-
Notifications
You must be signed in to change notification settings - Fork 2
Expand file tree
/
Copy pathmain.c
More file actions
executable file
·1119 lines (853 loc) · 23.4 KB
/
main.c
File metadata and controls
executable file
·1119 lines (853 loc) · 23.4 KB
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
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*
* XenLoop -- A High Performance Inter-VM Network Loopback
*
* Installation and Usage instructions
*
* Authors:
* Jian Wang - Binghamton University ([email protected])
* Kartik Gopalan - Binghamton University ([email protected])
*
* Copyright (C) 2007-2009 Kartik Gopalan, Jian Wang
*
* Permission is hereby granted, free of charge, to any person
* obtaining a copy of this software and associated documentation
* files (the "Software"), to deal in the Software without
* restriction, including without limitation the rights to use,
* copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following
* conditions:
*
* The above copyright notice and this permission notice shall be
* included in all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
* OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
* HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
* WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*/
#include <linux/init.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/genhd.h>
#include <asm/xen/hypercall.h>
#include <xen/grant_table.h>
#include <xen/events.h>
#include <xen/xenbus.h>
#include <net/sock.h>
#include <net/inet_sock.h>
#include <net/protocol.h>
#include <linux/in.h>
#include <linux/skbuff.h>
#include <net/neighbour.h>
#include <net/dst.h>
#include <linux/if_ether.h>
#include <net/inet_common.h>
#include <linux/inetdevice.h>
#include <linux/mm.h>
#include <linux/time.h>
#include <linux/genhd.h>
#include <linux/netfilter.h>
#include <linux/netfilter_ipv4.h>
#include <linux/kthread.h>
#include <linux/wait.h>
#include <linux/timer.h>
#include <linux/spinlock.h>
#include "main.h"
#include "debug.h"
#include "bififo.h"
#include "maptable.h"
#include <linux/if_arp.h>
#include <uapi/linux/netfilter_arp.h>
extern int init_hash_table(HashTable *, char *);
extern void clean_table(HashTable *);
extern void insert_table(HashTable *, void *, u8);
extern void* lookup_table(HashTable *, void *);
extern void update_table(HashTable *,u8 *, int);
extern void mark_suspend(HashTable *);
extern int has_suspend_entry(HashTable *);
extern void clean_suspended_entries(HashTable * ht);
extern void notify_all_bfs(HashTable * ht);
extern void check_timeout(HashTable * ht);
static domid_t my_domid;
static u8 my_macs[MAX_MAC_NUM][ETH_ALEN];
static u8 num_of_macs = 0;
static u8 freezed = 0;
struct net_device *NIC = NULL;
static int if_drops = 0;
static skb_queue_t out_queue;
static skb_queue_t pending_free;
static int xenloop_connect(message_t *msg, Entry *e);
