| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Sat, 01 May 2010 07:57:04 +0200
From: Eric Dumazet <eric.dumazet@...il.com>
To: hadi@...erus.ca
Cc: Changli Gao <xiaosuo@...il.com>,
David Miller <davem@...emloft.net>, therbert@...gle.com,
shemminger@...tta.com, netdev@...r.kernel.org,
Eilon Greenstein <eilong@...adcom.com>,
Brian Bloniarz <bmb@...enacr.com>
Subject: Re: [PATCH net-next-2.6] net: speedup udp receive path
Le vendredi 30 avril 2010 à 20:06 -0400, jamal a écrit :
> Yes, Nehalem.
> RPS off is better (~700Kpp) than RPS on(~650kpps). Are you seeing the
> same trend on the old hardware?
>
Of course not ! Or else RPS would be useless :(
I changed your program a bit to use EV_PERSIST, (to avoid epoll_ctl()
overhead for each packet...)
RPS off : 220.000 pps
RPS on (ee mask) : 700.000 pps (with a slightly modified tg3 driver)
96% of delivered packets
This is on tg3 adapter, and tg3 has copybreak feature : small packets
are copied into skb of the right size.
define TG3_RX_COPY_THRESHOLD 256 -> 40 ...
We really should disable this feature for RPS workload,
unfortunatly ethtool cannot tweak this.
So profile of cpu 0 (RPS ON) looks like :
------------------------------------------------------------------------------------------------------------------------
PerfTop: 1001 irqs/sec kernel:99.7% [1000Hz cycles], (all, cpu: 0)
------------------------------------------------------------------------------------------------------------------------
samples pcnt function DSO
_______ _____ ______________________ _______
819.00 12.6% __alloc_skb vmlinux
592.00 9.1% eth_type_trans vmlinux
509.00 7.8% _raw_spin_lock vmlinux
475.00 7.3% __kmalloc_track_caller vmlinux
358.00 5.5% tg3_read32 vmlinux
345.00 5.3% __netdev_alloc_skb vmlinux
329.00 5.0% kmem_cache_alloc vmlinux
307.00 4.7% _raw_spin_lock_irqsave vmlinux
284.00 4.4% bnx2_interrupt vmlinux
277.00 4.2% skb_pull vmlinux
248.00 3.8% tg3_poll_work vmlinux
202.00 3.1% __slab_alloc vmlinux
197.00 3.0% get_rps_cpu vmlinux
106.00 1.6% enqueue_to_backlog vmlinux
87.00 1.3% _raw_spin_lock_bh vmlinux
80.00 1.2% __copy_to_user_ll vmlinux
77.00 1.2% nommu_map_page vmlinux
77.00 1.2% __napi_gro_receive vmlinux
65.00 1.0% tg3_alloc_rx_skb vmlinux
60.00 0.9% skb_gro_reset_offset vmlinux
57.00 0.9% skb_put vmlinux
57.00 0.9% __slab_free vmlinux
/*
* Usage: udpsnkfrk [ -p baseport] nbports
*/
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <string.h>
#include <stdio.h>
#include <errno.h>
#include <unistd.h>
#include <stdlib.h>
#include <fcntl.h>
#include <event.h>
struct worker_data {
struct event *snk_ev;
struct event_base *base;
struct timeval t;
unsigned long pack_count;
unsigned long bytes_count;
unsigned long tout;
int fd; /* move to avoid hole on 64-bit */
int pad1;
unsigned long _padd[99]; /* avoid false sharing */
};
void usage(int code)
{
fprintf(stderr, "Usage: udpsink [-p baseport] nbports\n");
exit(code);
}
void process_recv(int fd, short ev, void *arg)
{
char buffer[4096];
struct sockaddr_in addr;
socklen_t len = sizeof(addr);
struct worker_data *wdata = (struct worker_data *)arg;
int lu = 0;
if (ev == EV_TIMEOUT) {
wdata->tout++;
if ((event_add(wdata->snk_ev, &wdata->t)) < 0) {
perror("cb event_add");
return;
}
} else {
do {
lu = recvfrom(wdata->fd, buffer, sizeof(buffer), 0,
(struct sockaddr *)&addr, &len);
if (lu > 0) {
wdata->pack_count++;
wdata->bytes_count += lu;
}
} while (lu > 0);
}
}
int prep_thread(struct worker_data *wdata)
{
wdata->t.tv_sec = 1;
wdata->t.tv_usec = random() % 50000L;
wdata->base = event_init();
event_set(wdata->snk_ev, wdata->fd, EV_READ|EV_PERSIST, process_recv, wdata);
event_base_set(wdata->base, wdata->snk_ev);
if ((event_add(wdata->snk_ev, &wdata->t)) < 0) {
perror("event_add");
return -1;
}
return 0;
}
void *worker_func(void *arg)
{
struct worker_data *wdata = (struct worker_data *)arg;
return (void *)event_base_loop(wdata->base, 0);
}
int main(int argc, char *argv[])
{
int c;
int baseport = 4000;
int nbthreads;
struct worker_data *wdata;
unsigned long ototal = 0;
int concurrent = 0;
int verbose = 0;
int i;
while ((c = getopt(argc, argv, "cvp:")) != -1) {
if (c == 'p')
baseport = atoi(optarg);
else if (c == 'c')
concurrent = 1;
else if (c == 'v')
verbose++;
else
usage(1);
}
if (optind == argc)
usage(1);
nbthreads = atoi(argv[optind]);
wdata = calloc(sizeof(struct worker_data), nbthreads);
if (!wdata) {
perror("calloc");
return 1;
}
for (i = 0; i < nbthreads; i++) {
struct sockaddr_in addr;
pthread_t tid;
if (i && concurrent) {
wdata[i].fd = wdata[0].fd;
} else {
wdata[i].snk_ev = malloc(sizeof(struct event));
if (!wdata[i].snk_ev)
return 1;
memset(wdata[i].snk_ev, 0, sizeof(struct event));
wdata[i].fd = socket(PF_INET, SOCK_DGRAM, 0);
if (wdata[i].fd == -1) {
free(wdata[i].snk_ev);
perror("socket");
return 1;
}
memset(&addr, 0, sizeof(addr));
addr.sin_family = AF_INET;
// addr.sin_addr.s_addr = inet_addr(argv[optind]);
addr.sin_port = htons(baseport + i);
if (bind
(wdata[i].fd, (struct sockaddr *)&addr,
sizeof(addr)) < 0) {
free(wdata[i].snk_ev);
perror("bind");
return 1;
}
fcntl(wdata[i].fd, F_SETFL, O_NDELAY);
}
if (prep_thread(wdata + i)) {
printf("failed to allocate thread %d, exit\n", i);
exit(0);
}
pthread_create(&tid, NULL, worker_func, wdata + i);
}
for (;;) {
unsigned long total;
long delta;
sleep(1);
total = 0;
for (i = 0; i < nbthreads; i++) {
total += wdata[i].pack_count;
}
delta = total - ototal;
if (delta) {
printf("%lu pps (%lu", delta, total);
if (verbose) {
for (i = 0; i < nbthreads; i++) {
if (wdata[i].pack_count)
printf(" %d:%lu", i,
wdata[i].pack_count);
}
}
printf(")\n");
}
ototal = total;
}
}
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists