From b1e93a68ca41e7e73766f95ba32ca05cf9052e15 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 16 Mar 2009 15:06:42 +0100 Subject: netfilter: conntrack: don't deliver events for racy packets This patch skips the delivery of conntrack events if the packet was drop due to a race condition in the conntrack insertion. Signed-off-by: Pablo Neira Ayuso Signed-off-by: Patrick McHardy --- include/net/netfilter/nf_conntrack_core.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h index c25068e38516..5a449b44ba33 100644 --- a/include/net/netfilter/nf_conntrack_core.h +++ b/include/net/netfilter/nf_conntrack_core.h @@ -62,7 +62,8 @@ static inline int nf_conntrack_confirm(struct sk_buff *skb) if (ct && ct != &nf_conntrack_untracked) { if (!nf_ct_is_confirmed(ct) && !nf_ct_is_dying(ct)) ret = __nf_conntrack_confirm(skb); - nf_ct_deliver_cached_events(ct); + if (likely(ret == NF_ACCEPT)) + nf_ct_deliver_cached_events(ct); } return ret; } -- cgit v1.2.3 From 626ba8fbac9156a94a80be46ffd2f2ce9e4e89a0 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 16 Mar 2009 15:50:51 +0100 Subject: netfilter: ctnetlink: fix crash during expectation creation This patch fixes a possible crash due to the missing initialization of the expectation class when nf_ct_expect_related() is called. Reported-by: BORBELY Zoltan Signed-off-by: Pablo Neira Ayuso Signed-off-by: Patrick McHardy --- net/netfilter/nf_conntrack_netlink.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index cb78aa00399e..ed6d873ad384 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -1780,6 +1780,7 @@ ctnetlink_create_expect(struct nlattr *cda[], u_int8_t u3, u32 pid, int report) goto out; } + exp->class = 0; exp->expectfn = NULL; exp->flags = 0; exp->master = ct; -- cgit v1.2.3 From ec8d540969da9a70790e9028d57b5b577dd7aa77 Mon Sep 17 00:00:00 2001 From: Christoph Paasch Date: Mon, 16 Mar 2009 15:51:29 +0100 Subject: netfilter: conntrack: fix dropping packet after l4proto->packet() We currently use the negative value in the conntrack code to encode the packet verdict in the error. As NF_DROP is equal to 0, inverting NF_DROP makes no sense and, as a result, no packets are ever dropped. Signed-off-by: Christoph Paasch Signed-off-by: Pablo Neira Ayuso Signed-off-by: Patrick McHardy --- net/netfilter/nf_conntrack_core.c | 2 +- net/netfilter/nf_conntrack_proto_tcp.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index 90ce9ddb9451..f4935e344b61 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -726,7 +726,7 @@ nf_conntrack_in(struct net *net, u_int8_t pf, unsigned int hooknum, NF_CT_ASSERT(skb->nfct); ret = l4proto->packet(ct, skb, dataoff, ctinfo, pf, hooknum); - if (ret < 0) { + if (ret <= 0) { /* Invalid: inverse of the return code tells * the netfilter core what to do */ pr_debug("nf_conntrack_in: Can't track with proto module\n"); diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index a1edb9c1adee..f3fd154d1ddd 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -859,7 +859,7 @@ static int tcp_packet(struct nf_conn *ct, */ if (nf_ct_kill(ct)) return -NF_REPEAT; - return -NF_DROP; + return NF_DROP; } /* Fall through */ case TCP_CONNTRACK_IGNORE: @@ -892,7 +892,7 @@ static int tcp_packet(struct nf_conn *ct, nf_log_packet(pf, 0, skb, NULL, NULL, NULL, "nf_ct_tcp: killing out of sync session "); nf_ct_kill(ct); - return -NF_DROP; + return NF_DROP; } ct->proto.tcp.last_index = index; ct->proto.tcp.last_dir = dir; -- cgit v1.2.3 From d1238d5337e8e53cddea77c2a26d26b6eb5a982f Mon Sep 17 00:00:00 2001 From: Christoph Paasch Date: Mon, 16 Mar 2009 15:52:11 +0100 Subject: netfilter: conntrack: check for NEXTHDR_NONE before header sanity checking NEXTHDR_NONE doesn't has an IPv6 option header, so the first check for the length will always fail and results in a confusing message "too short" if debugging enabled. With this patch, we check for NEXTHDR_NONE before length sanity checkings are done. Signed-off-by: Christoph Paasch Signed-off-by: Pablo Neira Ayuso Signed-off-by: Patrick McHardy --- net/ipv6/netfilter/nf_conntrack_reasm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index ed4d79a9e4a6..058a5e4a60c3 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -528,14 +528,14 @@ find_prev_fhdr(struct sk_buff *skb, u8 *prevhdrp, int *prevhoff, int *fhoff) if (!ipv6_ext_hdr(nexthdr)) { return -1; } - if (len < (int)sizeof(struct ipv6_opt_hdr)) { - pr_debug("too short\n"); - return -1; - } if (nexthdr == NEXTHDR_NONE) { pr_debug("next header is none\n"); return -1; } + if (len < (int)sizeof(struct ipv6_opt_hdr)) { + pr_debug("too short\n"); + return -1; + } if (skb_copy_bits(skb, start, &hdr, sizeof(hdr))) BUG(); if (nexthdr == NEXTHDR_AUTH) -- cgit v1.2.3 From 6f16bf3bdb94b567e2b6663378efb2dbf40db133 Mon Sep 17 00:00:00 2001 From: John W. Linville Date: Wed, 11 Mar 2009 11:05:25 -0400 Subject: lib80211: silence excessive crypto debugging messages When they were part of the now defunct ieee80211 component, these messages were only visible when special debugging settings were enabled. Let's mirror that with a new lib80211 debugging Kconfig option. Signed-off-by: John W. Linville --- net/wireless/Kconfig | 10 ++++++++++ net/wireless/lib80211_crypt_ccmp.c | 2 ++ net/wireless/lib80211_crypt_tkip.c | 4 ++++ 3 files changed, 16 insertions(+) diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig index e28e2b8fa436..092ae6faccca 100644 --- a/net/wireless/Kconfig +++ b/net/wireless/Kconfig @@ -102,3 +102,13 @@ config LIB80211_CRYPT_CCMP config LIB80211_CRYPT_TKIP tristate + +config LIB80211_DEBUG + bool "lib80211 debugging messages" + depends on LIB80211 + default n + ---help--- + You can enable this if you want verbose debugging messages + from lib80211. + + If unsure, say N. diff --git a/net/wireless/lib80211_crypt_ccmp.c b/net/wireless/lib80211_crypt_ccmp.c index db428194c16a..2301dc1edc4c 100644 --- a/net/wireless/lib80211_crypt_ccmp.c +++ b/net/wireless/lib80211_crypt_ccmp.c @@ -337,6 +337,7 @@ static int lib80211_ccmp_decrypt(struct sk_buff *skb, int hdr_len, void *priv) pos += 8; if (ccmp_replay_check(pn, key->rx_pn)) { +#ifdef CONFIG_LIB80211_DEBUG if (net_ratelimit()) { printk(KERN_DEBUG "CCMP: replay detected: STA=%pM " "previous PN %02x%02x%02x%02x%02x%02x " @@ -346,6 +347,7 @@ static int lib80211_ccmp_decrypt(struct sk_buff *skb, int hdr_len, void *priv) key->rx_pn[3], key->rx_pn[4], key->rx_pn[5], pn[0], pn[1], pn[2], pn[3], pn[4], pn[5]); } +#endif key->dot11RSNAStatsCCMPReplays++; return -4; } diff --git a/net/wireless/lib80211_crypt_tkip.c b/net/wireless/lib80211_crypt_tkip.c index 7e8e22bfed90..c36287399d7e 100644 --- a/net/wireless/lib80211_crypt_tkip.c +++ b/net/wireless/lib80211_crypt_tkip.c @@ -465,12 +465,14 @@ static int lib80211_tkip_decrypt(struct sk_buff *skb, int hdr_len, void *priv) pos += 8; if (tkip_replay_check(iv32, iv16, tkey->rx_iv32, tkey->rx_iv16)) { +#ifdef CONFIG_LIB80211_DEBUG if (net_ratelimit()) { printk(KERN_DEBUG "TKIP: replay detected: STA=%pM" " previous TSC %08x%04x received TSC " "%08x%04x\n", hdr->addr2, tkey->rx_iv32, tkey->rx_iv16, iv32, iv16); } +#endif tkey->dot11RSNAStatsTKIPReplays++; return -4; } @@ -505,10 +507,12 @@ static int lib80211_tkip_decrypt(struct sk_buff *skb, int hdr_len, void *priv) * it needs to be recalculated for the next packet. */ tkey->rx_phase1_done = 0; } +#ifdef CONFIG_LIB80211_DEBUG if (net_ratelimit()) { printk(KERN_DEBUG "TKIP: ICV error detected: STA=" "%pM\n", hdr->addr2); } +#endif tkey->dot11RSNAStatsTKIPICVErrors++; return -5; } -- cgit v1.2.3 From 6158425be398936af1fd04451f78ffad01529cb0 Mon Sep 17 00:00:00 2001 From: Luis R. Rodriguez Date: Thu, 12 Mar 2009 18:18:49 -0400 Subject: ath9k: implement IO serialization All 802.11n PCI devices (Cardbus, PCI, mini-PCI) require serialization of IO when on non-uniprocessor systems. PCI express devices not not require this. This should fix our only last standing open ath9k kernel.org bugzilla bug report: http://bugzilla.kernel.org/show_bug.cgi?id=12110 A port is probably required to older kernels and I can work on that. Cc: stable@kernel.org Signed-off-by: Luis R. Rodriguez Signed-off-by: John W. Linville --- drivers/net/wireless/ath9k/ath9k.h | 4 ++-- drivers/net/wireless/ath9k/core.h | 33 +++++++++++++++++++++++++++++++++ drivers/net/wireless/ath9k/hw.c | 19 +++++++++++++++++++ drivers/net/wireless/ath9k/main.c | 1 + 4 files changed, 55 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath9k/ath9k.h b/drivers/net/wireless/ath9k/ath9k.h index d27813502953..6650f609ece4 100644 --- a/drivers/net/wireless/ath9k/ath9k.h +++ b/drivers/net/wireless/ath9k/ath9k.h @@ -587,8 +587,8 @@ struct ath9k_country_entry { u8 iso[3]; }; -#define REG_WRITE(_ah, _reg, _val) iowrite32(_val, _ah->ah_sh + _reg) -#define REG_READ(_ah, _reg) ioread32(_ah->ah_sh + _reg) +#define REG_WRITE(_ah, _reg, _val) ath9k_iowrite32((_ah), (_reg), (_val)) +#define REG_READ(_ah, _reg) ath9k_ioread32((_ah), (_reg)) #define SM(_v, _f) (((_v) << _f##_S) & _f) #define MS(_v, _f) (((_v) & _f) >> _f##_S) diff --git a/drivers/net/wireless/ath9k/core.h b/drivers/net/wireless/ath9k/core.h index 4ca2aed236e0..139566cbbf65 100644 --- a/drivers/net/wireless/ath9k/core.h +++ b/drivers/net/wireless/ath9k/core.h @@ -701,6 +701,7 @@ struct ath_softc { struct ath_hal *sc_ah; void __iomem *mem; spinlock_t sc_resetlock; + spinlock_t sc_serial_rw; struct mutex mutex; u8 sc_curbssid[ETH_ALEN]; @@ -751,4 +752,36 @@ int ath_get_hal_qnum(u16 queue, struct ath_softc *sc); int ath_get_mac80211_qnum(u32 queue, struct ath_softc *sc); int ath_cabq_update(struct ath_softc *); +/* + * Read and write, they both share the same lock. We do this to serialize + * reads and writes on Atheros 802.11n PCI devices only. This is required + * as the FIFO on these devices can only accept sanely 2 requests. After + * that the device goes bananas. Serializing the reads/writes prevents this + * from happening. + */ + +static inline void ath9k_iowrite32(struct ath_hal *ah, u32 reg_offset, u32 val) +{ + if (ah->ah_config.serialize_regmode == SER_REG_MODE_ON) { + unsigned long flags; + spin_lock_irqsave(&ah->ah_sc->sc_serial_rw, flags); + iowrite32(val, ah->ah_sc->mem + reg_offset); + spin_unlock_irqrestore(&ah->ah_sc->sc_serial_rw, flags); + } else + iowrite32(val, ah->ah_sc->mem + reg_offset); +} + +static inline unsigned int ath9k_ioread32(struct ath_hal *ah, u32 reg_offset) +{ + u32 val; + if (ah->ah_config.serialize_regmode == SER_REG_MODE_ON) { + unsigned long flags; + spin_lock_irqsave(&ah->ah_sc->sc_serial_rw, flags); + val = ioread32(ah->ah_sc->mem + reg_offset); + spin_unlock_irqrestore(&ah->ah_sc->sc_serial_rw, flags); + } else + val = ioread32(ah->ah_sc->mem + reg_offset); + return val; +} + #endif /* CORE_H */ diff --git a/drivers/net/wireless/ath9k/hw.c b/drivers/net/wireless/ath9k/hw.c index 34474edefc97..5c870436fc5f 100644 --- a/drivers/net/wireless/ath9k/hw.c +++ b/drivers/net/wireless/ath9k/hw.c @@ -437,6 +437,25 @@ static void ath9k_hw_set_defaults(struct ath_hal *ah) } ah->ah_config.intr_mitigation = 1; + + /* + * We need this for PCI devices only (Cardbus, PCI, miniPCI) + * _and_ if on non-uniprocessor systems (Multiprocessor/HT). + * This means we use it for all AR5416 devices, and the few + * minor PCI AR9280 devices out there. + * + * Serialization is required because these devices do not handle + * well the case of two concurrent reads/writes due to the latency + * involved. During one read/write another read/write can be issued + * on another CPU while the previous read/write may still be working + * on our hardware, if we hit this case the hardware poops in a loop. + * We prevent this by serializing reads and writes. + * + * This issue is not present on PCI-Express devices or pre-AR5416 + * devices (legacy, 802.11abg). + */ + if (num_possible_cpus() > 1) + ah->ah_config.serialize_regmode = SER_REG_MODE_AUTO; } static struct ath_hal_5416 *ath9k_hw_newstate(u16 devid, diff --git a/drivers/net/wireless/ath9k/main.c b/drivers/net/wireless/ath9k/main.c index 0e80990d8e84..3c04044a60bd 100644 --- a/drivers/net/wireless/ath9k/main.c +++ b/drivers/net/wireless/ath9k/main.c @@ -1336,6 +1336,7 @@ static int ath_init(u16 devid, struct ath_softc *sc) printk(KERN_ERR "Unable to create debugfs files\n"); spin_lock_init(&sc->sc_resetlock); + spin_lock_init(&sc->sc_serial_rw); mutex_init(&sc->mutex); tasklet_init(&sc->intr_tq, ath9k_tasklet, (unsigned long)sc); tasklet_init(&sc->bcon_tasklet, ath9k_beacon_tasklet, -- cgit v1.2.3 From 5ec905a8df3fa877566ba98298433fbfb3d688cc Mon Sep 17 00:00:00 2001 From: Luis R. Rodriguez Date: Thu, 12 Mar 2009 18:18:50 -0400 Subject: ath9k: AR9280 PCI devices must serialize IO as well Cc: stable@kernel.org Signed-off-by: Luis R. Rodriguez Signed-off-by: John W. Linville --- drivers/net/wireless/ath9k/hw.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath9k/hw.c b/drivers/net/wireless/ath9k/hw.c index 5c870436fc5f..c38a00bbce64 100644 --- a/drivers/net/wireless/ath9k/hw.c +++ b/drivers/net/wireless/ath9k/hw.c @@ -687,7 +687,8 @@ static struct ath_hal *ath9k_hw_do_attach(u16 devid, struct ath_softc *sc, } if (ah->ah_config.serialize_regmode == SER_REG_MODE_AUTO) { - if (ah->ah_macVersion == AR_SREV_VERSION_5416_PCI) { + if (ah->ah_macVersion == AR_SREV_VERSION_5416_PCI || + (AR_SREV_9280(ah) && !ah->ah_isPciExpress)) { ah->ah_config.serialize_regmode = SER_REG_MODE_ON; } else { -- cgit v1.2.3 From 055249d20de06c290fe7625be0a7164bef3958f5 Mon Sep 17 00:00:00 2001 From: Jouni Malinen Date: Fri, 13 Mar 2009 13:59:39 +0200 Subject: mac80211: Fix panic on fragmentation with power saving It was possible to hit a kernel panic on NULL pointer dereference in dev_queue_xmit() when sending power save buffered frames to a STA that woke up from sleep. This happened when the buffered frame was requeued for transmission in ap_sta_ps_end(). In order to avoid the panic, copy the skb->dev and skb->iif values from the first fragment to all other fragments. Signed-off-by: Jouni Malinen Signed-off-by: John W. Linville --- net/mac80211/tx.