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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
|
/*
* Copyright c Realtek Semiconductor Corporation, 2002
* All rights reserved.
*
* Program : The mbuf module header file
* Abstract :
* Author : David Chun-Feng Liu (cfliu@realtek.com.tw)
*
*/
#ifndef _MBUF_H_
#define _MBUF_H_
/*********************************************************************************
SECTION 1: mbuf module default settings
**********************************************************************************/
#define PKT_MBUF_CLUSTER_LEN 1650
#define PKT_BUF_SZ PKT_MBUF_CLUSTER_LEN /* Size of each temporary Rx buffer.*/
/*********************************************************************************
SECTION 2: mbuf module data structure definitions
**********************************************************************************/
#define BUF_FREE 0x00 /* Buffer is Free */
#define BUF_USED 0x80 /* Buffer is occupied */
#define BUF_ASICHOLD 0x80 /* Buffer is hold by ASIC */
#define BUF_DRIVERHOLD 0xc0 /* Buffer is hold by driver */
#define MBUFTYPE_DATA 0x01 //not in use now. Keep for backward compatablity.
/*@struct m_buf | The mbuf header associated with each cluster */
struct rtl_mBuf
{
struct rtl_mBuf *m_next;
struct rtl_pktHdr *m_pkthdr; /* Points to the pkthdr structure */
uint16 m_len; /* data bytes used in this cluster */
int8 m_flags; /* mbuf flags; see below */
#define MBUF_FREE BUF_FREE /* Free. Not occupied. should be on free list */
#define MBUF_USED BUF_USED /* Buffer is occupied */
#define MBUF_EXT 0x10 /* has associated with an external cluster, this is always set. */
#define MBUF_PKTHDR 0x08 /* is the 1st mbuf of this packet */
#define MBUF_EOR 0x04 /* is the last mbuf of this packet. Set only by ASIC*/
uint8 *m_data; /* location of data in the cluster */
uint8 *m_extbuf; /* start of buffer*/
uint16 m_extsize; /* sizeof the cluster */
int8 m_reserved[2]; /* padding */
void *skb;
};
/*@struct rtl_pktHdr | pkthdr records packet specific information. Each pkthdr is exactly 32 bytes.
first 20 bytes are for ASIC, the rest 12 bytes are for driver and software usage.
*/
struct rtl_pktHdr
{
union
{
struct rtl_pktHdr *pkthdr_next; /* next pkthdr in free list */
struct rtl_mBuf *mbuf_first; /* 1st mbuf of this pkt */
}PKTHDRNXT;
#define ph_nextfree PKTHDRNXT.pkthdr_next
#define ph_mbuf PKTHDRNXT.mbuf_first
uint16 ph_len; /* total packet length */
uint16 ph_reserved1: 1; /* reserved */
uint16 ph_queueId: 3; /* bit 2~0: Queue ID */
uint16 ph_extPortList: 4; /* dest extension port list. must be 0 for TX */
/* for ph_extPortList */
#define PKTHDR_EXTPORT_MAGIC 0xA530
#define PKTHDR_EXTPORT_MAGIC2 0xA531
#define PKTHDR_EXTPORT_P1 6
#define PKTHDR_EXTPORT_P2 7
#define PKTHDR_EXTPORT_P3 8
#define PKTHDR_EXTPORT_LIST_P0 0
#define PKTHDR_EXTPORT_LIST_P1 1
#define PKTHDR_EXTPORT_LIST_P2 2
#define PKTHDR_EXTPORT_LIST_CPU 3
#define PKTHDR_EXTPORTMASK_P0 (0x1 << (PKTHDR_EXTPORT_LIST_P0))
#define PKTHDR_EXTPORTMASK_P1 (0x1 << (PKTHDR_EXTPORT_LIST_P1))
#define PKTHDR_EXTPORTMASK_P2 (0x1 << (PKTHDR_EXTPORT_LIST_P2))
#define PKTHDR_EXTPORTMASK_CPU (0x1 << (PKTHDR_EXTPORT_LIST_CPU))
#define PKTHDR_EXTPORTMASK_ALL ( PKTHDR_EXTPORTMASK_P0 |\
PKTHDR_EXTPORTMASK_P1 |\
PKTHDR_EXTPORTMASK_P2 |\
PKTHDR_EXTPORTMASK_CPU \
)
uint16 ph_reserved2: 3; /* reserved */
uint16 ph_hwFwd: 1; /* hwFwd - copy from HSA bit 200 */
uint16 ph_isOriginal: 1; /* isOriginal - DP included cpu port or more than one ext port */
uint16 ph_l2Trans: 1; /* l2Trans - copy from HSA bit 129 */
uint16 ph_srcExtPortNum: 2; /* Both in RX & TX. Source extension port number. */
uint16 ph_type: 3;
#define ph_proto ph_type
#define PKTHDR_ETHERNET 0
#define PKTHDR_IP 2
#define PKTHDR_ICMP 3
#define PKTHDR_IGMP 4
#define PKTHDR_TCP 5
#define PKTHDR_UDP 6
uint16 ph_vlanTagged: 1; /* the tag status after ALE */
uint16 ph_LLCTagged: 1; /* the tag status after ALE */
uint16 ph_pppeTagged: 1; /* the tag status after ALE */
uint16 ph_pppoeIdx: 3;
uint16 ph_linkID: 7; /* for WLAN WDS multiple tunnel */
uint16 ph_reason; /* indicates wht the packet is received by CPU */
uint16 ph_flags; /* NEW:Packet header status bits */
#define PKTHDR_FREE (BUF_FREE << 8) /* Free. Not occupied. should be on free list */
#define PKTHDR_USED (BUF_USED << 8)
#define PKTHDR_ASICHOLD (BUF_ASICHOLD<<8) /* Hold by ASIC */
#define PKTHDR_DRIVERHOLD (BUF_DRIVERHOLD<<8) /* Hold by driver */
#define PKTHDR_CPU_OWNED 0x4000
#define PKT_INCOMING 0x1000 /* Incoming: packet is incoming */
#define PKT_OUTGOING 0x0800 /* Outgoing: packet is outgoing */
#define PKT_BCAST 0x0100 /*send/received as link-level broadcast */
#define PKT_MCAST 0x0080 /*send/received as link-level multicast */
#define PKTHDR_BRIDGING 0x0040 /* when PKTHDR_HWLOOKUP is on. 1: Hardware assist to do L2 bridging only, 0:hardware assist to do NAPT*/
#define PKTHDR_HWLOOKUP 0x0020 /* valid when ph_extPortList!=0. 1:Hardware table lookup assistance*/
#define PKTHDR_PPPOE_AUTOADD 0x0004 /* PPPoE header auto-add */
#define CSUM_TCPUDP_OK 0x0001 /*Incoming:TCP or UDP cksum checked */
#define CSUM_IP_OK 0x0002 /* Incoming: IP header cksum has checked */
#define CSUM_TCPUDP 0x0001 /*Outgoing:TCP or UDP cksum offload to ASIC*/
#define CSUM_IP 0x0002 /* Outgoing: IP header cksum offload to ASIC*/
uint8 ph_orgtos; /* RX: original TOS of IP header's value before remarking, TX: undefined */
uint8 ph_portlist; /* RX: source port number, TX: destination portmask */
uint16 ph_vlanId_resv: 1;
uint16 ph_txPriority: 3;
uint16 ph_vlanId: 12;
// uint16 ph_flags2;
union
{
uint16 _flags2; /* RX: bit 15: Reserved, bit14~12: Original Priority, bit 11~0: Original VLAN ID */
/* TX: bit 15~6: Reserved, bit 5~0: Per Port Tag mask setting for TX(bit 5:MII, bit 4~0: Physical Port) */
struct
{
/* RX: bit 15: Reserved, bit14~12: Original Priority, bit 11~0: Original VLAN ID */
