Main_Page.mwiki

{{title|Libsimdpp reference}}
{| class="mainpagetable" cellspacing=0 style="width:100%; white-space:nowrap;"
|- class="row rowtop"
| colspan=4 | <span style="font-size: 2em;">Libsimdpp documentation</span>
|- class="row"
| colspan=4  |
<div style="white-space:normal;">
libsimdpp is a portable header-only zero-overhead C++ wrapper around
single-instruction multiple-data (SIMD) intrinsics found in many compilers. The
library presents a single interface over several instruction sets in such a way
that the same source code may be compiled for different instruction sets. The
resulting object files then may be hooked into internal dynamic dispatch
mechanism.


The library resolves differences between instruction sets by implementing the
missing functionality as a combination of several intrinsics. Moreover, the
library supplies a lot of additional, commonly used functionality, such as
various variants of matrix transpositions, interleaving loads/stores, optimized
compile-time shuffling instructions, etc. Each of these are implemented in the
most efficient manner for the target instruction set. Finally, it's possible
to fall back to native intrinsics when necessary, without compromising
maintainability.


The library sits somewhere in the middle between programming directly in
intrinsics and even higher-level SIMD libraries. As much control as possible
is given to the developer, so that it's possible to exactly predict what code
the compiler will generate.
</div>
|- class="row"
|

'''Instruction sets'''
<div class="mainpagediv">
[[arch/selection|Instruction set selection]]<br/>
[[arch/dispatch|Dynamic dispatch]]<br/>
</div>

'''[[depr|Deprecations]]'''

'''[[types|Types]]'''
<div class="mainpagediv">
[[types/overview|Overview]]<br/>
[[types/promotion|Vector type promotion]]<br/>
</div>

'''[[cap|Capabilities]]'''

'''[[util | Utilities]]'''
<div class="mainpagediv">
{{ltt|util#zzz|SIMDPP_ALIGN}}<br/>
{{ltt|util/aligned_allocator}}<br/>
[[util#Vector_size_macros|Vector size macros]]<br/>
[[util#Enabled_instruction_set | Enabled instruction set]]
</div>

'''[[conv | Conversions]]'''
<div class="mainpagediv">
{{ltt|conv/to_int8}}&nbsp;&nbsp;{{ltt|conv/to_uint8}}<br/>
{{ltt|conv/to_int16}}&nbsp;&nbsp;{{ltt|conv/to_uint16}}<br/>
{{ltt|conv/to_int32}}&nbsp;&nbsp;{{ltt|conv/to_uint32}}<br/>
{{ltt|conv/to_int64}}&nbsp;&nbsp;{{ltt|conv/to_uint64}}<br/>
{{ltt|conv/to_float32}}&nbsp;&nbsp;{{ltt|conv/to_float64}}<br/>
{{ltt|conv/to_mask}}
</div>

|

'''[[bitwise | Bitwise operations]]'''
<div class="mainpagediv">
{{ltt|bitwise/bit_and}}&nbsp;&nbsp;{{ltt|bitwise/bit_andnot}} <br/>
{{ltt|bitwise/bit_or}}&nbsp;&nbsp;{{ltt|bitwise/bit_xor}} <br/>
{{ltt|bitwise/bit_not}} <br/>
</div>

'''[[fp | Floating-point operations]]'''
<div class="mainpagediv">
{{ltt|fp/add}}&nbsp;&nbsp;{{ltt|fp/sub}} <br/>
{{ltt|fp/mul}}&nbsp;&nbsp;{{ltt|fp/div}} <br/>
{{ltt|fp/fmadd}}&nbsp;&nbsp;{{ltt|fp/fmsub}} <br/>
{{ltt|fp/neg}} <br/>
{{ltt|fp/round}}&nbsp;&nbsp;{{ltt|fp/trunc}} <br/>
{{ltt|fp/floor}}&nbsp;&nbsp;{{ltt|fp/ceil}} <br/>
{{ltt|fp/cmp_eq}}&nbsp;&nbsp;{{ltt|fp/cmp_neq}} <br/>
{{ltt|fp/cmp_lt}}&nbsp;&nbsp;{{ltt|fp/cmp_gt}} <br/>
{{ltt|fp/cmp_le}}&nbsp;&nbsp;{{ltt|fp/cmp_ge}} <br/>
{{ltt|fp/abs}}&nbsp;&nbsp;{{ltt|fp/sign}} <br/>
{{ltt|fp/min}}&nbsp;&nbsp;{{ltt|fp/max}} <br/>
{{ltt|fp/isnan}}&nbsp;&nbsp;{{ltt|fp/isnan2}} <br/>
{{ltt|fp/sqrt}} <br/>
{{ltt|fp/rcp_e}}&nbsp;&nbsp;{{ltt|fp/rcp_rh}} <br/>
{{ltt|fp/rsqrt_e}}&nbsp;&nbsp;{{ltt|fp/rsqrt_rh}} <br/>
{{ltt|fp/reduce_add}}&nbsp;&nbsp;{{ltt|fp/reduce_mul}} <br/>
{{ltt|fp/reduce_min}}&nbsp;&nbsp;{{ltt|fp/reduce_max}} <br/>
</div>