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 94de5033f0b6..37e3d5ef7e3f 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -752,6 +752,8 @@ ieee80211_tx_h_fragment(struct ieee80211_tx_data *tx) skb_copy_queue_mapping(frag, first); frag->do_not_encrypt = first->do_not_encrypt; + frag->dev = first->dev; + frag->iif = first->iif; pos += copylen; left -= copylen; -- cgit v1.2.3 From 640c65eae673d2caf6e7bf61c1eb4e9513b88fda Mon Sep 17 00:00:00 2001 From: Jouni Malinen Date: Mon, 16 Mar 2009 21:47:33 +0200 Subject: zd1211rw: Do not panic on device eject when associated zd_op_tx() must not return an arbitrary error value since that can leave mac80211 trying to retransmit the frame and with the extra data pushed into the beginning of the skb on every attempt, this will end up causing a kernel panic (skb_under_panic from skb_push call). This can happen, e.g., when ejecting the device when associated. Signed-off-by: Jouni Malinen Signed-off-by: John W. Linville --- drivers/net/wireless/zd1211rw/zd_mac.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/zd1211rw/zd_mac.c b/drivers/net/wireless/zd1211rw/zd_mac.c index a611ad857983..847057d682b1 100644 --- a/drivers/net/wireless/zd1211rw/zd_mac.c +++ b/drivers/net/wireless/zd1211rw/zd_mac.c @@ -575,13 +575,17 @@ static int zd_op_tx(struct ieee80211_hw *hw, struct sk_buff *skb) r = fill_ctrlset(mac, skb); if (r) - return r; + goto fail; info->rate_driver_data[0] = hw; r = zd_usb_tx(&mac->chip.usb, skb); if (r) - return r; + goto fail; + return 0; + +fail: + dev_kfree_skb(skb); return 0; } -- cgit v1.2.3 From 303c6a0251852ecbdc5c15e466dcaff5971f7517 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Tue, 17 Mar 2009 13:11:29 -0700 Subject: gro: Fix legacy path napi_complete crash On the legacy netif_rx path, I incorrectly tried to optimise the napi_complete call by using __napi_complete before we reenable IRQs. This simply doesn't work since we need to flush the held GRO packets first. This patch fixes it by doing the obvious thing of reenabling IRQs first and then calling napi_complete. Reported-by: Frank Blaschka Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/core/dev.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index f1129706ce7b..2565f6d1d661 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2588,9 +2588,9 @@ static int process_backlog(struct napi_struct *napi, int quota) local_irq_disable(); skb = __skb_dequeue(&queue->input_pkt_queue); if (!skb) { - __napi_complete(napi); local_irq_enable(); - break; + napi_complete(napi); + goto out; } local_irq_enable(); @@ -2599,6 +2599,7 @@ static int process_backlog(struct napi_struct *napi, int quota) napi_gro_flush(napi); +out: return work; } -- cgit v1.2.3 From 690103137267e9ed893febf7ff061af63e8235a9 Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Wed, 18 Mar 2009 18:11:51 -0700 Subject: bnx2: Fix problem of using wrong IRQ handler. The MSI-X handler was chosen before the call to pci_enable_msix(). If MSI-X was not available, the wrong MSI-X handler would be used in INTA mode. This would cause a screaming interrupt problem because INTA would not be cleared by the MSI-X handler. Fixed by assigning MSI-X handler after pci_enable_msix() returns successfully. Also update version to 1.9.3. Thomas Chenault helped us find this problem. Signed-off-by: Michael Chan Signed-off-by: David S. Miller --- drivers/net/bnx2.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c index 6500b7c4739f..6b6530ffdf19 100644 --- a/drivers/net/bnx2.c +++ b/drivers/net/bnx2.c @@ -57,8 +57,8 @@ #define DRV_MODULE_NAME "bnx2" #define PFX DRV_MODULE_NAME ": " -#define DRV_MODULE_VERSION "1.9.2" -#define DRV_MODULE_RELDATE "Feb 11, 2009" +#define DRV_MODULE_VERSION "1.9.3" +#define DRV_MODULE_RELDATE "March 17, 2009" #define RUN_AT(x) (jiffies + (x)) @@ -5843,9 +5843,6 @@ bnx2_enable_msix(struct bnx2 *bp, int msix_vecs) for (i = 0; i < BNX2_MAX_MSIX_VEC; i++) { msix_ent[i].entry = i; msix_ent[i].vector = 0; - - snprintf(bp->irq_tbl[i].name, len, "%s-%d", dev->name, i); - bp->irq_tbl[i].handler = bnx2_msi_1shot; } rc = pci_enable_msix(bp->pdev, msix_ent, BNX2_MAX_MSIX_VEC); @@ -5854,8 +5851,11 @@ bnx2_enable_msix(struct bnx2 *bp, int msix_vecs) bp->irq_nvecs = msix_vecs; bp->flags |= BNX2_FLAG_USING_MSIX | BNX2_FLAG_ONE_SHOT_MSI; - for (i = 0; i < BNX2_MAX_MSIX_VEC; i++) + for (i = 0; i < BNX2_MAX_MSIX_VEC; i++) { bp->irq_tbl[i].vector = msix_ent[i].vector; + snprintf(bp->irq_tbl[i].name, len, "%s-%d", dev->name, i); + bp->irq_tbl[i].handler = bnx2_msi_1shot; + } } static void -- cgit v1.2.3 From 17d04500e2528217de5fe967599f98ee84348a9c Mon Sep 17 00:00:00 2001 From: Jay Vosburgh Date: Wed, 18 Mar 2009 18:38:25 -0700 Subject: bonding: Fix updating of speed/duplex changes This patch corrects an omission from the following commit: commit f0c76d61779b153dbfb955db3f144c62d02173c2 Author: Jay Vosburgh Date: Wed Jul 2 18:21:58 2008 -0700 bonding: refactor mii monitor The un-refactored code checked the link speed and duplex of every slave on every pass; the refactored code did not do so. The 802.3ad and balance-alb/tlb modes utilize the speed and duplex information, and require it to be kept up to date. This patch adds a notifier check to perform the appropriate updating when the slave device speed changes. Signed-off-by: Jay Vosburgh Signed-off-by: David S. Miller --- drivers/net/bonding/bond_main.c | 25 ++++++++++++++++++++----- 1 file changed, 20 insertions(+), 5 deletions(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index e0578fe8c0db..3d76686dceca 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -3537,11 +3537,26 @@ static int bond_slave_netdev_event(unsigned long event, struct net_device *slave } break; case NETDEV_CHANGE: - /* - * TODO: is this what we get if somebody - * sets up a hierarchical bond, then rmmod's - * one of the slave bonding devices? - */ + if (bond->params.mode == BOND_MODE_8023AD || bond_is_lb(bond)) { + struct slave *slave; + + slave = bond_get_slave_by_dev(bond, slave_dev); + if (slave) { + u16 old_speed = slave->speed; + u16 old_duplex = slave->duplex; + + bond_update_speed_duplex(slave); + + if (bond_is_lb(bond)) + break; + + if (old_speed != slave->speed) + bond_3ad_adapter_speed_changed(slave); + if (old_duplex != slave->duplex) + bond_3ad_adapter_duplex_changed(slave); + } + } + break; case NETDEV_DOWN: /* -- cgit v1.2.3 From 4783256ef92f5aecd6d54693b16386f2a0021c2a Mon Sep 17 00:00:00 2001 From: Pantelis Koukousoulas Date: Wed, 18 Mar 2009 18:40:02 -0700 Subject: virtio_net: Make virtio_net support carrier detection Impact: Make NetworkManager work with virtio_net For now the semantics are simple: There is always carrier. This allows a seamless experience with e.g., qemu/kvm where NetworkManager just configures and sets up everything automagically. If/when a generally agreed-upon way to control carrier on/off in the emulator/hypervisor level emerges, it will be trivial to extend the driver to support that too, but for now even this 2-liner makes user experience that much better. Signed-off-by: Pantelis Koukousoulas Signed-off-by: Rusty Russell Signed-off-by: David S. Miller --- drivers/net/virtio_net.