uint16 _reserved:1;
uint16 _rxPktPriority:3; /* Rx packet's original priority */
uint16 _svlanId:12; /* Source (Original) VLAN ID */
} _rx;
struct
{
/* TX: bit 15~6: Reserved, bit 5~0: Per Port Tag mask setting for TX(bit 5:MII, bit 4~0: Physical Port) */
uint16 _reserved:10;
uint16 _txCVlanTagAutoAdd:6; /* BitMask to indicate the port which would need to add VLAN tag */
} _tx;
} _flags2;
#define ph_dvlanId ph_vlanId
#define ph_rxPriority ph_txPriority
#define PKTHDR_TXPRIORITY_MIN 0
#define PKTHDR_TXPRIORITY_MAX 7
#define ph_flags2 _flags2._flags2
#define ph_svlanId _flags2._rx._svlanId
#define ph_rxPktPriority _flags2._rx._rxPktPriority
#define ph_txCVlanTagAutoAdd _flags2._tx._txCVlanTagAutoAdd
#define PKTHDR_TXCVID(vid) (vid & 0xfff)
#define PKTHDR_VLAN_P0_AUTOADD (0x0001<<0)
#define PKTHDR_VLAN_P1_AUTOADD (0x0001<<1)
#define PKTHDR_VLAN_P2_AUTOADD (0x0001<<2)
#define PKTHDR_VLAN_P3_AUTOADD (0x0001<<3)
#define PKTHDR_VLAN_P4_AUTOADD (0x0001<<4)
#define PKTHDR_VLAN_P5_AUTOADD (0x0001<<5)
#define PKTHDR_VLAN_AUTOADD ( (PKTHDR_VLAN_P0_AUTOADD)| \
(PKTHDR_VLAN_P1_AUTOADD)| \
(PKTHDR_VLAN_P2_AUTOADD)| \
(PKTHDR_VLAN_P3_AUTOADD)| \
(PKTHDR_VLAN_P4_AUTOADD)| \
(PKTHDR_VLAN_P5_AUTOADD) )
};
//property for ph_unnumber : cw_du
#define PHUNNUMBER_SRCGLOBEIP 0x01
#define PHUNNUMBER_DIRECTION 0x02 // 1 : WAN->LAN; 0: LAN->WAN
#define PHUNNUMBER_UNNUMBEREDSRCIP 0x04
#define PHUNNUMBER_ADVRTMATCHED 0x08
#define PHUNNUMBER_ALLBITS 0xff
#define PKTHDR_PHUNNUMBER_SET(pkthdrPtr, property) \
do{\
assert(pkthdrPtr);\
pkthdrPtr->ph_unnumber |=property;\
}while(0)
#define PKTHDR_PHUNNUMBER_CLR(pkthdrPtr, property) \
do{\
assert(pkthdrPtr);\
pkthdrPtr->ph_unnumber &= ~property;\
}while(0)
#define PKTHDR_PHUNNUMBER_TEST(pkthdrPtr, property) ((pkthdrPtr->ph_unnumber & property) ? 1: 0)
struct rtl_mBufStatus
{
uint32 m_totalmbufs; //Total mbufs allocated during initialization
uint32 m_totalclusters; //Total clusters allocated during initialization
uint32 m_totalpkthdrs; //Total pkthdrs allocated during initialization
uint32 m_freembufs; /* free mbufs in pool now*/
uint32 m_freeclusters; /* free clusters in pool now*/
uint32 m_freepkthdrs; /* free pkthdrs in pool now*/
uint32 m_msize; /* length of an mbuf */
uint32 m_mclbytes; /* length of an mbuf cluster */
uint32 m_pkthdrsize; /* length of an pkthdr */
uint32 m_wait; /* times waited for space, includes mbuf, pkthdr and cluster */
};
/*********************************************************************************
SECTION 3: mbuf module exported variables, symbols and macros
**********************************************************************************/
#define MBUF_COPYALL 1000000000 /* length for m_copy to copy all */
#define MBUF_WAITOK 0x01
#define MBUF_DONTWAIT 0x02 /* Don't wait if there is no buffer available */
#define MBUF_ONLY 0x04 /* Don't allocate a cluster in mBuf_get */
#define MBUF_ALLOCPKTHDR 0x08 /* Allocate a packet header with mbuf chain in mBuf_getm, mBuf_cloneMbufChain, mBuf_dupMbufChain*/
#define MBUF_GETNEWMBUF 0x10 /* In mBuf_prepend, alloate new mbufs directly */
#define MBUF_CHECKPKTHDR(m) ((m)&&(ISSET((m)->m_flags, MBUF_USED) && ((m)->m_pkthdr))?1:0)
#define MBUF_GETPKTHDRFIELD16(field) (*((uint16 *)(field)))
#define MBUF_SETPKTHDRFIELD16(field, value) *((uint16 *)(field)) = (value)
//The size of each cluster.
extern int32 m_clusterSize;
/*********************************************************************************
SECTION 4: mbuf module exported API prototype
**********************************************************************************/
/* mbuf module exported APIs */
/* @doc MBUF_API
@module mbuf.h - Mbuf module API documentation |
This document illustrates the API interface of the mbuf module.
@normal Chun-Feng Liu (cfliu@realtek.com.tw) <date>
Copyright <cp>2001 Realtek<tm> Semiconductor Cooperation, All Rights Reserved.
@head3 List of Symbols |
Here is a list of all functions and variables in this module.
@index | MBUF_API
*/
extern int32 mBuf_init(uint32, uint32, uint32, uint32, uint32); //was tunable_mbinit()
#if defined(RTL865X_TEST)||defined(RTL865X_MODEL_USER)
extern int32 mBuf_Reinit(void);
#endif /* RTL865X_TEST */
/*
@func void | mBuf_init | mbuf module initialization
@parm uint32 | mbufs | Number of mbufs
@parm uint32 | clusters | Number of clusters
@parm uint32 | pkthdrs | Number of packet headers
@parm uint32 | clusterSize | Size of each cluster. must be power of 2
@parm uint32 | msgLogId | Debuging message log list id
@rdesc None
@rvalue SUCCESS | The mbuf module and its memory pool is initiated as requested
@rvalue FAILED | Failed to initialize the mbuf module.
@comm mbuf module initialization. Allocate mbuf and related structure pools. Value for <p clusterSize> must be power of 2.
If <p clusters> is 0, clusters would be allocated externally through registered OS buffer allocation glue function. If clusters
are allocated this way, MBUF_WAITOK flag can't be used for mBuf_get, mBuf_getm, mBuf_dupMbufChain, mBuf_dupPacket since
mbuf module knows nothing about external cluter pool managed by OS hence doesn't know when to wakeup waiting threads.
*/
extern int32 mBuf_getBufStat(struct rtl_mBufStatus *mbs);
/*
@func int32 | mBuf_getBufStat | Returns current status of the mbuf module
@parm struct rtl_mBufStatus * | mbs | A pointer to the mbstat structure for mBuf_getBufStat() to fill in
@rdesc Returns current status of the mbuf module
@rvalue <p FAILED> | Failed to get mbuf module status. Maybe mbs is NULL or mbuf module not yet initiated.
@rvalue <p SUCCESS> | Mbuf module status is returned with the mbstat structure given.
*/
extern int32 mBuf_leadingSpace(struct rtl_mBuf *m);
/*
@func int32 | mBuf_leadingSpace | Calculate the number of leading free data bytes.
@parm struct rtl_mBuf * | m | Pointer to the mbuf chain.
