From 7e43f1128d4c4bd91786ca6abff45a91e88f9776 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Sat, 2 Aug 2008 23:27:37 -0700 Subject: pkt_sched: Make sure RTNL is held in qdisc_root_lock(). It is the only legal environment in which this can be used. Add some commentary explaining the situation. Signed-off-by: David S. Miller --- include/net/sch_generic.h | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'include') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index b5f40d7ef724..c5bb13065051 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -193,10 +193,22 @@ static inline struct Qdisc *qdisc_root(struct Qdisc *qdisc) return qdisc->dev_queue->qdisc; } +/* The qdisc root lock is a mechanism by which to top level + * of a qdisc tree can be locked from any qdisc node in the + * forest. This allows changing the configuration of some + * aspect of the qdisc tree while blocking out asynchronous + * qdisc access in the packet processing paths. + * + * It is only legal to do this when the root will not change + * on us. Otherwise we'll potentially lock the wrong qdisc + * root. This is enforced by holding the RTNL semaphore, which + * all users of this lock accessor must do. + */ static inline spinlock_t *qdisc_root_lock(struct Qdisc *qdisc) { struct Qdisc *root = qdisc_root(qdisc); + ASSERT_RTNL(); return qdisc_lock(root); } -- cgit v1.2.3 From f880374c2fe37aad3fa62253a4bc125d7a933aad Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sun, 3 Aug 2008 21:15:08 -0700 Subject: sctp: Drop ipfargok in sctp_xmit function The ipfragok flag controls whether the packet may be fragmented either on the local host on beyond. The latter is only valid on IPv4. In fact, we never want to do the latter even on IPv4 when PMTU is enabled. This is because even though we can't fragment packets within SCTP due to the prtocol's inherent faults, we can still fragment it at IP layer. By setting the DF bit we will improve the PMTU process. RFC 2960 only says that we SHOULD clear the DF bit in this case, so we're compliant even if we set the DF bit. In fact RFC 4960 no longer has this statement. Once we make this change, we only need to control the local fragmentation. There is already a bit in the skb which controls that, local_df. So this patch sets that instead of using the ipfragok argument. The only complication is that there isn't a struct sock object per transport, so for IPv4 we have to resort to changing the pmtudisc field for every packet. This should be safe though as the protocol is single-threaded. Note that after this patch we can remove ipfragok from the rest of the stack too. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- include/net/sctp/structs.h | 3 +-- net/sctp/ipv6.c | 8 +++++--- net/sctp/output.c | 6 ++---- net/sctp/protocol.c | 9 +++++++-- 4 files changed, 15 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 535a18f57a13..ab1c472ea753 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -524,8 +524,7 @@ static inline void sctp_ssn_skip(struct sctp_stream *stream, __u16 id, */ struct sctp_af { int (*sctp_xmit) (struct sk_buff *skb, - struct sctp_transport *, - int ipfragok); + struct sctp_transport *); int (*setsockopt) (struct sock *sk, int level, int optname, diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index a238d6834b33..483a01d0740a 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -195,8 +195,7 @@ out: } /* Based on tcp_v6_xmit() in tcp_ipv6.c. */ -static int sctp_v6_xmit(struct sk_buff *skb, struct sctp_transport *transport, - int ipfragok) +static int sctp_v6_xmit(struct sk_buff *skb, struct sctp_transport *transport) { struct sock *sk = skb->sk; struct ipv6_pinfo *np = inet6_sk(sk); @@ -231,7 +230,10 @@ static int sctp_v6_xmit(struct sk_buff *skb, struct sctp_transport *transport, SCTP_INC_STATS(SCTP_MIB_OUTSCTPPACKS); - return ip6_xmit(sk, skb, &fl, np->opt, ipfragok); + if (!