skia2/include/private/SkVx.h
Mike Klein 952f8f17e4 Reland "update skvx scalar-fallback strategy"
This is a reland of 4985db413d

...with a better implementation of map().  I don't understand
why we had to revert, but it had something with calling the
function pointer in map_(), so maybe this will help.

I've flattened the map_() / map() merge CL into this one,
and marked the resulting map() as no_sanitize("cfi").  I
don't see anything wrong, so I think it's a false positive.

Original change's description:
> update skvx scalar-fallback strategy
>
> Turns out Clang's a lot better at auto-vectorizing "obvious" scalar code
> into obvious vector code when it's written out the long way, e.g.
>
>      F32x4 x = ...;
>      x = { sqrtf(x[0]), sqrtf(x[1]), sqrtf(x[2]), sqrtf(x[3]) };
>
> vectorizes into sqrtps a lot more reliably than our recurse-onto-scalars
> strategy, and also better than the other naive approach,
>
>      F32x4 x = ...;
>      for (int i = 0; i < 4; i++) { x[i] = sqrtf(x[i]); }
>
> So here I've added a map(V, fn) -> V' using C++14 tricks to let the
> compiler handle the expansion of x = { fn(x[0]), fn(x[1]), ...
> fn(x[N-1]) } for any N, and implemented most skvx scalar fallback code
> using that.
>
> With these now vectorizing well at any N, we can remove any
> specializations we'd written for particular N, really tidying up.
>
> Over in the SkVM interpreter, this is a big improvement for ceil and
> floor, which were being done 2 floats at a time instead of 8.  They're
> now slimmed way down to
>
>    shlq       $6, %r13
>    vroundps   $K, (%r12,%r13), %ymm0
>    vroundps   $K, 32(%r12,%r13), %ymm1
>    jmp        ...
>
> where K is 9 or 10 depending on the op.
>
> I haven't found a scalar function that Clang will vectorize to vcvtps2pd
> (the rounding one, not truncating vcvttps2pd), so I've kept lrint()
> written the long way, updated to the style I've been using lately with
> specializations inline.
>
> Change-Id: Ia97abe3c876008228bf62b1daacd6f6140408fc4
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/317375
> Reviewed-by: Herb Derby <herb@google.com>
> Commit-Queue: Mike Klein <mtklein@google.com>

Cq-Include-Trybots: luci.chromium.try:linux_chromium_cfi_rel_ng
Bug: chromium:1129408
Change-Id: Ia9c14074b9a14a67dd221f4925894d35a551f9d7
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/317551
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
2020-09-17 19:58:34 +00:00

717 lines
30 KiB
C++

/*
* Copyright 2019 Google Inc.
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#ifndef SKVX_DEFINED
#define SKVX_DEFINED
// skvx::Vec<N,T> are SIMD vectors of N T's, a v1.5 successor to SkNx<N,T>.
//
// This time we're leaning a bit less on platform-specific intrinsics and a bit
// more on Clang/GCC vector extensions, but still keeping the option open to
// drop in platform-specific intrinsics, actually more easily than before.
//
// We've also fixed a few of the caveats that used to make SkNx awkward to work
// with across translation units. skvx::Vec<N,T> always has N*sizeof(T) size
// and alignment and is safe to use across translation units freely.
// Ideally we'd only align to T, but that tanks ARMv7 NEON codegen.
// Please try to keep this file independent of Skia headers.
#include <algorithm> // std::min, std::max
#include <cmath> // std::ceil, std::floor, std::trunc, std::round, std::sqrt, etc.
#include <cstdint> // intXX_t
#include <cstring> // memcpy()
#include <initializer_list> // std::initializer_list
#if defined(__SSE__) || defined(__AVX__) || defined(__AVX2__)
#include <immintrin.h>
#elif defined(__ARM_NEON)
#include <arm_neon.h>
#elif defined(__wasm_simd128__)
#include <wasm_simd128.h>
#endif
// To avoid ODR violations, all methods must be force-inlined...
#if defined(_MSC_VER)
#define SKVX_ALWAYS_INLINE __forceinline
#else
#define SKVX_ALWAYS_INLINE __attribute__((always_inline))
#endif
// ... and all standalone functions must be static. Please use these helpers:
#define SI static inline
#define SIT template < typename T> SI
#define SIN template <int N > SI
#define SINT template <int N, typename T> SI
#define SINTU template <int N, typename T, typename U, \
typename=std::enable_if_t<std::is_convertible<U,T>::value>> SI
namespace skvx {
// All Vec have the same simple memory layout, the same as `T vec[N]`.
template <int N, typename T>
struct alignas(N*sizeof(T)) Vec {
static_assert((N & (N-1)) == 0, "N must be a power of 2.");
static_assert(sizeof(T) >= alignof(T), "What kind of crazy T is this?");
Vec<N/2,T> lo, hi;
// Methods belong here in the class declaration of Vec only if:
// - they must be here, like constructors or operator[];
// - they'll definitely never want a specialized implementation.