@rdesc Returns the number of free leading space
@rvalue n | The number of free leading data bytes in cluster.
@comm
Calculate the number of leading free data bytes.
@xref <c mBuf_trailingSpace>
*/
extern int32 mBuf_trailingSpace(struct rtl_mBuf *m);
/*
@func int32 | mBuf_trailingSpace | Calculate the number of trailing free data bytes.
@parm struct rtl_mBuf * | m | Pointer to the mbuf chain.
@rdesc Returns the number of free trailing space
@rvalue n | The number of free trailing data bytes in cluster.
@comm
Calculate the number of trailing free data bytes.
@xref <c mBuf_leadingSpace>
*/
extern int32 mBuf_clusterIsWritable(struct rtl_mBuf *m);
/*
@func int32 | mBuf_clusterIsWritable | Determine whether <p m>'s cluster is writable or not.
@parm struct rtl_mBuf * | m | Pointer to an mbuf.
@rdesc Returns TRUE or FALSE
@rvalue TRUE | The cluster is writable.
@rvalue FALSE | The cluster is not writable.
@comm
Determine whether <p m>'s cluster is writable or not. New mbufs allocated due to mBuf_clonePacket, mBuf_split, mBuf_cloneMbufChain
should not modify its cluster becoz it is not the owner of these clusters. This function always return TRUE if the 'clusters' parameter
during mBuf_init() was initialized to 0.
*/
extern uint32 mBuf_getPktlen(struct rtl_mBuf *m);
/*
@func uint32 | mBuf_getPktlen | Get total number of data bytes in the packet which <p m> belongs to
@parm struct rtl_mBuf * |m | Indicate the packet
@rdesc Returns data length of the packet
@rvalue -1 | Failed.
@rvalue <p length>| The actual length of packet
@comm
Get total number of data bytes in the packet which <p m> belongs to
@xref <c MBUF_GETLEN>
*/
extern struct rtl_mBuf *mBuf_data2Mbuf(int8 * x); //was dtom
/*
@func struct rtl_mBuf * | mBuf_data2Mbuf | Given data address <p x>, return mbuf address <p m>
@parm int8 * |x | Data address
@rdesc Returns mbuf address <p m> which owns the cluster block where <p x> resides.
@rvalue <p NULL> | Can't find the address. The data address <p x> given might be within a zombie cluster whose owning mbuf has already been freed.
@rvalue <p m> | The address of owning mbuf
@comm
Finds the mbuf address of given data address <p x>
This function can't be called if the 'clusters' parameter during mBuf_init() was initialized to 0.
@devnote Original BSD code define this as a dtom macro. In our implmentation, we make it a function.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_data2Mbuf(x);
if (n != NULL) {
printf("There are %d bytes in the cluster", n->m_len);
}
@xref <c MBUF2DATAPTR>
*/
extern struct rtl_mBuf *mBuf_get(int32 how, int32 unused, uint32 Nbuf);
/*
@func struct rtl_mBuf * | mBuf_get | mbuf space allocation routines.
@parm int32 | how | <p MBUF_WAITOK>, <p MBUF_DONTWAIT>, and/or <p MBUF_ONLY>
@parm int32 | unused | unused now. Keep for backward compatibility.
@parm uint32 | Nbuf | The number of mbufs requesting.
@rdesc Returns the address of allocated mbuf head
@rvalue <p NULL> | Can't allocate all <p Nbuf>s at once.
@rvalue <p n> | The address of first mbuf allocated. All <p Nbuf>s have been linked together.
@comm
Get <p Nbuf> mbufs and clusters. All mbufs are chained via the <p m_next> pointer inside each mbuf. Note that memory content in all allocated clusters are NOT initialized.
If <p how> has MBUF_DONTWAIT bit set, and we can't get all <p Nbuf>s immediately, return NULL.
If <p how> has MBUF_WAITOK bit set, and we can't get all <p Nbuf>s right away, block waiting until our request is satisfied
If <p how> has MBUF_ONLY bit set, then no clusters would be allocated. Only mbufs are allocated. Can be used with MBUF_DONTWAIT or MBUF_WAITOK
@devnote
1.Side effect: <p mBuf_get> wakes up any sleeping threads if they are block waiting for free mbufs or clusters.
2. If the 'clusters' parameter during mBuf_init() was initialized to 0, mBuf_get doesn't accept MBUF_ONLY flag
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *m;
m= mBuf_get( MBUF_DONTWAIT, 0, 5); //allocate 5 mbufs and clusters without blocking.
if (!m)
return NULL;
@xref
<c mBuf_getCleared> , <c mBuf_getm>, <c mBuf_driverGet>
*/
extern struct rtl_mBuf *mBuf_getCleared(int32 how, int32 unused, uint32 Nbuf);
/*
@func struct rtl_mBuf * | mBuf_getCleared| Same as mBuf_get, but initialize all clusters to zero.
@parm int32 |how | <p MBUF_WAIT>, <p MBUF_DONTWAIT> and/or <p MBUF_ONLY>
@parm int32 |unused | unused now. Keep for backward compatibility.
@parm uint32 | Nbuf | The number of mbufs requesting.
@rdesc Returns the address of allocated mbuf head
@rvalue <p NULL> | Can't allocate all <p Nbuf>s at once.
@rvalue <p n> | The address of first mbuf allocated. All <p Nbuf>s have been linked together and cleared to 0.
@comm
Same as <p mget>(). However, content in all mbufs and clusters have been cleared to 0.
@ex The following example demonstrates how to use this function |
register struct rtl_mBuf *m;
m= mBuf_getCleared(MBUF_DONTWAIT | MBUF_ONLY, 0, 5); //allocate 5 mbufs without blocking.
if (!m)
return NULL;
@xref
<c mBuf_get>, <c mBuf_getm>, <c mBuf_driverGet>
*/
extern uint32 mBuf_driverGet(uint32 Nmbuf,struct rtl_mBuf **ppFirstMbuf, struct rtl_mBuf **ppLastMbuf);
/*
@func uint32 | mBuf_driverGet| Driver specific. Get multiple mbufs without blocking
@parm uint32 |Nmbuf | Number of mbufs requesting
@parm struct rtl_mBuf ** |ppFirstMbuf | Returns a pointer which points to the first mbuf allocated.
@parm struct rtl_mBuf ** | ppLastMbuf | Returns a pointer which points to the last mbuf allocated.
@rdesc Returns number of mbufs successfully allocated.
@comm
An optimized mBuf_get() for driver. <p ppFirstMbuf> and <p ppLastMbuf> should not be NULL!!
mbuf and cluster not initialized to zero. Only required fields in mBuf_driverGet have been properly filled.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf * localTempMbufHead, * localTempMbufTail;
int32 allocated;
requested = 5;
allocated= mBuf_driverGet(requested, &localTempMbufHead, &localTempMbufTail);
if (allocated!=requested)
printf("Can't allocate all 5 mbufs in driver\n")
@xref
<c mBuf_get>, <c mBuf_getm>, <c mBuf_getCleared>
*/
extern struct rtl_mBuf *mBuf_getm(struct rtl_mBuf *m, uint32 len, int32 how, int32 unused);
/*
@func struct rtl_mBuf * | mBuf_getm| allocate <p len>-worth of mbuf clusters and return a pointer to the top
of the allocated chain.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain. Could be NULL.
@parm uint32 |len | Number of data bytes requesting.
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT> and/or <p M_ALLOCPKTHDR>.
@parm int32 |unused | unused now. Keep for backward compatibility.
@rdesc Returns return a pointer to the head of the allocated chain.
@rvalue <p NULL> | Allocation failed.
@rvalue <p n> | The memory address of first mbuf allocated.
