Mercurial > vec

comparison README @ 15:e05c257c6a23

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
*: huge refactor, add many new x86 intrinsics and the like ALSO!! now intrinsics are enabled at runtime, depending on what is detected. altivec *should* still work but I only tested compiling it. the major version has been updated to 2.0 for this...
author Paper <paper@tflc.us>
date Wed, 20 Nov 2024 04:10:37 -0500
parents f12b5dd4e18c
children e26874655738
comparison
equal deleted inserted replaced
14:981cf0bc7f3a 15:e05c257c6a23
1 vec - a tiny SIMD vector header-only library written in C99 1 vec - a tiny SIMD vector header-only library written in C99
2 2
3 it comes with an extremely basic (and somewhat lacking) API, 3 it comes with an extremely basic API that is similar to other intrinsics
4 where there are eight supported vector types, all 128-bit: 4 libraries; each type is in the exact same format:
5 5
6 vint8x16 - 16 signed 8-bit integers 6 v[sign][bits]x[size]
7 vint16x8 - 8 signed 16-bit integers 7 where `sign' is either nothing (for signed) or `u' (for unsigned),
8 vint32x4 - 4 signed 32-bit integers 8 `bits' is the bit size of the integer format,
9 vint64x2 - 2 signed 64-bit integers 9 and `size' is the how many integers are in the vector
10 vuint8x16 - 16 unsigned 8-bit integers
11 vuint16x8 - 8 unsigned 16-bit integers
12 vuint32x4 - 4 unsigned 32-bit integers
13 vuint32x4 - 2 unsigned 64-bit integers
14 10
15 all of these have many operations that are prefixed with the 11 vec provides types for 64-bit, 128-bit, 256-bit, and 512-bit SIMD intrinsics
16 name of the type and an underscore, for example: 12 on processors where vec has an implementation and falls back to array-based
13 implementations where they are not.
14
15 all of these have many operations that are prefixed with the name of the
16 type and an underscore, for example:
17 17
18 vint8x16 vint8x16_splat(uint8_t x) 18 vint8x16 vint8x16_splat(uint8_t x)
19 - creates a vint8x16 where all of the values are filled 19 - creates a vint8x16 where all of the values are filled
20 with the value of `x' 20 with the value of `x'
21 21
104 v[u]intAxB cmpge(v[u]intAxB vec1, v[u]intAxB vec2) 104 v[u]intAxB cmpge(v[u]intAxB vec1, v[u]intAxB vec2)
105 turns on all bits of the corresponding value in 105 turns on all bits of the corresponding value in
106 the result vector if the value in `vec1' is greater 106 the result vector if the value in `vec1' is greater
107 than or equal to the corresponding value in `vec2', 107 than or equal to the corresponding value in `vec2',
108 else all of the bits are turned off. 108 else all of the bits are turned off.
109
110 to initialize vec, you MUST call `vec_init()' when your programs starts up.
111
112 note that `vec_init()' is NOT thread-safe, and things can and will
113 blow up if you call it simultaneously from different threads (i.e. you
114 try to only initialize it when you need to... please just initialize
115 it on startup so you don't have to worry about that!!!)