static int xenloop_listen(Entry *e);
static struct task_struct *suspend_thread = NULL;
DECLARE_WAIT_QUEUE_HEAD(swq);
static struct task_struct *pending_thread = NULL;
DECLARE_WAIT_QUEUE_HEAD(pending_wq);
// additional map table functions for IP hash map
extern void insert_table_ip(HashTable* ht, u32 ip, Entry* old_entry);
extern void * lookup_table_ip(HashTable * ht, u32 ip);
extern void remove_entry_mac(HashTable* ht, void* mac);
extern int init_hash_table_ip(HashTable* ht);
HashTable mac_domid_map;
HashTable ip_domid_map;
static char* nic = NULL;
module_param(nic,charp,0660);
static int write_xenstore(int status)
{
int err = 1;
err = xenbus_printf(XBT_NIL, "xenloop", "xenloop","%d", status);
if (err) {
EPRINTK( "writing xenstore xenloop status failed, err = %d \n", err);
}
return err;
}
static domid_t get_my_domid(void)
{
char *domidstr;
domid_t domid;
domidstr = xenbus_read(XBT_NIL, "domid", "", NULL);
if ( IS_ERR(domidstr) ) {
EPRINTK("xenbus_read error\n");
return PTR_ERR(domidstr);
}
domid = (domid_t) simple_strtoul(domidstr, NULL, 10);
kfree(domidstr);
return domid;
}
int store_mac(char* mac)
{
char *pEnd = mac;
int i;
for (i=0; i < (ETH_ALEN-1); i++) {
my_macs[(int)num_of_macs][i] = simple_strtol(pEnd, &pEnd, 16);
pEnd++;
}
my_macs[(int)num_of_macs][ETH_ALEN-1] = simple_strtol(pEnd, NULL, 16);
num_of_macs++;
return 0;
}
static int probe_vifs(void)
{
int err = 0;
char **dir;
char * path, *macstr;
unsigned int i, dir_n;
dir = xenbus_directory(XBT_NIL, "device/vif", "", &dir_n);
if (IS_ERR(dir))
return PTR_ERR(dir);
for (i = 0; i < dir_n; i++) {
path = kasprintf(GFP_KERNEL, "device/vif/%s", dir[i]);
if (!path) {
EPRINTK("kasprintf failed dir[%d]=%s \n", i, dir[i]);
err = -ENOMEM;
goto out;
}
macstr = xenbus_read(XBT_NIL, path, "mac", NULL);
if ( IS_ERR(macstr) ) {
EPRINTK("xenbus_read error path=%s \n", path);
err = PTR_ERR(macstr);
kfree(path);
goto out;
}
store_mac(macstr);
DB("device/vif/%s/mac path=%s ==> %s\n", dir[i], path, macstr);
kfree(macstr);
kfree(path);
}
out:
kfree(dir);
return err;
}
void session_update(message_t* msg)
{
int i, found = 0;
u8 mac_count = msg->mac_count;
Entry * e;
for(i=0; i<mac_count; i++) {
if (memcmp(msg->mac[i], my_macs[0], ETH_ALEN) == 0) {
found = 1;
break;
}
}
if (!found) return;
for(i=0; i<mac_count; i++) {
if (memcmp(msg->mac[i], my_macs[0], ETH_ALEN) == 0)
continue;
if (!(e = lookup_table(&mac_domid_map, msg->mac[i]))) {
insert_table(&mac_domid_map, msg->mac[i], msg->guest_domids[i]);
DPRINTK("Added one new guest mac = " MAC_FMT " Domid=%d.\n", \
MAC_NTOA(msg->mac[i]), msg->guest_domids[i]);
} else
e->timestamp = jiffies;
}
update_table(&mac_domid_map, (u8*)msg->mac, msg->mac_count);
}
Entry *pre_check_msg(message_t *msg)
{
Entry *e = NULL;
if(msg->mac_count > 1){
EPRINTK("warning more than one mac\n");
}
if (!(e = lookup_table(&mac_domid_map, msg->mac[0]))) {
EPRINTK("lookup table failed\n");
}
return e;
}
int session_recv(struct sk_buff * skb, net_device * dev, packet_type * pt, net_device * d)
{
int ret = NET_RX_SUCCESS;
message_t * msg = NULL;
Entry *e;
TRACE_ENTRY;
BUG_ON(!skb);
msg = (message_t *)skb->data;
BUG_ON(!msg);
skb_linearize(skb);
switch(msg->type) {
case XENLOOP_MSG_TYPE_SESSION_DISCOVER:
if (!freezed)
session_update(msg);
break;
case XENLOOP_MSG_TYPE_CREATE_CHN:
e = pre_check_msg(msg);
if(!e) goto out;
ret = xenloop_connect(msg, e);
break;
case XENLOOP_MSG_TYPE_CREATE_ACK:
e = pre_check_msg(msg);
if(!e) goto out;
e->status = XENLOOP_STATUS_CONNECTED;
if(e->del_timer) {
del_timer(&e->ack_timer);
}
DPRINTK("LISTENER status changed to XENLOOP_STATUS_CONNECTED!!!\n");
break;
default:
EPRINTK("session_recv(): unknown msg type %d\n", msg->type);
}
out:
kfree_skb(skb);
TRACE_EXIT;
return ret;
}
static packet_type xenloop_ptype = {
.type = __constant_htons(ETH_P_TIDC),
.func = session_recv,
.dev = NULL,
.af_packet_priv = NULL,
};
inline void net_send(struct sk_buff * skb, u8 * dest)
{
ethhdr * eth;
int ret;
skb->network_header = 0;
skb->len = headers;
skb->data_len = 0;
skb_shinfo(skb)->nr_frags = 0;
skb_shinfo(skb)->frag_list = NULL;
skb->tail = headers;
skb->dev = NIC;
skb->protocol = htons(ETH_P_TIDC);
eth = (ethhdr *) skb->data;
eth->h_proto = htons(ETH_P_TIDC);
memcpy(eth->h_dest, dest, ETH_ALEN);
memcpy(eth->h_source, NIC->dev_addr, ETH_ALEN);
if((skb_shinfo(skb) == NULL)) {
WARN_ON(1);
TRACE_ERROR;
}
SKB_LINEAR_ASSERT(skb);
if((ret = dev_queue_xmit(skb))) {
DB("Non-zero return code: %d %s", ret,
skb_shinfo(skb) ? "good" : "bad");
if_drops++;
TRACE_ERROR;
}
}
void send_create_chn_msg(int gref_in, int gref_out, int remote_port, u8 *dest_mac)
{
message_t *m;
struct sk_buff *skb;
TRACE_ENTRY;
skb = alloc_skb(headers, GFP_ATOMIC);
BUG_ON(!skb);
m = (message_t *) (skb->data + LINK_HDR);
memset(m, 0, MSGSIZE);
m->type = XENLOOP_MSG_TYPE_CREATE_CHN;
m->domid= my_domid;
m->mac_count = num_of_macs;
memcpy(m->mac, my_macs, num_of_macs*ETH_ALEN);
m->gref_in = gref_in;
m->gref_out = gref_out;
m->remote_port = remote_port;
net_send(skb, dest_mac);
TRACE_EXIT;
}
void send_create_ack_msg(u8 *dest_mac)
{
message_t *m;
struct sk_buff *skb;
TRACE_ENTRY;
BUG_ON(!(skb = alloc_skb(headers, GFP_ATOMIC)));
m = (message_t *) (skb->data + LINK_HDR);
memset(m, 0, MSGSIZE);
m->type = XENLOOP_MSG_TYPE_CREATE_ACK;
m->domid= my_domid;
m->mac_count = num_of_macs;
memcpy(m->mac, my_macs, num_of_macs*ETH_ALEN);
net_send(skb, dest_mac);
TRACE_EXIT;
}
static void ack_timeout(struct timer_list* tm)
{
// the first member of the struct is the address of the struct Entry
Entry* e = container_of(tm, struct Entry, ack_timer);
bf_handle_t *bfl;
TRACE_ENTRY;
BUG_ON(!e);
BUG_ON(!e->listen_flag);
if( e->status == XENLOOP_STATUS_CONNECTED )
return;
BUG_ON(e->status != XENLOOP_STATUS_LISTEN);
bfl = e->bfh;
BUG_ON(!bfl);