(transport->param_flags & SPP_PMTUD_ENABLE)) + skb->local_df = 1; + + return ip6_xmit(sk, skb, &fl, np->opt, 0); } /* Returns the dst cache entry for the given source and destination ip diff --git a/net/sctp/output.c b/net/sctp/output.c index 45684646b1db..0dc4a7dfb234 100644 --- a/net/sctp/output.c +++ b/net/sctp/output.c @@ -586,10 +586,8 @@ int sctp_packet_transmit(struct sctp_packet *packet) SCTP_DEBUG_PRINTK("***sctp_transmit_packet*** skb len %d\n", nskb->len); - if (tp->param_flags & SPP_PMTUD_ENABLE) - (*tp->af_specific->sctp_xmit)(nskb, tp, packet->ipfragok); - else - (*tp->af_specific->sctp_xmit)(nskb, tp, 1); + nskb->local_df = packet->ipfragok; + (*tp->af_specific->sctp_xmit)(nskb, tp); out: packet->size = packet->overhead; diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index a6e0818bcff5..0b65354aaf64 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -862,16 +862,21 @@ static int sctp_inet_supported_addrs(const struct sctp_sock *opt, /* Wrapper routine that calls the ip transmit routine. */ static inline int sctp_v4_xmit(struct sk_buff *skb, - struct sctp_transport *transport, int ipfragok) + struct sctp_transport *transport) { + struct inet_sock *inet = inet_sk(skb->sk); + SCTP_DEBUG_PRINTK("%s: skb:%p, len:%d, " "src:%u.%u.%u.%u, dst:%u.%u.%u.%u\n", __func__, skb, skb->len, NIPQUAD(skb->rtable->rt_src), NIPQUAD(skb->rtable->rt_dst)); + inet->pmtudisc = transport->param_flags & SPP_PMTUD_ENABLE ? + IP_PMTUDISC_DO : IP_PMTUDISC_DONT; + SCTP_INC_STATS(SCTP_MIB_OUTSCTPPACKS); - return ip_queue_xmit(skb, ipfragok); + return ip_queue_xmit(skb, 0); } static struct sctp_af sctp_af_inet; -- cgit v1.2.3 From 98f7dfd86cbbd377e2cbc293529681b914296f68 Mon Sep 17 00:00:00 2001 From: Emmanuel Grumbach Date: Fri, 18 Jul 2008 13:52:59 +0800 Subject: mac80211: pass dtim_period to low level driver This patch adds the dtim_period in ieee80211_bss_conf, this allows the low level driver to know the dtim_period, and to plan power save accordingly. Signed-off-by: Emmanuel Grumbach Signed-off-by: Tomas Winkler Signed-off-by: Zhu Yi Acked-by: Johannes Berg Signed-off-by: John W. Linville --- include/linux/ieee80211.h | 13 +++++++++++++ include/net/mac80211.h | 4 +++- net/mac80211/ieee80211_i.h | 1 + net/mac80211/mlme.c | 11 +++++++++++ 4 files changed, 28 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h index a1630ba0b87c..7f4df7c7659d 100644 --- a/include/linux/ieee80211.h +++ b/include/linux/ieee80211.h @@ -506,6 +506,19 @@ struct ieee80211_channel_sw_ie { u8 count; } __attribute__ ((packed)); +/** + * struct ieee80211_tim + * + * This structure refers to "Traffic Indication Map information element" + */ +struct ieee80211_tim_ie { + u8 dtim_count; + u8 dtim_period; + u8 bitmap_ctrl; + /* variable size: 1 - 251 bytes */ + u8 virtual_map[0]; +} __attribute__ ((packed)); + struct ieee80211_mgmt { __le16 frame_control; __le16 duration; diff --git a/include/net/mac80211.h b/include/net/mac80211.h index b52721008be8..9d99f2e0a204 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -177,9 +177,10 @@ enum ieee80211_bss_change { * @aid: association ID number, valid only when @assoc is true * @use_cts_prot: use CTS protection * @use_short_preamble: use 802.11b short preamble + * @dtim_period: num of beacons before the next DTIM, for PSM * @timestamp: beacon timestamp * @beacon_int: beacon interval - * @assoc_capability: capabbilities taken from assoc resp + * @assoc_capability: capabilities taken from assoc resp * @assoc_ht: association in HT mode * @ht_conf: ht capabilities * @ht_bss_conf: ht extended capabilities @@ -191,6 +192,7 @@ struct ieee80211_bss_conf { /* erp related data */ bool use_cts_prot; bool use_short_preamble; + u8 dtim_period; u16 beacon_int; u16 assoc_capability; u64 timestamp; diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index a2e200f9811e..ec59345af65b 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -82,6 +82,7 @@ struct ieee80211_sta_bss { u8 bssid[ETH_ALEN]; u8 ssid[IEEE80211_MAX_SSID_LEN]; + u8 dtim_period; u16 capability; /* host byte order */ enum ieee80211_band band; int freq; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index acb04133a95d..591e6331c427 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -551,6 +551,7 @@ static void ieee80211_set_associated(struct net_device *dev, /* set timing information */ sdata->bss_conf.beacon_int = bss->beacon_int; sdata->bss_conf.timestamp = bss->timestamp; + sdata->bss_conf.dtim_period = bss->dtim_period; changed |= ieee80211_handle_bss_capability(sdata, bss); @@ -2688,6 +2689,16 @@ static void ieee80211_rx_bss_info(struct net_device *dev, bss->beacon_int = le16_to_cpu(mgmt->u.beacon.beacon_int); bss->capability = le16_to_cpu(mgmt->u.beacon.capab_info); + if (elems->tim) { + struct ieee80211_tim_ie *tim_ie = + (struct ieee80211_tim_ie *)elems->tim; + bss->dtim_period = tim_ie->dtim_period; + } + + /* set default value for buggy APs */ + if (!elems->tim || bss->dtim_period == 0) + bss->dtim_period = 1; + bss->supp_rates_len = 0; if (elems->supp_rates) { clen = IEEE80211_MAX_SUPP_RATES - bss->supp_rates_len; -- cgit v1.2.3 From ea95bba41e69c616bb1512cf59d22f33266b8568 Mon Sep 17 00:00:00 2001 From: Tomas Winkler Date: Fri, 18 Jul 2008 13:53:00 +0800 Subject: mac80211: make listen_interval be limited by low level driver This patch makes possible for a driver to specify maximal listen interval The possibility for user to configure listen interval is not implemented yet, currently the maximum provided by the driver or 1 is used. Mac80211 uses config handler to set listen interval for to the driver. Signed-off-by: Tomas Winkler Signed-off-by: Emmanuel Grumbach Signed-off-by: Zhu Yi Signed-off-by: John W. Linville --- include/net/mac80211.h | 9 ++++++++- net/mac80211/main.c | 5 +++++ net/mac80211/mlme.c | 6 ++++-- 3 files changed, 17 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 9d99f2e0a204..b397e4d984c7 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -432,6 +432,7 @@ enum ieee80211_conf_flags { * @radio_enabled: when zero, driver is required to switch off the radio. * TODO make a flag * @beacon_int: beacon interval (TODO make interface config) + * @listen_interval: listen interval in units of beacon interval * @flags: configuration flags defined above * @power_level: requested transmit power (in dBm) * @max_antenna_gain: maximum antenna gain (in dBi) @@ -446,6 +447,7 @@ struct ieee80211_conf { int radio_enabled; int beacon_int; + u16 listen_interval; u32 flags; int power_level; int max_antenna_gain; @@ -787,6 +789,9 @@ enum ieee80211_hw_flags { * @max_signal: Maximum value for signal (rssi) in RX information, used * only when @IEEE80211_HW_SIGNAL_UNSPEC or @IEEE80211_HW_SIGNAL_DB * + * @max_listen_interval: max listen interval in units of beacon interval + * that HW supports + * * @queues: number of available hardware transmit queues for * data packets. WMM/QoS requires at least four, these * queues need to have configurable access parameters. @@ -814,7 +819,9 @@ struct ieee80211_hw { unsigned int extra_tx_headroom; int channel_change_time; int vif_data_size; - u16 queues, ampdu_queues; + u16 queues; + u16 ampdu_queues; + u16 max_listen_interval; s8 max_signal; }; diff --git a/net/mac80211/main.c b/net/mac80211/main.c index a4c5b90de769..0c02c471bca2 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -1689,6 +1689,11 @@ int ieee80211_register_hw(struct ieee80211_hw *hw) if (local->hw.conf.beacon_int < 10) local->hw.conf.beacon_int = 100; + if (local->hw.max_listen_interval == 0) + local->hw.max_listen_interval = 1; + + local->hw.conf.listen_interval = local->hw.max_listen_interval; + local->wstats_flags |= local->hw.flags & (IEEE80211_HW_SIGNAL_UNSPEC | IEEE80211_HW_SIGNAL_DB | IEEE80211_HW_SIGNAL_DBM) ? diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 591e6331c427..779affd8b8fe 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -774,7 +774,8 @@ static void ieee80211_send_assoc(struct net_device *dev, mgmt->frame_control = IEEE80211_FC(IEEE80211_FTYPE_MGMT, IEEE80211_STYPE_REASSOC_REQ); mgmt->u.reassoc_req.capab_info = cpu_to_le16(capab); - mgmt->u.reassoc_req.listen_interval = cpu_to_le16(1); + mgmt->u.reassoc_req.listen_interval = + cpu_to_le16(local->hw.conf.listen_interval); memcpy(mgmt->u.reassoc_req.current_ap, ifsta->prev_bssid, ETH_ALEN); } else { @@ -782,7 +783,8 @@ static void ieee80211_send_assoc(struct net_device *dev, mgmt->frame_control = IEEE80211_FC(IEEE80211_FTYPE_MGMT, IEEE80211_STYPE_ASSOC_REQ); mgmt->u.assoc_req.capab_info = cpu_to_le16(capab); - mgmt->u.assoc_req.listen_interval = cpu_to_le16(1); + mgmt->u.reassoc_req.listen_interval = + cpu_to_le16(local->hw.conf.listen_interval); } /* SSID */ -- cgit v1.2.3 From 378a2f090f7a478704a372a4869b8a9ac206234e Mon Sep 17 00:00:00 2001 From: Jarek Poplawski Date: Mon, 4 Aug 2008 22:31:03 -0700 Subject: net_sched: Add qdisc __NET_XMIT_STOLEN flag Patrick McHardy noticed: "The other problem that affects all qdiscs supporting actions is TC_ACT_QUEUED/TC_ACT_STOLEN getting mapped to NET_XMIT_SUCCESS even though the packet is not queued, corrupting upper qdiscs' qlen counters." and later explained: "The reason why it translates it at all seems to be to not increase the drops counter. Within a single qdisc this could be avoided by other means easily, upper qdiscs would still increase the counter when we return anything besides NET_XMIT_SUCCESS though. This means we need a new NET_XMIT return value to indicate this to the upper qdiscs. So I'd suggest to introduce NET_XMIT_STOLEN, return that to upper qdiscs and translate it to NET_XMIT_SUCCESS in dev_queue_xmit, similar to NET_XMIT_BYPASS." David Miller noticed: "Maybe these NET_XMIT_* values being passed around should be a set of bits. They could be composed of base meanings, combined with specific attributes. So you could say "NET_XMIT_DROP | __NET_XMIT_NO_DROP_COUNT" The attributes get masked out by the top-level ->enqueue() caller, such that the base meanings are the only thing that make their way up into the stack. If it's only about communication within the qdisc tree, let's simply code it that way." This patch is trying to realize these ideas. Signed-off-by: Jarek Poplawski Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 + include/net/sch_generic.h | 14 +++++++++++++- net/sched/sch_atm.c | 12 +++++++----- net/sched/sch_cbq.c | 23 +++++++++++++++-------- net/sched/sch_dsmark.c | 8 +++++--- net/sched/sch_hfsc.c | 8 +++++--- net/sched/sch_htb.c | 18 +++++++++++------- net/sched/sch_netem.c | 3 ++- net/sched/sch_prio.c | 8 +++++--- net/sched/sch_red.c | 2 +- net/sched/sch_sfq.c | 2 +- net/sched/sch_tbf.c | 3 ++- 12 files changed, 68 insertions(+), 34 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index ee583f642a9f..abbf5d52ec86 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -64,6 +64,7 @@ struct wireless_dev; #define NET_XMIT_BYPASS 4 /* packet does not leave via dequeue; (TC use only - dev_queue_xmit returns this as NET_XMIT_SUCCESS) */ +#define NET_XMIT_MASK 0xFFFF /* qdisc flags in net/sch_generic.h */ /* Backlog congestion levels */ #define NET_RX_SUCCESS 0 /* keep 'em coming, baby */ diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index c5bb13065051..f15b045a85e9 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -343,6 +343,18 @@ static inline unsigned int qdisc_pkt_len(struct sk_buff *skb) return qdisc_skb_cb(skb)->pkt_len; } +#ifdef CONFIG_NET_CLS_ACT +/* additional qdisc xmit flags */ +enum net_xmit_qdisc_t { + __NET_XMIT_STOLEN = 0x00010000, +}; + +#define net_xmit_drop_count(e) ((e) & __NET_XMIT_STOLEN ? 0 : 1) + +#else +#define net_xmit_drop_count(e) (1) +#endif + static inline int qdisc_enqueue(struct sk_buff *skb, struct Qdisc *sch) { #ifdef CONFIG_NET_SCHED @@ -355,7 +367,7 @@ static inline int qdisc_enqueue(struct sk_buff *skb, struct Qdisc *sch) static inline int qdisc_enqueue_root(struct sk_buff *skb, struct Qdisc *sch) { qdisc_skb_cb(skb)->pkt_len = skb->len; - return qdisc_enqueue(skb, sch); + return qdisc_enqueue(skb, sch) & NET_XMIT_MASK; } static inline int __qdisc_enqueue_tail(struct sk_buff *skb, struct Qdisc *sch, diff --git a/net/sched/sch_atm.c b/net/sched/sch_atm.c index 6b517b9dac5b..27dd773481bc 100644 --- a/net/sched/sch_atm.c +++ b/net/sched/sch_atm.c @@ -415,7 +415,7 @@ static int atm_tc_enqueue(struct sk_buff *skb, struct Qdisc *sch) case TC_ACT_QUEUED: case TC_ACT_STOLEN: kfree_skb(skb); - return NET_XMIT_SUCCESS; + return NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: kfree_skb(skb); goto drop; @@ -432,9 +432,11 @@ static int atm_tc_enqueue(struct sk_buff *skb, struct Qdisc *sch) ret = qdisc_enqueue(skb, flow->q); if (ret != 