ref: fb5e508e64f47d9cc8289277134a302853a58a0f
dir: /codec/common/mips/deblock_msa.c/
/*!
* \copy
* Copyright (C) 2019 Loongson Technology Co. Ltd.
* Contributed by Gu Xiwei(guxiwei-hf@loongson.cn)
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
* COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
* ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*
*
* \file deblock_msa.c
*
* \brief MIPS MSA optimizations
*
* \date 15/05/2020 Created
*
*************************************************************************************
*/
#include <stdint.h>
#include "msa_macros.h"
void DeblockLumaLt4V_msa(uint8_t *pPix, int32_t iStride, int32_t iAlpha,
int32_t iBeta, int8_t *pTc) {
v16u8 p0, p1, p2, q0, q1, q2;
v16i8 iTc, negiTc, negTc, flags, f;
v8i16 p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, q0_l, q0_r, q1_l, q1_r, q2_l, q2_r;
v8i16 tc_l, tc_r, negTc_l, negTc_r;
v8i16 iTc_l, iTc_r, negiTc_l, negiTc_r;
// Use for temporary variable
v8i16 t0, t1, t2, t3;
v16u8 alpha, beta;
v16u8 bDetaP0Q0, bDetaP1P0, bDetaQ1Q0, bDetaP2P0, bDetaQ2Q0;
v16i8 const_1_b = __msa_ldi_b(1);
v8i16 const_1_h = __msa_ldi_h(1);
v8i16 const_4_h = __msa_ldi_h(4);
v8i16 const_not_255_h = __msa_ldi_h(~255);
v16i8 zero = { 0 };
v16i8 tc = { pTc[0 >> 2], pTc[1 >> 2], pTc[2 >> 2], pTc[3 >> 2],
pTc[4 >> 2], pTc[5 >> 2], pTc[6 >> 2], pTc[7 >> 2],
pTc[8 >> 2], pTc[9 >> 2], pTc[10 >> 2], pTc[11 >> 2],
pTc[12 >> 2], pTc[13 >> 2], pTc[14 >> 2], pTc[15 >> 2] };
negTc = zero - tc;
iTc = tc;
// Load data from pPix
MSA_LD_V4(v16u8, pPix - 3 * iStride, iStride, p2, p1, p0, q0);
MSA_LD_V2(v16u8, pPix + iStride, iStride, q1, q2);
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP2P0 = __msa_asub_u_b(p2, p0);
bDetaQ2Q0 = __msa_asub_u_b(q2, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
bDetaP2P0 = (v16u8)__msa_clt_u_b(bDetaP2P0, beta);
bDetaQ2Q0 = (v16u8)__msa_clt_u_b(bDetaQ2Q0, beta);
// Unsigned extend p0, p1, p2, q0, q1, q2 from 8 bits to 16 bits
MSA_ILVRL_B4(v8i16, zero, p0, zero, p1,
p0_r, p0_l, p1_r, p1_l);
MSA_ILVRL_B4(v8i16, zero, p2, zero, q0,
p2_r, p2_l, q0_r, q0_l);
MSA_ILVRL_B4(v8i16, zero, q1, zero, q2,
q1_r, q1_l, q2_r, q2_l);
// Signed extend tc, negTc from 8 bits to 16 bits
flags = __msa_clt_s_b(tc, zero);
MSA_ILVRL_B2(v8i16, flags, tc, tc_r, tc_l);
flags = __msa_clt_s_b(negTc, zero);
MSA_ILVRL_B2(v8i16, flags, negTc, negTc_r, negTc_l);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
flags = f & (v16i8)bDetaP2P0;
flags = __msa_ceq_b(flags, zero);
iTc += ((~flags) & const_1_b);
flags = f & (v16i8)bDetaQ2Q0;
flags = __msa_ceq_b(flags, zero);
iTc += ((~flags) & const_1_b);
negiTc = zero - iTc;
// Signed extend iTc, negiTc from 8 bits to 16 bits
flags = __msa_clt_s_b(iTc, zero);
MSA_ILVRL_B2(v8i16, flags, iTc, iTc_r, iTc_l);
flags = __msa_clt_s_b(negiTc, zero);
MSA_ILVRL_B2(v8i16, flags, negiTc, negiTc_r, negiTc_l);
// Calculate the left part
// p1
t0 = (p2_l + ((p0_l + q0_l + const_1_h) >> 1) - (p1_l << 1)) >> 1;
t0 = __msa_max_s_h(negTc_l, t0);
t0 = __msa_min_s_h(tc_l, t0);
t1 = p1_l + t0;
// q1
t0 = (q2_l + ((p0_l + q0_l + const_1_h) >> 1) - (q1_l << 1)) >> 1;
t0 = __msa_max_s_h(negTc_l, t0);
t0 = __msa_min_s_h(tc_l, t0);
t2 = q1_l + t0;
// iDeta
t0 = (((q0_l - p0_l) << 2) + (p1_l - q1_l) + const_4_h) >> 3;
t0 = __msa_max_s_h(negiTc_l, t0);
t0 = __msa_min_s_h(iTc_l, t0);
p1_l = t1;
q1_l = t2;
// p0
t1 = p0_l + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_l = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_l - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_l = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// Calculate the right part
// p1
t0 = (p2_r + ((p0_r + q0_r + const_1_h) >> 1) - (p1_r << 1)) >> 1;
t0 = __msa_max_s_h(negTc_r, t0);
t0 = __msa_min_s_h(tc_r, t0);
t1 = p1_r + t0;
// q1
t0 = (q2_r + ((p0_r + q0_r + const_1_h) >> 1) - (q1_r << 1)) >> 1;
t0 = __msa_max_s_h(negTc_r, t0);
t0 = __msa_min_s_h(tc_r, t0);
t2 = q1_r + t0;
// iDeta
t0 = (((q0_r - p0_r) << 2) + (p1_r - q1_r) + const_4_h) >> 3;
t0 = __msa_max_s_h(negiTc_r, t0);
t0 = __msa_min_s_h(iTc_r, t0);
p1_r = t1;
q1_r = t2;
// p0
t1 = p0_r + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_r = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_r - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_r = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// Combined left and right
MSA_PCKEV_B4(v8i16, p1_l, p1_r, p0_l, p0_r, q0_l, q0_r, q1_l, q1_r,
t0, t1, t2, t3);
flags = (v16i8)__msa_cle_s_b(zero, tc);
flags &= f;
p0 = (v16u8)(((v16i8)t1 & flags) + (p0 & (~flags)));
q0 = (v16u8)(((v16i8)t2 & flags) + (q0 & (~flags)));
// Using t1, t2 as temporary flags
t1 = (v8i16)(flags & (~(__msa_ceq_b((v16i8)bDetaP2P0, zero))));
p1 = (v16u8)(t0 & t1) + (p1 & (v16u8)(~t1));
t2 = (v8i16)(flags & (~(__msa_ceq_b((v16i8)bDetaQ2Q0, zero))));
q1 = (v16u8)(t3 & t2) + (q1 & (v16u8)(~t2));
// Store data to pPix
MSA_ST_V4(v16u8, p1, p0, q0, q1, pPix - 2 * iStride, iStride);
}
void DeblockLumaEq4V_msa(uint8_t *pPix, int32_t iStride, int32_t iAlpha,
int32_t iBeta) {
v16u8 p0, p1, p2, p3, q0, q1, q2, q3;
v8i16 p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, p3_l, p3_r,
q0_l, q0_r, q1_l, q1_r, q2_l, q2_r, q3_l, q3_r;
v8i16 t0, t1, t2, t0_con1;
v8i16 s0, s1, s2, s0_con1;
v16u8 alpha, beta;
v16u8 iDetaP0Q0, bDetaP1P0, bDetaQ1Q0, bDetaP2P0, bDetaQ2Q0;
// Condition mask
v16u8 mask0, mask1;
v16i8 const_2_b = __msa_ldi_b(2);
v8i16 const_2_h = __msa_ldi_h(2);
v8i16 const_4_h = __msa_ldi_h(4);
v16i8 zero = { 0 };
// Load data from pPix
MSA_LD_V8(v16u8, pPix - 4 * iStride, iStride, p3, p2, p1, p0,
q0, q1, q2, q3);
// iAlpha and beta are uint8_t type
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
// iDetaP0Q0 is not bool type
iDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP2P0 = __msa_asub_u_b(p2, p0);
bDetaQ2Q0 = __msa_asub_u_b(q2, q0);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
bDetaP2P0 = (v16u8)__msa_clt_u_b(bDetaP2P0, beta);
bDetaQ2Q0 = (v16u8)__msa_clt_u_b(bDetaQ2Q0, beta);
// Unsigned extend p0, p1, p2, p3, q0, q1, q2, q3 from 8 bits to 16 bits
MSA_ILVRL_B4(v8i16, zero, p0, zero, p1,
p0_r, p0_l, p1_r, p1_l);
MSA_ILVRL_B4(v8i16, zero, p2, zero, p3,
p2_r, p2_l, p3_r, p3_l);
MSA_ILVRL_B4(v8i16, zero, q0, zero, q1,
q0_r, q0_l, q1_r, q1_l);
MSA_ILVRL_B4(v8i16, zero, q2, zero, q3,
q2_r, q2_l, q3_r, q3_l);
// Calculate condition mask
// (iDetaP0Q0 < iAlpha) && bDetaP1P0 && bDetaQ1Q0
mask0 = (v16u8)__msa_clt_u_b(iDetaP0Q0, alpha);
mask0 &= bDetaP1P0;
mask0 &= bDetaQ1Q0;
// iDetaP0Q0 < ((iAlpha >> 2) + 2)
mask1 = (v16u8)((alpha >> 2) + const_2_b);
mask1 = (v16u8)__msa_clt_u_b(iDetaP0Q0, mask1);
// Calculate the left part
// p0
t0 = (p2_l + (p1_l << 1) + (p0_l << 1) + (q0_l << 1) + q1_l + const_4_h) >> 3;
// p1
t1 = (p2_l + p1_l + p0_l + q0_l + const_2_h) >> 2;
// p2
t2 = ((p3_l << 1) + p2_l + (p2_l << 1) + p1_l + p0_l + q0_l + const_4_h) >> 3;
// p0 condition 1
t0_con1 = ((p1_l << 1) + p0_l + q1_l + const_2_h) >> 2;
// q0
s0 = (p1_l + (p0_l << 1) + (q0_l << 1) + (q1_l << 1) + q2_l + const_4_h) >> 3;
// q1
s1 = (p0_l + q0_l + q1_l + q2_l + const_2_h) >> 2;
// q2
s2 = ((q3_l << 1) + q2_l + (q2_l << 1) + q1_l + q0_l + p0_l + const_4_h) >> 3;
// q0 condition 1
s0_con1 = ((q1_l << 1) + q0_l + p1_l + const_2_h) >> 2;
// Move back
p0_l = t0;
p1_l = t1;
p2_l = t2;
q0_l = s0;
q1_l = s1;
q2_l = s2;
// Use p3_l, q3_l as tmp
p3_l = t0_con1;
q3_l = s0_con1;
// Calculate the right part
// p0
t0 = (p2_r + (p1_r << 1) + (p0_r << 1) + (q0_r << 1) + q1_r + const_4_h) >> 3;
// p1
t1 = (p2_r + p1_r + p0_r + q0_r + const_2_h) >> 2;
// p2
t2 = ((p3_r << 1) + p2_r + (p2_r << 1) + p1_r + p0_r + q0_r + const_4_h) >> 3;
// p0 condition 1
t0_con1 = ((p1_r << 1) + p0_r + q1_r + const_2_h) >> 2;
// q0
s0 = (p1_r + (p0_r << 1) + (q0_r << 1) + (q1_r << 1) + q2_r + const_4_h) >> 3;
// q1
s1 = (p0_r + q0_r + q1_r + q2_r + const_2_h) >> 2;
// q2
s2 = ((q3_r << 1) + q2_r + (q2_r << 1) + q1_r + q0_r + p0_r + const_4_h) >> 3;
// q0 condition 1
s0_con1 = ((q1_r << 1) + q0_r + p1_r + const_2_h) >> 2;
// Move back
p0_r = t0;
p1_r = t1;
p2_r = t2;
q0_r = s0;
q1_r = s1;
q2_r = s2;
// Use p3_r, q3_r as tmp
p3_r = t0_con1;
q3_r = s0_con1;
// Combined left and right
MSA_PCKEV_B4(v8i16, p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, q0_l, q0_r,
t0, t1, t2, s0);
MSA_PCKEV_B4(v8i16, q1_l, q1_r, q2_l, q2_r, p3_l, p3_r, q3_l, q3_r,
s1, s2, t0_con1, s0_con1);
t0 = (v8i16)(((v16u8)t0 & mask0 & mask1 & bDetaP2P0) + ((v16u8)t0_con1 &
mask0 & mask1 & (~bDetaP2P0)) + ((v16u8)t0_con1 & mask0 & (~mask1)));
t1 = (v8i16)((v16u8)t1 & mask0 & mask1 & bDetaP2P0);
t2 = (v8i16)((v16u8)t2 & mask0 & mask1 & bDetaP2P0);
s0 = (v8i16)(((v16u8)s0 & mask0 & mask1 & bDetaQ2Q0) + ((v16u8)s0_con1 &
mask0 & mask1 & (~bDetaQ2Q0)) + ((v16u8)s0_con1 & mask0 & (~mask1)));
s1 = (v8i16)((v16u8)s1 & mask0 & mask1 & bDetaQ2Q0);
s2 = (v8i16)((v16u8)s2 & mask0 & mask1 & bDetaQ2Q0);
p0 = (v16u8)t0 + (p0 & (~mask0));
p1 = (v16u8)t1 + (p1 & ~(mask0 & mask1 & bDetaP2P0));
p2 = (v16u8)t2 + (p2 & ~(mask0 & mask1 & bDetaP2P0));
q0 = (v16u8)s0 + (q0 & (~mask0));
q1 = (v16u8)s1 + (q1 & ~(mask0 & mask1 & bDetaQ2Q0));
q2 = (v16u8)s2 + (q2 & ~(mask0 & mask1 & bDetaQ2Q0));
// Store data to pPix
MSA_ST_V4(v16u8, p2, p1, p0, q0, pPix - 3 * iStride, iStride);
MSA_ST_V2(v16u8, q1, q2, pPix + iStride, iStride);
}
void DeblockLumaLt4H_msa(uint8_t* pPix, int32_t iStride, int32_t iAlpha,
int32_t iBeta, int8_t* pTc) {
v16u8 p0, p1, p2, q0, q1, q2;
v16i8 iTc, negiTc, negTc, flags, f;
v8i16 p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, q0_l, q0_r, q1_l, q1_r, q2_l, q2_r;
v8i16 tc_l, tc_r, negTc_l, negTc_r;
v8i16 iTc_l, iTc_r, negiTc_l, negiTc_r;
// Use for temporary variable
v8i16 t0, t1, t2, t3;
v16u8 alpha, beta;
v16u8 bDetaP0Q0, bDetaP1P0, bDetaQ1Q0, bDetaP2P0, bDetaQ2Q0;
v16i8 const_1_b = __msa_ldi_b(1);
v8i16 const_1_h = __msa_ldi_h(1);
v8i16 const_4_h = __msa_ldi_h(4);
v8i16 const_not_255_h = __msa_ldi_h(~255);
v16i8 zero = { 0 };
v16i8 tc = { pTc[0 >> 2], pTc[1 >> 2], pTc[2 >> 2], pTc[3 >> 2],
pTc[4 >> 2], pTc[5 >> 2], pTc[6 >> 2], pTc[7 >> 2],
pTc[8 >> 2], pTc[9 >> 2], pTc[10 >> 2], pTc[11 >> 2],
pTc[12 >> 2], pTc[13 >> 2], pTc[14 >> 2], pTc[15 >> 2] };
negTc = zero - tc;
iTc = tc;
// Load data from pPix
MSA_LD_V8(v8i16, pPix - 3, iStride, t0, t1, t2, t3, q1_l, q1_r, q2_l, q2_r);
MSA_LD_V8(v8i16, pPix + 8 * iStride - 3, iStride, p0_l, p0_r, p1_l, p1_r,
p2_l, p2_r, q0_l, q0_r);
// Transpose 16x8 to 8x16, we just need p0, p1, p2, q0, q1, q2
MSA_TRANSPOSE16x8_B(v16u8, t0, t1, t2, t3, q1_l, q1_r, q2_l, q2_r,
p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, q0_l, q0_r,
p2, p1, p0, q0, q1, q2, alpha, beta);
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP2P0 = __msa_asub_u_b(p2, p0);
bDetaQ2Q0 = __msa_asub_u_b(q2, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
bDetaP2P0 = (v16u8)__msa_clt_u_b(bDetaP2P0, beta);
bDetaQ2Q0 = (v16u8)__msa_clt_u_b(bDetaQ2Q0, beta);
// Unsigned extend p0, p1, p2, q0, q1, q2 from 8 bits to 16 bits
MSA_ILVRL_B4(v8i16, zero, p0, zero, p1,
p0_r, p0_l, p1_r, p1_l);
MSA_ILVRL_B4(v8i16, zero, p2, zero, q0,
p2_r, p2_l, q0_r, q0_l);
MSA_ILVRL_B4(v8i16, zero, q1, zero, q2,
q1_r, q1_l, q2_r, q2_l);
// Signed extend tc, negTc from 8 bits to 16 bits
flags = __msa_clt_s_b(tc, zero);
MSA_ILVRL_B2(v8i16, flags, tc, tc_r, tc_l);
flags = __msa_clt_s_b(negTc, zero);
MSA_ILVRL_B2(v8i16, flags, negTc, negTc_r, negTc_l);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
flags = f & (v16i8)bDetaP2P0;
flags = __msa_ceq_b(flags, zero);
iTc += ((~flags) & const_1_b);
flags = f & (v16i8)bDetaQ2Q0;
flags = __msa_ceq_b(flags, zero);
iTc += ((~flags) & const_1_b);
negiTc = zero - iTc;
// Signed extend iTc, negiTc from 8 bits to 16 bits
flags = __msa_clt_s_b(iTc, zero);
MSA_ILVRL_B2(v8i16, flags, iTc, iTc_r, iTc_l);
flags = __msa_clt_s_b(negiTc, zero);
MSA_ILVRL_B2(v8i16, flags, negiTc, negiTc_r, negiTc_l);
// Calculate the left part
// p1
t0 = (p2_l + ((p0_l + q0_l + const_1_h) >> 1) - (p1_l << 1)) >> 1;
t0 = __msa_max_s_h(negTc_l, t0);
t0 = __msa_min_s_h(tc_l, t0);
t1 = p1_l + t0;
// q1
t0 = (q2_l + ((p0_l + q0_l + const_1_h) >> 1) - (q1_l << 1)) >> 1;
t0 = __msa_max_s_h(negTc_l, t0);
t0 = __msa_min_s_h(tc_l, t0);
t2 = q1_l + t0;
// iDeta
t0 = (((q0_l - p0_l) << 2) + (p1_l - q1_l) + const_4_h) >> 3;
t0 = __msa_max_s_h(negiTc_l, t0);
t0 = __msa_min_s_h(iTc_l, t0);
p1_l = t1;
q1_l = t2;
// p0
t1 = p0_l + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_l = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_l - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_l = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// Calculate the right part
// p1
t0 = (p2_r + ((p0_r + q0_r + const_1_h) >> 1) - (p1_r << 1)) >> 1;
t0 = __msa_max_s_h(negTc_r, t0);
t0 = __msa_min_s_h(tc_r, t0);
t1 = p1_r + t0;
// q1
t0 = (q2_r + ((p0_r + q0_r + const_1_h) >> 1) - (q1_r << 1)) >> 1;
t0 = __msa_max_s_h(negTc_r, t0);
t0 = __msa_min_s_h(tc_r, t0);
t2 = q1_r + t0;
// iDeta
t0 = (((q0_r - p0_r) << 2) + (p1_r - q1_r) + const_4_h) >> 3;
t0 = __msa_max_s_h(negiTc_r, t0);
t0 = __msa_min_s_h(iTc_r, t0);
p1_r = t1;
q1_r = t2;
// p0
t1 = p0_r + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_r = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_r - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_r = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// Combined left and right
MSA_PCKEV_B4(v8i16, p1_l, p1_r, p0_l, p0_r, q0_l, q0_r, q1_l, q1_r,