@comm
If <p m> is non-null, then it is an mbuf chain to which we want <p len> bytes worth of clusters and their associating mbufs be attached, and so
if allocation is successful, a pointer to m is returned. i.e <p n> = <p m>)
If <p m> is null, <p len> bytes worth of clusters and their associating mbufs would be allocated.
If <p how> has M_ALLOCPKTHDR bit set, <p mBuf_getm> also allocates a pkthdr with the mbuf chain.
If <p how> has MBUF_DONTWAIT bit set, and we can't get a pkthdr immediately, return NULL.
If <p how> has MBUF_WAITOK bit set, and we can't get a pkthdr right away, block waiting until our request is satisfied.
M_BUFONLY can't be used with mBuf_getm, in this case, you should use mBuf_get() and calculate how many mbufs you need on your own.
@ex The following example demonstrates how to use this function |
m = mBuf_getm(m, size, MBUF_WAITOK, 0);
if (m == NULL)
return ENOBUFS;
@xref
<c mBuf_get> , <c mBuf_driverGet>, <c mBuf_getCleared>
*/
extern struct rtl_mBuf *mBuf_getPkthdr(struct rtl_mBuf *m, int32 how);
/*
@func struct rtl_mBuf * | mBuf_getPkthdr| Allocate a packet header for mbuf chain <p m>
@parm struct rtl_mBuf * | m | Pointer to the mbuf chain.
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT>
@rdesc Returns the address of mbuf <p m> that holds a packet header.
@rvalue <p NULL> | Can't allocate a pkthdr for mbuf <p m>.
@rvalue <p m> | A pkthdr is allocated to mbuf <p m>.
@comm
Given an mbuf <p m>, allocate a packet header (pkthdr) structure for <p m> and makes <p m> the owner of this pkthdr.
If there is already a pkthdr owned by some mbuf after m on the same mbuf chain, m becomes the owner of that pkthdr without getting a new one. Corresponding pkthdr fields are modified accordingly.
If <p how> has MBUF_DONTWAIT bit set, and we can't get a pkthdr immediately, return NULL.
If <p how> has MBUF_WAITOK bit set, and we can't get a pkthdr right away, block waiting until our request is satisfied.
mBuf_getPkthdr may wake up other sleeping threads waiting for pkthdrs if any
@ex The following example demonstrates how to use this function |
if (!mBuf_getPkthdr(pNewMbuf, MBUF_DONTWAIT)) //get a pkthdr
goto nospace;
@xref
<c mBuf_driverGetPkthdr>
*/
extern uint32 mBuf_driverGetPkthdr(uint32 Npkthdr,struct rtl_pktHdr **ppHeadPkthdr, struct rtl_pktHdr **ppTailPkthdr);
/*
@func uint32 | mBuf_driverGetPkthdr| Driver specific. Get multiple pkthdrs without blocking
@parm uint32 |Npkthdr | Number of pkthdrs requesting
@parm struct rtl_pktHdr ** |ppHeadPkthdr | Returns a pointer which points to the first pkthdr allocated.
@parm struct rtl_pktHdr ** | ppTailPkthdr | Returns a pointer which points to the last pkthdr allocated.
@rdesc Returns number of pkthdrs successfully allocated.
@comm
This function is dedicated for driver. <p ppHeadPkthdr> and <p ppTailPkthdr> should not be NULL!!
@ex The following example demonstrates how to use this function |
struct rtl_pktHdr * localTempPkthdrHead, * localTempPkthdrTail;
int32 allocated;
requested = 5;
allocated= mBuf_driverGetPkthdr(requested, &localTempPkthdrHead, &localTempPkthdrTail);
if (allocated!=requested)
printf("Can't allocate all 5 mbufs in driver\n")
@xref
<c mBuf_getPkthdr>
*/
extern void mBuf_freeMbuf(struct rtl_mBuf *m);
/*
@func struct rtl_mBuf * | mBuf_freeMbuf | Free a single mbuf <p m>..
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@rdesc None
@comm
Free a single mbuf. Leave assocated pkthdr or cluster, if any, intact.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
mBuf_freeMbuf(m);
@xref
<c mBuf_freeMbufChain> , <c mBuf_driverFreeMbufChain>, <c mBuf_freePkthdr>, <c mBuf_driverFreePkthdr>
*/
extern int32 mBuf_attachCluster(struct rtl_mBuf *m, void *buffer, uint32 id, uint32 size, uint16 datalen, uint16 align);
/*
@func struct rtl_mBuf * | mBuf_attachCluster | Attach a cluster to mbuf <p m>
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm void * | buffer | Address of cluster to be associated with mbuf <p m>.
@parm uint32 | id | Identifier of cluster to be associated with mbuf <p m>.
@parm uint32 | size | Size of cluster to be associated with mbuf <p m>.
@parm uint32 | datalen | Byte count of data already in packet.
@parm uint16 | align | byte offset of first data byte in cluster <p buffer>,if any,from which <p m> attaches to. Default is 0.
@rvalue <p SUCCESS> | cluster has been successfully attached with mbuf <p m>.
@rvalue <p FAILED> | Failed to attach cluster with mbuf <p m>
@comm
Attach a cluster to mbuf <p m>. If there is already a cluster attached with mbuf <p m>, return FAILED.
If <p align> is non-zero, mbuf <p m> sets it's data pointer from which <p m>'s data pointer attaches.
@xref
<c mBuf_get> , <c mBuf_driverGetMbufChain>, <c mBuf_GetPkthdr>, <c mBuf_getm>
*/
extern int32 mBuf_freeOneMbufPkthdr(struct rtl_mBuf *m, void **buffer, uint32 *id, uint16 *size);
/*
@func struct int32 | mBuf_freeOneMbufPkthdr | Free a single mbuf <p m> and associated pkthdr, if any. Returns information of cluster associated.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm void * | buffer | placeholder for address of cluster associated on return.
@parm void * | id | placeholder of identifier of cluster associated on return.
@parm uint16 * | size | placeholder of size of cluster associated on return.
@rdesc Returns number of mbufs being freed
@rvalue <p n> | number of mbufs being freed
@comm
Free a single mbuf and associated pkthdr, if any. Place the successor, if any, in <p n>.
Note that this function DOES NOT free the associated cluster.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
void *buffer ;
uint32 size, id;
n = mBuf_freeOne(m,&buffer,&id, &size);
@xref
<c mBuf_freeMbufChain> , <c mBuf_driverFreeMbufChain>, <c mBuf_freePkthdr>, <c mBuf_driverFreePkthdr>, <c mBuf_freeOne>
*/
extern struct rtl_mBuf *mBuf_freeOne(struct rtl_mBuf *m);
/*
@func struct rtl_mBuf * | mBuf_freeOne | Free a single mbuf <p m> and associated cluster storage.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@rdesc Returns address of next mbuf if any
@rvalue <p NULL> | <p m> has been freed and there isn't any successing mbufs.
@rvalue <p n> | The address of next mbuf after <p m>.
@comm
Free a single mbuf and associated external storage. Place the successor, if any, in <p n>.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_freeOne(m);
@xref
<c mBuf_freeMbufChain> , <c mBuf_driverFreeMbufChain>, <c mBuf_freePkthdr>, <c mBuf_driverFreePkthdr>
*/
extern uint32 mBuf_freeMbufChain(register struct rtl_mBuf *m);
/*
@func struct rtl_mBuf * | mBuf_freeMbufChain| Free the whole mbuf chain started from <p m>
@parm struct rtl_mBuf * | m | Pointer to the mbuf chain.
@rdesc Returns number of mbufs being freed in mbuf chain <p m>.