0) { drop: __maybe_unused - sch->qstats.drops++; - if (flow) - flow->qstats.drops++; + if (net_xmit_drop_count(ret)) { + sch->qstats.drops++; + if (flow) + flow->qstats.drops++; + } return ret; } sch->bstats.bytes += qdisc_pkt_len(skb); @@ -530,7 +532,7 @@ static int atm_tc_requeue(struct sk_buff *skb, struct Qdisc *sch) if (!ret) { sch->q.qlen++; sch->qstats.requeues++; - } else { + } else if (net_xmit_drop_count(ret)) { sch->qstats.drops++; p->link.qstats.drops++; } diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c index 14954bf4a683..765ae5659000 100644 --- a/net/sched/sch_cbq.c +++ b/net/sched/sch_cbq.c @@ -256,7 +256,7 @@ cbq_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) switch (result) { case TC_ACT_QUEUED: case TC_ACT_STOLEN: - *qerr = NET_XMIT_SUCCESS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: return NULL; case TC_ACT_RECLASSIFY: @@ -397,9 +397,11 @@ cbq_enqueue(struct sk_buff *skb, struct Qdisc *sch) return ret; } - sch->qstats.drops++; - cbq_mark_toplevel(q, cl); - cl->qstats.drops++; + if (net_xmit_drop_count(ret)) { + sch->qstats.drops++; + cbq_mark_toplevel(q, cl); + cl->qstats.drops++; + } return ret; } @@ -430,8 +432,10 @@ cbq_requeue(struct sk_buff *skb, struct Qdisc *sch) cbq_activate_class(cl); return 0; } - sch->qstats.drops++; - cl->qstats.drops++; + if (net_xmit_drop_count(ret)) { + sch->qstats.drops++; + cl->qstats.drops++; + } return ret; } @@ -664,13 +668,15 @@ static int cbq_reshape_fail(struct sk_buff *skb, struct Qdisc *child) q->rx_class = NULL; if (cl && (cl = cbq_reclassify(skb, cl)) != NULL) { + int ret; cbq_mark_toplevel(q, cl); q->rx_class = cl; cl->q->__parent = sch; - if (qdisc_enqueue(skb, cl->q) == 0) { + ret = qdisc_enqueue(skb, cl->q); + if (ret == NET_XMIT_SUCCESS) { sch->q.qlen++; sch->bstats.packets++; sch->bstats.bytes += qdisc_pkt_len(skb); @@ -678,7 +684,8 @@ static int cbq_reshape_fail(struct sk_buff *skb, struct Qdisc *child) cbq_activate_class(cl); return 0; } - sch->qstats.drops++; + if (net_xmit_drop_count(ret)) + sch->qstats.drops++; return 0; } diff --git a/net/sched/sch_dsmark.c b/net/sched/sch_dsmark.c index a935676987e2..7170275d9f99 100644 --- a/net/sched/sch_dsmark.c +++ b/net/sched/sch_dsmark.c @@ -236,7 +236,7 @@ static int dsmark_enqueue(struct sk_buff *skb, struct Qdisc *sch) case TC_ACT_QUEUED: case TC_ACT_STOLEN: kfree_skb(skb); - return NET_XMIT_SUCCESS; + return NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: goto drop; @@ -254,7 +254,8 @@ static int dsmark_enqueue(struct sk_buff *skb, struct Qdisc *sch) err = qdisc_enqueue(skb, p->q); if (err != NET_XMIT_SUCCESS) { - sch->qstats.drops++; + if (net_xmit_drop_count(err)) + sch->qstats.drops++; return err; } @@ -321,7 +322,8 @@ static int dsmark_requeue(struct sk_buff *skb, struct Qdisc *sch) err = p->q->ops->requeue(skb, p->q); if (err != NET_XMIT_SUCCESS) { - sch->qstats.drops++; + if (net_xmit_drop_count(err)) + sch->qstats.drops++; return err; } diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index 0ae7d19dcba8..5cf9ae716118 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -1166,7 +1166,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) switch (result) { case TC_ACT_QUEUED: case TC_ACT_STOLEN: - *qerr = NET_XMIT_SUCCESS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: return NULL; } @@ -1586,8 +1586,10 @@ hfsc_enqueue(struct sk_buff *skb, struct Qdisc *sch) err = qdisc_enqueue(skb, cl->qdisc); if (unlikely(err != NET_XMIT_SUCCESS)) { - cl->qstats.drops++; - sch->qstats.drops++; + if (net_xmit_drop_count(err)) { + cl->qstats.drops++; + sch->qstats.drops++; + } return err; } diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 75a40951c4f2..538d79b489ae 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -221,7 +221,7 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, switch (result) { case TC_ACT_QUEUED: case TC_ACT_STOLEN: - *qerr = NET_XMIT_SUCCESS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: return NULL; } @@ -572,9 +572,11 @@ static int htb_enqueue(struct sk_buff *skb, struct Qdisc *sch) kfree_skb(skb); return ret; #endif - } else if (qdisc_enqueue(skb, cl->un.leaf.q) != NET_XMIT_SUCCESS) { - sch->qstats.drops++; - cl->qstats.drops++; + } else if ((ret = qdisc_enqueue(skb, cl->un.leaf.q)) != NET_XMIT_SUCCESS) { + if (net_xmit_drop_count(ret)) { + sch->qstats.drops++; + cl->qstats.drops++; + } return NET_XMIT_DROP; } else { cl->bstats.packets += @@ -615,10 +617,12 @@ static int htb_requeue(struct sk_buff *skb, struct Qdisc *sch) kfree_skb(skb); return ret; #endif - } else if (cl->un.leaf.q->ops->requeue(skb, cl->un.leaf.q) != + } else if ((ret = cl->un.leaf.q->ops->requeue(skb, cl->un.leaf.q)) != NET_XMIT_SUCCESS) { - sch->qstats.drops++; - cl->qstats.drops++; + if (net_xmit_drop_count(ret)) { + sch->qstats.drops++; + cl->qstats.drops++; + } return NET_XMIT_DROP; } else htb_activate(q, cl); diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index a59085700678..6cd6f2bc749e 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -240,8 +240,9 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch) sch->q.qlen++; sch->bstats.bytes += qdisc_pkt_len(skb); sch->bstats.packets++; - } else + } else if (net_xmit_drop_count(ret)) { sch->qstats.drops++; + } pr_debug("netem: enqueue ret %d\n", ret); return ret; diff --git a/net/sched/sch_prio.c b/net/sched/sch_prio.c index f849243eb095..adb1a52b77d3 100644 --- a/net/sched/sch_prio.c +++ b/net/sched/sch_prio.c @@ -45,7 +45,7 @@ prio_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) switch (err) { case TC_ACT_STOLEN: case TC_ACT_QUEUED: - *qerr = NET_XMIT_SUCCESS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: return NULL; } @@ -88,7 +88,8 @@ prio_enqueue(struct sk_buff *skb, struct Qdisc *sch) sch->q.qlen++; return NET_XMIT_SUCCESS; } - sch->qstats.drops++; + if (net_xmit_drop_count(ret)) + sch->qstats.drops++; return ret; } @@ -114,7 +115,8 @@ prio_requeue(struct sk_buff *skb, struct Qdisc* sch) sch->qstats.requeues++; return 0; } - sch->qstats.drops++; + if (net_xmit_drop_count(ret)) + sch->qstats.drops++; return NET_XMIT_DROP; } diff --git a/net/sched/sch_red.c b/net/sched/sch_red.c index 3f2d1d7f3bbd..5da05839e225 100644 --- a/net/sched/sch_red.c +++ b/net/sched/sch_red.c @@ -97,7 +97,7 @@ static int red_enqueue(struct sk_buff *skb, struct Qdisc* sch) sch->bstats.bytes += qdisc_pkt_len(skb); sch->bstats.packets++; sch->q.qlen++; - } else { + } else if (net_xmit_drop_count(ret)) { q->stats.pdrop++; sch->qstats.drops++; } diff --git a/net/sched/sch_sfq.c b/net/sched/sch_sfq.c index 8589da666568..3a456e1b829a 100644 --- a/net/sched/sch_sfq.c +++ b/net/sched/sch_sfq.c @@ -178,7 +178,7 @@ static unsigned int sfq_classify(struct sk_buff *skb, struct Qdisc *sch, switch (result) { case TC_ACT_STOLEN: case TC_ACT_QUEUED: - *qerr = NET_XMIT_SUCCESS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; case TC_ACT_SHOT: return 0; } diff --git a/net/sched/sch_tbf.c b/net/sched/sch_tbf.c index b296672f7632..7d3b7ff3bf07 100644 --- a/net/sched/sch_tbf.c +++ b/net/sched/sch_tbf.c @@ -135,7 +135,8 @@ static int tbf_enqueue(struct sk_buff *skb, struct Qdisc* sch) ret = qdisc_enqueue(skb, q->qdisc); if (ret != 0) { - sch->qstats.drops++; + if (net_xmit_drop_count(ret)) + sch->qstats.drops++; return ret; } -- cgit v1.2.3 From c27f339af90bb874a7a9c680b17abfd32d4a727b Mon Sep 17 00:00:00 2001 From: Jarek Poplawski Date: Mon, 4 Aug 2008 22:39:11 -0700 Subject: net_sched: Add qdisc __NET_XMIT_BYPASS flag Patrick McHardy noticed that it would be nice to handle NET_XMIT_BYPASS by NET_XMIT_SUCCESS with an internal qdisc flag __NET_XMIT_BYPASS and to remove the mapping from dev_queue_xmit(). David Miller spotted a serious bug in the first version of this patch. Signed-off-by: Jarek Poplawski Signed-off-by: David S. Miller --- include/net/sch_generic.h | 6 +++--- net/core/dev.c | 1 - net/sched/sch_atm.c | 2 +- net/sched/sch_cbq.c | 4 ++-- net/sched/sch_dsmark.c | 2 +- net/sched/sch_hfsc.c | 4 ++-- net/sched/sch_htb.c | 6 +++--- net/sched/sch_netem.c | 2 +- net/sched/sch_prio.c | 6 +++--- net/sched/sch_sfq.c | 6 +++--- 10 files changed, 19 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index f15b045a85e9..a7abfda3e447 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -343,14 +343,14 @@ static inline unsigned int qdisc_pkt_len(struct sk_buff *skb) return qdisc_skb_cb(skb)->pkt_len; } -#ifdef CONFIG_NET_CLS_ACT -/* additional qdisc xmit flags */ +/* additional qdisc xmit flags (NET_XMIT_MASK in linux/netdevice.h) */ enum net_xmit_qdisc_t { __NET_XMIT_STOLEN = 0x00010000, + __NET_XMIT_BYPASS = 0x00020000, }; +#ifdef CONFIG_NET_CLS_ACT #define net_xmit_drop_count(e) ((e) & __NET_XMIT_STOLEN ? 