struct htb_class *cl) * Make sure that leaf is active. In the other words it can't be called * with non-active leaf. It also removes class from the drop list. */ -static __inline__ void -htb_deactivate(struct htb_sched *q,struct htb_class *cl) +static inline void htb_deactivate(struct htb_sched *q, struct htb_class *cl) { BUG_TRAP(cl->prio_activity); - HTB_CHCL(cl); - htb_deactivate_prios(q,cl); + + htb_deactivate_prios(q, cl); cl->prio_activity = 0; list_del_init(&cl->un.leaf.drop_list); } static int htb_enqueue(struct sk_buff *skb, struct Qdisc *sch) { - int ret; - struct htb_sched *q = qdisc_priv(sch); - struct htb_class *cl = htb_classify(skb,sch,&ret); - - if (cl == HTB_DIRECT) { - /* enqueue to helper queue */ - if (q->direct_queue.qlen < q->direct_qlen) { - __skb_queue_tail(&q->direct_queue, skb); - q->direct_pkts++; - } else { - kfree_skb(skb); - sch->qstats.drops++; - return NET_XMIT_DROP; - } + int ret; + struct htb_sched *q = qdisc_priv(sch); + struct htb_class *cl = htb_classify(skb, sch, &ret); + + if (cl == HTB_DIRECT) { + /* enqueue to helper queue */ + if (q->direct_queue.qlen < q->direct_qlen) { + __skb_queue_tail(&q->direct_queue, skb); + q->direct_pkts++; + } else { + kfree_skb(skb); + sch->qstats.drops++; + return NET_XMIT_DROP; + } #ifdef CONFIG_NET_CLS_ACT - } else if (!cl) { - if (ret == NET_XMIT_BYPASS) - sch->qstats.drops++; - kfree_skb (skb); - return ret; + } else if (!cl) { + if (ret == NET_XMIT_BYPASS) + sch->qstats.drops++; + kfree_skb(skb); + return ret; #endif - } else if (cl->un.leaf.q->enqueue(skb, cl->un.leaf.q) != NET_XMIT_SUCCESS) { - sch->qstats.drops++; - cl->qstats.drops++; - return NET_XMIT_DROP; - } else { - cl->bstats.packets++; cl->bstats.bytes += skb->len; - htb_activate (q,cl); - } - - sch->q.qlen++; - sch->bstats.packets++; sch->bstats.bytes += skb->len; - HTB_DBG(1,1,"htb_enq_ok cl=%X skb=%p\n",(cl && cl != HTB_DIRECT)?cl->classid:0,skb); - return NET_XMIT_SUCCESS; + } else if (cl->un.leaf.q->enqueue(skb, cl->un.leaf.q) != + NET_XMIT_SUCCESS) { + sch->qstats.drops++; + cl->qstats.drops++; + return NET_XMIT_DROP; + } else { + cl->bstats.packets++; + cl->bstats.bytes += skb->len; + htb_activate(q, cl); + } + + sch->q.qlen++; + sch->bstats.packets++; + sch->bstats.bytes += skb->len; + return NET_XMIT_SUCCESS; } /* TODO: requeuing packet charges it to policers again !! */ static int htb_requeue(struct sk_buff *skb, struct Qdisc *sch) { - struct htb_sched *q = qdisc_priv(sch); - int ret = NET_XMIT_SUCCESS; - struct htb_class *cl = htb_classify(skb,sch, &ret); - struct sk_buff *tskb; - - if (cl == HTB_DIRECT || !cl) { - /* enqueue to helper queue */ - if (q->direct_queue.qlen < q->direct_qlen && cl) { - __skb_queue_head(&q->direct_queue, skb); - } else { - __skb_queue_head(&q->direct_queue, skb); - tskb = __skb_dequeue_tail(&q->direct_queue); - kfree_skb (tskb); - sch->qstats.drops++; - return NET_XMIT_CN; - } - } else if (cl->un.leaf.q->ops->requeue(skb, cl->un.leaf.q) != NET_XMIT_SUCCESS) { - sch->qstats.drops++; - cl->qstats.drops++; - return NET_XMIT_DROP; - } else - htb_activate (q,cl); - - sch->q.qlen++; - sch->qstats.requeues++; - HTB_DBG(1,1,"htb_req_ok cl=%X skb=%p\n",(cl && cl != HTB_DIRECT)?cl->classid:0,skb); - return NET_XMIT_SUCCESS; + struct htb_sched *q = qdisc_priv(sch); + int ret = NET_XMIT_SUCCESS; + struct htb_class *cl = htb_classify(skb, sch, &ret); + struct sk_buff *tskb; + + if (cl == HTB_DIRECT || !cl) { + /* enqueue to helper queue */ + if (q->direct_queue.qlen < q->direct_qlen && cl) { + __skb_queue_head(&q->direct_queue, skb); + } else { + __skb_queue_head(&q->direct_queue, skb); + tskb = __skb_dequeue_tail(&q->direct_queue); + kfree_skb(tskb); + sch->qstats.drops++; + return NET_XMIT_CN; + } + } else if (cl->un.leaf.q->ops->requeue(skb, cl->un.leaf.q) != + NET_XMIT_SUCCESS) { + sch->qstats.drops++; + cl->qstats.drops++; + return NET_XMIT_DROP; + } else + htb_activate(q, cl); + + sch->q.qlen++; + sch->qstats.requeues++; + return NET_XMIT_SUCCESS; } static void htb_timer(unsigned long arg) { - struct Qdisc *sch = (struct Qdisc*)arg; - sch->flags &= ~TCQ_F_THROTTLED; - wmb(); - netif_schedule(sch->dev); + struct Qdisc *sch = (struct Qdisc *)arg; + sch->flags &= ~TCQ_F_THROTTLED; + wmb(); + netif_schedule(sch->dev); } #ifdef HTB_RATECM #define RT_GEN(D,R) R+=D-(R/HTB_EWMAC);D=0 static void htb_rate_timer(unsigned long arg) { - struct Qdisc *sch = (struct Qdisc*)arg; + struct Qdisc *sch = (struct Qdisc *)arg; struct htb_sched *q = qdisc_priv(sch); - struct list_head *p; + struct hlist_node *p; + struct htb_class *cl; + /* lock queue so that we can muck with it */ - HTB_QLOCK(sch); - HTB_DBG(10,1,"htb_rttmr j=%ld\n",jiffies); + spin_lock_bh(&sch->dev->queue_lock); q->rttim.expires = jiffies + HZ; add_timer(&q->rttim); /* scan and recompute one bucket at time */ - if (++q->recmp_bucket >= HTB_HSIZE) + if (++q->recmp_bucket >= HTB_HSIZE) q->recmp_bucket = 0; - list_for_each (p,q->hash+q->recmp_bucket) { - struct htb_class *cl = list_entry(p,struct htb_class,hlist); - HTB_DBG(10,2,"htb_rttmr_cl cl=%X sbyte=%lu spkt=%lu\n", - cl->classid,cl->sum_bytes,cl->sum_packets); - RT_GEN (cl->sum_bytes,cl->rate_bytes); - RT_GEN (cl->sum_packets,cl->rate_packets); + + hlist_for_each_entry(cl,p, q->hash + q->recmp_bucket, hlist) { + RT_GEN(cl->sum_bytes, cl->rate_bytes); + RT_GEN(cl->sum_packets, cl->rate_packets); } - HTB_QUNLOCK(sch); + spin_unlock_bh(&sch->dev->queue_lock); } #endif @@ -823,12 +723,11 @@ static void htb_rate_timer(unsigned long arg) * CAN_SEND) because we can use more precise clock that event queue here. * In such case we remove class from event queue first. */ -static void htb_charge_class(struct htb_sched *q,struct htb_class *cl, - int level,int bytes) -{ - long toks,diff; +static void htb_charge_class(struct htb_sched *q, struct htb_class *cl, + int level, int bytes) +{ + long toks, diff; enum htb_cmode old_mode; - HTB_DBG(5,1,"htb_chrg_cl cl=%X lev=%d len=%d\n",cl->classid,level,bytes); #define HTB_ACCNT(T,B,R) toks = diff + cl->T; \ if (toks > cl->B) toks = cl->B; \ @@ -837,47 +736,31 @@ static void htb_charge_class(struct htb_sched *q,struct htb_class *cl, cl->T = toks while (cl) { - HTB_CHCL(cl); - diff = PSCHED_TDIFF_SAFE(q->now, cl->t_c, (u32)cl->mbuffer); -#ifdef HTB_DEBUG - if (diff > cl->mbuffer || diff < 0 || PSCHED_TLESS(q->now, cl->t_c)) { - if (net_ratelimit()) - printk(KERN_ERR "HTB: bad diff in charge, cl=%X diff=%lX now=%Lu then=%Lu j=%lu\n", - cl->classid, diff, -#ifdef CONFIG_NET_SCH_CLK_GETTIMEOFDAY - q->now.tv_sec * 1000000ULL + q->now.tv_usec, - cl->t_c.tv_sec * 1000000ULL + cl->t_c.tv_usec, -#else - (unsigned long long) q->now, - (unsigned long long) cl->t_c, -#endif - q->jiffies); - diff = 1000; - } -#endif + diff = PSCHED_TDIFF_SAFE(q->now, cl->t_c, (u32) cl->mbuffer); if (cl->level >= level) { - if (cl->level == level) cl->xstats.lends++; - HTB_ACCNT (tokens,buffer,rate); + if (cl->level == level) + cl->xstats.lends++; + HTB_ACCNT(tokens, buffer, rate); } else { cl->xstats.borrows++; - cl->tokens += diff; /* we moved t_c; update tokens */ + cl->tokens += diff; /* we moved t_c; update tokens */ } - HTB_ACCNT (ctokens,cbuffer,ceil); + HTB_ACCNT(ctokens, cbuffer, ceil); cl->t_c = q->now; - HTB_DBG(5,2,"htb_chrg_clp cl=%X diff=%ld tok=%ld ctok=%ld\n",cl->classid,diff,cl->tokens,cl->ctokens); - old_mode = cl->cmode; diff = 0; - htb_change_class_mode(q,cl,&diff); + old_mode = cl->cmode; + diff = 0; + htb_change_class_mode(q, cl, &diff); if (old_mode != cl->cmode) { if (old_mode != HTB_CAN_SEND) - htb_safe_rb_erase(&cl->pq_node,q->wait_pq+cl->level); + htb_safe_rb_erase(&cl->pq_node, q->wait_pq + cl->level); if (cl->cmode != HTB_CAN_SEND) - htb_add_to_wait_tree (q,cl,diff,1); + htb_add_to_wait_tree(q, cl, diff); } - #ifdef HTB_RATECM /* update rate counters */ - cl->sum_bytes += bytes; cl->sum_packets++; + cl->sum_bytes += bytes; + cl->sum_packets++; #endif /* update byte stats except for leaves which are already updated */ @@ -896,60 +779,46 @@ static void htb_charge_class(struct htb_sched *q,struct htb_class *cl, * next pending event (0 for no event in pq). * Note: Aplied are events whose have cl->pq_key <= jiffies. */ -static long htb_do_events(struct htb_sched *q,int level) +static long htb_do_events(struct htb_sched *q, int level) { int i; - HTB_DBG(8,1,"htb_do_events l=%d root=%p rmask=%X\n", - level,q->wait_pq[level].rb_node,q->row_mask[level]); + for (i = 0; i < 500; i++) { struct htb_class *cl; long diff; struct rb_node *p = q->wait_pq[level].rb_node; - if (!p) return 0; - while (p->rb_left) p = p->rb_left; + if (!p) + return 0; + while (p->rb_left) + p = p->rb_left; cl = rb_entry(p, struct htb_class, pq_node); if (time_after(cl->pq_key, q->jiffies)) { - HTB_DBG(8,3,"htb_do_ev_ret delay=%ld\n",cl->pq_key - q->jiffies); return cl->pq_key - q->jiffies; } - htb_safe_rb_erase(p,q->wait_pq+level); - diff = PSCHED_TDIFF_SAFE(q->now, cl->t_c, (u32)cl->mbuffer); -#ifdef HTB_DEBUG - if (diff > cl->mbuffer || diff < 0 || PSCHED_TLESS(q->now, cl->t_c)) { - if (net_ratelimit()) - printk(KERN_ERR "HTB: bad diff in events, cl=%X diff=%lX now=%Lu then=%Lu j=%lu\n", - cl->classid, diff, -#ifdef CONFIG_NET_SCH_CLK_GETTIMEOFDAY - q->now.tv_sec * 1000000ULL + q->now.tv_usec, - cl->t_c.tv_sec * 1000000ULL + cl->t_c.tv_usec, -#else - (unsigned long long) q->now, - (unsigned long long) cl->t_c, -#endif - q->jiffies); - diff = 1000; - } -#endif - htb_change_class_mode(q,cl,&diff); + htb_safe_rb_erase(p, q->wait_pq + level); + diff = PSCHED_TDIFF_SAFE(q->now, cl->t_c, (u32) cl->mbuffer); + htb_change_class_mode(q, cl, &diff); if (cl->cmode != HTB_CAN_SEND) - htb_add_to_wait_tree (q,cl,diff,2); + htb_add_to_wait_tree(q, cl, diff); } if (net_ratelimit()) printk(KERN_WARNING "htb: too many events !