@rvalue <p n> | The number of mbufs being freed. If <p n> is 0, <p m> is not freed.
@comm
Free the whole mbuf chain starting from <p m>.
@devnote Return type changed from void to int32 to provide more information.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *m;
int32 n;
m = mBuf_get(MBUF_DONTWAIT, 0, 5);
n = mBuf_freeMbufChain(m);
if (n!=5){
//Do error handling...
}
@xref
<c mBuf_freeOne> , <c mBuf_driverFreeMbufChain>, <c mBuf_freePkthdr>, <c mBuf_driverFreePkthdr>
*/
extern uint32 mBuf_driverFreeMbufChain(struct rtl_mBuf *pFirstMbuf);
/*
@func uint32 | mBuf_driverFreeMbufChain| Free the whole mbuf chain started from <p m>
@parm struct rtl_mBuf * | m | Pointer to the mbuf chain.
@rdesc Returns number of mbufs being freed in mbuf chain <p m>.
@rvalue <p n> | The number of mbufs being freed. If <p n> is 0, <p m> is not freed.
@comm
Free the whole mbuf chain starting from <p m>. To be used only in driver.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf * localTempMbufHead, * localTempMbufTail;
int32 allocated;
requested = 5;
allocated= mBuf_driverGet(requested, &localTempMbufHead, &localTempMbufTail);
if (allocated==requested){
n = mBuf_driverFreeMbufChain(localTempMbufHead);
if(n==allocated)
printf("All mbufs successfully freed in driver\n");
}
@xref
<c mBuf_freeOne> , <c mBuf_driverFreeMbufChain>, <c mBuf_freePkthdr>, <c mBuf_driverFreePkthdr>
*/
extern void mBuf_freePkthdr(struct rtl_pktHdr *ph);
/*
@func struct rtl_pktHdr * | mBuf_freePkthdr| Free a pkthdr alone.
@parm struct rtl_pktHdr * |ph | The pkthdr to be freed.
@rdesc No return value
@comm Free a pkthdr alone. Callers should be aware that this is NOT the normal way to free a pkthdr with mbufs attached.
You should use <p mBuf_freeOne> or <p mBuf_freeMbufChain> to free pkthdrs attached with mbuf chains. This function is usd once only in TCP module.
@devnote
Caller of this function should remove the links between mbufs and pkthdrs on their own before <p mBuf_freePkthdr> is called.
@xref <c mBuf_freeOne> , <c mBuf_freeMbufChain>, <c mBuf_driverFreeMbufChain>, <c mBuf_driverFreePkthdr>
*/
uint32 mBuf_driverFreePkthdr(struct rtl_pktHdr *ph, uint32 Npkthdr, struct rtl_pktHdr **ppHeadPkthdr);
/*
@func uint32 | mBuf_driverFreePkthdr| Driver specific. Free multiple pkthdrs without blocking
@parm struct rtl_pktHdr * |ph | The first pkthdr to be freed
@parm uint32 |Npkthdr | Number of pkthdrs to be freed
@parm struct rtl_pktHdr ** |ppHeadPkthdr | Returns next un-freed pkthdr, if any.
@rdesc Returns number of pkthdrs successfully freed.
@comm
This function is dedicated for driver. <p ph> should not be NULL!!
@xref
<c mBuf_freeOne> , <c mBuf_freeMbufChain>, <c mBuf_driverFreeMbufChain>, <c mBuf_freePkthdr>
*/
extern struct rtl_mBuf *mBuf_adjHead(struct rtl_mBuf *, uint32 req_len);
/*
@func struct rtl_mBuf * | mBuf_adjHead | Remove <p req_len> bytes of data from head the mbuf chain pointed to by <p m>.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain. <p m> doesn't need to be the first mbuf in chain.
@parm int32 | req_len | Number of data bytes to be removed
@rdesc m_adj returns the address of first mbuf after trimming data.
@rvalue <p NULL> | Can't adjust length of mbuf. <p m> might 1)be NULL 2)have no clusters.
@rvalue <p m> | When success, the first mbuf <p m> where user can read/write data is returned.
@comm
Remove <p req_len> bytes of data from the head of mbuf chain <p m>. If user removed all
data bytes in the whole packet, <p m> is returned.
@devnote If any clusters, after m_adj, is totally unused (ie. m_len=0),
the m_data pointer would be reset to m_extbuf and the mbuf would NOT be freed.
@xref
<c mBuf_adjTail>, <c mBuf_trimHead>, <c mBuf_trimTail>
*/
extern struct rtl_mBuf *mBuf_adjTail(struct rtl_mBuf *, uint32 req_len);
/*
@func struct rtl_mBuf * | mBuf_adjTail | Remove <p req_len> bytes of data from tail of the mbuf chain pointed to by <p m>.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain. <p m> doesn't need to be the first mbuf in chain.
@parm int32 | req_len | Number of data bytes to be removed
@rdesc m_adj returns the address of mbuf which contains the last data byte.
@rvalue <p NULL> | Can't adjust length of mbuf. <p m> might 1)be NULL 2)have no clusters or <p req_len> is 0
@rvalue <p m> | When success, the address of the mbuf which hold the last data byte is returned.
@comm
Remove <p req_len> bytes of data from the tail of mbuf chain <p m>. If user removed all
data bytes in the whole packet, <p m> is returned.
@devnote If any clusters, after m_adj, is totally unused (ie. m_len=0),
the m_data pointer would be reset to m_extbuf and the mbuf would NOT be freed.
@xref
<c mBuf_adjHead>, <c mBuf_trimHead>, <c mBuf_trimTail>
*/
extern struct rtl_mBuf *mBuf_trimHead(struct rtl_mBuf *, uint32 req_len);
/*
@func struct rtl_mBuf * | mBuf_trimHead | Same as mBuf_adjHead, but also frees unused mbufs and clusters
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain. <p m> doesn't need to be the first mbuf in chain.
@parm int32 | req_len | Number of data bytes to be trimmed from head.
@rdesc mBuf_trimHead returns the address of first mbuf after trimming.
@rvalue <p NULL> | Can't adjust length of mbuf or the whole mbuf chain is trimmed and freed.
@rvalue <p m> | When success, the first mbuf <p m> where user can read/write data is returned.
@comm
Same as mBuf_adjHead, but also frees unused mbufs and clusters
@devnote mBuf_trimHead is implemented with mBuf_split and mBuf_freeMbufChain. It first splits the mbuf
to two mbuf chains from indicated position and frees the first one.
There is a possible risk that mBuf_trimHead may fail when the mbuf chain <p m> consumes all mbufs and no free mbufs
is available for mBuf_split to work correctly.
@xref
<c mBuf_adjHead>, <c mBuf_adjTail>, <c mBuf_trimTail>
*/
extern struct rtl_mBuf *mBuf_trimTail(struct rtl_mBuf *, uint32 req_len);
/*
@func struct rtl_mBuf * | mBuf_trimTail | Same as mBuf_adjTail, but also frees unused mbufs and clusters
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain. <p m> doesn't need to be the first mbuf in chain.
@parm int32 | req_len | Number of data bytes to be trimmed from tail.
@rdesc mBuf_trimTail returns the address <p m>.
@rvalue <p NULL> | Can't adjust length of mbuf or the whole mbuf chain is trimmed and freed.
@rvalue <p m> | When success, returns the first mbuf <p m> with data.
@comm
Same as mBuf_adjTail, but also frees unused mbufs and clusters
@devnote mBuf_trimTail is implemented with mBuf_split and mBuf_freeMbufChain. It first splits the mbuf
to two mbuf chains from indicated position and frees the latter one.
There is a possible risk that mBuf_trimTail may fail when the mbuf chain <p m> consumes all mbufs and no free mbufs
is available for mBuf_split to work correctly.
