The-Homebrew-Cloud/Cemu
2
0
Fork 0
mirror of https://github.com/cemu-project/Cemu.git synced 2024-11-23 09:39:18 +01:00
Code Issues Packages Projects Releases Wiki Activity
0b559fb872
Cemu/dependencies/ih264d/common/x86/ih264_deblk_chroma_ssse3.c

1106 lines
55 KiB
C
Raw Normal View History

Add all the files
2022-08-22 22:21:23 +02:00
/******************************************************************************
*
* Copyright (C) 2015 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at:
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
*****************************************************************************
* Originally developed and contributed by Ittiam Systems Pvt. Ltd, Bangalore
*/
/*****************************************************************************/
/* */
/* File Name : ih264_deblk_chroma_ssse3.c */
/* */
/* Description : Contains function definitions for deblocking */
/* */
/* List of Functions : ih264_deblk_chroma_vert_bs4_ssse3() */
/* ih264_deblk_chroma_horz_bs4_ssse3() */
/* ih264_deblk_chroma_vert_bslt4_ssse3() */
/* ih264_deblk_chroma_horz_bslt4_ssse3() */
/* ih264_deblk_chroma_vert_bs4_mbaff_ssse3() */
/* ih264_deblk_chroma_vert_bslt4_mbaff_ssse3() */
/* */
/* Issues / Problems : None */
/* */
/* Revision History : */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Added chrom deblocking ssse3 */
/* intrinsics */
/* */
/*****************************************************************************/
/*****************************************************************************/
/* File Includes */
/*****************************************************************************/
/* System include files */
#include <stdio.h>
/* User include files */
#include "ih264_typedefs.h"
#include "ih264_platform_macros.h"
#include "ih264_deblk_edge_filters.h"
#include "ih264_macros.h"
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
#ifdef __GNUC__
#define ATTRIBUTE_SSSE3 __attribute__((target("ssse3")))
#else
#define ATTRIBUTE_SSSE3
#endif
Add all the files
2022-08-22 22:21:23 +02:00
/*****************************************************************************/
/* Function Definitions */
/*****************************************************************************/
/*****************************************************************************/
/* */
/* Function Name : ih264_deblk_chroma_vert_bs4_ssse3() */
/* */
/* Description : This function performs filtering of a chroma block */
/* vertical edge when the boundary strength is set to 4 in */
/* high profile. */
/* */
/* Inputs : pu1_src - pointer to the src sample q0 of U */
/* src_strd - source stride */
/* alpha_cb - alpha value for the boundary in U */
/* beta_cb - beta value for the boundary in U */
/* alpha_cr - alpha value for the boundary in V */
/* beta_cr - beta value for the boundary in V */
/* */
/* Globals : None */
/* */
/* Processing : This operation is described in Sec. 8.7.2.4 under the */
/* title "Filtering process for edges for bS equal to 4" in */
/* ITU T Rec H.264 with alpha and beta values different in */
/* U and V. */
/* */
/* Outputs : None */
/* */
/* Returns : None */
/* */
/* Issues : None */
/* */
/* Revision History: */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Initial version */
/* */
/*****************************************************************************/
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
ATTRIBUTE_SSSE3
Add all the files
2022-08-22 22:21:23 +02:00
void ih264_deblk_chroma_vert_bs4_ssse3(UWORD8 *pu1_src,
WORD32 src_strd,
WORD32 alpha_cb,
WORD32 beta_cb,
WORD32 alpha_cr,
WORD32 beta_cr)
{
UWORD8 *pu1_src_uv = pu1_src; /* Pointer to the src sample q0 of plane U*/
WORD32 alpha_cbcr = (alpha_cr << 16) + alpha_cb;
WORD32 beta_cbcr = (beta_cr << 16) + beta_cb;
__m128i linea, lineb, linec, lined, linee, linef, lineg, lineh;
__m128i temp1, temp2, temp3, temp4;
__m128i q0_uv_16x8, p0_uv_16x8, q1_uv_16x8, p1_uv_16x8;
__m128i q0_uv_8x16, p0_uv_8x16, q1_uv_8x16, p1_uv_8x16;
__m128i flag1, flag2;
__m128i diff, alpha_cbcr_16x8, beta_cbcr_16x8;
__m128i zero = _mm_setzero_si128();
__m128i p0_uv_8x16_1, p0_uv_8x16_2, q0_uv_8x16_1, q0_uv_8x16_2;
/* Load and transpose the pixel values */
linea = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4));
lineb = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + src_strd));
linec = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd));
lined = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd));
linee = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 4 * src_strd));
linef = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 5 * src_strd));
lineg = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 6 * src_strd));
lineh = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 7 * src_strd));
temp1 = _mm_unpacklo_epi16(linea, lineb);
temp2 = _mm_unpacklo_epi16(linec, lined);
temp3 = _mm_unpacklo_epi16(linee, linef);
temp4 = _mm_unpacklo_epi16(lineg, lineh);
p1_uv_8x16 = _mm_unpacklo_epi32(temp1, temp2);
p0_uv_8x16 = _mm_unpacklo_epi32(temp3, temp4);
q0_uv_8x16 = _mm_unpackhi_epi32(temp1, temp2);
q1_uv_8x16 = _mm_unpackhi_epi32(temp3, temp4);
p1_uv_16x8 = _mm_unpacklo_epi64(p1_uv_8x16, p0_uv_8x16);
p0_uv_16x8 = _mm_unpackhi_epi64(p1_uv_8x16, p0_uv_8x16);
q0_uv_16x8 = _mm_unpacklo_epi64(q0_uv_8x16, q1_uv_8x16);
q1_uv_16x8 = _mm_unpackhi_epi64(q0_uv_8x16, q1_uv_8x16);
/* End of transpose */
q0_uv_8x16 = _mm_unpacklo_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpacklo_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpacklo_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpacklo_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag1 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
temp1 = _mm_slli_epi16(p1_uv_8x16, 1);
temp2 = _mm_add_epi16(p0_uv_8x16, q1_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
p0_uv_8x16_1 = _mm_srai_epi16(temp1, 2);
temp1 = _mm_slli_epi16(q1_uv_8x16, 1);
temp2 = _mm_add_epi16(p1_uv_8x16, q0_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
q0_uv_8x16_1 = _mm_srai_epi16(temp1, 2);
q0_uv_8x16 = _mm_unpackhi_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpackhi_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpackhi_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpackhi_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag2 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
temp1 = _mm_slli_epi16(p1_uv_8x16, 1);
temp2 = _mm_add_epi16(p0_uv_8x16, q1_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
p0_uv_8x16_2 = _mm_srai_epi16(temp1, 2);
temp1 = _mm_slli_epi16(q1_uv_8x16, 1);
temp2 = _mm_add_epi16(p1_uv_8x16, q0_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
q0_uv_8x16_2 = _mm_srai_epi16(temp1, 2);
p0_uv_8x16_2 = _mm_packus_epi16(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_2 = _mm_packus_epi16(q0_uv_8x16_1, q0_uv_8x16_2);
flag1 = _mm_packs_epi16(flag1, flag2);
p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1);
p0_uv_16x8 = _mm_add_epi8(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1);
q0_uv_16x8 = _mm_add_epi8(q0_uv_8x16_1, q0_uv_8x16_2);
/* Inverse-transpose and store back */
temp1 = _mm_unpacklo_epi16(p1_uv_16x8, p0_uv_16x8);
temp2 = _mm_unpackhi_epi16(p1_uv_16x8, p0_uv_16x8);
temp3 = _mm_unpacklo_epi16(q0_uv_16x8, q1_uv_16x8);
temp4 = _mm_unpackhi_epi16(q0_uv_16x8, q1_uv_16x8);
linea = _mm_unpacklo_epi32(temp1, temp3);
lineb = _mm_srli_si128(linea, 8);
linec = _mm_unpackhi_epi32(temp1, temp3);
lined = _mm_srli_si128(linec, 8);
linee = _mm_unpacklo_epi32(temp2, temp4);
linef = _mm_srli_si128(linee, 8);
lineg = _mm_unpackhi_epi32(temp2, temp4);
lineh = _mm_srli_si128(lineg, 8);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4), linea);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + src_strd), lineb);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd), linec);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd), lined);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 4 * src_strd), linee);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 5 * src_strd), linef);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 6 * src_strd), lineg);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 7 * src_strd), lineh);
}
/*****************************************************************************/
/* */
/* Function Name : ih264_deblk_chroma_horz_bs4_ssse3() */
/* */
/* Description : This function performs filtering of a chroma block */
/* horizontal edge when the boundary strength is set to 4 */
/* in high profile. */
/* */
/* Inputs : pu1_src - pointer to the src sample q0 of U */
/* src_strd - source stride */
/* alpha_cb - alpha value for the boundary in U */
/* beta_cb - beta value for the boundary in U */
/* alpha_cr - alpha value for the boundary in V */
/* beta_cr - beta value for the boundary in V */
/* */
/* Globals : None */
/* */
/* Processing : This operation is described in Sec. 8.7.2.4 under the */
/* title "Filtering process for edges for bS equal to 4" in */
/* ITU T Rec H.264 with alpha and beta values different in */
/* U and V. */
/* */
/* Outputs : None */
/* */
/* Returns : None */
/* */
/* Issues : None */
/* */
/* Revision History: */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Initial version */
/* */
/*****************************************************************************/
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
ATTRIBUTE_SSSE3
Add all the files
2022-08-22 22:21:23 +02:00
void ih264_deblk_chroma_horz_bs4_ssse3(UWORD8 *pu1_src,
WORD32 src_strd,
WORD32 alpha_cb,
WORD32 beta_cb,
WORD32 alpha_cr,
WORD32 beta_cr)
{
UWORD8 *pu1_src_uv = pu1_src; /* Pointer to the src sample q0 of plane U*/
WORD16 i16_posP1, i16_posP0, i16_posQ1;
UWORD8 *pu1_HorzPixelUV; /*! < Pointer to the first pixel of the boundary */
WORD32 alpha_cbcr = (alpha_cr << 16) + alpha_cb;
WORD32 beta_cbcr = (beta_cr << 16) + beta_cb;
__m128i q0_uv_16x8, p0_uv_16x8, q1_uv_16x8, p1_uv_16x8;
__m128i q0_uv_8x16, p0_uv_8x16, q1_uv_8x16, p1_uv_8x16;
__m128i flag1, flag2;
__m128i diff, alpha_cbcr_16x8, beta_cbcr_16x8;
__m128i zero = _mm_setzero_si128();
__m128i p0_uv_8x16_1, p0_uv_8x16_2, q0_uv_8x16_1, q0_uv_8x16_2;
__m128i temp1, temp2;
pu1_HorzPixelUV = pu1_src_uv - (src_strd << 1);
i16_posQ1 = src_strd;
i16_posP0 = src_strd;
i16_posP1 = 0;
q0_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_src_uv));
q1_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_src_uv + i16_posQ1));
p1_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_HorzPixelUV + i16_posP1));
p0_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_HorzPixelUV + i16_posP0));
q0_uv_8x16 = _mm_unpacklo_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpacklo_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpacklo_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpacklo_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag1 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
temp1 = _mm_slli_epi16(p1_uv_8x16, 1);
temp2 = _mm_add_epi16(p0_uv_8x16, q1_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
p0_uv_8x16_1 = _mm_srai_epi16(temp1, 2);
temp1 = _mm_slli_epi16(q1_uv_8x16, 1);
temp2 = _mm_add_epi16(p1_uv_8x16, q0_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
q0_uv_8x16_1 = _mm_srai_epi16(temp1, 2);
q0_uv_8x16 = _mm_unpackhi_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpackhi_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpackhi_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpackhi_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag2 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
temp1 = _mm_slli_epi16(p1_uv_8x16, 1);
temp2 = _mm_add_epi16(p0_uv_8x16, q1_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
p0_uv_8x16_2 = _mm_srai_epi16(temp1, 2);
temp1 = _mm_slli_epi16(q1_uv_8x16, 1);
temp2 = _mm_add_epi16(p1_uv_8x16, q0_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
q0_uv_8x16_2 = _mm_srai_epi16(temp1, 2);
p0_uv_8x16_2 = _mm_packus_epi16(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_2 = _mm_packus_epi16(q0_uv_8x16_1, q0_uv_8x16_2);
flag1 = _mm_packs_epi16(flag1, flag2);
p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1);
p0_uv_8x16_1 = _mm_add_epi8(p0_uv_8x16_1, p0_uv_8x16_2);
_mm_storeu_si128((__m128i *)(pu1_HorzPixelUV + i16_posP0), p0_uv_8x16_1);
q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1);
q0_uv_8x16_1 = _mm_add_epi8(q0_uv_8x16_1, q0_uv_8x16_2);
_mm_storeu_si128((__m128i *)(pu1_src_uv), q0_uv_8x16_1);
}
/*****************************************************************************/
/* */
/* Function Name : ih264_deblk_chroma_vert_bslt4_ssse3() */
/* */
/* Description : This function performs filtering of a chroma block */
/* vertical edge when the boundary strength is less than 4 */
/* in high profile. */
/* */
/* Inputs : pu1_src - pointer to the src sample q0 of U */
/* src_strd - source stride */
/* alpha_cb - alpha value for the boundary in U */
/* beta_cb - beta value for the boundary in U */
/* alpha_cr - alpha value for the boundary in V */
/* beta_cr - beta value for the boundary in V */
/* u4_bs - packed Boundary strength array */
/* pu1_cliptab_cb - tc0_table for U */
/* pu1_cliptab_cr - tc0_table for V */
/* */
/* Globals : None */
/* */
/* Processing : This operation is described in Sec. 8.7.2.3 under the */
/* title "Filtering process for edges for bS less than 4" */
/* in ITU T Rec H.264 with alpha and beta values different */
/* in U and V. */
/* */
/* Outputs : None */
/* */
/* Returns : None */
/* */
/* Issues : None */
/* */
/* Revision History: */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Initial version */
/* */
/*****************************************************************************/
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
ATTRIBUTE_SSSE3
Add all the files
2022-08-22 22:21:23 +02:00
void ih264_deblk_chroma_vert_bslt4_ssse3(UWORD8 *pu1_src,
WORD32 src_strd,
WORD32 alpha_cb,
WORD32 beta_cb,
WORD32 alpha_cr,
WORD32 beta_cr,
UWORD32 u4_bs,
const UWORD8 *pu1_cliptab_cb,
const UWORD8 *pu1_cliptab_cr)
{
UWORD8 *pu1_src_uv = pu1_src; /* Pointer to the src sample q0 of plane U*/
UWORD8 u1_Bs0, u1_Bs1, u1_Bs2, u1_Bs3;
WORD32 alpha_cbcr = (alpha_cr << 16) + alpha_cb;
WORD32 beta_cbcr = (beta_cr << 16) + beta_cb;
__m128i linea, lineb, linec, lined, linee, linef, lineg, lineh;
__m128i temp1, temp2, temp3, temp4;
__m128i q0_uv_16x8, p0_uv_16x8, q1_uv_16x8, p1_uv_16x8;
__m128i q0_uv_8x16, p0_uv_8x16, q1_uv_8x16, p1_uv_8x16;
__m128i flag_bs, flag1, flag2;
__m128i diff, diff1, alpha_cbcr_16x8, beta_cbcr_16x8, in_macro;
__m128i zero = _mm_setzero_si128();
__m128i C0_uv_8x16;
__m128i p0_uv_8x16_1, p0_uv_8x16_2, q0_uv_8x16_1, q0_uv_8x16_2;
u1_Bs0 = (u4_bs >> 24) & 0xff;
u1_Bs1 = (u4_bs >> 16) & 0xff;
u1_Bs2 = (u4_bs >> 8) & 0xff;
u1_Bs3 = (u4_bs >> 0) & 0xff;
flag_bs = _mm_set_epi8(u1_Bs3, u1_Bs3, u1_Bs3, u1_Bs3, u1_Bs2, u1_Bs2,
u1_Bs2, u1_Bs2, u1_Bs1, u1_Bs1, u1_Bs1, u1_Bs1,
u1_Bs0, u1_Bs0, u1_Bs0, u1_Bs0);
flag_bs = _mm_cmpeq_epi8(flag_bs, zero); //Set flag to 1s and 0s
flag_bs = _mm_xor_si128(flag_bs, _mm_set1_epi8(0xFF)); //Invert for required mask
/* Load and transpose the pixel values */
linea = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4));
lineb = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + src_strd));
linec = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd));
lined = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd));
linee = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 4 * src_strd));
linef = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 5 * src_strd));
lineg = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 6 * src_strd));
lineh = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 7 * src_strd));
temp1 = _mm_unpacklo_epi16(linea, lineb);
temp2 = _mm_unpacklo_epi16(linec, lined);
temp3 = _mm_unpacklo_epi16(linee, linef);
temp4 = _mm_unpacklo_epi16(lineg, lineh);
p1_uv_8x16 = _mm_unpacklo_epi32(temp1, temp2);
p0_uv_8x16 = _mm_unpacklo_epi32(temp3, temp4);
q0_uv_8x16 = _mm_unpackhi_epi32(temp1, temp2);
q1_uv_8x16 = _mm_unpackhi_epi32(temp3, temp4);
p1_uv_16x8 = _mm_unpacklo_epi64(p1_uv_8x16, p0_uv_8x16);
p0_uv_16x8 = _mm_unpackhi_epi64(p1_uv_8x16, p0_uv_8x16);
q0_uv_16x8 = _mm_unpacklo_epi64(q0_uv_8x16, q1_uv_8x16);
q1_uv_16x8 = _mm_unpackhi_epi64(q0_uv_8x16, q1_uv_8x16);
/* End of transpose */
q0_uv_8x16 = _mm_unpacklo_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpacklo_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpacklo_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpacklo_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag1 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(q0_uv_8x16, p0_uv_8x16);
diff = _mm_slli_epi16(diff, 2);
diff1 = _mm_subs_epi16(p1_uv_8x16, q1_uv_8x16);
diff = _mm_add_epi16(diff, diff1);
diff = _mm_add_epi16(diff, _mm_set1_epi16(4));
in_macro = _mm_srai_epi16(diff, 3);
C0_uv_8x16 = _mm_set_epi16(pu1_cliptab_cr[u1_Bs1], pu1_cliptab_cb[u1_Bs1],
pu1_cliptab_cr[u1_Bs1], pu1_cliptab_cb[u1_Bs1],
pu1_cliptab_cr[u1_Bs0], pu1_cliptab_cb[u1_Bs0],
pu1_cliptab_cr[u1_Bs0], pu1_cliptab_cb[u1_Bs0]);
C0_uv_8x16 = _mm_add_epi16(C0_uv_8x16, _mm_set1_epi16(1));
in_macro = _mm_min_epi16(C0_uv_8x16, in_macro); //CLIP3
C0_uv_8x16 = _mm_subs_epi16(zero, C0_uv_8x16);
in_macro = _mm_max_epi16(C0_uv_8x16, in_macro);
p0_uv_8x16_1 = _mm_add_epi16(p0_uv_8x16, in_macro);
q0_uv_8x16_1 = _mm_sub_epi16(q0_uv_8x16, in_macro);
q0_uv_8x16 = _mm_unpackhi_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpackhi_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpackhi_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpackhi_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag2 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(q0_uv_8x16, p0_uv_8x16);
diff = _mm_slli_epi16(diff, 2);
diff1 = _mm_subs_epi16(p1_uv_8x16, q1_uv_8x16);
diff = _mm_add_epi16(diff, diff1);
diff = _mm_add_epi16(diff, _mm_set1_epi16(4));
in_macro = _mm_srai_epi16(diff, 3);
C0_uv_8x16 = _mm_set_epi16(pu1_cliptab_cr[u1_Bs3], pu1_cliptab_cb[u1_Bs3],
pu1_cliptab_cr[u1_Bs3], pu1_cliptab_cb[u1_Bs3],
pu1_cliptab_cr[u1_Bs2], pu1_cliptab_cb[u1_Bs2],
pu1_cliptab_cr[u1_Bs2], pu1_cliptab_cb[u1_Bs2]);
C0_uv_8x16 = _mm_add_epi16(C0_uv_8x16, _mm_set1_epi16(1));
in_macro = _mm_min_epi16(C0_uv_8x16, in_macro); //CLIP3
C0_uv_8x16 = _mm_subs_epi16(zero, C0_uv_8x16);
in_macro = _mm_max_epi16(C0_uv_8x16, in_macro);
p0_uv_8x16_2 = _mm_add_epi16(p0_uv_8x16, in_macro);
q0_uv_8x16_2 = _mm_sub_epi16(q0_uv_8x16, in_macro);
p0_uv_8x16_2 = _mm_packus_epi16(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_2 = _mm_packus_epi16(q0_uv_8x16_1, q0_uv_8x16_2);
flag1 = _mm_packs_epi16(flag1, flag2);
flag1 = _mm_and_si128(flag1, flag_bs); //Final flag (BS condition + other 3 conditions)
p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1);
p0_uv_16x8 = _mm_add_epi8(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1);
q0_uv_16x8 = _mm_add_epi8(q0_uv_8x16_1, q0_uv_8x16_2);
/* Inverse-transpose and store back */
temp1 = _mm_unpacklo_epi16(p1_uv_16x8, p0_uv_16x8);
temp2 = _mm_unpackhi_epi16(p1_uv_16x8, p0_uv_16x8);
temp3 = _mm_unpacklo_epi16(q0_uv_16x8, q1_uv_16x8);
temp4 = _mm_unpackhi_epi16(q0_uv_16x8, q1_uv_16x8);
linea = _mm_unpacklo_epi32(temp1, temp3);
lineb = _mm_srli_si128(linea, 8);
linec = _mm_unpackhi_epi32(temp1, temp3);
lined = _mm_srli_si128(linec, 8);
linee = _mm_unpacklo_epi32(temp2, temp4);
linef = _mm_srli_si128(linee, 8);
lineg = _mm_unpackhi_epi32(temp2, temp4);
lineh = _mm_srli_si128(lineg, 8);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4), linea);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + src_strd), lineb);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd), linec);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd), lined);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 4 * src_strd), linee);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 5 * src_strd), linef);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 6 * src_strd), lineg);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 7 * src_strd), lineh);
}
/*****************************************************************************/
/* */
/* Function Name : ih264_deblk_chroma_horz_bslt4_ssse3() */
/* */
/* Description : This function performs filtering of a chroma block */
/* horizontal edge when the boundary strength is less than */
/* 4 in high profile. */
/* */
/* Inputs : pu1_src - pointer to the src sample q0 of U */
/* src_strd - source stride */
/* alpha_cb - alpha value for the boundary in U */
/* beta_cb - beta value for the boundary in U */
/* alpha_cr - alpha value for the boundary in V */
/* beta_cr - beta value for the boundary in V */
/* u4_bs - packed Boundary strength array */
/* pu1_cliptab_cb - tc0_table for U */
/* pu1_cliptab_cr - tc0_table for V */
/* */
/* Globals : None */
/* */
/* Processing : This operation is described in Sec. 8.7.2.3 under the */
/* title "Filtering process for edges for bS less than 4" */
/* in ITU T Rec H.264 with alpha and beta values different */
/* in U and V. */
/* */
/* Outputs : None */
/* */
/* Returns : None */
/* */
/* Issues : None */
/* */
/* Revision History: */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Initial version */
/* */
/*****************************************************************************/
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
ATTRIBUTE_SSSE3
Add all the files
2022-08-22 22:21:23 +02:00
void ih264_deblk_chroma_horz_bslt4_ssse3(UWORD8 *pu1_src,
WORD32 src_strd,
WORD32 alpha_cb,
WORD32 beta_cb,
WORD32 alpha_cr,
WORD32 beta_cr,
UWORD32 u4_bs,
const UWORD8 *pu1_cliptab_cb,
const UWORD8 *pu1_cliptab_cr)
{
UWORD8 *pu1_src_uv = pu1_src; /* Pointer to the src sample q0 of plane U*/
WORD16 i16_posP1, i16_posP0, i16_posQ1;
UWORD8 u1_Bs0, u1_Bs1, u1_Bs2, u1_Bs3;
UWORD8 *pu1_HorzPixelUV; /*! < Pointer to the first pixel of the boundary */
WORD32 alpha_cbcr = (alpha_cr << 16) + alpha_cb;
WORD32 beta_cbcr = (beta_cr << 16) + beta_cb;
__m128i q0_uv_16x8, p0_uv_16x8, q1_uv_16x8, p1_uv_16x8;
__m128i q0_uv_8x16, p0_uv_8x16, q1_uv_8x16, p1_uv_8x16;
__m128i flag_bs, flag1, flag2;
__m128i diff, diff1, alpha_cbcr_16x8, beta_cbcr_16x8, in_macro;
__m128i zero = _mm_setzero_si128();
__m128i C0_uv_8x16;
__m128i p0_uv_8x16_1, p0_uv_8x16_2, q0_uv_8x16_1, q0_uv_8x16_2;
pu1_HorzPixelUV = pu1_src_uv - (src_strd << 1);
i16_posQ1 = src_strd;
i16_posP0 = src_strd;
i16_posP1 = 0;
u1_Bs0 = (u4_bs >> 24) & 0xff;
u1_Bs1 = (u4_bs >> 16) & 0xff;
u1_Bs2 = (u4_bs >> 8) & 0xff;
u1_Bs3 = (u4_bs >> 0) & 0xff;
flag_bs = _mm_set_epi8(u1_Bs3, u1_Bs3, u1_Bs3, u1_Bs3, u1_Bs2, u1_Bs2,
u1_Bs2, u1_Bs2, u1_Bs1, u1_Bs1, u1_Bs1, u1_Bs1,
u1_Bs0, u1_Bs0, u1_Bs0, u1_Bs0);
flag_bs = _mm_cmpeq_epi8(flag_bs, zero); //Set flag to 1s and 0s
flag_bs = _mm_xor_si128(flag_bs, _mm_set1_epi8(0xFF)); //Invert for required mask
q0_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_src_uv));
q1_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_src_uv + i16_posQ1));
p1_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_HorzPixelUV + i16_posP1));
p0_uv_16x8 = _mm_loadu_si128((__m128i *)(pu1_HorzPixelUV + i16_posP0));
q0_uv_8x16 = _mm_unpacklo_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpacklo_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpacklo_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpacklo_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag1 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(q0_uv_8x16, p0_uv_8x16);
diff = _mm_slli_epi16(diff, 2);
diff1 = _mm_subs_epi16(p1_uv_8x16, q1_uv_8x16);
diff = _mm_add_epi16(diff, diff1);
diff = _mm_add_epi16(diff, _mm_set1_epi16(4));
in_macro = _mm_srai_epi16(diff, 3);
C0_uv_8x16 = _mm_set_epi16(pu1_cliptab_cr[u1_Bs1], pu1_cliptab_cb[u1_Bs1],
pu1_cliptab_cr[u1_Bs1], pu1_cliptab_cb[u1_Bs1],
pu1_cliptab_cr[u1_Bs0], pu1_cliptab_cb[u1_Bs0],
pu1_cliptab_cr[u1_Bs0], pu1_cliptab_cb[u1_Bs0]);
C0_uv_8x16 = _mm_add_epi16(C0_uv_8x16, _mm_set1_epi16(1));
in_macro = _mm_min_epi16(C0_uv_8x16, in_macro); //CLIP3
C0_uv_8x16 = _mm_subs_epi16(zero, C0_uv_8x16);
in_macro = _mm_max_epi16(C0_uv_8x16, in_macro);
p0_uv_8x16_1 = _mm_add_epi16(p0_uv_8x16, in_macro);
q0_uv_8x16_1 = _mm_sub_epi16(q0_uv_8x16, in_macro);
q0_uv_8x16 = _mm_unpackhi_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpackhi_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpackhi_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpackhi_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag2 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(q0_uv_8x16, p0_uv_8x16);
diff = _mm_slli_epi16(diff, 2);
diff1 = _mm_subs_epi16(p1_uv_8x16, q1_uv_8x16);
diff = _mm_add_epi16(diff, diff1);
diff = _mm_add_epi16(diff, _mm_set1_epi16(4));
in_macro = _mm_srai_epi16(diff, 3);
C0_uv_8x16 = _mm_set_epi16(pu1_cliptab_cr[u1_Bs3], pu1_cliptab_cb[u1_Bs3],
pu1_cliptab_cr[u1_Bs3], pu1_cliptab_cb[u1_Bs3],
pu1_cliptab_cr[u1_Bs2], pu1_cliptab_cb[u1_Bs2],
pu1_cliptab_cr[u1_Bs2], pu1_cliptab_cb[u1_Bs2]);
C0_uv_8x16 = _mm_add_epi16(C0_uv_8x16, _mm_set1_epi16(1));
in_macro = _mm_min_epi16(C0_uv_8x16, in_macro); //CLIP3
C0_uv_8x16 = _mm_subs_epi16(zero, C0_uv_8x16);
in_macro = _mm_max_epi16(C0_uv_8x16, in_macro);
p0_uv_8x16_2 = _mm_add_epi16(p0_uv_8x16, in_macro);
q0_uv_8x16_2 = _mm_sub_epi16(q0_uv_8x16, in_macro);
p0_uv_8x16_2 = _mm_packus_epi16(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_2 = _mm_packus_epi16(q0_uv_8x16_1, q0_uv_8x16_2);
flag1 = _mm_packs_epi16(flag1, flag2);
flag1 = _mm_and_si128(flag1, flag_bs); //Final flag (BS condition + other 3 conditions)
p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1);
p0_uv_8x16_1 = _mm_add_epi8(p0_uv_8x16_1, p0_uv_8x16_2);
_mm_storeu_si128((__m128i *)(pu1_HorzPixelUV + i16_posP0), p0_uv_8x16_1);
q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1);
q0_uv_8x16_1 = _mm_add_epi8(q0_uv_8x16_1, q0_uv_8x16_2);
_mm_storeu_si128((__m128i *)(pu1_src_uv), q0_uv_8x16_1);
}
/*****************************************************************************/
/* */
/* Function Name : ih264_deblk_chroma_vert_bs4_mbaff_ssse3() */
/* */
/* Description : This function performs filtering of a chroma block */
/* vertical edge when boundary strength is set to 4 in high */
/* profile. */
/* */
/* Inputs : pu1_src - pointer to the src sample q0 of U */
/* src_strd - source stride */
/* alpha_cb - alpha value for the boundary in U */
/* beta_cb - beta value for the boundary in U */
/* alpha_cr - alpha value for the boundary in V */
/* beta_cr - beta value for the boundary in V */
/* u4_bs - packed Boundary strength array */
/* pu1_cliptab_cb - tc0_table for U */
/* pu1_cliptab_cr - tc0_table for V */
/* */
/* Globals : None */
/* */
/* Processing : When the function is called twice, this operation is as */
/* described in Sec. 8.7.2.4 under the title "Filtering */
/* process for edges for bS equal to 4" in ITU T Rec H.264 */
/* with alpha and beta values different in U and V. */
/* */
/* Outputs : None */
/* */
/* Returns : None */
/* */
/* Issues : None */
/* */
/* Revision History: */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Initial version */
/* */
/*****************************************************************************/
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
ATTRIBUTE_SSSE3
Add all the files
2022-08-22 22:21:23 +02:00
void ih264_deblk_chroma_vert_bs4_mbaff_ssse3(UWORD8 *pu1_src,
WORD32 src_strd,
WORD32 alpha_cb,
WORD32 beta_cb,
WORD32 alpha_cr,
WORD32 beta_cr)
{
UWORD8 *pu1_src_uv = pu1_src; /* Pointer to the src sample q0 of plane U*/
WORD32 alpha_cbcr = (alpha_cr << 16) + alpha_cb;
WORD32 beta_cbcr = (beta_cr << 16) + beta_cb;
__m128i linea, lineb, linec, lined;
__m128i temp1, temp2;
__m128i q0_uv_16x8, p0_uv_16x8, q1_uv_16x8, p1_uv_16x8;
__m128i q0_uv_8x16, p0_uv_8x16, q1_uv_8x16, p1_uv_8x16;
__m128i flag1;
__m128i diff, alpha_cbcr_16x8, beta_cbcr_16x8;
__m128i zero = _mm_setzero_si128();
__m128i p0_uv_8x16_1, p0_uv_8x16_2, q0_uv_8x16_1, q0_uv_8x16_2;
/* Load and transpose the pixel values */
linea = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4));
lineb = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + src_strd));
linec = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd));
lined = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd));
temp1 = _mm_unpacklo_epi16(linea, lineb);
temp2 = _mm_unpacklo_epi16(linec, lined);
p1_uv_16x8 = _mm_unpacklo_epi32(temp1, temp2);
p0_uv_16x8 = _mm_srli_si128(p1_uv_16x8, 8);
q0_uv_16x8 = _mm_unpackhi_epi32(temp1, temp2);
q1_uv_16x8 = _mm_srli_si128(q0_uv_16x8, 8);
/* End of transpose */
q0_uv_8x16 = _mm_unpacklo_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpacklo_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpacklo_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpacklo_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag1 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
temp1 = _mm_slli_epi16(p1_uv_8x16, 1);
temp2 = _mm_add_epi16(p0_uv_8x16, q1_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
p0_uv_8x16_1 = _mm_srai_epi16(temp1, 2);
temp1 = _mm_slli_epi16(q1_uv_8x16, 1);
temp2 = _mm_add_epi16(p1_uv_8x16, q0_uv_8x16);
temp1 = _mm_add_epi16(temp1, _mm_set1_epi16(2));
temp1 = _mm_add_epi16(temp1, temp2);
q0_uv_8x16_1 = _mm_srai_epi16(temp1, 2);
p0_uv_8x16_2 = _mm_packus_epi16(p0_uv_8x16_1, p0_uv_8x16_1);
q0_uv_8x16_2 = _mm_packus_epi16(q0_uv_8x16_1, q0_uv_8x16_1);
flag1 = _mm_packs_epi16(flag1, flag1);
p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1);
p0_uv_16x8 = _mm_add_epi8(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1);
q0_uv_16x8 = _mm_add_epi8(q0_uv_8x16_1, q0_uv_8x16_2);
/* Inverse-transpose and store back */
temp1 = _mm_unpacklo_epi16(p1_uv_16x8, p0_uv_16x8);
temp2 = _mm_unpacklo_epi16(q0_uv_16x8, q1_uv_16x8);
linea = _mm_unpacklo_epi32(temp1, temp2);
lineb = _mm_srli_si128(linea, 8);
linec = _mm_unpackhi_epi32(temp1, temp2);
lined = _mm_srli_si128(linec, 8);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4), linea);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + src_strd), lineb);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd), linec);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd), lined);
}
/*****************************************************************************/
/* */
/* Function Name : ih264_deblk_chroma_vert_bslt4_mbaff_ssse3() */
/* */
/* Description : This function performs filtering of a chroma block */
/* vertical edge when boundary strength is less than 4 in */
/* high profile. */
/* */
/* Inputs : pu1_src - pointer to the src sample q0 of U */
/* src_strd - source stride */
/* alpha_cb - alpha value for the boundary in U */
/* beta_cb - beta value for the boundary in U */
/* alpha_cr - alpha value for the boundary in V */
/* beta_cr - beta value for the boundary in V */
/* u4_bs - packed Boundary strength array */
/* pu1_cliptab_cb - tc0_table for U */
/* pu1_cliptab_cr - tc0_table for V */
/* */
/* Globals : None */
/* */
/* Processing : When the function is called twice, this operation is as */
/* described in Sec. 8.7.2.4 under the title "Filtering */
/* process for edges for bS less than 4" in ITU T Rec H.264 */
/* with alpha and beta values different in U and V. */
/* */
/* Outputs : None */
/* */
/* Returns : None */
/* */
/* Issues : None */
/* */
/* Revision History: */
/* */
/* DD MM YYYY Author(s) Changes (Describe the changes made) */
/* 12 02 2015 Naveen Kumar P Initial version */
/* */
/*****************************************************************************/
[ih264] per-function target attribute on clang and GCC (#328)
2022-10-04 15:24:14 +02:00
ATTRIBUTE_SSSE3
Add all the files
2022-08-22 22:21:23 +02:00
void ih264_deblk_chroma_vert_bslt4_mbaff_ssse3(UWORD8 *pu1_src,
WORD32 src_strd,
WORD32 alpha_cb,
WORD32 beta_cb,
WORD32 alpha_cr,
WORD32 beta_cr,
UWORD32 u4_bs,
const UWORD8 *pu1_cliptab_cb,
const UWORD8 *pu1_cliptab_cr)
{
UWORD8 *pu1_src_uv = pu1_src; /* Pointer to the src sample q0 of plane U*/
UWORD8 u1_Bs0, u1_Bs1, u1_Bs2, u1_Bs3;
WORD32 alpha_cbcr = (alpha_cr << 16) + alpha_cb;
WORD32 beta_cbcr = (beta_cr << 16) + beta_cb;
__m128i linea, lineb, linec, lined;
__m128i temp1, temp2;
__m128i q0_uv_16x8, p0_uv_16x8, q1_uv_16x8, p1_uv_16x8;
__m128i q0_uv_8x16, p0_uv_8x16, q1_uv_8x16, p1_uv_8x16;
__m128i flag_bs, flag1;
__m128i diff, diff1, alpha_cbcr_16x8, beta_cbcr_16x8, in_macro;
__m128i zero = _mm_setzero_si128();
__m128i C0_uv_8x16;
__m128i p0_uv_8x16_1, p0_uv_8x16_2, q0_uv_8x16_1, q0_uv_8x16_2;
u1_Bs0 = (u4_bs >> 24) & 0xff;
u1_Bs1 = (u4_bs >> 16) & 0xff;
u1_Bs2 = (u4_bs >> 8) & 0xff;
u1_Bs3 = (u4_bs >> 0) & 0xff;
flag_bs = _mm_set_epi8(0, 0, 0, 0, 0, 0, 0, 0, u1_Bs3, u1_Bs3, u1_Bs2,
u1_Bs2, u1_Bs1, u1_Bs1, u1_Bs0, u1_Bs0);
flag_bs = _mm_cmpeq_epi8(flag_bs, zero); //Set flag to 1s and 0s
flag_bs = _mm_xor_si128(flag_bs, _mm_set1_epi8(0xFF)); //Invert for required mask
/* Load and transpose the pixel values */
linea = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4));
lineb = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + src_strd));
linec = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd));
lined = _mm_loadl_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd));
temp1 = _mm_unpacklo_epi16(linea, lineb);
temp2 = _mm_unpacklo_epi16(linec, lined);
p1_uv_16x8 = _mm_unpacklo_epi32(temp1, temp2);
p0_uv_16x8 = _mm_srli_si128(p1_uv_16x8, 8);
q0_uv_16x8 = _mm_unpackhi_epi32(temp1, temp2);
q1_uv_16x8 = _mm_srli_si128(q0_uv_16x8, 8);
/* End of transpose */
q0_uv_8x16 = _mm_unpacklo_epi8(q0_uv_16x8, zero);
q1_uv_8x16 = _mm_unpacklo_epi8(q1_uv_16x8, zero);
p1_uv_8x16 = _mm_unpacklo_epi8(p1_uv_16x8, zero);
p0_uv_8x16 = _mm_unpacklo_epi8(p0_uv_16x8, zero);
diff = _mm_subs_epi16(p0_uv_8x16, q0_uv_8x16); //Condn 1
diff = _mm_abs_epi16(diff);
alpha_cbcr_16x8 = _mm_set1_epi32(alpha_cbcr);
flag1 = _mm_cmpgt_epi16(alpha_cbcr_16x8, diff);
diff = _mm_subs_epi16(q1_uv_8x16, q0_uv_8x16); //Condtn 2
diff = _mm_abs_epi16(diff);
beta_cbcr_16x8 = _mm_set1_epi32(beta_cbcr);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(p1_uv_8x16, p0_uv_8x16); //Condtn 3
diff = _mm_abs_epi16(diff);
flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff));
diff = _mm_subs_epi16(q0_uv_8x16, p0_uv_8x16);
diff = _mm_slli_epi16(diff, 2);
diff1 = _mm_subs_epi16(p1_uv_8x16, q1_uv_8x16);
diff = _mm_add_epi16(diff, diff1);
diff = _mm_add_epi16(diff, _mm_set1_epi16(4));
in_macro = _mm_srai_epi16(diff, 3);
C0_uv_8x16 = _mm_set_epi16(pu1_cliptab_cr[u1_Bs3], pu1_cliptab_cb[u1_Bs3],
pu1_cliptab_cr[u1_Bs2], pu1_cliptab_cb[u1_Bs2],
pu1_cliptab_cr[u1_Bs1], pu1_cliptab_cb[u1_Bs1],
pu1_cliptab_cr[u1_Bs0], pu1_cliptab_cb[u1_Bs0]);
C0_uv_8x16 = _mm_add_epi16(C0_uv_8x16, _mm_set1_epi16(1));
in_macro = _mm_min_epi16(C0_uv_8x16, in_macro); //CLIP3
C0_uv_8x16 = _mm_subs_epi16(zero, C0_uv_8x16);
in_macro = _mm_max_epi16(C0_uv_8x16, in_macro);
p0_uv_8x16_1 = _mm_add_epi16(p0_uv_8x16, in_macro);
q0_uv_8x16_1 = _mm_sub_epi16(q0_uv_8x16, in_macro);
p0_uv_8x16_2 = _mm_packus_epi16(p0_uv_8x16_1, p0_uv_8x16_1);
q0_uv_8x16_2 = _mm_packus_epi16(q0_uv_8x16_1, q0_uv_8x16_1);
flag1 = _mm_packs_epi16(flag1, flag1);
flag1 = _mm_and_si128(flag1, flag_bs); //Final flag (BS condition + other 3 conditions)
p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1);
p0_uv_16x8 = _mm_add_epi8(p0_uv_8x16_1, p0_uv_8x16_2);
q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8,
_mm_xor_si128(flag1, _mm_set1_epi8(0xFF)));
q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1);
q0_uv_16x8 = _mm_add_epi8(q0_uv_8x16_1, q0_uv_8x16_2);
/* Inverse-transpose and store back */
temp1 = _mm_unpacklo_epi16(p1_uv_16x8, p0_uv_16x8);
temp2 = _mm_unpacklo_epi16(q0_uv_16x8, q1_uv_16x8);
linea = _mm_unpacklo_epi32(temp1, temp2);
lineb = _mm_srli_si128(linea, 8);
linec = _mm_unpackhi_epi32(temp1, temp2);
lined = _mm_srli_si128(linec, 8);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4), linea);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + src_strd), lineb);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 2 * src_strd), linec);
_mm_storel_epi64((__m128i *)(pu1_src_uv - 4 + 3 * src_strd), lined);
}
Reference in New Issue Copy Permalink