/
l3switch_rx.c
407 lines (375 loc) · 12.4 KB
/
l3switch_rx.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
#include"l3switch_main.h"
#define ETH_TYPE_ARP 0x0806
#define ETH_TYPE_IP 0x0800
#define ETH_TYPE_IPV6 0x86dd //we decap here,if IPv6 is supported,we enter ipv6 module for details
extern struct port_usr_para_conf gPortParaConf[MAX_PORT_NB];
extern struct lcore_usr_para_conf gLcoreParaConf[MAX_LCORE_NB];
extern struct netentry gNetEntry[MAX_NETENTRY_ENTRY_NB];
extern struct mac_ip_exception gMACIPExcept[MAX_MAC_IP_EXCEPTION_LENGTH];
extern struct net_flow_item *gFlow;
extern enum PORT_POLICY_MAP_CONST port_policy_map[MAX_PORT_NB][MAX_PORT_NB];
extern int sched_mod_map[MAX_PORT_NB][MAX_PORT_NB];
extern struct qos_sched_mod sched_mod_list[MAX_SCHED_MOD_NB];
extern int port_brdcst_map[MAX_PORT_NB][MAX_PORT_NB];
extern int port_def_frd_map[MAX_PORT_NB];
extern struct rte_mempool* gmempool;
extern int gNBPort;
extern int gNBFlow;
extern int gNBMACExcept;
/*
functional description:
layer-2 decapsulate to decide which mod should recv it,
input mod:any
output mod:
RX_MOD_ARP
RX_MOD_IP
RX_MOD_IPv6
RX_MOD_DROP:we do not deal with other kind of pkt
mod stack pos:first module
date :2014-05-10
author:jzheng
*/
dbg_local enum RX_MOD_INDEX rx_module_l2decap(dbg_local struct rte_mbuf* pktbuf,dbg_local enum RX_MOD_INDEX imodid)
{
enum RX_MOD_INDEX nextmodid=RX_MOD_DROP;//default mod is droping
struct ether_hdr*eh;
uint16_t eth_type;
int iport_in;//for RX statistic
iport_in=pktbuf->pkt.in_port;
rte_atomic64_inc(&gPortParaConf[iport_in].sd_port.ui_cnt_rx);
rte_atomic64_add(&gPortParaConf[iport_in].sd_port.ui_byte_rx,pktbuf->pkt.pkt_len-sizeof(struct ether_hdr));
rte_atomic64_add(&gPortParaConf[iport_in].sd_port.ui_cnt_timer_rx,pktbuf->pkt.pkt_len-sizeof(struct ether_hdr));
eh=rte_pktmbuf_mtod(pktbuf,struct ether_hdr*);
eth_type=HTONS(eh->ether_type);
if(eth_type==ETH_TYPE_ARP)
nextmodid=RX_MOD_ARP;
else if(eth_type==ETH_TYPE_IP)
nextmodid=RX_MOD_IP;
else if(eth_type==ETH_TYPE_IPV6)
nextmodid=RX_MOD_IPv6;
else
nextmodid=RX_MOD_DROP;
//printf("flag:%d %04x\n",nextmodid,eth_type);
return nextmodid;
}
/*
functional description:
input mod:RX_MOD_ARP
output mod:
RX_MOD_DROP
RX_MOD_FRWARD
mod stack pos:
below l2decap-mod
above any action-mod
date :2014-05-14
author:jzheng
*/
dbg_local enum RX_MOD_INDEX rx_module_arp(dbg_local struct rte_mbuf*pktbuf,dbg_local enum RX_MOD_INDEX imodid)
{//check arp request /response format and content legality,we just check ,not act as ARP proxy though we do snoop mac-ip-port tables,and we even learn mac-ip mapping even when depoyed in transparent environment
dbg_local enum RX_MOD_INDEX nextmodid=imodid;
dbg_local struct ether_hdr * eh;
dbg_local struct ether_arp * ea;
dbg_local int idx;
dbg_local int iFlowIdx;
if(imodid!=RX_MOD_ARP)
goto local_ret;
eh=rte_pktmbuf_mtod(pktbuf,struct ether_hdr*);
ea=(struct ether_arp*)(sizeof(struct ether_hdr)+(char*)eh);
