diff options
Diffstat (limited to 'shared/nm-glib-aux/nm-hash-utils.h')
| -rw-r--r-- | shared/nm-glib-aux/nm-hash-utils.h | 315 |
1 files changed, 315 insertions, 0 deletions
diff --git a/shared/nm-glib-aux/nm-hash-utils.h b/shared/nm-glib-aux/nm-hash-utils.h new file mode 100644 index 00000000..3f622f99 --- /dev/null +++ b/shared/nm-glib-aux/nm-hash-utils.h @@ -0,0 +1,315 @@ +/* -*- Mode: C; tab-width: 4; indent-tabs-mode: t; c-basic-offset: 4 -*- */ +/* NetworkManager -- Network link manager + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the + * Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, + * Boston, MA 02110-1301 USA. + * + * (C) Copyright 2017 Red Hat, Inc. + */ + +#ifndef __NM_HASH_UTILS_H__ +#define __NM_HASH_UTILS_H__ + +#include "c-siphash/src/c-siphash.h" +#include "nm-macros-internal.h" + +/*****************************************************************************/ + +void nm_hash_siphash42_init (CSipHash *h, guint static_seed); + +/* Siphash24 of binary buffer @arr and @len, using the randomized seed from + * other NMHash functions. + * + * Note, that this is guaranteed to use siphash42 under the hood (contrary to + * all other NMHash API, which leave this undefined). That matters at the point, + * where the caller needs to be sure that a reasonably strong hasing algorithm + * is used. (Yes, NMHash is all about siphash24, but otherwise that is not promised + * anywhere). + * + * Another difference is, that this returns guint64 (not guint like other NMHash functions). + * + * Another difference is, that this may also return zero (not like nm_hash_complete()). + * + * Then, why not use c_siphash_hash() directly? Because this also uses the randomized, + * per-run hash-seed like nm_hash_init(). So, you get siphash24 with a random + * seed (which is cached for the current run of the program). + */ +static inline guint64 +nm_hash_siphash42 (guint static_seed, const void *ptr, gsize n) +{ + CSipHash h; + + nm_hash_siphash42_init (&h, static_seed); + c_siphash_append (&h, ptr, n); + return c_siphash_finalize (&h); +} + +/*****************************************************************************/ + +struct _NMHashState { + CSipHash _state; +}; + +typedef struct _NMHashState NMHashState; + +guint nm_hash_static (guint static_seed); + +static inline void +nm_hash_init (NMHashState *state, guint static_seed) +{ + nm_assert (state); + + nm_hash_siphash42_init (&state->_state, static_seed); +} + +static inline guint64 +nm_hash_complete_u64 (NMHashState *state) +{ + nm_assert (state); + + /* this returns the native u64 hash value. Note that this differs + * from nm_hash_complete() in two ways: + * + * - the type, guint64 vs. guint. + * - nm_hash_complete() never returns zero. + * + * In practice, nm_hash*() API is implemented via siphash24, so this returns + * the siphash24 value. But that is not guaranteed by the API, and if you need + * siphash24 directly, use c_siphash_*() and nm_hash_siphash42*() API. */ + return c_siphash_finalize (&state->_state); +} + +static inline guint +nm_hash_complete (NMHashState *state) +{ + guint64 h; + + h = nm_hash_complete_u64 (state); + + /* we don't ever want to return a zero hash. + * + * NMPObject requires that in _idx_obj_part(), and it's just a good idea. */ + return (((guint) (h >> 32)) ^ ((guint) h)) ?: 1396707757u; +} + +static inline void +nm_hash_update (NMHashState *state, const void *ptr, gsize n) +{ + nm_assert (state); + nm_assert (ptr); + nm_assert (n > 0); + + /* Note: the data passed in here might be sensitive data (secrets), + * that we should nm_explicty_zero() afterwards. However, since + * we are using siphash24 with a random key, that