diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-05-05 17:41:34 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-05-05 17:41:34 +0000 |
commit | f7b5fc832288fae3f5a0f3d24b96782252aabbaf (patch) | |
tree | 711bd849ca1348d3c97dc9bbbd157b80e9c410aa /extract.h | |
parent | Initial commit. (diff) | |
download | tcpdump-f7b5fc832288fae3f5a0f3d24b96782252aabbaf.tar.xz tcpdump-f7b5fc832288fae3f5a0f3d24b96782252aabbaf.zip |
Adding upstream version 4.99.3.upstream/4.99.3upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r-- | extract.h | 910 |
1 files changed, 910 insertions, 0 deletions
diff --git a/extract.h b/extract.h new file mode 100644 index 0000000..d7d803d --- /dev/null +++ b/extract.h @@ -0,0 +1,910 @@ +/* + * Copyright (c) 1992, 1993, 1994, 1995, 1996 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that: (1) source code distributions + * retain the above copyright notice and this paragraph in its entirety, (2) + * distributions including binary code include the above copyright notice and + * this paragraph in its entirety in the documentation or other materials + * provided with the distribution, and (3) all advertising materials mentioning + * features or use of this software display the following acknowledgement: + * ``This product includes software developed by the University of California, + * Lawrence Berkeley Laboratory and its contributors.'' Neither the name of + * the University nor the names of its contributors may be used to endorse + * or promote products derived from this software without specific prior + * written permission. + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED + * WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. + */ + +#ifndef EXTRACT_H +#define EXTRACT_H + +#include <string.h> + +/* + * For 8-bit values; needed to fetch a one-byte value. Byte order + * isn't relevant, and alignment isn't an issue. + */ +#define EXTRACT_U_1(p) ((uint8_t)(*(p))) +#define EXTRACT_S_1(p) ((int8_t)(*(p))) + +/* + * Inline functions or macros to extract possibly-unaligned big-endian + * integral values. + */ +#include "funcattrs.h" +#include "netdissect.h" +#include "diag-control.h" + +/* + * If we have versions of GCC or Clang that support an __attribute__ + * to say "if we're building with unsigned behavior sanitization, + * don't complain about undefined behavior in this function", we + * label these functions with that attribute - we *know* it's undefined + * in the C standard, but we *also* know it does what we want with + * the ISA we're targeting and the compiler we're using. + * + * For GCC 4.9.0 and later, we use __attribute__((no_sanitize_undefined)); + * pre-5.0 GCC doesn't have __has_attribute, and I'm not sure whether + * GCC or Clang first had __attribute__((no_sanitize(XXX)). + * + * For Clang, we check for __attribute__((no_sanitize(XXX)) with + * __has_attribute, as there are versions of Clang that support + * __attribute__((no_sanitize("undefined")) but don't support + * __attribute__((no_sanitize_undefined)). + * + * We define this here, rather than in funcattrs.h, because we + * only want it used here, we don't want it to be broadly used. + * (Any printer will get this defined, but this should at least + * make it harder for people to find.) + */ +#if defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 409) +#define UNALIGNED_OK __attribute__((no_sanitize_undefined)) +#elif __has_attribute(no_sanitize) +#define UNALIGNED_OK __attribute__((no_sanitize("undefined"))) +#else +#define UNALIGNED_OK +#endif + +#if (defined(__i386__) || defined(_M_IX86) || defined(__X86__) || defined(__x86_64__) || defined(_M_X64)) || \ + (defined(__m68k__) && (!defined(__mc68000__) && !defined(__mc68010__))) || \ + (defined(__ppc__) || defined(__ppc64__) || defined(_M_PPC) || defined(_ARCH_PPC) || defined(_ARCH_PPC64)) || \ + (defined(__s390__) || defined(__s390x__) || defined(__zarch__)) +/* + * The processor natively handles unaligned loads, so we can just + * cast the pointer and fetch through it. + * + * XXX - are those all the x86 tests we need? + * XXX - are those the only 68k tests we need not to generated + * unaligned accesses if the target is the 68000 or 68010? + * XXX - are there any tests we don't need, because some definitions