Default: 87380*2 bytes.
tcp_mem - vector of 3 INTEGERs: min, pressure, max
- low: below this number of pages TCP is not bothered about its
+ min: below this number of pages TCP is not bothered about its
memory appetite.
pressure: when amount of memory allocated by TCP exceeds this number
of pages, TCP moderates its memory consumption and enters memory
pressure mode, which is exited when memory consumption falls
- under "low".
+ under "min".
- high: number of pages allowed for queueing by all TCP sockets.
+ max: number of pages allowed for queueing by all TCP sockets.
Defaults are calculated at boot time from amount of available
memory.
#define DRV_MODULE_NAME "bnx2"
#define PFX DRV_MODULE_NAME ": "
-#define DRV_MODULE_VERSION "1.4.43"
-#define DRV_MODULE_RELDATE "June 28, 2006"
+#define DRV_MODULE_VERSION "1.4.44"
+#define DRV_MODULE_RELDATE "August 10, 2006"
#define RUN_AT(x) (jiffies + (x))
static inline u32 bnx2_tx_avail(struct bnx2 *bp)
{
- u32 diff = TX_RING_IDX(bp->tx_prod) - TX_RING_IDX(bp->tx_cons);
+ u32 diff;
+ smp_mb();
+ diff = TX_RING_IDX(bp->tx_prod) - TX_RING_IDX(bp->tx_cons);
if (diff > MAX_TX_DESC_CNT)
diff = (diff & MAX_TX_DESC_CNT) - 1;
return (bp->tx_ring_size - diff);
struct rx_bd *rxbd = &bp->rx_desc_ring[RX_RING(index)][RX_IDX(index)];
unsigned long align;
- skb = dev_alloc_skb(bp->rx_buf_size);
+ skb = netdev_alloc_skb(bp->dev, bp->rx_buf_size);
if (skb == NULL) {
return -ENOMEM;
}
skb_reserve(skb, 8 - align);
}
- skb->dev = bp->dev;
mapping = pci_map_single(bp->pdev, skb->data, bp->rx_buf_use_size,
PCI_DMA_FROMDEVICE);
}
bp->tx_cons = sw_cons;
+ /* Need to make the tx_cons update visible to bnx2_start_xmit()
+ * before checking for netif_queue_stopped(). Without the
+ * memory barrier, there is a small possibility that bnx2_start_xmit()
+ * will miss it and cause the queue to be stopped forever.
+ */
+ smp_mb();
- if (unlikely(netif_queue_stopped(bp->dev))) {
- spin_lock(&bp->tx_lock);
+ if (unlikely(netif_queue_stopped(bp->dev)) &&
+ (bnx2_tx_avail(bp) > bp->tx_wake_thresh)) {
+ netif_tx_lock(bp->dev);
if ((netif_queue_stopped(bp->dev)) &&
- (bnx2_tx_avail(bp) > MAX_SKB_FRAGS)) {
-
+ (bnx2_tx_avail(bp) > bp->tx_wake_thresh))
netif_wake_queue(bp->dev);
- }
- spin_unlock(&bp->tx_lock);
+ netif_tx_unlock(bp->dev);
}
}
if ((bp->dev->mtu > 1500) && (len <= RX_COPY_THRESH)) {
struct sk_buff *new_skb;
- new_skb = dev_alloc_skb(len + 2);
+ new_skb = netdev_alloc_skb(bp->dev, len + 2);
if (new_skb == NULL)
goto reuse_rx;
skb_reserve(new_skb, 2);
skb_put(new_skb, len);
- new_skb->dev = bp->dev;
bnx2_reuse_rx_skb(bp, skb,
sw_ring_cons, sw_ring_prod);
struct tx_bd *txbd;
u32 val;
+ bp->tx_wake_thresh = bp->tx_ring_size / 2;
+
txbd = &bp->tx_desc_ring[MAX_TX_DESC_CNT];
txbd->tx_bd_haddr_hi = (u64) bp->tx_desc_mapping >> 32;
return -EINVAL;
pkt_size = 1514;
- skb = dev_alloc_skb(pkt_size);
+ skb = netdev_alloc_skb(bp->dev, pkt_size);
if (!skb)
return -ENOMEM;
packet = skb_put(skb, pkt_size);
#endif
/* Called with netif_tx_lock.