@xref
<c mBuf_adjHead>, <c mBuf_adjTail>, <c mBuf_trimTail>
*/
extern int32 mBuf_copyToMbuf(struct rtl_mBuf *, uint32 offset, uint32 len, int8 *cp);
/*
@func int32 | mBuf_copyToMbuf | Copy <p len> bytes of data from user's buffer <p cp> to mbuf chain <p m>, started form offset <p offset>.
@parm struct rtl_mBuf * | m | Address of the mbuf chain.
@parm uint32 |offset | Starting byte to be copied in mbuf chain <p m>. Start from 0
@parm uint32 |len | Number of bytes to be copied from <p cp>
@parm int8 * |cp | Address of user's buffer.
@rdesc Returns number of bytes successfully copied
@rvalue <p -1> | Failed.
@rvalue <p n> | <p n> bytes have been copied from <p cp> into indicated mbuf chain <p m>
@comm
Copy <p len> bytes from user's buffer <p cp>, to the indicated mbuf chain
<p m> beginning from the <p offset>-th data byte in mbuf chain.
If there aren't at least 'offset' bytes in mbuf chain, -1 is returned.
mBuf_copyToMbuf() extends mbuf chain if neccessary.
@comm
Be careful, when clusters are allocated externally by OS, mbuf module doesn't know which mbuf
is the first referee and owns the write priviledge to cluster. Therefore, write priviledge
is granted to ALL cluster referees.
@ex The following example demonstrates how to use this function |
#define SIZE 100
int32 i;
int8 my_buffer[SIZE];
for(i=0; i<SIZE; i++)
my_buffer[i] = i;
if (100!=mBuf_copyToMbuf(m, 10, 100, my_buffer)) //copy 100 bytes from my_buffer to mbuf m, started from the 10th bytes in mbuf
printf("Can't copy all data!");
@xref
<c mBuf_copyToUserBuffer>
*/
extern int32 mBuf_copyToUserBuffer(struct rtl_mBuf *m, uint32 off, uint32 len, int8 * cp);
/*
@func int32 | mBuf_copyToUserBuffer | Copy some data from an mbuf chain <p m> to user's buffer <p cp>.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm uint32 |offset | The number of starting data byte in mbuf chain <p m>. Start from 0.
@parm uint32 |len | The number of data bytes to be copied. If <p len>=M_COPYALL, then copy all data till the end of mbuf chain.
@parm int8 * |cp | User specified buffer address.
@rdesc Returns number of bytes successfully received and copied to user's buffer
@rvalue <p -1> | Failed.
@rvalue <p n> | <p n> bytes have been copied from <p m> into indicated buffer <p cp> successfully.
@comm
Copy data from an mbuf chain starting from the <p offset>-th data byte,
continuing for <p len> bytes, into the indicated buffer <p cp>. User should be sure that there is enough free space in
the specified buffer <p cp>.
@ex The following example demonstrates how to use this function |
int8 my_buffer[100];
mBuf_copyToUserBuffer(m, 10, 100, my_buffer); //copy 100 bytes from the 10-th byte in mbuf to user's cp buffer.
@xref
<c mBuf_copyToMbuf>
*/
extern struct rtl_mBuf *mBuf_cloneMbufChain(struct rtl_mBuf *pThisMbuf, int32 iOffset,
int32 iLength, int32 iWait);
/*
@func struct rtl_mBuf * | mBuf_cloneMbufChain | Clone a part of mbuf chain <p m>
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm int32 |offset | The number of starting data byte in mbuf chain <p m>. Start from 0
@parm int32 |len | The number of data bytes to be cloned. If <p len>=M_COPYALL, then clone all data till the end of mbuf chain.
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT> and/or <p M_ALLOCPKTHDR>
@rdesc Returns address of the new, cloned mbuf chain <p n>.
@rvalue <p NULL> | Can't clone mbuf chain <p m>.
@rvalue <p n> | The address of cloned mbuf chain <p n>. Data in <p n> is read only.
@comm
Make a clone of an mbuf chain starting from <p offset>-th byte from the beginning,
continuing for <p len> bytes. If <p len> is M_COPYALL, then clone to end of the whole mbuf chain. (You may choose to use the optimized mBuf_clonePacket() in this case)
The <p how> parameter is a choice of MBUF_WAIT or MBUF_DONTWAIT and/or M_ALLOCPKTHDR by the caller.
If M_ALLOCPKTHDR flag is given, a new pkthdr would be allocated for duplicated packet.
Note that the clone is read-only, The new mbuf chain <p n> only shares cluster data with <p m>.
<p n> can only read cluster data, but not write. <p m> still owns write priviledge.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_cloneMbufChain(m, 20, M_COPYALL, MBUF_WAIT);
if (n != NULL) {
// do following processing
}
@xref
<c mBuf_clonePacket>, <c mBuf_dupPacket>, <c mBuf_cloneMbufChain>
*/
extern struct rtl_mBuf *mBuf_dupMbufChain(struct rtl_mBuf *pMbufChain, int32 iOffset, int32 iLength, int32 flag);
/*
@func struct rtl_mBuf * | mBuf_dupMbufChain | Duplicate a part of mbuf chain <p m>
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm int32 |offset | The number of starting data byte in mbuf chain <p m>. Start from 0
@parm int32 |len | The number of data bytes to be cloned. If <p len>=M_COPYALL, then duplicate all data till the end of mbuf chain.
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT> and/or <p M_ALLOCPKTHDR>
@rdesc Returns address of the new, duplicated mbuf chain <p n>.
@rvalue <p NULL> | Can't duplicate mbuf chain <p m>.
@rvalue <p n> | The address of duplicated mbuf chain <p n>. Data in <p n> is writable.
@comm
Duplicate an mbuf chain starting from <p offset>-th byte from the beginning,
continuing for <p len> bytes. If <p len> is M_COPYALL, then duplicate to end of the whole mbuf chain. (You may choose to use the optimized mBuf_dupPacket() in this case)
The <p how> parameter is a choice of MBUF_WAIT or MBUF_DONTWAIT and/or M_ALLOCPKTHDR by the caller.
If M_ALLOCPKTHDR flag is given, a new pkthdr would be allocated for duplicated packet.
Note that the duplicated mbuf is writable..
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_dupMbufChain(m, 20, M_COPYALL, MBUF_WAIT);
if (n != NULL) {
// do following processing
}
@xref
<c mBuf_clonePacket>, <c mBuf_dupPacket>, <c mBuf_cloneMbufChain>
*/
extern struct rtl_mBuf *mBuf_clonePacket(struct rtl_mBuf *pMbuf, int32 iHow);
/*
@func struct rtl_mBuf * | mBuf_clonePacket | Clone the entire packet <p m>
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT>
@rdesc Returns address of the new mbuf chain <p n>.
@rvalue <p NULL> | Can't clone mbuf chain <p m>.
@rvalue <p n> | The address of copied mbuf chain <p n>. Data in <p n> is read only.
@comm
Clone an entire mbuf chain <p m>, including the packet header (which must be present).
An optimization of the common case `mBuf_cloneMbufChain(m, 0, M_COPYALL, how)'.
The new cloned packet always allocates a new pkthdr.
Note that the copy is read-only, because clusters are not copied, only their reference counts are incremented.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_clonePacket(m, MBUF_WAIT);
if (n != NULL) {
// do following processing
}
@xref
<c mBuf_dupPacket>, <c mBuf_cloneMbufChain>, <c mBuf_dupMbufChain>
*/
extern struct rtl_mBuf *mBuf_dupPacket(struct rtl_mBuf *pMbuf, int32 iHow);
/*
@func struct rtl_mBuf * | mBuf_dupPacket | Duplicate an mbuf chain <p m>, including its data, into a completely new chain <p n>
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT>
@rdesc Returns address of the new mbuf chain <p n>.