if(e->retry_count < MAX_RETRY_COUNT ) {
send_create_chn_msg(BF_GREF_IN(bfl), \
BF_GREF_OUT(bfl), \
BF_EVT_PORT(bfl), \
e->mac);
e->retry_count++;
mod_timer(&e->ack_timer, jiffies + XENLOOP_ACK_TIMEOUT*HZ);
} else {
if (check_descriptor(e->bfh)) {
BF_SUSPEND_IN(e->bfh) = 1;
BF_SUSPEND_OUT(e->bfh) = 1;
}
e->status = XENLOOP_STATUS_SUSPEND;
wake_up_interruptible(&swq);
}
TRACE_EXIT;
}
static int xenloop_listen(Entry *e)
{
static DEFINE_SPINLOCK(listen_lock);
unsigned long flag;
domid_t remote_domid = e->domid;
bf_handle_t *bfl = NULL;
int i;
bf_data_t *pbf;
TRACE_ENTRY;
spin_lock_irqsave(&listen_lock, flag);
if( e->status != XENLOOP_STATUS_INIT) {
spin_unlock_irqrestore(&listen_lock, flag);
TRACE_EXIT;
return 0;
}
e->status = XENLOOP_STATUS_LISTEN;
spin_unlock_irqrestore(&listen_lock, flag);
bfl = bf_create(remote_domid, XENLOOP_ENTRY_ORDER);
if(!bfl) {
e->status = XENLOOP_STATUS_INIT;
EPRINTK("bf_create failed\n");
TRACE_ERROR;
return -1;
}
for(i=0; i<=xf_size(bfl->in); i++) {
pbf = xf_entry(bfl->in, bf_data_t, i);
pbf->status = BF_FREE;
}
for(i=0; i<=xf_size(bfl->out); i++) {
pbf = xf_entry(bfl->out, bf_data_t, i);
pbf->status = BF_FREE;
}
e->listen_flag = 1;
e->bfh = bfl;
send_create_chn_msg(BF_GREF_IN(bfl),
BF_GREF_OUT(bfl),
BF_EVT_PORT(bfl),
e->mac);
timer_setup(&e->ack_timer, ack_timeout, 0);
e->del_timer = 1;
e->ack_timer.expires = jiffies + XENLOOP_ACK_TIMEOUT*HZ;
add_timer(&e->ack_timer);
TRACE_EXIT;
return 0;
}
static int xenloop_connect(message_t *msg, Entry *e)
{
domid_t remote_domid = e->domid;
bf_handle_t *bfc = NULL;
TRACE_ENTRY;
BUG_ON(!msg);
if(e->status == XENLOOP_STATUS_CONNECTED) {
send_create_ack_msg(e->mac);
TRACE_EXIT;
return 0;
}
if(msg->gref_in <= 0 || msg->gref_out <= 0 || msg->remote_port <= 0) {
EPRINTK("gref_in %d gref_out %d remote_port %d\n", msg->gref_in, msg->gref_out, msg->remote_port);
goto err;
}
bfc = bf_connect(remote_domid, msg->gref_out, msg->gref_in,\
msg->remote_port);
if(!bfc) {
EPRINTK("bf_connect failed\n");
goto err;
}
e->listen_flag = 0;
e->bfh = bfc;
e->status = XENLOOP_STATUS_CONNECTED;
DPRINTK("CONNECTOR status changed to XENLOOP_STATUS_CONNECTED!!!\n");
send_create_ack_msg(e->mac);
TRACE_EXIT;
return 0;
err:
TRACE_ERROR;
return -1;
}
static int xmit_large_pkt(struct sk_buff *skb, xf_handle_t *xfh)
{
bf_data_t *mdata;
char *pback, *pfront, *pfifo;
int num_entries, ret, len=0, len1=0, len2=0;
TRACE_ENTRY;
BUG_ON(!skb);
BUG_ON(!xfh);
if( skb->len + sizeof(bf_data_t) > xf_free(xfh)*sizeof(bf_data_t) ) {
TRACE_EXIT;
return -1;
}
mdata = xf_entry(xfh, bf_data_t, xf_size(xfh));
BUG_ON(!mdata);
mdata->status = BF_WAITING;
mdata->type = BF_PACKET;
mdata->pkt_info = skb->len;
num_entries = skb->len/sizeof(bf_data_t);
if (skb->len % sizeof(bf_data_t))
num_entries++;
pfifo = (char *)xfh->fifo;
pfront = (char *)xf_entry(xfh, bf_data_t, 0);
pback = (char *) xf_entry(xfh, bf_data_t, xf_size(xfh) + 1);
