[03eabf4] | 1 | // -*- Mode: C -*- |
---|
| 2 | // |
---|
| 3 | // Cforall Version 1.0.0 Copyright (C) 2020 University of Waterloo |
---|
| 4 | // |
---|
| 5 | // The contents of this file are covered under the licence agreement in the |
---|
| 6 | // file "LICENCE" distributed with Cforall. |
---|
| 7 | // |
---|
| 8 | // bitmanip.hfa -- |
---|
| 9 | // |
---|
| 10 | // Author : Peter A. Buhr |
---|
| 11 | // Created On : Sat Mar 14 18:12:27 2020 |
---|
| 12 | // Last Modified By : Peter A. Buhr |
---|
[da3b790] | 13 | // Last Modified On : Mon Aug 10 09:21:02 2020 |
---|
| 14 | // Update Count : 139 |
---|
[03eabf4] | 15 | // |
---|
| 16 | |
---|
| 17 | #pragma once |
---|
| 18 | |
---|
| 19 | // Reference: Bit Twiddling Hacks: http://graphics.stanford.edu/%7Eseander/bithacks.html#CountBitsSetNaive |
---|
| 20 | |
---|
| 21 | // Bits are numbered 1-N. |
---|
| 22 | |
---|
[da3b790] | 23 | #include <assert.h> |
---|
[03eabf4] | 24 | |
---|
[76bf23c] | 25 | #define __bitsizeof( n ) (sizeof(n) * __CHAR_BIT__) |
---|
| 26 | |
---|
[03eabf4] | 27 | static inline { |
---|
[66f3bae] | 28 | // Count leading 0 bits. |
---|
| 29 | unsigned int leading0s( unsigned char n ) { return n != 0 ? __builtin_clz( n ) - (__bitsizeof(unsigned int) - __bitsizeof(n)) : __bitsizeof(n); } |
---|
| 30 | unsigned int leading0s( unsigned short int n ) { return n != 0 ? __builtin_clz( n ) - (__bitsizeof(unsigned int) - __bitsizeof(n)) : __bitsizeof(n); } |
---|
| 31 | unsigned int leading0s( unsigned int n ) { return n != 0 ? __builtin_clz( n ) : __bitsizeof(n); } |
---|
| 32 | unsigned int leading0s( unsigned long int n ) { return n != 0 ? __builtin_clzl( n ) : __bitsizeof(n); } |
---|
| 33 | unsigned int leading0s( unsigned long long int n ) { return n != 0 ? __builtin_clzll( n ) : __bitsizeof(n); } |
---|
| 34 | |
---|
| 35 | // Count trailing 0 bits. |
---|
| 36 | unsigned int trailing0s( unsigned char n ) { return n != 0 ? __builtin_ctz( n ) : __bitsizeof(n); } |
---|
| 37 | unsigned int trailing0s( unsigned short int n ) { return n != 0 ? __builtin_ctz( n ) : __bitsizeof(n); } |
---|
| 38 | unsigned int trailing0s( unsigned int n ) { return n != 0 ? __builtin_ctz( n ) : __bitsizeof(n); } |
---|
| 39 | unsigned int trailing0s( unsigned long int n ) { return n != 0 ? __builtin_ctzl( n ) : __bitsizeof(n); } |
---|
| 40 | unsigned int trailing0s( unsigned long long int n ) { return n != 0 ? __builtin_ctzll( n ) : __bitsizeof(n); } |
---|
| 41 | |
---|
| 42 | // Count all 1 bits. |
---|
| 43 | unsigned int all1s( unsigned char n ) { return __builtin_popcount( n ); } |
---|
| 44 | unsigned int all1s( unsigned short int n ) { return __builtin_popcount( n ); } |
---|
| 45 | unsigned int all1s( unsigned int n ) { return __builtin_popcount( n ); } |
---|
| 46 | unsigned int all1s( unsigned long int n ) { return __builtin_popcountl( n ); } |
---|
| 47 | unsigned int all1s( unsigned long long int n ) { return __builtin_popcountll( n ); } |
---|
| 48 | |
---|
| 49 | // Count all 0 bits. |
---|
[ee06db5c] | 50 | unsigned int all0s( unsigned char n ) { return __builtin_popcount( (typeof(n))~n ); } |
---|
| 51 | unsigned int all0s( unsigned short int n ) { return __builtin_popcount( (typeof(n))~n ); } |
---|
| 52 | unsigned int all0s( unsigned int n ) { return __builtin_popcount( ~n ); } |
---|
| 53 | unsigned int all0s( unsigned long int n ) { return __builtin_popcountl( ~n ); } |
---|
| 54 | unsigned int all0s( unsigned long long int n ) { return __builtin_popcountll( ~n ); } |
---|
[66f3bae] | 55 | |
---|
| 56 | // Find least significiant zero bit. (ffs) |
---|
| 57 | unsigned int low0( unsigned char n ) { return __builtin_ffs( (typeof(n))~n ); } |
---|
| 58 | unsigned int low0( unsigned short int n ) { return __builtin_ffs( (typeof(n))~n ); } |
---|
| 59 | unsigned int low0( unsigned int n ) { return __builtin_ffs( ~n ); } |
---|
| 60 | unsigned int low0( unsigned long int n ) { return __builtin_ffsl( ~n ); } |
---|
| 61 | unsigned int low0( unsigned long long int n ) { return __builtin_ffsll( ~n ); } |
---|
| 62 | |
---|
| 63 | // Find least significiant one bit. |
---|
| 64 | unsigned int low1( unsigned int n ) { return __builtin_ffs( n ); } |
---|
| 65 | unsigned int low1( unsigned long int n ) { return __builtin_ffsl( n ); } |
---|
| 66 | unsigned int low1( unsigned long long int n ) { return __builtin_ffsll( n ); } |
---|
| 67 | |
---|
| 68 | // Find most significiant zero bit. |
---|
| 69 | unsigned int high0( unsigned char n ) { return n == (typeof(n))-1 ? 0 : __bitsizeof(unsigned int) - __builtin_clz( (typeof(n))~n ); } |
---|