// Other operations on Vec should be defined outside the type.
SKVX_ALWAYS_INLINE Vec() = default;
template <typename U, typename=std::enable_if_t<std::is_convertible<U,T>::value>>
SKVX_ALWAYS_INLINE
Vec(U x) : lo(x), hi(x) {}
SKVX_ALWAYS_INLINE Vec(std::initializer_list<T> xs) {
T vals[N] = {0};
memcpy(vals, xs.begin(), std::min(xs.size(), (size_t)N)*sizeof(T));
lo = Vec<N/2,T>::Load(vals + 0);
hi = Vec<N/2,T>::Load(vals + N/2);
}
SKVX_ALWAYS_INLINE T operator[](int i) const { return i < N/2 ? lo[i] : hi[i-N/2]; }
SKVX_ALWAYS_INLINE T& operator[](int i) { return i < N/2 ? lo[i] : hi[i-N/2]; }
SKVX_ALWAYS_INLINE static Vec Load(const void* ptr) {
Vec v;
memcpy(&v, ptr, sizeof(Vec));
return v;
}
SKVX_ALWAYS_INLINE void store(void* ptr) const {
memcpy(ptr, this, sizeof(Vec));
}
};
template <typename T>
struct Vec<1,T> {
T val;
SKVX_ALWAYS_INLINE Vec() = default;
template <typename U, typename=std::enable_if_t<std::is_convertible<U,T>::value>>
SKVX_ALWAYS_INLINE
Vec(U x) : val(x) {}
SKVX_ALWAYS_INLINE Vec(std::initializer_list<T> xs) : val(xs.size() ? *xs.begin() : 0) {}
SKVX_ALWAYS_INLINE T operator[](int) const { return val; }
SKVX_ALWAYS_INLINE T& operator[](int) { return val; }
SKVX_ALWAYS_INLINE static Vec Load(const void* ptr) {
Vec v;
memcpy(&v, ptr, sizeof(Vec));
return v;
}
SKVX_ALWAYS_INLINE void store(void* ptr) const {
memcpy(ptr, this, sizeof(Vec));
}
};
template <typename D, typename S>
SI D unchecked_bit_pun(const S& s) {
D d;
memcpy(&d, &s, sizeof(D));
return d;
}
template <typename D, typename S>
SI D bit_pun(const S& s) {
static_assert(sizeof(D) == sizeof(S), "");
return unchecked_bit_pun<D>(s);
}
// Translate from a value type T to its corresponding Mask, the result of a comparison.
template <typename T> struct Mask { using type = T; };
template <> struct Mask<float > { using type = int32_t; };
template <> struct Mask<double> { using type = int64_t; };
template <typename T> using M = typename Mask<T>::type;
// Join two Vec<N,T> into one Vec<2N,T>.
SINT Vec<2*N,T> join(const Vec<N,T>& lo, const Vec<N,T>& hi) {
Vec<2*N,T> v;
v.lo = lo;
v.hi = hi;
return v;
}
// We have two default strategies for implementing most operations:
// 1) lean on Clang/GCC vector extensions when available;
// 2) recurse to scalar portable implementations when not.
// At the end we can drop in platform-specific implementations that override either default.
#if !defined(SKNX_NO_SIMD) && (defined(__clang__) || defined(__GNUC__))
// VExt<N,T> types have the same size as Vec<N,T> and support most operations directly.
// N.B. VExt<N,T> alignment is N*alignof(T), stricter than Vec<N,T>'s alignof(T).
#if defined(__clang__)
template <int N, typename T>
using VExt = T __attribute__((ext_vector_type(N)));
#elif defined(__GNUC__)
template <int N, typename T>
struct VExtHelper {
typedef T __attribute__((vector_size(N*sizeof(T)))) type;
};
template <int N, typename T>
using VExt = typename VExtHelper<N,T>::type;
// For some reason some (new!) versions of GCC cannot seem to deduce N in the generic
// to_vec<N,T>() below for N=4 and T=float. This workaround seems to help...
SI Vec<4,float> to_vec(VExt<4,float> v) { return bit_pun<Vec<4,float>>(v); }
#endif
SINT VExt<N,T> to_vext(const Vec<N,T>& v) { return bit_pun<VExt<N,T>>(v); }
SINT Vec <N,T> to_vec(const VExt<N,T>& v) { return bit_pun<Vec <N,T>>(v); }
SINT Vec<N,T> operator+(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) + to_vext(y));
}
SINT Vec<N,T> operator-(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) - to_vext(y));
}
SINT Vec<N,T> operator*(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) * to_vext(y));
}
SINT Vec<N,T> operator/(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) / to_vext(y));
}
SINT Vec<N,T> operator^(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) ^ to_vext(y));
}
SINT Vec<N,T> operator&(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) & to_vext(y));
}
SINT Vec<N,T> operator|(const Vec<N,T>& x, const Vec<N,T>& y) {
return to_vec<N,T>(to_vext(x) | to_vext(y));
}
SINT Vec<N,T> operator!(const Vec<N,T>& x) { return to_vec<N,T>(!to_vext(x)); }
SINT Vec<N,T> operator-(const Vec<N,T>& x) { return to_vec<N,T>(-to_vext(x)); }
SINT Vec<N,T> operator~(const Vec<N,T>& x) { return to_vec<N,T>(~to_vext(x)); }
SINT Vec<N,T> operator<<(const Vec<N,T>& x, int k) { return to_vec<N,T>(to_vext(x) << k); }
SINT Vec<N,T> operator>>(const Vec<N,T>& x, int k) { return to_vec<N,T>(to_vext(x) >> k); }
SINT Vec<N,M<T>> operator==(const Vec<N,T>& x, const Vec<N,T>& y) {
return bit_pun<Vec<N,M<T>>>(to_vext(x) == to_vext(y));
}
SINT Vec<N,M<T>> operator!=(const Vec<N,T>& x, const Vec<N,T>& y) {
return bit_pun<Vec<N,M<T>>>(to_vext(x) != to_vext(y));
}
SINT Vec<N,M<T>> operator<=(const Vec<N,T>& x, const Vec<N,T>& y) {
return bit_pun<Vec<N,M<T>>>(to_vext(x) <= to_vext(y));
}
SINT Vec<N,M<T>> operator>=(const Vec<N,T>& x, const Vec<N,T>& y) {
return bit_pun<Vec<N,M<T>>>(to_vext(x) >= to_vext(y));
}
SINT Vec<N,M<T>> operator< (const Vec<N,T>& x, const Vec<N,T>& y) {
return bit_pun<Vec<N,M<T>>>(to_vext(x) < to_vext(y));
}
SINT Vec<N,M<T>> operator> (const Vec<N,T>& x, const Vec<N,T>& y) {
return bit_pun<Vec<N,M<T>>>(to_vext(x) > to_vext(y));
}
#else
// Either SKNX_NO_SIMD is defined, or Clang/GCC vector extensions are not available.
// We'll implement things portably, in a way that should be easily autovectorizable.
// N == 1 scalar implementations.
SIT Vec<1,T> operator+(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val + y.val; }
SIT Vec<1,T> operator-(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val - y.val; }
SIT Vec<1,T> operator*(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val * y.val; }
SIT Vec<1,T> operator/(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val / y.val; }
SIT Vec<1,T> operator^(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val ^ y.val; }
SIT Vec<1,T> operator&(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val & y.val; }
SIT Vec<1,T> operator|(const Vec<1,T>& x, const Vec<1,T>& y) { return x.val | y.val; }
SIT Vec<1,T> operator!(const Vec<1,T>& x) { return !x.val; }
SIT Vec<1,T> operator-(const Vec<1,T>& x) { return -x.val; }
SIT Vec<1,T> operator~(const Vec<1,T>& x) { return ~x.val; }
SIT Vec<1,T> operator<<(const Vec<1,T>& x, int k) { return x.val << k; }
SIT Vec<1,T> operator>>(const Vec<1,T>& x, int k) { return x.val >> k; }
SIT Vec<1,M<T>> operator==(const Vec<1,T>& x, const Vec<1,T>& y) {
return x.val == y.val ? ~0 : 0;
}
SIT Vec<1,M<T>> operator!=(const Vec<1,T>& x, const Vec<1,T>& y) {
return x.val != y.val ? ~0 : 0;
}
SIT Vec<1,M<T>> operator<=(const Vec<1,T>& x, const Vec<1,T>& y) {
return x.val <= y.val ? ~0 : 0;
}
SIT Vec<1,M<T>> operator>=(const Vec<1,T>& x, const Vec<1,T>& y) {
return x.val >= y.val ? ~0 : 0;
}
SIT Vec<1,M<T>> operator< (const Vec<1,T>& x, const Vec<1,T>& y) {
return x.val < y.val ? ~0 : 0;
}
SIT Vec<1,M<T>> operator> (const Vec<1,T>& x, const Vec<1,T>& y) {
return x.val > y.val ? ~0 : 0;
}
// All default N != 1 implementations just recurse on lo and hi halves.
SINT Vec<N,T> operator+(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo + y.lo, x.hi + y.hi);
}
SINT Vec<N,T> operator-(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo - y.lo, x.hi - y.hi);
}
SINT Vec<N,T> operator*(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo * y.lo, x.hi * y.hi);
}
SINT Vec<N,T> operator/(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo / y.lo, x.hi / y.hi);
}
SINT Vec<N,T> operator^(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo ^ y.lo, x.hi ^ y.hi);
}
SINT Vec<N,T> operator&(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo & y.lo, x.hi & y.hi);
}
SINT Vec<N,T> operator|(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo | y.lo, x.hi | y.hi);
}
SINT Vec<N,T> operator!(const Vec<N,T>& x) { return join(!x.lo, !x.hi); }
SINT Vec<N,T> operator-(const Vec<N,T>& x) { return join(-x.lo, -x.hi); }
SINT Vec<N,T> operator~(const Vec<N,T>& x) { return join(~x.lo, ~x.hi); }
SINT Vec<N,T> operator<<(const Vec<N,T>& x, int k) { return join(x.lo << k, x.hi << k); }
SINT Vec<N,T> operator>>(const Vec<N,T>& x, int k) { return join(x.lo >> k, x.hi >> k); }
SINT Vec<N,M<T>> operator==(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo == y.lo, x.hi == y.hi);
}
SINT Vec<N,M<T>> operator!=(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo != y.lo, x.hi != y.hi);
}
SINT Vec<N,M<T>> operator<=(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo <= y.lo, x.hi <= y.hi);
}
SINT Vec<N,M<T>> operator>=(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo >= y.lo, x.hi >= y.hi);
}
SINT Vec<N,M<T>> operator< (const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo < y.lo, x.hi < y.hi);
}
SINT Vec<N,M<T>> operator> (const Vec<N,T>& x, const Vec<N,T>& y) {
return join(x.lo > y.lo, x.hi > y.hi);
}
#endif
// Some operations we want are not expressible with Clang/GCC vector
// extensions, so we implement them using the recursive approach.
// N == 1 scalar implementations.
SIT Vec<1,T> if_then_else(const Vec<1,M<T>>& cond, const Vec<1,T>& t, const Vec<1,T>& e) {
// In practice this scalar implementation is unlikely to be used. See if_then_else() below.
return bit_pun<Vec<1,T>>(( cond & bit_pun<Vec<1, M<T>>>(t)) |
(~cond & bit_pun<Vec<1, M<T>>>(e)) );
}
SIT bool any(const Vec<1,T>& x) { return x.val != 0; }
SIT bool all(const Vec<1,T>& x) { return x.val != 0; }
SIT Vec<1,T> pow(const Vec<1,T>& x, const Vec<1,T>& y) { return std::pow(x.val, y.val); }
// All default N != 1 implementations just recurse on lo and hi halves.
// Clang can reason about naive_if_then_else() and optimize through it better
// than if_then_else(), so it's sometimes useful to call it directly when we
// think an entire expression should optimize away, e.g. min()/max().
SINT Vec<N,T> naive_if_then_else(const Vec<N,M<T>>& cond, const Vec<N,T>& t, const Vec<N,T>& e) {
return bit_pun<Vec<N,T>>(( cond & bit_pun<Vec<N, M<T>>>(t)) |
(~cond & bit_pun<Vec<N, M<T>>>(e)) );
}
SINT Vec<N,T> if_then_else(const Vec<N,M<T>>& cond, const Vec<N,T>& t, const Vec<N,T>& e) {
// Specializations inline here so they can generalize what types the apply to.
// (This header is used in C++14 contexts, so we have to kind of fake constexpr if.)
#if defined(__AVX2__)
if /*constexpr*/ (N*sizeof(T) == 32) {
return unchecked_bit_pun<Vec<N,T>>(_mm256_blendv_epi8(unchecked_bit_pun<__m256i>(e),
unchecked_bit_pun<__m256i>(t),
unchecked_bit_pun<__m256i>(cond)));
}
#endif
#if defined(__SSE4_1__)
if /*constexpr*/ (N*sizeof(T) == 16) {
return unchecked_bit_pun<Vec<N,T>>(_mm_blendv_epi8(unchecked_bit_pun<__m128i>(e),
unchecked_bit_pun<__m128i>(t),
unchecked_bit_pun<__m128i>(cond)));
}
#endif
#if defined(__ARM_NEON)
if /*constexpr*/ (N*sizeof(T) == 16) {
return unchecked_bit_pun<Vec<N,T>>(vbslq_u8(unchecked_bit_pun<uint8x16_t>(cond),
unchecked_bit_pun<uint8x16_t>(t),
unchecked_bit_pun<uint8x16_t>(e)));
}
#endif
// Recurse for large vectors to try to hit the specializations above.
if /*constexpr*/ (N*sizeof(T) > 16) {
return join(if_then_else(cond.lo, t.lo, e.lo),
if_then_else(cond.hi, t.hi, e.hi));
}
// This default can lead to better code than the recursing onto scalars.
return naive_if_then_else(cond, t, e);
}
SINT bool any(const Vec<N,T>& x) { return any(x.lo) || any(x.hi); }
SINT bool all(const Vec<N,T>& x) { return all(x.lo) && all(x.hi); }
SINT Vec<N,T> pow(const Vec<N,T>& x, const Vec<N,T>& y) {
return join(pow(x.lo, y.lo), pow(x.hi, y.hi));
}
// Scalar/vector operations just splat the scalar to a vector...
