include/linux/stringhash.h

0001 /* SPDX-License-Identifier: GPL-2.0 */
0002 #ifndef __LINUX_STRINGHASH_H
0003 #define __LINUX_STRINGHASH_H
0004
0005 #include <linux/compiler.h> /* For __pure */
0006 #include <linux/types.h>    /* For u32, u64 */
0007 #include <linux/hash.h>
0008
0009 /*
0010  * Routines for hashing strings of bytes to a 32-bit hash value.
0011  *
0012  * These hash functions are NOT GUARANTEED STABLE between kernel
0013  * versions, architectures, or even repeated boots of the same kernel.
0014  * (E.g. they may depend on boot-time hardware detection or be
0015  * deliberately randomized.)
0016  *
0017  * They are also not intended to be secure against collisions caused by
0018  * malicious inputs; much slower hash functions are required for that.
0019  *
0020  * They are optimized for pathname components, meaning short strings.
0021  * Even if a majority of files have longer names, the dynamic profile of
0022  * pathname components skews short due to short directory names.
0023  * (E.g. /usr/lib/libsesquipedalianism.so.3.141.)
0024  */
0025
0026 /*
0027  * Version 1: one byte at a time.  Example of use:
0028  *
0029  * unsigned long hash = init_name_hash;
0030  * while (*p)
0031  *  hash = partial_name_hash(tolower(*p++), hash);
0032  * hash = end_name_hash(hash);
0033  *
0034  * Although this is designed for bytes, fs/hfsplus/unicode.c
0035  * abuses it to hash 16-bit values.
0036  */
0037
0038 /* Hash courtesy of the R5 hash in reiserfs modulo sign bits */
0039 #define init_name_hash(salt)        (unsigned long)(salt)
0040
0041 /* partial hash update function. Assume roughly 4 bits per character */
0042 static inline unsigned long
0043 partial_name_hash(unsigned long c, unsigned long prevhash)
0044 {
0045     return (prevhash + (c << 4) + (c >> 4)) * 11;
0046 }
0047
0048 /*
0049  * Finally: cut down the number of bits to a int value (and try to avoid
0050  * losing bits).  This also has the property (wanted by the dcache)
0051  * that the msbits make a good hash table index.
0052  */
0053 static inline unsigned int end_name_hash(unsigned long hash)
0054 {
0055     return hash_long(hash, 32);
0056 }
0057
0058 /*
0059  * Version 2: One word (32 or 64 bits) at a time.
0060  * If CONFIG_DCACHE_WORD_ACCESS is defined (meaning <asm/word-at-a-time.h>
0061  * exists, which describes major Linux platforms like x86 and ARM), then
0062  * this computes a different hash function much faster.
0063  *
0064  * If not set, this falls back to a wrapper around the preceding.
0065  */
0066 extern unsigned int __pure full_name_hash(const void *salt, const char *, unsigned int);
0067
0068 /*
0069  * A hash_len is a u64 with the hash of a string in the low
0070  * half and the length in the high half.
0071  */
0072 #define hashlen_hash(hashlen) ((u32)(hashlen))
0073 #define hashlen_len(hashlen)  ((u32)((hashlen) >> 32))
0074 #define hashlen_create(hash, len) ((u64)(len)<<32 | (u32)(hash))
0075
0076 /* Return the "hash_len" (hash and length) of a null-terminated string */
0077 extern u64 __pure hashlen_string(const void *salt, const char *name);
0078
0079 #endif  /* __LINUX_STRINGHASH_H */