- * hard_start_xmit is pseudo-lockless - a lock is only required when
- * the tx queue is full. This way, we get the benefit of lockless
- * operations most of the time without the complexities to handle
- * netif_stop_queue/wake_queue race conditions.
+ * bnx2_tx_int() runs without netif_tx_lock unless it needs to call
+ * netif_wake_queue().
*/
static int
bnx2_start_xmit(struct sk_buff *skb, struct net_device *dev)
dev->trans_start = jiffies;
if (unlikely(bnx2_tx_avail(bp) <= MAX_SKB_FRAGS)) {
- spin_lock(&bp->tx_lock);
netif_stop_queue(dev);
-
- if (bnx2_tx_avail(bp) > MAX_SKB_FRAGS)
+ if (bnx2_tx_avail(bp) > bp->tx_wake_thresh)
netif_wake_queue(dev);
- spin_unlock(&bp->tx_lock);
}
return NETDEV_TX_OK;
bp->pdev = pdev;
spin_lock_init(&bp->phy_lock);
- spin_lock_init(&bp->tx_lock);
INIT_WORK(&bp->reset_task, bnx2_reset_task, bp);
dev->base_addr = dev->mem_start = pci_resource_start(pdev, 0);
bp->mac_addr[5] = (u8) reg;
bp->tx_ring_size = MAX_TX_DESC_CNT;
- bnx2_set_rx_ring_size(bp, 100);
+ bnx2_set_rx_ring_size(bp, 255);
bp->rx_csum = 1;
u32 tx_prod_bseq __attribute__((aligned(L1_CACHE_BYTES)));
u16 tx_prod;
- struct tx_bd *tx_desc_ring;
- struct sw_bd *tx_buf_ring;
- int tx_ring_size;
-
u16 tx_cons __attribute__((aligned(L1_CACHE_BYTES)));
u16 hw_tx_cons;
struct sw_bd *rx_buf_ring;
struct rx_bd *rx_desc_ring[MAX_RX_RINGS];
- /* Only used to synchronize netif_stop_queue/wake_queue when tx */
- /* ring is full */
- spinlock_t tx_lock;
+ /* TX constants */
+ struct tx_bd *tx_desc_ring;
+ struct sw_bd *tx_buf_ring;
+ int tx_ring_size;
+ u32 tx_wake_thresh;
/* End of fields used in the performance code paths. */
void *ptr[CARDMAP_WIDTH];
};
static void *cardmap_get(struct cardmap *map, unsigned int nr);
-static void cardmap_set(struct cardmap **map, unsigned int nr, void *ptr);
+static int cardmap_set(struct cardmap **map, unsigned int nr, void *ptr);
static unsigned int cardmap_find_first_free(struct cardmap *map);
static void cardmap_destroy(struct cardmap **map);
{
struct channel *pch;
- pch = kmalloc(sizeof(struct channel), GFP_KERNEL);
+ pch = kzalloc(sizeof(struct channel), GFP_KERNEL);
if (pch == 0)
return -ENOMEM;
- memset(pch, 0, sizeof(struct channel));
pch->ppp = NULL;
pch->chan = chan;
chan->ppp = pch;
int ret = -ENOMEM;
int i;
- ppp = kmalloc(sizeof(struct ppp), GFP_KERNEL);
+ ppp = kzalloc(sizeof(struct ppp), GFP_KERNEL);
if (!ppp)
goto out;
dev = alloc_netdev(0, "", ppp_setup);
if (!dev)
goto out1;
- memset(ppp, 0, sizeof(struct ppp));
ppp->mru = PPP_MRU;
init_ppp_file(&ppp->file, INTERFACE);
}
atomic_inc(&ppp_unit_count);
- cardmap_set(&all_ppp_units, unit, ppp);
+ ret = cardmap_set(&all_ppp_units, unit, ppp);
+ if (ret != 0)
+ goto out3;
+
mutex_unlock(&all_ppp_mutex);
*retp = 0;
return ppp;
+out3:
+ atomic_dec(&ppp_unit_count);
out2:
mutex_unlock(&all_ppp_mutex);
free_netdev(dev);
return NULL;
}
-static void cardmap_set(struct cardmap **pmap, unsigned int nr, void *ptr)
+static int cardmap_set(struct cardmap **pmap, unsigned int nr, void *ptr)
