/*

* Simd Library (http://ermig1979.github.io/Simd).

*

* Copyright (c) 2011-2026 Yermalayeu Ihar.

*

* Permission is hereby granted, free of charge, to any person obtaining a copy

* of this software and associated documentation files (the "Software"), to deal

* in the Software without restriction, including without limitation the rights

* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell

* copies of the Software, and to permit persons to whom the Software is

* furnished to do so, subject to the following conditions:

*

* The above copyright notice and this permission notice shall be included in

* all copies or substantial portions of the Software.

*

* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR

* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,

* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE

* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER

* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,

* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE

* SOFTWARE.

*/

#include "Simd/SimdMemory.h"

#include "Simd/SimdStore.h"

#include "Simd/SimdConversion.h"

#include "Simd/SimdInterleave.h"

namespace Simd

{

#ifdef SIMD_AVX2_ENABLE

    namespace Avx2

    {

        SIMD_INLINE __m256i MulDiv32Hsv(__m256i dividend, __m256i divisor, const __m256& scale)

        {

            return _mm256_cvttps_epi32(_mm256_div_ps(_mm256_mul_ps(scale, _mm256_cvtepi32_ps(dividend)), _mm256_cvtepi32_ps(divisor)));

        }

        SIMD_INLINE __m256i MulDiv16Hsv(__m256i dividend, __m256i divisor, const __m256& scale)

        {

            const __m256i lo = MulDiv32Hsv(_mm256_unpacklo_epi16(dividend, K_ZERO), _mm256_unpacklo_epi16(divisor, K_ZERO), scale);

            const __m256i hi = MulDiv32Hsv(_mm256_unpackhi_epi16(dividend, K_ZERO), _mm256_unpackhi_epi16(divisor, K_ZERO), scale);

            return _mm256_packs_epi32(lo, hi);

        }

        SIMD_INLINE void BgrToHsv16(__m256i blue, __m256i green, __m256i red,

            __m256i& hue, __m256i& sat, __m256i& val,

            const __m256& KF_255_DIV_6, const __m256& K_255F)

        {

            __m256i max = MaxI16(red, green, blue);

            __m256i min = MinI16(red, green, blue);

            __m256i range = _mm256_sub_epi16(max, min);

            // Hue: determine which channel is the maximum

            __m256i redMaxMask = _mm256_cmpeq_epi16(red, max);

            __m256i greenMaxMask = _mm256_andnot_si256(redMaxMask, _mm256_cmpeq_epi16(green, max));

            __m256i blueMaxMask = _mm256_andnot_si256(_mm256_or_si256(redMaxMask, greenMaxMask), K_INV_ZERO);

            __m256i hueDividend = _mm256_or_si256(

                _mm256_and_si256(redMaxMask,

                    _mm256_add_epi16(_mm256_sub_epi16(green, blue), _mm256_mullo_epi16(range, K16_0006))),

                _mm256_or_si256(

                    _mm256_and_si256(greenMaxMask,

                        _mm256_add_epi16(_mm256_sub_epi16(blue, red), _mm256_mullo_epi16(range, K16_0002))),

                    _mm256_and_si256(blueMaxMask,

                        _mm256_add_epi16(_mm256_sub_epi16(red, green), _mm256_mullo_epi16(range, K16_0004)))));

            __m256i safeRange = _mm256_max_epi16(range, K16_0001);

            hue = _mm256_andnot_si256(_mm256_cmpeq_epi16(range, K_ZERO),

                _mm256_and_si256(MulDiv16Hsv(hueDividend, safeRange, KF_255_DIV_6), K16_00FF));

            // Value: V = max

            val = max;

            // Saturation: S = 255 * range / max, zero when max == 0

            __m256i safeMax = _mm256_max_epi16(max, K16_0001);

            sat = _mm256_andnot_si256(_mm256_cmpeq_epi16(max, K_ZERO),

                MulDiv16Hsv(range, safeMax, K_255F));

        }

        template <bool align> SIMD_INLINE void BgrToHsv32(const uint8_t* bgr, uint8_t* hsv,

            const __m256& KF_255_DIV_6, const __m256& K_255F)

        {

            __m256i bgr_data[3];

            bgr_data[0] = Load<align>((__m256i*)bgr + 0);

            bgr_data[1] = Load<align>((__m256i*)bgr + 1);

            bgr_data[2] = Load<align>((__m256i*)bgr + 2);

            __m256i blue8 = BgrToBlue(bgr_data);

            __m256i green8 = BgrToGreen(bgr_data);

            __m256i red8 = BgrToRed(bgr_data);

            __m256i blue_lo = _mm256_unpacklo_epi8(blue8, K_ZERO);

            __m256i blue_hi = _mm256_unpackhi_epi8(blue8, K_ZERO);

            __m256i green_lo = _mm256_unpacklo_epi8(green8, K_ZERO);

            __m256i green_hi = _mm256_unpackhi_epi8(green8, K_ZERO);

            __m256i red_lo = _mm256_unpacklo_epi8(red8, K_ZERO);

            __m256i red_hi = _mm256_unpackhi_epi8(red8, K_ZERO);

            __m256i hue_lo, sat_lo, val_lo, hue_hi, sat_hi, val_hi;

            BgrToHsv16(blue_lo, green_lo, red_lo, hue_lo, sat_lo, val_lo, KF_255_DIV_6, K_255F);

            BgrToHsv16(blue_hi, green_hi, red_hi, hue_hi, sat_hi, val_hi, KF_255_DIV_6, K_255F);

            __m256i hue8 = _mm256_packus_epi16(hue_lo, hue_hi);

            __m256i sat8 = _mm256_packus_epi16(sat_lo, sat_hi);

            __m256i val8 = _mm256_packus_epi16(val_lo, val_hi);

            Store<align>((__m256i*)hsv + 0, InterleaveBgr<0>(hue8, sat8, val8));

            Store<align>((__m256i*)hsv + 1, InterleaveBgr<1>(hue8, sat8, val8));

            Store<align>((__m256i*)hsv + 2, InterleaveBgr<2>(hue8, sat8, val8));

        }

Unexecuted instantiation: void Simd::Avx2::BgrToHsv32<true>(unsigned char const*, unsigned char*, float __vector(8) const&, float __vector(8) const&)

Unexecuted instantiation: void Simd::Avx2::BgrToHsv32<false>(unsigned char const*, unsigned char*, float __vector(8) const&, float __vector(8) const&)

        template <bool align> void BgrToHsv(const uint8_t* bgr, size_t width, size_t height, size_t bgrStride, uint8_t* hsv, size_t hsvStride)

        {

            assert(width >= A);

            if (align)

                assert(Aligned(bgr) && Aligned(bgrStride) && Aligned(hsv) && Aligned(hsvStride));

            size_t alignedWidth = AlignLo(width, A);

            const __m256 KF = _mm256_set1_ps(Base::KF_255_DIV_6);

            const __m256 K255 = _mm256_set1_ps(255.0f);

            for (size_t row = 0; row < height; ++row)

            {

                for (size_t col = 0; col < alignedWidth; col += A)

                    BgrToHsv32<align>(bgr + 3 * col, hsv + 3 * col, KF, K255);

                if (width != alignedWidth)

                    BgrToHsv32<false>(bgr + 3 * (width - A), hsv + 3 * (width - A), KF, K255);

                bgr += bgrStride;

                hsv += hsvStride;

            }

        }

Unexecuted instantiation: void Simd::Avx2::BgrToHsv<true>(unsigned char const*, unsigned long, unsigned long, unsigned long, unsigned char*, unsigned long)

Unexecuted instantiation: void Simd::Avx2::BgrToHsv<false>(unsigned char const*, unsigned long, unsigned long, unsigned long, unsigned char*, unsigned long)

        void BgrToHsv(const uint8_t* bgr, size_t width, size_t height, size_t bgrStride, uint8_t* hsv, size_t hsvStride)

        {

            if (Aligned(bgr) && Aligned(bgrStride) && Aligned(hsv) && Aligned(hsvStride))

                BgrToHsv<true>(bgr, width, height, bgrStride, hsv, hsvStride);

            else

                BgrToHsv<false>(bgr, width, height, bgrStride, hsv, hsvStride);

        }

    }

#endif

}