t0, t1, t2, t3);
flags = (v16i8)__msa_cle_s_b(zero, tc);
flags &= f;
p0 = (v16u8)(((v16i8)t1 & flags) + (p0 & (~flags)));
q0 = (v16u8)(((v16i8)t2 & flags) + (q0 & (~flags)));
// Using t1, t2 as temporary flags
t1 = (v8i16)(flags & (~(__msa_ceq_b((v16i8)bDetaP2P0, zero))));
p1 = (v16u8)(t0 & t1) + (p1 & (v16u8)(~t1));
t2 = (v8i16)(flags & (~(__msa_ceq_b((v16i8)bDetaQ2Q0, zero))));
q1 = (v16u8)(t3 & t2) + (q1 & (v16u8)(~t2));
MSA_ILVRL_B4(v8i16, p0, p1, q1, q0, t0, t1, t2, t3);
MSA_ILVRL_H4(v16u8, t2, t0, t3, t1, p1, p0, q0, q1);
// Store data to pPix
MSA_ST_W8(p1, p0, 0, 1, 2, 3, 0, 1, 2, 3, pPix - 2, iStride);
MSA_ST_W8(q0, q1, 0, 1, 2, 3, 0, 1, 2, 3, pPix + 8 * iStride - 2, iStride);
}
void DeblockLumaEq4H_msa(uint8_t *pPix, int32_t iStride, int32_t iAlpha,
int32_t iBeta) {
v16u8 p0, p1, p2, p3, q0, q1, q2, q3;
v8i16 p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, p3_l, p3_r,
q0_l, q0_r, q1_l, q1_r, q2_l, q2_r, q3_l, q3_r;
v8i16 t0, t1, t2, t0_con1;
v8i16 s0, s1, s2, s0_con1;
v16u8 alpha, beta;
v16u8 iDetaP0Q0, bDetaP1P0, bDetaQ1Q0, bDetaP2P0, bDetaQ2Q0;
// Condition mask
v16u8 mask0, mask1;
v16i8 const_2_b = __msa_ldi_b(2);
v8i16 const_2_h = __msa_ldi_h(2);
v8i16 const_4_h = __msa_ldi_h(4);
v16i8 zero = { 0 };
// Load data from pPix
MSA_LD_V8(v8i16, pPix - 4, iStride, p0_l, p0_r, p1_l, p1_r,
p2_l, p2_r, p3_l, p3_r);
MSA_LD_V8(v8i16, pPix + 8 * iStride - 4, iStride,
q0_l, q0_r, q1_l, q1_r, q2_l, q2_r, q3_l, q3_r);
// Transpose 16x8 to 8x16, we just need p0, p1, p2, p3, q0, q1, q2, q3
MSA_TRANSPOSE16x8_B(v16u8, p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, p3_l, p3_r,
q0_l, q0_r, q1_l, q1_r, q2_l, q2_r, q3_l, q3_r,
p3, p2, p1, p0, q0, q1, q2, q3);
// iAlpha and beta are uint8_t type
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
// iDetaP0Q0 is not bool type
iDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP2P0 = __msa_asub_u_b(p2, p0);
bDetaQ2Q0 = __msa_asub_u_b(q2, q0);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
bDetaP2P0 = (v16u8)__msa_clt_u_b(bDetaP2P0, beta);
bDetaQ2Q0 = (v16u8)__msa_clt_u_b(bDetaQ2Q0, beta);
// Unsigned extend p0, p1, p2, p3, q0, q1, q2, q3 from 8 bits to 16 bits
MSA_ILVRL_B4(v8i16, zero, p0, zero, p1,
p0_r, p0_l, p1_r, p1_l);
MSA_ILVRL_B4(v8i16, zero, p2, zero, p3,
p2_r, p2_l, p3_r, p3_l);
MSA_ILVRL_B4(v8i16, zero, q0, zero, q1,
q0_r, q0_l, q1_r, q1_l);
MSA_ILVRL_B4(v8i16, zero, q2, zero, q3,
q2_r, q2_l, q3_r, q3_l);
// Calculate condition mask
// (iDetaP0Q0 < iAlpha) && bDetaP1P0 && bDetaQ1Q0
mask0 = (v16u8)__msa_clt_u_b(iDetaP0Q0, alpha);
mask0 &= bDetaP1P0;
mask0 &= bDetaQ1Q0;
// iDetaP0Q0 < ((iAlpha >> 2) + 2)
mask1 = (v16u8)((alpha >> 2) + const_2_b);
mask1 = (v16u8)__msa_clt_u_b(iDetaP0Q0, mask1);
// Calculate the left part
// p0
t0 = (p2_l + (p1_l << 1) + (p0_l << 1) + (q0_l << 1) + q1_l + const_4_h) >> 3;
// p1
t1 = (p2_l + p1_l + p0_l + q0_l + const_2_h) >> 2;
// p2
t2 = ((p3_l << 1) + p2_l + (p2_l << 1) + p1_l + p0_l + q0_l + const_4_h) >> 3;
// p0 condition 1
t0_con1 = ((p1_l << 1) + p0_l + q1_l + const_2_h) >> 2;
// q0
s0 = (p1_l + (p0_l << 1) + (q0_l << 1) + (q1_l << 1) + q2_l + const_4_h) >> 3;
// q1
s1 = (p0_l + q0_l + q1_l + q2_l + const_2_h) >> 2;
// q2
s2 = ((q3_l << 1) + q2_l + (q2_l << 1) + q1_l + q0_l + p0_l + const_4_h) >> 3;
// q0 condition 1
s0_con1 = ((q1_l << 1) + q0_l + p1_l + const_2_h) >> 2;
