From fa86d322d89995fef1bfb5cc768b89d8c22ea0d9 Mon Sep 17 00:00:00 2001 From: Pavel Emelyanov Date: Mon, 24 Mar 2008 14:48:59 -0700 Subject: [NEIGH]: Fix race between pneigh deletion and ipv6's ndisc_recv_ns (v3). Proxy neighbors do not have any reference counting, so any caller of pneigh_lookup (unless it's a netlink triggered add/del routine) should _not_ perform any actions on the found proxy entry. There's one exception from this rule - the ipv6's ndisc_recv_ns() uses found entry to check the flags for NTF_ROUTER. This creates a race between the ndisc and pneigh_delete - after the pneigh is returned to the caller, the nd_tbl.lock is dropped and the deleting procedure may proceed. One of the fixes would be to add a reference counting, but this problem exists for ndisc only. Besides such a patch would be too big for -rc4. So I propose to introduce a __pneigh_lookup() which is supposed to be called with the lock held and use it in ndisc code to check the flags on alive pneigh entry. Changes from v2: As David noticed, Exported the __pneigh_lookup() to ipv6 module. The checkpatch generates a warning on it, since the EXPORT_SYMBOL does not follow the symbol itself, but in this file all the exports come at the end, so I decided no to break this harmony. Changes from v1: Fixed comments from YOSHIFUJI - indentation of prototype in header and the pndisc_check_router() name - and a compilation fix, pointed by Daniel - the is_routed was (falsely) considered as uninitialized by gcc. Signed-off-by: Pavel Emelyanov Signed-off-by: David S. Miller diff --git a/include/net/neighbour.h b/include/net/neighbour.h index ebbfb50..64a5f01 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -218,6 +218,10 @@ extern unsigned long neigh_rand_reach_time(unsigned long base); extern void pneigh_enqueue(struct neigh_table *tbl, struct neigh_parms *p, struct sk_buff *skb); extern struct pneigh_entry *pneigh_lookup(struct neigh_table *tbl, struct net *net, const void *key, struct net_device *dev, int creat); +extern struct pneigh_entry *__pneigh_lookup(struct neigh_table *tbl, + struct net *net, + const void *key, + struct net_device *dev); extern int pneigh_delete(struct neigh_table *tbl, struct net *net, const void *key, struct net_device *dev); extern void neigh_app_ns(struct neighbour *n); diff --git a/net/core/neighbour.c b/net/core/neighbour.c index d9a02b2..19b8e00 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -466,6 +466,28 @@ out_neigh_release: goto out; } +struct pneigh_entry *__pneigh_lookup(struct neigh_table *tbl, + struct net *net, const void *pkey, struct net_device *dev) +{ + struct pneigh_entry *n; + int key_len = tbl->key_len; + u32 hash_val = *(u32 *)(pkey + key_len - 4); + + hash_val ^= (hash_val >> 16); + hash_val ^= hash_val >> 8; + hash_val ^= hash_val >> 4; + hash_val &= PNEIGH_HASHMASK; + + for (n = tbl->phash_buckets[hash_val]; n; n = n->next) { + if (!memcmp(n->key, pkey, key_len) && + (n->net == net) && + (n->dev == dev || !n->dev)) + break; + } + + return n; +} + struct pneigh_entry * pneigh_lookup(struct neigh_table *tbl, struct net *net, const void *pkey, struct net_device *dev, int creat) @@ -2803,6 +2825,7 @@ EXPORT_SYMBOL(neigh_table_init_no_netlink); EXPORT_SYMBOL(neigh_update); EXPORT_SYMBOL(pneigh_enqueue); EXPORT_SYMBOL(pneigh_lookup); +EXPORT_SYMBOL_GPL(__pneigh_lookup); #ifdef CONFIG_ARPD EXPORT_SYMBOL(neigh_app_ns); diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index 51557c2..452a2ac 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -676,6 +676,20 @@ static void ndisc_solicit(struct neighbour *neigh, struct sk_buff *skb) } } +static struct pneigh_entry *pndisc_check_router(struct net_device *dev, + struct in6_addr *addr, int *is_router) +{ + struct pneigh_entry *n; + + read_lock_bh(&nd_tbl.lock); + n = __pneigh_lookup(&nd_tbl, &init_net, addr, dev); + if (n != NULL) + *is_router = (n->flags & NTF_ROUTER); + read_unlock_bh(&nd_tbl.lock); + + return n; +} + static void ndisc_recv_ns(struct sk_buff *skb) { struct nd_msg *msg = (struct nd_msg *)skb_transport_header(skb); @@ -692,7 +706,7 @@ static void ndisc_recv_ns(struct sk_buff *skb) struct pneigh_entry *pneigh = NULL; int dad = ipv6_addr_any(saddr); int inc; - int is_router; + int is_router = 0; if (ipv6_addr_is_multicast(&msg->target)) { ND_PRINTK2(KERN_WARNING @@ -790,8 +804,8 @@ static void ndisc_recv_ns(struct sk_buff *skb) if (ipv6_chk_acast_addr(dev, &msg->target) || (idev->cnf.forwarding && (ipv6_devconf.proxy_ndp || idev->cnf.proxy_ndp) && - (pneigh = pneigh_lookup(&nd_tbl, &init_net, - &msg->target, dev, 0)) != NULL)) { + (pneigh = pndisc_check_router(dev, &msg->target, + &is_router)) != NULL)) { if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) && skb->pkt_type != PACKET_HOST && inc != 0 && @@ -812,7 +826,7 @@ static void ndisc_recv_ns(struct sk_buff *skb) goto out; } - is_router = !!(pneigh ? pneigh->flags & NTF_ROUTER : idev->cnf.forwarding); + is_router = !!(pneigh ? is_router : idev->cnf.forwarding); if (dad) { struct in6_addr maddr; -- cgit v0.10.2 From df9dcb4588aca9cc243cf1f3f454361a84e1cbdb Mon Sep 17 00:00:00 2001 From: Kazunori MIYAZAWA Date: Mon, 24 Mar 2008 14:51:51 -0700 Subject: [IPSEC]: Fix inter address family IPsec tunnel handling. Signed-off-by: Kazunori MIYAZAWA Signed-off-by: David S. Miller diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 619c53b..4e6f956 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -204,6 +204,7 @@ struct xfrm_state * transformer. */ const struct xfrm_type *type; struct xfrm_mode *inner_mode; + struct xfrm_mode *inner_mode_iaf; struct xfrm_mode *outer_mode; /* Security context */ @@ -387,6 +388,27 @@ enum { extern int xfrm_register_mode(struct xfrm_mode *mode, int family); extern int xfrm_unregister_mode(struct xfrm_mode *mode, int family); +static inline int xfrm_af2proto(unsigned int family) +{ + switch(family) { + case AF_INET: + return IPPROTO_IPIP; + case AF_INET6: + return IPPROTO_IPV6; + default: + return 0; + } +} + +static inline struct xfrm_mode *xfrm_ip2inner_mode(struct xfrm_state *x, int ipproto) +{ + if ((ipproto == IPPROTO_IPIP && x->props.family == AF_INET) || + (ipproto == IPPROTO_IPV6 && x->props.family == AF_INET6)) + return x->inner_mode; + else + return x->inner_mode_iaf; +} + struct xfrm_tmpl { /* id in template is interpreted as: @@ -1253,6 +1275,7 @@ extern int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, extern int xfrm_input_resume(struct sk_buff *skb, int nexthdr); extern int xfrm_output_resume(struct sk_buff *skb, int err); extern int xfrm_output(struct sk_buff *skb); +extern int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb); extern int xfrm4_extract_header(struct sk_buff *skb); extern int xfrm4_extract_input(struct xfrm_state *x, struct sk_buff *skb); extern int xfrm4_rcv_encap(struct sk_buff *skb, int nexthdr, __be32 spi, diff --git a/net/ipv4/xfrm4_mode_tunnel.c b/net/ipv4/xfrm4_mode_tunnel.c index 8dee617..584e6d7 100644 --- a/net/ipv4/xfrm4_mode_tunnel.c +++ b/net/ipv4/xfrm4_mode_tunnel.c @@ -41,7 +41,7 @@ static int xfrm4_mode_tunnel_output(struct xfrm_state *x, struct sk_buff *skb) top_iph->ihl = 5; top_iph->version = 4; - top_iph->protocol = x->inner_mode->afinfo->proto; + top_iph->protocol = xfrm_af2proto(skb->dst->ops->family); /* DS disclosed */ top_iph->tos = INET_ECN_encapsulate(XFRM_MODE_SKB_CB(skb)->tos, diff --git a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c index d5a58a8..8c3180a 100644 --- a/net/ipv4/xfrm4_output.c +++ b/net/ipv4/xfrm4_output.c @@ -56,7 +56,7 @@ int xfrm4_prepare_output(struct xfrm_state *x, struct sk_buff *skb) { int err; - err = x->inner_mode->afinfo->extract_output(x, skb); + err = xfrm_inner_extract_output(x, skb); if (err) return err; diff --git a/net/ipv6/xfrm6_mode_tunnel.c b/net/ipv6/xfrm6_mode_tunnel.c index 0c742fa..e20529b 100644 --- a/net/ipv6/xfrm6_mode_tunnel.c +++ b/net/ipv6/xfrm6_mode_tunnel.c @@ -45,7 +45,7 @@ static int xfrm6_mode_tunnel_output(struct xfrm_state *x, struct sk_buff *skb) memcpy(top_iph->flow_lbl, XFRM_MODE_SKB_CB(skb)->flow_lbl, sizeof(top_iph->flow_lbl)); - top_iph->nexthdr = x->inner_mode->afinfo->proto; + top_iph->nexthdr = xfrm_af2proto(skb->dst->ops->family); dsfield = XFRM_MODE_SKB_CB(skb)->tos; dsfield = INET_ECN_encapsulate(dsfield, dsfield); diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c index 79ccfb0..0af823cf 100644 --- a/net/ipv6/xfrm6_output.c +++ b/net/ipv6/xfrm6_output.c @@ -62,7 +62,7 @@ int xfrm6_prepare_output(struct xfrm_state *x, struct sk_buff *skb) { int err; - err = x->inner_mode->afinfo->extract_output(x, skb); + err = xfrm_inner_extract_output(x, skb); if (err) return err; diff --git a/net/key/af_key.c b/net/key/af_key.c index 8b5f486..e9ef9af 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -1219,7 +1219,7 @@ static struct xfrm_state * pfkey_msg2xfrm_state(struct sadb_msg *hdr, x->sel.prefixlen_s = addr->sadb_address_prefixlen; } - if (!x->sel.family) + if (x->props.mode == XFRM_MODE_TRANSPORT) x->sel.family = x->props.family; if (ext_hdrs[SADB_X_EXT_NAT_T_TYPE-1]) { diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c index 62188c6..7527940 100644 --- a/net/xfrm/xfrm_input.c +++ b/net/xfrm/xfrm_input.c @@ -84,14 +84,21 @@ int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, __be32 *spi, __be32 *seq) int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb) { + struct xfrm_mode *inner_mode = x->inner_mode; int err; err = x->outer_mode->afinfo->extract_input(x, skb); if (err) return err; - skb->protocol = x->inner_mode->afinfo->eth_proto; - return x->inner_mode->input2(x, skb); + if (x->sel.family == AF_UNSPEC) { + inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol); + if (inner_mode == NULL) + return -EAFNOSUPPORT; + } + + skb->protocol = inner_mode->afinfo->eth_proto; + return inner_mode->input2(x, skb); } EXPORT_SYMBOL(xfrm_prepare_input); @@ -101,6 +108,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type) __be32 seq; struct xfrm_state *x; xfrm_address_t *daddr; + struct xfrm_mode *inner_mode; unsigned int family; int decaps = 0; int async = 0; @@ -207,7 +215,15 @@ resume: XFRM_MODE_SKB_CB(skb)->protocol = nexthdr; - if (x->inner_mode->input(x, skb)) { + inner_mode = x->inner_mode; + + if (x->sel.family == AF_UNSPEC) { + inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol); + if (inner_mode == NULL) + goto drop; + } + + if (inner_mode->input(x, skb)) { XFRM_INC_STATS(LINUX_MIB_XFRMINSTATEMODEERROR); goto drop; } diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c index 569d377..2519129 100644 --- a/net/xfrm/xfrm_output.c +++ b/net/xfrm/xfrm_output.c @@ -124,7 +124,7 @@ int xfrm_output_resume(struct sk_buff *skb, int err) if (!x) return dst_output(skb); - err = nf_hook(x->inner_mode->afinfo->family, + err = nf_hook(skb->dst->ops->family, NF_INET_POST_ROUTING, skb, NULL, skb->dst->dev, xfrm_output2); if (unlikely(err != 1)) @@ -193,4 +193,20 @@ int xfrm_output(struct sk_buff *skb) return xfrm_output2(skb); } + +int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb) +{ + struct xfrm_mode *inner_mode; + if (x->sel.family == AF_UNSPEC) + inner_mode = xfrm_ip2inner_mode(x, + xfrm_af2proto(skb->dst->ops->family)); + else + inner_mode = x->inner_mode; + + if (inner_mode == NULL) + return -EAFNOSUPPORT; + return inner_mode->afinfo->extract_output(x, skb); +} + EXPORT_SYMBOL_GPL(xfrm_output); +EXPORT_SYMBOL_GPL(xfrm_inner_extract_output); diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index 7ba65e8..58f1f93 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -388,6 +388,8 @@ static void xfrm_state_gc_destroy(struct xfrm_state *x) kfree(x->coaddr); if (x->inner_mode) xfrm_put_mode(x->inner_mode); + if (x->inner_mode_iaf) + xfrm_put_mode(x->inner_mode_iaf); if (x->outer_mode) xfrm_put_mode(x->outer_mode); if (x->type) { @@ -523,6 +525,8 @@ struct xfrm_state *xfrm_state_alloc(void) x->lft.hard_packet_limit = XFRM_INF; x->replay_maxage = 0; x->replay_maxdiff = 0; + x->inner_mode = NULL; + x->inner_mode_iaf = NULL; spin_lock_init(&x->lock); } return x; @@ -796,7 +800,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr, selector. */ if (x->km.state == XFRM_STATE_VALID) { - if (!xfrm_selector_match(&x->sel, fl, x->sel.family) || + if ((x->sel.family && !xfrm_selector_match(&x->sel, fl, x->sel.family)) || !security_xfrm_state_pol_flow_match(x, pol, fl)) continue; if (!best || @@ -1944,6 +1948,7 @@ int xfrm_state_mtu(struct xfrm_state *x, int mtu) int xfrm_init_state(struct xfrm_state *x) { struct xfrm_state_afinfo *afinfo; + struct xfrm_mode *inner_mode; int family = x->props.family; int err; @@ -1962,13 +1967,48 @@ int xfrm_init_state(struct xfrm_state *x) goto error; err = -EPROTONOSUPPORT; - x->inner_mode = xfrm_get_mode(x->props.mode, x->sel.family); - if (x->inner_mode == NULL) - goto error; - if (!(x->inner_mode->flags & XFRM_MODE_FLAG_TUNNEL) && - family != x->sel.family) - goto error; + if (x->sel.family != AF_UNSPEC) { + inner_mode = xfrm_get_mode(x->props.mode, x->sel.family); + if (inner_mode == NULL) + goto error; + + if (!(inner_mode->flags & XFRM_MODE_FLAG_TUNNEL) && + family != x->sel.family) { + xfrm_put_mode(inner_mode); + goto error; + } + + x->inner_mode = inner_mode; + } else { + struct xfrm_mode *inner_mode_iaf; + + inner_mode = xfrm_get_mode(x->props.mode, AF_INET); + if (inner_mode == NULL) + goto error; + + if (!(inner_mode->flags & XFRM_MODE_FLAG_TUNNEL)) { + xfrm_put_mode(inner_mode); + goto error; + } + + inner_mode_iaf = xfrm_get_mode(x->props.mode, AF_INET6); + if (inner_mode_iaf == NULL) + goto error; + + if (!