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
|
// Copyright 2018 The gVisor Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package header
import (
"encoding/binary"
"github.com/google/btree"
"gvisor.dev/gvisor/pkg/tcpip"
"gvisor.dev/gvisor/pkg/tcpip/checksum"
"gvisor.dev/gvisor/pkg/tcpip/seqnum"
)
// These constants are the offsets of the respective fields in the TCP header.
const (
TCPSrcPortOffset = 0
TCPDstPortOffset = 2
TCPSeqNumOffset = 4
TCPAckNumOffset = 8
TCPDataOffset = 12
TCPFlagsOffset = 13
TCPWinSizeOffset = 14
TCPChecksumOffset = 16
TCPUrgentPtrOffset = 18
)
const (
// MaxWndScale is maximum allowed window scaling, as described in
// RFC 1323, section 2.3, page 11.
MaxWndScale = 14
// TCPMaxSACKBlocks is the maximum number of SACK blocks that can
// be encoded in a TCP option field.
TCPMaxSACKBlocks = 4
)
// TCPFlags is the dedicated type for TCP flags.
type TCPFlags uint8
// Intersects returns true iff there are flags common to both f and o.
func (f TCPFlags) Intersects(o TCPFlags) bool {
return f&o != 0
}
// Contains returns true iff all the flags in o are contained within f.
func (f TCPFlags) Contains(o TCPFlags) bool {
return f&o == o
}
// String implements Stringer.String.
func (f TCPFlags) String() string {
flagsStr := []byte("FSRPAUEC")
for i := range flagsStr {
if f&(1<<uint(i)) == 0 {
flagsStr[i] = ' '
}
}
return string(flagsStr)
}
// Flags that may be set in a TCP segment.
const (
TCPFlagFin TCPFlags = 1 << iota
TCPFlagSyn
TCPFlagRst
TCPFlagPsh
TCPFlagAck
TCPFlagUrg
TCPFlagEce
TCPFlagCwr
)
// Options that may be present in a TCP segment.
const (
TCPOptionEOL = 0
TCPOptionNOP = 1
TCPOptionMSS = 2
TCPOptionWS = 3
TCPOptionTS = 8
TCPOptionSACKPermitted = 4
TCPOptionSACK = 5
)
// Option Lengths.
const (
TCPOptionMSSLength = 4
TCPOptionTSLength = 10
TCPOptionWSLength = 3
TCPOptionSackPermittedLength = 2
)
// TCPFields contains the fields of a TCP packet. It is used to describe the
// fields of a packet that needs to be encoded.
type TCPFields struct {
// SrcPort is the "source port" field of a TCP packet.
SrcPort uint16
// DstPort is the "destination port" field of a TCP packet.
DstPort uint16
// SeqNum is the "sequence number" field of a TCP packet.
SeqNum uint32
// AckNum is the "acknowledgement number" field of a TCP packet.
AckNum uint32
// DataOffset is the "data offset" field of a TCP packet. It is the length of
// the TCP header in bytes.
DataOffset uint8
// Flags is the "flags" field of a TCP packet.
Flags TCPFlags
// WindowSize is the "window size" field of a TCP packet.
WindowSize uint16
// Checksum is the "checksum" field of a TCP packet.
Checksum uint16
// UrgentPointer is the "urgent pointer" field of a TCP packet.
UrgentPointer uint16
}
// TCPSynOptions is used to return the parsed TCP Options in a syn
// segment.
//
// +stateify savable
type TCPSynOptions struct {
// MSS is the maximum segment size provided by the peer in the SYN.
MSS uint16
// WS is the window scale option provided by the peer in the SYN.
//
// Set to -1 if no window scale option was provided.
WS int
// TS is true if the timestamp option was provided in the syn/syn-ack.
TS bool
// TSVal is the value of the TSVal field in the timestamp option.
TSVal uint32
// TSEcr is the value of the TSEcr field in the timestamp option.
TSEcr uint32
// SACKPermitted is true if the SACK option was provided in the SYN/SYN-ACK.
SACKPermitted bool
// Flags if specified are set on the outgoing SYN. The SYN flag is
// always set.
Flags TCPFlags
}
// SACKBlock represents a single contiguous SACK block.