\n"); - return HZ/10; + return HZ / 10; } /* Returns class->node+prio from id-tree where classe's id is >= id. NULL is no such one exists. */ -static struct rb_node * -htb_id_find_next_upper(int prio,struct rb_node *n,u32 id) +static struct rb_node *htb_id_find_next_upper(int prio, struct rb_node *n, + u32 id) { struct rb_node *r = NULL; while (n) { - struct htb_class *cl = rb_entry(n,struct htb_class,node[prio]); - if (id == cl->classid) return n; - + struct htb_class *cl = + rb_entry(n, struct htb_class, node[prio]); + if (id == cl->classid) + return n; + if (id > cl->classid) { n = n->rb_right; } else { @@ -965,49 +834,49 @@ htb_id_find_next_upper(int prio,struct rb_node *n,u32 id) * * Find leaf where current feed pointers points to. */ -static struct htb_class * -htb_lookup_leaf(HTB_ARGQ struct rb_root *tree,int prio,struct rb_node **pptr,u32 *pid) +static struct htb_class *htb_lookup_leaf(struct rb_root *tree, int prio, + struct rb_node **pptr, u32 * pid) { int i; struct { struct rb_node *root; struct rb_node **pptr; u32 *pid; - } stk[TC_HTB_MAXDEPTH],*sp = stk; - + } stk[TC_HTB_MAXDEPTH], *sp = stk; + BUG_TRAP(tree->rb_node); sp->root = tree->rb_node; sp->pptr = pptr; sp->pid = pid; for (i = 0; i < 65535; i++) { - HTB_DBG(4,2,"htb_lleaf ptr=%p pid=%X\n",*sp->pptr,*sp->pid); - - if (!*sp->pptr && *sp->pid) { + if (!*sp->pptr && *sp->pid) { /* ptr was invalidated but id is valid - try to recover the original or next ptr */ - *sp->pptr = htb_id_find_next_upper(prio,sp->root,*sp->pid); + *sp->pptr = + htb_id_find_next_upper(prio, sp->root, *sp->pid); } - *sp->pid = 0; /* ptr is valid now so that remove this hint as it - can become out of date quickly */ - if (!*sp->pptr) { /* we are at right end; rewind & go up */ + *sp->pid = 0; /* ptr is valid now so that remove this hint as it + can become out of date quickly */ + if (!*sp->pptr) { /* we are at right end; rewind & go up */ *sp->pptr = sp->root; - while ((*sp->pptr)->rb_left) + while ((*sp->pptr)->rb_left) *sp->pptr = (*sp->pptr)->rb_left; if (sp > stk) { sp--; - BUG_TRAP(*sp->pptr); if(!*sp->pptr) return NULL; - htb_next_rb_node (sp->pptr); + BUG_TRAP(*sp->pptr); + if (!*sp->pptr) + return NULL; + htb_next_rb_node(sp->pptr); } } else { struct htb_class *cl; - cl = rb_entry(*sp->pptr,struct htb_class,node[prio]); - HTB_CHCL(cl); - if (!cl->level) + cl = rb_entry(*sp->pptr, struct htb_class, node[prio]); + if (!cl->level) return cl; (++sp)->root = cl->un.inner.feed[prio].rb_node; - sp->pptr = cl->un.inner.ptr+prio; - sp->pid = cl->un.inner.last_ptr_id+prio; + sp->pptr = cl->un.inner.ptr + prio; + sp->pid = cl->un.inner.last_ptr_id + prio; } } BUG_TRAP(0); @@ -1016,21 +885,21 @@ htb_lookup_leaf(HTB_ARGQ struct rb_root *tree,int prio,struct rb_node **pptr,u32 /* dequeues packet at given priority and level; call only if you are sure that there is active class at prio/level */ -static struct sk_buff * -htb_dequeue_tree(struct htb_sched *q,int prio,int level) +static struct sk_buff *htb_dequeue_tree(struct htb_sched *q, int prio, + int level) { struct sk_buff *skb = NULL; - struct htb_class *cl,*start; + struct htb_class *cl, *start; /* look initial class up in the row */ - start = cl = htb_lookup_leaf (HTB_PASSQ q->row[level]+prio,prio, - q->ptr[level]+prio,q->last_ptr_id[level]+prio); - + start = cl = htb_lookup_leaf(q->row[level] + prio, prio, + q->ptr[level] + prio, + q->last_ptr_id[level] + prio); + do { next: - BUG_TRAP(cl); - if (!cl) return NULL; - HTB_DBG(4,1,"htb_deq_tr prio=%d lev=%d cl=%X defic=%d\n", - prio,level,cl->classid,cl->un.leaf.deficit[level]); + BUG_TRAP(cl); + if (!cl) + return NULL; /* class can be empty - it is unlikely but can be true if leaf qdisc drops packets in enqueue routine or if someone used @@ -1038,64 +907,69 @@ next: simply deactivate and skip such class */ if (unlikely(cl->un.leaf.q->q.qlen == 0)) { struct htb_class *next; - htb_deactivate(q,cl); + htb_deactivate(q, cl); /* row/level might become empty */ if ((q->row_mask[level] & (1 << prio)) == 0) - return NULL; - - next = htb_lookup_leaf (HTB_PASSQ q->row[level]+prio, - prio,q->ptr[level]+prio,q->last_ptr_id[level]+prio); + return NULL; + + next = htb_lookup_leaf(q->row[level] + prio, + prio, q->ptr[level] + prio, + q->last_ptr_id[level] + prio); - if (cl == start) /* fix start if we just deleted it */ + if (cl == start) /* fix start if we just deleted it */ start = next; cl = next; goto next; } - - if (likely((skb = cl->un.leaf.q->dequeue(cl->un.leaf.q)) != NULL)) + + skb = cl->un.leaf.q->dequeue(cl->un.leaf.q); + if (likely(skb != NULL)) break; if (!cl->warned) { - printk(KERN_WARNING "htb: class %X isn't work conserving ?!\n",cl->classid); + printk(KERN_WARNING + "htb: class %X isn't work conserving ?!