(inner_mode_iaf->flags & XFRM_MODE_FLAG_TUNNEL)) { + xfrm_put_mode(inner_mode_iaf); + goto error; + } + + if (x->props.family == AF_INET) { + x->inner_mode = inner_mode; + x->inner_mode_iaf = inner_mode_iaf; + } else { + x->inner_mode = inner_mode_iaf; + x->inner_mode_iaf = inner_mode; + } + } x->type = xfrm_get_type(x->id.proto, family); if (x->type == NULL) diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index f971ca5..5d96f27 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -288,12 +288,9 @@ static void copy_from_user_state(struct xfrm_state *x, struct xfrm_usersa_info * memcpy(&x->props.saddr, &p->saddr, sizeof(x->props.saddr)); x->props.flags = p->flags; - /* - * Set inner address family if the KM left it as zero. - * See comment in validate_tmpl. - */ - if (!x->sel.family) + if (x->props.mode == XFRM_MODE_TRANSPORT) x->sel.family = p->family; + } /* -- cgit v0.10.2 From 7c44b6e922052989ca2af62aa4ff9ff30f04e092 Mon Sep 17 00:00:00 2001 From: Pascal Terjan Date: Thu, 13 Mar 2008 19:13:24 +0100 Subject: iwlwifi: fix a typo in Kconfig message Signed-off-by: Pascal Terjan Signed-off-by: Reinette Chatre Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/iwlwifi/Kconfig b/drivers/net/wireless/iwlwifi/Kconfig index d1af938..b79a35a 100644 --- a/drivers/net/wireless/iwlwifi/Kconfig +++ b/drivers/net/wireless/iwlwifi/Kconfig @@ -20,7 +20,7 @@ config IWL4965 runs. If you want to compile the driver as a module ( = code which can be - inserted in and remvoed from the running kernel whenever you want), + inserted in and removed from the running kernel whenever you want), say M here and read . The module will be called iwl4965.ko. @@ -101,7 +101,7 @@ config IWL3945 runs. If you want to compile the driver as a module ( = code which can be - inserted in and remvoed from the running kernel whenever you want), + inserted in and removed from the running kernel whenever you want), say M here and read . The module will be called iwl3945.ko. -- cgit v0.10.2 From 0a74892b6df8f1e7191dffd5a2b0e0e2ca7e73fb Mon Sep 17 00:00:00 2001 From: Masakazu Mokuno Date: Sat, 15 Mar 2008 21:38:29 +0100 Subject: rt2x00: Add id for Corega CG-WLUSB2GPX This adds the id for Corega CG-WLUSB2GPX. Signed-off-by: Masakazu Mokuno Signed-off-by: Ivo van Doorn Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/rt2x00/rt73usb.c b/drivers/net/wireless/rt2x00/rt73usb.c index 8103d41..3909cf4 100644 --- a/drivers/net/wireless/rt2x00/rt73usb.c +++ b/drivers/net/wireless/rt2x00/rt73usb.c @@ -2095,6 +2095,8 @@ static struct usb_device_id rt73usb_device_table[] = { { USB_DEVICE(0x1371, 0x9032), USB_DEVICE_DATA(&rt73usb_ops) }, /* Conceptronic */ { USB_DEVICE(0x14b2, 0x3c22), USB_DEVICE_DATA(&rt73usb_ops) }, + /* Corega */ + { USB_DEVICE(0x07aa, 0x002e), USB_DEVICE_DATA(&rt73usb_ops) }, /* D-Link */ { USB_DEVICE(0x07d1, 0x3c03), USB_DEVICE_DATA(&rt73usb_ops) }, { USB_DEVICE(0x07d1, 0x3c04), USB_DEVICE_DATA(&rt73usb_ops) }, -- cgit v0.10.2 From ea995abfed7f0726aaa22580aaf10b2cf5d91be5 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 16 Mar 2008 22:43:06 +0000 Subject: wavelan_cs arm fix Even when all fields are unsigned char, struct still might have alignment > 1. Does so on arm, unless you explicitly say that it's packed... Signed-off-by: Al Viro Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/wavelan_cs.h b/drivers/net/wireless/wavelan_cs.h index fabc63e..2e4bfe4 100644 --- a/drivers/net/wireless/wavelan_cs.h +++ b/drivers/net/wireless/wavelan_cs.h @@ -309,7 +309,7 @@ struct mmw_t #define MMW_EXT_ANT_INTERNAL 0x00 /* Internal antenna */ #define MMW_EXT_ANT_EXTERNAL 0x03 /* External antenna */ #define MMW_EXT_ANT_IQ_TEST 0x1C /* IQ test pattern (set to 0) */ -}; +} __attribute__((packed)); /* Size for structure checking (if padding is correct) */ #define MMW_SIZE 37 -- cgit v0.10.2 From a9f46786ec5bb291c9b5e6cc6f83ebbd3fceea51 Mon Sep 17 00:00:00 2001 From: Rick Farrington Date: Tue, 18 Mar 2008 14:57:49 -0700 Subject: iwlwifi: mac start synchronization issue This patch fixes a synchronization problem on the 4965 and 3945 with the mac start callback routine. The problem is that this function exits BEFORE the 'xxx_alive_start' has completed. This can lead to a problem if a subsequent MAC callback attempts to issue a firmware command. Signed-off-by: Rick Farrington Signed-off-by: Reinette Chatre Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c index 40b71bc..39cc13f 100644 --- a/drivers/net/wireless/iwlwifi/iwl3945-base.c +++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c @@ -6206,11 +6206,11 @@ static void iwl3945_alive_start(struct iwl3945_priv *priv) /* At this point, the NIC is initialized and operational */ priv->notif_missed_beacons = 0; - set_bit(STATUS_READY, &priv->status); iwl3945_reg_txpower_periodic(priv); IWL_DEBUG_INFO("ALIVE processing complete.\n"); + set_bit(STATUS_READY, &priv->status); wake_up_interruptible(&priv->wait_command_queue); if (priv->error_recovering) diff --git a/drivers/net/wireless/iwlwifi/iwl4965-base.c b/drivers/net/wireless/iwlwifi/iwl4965-base.c index a23d479..7020822 100644 --- a/drivers/net/wireless/iwlwifi/iwl4965-base.c +++ b/drivers/net/wireless/iwlwifi/iwl4965-base.c @@ -6628,11 +6628,11 @@ static void iwl4965_alive_start(struct iwl4965_priv *priv) /* At this point, the NIC is initialized and operational */ priv->notif_missed_beacons = 0; - set_bit(STATUS_READY, &priv->status); iwl4965_rf_kill_ct_config(priv); IWL_DEBUG_INFO("ALIVE processing complete.\n"); + set_bit(STATUS_READY, &priv->status); wake_up_interruptible(&priv->wait_command_queue); if (priv->error_recovering) -- cgit v0.10.2 From 9fe0a8c838f0584bca51bcc2f81e5c2c933a5880 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 20 Mar 2008 16:19:04 -0400 Subject: arlan: fix warning when PROC_FS=n drivers/net/wireless/arlan-proc.c:1216: warning: 'arlan_root_table' defined but not used Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/arlan-proc.c b/drivers/net/wireless/arlan-proc.c index c6e70db..2ab1d59 100644 --- a/drivers/net/wireless/arlan-proc.c +++ b/drivers/net/wireless/arlan-proc.c @@ -1202,13 +1202,6 @@ static ctl_table arlan_table[MAX_ARLANS + 1] = { .ctl_name = 0 } }; #endif -#else - -static ctl_table arlan_table[MAX_ARLANS + 1] = -{ - { .ctl_name = 0 } -}; -#endif // static int mmtu = 1234; @@ -1233,7 +1226,6 @@ static ctl_table arlan_root_table[] = //}; -#ifdef CONFIG_PROC_FS static struct ctl_table_header *arlan_device_sysctl_header; int __init init_arlan_proc(void) -- cgit v0.10.2 From c2c25fb8144e9e77910294ee76a888e5c8477793 Mon Sep 17 00:00:00 2001 From: Reinette Chatre Date: Fri, 21 Mar 2008 13:53:40 -0700 Subject: MAINTAINERS: update iwlwifi git url Signed-off-by: Reinette Chatre Signed-off-by: John W. Linville diff --git a/MAINTAINERS b/MAINTAINERS index 76843d0..f148fa6 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2110,7 +2110,7 @@ M: reinette.chatre@intel.com L: linux-wireless@vger.kernel.org L: ipw3945-devel@lists.sourceforge.net W: http://intellinuxwireless.org -T: git git://intellinuxwireless.org/repos/iwlwifi +T: git git://git.kernel.org/pub/scm/linux/kernel/git/rchatre/iwlwifi-2.6.git S: Supported IOC3 ETHERNET DRIVER -- cgit v0.10.2 From c83dbf687f1e0f6eccb96cd08438285e280a3876 Mon Sep 17 00:00:00 2001 From: Reinette Chatre Date: Fri, 21 Mar 2008 13:53:41 -0700 Subject: iwlwifi: fix __devexit_p points to __devexit functions The iwlxxxx_pci_remove functions are not needed when drivers are not compiled as modules - they can thus be discarded at kernel link time. This is already captured by having them as __devexit_p in the pci_driver struct - these are supposed to be pointers to __devexit functions, but was not. This is now fixed. This problem was reported by Toralf Forster when testing the compilation of 2.6.25-rc6. Signed-off-by: Reinette Chatre CC: Toralf Forster Signed-off-by: Tomas Winkler Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c index 39cc13f..cbaeaf1 100644 --- a/drivers/net/wireless/iwlwifi/iwl3945-base.c +++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c @@ -8706,7 +8706,7 @@ static int iwl3945_pci_probe(struct pci_dev *pdev, const struct pci_device_id *e return err; } -static void iwl3945_pci_remove(struct pci_dev *pdev) +static void __devexit iwl3945_pci_remove(struct pci_dev *pdev) { struct iwl3945_priv *priv = pci_get_drvdata(pdev); struct list_head *p, *q; diff --git a/drivers/net/wireless/iwlwifi/iwl4965-base.c b/drivers/net/wireless/iwlwifi/iwl4965-base.c index 7020822..60ec29e 100644 --- a/drivers/net/wireless/iwlwifi/iwl4965-base.c +++ b/drivers/net/wireless/iwlwifi/iwl4965-base.c @@ -9282,7 +9282,7 @@ static int iwl4965_pci_probe(struct pci_dev *pdev, const struct pci_device_id *e return err; } -static void iwl4965_pci_remove(struct pci_dev *pdev) +static void __devexit iwl4965_pci_remove(struct pci_dev *pdev) { struct iwl4965_priv *priv = pci_get_drvdata(pdev); struct list_head *p, *q; -- cgit v0.10.2 From ffa9256aaa28af24ea9d2a7d81b3b93b571226a3 Mon Sep 17 00:00:00 2001 From: Michael Buesch Date: Sat, 22 Mar 2008 22:04:45 +0100 Subject: b43: Fix DMA mapping leakage This fixes a DMA mapping leakage in the case where we reject a DMA buffer because of its address. Signed-off-by: Michael Buesch Cc: Christian Casteyde Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/b43/dma.c b/drivers/net/wireless/b43/dma.c index 3dfb28a..766d955 100644 --- a/drivers/net/wireless/b43/dma.c +++ b/drivers/net/wireless/b43/dma.c @@ -560,7 +560,7 @@ static int b43_dmacontroller_tx_reset(struct b43_wldev *dev, u16 mmio_base, /* Check if a DMA mapping address is invalid. */ static bool b43_dma_mapping_error(struct b43_dmaring *ring, dma_addr_t addr, - size_t buffersize) + size_t buffersize, bool dma_to_device) { if (unlikely(dma_mapping_error(addr))) return 1; @@ -568,11 +568,11 @@ static bool b43_dma_mapping_error(struct b43_dmaring *ring, switch (ring->type) { case B43_DMA_30BIT: if ((u64)addr + buffersize > (1ULL << 30)) - return 1; + goto address_error; break; case B43_DMA_32BIT: if ((u64)addr + buffersize > (1ULL << 32)) - return 1; + goto address_error; break; case B43_DMA_64BIT: /* Currently we can't have addresses beyond @@ -582,6 +582,12 @@ static bool b43_dma_mapping_error(struct b43_dmaring *ring, /* The address is OK. */ return 0; + +address_error: + /* We can't support this address. Unmap it again. */ + unmap_descbuffer(ring, addr, buffersize, dma_to_device); + + return 1; } static int setup_rx_descbuffer(struct b43_dmaring *ring, @@ -599,7 +605,7 @@ static int setup_rx_descbuffer(struct b43_dmaring *ring, if (unlikely(!skb)) return -ENOMEM; dmaaddr = map_descbuffer(ring, skb->data, ring->rx_buffersize, 0); - if (b43_dma_mapping_error(ring, dmaaddr, ring->rx_buffersize)) { + if (b43_dma_mapping_error(ring, dmaaddr, ring->rx_buffersize, 0)) { /* ugh. try to realloc in zone_dma */ gfp_flags |= GFP_DMA; @@ -612,7 +618,7 @@ static int setup_rx_descbuffer(struct b43_dmaring *ring, ring->rx_buffersize, 0); } - if (b43_dma_mapping_error(ring, dmaaddr, ring->rx_buffersize)) { + if (b43_dma_mapping_error(ring, dmaaddr, ring->rx_buffersize, 0)) { dev_kfree_skb_any(skb); return -EIO; } @@ -852,7 +858,8 @@ struct b43_dmaring *b43_setup_dmaring(struct b43_wldev *dev, b43_txhdr_size(dev), DMA_TO_DEVICE); - if (b43_dma_mapping_error(ring, dma_test, b43_txhdr_size(dev))) { + if (b43_dma_mapping_error(ring, dma_test, + b43_txhdr_size(dev), 1)) { /* ugh realloc */ kfree(ring->txhdr_cache); ring->txhdr_cache = kcalloc(nr_slots, @@ -867,7 +874,7 @@ struct b43_dmaring *b43_setup_dmaring(struct b43_wldev *dev, DMA_TO_DEVICE); if (b43_dma_mapping_error(ring, dma_test, - b43_txhdr_size(dev))) + b43_txhdr_size(dev), 1)) goto err_kfree_txhdr_cache; } @@ -1189,7 +1196,7 @@ static int dma_tx_fragment(struct b43_dmaring *ring, meta_hdr->dmaaddr = map_descbuffer(ring, (unsigned char *)header, hdrsize, 1); - if (b43_dma_mapping_error(ring, meta_hdr->dmaaddr, hdrsize)) { + if (b43_dma_mapping_error(ring, meta_hdr->dmaaddr, hdrsize, 1)) { ring->current_slot = old_top_slot; ring->used_slots = old_used_slots; return -EIO; @@ -1208,7 +1215,7 @@ static int dma_tx_fragment(struct b43_dmaring *ring, meta->dmaaddr = map_descbuffer(ring, skb->data, skb->len, 1); /* create a bounce buffer in zone_dma on mapping failure. */ - if (b43_dma_mapping_error(ring, meta->dmaaddr, skb->len)) { + if (b43_dma_mapping_error(ring, meta->dmaaddr, skb->len, 1)) { bounce_skb = __dev_alloc_skb(skb->len, GFP_ATOMIC | GFP_DMA); if (!bounce_skb) { ring->current_slot = old_top_slot; @@ -1222,7 +1229,7 @@ static int dma_tx_fragment(struct b43_dmaring *ring, skb = bounce_skb; meta->skb = skb; meta->dmaaddr = map_descbuffer(ring, skb->data, skb->len, 1); - if (b43_dma_mapping_error(ring, meta->dmaaddr, skb->len)) { + if (b43_dma_mapping_error(ring, meta->dmaaddr, skb->len, 1)) { ring->current_slot = old_top_slot; ring->used_slots = old_used_slots; err = -EIO; -- cgit v0.10.2 From 7a193a5df7580957c918269e4ba2fccbe4141cb4 Mon Sep 17 00:00:00 2001 From: Michael Buesch Date: Sun, 23 Mar 2008 01:08:22 +0100 Subject: b43: Remove irqs_disabled() sanity checks Remove all irqs_disabled() sanity checks, as they are not safe on a RT-enabled kernel and will trigger bogus warnings. Signed-off-by: Michael Buesch Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/b43/debugfs.c b/drivers/net/wireless/b43/debugfs.c index e38ed0f..7fca2eb 100644 --- a/drivers/net/wireless/b43/debugfs.c +++ b/drivers/net/wireless/b43/debugfs.c @@ -618,6 +618,7 @@ void b43_debugfs_remove_device(struct b43_wldev *dev) kfree(e); } +/* Called with IRQs disabled. */ void b43_debugfs_log_txstat(struct b43_wldev *dev, const struct b43_txstatus *status) { @@ -629,8 +630,7 @@ void b43_debugfs_log_txstat(struct b43_wldev *dev, if (!e) return; log = &e->txstatlog; - B43_WARN_ON(!irqs_disabled()); - spin_lock(&log->lock); + spin_lock(&log->lock); /* IRQs are already disabled. */ i = log->end + 1; if (i == B43_NR_LOGGED_TXSTATUS) i = 0; diff --git a/drivers/net/wireless/b43/dma.c b/drivers/net/wireless/b43/dma.c index 766d955..cfbc1a2 100644 --- a/drivers/net/wireless/b43/dma.c +++ b/drivers/net/wireless/b43/dma.c @@ -1344,6 +1344,7 @@ out_unlock: return err; } +/* Called with IRQs disabled. */ void b43_dma_handle_txstatus(struct b43_wldev *dev, const struct b43_txstatus *status) { @@ -1356,8 +1357,8 @@ void b43_dma_handle_txstatus(struct b43_wldev *dev, ring = parse_cookie(dev, status->cookie, &slot); if (unlikely(!ring)) return; - B43_WARN_ON(!irqs_disabled()); - spin_lock(&ring->lock); + + spin_lock(&ring->lock); /* IRQs are already disabled. */ B43_WARN_ON(!ring->tx); ops = ring->ops; diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index 51dfce1..c73a75b 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -2049,7 +2049,6 @@ void b43_mac_enable(struct b43_wldev *dev) { dev->mac_suspended--; B43_WARN_ON(dev->mac_suspended < 0); - B43_WARN_ON(irqs_disabled()); if (dev->mac_suspended == 0) { b43_write32(dev, B43_MMIO_MACCTL, b43_read32(dev, B43_MMIO_MACCTL) @@ -2075,7 +2074,6 @@ void b43_mac_suspend(struct b43_wldev *dev) u32 tmp; might_sleep(); - B43_WARN_ON(irqs_disabled()); B43_WARN_ON(dev->mac_suspended < 0); if (dev->mac_suspended == 0) { -- cgit v0.10.2 From 26c080bf8308449330037f91daa3ac0a7c41023e Mon Sep 17 00:00:00 2001 From: Jay Schulist Date: Tue, 18 Mar 2008 17:53:09 -0400 Subject: smctr.c: fix logical-bitwise-or confusion This patch to drivers/net/tokenring/smctr.c fixes a "bitwise vs logical" or error. Signed-off-by: Jay Schulist Signed-off-by: Jeff Garzik diff --git a/drivers/net/tokenring/smctr.c b/drivers/net/tokenring/smctr.c index 8909050..5f1c507 100644 --- a/drivers/net/tokenring/smctr.c +++ b/drivers/net/tokenring/smctr.c @@ -3413,7 +3413,7 @@ static int smctr_make_tx_status_code(struct net_device *dev, tsv->svi = TRANSMIT_STATUS_CODE; tsv->svl = S_TRANSMIT_STATUS_CODE; - tsv->svv[0] = ((tx_fstatus & 0x0100 >> 6) || IBM_PASS_SOURCE_ADDR); + tsv->svv[0] = ((tx_fstatus & 0x0100 >> 6) | IBM_PASS_SOURCE_ADDR); /* Stripped frame status of Transmitted Frame */ tsv->svv[1] = tx_fstatus & 0xff; -- cgit v0.10.2 From 2bf86b7aa8e74bf81a9872f7b610f49b610a4649 Mon Sep 17 00:00:00 2001 From: Jay Vosburgh Date: Fri, 21 Mar 2008 22:29:33 -0700 Subject: bonding: Fix locking in 802.3ad mode The 802.3ad state machine lock can be acquired in both softirq and not softirq context, but was not held at _bh to prevent a deadlock (which could occur if a LACPDU arrived and was processed while the lock was held). Corrected this, now hold the state machine lock at _bh to prevent deadlock. Bug reported by Todd Fleisher . Signed-off-by: Jay Vosburgh Signed-off-by: Jeff Garzik diff --git a/drivers/net/bonding/bond_3ad.c b/drivers/net/bonding/bond_3ad.c index cb3c6fa..d16e0e1 100644 --- a/drivers/net/bonding/bond_3ad.c +++ b/drivers/net/bonding/bond_3ad.c @@ -310,7 +310,7 @@ static inline int __check_agg_selection_timer(struct port *port) */ static inline void __get_rx_machine_lock(struct port *port) { - spin_lock(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); + spin_lock_bh(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); } /** @@ -320,7 +320,7 @@ static inline void __get_rx_machine_lock(struct port *port) */ static inline void __release_rx_machine_lock(struct port *port) { - spin_unlock(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); + spin_unlock_bh(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); } /** -- cgit v0.10.2 From 966bc6f434df4a02108d01dda8cd52951fe853da Mon Sep 17 00:00:00 2001 From: Jay Vosburgh Date: Fri, 21 Mar 2008 22:29:34 -0700 Subject: bonding: fix two compiler warnings Fix two compiler warnings that are new with recent versions of gcc (apparently 4.2 and up). One is fixed by refactoring; this change was supplied by Stephen Hemminger. The other was fixed by labelling the variable as uninitialized_var() after confirming via inspection that it cannot actually be used uninitialized. Signed-off-by: Jay Vosburgh Signed-off-by: Stephen Hemminger Signed-off-by: Jeff Garzik diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index b57bc94..3f58c3d 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -678,12 +678,8 @@ static struct slave *rlb_choose_channel(struct sk_buff *skb, struct bonding *bon } if (!list_empty(&bond->vlan_list)) { - unsigned short vlan_id; - int res = vlan_get_tag(skb, &vlan_id); - if (!res) { + if (!vlan_get_tag(skb, &client_info->vlan_id)) client_info->tag = 1; - client_info->vlan_id = vlan_id; - } } if (!client_info->assigned) { diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 0942d82..2056a87 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -383,7 +383,7 @@ struct vlan_entry *bond_next_vlan(struct bonding *bond, struct vlan_entry *curr) */ int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb, struct net_device *slave_dev) { - unsigned short vlan_id; + unsigned short uninitialized_var(vlan_id); if (!list_empty(&bond->vlan_list) && !