are for + * compilers that also predefine the GCC symbols? + * XXX - do we need to test for both 32-bit and 64-bit versions of those + * architectures in all cases? + */ +UNALIGNED_OK static inline uint16_t +EXTRACT_BE_U_2(const void *p) +{ + return ((uint16_t)ntohs(*(const uint16_t *)(p))); +} + +UNALIGNED_OK static inline int16_t +EXTRACT_BE_S_2(const void *p) +{ + return ((int16_t)ntohs(*(const int16_t *)(p))); +} + +UNALIGNED_OK static inline uint32_t +EXTRACT_BE_U_4(const void *p) +{ + return ((uint32_t)ntohl(*(const uint32_t *)(p))); +} + +UNALIGNED_OK static inline int32_t +EXTRACT_BE_S_4(const void *p) +{ + return ((int32_t)ntohl(*(const int32_t *)(p))); +} + +UNALIGNED_OK static inline uint64_t +EXTRACT_BE_U_8(const void *p) +{ + return ((uint64_t)(((uint64_t)ntohl(*((const uint32_t *)(p) + 0))) << 32 | + ((uint64_t)ntohl(*((const uint32_t *)(p) + 1))) << 0)); + +} + +UNALIGNED_OK static inline int64_t +EXTRACT_BE_S_8(const void *p) +{ + return ((int64_t)(((int64_t)ntohl(*((const uint32_t *)(p) + 0))) << 32 | + ((uint64_t)ntohl(*((const uint32_t *)(p) + 1))) << 0)); + +} + +/* + * Extract an IPv4 address, which is in network byte order, and not + * necessarily aligned, and provide the result in host byte order. + */ +UNALIGNED_OK static inline uint32_t +EXTRACT_IPV4_TO_HOST_ORDER(const void *p) +{ + return ((uint32_t)ntohl(*(const uint32_t *)(p))); +} +#elif ND_IS_AT_LEAST_GNUC_VERSION(2,0) && \ + (defined(__alpha) || defined(__alpha__) || \ + defined(__mips) || defined(__mips__)) +/* + * This is MIPS or Alpha, which don't natively handle unaligned loads, + * but which have instructions that can help when doing unaligned + * loads, and this is GCC 2.0 or later or a compiler that claims to + * be GCC 2.0 or later, which we assume that mean we have + * __attribute__((packed)), which we can use to convince the compiler + * to generate those instructions. + * + * Declare packed structures containing a uint16_t and a uint32_t, + * cast the pointer to point to one of those, and fetch through it; + * the GCC manual doesn't appear to explicitly say that + * __attribute__((packed)) causes the compiler to generate unaligned-safe + * code, but it appears to do so. + * + * We do this in case the compiler can generate code using those + * instructions to do an unaligned load and pass stuff to "ntohs()" or + * "ntohl()", which might be better than the code to fetch the + * bytes one at a time and assemble them. (That might not be the + * case on a little-endian platform, such as DEC's MIPS machines and + * Alpha machines, where "ntohs()" and "ntohl()" might not be done + * inline.) + * + * We do this only for specific architectures because, for example, + * at least some versions of GCC, when compiling for 64-bit SPARC, + * generate code that assumes alignment if we do this. + * + * XXX - add other architectures and compilers as possible and + * appropriate. + * + * HP's C compiler, indicated by __HP_cc being defined, supports + * "#pragma unaligned N" in version A.05.50 and later, where "N" + * specifies a number of bytes at which the typedef on the next + * line is aligned, e.g. + * + * #pragma unalign 1 + * typedef uint16_t unaligned_uint16_t; + * + * to define unaligned_uint16_t as a 16-bit unaligned data type. + * This could be presumably used, in sufficiently recent versions of + * the compiler, with macros similar to those below. This would be + * useful only if that compiler could generate better code for PA-RISC + * or Itanium than would be generated by a bunch of shifts-and-ORs. + * + * DEC C, indicated by __DECC being defined, has, at least on Alpha, + * an __unaligned qualifier that can be applied to pointers to get the + * compiler to generate code that does unaligned loads and stores when + * dereferencing the pointer in question. + * + * XXX - what if the native C compiler doesn't support + * __attribute__((packed))? How can we get it to generate unaligned + * accesses for *specific* items? + */ +typedef struct { + uint16_t val; +} __attribute__((packed)) unaligned_uint16_t; + +typedef struct { + int16_t val; +} __attribute__((packed)) unaligned_int16_t; + +typedef struct { + uint32_t val; +} __attribute__((packed)) unaligned_uint32_t; + +typedef struct { + int32_t