// Move back
p0_l = t0;
p1_l = t1;
p2_l = t2;
q0_l = s0;
q1_l = s1;
q2_l = s2;
// Use p3_l, q3_l as tmp
p3_l = t0_con1;
q3_l = s0_con1;
// Calculate the right part
// p0
t0 = (p2_r + (p1_r << 1) + (p0_r << 1) + (q0_r << 1) + q1_r + const_4_h) >> 3;
// p1
t1 = (p2_r + p1_r + p0_r + q0_r + const_2_h) >> 2;
// p2
t2 = ((p3_r << 1) + p2_r + (p2_r << 1) + p1_r + p0_r + q0_r + const_4_h) >> 3;
// p0 condition 1
t0_con1 = ((p1_r << 1) + p0_r + q1_r + const_2_h) >> 2;
// q0
s0 = (p1_r + (p0_r << 1) + (q0_r << 1) + (q1_r << 1) + q2_r + const_4_h) >> 3;
// q1
s1 = (p0_r + q0_r + q1_r + q2_r + const_2_h) >> 2;
// q2
s2 = ((q3_r << 1) + q2_r + (q2_r << 1) + q1_r + q0_r + p0_r + const_4_h) >> 3;
// q0 condition 1
s0_con1 = ((q1_r << 1) + q0_r + p1_r + const_2_h) >> 2;
// Move back
p0_r = t0;
p1_r = t1;
p2_r = t2;
q0_r = s0;
q1_r = s1;
q2_r = s2;
// Use p3_r, q3_r as tmp
p3_r = t0_con1;
q3_r = s0_con1;
// Combined left and right
MSA_PCKEV_B4(v8i16, p0_l, p0_r, p1_l, p1_r, p2_l, p2_r, q0_l, q0_r,
t0, t1, t2, s0);
MSA_PCKEV_B4(v8i16, q1_l, q1_r, q2_l, q2_r, p3_l, p3_r, q3_l, q3_r,
s1, s2, t0_con1, s0_con1);
t0 = (v8i16)(((v16u8)t0 & mask0 & mask1 & bDetaP2P0) + ((v16u8)t0_con1 &
mask0 & mask1 & (~bDetaP2P0)) + ((v16u8)t0_con1 & mask0 & (~mask1)));
t1 = (v8i16)((v16u8)t1 & mask0 & mask1 & bDetaP2P0);
t2 = (v8i16)((v16u8)t2 & mask0 & mask1 & bDetaP2P0);
s0 = (v8i16)(((v16u8)s0 & mask0 & mask1 & bDetaQ2Q0) + ((v16u8)s0_con1 &
mask0 & mask1 & (~bDetaQ2Q0)) + ((v16u8)s0_con1 & mask0 & (~mask1)));
s1 = (v8i16)((v16u8)s1 & mask0 & mask1 & bDetaQ2Q0);
s2 = (v8i16)((v16u8)s2 & mask0 & mask1 & bDetaQ2Q0);
p0 = (v16u8)t0 + (p0 & (~mask0));
p1 = (v16u8)t1 + (p1 & ~(mask0 & mask1 & bDetaP2P0));
p2 = (v16u8)t2 + (p2 & ~(mask0 & mask1 & bDetaP2P0));
q0 = (v16u8)s0 + (q0 & (~mask0));
q1 = (v16u8)s1 + (q1 & ~(mask0 & mask1 & bDetaQ2Q0));
q2 = (v16u8)s2 + (q2 & ~(mask0 & mask1 & bDetaQ2Q0));
MSA_ILVRL_B4(v8i16, p1, p2, q0, p0, t0, s0, t1, s1);
MSA_ILVRL_B2(v8i16, q2, q1, t2, s2);
MSA_ILVRL_H4(v16u8, t1, t0, s1, s0, p2, p1, p0, q0);
// Store data to pPix
MSA_ST_W8(p2, p1, 0, 1, 2, 3, 0, 1, 2, 3, pPix - 3, iStride);
MSA_ST_W8(p0, q0, 0, 1, 2, 3, 0, 1, 2, 3, pPix + 8 * iStride - 3, iStride);
MSA_ST_H8(t2, 0, 1, 2, 3, 4, 5, 6, 7, pPix + 1, iStride);
MSA_ST_H8(s2, 0, 1, 2, 3, 4, 5, 6, 7, pPix + 8 * iStride + 1, iStride);
}
void DeblockChromaLt4V_msa(uint8_t* pPixCb, uint8_t* pPixCr, int32_t iStride,
int32_t iAlpha, int32_t iBeta, int8_t* pTc) {
v16u8 p0, p1, q0, q1;
v8i16 p0_e, p1_e, q0_e, q1_e;
v16i8 negTc, flags, f;
v8i16 tc_e, negTc_e;
// Use for temporary variable
v8i16 t0, t1, t2, t3;
v16u8 alpha, beta;
v16u8 bDetaP0Q0, bDetaP1P0, bDetaQ1Q0;
v8i16 const_4_h = __msa_ldi_h(4);
v8i16 const_not_255_h = __msa_ldi_h(~255);
v16i8 zero = { 0 };
v16i8 tc = { pTc[0 >> 1], pTc[1 >> 1], pTc[2 >> 1], pTc[3 >> 1],
pTc[4 >> 1], pTc[5 >> 1], pTc[6 >> 1], pTc[7 >> 1] };
negTc = zero - tc;
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
// Signed extend tc, negTc from 8 bits to 16 bits
flags = __msa_clt_s_b(tc, zero);
MSA_ILVR_B(v8i16, flags, tc, tc_e);
flags = __msa_clt_s_b(negTc, zero);
MSA_ILVR_B(v8i16, flags, negTc, negTc_e);
// Cb
// Load data from pPixCb
MSA_LD_V4(v16u8, pPixCb - 2 * iStride, iStride, p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// iDeta
t0 = (((q0_e - p0_e) << 2) + (p1_e - q1_e) + const_4_h) >> 3;