{
struct cardmap *p;
int i;
if (p == NULL || (nr >> p->shift) >= CARDMAP_WIDTH) {
do {
/* need a new top level */
- struct cardmap *np = kmalloc(sizeof(*np), GFP_KERNEL);
- memset(np, 0, sizeof(*np));
+ struct cardmap *np = kzalloc(sizeof(*np), GFP_KERNEL);
+ if (!np)
+ goto enomem;
np->ptr[0] = p;
if (p != NULL) {
np->shift = p->shift + CARDMAP_ORDER;
while (p->shift > 0) {
i = (nr >> p->shift) & CARDMAP_MASK;
if (p->ptr[i] == NULL) {
- struct cardmap *np = kmalloc(sizeof(*np), GFP_KERNEL);
- memset(np, 0, sizeof(*np));
+ struct cardmap *np = kzalloc(sizeof(*np), GFP_KERNEL);
+ if (!np)
+ goto enomem;
np->shift = p->shift - CARDMAP_ORDER;
np->parent = p;
p->ptr[i] = np;
set_bit(i, &p->inuse);
else
clear_bit(i, &p->inuse);
+ return 0;
+ enomem:
+ return -ENOMEM;
}
static unsigned int cardmap_find_first_free(struct cardmap *map)
{
struct net_device_stats *stats;
+ if (skb_bond_should_drop(skb)) {
+ dev_kfree_skb_any(skb);
+ return NET_RX_DROP;
+ }
+
skb->dev = grp->vlan_devices[vlan_tag & VLAN_VID_MASK];
if (skb->dev == NULL) {
dev_kfree_skb_any(skb);
#define NETIF_F_TSO_ECN (SKB_GSO_TCP_ECN << NETIF_F_GSO_SHIFT)
#define NETIF_F_TSO6 (SKB_GSO_TCPV6 << NETIF_F_GSO_SHIFT)
+ /* List of features with software fallbacks. */
+#define NETIF_F_GSO_SOFTWARE (NETIF_F_TSO | NETIF_F_TSO_ECN | NETIF_F_TSO6)
+
#define NETIF_F_GEN_CSUM (NETIF_F_NO_CSUM | NETIF_F_HW_CSUM)
#define NETIF_F_ALL_CSUM (NETIF_F_IP_CSUM | NETIF_F_GEN_CSUM)
unlikely(skb->ip_summed != CHECKSUM_HW));
}
+/* On bonding slaves other than the currently active slave, suppress
+ * duplicates except for 802.3ad ETH_P_SLOW and alb non-mcast/bcast.
+ */
+static inline int skb_bond_should_drop(struct sk_buff *skb)
+{
+ struct net_device *dev = skb->dev;
+ struct net_device *master = dev->master;
+
+ if (master &&
+ (dev->priv_flags & IFF_SLAVE_INACTIVE)) {
+ if (master->priv_flags & IFF_MASTER_ALB) {
+ if (skb->pkt_type != PACKET_BROADCAST &&
+ skb->pkt_type != PACKET_MULTICAST)
+ return 0;
+ }
+ if (master->priv_flags & IFF_MASTER_8023AD &&
+ skb->protocol == __constant_htons(ETH_P_SLOW))
+ return 0;
+
+ return 1;
+ }
+ return 0;
+}
+
#endif /* __KERNEL__ */
#endif /* _LINUX_DEV_H */
goto out;
}
-void __exit atm_proc_exit(void)
+void atm_proc_exit(void)
{
atm_proc_dirs_remove();
}
checksum = 0;
if (feature & NETIF_F_GSO)
- feature |= NETIF_F_TSO;
+ feature |= NETIF_F_GSO_SOFTWARE;
feature |= NETIF_F_GSO;
features &= feature;
}
+ if (!(checksum & NETIF_F_ALL_CSUM))
+ features &= ~NETIF_F_SG;
+ if (!(features & NETIF_F_SG))
+ features &= ~NETIF_F_GSO_MASK;
+
br->dev->features = features | checksum | NETIF_F_LLTX |
NETIF_F_GSO_ROBUST;
}
#include <linux/audit.h>
#include <linux/dmaengine.h>
#include <linux/err.h>
+#include <linux/ctype.h>
/*
* The list of packet types we will receive (as opposed to discard)
* @name: name string
*
* Network device names need to be valid file names to
- * to allow sysfs to work
+ * to allow sysfs to work. We also disallow any kind of
+ * whitespace.
