OpenCores
URL https://opencores.org/ocsvn/fade_ether_protocol/fade_ether_protocol/trunk

Subversion Repositories fade_ether_protocol

Compare Revisions

  • This comparison shows the changes necessary to convert path 
    /fade_ether_protocol
    from Rev 29 to Rev 30
    ↔ Reverse comparison
  •

Rev 29 → Rev 30

/trunk/experimental_jumbo_frames_version/linux/fpga_l3_fade.c
269,7 → 269,7
memset(my_data,0xa5,MY_ACK_LEN-10);
dev_queue_xmit(newskb);
//Sleep with short timeout, waiting for response
if(wait_event_timeout(usercmd_queue,sd->cmd_ack==2,timeout)) {
if(wait_event_interruptible_timeout(usercmd_queue,sd->cmd_ack==2,timeout)) {
//Response received
//If target buffer provided, copy data to the userspace buffer
if(resp) memcpy(resp,sd->cmd_resp,12);
/trunk/experimental_jumbo_frames_version/linux/receiver2t_cmd.c
0,0 → 1,255
/*
* fpga_l3_fade - driver for L3 communication protocol with FPGA based system
* Copyright (C) 2012 by Wojciech M. Zabolotny
* Institute of Electronic Systems, Warsaw University of Technology
*
* This code is PUBLIC DOMAIN
*/
 
#include<termios.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <sys/ioctl.h>
#include <fcntl.h>
#include <poll.h>
#include <unistd.h>
#include <stdio.h>
#include <strings.h>
#include <stdlib.h>
#include <stdint.h>
#include <endian.h>
#include <sys/socket.h>
#include <linux/serial.h>
#include <sched.h>
#include "fpga_l3_fade.h"
#include <sys/time.h>
#include <pthread.h>
 
int frs[3]={-1,-1,-1};
int active[3]={0,0,0};
int leave[3]={0,0,0};
pthread_t ucmd_thread;
double tend=0.0;
#define nic0 "eth0"
 