val; +} __attribute__((packed)) unaligned_int32_t; + +UNALIGNED_OK static inline uint16_t +EXTRACT_BE_U_2(const void *p) +{ + return ((uint16_t)ntohs(((const unaligned_uint16_t *)(p))->val)); +} + +UNALIGNED_OK static inline int16_t +EXTRACT_BE_S_2(const void *p) +{ + return ((int16_t)ntohs(((const unaligned_int16_t *)(p))->val)); +} + +UNALIGNED_OK static inline uint32_t +EXTRACT_BE_U_4(const void *p) +{ + return ((uint32_t)ntohl(((const unaligned_uint32_t *)(p))->val)); +} + +UNALIGNED_OK static inline int32_t +EXTRACT_BE_S_4(const void *p) +{ + return ((int32_t)ntohl(((const unaligned_int32_t *)(p))->val)); +} + +UNALIGNED_OK static inline uint64_t +EXTRACT_BE_U_8(const void *p) +{ + return ((uint64_t)(((uint64_t)ntohl(((const unaligned_uint32_t *)(p) + 0)->val)) << 32 | + ((uint64_t)ntohl(((const unaligned_uint32_t *)(p) + 1)->val)) << 0)); +} + +UNALIGNED_OK static inline int64_t +EXTRACT_BE_S_8(const void *p) +{ + return ((int64_t)(((uint64_t)ntohl(((const unaligned_uint32_t *)(p) + 0)->val)) << 32 | + ((uint64_t)ntohl(((const unaligned_uint32_t *)(p) + 1)->val)) << 0)); +} + +/* + * Extract an IPv4 address, which is in network byte order, and not + * necessarily aligned, and provide the result in host byte order. + */ +UNALIGNED_OK static inline uint32_t +EXTRACT_IPV4_TO_HOST_ORDER(const void *p) +{ + return ((uint32_t)ntohl(((const unaligned_uint32_t *)(p))->val)); +} +#else +/* + * This architecture doesn't natively support unaligned loads, and either + * this isn't a GCC-compatible compiler, we don't have __attribute__, + * or we do but we don't know of any better way with this instruction + * set to do unaligned loads, so do unaligned loads of big-endian + * quantities the hard way - fetch the bytes one at a time and + * assemble them. + * + * XXX - ARM is a special case. ARMv1 through ARMv5 didn't suppory + * unaligned loads; ARMv6 and later support it *but* have a bit in + * the system control register that the OS can set and that causes + * unaligned loads to fault rather than succeeding. + * + * At least some OSes may set that flag, so we do *not* treat ARM + * as supporting unaligned loads. If your OS supports them on ARM, + * and you want to use them, please update the tests in the #if above + * to check for ARM *and* for your OS. + */ +#define EXTRACT_BE_U_2(p) \ + ((uint16_t)(((uint16_t)(*((const uint8_t *)(p) + 0)) << 8) | \ + ((uint16_t)(*((const uint8_t *)(p) + 1)) << 0))) +#define EXTRACT_BE_S_2(p) \ + ((int16_t)(((uint16_t)(*((const uint8_t *)(p) + 0)) << 8) | \ + ((uint16_t)(*((const uint8_t *)(p) + 1)) << 0))) +#define EXTRACT_BE_U_4(p) \ + ((uint32_t)(((uint32_t)(*((const uint8_t *)(p) + 0)) << 24) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 3)) << 0))) +#define EXTRACT_BE_S_4(p) \ + ((int32_t)(((uint32_t)(*((const uint8_t *)(p) + 0)) << 24) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 3)) << 0))) +#define EXTRACT_BE_U_8(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 56) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 7)) << 0))) +#define EXTRACT_BE_S_8(p) \ + ((int64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 56) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 7)) << 0))) + +/* + * Extract an IPv4 address, which is in network byte order, and not + * necessarily aligned, and provide the result in host byte order. + */ +#define EXTRACT_IPV4_TO_HOST_ORDER(p) \ + ((uint32_t)(((uint32_t)(*((const uint8_t *)(p) + 0)) << 24) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 3)) << 0))) +#endif /* unaligned access checks */ + +/* + * Extract numerical values in *host* byte order. (Some metadata + * headers are in the byte order of the host that wrote the file, + * and libpcap translate them to the byte order of the host + * reading the file. This means that if a program on that host + * reads with libpcap and writes to a new file, the new file will + * be written in the byte order of the host writing the file. Thus, + * the magic number in pcap files and byte-order magic in pcapng + * files can be used to determine the byte order in those metadata + * headers.) + * + * XXX - on platforms that can do unaligned accesses, just cast and + * dereference the pointer. + */ +static inline uint16_t +EXTRACT_HE_U_2(const void *p) +{ + uint16_t