*/
int dev_valid_name(const char *name)
{
- return !(*name == '\0'
- || !strcmp(name, ".")
- || !strcmp(name, "..")
- || strchr(name, '/'));
+ if (*name == '\0')
+ return 0;
+ if (!strcmp(name, ".") || !strcmp(name, ".."))
+ return 0;
+
+ while (*name) {
+ if (*name == '/' || isspace(*name))
+ return 0;
+ name++;
+ }
+ return 1;
}
/**
struct net_device *dev = skb->dev;
if (dev->master) {
- /*
- * On bonding slaves other than the currently active
- * slave, suppress duplicates except for 802.3ad
- * ETH_P_SLOW and alb non-mcast/bcast.
- */
- if (dev->priv_flags & IFF_SLAVE_INACTIVE) {
- if (dev->master->priv_flags & IFF_MASTER_ALB) {
- if (skb->pkt_type != PACKET_BROADCAST &&
- skb->pkt_type != PACKET_MULTICAST)
- goto keep;
- }
-
- if (dev->master->priv_flags & IFF_MASTER_8023AD &&
- skb->protocol == __constant_htons(ETH_P_SLOW))
- goto keep;
-
+ if (skb_bond_should_drop(skb)) {
kfree_skb(skb);
return NULL;
}
-keep:
skb->dev = dev->master;
}
static int net_random_reseed(void)
{
int i;
- unsigned long seed[NR_CPUS];
+ unsigned long seed;
- get_random_bytes(seed, sizeof(seed));
for_each_possible_cpu(i) {
struct nrnd_state *state = &per_cpu(net_rand_state,i);
- __net_srandom(state, seed[i]);
+
+ get_random_bytes(&seed, sizeof(seed));
+ __net_srandom(state, seed);
}
return 0;
}
void fib_release_info(struct fib_info *fi)
{
- write_lock(&fib_info_lock);
+ write_lock_bh(&fib_info_lock);
if (fi && --fi->fib_treeref == 0) {
hlist_del(&fi->fib_hash);
if (fi->fib_prefsrc)
fi->fib_dead = 1;
fib_info_put(fi);
}
- write_unlock(&fib_info_lock);
+ write_unlock_bh(&fib_info_lock);
}
static __inline__ int nh_comp(const struct fib_info *fi, const struct fib_info *ofi)
unsigned int old_size = fib_hash_size;
unsigned int i, bytes;
- write_lock(&fib_info_lock);
+ write_lock_bh(&fib_info_lock);
old_info_hash = fib_info_hash;
old_laddrhash = fib_info_laddrhash;
fib_hash_size = new_size;
}
fib_info_laddrhash = new_laddrhash;
- write_unlock(&fib_info_lock);
+ write_unlock_bh(&fib_info_lock);
bytes = old_size * sizeof(struct hlist_head *);
fib_hash_free(old_info_hash, bytes);
fi->fib_treeref++;
atomic_inc(&fi->fib_clntref);
- write_lock(&fib_info_lock);
+ write_lock_bh(&fib_info_lock);
hlist_add_head(&fi->fib_hash,
&fib_info_hash[fib_info_hashfn(fi)]);
if (fi->fib_prefsrc) {
head = &fib_info_devhash[hash];
hlist_add_head(&nh->nh_hash, head);
} endfor_nexthops(fi)
- write_unlock(&fib_info_lock);
+ write_unlock_bh(&fib_info_lock);
return fi;
err_inval:
struct in_device *in_dev;
u32 group = imr->imr_multiaddr.s_addr;
u32 ifindex;
+ int ret = -EADDRNOTAVAIL;
rtnl_lock();
in_dev = ip_mc_find_dev(imr);
- if (!in_dev) {
- rtnl_unlock();
- return -ENODEV;
- }
ifindex = imr->imr_ifindex;
for (imlp = &inet->mc_list; (iml = *imlp) != NULL; imlp = &iml->next) {
- if (iml->multi.imr_multiaddr.s_addr == group &&
- iml->multi.imr_ifindex == ifindex) {
- (void) ip_mc_leave_src(sk, iml, in_dev);