| 70 | unsigned int high0( unsigned short int n ) { return n == (typeof(n))-1 ? 0 : __bitsizeof(unsigned int) - __builtin_clz( (typeof(n))~n ); } |
---|
| 71 | unsigned int high0( unsigned int n ) { return n == -1 ? 0 : __bitsizeof(n) - __builtin_clz( ~n ); } |
---|
| 72 | unsigned int high0( unsigned long int n ) { return n == -1 ? 0 : __bitsizeof(n) - __builtin_clzl( ~n ); } |
---|
| 73 | unsigned int high0( unsigned long long int n ) { return n == -1 ? 0 : __bitsizeof(n) - __builtin_clzll( ~n ); } |
---|
| 74 | |
---|
| 75 | // Find most significiant one bit. |
---|
| 76 | unsigned int high1( unsigned char n ) { return n == 0 ? 0 : __bitsizeof(unsigned int) - __builtin_clz( n ); } |
---|
| 77 | unsigned int high1( unsigned short int n ) { return n == 0 ? 0 : __bitsizeof(unsigned int) - __builtin_clz( n ); } |
---|
| 78 | unsigned int high1( unsigned int n ) { return n == 0 ? 0 : __bitsizeof(n) - __builtin_clz( n ); } |
---|
| 79 | unsigned int high1( unsigned long int n ) { return n == 0 ? 0 : __bitsizeof(n) - __builtin_clzl( n ); } |
---|
| 80 | unsigned int high1( unsigned long long int n ) { return n == 0 ? 0 : __bitsizeof(n) - __builtin_clzll( n ); } |
---|
| 81 | |
---|
| 82 | // Check for power of 2, clears bits below n, rounding down to the next lower multiple of n. 0 is not a power of 2 |
---|
| 83 | // but this computation returns true because of the two's complement, so it is a special case. |
---|
| 84 | bool is_pow2( unsigned char n ) { return n == 0 ? false : (n & (n - 1)) == 0; } |
---|
| 85 | bool is_pow2( unsigned short int n ) { return n == 0 ? false : (n & (n - 1)) == 0; } |
---|
| 86 | bool is_pow2( unsigned int n ) { return n == 0 ? false : (n & (n - 1)) == 0; } |
---|
| 87 | bool is_pow2( unsigned long int n ) { return n == 0 ? false : (n & (n - 1)) == 0; } |
---|
| 88 | bool is_pow2( unsigned long long int n ) { return n == 0 ? false : (n & (n - 1)) == 0; } |
---|
| 89 | |
---|
| 90 | // Returns n aligned at the floor of align, clear bits above or equal to align, giving n % align. |
---|
[da3b790] | 91 | signed char floor2( signed char n, signed char align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 92 | unsigned char floor2( unsigned char n, unsigned char align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 93 | short int floor2( short int n, short int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 94 | unsigned short int floor2( unsigned short int n, unsigned short int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 95 | int floor2( int n, int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 96 | unsigned int floor2( unsigned int n, unsigned int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 97 | long int floor2( long int n, long int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 98 | unsigned long int floor2( unsigned long int n, unsigned long int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 99 | long long int floor2( long long int n, long long int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
| 100 | unsigned long long int floor2( unsigned long long int n, unsigned long long int align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
[21923bd] | 101 | |
---|
| 102 | // forall( otype T | { T ?&?( T, T ); T -?( T ); } ) |
---|
[da3b790] | 103 | // T floor2( T n, T align ) { verify( is_pow2( align ) ); return n & -align; } |
---|
[21923bd] | 104 | |
---|
[da3b790] | 105 | signed char floor( signed char n, signed char align ) { return n / align * align; } |
---|
[66f3bae] | 106 | unsigned char floor( unsigned char n, unsigned char align ) { return n / align * align; } |
---|
| 107 | short int floor( short int n, short int align ) { return n / align * align; } |
---|
| 108 | unsigned short int floor( unsigned short int n, unsigned short int align ) { return n / align * align; } |
---|
| 109 | int floor( int n, int align ) { return n / align * align; } |
---|
| 110 | unsigned int floor( unsigned int n, unsigned int align ) { return n / align * align; } |
---|
| 111 | long int floor( long int n, long int align ) { return n / align * align; } |
---|
| 112 | unsigned long int floor( unsigned long int n, unsigned long int align ) { return n / align * align; } |
---|
[21923bd] | 113 | long long int floor( long long int n, long long int align ) { return n / align * align; } |
---|
[66f3bae] | 114 | unsigned long long int floor( unsigned long long int n, unsigned long long int align ) { return n / align * align; } |
---|
[21923bd] | 115 | |
---|
| 116 | // forall( otype T | { T ?/?( T, T ); T ?*?( T, T ); } ) |
---|
| 117 | // T floor( T n, T align ) { return n / align * align; } |
