In 'net' change the explicit use of for-loops and NR_CPUS into the
general for_each_cpu() or for_each_online_cpu() constructs, as
appropriate. This widens the scope of potential future optimizations
of the general constructs, as well as takes advantage of the existing
optimizations of first_cpu() and next_cpu(), which is advantageous
when the true CPU count is much smaller than NR_CPUS.
Signed-off-by: John Hawkes <[email protected]>
Index: linux/net/core/neighbour.c
===================================================================
--- linux.orig/net/core/neighbour.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/core/neighbour.c 2005-10-13 09:30:31.000000000 -0700
@@ -1641,12 +1641,9 @@
memset(&ndst, 0, sizeof(ndst));
- for (cpu = 0; cpu < NR_CPUS; cpu++) {
+ for_each_cpu(cpu) {
struct neigh_statistics *st;
- if (!cpu_possible(cpu))
- continue;
-
st = per_cpu_ptr(tbl->stats, cpu);
ndst.ndts_allocs += st->allocs;
ndst.ndts_destroys += st->destroys;
Index: linux/net/core/pktgen.c
===================================================================
--- linux.orig/net/core/pktgen.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/core/pktgen.c 2005-10-13 09:30:31.000000000 -0700
@@ -3065,12 +3065,9 @@
/* Register us to receive netdevice events */
register_netdevice_notifier(&pktgen_notifier_block);
- for (cpu = 0; cpu < NR_CPUS ; cpu++) {
+ for_each_online_cpu(cpu) {
char buf[30];
- if (!cpu_online(cpu))
- continue;
-
sprintf(buf, "kpktgend_%i", cpu);
pktgen_create_thread(buf, cpu);
}
Index: linux/net/ipv4/icmp.c
===================================================================
--- linux.orig/net/ipv4/icmp.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/ipv4/icmp.c 2005-10-13 09:30:31.000000000 -0700
@@ -1108,12 +1108,9 @@
struct inet_sock *inet;
int i;
- for (i = 0; i < NR_CPUS; i++) {
+ for_each_cpu(i) {
int err;
- if (!cpu_possible(i))
- continue;
-
err = sock_create_kern(PF_INET, SOCK_RAW, IPPROTO_ICMP,
&per_cpu(__icmp_socket, i));
Index: linux/net/ipv4/proc.c
===================================================================
--- linux.orig/net/ipv4/proc.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/ipv4/proc.c 2005-10-13 09:30:31.000000000 -0700
@@ -90,9 +90,7 @@
unsigned long res = 0;
int i;
- for (i = 0; i < NR_CPUS; i++) {
- if (!cpu_possible(i))
- continue;
+ for_each_cpu(i) {
res += *(((unsigned long *) per_cpu_ptr(mib[0], i)) + offt);
res += *(((unsigned long *) per_cpu_ptr(mib[1], i)) + offt);
}
Index: linux/net/ipv6/proc.c
===================================================================
--- linux.orig/net/ipv6/proc.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/ipv6/proc.c 2005-10-13 09:30:31.000000000 -0700
@@ -140,9 +140,7 @@
unsigned long res = 0;
int i;
- for (i = 0; i < NR_CPUS; i++) {
- if (!cpu_possible(i))
- continue;
+ for_each_cpu(i) {
res += *(((unsigned long *)per_cpu_ptr(mib[0], i)) + offt);
res += *(((unsigned long *)per_cpu_ptr(mib[1], i)) + offt);
}
Index: linux/net/ipv6/icmp.c
===================================================================
--- linux.orig/net/ipv6/icmp.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/ipv6/icmp.c 2005-10-13 09:30:31.000000000 -0700
@@ -700,10 +700,7 @@
struct sock *sk;
int err, i, j;
- for (i = 0; i < NR_CPUS; i++) {
- if (!cpu_possible(i))
- continue;
-
+ for_each_cpu(i) {
err = sock_create_kern(PF_INET6, SOCK_RAW, IPPROTO_ICMPV6,
&per_cpu(__icmpv6_socket, i));
if (err < 0) {
@@ -749,9 +746,7 @@
{
int i;
- for (i = 0; i < NR_CPUS; i++) {
- if (!cpu_possible(i))
- continue;
+ for_each_cpu(i) {
sock_release(per_cpu(__icmpv6_socket, i));
}
inet6_del_protocol(&icmpv6_protocol, IPPROTO_ICMPV6);
Index: linux/net/sctp/proc.c
===================================================================
--- linux.orig/net/sctp/proc.c 2005-10-13 09:29:54.000000000 -0700
+++ linux/net/sctp/proc.c 2005-10-13 09:30:31.000000000 -0700
@@ -69,9 +69,7 @@
unsigned long res = 0;
int i;
- for (i = 0; i < NR_CPUS; i++) {
- if (!cpu_possible(i))
- continue;
+ for_each_cpu(i) {
res +=
*((unsigned long *) (((void *) per_cpu_ptr(mib[0], i)) +
sizeof (unsigned long) * nr));
On 10/13/05, [email protected] <[email protected]> wrote:
> In 'net' change the explicit use of for-loops and NR_CPUS into the
> general for_each_cpu() or for_each_online_cpu() constructs, as
> appropriate. This widens the scope of potential future optimizations
> of the general constructs, as well as takes advantage of the existing
> optimizations of first_cpu() and next_cpu(), which is advantageous
> when the true CPU count is much smaller than NR_CPUS.
>
> Signed-off-by: John Hawkes <[email protected]>
Thanks, applied and pushed to Linus,
- Arnaldo