//
// +stateify savable
type SACKBlock struct {
// Start indicates the lowest sequence number in the block.
Start seqnum.Value
// End indicates the sequence number immediately following the last
// sequence number of this block.
End seqnum.Value
}
// Less returns true if r.Start < b.Start.
func (r SACKBlock) Less(b btree.Item) bool {
return r.Start.LessThan(b.(SACKBlock).Start)
}
// Contains returns true if b is completely contained in r.
func (r SACKBlock) Contains(b SACKBlock) bool {
return r.Start.LessThanEq(b.Start) && b.End.LessThanEq(r.End)
}
// TCPOptions are used to parse and cache the TCP segment options for a non
// syn/syn-ack segment.
//
// +stateify savable
type TCPOptions struct {
// TS is true if the TimeStamp option is enabled.
TS bool
// TSVal is the value in the TSVal field of the segment.
TSVal uint32
// TSEcr is the value in the TSEcr field of the segment.
TSEcr uint32
// SACKBlocks are the SACK blocks specified in the segment.
SACKBlocks []SACKBlock
}
// TCP represents a TCP header stored in a byte array.
type TCP []byte
const (
// TCPMinimumSize is the minimum size of a valid TCP packet.
TCPMinimumSize = 20
// TCPOptionsMaximumSize is the maximum size of TCP options.
TCPOptionsMaximumSize = 40
// TCPHeaderMaximumSize is the maximum header size of a TCP packet.
TCPHeaderMaximumSize = TCPMinimumSize + TCPOptionsMaximumSize
// TCPTotalHeaderMaximumSize is the maximum size of headers from all layers in
// a TCP packet. It analogous to MAX_TCP_HEADER in Linux.
//
// TODO(b/319936470): Investigate why this needs to be at least 140 bytes. In
// Linux this value is at least 160, but in theory we should be able to use
// 138. In practice anything less than 140 starts to break GSO on gVNIC
// hardware.
TCPTotalHeaderMaximumSize = 160
// TCPProtocolNumber is TCP's transport protocol number.
TCPProtocolNumber tcpip.TransportProtocolNumber = 6
// TCPMinimumMSS is the minimum acceptable value for MSS. This is the
// same as the value TCP_MIN_MSS defined net/tcp.h.
TCPMinimumMSS = IPv4MaximumHeaderSize + TCPHeaderMaximumSize + MinIPFragmentPayloadSize - IPv4MinimumSize - TCPMinimumSize
// TCPMinimumSendMSS is the minimum value for MSS in a sender. This is the
// same as the value TCP_MIN_SND_MSS in net/tcp.h.
TCPMinimumSendMSS = TCPOptionsMaximumSize + MinIPFragmentPayloadSize
// TCPMaximumMSS is the maximum acceptable value for MSS.
TCPMaximumMSS = 0xffff
// TCPDefaultMSS is the MSS value that should be used if an MSS option
// is not received from the peer. It's also the value returned by
// TCP_MAXSEG option for a socket in an unconnected state.
//
// Per RFC 1122, page 85: "If an MSS option is not received at
// connection setup, TCP MUST assume a default send MSS of 536."
TCPDefaultMSS = 536
)
// SourcePort returns the "source port" field of the TCP header.
func (b TCP) SourcePort() uint16 {
return binary.BigEndian.Uint16(b[TCPSrcPortOffset:])
}
// DestinationPort returns the "destination port" field of the TCP header.
func (b TCP) DestinationPort() uint16 {
return binary.BigEndian.Uint16(b[TCPDstPortOffset:])
}
// SequenceNumber returns the "sequence number" field of the TCP header.
func (b TCP) SequenceNumber() uint32 {
return binary.BigEndian.Uint32(b[TCPSeqNumOffset:])
}
// AckNumber returns the "ack number" field of the TCP header.
func (b TCP) AckNumber() uint32 {
return binary.BigEndian.Uint32(b[TCPAckNumOffset:])
}
// DataOffset returns the "data offset" field of the TCP header. The return
// value is the length of the TCP header in bytes.
func (b TCP) DataOffset() uint8 {
return (b[TCPDataOffset] >> 4) * 4
}
// Payload returns the data in the TCP packet.
func (b TCP) Payload() []byte {
return b[b.DataOffset():]
}
// Flags returns the flags field of the TCP header.
func (b TCP) Flags() TCPFlags {
return TCPFlags(b[TCPFlagsOffset])
}
// WindowSize returns the "window size" field of the TCP header.
func (b TCP) WindowSize() uint16 {
return binary.BigEndian.Uint16(b[TCPWinSizeOffset:])
}
// Checksum returns the "checksum" field of the TCP header.
func (b TCP) Checksum() uint16 {
return binary.BigEndian.Uint16(b[TCPChecksumOffset:])
}
// UrgentPointer returns the "urgent pointer" field of the TCP header.
func (b TCP) UrgentPointer() uint16 {
return binary.BigEndian.Uint16(b[TCPUrgentPtrOffset:])
}
// SetSourcePort sets the "source port" field of the TCP header.
func (b TCP) SetSourcePort(port uint16) {
binary.BigEndian.PutUint16(b[TCPSrcPortOffset:], port)
}
// SetDestinationPort sets the "destination port" field of the TCP header.
func (b TCP) SetDestinationPort(port uint16) {
binary.BigEndian.PutUint16(b[TCPDstPortOffset:], port)
}
// SetChecksum sets the checksum field of the TCP header.
func (b TCP) SetChecksum(xsum uint16) {
checksum.Put(b[TCPChecksumOffset:], xsum)
}
// SetDataOffset sets the data offset field of the TCP header. headerLen should
// be the length of the TCP header in bytes.
func (b TCP) SetDataOffset(headerLen uint8) {
b[TCPDataOffset] = (headerLen / 4) << 4
}
// SetSequenceNumber sets the sequence number field of the TCP header.
func (b TCP) SetSequenceNumber(seqNum uint32) {
binary.BigEndian.PutUint32(b[TCPSeqNumOffset:], seqNum)
}
// SetAckNumber sets the ack number field of the TCP header.
func (b TCP) SetAckNumber(ackNum uint32) {
binary.BigEndian.PutUint32(b[TCPAckNumOffset:], ackNum)
}
// SetFlags sets the flags field of the TCP header.
func (b TCP) SetFlags(flags uint8) {
b[TCPFlagsOffset] = flags
}
// SetWindowSize sets the window size field of the TCP header.
func (b TCP) SetWindowSize(rcvwnd uint16) {
binary.BigEndian.PutUint16(b[TCPWinSizeOffset:], rcvwnd)
}
// SetUrgentPointer sets the window size field of the TCP header.
func (b TCP) SetUrgentPointer(urgentPointer uint16) {
binary.BigEndian.PutUint16(b[TCPUrgentPtrOffset:], urgentPointer)
}
// CalculateChecksum calculates the checksum of the TCP segment.
// partialChecksum is the checksum of the network-layer pseudo-header
// and the checksum of the segment data.
func (b TCP) CalculateChecksum(partialChecksum uint16) uint16 {
// Calculate the rest of the checksum.
return checksum.Checksum(b[:b.DataOffset()], partialChecksum)
}
// IsChecksumValid returns true iff the TCP header's checksum is valid.
func (b TCP) IsChecksumValid(src, dst tcpip.Address, payloadChecksum, payloadLength uint16) bool {
xsum := PseudoHeaderChecksum(TCPProtocolNumber, src, dst, uint16(b.DataOffset())+payloadLength)
xsum = checksum.Combine(xsum, payloadChecksum)
return b.CalculateChecksum(xsum) == 0xffff
}
// Options returns a slice that holds the unparsed TCP options in the segment.
func (b TCP) Options() []byte {
return b[TCPMinimumSize:b.DataOffset()]
}
// ParsedOptions returns a TCPOptions structure which parses and caches the TCP
// option values in the TCP segment. NOTE: Invoking this function repeatedly is
// expensive as it reparses the options on each invocation.
func (b TCP) ParsedOptions() TCPOptions {
return ParseTCPOptions(b.Options())
}
func (b TCP) encodeSubset(seq, ack uint32, flags TCPFlags, rcvwnd uint16) {
binary.BigEndian.PutUint32(b[TCPSeqNumOffset:], seq)
binary.BigEndian.PutUint32(b[TCPAckNumOffset:], ack)
b[TCPFlagsOffset] = uint8(flags)
binary.BigEndian.PutUint16(b[TCPWinSizeOffset:], rcvwnd)
}
// Encode encodes all the fields of the TCP header.
func (b TCP) Encode(t *TCPFields) {
b.encodeSubset(t.SeqNum, t.AckNum, t.Flags, t.WindowSize)
b.SetSourcePort(t.SrcPort)
b.SetDestinationPort(t.DstPort)
b.SetDataOffset(t.DataOffset)
b.SetChecksum(t.Checksum)
b.SetUrgentPointer(t.UrgentPointer)
}
// EncodePartial updates a subset of the fields of the TCP header. It is useful
// in cases when similar segments are produced.
func (b TCP) EncodePartial(partialChecksum, length uint16, seqnum, acknum uint32, flags TCPFlags, rcvwnd uint16) {
// Add the total length and "flags" field contributions to the checksum.
// We don't use the flags field directly from the header because it's a
// one-byte field with an odd offset, so it would be accounted for
// incorrectly by the Checksum routine.
tmp := make([]byte, 4)
binary.BigEndian.PutUint16(tmp, length)
binary.BigEndian.PutUint16(tmp[2:], uint16(flags))
xsum := checksum.Checksum(tmp, partialChecksum)
// Encode the passed-in fields.
b.encodeSubset(seqnum, acknum, flags, rcvwnd)
// Add the contributions of the passed-in fields to the checksum.
xsum = checksum.Checksum(b[TCPSeqNumOffset:TCPSeqNumOffset+8], xsum)
xsum = checksum.Checksum(b[TCPWinSizeOffset:TCPWinSizeOffset+2], xsum)
// Encode the checksum.
b.SetChecksum(^xsum)
}
// SetSourcePortWithChecksumUpdate implements ChecksummableTransport.
func (b TCP) SetSourcePortWithChecksumUpdate(new uint16) {
old := b.SourcePort()
b.SetSourcePort(new)
b.SetChecksum(^checksumUpdate2ByteAlignedUint16(^b.Checksum(), old, new))
}
// SetDestinationPortWithChecksumUpdate implements ChecksummableTransport.
func (b TCP) SetDestinationPortWithChecksumUpdate(new uint16) {
old := b.DestinationPort()
b.SetDestinationPort(new)
b.SetChecksum(^checksumUpdate2ByteAlignedUint16(^b.Checksum(), old, new))
}
// UpdateChecksumPseudoHeaderAddress implements ChecksummableTransport.
func (b TCP) UpdateChecksumPseudoHeaderAddress(old, new tcpip.Address, fullChecksum bool) {
xsum := b.Checksum()
if fullChecksum {
xsum = ^xsum
}
xsum = checksumUpdate2ByteAlignedAddress(xsum, old, new)
if fullChecksum {
xsum = ^xsum
}
b.SetChecksum(xsum)
}
// ParseSynOptions parses the options received in a SYN segment and returns the
// relevant ones. opts should point to the option part of the TCP header.
func ParseSynOptions(opts []byte, isAck bool) TCPSynOptions {
limit := len(opts)
synOpts := TCPSynOptions{
// Per RFC 1122, page 85: "If an MSS option is not received at
// connection setup, TCP MUST assume a default send MSS of 536."
MSS: TCPDefaultMSS,
// If no window scale option is specified, WS in options is
// returned as -1; this is because the absence of the option
// indicates that the we cannot use window scaling on the
// receive end either.
WS: -1,
}
for i := 0; i < limit; {
switch opts[i] {
case TCPOptionEOL:
i = limit
case TCPOptionNOP:
i++
case TCPOptionMSS:
if i+4 > limit || opts[i+1] != 4 {
return synOpts
}
mss := uint16(opts[i+2])<<8 | uint16(opts[i+3])
if mss == 0 {
return synOpts
}
synOpts.MSS = mss
if mss < TCPMinimumSendMSS {
synOpts.MSS = TCPMinimumSendMSS
}
i += 4
case TCPOptionWS:
if i+3 > limit || opts[i+1] != 3 {
return synOpts
}
ws := int(opts[i+2])
if ws > MaxWndScale {
ws = MaxWndScale
}
synOpts.WS = ws
i += 3
case TCPOptionTS:
if i+10 > limit || opts[i+1] != 10 {
return synOpts
}
synOpts.TSVal = binary.BigEndian.Uint32(opts[i+2:])
if isAck {
// If the segment is a SYN-ACK then store the Timestamp Echo Reply
// in the segment.