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index c68808336c8c..e67d16c2e5f3 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -612,6 +612,7 @@ static struct ethtool_ops virtnet_ethtool_ops = { .set_tx_csum = virtnet_set_tx_csum, .set_sg = ethtool_op_set_sg, .set_tso = ethtool_op_set_tso, + .get_link = ethtool_op_get_link, }; #define MIN_MTU 68 @@ -739,6 +740,8 @@ static int virtnet_probe(struct virtio_device *vdev) goto unregister; } + netif_carrier_on(dev); + pr_debug("virtnet: registered device %s\n", dev->name); return 0; -- cgit v1.2.3 From 69145635d4db0a0382885b14634aa5b721f3aa1a Mon Sep 17 00:00:00 2001 From: Kyle McMartin Date: Wed, 18 Mar 2009 18:49:01 -0700 Subject: tulip: fix crash on iface up with shirq debug Tulip is currently doing request_irq before it has done its initialization. This is usually not a problem because it hasn't enable interrupts yet, but with DEBUG_SHIRQ on, we call the irq handler when registering the interrupt as a sanity check. This can result in a NULL ptr dereference, so call tulip_init_ring before request_irq, and add a free_ring function to do the freeing now shared with tulip_close. Tested with a shell loop running ifup, ifdown in a loop a few hundred times with DEBUG_SHIRQ on. Signed-off-by: Kyle McMartin Signed-off-by: David S. Miller --- drivers/net/tulip/tulip_core.c | 45 +++++++++++++++++++++++++++--------------- 1 file changed, 29 insertions(+), 16 deletions(-) diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c index bee75fa87a9c..2abb5d3becc6 100644 --- a/drivers/net/tulip/tulip_core.c +++ b/drivers/net/tulip/tulip_core.c @@ -255,6 +255,7 @@ const char tulip_media_cap[32] = static void tulip_tx_timeout(struct net_device *dev); static void tulip_init_ring(struct net_device *dev); +static void tulip_free_ring(struct net_device *dev); static int tulip_start_xmit(struct sk_buff *skb, struct net_device *dev); static int tulip_open(struct net_device *dev); static int tulip_close(struct net_device *dev); @@ -502,16 +503,21 @@ tulip_open(struct net_device *dev) { int retval; - if ((retval = request_irq(dev->irq, &tulip_interrupt, IRQF_SHARED, dev->name, dev))) - return retval; - tulip_init_ring (dev); + retval = request_irq(dev->irq, &tulip_interrupt, IRQF_SHARED, dev->name, dev); + if (retval) + goto free_ring; + tulip_up (dev); netif_start_queue (dev); return 0; + +free_ring: + tulip_free_ring (dev); + return retval; } @@ -768,23 +774,11 @@ static void tulip_down (struct net_device *dev) tulip_set_power_state (tp, 0, 1); } - -static int tulip_close (struct net_device *dev) +static void tulip_free_ring (struct net_device *dev) { struct tulip_private *tp = netdev_priv(dev); - void __iomem *ioaddr = tp->base_addr; int i; - netif_stop_queue (dev); - - tulip_down (dev); - - if (tulip_debug > 1) - printk (KERN_DEBUG "%s: Shutting down ethercard, status was %2.2x.\n", - dev->name, ioread32 (ioaddr + CSR5)); - - free_irq (dev->irq, dev); - /* Free all the skbuffs in the Rx queue. */ for (i = 0; i < RX_RING_SIZE; i++) { struct sk_buff *skb = tp->rx_buffers[i].skb; @@ -803,6 +797,7 @@ static int tulip_close (struct net_device *dev) dev_kfree_skb (skb); } } + for (i = 0; i < TX_RING_SIZE; i++) { struct sk_buff *skb = tp->tx_buffers[i].skb; @@ -814,6 +809,24 @@ static int tulip_close (struct net_device *dev) tp->tx_buffers[i].skb = NULL; tp->tx_buffers[i].mapping = 0; } +} + +static int tulip_close (struct net_device *dev) +{ + struct tulip_private *tp = netdev_priv(dev); + void __iomem *ioaddr = tp->base_addr; + + netif_stop_queue (dev); + + tulip_down (dev); + + if (tulip_debug > 1) + printk (KERN_DEBUG "%s: Shutting down ethercard, status was %2.2x.\n", + dev->name, ioread32 (ioaddr + CSR5)); + + free_irq (dev->irq, dev); + + tulip_free_ring (dev); return 0; } -- cgit v1.2.3 From e9c6a586f5335938b11be0a2554af58a2860e4ef Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 18 Mar 2009 18:51:06 -0700 Subject: net: Document /proc/sys/net/core/netdev_budget The NAPI poll parameter netdev_budget is not documented in kernel-docs. Since it may have a substantial effect on at least some network loads, it should be. Signed-off-by: Stanislaw Gruszka Signed-off-by: David S. Miller --- Documentation/filesystems/proc.txt | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index a87be42f8211..830bad7cce0f 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -1478,6 +1478,13 @@ of problems on the network like duplicate address or bad checksums. Normally, this should be enabled, but if the problem persists the messages can be disabled. +netdev_budget +------------- + +Maximum number of packets taken from all interfaces in one polling cycle (NAPI +poll). In one polling cycle interfaces which are registered to polling are +probed in a round-robin manner. The limit of packets in one such probe can be +set per-device via sysfs class/net//weight . netdev_max_backlog ------------------ -- cgit v1.2.3 From 1b1d8f73a44fe1796a73a97ebcb08cad18b97f07 Mon Sep 17 00:00:00 2001 From: Bjørn Mork Date: Wed, 18 Mar 2009 18:56:54 -0700 Subject: ipv6: fix display of local and remote sit endpoints This fixes the regressions cause by commit 1326c3d5a4b792a2b15877feb7fb691f8945d203 (v2.6.28-rc6-461-g23a12b1) broke the display of local and remote addresses of an SIT tunnel in iproute2. nt->parms is used by ipip6_tunnel_init() and therefore need to be initialized first. Tracked as http://bugzilla.kernel.org/show_bug.cgi?id=12868 Reported-by: Jan Engelhardt Signed-off-by: Bjørn Mork Signed-off-by: David S. Miller --- net/ipv6/sit.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c index d3467e563f02..5cee2bcbcece 100644 --- a/net/ipv6/sit.c +++ b/net/ipv6/sit.c @@ -188,9 +188,9 @@ static struct ip_tunnel * ipip6_tunnel_locate(struct net *net, } nt = netdev_priv(dev); - ipip6_tunnel_init(dev); nt->parms = *parms; + ipip6_tunnel_init(dev); if (parms->i_flags & SIT_ISATAP) dev->priv_flags |= IFF_ISATAP; -- cgit v1.2.3 From cb0dc77de0d23615a845e45844a2e22fc224d7fe Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 18 Mar 2009 19:12:42 -0700 Subject: net: fix sctp breakage broken by commit 5e739d1752aca4e8f3e794d431503bfca3162df4; AFAICS should be -stable fodder as well... Signed-off-by: Al Viro Aced-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/endpointola.