fedora core 6 1.2949 + vserver 2.2.0

[linux-2.6.git] / net / ipv4 / tcp_cong.c

diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c
index e688c68..5ca7723 100644 (file)
--- a/net/ipv4/tcp_cong.c
+++ b/net/ipv4/tcp_cong.c
@@ -6,7 +6,6 @@
  * Copyright (C) 2005 Stephen Hemminger <shemminger@osdl.org>
  */
 
-#include <linux/config.h>
 #include <linux/module.h>
 #include <linux/mm.h>
 #include <linux/types.h>
@@ -38,7 +37,7 @@ int tcp_register_congestion_control(struct tcp_congestion_ops *ca)
        int ret = 0;
 
        /* all algorithms must implement ssthresh and cong_avoid ops */
-       if (!ca->ssthresh || !ca->cong_avoid || !ca->min_cwnd) {
+       if (!ca->ssthresh || !ca->cong_avoid) {
                printk(KERN_ERR "TCP %s does not implement required ops\n",
                       ca->name);
                return -EINVAL;
@@ -49,7 +48,7 @@ int tcp_register_congestion_control(struct tcp_congestion_ops *ca)
                printk(KERN_NOTICE "TCP %s already registered\n", ca->name);
                ret = -EEXIST;
        } else {
-               list_add_rcu(&ca->list, &tcp_cong_list);
+               list_add_tail_rcu(&ca->list, &tcp_cong_list);
                printk(KERN_INFO "TCP %s registered\n", ca->name);
        }
        spin_unlock(&tcp_cong_list_lock);
@@ -114,7 +113,7 @@ int tcp_set_default_congestion_control(const char *name)
        spin_lock(&tcp_cong_list_lock);
        ca = tcp_ca_find(name);
 #ifdef CONFIG_KMOD
-       if (!ca) {
+       if (!ca && capable(CAP_SYS_MODULE)) {
                spin_unlock(&tcp_cong_list_lock);
 
                request_module("tcp_%s", name);
@@ -124,6 +123,7 @@ int tcp_set_default_congestion_control(const char *name)
 #endif
 
        if (ca) {
+               ca->non_restricted = 1; /* default is always allowed */
                list_move(&ca->list, &tcp_cong_list);
                ret = 0;
        }
@@ -132,6 +132,30 @@ int tcp_set_default_congestion_control(const char *name)
        return ret;
 }
 
+/* Set default value from kernel configuration at bootup */
+static int __init tcp_congestion_default(void)
+{
+       return tcp_set_default_congestion_control(CONFIG_DEFAULT_TCP_CONG);
+}
+late_initcall(tcp_congestion_default);
+
+
+/* Build string with list of available congestion control values */
+void tcp_get_available_congestion_control(char *buf, size_t maxlen)
+{
+       struct tcp_congestion_ops *ca;
+       size_t offs = 0;
+
+       rcu_read_lock();
+       list_for_each_entry_rcu(ca, &tcp_cong_list, list) {
+               offs += snprintf(buf + offs, maxlen - offs,
+                                "%s%s",
+                                offs == 0 ? "" : " ", ca->name);
+
+       }
+       rcu_read_unlock();
+}
+
 /* Get current default congestion control */
 void tcp_get_default_congestion_control(char *name)
 {
@@ -145,6 +169,64 @@ void tcp_get_default_congestion_control(char *name)
        rcu_read_unlock();
 }
 