uint32_t uip_src=HTONL(MAKEUINT32FROMUINT8ARR(ea->arp_spa));//trans to little endian
// printf("flag:%08x,,%x\n",uip_src,pktbuf->pkt.in_port);
switch(HTONS(ea->ea_hdr.ar_op))
{
case 0x0001://active learning
nextmodid=RX_MOD_FORWARD;
if(!IS_MAC_BROADCAST(eh->d_addr.addr_bytes))//disable uni-cast arp request packet,because rarely no such formated packet will be sent into an Ethernet
goto exception_tag;
if(!IS_MAC_EQUAL(eh->s_addr.addr_bytes,ea->arp_sha))
goto exception_tag;
#if 0
for(idx=0;idx<gNBMACExcept;idx++)//check mandatory mac-ip entrusted domain white list
if(uip_src==gMACIPExcept[idx].uip){
if((!IS_MAC_EQUAL(ea->arp_sha,gMACIPExcept[idx].uimac))||(pktbuf->pkt.in_port!=gMACIPExcept[idx].uiport))
goto exception_tag;
break;
}
#endif
iFlowIdx=find_net_entry(uip_src,1);
if(iFlowIdx==-1)
goto exception_tag;
#if 0
if(gFlow[iFlowIdx].uiInnerIP!=uip_src)//redundant ip check
goto exception_tag;
#endif
if(gFlow[iFlowIdx].b_flow_enabled==FALSE)
goto exception_tag;
if(gFlow[iFlowIdx].b_mac_learned==FALSE){
if(gFlow[iFlowIdx].b_dynamic){//can be update for dynamic at 1st time
COPYMAC(gFlow[iFlowIdx].eaHostMAC.addr_bytes,ea->arp_sha);
gFlow[iFlowIdx].portid=pktbuf->pkt.in_port;
gFlow[iFlowIdx].b_mac_learned=TRUE;
}
}else{
if((!IS_MAC_EQUAL(gFlow[iFlowIdx].eaHostMAC.addr_bytes,ea->arp_sha))||(gFlow[iFlowIdx].portid!=pktbuf->pkt.in_port)){//conflict detected,warning will be raised
if(gFlow[iFlowIdx].b_preempty_enabled){
COPYMAC(gFlow[iFlowIdx].eaHostMAC.addr_bytes,ea->arp_sha);
gFlow[iFlowIdx].portid=pktbuf->pkt.in_port;
}else{//if preempty policy is disabled,,this arp request learning do not take any signifinance
if(gFlow[iFlowIdx].b_dynamic==FALSE)//and if IP to be protected it is,,then ARP spoofing detected,,log and warning
goto exception_tag;//we discard this kind of packet
}
}
}
break;
case 0x0002://we also learning mapping-relationship from arp-response packet
nextmodid=RX_MOD_FORWARD;
#if 0
if(IS_MAC_BROADCAST(eh->s_addr.addr_bytes))
goto exception_tag;
#endif
if(IS_MAC_BROADCAST(eh->d_addr.addr_bytes))
goto exception_tag;
if(!IS_MAC_EQUAL(eh->s_addr.addr_bytes,ea->arp_sha))
goto exception_tag;
if(!IS_MAC_EQUAL(eh->d_addr.addr_bytes,ea->arp_tha))
goto exception_tag;
iFlowIdx=find_net_entry(uip_src,1);
if(iFlowIdx==-1)
goto exception_tag;
if(gFlow[iFlowIdx].b_flow_enabled==FALSE)
goto exception_tag;
if(gFlow[iFlowIdx].b_mac_learned==FALSE){
if(gFlow[iFlowIdx].b_dynamic){
COPYMAC(gFlow[iFlowIdx].eaHostMAC.addr_bytes,ea->arp_sha);
gFlow[iFlowIdx].portid=pktbuf->pkt.in_port;
gFlow[iFlowIdx].b_mac_learned=TRUE;
}
}else{
if((!IS_MAC_EQUAL(gFlow[iFlowIdx].eaHostMAC.addr_bytes,ea->arp_sha))||(gFlow[iFlowIdx].portid!=pktbuf->pkt.in_port)){
if(gFlow[iFlowIdx].b_preempty_enabled){
COPYMAC(gFlow[iFlowIdx].eaHostMAC.addr_bytes,ea->arp_sha);
gFlow[iFlowIdx].portid=pktbuf->pkt.in_port;
}else{
if(gFlow[iFlowIdx].b_dynamic==FALSE)
goto exception_tag;
}
}
}
break;
default:
goto exception_tag;
break;
}
local_ret:
return nextmodid;
exception_tag://here we set next module id
nextmodid=RX_MOD_DROP;
goto local_ret;
}
/*
functional description:
IP RX module check,even we do nothing here
input mod:
RX_MOD_IP
output mod:
RX_MOD_DROP
RX_MOD_FRWARD
module stack pos:
below l2decap-mod
above any action-mod
date :2014-05-28
author:jzheng
*/
dbg_local enum RX_MOD_INDEX rx_module_ip(dbg_local struct rte_mbuf*pktbuf,dbg_local enum RX_MOD_INDEX imodid)
{
dbg_local enum RX_MOD_INDEX nextmodid=imodid;
if(imodid!=RX_MOD_IP)
goto local_ret;
nextmodid=RX_MOD_FORWARD;
//for now we do not filter such pkts with their invalid dst mac-ip ,
local_ret:
return nextmodid;
}
/*
functional description:
reservred IPv6 Entry Interface,actually we just forward or block such IPv6 Packets
input mod:
RX_MOD_IPv6
output mod:
RX_MOD_DROP
RX_MOD_FRWARD
module stack pos:
below l2decap-mod
above any action-mod
date :2014-05-28
author:jzheng
*/
dbg_local enum RX_MOD_INDEX rx_module_ipv6(dbg_local struct rte_mbuf*pktbuf,dbg_local enum RX_MOD_INDEX imodid)
{
dbg_local enum RX_MOD_INDEX nextmodid=imodid;
if(imodid!=RX_MOD_IPv6)
goto local_ret;
nextmodid=RX_MOD_DROP;
local_ret:
return nextmodid;
}
/*
functional description:
packet forward(enqueue) module,we push packet into QoS queue or TX queue
input mod:
RX_MOD_FORWARD
output mod:
RX_MOD_DROP
RX_MOD_IDLE
module stack pos:
as 1st action ,before drop module,and below any policy module
*/
dbg_local enum RX_MOD_INDEX rx_module_forward(dbg_local struct rte_mbuf*pktbuf,dbg_local enum RX_MOD_INDEX imodid)
{
dbg_local int iPortIn;
dbg_local int iPortOut;
dbg_local int iFlowIdx;
dbg_local int iDstIdx;
dbg_local int uiSIP;
dbg_local int uiDIP;
dbg_local int idx;
dbg_local int rc;
dbg_local int iUsed;
dbg_local int iMod;
dbg_local struct ether_hdr *eh;
dbg_local struct ether_arp *ea;
dbg_local struct iphdr *iph;
dbg_local int iPortForwardFlag[MAX_PORT_NB];
dbg_local struct rte_mbuf*rmPortForwardMbuf[MAX_PORT_NB];
dbg_local enum RX_MOD_INDEX nextmodid=imodid;
if(imodid!=RX_MOD_FORWARD)
goto local_ret;
//phaze 1,we filter packet with illegal l2 and l3 address,and drop it
//printf("input port:%d\n",(int)pktbuf->pkt.in_port);
eh=rte_pktmbuf_mtod(pktbuf,struct ether_hdr*);
uiSIP=0xffffffff;
switch(HTONS(eh->ether_type))
{
case ETH_TYPE_ARP:
ea=(struct ether_arp *)(sizeof(struct ether_hdr)+(char*)eh);
uiSIP=HTONL(MAKEUINT32FROMUINT8ARR(ea->arp_spa));
uiDIP=HTONL(MAKEUINT32FROMUINT8ARR(ea->arp_tpa));
break;
case ETH_TYPE_IP:
iph=(struct iphdr *)(sizeof(struct ether_hdr)+(char*)eh);
uiSIP=HTONL(iph->ip_src);
uiDIP=HTONL(iph->ip_dst);
break;
default:
goto exception_tag;
break;
}
iFlowIdx=find_net_entry(uiSIP,TRUE);
//pkt source legality checking
if(iFlowIdx==-1)
goto exception_tag;
if(gFlow[iFlowIdx].b_mac_learned==FALSE)
goto exception_tag;
if(gFlow[iFlowIdx].b_flow_enabled==FALSE)
goto exception_tag;
if(gFlow[iFlowIdx].portid!=pktbuf->pkt.in_port)
