下面代码描述了在用户层访问smi／mdio总线， 读写phy芯片寄存器的通用代码。Linux内核2.6以上通用。将下面代码编译后，将可执行文件a.out 重命名为mdio

mdio eth0 1 读取phy寄存器1的数值
mdio eth0 0 0x1120 将0x1120写入 phy寄存器1

eth0 为mac层控制器的名称， 一般为eth0 或mgmt0。

#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include

#define reteck(ret) \
if(ret < 0){ \
printf("%m! \"%s\" : line: %d\n", __func__, __LINE__); \
goto lab; \
}

#define help() \
printf("mdio:\n"); \
printf("read operation: mdio reg_addr\n"); \
printf("write operation: mdio reg_addr value\n"); \
printf("For example:\n"); \
printf("mdio eth0 1\n"); \
printf("mdio eth0 0 0x12\n\n"); \
exit(0);

int sockfd;

int main(int argc, char *argv[]){

if(argc == 1 || !strcmp(argv[1], "-h")){
help();
}

struct mii_ioctl_data *mii = NULL;
struct ifreq ifr;
int ret;

memset(&ifr, 0, sizeof(ifr));
strncpy(ifr.ifr_name, argv[1], IFNAMSIZ - 1);

sockfd = socket(PF_LOCAL, SOCK_DGRAM, 0);
reteck(sockfd);

//get phy address in smi bus
ret = ioctl(sockfd, SIOCGMIIPHY, &ifr);
reteck(ret);

mii = (struct mii_ioctl_data*)&ifr.ifr_data;

if(argc == 3){

mii->reg_num = (uint16_t)strtoul(argv[2], NULL, 0);

ret = ioctl(sockfd, SIOCGMIIREG, &ifr);
reteck(ret);

printf("read phy addr: 0x%x reg: 0x%x value : 0x%x\n\n", mii->phy_id, mii->reg_num, mii->val_out);
}else if(argc == 4){

mii->reg_num = (uint16_t)strtoul(argv[2], NULL, 0);
mii->val_in = (uint16_t)strtoul(argv[3], NULL, 0);

ret = ioctl(sockfd, SIOCSMIIREG, &ifr);
reteck(ret);

printf("write phy addr: 0x%x reg: 0x%x value : 0x%x\n\n", mii->phy_id, mii->reg_num, mii->val_in);
}

lab:
close(sockfd);
return 0;
}