+/* Built list of non-restricted congestion control values */
+void tcp_get_allowed_congestion_control(char *buf, size_t maxlen)
+{
+       struct tcp_congestion_ops *ca;
+       size_t offs = 0;
+
+       *buf = '\0';
+       rcu_read_lock();
+       list_for_each_entry_rcu(ca, &tcp_cong_list, list) {
+               if (!ca->non_restricted)
+                       continue;
+               offs += snprintf(buf + offs, maxlen - offs,
+                                "%s%s",
+                                offs == 0 ? "" : " ", ca->name);
+
+       }
+       rcu_read_unlock();
+}
+
+/* Change list of non-restricted congestion control */
+int tcp_set_allowed_congestion_control(char *val)
+{
+       struct tcp_congestion_ops *ca;
+       char *clone, *name;
+       int ret = 0;
+
+       clone = kstrdup(val, GFP_USER);
+       if (!clone)
+               return -ENOMEM;
+
+       spin_lock(&tcp_cong_list_lock);
+       /* pass 1 check for bad entries */
+       while ((name = strsep(&clone, " ")) && *name) {
+               ca = tcp_ca_find(name);
+               if (!ca) {
+                       ret = -ENOENT;
+                       goto out;
+               }
+       }
+
+       /* pass 2 clear */
+       list_for_each_entry_rcu(ca, &tcp_cong_list, list)
+               ca->non_restricted = 0;
+
+       /* pass 3 mark as allowed */
+       while ((name = strsep(&val, " ")) && *name) {
+               ca = tcp_ca_find(name);
+               WARN_ON(!ca);
+               if (ca)
+                       ca->non_restricted = 1;
+       }
+out:
+       spin_unlock(&tcp_cong_list_lock);
+
+       return ret;
+}
+
+
 /* Change congestion control for socket */
 int tcp_set_congestion_control(struct sock *sk, const char *name)
 {
@@ -154,12 +236,25 @@ int tcp_set_congestion_control(struct sock *sk, const char *name)
 
        rcu_read_lock();
        ca = tcp_ca_find(name);
+       /* no change asking for existing value */
        if (ca == icsk->icsk_ca_ops)
                goto out;
 
+#ifdef CONFIG_KMOD
+       /* not found attempt to autoload module */
+       if (!ca && capable(CAP_SYS_MODULE)) {
+               rcu_read_unlock();
+               request_module("tcp_%s", name);
+               rcu_read_lock();
+               ca = tcp_ca_find(name);
+       }
+#endif
        if (!ca)
                err = -ENOENT;
 
+       else if (!(ca->non_restricted || capable(CAP_NET_ADMIN)))
+               err = -EPERM;
+
        else if (!try_module_get(ca->owner))
                err = -EBUSY;
 
@@ -190,7 +285,7 @@ void tcp_slow_start(struct tcp_sock *tp)
                        return;
 
                /* We MAY increase by 2 if discovered delayed ack */
-               if (sysctl_tcp_abc > 1 && tp->bytes_acked > 2*tp->mss_cache) {
+               if (sysctl_tcp_abc > 1 && tp->bytes_acked >= 2*tp->mss_cache) {
                        if (tp->snd_cwnd < tp->snd_cwnd_clamp)
                                tp->snd_cwnd++;
                }
@@ -223,7 +318,7 @@ void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 rtt, u32 in_flight,
 
        /* In dangerous area, increase slowly. */
        else if (sysctl_tcp_abc) {
-               /* RFC3465: Apppriate Byte Count
+               /* RFC3465: Appropriate Byte Count
                 * increase once for each full cwnd acked
                 */
                if (tp->bytes_acked >= tp->snd_cwnd*tp->mss_cache) {
@@ -251,8 +346,8 @@ u32 tcp_reno_ssthresh(struct sock *sk)
 }
 EXPORT_SYMBOL_GPL(tcp_reno_ssthresh);
 
-/* Lower bound on congestion window. */
-u32 tcp_reno_min_cwnd(struct sock *sk)
+/* Lower bound on congestion window with halving. */
+u32 tcp_reno_min_cwnd(const struct sock *sk)
 {
        const struct tcp_sock *tp = tcp_sk(sk);
        return tp->snd_ssthresh/2;
@@ -261,6 +356,7 @@ EXPORT_SYMBOL_GPL(tcp_reno_min_cwnd);
 
 struct tcp_congestion_ops tcp_reno = {
        .name           = "reno",
+       .non_restricted = 1,
        .owner          = THIS_MODULE,
        .ssthresh       = tcp_reno_ssthresh,
        .cong_avoid     = tcp_reno_cong_avoid,