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/sctp/endpointola.c b/net/sctp/endpointola.c index 4c8d9f45ce09..905fda582b92 100644 --- a/net/sctp/endpointola.c +++ b/net/sctp/endpointola.c @@ -111,7 +111,8 @@ static struct sctp_endpoint *sctp_endpoint_init(struct sctp_endpoint *ep, if (sctp_addip_enable) { auth_chunks->chunks[0] = SCTP_CID_ASCONF; auth_chunks->chunks[1] = SCTP_CID_ASCONF_ACK; - auth_chunks->param_hdr.length += htons(2); + auth_chunks->param_hdr.length = + htons(sizeof(sctp_paramhdr_t) + 2); } } -- cgit v1.2.3 From e4a389a9b5c892446b5de2038bdc0cca8703c615 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 18 Mar 2009 23:12:13 -0700 Subject: net: kfree(napi->skb) => kfree_skb struct sk_buff pointers should be freed with kfree_skb. Signed-off-by: Roel Kluin Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index 2565f6d1d661..e3fe5c705606 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2672,7 +2672,7 @@ void netif_napi_del(struct napi_struct *napi) struct sk_buff *skb, *next; list_del_init(&napi->dev_list); - kfree(napi->skb); + kfree_skb(napi->skb); for (skb = napi->gro_list; skb; skb = next) { next = skb->next; -- cgit v1.2.3 From 2bad35b7c9588eb5e65c03bcae54e7eb6b1a6504 Mon Sep 17 00:00:00 2001 From: Jorge Boncompte [DTI2] Date: Wed, 18 Mar 2009 23:26:11 -0700 Subject: netns: oops in ip[6]_frag_reasm incrementing stats dev can be NULL in ip[6]_frag_reasm for skb's coming from RAW sockets. Quagga's OSPFD sends fragmented packets on a RAW socket, when netfilter conntrack reassembles them on the OUTPUT path you hit this code path. You can test it with something like "hping2 -0 -d 2000 -f AA.BB.CC.DD" With help from Jarek Poplawski. Signed-off-by: Jorge Boncompte [DTI2] Signed-off-by: David S. Miller --- net/ipv4/ip_fragment.c | 3 ++- net/ipv6/reassembly.c | 7 +++---- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c index 6659ac000eeb..7985346653bd 100644 --- a/net/ipv4/ip_fragment.c +++ b/net/ipv4/ip_fragment.c @@ -463,6 +463,7 @@ err: static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev, struct net_device *dev) { + struct net *net = container_of(qp->q.net, struct net, ipv4.frags); struct iphdr *iph; struct sk_buff *fp, *head = qp->q.fragments; int len; @@ -548,7 +549,7 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev, iph = ip_hdr(head); iph->frag_off = 0; iph->tot_len = htons(len); - IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_REASMOKS); + IP_INC_STATS_BH(net, IPSTATS_MIB_REASMOKS); qp->q.fragments = NULL; return 0; diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c index 3c575118fca5..e9ac7a12f595 100644 --- a/net/ipv6/reassembly.c +++ b/net/ipv6/reassembly.c @@ -452,6 +452,7 @@ err: static int ip6_frag_reasm(struct frag_queue *fq, struct sk_buff *prev, struct net_device *dev) { + struct net *net = container_of(fq->q.net, struct net, ipv6.frags); struct sk_buff *fp, *head = fq->q.fragments; int payload_len; unsigned int nhoff; @@ -551,8 +552,7 @@ static int ip6_frag_reasm(struct frag_queue *fq, struct sk_buff *prev, head->csum); rcu_read_lock(); - IP6_INC_STATS_BH(dev_net(dev), - __in6_dev_get(dev), IPSTATS_MIB_REASMOKS); + IP6_INC_STATS_BH(net, __in6_dev_get(dev), IPSTATS_MIB_REASMOKS); rcu_read_unlock(); fq->q.fragments = NULL; return 1; @@ -566,8 +566,7 @@ out_oom: printk(KERN_DEBUG "ip6_frag_reasm: no memory for reassembly\n"); out_fail: rcu_read_lock(); - IP6_INC_STATS_BH(dev_net(dev), - __in6_dev_get(dev), IPSTATS_MIB_REASMFAILS); + IP6_INC_STATS_BH(net, __in6_dev_get(dev), IPSTATS_MIB_REASMFAILS); rcu_read_unlock(); return -1; } -- cgit v1.2.3 From 0e0fde3c8d65524b8dfd834332d6e4a92711a66a Mon Sep 17 00:00:00 2001 From: Nobuhiro Iwamatsu Date: Mon, 16 Mar 2009 19:50:57 +0000 Subject: sh_eth: Change handling of IRQ Handling of IRQ of the SH7763/SH7764 CPU which sh_eth supported was changed. This revises it for this change. Signed-off-by: Nobuhiro Iwamatsu Signed-off-by: David S. Miller --- drivers/net/sh_eth.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c index 7f8e514eb5e9..7b1882765a0c 100644 --- a/drivers/net/sh_eth.c +++ b/drivers/net/sh_eth.c @@ -687,6 +687,7 @@ static irqreturn_t sh_eth_interrupt(int irq, void *netdev) { struct net_device *ndev = netdev; struct sh_eth_private *mdp = netdev_priv(ndev); + irqreturn_t ret = IRQ_NONE; u32 ioaddr, boguscnt = RX_RING_SIZE; u32 intr_status = 0; @@ -696,7 +697,13 @@ static irqreturn_t sh_eth_interrupt(int irq, void *netdev) /* Get interrpt stat */ intr_status = ctrl_inl(ioaddr + EESR); /* Clear interrupt */ - ctrl_outl(intr_status, ioaddr + EESR); + if (intr_status & (EESR_FRC | EESR_RMAF | EESR_RRF | + EESR_RTLF | EESR_RTSF | EESR_PRE | EESR_CERF | + TX_CHECK | EESR_ERR_CHECK)) { + ctrl_outl(intr_status, ioaddr + EESR); + ret = IRQ_HANDLED; + } else + goto other_irq; if (intr_status & (EESR_FRC | /* Frame recv*/ EESR_RMAF | /* Multi cast address recv*/ @@ -723,9 +730,10 @@ static irqreturn_t sh_eth_interrupt(int irq, void *netdev) ndev->name, intr_status); } +other_irq: spin_unlock(&mdp->lock); - return IRQ_HANDLED; + return ret; } static void sh_eth_timer(unsigned long data) @@ -844,7 +852,13 @@ static int sh_eth_open(struct net_device *ndev) int ret = 0; struct sh_eth_private *mdp = netdev_priv(ndev); - ret = request_irq(ndev->irq, &sh_eth_interrupt, 0, ndev->name, ndev); + ret = request_irq(ndev->irq, &sh_eth_interrupt, +#if defined(CONFIG_CPU_SUBTYPE_SH7763) || defined(CONFIG_CPU_SUBTYPE_SH7764) + IRQF_SHARED, +#else + 0, +#endif + ndev->name, ndev); if (ret) { printk(KERN_ERR "Can not assign IRQ number to %s\n", CARDNAME); return ret; -- cgit v1.2.3 From 2e2a6a9f710255c87cef670fb71fc9e74bef1da2 Mon Sep 17 00:00:00 2001 From: Nobuhiro Iwamatsu Date: Mon, 16 Mar 2009 19:52:23 +0000 Subject: sh_eth: Fix mistake of the address of SH7763 Address of SH_TSU_ADDR and ARSTR of SH7763 was wrong. This revise it. Signed-off-by: Nobuhiro Iwamatsu Signed-off-by: David S. Miller --- drivers/net/sh_eth.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/sh_eth.h b/drivers/net/sh_eth.h index 73bc7181cc18..1537e13e623d 100644 --- a/drivers/net/sh_eth.h +++ b/drivers/net/sh_eth.h @@ -43,8 +43,8 @@ #define SH7763_SKB_ALIGN 32 /* Chip Base Address */ -# define SH_TSU_ADDR 0xFFE01800 -# define ARSTR 0xFFE01800 +# define SH_TSU_ADDR 0xFEE01800 +# define ARSTR SH_TSU_ADDR /* Chip Registers */ /* E-DMAC */ -- cgit v1.2.3 From 44c1d6f99d4c86638bffabf0b7a232d0fe7ae574 Mon Sep 17 00:00:00 2001 From: Steve Glendinning Date: Wed, 18 Mar 2009 23:37:18 -0700 Subject: smsc911x: reset last known duplex and carrier on open smsc911x_phy_adjust_link is called periodically by the phy layer (as it's run in polling mode), and it only updates the hardware when it sees a change in duplex or carrier. This patch clears the last known values every time the interface is brought up, instead of only when the module is loaded. Without this patch the adjust_link function never updates the hardware after an ifconfig down; ifconfig up. On a full duplex link this causes the tx error counter to increment, even though packets are correctly transmitted, as the default MAC_CR register setting is for half duplex. The tx errors are "no carrier" errors, which should be ignored in full-duplex mode. When MAC_CR is set to "full duplex" mode they are correctly ignored by the hardware. Note that even with this patch the tx error counter can increment if packets are transmitted between "ifconfig up" and the first phy poll interval. An improved solution would use the phy interrupt with phylib, but I haven't managed to make this work 100% robustly yet. Signed-off-by: Steve Glendinning Signed-off-by: David S. Miller --- drivers/net/smsc911x.