+ if (iml->multi.imr_multiaddr.s_addr != group)
+ continue;
+ if (ifindex) {
+ if (iml->multi.imr_ifindex != ifindex)
+ continue;
+ } else if (imr->imr_address.s_addr && imr->imr_address.s_addr !=
+ iml->multi.imr_address.s_addr)
+ continue;
+
+ (void) ip_mc_leave_src(sk, iml, in_dev);
- *imlp = iml->next;
+ *imlp = iml->next;
+ if (in_dev)
ip_mc_dec_group(in_dev, group);
- rtnl_unlock();
- sock_kfree_s(sk, iml, sizeof(*iml));
- return 0;
- }
+ rtnl_unlock();
+ sock_kfree_s(sk, iml, sizeof(*iml));
+ return 0;
}
+ if (!in_dev)
+ ret = -ENODEV;
rtnl_unlock();
- return -EADDRNOTAVAIL;
+ return ret;
}
int ip_mc_source(int add, int omode, struct sock *sk, struct
struct in_device *in_dev;
inet->mc_list = iml->next;
- if ((in_dev = inetdev_by_index(iml->multi.imr_ifindex)) != NULL) {
- (void) ip_mc_leave_src(sk, iml, in_dev);
+ in_dev = inetdev_by_index(iml->multi.imr_ifindex);
+ (void) ip_mc_leave_src(sk, iml, in_dev);
+ if (in_dev != NULL) {
ip_mc_dec_group(in_dev, iml->multi.imr_multiaddr.s_addr);
in_dev_put(in_dev);
}
sock_kfree_s(sk, iml, sizeof(*iml));
-
}
rtnl_unlock();
}
cb->args[0], *id);
read_lock_bh(&ip_conntrack_lock);
+ last = (struct ip_conntrack *)cb->args[1];
for (; cb->args[0] < ip_conntrack_htable_size; cb->args[0]++) {
restart:
- last = (struct ip_conntrack *)cb->args[1];
list_for_each_prev(i, &ip_conntrack_hash[cb->args[0]]) {
h = (struct ip_conntrack_tuple_hash *) i;
if (DIRECTION(h) != IP_CT_DIR_ORIGINAL)
continue;
ct = tuplehash_to_ctrack(h);
- if (last != NULL) {
- if (ct == last) {
- ip_conntrack_put(last);
- cb->args[1] = 0;
- last = NULL;
- } else
+ if (cb->args[1]) {
+ if (ct != last)
continue;
+ cb->args[1] = 0;
}
if (ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).pid,
cb->nlh->nlmsg_seq,
goto out;
}
}
- if (last != NULL) {
- ip_conntrack_put(last);
+ if (cb->args[1]) {
cb->args[1] = 0;
goto restart;
}
}
out:
read_unlock_bh(&ip_conntrack_lock);
+ if (last)
+ ip_conntrack_put(last);
DEBUGP("leaving, last bucket=%lu id=%u\n", cb->args[0], *id);
-
return skb->len;
}
const char *indev, *outdev;
void *table_base;
struct ipt_entry *e, *back;
- struct xt_table_info *private = table->private;
+ struct xt_table_info *private;
/* Initialization */
ip = (*pskb)->nh.iph;
read_lock_bh(&table->lock);
IP_NF_ASSERT(table->valid_hooks & (1 << hook));
+ private = table->private;
table_base = (void *)private->entries[smp_processor_id()];
e = get_entry(table_base, private->hook_entry[hook]);
return 0;
}
+/*
+ * Special lock-class for __icmpv6_socket:
+ */
+static struct lock_class_key icmpv6_socket_sk_dst_lock_key;
+
int __init icmpv6_init(struct net_proto_family *ops)
{
struct sock *sk;
sk = per_cpu(__icmpv6_socket, i)->sk;
sk->sk_allocation = GFP_ATOMIC;
+ /*
+ * Split off their lock-class, because sk->sk_dst_lock
+ * gets used from softirqs, which is safe for
+ * __icmpv6_socket (because those never get directly used
+ * via userspace syscalls), but unsafe for normal sockets.