BUG_ON(!pfifo);
BUG_ON(!pfront);
BUG_ON(!pback);
if( pback >= pfront ) {
len1 = (pfifo + xfh->descriptor->max_data_entries*sizeof(bf_data_t)) - pback;
len = (len1 >= skb->len) ? skb->len : len1;
if(skb_copy_bits(skb, 0, pback, len))
BUG();
len2 = skb->len - len;
if( len2 > 0 ) {
if(skb_copy_bits(skb, len, pfifo, len2))
BUG();
}
} else {
if(skb_copy_bits(skb, 0, pback, skb->len))
BUG();
}
ret = xf_pushn(xfh, num_entries + 1);
BUG_ON( ret < 0 );
TRACE_EXIT;
return 0;
}
void enqueue(skb_queue_t *Q, struct sk_buff *skb)
{
if (Q->count++ == 0) {
Q->head = skb;
Q->tail = skb;
} else {
Q->tail->next = skb;
Q->tail = skb;
}
}
void dequeue(skb_queue_t *Q)
{
if (Q->head != Q->tail)
Q->head = Q->head->next;
if (--Q->count == 0) {
Q->head = NULL;
Q->tail = NULL;
}
}
void clean_pending(skb_queue_t *Q) {
struct sk_buff *skb;
while (Q->count > 0) {
skb = Q->head;
dequeue(Q);
kfree_skb(skb);
}
}
// TODO just pass Entry e to this function?
// sometimes this get's passed NULL to just empty the queue out
// queue will fill up if xmit_large_pkt returns some error
inline int xmit_packets(struct sk_buff *skb)
{
static DEFINE_SPINLOCK(xmit_lock);
int ret = 0;
unsigned long flags;
TRACE_ENTRY;
BUG_ON( in_irq() );
spin_lock_irqsave( &xmit_lock, flags );
if(skb) {
if ( skb->len + sizeof(bf_data_t) < (1 << XENLOOP_ENTRY_ORDER)*sizeof(bf_data_t) )
enqueue(&out_queue, skb);
else {
DB("Packet size greater than total fifo size\n");
ret = -1;
}
}
while (out_queue.count > 0) {
int rc;
Entry *e;
skb = out_queue.head;
BUG_ON(!skb);
// TODO store skb and entry together? so we don't have to do this lookup
// this lookup is fairly negligle though
e = lookup_table_ip(&ip_domid_map, ip_hdr(skb)->daddr);
BUG_ON(!e);
rc = xmit_large_pkt(skb, e->bfh->out);
if (rc < 0) {
// EPRINTK("xmit_large_pkt failed: %d\n", rc);
// TODO this seems dangerous, I think calling bf_notify here could lock the CPU (since we're in irqsave)
// we're disabling interrupts and then calling a hypercall in bf_notify, we'll probably lose the return and get stuck :(
// why do we even need a notify here in the first place? the xmit_pending thread will just call this function again
// we haven't transmitted any data, so why tell the other guest we did? seems silly, but maybe I'm wrong
// NOTE: the original XenLoop calls bf_notify here
// to me, this seems dangerous since it uses a hypercall to send a signal on the event channel, but we've disabled interrupts
// I'm not sure this is 100% dangerous, but it seems like it could lock up the CPU since we'll never get a response to the hypercall with interrupts masked
// bf_notify(e->bfh->port);
wake_up_interruptible(&pending_wq);
break;
}
dequeue(&out_queue);
kfree_skb(skb);
}
spin_unlock_irqrestore( &xmit_lock, flags );
// TODO why don't we onlt call notify on the bififos we updated? we'd have to track that
