PF_Ring性能提升使用说明

pf_ring通过用户态、内核态共享内存达到提高包处理效率的目的。

1、PF_Ring加载的3种模式，模式0和调用libpcap性能相同，模式1和模式2相当于0 coyp性能会提升，官网http://www.ntop.org/products/pf_ring/。

As of PF_RING 4.1, when inserting the pf_ring module it is possible to specify three operational modes:

• Transparent Mode 0: standard NAPI polling
  insmod pf_ring.ko transparent_mode=0
• Transparent Mode 1: PF_RING-aware driver copies the packets into PF_RING, while the same packet is still passed to NAPI
  insmod pf_ring.ko transparent_mode=1
• Transparent Mode 2: PF_RING-aware driver copies the packets into PF_RING, the packet is not passed to NAPI
  insmod pf_ring.ko transparent_mode=2

2、整理一下pf_ring是如何通过mmap来实现共享内存的。

mmap注册

kernel/pf_ring.c：7861

点击(此处)折叠或打开

static struct proto_ops ring_ops = {
.family = PF_RING,
.owner = THIS_MODULE,

/* Operations that make no sense on ring sockets. */
.connect = sock_no_connect,
.socketpair = sock_no_socketpair,
.accept = sock_no_accept,
.getname = sock_no_getname,
.listen = sock_no_listen,
.shutdown = sock_no_shutdown,
.sendpage = sock_no_sendpage,

/* Now the operations that really occur. */
.release = ring_release,
.bind = ring_bind,
.mmap = ring_mmap,
.poll = ring_poll,
.setsockopt = ring_setsockopt,
.getsockopt = ring_getsockopt,
.ioctl = ring_ioctl,
.recvmsg = ring_recvmsg,
.sendmsg = ring_sendmsg,
};

这样，当用户态调用mmap时，将会调用ring_mmap函数。

用户态调用

userland/lib/pfring_mod.c:162

点击(此处)折叠或打开

ring->buffer = (char *)mmap(NULL, PAGE_SIZE, PROT_READ|PROT_WRITE,
MAP_SHARED, ring->fd, 0);

if(ring->buffer == MAP_FAILED) {
printf("mmap() failed: try with a smaller snaplen\n");
close(ring->fd);
return -1;
}

ring->slots_info = (FlowSlotInfo *)ring->buffer;
if(ring->slots_info->version != RING_FLOWSLOT_VERSION) {
printf("Wrong RING version: "
"kernel is %i, libpfring was compiled with %i\n",
ring->slots_info->version, RING_FLOWSLOT_VERSION);
close(ring->fd);
return -1;
}
memSlotsLen = ring->slots_info->tot_mem;
munmap(ring->buffer, PAGE_SIZE);

ring->buffer = (char *)mmap(NULL, memSlotsLen,
PROT_READ|PROT_WRITE,
MAP_SHARED, ring->fd, 0);

/* printf("mmap len %u\n", memSlotsLen); */

if(ring->buffer == MAP_FAILED) {
printf("mmap() failed");
close(ring->fd);
return -1;
}

ring->slots_info = (FlowSlotInfo *)ring->buffer;
ring->slots = (char *)(ring->buffer+sizeof(FlowSlotInfo));

首先通过mmap出一个页，读出ring的相关信息（包括memSlot长度），进而将所有的共享区域全部map出来。

内核态实现

用户态调用了mmap后，内核态对应的实现函数是ring_mmap。

调用链是ring_alloc_mem => ring_alloc_mem => allocate_shared_memory => vmalloc_user

分配完映射的内存后

ring_mmap => do_memory_mmap => remap_vmalloc_range

完成映射过程.