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/smsc911x.c b/drivers/net/smsc911x.c index 9a78daec2fe9..d1590ac55e4b 100644 --- a/drivers/net/smsc911x.c +++ b/drivers/net/smsc911x.c @@ -1225,6 +1225,10 @@ static int smsc911x_open(struct net_device *dev) dev_info(&dev->dev, "SMSC911x/921x identified at %#08lx, IRQ: %d\n", (unsigned long)pdata->ioaddr, dev->irq); + /* Reset the last known duplex and carrier */ + pdata->last_duplex = -1; + pdata->last_carrier = -1; + /* Bring the PHY up */ phy_start(pdata->phy_dev); -- cgit v1.2.3 From 170ebf85160dd128e1c4206cc197cce7d1424705 Mon Sep 17 00:00:00 2001 From: Tilman Schmidt Date: Wed, 18 Mar 2009 23:44:23 -0700 Subject: bas_gigaset: correctly allocate USB interrupt transfer buffer Every USB transfer buffer has to be allocated individually by kmalloc. Impact: bugfix, no functional change Signed-off-by: Tilman Schmidt Tested-by: Kolja Waschk Signed-off-by: David S. Miller --- drivers/isdn/gigaset/bas-gigaset.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/drivers/isdn/gigaset/bas-gigaset.c b/drivers/isdn/gigaset/bas-gigaset.c index 18dd8aacbe8d..831ddce1467b 100644 --- a/drivers/isdn/gigaset/bas-gigaset.c +++ b/drivers/isdn/gigaset/bas-gigaset.c @@ -46,6 +46,9 @@ MODULE_PARM_DESC(cidmode, "Call-ID mode"); /* length limit according to Siemens 3070usb-protokoll.doc ch. 2.1 */ #define IF_WRITEBUF 264 +/* interrupt pipe message size according to ibid. ch. 2.2 */ +#define IP_MSGSIZE 3 + /* Values for the Gigaset 307x */ #define USB_GIGA_VENDOR_ID 0x0681 #define USB_3070_PRODUCT_ID 0x0001 @@ -110,7 +113,7 @@ struct bas_cardstate { unsigned char *rcvbuf; /* AT reply receive buffer */ struct urb *urb_int_in; /* URB for interrupt pipe */ - unsigned char int_in_buf[3]; + unsigned char *int_in_buf; spinlock_t lock; /* locks all following */ int basstate; /* bitmap (BS_*) */ @@ -657,7 +660,7 @@ static void read_int_callback(struct urb *urb) } /* drop incomplete packets even if the missing bytes wouldn't matter */ - if (unlikely(urb->actual_length < 3)) { + if (unlikely(urb->actual_length < IP_MSGSIZE)) { dev_warn(cs->dev, "incomplete interrupt packet (%d bytes)\n", urb->actual_length); goto resubmit; @@ -2127,6 +2130,7 @@ static void gigaset_reinitbcshw(struct bc_state *bcs) static void gigaset_freecshw(struct cardstate *cs) { /* timers, URBs and rcvbuf are disposed of in disconnect */ + kfree(cs->hw.bas->int_in_buf); kfree(cs->hw.bas); cs->hw.bas = NULL; } @@ -2140,6 +2144,12 @@ static int gigaset_initcshw(struct cardstate *cs) pr_err("out of memory\n"); return 0; } + ucs->int_in_buf = kmalloc(IP_MSGSIZE, GFP_KERNEL); + if (!ucs->int_in_buf) { + kfree(ucs); + pr_err("out of memory\n"); + return 0; + } ucs->urb_cmd_in = NULL; ucs->urb_cmd_out = NULL; @@ -2292,7 +2302,7 @@ static int gigaset_probe(struct usb_interface *interface, usb_fill_int_urb(ucs->urb_int_in, udev, usb_rcvintpipe(udev, (endpoint->bEndpointAddress) & 0x0f), - ucs->int_in_buf, 3, read_int_callback, cs, + ucs->int_in_buf, IP_MSGSIZE, read_int_callback, cs, endpoint->bInterval); if ((rc = usb_submit_urb(ucs->urb_int_in, GFP_KERNEL)) != 0) { dev_err(cs->dev, "could not submit interrupt URB: %s\n", -- cgit v1.2.3 From ea1dae11e0baca5d633207fe50fc3cd30a5d68ee Mon Sep 17 00:00:00 2001 From: Sathya Perla Date: Thu, 19 Mar 2009 23:56:20 -0700 Subject: be2net: replenish when posting to rx-queue is starved in out of mem conditions This is a patch to replenish the rx-queue when it is in a starved state (due to out-of-mem conditions) Signed-off-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/benet/be.h | 1 + drivers/net/benet/be_main.c | 50 ++++++++++++++++++++++++++------------------- 2 files changed, 30 insertions(+), 21 deletions(-) diff --git a/drivers/net/benet/be.h b/drivers/net/benet/be.h index 63d593d53153..f327be57ca96 100644 --- a/drivers/net/benet/be.h +++ b/drivers/net/benet/be.h @@ -194,6 +194,7 @@ struct be_adapter { struct be_eq_obj rx_eq; struct be_rx_obj rx_obj; u32 big_page_size; /* Compounded page size shared by rx wrbs */ + bool rx_post_starved; /* Zero rx frags have been posted to BE */ struct vlan_group *vlan_grp; u16 num_vlans; diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c index 897a63de5bdb..80fe1e055d91 100644 --- a/drivers/net/benet/be_main.c +++ b/drivers/net/benet/be_main.c @@ -273,26 +273,6 @@ static void be_rx_eqd_update(struct be_adapter *adapter) rx_eq->cur_eqd = eqd; } -static void be_worker(struct work_struct *work) -{ - struct be_adapter *adapter = - container_of(work, struct be_adapter, work.work); - int status; - - /* Check link */ - be_link_status_update(adapter); - - /* Get Stats */ - status = be_cmd_get_stats(&adapter->ctrl, &adapter->stats.cmd); - if (!status) - netdev_stats_update(adapter); - - /* Set EQ delay */ - be_rx_eqd_update(adapter); - - schedule_delayed_work(&adapter->work, msecs_to_jiffies(1000)); -} - static struct net_device_stats *be_get_stats(struct net_device *dev) { struct be_adapter *adapter = netdev_priv(dev); @@ -900,8 +880,11 @@ static void be_post_rx_frags(struct be_adapter *adapter) page_info->last_page_user = true; if (posted) { - be_rxq_notify(&adapter->ctrl, rxq->id, posted); atomic_add(posted, &rxq->used); + be_rxq_notify(&adapter->ctrl, rxq->id, posted); + } else if (atomic_read(&rxq->used) == 0) { + /* Let be_worker replenish when memory is available */ + adapter->rx_post_starved = true; } return; @@ -1305,6 +1288,31 @@ int be_poll_tx(struct napi_struct *napi, int budget) return 1; } +static void be_worker(struct work_struct *work) +{ + struct be_adapter *adapter = + container_of(work, struct be_adapter, work.work); + int status; + + /* Check link */ + be_link_status_update(adapter); + + /* Get Stats */ + status = be_cmd_get_stats(&adapter->ctrl, &adapter->stats.cmd); + if (!status) + netdev_stats_update(adapter); + + /* Set EQ delay */ + be_rx_eqd_update(adapter); + + if (adapter->rx_post_starved) { + adapter->rx_post_starved = false; + be_post_rx_frags(adapter); + } + + schedule_delayed_work(&adapter->work, msecs_to_jiffies(1000)); +} + static void be_msix_enable(struct be_adapter *adapter) { int i, status; -- cgit v1.2.3 From 1ab1ab7543de53c945ea24140409ef67ed173eb4 Mon Sep 17 00:00:00 2001 From: Sathya Perla Date: Thu, 19 Mar 2009 23:56:46 -0700 Subject: be2net: fix to restore vlan ids into BE2 during a IF DOWN->UP cycle This is a patch to reconfigure vlan-ids during an i/f down/up cycle Signed-off-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/benet/be_main.