/*
* Copyright (C) 2020 Linux Studio Plugins Project
* (C) 2020 Vladimir Sadovnikov
*
* This file is part of lsp-dsp-lib
* Created on: 31 мар. 2020 г.
*
* lsp-dsp-lib is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* any later version.
*
* lsp-dsp-lib is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with lsp-dsp-lib. If not, see .
*/
#ifndef PRIVATE_DSP_ARCH_X86_SSE2_GRAPHICS_EFFECTS_H_
#define PRIVATE_DSP_ARCH_X86_SSE2_GRAPHICS_EFFECTS_H_
#ifndef PRIVATE_DSP_ARCH_X86_SSE2_IMPL
#error "This header should not be included directly"
#endif /* PRIVATE_DSP_ARCH_X86_SSE2_IMPL */
#define X4_TRANSPOSE \
__ASM_EMIT("movaps %%xmm2, %%xmm4") \
__ASM_EMIT("punpckldq %%xmm3, %%xmm2") \
__ASM_EMIT("punpckhdq %%xmm3, %%xmm4") \
__ASM_EMIT("movaps %%xmm0, %%xmm3") \
__ASM_EMIT("punpckldq %%xmm1, %%xmm0") \
__ASM_EMIT("punpckhdq %%xmm1, %%xmm3") \
__ASM_EMIT("movaps %%xmm0, %%xmm1") \
__ASM_EMIT("punpcklqdq %%xmm2, %%xmm0") \
__ASM_EMIT("punpckhqdq %%xmm2, %%xmm1") \
__ASM_EMIT("movaps %%xmm3, %%xmm2") \
__ASM_EMIT("punpcklqdq %%xmm4, %%xmm2") \
__ASM_EMIT("punpckhqdq %%xmm4, %%xmm3")
namespace lsp
{
namespace sse2
{
IF_ARCH_X86(
static const float EFF_HSLA_HUE_XC[] __lsp_aligned16 =
{
LSP_DSP_VEC4(1.0f)
};
)
#define EFF_HSLA_HUE_CORE \
/* xmm0 = v, xmm6 = T, xmm7 = KT */ \
__ASM_EMIT("movaps %[XC], %%xmm5") /* xmm5 = 1 */ \
__ASM_EMIT("xorps %%xmm4, %%xmm4") /* xmm4 = 0 */ \
__ASM_EMIT("movaps %%xmm5, %%xmm1") /* xmm1 = 1 */ \
__ASM_EMIT("cmpps $6, %%xmm0, %%xmm4") /* xmm4 = [0 > v] */ \
__ASM_EMIT("movups 0x00(%[eff]), %%xmm2") /* xmm2 = h s l a */ \
__ASM_EMIT("subps %%xmm0, %%xmm1") /* xmm1 = 1 - v */ \
__ASM_EMIT("shufps $0x00, %%xmm2, %%xmm2") /* xmm2 = EH */ \
__ASM_EMIT("addps %%xmm5, %%xmm0") /* xmm0 = 1 + v */ \
__ASM_EMIT("andps %%xmm4, %%xmm0") /* xmm0 = (1+v) & [0 > v] */ \
__ASM_EMIT("andnps %%xmm1, %%xmm4") /* xmm4 = (1-v) & [0 <= v] */ \
__ASM_EMIT("orps %%xmm4, %%xmm0") /* xmm0 = V = ((1+v) & [0 > v]) | ((1-v) & [0 <= v]) */ \
/* xmm0 = V */ \
/* xmm2 = EH */ \
/* xmm6 = T */ \
/* xmm7 = KT */ \
__ASM_EMIT("movaps %%xmm0, %%xmm1") /* xmm1 = V */ \
__ASM_EMIT("xorps %%xmm4, %%xmm4") /* xmm4 = 0 */ \
__ASM_EMIT("subps %%xmm6, %%xmm1") /* xmm1 = V - T */ \
__ASM_EMIT("addps %%xmm2, %%xmm0") /* xmm0 = EH + V */ \
__ASM_EMIT("movaps %%xmm1, %%xmm3") /* xmm2 = V - T */ \
__ASM_EMIT("addps %%xmm6, %%xmm2") /* xmm3 = EH + T */ \
__ASM_EMIT("mulps %%xmm7, %%xmm3") /* xmm3 = (V-T)*KT */ \
__ASM_EMIT("cmpps $2, %%xmm1, %%xmm4") /* xmm4 = [(V - T) >= 0] */ \
__ASM_EMIT("andps %%xmm4, %%xmm3") /* xmm3 = A = ((V-T)*KT) & [(V - T) >= 0] */ \
__ASM_EMIT("andps %%xmm4, %%xmm2") /* xmm2 = (EH + T) & [(V - T) >= 0] */ \
__ASM_EMIT("andnps %%xmm0, %%xmm4") /* xmm4 = (EH + V) & [(V - T) < 0] */ \
__ASM_EMIT("orps %%xmm2, %%xmm4") /* xmm4 = NH = ((EH + T) & [(V - T) >= 0]) | ((EH + V) & [(V - T) < 0]) */ \
__ASM_EMIT("movaps %[XC], %%xmm0") /* xmm0 = 1 */ \
__ASM_EMIT("movaps %%xmm4, %%xmm1") /* xmm1 = NH */ \
__ASM_EMIT("movaps %%xmm0, %%xmm5") /* xmm5 = 1 */ \
__ASM_EMIT("cmpps $6, %%xmm4, %%xmm0") /* xmm0 = [1 > NH] */ \
__ASM_EMIT("subps %%xmm5, %%xmm4") /* xmm4 = NH - 1 */ \
__ASM_EMIT("movups 0x00(%[eff]), %%xmm2") /* xmm2 = h s l a */ \
__ASM_EMIT("andps %%xmm0, %%xmm1") /* xmm1 = NH & [1 > H] */ \
__ASM_EMIT("andnps %%xmm4, %%xmm0") /* xmm0 = (NH-1) & [1 <= H] */ \
__ASM_EMIT("orps %%xmm1, %%xmm0") /* xmm0 = H = (NH & [1 > H]) | ((NH-1) & [1 <= H]) */ \
__ASM_EMIT("movaps %%xmm2, %%xmm1") /* xmm1 = h s l a */ \
__ASM_EMIT("shufps $0x55, %%xmm1, %%xmm1") /* xmm1 = S */ \
__ASM_EMIT("shufps $0xaa, %%xmm2, %%xmm2") /* xmm2 = L */ \
\
X4_TRANSPOSE
void eff_hsla_hue(float *dst, const float *v, const dsp::hsla_hue_eff_t *eff, size_t count)
{
ARCH_X86_ASM(
__ASM_EMIT("movaps %[XC], %%xmm6") /* xmm6 = 1 */
__ASM_EMIT("movss 0x10(%[eff]), %%xmm4") /* xmm4 = t 0 0 0 */
__ASM_EMIT("movaps %%xmm6, %%xmm7") /* xmm7 = 1 */
__ASM_EMIT("shufps $0x00, %%xmm4, %%xmm4") /* xmm4 = t */
__ASM_EMIT("subps %%xmm4, %%xmm6") /* xmm6 = T = 1 - t */
__ASM_EMIT("divps %%xmm4, %%xmm7") /* xmm7 = KT = 1 / t */
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jb 2f")
//-----------------------------------------------------------------
// 4x blocks
__ASM_EMIT("1:")
__ASM_EMIT("movups 0x00(%[src]), %%xmm0") /* xmm0 = v */
EFF_HSLA_HUE_CORE
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("movups %%xmm2, 0x20(%[dst])")
__ASM_EMIT("movups %%xmm3, 0x30(%[dst])")
__ASM_EMIT("add $0x10, %[src]")
__ASM_EMIT("add $0x40, %[dst]")
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jae 1b")
__ASM_EMIT("2:")
__ASM_EMIT("add $4, %[count]")
__ASM_EMIT("jle 10f")
//-----------------------------------------------------------------
// 1x - 3x block
// Load last variable-sized chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 4f")
__ASM_EMIT("movss 0x00(%[src]), %%xmm0")
__ASM_EMIT("add $0x04, %[src]")
__ASM_EMIT("movlhps %%xmm0, %%xmm0")
__ASM_EMIT("4:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 6f")
__ASM_EMIT("movlps 0x00(%[src]), %%xmm0")
__ASM_EMIT("6:")
EFF_HSLA_HUE_CORE
// Store last chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 8f")
__ASM_EMIT("movups %%xmm2, 0x00(%[dst])")
__ASM_EMIT("add $0x10, %[dst]")
__ASM_EMIT("8:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 10f")
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("10:")