synOpts.TSEcr = binary.BigEndian.Uint32(opts[i+6:])
}
synOpts.TS = true
i += 10
case TCPOptionSACKPermitted:
if i+2 > limit || opts[i+1] != 2 {
return synOpts
}
synOpts.SACKPermitted = true
i += 2
default:
// We don't recognize this option, just skip over it.
if i+2 > limit {
return synOpts
}
l := int(opts[i+1])
// If the length is incorrect or if l+i overflows the
// total options length then return false.
if l < 2 || i+l > limit {
return synOpts
}
i += l
}
}
return synOpts
}
// ParseTCPOptions extracts and stores all known options in the provided byte
// slice in a TCPOptions structure.
func ParseTCPOptions(b []byte) TCPOptions {
opts := TCPOptions{}
limit := len(b)
for i := 0; i < limit; {
switch b[i] {
case TCPOptionEOL:
i = limit
case TCPOptionNOP:
i++
case TCPOptionTS:
if i+10 > limit || (b[i+1] != 10) {
return opts
}
opts.TS = true
opts.TSVal = binary.BigEndian.Uint32(b[i+2:])
opts.TSEcr = binary.BigEndian.Uint32(b[i+6:])
i += 10
case TCPOptionSACK:
if i+2 > limit {
// Malformed SACK block, just return and stop parsing.
return opts
}
sackOptionLen := int(b[i+1])
if i+sackOptionLen > limit || (sackOptionLen-2)%8 != 0 {
// Malformed SACK block, just return and stop parsing.
return opts
}
numBlocks := (sackOptionLen - 2) / 8
opts.SACKBlocks = []SACKBlock{}
for j := 0; j < numBlocks; j++ {
start := binary.BigEndian.Uint32(b[i+2+j*8:])
end := binary.BigEndian.Uint32(b[i+2+j*8+4:])
opts.SACKBlocks = append(opts.SACKBlocks, SACKBlock{
Start: seqnum.Value(start),
End: seqnum.Value(end),
})
}
i += sackOptionLen
default:
// We don't recognize this option, just skip over it.
if i+2 > limit {
return opts
}
l := int(b[i+1])
// If the length is incorrect or if l+i overflows the
// total options length then return false.
if l < 2 || i+l > limit {
return opts
}
i += l
}
}
return opts
}
// EncodeMSSOption encodes the MSS TCP option with the provided MSS values in
// the supplied buffer. If the provided buffer is not large enough then it just
// returns without encoding anything. It returns the number of bytes written to
// the provided buffer.
func EncodeMSSOption(mss uint32, b []byte) int {
if len(b) < TCPOptionMSSLength {
return 0
}
b[0], b[1], b[2], b[3] = TCPOptionMSS, TCPOptionMSSLength, byte(mss>>8), byte(mss)
return TCPOptionMSSLength
}
// EncodeWSOption encodes the WS TCP option with the WS value in the
// provided buffer. If the provided buffer is not large enough then it just
// returns without encoding anything. It returns the number of bytes written to
// the provided buffer.
func EncodeWSOption(ws int, b []byte) int {
if len(b) < TCPOptionWSLength {
return 0
}
b[0], b[1], b[2] = TCPOptionWS, TCPOptionWSLength, uint8(ws)
return int(b[1])
}
// EncodeTSOption encodes the provided tsVal and tsEcr values as a TCP timestamp
// option into the provided buffer. If the buffer is smaller than expected it
// just returns without encoding anything. It returns the number of bytes
// written to the provided buffer.
func EncodeTSOption(tsVal, tsEcr uint32, b []byte) int {
if len(b) < TCPOptionTSLength {
return 0
}
b[0], b[1] = TCPOptionTS, TCPOptionTSLength
binary.BigEndian.PutUint32(b[2:], tsVal)
binary.BigEndian.PutUint32(b[6:], tsEcr)
return int(b[1])
}
// EncodeSACKPermittedOption encodes a SACKPermitted option into the provided
// buffer. If the buffer is smaller than required it just returns without
// encoding anything. It returns the number of bytes written to the provided
// buffer.