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c index 80fe1e055d91..0ecaffb70e58 100644 --- a/drivers/net/benet/be_main.c +++ b/drivers/net/benet/be_main.c @@ -473,7 +473,7 @@ static int be_change_mtu(struct net_device *netdev, int new_mtu) * program them in BE. If more than BE_NUM_VLANS_SUPPORTED are configured, * set the BE in promiscuous VLAN mode. */ -static void be_vids_config(struct net_device *netdev) +static void be_vid_config(struct net_device *netdev) { struct be_adapter *adapter = netdev_priv(netdev); u16 vtag[BE_NUM_VLANS_SUPPORTED]; @@ -516,7 +516,7 @@ static void be_vlan_add_vid(struct net_device *netdev, u16 vid) adapter->num_vlans++; adapter->vlan_tag[vid] = 1; - be_vids_config(netdev); + be_vid_config(netdev); } static void be_vlan_rem_vid(struct net_device *netdev, u16 vid) @@ -527,7 +527,7 @@ static void be_vlan_rem_vid(struct net_device *netdev, u16 vid) adapter->vlan_tag[vid] = 0; vlan_group_set_device(adapter->vlan_grp, vid, NULL); - be_vids_config(netdev); + be_vid_config(netdev); } static void be_set_multicast_filter(struct net_device *netdev) @@ -1430,6 +1430,8 @@ static int be_open(struct net_device *netdev) if (status != 0) goto do_none; + be_vid_config(netdev); + status = be_cmd_set_flow_control(ctrl, true, true); if (status != 0) goto if_destroy; @@ -1864,8 +1866,6 @@ static int be_resume(struct pci_dev *pdev) pci_set_power_state(pdev, 0); pci_restore_state(pdev); - be_vids_config(netdev); - if (netif_running(netdev)) { rtnl_lock(); be_open(netdev); -- cgit v1.2.3 From 5ed0102fbf36f58091089907213b4bd191ca2e0c Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Thu, 19 Mar 2009 23:58:01 -0700 Subject: sungem: missing net_device_ops Sungem driver only got partially converted to net_device_ops. Since this could cause bugs, please push this to 2.6.29 Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/net/sungem.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c index 0fcb7503363d..c9c7650826c0 100644 --- a/drivers/net/sungem.c +++ b/drivers/net/sungem.c @@ -2998,8 +2998,11 @@ static const struct net_device_ops gem_netdev_ops = { .ndo_do_ioctl = gem_ioctl, .ndo_tx_timeout = gem_tx_timeout, .ndo_change_mtu = gem_change_mtu, - .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, + .ndo_set_mac_address = gem_set_mac_address, +#ifdef CONFIG_NET_POLL_CONTROLLER + .ndo_poll_controller = gem_poll_controller, +#endif }; static int __devinit gem_init_one(struct pci_dev *pdev, @@ -3161,10 +3164,6 @@ static int __devinit gem_init_one(struct pci_dev *pdev, dev->watchdog_timeo = 5 * HZ; dev->irq = pdev->irq; dev->dma = 0; - dev->set_mac_address = gem_set_mac_address; -#ifdef CONFIG_NET_POLL_CONTROLLER - dev->poll_controller = gem_poll_controller; -#endif /* Set that now, in case PM kicks in now */ pci_set_drvdata(pdev, dev); -- cgit v1.2.3 From f3f9258678b081c3ef2f036aef450cd2053ef419 Mon Sep 17 00:00:00 2001 From: Jouni Malinen Date: Fri, 20 Mar 2009 17:57:36 +0200 Subject: nl80211: Check that function pointer != NULL before using it NL80211_CMD_GET_MESH_PARAMS and NL80211_CMD_SET_MESH_PARAMS handlers did not verify whether a function pointer is NULL (not supported by the driver) before trying to call the function. The former nl80211 command is available for unprivileged users, too, so this can potentially allow normal users to kill networking (or worse..) if mac80211 is built without CONFIG_MAC80211_MESH=y. Signed-off-by: Jouni Malinen Signed-off-by: John W. Linville --- net/wireless/nl80211.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 1e728fff474e..31b807af3235 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -1908,6 +1908,11 @@ static int nl80211_get_mesh_params(struct sk_buff *skb, if (err) return err; + if (!drv->ops->get_mesh_params) { + err = -EOPNOTSUPP; + goto out; + } + /* Get the mesh params */ rtnl_lock(); err = drv->ops->get_mesh_params(&drv->wiphy, dev, &cur_params); @@ -2017,6 +2022,11 @@ static int nl80211_set_mesh_params(struct sk_buff *skb, struct genl_info *info) if (err) return err; + if (!drv->ops->set_mesh_params) { + err = -EOPNOTSUPP; + goto out; + } + /* This makes sure that there aren't more than 32 mesh config * parameters (otherwise our bitfield scheme would not work.) */ BUILD_BUG_ON(NL80211_MESHCONF_ATTR_MAX > 32); @@ -2061,6 +2071,7 @@ static int nl80211_set_mesh_params(struct sk_buff *skb, struct genl_info *info) err = drv->ops->set_mesh_params(&drv->wiphy, dev, &cfg, mask); rtnl_unlock(); + out: /* cleanup */ cfg80211_put_dev(drv); dev_put(dev); -- cgit v1.2.3 From e2fc4d19292ef2eb208f76976ddc3320cc5839b6 Mon Sep 17 00:00:00 2001 From: Maciej Sosnowski Date: Sat, 21 Mar 2009 13:31:23 -0700 Subject: dca: add missing copyright/license headers In two dca files copyright and license headers are missing. This patch adds them there. Signed-off-by: Maciej Sosnowski Signed-off-by: David S. Miller --- drivers/dca/dca-sysfs.c | 21 +++++++++++++++++++++ include/linux/dca.h | 20 ++++++++++++++++++++ 2 files changed, 41 insertions(+) diff --git a/drivers/dca/dca-sysfs.c b/drivers/dca/dca-sysfs.c index bb538b9690e0..ee916c9857ee 100644 --- a/drivers/dca/dca-sysfs.c +++ b/drivers/dca/dca-sysfs.c @@ -1,3 +1,24 @@ +/* + * Copyright(c) 2007 - 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ + #include #include #include diff --git a/include/linux/dca.h b/include/linux/dca.h index b00a753eda53..9c20c7e87d0a 100644 --- a/include/linux/dca.h +++ b/include/linux/dca.h @@ -1,3 +1,23 @@ +/* + * Copyright(c) 2007 - 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ #ifndef DCA_H #define DCA_H /* DCA Provider API */ -- cgit v1.2.3 From 4b97926ddf51b3919c859e2086fef3c8c3c46c61 Mon Sep 17 00:00:00 2001 From: Ilya Yanok Date: Sat, 21 Mar 2009 16:58:47 -0700 Subject: dnet: DNET should depend on HAS_IOMEM Signed-off-by: Ilya Yanok Signed-off-by: David S. Miller --- drivers/net/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index 435e2e3a82c8..62d732a886f1 100644 --- a/drivers/net/Kconfig +++ b/drivers/net/Kconfig @@ -1042,7 +1042,7 @@ config NI65 config DNET tristate "Dave ethernet support (DNET)" - depends on NET_ETHERNET + depends on NET_ETHERNET && HAS_IOMEM select PHYLIB help The Dave ethernet interface (DNET) is found on Qong Board FPGA. -- cgit v1.2.3 From 6580f57d485f70851218813fa053d971915f61fb Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Sun, 22 Mar 2009 21:22:48 -0700 Subject: net: update dnet.c for bus_id removal Signed-off-by: Stephen Rothwell Signed-off-by: David S. Miller --- drivers/net/dnet.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c index 5c347f70cb67..1b4063222a82 100644 --- a/drivers/net/dnet.c +++ b/drivers/net/dnet.c @@ -280,11 +280,11 @@ static int dnet_mii_probe(struct net_device *dev) /* attach the mac to the phy */ if (bp->capabilities & DNET_HAS_RMII) { - phydev = phy_connect(dev, phydev->dev.bus_id, + phydev = phy_connect(dev, dev_name(&phydev->dev), &dnet_handle_link_change, 0, PHY_INTERFACE_MODE_RMII); } else { - phydev = phy_connect(dev, phydev->dev.bus_id, + phydev = phy_connect(dev, dev_name(&phydev->dev), &dnet_handle_link_change, 0, PHY_INTERFACE_MODE_MII); } @@ -927,7 +927,7 @@ static int __devinit dnet_probe(struct platform_device *pdev) phydev = bp->phy_dev; dev_info(&pdev->dev, "attached PHY driver [%s] " "(mii_bus:phy_addr=%s, irq=%d)\n", - phydev->drv->name, phydev->dev.bus_id, phydev->irq); + phydev->drv->name, dev_name(&phydev->dev), phydev->irq); return 0; -- cgit v1.2.3 From e3162d381fc359ebe5c98a3e216888a7cb200051 Mon Sep 17 00:00:00 2001 From: David Brownell Date: Sun, 22 Mar 2009 21:28:39 -0700 Subject: dm9000: locking bugfix This fixes a locking bug in the dm9000 driver. It calls request_irq() without setting IRQF_DISABLED ... which is correct for handlers that support IRQ sharing, since that behavior is not guaranteed for shared IRQs. However, its IRQ handler then wrongly assumes that IRQs are blocked. So the fix just uses the right spinlock primitives in the IRQ handler. NOTE: this is a classic example of the type of bug which lockdep currently masks by forcibly setting IRQF_DISABLED on IRQ handlers that did not request that flag. Signed-off-by: David Brownell Signed-off-by: David S. Miller --- drivers/net/dm9000.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/dm9000.c b/drivers/net/dm9000.c index bcf92917bbf3..254ec62b5f58 100644 --- a/drivers/net/dm9000.c +++ b/drivers/net/dm9000.c @@ -930,13 +930,15 @@ static irqreturn_t dm9000_interrupt(int irq, void *dev_id) struct net_device *dev = dev_id; board_info_t *db = netdev_priv(dev); int int_status; + unsigned long flags; u8 reg_save; dm9000_dbg(db, 3, "entering %s\n", __func__); /* A real interrupt coming */ - spin_lock(&db->lock); + /* holders of db->lock must always block IRQs */ + spin_lock_irqsave(&db->lock, flags); /* Save previous register address */ reg_save = readb(db->io_addr); @@ -972,7 +974,7 @@ static irqreturn_t dm9000_interrupt(int irq, void *dev_id) /* Restore previous register address */ writeb(reg_save, db->io_addr); - spin_unlock(&db->lock); + spin_unlock_irqrestore(&db->lock, flags); return IRQ_HANDLED; } -- cgit v1.2.3 From 61fa9dcf9329cb92c220f7b656410fbe5e72f933 Mon Sep 17 00:00:00 2001 From: Anton Vorontsov Date: Sun, 22 Mar 2009 21:30:52 -0700 Subject: ucc_geth: Fix oops when using fixed-link support commit b1c4a9dddf09fe99b8f88252718ac5b357363dc4 ("ucc_geth: Change uec phy id to the same format as gianfar's") introduced a regression in the ucc_geth driver that causes this oops when fixed-link is used: Unable to handle kernel paging request for data at address 0x00000000 Faulting instruction address: 0xc0151270 Oops: Kernel access of bad area, sig: 11 [#1] TMCUTU NIP: c0151270 LR: c0151270 CTR: c0017760 REGS: cf81fa60 TRAP: 0300 Not tainted (2.6.29-rc8) MSR: 00009032 CR: 24024042 XER: 20000000 DAR: 00000000, DSISR: 20000000 TASK = cf81cba0[1] 'swapper' THREAD: cf81e000 GPR00: c0151270 cf81fb10 cf81cba0 00000000 c0272e20 c025f354 00001e80 cf86b08c GPR08: d1068200 cffffb74 06000000 d106c200 42024042 10085148 0fffd000 0ffc81a0 GPR16: 00000001 00000001 00000000 007ffeb0 00000000 0000c000 cf83f36c cf83f000 GPR24: 00000030 cf83f360 cf81fb20 00000000 d106c200 20000000 00001e80 cf83f360 NIP [c0151270] ucc_geth_open+0x330/0x1efc LR [c0151270] ucc_geth_open+0x330/0x1efc Call Trace: [cf81fb10] [c0151270] ucc_geth_open+0x330/0x1efc (unreliable) [cf81fba0] [c0187638] dev_open+0xbc/0x12c [cf81fbc0] [c0187e38] dev_change_flags+0x8c/0x1b0 This patch fixes the issue by removing offending (and somewhat duplicate) code from init_phy() routine, and changes _probe() function to use uec_mdio_bus_name(). Also, since we fully construct phy_bus_id in the _probe() routine, we no longer need ->phy_address and ->mdio_bus fields in ucc_geth_info structure. I wish the patch would be a bit shorter, but it seems like the only way to fix the issue in a sane way. Luckily, the patch has been tested with real PHYs and fixed-link, so no further regressions expected. Reported-by: Joakim Tjernlund Signed-off-by: Anton Vorontsov Tested-by: Joakim Tjernlund Signed-off-by: David S. Miller --- drivers/net/ucc_geth.c | 34 ++++++++++------------------------ drivers/net/ucc_geth.h | 3 +-- 2 files changed, 11 insertions(+), 26 deletions(-) diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c index e87986867ba5..1f61e42c641d 100644 --- a/drivers/net/ucc_geth.c +++ b/drivers/net/ucc_geth.c @@ -1536,32 +1536,15 @@ static void adjust_link(struct net_device *dev) static int init_phy(struct net_device *dev) { struct ucc_geth_private *priv = netdev_priv(dev); - struct device_node *np = priv->node; - struct device_node *phy, *mdio; - const phandle *ph; - char bus_name[MII_BUS_ID_SIZE]; - const unsigned int *id; + struct ucc_geth_info *ug_info = priv->ug_info; struct phy_device *phydev; - char phy_id[BUS_ID_SIZE]; priv->oldlink = 0; priv->oldspeed = 0; priv->oldduplex = -1; - ph = of_get_property(np, "phy-handle", NULL); - phy = of_find_node_by_phandle(*ph); - mdio = of_get_parent(phy); - - id = of_get_property(phy, "reg", NULL); - - of_node_put(phy); - of_node_put(mdio); - - uec_mdio_bus_name(bus_name, mdio); - snprintf(phy_id, sizeof(phy_id), "%s:%02x", - bus_name, *id); - - phydev = phy_connect(dev, phy_id, &adjust_link, 0, priv->phy_interface); + phydev = phy_connect(dev, ug_info->phy_bus_id, &adjust_link, 0, + priv->phy_interface); if (IS_ERR(phydev)) { printk("%s: Could not attach to PHY\n", dev->name); @@ -3629,10 +3612,12 @@ static int ucc_geth_probe(struct of_device* ofdev, const struct of_device_id *ma ug_info->uf_info.irq = irq_of_parse_and_map(np, 0); fixed_link = of_get_property(np, "fixed-link", NULL); if (fixed_link) { - snprintf(ug_info->mdio_bus, MII_BUS_ID_SIZE, "0"); - ug_info->phy_address = fixed_link[0]; + snprintf(ug_info->phy_bus_id, sizeof(ug_info->phy_bus_id), + PHY_ID_FMT, "0", fixed_link[0]); phy = NULL; } else { + char bus_name[MII_BUS_ID_SIZE]; + ph = of_get_property(np, "phy-handle", NULL); phy = of_find_node_by_phandle(*ph); @@ -3643,7 +3628,6 @@ static int ucc_geth_probe(struct of_device* ofdev, const struct of_device_id *ma prop = of_get_property(phy, "reg", NULL); if (prop == NULL) return -1; - ug_info->phy_address = *prop; /* Set the bus id */ mdio = of_get_parent(phy); @@ -3657,7 +3641,9 @@ static int ucc_geth_probe(struct of_device* ofdev, const struct of_device_id *ma if (err) return -1; - snprintf(ug_info->mdio_bus, MII_BUS_ID_SIZE, "%x", res.start); + uec_mdio_bus_name(bus_name, mdio); + snprintf(ug_info->phy_bus_id, sizeof(ug_info->phy_bus_id), + "%s:%02x", bus_name, *prop); } /* get the phy interface type, or default to MII */ diff --git a/drivers/net/ucc_geth.h b/drivers/net/ucc_geth.h index 16cbe42ba43c..611bdef2402b 100644 --- a/drivers/net/ucc_geth.h +++ b/drivers/net/ucc_geth.h @@ -1091,8 +1091,7 @@ struct ucc_geth_info { u32 eventRegMask; u16 pausePeriod; u16 extensionField; - u8 phy_address; - char mdio_bus[MII_BUS_ID_SIZE]; + char phy_bus_id[BUS_ID_SIZE]; u8 weightfactor[NUM_TX_QUEUES]; u8 interruptcoalescingmaxvalue[NUM_RX_QUEUES]; u8 l2qt[UCC_GETH_VLAN_PRIORITY_MAX]; -- cgit v1.2.3