(slave_dev->features & NETIF_F_HW_VLAN_TX) && -- cgit v0.10.2 From 92b41daa45a505268b11de9b7cbde2c13c0223b5 Mon Sep 17 00:00:00 2001 From: Libor Pechacek Date: Fri, 21 Mar 2008 22:29:35 -0700 Subject: bonding: Fix sysfs attribute handling For bonding interfaces any attempt to read the sysfs directory contents after module removal results in an oops. The fix is to release sysfs attributes for the interfaces upon module unload. Signed-off-by: Libor Pechacek Acked-by: Jay Vosburgh Signed-off-by: Jeff Garzik diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 2056a87..0f06753 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -4528,8 +4528,7 @@ static void bond_free_all(void) netif_tx_unlock_bh(bond_dev); /* Release the bonded slaves */ bond_release_all(bond_dev); - bond_deinit(bond_dev); - unregister_netdevice(bond_dev); + bond_destroy(bond); } #ifdef CONFIG_PROC_FS -- cgit v0.10.2 From 2cfb8b71ccb3ebca76cff48241527eba39226883 Mon Sep 17 00:00:00 2001 From: Jay Vosburgh Date: Fri, 21 Mar 2008 22:29:36 -0700 Subject: bonding: update version Update version to 3.2.5. Signed-off-by: Jay Vosburgh Signed-off-by: Jeff Garzik diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h index 67ccad6..a3c74e2 100644 --- a/drivers/net/bonding/bonding.h +++ b/drivers/net/bonding/bonding.h @@ -22,8 +22,8 @@ #include "bond_3ad.h" #include "bond_alb.h" -#define DRV_VERSION "3.2.4" -#define DRV_RELDATE "January 28, 2008" +#define DRV_VERSION "3.2.5" +#define DRV_RELDATE "March 21, 2008" #define DRV_NAME "bonding" #define DRV_DESCRIPTION "Ethernet Channel Bonding Driver" -- cgit v0.10.2 From 7deb07b1befc9f09dbdaed66c9613a3d08f1c061 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 16 Mar 2008 22:43:06 +0000 Subject: igb: endianness fix le16_to_cpu() should be done before mask and shift... Signed-off-by: Al Viro Signed-off-by: Jeff Garzik diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c index 928ce82..9f3168d 100644 --- a/drivers/net/igb/igb_main.c +++ b/drivers/net/igb/igb_main.c @@ -3452,8 +3452,8 @@ static bool igb_clean_rx_irq_adv(struct igb_adapter *adapter, * that case, it fills the header buffer and spills the rest * into the page. */ - hlen = le16_to_cpu((rx_desc->wb.lower.lo_dword.hdr_info & - E1000_RXDADV_HDRBUFLEN_MASK) >> E1000_RXDADV_HDRBUFLEN_SHIFT); + hlen = (le16_to_cpu(rx_desc->wb.lower.lo_dword.hdr_info) & + E1000_RXDADV_HDRBUFLEN_MASK) >> E1000_RXDADV_HDRBUFLEN_SHIFT; if (hlen > adapter->rx_ps_hdr_size) hlen = adapter->rx_ps_hdr_size; -- cgit v0.10.2 From 6d8126f988556d593f13be32cbe60dacf19c2d2c Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 16 Mar 2008 22:23:24 +0000 Subject: igb trivial annotations Signed-off-by: Al Viro Signed-off-by: Jeff Garzik diff --git a/drivers/net/igb/e1000_82575.h b/drivers/net/igb/e1000_82575.h index 6604d96..76ea846 100644 --- a/drivers/net/igb/e1000_82575.h +++ b/drivers/net/igb/e1000_82575.h @@ -61,28 +61,28 @@ /* Receive Descriptor - Advanced */ union e1000_adv_rx_desc { struct { - u64 pkt_addr; /* Packet buffer address */ - u64 hdr_addr; /* Header buffer address */ + __le64 pkt_addr; /* Packet buffer address */ + __le64 hdr_addr; /* Header buffer address */ } read; struct { struct { struct { - u16 pkt_info; /* RSS type, Packet type */ - u16 hdr_info; /* Split Header, - * header buffer length */ + __le16 pkt_info; /* RSS type, Packet type */ + __le16 hdr_info; /* Split Header, + * header buffer length */ } lo_dword; union { - u32 rss; /* RSS Hash */ + __le32 rss; /* RSS Hash */ struct { - u16 ip_id; /* IP id */ - u16 csum; /* Packet Checksum */ + __le16 ip_id; /* IP id */ + __le16 csum; /* Packet Checksum */ } csum_ip; } hi_dword; } lower; struct { - u32 status_error; /* ext status/error */ - u16 length; /* Packet length */ - u16 vlan; /* VLAN tag */ + __le32 status_error; /* ext status/error */ + __le16 length; /* Packet length */ + __le16 vlan; /* VLAN tag */ } upper; } wb; /* writeback */ }; @@ -97,14 +97,14 @@ union e1000_adv_rx_desc { /* Transmit Descriptor - Advanced */ union e1000_adv_tx_desc { struct { - u64 buffer_addr; /* Address of descriptor's data buf */ - u32 cmd_type_len; - u32 olinfo_status; + __le64 buffer_addr; /* Address of descriptor's data buf */ + __le32 cmd_type_len; + __le32 olinfo_status; } read; struct { - u64 rsvd; /* Reserved */ - u32 nxtseq_seed; - u32 status; + __le64 rsvd; /* Reserved */ + __le32 nxtseq_seed; + __le32 status; } wb; }; @@ -119,10 +119,10 @@ union e1000_adv_tx_desc { /* Context descriptors */ struct e1000_adv_tx_context_desc { - u32 vlan_macip_lens; - u32 seqnum_seed; - u32 type_tucmd_mlhl; - u32 mss_l4len_idx; + __le32 vlan_macip_lens; + __le32 seqnum_seed; + __le32 type_tucmd_mlhl; + __le32 mss_l4len_idx; }; #define E1000_ADVTXD_MACLEN_SHIFT 9 /* Adv ctxt desc mac len shift */ diff --git a/drivers/net/igb/e1000_hw.h b/drivers/net/igb/e1000_hw.h index 161fb68..7b2c70a 100644 --- a/drivers/net/igb/e1000_hw.h +++ b/drivers/net/igb/e1000_hw.h @@ -143,35 +143,35 @@ enum e1000_fc_type { /* Receive Descriptor */ struct e1000_rx_desc { - u64 buffer_addr; /* Address of the descriptor's data buffer */ - u16 length; /* Length of data DMAed into data buffer */ - u16 csum; /* Packet checksum */ + __le64 buffer_addr; /* Address of the descriptor's data buffer */ + __le16 length; /* Length of data DMAed into data buffer */ + __le16 csum; /* Packet checksum */ u8 status; /* Descriptor status */ u8 errors; /* Descriptor Errors */ - u16 special; + __le16 special; }; /* Receive Descriptor - Extended */ union e1000_rx_desc_extended { struct { - u64 buffer_addr; - u64 reserved; + __le64 buffer_addr; + __le64 reserved; } read; struct { struct { - u32 mrq; /* Multiple Rx Queues */ + __le32 mrq; /* Multiple Rx Queues */ union { - u32 rss; /* RSS Hash */ + __le32 rss; /* RSS Hash */ struct { - u16 ip_id; /* IP id */ - u16 csum; /* Packet Checksum */ + __le16 ip_id; /* IP id */ + __le16 csum; /* Packet Checksum */ } csum_ip; } hi_dword; } lower; struct { - u32 status_error; /* ext status/error */ - u16 length; - u16 vlan; /* VLAN tag */ + __le32 status_error; /* ext status/error */ + __le16 length; + __le16 vlan; /* VLAN tag */ } upper; } wb; /* writeback */ }; @@ -181,49 +181,49 @@ union e1000_rx_desc_extended { union e1000_rx_desc_packet_split { struct { /* one buffer for protocol header(s), three data buffers */ - u64 buffer_addr[MAX_PS_BUFFERS]; + __le64 buffer_addr[MAX_PS_BUFFERS]; } read; struct { struct { - u32 mrq; /* Multiple Rx Queues */ + __le32 mrq; /* Multiple Rx Queues */ union { - u32 rss; /* RSS Hash */ + __le32 rss; /* RSS Hash */ struct { - u16 ip_id; /* IP id */ - u16 csum; /* Packet Checksum */ + __le16 ip_id; /* IP id */ + __le16 csum; /* Packet Checksum */ } csum_ip; } hi_dword; } lower; struct { - u32 status_error; /* ext status/error */ - u16 length0; /* length of buffer 0 */ - u16 vlan; /* VLAN tag */ + __le32 status_error; /* ext status/error */ + __le16 length0; /* length of buffer 0 */ + __le16 vlan; /* VLAN tag */ } middle; struct { - u16 header_status; - u16 length[3]; /* length of buffers 1-3 */ + __le16 header_status; + __le16 length[3]; /* length of buffers 1-3 */ } upper; - u64 reserved; + __le64 reserved; } wb; /* writeback */ }; /* Transmit Descriptor */ struct e1000_tx_desc { - u64 buffer_addr; /* Address of the descriptor's data buffer */ + __le64 buffer_addr; /* Address of the descriptor's data buffer */ union { - u32 data; + __le32 data; struct { - u16 length; /* Data buffer length */ + __le16 length; /* Data buffer length */ u8 cso; /* Checksum offset */ u8 cmd; /* Descriptor control */ } flags; } lower; union { - u32 data; + __le32 data; struct { u8 status; /* Descriptor status */ u8 css; /* Checksum start */ - u16 special; + __le16 special; } fields; } upper; }; @@ -231,49 +231,49 @@ struct e1000_tx_desc { /* Offload Context Descriptor */ struct e1000_context_desc { union { - u32 ip_config; + __le32 ip_config; struct { u8 ipcss; /* IP checksum start */ u8 ipcso; /* IP checksum offset */ - u16 ipcse; /* IP checksum end */ + __le16 ipcse; /* IP checksum end */ } ip_fields; } lower_setup; union { - u32 tcp_config; + __le32 tcp_config; struct { u8 tucss; /* TCP checksum start */ u8 tucso; /* TCP checksum offset */ - u16 tucse; /* TCP checksum end */ + __le16 tucse; /* TCP checksum end */ } tcp_fields; } upper_setup; - u32 cmd_and_length; + __le32 cmd_and_length; union { - u32 data; + __le32 data; struct { u8 status; /* Descriptor status */ u8 hdr_len; /* Header length */ - u16 mss; /* Maximum segment size */ + __le16 mss; /* Maximum segment size */ } fields; } tcp_seg_setup; }; /* Offload data descriptor */ struct e1000_data_desc { - u64 buffer_addr; /* Address of the descriptor's buffer address */ + __le64 buffer_addr; /* Address of the descriptor's buffer address */ union { - u32 data; + __le32 data; struct { - u16 length; /* Data buffer length */ + __le16 length; /* Data buffer length */ u8 typ_len_ext; u8 cmd; } flags; } lower; union { - u32 data; + __le32 data; struct { u8 status; /* Descriptor status */ u8 popts; /* Packet Options */ - u16 special; + __le16 special; } fields; } upper; }; diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c index 9f3168d..aaee02e 100644 --- a/drivers/net/igb/igb_main.c +++ b/drivers/net/igb/igb_main.c @@ -3254,6 +3254,13 @@ quit_polling: return 1; } + +static inline u32 get_head(struct igb_ring *tx_ring) +{ + void *end = (struct e1000_tx_desc *)tx_ring->desc + tx_ring->count; + return le32_to_cpu(*(volatile __le32 *)end); +} + /** * igb_clean_tx_irq - Reclaim resources after transmit completes * @adapter: board private structure @@ -3275,9 +3282,7 @@ static bool igb_clean_tx_irq(struct igb_adapter *adapter, unsigned int total_bytes = 0, total_packets = 0; rmb(); - head = *(volatile u32 *)((struct e1000_tx_desc *)tx_ring->desc - + tx_ring->count); - head = le32_to_cpu(head); + head = get_head(tx_ring); i = tx_ring->next_to_clean; while (1) { while (i != head) { @@ -3312,9 +3317,7 @@ static bool igb_clean_tx_irq(struct igb_adapter *adapter, } oldhead = head; rmb(); - head = *(volatile u32 *)((struct e1000_tx_desc *)tx_ring->desc - + tx_ring->count); - head = le32_to_cpu(head); + head = get_head(tx_ring); if (head == oldhead) goto done_cleaning; } /* while (1) */ @@ -3388,7 +3391,7 @@ done_cleaning: * @vlan: descriptor vlan field as written by hardware (no le/be conversion) * @skb: pointer to sk_buff to be indicated to stack **/ -static void igb_receive_skb(struct igb_adapter *adapter, u8 status, u16 vlan, +static void igb_receive_skb(struct igb_adapter *adapter, u8 status, __le16 vlan, struct sk_buff *skb) { if (adapter->vlgrp && (status & E1000_RXD_STAT_VP)) -- cgit v0.10.2 From 1172899a30d0c26df934d63cc98241cd53fe7130 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 16 Mar 2008 22:21:44 +0000 Subject: e100: endianness annotations Signed-off-by: Al Viro Signed-off-by: Jeff Garzik diff --git a/drivers/net/e100.c b/drivers/net/e100.c index cdf3090..2d139ec 100644 --- a/drivers/net/e100.c +++ b/drivers/net/e100.c @@ -960,7 +960,7 @@ static void e100_get_defaults(struct nic *nic) /* Template for a freshly allocated RFD */ nic->blank_rfd.command = 0; - nic->blank_rfd.rbd = 0xFFFFFFFF; + nic->blank_rfd.rbd = cpu_to_le32(0xFFFFFFFF); nic->blank_rfd.size = cpu_to_le16(VLAN_ETH_FRAME_LEN); /* MII setup */ -- cgit v0.10.2 From c7793ace780942831f9b2b2b0518e810ecbd4d6f Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Mon, 17 Mar 2008 10:12:27 -0700 Subject: ixgb: remove unused variable The variable num_group_tail_writes is initialized but never used otherwise. The semantic patch that makes this change is as follows: (http://www.emn.fr/x-info/coccinelle/) // @@ type T; identifier i; constant C; @@ ( extern T i; | - T i; <+... when != i - i = C; ...+> ) // Signed-off-by: Julia Lawall Signed-off-by: Auke Kok Signed-off-by: Jeff Garzik diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c index 269e6f8..6738b4d 100644 --- a/drivers/net/ixgb/ixgb_main.c +++ b/drivers/net/ixgb/ixgb_main.c @@ -2088,14 +2088,12 @@ ixgb_alloc_rx_buffers(struct ixgb_adapter *adapter) struct ixgb_buffer *buffer_info; struct sk_buff *skb; unsigned int i; - int num_group_tail_writes; long cleancount; i = rx_ring->next_to_use; buffer_info = &rx_ring->buffer_info[i]; cleancount = IXGB_DESC_UNUSED(rx_ring); - num_group_tail_writes = IXGB_RX_BUFFER_WRITE; /* leave three descriptors unused */ while(--cleancount > 2) { -- cgit v0.10.2 From 9e6db60825ef7e7999abc610ce256ba768e58162 Mon Sep 17 00:00:00 2001 From: Bryan Wu Date: Mon, 17 Mar 2008 14:18:31 -0700 Subject: smc91x: fix build breakage from the SMC_GET_MAC_ADDR API upgrade Cc: Jeff Garzik Cc: Deepak Saxena Cc: Nicolas Pitre Cc: Russell King Cc: Paul Mundt Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik diff --git a/drivers/net/smc91x.h b/drivers/net/smc91x.h index 51d4134..98a832a 100644 --- a/drivers/net/smc91x.h +++ b/drivers/net/smc91x.h @@ -92,14 +92,14 @@ #define SMC_insw(a, r, p, l) insw ((unsigned long *)((a) + (r)), p, l) # endif /* check if the mac in reg is valid */ -#define SMC_GET_MAC_ADDR(addr) \ +#define SMC_GET_MAC_ADDR(lp, addr) \ do { \ unsigned int __v; \ - __v = SMC_inw(ioaddr, ADDR0_REG); \ + __v = SMC_inw(ioaddr, ADDR0_REG(lp)); \ addr[0] = __v; addr[1] = __v >> 8; \ - __v = SMC_inw(ioaddr, ADDR1_REG); \ + __v = SMC_inw(ioaddr, ADDR1_REG(lp)); \ addr[2] = __v; addr[3] = __v >> 8; \ - __v = SMC_inw(ioaddr, ADDR2_REG); \ + __v = SMC_inw(ioaddr, ADDR2_REG(lp)); \ addr[4] = __v; addr[5] = __v >> 8; \ if (*(u32 *)(&addr[0]) == 0xFFFFFFFF) { \ random_ether_addr(addr); \ -- cgit v0.10.2 From 443be7960be77f3345b44491c700ae4471b0fe57 Mon Sep 17 00:00:00 2001 From: Dhananjay Phadke Date: Mon, 17 Mar 2008 19:59:48 -0700 Subject: netxen: improve msi support Recent netxen firmware has new scheme of generating MSI interrupts, it raises interrupt and blocks itself, waiting for driver to unmask. This reduces chance of spurious interrupts. The driver will be able to deal with older firmware as well. Signed-off-by: Dhananjay Phadke Tested-by: Vernon Mauery Signed-off-by: Jeff Garzik diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h index 2bc5eaa..876cd06 100644 --- a/drivers/net/netxen/netxen_nic.h +++ b/drivers/net/netxen/netxen_nic.h @@ -942,6 +942,7 @@ struct netxen_adapter { struct pci_dev *ctx_desc_pdev; dma_addr_t ctx_desc_phys_addr; int intr_scheme; + int msi_mode; int (*enable_phy_interrupts) (struct netxen_adapter *); int (*disable_phy_interrupts) (struct netxen_adapter *); void (*handle_phy_intr) (struct netxen_adapter *); diff --git a/drivers/net/netxen/netxen_nic_hdr.h b/drivers/net/netxen/netxen_nic_hdr.h index d72f8f8..160f605 100644 --- a/drivers/net/netxen/netxen_nic_hdr.h +++ b/drivers/net/netxen/netxen_nic_hdr.h @@ -456,6 +456,12 @@ enum { #define ISR_INT_MASK_SLOW (NETXEN_PCIX_PS_REG(PCIX_INT_MASK)) #define ISR_INT_TARGET_STATUS (NETXEN_PCIX_PS_REG(PCIX_TARGET_STATUS)) #define ISR_INT_TARGET_MASK (NETXEN_PCIX_PS_REG(PCIX_TARGET_MASK)) +#define ISR_INT_TARGET_STATUS_F1 (NETXEN_PCIX_PS_REG(PCIX_TARGET_STATUS_F1)) +#define ISR_INT_TARGET_MASK_F1 (NETXEN_PCIX_PS_REG(PCIX_TARGET_MASK_F1)) +#define ISR_INT_TARGET_STATUS_F2 (NETXEN_PCIX_PS_REG(PCIX_TARGET_STATUS_F2)) +#define ISR_INT_TARGET_MASK_F2 (NETXEN_PCIX_PS_REG(PCIX_TARGET_MASK_F2)) +#define ISR_INT_TARGET_STATUS_F3 (NETXEN_PCIX_PS_REG(PCIX_TARGET_STATUS_F3)) +#define ISR_INT_TARGET_MASK_F3 (NETXEN_PCIX_PS_REG(PCIX_TARGET_MASK_F3)) #define NETXEN_PCI_MAPSIZE 128 #define NETXEN_PCI_DDR_NET (0x00000000UL) @@ -662,6 +668,12 @@ enum { #define PCIX_TARGET_STATUS (0x10118) #define PCIX_TARGET_MASK (0x10128) +#define PCIX_TARGET_STATUS_F1 (0x10160) +#define PCIX_TARGET_MASK_F1 (0x10170) +#define PCIX_TARGET_STATUS_F2 (0x10164) +#define PCIX_TARGET_MASK_F2 (0x10174) +#define PCIX_TARGET_STATUS_F3 (0x10168) +#define PCIX_TARGET_MASK_F3 (0x10178) #define PCIX_MSI_F0 (0x13000) #define PCIX_MSI_F1 (0x13004) diff --git a/drivers/net/netxen/netxen_nic_hw.c b/drivers/net/netxen/netxen_nic_hw.c index 0135570..05748ca 100644 --- a/drivers/net/netxen/netxen_nic_hw.c +++ b/drivers/net/netxen/netxen_nic_hw.c @@ -398,6 +398,8 @@ int netxen_nic_hw_resources(struct netxen_adapter *adapter) NETXEN_CRB_NORMALIZE(adapter, CRB_NIC_CAPABILITIES_FW)); printk(KERN_NOTICE "%s: FW capabilities:0x%x\n", netxen_nic_driver_name, adapter->intr_scheme); + adapter->msi_mode = readl( + NETXEN_CRB_NORMALIZE(adapter, CRB_NIC_MSI_MODE_FW)); DPRINTK(INFO, "Receive Peg ready too. starting stuff\n"); addr = netxen_alloc(adapter->ahw.pdev, diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c index 9e38bcb..43eb1f6 100644 --- a/drivers/net/netxen/netxen_nic_init.c +++ b/drivers/net/netxen/netxen_nic_init.c @@ -145,6 +145,8 @@ int netxen_init_firmware(struct netxen_adapter *adapter) /* Window 1 call */ writel(INTR_SCHEME_PERPORT, NETXEN_CRB_NORMALIZE(adapter, CRB_NIC_CAPABILITIES_HOST)); + writel(MSI_MODE_MULTIFUNC, + NETXEN_CRB_NORMALIZE(adapter, CRB_NIC_MSI_MODE_HOST)); writel(MPORT_MULTI_FUNCTION_MODE, NETXEN_CRB_NORMALIZE(adapter, CRB_MPORT_MODE)); writel(PHAN_INITIALIZE_ACK, diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index 9737eae..cd665da 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -149,33 +149,31 @@ static void netxen_nic_update_cmd_consumer(struct netxen_adapter *adapter, #define ADAPTER_LIST_SIZE 12 +static uint32_t msi_tgt_status[4] = { + ISR_INT_TARGET_STATUS, ISR_INT_TARGET_STATUS_F1, + ISR_INT_TARGET_STATUS_F2, ISR_INT_TARGET_STATUS_F3 +}; + +static uint32_t sw_int_mask[4] = { + CRB_SW_INT_MASK_0, CRB_SW_INT_MASK_1, + CRB_SW_INT_MASK_2, CRB_SW_INT_MASK_3 +}; + static void netxen_nic_disable_int(struct netxen_adapter *adapter) { - uint32_t mask = 0x7ff; + u32 mask = 0x7ff; int retries = 32; + int port = adapter->portnum; + int pci_fn = adapter->ahw.pci_func; - DPRINTK(1, INFO, "Entered ISR Disable \n"); - - switch (adapter->portnum) { - case 0: - writel(0x0, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_0)); - break; - case 1: - writel(0x0, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_1)); - break; - case 2: - writel(0x0, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_2)); - break; - case 3: - writel(0x0, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_3)); - break; + if (adapter->msi_mode != MSI_MODE_MULTIFUNC) { + writel(0x0, NETXEN_CRB_NORMALIZE(adapter, sw_int_mask[port])); } if (adapter->intr_scheme != -1 && adapter->intr_scheme != INTR_SCHEME_PERPORT) writel(mask,PCI_OFFSET_SECOND_RANGE(adapter, ISR_INT_MASK)); - /* Window = 0 or 1 */ if (!(adapter->flags & NETXEN_NIC_MSI_ENABLED)) { do { writel(0xffffffff, @@ -190,14 +188,18 @@ static void netxen_nic_disable_int(struct netxen_adapter *adapter) printk(KERN_NOTICE "%s: Failed to disable interrupt completely\n", netxen_nic_driver_name); } + } else { + if (adapter->msi_mode == MSI_MODE_MULTIFUNC) { + writel(0xffffffff, PCI_OFFSET_SECOND_RANGE(adapter, + msi_tgt_status[pci_fn])); + } } - - DPRINTK(1, INFO, "Done with Disable Int\n"); } static void netxen_nic_enable_int(struct netxen_adapter *adapter) { u32 mask; + int port = adapter->portnum; DPRINTK(1, INFO, "Entered ISR Enable \n"); @@ -218,20 +220,7 @@ static void netxen_nic_enable_int(struct netxen_adapter *adapter) writel(mask, PCI_OFFSET_SECOND_RANGE(adapter, ISR_INT_MASK)); } - switch (adapter->portnum) { - case 0: - writel(0x1, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_0)); - break; - case 1: - writel(0x1, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_1)); - break; - case 2: - writel(0x1, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_2)); - break; - case 3: - writel(0x1, NETXEN_CRB_NORMALIZE(adapter, CRB_SW_INT_MASK_3)); - break; - } + writel(0x1, NETXEN_CRB_NORMALIZE(adapter, sw_int_mask[port])); if (!(adapter->flags & NETXEN_NIC_MSI_ENABLED)) { mask = 0xbff; @@ -401,6 +390,7 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) /* this will be read from FW later */ adapter->intr_scheme = -1; + adapter->msi_mode = -1; /* This will be reset for mezz cards */ adapter->portnum = pci_func_id; diff --git a/drivers/net/netxen/netxen_nic_phan_reg.h b/drivers/net/netxen/netxen_nic_phan_reg.h index ffa3b72..a566b50 100644 --- a/drivers/net/netxen/netxen_nic_phan_reg.h +++ b/drivers/net/netxen/netxen_nic_phan_reg.h @@ -126,8 +126,11 @@ */ #define CRB_NIC_CAPABILITIES_HOST NETXEN_NIC_REG(0x1a8) #define CRB_NIC_CAPABILITIES_FW NETXEN_NIC_REG(0x1dc) +#define CRB_NIC_MSI_MODE_HOST NETXEN_NIC_REG(0x270) +#define CRB_NIC_MSI_MODE_FW NETXEN_NIC_REG(0x274) #define INTR_SCHEME_PERPORT 0x1 +#define MSI_MODE_MULTIFUNC 0x1 /* used for ethtool tests */ #define CRB_SCRATCHPAD_TEST NETXEN_NIC_REG(0x280) -- cgit v0.10.2 From 05aaa02d799e8e9548d57ac92fcb05e783027341 Mon Sep 17 00:00:00 2001 From: Dhananjay Phadke Date: Mon, 17 Mar 2008 19:59:49 -0700 Subject: netxen: napi and irq cleanup o separate and simpler irq handler for msi interrupts, avoids few checks than legacy mode. o avoid redudant tx_has_work() and rx_has_work() checks in interrupt and napi, which can uncork irq based on racy (lockless) access to tx and rx ring indices. If we get interrupt, there's sufficient reason to schedule napi. o replenish rx ring more often, remove self-imposed threshold rcv_free that prevents posting rx desc to card. This improves performance in low memory. Signed-off-by: Dhananjay Phadke Tested-by: Vernon Mauery Signed-off-by: Jeff Garzik diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h index 876cd06..8b6546c 100644 --- a/drivers/net/netxen/netxen_nic.h +++ b/drivers/net/netxen/netxen_nic.h @@ -842,7 +842,6 @@ struct netxen_rcv_desc_ctx { u32 flags; u32 producer; u32 rcv_pending; /* Num of bufs posted in phantom */ - u32 rcv_free; /* Num of bufs in free list */ dma_addr_t phys_addr; struct pci_dev *phys_pdev; struct rcv_desc *desc_head; /* address of rx ring in Phantom */ @@ -1076,12 +1075,10 @@ void netxen_tso_check(struct netxen_adapter *adapter, struct cmd_desc_type0 *desc, struct sk_buff *skb); int netxen_nic_hw_resources(struct netxen_adapter *adapter); void netxen_nic_clear_stats(struct netxen_adapter *adapter); -int netxen_nic_rx_has_work(struct netxen_adapter *adapter); -int netxen_nic_tx_has_work(struct netxen_adapter *adapter); void netxen_watchdog_task(struct work_struct *work); void netxen_post_rx_buffers(struct netxen_adapter *adapter, u32 ctx, u32 ringid); -int netxen_process_cmd_ring(unsigned long data); +int netxen_process_cmd_ring(struct netxen_adapter *adapter); u32 netxen_process_rcv_ring(struct netxen_adapter *adapter, int ctx, int max); void netxen_nic_set_multi(struct net_device *netdev); int netxen_nic_change_mtu(struct net_device *netdev, int new_mtu); diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c index 43eb1f6..64fc18d 100644 --- a/drivers/net/netxen/netxen_nic_init.c +++ b/drivers/net/netxen/netxen_nic_init.c @@ -185,7 +185,6 @@ void netxen_initialize_adapter_sw(struct netxen_adapter *adapter) for (ring = 0; ring < NUM_RCV_DESC_RINGS; ring++) { struct netxen_rx_buffer *rx_buf; rcv_desc = &adapter->recv_ctx[ctxid].rcv_desc[ring]; - rcv_desc->rcv_free = rcv_desc->max_rx_desc_count; rcv_desc->begin_alloc = 0; rx_buf = rcv_desc->rx_buf_arr; num_rx_bufs = rcv_desc->max_rx_desc_count; @@ -976,28 +975,6 @@ int netxen_phantom_init(struct netxen_adapter *adapter, int pegtune_val) return 0; } -int netxen_nic_rx_has_work(struct netxen_adapter *adapter) -{ - int ctx; - - for (ctx = 0; ctx < MAX_RCV_CTX; ++ctx) { - struct netxen_recv_context *recv_ctx = - &(adapter->recv_ctx[ctx]); - u32 consumer; - struct status_desc *desc_head; - struct status_desc *desc; - - consumer = recv_ctx->status_rx_consumer; - desc_head = recv_ctx->rcv_status_desc_head; - desc = &desc_head[consumer]; - - if (netxen_get_sts_owner(desc) & STATUS_OWNER_HOST) - return 1; - } - - return 0; -} - static int netxen_nic_check_temp(struct netxen_adapter *adapter) { struct net_device *netdev = adapter->netdev; @@ -1040,7 +1017,6 @@ static int netxen_nic_check_temp(struct netxen_adapter *adapter) void netxen_watchdog_task(struct work_struct *work) { - struct net_device *netdev; struct netxen_adapter *adapter = container_of(work, struct netxen_adapter, watchdog_task); @@ -1050,20 +1026,6 @@ void netxen_watchdog_task(struct work_struct *work) if (adapter->handle_phy_intr) adapter->handle_phy_intr(adapter); - netdev = adapter->netdev; - if ((netif_running(netdev)) && !netif_carrier_ok(netdev) && - netxen_nic_link_ok(adapter) ) { - printk(KERN_INFO "%s %s (port %d), Link is up\n", - netxen_nic_driver_name, netdev->name, adapter->portnum); - netif_carrier_on(netdev); - netif_wake_queue(netdev); - } else if(!(netif_running(netdev)) && netif_carrier_ok(netdev)) { - printk(KERN_ERR "%s %s Link is Down\n", - netxen_nic_driver_name, netdev->name); - netif_carrier_off(netdev); - netif_stop_queue(netdev); - } - mod_timer(&adapter->watchdog_timer, jiffies + 2 * HZ); } @@ -1177,7 +1139,6 @@ static void netxen_process_rcv(struct netxen_adapter *adapter, int ctxid, netdev->last_rx = jiffies; - rcv_desc->rcv_free++; rcv_desc->rcv_pending--; /* @@ -1202,13 +1163,6 @@ u32 netxen_process_rcv_ring(struct netxen_adapter *adapter, int ctxid, int max) u32 producer = 0; int count = 0, ring; - DPRINTK(INFO, "procesing receive\n"); - /* - * we assume in this case that there is only one port and that is - * port #1...changes need to be done in firmware to indicate port - * number as part of the descriptor. This way we will be able to get - * the netdev which is associated with that device. - */ while (count < max) { desc = &desc_head[consumer]; if (!(netxen_get_sts_owner(desc) & STATUS_OWNER_HOST)) { @@ -1221,10 +1175,8 @@ u32 netxen_process_rcv_ring(struct netxen_adapter *adapter, int ctxid, int max) consumer = (consumer + 1) & (adapter->max_rx_desc_count - 1); count++; } - if (count) { - for (ring = 0; ring < NUM_RCV_DESC_RINGS; ring++) { - netxen_post_rx_buffers_nodb(adapter, ctxid, ring); - } + for (ring = 0; ring < NUM_RCV_DESC_RINGS; ring++) { + netxen_post_rx_buffers_nodb(adapter, ctxid, ring); } /* update the consumer index in phantom */ @@ -1235,20 +1187,18 @@ u32 netxen_process_rcv_ring(struct netxen_adapter *adapter, int ctxid, int max) /* Window = 1 */ writel(consumer, NETXEN_CRB_NORMALIZE(adapter, - recv_crb_registers[adapter->portnum]. + recv_crb_registers[adapter->portnum]. crb_rcv_status_consumer)); - wmb(); } return count; } /* Process Command status ring */ -int netxen_process_cmd_ring(unsigned long data) +int netxen_process_cmd_ring(struct netxen_adapter *adapter) { u32 last_consumer; u32 consumer; - struct netxen_adapter *adapter = (struct netxen_adapter *)data; int count1 = 0; int count2 = 0; struct netxen_cmd_buffer *buffer; @@ -1435,8 +1385,6 @@ void netxen_post_rx_buffers(struct netxen_adapter *adapter, u32 ctx, u32 ringid) rcv_desc->begin_alloc = index; rcv_desc->rcv_pending += count; rcv_desc->producer = producer; - if (rcv_desc->rcv_free >= 32) { - rcv_desc->rcv_free = 0; /* Window = 1 */ writel((producer - 1) & (rcv_desc->max_rx_desc_count - 1), @@ -1460,8 +1408,6 @@ void netxen_post_rx_buffers(struct netxen_adapter *adapter, u32 ctx, u32 ringid) writel(msg, DB_NORMALIZE(adapter, NETXEN_RCV_PRODUCER_OFFSET)); - wmb(); - } } } @@ -1525,8 +1471,6 @@ static void netxen_post_rx_buffers_nodb(struct netxen_adapter *adapter, rcv_desc->begin_alloc = index; rcv_desc->rcv_pending += count; rcv_desc->producer = producer; - if (rcv_desc->rcv_free >= 32) { - rcv_desc->rcv_free = 0; /* Window = 1 */ writel((producer - 1) & (rcv_desc->max_rx_desc_count - 1), @@ -1536,21 +1480,9 @@ static void netxen_post_rx_buffers_nodb(struct netxen_adapter *adapter, rcv_desc_crb[ringid]. crb_rcv_producer_offset)); wmb(); - } } } -int netxen_nic_tx_has_work(struct netxen_adapter *adapter) -{ - if (find_diff_among(adapter->last_cmd_consumer, - adapter->cmd_producer, - adapter->max_tx_desc_count) > 0) - return 1; - - return 0; -} - - void netxen_nic_clear_stats(struct netxen_adapter *adapter) { memset(&adapter->stats, 0, sizeof(adapter->stats)); diff --git a/drivers/net/netxen/netxen_nic_isr.c b/drivers/net/netxen/netxen_nic_isr.c index 48a404a..1a2333a 100644 --- a/drivers/net/netxen/netxen_nic_isr.c +++ b/drivers/net/netxen/netxen_nic_isr.c @@ -193,14 +193,14 @@ int netxen_nic_link_ok(struct netxen_adapter *adapter) void netxen_nic_xgbe_handle_phy_intr(struct netxen_adapter *adapter) { struct net_device *netdev = adapter->netdev; - u32 val, val1; + u32 val; /* WINDOW = 1 */ val = readl(NETXEN_CRB_NORMALIZE(adapter, CRB_XG_STATE)); val >>= (physical_port[adapter->portnum] * 8); - val1 = val & 0xff; + val &= 0xff; - if (adapter->ahw.xg_linkup == 1 && val1 != XG_LINK_UP) { + if (adapter->ahw.xg_linkup == 1 && val != XG_LINK_UP) { printk(KERN_INFO "%s: %s NIC Link is down\n", netxen_nic_driver_name, netdev->name); adapter->ahw.xg_linkup = 0; @@ -208,16 +208,7 @@ void netxen_nic_xgbe_handle_phy_intr(struct netxen_adapter *adapter) netif_carrier_off(netdev); netif_stop_queue(netdev); } - /* read twice to clear sticky bits */ - /* WINDOW = 0 */ - netxen_nic_read_w0(adapter, NETXEN_NIU_XG_STATUS, &val1); - netxen_nic_read_w0(adapter, NETXEN_NIU_XG_STATUS, &val1); - - if ((val & 0xffb) != 0xffb) { - printk(KERN_INFO "%s ISR: Sync/Align BAD: 0x%08x\n", - netxen_nic_driver_name, val1); - } - } else if (adapter->ahw.xg_linkup == 0 && val1 == XG_LINK_UP) { + } else if (adapter->ahw.xg_linkup == 0 && val == XG_LINK_UP) { printk(KERN_INFO "%s: %s NIC Link is up\n", netxen_nic_driver_name, netdev->name); adapter->ahw.xg_linkup = 1; diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index cd665da..9595520 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -63,12 +63,12 @@ static int netxen_nic_xmit_frame(struct sk_buff *, struct net_device *); static void netxen_tx_timeout(struct net_device *netdev); static void netxen_tx_timeout_task(struct work_struct *work); static void netxen_watchdog(unsigned long); -static int netxen_handle_int(struct netxen_adapter *, struct net_device *); static int netxen_nic_poll(struct napi_struct *napi, int budget); #ifdef CONFIG_NET_POLL_CONTROLLER static void netxen_nic_poll_controller(struct net_device *netdev); #endif static irqreturn_t netxen_intr(int irq, void *data); +static irqreturn_t netxen_msi_intr(int irq, void *data); int physical_port[] = {0, 1, 2, 3}; @@ -405,7 +405,7 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) netdev->set_mac_address = netxen_nic_set_mac; netdev->change_mtu = netxen_nic_change_mtu; netdev->tx_timeout = netxen_tx_timeout; - netdev->watchdog_timeo = HZ; + netdev->watchdog_timeo = 2*HZ; netxen_nic_change_mtu(netdev, netdev->mtu); @@ -823,6 +823,8 @@ static int netxen_nic_open(struct net_device *netdev) struct netxen_adapter *adapter = (struct netxen_adapter *)netdev->priv; int err = 0; int ctx, ring; + irq_handler_t handler; + unsigned long flags = IRQF_SAMPLE_RANDOM; if (adapter->is_up != NETXEN_ADAPTER_UP_MAGIC) { err = netxen_init_firmware(adapter); @@ -846,9 +848,14 @@ static int netxen_nic_open(struct net_device *netdev) netxen_post_rx_buffers(adapter, ctx, ring); } adapter->irq = adapter->ahw.pdev->irq; - err = request_irq(adapter->ahw.pdev->irq, netxen_intr, - IRQF_SHARED|IRQF_SAMPLE_RANDOM, netdev->name, - adapter); + if (adapter->flags & NETXEN_NIC_MSI_ENABLED) + handler = netxen_msi_intr; + else { + flags |= IRQF_SHARED; + handler = netxen_intr; + } + err = request_irq(adapter->irq, handler, + flags, netdev->name, adapter); if (err) { printk(KERN_ERR "request_irq failed with: %d\n", err); netxen_free_hw_resources(adapter); @@ -857,21 +864,12 @@ static int netxen_nic_open(struct net_device *netdev) adapter->is_up = NETXEN_ADAPTER_UP_MAGIC; } - if (!adapter->driver_mismatch) - mod_timer(&adapter->watchdog_timer, jiffies); - - napi_enable(&adapter->napi); - - netxen_nic_enable_int(adapter); - /* Done here again so that even if phantom sw overwrote it, * we set it */ if (adapter->init_port && adapter->init_port(adapter, adapter->portnum) != 0) { - del_timer_sync(&adapter->watchdog_timer); printk(KERN_ERR "%s: Failed to initialize port %d\n", netxen_nic_driver_name, adapter->portnum); - napi_disable(&adapter->napi); return -EIO; } if (adapter->macaddr_set) @@ -884,6 +882,12 @@ static int netxen_nic_open(struct net_device *netdev) adapter->set_mtu(adapter, netdev->mtu); if (!adapter->driver_mismatch) + mod_timer(&adapter->watchdog_timer, jiffies); + + napi_enable(&adapter->napi); + netxen_nic_enable_int(adapter); + + if (!adapter->driver_mismatch) netif_start_queue(netdev); return 0; @@ -1196,81 +1200,50 @@ static void netxen_tx_timeout_task(struct work_struct *work) netif_wake_queue(adapter->netdev); } -static int -netxen_handle_int(struct netxen_adapter *adapter, struct net_device *netdev) +static inline void +netxen_handle_int(struct netxen_adapter *adapter) { - u32 ret = 0; - - DPRINTK(INFO, "Entered handle ISR\n"); - adapter->stats.ints++; - netxen_nic_disable_int(adapter); - - if (netxen_nic_rx_has_work(adapter) || netxen_nic_tx_has_work(adapter)) { - if (netif_rx_schedule_prep(netdev, &adapter->napi)) { - /* - * Interrupts are already disabled. - */ - __netif_rx_schedule(netdev, &adapter->napi); - } else { - static unsigned int intcount = 0; - if ((++intcount & 0xfff) == 0xfff) - DPRINTK(KERN_ERR - "%s: %s interrupt %d while in poll\n", - netxen_nic_driver_name, netdev->name, - intcount); - } - ret = 1; - } - - if (ret == 0) { - netxen_nic_enable_int(adapter); - } - - return ret; + napi_schedule(&adapter->napi); } -/* - * netxen_intr - Interrupt Handler - * @irq: interrupt number - * data points to adapter stucture (which may be handling more than 1 port - */ irqreturn_t netxen_intr(int irq, void *data) { struct netxen_adapter *adapter = data; - struct net_device *netdev = adapter->netdev; u32 our_int = 0; - if (!(adapter->flags & NETXEN_NIC_MSI_ENABLED)) { - our_int = readl(NETXEN_CRB_NORMALIZE(adapter, CRB_INT_VECTOR)); - /* not our interrupt */ - if ((our_int & (0x80 << adapter->portnum)) == 0) - return IRQ_NONE; - } + our_int = readl(NETXEN_CRB_NORMALIZE(adapter, CRB_INT_VECTOR)); + /* not our interrupt */ + if ((our_int & (0x80 << adapter->portnum)) == 0) + return IRQ_NONE; if (adapter->intr_scheme == INTR_SCHEME_PERPORT) { /* claim interrupt */ - if (!(adapter->flags & NETXEN_NIC_MSI_ENABLED)) { - writel(our_int & ~((u32)(0x80 << adapter->portnum)), + writel(our_int & ~((u32)(0x80 << adapter->portnum)), NETXEN_CRB_NORMALIZE(adapter, CRB_INT_VECTOR)); - } } - if (netif_running(netdev)) - netxen_handle_int(adapter, netdev); + netxen_handle_int(adapter); return IRQ_HANDLED; } +irqreturn_t netxen_msi_intr(int irq, void *data) +{ + struct netxen_adapter *adapter = data; + + netxen_handle_int(adapter); + return IRQ_HANDLED; +} + static int netxen_nic_poll(struct napi_struct *napi, int budget) { struct netxen_adapter *adapter = container_of(napi, struct netxen_adapter, napi); - struct net_device *netdev = adapter->netdev; - int done = 1; + int tx_complete; int ctx; int work_done; - DPRINTK(INFO, "polling for %d descriptors\n", *budget); + tx_complete = netxen_process_cmd_ring(adapter); work_done = 0; for (ctx = 0; ctx < MAX_RCV_CTX; ++ctx) { @@ -1290,16 +1263,8 @@ static int netxen_nic_poll(struct napi_struct *napi, int budget) budget / MAX_RCV_CTX); } - if (work_done >= budget) - done = 0; - - if (netxen_process_cmd_ring((unsigned long)adapter) == 0) - done = 0; - - DPRINTK(INFO, "new work_done: %d work_to_do: %d\n", - work_done, work_to_do); - if (done) { - netif_rx_complete(netdev, napi); + if ((work_done < budget) && tx_complete) { + netif_rx_complete(adapter->netdev, &adapter->napi); netxen_nic_enable_int(adapter); } -- cgit v0.10.2 From ba53e6b4878e07411826312c59bfe49561594b6e Mon Sep 17 00:00:00 2001 From: Dhananjay Phadke Date: Mon, 17 Mar 2008 19:59:50 -0700 Subject: netxen: remove low level tx lock o eliminate tx lock in netxen adapter struct, instead pound on netdev tx lock appropriately. o remove old "concurrent transmit" code that unnecessarily drops and reacquires tx lock in hard_xmit_frame(), this is already serialized the netdev xmit lock. o reduce scope of tx lock in tx cleanup. tx cleanup operates on different section of the ring than transmitting cpus and is guarded by producer and consumer indices. This fixes a race caused by rx softirq preemption on realtime kernels. Signed-off-by: Dhananjay Phadke Tested-by: Vernon Mauery Signed-off-by: Jeff Garzik diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h index 8b6546c..070421b 100644 --- a/drivers/net/netxen/netxen_nic.h +++ b/drivers/net/netxen/netxen_nic.h @@ -85,7 +85,7 @@ (sizeof(struct netxen_cmd_buffer) * adapter->max_tx_desc_count) #define RCV_BUFFSIZE \ (sizeof(struct netxen_rx_buffer) * rcv_desc->max_rx_desc_count) -#define find_diff_among(a,b,range) ((a)<=(b)?((b)-(a)):((b)+(range)-(a))) +#define find_diff_among(a,b,range) ((a)<(b)?((b)-(a)):((b)+(range)-(a))) #define NETXEN_NETDEV_STATUS 0x1 #define NETXEN_RCV_PRODUCER_OFFSET 0 @@ -204,7 +204,7 @@ enum { ? RCV_DESC_LRO : \ (RCV_DESC_NORMAL))) -#define MAX_CMD_DESCRIPTORS 1024 +#define MAX_CMD_DESCRIPTORS 4096 #define MAX_RCV_DESCRIPTORS 16384 #define MAX_CMD_DESCRIPTORS_HOST (MAX_CMD_DESCRIPTORS / 4) #define MAX_RCV_DESCRIPTORS_1G (MAX_RCV_DESCRIPTORS / 4) @@ -824,9 +824,7 @@ struct netxen_adapter_stats { u64 uphcong; u64 upmcong; u64 updunno; - u64 skbfreed; u64 txdropped; - u64 txnullskb; u64 csummed; u64 no_rcv; u64 rxbytes; @@ -888,8 +886,6 @@ struct netxen_adapter { int mtu; int portnum; - spinlock_t tx_lock; - spinlock_t lock; struct work_struct watchdog_task; struct timer_list watchdog_timer; struct work_struct tx_timeout_task; @@ -898,16 +894,12 @@ struct netxen_adapter { u32 cmd_producer; __le32 *cmd_consumer; - u32 last_cmd_consumer; + u32 max_tx_desc_count; u32 max_rx_desc_count; u32 max_jumbo_rx_desc_count; u32 max_lro_rx_desc_count; - /* Num of instances active on cmd buffer ring */ - u32 proc_cmd_buf_counter; - - u32 num_threads, total_threads; /*Use to keep track of xmit threads */ u32 flags; u32 irq; diff --git a/drivers/net/netxen/netxen_nic_ethtool.c b/drivers/net/netxen/netxen_nic_ethtool.c index 7a876f4..d324ea3 100644 --- a/drivers/net/netxen/netxen_nic_ethtool.c +++ b/drivers/net/netxen/netxen_nic_ethtool.c @@ -70,9 +70,7 @@ static const struct netxen_nic_stats netxen_nic_gstrings_stats[] = { {"uphcong", NETXEN_NIC_STAT(stats.uphcong)}, {"upmcong", NETXEN_NIC_STAT(stats.upmcong)}, {"updunno", NETXEN_NIC_STAT(stats.updunno)}, - {"skb_freed", NETXEN_NIC_STAT(stats.skbfreed)}, {"tx_dropped", NETXEN_NIC_STAT(stats.txdropped)}, - {"tx_null_skb", NETXEN_NIC_STAT(stats.txnullskb)}, {"csummed", NETXEN_NIC_STAT(stats.csummed)}, {"no_rcv", NETXEN_NIC_STAT(stats.no_rcv)}, {"rx_bytes", NETXEN_NIC_STAT(stats.rxbytes)}, diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c index 64fc18d..fe64618 100644 --- a/drivers/net/netxen/netxen_nic_init.c +++ b/drivers/net/netxen/netxen_nic_init.c @@ -1197,96 +1197,50 @@ u32 netxen_process_rcv_ring(struct netxen_adapter *adapter, int ctxid, int max) /* Process Command status ring */ int netxen_process_cmd_ring(struct netxen_adapter *adapter) { - u32 last_consumer; - u32 consumer; - int count1 = 0; - int count2 = 0; + u32 last_consumer, consumer; + int count = 0, i; struct netxen_cmd_buffer *buffer; - struct pci_dev *pdev; + struct pci_dev *pdev = adapter->pdev; + struct net_device *netdev = adapter->netdev; struct netxen_skb_frag *frag; - u32 i; - int done; + int done = 0; - spin_lock(&adapter->tx_lock); last_consumer = adapter->last_cmd_consumer; - DPRINTK(INFO, "procesing xmit complete\n"); - /* we assume in this case that there is only one port and that is - * port #1...changes need to be done in firmware to indicate port - * number as part of the descriptor. This way we will be able to get - * the netdev which is associated with that device. - */ - consumer = le32_to_cpu(*(adapter->cmd_consumer)); - if (last_consumer == consumer) { /* Ring is empty */ - DPRINTK(INFO, "last_consumer %d == consumer %d\n", - last_consumer, consumer); - spin_unlock(&adapter->tx_lock); - return 1; - } - - adapter->proc_cmd_buf_counter++; - /* - * Not needed - does not seem to be used anywhere. - * adapter->cmd_consumer = consumer; - */ - spin_unlock(&adapter->tx_lock); - while ((last_consumer != consumer) && (count1 < MAX_STATUS_HANDLE)) { + while (last_consumer != consumer) { buffer = &adapter->cmd_buf_arr[last_consumer]; - pdev = adapter->pdev; if (buffer->skb) { frag = &buffer->frag_array[0]; pci_unmap_single(pdev, frag->dma, frag->length, PCI_DMA_TODEVICE); frag->dma = 0ULL; for (i = 1; i < buffer->frag_count; i++) { - DPRINTK(INFO, "getting fragment no %d\n", i); frag++; /* Get the next frag */ pci_unmap_page(pdev, frag->dma, frag->length, PCI_DMA_TODEVICE); frag->dma = 0ULL; } - adapter->stats.skbfreed++; + adapter->stats.xmitfinished++; dev_kfree_skb_any(buffer->skb); buffer->skb = NULL; - } else if (adapter->proc_cmd_buf_counter == 1) { - adapter->stats.txnullskb++; - } - if (unlikely(netif_queue_stopped(adapter->netdev) - && netif_carrier_ok(adapter->netdev)) - && ((jiffies - adapter->netdev->trans_start) > - adapter->netdev->watchdog_timeo)) { - SCHEDULE_WORK(&adapter->tx_timeout_task); } last_consumer = get_next_index(last_consumer, adapter->max_tx_desc_count); - count1++; + if (++count >= MAX_STATUS_HANDLE) + break; } - count2 = 0; - spin_lock(&adapter->tx_lock); - if ((--adapter->proc_cmd_buf_counter) == 0) { + if (count) { adapter->last_cmd_consumer = last_consumer; - while ((adapter->last_cmd_consumer != consumer) - && (count2 < MAX_STATUS_HANDLE)) { - buffer = - &adapter->cmd_buf_arr[adapter->last_cmd_consumer]; - count2++; - if (buffer->skb) - break; - else - adapter->last_cmd_consumer = - get_next_index(adapter->last_cmd_consumer, - adapter->max_tx_desc_count); - } - } - if (count1 || count2) { - if (netif_queue_stopped(adapter->netdev) - && (adapter->flags & NETXEN_NETDEV_STATUS)) { - netif_wake_queue(adapter->netdev); - adapter->flags &= ~NETXEN_NETDEV_STATUS; + smp_mb(); + if (netif_queue_stopped(netdev) && netif_running(netdev)) { + netif_tx_lock(netdev); + netif_wake_queue(netdev); + smp_mb(); + netif_tx_unlock(netdev); } } /* @@ -1302,16 +1256,9 @@ int netxen_process_cmd_ring(struct netxen_adapter *adapter) * There is still a possible race condition and the host could miss an * interrupt. The card has to take care of this. */ - if (adapter->last_cmd_consumer == consumer && - (((adapter->cmd_producer + 1) % - adapter->max_tx_desc_count) == adapter->last_cmd_consumer)) { - consumer = le32_to_cpu(*(adapter->cmd_consumer)); - } - done = (adapter->last_cmd_consumer == consumer); + consumer = le32_to_cpu(*(adapter->cmd_consumer)); + done = (last_consumer == consumer); - spin_unlock(&adapter->tx_lock); - DPRINTK(INFO, "last consumer is %d in %s\n", last_consumer, - __FUNCTION__); return (done); } diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index 9595520..dc4d593 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -317,7 +317,6 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) adapter->ahw.pdev = pdev; adapter->ahw.pci_func = pci_func_id; - spin_lock_init(&adapter->tx_lock); /* remap phys address */ mem_base = pci_resource_start(pdev, 0); /* 0 is for BAR 0 */ @@ -533,7 +532,6 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) adapter->watchdog_timer.data = (unsigned long)adapter; INIT_WORK(&adapter->watchdog_task, netxen_watchdog_task); adapter->ahw.pdev = pdev; - adapter->proc_cmd_buf_counter = 0; adapter->ahw.revision_id = pdev->revision; /* make sure Window == 1 */ @@ -952,41 +950,17 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) struct netxen_skb_frag *buffrag; unsigned int i; - u32 producer = 0; + u32 producer, consumer; u32 saved_producer = 0; struct cmd_desc_type0 *hwdesc; int k; struct netxen_cmd_buffer *pbuf = NULL; - static int dropped_packet = 0; int frag_count; - u32 local_producer = 0; - u32 max_tx_desc_count = 0; - u32 last_cmd_consumer = 0; int no_of_desc; + u32 num_txd = adapter->max_tx_desc_count; - adapter->stats.xmitcalled++; frag_count = skb_shinfo(skb)->nr_frags + 1; - if (unlikely(skb->len <= 0)) { - dev_kfree_skb_any(skb); - adapter->stats.badskblen++; - return NETDEV_TX_OK; - } - - if (frag_count > MAX_BUFFERS_PER_CMD) { - printk("%s: %s netxen_nic_xmit_frame: frag_count (%d) " - "too large, can handle only %d frags\n", - netxen_nic_driver_name, netdev->name, - frag_count, MAX_BUFFERS_PER_CMD); - adapter->stats.txdropped++; - if ((++dropped_packet & 0xff) == 0xff) - printk("%s: %s droppped packets = %d\n", - netxen_nic_driver_name, netdev->name, - dropped_packet); - - return NETDEV_TX_OK; - } - /* There 4 fragments per descriptor */ no_of_desc = (frag_count + 3) >> 2; if (netdev->features & NETIF_F_TSO) { @@ -1001,27 +975,16 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) } } - spin_lock_bh(&adapter->tx_lock); - if (adapter->total_threads >= MAX_XMIT_PRODUCERS) { - goto out_requeue; + producer = adapter->cmd_producer; + smp_mb(); + consumer = adapter->last_cmd_consumer; + if ((no_of_desc+2) > find_diff_among(producer, consumer, num_txd)) { + netif_stop_queue(netdev); + smp_mb(); + return NETDEV_TX_BUSY; } - local_producer = adapter->cmd_producer; - k = adapter->cmd_producer; - max_tx_desc_count = adapter->max_tx_desc_count; - last_cmd_consumer = adapter->last_cmd_consumer; - if ((k + no_of_desc) >= - ((last_cmd_consumer <= k) ? last_cmd_consumer + max_tx_desc_count : - last_cmd_consumer)) { - goto out_requeue; - } - k = get_index_range(k, max_tx_desc_count, no_of_desc); - adapter->cmd_producer = k; - adapter->total_threads++; - adapter->num_threads++; - spin_unlock_bh(&adapter->tx_lock); /* Copy the descriptors into the hardware */ - producer = local_producer; saved_producer = producer; hwdesc = &hw->cmd_desc_head[producer]; memset(hwdesc, 0, sizeof(struct cmd_desc_type0)); @@ -1061,8 +1024,7 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) /* move to next desc. if there is a need */ if ((i & 0x3) == 0) { k = 0; - producer = get_next_index(producer, - adapter->max_tx_desc_count); + producer = get_next_index(producer, num_txd); hwdesc = &hw->cmd_desc_head[producer]; memset(hwdesc, 0, sizeof(struct cmd_desc_type0)); pbuf = &adapter->cmd_buf_arr[producer]; @@ -1080,7 +1042,6 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) buffrag->dma = temp_dma; buffrag->length = temp_len; - DPRINTK(INFO, "for loop. i=%d k=%d\n", i, k); switch (k) { case 0: hwdesc->buffer1_length = cpu_to_le16(temp_len); @@ -1101,7 +1062,7 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) } frag++; } - producer = get_next_index(producer, adapter->max_tx_desc_count); + producer = get_next_index(producer, num_txd); /* might change opcode to TX_TCP_LSO */ netxen_tso_check(adapter, &hw->cmd_desc_head[saved_producer], skb); @@ -1128,7 +1089,7 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) /* copy the first 64 bytes */ memcpy(((void *)hwdesc) + 2, (void *)(skb->data), first_hdr_len); - producer = get_next_index(producer, max_tx_desc_count); + producer = get_next_index(producer, num_txd); if (more_hdr) { hwdesc = &hw->cmd_desc_head[producer]; @@ -1141,35 +1102,19 @@ static int netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) hwdesc, (hdr_len - first_hdr_len)); - producer = get_next_index(producer, max_tx_desc_count); + producer = get_next_index(producer, num_txd); } } - spin_lock_bh(&adapter->tx_lock); + adapter->cmd_producer = producer; adapter->stats.txbytes += skb->len; - /* Code to update the adapter considering how many producer threads - are currently working */ - if ((--adapter->num_threads) == 0) { - /* This is the last thread */ - u32 crb_producer = adapter->cmd_producer; - netxen_nic_update_cmd_producer(adapter, crb_producer); - wmb(); - adapter->total_threads = 0; - } + netxen_nic_update_cmd_producer(adapter, adapter->cmd_producer); - adapter->stats.xmitfinished++; + adapter->stats.xmitcalled++; netdev->trans_start = jiffies; - spin_unlock_bh(&adapter->tx_lock); return NETDEV_TX_OK; - -out_requeue: - netif_stop_queue(netdev); - adapter->flags |= NETXEN_NETDEV_STATUS; - - spin_unlock_bh(&adapter->tx_lock); - return NETDEV_TX_BUSY; } static void netxen_watchdog(unsigned long v) @@ -1194,9 +1139,13 @@ static void netxen_tx_timeout_task(struct work_struct *work) printk(KERN_ERR "%s %s: transmit timeout, resetting.