is not really + * necessary. Something to keep in mind, if we ever move away from + * this hash implementation. */ + c_siphash_append (&state->_state, ptr, n); +} + +#define nm_hash_update_val(state, val) \ + G_STMT_START { \ + typeof (val) _val = (val); \ + \ + nm_hash_update ((state), &_val, sizeof (_val)); \ + } G_STMT_END + +#define nm_hash_update_valp(state, val) \ + nm_hash_update ((state), (val), sizeof (*(val))) \ + +static inline void +nm_hash_update_bool (NMHashState *state, bool val) +{ + nm_hash_update (state, &val, sizeof (val)); +} + +#define _NM_HASH_COMBINE_BOOLS_x_1( t, y) ((y) ? ((t) (1ull << 0)) : ((t) 0ull)) +#define _NM_HASH_COMBINE_BOOLS_x_2( t, y, ...) ((y) ? ((t) (1ull << 1)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_1 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_3( t, y, ...) ((y) ? ((t) (1ull << 2)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_2 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_4( t, y, ...) ((y) ? ((t) (1ull << 3)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_3 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_5( t, y, ...) ((y) ? ((t) (1ull << 4)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_4 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_6( t, y, ...) ((y) ? ((t) (1ull << 5)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_5 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_7( t, y, ...) ((y) ? ((t) (1ull << 6)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_6 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_8( t, y, ...) ((y) ? ((t) (1ull << 7)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_7 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_9( t, y, ...) ((y) ? ((t) (1ull << 8)) : ((t) 0ull)) | (G_STATIC_ASSERT_EXPR (sizeof (t) >= 2), (_NM_HASH_COMBINE_BOOLS_x_8 (t, __VA_ARGS__))) +#define _NM_HASH_COMBINE_BOOLS_x_10(t, y, ...) ((y) ? ((t) (1ull << 9)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_9 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_x_11(t, y, ...) ((y) ? ((t) (1ull << 10)) : ((t) 0ull)) | _NM_HASH_COMBINE_BOOLS_x_10 (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_n2(t, n, ...) _NM_HASH_COMBINE_BOOLS_x_##n (t, __VA_ARGS__) +#define _NM_HASH_COMBINE_BOOLS_n(t, n, ...) _NM_HASH_COMBINE_BOOLS_n2(t, n, __VA_ARGS__) + +#define NM_HASH_COMBINE_BOOLS(type, ...) ((type) (_NM_HASH_COMBINE_BOOLS_n(type, NM_NARG (__VA_ARGS__), __VA_ARGS__))) + +#define nm_hash_update_bools(state, ...) \ + nm_hash_update_val (state, NM_HASH_COMBINE_BOOLS (guint8, __VA_ARGS__)) + +#define _NM_HASH_COMBINE_VALS_typ_x_1( y) typeof (y) _v1; +#define _NM_HASH_COMBINE_VALS_typ_x_2( y, ...) typeof (y) _v2; _NM_HASH_COMBINE_VALS_typ_x_1 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_3( y, ...) typeof (y) _v3; _NM_HASH_COMBINE_VALS_typ_x_2 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_4( y, ...) typeof (y) _v4; _NM_HASH_COMBINE_VALS_typ_x_3 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_5( y, ...) typeof (y) _v5; _NM_HASH_COMBINE_VALS_typ_x_4 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_6( y, ...) typeof (y) _v6; _NM_HASH_COMBINE_VALS_typ_x_5 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_7( y, ...) typeof (y) _v7; _NM_HASH_COMBINE_VALS_typ_x_6 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_8( y, ...) typeof (y) _v8; _NM_HASH_COMBINE_VALS_typ_x_7 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_9( y, ...) typeof (y) _v9; _NM_HASH_COMBINE_VALS_typ_x_8 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_10(y, ...) typeof (y) _v10; _NM_HASH_COMBINE_VALS_typ_x_9 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_11(y, ...) typeof (y) _v11; _NM_HASH_COMBINE_VALS_typ_x_10 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_12(y, ...) typeof (y) _v12; _NM_HASH_COMBINE_VALS_typ_x_11 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_13(y, ...) typeof (y) _v13; _NM_HASH_COMBINE_VALS_typ_x_12 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_14(y, ...) typeof (y) _v14; _NM_HASH_COMBINE_VALS_typ_x_13 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_15(y, ...) typeof (y) _v15; _NM_HASH_COMBINE_VALS_typ_x_14 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_16(y, ...) typeof (y) _v16; _NM_HASH_COMBINE_VALS_typ_x_15 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_17(y, ...) typeof (y) _v17; _NM_HASH_COMBINE_VALS_typ_x_16 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_18(y, ...) typeof (y) _v18; _NM_HASH_COMBINE_VALS_typ_x_17 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_19(y, ...) typeof (y) _v19; _NM_HASH_COMBINE_VALS_typ_x_18 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_x_20(y, ...) typeof (y) _v20; _NM_HASH_COMBINE_VALS_typ_x_19 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_n2(n, ...) _NM_HASH_COMBINE_VALS_typ_x_##n (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_typ_n(n, ...) _NM_HASH_COMBINE_VALS_typ_n2(n, __VA_ARGS__) + +#define _NM_HASH_COMBINE_VALS_val_x_1( y) ._v1 = (y), +#define _NM_HASH_COMBINE_VALS_val_x_2( y, ...) ._v2 = (y), _NM_HASH_COMBINE_VALS_val_x_1 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_3( y, ...) ._v3 = (y), _NM_HASH_COMBINE_VALS_val_x_2 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_4( y, ...) ._v4 = (y), _NM_HASH_COMBINE_VALS_val_x_3 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_5( y, ...) ._v5 = (y), _NM_HASH_COMBINE_VALS_val_x_4 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_6( y, ...) ._v6 = (y), _NM_HASH_COMBINE_VALS_val_x_5 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_7( y, ...) ._v7 = (y), _NM_HASH_COMBINE_VALS_val_x_6 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_8( y, ...) ._v8 = (y), _NM_HASH_COMBINE_VALS_val_x_7 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_9( y, ...) ._v9 = (y), _NM_HASH_COMBINE_VALS_val_x_8 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_10(y, ...) ._v10 = (y), _NM_HASH_COMBINE_VALS_val_x_9 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_11(y, ...) ._v11 = (y), _NM_HASH_COMBINE_VALS_val_x_10 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_12(y, ...) ._v12 = (y), _NM_HASH_COMBINE_VALS_val_x_11 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_13(y, ...) ._v13 = (y), _NM_HASH_COMBINE_VALS_val_x_12 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_14(y, ...) ._v14 = (y), _NM_HASH_COMBINE_VALS_val_x_13 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_15(y, ...) ._v15 = (y), _NM_HASH_COMBINE_VALS_val_x_14 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_16(y, ...) ._v16 = (y), _NM_HASH_COMBINE_VALS_val_x_15 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_17(y, ...) ._v17 = (y), _NM_HASH_COMBINE_VALS_val_x_16 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_18(y, ...) ._v18 = (y), _NM_HASH_COMBINE_VALS_val_x_17 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_19(y, ...) ._v19 = (y), _NM_HASH_COMBINE_VALS_val_x_18 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_x_20(y, ...) ._v20 = (y), _NM_HASH_COMBINE_VALS_val_x_19 (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_n2(n, ...) _NM_HASH_COMBINE_VALS_val_x_##n (__VA_ARGS__) +#define _NM_HASH_COMBINE_VALS_val_n(n, ...) _NM_HASH_COMBINE_VALS_val_n2(n, __VA_ARGS__) + +/* NM_HASH_COMBINE_VALS() is faster then nm_hash_update_val() as it combines multiple + * calls to