val; + + UNALIGNED_MEMCPY(&val, p, sizeof(uint16_t)); + return val; +} + +static inline int16_t +EXTRACT_HE_S_2(const void *p) +{ + int16_t val; + + UNALIGNED_MEMCPY(&val, p, sizeof(int16_t)); + return val; +} + +static inline uint32_t +EXTRACT_HE_U_4(const void *p) +{ + uint32_t val; + + UNALIGNED_MEMCPY(&val, p, sizeof(uint32_t)); + return val; +} + +static inline int32_t +EXTRACT_HE_S_4(const void *p) +{ + int32_t val; + + UNALIGNED_MEMCPY(&val, p, sizeof(int32_t)); + return val; +} + +/* + * Extract an IPv4 address, which is in network byte order, and which + * is not necessarily aligned on a 4-byte boundary, and provide the + * result in network byte order. + * + * This works the same way regardless of the host's byte order. + */ +static inline uint32_t +EXTRACT_IPV4_TO_NETWORK_ORDER(const void *p) +{ + uint32_t addr; + + UNALIGNED_MEMCPY(&addr, p, sizeof(uint32_t)); + return addr; +} + +/* + * Non-power-of-2 sizes. + */ +#define EXTRACT_BE_U_3(p) \ + ((uint32_t)(((uint32_t)(*((const uint8_t *)(p) + 0)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 0))) + +#define EXTRACT_BE_S_3(p) \ + (((*((const uint8_t *)(p) + 0)) & 0x80) ? \ + ((int32_t)(((uint32_t)(*((const uint8_t *)(p) + 0)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 0))) : \ + ((int32_t)(0xFF000000U | \ + ((uint32_t)(*((const uint8_t *)(p) + 0)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 0)))) + +#define EXTRACT_BE_U_5(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 0))) + +#define EXTRACT_BE_S_5(p) \ + (((*((const uint8_t *)(p) + 0)) & 0x80) ? \ + ((int64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 0))) : \ + ((int64_t)(INT64_T_CONSTANT(0xFFFFFF0000000000U) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 0)))) + +#define EXTRACT_BE_U_6(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 0))) + +#define EXTRACT_BE_S_6(p) \ + (((*((const uint8_t *)(p) + 0)) & 0x80) ? \ + ((int64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 0))) : \ + ((int64_t)(INT64_T_CONSTANT(0xFFFFFFFF00000000U) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 0)))) + +#define EXTRACT_BE_U_7(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 0))) + +#define EXTRACT_BE_S_7(p) \ + (((*((const uint8_t *)(p) + 0)) & 0x80) ? \ + ((int64_t)(((uint64_t)(*((const uint8_t *)(p) + 0)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 0))) : \ + ((int64_t)(INT64_T_CONSTANT(0xFFFFFFFFFF000000U) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 0)))) + +/* + * Macros to extract possibly-unaligned little-endian integral values. + * XXX - do loads on little-endian machines that support unaligned loads? + */ +#define EXTRACT_LE_U_2(p) \ + ((uint16_t)(((uint16_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint16_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_S_2(p) \ + ((int16_t)(((uint16_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint16_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_U_4(p) \ + ((uint32_t)(((uint32_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_S_4(p) \ + ((int32_t)(((uint32_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint32_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_U_8(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 7)) << 56) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_S_8(p) \ + ((int64_t)(((uint64_t)(*((const uint8_t *)(p) + 7)) << 56) | \ + ((uint64_t)(*((const uint8_t *)(p) + 6)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 0))) + +/* + * Non-power-of-2 sizes. + */ + +#define EXTRACT_LE_U_3(p) \ + ((uint32_t)(((uint32_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_S_3(p) \ + ((int32_t)(((uint32_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint32_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint32_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_U_5(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 4)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_U_6(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 5)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 0))) +#define EXTRACT_LE_U_7(p) \ + ((uint64_t)(((uint64_t)(*((const uint8_t *)(p) + 6)) << 48) | \ + ((uint64_t)(*((const uint8_t *)(p) + 5)) << 40) | \ + ((uint64_t)(*((const uint8_t *)(p) + 4)) << 32) | \ + ((uint64_t)(*((const uint8_t *)(p) + 3)) << 24) | \ + ((uint64_t)(*((const uint8_t *)(p) + 2)) << 16) | \ + ((uint64_t)(*((const uint8_t *)(p) + 1)) << 8) | \ + ((uint64_t)(*((const uint8_t *)(p) + 0)) << 0))) + +/* + * Macros to check the presence of the values in question. + */ +#define ND_TTEST_1(p) ND_TTEST_LEN((p), 1) +#define ND_TCHECK_1(p) ND_TCHECK_LEN((p), 1) + +#define ND_TTEST_2(p) ND_TTEST_LEN((p), 2) +#define ND_TCHECK_2(p) ND_TCHECK_LEN((p), 2) + +#define ND_TTEST_3(p) ND_TTEST_LEN((p), 3) +#define ND_TCHECK_3(p) ND_TCHECK_LEN((p), 3) + +#define ND_TTEST_4(p) ND_TTEST_LEN((p), 4) +#define ND_TCHECK_4(p) ND_TCHECK_LEN((p), 4) + +#define ND_TTEST_5(p) ND_TTEST_LEN((p), 5) +#define ND_TCHECK_5(p) ND_TCHECK_LEN((p), 5) + +#define ND_TTEST_6(p) ND_TTEST_LEN((p), 6) +#define ND_TCHECK_6(p) ND_TCHECK_LEN((p), 6) + +#define ND_TTEST_7(p) ND_TTEST_LEN((p), 7) +#define ND_TCHECK_7(p) ND_TCHECK_LEN((p), 7) + +#define ND_TTEST_8(p) ND_TTEST_LEN((p), 8) +#define ND_TCHECK_8(p) ND_TCHECK_LEN((p), 8) + +#define ND_TTEST_16(p) ND_TTEST_LEN((p), 16) +#define ND_TCHECK_16(p) ND_TCHECK_LEN((p), 16) + +/* get_u_1 and get_s_1 */ + +static inline uint8_t +get_u_1(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_1(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_U_1(p); +} + +static inline int8_t +get_s_1(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_1(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_S_1(p); +} + +/* get_be_u_N */ + +static inline uint16_t +get_be_u_2(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_2(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_2(p); +} + +static inline uint32_t +get_be_u_3(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_3(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_3(p); +} + +static inline uint32_t +get_be_u_4(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_4(p); +} + +static inline uint64_t +get_be_u_5(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_5(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_5(p); +} + +static inline uint64_t +get_be_u_6(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_6(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_6(p); +} + +static inline uint64_t +get_be_u_7(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_7(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_7(p); +} + +static inline uint64_t +get_be_u_8(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_8(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_U_8(p); +} + +/* get_be_s_N */ + +static inline int16_t +get_be_s_2(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_2(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_2(p); +} + +static inline int32_t +get_be_s_3(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_3(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_3(p); +} + +static inline int32_t +get_be_s_4(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_4(p); +} + +static inline int64_t +get_be_s_5(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_5(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_5(p); +} + +static inline int64_t +get_be_s_6(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_6(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_6(p); +} + +static inline int64_t +get_be_s_7(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_7(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_7(p); +} + +static inline int64_t +get_be_s_8(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_8(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_BE_S_8(p); +} + +/* get_he_u_N */ + +static inline uint16_t +get_he_u_2(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_2(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_HE_U_2(p); +} + +static inline uint32_t +get_he_u_4(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_HE_U_4(p); +} + +/* get_he_s_N */ + +static inline int16_t +get_he_s_2(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_2(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_HE_S_2(p); +} + +static inline int32_t +get_he_s_4(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_HE_S_4(p); +} + +/* get_le_u_N */ + +static inline uint16_t +get_le_u_2(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_2(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_2(p); +} + +static