// we can't call notify while IRQs are masked for the same reason as above, it could lock the CPU
// calling bf_notify in the above loop seemed to confirm this theory as the CPU locked up
notify_all_bfs(&ip_domid_map);
TRACE_EXIT;
return ret;
}
// hook outgoing packets
static unsigned int iphook_out(
void* priv,
struct sk_buff *skb,
const struct nf_hook_state* state) {
Entry * e;
int ret = NF_ACCEPT;
// DPRINTK("Hooked out IP: %\n", htonl(ip_hdr(skb)->daddr));
if(!(e = lookup_table_ip(&ip_domid_map, ip_hdr(skb)->daddr))) {
// DPRINTK("Not in table, using normal routing\n");
return NF_ACCEPT;
}
TRACE_ENTRY;
if (check_descriptor(e->bfh) && (BF_SUSPEND_IN(e->bfh) || BF_SUSPEND_OUT(e->bfh))) {
e->status = XENLOOP_STATUS_SUSPEND;
wake_up_interruptible(&swq);
return NF_ACCEPT;
}
switch (e->status) {
case XENLOOP_STATUS_INIT:
if( my_domid < e->domid) {
xenloop_listen(e);
}
TRACE_EXIT;
return NF_ACCEPT;
case XENLOOP_STATUS_CONNECTED:
if( xmit_packets(skb) < 0 ) {
EPRINTK("Couldn't send packet via bififo. Using network instead\n");
ret = NF_ACCEPT;
goto out;
}
// DPRINTK("packet transmitted through Xenloop\n");
ret = NF_STOLEN;
break;
case XENLOOP_STATUS_LISTEN:
default:
TRACE_EXIT;
return ret;
}
out:
TRACE_EXIT;
return ret;
}
static unsigned int iphook_in(
void* priv,
struct sk_buff* skb,
const struct nf_hook_state* state)
{
Entry * e;
int ret = NF_ACCEPT;
if(!(e = lookup_table_ip(&ip_domid_map, ip_hdr(skb)->daddr))) {
return ret;
}
if ((e->status == XENLOOP_STATUS_INIT) && (my_domid < e->domid))
xenloop_listen(e);
TRACE_EXIT;
return NF_ACCEPT;
}
// hook incoming ARP packets
// if it's resolving a MAC address the dom0 has told us about, add it's IP to the table we check
static unsigned int arphook_in(void* priv, struct sk_buff* skb,
const struct nf_hook_state* state) {
int ret = NF_ACCEPT;
struct arphdr* hdr;
Entry* e;
u32 ip;
u8* mac;
hdr = arp_hdr(skb);
if(hdr->ar_pro != htons(ETH_P_IP)) {
return ret;
}
mac = (u8*)(&(hdr->ar_op)) + 2;
if(!(e = lookup_table(&mac_domid_map, (void*)(&(hdr->ar_op)) + 2))) {
return ret;
}
memcpy((void*)&ip, (void*)(&(hdr->ar_op)) + 2 + ETH_ALEN, 4);
if(NULL == lookup_table_ip(&ip_domid_map, ip)) {
insert_table_ip(&ip_domid_map, ip, e);
DPRINTK("Added IP: %u to table\n", ip);
}
// NOTE: we now the same entry in our MAC and IP table,
// the IP table stores a pointer to the Entry allocated by the mac table
return ret;
}
struct nf_hook_ops iphook_in_ops = {
.hook = iphook_in,
.pf = PF_INET, // NOTE: this should be NFPROTO_IPV4, which is the same as PF_INET (and the same as AF_INET) however this is hardcoded for both
.hooknum = NF_INET_PRE_ROUTING,
.priority = 10,
};
struct nf_hook_ops iphook_out_ops = {
.hook = iphook_out,
.pf = PF_INET, // NOTE: should be NFPROTO_IPV4, same reason as with 'iphook_in_ops' above
.hooknum = NF_INET_LOCAL_OUT, // NOTE: this used to be NF_INET_POST_ROUTING, but since we do IP lookup instead of MAC lookup we can hook further up in the stack