\n", + cl->classid); cl->warned = 1; } q->nwc_hit++; - htb_next_rb_node((level?cl->parent->un.inner.ptr:q->ptr[0])+prio); - cl = htb_lookup_leaf (HTB_PASSQ q->row[level]+prio,prio,q->ptr[level]+prio, - q->last_ptr_id[level]+prio); + htb_next_rb_node((level ? cl->parent->un.inner.ptr : q-> + ptr[0]) + prio); + cl = htb_lookup_leaf(q->row[level] + prio, prio, + q->ptr[level] + prio, + q->last_ptr_id[level] + prio); } while (cl != start); if (likely(skb != NULL)) { if ((cl->un.leaf.deficit[level] -= skb->len) < 0) { - HTB_DBG(4,2,"htb_next_cl oldptr=%p quant_add=%d\n", - level?cl->parent->un.inner.ptr[prio]:q->ptr[0][prio],cl->un.leaf.quantum); cl->un.leaf.deficit[level] += cl->un.leaf.quantum; - htb_next_rb_node((level?cl->parent->un.inner.ptr:q->ptr[0])+prio); + htb_next_rb_node((level ? cl->parent->un.inner.ptr : q-> + ptr[0]) + prio); } /* this used to be after charge_class but this constelation gives us slightly better performance */ if (!cl->un.leaf.q->q.qlen) - htb_deactivate (q,cl); - htb_charge_class (q,cl,level,skb->len); + htb_deactivate(q, cl); + htb_charge_class(q, cl, level, skb->len); } return skb; } -static void htb_delay_by(struct Qdisc *sch,long delay) +static void htb_delay_by(struct Qdisc *sch, long delay) { struct htb_sched *q = qdisc_priv(sch); - if (delay <= 0) delay = 1; - if (unlikely(delay > 5*HZ)) { + if (delay <= 0) + delay = 1; + if (unlikely(delay > 5 * HZ)) { if (net_ratelimit()) printk(KERN_INFO "HTB delay %ld > 5sec\n", delay); - delay = 5*HZ; + delay = 5 * HZ; } /* why don't use jiffies here ? because expires can be in past */ mod_timer(&q->timer, q->jiffies + delay); sch->flags |= TCQ_F_THROTTLED; sch->qstats.overlimits++; - HTB_DBG(3,1,"htb_deq t_delay=%ld\n",delay); } static struct sk_buff *htb_dequeue(struct Qdisc *sch) @@ -1104,22 +978,19 @@ static struct sk_buff *htb_dequeue(struct Qdisc *sch) struct htb_sched *q = qdisc_priv(sch); int level; long min_delay; -#ifdef HTB_DEBUG - int evs_used = 0; -#endif q->jiffies = jiffies; - HTB_DBG(3,1,"htb_deq dircnt=%d qlen=%d\n",skb_queue_len(&q->direct_queue), - sch->q.qlen); /* try to dequeue direct packets as high prio (!) to minimize cpu work */ - if ((skb = __skb_dequeue(&q->direct_queue)) != NULL) { + skb = __skb_dequeue(&q->direct_queue); + if (skb != NULL) { sch->flags &= ~TCQ_F_THROTTLED; sch->q.qlen--; return skb; } - if (!sch->q.qlen) goto fin; + if (!sch->q.qlen) + goto fin; PSCHED_GET_TIME(q->now); min_delay = LONG_MAX; @@ -1129,21 +1000,19 @@ static struct sk_buff *htb_dequeue(struct Qdisc *sch) int m; long delay; if (time_after_eq(q->jiffies, q->near_ev_cache[level])) { - delay = htb_do_events(q,level); - q->near_ev_cache[level] = q->jiffies + (delay ? delay : HZ); -#ifdef HTB_DEBUG - evs_used++; -#endif + delay = htb_do_events(q, level); + q->near_ev_cache[level] = + q->jiffies + (delay ? delay : HZ); } else - delay = q->near_ev_cache[level] - q->jiffies; - - if (delay && min_delay > delay) + delay = q->near_ev_cache[level] - q->jiffies; + + if (delay && min_delay > delay) min_delay = delay; m = ~q->row_mask[level]; while (m != (int)(-1)) { - int prio = ffz (m); + int prio = ffz(m); m |= 1 << prio; - skb = htb_dequeue_tree(q,prio,level); + skb = htb_dequeue_tree(q, prio, level); if (likely(skb != NULL)) { sch->q.qlen--; sch->flags &= ~TCQ_F_THROTTLED; @@ -1151,40 +1020,28 @@ static struct sk_buff *htb_dequeue(struct Qdisc *sch) } } } -#ifdef HTB_DEBUG - if (!q->nwc_hit && min_delay >= 10*HZ && net_ratelimit()) { - if (min_delay == LONG_MAX) { - printk(KERN_ERR "HTB: dequeue bug (%d,%lu,%lu), report it please !\n", - evs_used,q->jiffies,jiffies); - htb_debug_dump(q); - } else - printk(KERN_WARNING "HTB: mindelay=%ld, some class has " - "too small rate\n",min_delay); - } -#endif - htb_delay_by (sch,min_delay > 5*HZ ? 5*HZ : min_delay); + htb_delay_by(sch, min_delay > 5 * HZ ? 5 * HZ : min_delay); fin: - HTB_DBG(3,1,"htb_deq_end %s j=%lu skb=%p\n",sch->dev->name,q->jiffies,skb); return skb; } /* try to drop from each class (by prio) until one succeed */ -static unsigned int htb_drop(struct Qdisc* sch) +static unsigned int htb_drop(struct Qdisc *sch) { struct htb_sched *q = qdisc_priv(sch); int prio; for (prio = TC_HTB_NUMPRIO - 1; prio >= 0; prio--) { struct list_head *p; - list_for_each (p,q->drops+prio) { + list_for_each(p, q->drops + prio) { struct htb_class *cl = list_entry(p, struct htb_class, un.leaf.drop_list); unsigned int len; - if (cl->un.leaf.q->ops->drop && - (len = cl->un.leaf.q->ops->drop(cl->un.leaf.q))) { + if (cl->un.leaf.q->ops->drop && + (len = cl->un.leaf.q->ops->drop(cl->un.leaf.q))) { sch->q.qlen--; if (!cl->un.leaf.q->q.qlen) - htb_deactivate (q,cl); + htb_deactivate(q, cl); return len; } } @@ -1194,29 +1051,25 @@ static unsigned int htb_drop(struct Qdisc* sch) /* reset all classes */ /* always caled under BH & queue lock */ -static void htb_reset(struct Qdisc* sch) +static void htb_reset(struct Qdisc *sch) { struct htb_sched *q = qdisc_priv(sch); int i; - HTB_DBG(0,1,"htb_reset sch=%p, handle=%X\n",sch,sch->handle); for (i = 0; i < HTB_HSIZE; i++) { - struct list_head *p; - list_for_each (p,q->hash+i) { - struct htb_class *cl = list_entry(p,struct htb_class,hlist); + struct hlist_node *p; + struct htb_class *cl; + + hlist_for_each_entry(cl, p, q->hash + i, hlist) { if (cl->level) - memset(&cl->un.inner,0,sizeof(cl->un.inner)); + memset(&cl->un.inner, 0, sizeof(cl->un.inner)); else { - if (cl->un.leaf.q) + if (cl->un.leaf.q) qdisc_reset(cl->un.leaf.q); INIT_LIST_HEAD(&cl->un.leaf.drop_list); } cl->prio_activity = 0; cl->cmode = HTB_CAN_SEND; -#ifdef HTB_DEBUG - cl->pq_node.rb_color = -1; - memset(cl->node,255,sizeof(cl->node)); -#endif } } @@ -1224,12 +1077,12 @@ static void htb_reset(struct Qdisc* sch) del_timer(&q->timer); __skb_queue_purge(&q->direct_queue); sch->q.qlen = 0; - memset(q->row,0,sizeof(q->row)); - memset(q->row_mask,0,sizeof(q->row_mask)); - memset(q->wait_pq,0,sizeof(q->wait_pq)); - memset(q->ptr,0,sizeof(q->ptr)); + memset(q->row, 0, sizeof(q->row)); + memset(q->row_mask, 0, sizeof(q->row_mask)); + memset(q->wait_pq, 0, sizeof(q->wait_pq)); + memset(q->ptr, 0, sizeof(q->ptr)); for (i = 0; i < TC_HTB_NUMPRIO; i++) - INIT_LIST_HEAD(q->drops+i); + INIT_LIST_HEAD(q->drops + i); } static int htb_init(struct Qdisc *sch, struct rtattr *opt) @@ -1238,36 +1091,31 @@ static int htb_init(struct Qdisc *sch, struct rtattr *opt) struct rtattr *tb[TCA_HTB_INIT]; struct tc_htb_glob *gopt; int i; -#ifdef HTB_DEBUG - printk(KERN_INFO "HTB init, kernel part version %d.%d\n", - HTB_VER >> 16,HTB_VER & 0xffff); -#endif if (!opt || rtattr_parse_nested(tb, TCA_HTB_INIT, opt) || - tb[TCA_HTB_INIT-1] == NULL || - RTA_PAYLOAD(tb[TCA_HTB_INIT-1]) < sizeof(*gopt)) { + tb[TCA_HTB_INIT - 1] == NULL || + RTA_PAYLOAD(tb[TCA_HTB_INIT - 1]) < sizeof(*gopt)) { printk(KERN_ERR "HTB: hey probably you have bad tc tool ?\n"); return -EINVAL; } - gopt = RTA_DATA(tb[TCA_HTB_INIT-1]); + gopt = RTA_DATA(tb[TCA_HTB_INIT - 1]); if (gopt->version != HTB_VER >> 16) { - printk(KERN_ERR "HTB: need tc/htb version %d (minor is %d), you have %d\n", - HTB_VER >> 16,HTB_VER & 0xffff,gopt->version); + printk(KERN_ERR + "HTB: need tc/htb version %d (minor is %d), you have %d\n", + HTB_VER >> 16, HTB_VER & 0xffff, gopt->version); return -EINVAL; } - q->debug = gopt->debug; - HTB_DBG(0,1,"htb_init sch=%p handle=%X r2q=%d\n",sch,sch->handle,gopt->rate2quantum); INIT_LIST_HEAD(&q->root); for (i = 0; i < HTB_HSIZE; i++) - INIT_LIST_HEAD(q->hash+i); + INIT_HLIST_HEAD(q->hash + i); for (i = 0; i < TC_HTB_NUMPRIO; i++) - INIT_LIST_HEAD(q->drops+i); + INIT_LIST_HEAD(q->drops + i); init_timer(&q->timer); skb_queue_head_init(&q->direct_queue); q->direct_qlen = sch->dev->tx_queue_len; - if (q->direct_qlen < 2) /* some devices have zero tx_queue_len */ + if (q->direct_qlen < 2) /* some devices have zero tx_queue_len */ q->direct_qlen = 2; q->timer.function = htb_timer; q->timer.data = (unsigned long)sch; @@ -1289,80 +1137,72 @@ static int htb_init(struct Qdisc *sch, struct rtattr *opt) static int htb_dump(struct Qdisc *sch, struct sk_buff *skb) { struct htb_sched *q = qdisc_priv(sch); - unsigned char *b = skb->tail; + unsigned char *b = skb->tail; struct rtattr *rta; struct tc_htb_glob gopt; - HTB_DBG(0,1,"htb_dump sch=%p, handle=%X\n",sch,sch->handle); - HTB_QLOCK(sch); + spin_lock_bh(&sch->dev->queue_lock); gopt.direct_pkts = q->direct_pkts; -#ifdef HTB_DEBUG - if (HTB_DBG_COND(0,2)) - htb_debug_dump(q); -#endif gopt.version = HTB_VER; gopt.rate2quantum = q->rate2quantum; gopt.defcls = q->defcls; - gopt.debug = q->debug; - rta = (struct rtattr*)b; + gopt.debug = 0; + rta = (struct rtattr *)b; RTA_PUT(skb, TCA_OPTIONS, 0, NULL); RTA_PUT(skb, TCA_HTB_INIT, sizeof(gopt), &gopt); rta->rta_len = skb->tail - b; - HTB_QUNLOCK(sch); + spin_unlock_bh(&sch->dev->queue_lock); return skb->len; rtattr_failure: - HTB_QUNLOCK(sch); + spin_unlock_bh(&sch->dev->queue_lock); skb_trim(skb, skb->tail - skb->data); return -1; } static int htb_dump_class(struct Qdisc *sch, unsigned long arg, - struct sk_buff *skb, struct tcmsg *tcm) + struct sk_buff *skb, struct tcmsg *tcm) { -#ifdef HTB_DEBUG - struct htb_sched *q = qdisc_priv(sch); -#endif - struct htb_class *cl = (struct htb_class*)arg; - unsigned char *b = skb->tail; + struct htb_class *cl = (struct htb_class *)arg; + unsigned char *b = skb->tail; struct rtattr *rta; struct tc_htb_opt opt; - HTB_DBG(0,1,"htb_dump_class handle=%X clid=%X\n",sch->handle,cl->classid); - - HTB_QLOCK(sch); + spin_lock_bh(&sch->dev->queue_lock); tcm->tcm_parent = cl->parent ? cl->parent->classid : TC_H_ROOT; tcm->tcm_handle = cl->classid; if (!cl->level && cl->un.leaf.q) tcm->tcm_info = cl->un.leaf.q->handle; - rta = (struct rtattr*)b; + rta = (struct rtattr *)b; RTA_PUT(skb, TCA_OPTIONS, 0, NULL); - memset (&opt,0,sizeof(opt)); + memset(&opt, 0, sizeof(opt)); - opt.rate = cl->rate->rate; opt.buffer = cl->buffer; - opt.ceil = cl->ceil->rate; opt.cbuffer = cl->cbuffer; - opt.quantum = cl->un.leaf.quantum; opt.prio = cl->un.leaf.prio; - opt.level = cl->level; + opt.rate = cl->rate->rate; + opt.buffer = cl->buffer; + opt.ceil = cl->ceil->rate; + opt.cbuffer = cl->cbuffer; + opt.quantum = cl->un.leaf.quantum; + opt.prio = cl->un.leaf.prio; + opt.level = cl->level; RTA_PUT(skb, TCA_HTB_PARMS, sizeof(opt), &opt); rta->rta_len = skb->tail - b; - HTB_QUNLOCK(sch); + spin_unlock_bh(&sch->dev->queue_lock); return