: [dst] "+r" (dst), [src] "+r" (v), [count] "+r" (count)
: [eff] "r" (eff),
[XC] "m" (EFF_HSLA_HUE_XC)
: "cc", "memory",
"%xmm0", "%xmm1", "%xmm2", "%xmm3",
"%xmm4", "%xmm5", "%xmm6", "%xmm7"
);
/*
float value, hue, alpha;
float t = 1.0f - eff->thresh;
float kt = 1.0f / eff->thresh;
for (size_t i=0; i value) ? 1.0f + value : 1.0f - value;
if ((value - t) >= 0)
{
hue = eff->h + t;
alpha = ((value - t) * kt);
}
else
{
hue = eff->h + value;
alpha = 0.0f;
}
dst[0] = (hue < 1.0f) ? hue : hue - 1.0f;
dst[1] = eff->s;
dst[2] = eff->l;
dst[3] = alpha;
}*/
}
#undef EFF_HSLA_HUE_CORE
IF_ARCH_X86(
static const float EFF_HSLA_ALPHA_XC[] __lsp_aligned16 =
{
LSP_DSP_VEC4(1.0f)
};
)
#define EFF_HSLA_ALPHA_CORE \
/* xmm3 = v */ \
__ASM_EMIT("xorps %%xmm4, %%xmm4") /* xmm4 = 0 */ \
__ASM_EMIT("movaps %[XC], %%xmm5") /* xmm5 = 1 */ \
__ASM_EMIT("cmpps $6, %%xmm3, %%xmm4") /* xmm4 = [0 > v] */ \
__ASM_EMIT("movups 0x00(%[eff]), %%xmm0") /* xmm0 = hsla */ \
__ASM_EMIT("movaps %%xmm3, %%xmm6") /* xmm6 = v */ \
__ASM_EMIT("movaps %%xmm0, %%xmm1") /* xmm1 = hsla */ \
__ASM_EMIT("addps %%xmm5, %%xmm3") /* xmm3 = 1 + v */ \
__ASM_EMIT("movaps %%xmm1, %%xmm2") /* xmm2 = hsla */ \
__ASM_EMIT("subps %%xmm6, %%xmm5") /* xmm5 = 1 - v */ \
__ASM_EMIT("shufps $0x00, %%xmm0, %%xmm0") /* xmm0 = H */ \
__ASM_EMIT("andps %%xmm4, %%xmm3") /* xmm4 = (1+v) & [0 > v] */ \
__ASM_EMIT("shufps $0x55, %%xmm1, %%xmm1") /* xmm1 = S */ \
__ASM_EMIT("andnps %%xmm5, %%xmm4") /* xmm4 = (1-v) & [0 <= v] */ \
__ASM_EMIT("shufps $0xaa, %%xmm2, %%xmm2") /* xmm1 = L */ \
__ASM_EMIT("orps %%xmm4, %%xmm3") /* xmm3 = A = ((1+v) & [0 > v]) | ((1-v) & [0 <= v]) */ \
\
X4_TRANSPOSE
/*
value = v[i];
value = (0.0f > value) ? 1.0f + value : 1.0f - value;
dst[0] = eff->h;
dst[1] = eff->s;
dst[2] = eff->l;
dst[3] = value; // Fill alpha channel
*/
void eff_hsla_alpha(float *dst, const float *v, const dsp::hsla_alpha_eff_t *eff, size_t count)
{
ARCH_X86_ASM(
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jb 2f")
//-----------------------------------------------------------------
// 4x blocks
__ASM_EMIT("1:")
__ASM_EMIT("movups 0x00(%[src]), %%xmm3") /* xmm3 = v */
EFF_HSLA_ALPHA_CORE
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("movups %%xmm2, 0x20(%[dst])")
__ASM_EMIT("movups %%xmm3, 0x30(%[dst])")
__ASM_EMIT("add $0x10, %[src]")
__ASM_EMIT("add $0x40, %[dst]")
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jae 1b")
__ASM_EMIT("2:")
__ASM_EMIT("add $4, %[count]")
__ASM_EMIT("jle 10f")
//-----------------------------------------------------------------
// 1x - 3x block
// Load last variable-sized chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 4f")
__ASM_EMIT("movss 0x00(%[src]), %%xmm3")
__ASM_EMIT("add $0x04, %[src]")
__ASM_EMIT("movlhps %%xmm3, %%xmm3")
__ASM_EMIT("4:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 6f")
__ASM_EMIT("movlps 0x00(%[src]), %%xmm3")
__ASM_EMIT("6:")
EFF_HSLA_ALPHA_CORE
// Store last chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 8f")
__ASM_EMIT("movups %%xmm2, 0x00(%[dst])")
__ASM_EMIT("add $0x10, %[dst]")
__ASM_EMIT("8:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 10f")
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("10:")
: [dst] "+r" (dst), [src] "+r" (v), [count] "+r" (count)
: [eff] "r" (eff),
[XC] "m" (EFF_HSLA_ALPHA_XC)
: "cc", "memory",
"%xmm0", "%xmm1", "%xmm2", "%xmm3",
"%xmm4", "%xmm5", "%xmm6", "%xmm7"
);
}
#undef EFF_HSLA_ALPHA_CORE
IF_ARCH_X86(
static const uint32_t EFF_HSLA_SAT_XC[] __lsp_aligned16 =
{
LSP_DSP_VEC4(0x3f800000),
LSP_DSP_VEC4(0x7fffffff)
};
)
#define EFF_HSLA_SAT_CORE \
/* xmm1 = v, xmm6 = T, xmm7 = KT */ \
__ASM_EMIT("movaps %%xmm6, %%xmm0") /* xmm0 = T */ \
__ASM_EMIT("andps 0x10(%[XC]), %%xmm1") /* xmm1 = V */ \
__ASM_EMIT("xorps %%xmm5, %%xmm5") /* xmm5 = 0 */ \
__ASM_EMIT("subps %%xmm1, %%xmm0") /* xmm0 = T-V */ \
__ASM_EMIT("cmpps $6, %%xmm0, %%xmm5") /* xmm5 = [ 0 >= T - V] */ \
__ASM_EMIT("mulps %%xmm7, %%xmm0") /* xmm0 = (T-V)*KT */ \
__ASM_EMIT("movaps %%xmm5, %%xmm3") /* xmm3 = [ 0 >= T - V] */ \
__ASM_EMIT("movups 0x00(%[eff]), %%xmm2") /* xmm2 = hsla */ \
__ASM_EMIT("andps %%xmm5, %%xmm1") /* xmm1 = V & [ 0 >= T - V] */ \
__ASM_EMIT("andnps %%xmm0, %%xmm3") /* xmm3 = A = (T-V)*KT & [ 0 < T - V] */ \
__ASM_EMIT("andnps %%xmm6, %%xmm5") /* xmm5 = T & [0 < T-V] */ \
__ASM_EMIT("movaps %%xmm2, %%xmm0") /* xmm0 = hsla */ \
__ASM_EMIT("orps %%xmm5, %%xmm1") /* xmm1 = KS = (V & [ 0 >= T - V]) | (T & [0 < T-V]) */ \
__ASM_EMIT("movaps %%xmm2, %%xmm4") /* xmm4 = hsla */ \
__ASM_EMIT("shufps $0x00, %%xmm0, %%xmm0") /* xmm0 = H */ \
__ASM_EMIT("shufps $0x55, %%xmm4, %%xmm4") /* xmm4 = s */ \
__ASM_EMIT("shufps $0xaa, %%xmm2, %%xmm2") /* xmm2 = L */ \
__ASM_EMIT("mulps %%xmm4, %%xmm1") /* xmm1 = S = s * KS */ \
\
X4_TRANSPOSE
/*
kt = 1.0f / eff->thresh;
value = (value >= 0.0f) ? value : -value;
if (0 >= (eff->thresh - value))
{
dst[1] = eff->s * value;
dst[3] = 0.0f;
}
else
{
dst[1] = eff->s * eff->thresh;
dst[3] = (eff->thresh - value) * kt;
}
dst[0] = eff->h;
dst[2] = eff->l;
*/
void eff_hsla_sat(float *dst, const float *v, const dsp::hsla_sat_eff_t *eff, size_t count)
{
ARCH_X86_ASM(
__ASM_EMIT("movss 0x10(%[eff]), %%xmm6") /* xmm6 = t 0 0 0 */
__ASM_EMIT("movaps 0x00(%[XC]), %%xmm7") /* xmm7 = 1 */
__ASM_EMIT("shufps $0x00, %%xmm6, %%xmm6") /* xmm6 = T */
__ASM_EMIT("divps %%xmm6, %%xmm7") /* xmm7 = KT = 1 / t */
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jb 2f")
//-----------------------------------------------------------------
// 4x blocks
__ASM_EMIT("1:")
__ASM_EMIT("movups 0x00(%[src]), %%xmm1") /* xmm1 = v */
EFF_HSLA_SAT_CORE
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("movups %%xmm2, 0x20(%[dst])")
__ASM_EMIT("movups %%xmm3, 0x30(%[dst])")
__ASM_EMIT("add $0x10, %[src]")
__ASM_EMIT("add $0x40, %[dst]")
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jae 1b")
__ASM_EMIT("2:")
__ASM_EMIT("add $4, %[count]")
__ASM_EMIT("jle 10f")
//-----------------------------------------------------------------
// 1x - 3x block
// Load last variable-sized chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 4f")
__ASM_EMIT("movss 0x00(%[src]), %%xmm1")
__ASM_EMIT("add $0x04, %[src]")
__ASM_EMIT("movlhps %%xmm1, %%xmm1")
__ASM_EMIT("4:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 6f")
__ASM_EMIT("movlps 0x00(%[src]), %%xmm1")
__ASM_EMIT("6:")
EFF_HSLA_SAT_CORE
// Store last chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 8f")
__ASM_EMIT("movups %%xmm2, 0x00(%[dst])")
__ASM_EMIT("add $0x10, %[dst]")
__ASM_EMIT("8:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 10f")
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("10:")
: [dst] "+r" (dst), [src] "+r" (v), [count] "+r" (count)
: [eff] "r" (eff),
[XC] "r" (EFF_HSLA_SAT_XC)
: "cc", "memory",
"%xmm0", "%xmm1", "%xmm2", "%xmm3",
"%xmm4", "%xmm5", "%xmm6", "%xmm7"
);
}
#undef EFF_HSLA_SAT_CORE
IF_ARCH_X86(
static const uint32_t EFF_HSLA_LIGHT_XC[] __lsp_aligned16 =
{
LSP_DSP_VEC4(0x3f800000),
LSP_DSP_VEC4(0x7fffffff)
};
)
#define EFF_HSLA_LIGHT_CORE \
/* xmm2 = v, xmm6 = T, xmm7 = KT */ \
__ASM_EMIT("movaps %%xmm6, %%xmm0") /* xmm0 = T */ \
__ASM_EMIT("andps 0x10(%[XC]), %%xmm2") /* xmm2 = V */ \
__ASM_EMIT("xorps %%xmm5, %%xmm5") /* xmm5 = 0 */ \
__ASM_EMIT("subps %%xmm2, %%xmm0") /* xmm0 = T-V */ \
__ASM_EMIT("cmpps $6, %%xmm0, %%xmm5") /* xmm5 = [ 0 >= T - V] */ \
__ASM_EMIT("mulps %%xmm7, %%xmm0") /* xmm0 = (T-V)*KT */ \
__ASM_EMIT("movaps %%xmm5, %%xmm3") /* xmm3 = [ 0 >= T - V] */ \
__ASM_EMIT("movups 0x00(%[eff]), %%xmm1") /* xmm1 = hsla */ \
__ASM_EMIT("andps %%xmm5, %%xmm2") /* xmm2 = V & [ 0 >= T - V] */ \
__ASM_EMIT("andnps %%xmm0, %%xmm3") /* xmm3 = A = (T-V)*KT & [ 0 < T - V] */ \
__ASM_EMIT("andnps %%xmm6, %%xmm5") /* xmm5 = T & [0 < T-V] */ \
__ASM_EMIT("movaps %%xmm1, %%xmm0") /* xmm0 = hsla */ \
__ASM_EMIT("orps %%xmm5, %%xmm2") /* xmm2 = KL = (V & [ 0 >= T - V]) | (T & [0 < T-V]) */ \
__ASM_EMIT("movaps %%xmm1, %%xmm4") /* xmm4 = hsla */ \
__ASM_EMIT("shufps $0x00, %%xmm0, %%xmm0") /* xmm0 = H */ \
__ASM_EMIT("shufps $0xaa, %%xmm4, %%xmm4") /* xmm4 = l */ \
__ASM_EMIT("shufps $0x55, %%xmm1, %%xmm1") /* xmm1 = S */ \
__ASM_EMIT("mulps %%xmm4, %%xmm2") /* xmm2 = L = l * KL */ \
\
X4_TRANSPOSE
/*
value = (value >= 0.0f) ? value : -value;
if (0 >= (eff->thresh - value))
{
dst[2] = eff->l * value;
dst[3] = 0.0f;
}
else
{
dst[2] = eff->l * eff->thresh;
dst[3] = (eff->thresh - value) * kt;
}
dst[0] = eff->h;
dst[1] = eff->s;
*/
void eff_hsla_light(float *dst, const float *v, const dsp::hsla_light_eff_t *eff, size_t count)
{
ARCH_X86_ASM(
__ASM_EMIT("movss 0x10(%[eff]), %%xmm6") /* xmm6 = t 0 0 0 */
__ASM_EMIT("movaps 0x00(%[XC]), %%xmm7") /* xmm7 = 1 */
__ASM_EMIT("shufps $0x00, %%xmm6, %%xmm6") /* xmm6 = T */
__ASM_EMIT("divps %%xmm6, %%xmm7") /* xmm7 = KT = 1 / t */
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jb 2f")
//-----------------------------------------------------------------
// 4x blocks
__ASM_EMIT("1:")
__ASM_EMIT("movups 0x00(%[src]), %%xmm2") /* xmm1 = v */
EFF_HSLA_LIGHT_CORE
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("movups %%xmm2, 0x20(%[dst])")
__ASM_EMIT("movups %%xmm3, 0x30(%[dst])")
__ASM_EMIT("add $0x10, %[src]")
__ASM_EMIT("add $0x40, %[dst]")
__ASM_EMIT("sub $4, %[count]")
__ASM_EMIT("jae 1b")
__ASM_EMIT("2:")
__ASM_EMIT("add $4, %[count]")
__ASM_EMIT("jle 10f")
//-----------------------------------------------------------------
// 1x - 3x block
// Load last variable-sized chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 4f")
__ASM_EMIT("movss 0x00(%[src]), %%xmm2")
__ASM_EMIT("add $0x04, %[src]")
__ASM_EMIT("movlhps %%xmm2, %%xmm2")
__ASM_EMIT("4:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 6f")
__ASM_EMIT("movlps 0x00(%[src]), %%xmm2")
__ASM_EMIT("6:")
EFF_HSLA_LIGHT_CORE
// Store last chunk
__ASM_EMIT("test $1, %[count]")
__ASM_EMIT("jz 8f")
__ASM_EMIT("movups %%xmm2, 0x00(%[dst])")
__ASM_EMIT("add $0x10, %[dst]")
__ASM_EMIT("8:")
__ASM_EMIT("test $2, %[count]")
__ASM_EMIT("jz 10f")
__ASM_EMIT("movups %%xmm0, 0x00(%[dst])")
__ASM_EMIT("movups %%xmm1, 0x10(%[dst])")
__ASM_EMIT("10:")
: [dst] "+r" (dst), [src] "+r" (v), [count] "+r" (count)
: [eff] "r" (eff),
[XC] "r" (EFF_HSLA_LIGHT_XC)
: "cc", "memory",
"%xmm0", "%xmm1", "%xmm2", "%xmm3",
"%xmm4", "%xmm5", "%xmm6", "%xmm7"
);
}
#undef EFF_HSLA_LIGHT_CORE
}
}
#undef X4_TRANSPOSE
#endif /* PRIVATE_DSP_ARCH_X86_SSE2_GRAPHICS_EFFECTS_H_ */