nm_hash_update() using a packed structure. */ +#define NM_HASH_COMBINE_VALS(var, ...) \ + const struct _nm_packed { \ + _NM_HASH_COMBINE_VALS_typ_n (NM_NARG (__VA_ARGS__), __VA_ARGS__) \ + } var _nm_alignas (guint64) = { \ + _NM_HASH_COMBINE_VALS_val_n (NM_NARG (__VA_ARGS__), __VA_ARGS__) \ + } + +/* nm_hash_update_vals() is faster then nm_hash_update_val() as it combines multiple + * calls to nm_hash_update() using a packed structure. */ +#define nm_hash_update_vals(state, ...) \ + G_STMT_START { \ + NM_HASH_COMBINE_VALS (_val, __VA_ARGS__); \ + \ + nm_hash_update ((state), &_val, sizeof (_val)); \ + } G_STMT_END + +static inline void +nm_hash_update_mem (NMHashState *state, const void *ptr, gsize n) +{ + /* This also hashes the length of the data. That means, + * hashing two consecutive binary fields (of arbitrary + * length), will hash differently. That is, + * [[1,1], []] differs from [[1],[1]]. + * + * If you have a constant length (sizeof), use nm_hash_update() + * instead. */ + nm_hash_update (state, &n, sizeof (n)); + if (n > 0) + nm_hash_update (state, ptr, n); +} + +static inline void +nm_hash_update_str0 (NMHashState *state, const char *str) +{ + if (str) + nm_hash_update_mem (state, str, strlen (str)); + else { + gsize n = G_MAXSIZE; + + nm_hash_update (state, &n, sizeof (n)); + } +} + +static inline void +nm_hash_update_str (NMHashState *state, const char *str) +{ + nm_assert (str); + nm_hash_update (state, str, strlen (str) + 1); +} + +#if _NM_CC_SUPPORT_GENERIC +/* Like nm_hash_update_str(), but restricted to arrays only. nm_hash_update_str() only works + * with a @str argument that cannot be NULL. If you have a string pointer, that is never NULL, use + * nm_hash_update() instead. */ +#define nm_hash_update_strarr(state, str) \ + (_Generic (&(str), \ + const char (*) [sizeof (str)]: nm_hash_update_str ((state), (str)), \ + char (*) [sizeof (str)]: nm_hash_update_str ((state), (str))) \ + ) +#else +#define nm_hash_update_strarr(state, str) nm_hash_update_str ((state), (str)) +#endif + +guint nm_hash_ptr (gconstpointer ptr); +guint nm_direct_hash (gconstpointer str); + +guint nm_hash_str (const char *str); +guint nm_str_hash (gconstpointer str); + +#define nm_hash_val(static_seed, val) \ + ({ \ + NMHashState _h; \ + \ + nm_hash_init (&_h, (static_seed)); \ + nm_hash_update_val (&_h, (val)); \ + nm_hash_complete (&_h); \ + }) + +/*****************************************************************************/ + +/* nm_pstr_*() are for hashing keys that are pointers to strings, + * that is, "const char *const*" types, using strcmp(). */ + +guint nm_pstr_hash (gconstpointer p); + +gboolean nm_pstr_equal (gconstpointer a, gconstpointer b); + +/*****************************************************************************/ + +#define NM_HASH_OBFUSCATE_PTR_FMT "%016llx" + +/* sometimes we want to log a pointer directly, for providing context/information about + * the message that get logged. Logging pointer values directly defeats ASLR, so we should + * not do that. This returns a "unsigned long long" value that can be used + * instead. + * + * Note that there is a chance that two different pointer values hash to the same obfuscated + * value. So beware of that when reviewing logs. However, such a collision is very unlikely. */ +#define nm_hash_obfuscate_ptr(static_seed, val) \ + ({ \ + NMHashState _h; \ + const void *_val_obf_ptr = (val); \ + \ + nm_hash_init (&_h, (static_seed)); \ + nm_hash_update_val (&_h, _val_obf_ptr); \ + (unsigned long long) nm_hash_complete_u64 (&_h); \ + }) + +/*****************************************************************************/ + +#endif /* __NM_HASH_UTILS_H__ */ |