'''[[operator | Operators]]'''
<div class="mainpagediv">
{{tt|+}}&nbsp;&nbsp;{{tt|-}}&nbsp;&nbsp;
{{tt|*}}&nbsp;&nbsp;{{tt|/}}&nbsp;&nbsp;
{{tt|&lt;&lt;}}&nbsp;&nbsp;{{tt|&gt;&gt;}}<br/>
{{tt|&}}&nbsp;&nbsp;{{tt|{{!}}}}&nbsp;&nbsp;
{{tt|^}}&nbsp;&nbsp;{{tt|~}}<br/>
{{tt|{{==}}}}&nbsp;&nbsp;{{tt|!{{=}}}}&nbsp;&nbsp;
{{tt|>}}&nbsp;&nbsp;{{tt|>{{=}}}}&nbsp;&nbsp;
{{tt|<}}&nbsp;&nbsp;{{tt|<{{=}}}}<br/>
</div>
|

'''[[int | Integer operations]]'''

<div class="mainpagediv">
{{ltt|int/add}}&nbsp;&nbsp;{{ltt|int/sub}} <br/>
{{ltt|int/add_sat}}&nbsp;&nbsp;{{ltt|int/sub_sat}} <br/>
{{ltt|int/mul_lo}}&nbsp;&nbsp;{{ltt|int/mul_hi}} <br/>
{{ltt|int/mull}}&nbsp;&nbsp;{{ltt|int/neg}} <br/>
{{ltt|int/cmp_eq}}&nbsp;&nbsp;{{ltt|int/cmp_neq}} <br/>
{{ltt|int/cmp_lt}}&nbsp;&nbsp;{{ltt|int/cmp_gt}} <br/>
{{ltt|int/cmp_le}}&nbsp;&nbsp;{{ltt|int/cmp_ge}} <br/>
{{ltt|int/abs}} <br/>
{{ltt|int/avg}}&nbsp;&nbsp;{{ltt|int/avg_trunc}} <br/>
{{ltt|int/min}}&nbsp;&nbsp;{{ltt|int/max}} <br/>
{{ltt|int/shift_l}}&nbsp;&nbsp;{{ltt|int/shift_r}} <br/>
{{ltt|int/reduce_add}}&nbsp;&nbsp;{{ltt|int/reduce_mul}} <br/>
{{ltt|int/reduce_min}}&nbsp;&nbsp;{{ltt|int/reduce_max}} <br/>
{{ltt|int/reduce_and}}&nbsp;&nbsp;{{ltt|int/reduce_or}} <br/>
</div>

'''[[mem | Memory access]]'''
<div class="mainpagediv">
{{ltt|mem/load}}&nbsp;&nbsp;{{ltt|mem/load_u}} <br/>
{{ltt|mem/load_splat}} <br/>
{{ltt|mem/load_packed2}}<br/>
{{ltt|mem/load_packed3}} <br/>
{{ltt|mem/load_packed4}} <br/>
{{ltt|mem/store}}&nbsp;&nbsp;{{ltt|mem/stream}} <br/>
{{ltt|mem/store_first}}&nbsp;&nbsp;{{ltt|mem/store_last}} <br/>
{{ltt|mem/store_packed2}}<br/>
{{ltt|mem/store_packed3}} <br/>
{{ltt|mem/store_packed4}} <br/>
</div>