+ */
+ lockdep_set_class(&sk->sk_dst_lock,
+ &icmpv6_socket_sk_dst_lock_key);
/* Enough space for 2 64K ICMP packets, including
* sk_buff struct overhead.
if ((dev = dev_get_by_index(mc_lst->ifindex)) != NULL) {
struct inet6_dev *idev = in6_dev_get(dev);
+ (void) ip6_mc_leave_src(sk, mc_lst, idev);
if (idev) {
- (void) ip6_mc_leave_src(sk,mc_lst,idev);
__ipv6_dev_mc_dec(idev, &mc_lst->addr);
in6_dev_put(idev);
}
dev_put(dev);
- }
+ } else
+ (void) ip6_mc_leave_src(sk, mc_lst, NULL);
sock_kfree_s(sk, mc_lst, sizeof(*mc_lst));
return 0;
}
if (dev) {
struct inet6_dev *idev = in6_dev_get(dev);
+ (void) ip6_mc_leave_src(sk, mc_lst, idev);
if (idev) {
- (void) ip6_mc_leave_src(sk, mc_lst, idev);
__ipv6_dev_mc_dec(idev, &mc_lst->addr);
in6_dev_put(idev);
}
dev_put(dev);
- }
+ } else
+ (void) ip6_mc_leave_src(sk, mc_lst, NULL);
sock_kfree_s(sk, mc_lst, sizeof(*mc_lst));
cb->args[0], *id);
read_lock_bh(&nf_conntrack_lock);
+ last = (struct nf_conn *)cb->args[1];
for (; cb->args[0] < nf_conntrack_htable_size; cb->args[0]++) {
restart:
- last = (struct nf_conn *)cb->args[1];
list_for_each_prev(i, &nf_conntrack_hash[cb->args[0]]) {
h = (struct nf_conntrack_tuple_hash *) i;
if (DIRECTION(h) != IP_CT_DIR_ORIGINAL)
* then dump everything. */
if (l3proto && L3PROTO(ct) != l3proto)
continue;
- if (last != NULL) {
- if (ct == last) {
- nf_ct_put(last);
- cb->args[1] = 0;
- last = NULL;
- } else
+ if (cb->args[1]) {
+ if (ct != last)
continue;
+ cb->args[1] = 0;
}
if (ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).pid,
cb->nlh->nlmsg_seq,
goto out;
}
}
- if (last != NULL) {
- nf_ct_put(last);
+ if (cb->args[1]) {
cb->args[1] = 0;
goto restart;
}
}
out:
read_unlock_bh(&nf_conntrack_lock);
+ if (last)
+ nf_ct_put(last);
DEBUGP("leaving, last bucket=%lu id=%u\n", cb->args[0], *id);
-
return skb->len;
}
#include <linux/module.h>
#include <linux/skbuff.h>
+#include <linux/netfilter_bridge.h>
#include <linux/netfilter/xt_physdev.h>
#include <linux/netfilter/x_tables.h>
#include <linux/netfilter_bridge.h>
{
printk("u32 classifier\n");
#ifdef CONFIG_CLS_U32_PERF
- printk(" Perfomance counters on\n");
+ printk(" Performance counters on\n");
#endif
#ifdef CONFIG_NET_CLS_POLICE
printk(" OLD policer on \n");