---|
[da36d25] | 118 | |
---|
[66f3bae] | 119 | // Returns n aligned at the ceiling of align, negate, round down, negate is the same as round up. |
---|
[da3b790] | 120 | signed char ceiling2( signed char n, signed char align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 121 | unsigned char ceiling2( unsigned char n, unsigned char align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 122 | short int ceiling2( short int n, short int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 123 | unsigned short int ceiling2( unsigned short int n, unsigned short int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 124 | int ceiling2( int n, int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 125 | unsigned int ceiling2( unsigned int n, unsigned int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 126 | long int ceiling2( long int n, long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 127 | unsigned long int ceiling2( unsigned long int n, unsigned long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 128 | long long int ceiling2( long long int n, long long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 129 | unsigned long long int ceiling2( unsigned long long int n, unsigned long long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
[21923bd] | 130 | |
---|
| 131 | // forall( otype T | { T floor2( T, T ); T -?( T ); } ) |
---|
[da3b790] | 132 | // T ceiling2( T n, T align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); } |
---|
| 133 | |
---|
| 134 | signed char ceiling_div( signed char n, char align ) { return (n + (align - 1)) / align; } |
---|
| 135 | unsigned char ceiling_div( unsigned char n, unsigned char align ) { return (n + (align - 1)) / align; } |
---|
| 136 | short int ceiling_div( short int n, short int align ) { return (n + (align - 1)) / align; } |
---|
| 137 | unsigned short int ceiling_div( unsigned short int n, unsigned short int align ) { return (n + (align - 1)) / align; } |
---|
| 138 | int ceiling_div( int n, int align ) { return (n + (align - 1)) / align; } |
---|
| 139 | unsigned int ceiling_div( unsigned int n, unsigned int align ) { return (n + (align - 1)) / align; } |
---|
| 140 | long int ceiling_div( long int n, long int align ) { return (n + (align - 1)) / align; } |
---|
| 141 | unsigned long int ceiling_div( unsigned long int n, unsigned long int align ) { return (n + (align - 1)) / align; } |
---|
| 142 | long long int ceiling_div( long long int n, long long int align ) { return (n + (align - 1)) / align; } |
---|
| 143 | unsigned long long int ceiling_div( unsigned long long int n, unsigned long long int align ) { return (n + (align - 1)) / align; } |
---|
| 144 | |
---|
| 145 | // forall( otype T | { T ?+?( T, T ); T ?-?( T, T ); T ?%?( T, T ); } ) |
---|
| 146 | // T ceiling_div( T n, T align ) { verify( is_pow2( align ) );return (n + (align - 1)) / align; } |
---|
| 147 | |
---|
| 148 | // gcc notices the div/mod pair and saves both so only one div. |
---|
| 149 | signed char ceiling( signed char n, signed char align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 150 | unsigned char ceiling( unsigned char n, unsigned char align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 151 | short int ceiling( short int n, short int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 152 | unsigned short int ceiling( unsigned short int n, unsigned short int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 153 | int ceiling( int n, int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 154 | unsigned int ceiling( unsigned int n, unsigned int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 155 | long int ceiling( long int n, long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 156 | unsigned long int ceiling( unsigned long int n, unsigned long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0) , align); } |
---|
| 157 | long long int ceiling( long long int n, long long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
| 158 | unsigned long long int ceiling( unsigned long long int n, unsigned long long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); } |
---|
[21923bd] | 159 | |
---|
| 160 | // forall( otype T | { void ?{}( T &, one_t ); T ?+?( T, T ); T ?-?( T, T ); T ?/?( T, T ); } ) |
---|
[da3b790] | 161 | // T ceiling( T n, T align ) { return return floor( n + (n % align != 0 ? align - 1 : 0), align ); *} |
---|
[76bf23c] | 162 | } // distribution |
---|
[03eabf4] | 163 | |
---|
| 164 | // Local Variables: // |
---|
| 165 | // tab-width: 4 // |
---|
| 166 | // End: // |
---|