func EncodeSACKPermittedOption(b []byte) int {
if len(b) < TCPOptionSackPermittedLength {
return 0
}
b[0], b[1] = TCPOptionSACKPermitted, TCPOptionSackPermittedLength
return int(b[1])
}
// EncodeSACKBlocks encodes the provided SACK blocks as a TCP SACK option block
// in the provided slice. It tries to fit in as many blocks as possible based on
// number of bytes available in the provided buffer. It returns the number of
// bytes written to the provided buffer.
func EncodeSACKBlocks(sackBlocks []SACKBlock, b []byte) int {
if len(sackBlocks) == 0 {
return 0
}
l := len(sackBlocks)
if l > TCPMaxSACKBlocks {
l = TCPMaxSACKBlocks
}
if ll := (len(b) - 2) / 8; ll < l {
l = ll
}
if l == 0 {
// There is not enough space in the provided buffer to add
// any SACK blocks.
return 0
}
b[0] = TCPOptionSACK
b[1] = byte(l*8 + 2)
for i := 0; i < l; i++ {
binary.BigEndian.PutUint32(b[i*8+2:], uint32(sackBlocks[i].Start))
binary.BigEndian.PutUint32(b[i*8+6:], uint32(sackBlocks[i].End))
}
return int(b[1])
}
// EncodeNOP adds an explicit NOP to the option list.
func EncodeNOP(b []byte) int {
if len(b) == 0 {
return 0
}
b[0] = TCPOptionNOP
return 1
}
// AddTCPOptionPadding adds the required number of TCPOptionNOP to quad align
// the option buffer. It adds padding bytes after the offset specified and
// returns the number of padding bytes added. The passed in options slice
// must have space for the padding bytes.
func AddTCPOptionPadding(options []byte, offset int) int {
paddingToAdd := -offset & 3
// Now add any padding bytes that might be required to quad align the
// options.
for i := offset; i < offset+paddingToAdd; i++ {
options[i] = TCPOptionNOP
}
return paddingToAdd
}
// Acceptable checks if a segment that starts at segSeq and has length segLen is
// "acceptable" for arriving in a receive window that starts at rcvNxt and ends
// before rcvAcc, according to the table on page 26 and 69 of RFC 793.
func Acceptable(segSeq seqnum.Value, segLen seqnum.Size, rcvNxt, rcvAcc seqnum.Value) bool {
if rcvNxt == rcvAcc {
return segLen == 0 && segSeq == rcvNxt
}
if segLen == 0 {
// rcvWnd is incremented by 1 because that is Linux's behavior despite the
// RFC.
return segSeq.InRange(rcvNxt, rcvAcc.Add(1))
}
// Page 70 of RFC 793 allows packets that can be made "acceptable" by trimming
// the payload, so we'll accept any payload that overlaps the receive window.
// segSeq < rcvAcc is more correct according to RFC, however, Linux does it
// differently, it uses segSeq <= rcvAcc, we'd want to keep the same behavior
// as Linux.
return rcvNxt.LessThan(segSeq.Add(segLen)) && segSeq.LessThanEq(rcvAcc)
}
// TCPValid returns true if the pkt has a valid TCP header. It checks whether:
// - The data offset is too small.
// - The data offset is too large.
// - The checksum is invalid.
//
// TCPValid corresponds to net/netfilter/nf_conntrack_proto_tcp.c:tcp_error.
func TCPValid(hdr TCP, payloadChecksum func() uint16, payloadSize uint16, srcAddr, dstAddr tcpip.Address, skipChecksumValidation bool) (csum uint16, csumValid, ok bool) {
if offset := int(hdr.DataOffset()); offset < TCPMinimumSize || offset > len(hdr) {
return
}
if skipChecksumValidation {
csumValid = true
} else {
csum = hdr.Checksum()
csumValid = hdr.IsChecksumValid(srcAddr, dstAddr, payloadChecksum(), payloadSize)
}
return csum, csumValid, true
}
|