SINTU Vec<N,T> operator+ (U x, const Vec<N,T>& y) { return Vec<N,T>(x) + y; }
SINTU Vec<N,T> operator- (U x, const Vec<N,T>& y) { return Vec<N,T>(x) - y; }
SINTU Vec<N,T> operator* (U x, const Vec<N,T>& y) { return Vec<N,T>(x) * y; }
SINTU Vec<N,T> operator/ (U x, const Vec<N,T>& y) { return Vec<N,T>(x) / y; }
SINTU Vec<N,T> operator^ (U x, const Vec<N,T>& y) { return Vec<N,T>(x) ^ y; }
SINTU Vec<N,T> operator& (U x, const Vec<N,T>& y) { return Vec<N,T>(x) & y; }
SINTU Vec<N,T> operator| (U x, const Vec<N,T>& y) { return Vec<N,T>(x) | y; }
SINTU Vec<N,M<T>> operator==(U x, const Vec<N,T>& y) { return Vec<N,T>(x) == y; }
SINTU Vec<N,M<T>> operator!=(U x, const Vec<N,T>& y) { return Vec<N,T>(x) != y; }
SINTU Vec<N,M<T>> operator<=(U x, const Vec<N,T>& y) { return Vec<N,T>(x) <= y; }
SINTU Vec<N,M<T>> operator>=(U x, const Vec<N,T>& y) { return Vec<N,T>(x) >= y; }
SINTU Vec<N,M<T>> operator< (U x, const Vec<N,T>& y) { return Vec<N,T>(x) < y; }
SINTU Vec<N,M<T>> operator> (U x, const Vec<N,T>& y) { return Vec<N,T>(x) > y; }
SINTU Vec<N,T> pow(U x, const Vec<N,T>& y) { return pow(Vec<N,T>(x), y); }
// ... and same deal for vector/scalar operations.
SINTU Vec<N,T> operator+ (const Vec<N,T>& x, U y) { return x + Vec<N,T>(y); }
SINTU Vec<N,T> operator- (const Vec<N,T>& x, U y) { return x - Vec<N,T>(y); }
SINTU Vec<N,T> operator* (const Vec<N,T>& x, U y) { return x * Vec<N,T>(y); }
SINTU Vec<N,T> operator/ (const Vec<N,T>& x, U y) { return x / Vec<N,T>(y); }
SINTU Vec<N,T> operator^ (const Vec<N,T>& x, U y) { return x ^ Vec<N,T>(y); }
SINTU Vec<N,T> operator& (const Vec<N,T>& x, U y) { return x & Vec<N,T>(y); }
SINTU Vec<N,T> operator| (const Vec<N,T>& x, U y) { return x | Vec<N,T>(y); }
SINTU Vec<N,M<T>> operator==(const Vec<N,T>& x, U y) { return x == Vec<N,T>(y); }
SINTU Vec<N,M<T>> operator!=(const Vec<N,T>& x, U y) { return x != Vec<N,T>(y); }
SINTU Vec<N,M<T>> operator<=(const Vec<N,T>& x, U y) { return x <= Vec<N,T>(y); }
SINTU Vec<N,M<T>> operator>=(const Vec<N,T>& x, U y) { return x >= Vec<N,T>(y); }
SINTU Vec<N,M<T>> operator< (const Vec<N,T>& x, U y) { return x < Vec<N,T>(y); }
SINTU Vec<N,M<T>> operator> (const Vec<N,T>& x, U y) { return x > Vec<N,T>(y); }
SINTU Vec<N,T> pow(const Vec<N,T>& x, U y) { return pow(x, Vec<N,T>(y)); }
// The various op= operators, for vectors...
SINT Vec<N,T>& operator+=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x + y); }
SINT Vec<N,T>& operator-=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x - y); }
SINT Vec<N,T>& operator*=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x * y); }
SINT Vec<N,T>& operator/=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x / y); }
SINT Vec<N,T>& operator^=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x ^ y); }
SINT Vec<N,T>& operator&=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x & y); }
SINT Vec<N,T>& operator|=(Vec<N,T>& x, const Vec<N,T>& y) { return (x = x | y); }
// ... for scalars...
SINTU Vec<N,T>& operator+=(Vec<N,T>& x, U y) { return (x = x + Vec<N,T>(y)); }
SINTU Vec<N,T>& operator-=(Vec<N,T>& x, U y) { return (x = x - Vec<N,T>(y)); }
SINTU Vec<N,T>& operator*=(Vec<N,T>& x, U y) { return (x = x * Vec<N,T>(y)); }
SINTU Vec<N,T>& operator/=(Vec<N,T>& x, U y) { return (x = x / Vec<N,T>(y)); }
SINTU Vec<N,T>& operator^=(Vec<N,T>& x, U y) { return (x = x ^ Vec<N,T>(y)); }
SINTU Vec<N,T>& operator&=(Vec<N,T>& x, U y) { return (x = x & Vec<N,T>(y)); }
SINTU Vec<N,T>& operator|=(Vec<N,T>& x, U y) { return (x = x | Vec<N,T>(y)); }
// ... and for shifts.