inline uint32_t +get_le_u_3(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_3(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_3(p); +} + +static inline uint32_t +get_le_u_4(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_4(p); +} + +static inline uint64_t +get_le_u_5(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_5(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_5(p); +} + +static inline uint64_t +get_le_u_6(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_6(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_6(p); +} + +static inline uint64_t +get_le_u_7(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_7(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_7(p); +} + +static inline uint64_t +get_le_u_8(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_8(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_U_8(p); +} + +/* get_le_s_N */ + +static inline int16_t +get_le_s_2(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_2(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_S_2(p); +} + +static inline int32_t +get_le_s_3(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_3(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_S_3(p); +} + +static inline int32_t +get_le_s_4(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_S_4(p); +} + +static inline int64_t +get_le_s_8(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_8(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_LE_S_8(p); +} + +/* get_ipv4_to_{host|network]_order */ + +static inline uint32_t +get_ipv4_to_host_order(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_IPV4_TO_HOST_ORDER(p); +} + +static inline uint32_t +get_ipv4_to_network_order(netdissect_options *ndo, const u_char *p) +{ + if (!ND_TTEST_4(p)) + nd_trunc_longjmp(ndo); + return EXTRACT_IPV4_TO_NETWORK_ORDER(p); +} + +static inline void +get_cpy_bytes(netdissect_options *ndo, u_char *dst, const u_char *p, size_t len) +{ + if (!ND_TTEST_LEN(p, len)) + nd_trunc_longjmp(ndo); + UNALIGNED_MEMCPY(dst, p, len); +} + +#define GET_U_1(p) get_u_1(ndo, (const u_char *)(p)) +#define GET_S_1(p) get_s_1(ndo, (const u_char *)(p)) + +#define GET_BE_U_2(p) get_be_u_2(ndo, (const u_char *)(p)) +#define GET_BE_U_3(p) get_be_u_3(ndo, (const u_char *)(p)) +#define GET_BE_U_4(p) get_be_u_4(ndo, (const u_char *)(p)) +#define GET_BE_U_5(p) get_be_u_5(ndo, (const u_char *)(p)) +#define GET_BE_U_6(p) get_be_u_6(ndo, (const u_char *)(p)) +#define GET_BE_U_7(p) get_be_u_7(ndo, (const u_char *)(p)) +#define GET_BE_U_8(p) get_be_u_8(ndo, (const u_char *)(p)) + +#define GET_BE_S_2(p) get_be_s_2(ndo, (const u_char *)(p)) +#define GET_BE_S_3(p) get_be_s_3(ndo, (const u_char *)(p)) +#define GET_BE_S_4(p) get_be_s_4(ndo, (const u_char *)(p)) +#define GET_BE_S_5(p) get_be_s_5(ndo, (const u_char *)(p)) +#define GET_BE_S_6(p) get_be_s_6(ndo, (const u_char *)(p)) +#define GET_BE_S_7(p) get_be_s_7(ndo, (const u_char *)(p)) +#define GET_BE_S_8(p) get_be_s_8(ndo, (const u_char *)(p)) + +#define GET_HE_U_2(p) get_he_u_2(ndo, (const u_char *)(p)) +#define GET_HE_U_4(p) get_he_u_4(ndo, (const u_char *)(p)) + +#define GET_HE_S_2(p) get_he_s_2(ndo, (const u_char *)(p)) +#define GET_HE_S_4(p) get_he_s_4(ndo, (const u_char *)(p)) + +#define GET_LE_U_2(p) get_le_u_2(ndo, (const u_char *)(p)) +#define GET_LE_U_3(p) get_le_u_3(ndo, (const u_char *)(p)) +#define GET_LE_U_4(p) get_le_u_4(ndo, (const u_char *)(p)) +#define GET_LE_U_5(p) get_le_u_5(ndo, (const u_char *)(p)) +#define GET_LE_U_6(p) get_le_u_6(ndo, (const u_char *)(p)) +#define GET_LE_U_7(p) get_le_u_7(ndo, (const u_char *)(p)) +#define GET_LE_U_8(p) get_le_u_8(ndo, (const u_char *)(p)) + +#define GET_LE_S_2(p) get_le_s_2(ndo, (const u_char *)(p)) +#define GET_LE_S_3(p) get_le_s_3(ndo, (const u_char *)(p)) +#define GET_LE_S_4(p) get_le_s_4(ndo, (const u_char *)(p)) +#define GET_LE_S_8(p) get_le_s_8(ndo, (const u_char *)(p)) + +#define GET_IPV4_TO_HOST_ORDER(p) get_ipv4_to_host_order(ndo, (const u_char *)(p)) +#define GET_IPV4_TO_NETWORK_ORDER(p) get_ipv4_to_network_order(ndo, (const u_char *)(p)) + +#define GET_CPY_BYTES(dst, p, len) get_cpy_bytes(ndo, (u_char *)(dst), (const u_char *)(p), len) + +#endif /* EXTRACT_H */ |