\n", netxen_nic_driver_name, adapter->netdev->name); - netxen_nic_close(adapter->netdev); - netxen_nic_open(adapter->netdev); + netxen_nic_disable_int(adapter); + napi_disable(&adapter->napi); + adapter->netdev->trans_start = jiffies; + + napi_enable(&adapter->napi); + netxen_nic_enable_int(adapter); netif_wake_queue(adapter->netdev); } -- cgit v0.10.2 From d1847a722e79bbfc557755d78f44d3e2c8ae5ea9 Mon Sep 17 00:00:00 2001 From: Dhananjay Phadke Date: Mon, 17 Mar 2008 19:59:51 -0700 Subject: netxen: fix rx dropped stats Don't count rx dropped packets based on return value of netif_receive_skb(), which is misleading. Signed-off-by: Dhananjay Phadke Tested-by: Vernon Mauery Signed-off-by: Jeff Garzik diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h index 070421b..7f20a03 100644 --- a/drivers/net/netxen/netxen_nic.h +++ b/drivers/net/netxen/netxen_nic.h @@ -818,12 +818,7 @@ struct netxen_adapter_stats { u64 badskblen; u64 nocmddescriptor; u64 polled; - u64 uphappy; - u64 updropped; - u64 uplcong; - u64 uphcong; - u64 upmcong; - u64 updunno; + u64 rxdropped; u64 txdropped; u64 csummed; u64 no_rcv; diff --git a/drivers/net/netxen/netxen_nic_ethtool.c b/drivers/net/netxen/netxen_nic_ethtool.c index d324ea3..6e98d83 100644 --- a/drivers/net/netxen/netxen_nic_ethtool.c +++ b/drivers/net/netxen/netxen_nic_ethtool.c @@ -64,12 +64,6 @@ static const struct netxen_nic_stats netxen_nic_gstrings_stats[] = { {"bad_skb_len", NETXEN_NIC_STAT(stats.badskblen)}, {"no_cmd_desc", NETXEN_NIC_STAT(stats.nocmddescriptor)}, {"polled", NETXEN_NIC_STAT(stats.polled)}, - {"uphappy", NETXEN_NIC_STAT(stats.uphappy)}, - {"updropped", NETXEN_NIC_STAT(stats.updropped)}, - {"uplcong", NETXEN_NIC_STAT(stats.uplcong)}, - {"uphcong", NETXEN_NIC_STAT(stats.uphcong)}, - {"upmcong", NETXEN_NIC_STAT(stats.upmcong)}, - {"updunno", NETXEN_NIC_STAT(stats.updunno)}, {"tx_dropped", NETXEN_NIC_STAT(stats.txdropped)}, {"csummed", NETXEN_NIC_STAT(stats.csummed)}, {"no_rcv", NETXEN_NIC_STAT(stats.no_rcv)}, diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c index fe64618..d9713d9 100644 --- a/drivers/net/netxen/netxen_nic_init.c +++ b/drivers/net/netxen/netxen_nic_init.c @@ -1089,7 +1089,7 @@ static void netxen_process_rcv(struct netxen_adapter *adapter, int ctxid, skb = (struct sk_buff *)buffer->skb; if (likely(adapter->rx_csum && - netxen_get_sts_status(sts_data) == STATUS_CKSUM_OK)) { + netxen_get_sts_status(sts_data) == STATUS_CKSUM_OK)) { adapter->stats.csummed++; skb->ip_summed = CHECKSUM_UNNECESSARY; } else @@ -1106,37 +1106,6 @@ static void netxen_process_rcv(struct netxen_adapter *adapter, int ctxid, skb->protocol = eth_type_trans(skb, netdev); ret = netif_receive_skb(skb); - - /* - * RH: Do we need these stats on a regular basis. Can we get it from - * Linux stats. - */ - switch (ret) { - case NET_RX_SUCCESS: - adapter->stats.uphappy++; - break; - - case NET_RX_CN_LOW: - adapter->stats.uplcong++; - break; - - case NET_RX_CN_MOD: - adapter->stats.upmcong++; - break; - - case NET_RX_CN_HIGH: - adapter->stats.uphcong++; - break; - - case NET_RX_DROP: - adapter->stats.updropped++; - break; - - default: - adapter->stats.updunno++; - break; - } - netdev->last_rx = jiffies; rcv_desc->rcv_pending--; diff --git a/drivers/net/netxen/netxen_nic_isr.c b/drivers/net/netxen/netxen_nic_isr.c index 1a2333a..c81313b 100644 --- a/drivers/net/netxen/netxen_nic_isr.c +++ b/drivers/net/netxen/netxen_nic_isr.c @@ -59,7 +59,7 @@ struct net_device_stats *netxen_nic_get_stats(struct net_device *netdev) /* packet transmit problems */ stats->tx_errors = adapter->stats.nocmddescriptor; /* no space in linux buffers */ - stats->rx_dropped = adapter->stats.updropped; + stats->rx_dropped = adapter->stats.rxdropped; /* no space available in linux */ stats->tx_dropped = adapter->stats.txdropped; -- cgit v0.10.2 From be937f1f89cacd833bd242c35af911b971225cf0 Mon Sep 17 00:00:00 2001 From: Alexandr Smirnov Date: Wed, 19 Mar 2008 00:37:24 +0300 Subject: Marvell PHY m88e1111 driver fix Marvell PHY m88e1111 (not sure about other models, but think they too) works in two modes: fiber and copper. In Marvell PHY driver (that we have in current community kernels) code supported only copper mode, and this is not configurable, bits for copper mode are simply written in registers during PHY initialization. This patch adds support for both modes. Signed-off-by: Alexandr Smirnov Acked-by: Andy Fleming Signed-off-by: Jeff Garzik diff --git a/drivers/net/phy/marvell.c b/drivers/net/phy/marvell.c index f057407..b5f4c28 100644 --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@ -58,9 +58,25 @@ #define MII_M1111_RX_DELAY 0x80 #define MII_M1111_TX_DELAY 0x2 #define MII_M1111_PHY_EXT_SR 0x1b -#define MII_M1111_HWCFG_MODE_MASK 0xf -#define MII_M1111_HWCFG_MODE_RGMII 0xb + +#define MII_M1111_HWCFG_MODE_MASK 0xf +#define MII_M1111_HWCFG_MODE_COPPER_RGMII 0xb +#define MII_M1111_HWCFG_MODE_FIBER_RGMII 0x3 #define MII_M1111_HWCFG_MODE_SGMII_NO_CLK 0x4 +#define MII_M1111_HWCFG_FIBER_COPPER_AUTO 0x8000 +#define MII_M1111_HWCFG_FIBER_COPPER_RES 0x2000 + +#define MII_M1111_COPPER 0 +#define MII_M1111_FIBER 1 + +#define MII_M1011_PHY_STATUS 0x11 +#define MII_M1011_PHY_STATUS_1000 0x8000 +#define MII_M1011_PHY_STATUS_100 0x4000 +#define MII_M1011_PHY_STATUS_SPD_MASK 0xc000 +#define MII_M1011_PHY_STATUS_FULLDUPLEX 0x2000 +#define MII_M1011_PHY_STATUS_RESOLVED 0x0800 +#define MII_M1011_PHY_STATUS_LINK 0x0400 + MODULE_DESCRIPTION("Marvell PHY driver"); MODULE_AUTHOR("Andy Fleming"); @@ -141,12 +157,22 @@ static int marvell_config_aneg(struct phy_device *phydev) static int m88e1111_config_init(struct phy_device *phydev) { int err; + int temp; + int mode; + + /* Enable Fiber/Copper auto selection */ + temp = phy_read(phydev, MII_M1111_PHY_EXT_SR); + temp |= MII_M1111_HWCFG_FIBER_COPPER_AUTO; + phy_write(phydev, MII_M1111_PHY_EXT_SR, temp); + + temp = phy_read(phydev, MII_BMCR); + temp |= BMCR_RESET; + phy_write(phydev, MII_BMCR, temp); if ((phydev->interface == PHY_INTERFACE_MODE_RGMII) || (phydev->interface == PHY_INTERFACE_MODE_RGMII_ID) || (phydev->interface == PHY_INTERFACE_MODE_RGMII_RXID) || (phydev->interface == PHY_INTERFACE_MODE_RGMII_TXID)) { - int temp; temp = phy_read(phydev, MII_M1111_PHY_EXT_CR); if (temp < 0) @@ -171,7 +197,13 @@ static int m88e1111_config_init(struct phy_device *phydev) return temp; temp &= ~(MII_M1111_HWCFG_MODE_MASK); - temp |= MII_M1111_HWCFG_MODE_RGMII; + + mode = phy_read(phydev, MII_M1111_PHY_EXT_CR); + + if (mode & MII_M1111_HWCFG_FIBER_COPPER_RES) + temp |= MII_M1111_HWCFG_MODE_FIBER_RGMII; + else + temp |= MII_M1111_HWCFG_MODE_COPPER_RGMII; err = phy_write(phydev, MII_M1111_PHY_EXT_SR, temp); if (err < 0) @@ -262,6 +294,93 @@ static int m88e1145_config_init(struct phy_device *phydev) return 0; } +/* marvell_read_status + * + * Generic status code does not detect Fiber correctly! + * Description: + * Check the link, then figure out the current state + * by comparing what we advertise with what the link partner + * advertises. Start by checking the gigabit possibilities, + * then move on to 10/100. + */ +static int marvell_read_status(struct phy_device *phydev) +{ + int adv; + int err; + int lpa; + int status = 0; + + /* Update the link, but return if there + * was an error */ + err = genphy_update_link(phydev); + if (err) + return err; + + if (AUTONEG_ENABLE == phydev->autoneg) { + status = phy_read(phydev, MII_M1011_PHY_STATUS); + if (status < 0) + return status; + + lpa = phy_read(phydev, MII_LPA); + if (lpa < 0) + return lpa; + + adv = phy_read(phydev, MII_ADVERTISE); + if (adv < 0) + return adv; + + lpa &= adv; + + if (status & MII_M1011_PHY_STATUS_FULLDUPLEX) + phydev->duplex = DUPLEX_FULL; + else + phydev->duplex = DUPLEX_HALF; + + status = status & MII_M1011_PHY_STATUS_SPD_MASK; + phydev->pause = phydev->asym_pause = 0; + + switch (status) { + case MII_M1011_PHY_STATUS_1000: + phydev->speed = SPEED_1000; + break; + + case MII_M1011_PHY_STATUS_100: + phydev->speed = SPEED_100; + break; + + default: + phydev->speed = SPEED_10; + break; + } + + if (phydev->duplex == DUPLEX_FULL) { + phydev->pause = lpa & LPA_PAUSE_CAP ? 1 : 0; + phydev->asym_pause = lpa & LPA_PAUSE_ASYM ? 1 : 0; + } + } else { + int bmcr = phy_read(phydev, MII_BMCR); + + if (bmcr < 0) + return bmcr; + + if (bmcr & BMCR_FULLDPLX) + phydev->duplex = DUPLEX_FULL; + else + phydev->duplex = DUPLEX_HALF; + + if (bmcr & BMCR_SPEED1000) + phydev->speed = SPEED_1000; + else if (bmcr & BMCR_SPEED100) + phydev->speed = SPEED_100; + else + phydev->speed = SPEED_10; + + phydev->pause = phydev->asym_pause = 0; + } + + return 0; +} + static struct phy_driver marvell_drivers[] = { { .phy_id = 0x01410c60, @@ -296,7 +415,7 @@ static struct phy_driver marvell_drivers[] = { .flags = PHY_HAS_INTERRUPT, .config_init = &m88e1111_config_init, .config_aneg = &marvell_config_aneg, - .read_status = &genphy_read_status, + .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, .config_intr = &marvell_config_intr, .driver = { .owner = THIS_MODULE }, -- cgit v0.10.2 From b47b4b22e2195146baf089b637ec8f6654953018 Mon Sep 17 00:00:00 2001 From: Peter Korsgaard Date: Tue, 18 Mar 2008 23:16:53 +0100 Subject: dm9601: add Hirose USB-100 device ID The Hirose USB-100 adapter uses a dm9601 chip. Reported by Robert Brockway. Signed-off-by: Peter Korsgaard Signed-off-by: Jeff Garzik diff --git a/drivers/net/usb/dm9601.c b/drivers/net/usb/dm9601.c index 4b131a6..723ac47 100644 --- a/drivers/net/usb/dm9601.c +++ b/drivers/net/usb/dm9601.c @@ -562,6 +562,10 @@ static const struct usb_device_id products[] = { USB_DEVICE(0x0a46, 0x8515), /* ADMtek ADM8515 USB NIC */ .driver_info = (unsigned long)&dm9601_info, }, + { + USB_DEVICE(0x0a47, 0x9601), /* Hirose USB-100 */ + .driver_info = (unsigned long)&dm9601_info, + }, {}, // END }; -- cgit v0.10.2 From 33eddedb9cae602e093348421205b1619cccdbb4 Mon Sep 17 00:00:00 2001 From: Peter Korsgaard Date: Tue, 18 Mar 2008 23:16:54 +0100 Subject: dm9601: configure MAC to drop invalid (crc/length) packets Signed-off-by: Peter Korsgaard Signed-off-by: Jeff Garzik diff --git a/drivers/net/usb/dm9601.c b/drivers/net/usb/dm9601.c index 723ac47..0343b00 100644 --- a/drivers/net/usb/dm9601.c +++ b/drivers/net/usb/dm9601.c @@ -341,7 +341,7 @@ static void dm9601_set_multicast(struct net_device *net) /* We use the 20 byte dev->data for our 8 byte filter buffer * to avoid allocating memory that is tricky to free later */ u8 *hashes = (u8 *) & dev->data; - u8 rx_ctl = 0x01; + u8 rx_ctl = 0x31; memset(hashes, 0x00, DM_MCAST_SIZE); hashes[DM_MCAST_SIZE - 1] |= 0x80; /* broadcast address */ -- cgit v0.10.2 From 23d245b66ec8bec21f41ea484e05f470bea764e8 Mon Sep 17 00:00:00 2001 From: Peter Korsgaard Date: Tue, 18 Mar 2008 23:17:16 +0100 Subject: dm9000: Support promisc and all-multi modes Signed-off-by: Peter Korsgaard Signed-off-by: Jeff Garzik diff --git a/drivers/net/dm9000.c b/drivers/net/dm9000.c index 1fe305c..b09a53d 100644 --- a/drivers/net/dm9000.c +++ b/drivers/net/dm9000.c @@ -798,8 +798,6 @@ dm9000_init_dm9000(struct net_device *dev) /* Set address filter table */ dm9000_hash_table(dev); - /* Activate DM9000 */ - iow(db, DM9000_RCR, RCR_DIS_LONG | RCR_DIS_CRC | RCR_RXEN); /* Enable TX/RX interrupt mask */ iow(db, DM9000_IMR, IMR_PAR | IMR_PTM | IMR_PRM); @@ -1197,6 +1195,7 @@ dm9000_hash_table(struct net_device *dev) int i, oft; u32 hash_val; u16 hash_table[4]; + u8 rcr = RCR_DIS_LONG | RCR_DIS_CRC | RCR_RXEN; unsigned long flags; dm9000_dbg(db, 1, "entering %s\n", __func__); @@ -1213,6 +1212,12 @@ dm9000_hash_table(struct net_device *dev) /* broadcast address */ hash_table[3] = 0x8000; + if (dev->flags & IFF_PROMISC) + rcr |= RCR_PRMSC; + + if (dev->flags & IFF_ALLMULTI) + rcr |= RCR_ALL; + /* the multicast address in Hash Table : 64 bits */ for (i = 0; i < mc_cnt; i++, mcptr = mcptr->next) { hash_val = ether_crc_le(6, mcptr->dmi_addr) & 0x3f; @@ -1225,6 +1230,7 @@ dm9000_hash_table(struct net_device *dev) iow(db, oft++, hash_table[i] >> 8); } + iow(db, DM9000_RCR, rcr); spin_unlock_irqrestore(&db->lock, flags); } -- cgit v0.10.2 From dc01c447123b489af7b4d0c58a15abcec36a40e6 Mon Sep 17 00:00:00 2001 From: Thomas Klein Date: Wed, 19 Mar 2008 13:55:43 +0100 Subject: ehea: Fix IPv6 support Indicate that HEA calculates IPv4 checksums only Signed-off-by: Thomas Klein Signed-off-by: Jeff Garzik diff --git a/drivers/net/ehea/ehea.h b/drivers/net/ehea/ehea.h index 7c4ead3..93b7fb2 100644 --- a/drivers/net/ehea/ehea.h +++ b/drivers/net/ehea/ehea.h @@ -40,7 +40,7 @@ #include #define DRV_NAME "ehea" -#define DRV_VERSION "EHEA_0087" +#define DRV_VERSION "EHEA_0089" /* eHEA capability flags */ #define DLPAR_PORT_ADD_REM 1 diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c index 21af674..07c742d 100644 --- a/drivers/net/ehea/ehea_main.c +++ b/drivers/net/ehea/ehea_main.c @@ -3108,7 +3108,7 @@ struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter, dev->vlan_rx_add_vid = ehea_vlan_rx_add_vid; dev->vlan_rx_kill_vid = ehea_vlan_rx_kill_vid; dev->features = NETIF_F_SG | NETIF_F_FRAGLIST | NETIF_F_TSO - | NETIF_F_HIGHDMA | NETIF_F_HW_CSUM | NETIF_F_HW_VLAN_TX + | NETIF_F_HIGHDMA | NETIF_F_IP_CSUM | NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_FILTER | NETIF_F_LLTX; dev->tx_timeout = &ehea_tx_watchdog; -- cgit v0.10.2 From b1186dee3e785679876f6b629609ec080842edda Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Thu, 20 Mar 2008 13:30:48 -0700 Subject: cxgb3: Fix lockdep problems with sge.reg_lock Using iWARP with a Chelsio T3 NIC generates the following lockdep warning: ================================= [ INFO: inconsistent lock state ] 2.6.25-rc6 #50 --------------------------------- inconsistent {softirq-on-W} -> {in-softirq-W} usage. swapper/0 [HC0[0]:SC1[1]:HE0:SE0] takes: (&adap->sge.reg_lock){-+..