0 : 1) - #else #define net_xmit_drop_count(e) (1) #endif diff --git a/net/core/dev.c b/net/core/dev.c index fc6c9881eca8..01993ad74e76 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1805,7 +1805,6 @@ gso: spin_unlock(root_lock); - rc = rc == NET_XMIT_BYPASS ? NET_XMIT_SUCCESS : rc; goto out; } diff --git a/net/sched/sch_atm.c b/net/sched/sch_atm.c index 27dd773481bc..43d37256c15e 100644 --- a/net/sched/sch_atm.c +++ b/net/sched/sch_atm.c @@ -457,7 +457,7 @@ drop: __maybe_unused return 0; } tasklet_schedule(&p->task); - return NET_XMIT_BYPASS; + return NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; } /* diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c index 765ae5659000..4e261ce62f48 100644 --- a/net/sched/sch_cbq.c +++ b/net/sched/sch_cbq.c @@ -230,7 +230,7 @@ cbq_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) (cl = cbq_class_lookup(q, prio)) != NULL) return cl; - *qerr = NET_XMIT_BYPASS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; for (;;) { int result = 0; defmap = head->defaults; @@ -377,7 +377,7 @@ cbq_enqueue(struct sk_buff *skb, struct Qdisc *sch) q->rx_class = cl; #endif if (cl == NULL) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; diff --git a/net/sched/sch_dsmark.c b/net/sched/sch_dsmark.c index 7170275d9f99..edd1298f85f6 100644 --- a/net/sched/sch_dsmark.c +++ b/net/sched/sch_dsmark.c @@ -268,7 +268,7 @@ static int dsmark_enqueue(struct sk_buff *skb, struct Qdisc *sch) drop: kfree_skb(skb); sch->qstats.drops++; - return NET_XMIT_BYPASS; + return NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; } static struct sk_buff *dsmark_dequeue(struct Qdisc *sch) diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index 5cf9ae716118..c2b8d9cce3d2 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -1159,7 +1159,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) if (cl->level == 0) return cl; - *qerr = NET_XMIT_BYPASS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; tcf = q->root.filter_list; while (tcf && (result = tc_classify(skb, tcf, &res)) >= 0) { #ifdef CONFIG_NET_CLS_ACT @@ -1578,7 +1578,7 @@ hfsc_enqueue(struct sk_buff *skb, struct Qdisc *sch) cl = hfsc_classify(skb, sch, &err); if (cl == NULL) { - if (err == NET_XMIT_BYPASS) + if (err & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return err; diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 538d79b489ae..be35422711a3 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -214,7 +214,7 @@ static struct htb_class *htb_classify(struct sk_buff *skb, struct Qdisc *sch, if ((cl = htb_find(skb->priority, sch)) != NULL && cl->level == 0) return cl; - *qerr = NET_XMIT_BYPASS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; tcf = q->filter_list; while (tcf && (result = tc_classify(skb, tcf, &res)) >= 0) { #ifdef CONFIG_NET_CLS_ACT @@ -567,7 +567,7 @@ static int htb_enqueue(struct sk_buff *skb, struct Qdisc *sch) } #ifdef CONFIG_NET_CLS_ACT } else if (!cl) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; @@ -612,7 +612,7 @@ static int htb_requeue(struct sk_buff *skb, struct Qdisc *sch) } #ifdef CONFIG_NET_CLS_ACT } else if (!cl) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index 6cd6f2bc749e..fb0294d0b55e 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -176,7 +176,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch) if (count == 0) { sch->qstats.drops++; kfree_skb(skb); - return NET_XMIT_BYPASS; + return NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; } skb_orphan(skb); diff --git a/net/sched/sch_prio.c b/net/sched/sch_prio.c index adb1a52b77d3..eac197610edf 100644 --- a/net/sched/sch_prio.c +++ b/net/sched/sch_prio.c @@ -38,7 +38,7 @@ prio_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) struct tcf_result