include/linux/stringhash.h

   1 #ifndef __LINUX_STRINGHASH_H
   2 #define __LINUX_STRINGHASH_H
   3
   4 #include <linux/compiler.h>     /* For __pure */
   5 #include <linux/types.h>        /* For u32, u64 */
   6
   7 /*
   8  * Routines for hashing strings of bytes to a 32-bit hash value.
   9  *
  10  * These hash functions are NOT GUARANTEED STABLE between kernel
  11  * versions, architectures, or even repeated boots of the same kernel.
  12  * (E.g. they may depend on boot-time hardware detection or be
  13  * deliberately randomized.)
  14  *
  15  * They are also not intended to be secure against collisions caused by
  16  * malicious inputs; much slower hash functions are required for that.
  17  *
  18  * They are optimized for pathname components, meaning short strings.
  19  * Even if a majority of files have longer names, the dynamic profile of
  20  * pathname components skews short due to short directory names.
  21  * (E.g. /usr/lib/libsesquipedalianism.so.3.141.)
  22  */
  23
  24 /*
  25  * Version 1: one byte at a time.  Example of use:
  26  *
  27  * unsigned long hash = init_name_hash;
  28  * while (*p)
  29  *      hash = partial_name_hash(tolower(*p++), hash);
  30  * hash = end_name_hash(hash);
  31  *
  32  * Although this is designed for bytes, fs/hfsplus/unicode.c
  33  * abuses it to hash 16-bit values.
  34  */
  35
  36 /* Hash courtesy of the R5 hash in reiserfs modulo sign bits */
  37 #define init_name_hash()                0
  38
  39 /* partial hash update function. Assume roughly 4 bits per character */
  40 static inline unsigned long
  41 partial_name_hash(unsigned long c, unsigned long prevhash)
  42 {
  43         return (prevhash + (c << 4) + (c >> 4)) * 11;
  44 }
  45
  46 /*
  47  * Finally: cut down the number of bits to a int value (and try to avoid
  48  * losing bits)
  49  */
  50 static inline unsigned long end_name_hash(unsigned long hash)
  51 {
  52         return (unsigned int)hash;
  53 }
  54
  55 /*
  56  * Version 2: One word (32 or 64 bits) at a time.
  57  * If CONFIG_DCACHE_WORD_ACCESS is defined (meaning <asm/word-at-a-time.h>
  58  * exists, which describes major Linux platforms like x86 and ARM), then
  59  * this computes a different hash function much faster.
  60  *
  61  * If not set, this falls back to a wrapper around the preceding.
  62  */
  63 extern unsigned int __pure full_name_hash(const char *, unsigned int);
  64
  65 /*
  66  * A hash_len is a u64 with the hash of a string in the low
  67  * half and the length in the high half.
  68  */
  69 #define hashlen_hash(hashlen) ((u32)(hashlen))
  70 #define hashlen_len(hashlen)  ((u32)((hashlen) >> 32))
  71 #define hashlen_create(hash, len) ((u64)(len)<<32 | (u32)(hash))
  72
  73 /* Return the "hash_len" (hash and length) of a null-terminated string */
  74 extern u64 __pure hashlen_string(const char *name);
  75
  76 #endif  /* __LINUX_STRINGHASH_H */