@rvalue <p NULL> | Can't duplicate mbuf chain <p m>.
@rvalue <p n> | The address of copied mbuf chain <p n>. Data in <p n> is writable.
@comm
Duplicate an mbuf chain <p m> into a completely new chain <p n>, including
copying data in any <p m>'s mbuf clusters.
The new duplicated packet always allocates a new pkthdr.
Use this instead of mBuf_clonePacket() when you need a writable copy of an mbuf chain.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_dupPacket(m, MBUF_WAIT);
if (n != NULL) {
// do following processing
}
@xref
<c mBuf_clonePacket> , <c mBuf_copyToUserBuffer>, <c mBuf_cloneMbufChain>
*/
extern struct rtl_mBuf *mBuf_prepend(struct rtl_mBuf *m, uint32 plen, int32 how);
/*
@func struct rtl_mBuf * | mBuf_prepend | Arrange to prepend space of size <p plen> to mbuf <p m>.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm uint32 |plen | Number of bytes to be prepended before <p m>
@parm int32 |how | <p MBUF_WAIT>, <p MBUF_DONTWAIT>, <p M_GETNEWMBUF>
@rdesc Returns address of the first mbuf.
@rvalue <p NULL> | Can't allocate memory or parameter problem.
@rvalue <p n> | The address of first mbuf after prepending.
@comm
Arrange to prepend space of size <p plen> to mbuf <p m>.
If new mbufs must be allocated, <p how> specifies whether to wait or not. When user calls mBuf_prepend, the
data length recorded in mbufs and pkthdr is incremented accordingly.
By default, mBuf_prepend uses any leading free space before allocating any new clusters.
But if <p M_GETNEWMBUF> is set, mBuf_prepend allocates new mbufs directly and leaves any leading free buffer space as is.
If <p how> has MBUF_DONTWAIT bit set, and we can't get all buffers immediately, return NULL.
If <p how> has MBUF_WAITOK bit set, and we can't get all buffers right away, block waiting until our request is satisfied.
M_BUFONLY can't be used with mBuf_prepend.
@ex The following example demonstrates how to use this function |
m0 = mBuf_prepend(m0, PPP_HDRLEN, MBUF_DONTWAIT | M_GETNEWMBUF);
if (m0 == 0) {
error = ENOBUFS;
goto bad;
}
@xref <c mBuf_cat>, <c mBuf_padding>
*/
extern struct rtl_mBuf *mBuf_padding(struct rtl_mBuf *m, uint32 plen, int32 how);
/*
@func struct rtl_mBuf * | mBuf_padding | Arrange to append space of size <p plen> to mbuf <p m>.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm uint32 |plen | Number of bytes to append at <p m>
@parm int32 |how | <p MBUF_WAIT> or <p MBUF_DONTWAIT>
@rdesc Returns address of the last mbuf padded.
@rvalue <p NULL> | Can't allocate memory or parameter problem.
@rvalue <p n> | The address of last mbuf after padding
@comm
Arrange to append space of size <p plen> to mbuf <p m>.
If new mbufs must be allocated, <p how> specifies whether to wait or not. When user calls mBuf_padding, the
data length recorded in mbufs and pkthdr are incremented accordingly.
M_BUFONLY can't be used with mBuf_padding.
@devnote API CHANGED: If <p how> is MBUF_DONTWAIT and allocation fails, NULL is returned.
Original mbuf <p m> is intact.
@ex The following example demonstrates how to use this function |
m0 = mBuf_padding(m, (ETHER_MIN_LEN-ETHER_CRC_LEN-len), MBUF_DONTWAIT);
if (m0 == 0) {
error = ENOBUFS;
goto bad;
}
@xref <c mBuf_split>, <c mBuf_cat>, <c mBuf_getm>, <c mBuf_prepend>, <c mBuf_split>
*/
extern struct rtl_mBuf *mBuf_cat(struct rtl_mBuf *m, struct rtl_mBuf *n);
/*
@func struct rtl_mBuf * | mBuf_cat | Concatenate mbuf chain <p n> to <p m>.
@parm struct rtl_mBuf * | m | mbuf chain to be appended.
@parm struct rtl_mBuf * | n | mbuf chain to be appended after <p m>
@rdesc Returns the address of <p m> or NULL if failed.
@rvalue <p NULL> | Can't concatenate two mbuf chain
@rvalue <p m> | When success, the address of <p m> is returned..
@comm
Concatenate mbuf chain <p n> to <p m>.
Total packet length of <p m> would be adjusted accordingly. However, <p n>'s control header, if any, would be freed
@xref <c mBuf_split>, <c mBuf_padding>, <p mBuf_getm>
*/
extern int32 mBuf_pullup(struct rtl_mBuf *, int32);
/*
@func int32 | mBuf_pullup| Rearange an mbuf chain so that <p len> bytes are contiguous
in <p m>'s first cluster.
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain.
@parm int32 |len | The number of databytes requested to put in <p m>'s first cluster. <p len> should not exceed the size of a cluster buffer (256 bytes).
@rdesc Returns total data bytes in the first mbuf cluster.
@rvalue <p n> | The number of data bytes in <p m>'s first cluster. If <p n> is less than <p len>,
that means <p mBuf_pullup> can't pull up all <p len> bytes requested because there isn't
so much data in mbuf chain.
@comm Original BSD4.4 note: Rearange an mbuf chain so that <p len> bytes are contiguous and in <p m>'s cluster.
Successing mbufs are adjusted accordingly. If the return value <p n> is smaller than requested <p len>,
maybe there isn't enough data bytes in the mbuf chain <p m>, or maybe the requested value <p len> exceeds
the maximum capacity for a single cluster.
Our implementation note:
(cfliu 2002.02.19)This function is important in traditional BSD networking stack because
original mbuf can't do dtom() if data is saved in cluster. This is not the
case here since we offer a back pointer from cluster to mbuf via cltag_mbuf field.
However, this function is still useful if we want to collect continous databytes from later clusters
to the specified mbuf's cluster. The maximum number of <p len> should be less than a cluster
can hold (ie. len less than or equal to m_clusterSIze)
(2002.08.11) If 1) requested length <p len> is smaller than m_clusterSize, and 2) there are enough data in this mbuf chain. However, the
trailing space in <p m> is not large enough, <p mBuf_pullup> will move m's data upward first and then copy requested data from latter clusters
@devnote This function is different from traditional BSD mbuf code. When pull up fails, the mbuf chain will not be freed,
and the return value has been changed to an integer instead of a pointer to mbuf structure.
@ex The following example demonstrates how to use this function |
int32 n;
n = mBuf_pullup(m, 128);
if (n < 128) {
//There isn't so much data in mbuf chain.
}
@xref
<c m_pulldown>
*/
struct rtl_mBuf *mBuf_split(register struct rtl_mBuf *m0, uint32 len0, int32 wait);
/*@func struct rtl_mBuf * | mBuf_split | Partition an mbuf chain in two pieces. Data is cloned
@parm struct rtl_mBuf * | m | Pointer to an mbuf chain
@parm int32 | len | The number of last data byte to remain in original mbuf chain <p m>
@parm int32 | wait | <p MBUF_DONTWAIT> or <p MBUF_WAIT>
@rdesc Returns a pointer to the splited mbuf chain <p n>
@rvalue NULL | Can't split the mbuf chain
@rvalue <p n> | Success. <p n> is the address of second mbuf chain. It holds all data bytes after the <p len>-th byte in <p m>.