}, at: [] cxgb_offload_ctl+0x3af/0x507 [cxgb3] The problem is that reg_lock is used with plain spin_lock() in drivers/net/cxgb3/sge.c but is used with spin_lock_irqsave() in drivers/net/cxgb3/cxgb3_offload.c. This is technically a false positive, since the uses in sge.c are only in the initialization and cleanup paths and cannot overlap with any use in interrupt context. The best fix is probably just to use spin_lock_irq() with reg_lock in sge.c. Even though it's not strictly required for correctness, it avoids triggering lockdep and the extra overhead of disabling interrupts is not important at all in the initialization and cleanup slow paths. Signed-off-by: Roland Dreier Signed-off-by: Jeff Garzik diff --git a/drivers/net/cxgb3/sge.c b/drivers/net/cxgb3/sge.c index db58687..98a6bbd 100644 --- a/drivers/net/cxgb3/sge.c +++ b/drivers/net/cxgb3/sge.c @@ -557,9 +557,9 @@ static void t3_free_qset(struct adapter *adapter, struct sge_qset *q) for (i = 0; i < SGE_RXQ_PER_SET; ++i) if (q->fl[i].desc) { - spin_lock(&adapter->sge.reg_lock); + spin_lock_irq(&adapter->sge.reg_lock); t3_sge_disable_fl(adapter, q->fl[i].cntxt_id); - spin_unlock(&adapter->sge.reg_lock); + spin_unlock_irq(&adapter->sge.reg_lock); free_rx_bufs(pdev, &q->fl[i]); kfree(q->fl[i].sdesc); dma_free_coherent(&pdev->dev, @@ -570,9 +570,9 @@ static void t3_free_qset(struct adapter *adapter, struct sge_qset *q) for (i = 0; i < SGE_TXQ_PER_SET; ++i) if (q->txq[i].desc) { - spin_lock(&adapter->sge.reg_lock); + spin_lock_irq(&adapter->sge.reg_lock); t3_sge_enable_ecntxt(adapter, q->txq[i].cntxt_id, 0); - spin_unlock(&adapter->sge.reg_lock); + spin_unlock_irq(&adapter->sge.reg_lock); if (q->txq[i].sdesc) { free_tx_desc(adapter, &q->txq[i], q->txq[i].in_use); @@ -586,9 +586,9 @@ static void t3_free_qset(struct adapter *adapter, struct sge_qset *q) } if (q->rspq.desc) { - spin_lock(&adapter->sge.reg_lock); + spin_lock_irq(&adapter->sge.reg_lock); t3_sge_disable_rspcntxt(adapter, q->rspq.cntxt_id); - spin_unlock(&adapter->sge.reg_lock); + spin_unlock_irq(&adapter->sge.reg_lock); dma_free_coherent(&pdev->dev, q->rspq.size * sizeof(struct rsp_desc), q->rspq.desc, q->rspq.phys_addr); @@ -2667,7 +2667,7 @@ int t3_sge_alloc_qset(struct adapter *adapter, unsigned int id, int nports, (16 * 1024) - SKB_DATA_ALIGN(sizeof(struct skb_shared_info)) : MAX_FRAME_SIZE + 2 + sizeof(struct cpl_rx_pkt); - spin_lock(&adapter->sge.reg_lock); + spin_lock_irq(&adapter->sge.reg_lock); /* FL threshold comparison uses < */ ret = t3_sge_init_rspcntxt(adapter, q->rspq.cntxt_id, irq_vec_idx, @@ -2711,7 +2711,7 @@ int t3_sge_alloc_qset(struct adapter *adapter, unsigned int id, int nports, goto err_unlock; } - spin_unlock(&adapter->sge.reg_lock); + spin_unlock_irq(&adapter->sge.reg_lock); q->adap = adapter; q->netdev = dev; @@ -2728,7 +2728,7 @@ int t3_sge_alloc_qset(struct adapter *adapter, unsigned int id, int nports, return 0; err_unlock: - spin_unlock(&adapter->sge.reg_lock); + spin_unlock_irq(&adapter->sge.reg_lock); err: t3_free_qset(adapter, q); return ret; -- cgit v0.10.2 From 9f5e60dd5ffca938da4cabc197af8b9405b5512e Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Sun, 23 Mar 2008 12:45:44 +0200 Subject: rndis_host: fix oops when query for OID_GEN_PHYSICAL_MEDIUM fails When query for OID_GEN_PHYSICAL_MEDIUM fails, uninitialized pointer 'phym' is being accessed in generic_rndis_bind(), resulting OOPS. Patch fixes phym to be initialized and setup correctly when rndis_query() for physical medium fails. Bug was introduced by following commit: commit 039ee17d1baabaa21783a0d5ab3e8c6d8c794bdf Author: Jussi Kivilinna Date: Sun Jan 27 23:34:33 2008 +0200 Reported-by: Dmitri Monakhov Signed-off-by: Jussi Kivilinna Acked-by: David Brownell Signed-off-by: Jeff Garzik diff --git a/drivers/net/usb/rndis_host.c b/drivers/net/usb/rndis_host.c index 727547a..369c731 100644 --- a/drivers/net/usb/rndis_host.c +++ b/drivers/net/usb/rndis_host.c @@ -283,7 +283,7 @@ generic_rndis_bind(struct usbnet *dev, struct usb_interface *intf, int flags) struct rndis_set_c *set_c; struct rndis_halt *halt; } u; - u32 tmp, *phym; + u32 tmp, phym_unspec, *phym; int reply_len; unsigned char *bp; @@ -363,12 +363,15 @@ generic_rndis_bind(struct usbnet *dev, struct usb_interface *intf, int flags) goto halt_fail_and_release; /* Check physical medium */ + phym = NULL; reply_len = sizeof *phym; retval = rndis_query(dev, intf, u.buf, OID_GEN_PHYSICAL_MEDIUM, 0, (void **) &phym, &reply_len); - if (retval != 0) + if (retval != 0 || !phym) { /* OID is optional so don't fail here. */ - *phym = RNDIS_PHYSICAL_MEDIUM_UNSPECIFIED; + phym_unspec = RNDIS_PHYSICAL_MEDIUM_UNSPECIFIED; + phym = &phym_unspec; + } if ((flags & FLAG_RNDIS_PHYM_WIRELESS) && *phym != RNDIS_PHYSICAL_MEDIUM_WIRELESS_LAN) { if (netif_msg_probe(dev)) -- cgit v0.10.2 From 6ef2977d414cc196baba0fb53509c5f8cd9154b4 Mon Sep 17 00:00:00 2001 From: Marin Mitov Date: Sun, 23 Mar 2008 10:20:09 +0200 Subject: skge napi->poll() locking bug According to: Documentation/networking/netdevices.txt: napi->poll: .......... Context: softirq will be called with interrupts disabled by netconsole. napi->poll() could be called either with interrupts enabled (in softirq context) or disabled (by netconsole), so the irq flag should be preserved. Inspired by Ingo's resent forcedeth patch :-) Signed-off-by: Marin Mitov Signed-off-by: Jeff Garzik diff --git a/drivers/net/skge.c b/drivers/net/skge.c index 186eb8e..ae52cba 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -3199,12 +3199,14 @@ static int skge_poll(struct napi_struct *napi, int to_do) skge_write8(hw, Q_ADDR(rxqaddr[skge->port], Q_CSR), CSR_START); if (work_done < to_do) { - spin_lock_irq(&hw->hw_lock); + unsigned long flags; + + spin_lock_irqsave(&hw->hw_lock, flags); __netif_rx_complete(dev, napi); hw->intr_mask |= napimask[skge->port]; skge_write32(hw, B0_IMSK, hw->intr_mask); skge_read32(hw, B0_IMSK); - spin_unlock_irq(&hw->hw_lock); + spin_unlock_irqrestore(&hw->hw_lock, flags); } return work_done; -- cgit v0.10.2 From 5ea79631c0c47d28831a0635e8af9da539d449cd Mon Sep 17 00:00:00 2001 From: Michael Buesch Date: Tue, 25 Mar 2008 18:04:46 +0100 Subject: b44: Truncate PHY address Some ROMs on embedded devices store incorrect values for the PHY address of the ethernet device. It looks like the number is sign-extended. Truncate the value by applying the PHY-address mask to it. The patch was tested on a bcm47xx embedded system (where the bug triggers) and a bcm4400 PCI card. Signed-off-by: Michael Buesch Signed-off-by: Jeff Garzik diff --git a/drivers/net/b44.c b/drivers/net/b44.c index ea2a2b5..25f1337 100644 --- a/drivers/net/b44.c +++ b/drivers/net/b44.c @@ -2082,6 +2082,11 @@ static int __devinit b44_get_invariants(struct b44 *bp) addr = sdev->bus->sprom.et0mac; bp->phy_addr = sdev->bus->sprom.et0phyaddr; } + /* Some ROMs have buggy PHY addresses with the high + * bits set (sign extension?). Truncate them to a + * valid PHY address. */ + bp->phy_addr &= 0x1F; + memcpy(bp->dev->dev_addr, addr, 6); if (!is_valid_ether_addr(&bp->dev->dev_addr[0])){ -- cgit v0.10.2 From f6f4bfa3561a2d8e949cab0d28e0db6d2b13641d Mon Sep 17 00:00:00 2001 From: Sreenivasa Honnur Date: Tue, 25 Mar 2008 15:11:56 -0400 Subject: S2io: Handle TX completions on the same CPU as the sender for MIS-X interrupts - Handling TX completions on the same cpu as the sender. Signed-off-by: Surjit Reang Signed-off-by: Jeff Garzik diff --git a/drivers/net/s2io.c b/drivers/net/s2io.c index c72787a..3c915b8 100644 --- a/drivers/net/s2io.c +++ b/drivers/net/s2io.c @@ -4172,6 +4172,9 @@ static int s2io_xmit(struct sk_buff *skb, struct net_device *dev) dev->trans_start = jiffies; spin_unlock_irqrestore(&fifo->tx_lock, flags); + if (sp->config.intr_type == MSI_X) + tx_intr_handler(fifo); + return 0; pci_map_failed: stats->pci_map_fail_cnt++; -- cgit v0.10.2 From f0c88f9c45f39acd017328515890481adcb32607 Mon Sep 17 00:00:00 2001 From: Jeff Garzik Date: Tue, 25 Mar 2008 23:53:24 -0400 Subject: netxen, phy/marvell, skge: minor checkpatch fixes Signed-off-by: Jeff Garzik diff --git a/drivers/net/netxen/netxen_nic_init.c b/drivers/net/netxen/netxen_nic_init.c index d9713d9..45fa33e 100644 --- a/drivers/net/netxen/netxen_nic_init.c +++ b/drivers/net/netxen/netxen_nic_init.c @@ -1144,9 +1144,8 @@ u32 netxen_process_rcv_ring(struct netxen_adapter *adapter, int ctxid, int max) consumer = (consumer + 1) & (adapter->max_rx_desc_count - 1); count++; } - for (ring = 0; ring < NUM_RCV_DESC_RINGS; ring++) { + for (ring = 0; ring < NUM_RCV_DESC_RINGS; ring++) netxen_post_rx_buffers_nodb(adapter, ctxid, ring); - } /* update the consumer index in phantom */ if (count) { diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index dc4d593..a8fb439 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -166,9 +166,8 @@ static void netxen_nic_disable_int(struct netxen_adapter *adapter) int port = adapter->portnum; int pci_fn = adapter->ahw.pci_func; - if (adapter->msi_mode != MSI_MODE_MULTIFUNC) { + if (adapter->msi_mode != MSI_MODE_MULTIFUNC) writel(0x0, NETXEN_CRB_NORMALIZE(adapter, sw_int_mask[port])); - } if (adapter->intr_scheme != -1 && adapter->intr_scheme != INTR_SCHEME_PERPORT) diff --git a/drivers/net/phy/marvell.c b/drivers/net/phy/marvell.c index b5f4c28..3353991 100644 --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@ -297,7 +297,7 @@ static int m88e1145_config_init(struct phy_device *phydev) /* marvell_read_status * * Generic status code does not detect Fiber correctly! - * Description: + * Description: * Check the link, then figure out the current state * by comparing what we advertise with what the link partner * advertises. Start by checking the gigabit possibilities, diff --git a/drivers/net/skge.c b/drivers/net/skge.c index ae52cba..2e26dce 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -3200,7 +3200,7 @@ static int skge_poll(struct napi_struct *napi, int to_do) if (work_done < to_do) { unsigned long flags; - + spin_lock_irqsave(&hw->hw_lock, flags); __netif_rx_complete(dev, napi); hw->intr_mask |= napimask[skge->port]; -- cgit v0.10.2 From 0ed21b321a13421e2dfeaa70a6c324e05e3e91e6 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Wed, 26 Mar 2008 00:15:17 -0700 Subject: [VLAN]: Don't copy ALLMULTI/PROMISC flags from underlying device Changing these flags requires to use dev_set_allmulti/dev_set_promiscuity or dev_change_flags. Setting it directly causes two unwanted effects: - the next dev_change_flags call will notice a difference between dev->gflags and the actual flags, enable promisc/allmulti mode and incorrectly update dev->gflags - this keeps the underlying device in promisc/allmulti mode until the VLAN device is deleted Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 8fbcefe..480ea90 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -660,7 +660,7 @@ static int vlan_dev_init(struct net_device *dev) int subclass = 0; /* IFF_BROADCAST|IFF_MULTICAST; ??? */ - dev->flags = real_dev->flags & ~IFF_UP; + dev->flags = real_dev->flags & ~(IFF_UP | IFF_PROMISC | IFF_ALLMULTI); dev->iflink = real_dev->ifindex; dev->state = (real_dev->state & ((1<<__LINK_STATE_NOCARRIER) | (1<<__LINK_STATE_DORMANT))) | -- cgit v0.10.2 From f70c176619dde7f32283fc5c94df8e9aaac12729 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Wed, 26 Mar 2008 00:16:29 -0700 Subject: [UML]: uml-net: don't set IFF_ALLMULTI in set_multicast_list IFF_ALLMULTI is an indication from the network stack to the driver to disable multicast filters, drivers should never set it directly. Since the UML networking device doesn't have any filtering capabilites, it doesn't the set_multicast_list function at all, it is kept so userspace can still issue SIOCADDMULTI/SIOCDELMULTI ioctls however. Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller diff --git a/arch/um/drivers/net_kern.c b/arch/um/drivers/net_kern.c index 1e8f41a..1d43bdf 100644 --- a/arch/um/drivers/net_kern.c +++ b/arch/um/drivers/net_kern.c @@ -256,11 +256,7 @@ static struct net_device_stats *uml_net_get_stats(struct net_device *dev) static void uml_net_set_multicast_list(struct net_device *dev) { - if (dev->flags & IFF_PROMISC) - return; - else if (dev->mc_count) - dev->flags |= IFF_ALLMULTI; - else dev->flags &= ~IFF_ALLMULTI; + return; } static void uml_net_tx_timeout(struct net_device *dev) -- cgit v0.10.2 From 8c7230f781749cd7261b504c0bfa188bb96e77ee Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 26 Mar 2008 00:55:50 -0700 Subject: [IRDA]: Store irnet_socket termios properly. It should be a "struct ktermios" not a "struct termios". Based upon a build warning reported by Stephen Rothwell. Signed-off-by: David S. Miller diff --git a/net/irda/irnet/irnet.h b/net/irda/irnet/irnet.h index bc2e15c..7873c39 100644 --- a/net/irda/irnet/irnet.h +++ b/net/irda/irnet/irnet.h @@ -405,7 +405,7 @@ typedef struct irnet_socket /* "pppd" interact directly with us on a /dev/ file */ struct file * file; /* File descriptor of this instance */ /* TTY stuff - to keep "pppd" happy */ - struct termios termios; /* Various tty flags */ + struct ktermios termios; /* Various tty flags */ /* Stuff for the control channel */ int event_index; /* Last read in the event log */ -- cgit v0.10.2 From 61ee6bd487b9cc160e533034eb338f2085dc7922 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Wed, 26 Mar 2008 02:12:11 -0700 Subject: [NET]: Fix multicast device ioctl checks SIOCADDMULTI/SIOCDELMULTI check whether the driver has a set_multicast_list method to determine whether it supports multicast. Drivers implementing secondary unicast support use set_rx_mode however. Check for both dev->set_multicast_mode and dev->set_rx_mode to determine multicast capabilities. Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller diff --git a/net/core/dev.c b/net/core/dev.c index fcdf03c..460e7f9 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3329,7 +3329,7 @@ static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd) return -EOPNOTSUPP; case SIOCADDMULTI: - if (!dev->set_multicast_list || + if ((!dev->set_multicast_list && !dev->set_rx_mode) || ifr->ifr_hwaddr.sa_family != AF_UNSPEC) return -EINVAL; if (!netif_device_present(dev)) @@ -3338,7 +3338,7 @@ static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd) dev->addr_len, 1); case SIOCDELMULTI: - if (!dev->set_multicast_list || + if ((!dev->set_multicast_list && !dev->set_rx_mode) || ifr->ifr_hwaddr.sa_family != AF_UNSPEC) return -EINVAL; if (!netif_device_present(dev)) -- cgit v0.10.2 From 5c2e2e239ebe44e3fdc5f2ae270d96c4ceee4e9a Mon Sep 17 00:00:00 2001 From: "Robert P. J. Day" Date: Wed, 26 Mar 2008 02:14:38 -0700 Subject: [AX25]: Remove obsolete references to BKL from TODO file. Given that there are no apparent calls to lock_kernel() or unlock_kernel() under net/ax25, delete the TODO reference related to that. Signed-off-by: Robert P. J. Day Signed-off-by: David S. Miller diff --git a/net/ax25/TODO b/net/ax25/TODO index 4089c49..69fb4e3 100644 --- a/net/ax25/TODO +++ b/net/ax25/TODO @@ -9,10 +9,6 @@ being used. Routes to a device being taken down might be deleted by ax25_rt_device_down but added by somebody else before the device has been deleted fully. -Massive amounts of lock_kernel / unlock_kernel are just a temporary solution to -get around the removal of SOCKOPS_WRAP. A serious locking strategy has to be -implemented. - The ax25_rt_find_route synopsys is pervert but I somehow had to deal with the race caused by the static variable in it's previous implementation. -- cgit v0.10.2 From 7c0ecc4c4f8fd90988aab8a95297b9c0038b6160 Mon Sep 17 00:00:00 2001 From: Pavel Emelyanov Date: Wed, 26 Mar 2008 02:27:09 -0700 Subject: [ICMP]: Dst entry leak in icmp_send host re-lookup code (v2). Commit 8b7817f3a959ed99d7443afc12f78a7e1fcc2063 ([IPSEC]: Add ICMP host relookup support) introduced some dst leaks on error paths: the rt pointer can be forgotten to be put. Fix it bu going to a proper label. Found after net namespace's lo refused to unregister :) Many thanks to Den for valuable help during debugging. Herbert pointed out, that xfrm_lookup() will put the rtable in case of error itself, so the first goto fix is redundant. Signed-off-by: Pavel Emelyanov Signed-off-by: Denis V. Lunev Signed-off-by: David S. Miller diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index a13c074..a944e80 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -591,7 +591,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) } if (xfrm_decode_session_reverse(skb_in, &fl, AF_INET)) - goto out_unlock; + goto ende; if (inet_addr_type(net, fl.fl4_src) == RTN_LOCAL) err = __ip_route_output_key(net, &rt2, &fl); @@ -601,7 +601,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) fl2.fl4_dst = fl.fl4_src; if (ip_route_output_key(net, &rt2, &fl2)) - goto out_unlock; + goto ende; /* Ugh! */ odst = skb_in->dst; @@ -614,7 +614,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info) } if (err) - goto out_unlock; + goto ende; err = xfrm_lookup((struct dst_entry **)&rt2, &fl, NULL, XFRM_LOOKUP_ICMP); -- cgit v0.10.2 From 732c8bd590625e8bc0b88313b82930e336b2bec4 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 26 Mar 2008 16:51:09 -0700 Subject: [IPSEC]: Fix BEET output The IPv6 BEET output function is incorrectly including the inner header in the payload to be protected. This causes a crash as the packet doesn't actually have that many bytes for a second header. The IPv4 BEET output on the other hand is broken when it comes to handling an inner IPv6 header since it always assumes an inner IPv4 header. This patch fixes both by making sure that neither BEET output function touches the inner header at all. All access is now done through the protocol-independent cb structure. Two new attributes are added to make this work, the IP header length and the IPv4 option length. They're filled in by the inner mode's output function. Thanks to Joakim Koskela for finding this problem. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 4e6f956..0d255ae 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -552,6 +552,9 @@ struct xfrm_mode_skb_cb { __be16 id; __be16 frag_off; + /* IP header length (excluding options or extension headers). */ + u8 ihl; + /* TOS for IPv4, class for IPv6. */ u8 tos; @@ -561,6 +564,9 @@ struct xfrm_mode_skb_cb { /* Protocol for IPv4, NH for IPv6. */ u8 protocol; + /* Option length for IPv4, zero for IPv6. */ + u8 optlen; + /* Used by IPv6 only, zero for IPv4. */ u8 flow_lbl[3]; }; diff --git a/net/ipv4/xfrm4_mode_beet.c b/net/ipv4/xfrm4_mode_beet.c index b47030b..9c798ab 100644 --- a/net/ipv4/xfrm4_mode_beet.c +++ b/net/ipv4/xfrm4_mode_beet.c @@ -39,13 +39,11 @@ static void xfrm4_beet_make_header(struct sk_buff *skb) static int xfrm4_beet_output(struct xfrm_state *x, struct sk_buff *skb) { struct ip_beet_phdr *ph; - struct iphdr *iph, *top_iph; + struct iphdr *top_iph; int hdrlen, optlen; - iph = ip_hdr(skb); - hdrlen = 0; - optlen = iph->ihl * 4 - sizeof(*iph); + optlen = XFRM_MODE_SKB_CB(skb)->optlen; if (unlikely(optlen)) hdrlen += IPV4_BEET_PHMAXLEN - (optlen & 4); @@ -53,11 +51,12 @@ static int xfrm4_beet_output(struct xfrm_state *x, struct sk_buff *skb) hdrlen); skb->mac_header = skb->network_header + offsetof(struct iphdr, protocol); - skb->transport_header = skb->network_header + sizeof(*iph); + skb->transport_header = skb->network_header + sizeof(*top_iph); xfrm4_beet_make_header(skb); - ph = (struct ip_beet_phdr *)__skb_pull(skb, sizeof(*iph) - hdrlen); + ph = (struct ip_beet_phdr *) + __skb_pull(skb, XFRM_MODE_SKB_CB(skb)->ihl - hdrlen); top_iph = ip_hdr(skb); diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c index fdeebe6..07735ed 100644 --- a/net/ipv4/xfrm4_state.c +++ b/net/ipv4/xfrm4_state.c @@ -52,10 +52,12 @@ int xfrm4_extract_header(struct sk_buff *skb) { struct iphdr *iph = ip_hdr(skb); + XFRM_MODE_SKB_CB(skb)->ihl = sizeof(*iph); XFRM_MODE_SKB_CB(skb)->id = iph->id; XFRM_MODE_SKB_CB(skb)->frag_off = iph->frag_off; XFRM_MODE_SKB_CB(skb)->tos = iph->tos; XFRM_MODE_SKB_CB(skb)->ttl = iph->ttl; + XFRM_MODE_SKB_CB(skb)->optlen = iph->ihl * 4 - sizeof(*iph); memset(XFRM_MODE_SKB_CB(skb)->flow_lbl, 0, sizeof(XFRM_MODE_SKB_CB(skb)->flow_lbl)); diff --git a/net/ipv6/xfrm6_mode_beet.c b/net/ipv6/xfrm6_mode_beet.c index 0527d11..d6ce400 100644 --- a/net/ipv6/xfrm6_mode_beet.c +++ b/net/ipv6/xfrm6_mode_beet.c @@ -45,6 +45,7 @@ static int xfrm6_beet_output(struct xfrm_state *x, struct sk_buff *skb) skb->mac_header = skb->network_header + offsetof(struct ipv6hdr, nexthdr); skb->transport_header = skb->network_header + sizeof(*top_iph); + __skb_pull(skb, XFRM_MODE_SKB_CB(skb)->ihl); xfrm6_beet_make_header(skb); diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c index dc817e0..ff1e1db 100644 --- a/net/ipv6/xfrm6_state.c +++ b/net/ipv6/xfrm6_state.c @@ -174,10 +174,12 @@ int xfrm6_extract_header(struct sk_buff *skb) { struct ipv6hdr *iph = ipv6_hdr(skb); + XFRM_MODE_SKB_CB(skb)->ihl = sizeof(*iph); XFRM_MODE_SKB_CB(skb)->id = 0; XFRM_MODE_SKB_CB(skb)->frag_off = htons(IP_DF); XFRM_MODE_SKB_CB(skb)->tos = ipv6_get_dsfield(iph); XFRM_MODE_SKB_CB(skb)->ttl = iph->hop_limit; + XFRM_MODE_SKB_CB(skb)->optlen = 0; memcpy(XFRM_MODE_SKB_CB(skb)->flow_lbl, iph->flow_lbl, sizeof(XFRM_MODE_SKB_CB(skb)->flow_lbl)); -- cgit v0.10.2 From 920fc941a9617f95ccb283037fe6f8a38d95bb69 Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 27 Mar 2008 16:08:03 -0700 Subject: [ESP]: Ensure IV is in linear part of the skb to avoid BUG() due to OOB access ESP does not account for the IV size when calling pskb_may_pull() to ensure everything it accesses directly is within the linear part of a potential fragment. This results in a BUG() being triggered when the both the IPv4 and IPv6 ESP stack is fed with an skb where the first fragment ends between the end of the esp header and the end of the IV. This bug was found by Dirk Nehring . Signed-off-by: Thomas Graf Signed-off-by: David S. Miller diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c index f3ceca3..4e73e57 100644 --- a/net/ipv4/esp4.c +++ b/net/ipv4/esp4.c @@ -336,7 +336,7 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb) struct scatterlist *asg; int err = -EINVAL; - if (!pskb_may_pull(skb, sizeof(*esph))) + if (!pskb_may_pull(skb, sizeof(*esph) + crypto_aead_ivsize(aead))) goto out; if (elen <= 0) diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c index 0ec1402..c6bb4c6 100644 --- a/net/ipv6/esp6.c +++ b/net/ipv6/esp6.c @@ -282,7 +282,7 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb) struct scatterlist *sg; struct scatterlist *asg; - if (!pskb_may_pull(skb, sizeof(*esph))) { + if (!pskb_may_pull(skb, sizeof(*esph) + crypto_aead_ivsize(aead))) { ret = -EINVAL; goto out; } -- cgit v0.10.2 From 8eeee8b152ae6bbe181518efaf62ba8e9c613693 Mon Sep 17 00:00:00 2001 From: "Denis V. Lunev" Date: Thu, 27 Mar 2008 16:55:53 -0700 Subject: [NETFILTER]: Replate direct proc_fops assignment with proc_create call. This elliminates infamous race during module loading when one could lookup proc entry without proc_fops assigned. Signed-off-by: Denis V. Lunev Signed-off-by: David S. Miller diff --git a/net/ipv4/netfilter/ip_queue.c b/net/ipv4/netfilter/ip_queue.c index fe05da4..4dc1628 100644 --- a/net/ipv4/netfilter/ip_queue.c +++ b/net/ipv4/netfilter/ip_queue.c @@ -588,11 +588,9 @@ static int __init ip_queue_init(void) } #ifdef CONFIG_PROC_FS - proc = create_proc_entry(IPQ_PROC_FS_NAME, 0, init_net.proc_net); - if (proc) { - proc->owner = THIS_MODULE; - proc->proc_fops = &ip_queue_proc_fops; - } else { + proc = proc_create(IPQ_PROC_FS_NAME, 0, init_net.proc_net, + &ip_queue_proc_fops); + if (!proc) { printk(KERN_ERR "ip_queue: failed to create proc entry\n"); goto cleanup_ipqnl; } diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c index c6cf84c..52926c8 100644 --- a/net/ipv4/netfilter/ipt_CLUSTERIP.c +++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c @@ -167,14 +167,13 @@ clusterip_config_init(struct ipt_clusterip_tgt_info *i, __be32 ip, /* create proc dir entry */ sprintf(buffer, "%u.%u.%u.%u", NIPQUAD(ip)); - c->pde = create_proc_entry(buffer, S_IWUSR|S_IRUSR, - clusterip_procdir); + c->pde = proc_create(buffer, S_IWUSR|S_IRUSR, + clusterip_procdir, &clusterip_proc_fops); if (!c->pde) { kfree(c); return NULL; } } - c->pde->proc_fops = &clusterip_proc_fops; c->pde->data = c; #endif diff --git a/net/ipv4/netfilter/ipt_recent.c b/net/ipv4/netfilter/ipt_recent.c index 8e8f042..50e0669 100644 --- a/net/ipv4/netfilter/ipt_recent.c +++ b/net/ipv4/netfilter/ipt_recent.c @@ -276,12 +276,11 @@ recent_mt_check(const char *tablename, const void *ip, for (i = 0; i < ip_list_hash_size; i++) INIT_LIST_HEAD(&t->iphash[i]); #ifdef CONFIG_PROC_FS - t->proc = create_proc_entry(t->name, ip_list_perms, proc_dir); + t->proc = proc_create(t->name, ip_list_perms, proc_dir, &recent_fops); if (t->proc == NULL) { kfree(t); goto out; } - t->proc->proc_fops = &recent_fops; t->proc->uid = ip_list_uid; t->proc->gid = ip_list_gid; t->proc->data = t; diff --git a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c index 089252e..f500b0f 100644 --- a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c +++ b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c @@ -395,13 +395,10 @@ int __init nf_conntrack_ipv4_compat_init(void) if (!proc_exp) goto err2; - proc_stat = create_proc_entry("ip_conntrack", S_IRUGO, init_net.proc_net_stat); + proc_stat = proc_create("ip_conntrack", S_IRUGO, + init_net.proc_net_stat, &ct_cpu_seq_fops); if (!proc_stat) goto err3; - - proc_stat->proc_fops = &ct_cpu_seq_fops; - proc_stat->owner = THIS_MODULE; - return 0; err3: diff --git a/net/ipv6/netfilter/ip6_queue.c b/net/ipv6/netfilter/ip6_queue.c index cc2f9af..8d366f7 100644 --- a/net/ipv6/netfilter/ip6_queue.c +++ b/net/ipv6/netfilter/ip6_queue.c @@ -591,11 +591,9 @@ static int __init ip6_queue_init(void) } #ifdef CONFIG_PROC_FS - proc = create_proc_entry(IPQ_PROC_FS_NAME, 0, init_net.proc_net); - if (proc) { - proc->owner = THIS_MODULE; - proc->proc_fops = &ip6_queue_proc_fops; - } else { + proc = proc_create(IPQ_PROC_FS_NAME, 0, init_net.proc_net, + &ip6_queue_proc_fops); + if (!proc) { printk(KERN_ERR "ip6_queue: failed to create proc entry\n"); goto cleanup_ipqnl; } diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c index e88e96a..8599068 100644 --- a/net/netfilter/nf_conntrack_standalone.c +++ b/net/netfilter/nf_conntrack_standalone.c @@ -395,7 +395,7 @@ EXPORT_SYMBOL_GPL(nf_ct_log_invalid); static int __init nf_conntrack_standalone_init(void) { #ifdef CONFIG_PROC_FS - struct proc_dir_entry *proc, *proc_stat; + struct proc_dir_entry *proc; #endif int ret = 0; @@ -407,12 +407,9 @@ static int __init nf_conntrack_standalone_init(void) proc = proc_net_fops_create(&init_net, "nf_conntrack", 0440, &ct_file_ops); if (!proc) goto cleanup_init; - proc_stat = create_proc_entry("nf_conntrack", S_IRUGO, init_net.proc_net_stat); - if (!proc_stat) + if (!proc_create("nf_conntrack", S_IRUGO, + init_net.proc_net_stat, &ct_cpu_seq_fops)) goto cleanup_proc; - - proc_stat->proc_fops = &ct_cpu_seq_fops; - proc_stat->owner = THIS_MODULE; #endif #ifdef CONFIG_SYSCTL nf_ct_sysctl_header = register_sysctl_paths(nf_ct_path, diff --git a/net/netfilter/nf_log.c b/net/netfilter/nf_log.c index cec9976..bc11d70 100644 --- a/net/netfilter/nf_log.c +++ b/net/netfilter/nf_log.c @@ -168,13 +168,9 @@ static const struct file_operations nflog_file_ops = { int __init netfilter_log_init(void) { #ifdef CONFIG_PROC_FS - struct proc_dir_entry *pde; - - pde = create_proc_entry("nf_log", S_IRUGO, proc_net_netfilter); - if (!pde) + if (!proc_create("nf_log", S_IRUGO, + proc_net_netfilter, &nflog_file_ops)) return -1; - - pde->proc_fops = &nflog_file_ops; #endif return 0; } diff --git a/net/netfilter/nf_queue.c b/net/netfilter/nf_queue.c index ddc80ea..bbd2689 100644 --- a/net/netfilter/nf_queue.c +++ b/net/netfilter/nf_queue.c @@ -348,12 +348,9 @@ static const struct file_operations nfqueue_file_ops = { int __init netfilter_queue_init(void) { #ifdef CONFIG_PROC_FS - struct proc_dir_entry *pde; - - pde = create_proc_entry("nf_queue", S_IRUGO, proc_net_netfilter); - if (!pde) + if (!proc_create("nf_queue", S_IRUGO, + proc_net_netfilter, &nfqueue_file_ops)) return -1; - pde->proc_fops = &nfqueue_file_ops; #endif return 0; } diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index bf3f19b..b8173af 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -923,9 +923,6 @@ static const struct file_operations nful_file_ops = { static int __init nfnetlink_log_init(void) { int i, status = -ENOMEM; -#ifdef CONFIG_PROC_FS - struct proc_dir_entry *proc_nful; -#endif for (i = 0; i < INSTANCE_BUCKETS; i++) INIT_HLIST_HEAD(&instance_table[i]); @@ -943,11 +940,9 @@ static int __init nfnetlink_log_init(void) } #ifdef CONFIG_PROC_FS - proc_nful = create_proc_entry("nfnetlink_log", 0440, - proc_net_netfilter); - if (!proc_nful) + if (!proc_create("nfnetlink_log", 0440, + proc_net_netfilter, &nful_file_ops)) goto cleanup_subsys; - proc_nful->proc_fops = &nful_file_ops; #endif return status; diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c index 012cb69..10522c0 100644 --- a/net/netfilter/nfnetlink_queue.c +++ b/net/netfilter/nfnetlink_queue.c @@ -896,9 +896,6 @@ static const struct file_operations nfqnl_file_ops = { static int __init nfnetlink_queue_init(void) { int i, status = -ENOMEM; -#ifdef CONFIG_PROC_FS - struct proc_dir_entry *proc_nfqueue; -#endif for (i = 0; i < INSTANCE_BUCKETS; i++) INIT_HLIST_HEAD(&instance_table[i]); @@ -911,11 +908,9 @@ static int __init nfnetlink_queue_init(void) } #ifdef CONFIG_PROC_FS - proc_nfqueue = create_proc_entry("nfnetlink_queue", 0440, - proc_net_netfilter); - if (!proc_nfqueue) + if (!proc_create("nfnetlink_queue", 0440, + proc_net_netfilter, &nfqnl_file_ops)) goto cleanup_subsys; - proc_nfqueue->proc_fops = &nfqnl_file_ops; #endif register_netdevice_notifier(&nfqnl_dev_notifier); diff --git a/net/netfilter/xt_hashlimit.c b/net/netfilter/xt_hashlimit.c index 5418ce5..dc29007 100644 --- a/net/netfilter/xt_hashlimit.c +++ b/net/netfilter/xt_hashlimit.c @@ -237,14 +237,14 @@ static int htable_create_v0(struct xt_hashlimit_info *minfo, int family) hinfo->family = family; hinfo->rnd_initialized = 0; spin_lock_init(&hinfo->lock); - hinfo->pde = create_proc_entry(minfo->name, 0, - family == AF_INET ? hashlimit_procdir4 : - hashlimit_procdir6); + hinfo->pde = proc_create(minfo->name, 0, + family == AF_INET ? hashlimit_procdir4 : + hashlimit_procdir6, + &dl_file_ops); if (!hinfo->pde) { vfree(hinfo); return -1; } - hinfo->pde->proc_fops = &dl_file_ops; hinfo->pde->data = hinfo; setup_timer(&hinfo->timer, htable_gc, (unsigned long )hinfo); @@ -301,14 +301,14 @@ static int htable_create(struct xt_hashlimit_mtinfo1 *minfo, hinfo->rnd_initialized = 0; spin_lock_init(&hinfo->lock); - hinfo->pde = create_proc_entry(minfo->name, 0, - family == AF_INET ? hashlimit_procdir4 : - hashlimit_procdir6); + hinfo->pde = proc_create(minfo->name, 0, + family == AF_INET ? hashlimit_procdir4 : + hashlimit_procdir6, + &dl_file_ops); if (hinfo->pde == NULL) { vfree(hinfo); return -1; } - hinfo->pde->proc_fops = &dl_file_ops; hinfo->pde->data = hinfo; setup_timer(&hinfo->timer, htable_gc, (unsigned long)hinfo); -- cgit v0.10.2 From 50fd4407b8bfbde7c1a0bfe4f24de7df37164342 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 27 Mar 2008 17:42:50 -0700 Subject: [NET]: Use local_irq_{save,restore}() in napi_complete(). Based upon a lockdep report. Since ->poll() can be invoked from netpoll with interrupts disabled, we must not unconditionally enable interrupts in napi_complete(). Instead we must use local_irq_{save,restore}(). Noticed by Peter Zijlstra: netpoll_poll() poll_napi() spin_trylock(&napi->poll_lock) poll_one_napi() napi->poll() := sky2_poll() napi_complete() local_irq_disable() local_irq_enable() <--- *BUG* irq_exit() do_softirq() net_rx_action() spin_lock(&napi->poll_lock) <--- Deadlock! Because we still hold the lock.... Signed-off-by: David S. Miller diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index a2f0032..fae6a7e 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -383,9 +383,11 @@ static inline void __napi_complete(struct napi_struct *n) static inline void napi_complete(struct napi_struct *n) { - local_irq_disable(); + unsigned long flags; + + local_irq_save(flags); __napi_complete(n); - local_irq_enable(); + local_irq_restore(flags); } /** -- cgit v0.10.2