|

'''[[shuffle | Shuffle operations]]'''
<div class="mainpagediv">
{{ltt|shuffle/blend}}&nbsp;&nbsp;{{ltt|shuffle/splat}} <br/>
{{ltt|shuffle/combine}}&nbsp;&nbsp;{{ltt|shuffle/split}} <br/>
{{ltt|shuffle/permute2}}&nbsp;&nbsp;{{ltt|shuffle/permute4}} <br/>
{{ltt|shuffle/shuffle2x2}}&nbsp;&nbsp;{{ltt|shuffle/shuffle4x2}} <br/>
{{ltt|shuffle/transpose2}}&nbsp;&nbsp;{{ltt|shuffle/transpose4}} <br/><br/>
&nbsp;&nbsp;Shuffles operating on 128-bit blocks <br/>
{{ltt | shuffle/alignX_128 | align16&nbsp;align8&nbsp;align4&nbsp;align2 }} <br/>
{{ltt | shuffle/splatX_128 | splat16&nbsp;splat8&nbsp;splat4&nbsp;splat2 }} <br/>
{{ltt | shuffle/moveX_l_128 | move16_l&nbsp;move8_l&nbsp;move4_l&nbsp;move2_l }} <br/>
{{ltt | shuffle/moveX_r_128 | move16_r&nbsp;move8_r&nbsp;move4_r&nbsp;move2_r }} <br/>
{{ltt | shuffle/zipX_hi_128 | zip16_hi&nbsp;zip8_hi&nbsp;zip4_hi&nbsp;zip2_hi }} <br/>
{{ltt | shuffle/zipX_lo_128 | zip16_lo&nbsp;zip8_lo&nbsp;zip4_lo&nbsp;zip2_lo }} <br/>
{{ltt | shuffle/unzipX_hi_128 | unzip16_hi&nbsp;unzip8_hi<br/>unzip4_hi&nbsp;unzip2_hi }} <br/>
{{ltt | shuffle/unzipX_lo_128 | unzip16_lo&nbsp;unzip8_lo<br/>unzip4_lo&nbsp;unzip2_lo }} <br/>
{{ltt | shuffle/make_shuffle_bytes16_mask}}<br/>
{{ltt | shuffle/shuffle_bytes16}}&nbsp;&nbsp;{{ltt | shuffle/shuffle_zbytes16}}<br/>
{{ltt | shuffle/permute_bytes16}}&nbsp;&nbsp;{{ltt | shuffle/permute_zbytes16}}<br/>
</div>

'''[[misc | Miscellaneous operations]]'''
<div class="mainpagediv">
{{ltt|misc/make_int}}&nbsp;&nbsp;{{ltt|misc/make_uint}}&nbsp;&nbsp;{{ltt|misc/make_float}} <br/>
{{ltt|misc/extract}}&nbsp;&nbsp;{{ltt|misc/insert}}&nbsp;&nbsp;{{ltt|misc/splat|splat{{small|(scalar)}}}} <br/>
{{ltt|misc/for_each}}&nbsp;&nbsp;{{ltt|misc/test_bits_any}}<br/>
{{ltt|misc/bit_cast}}<br/>
{{ltt|misc/prefetch_read}}&nbsp;&nbsp;{{ltt|misc/prefetch_write}}
</div>

'''[[Algorithm | STL like algorithm]]'''
<div class="mainpagediv">
{{ltt|algorithm/any_of}}<br/>
{{ltt|algorithm/all_of}}<br/>
{{ltt|algorithm/copy}}<br/>
{{ltt|algorithm/copy_n}}<br/>
{{ltt|algorithm/count}}<br/>
{{ltt|algorithm/count_if}}<br/>
{{ltt|algorithm/equal}}<br/>
{{ltt|algorithm/fill}}<br/>
{{ltt|algorithm/find}}<br/>
{{ltt|algorithm/find_if}}<br/>
{{ltt|algorithm/find_if_not}}<br/>
{{ltt|algorithm/lexicographical_compare}}<br/>
{{ltt|algorithm/max}}<br/>
{{ltt|algorithm/max_element}}<br/>
{{ltt|algorithm/min}}<br/>
{{ltt|algorithm/min_element}}<br/>
{{ltt|algorithm/none_of}}<br/>
{{ltt|algorithm/reduce}}<br/>
{{ltt|algorithm/replace}}<br/>
{{ltt|algorithm/replace_if}}<br/>
{{ltt|algorithm/transform}}<br/>
{{ltt|algorithm/transform_reduce}}<br/>
</div>

|- class="row rowbottom"
| colspan=4|
|}