@comm Partition an mbuf chain <p m> into two pieces, returning the tail mbuf chain <p n>. In case of failure, it returns NULL and
attempts to restore the chain to its original state. Data in clusters are NOT copied, but cloned only.
@ex The following example demonstrates how to use this function |
struct rtl_mBuf *n;
n = mBuf_split(m, off, MBUF_DONTWAIT);
if (n == NULL) {
goto bad;
}
@xref
<c mBuf_cat>, <c mBuf_padding>, <c mBuf_getm>
*/
void mBuf_getBMjmpTable(uint8 *pat, uint16 *jump_tbl,uint16 patLen, uint8 caseSensitive);
int32 mBuf_BMpatternMatch(struct rtl_mBuf *m, uint32 len, uint8 *delimiter, uint32 delimitLen, uint16 *jmp_tbl, uint8 caseSensitive);
#define MBUF_GETLEN(m) ((m)? ((m)->m_len : -1)
/*
@func MACRO | MBUF_GETLEN | Get total number of data bytes in the mbuf <p m>
@parm struct rtl_mBuf * |m | Indicate the packet
@rdesc Returns data length of the mbuf
@rvalue -1 | Failed.
@rvalue <p length>| The actual length of data in <p m>
@comm
Get total number of data bytes in the mbuf <p m>
@xref <c mBuf_getPktlen>
*/
struct rtl_mBuf *mBuf_attachHeader(void *buffer, uint32 id, uint32 bufsize,uint32 datalen, uint16 align);
struct rtl_mBuf *mBuf_attachHeaderJumbo(void *buffer, uint32 id, uint32 bufsize,uint32 datalen);
int32 mBuf_setNICRxRingSize(uint32 size);
extern int32 mBuf_reserve(struct rtl_mBuf * m, uint16 headroom);
#define MBUF_SET_PKTHDRFLAGS(m, Flags) do{\
assert((m));\
assert(ISSET((m)->m_flags, MBUF_USED));\
assert((m)->m_pkthdr);\
MBUF_SETPKTHDRFIELD16(((memaddr *)&(m)->m_pkthdr->ph_flags), (Flags));\
}while(0)
/*
@func MACRO | MBUF_SET_PKTHDRFLAGS | Set packet specific flags for ASIC processing
@parm struct rtl_mBuf * |m | Indicate the packet
@parm uint32 |Flags | flag to be set
@rdesc None
@comm
Set packet specific flags for ASIC processing. This overwrites original flag value. If you are adding flag bits rather than reseting it,
remember to save its old value first
@xref <c MBUF_GET_PKTHDRFLAGS>
*/
#define MBUF_GET_PKTHDRFLAGS(m) (MBUF_CHECKPKTHDR((m))? MBUF_GETPKTHDRFIELD16((memaddr *)(&m->m_pkthdr->ph_flags)):-1)
/*
@func MACRO | MBUF_GET_PKTHDRFLAGS | Get packet specific flags set by ASIC
@parm struct rtl_mBuf * |m | Indicate the packet
@rdesc Returns the result of execution
@rvalue -1 | Failed. <p m> might have no pkthdrs.
@rvalue FLAGS | Returns the flags.
@comm
Get packet specific flags set by ASIC
@xref <c MBUF_SET_PKTHDRFLAGS>
*/
#define MBUF_SET_PORTLIST(m,Portlist) do{\
assert((m));\
assert(ISSET((m)->m_flags, MBUF_USED));\
assert((m)->m_pkthdr);\
(m)->m_pkthdr->ph_portlist = (Portlist);\
}while(0)
/*
@func MACRO | MBUF_SET_PORTLIST | Set outgoing port list
@parm struct rtl_mBuf * |m | Indicate the packet
@parm uint32 |Portlist | port list
@rdesc None
@comm Set outgoing port list
@xref <c MBUF_GET_PORTLIST>
*/
#define MBUF_GET_PORTLIST(m) (MBUF_CHECKPKTHDR((m))? ((m)->m_pkthdr->ph_portlist):-1)
/*
@func MACRO | MBUF_GET_PORTLIST | get incoming port list
@parm struct rtl_mBuf * |m | Indicate the packet
@rdesc Returns the result of execution
@rvalue 0 | Failed. <p m> might have no pkthdrs.
@rvalue Portlist | Returns incoming portlist.
@comm
get incoming port list
@xref <c MBUF_SET_PORTLIST>
*/
#define MBUF_SET_TRAPREASON(m, Reason) do{\
assert((m));\
assert(ISSET((m)->m_flags, MBUF_USED));\
assert((m)->m_pkthdr);\
MBUF_SETPKTHDRFIELD16(((memaddr *)&(m)->m_pkthdr->ph_reason), (Reason));\
}while(0)
/*
@func MACRO | MBUF_SET_TRAPREASON | Set packet trapping reason for ASIC processing
@parm struct rtl_mBuf * |m | Indicate the packet
@parm uint16 |Reason | Trapping reason
@rdesc None
@comm
Set packet trapping reason for ASIC processing
@xref <c MBUF_GET_TRAPREASON>
*/
#define MBUF_GET_TRAPREASON(m) (MBUF_CHECKPKTHDR((m))? MBUF_GETPKTHDRFIELD16((memaddr *)(&m->m_pkthdr->ph_reason)):-1)
/*
@func MACRO | MBUF_GET_TRAPREASON | Get packet trapping reason set by ASIC
@parm struct rtl_mBuf * |m | Indicate the packet
@rdesc Returns the result of execution
@rvalue -1 | Failed. <p m> might have no pkthdrs.
@rvalue Reason | Returns trap reason
@comm
Get packet trapping reason set by ASIC
@xref <c MBUF_SET_TRAPREASON>
*/
#define MBUF2DATAPTR(m, t) ((t)((m)->m_data))
/*
@func MACRO | MBUF2DATAPTR | Given mbuf pointer 'm', returns address of m->m_data, and convert it to type 't'
@parm struct rtl_mBuf * |m | Pointer to an mbuf
@parm TYPE |t | A type to be casted
@rdesc Address of m->m_data
@comm
Given mbuf pointer 'm', returns address of m->m_data, and convert it to type 't'
@xref <c mBuf_data2Mbuf>
*/
#define MBUF_ALIGN(m, len) do { \
assert((m)->m_len==0);\
assert(len>0);\
assert(((m)->m_extsize - (len))>0);\
(m)->m_data += ( (m)->m_extsize - (len)) & ~(sizeof(memaddr) - 1); \
} while (0)
/*
@func MACRO | MBUF_ALIGN | Align the m->m_data pointer from end of the cluster, for 'len' bytes.
@parm struct rtl_mBuf * |m | Pointer to an mbuf
@parm uint32 |len | Size to be aligned.
@rdesc None
@comm
Set the m_data pointer of a NEWLY-allocated cluster to place
an object of the specified size at the end of the mbuf, longword aligned.
@xref <c MBUF_RESERVE>
*/
#define MBUF_RESERVE(m, len) do { \
assert(len>0);\
(m)->m_data += ( ((len) > (m)->m_extsize)? (m)->m_extsize : len); \
} while (0)
/*
@func MACRO | MBUF_RESERVE | Reserve 'len' bytes from the beginning of the newly allocated cluster
@parm struct rtl_mBuf * |m | Pointer to an mbuf
@parm uint32 |len | Size to be aligned.
@rdesc None
@comm
Forward move the m_data pointer of a NEWLY-allocated cluster for 'len' bytes.
m_data won't move beyond m_extbuf + m_extsize, make sure 'len' you give is a reasonable value
@xref <c MBUF_ALIGN>
*/
#endif /* !_MBUF_H_ */
|