goto exception_tag;
if(!IS_MAC_EQUAL(eh->s_addr.addr_bytes,gFlow[iFlowIdx].eaHostMAC.addr_bytes))
goto exception_tag;
iDstIdx=find_net_entry(uiDIP,TRUE);
/*printf("%02x,%02x,%02x,%02x,%02x,%02x,\n",eh->d_addr.addr_bytes[0]
,eh->d_addr.addr_bytes[1]
,eh->d_addr.addr_bytes[2]
,eh->d_addr.addr_bytes[3]
,eh->d_addr.addr_bytes[4]
,eh->d_addr.addr_bytes[5]
);*/
if(IS_MAC_BROADCAST(eh->d_addr.addr_bytes)){//link broadcast
iUsed=FALSE;
for(idx=0;idx<gNBPort;idx++)
if(port_brdcst_map[(int)pktbuf->pkt.in_port][idx]==TRUE){
iPortForwardFlag[idx]=1;
if(iUsed==FALSE){
rmPortForwardMbuf[idx]=pktbuf;
iUsed=TRUE;
}else{
rmPortForwardMbuf[idx]=rte_pktmbuf_clone(pktbuf,gmempool);
if(!rmPortForwardMbuf[idx])
iPortForwardFlag[idx]=0;
}
}else iPortForwardFlag[idx]=0;
}else{//link uicast
for(idx=0;idx<gNBPort;idx++)
iPortForwardFlag[idx]=0;
if(iDstIdx!=-1){
// printf(".......idstidx!=-1.%08x\n",gFlow[iDstIdx].uiInnerIP);
if(gFlow[iDstIdx].b_mac_learned==FALSE)
goto exception_tag;
// printf("Heror1\n");
if(gFlow[iDstIdx].b_flow_enabled==FALSE)
goto exception_tag;
// printf("Heror2\n");
iPortForwardFlag[(int)gFlow[iDstIdx].portid]=1;
rmPortForwardMbuf[(int)gFlow[iDstIdx].portid]=pktbuf;
}
else {
// printf(".......idstidx==-1\n");
if(port_def_frd_map[(int)pktbuf->pkt.in_port]!=-1){
iPortForwardFlag[port_def_frd_map[(int)pktbuf->pkt.in_port]]=1;
rmPortForwardMbuf[port_def_frd_map[(int)pktbuf->pkt.in_port]]=pktbuf;
}
else goto exception_tag;
}
}
//mirror--pkt--cloning
for(idx=0;idx<gNBPort;idx++)
{
if(!iPortForwardFlag[idx])continue;
switch(port_policy_map[(int)pktbuf->pkt.in_port][idx])
{
case PORT_POLICY_MAP_DIRECT:
//printf("...direct:%d->%d\n",(int)pktbuf->pkt.in_port,idx);
rc=EnqueueIntoPortQueue(idx,&rmPortForwardMbuf[idx],1);
//printf("...rc:%d\n",rc);
if(!rc) goto exception_tag;
break;
case PORT_POLICY_MAP_QOS:
//high nibble:dst port
//low nibble :src port
iMod=sched_mod_map[(int)rmPortForwardMbuf[idx]->pkt.in_port][idx];
rmPortForwardMbuf[idx]->pkt.in_port=MAKEBYTE(idx,rmPortForwardMbuf[idx]->pkt.in_port);
rc=rte_ring_mp_enqueue(sched_mod_list[iMod].rrFirLev,rmPortForwardMbuf[idx]);
if(rc==-ENOBUFS){
goto exception_tag;
}
break;
case PORT_POLICY_MAP_UNDEFINE:
goto exception_tag;
break;
}
}
local_ret:
return nextmodid;
exception_tag:
nextmodid=RX_MOD_DROP;
goto local_ret;
}
/*
functional description:
pkt which not fulfil forwarding rules will be drop here ,and sys rc will deallocated
input mod:RX_MOD_DROP
output mod:RX_MOD_IDLE
module stack pos:last module
date :2014-05-10
author:jzheng
*/
dbg_local enum RX_MOD_INDEX rx_module_drop(dbg_local struct rte_mbuf*pktbuf,dbg_local enum RX_MOD_INDEX imodid)
{
dbg_local enum RX_MOD_INDEX nextmodid=RX_MOD_IDLE;//default mod id can be assigned to any mod ,as this mod is last one
int iport_in;
if(imodid!=RX_MOD_DROP)//check entry legality
goto local_ret;
iport_in=pktbuf->pkt.in_port;
rte_pktmbuf_free(pktbuf);
local_ret:
return nextmodid;
}