Commit | Line | Data |
---|---|---|
f4bcbe79 GS |
1 | #ifndef __LINUX_STRINGHASH_H |
2 | #define __LINUX_STRINGHASH_H | |
3 | ||
4 | #include <linux/types.h> | |
5 | ||
6 | /* | |
7 | * Routines for hashing strings of bytes to a 32-bit hash value. | |
8 | * | |
9 | * These hash functions are NOT GUARANTEED STABLE between kernel | |
10 | * versions, architectures, or even repeated boots of the same kernel. | |
11 | * (E.g. they may depend on boot-time hardware detection or be | |
12 | * deliberately randomized.) | |
13 | * | |
14 | * They are also not intended to be secure against collisions caused by | |
15 | * malicious inputs; much slower hash functions are required for that. | |
16 | * | |
17 | * They are optimized for pathname components, meaning short strings. | |
18 | * Even if a majority of files have longer names, the dynamic profile of | |
19 | * pathname components skews short due to short directory names. | |
20 | * (E.g. /usr/lib/libsesquipedalianism.so.3.141.) | |
21 | */ | |
22 | ||
23 | /* | |
24 | * Version 1: one byte at a time. Example of use: | |
25 | * | |
26 | * unsigned long hash = init_name_hash; | |
27 | * while (*p) | |
28 | * hash = partial_name_hash(tolower(*p++), hash); | |
29 | * hash = end_name_hash(hash); | |
30 | * | |
31 | * Although this is designed for bytes, fs/hfsplus/unicode.c | |
32 | * abuses it to hash 16-bit values. | |
33 | */ | |
34 | ||
35 | /* Hash courtesy of the R5 hash in reiserfs modulo sign bits */ | |
36 | #define init_name_hash() 0 | |
37 | ||
38 | /* partial hash update function. Assume roughly 4 bits per character */ | |
39 | static inline unsigned long | |
40 | partial_name_hash(unsigned long c, unsigned long prevhash) | |
41 | { | |
42 | return (prevhash + (c << 4) + (c >> 4)) * 11; | |
43 | } | |
44 | ||
45 | /* | |
46 | * Finally: cut down the number of bits to a int value (and try to avoid | |
47 | * losing bits) | |
48 | */ | |
49 | static inline unsigned long end_name_hash(unsigned long hash) | |
50 | { | |
51 | return (unsigned int)hash; | |
52 | } | |
53 | ||
54 | /* | |
55 | * Version 2: One word (32 or 64 bits) at a time. | |
56 | * If CONFIG_DCACHE_WORD_ACCESS is defined (meaning <asm/word-at-a-time.h> | |
57 | * exists, which describes major Linux platforms like x86 and ARM), then | |
58 | * this computes a different hash function much faster. | |
59 | * | |
60 | * If not set, this falls back to a wrapper around the preceding. | |
61 | */ | |
62 | extern unsigned int full_name_hash(const unsigned char *, unsigned int); | |
63 | ||
64 | /* | |
65 | * A hash_len is a u64 with the hash of a string in the low | |
66 | * half and the length in the high half. | |
67 | */ | |
68 | #define hashlen_hash(hashlen) ((u32)(hashlen)) | |
69 | #define hashlen_len(hashlen) ((u32)((hashlen) >> 32)) | |
70 | #define hashlen_create(hash, len) ((u64)(len)<<32 | (u32)(hash)) | |
71 | ||
72 | #endif /* __LINUX_STRINGHASH_H */ |