skb->len; rtattr_failure: - HTB_QUNLOCK(sch); + spin_unlock_bh(&sch->dev->queue_lock); skb_trim(skb, b - skb->data); return -1; } static int -htb_dump_class_stats(struct Qdisc *sch, unsigned long arg, - struct gnet_dump *d) +htb_dump_class_stats(struct Qdisc *sch, unsigned long arg, struct gnet_dump *d) { - struct htb_class *cl = (struct htb_class*)arg; + struct htb_class *cl = (struct htb_class *)arg; #ifdef HTB_RATECM - cl->rate_est.bps = cl->rate_bytes/(HTB_EWMAC*HTB_HSIZE); - cl->rate_est.pps = cl->rate_packets/(HTB_EWMAC*HTB_HSIZE); + cl->rate_est.bps = cl->rate_bytes / (HTB_EWMAC * HTB_HSIZE); + cl->rate_est.pps = cl->rate_packets / (HTB_EWMAC * HTB_HSIZE); #endif if (!cl->level && cl->un.leaf.q) @@ -1379,21 +1219,22 @@ htb_dump_class_stats(struct Qdisc *sch, unsigned long arg, } static int htb_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new, - struct Qdisc **old) + struct Qdisc **old) { - struct htb_class *cl = (struct htb_class*)arg; + struct htb_class *cl = (struct htb_class *)arg; if (cl && !cl->level) { - if (new == NULL && (new = qdisc_create_dflt(sch->dev, - &pfifo_qdisc_ops)) == NULL) - return -ENOBUFS; + if (new == NULL && (new = qdisc_create_dflt(sch->dev, + &pfifo_qdisc_ops)) + == NULL) + return -ENOBUFS; sch_tree_lock(sch); if ((*old = xchg(&cl->un.leaf.q, new)) != NULL) { if (cl->prio_activity) - htb_deactivate (qdisc_priv(sch),cl); + htb_deactivate(qdisc_priv(sch), cl); /* TODO: is it correct ? Why CBQ doesn't do it ? */ - sch->q.qlen -= (*old)->q.qlen; + sch->q.qlen -= (*old)->q.qlen; qdisc_reset(*old); } sch_tree_unlock(sch); @@ -1402,20 +1243,16 @@ static int htb_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new, return -ENOENT; } -static struct Qdisc * htb_leaf(struct Qdisc *sch, unsigned long arg) +static struct Qdisc *htb_leaf(struct Qdisc *sch, unsigned long arg) { - struct htb_class *cl = (struct htb_class*)arg; + struct htb_class *cl = (struct htb_class *)arg; return (cl && !cl->level) ? cl->un.leaf.q : NULL; } static unsigned long htb_get(struct Qdisc *sch, u32 classid) { -#ifdef HTB_DEBUG - struct htb_sched *q = qdisc_priv(sch); -#endif - struct htb_class *cl = htb_find(classid,sch); - HTB_DBG(0,1,"htb_get clid=%X q=%p cl=%p ref=%d\n",classid,q,cl,cl?cl->refcnt:0); - if (cl) + struct htb_class *cl = htb_find(classid, sch); + if (cl) cl->refcnt++; return (unsigned long)cl; } @@ -1430,10 +1267,9 @@ static void htb_destroy_filters(struct tcf_proto **fl) } } -static void htb_destroy_class(struct Qdisc* sch,struct htb_class *cl) +static void htb_destroy_class(struct Qdisc *sch, struct htb_class *cl) { struct htb_sched *q = qdisc_priv(sch); - HTB_DBG(0,1,"htb_destrycls clid=%X ref=%d\n", cl?cl->classid:0,cl?cl->refcnt:0); if (!cl->level) { BUG_TRAP(cl->un.leaf.q); sch->q.qlen -= cl->un.leaf.q->q.qlen; @@ -1441,45 +1277,45 @@ static void htb_destroy_class(struct Qdisc* sch,struct htb_class *cl) } qdisc_put_rtab(cl->rate); qdisc_put_rtab(cl->ceil); - - htb_destroy_filters (&cl->filter_list); - - while (!list_empty(&cl->children)) - htb_destroy_class (sch,list_entry(cl->children.next, - struct htb_class,sibling)); + + htb_destroy_filters(&cl->filter_list); + + while (!list_empty(&cl->children)) + htb_destroy_class(sch, list_entry(cl->children.next, + struct htb_class, sibling)); /* note: this delete may happen twice (see htb_delete) */ - list_del(&cl->hlist); + if (!hlist_unhashed(&cl->hlist)) + hlist_del(&cl->hlist); list_del(&cl->sibling); - + if (cl->prio_activity) - htb_deactivate (q,cl); - + htb_deactivate(q, cl); + if (cl->cmode != HTB_CAN_SEND) - htb_safe_rb_erase(&cl->pq_node,q->wait_pq+cl->level); - + htb_safe_rb_erase(&cl->pq_node, q->wait_pq + cl->level); + kfree(cl); } /* always caled under BH & queue lock */ -static void htb_destroy(struct Qdisc* sch) +static void htb_destroy(struct Qdisc *sch) { struct htb_sched *q = qdisc_priv(sch); - HTB_DBG(0,1,"htb_destroy q=%p\n",q); - del_timer_sync (&q->timer); + del_timer_sync(&q->timer); #ifdef HTB_RATECM - del_timer_sync (&q->rttim); + del_timer_sync(&q->rttim); #endif /* This line used to be after htb_destroy_class call below and surprisingly it worked in 2.4. But it must precede it because filter need its target class alive to be able to call unbind_filter on it (without Oops). */ htb_destroy_filters(&q->filter_list); - - while (!list_empty(&q->root)) - htb_destroy_class (sch,list_entry(q->root.next, - struct htb_class,sibling)); + + while (!list_empty(&q->root)) + htb_destroy_class(sch, list_entry(q->root.next, + struct htb_class, sibling)); __skb_queue_purge(&q->direct_queue); } @@ -1487,24 +1323,25 @@ static void htb_destroy(struct Qdisc* sch) static int htb_delete(struct Qdisc *sch, unsigned long arg) { struct htb_sched *q = qdisc_priv(sch); - struct htb_class *cl = (struct htb_class*)arg; - HTB_DBG(0,1,"htb_delete q=%p cl=%X ref=%d\n",q,cl?cl->classid:0,cl?cl->refcnt:0); + struct htb_class *cl = (struct htb_class *)arg; // TODO: why don't allow to delete subtree ? references ? does // tc subsys quarantee us that in htb_destroy it holds no class // refs so that we can remove children safely there ? if (!list_empty(&cl->children) || cl->filter_cnt) return -EBUSY; - + sch_tree_lock(sch); - + /* delete from hash and active; remainder in destroy_class */ - list_del_init(&cl->hlist); + if (!hlist_unhashed(&cl->hlist)) + hlist_del(&cl->hlist); + if (cl->prio_activity) - htb_deactivate (q,cl); + htb_deactivate(q, cl); if (--cl->refcnt == 0) - htb_destroy_class(sch,cl); + htb_destroy_class(sch, cl); sch_tree_unlock(sch); return 0; @@ -1512,45 +1349,46 @@ static int htb_delete(struct Qdisc *sch, unsigned long arg) static void htb_put(struct Qdisc *sch, unsigned long arg) { -#ifdef HTB_DEBUG - struct htb_sched *q = qdisc_priv(sch); -#endif - struct htb_class *cl = (struct htb_class*)arg; - HTB_DBG(0,1,"htb_put q=%p cl=%X ref=%d\n",q,cl?cl->classid:0,cl?cl->refcnt:0); + struct htb_class *cl = (struct htb_class *)arg; if (--cl->refcnt == 0) - htb_destroy_class(sch,cl); + htb_destroy_class(sch, cl); } -static int htb_change_class(struct Qdisc *sch, u32 classid, - u32 parentid, struct rtattr **tca, unsigned long *arg) +static int htb_change_class(struct Qdisc *sch, u32 classid, + u32 parentid, struct rtattr **tca, + unsigned long *arg) { int err = -EINVAL; struct htb_sched *q = qdisc_priv(sch); - struct htb_class *cl = (struct htb_class*)*arg,*parent; - struct rtattr *opt = tca[TCA_OPTIONS-1]; + struct htb_class *cl = (struct htb_class *)*arg, *parent; + struct rtattr *opt = tca[TCA_OPTIONS - 1]; struct qdisc_rate_table *rtab = NULL, *ctab = NULL; struct rtattr *tb[TCA_HTB_RTAB]; struct tc_htb_opt *hopt; /* extract all subattrs from opt attr */ if (!opt || rtattr_parse_nested(tb, TCA_HTB_RTAB, opt) || - tb[TCA_HTB_PARMS-1] == NULL || - RTA_PAYLOAD(tb[TCA_HTB_PARMS-1]) < sizeof(*hopt)) + tb[TCA_HTB_PARMS - 1] == NULL || + RTA_PAYLOAD(tb[TCA_HTB_PARMS - 1]) < sizeof(*hopt)) goto failure; - - parent = parentid == TC_H_ROOT ? NULL : htb_find (parentid,sch); - hopt = RTA_DATA(tb[TCA_HTB_PARMS-1]); - HTB_DBG(0,1,"htb_chg cl=%p(%X), clid=%X, parid=%X, opt/prio=%d, rate=%u, buff=%d, quant=%d\n", cl,cl?cl->classid:0,classid,parentid,(int)hopt->prio,hopt->rate.rate,hopt->buffer,hopt->quantum); - rtab = qdisc_get_rtab(&hopt->rate, tb[TCA_HTB_RTAB-1]); - ctab = qdisc_get_rtab(&hopt->ceil, tb[TCA_HTB_CTAB-1]); - if (!rtab || !ctab) goto failure; + parent = parentid == TC_H_ROOT ? NULL : htb_find(parentid, sch); + + hopt = RTA_DATA(tb[TCA_HTB_PARMS - 1]); + + rtab = qdisc_get_rtab(&hopt->rate, tb[TCA_HTB_RTAB - 1]); + ctab = qdisc_get_rtab(&hopt->ceil, tb[TCA_HTB_CTAB - 1]); + if (!rtab || !ctab) + goto failure; - if (!cl) { /* new class */ + if (!cl) { /* new class */ struct Qdisc *new_q; + int prio; + /* check for valid classid */ - if (!classid || TC_H_MAJ(classid^sch->handle) || htb_find(classid,sch)) + if (!classid || TC_H_MAJ(classid ^ sch->handle) + || htb_find(classid, sch)) goto failure; /* check maximal depth */ @@ -1561,15 +1399,16 @@ static int htb_change_class(struct Qdisc *sch, u32 classid, err = -ENOBUFS; if ((cl = kzalloc(sizeof(*cl), GFP_KERNEL)) == NULL) goto failure; - + cl->refcnt = 1; INIT_LIST_HEAD(&cl->sibling); - INIT_LIST_HEAD(&cl->hlist); + INIT_HLIST_NODE(&cl->hlist); INIT_LIST_HEAD(&cl->children); INIT_LIST_HEAD(&cl->un.leaf.drop_list); -#ifdef HTB_DEBUG - cl->magic = HTB_CMAGIC; -#endif + RB_CLEAR_NODE(&cl->pq_node); + + for (prio = 0; prio < TC_HTB_NUMPRIO; prio++) + RB_CLEAR_NODE(&cl->node[prio]); /* create leaf qdisc early because it uses kmalloc(GFP_KERNEL) so that can't be used inside of sch_tree_lock @@ -1579,53 +1418,53 @@ static int htb_change_class(struct Qdisc *sch, u32 classid, if (parent && !parent->level) { /* turn parent into inner node */ sch->q.qlen -= parent->un.leaf.q->q.qlen; - qdisc_destroy (parent->un.leaf.q); - if (parent->prio_activity) - htb_deactivate (q,parent); + qdisc_destroy(parent->un.leaf.q); + if (parent->prio_activity) + htb_deactivate(q, parent); /* remove from evt list because of level change */ if (parent->cmode != HTB_CAN_SEND) { - htb_safe_rb_erase(&parent->pq_node,q->wait_pq /*+0*/); + htb_safe_rb_erase(&parent->pq_node, q->wait_pq); parent->cmode = HTB_CAN_SEND; } parent->level = (parent->parent ? parent->parent->level - : TC_HTB_MAXDEPTH) - 1; - memset (&parent->un.inner,0,sizeof(parent->un.inner)); + : TC_HTB_MAXDEPTH) - 1; + memset(&parent->un.inner, 0, sizeof(parent->un.inner)); } /* leaf (we) needs elementary qdisc */ cl->un.leaf.q = new_q ? new_q : &noop_qdisc; - cl->classid = classid; cl->parent = parent; + cl->classid = classid; + cl->parent = parent; /* set class to be in HTB_CAN_SEND state */ cl->tokens = hopt->buffer; cl->ctokens = hopt->cbuffer; - cl->mbuffer = PSCHED_JIFFIE2US(HZ*60); /* 1min */ + cl->mbuffer = PSCHED_JIFFIE2US(HZ * 60); /* 1min */ PSCHED_GET_TIME(cl->t_c); cl->cmode = HTB_CAN_SEND; /* attach to the hash list and parent's family */ - list_add_tail(&cl->hlist, q->hash+htb_hash(classid)); - list_add_tail(&cl->sibling, parent ? &parent->children : &q->root); -#ifdef HTB_DEBUG - { - int i; - for (i = 0; i < TC_HTB_NUMPRIO; i++) cl->node[i].rb_color = -1; - cl->pq_node.rb_color = -1; - } -#endif - } else sch_tree_lock(sch); + hlist_add_head(&cl->hlist, q->hash + htb_hash(classid)); + list_add_tail(&cl->sibling, + parent ? &parent->children : &q->root); + } else + sch_tree_lock(sch); /* it used to be a nasty bug here, we have to check that node - is really leaf before changing cl->un.leaf ! */ + is really leaf before changing cl->un.leaf ! */ if (!cl->level) { cl->un.leaf.quantum = rtab->rate.rate / q->rate2quantum; if (!hopt->quantum && cl->un.leaf.quantum < 1000) { - printk(KERN_WARNING "HTB: quantum of class %X is small. Consider r2q change.\n", cl->classid); + printk(KERN_WARNING + "HTB: quantum of class %X is small. Consider r2q change.\n", + cl->classid); cl->un.leaf.quantum = 1000; } if (!hopt->quantum && cl->un.leaf.quantum > 200000) { - printk(KERN_WARNING "HTB: quantum of class %X is big. Consider r2q change.\n", cl->classid); + printk(KERN_WARNING + "HTB: quantum of class %X is big. Consider r2q change.\n", + cl->classid); cl->un.leaf.quantum = 200000; } if (hopt->quantum) @@ -1636,16 +1475,22 @@ static int htb_change_class(struct Qdisc *sch, u32 classid, cl->buffer = hopt->buffer; cl->cbuffer = hopt->cbuffer; - if (cl->rate) qdisc_put_rtab(cl->rate); cl->rate = rtab; - if (cl->ceil) qdisc_put_rtab(cl->ceil); cl->ceil = ctab; + if (cl->rate) + qdisc_put_rtab(cl->rate); + cl->rate = rtab; + if (cl->ceil) + qdisc_put_rtab(cl->ceil); + cl->ceil = ctab; sch_tree_unlock(sch); *arg = (unsigned long)cl; return 0; failure: - if (rtab) qdisc_put_rtab(rtab); - if (ctab) qdisc_put_rtab(ctab); + if (rtab) + qdisc_put_rtab(rtab); + if (ctab) + qdisc_put_rtab(ctab); return err; } @@ -1654,28 +1499,28 @@ static struct tcf_proto **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; - HTB_DBG(0,2,"htb_tcf q=%p clid=%X fref=%d fl=%p\n",q,cl?cl->classid:0,cl?cl->filter_cnt:q->filter_cnt,*fl); + return fl; } static unsigned long htb_bind_filter(struct Qdisc *sch, unsigned long parent, - u32 classid) + u32 classid) { struct htb_sched *q = qdisc_priv(sch); - struct htb_class *cl = htb_find (classid,sch); - HTB_DBG(0,2,"htb_bind q=%p clid=%X cl=%p fref=%d\n",q,classid,cl,cl?cl->filter_cnt:q->filter_cnt); + struct htb_class *cl = htb_find(classid, sch); + /*if (cl && !cl->level) return 0; - The line above used to be there to prevent attaching filters to - leaves. But at least tc_index filter uses this just to get class - for other reasons so that we have to allow for it. - ---- - 19.6.2002 As Werner explained it is ok - bind filter is just - another way to "lock" the class - unlike "get" this lock can - be broken by class during destroy IIUC. + The line above used to be there to prevent attaching filters to + leaves. But at least tc_index filter uses this just to get class + for other reasons so that we have to allow for it. + ---- + 19.6.2002 As Werner explained it is ok - bind filter is just + another way to "lock" the class - unlike "get" this lock can + be broken by class during destroy IIUC. */ - if (cl) - cl->filter_cnt++; - else + if (cl) + cl->filter_cnt++; + else q->filter_cnt++; return (unsigned long)cl; } @@ -1684,10 +1529,10 @@ static void htb_unbind_filter(struct Qdisc *sch, unsigned long arg) { struct htb_sched *q = qdisc_priv(sch); struct htb_class *cl = (struct htb_class *)arg; - HTB_DBG(0,2,"htb_unbind q=%p cl=%p fref=%d\n",q,cl,cl?cl->filter_cnt:q->filter_cnt); - if (cl) - cl->filter_cnt--; - else + + if (cl) + cl->filter_cnt--; + else q->filter_cnt--; } @@ -1700,9 +1545,10 @@ static void htb_walk(struct Qdisc *sch, struct qdisc_walker *arg) return; for (i = 0; i < HTB_HSIZE; i++) { - struct list_head *p; - list_for_each (p,q->hash+i) { - struct htb_class *cl = list_entry(p,struct htb_class,hlist); + struct hlist_node *p; + struct htb_class *cl; + + hlist_for_each_entry(cl, p, q->hash + i, hlist) { if (arg->count < arg->skip) { arg->count++; continue; @@ -1750,12 +1596,13 @@ static struct Qdisc_ops htb_qdisc_ops = { static int __init htb_module_init(void) { - return register_qdisc(&htb_qdisc_ops); + return register_qdisc(&htb_qdisc_ops); } -static void __exit htb_module_exit(void) +static void __exit htb_module_exit(void) { - unregister_qdisc(&htb_qdisc_ops); + unregister_qdisc(&htb_qdisc_ops); } + module_init(htb_module_init) module_exit(htb_module_exit) MODULE_LICENSE("GPL"); diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index a08ec4c..45939ba 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -192,8 +192,8 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch) */ if (q->corrupt && q->corrupt >= get_crandom(&q->corrupt_cor)) { if (!(skb = skb_unshare(skb, GFP_ATOMIC)) - || (skb->ip_summed == CHECKSUM_HW - && skb_checksum_help(skb, 0))) { + || (skb->ip_summed == CHECKSUM_PARTIAL + && skb_checksum_help(skb))) { sch->qstats.drops++; return NET_XMIT_DROP; } |