SINT Vec<N,T>& operator<<=(Vec<N,T>& x, int bits) { return (x = x << bits); }
SINT Vec<N,T>& operator>>=(Vec<N,T>& x, int bits) { return (x = x >> bits); }
// cast() Vec<N,S> to Vec<N,D>, as if applying a C-cast to each lane.
template <typename D, typename S>
SI Vec<1,D> cast(const Vec<1,S>& src) { return (D)src.val; }
template <typename D, int N, typename S>
SI Vec<N,D> cast(const Vec<N,S>& src) {
#if !defined(SKNX_NO_SIMD) && defined(__clang__)
return to_vec(__builtin_convertvector(to_vext(src), VExt<N,D>));
#else
return join(cast<D>(src.lo), cast<D>(src.hi));
#endif
}
// min/max match logic of std::min/std::max, which is important when NaN is involved.
SIT T min(const Vec<1,T>& x) { return x.val; }
SIT T max(const Vec<1,T>& x) { return x.val; }
SINT T min(const Vec<N,T>& x) { return std::min(min(x.lo), min(x.hi)); }
SINT T max(const Vec<N,T>& x) { return std::max(max(x.lo), max(x.hi)); }
SINT Vec<N,T> min(const Vec<N,T>& x, const Vec<N,T>& y) { return naive_if_then_else(y < x, y, x); }
SINT Vec<N,T> max(const Vec<N,T>& x, const Vec<N,T>& y) { return naive_if_then_else(x < y, y, x); }
SINTU Vec<N,T> min(const Vec<N,T>& x, U y) { return min(x, Vec<N,T>(y)); }
SINTU Vec<N,T> max(const Vec<N,T>& x, U y) { return max(x, Vec<N,T>(y)); }
SINTU Vec<N,T> min(U x, const Vec<N,T>& y) { return min(Vec<N,T>(x), y); }
SINTU Vec<N,T> max(U x, const Vec<N,T>& y) { return max(Vec<N,T>(x), y); }
// Shuffle values from a vector pretty arbitrarily:
// skvx::Vec<4,float> rgba = {R,G,B,A};
// shuffle<2,1,0,3> (rgba) ~> {B,G,R,A}
// shuffle<2,1> (rgba) ~> {B,G}
// shuffle<2,1,2,1,2,1,2,1>(rgba) ~> {B,G,B,G,B,G,B,G}
// shuffle<3,3,3,3> (rgba) ~> {A,A,A,A}
// The only real restriction is that the output also be a legal N=power-of-two sknx::Vec.
template <int... Ix, int N, typename T>
SI Vec<sizeof...(Ix),T> shuffle(const Vec<N,T>& x) {
#if !defined(SKNX_NO_SIMD) && defined(__clang__)
return to_vec<sizeof...(Ix),T>(__builtin_shufflevector(to_vext(x), to_vext(x), Ix...));
#else
return { x[Ix]... };
#endif
}
// fma() delivers a fused mul-add, even if that's really expensive.
SI Vec<1,float> fma(const Vec<1,float>& x, const Vec<1,float>& y, const Vec<1,float>& z) {
return std::fma(x.val, y.val, z.val);
}
SIN Vec<N,float> fma(const Vec<N,float>& x, const Vec<N,float>& y, const Vec<N,float>& z) {
return join(fma(x.lo, y.lo, z.lo),
fma(x.hi, y.hi, z.hi));
}
template <int N, typename T, typename Fn, std::size_t... I>
#if defined(__clang__)
// CFI, specifically -fsanitize=cfi-icall, seems to give a false positive here,
// with errors like "control flow integrity check for type 'float (float)
// noexcept' failed during indirect function call... note: sqrtf.cfi_jt defined
// here". But we can be quite sure fn is the right type: it's all inferred!
// So, stifle CFI in this function.
__attribute__((no_sanitize("cfi")))
#endif
SI auto map(const skvx::Vec<N,T>& x, Fn&& fn,
std::index_sequence<I...> ix = {}) -> skvx::Vec<N, decltype(fn(x[0]))> {
if /*constexpr*/ (sizeof...(I) == 0) {
// When called as map(x, fn), bootstrap the index_sequence we want: 0,1,...,N-1.
return map(x, fn, std::make_index_sequence<N>{});
}
return { fn(x[I])... };
}
SIN Vec<N,float> atan(const Vec<N,float>& x) { return map(x, atanf); }
SIN Vec<N,float> ceil(const Vec<N,float>& x) { return map(x, ceilf); }
SIN Vec<N,float> floor(const Vec<N,float>& x) { return map(x, floorf); }
SIN Vec<N,float> trunc(const Vec<N,float>& x) { return map(x, truncf); }
SIN Vec<N,float> round(const Vec<N,float>& x) { return map(x, roundf); }
SIN Vec<N,float> sqrt(const Vec<N,float>& x) { return map(x, sqrtf); }
SIN Vec<N,float> abs(const Vec<N,float>& x) { return map(x, fabsf); }
SIN Vec<N,float> sin(const Vec<N,float>& x) { return map(x, sinf); }
SIN Vec<N,float> cos(const Vec<N,float>& x) { return map(x, cosf); }
SIN Vec<N,float> tan(const Vec<N,float>& x) { return map(x, tanf); }
SI Vec<1,int> lrint(const Vec<1,float>& x) {
return (int)lrintf(x.val);
}
SIN Vec<N,int> lrint(const Vec<N,float>& x) {
#if defined(__AVX__)
if /*constexpr*/ (N == 8) {
return unchecked_bit_pun<Vec<N,int>>(_mm256_cvtps_epi32(unchecked_bit_pun<__m256>(x)));
}
#endif
#if defined(__SSE__)
if /*constexpr*/ (N == 4) {
return unchecked_bit_pun<Vec<N,int>>(_mm_cvtps_epi32(unchecked_bit_pun<__m128>(x)));
}
#endif
return join(lrint(x.lo),
lrint(x.hi));
}
SIN Vec<N,float> rcp(const Vec<N,float>& x) { return 1/x; }
SIN Vec<N,float> rsqrt(const Vec<N,float>& x) { return rcp(sqrt(x)); }
SIN Vec<N,float> fract(const Vec<N,float>& x) { return x - floor(x); }
// The default cases for to_half/from_half are borrowed from skcms,
// and assume inputs are finite and treat/flush denorm half floats as/to zero.
// Key constants to watch for:
// - a float is 32-bit, 1-8-23 sign-exponent-mantissa, with 127 exponent bias;
// - a half is 16-bit, 1-5-10 sign-exponent-mantissa, with 15 exponent bias.
SIN Vec<N,uint16_t> to_half_finite_ftz(const Vec<N,float>& x) {
Vec<N,uint32_t> sem = bit_pun<Vec<N,uint32_t>>(x),
s = sem & 0x8000'0000,
em = sem ^ s,
is_denorm = em < 0x3880'0000;
return cast<uint16_t>(if_then_else(is_denorm, Vec<N,uint32_t>(0)
, (s>>16) + (em>>13) - ((127-15)<<10)));
}
SIN Vec<N,float> from_half_finite_ftz(const Vec<N,uint16_t>& x) {
Vec<N,uint32_t> wide = cast<uint32_t>(x),
s = wide & 0x8000,
em = wide ^ s;
auto is_denorm = bit_pun<Vec<N,int32_t>>(em < 0x0400);
return if_then_else(is_denorm, Vec<N,float>(0)
, bit_pun<Vec<N,float>>( (s<<16) + (em<<13) + ((127-15)<<23) ));
}
// Like if_then_else(), these N=1 base cases won't actually be used unless explicitly called.
SI Vec<1,uint16_t> to_half(const Vec<1,float>& x) { return to_half_finite_ftz(x); }
SI Vec<1,float> from_half(const Vec<1,uint16_t>& x) { return from_half_finite_ftz(x); }
SIN Vec<N,uint16_t> to_half(const Vec<N,float>& x) {
#if defined(__F16C__)
if /*constexpr*/ (N == 8) {
return unchecked_bit_pun<Vec<N,uint16_t>>(_mm256_cvtps_ph(unchecked_bit_pun<__m256>(x),
_MM_FROUND_CUR_DIRECTION));
}
#endif
#if defined(__aarch64__)
if /*constexpr*/ (N == 4) {
return unchecked_bit_pun<Vec<N,uint16_t>>(vcvt_f16_f32(unchecked_bit_pun<float32x4_t>(x)));
}
#endif
if /*constexpr*/ (N > 4) {
return join(to_half(x.lo),
to_half(x.hi));
}
return to_half_finite_ftz(x);
}
SIN Vec<N,float> from_half(const Vec<N,uint16_t>& x) {
#if defined(__F16C__)
if /*constexpr*/ (N == 8) {
return unchecked_bit_pun<Vec<N,float>>(_mm256_cvtph_ps(unchecked_bit_pun<__m128i>(x)));
}
#endif
#if defined(__aarch64__)
if /*constexpr*/ (N == 4) {
return unchecked_bit_pun<Vec<N,float>>(vcvt_f32_f16(unchecked_bit_pun<float16x4_t>(x)));
}
#endif
if /*constexpr*/ (N > 4) {
return join(from_half(x.lo),
from_half(x.hi));
}
return from_half_finite_ftz(x);
}
// div255(x) = (x + 127) / 255 is a bit-exact rounding divide-by-255, packing down to 8-bit.
SIN Vec<N,uint8_t> div255(const Vec<N,uint16_t>& x) {
return cast<uint8_t>( (x+127)/255 );
}
// approx_scale(x,y) approximates div255(cast<uint16_t>(x)*cast<uint16_t>(y)) within a bit,
// and is always perfect when x or y is 0 or 255.
SIN Vec<N,uint8_t> approx_scale(const Vec<N,uint8_t>& x, const Vec<N,uint8_t>& y) {
// All of (x*y+x)/256, (x*y+y)/256, and (x*y+255)/256 meet the criteria above.
// We happen to have historically picked (x*y+x)/256.
auto X = cast<uint16_t>(x),
Y = cast<uint16_t>(y);
return cast<uint8_t>( (X*Y+X)/256 );
}
#if !defined(SKNX_NO_SIMD) && defined(__ARM_NEON)
// With NEON we can do eight u8*u8 -> u16 in one instruction, vmull_u8 (read, mul-long).
SI Vec<8,uint16_t> mull(const Vec<8,uint8_t>& x,
const Vec<8,uint8_t>& y) {
return to_vec<8,uint16_t>(vmull_u8(to_vext(x),
to_vext(y)));
}
SIN std::enable_if_t<(N < 8), Vec<N,uint16_t>> mull(const Vec<N,uint8_t>& x,
const Vec<N,uint8_t>& y) {
// N < 8 --> double up data until N == 8, returning the part we need.
return mull(join(x,x),
join(y,y)).lo;
}
SIN std::enable_if_t<(N > 8), Vec<N,uint16_t>> mull(const Vec<N,uint8_t>& x,
const Vec<N,uint8_t>& y) {
// N > 8 --> usual join(lo,hi) strategy to recurse down to N == 8.
return join(mull(x.lo, y.lo),
mull(x.hi, y.hi));
}
#else
// Nothing special when we don't have NEON... just cast up to 16-bit and multiply.
SIN Vec<N,uint16_t> mull(const Vec<N,uint8_t>& x,
const Vec<N,uint8_t>& y) {
return cast<uint16_t>(x)
* cast<uint16_t>(y);
}
#endif
#if !defined(SKNX_NO_SIMD)
// Platform-specific specializations and overloads can now drop in here.
#if defined(__AVX__)
SI Vec<8,float> rsqrt(const Vec<8,float>& x) {
return bit_pun<Vec<8,float>>(_mm256_rsqrt_ps(bit_pun<__m256>(x)));
}
SI Vec<8,float> rcp(const Vec<8,float>& x) {
return bit_pun<Vec<8,float>>(_mm256_rcp_ps(bit_pun<__m256>(x)));
}
#endif
#if defined(__SSE__)
SI Vec<4,float> rsqrt(const Vec<4,float>& x) {
return bit_pun<Vec<4,float>>(_mm_rsqrt_ps(bit_pun<__m128>(x)));
}
SI Vec<4,float> rcp(const Vec<4,float>& x) {
return bit_pun<Vec<4,float>>(_mm_rcp_ps(bit_pun<__m128>(x)));
}
SI Vec<2,float> rsqrt(const Vec<2,float>& x) {
return shuffle<0,1>(rsqrt(shuffle<0,1,0,1>(x)));
}
SI Vec<2,float> rcp(const Vec<2,float>& x) {
return shuffle<0,1>( rcp(shuffle<0,1,0,1>(x)));
}
#endif
#if defined(__AVX2__)
SI Vec<4,float> fma(const Vec<4,float>& x, const Vec<4,float>& y, const Vec<4,float>& z) {
return bit_pun<Vec<4,float>>(_mm_fmadd_ps(bit_pun<__m128>(x),
bit_pun<__m128>(y),
bit_pun<__m128>(z)));
}
SI Vec<8,float> fma(const Vec<8,float>& x, const Vec<8,float>& y, const Vec<8,float>& z) {
return bit_pun<Vec<8,float>>(_mm256_fmadd_ps(bit_pun<__m256>(x),
bit_pun<__m256>(y),
bit_pun<__m256>(z)));
}
#elif defined(__aarch64__)
SI Vec<4,float> fma(const Vec<4,float>& x, const Vec<4,float>& y, const Vec<4,float>& z) {
// These instructions tend to work like z += xy, so the order here is z,x,y.
return bit_pun<Vec<4,float>>(vfmaq_f32(bit_pun<float32x4_t>(z),
bit_pun<float32x4_t>(x),
bit_pun<float32x4_t>(y)));
}
#endif
#if defined __wasm_simd128__
SI bool any(const Vec<4, int32_t>& x) { return wasm_i32x4_any_true(to_vext(x)); }
SI bool any(const Vec<4,uint32_t>& x) { return wasm_i32x4_any_true(to_vext(x)); }
SI bool all(const Vec<4, int32_t>& x) { return wasm_i32x4_all_true(to_vext(x)); }
SI bool all(const Vec<4,uint32_t>& x) { return wasm_i32x4_all_true(to_vext(x)); }
#endif
#endif // !defined(SKNX_NO_SIMD)
} // namespace skvx
#undef SINTU
#undef SINT
#undef SIT
#undef SI
#endif//SKVX_DEFINED