res; int err; - *qerr = NET_XMIT_BYPASS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; if (TC_H_MAJ(skb->priority) != sch->handle) { err = tc_classify(skb, q->filter_list, &res); #ifdef CONFIG_NET_CLS_ACT @@ -74,7 +74,7 @@ prio_enqueue(struct sk_buff *skb, struct Qdisc *sch) #ifdef CONFIG_NET_CLS_ACT if (qdisc == NULL) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; @@ -103,7 +103,7 @@ prio_requeue(struct sk_buff *skb, struct Qdisc* sch) qdisc = prio_classify(skb, sch, &ret); #ifdef CONFIG_NET_CLS_ACT if (qdisc == NULL) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; diff --git a/net/sched/sch_sfq.c b/net/sched/sch_sfq.c index 3a456e1b829a..6e041d10dbdb 100644 --- a/net/sched/sch_sfq.c +++ b/net/sched/sch_sfq.c @@ -171,7 +171,7 @@ static unsigned int sfq_classify(struct sk_buff *skb, struct Qdisc *sch, if (!q->filter_list) return sfq_hash(q, skb) + 1; - *qerr = NET_XMIT_BYPASS; + *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; result = tc_classify(skb, q->filter_list, &res); if (result >= 0) { #ifdef CONFIG_NET_CLS_ACT @@ -285,7 +285,7 @@ sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch) hash = sfq_classify(skb, sch, &ret); if (hash == 0) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; @@ -339,7 +339,7 @@ sfq_requeue(struct sk_buff *skb, struct Qdisc *sch) hash = sfq_classify(skb, sch, &ret); if (hash == 0) { - if (ret == NET_XMIT_BYPASS) + if (ret & __NET_XMIT_BYPASS) sch->qstats.drops++; kfree_skb(skb); return ret; -- cgit v1.2.3 From cc6533e98a7f3cb7fce9d740da49195c7aa523a4 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 4 Aug 2008 23:04:08 -0700 Subject: net: Kill plain NET_XMIT_BYPASS. dst_input() was doing something completely absurd, looping on skb->dst->input() if NET_XMIT_BYPASS was seen, but these functions never return such an error. And as a result plain ole' NET_XMIT_BYPASS has no more references and can be completely killed off. Signed-off-by: David S. Miller --- include/linux/netdevice.h | 3 --- include/net/dst.h | 12 +----------- 2 files changed, 1 insertion(+), 14 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index abbf5d52ec86..488c56e649b5 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -61,9 +61,6 @@ struct wireless_dev; #define NET_XMIT_DROP 1 /* skb dropped */ #define NET_XMIT_CN 2 /* congestion notification */ #define NET_XMIT_POLICED 3 /* skb is shot by police */ -#define NET_XMIT_BYPASS 4 /* packet does not leave via dequeue; - (TC use only - dev_queue_xmit - returns this as NET_XMIT_SUCCESS) */ #define NET_XMIT_MASK 0xFFFF /* qdisc flags in net/sch_generic.h */ /* Backlog congestion levels */ diff --git a/include/net/dst.h b/include/net/dst.h index c5c318a628f8..8a8b71e5f3f1 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -252,17 +252,7 @@ static inline int dst_output(struct sk_buff *skb) /* Input packet from network to transport. */ static inline int dst_input(struct sk_buff *skb) { - int err; - - for (;;) { - err = skb->dst->input(skb); - - if (likely(err == 0)) - return err; - /* Oh, Jamal... Seems, I will not forgive you this mess. :-) */ - if (unlikely(err != NET_XMIT_BYPASS)) - return err; - } + return skb->dst->input(skb); } static inline struct dst_entry *dst_check(struct dst_entry *dst, u32 cookie) -- cgit v1.2.3 From 95c3e8bfcdea8676e2d4d61910c379f4502049bf Mon Sep 17 00:00:00 2001 From: Rami Rosen Date: Tue, 5 Aug 2008 01:19:50 -0700 Subject: ipv4: remove unused field in struct flowi (include/net/flow.h). This patch removes an unused field (flags) from struct flowi; it seems that this "flags" field was used once in the past for multipath routing with FLOWI_FLAG_MULTIPATHOLDROUTE flag (which does no longer exist); however, the "flags" field of struct flowi is not used anymore. Signed-off-by: Rami Rosen Signed-off-by: David S. Miller --- include/net/flow.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/net/flow.h b/include/net/flow.h index ad16e0076c89..228b2477ceec 100644 --- a/include/net/flow.h +++ b/include/net/flow.h @@ -47,7 +47,6 @@ struct flowi { #define fl4_scope nl_u.ip4_u.scope __u8 proto; - __u8 flags; union { struct { __be16 sport; -- cgit v1.2.3