.priority = 10,
};
struct nf_hook_ops hook_arp_ops = {
.hook = arphook_in,
.pf = NFPROTO_ARP,
.hooknum = NF_ARP_IN,
.priority = 10,
};
int net_init(void)
{
int ret = 0;
TRACE_ENTRY;
NIC = dev_get_by_name(&init_net, nic);
if(!NIC) {
EPRINTK("Could not find network card %s\n", nic);
ret = -ENODEV;
goto out;
}
DB("Using interface %s, MTU: %d bytes\n", NIC->name, NIC->mtu);
ret = nf_register_net_hook(&init_net, &iphook_out_ops);
if (ret < 0) {
EPRINTK("can't register OUT hook.\n");
goto out;
}
ret = nf_register_net_hook(&init_net, &iphook_in_ops);
if (ret < 0) {
EPRINTK("can't register OUT hook.\n");
goto out;
}
ret = nf_register_net_hook(&init_net, &hook_arp_ops);
if (ret < 0) {
EPRINTK("can't register ARP hook.\n");
goto out;
}
dev_add_pack(&xenloop_ptype);
out:
TRACE_EXIT;
return ret;
}
void net_exit(void)
{
TRACE_ENTRY;
dev_remove_pack(&xenloop_ptype);
nf_unregister_net_hook(&init_net, &iphook_in_ops);
nf_unregister_net_hook(&init_net, &iphook_out_ops);
nf_unregister_net_hook(&init_net, &hook_arp_ops);
if(NIC) dev_put(NIC);
TRACE_EXIT;
}
void pre_migration(void)
{
TRACE_ENTRY;
write_xenstore(0);
freezed = 1;
mark_suspend(&mac_domid_map);
wake_up_interruptible(&swq);
TRACE_EXIT;
return;
}
void post_migration(void)
{
TRACE_ENTRY;
freezed = 0;
write_xenstore(1);
TRACE_EXIT;
return;
}
#define LONG_PENDING_TIMEOUT 1 // seconds
#define SHORT_PENDING_TIMEOUT 1 // jiffies
static int xmit_pending(void *useless)
{
unsigned long timeout;
TRACE_ENTRY;
while(!kthread_should_stop()) {
timeout = out_queue.count ? SHORT_PENDING_TIMEOUT : LONG_PENDING_TIMEOUT*HZ;
wait_event_interruptible_timeout(pending_wq, (out_queue.count > 0), timeout);
xmit_packets(NULL);
}
TRACE_EXIT;
return 0;
}
#define SUSPEND_TIMEOUT 5
static int check_suspend(void *useless) {
int ret;
TRACE_ENTRY;
while(!kthread_should_stop()) {
ret = wait_event_interruptible_timeout(swq, has_suspend_entry(&mac_domid_map), SUSPEND_TIMEOUT*HZ);
if (ret > 0) {
// we have something suspended that we need to cleanup
clean_suspended_entries(&mac_domid_map);
} else if (ret == 0) {
check_timeout(&mac_domid_map);
}
}
TRACE_EXIT;
return 0;
}
static void suspend_resume_handler(struct xenbus_watch *watch,
const char *path, const char* token)
{
char **dir;
unsigned int i, dir_n;
#define SR_UNDEFINED 0
#define SR_SUSPENDED 1
#define SR_RESUMED 2
static int cur_state = SR_UNDEFINED;
int prev_state = cur_state;
if( prev_state == SR_UNDEFINED ) {
cur_state = SR_RESUMED;
return;
}
dir = xenbus_directory(XBT_NIL, "control", "", &dir_n);
if (IS_ERR(dir)) {
EPRINTK("ERROR\n");
return;
}
cur_state = SR_RESUMED;
for (i = 0; i < dir_n; i++) {
if (strcmp(dir[i], "shutdown") != 0)
continue;
cur_state = SR_SUSPENDED;
break;
}
if( prev_state == cur_state)
goto out;
switch(cur_state) {
case SR_SUSPENDED:
pre_migration();
break;
case SR_RESUMED:
post_migration();
break;
}
out: