kernel_optimize_test/net/ipv6/sysctl_net_ipv6.c
Tom Herbert cb1ce2ef38 ipv6: Implement automatic flow label generation on transmit
Automatically generate flow labels for IPv6 packets on transmit.
The flow label is computed based on skb_get_hash. The flow label will
only automatically be set when it is zero otherwise (i.e. flow label
manager hasn't set one). This supports the transmit side functionality
of RFC 6438.

Added an IPv6 sysctl auto_flowlabels to enable/disable this behavior
system wide, and added IPV6_AUTOFLOWLABEL socket option to enable this
functionality per socket.

By default, auto flowlabels are disabled to avoid possible conflicts
with flow label manager, however if this feature proves useful we
may want to enable it by default.

It should also be noted that FreeBSD has already implemented automatic
flow labels (including the sysctl and socket option). In FreeBSD,
automatic flow labels default to enabled.

Performance impact:

Running super_netperf with 200 flows for TCP_RR and UDP_RR for
IPv6. Note that in UDP case, __skb_get_hash will be called for
every packet with explains slight regression. In the TCP case
the hash is saved in the socket so there is no regression.

Automatic flow labels disabled:

  TCP_RR:
    86.53% CPU utilization
    127/195/322 90/95/99% latencies
    1.40498e+06 tps

  UDP_RR:
    90.70% CPU utilization
    118/168/243 90/95/99% latencies
    1.50309e+06 tps

Automatic flow labels enabled:

  TCP_RR:
    85.90% CPU utilization
    128/199/337 90/95/99% latencies
    1.40051e+06

  UDP_RR
    92.61% CPU utilization
    115/164/236 90/95/99% latencies
    1.4687e+06

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-07 21:14:21 -07:00

174 lines
4.1 KiB
C

/*
* sysctl_net_ipv6.c: sysctl interface to net IPV6 subsystem.
*
* Changes:
* YOSHIFUJI Hideaki @USAGI: added icmp sysctl table.
*/
#include <linux/mm.h>
#include <linux/sysctl.h>
#include <linux/in6.h>
#include <linux/ipv6.h>
#include <linux/slab.h>
#include <linux/export.h>
#include <net/ndisc.h>
#include <net/ipv6.h>
#include <net/addrconf.h>
#include <net/inet_frag.h>
static struct ctl_table ipv6_table_template[] = {
{
.procname = "bindv6only",
.data = &init_net.ipv6.sysctl.bindv6only,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec
},
{
.procname = "anycast_src_echo_reply",
.data = &init_net.ipv6.sysctl.anycast_src_echo_reply,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec
},
{
.procname = "flowlabel_consistency",
.data = &init_net.ipv6.sysctl.flowlabel_consistency,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec
},
{
.procname = "auto_flowlabels",
.data = &init_net.ipv6.sysctl.auto_flowlabels,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec
},
{
.procname = "fwmark_reflect",
.data = &init_net.ipv6.sysctl.fwmark_reflect,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec
},
{ }
};
static struct ctl_table ipv6_rotable[] = {
{
.procname = "mld_max_msf",
.data = &sysctl_mld_max_msf,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec
},
{ }
};
static int __net_init ipv6_sysctl_net_init(struct net *net)
{
struct ctl_table *ipv6_table;
struct ctl_table *ipv6_route_table;
struct ctl_table *ipv6_icmp_table;
int err;
err = -ENOMEM;
ipv6_table = kmemdup(ipv6_table_template, sizeof(ipv6_table_template),
GFP_KERNEL);
if (!ipv6_table)
goto out;
ipv6_table[0].data = &net->ipv6.sysctl.bindv6only;
ipv6_table[1].data = &net->ipv6.sysctl.anycast_src_echo_reply;
ipv6_table[2].data = &net->ipv6.sysctl.flowlabel_consistency;
ipv6_table[3].data = &net->ipv6.sysctl.auto_flowlabels;
ipv6_route_table = ipv6_route_sysctl_init(net);
if (!ipv6_route_table)
goto out_ipv6_table;
ipv6_icmp_table = ipv6_icmp_sysctl_init(net);
if (!ipv6_icmp_table)
goto out_ipv6_route_table;
net->ipv6.sysctl.hdr = register_net_sysctl(net, "net/ipv6", ipv6_table);
if (!net->ipv6.sysctl.hdr)
goto out_ipv6_icmp_table;
net->ipv6.sysctl.route_hdr =
register_net_sysctl(net, "net/ipv6/route", ipv6_route_table);
if (!net->ipv6.sysctl.route_hdr)
goto out_unregister_ipv6_table;
net->ipv6.sysctl.icmp_hdr =
register_net_sysctl(net, "net/ipv6/icmp", ipv6_icmp_table);
if (!net->ipv6.sysctl.icmp_hdr)
goto out_unregister_route_table;
err = 0;
out:
return err;
out_unregister_route_table:
unregister_net_sysctl_table(net->ipv6.sysctl.route_hdr);
out_unregister_ipv6_table:
unregister_net_sysctl_table(net->ipv6.sysctl.hdr);
out_ipv6_icmp_table:
kfree(ipv6_icmp_table);
out_ipv6_route_table:
kfree(ipv6_route_table);
out_ipv6_table:
kfree(ipv6_table);
goto out;
}
static void __net_exit ipv6_sysctl_net_exit(struct net *net)
{
struct ctl_table *ipv6_table;
struct ctl_table *ipv6_route_table;
struct ctl_table *ipv6_icmp_table;
ipv6_table = net->ipv6.sysctl.hdr->ctl_table_arg;
ipv6_route_table = net->ipv6.sysctl.route_hdr->ctl_table_arg;
ipv6_icmp_table = net->ipv6.sysctl.icmp_hdr->ctl_table_arg;
unregister_net_sysctl_table(net->ipv6.sysctl.icmp_hdr);
unregister_net_sysctl_table(net->ipv6.sysctl.route_hdr);
unregister_net_sysctl_table(net->ipv6.sysctl.hdr);
kfree(ipv6_table);
kfree(ipv6_route_table);
kfree(ipv6_icmp_table);
}
static struct pernet_operations ipv6_sysctl_net_ops = {
.init = ipv6_sysctl_net_init,
.exit = ipv6_sysctl_net_exit,
};
static struct ctl_table_header *ip6_header;
int ipv6_sysctl_register(void)
{
int err = -ENOMEM;
ip6_header = register_net_sysctl(&init_net, "net/ipv6", ipv6_rotable);
if (ip6_header == NULL)
goto out;
err = register_pernet_subsys(&ipv6_sysctl_net_ops);
if (err)
goto err_pernet;
out:
return err;
err_pernet:
unregister_net_sysctl_table(ip6_header);
goto out;
}
void ipv6_sysctl_unregister(void)
{
unregister_net_sysctl_table(ip6_header);
unregister_pernet_subsys(&ipv6_sysctl_net_ops);
}