void * user_cmd_thread(void * a1)
{
uint16_t cmd;
int i;
int res;
struct timeval tv;
double tstart=0.0;
sleep(5);
for(i=0;i<=2;i++) {
if(active[i]) {
res = ioctl(frs[i],L3_V1_IOC_STARTMAC,NULL);
printf("Result of startmac for slave %d : %d\n",i,res);
}
}
gettimeofday(&tv, NULL);
tstart=tv.tv_sec+1.0e-6*tv.tv_usec;
//Send three commands to each slave
//Waiting 10 seconds
for(cmd=0x112; cmd<=0x114; cmd++) {
sleep(5);
//Send user command
for(i=0;i<=2;i++) {
if(active[i]) {
int j;
struct l3_v1_usercmd uc;
uc.cmd=cmd;
uc.arg=0x56789abc;
uc.timeout=2;
uc.nr_of_retries=20;
res = ioctl(frs[i],L3_V1_IOC_USERCMD,&uc);
printf("Result of usercmd for slave %d : %d\n",i,res);
for(j=0;j<12;j++) {
printf("%2.2x",(int)(uc.resp[j]));
}
printf("\n");
}
}
}
//Now check, how long it will take to execute certain number of
//user commands!
{
struct l3_v1_usercmd uc;
int j;
int res;
int n_of_tries=100000;
double tcmd1=0.0,tcmd2=0.0;
gettimeofday(&tv, NULL);
tcmd1=tv.tv_sec+1.0e-6*tv.tv_usec;
for(j=0;j<n_of_tries;j++) {
for(i=0;i<=2;i++) {
if(active[i]) {
uc.cmd=0x113;
uc.arg=0x56789abc;
uc.timeout=2;
uc.nr_of_retries=20;
res = ioctl(frs[i],L3_V1_IOC_USERCMD,&uc);
if(res<0) {
printf("Error in user command: %d in %d call\n", res, j);
exit(1);
}
}
}
}
gettimeofday(&tv, NULL);
tcmd2=tv.tv_sec+1.0e-6*tv.tv_usec-tcmd1;
printf("%d tries in %g seconds, %g commands per second\n",n_of_tries,tcmd2, n_of_tries/tcmd2);
}
//Wait for 500 seconds
sleep(500);
for(i=0;i<=2;i++) {
if(active[i]) {
res = ioctl(frs[i],L3_V1_IOC_STOPMAC,NULL);
printf("Result of usercmd for slave %d : %d\n",i,res);
}
}
gettimeofday(&tv, NULL);
tend=tv.tv_sec+1.0e-6*tv.tv_usec;
tend=tend-tstart;
}
void main(int argc, char * argv[])
{
struct l3_v1_buf_pointers bp;
struct l3_v1_slave sl[3] = {
{
.mac = {0xde, 0xad, 0xba, 0xbe, 0xbe,0xef},
.devname = nic0
},
{
.mac = {0xde, 0xad, 0xba, 0xbe, 0xbe,0xe1},
.devname = nic0
},
{
.mac = {0xde, 0xad, 0xba, 0xbe, 0xbe,0xe2},
.devname = nic0
}
};
int i,j;
int res;
int blen[3];
uint64_t data[3] = {0,0,0};
long long total_len[3]={0,0,0};
unsigned char * v[3];
int stop;
struct sched_param s;
s.sched_priority = 90;
//Read active channels
for(i=1;i<argc;i++) {
int n = atoi(argv[i]);
if ((n>=0) && (n<=2))
active[n]=1;
}
printf("sched=%d\n",sched_setscheduler(0,SCHED_RR,&s));
//Prepare all slaves to work
for(i=0;i<3;i++) {
char devname[30];
sprintf(devname,"/dev/l3_fpga%d",i);
frs[i]=open(devname,O_RDONLY);
if(frs[i]<0) {
printf("I can't open device %s\b",devname);
perror("");
exit(1);
}
//Get the length of the buffer
blen[i] = ioctl(frs[i],L3_V1_IOC_GETBUFLEN,NULL);
//Set the wakeup threshold
res=ioctl(frs[i],L3_V1_IOC_SETWAKEUP,2000000);
printf("length of buffer: %d, result of set wakeup: %d\n",blen[i],res);
v[i]=(unsigned char *)mmap(0,blen[i],PROT_READ,MAP_PRIVATE,frs[i],0);
if(!v[i]) {
printf("mmap for device %s failed\n",devname);
exit(1);
}
}
//Connect to appropriate FPGAs
for(i=0;i<=2;i++) {
if(active[i]) {
res = ioctl(frs[i],L3_V1_IOC_GETMAC,&sl[i]);
printf("Result of get for slave %d : %d\n",i,res);
if(res<0) {
printf("I couldn't bind Ethernet device %s\n",sl[i].devname);
exit(1);
}
} else {
leave[i]=1;
}
}
//Send FADE reset command, so that FADE core is in predictable state
for(i=0;i<=2;i++) {
if(active[i]) {
int j;
for(j=0;j<3;j++) {
res = ioctl(frs[i],L3_V1_IOC_RESETMAC,&sl[i]);
if(res<0) {
printf("I couldn't RESET Ethernet device %d\n",i);
exit(1);
}
}
} else {
leave[i]=1;
}
}
//Start the second thread, which uses user commands
pthread_create(&ucmd_thread, NULL, user_cmd_thread, NULL);
int first_served=0;
do{
struct pollfd pfd[3];
int ptr=0;
int len=0;
int pres;
for(i=0;i<3;i++) {
pfd[i].fd = active[i] ? frs[i] : -1;
pfd[i].events = POLLIN;
pfd[i].revents = 0;
}
//Wait for data using "poll"
pres = poll(pfd,3,1000);
if(pres<0) {
perror("Error in poll:");
exit(1);
}
first_served = (first_served+1) %3; //Rotate priority of slaves
for(j=0;j<3;j++) {
i=(j+first_served) % 3;
if(pfd[i].revents) {
int ofs=0;
len = ioctl(frs[i],L3_V1_IOC_READPTRS,&bp);
if(bp.eof) leave[i]=1;
//OK. The data are read, let's analyze them
while (bp.head != bp.tail) {
uint64_t c;
c = be64toh(*(uint64_t *)(v[i]+bp.tail));
bp.tail=(bp.tail+8) & (blen[i]-1); //Adjust tail pointer modulo blen[i]
if (__builtin_expect((c != data[i]), 0)) {
printf("Error! received: %llx expected: %llx position: %8.8x\n",c,data[i],total_len[i]+ofs);
exit(1);
}
data[i] += 0x1234567809abcdefL;
ofs++;
}
total_len[i] += len;
//printf("i=%d len=%d total=%lld head:%d tail: %d revents=%d eof=%d\n",i,len,total_len[i],bp.head, bp.tail,pfd[i].revents,(int)bp.eof);
ioctl(frs[i],L3_V1_IOC_WRITEPTRS,len);
}
}
fflush(stdout);
} while ((leave[0] && leave[1] && leave[2]) == 0);
pthread_join(ucmd_thread, NULL);
for(i=0;i<=2;i++) {
if(active[i]) {
res = ioctl(frs[i],L3_V1_IOC_FREEMAC,0);
}
munmap(v[i],blen[i]);
close(frs[i]);
}
fprintf(stderr,"act0:%d act1:%d act2:%d\n",active[0],active[1],active[2]);
for(i=0;i<3;i++) {
fprintf(stderr,"total data %d=%lld time=%g throughput=%g [Mb/s]\n",i,total_len[i], tend, total_len[i]/tend*8.0);
}
}

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.