t0 = __msa_max_s_h(negTc_e, t0);
t0 = __msa_min_s_h(tc_e, t0);
// p0
t1 = p0_e + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_e - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
flags = (v16i8)__msa_cle_s_b(zero, tc);
flags &= f;
p0 = (v16u8)(((v16i8)t0 & flags) + (p0 & (~flags)));
q0 = (v16u8)(((v16i8)t1 & flags) + (q0 & (~flags)));
// Store data to pPixCb
MSA_ST_D(p0, 0, pPixCb - iStride);
MSA_ST_D(q0, 0, pPixCb);
// Cr
// Load data from pPixCr
MSA_LD_V4(v16u8, pPixCr - 2 * iStride, iStride, p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// iDeta
t0 = (((q0_e - p0_e) << 2) + (p1_e - q1_e) + const_4_h) >> 3;
t0 = __msa_max_s_h(negTc_e, t0);
t0 = __msa_min_s_h(tc_e, t0);
// p0
t1 = p0_e + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_e - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
flags = (v16i8)__msa_cle_s_b(zero, tc);
flags &= f;
p0 = (v16u8)(((v16i8)t0 & flags) + (p0 & (~flags)));
q0 = (v16u8)(((v16i8)t1 & flags) + (q0 & (~flags)));
// Store data to pPixCr
MSA_ST_D(p0, 0, pPixCr - iStride);
MSA_ST_D(q0, 0, pPixCr);
}
void DeblockChromaEq4V_msa(uint8_t* pPixCb, uint8_t* pPixCr, int32_t iStride,
int32_t iAlpha, int32_t iBeta) {
v16u8 p0, p1, q0, q1;
v8i16 p0_e, p1_e, q0_e, q1_e;
v16i8 f;
// Use for temporary variable
v8i16 t0, t1;
v16u8 alpha, beta;
v16u8 bDetaP0Q0, bDetaP1P0, bDetaQ1Q0;
v8i16 const_2_h = __msa_ldi_h(2);
v16i8 zero = { 0 };
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
// Cb
// Load data from pPixCb
MSA_LD_V4(v16u8, pPixCb - 2 * iStride, iStride, p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// p0
p0_e = ((p1_e << 1) + p0_e + q1_e + const_2_h) >> 2;
// q0
q0_e = ((q1_e << 1) + q0_e + p1_e + const_2_h) >> 2;
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
p0 = (v16u8)(((v16i8)t0 & f) + (p0 & (~f)));
q0 = (v16u8)(((v16i8)t1 & f) + (q0 & (~f)));
// Store data to pPixCb
MSA_ST_D(p0, 0, pPixCb - iStride);
MSA_ST_D(q0, 0, pPixCb);
// Cr
// Load data from pPixCr
MSA_LD_V4(v16u8, pPixCr - 2 * iStride, iStride, p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// p0
p0_e = ((p1_e << 1) + p0_e + q1_e + const_2_h) >> 2;
// q0
q0_e = ((q1_e << 1) + q0_e + p1_e + const_2_h) >> 2;
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
p0 = (v16u8)(((v16i8)t0 & f) + (p0 & (~f)));
q0 = (v16u8)(((v16i8)t1 & f) + (q0 & (~f)));
// Store data to pPixCr
MSA_ST_D(p0, 0, pPixCr - iStride);
MSA_ST_D(q0, 0, pPixCr);
}
void DeblockChromaLt4H_msa(uint8_t* pPixCb, uint8_t* pPixCr, int32_t iStride,
int32_t iAlpha, int32_t iBeta, int8_t* pTc) {
v16u8 p0, p1, q0, q1;
v8i16 p0_e, p1_e, q0_e, q1_e;
v16i8 negTc, flags, f;
v8i16 tc_e, negTc_e;
// Use for temporary variable
v8i16 t0, t1, t2, t3;
v16u8 alpha, beta;
v16u8 bDetaP0Q0, bDetaP1P0, bDetaQ1Q0;
v8i16 const_4_h = __msa_ldi_h(4);
v8i16 const_not_255_h = __msa_ldi_h(~255);
v16i8 zero = { 0 };
v16i8 tc = { pTc[0 >> 1], pTc[1 >> 1], pTc[2 >> 1], pTc[3 >> 1],
pTc[4 >> 1], pTc[5 >> 1], pTc[6 >> 1], pTc[7 >> 1] };
negTc = zero - tc;
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
// Signed extend tc, negTc from 8 bits to 16 bits
flags = __msa_clt_s_b(tc, zero);
MSA_ILVR_B(v8i16, flags, tc, tc_e);
flags = __msa_clt_s_b(negTc, zero);
MSA_ILVR_B(v8i16, flags, negTc, negTc_e);
// Cb
// Load data from pPixCb
MSA_LD_V8(v8i16, pPixCb - 2, iStride, p1_e, p0_e, q0_e, q1_e,
t0, t1, t2, t3);
// Transpose 8x4 to 4x8, we just need p0, p1, q0, q1
MSA_TRANSPOSE8x4_B(v16u8, p1_e, p0_e, q0_e, q1_e, t0, t1, t2, t3,
p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// iDeta
t0 = (((q0_e - p0_e) << 2) + (p1_e - q1_e) + const_4_h) >> 3;
t0 = __msa_max_s_h(negTc_e, t0);
t0 = __msa_min_s_h(tc_e, t0);
// p0
t1 = p0_e + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_e - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
flags = (v16i8)__msa_cle_s_b(zero, tc);
flags &= f;
p0 = (v16u8)(((v16i8)t0 & flags) + (p0 & (~flags)));
q0 = (v16u8)(((v16i8)t1 & flags) + (q0 & (~flags)));
// Store data to pPixCb
MSA_ILVR_B(v16u8, q0, p0, p0);
MSA_ST_H8(p0, 0, 1, 2, 3, 4, 5, 6, 7, pPixCb - 1, iStride);
// Cr
// Load data from pPixCr
MSA_LD_V8(v8i16, pPixCr - 2, iStride, p1_e, p0_e, q0_e, q1_e,
t0, t1, t2, t3);
// Transpose 8x4 to 4x8, we just need p0, p1, q0, q1
MSA_TRANSPOSE8x4_B(v16u8, p1_e, p0_e, q0_e, q1_e, t0, t1, t2, t3,
p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// iDeta
t0 = (((q0_e - p0_e) << 2) + (p1_e - q1_e) + const_4_h) >> 3;
t0 = __msa_max_s_h(negTc_e, t0);
t0 = __msa_min_s_h(tc_e, t0);
// p0
t1 = p0_e + t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
p0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
// q0
t1 = q0_e - t0;
t2 = t1 & const_not_255_h;
t3 = __msa_cle_s_h((v8i16)zero, t1);
flags = (v16i8)__msa_ceq_h(t2, (v8i16)zero);
q0_e = (t1 & (v8i16)flags) + (t3 & (v8i16)(~flags));
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
flags = (v16i8)__msa_cle_s_b(zero, tc);
flags &= f;
p0 = (v16u8)(((v16i8)t0 & flags) + (p0 & (~flags)));
q0 = (v16u8)(((v16i8)t1 & flags) + (q0 & (~flags)));
// Store data to pPixCr
MSA_ILVR_B(v16u8, q0, p0, p0);
MSA_ST_H8(p0, 0, 1, 2, 3, 4, 5, 6, 7, pPixCr - 1, iStride);
}
void DeblockChromaEq4H_msa(uint8_t* pPixCb, uint8_t* pPixCr, int32_t iStride,
int32_t iAlpha, int32_t iBeta) {
v16u8 p0, p1, q0, q1;
v8i16 p0_e, p1_e, q0_e, q1_e;
v16i8 f;
// Use for temporary variable
v8i16 t0, t1, t2, t3;
v16u8 alpha, beta;
v16u8 bDetaP0Q0, bDetaP1P0, bDetaQ1Q0;
v8i16 const_2_h = __msa_ldi_h(2);
v16i8 zero = { 0 };
alpha = (v16u8)__msa_fill_b(iAlpha);
beta = (v16u8)__msa_fill_b(iBeta);
// Cb
// Load data from pPixCb
MSA_LD_V8(v8i16, pPixCb - 2, iStride, p1_e, p0_e, q0_e, q1_e,
t0, t1, t2, t3);
// Transpose 8x4 to 4x8, we just need p0, p1, q0, q1
MSA_TRANSPOSE8x4_B(v16u8, p1_e, p0_e, q0_e, q1_e, t0, t1, t2, t3,
p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// p0
p0_e = ((p1_e << 1) + p0_e + q1_e + const_2_h) >> 2;
// q0
q0_e = ((q1_e << 1) + q0_e + p1_e + const_2_h) >> 2;
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
p0 = (v16u8)(((v16i8)t0 & f) + (p0 & (~f)));
q0 = (v16u8)(((v16i8)t1 & f) + (q0 & (~f)));
// Store data to pPixCb
MSA_ILVR_B(v16u8, q0, p0, p0);
MSA_ST_H8(p0, 0, 1, 2, 3, 4, 5, 6, 7, pPixCb - 1, iStride);
// Cr
// Load data from pPixCr
MSA_LD_V8(v8i16, pPixCr - 2, iStride, p1_e, p0_e, q0_e, q1_e,
t0, t1, t2, t3);
// Transpose 8x4 to 4x8, we just need p0, p1, q0, q1
MSA_TRANSPOSE8x4_B(v16u8, p1_e, p0_e, q0_e, q1_e, t0, t1, t2, t3,
p1, p0, q0, q1);
bDetaP0Q0 = __msa_asub_u_b(p0, q0);
bDetaP1P0 = __msa_asub_u_b(p1, p0);
bDetaQ1Q0 = __msa_asub_u_b(q1, q0);
bDetaP0Q0 = (v16u8)__msa_clt_u_b(bDetaP0Q0, alpha);
bDetaP1P0 = (v16u8)__msa_clt_u_b(bDetaP1P0, beta);
bDetaQ1Q0 = (v16u8)__msa_clt_u_b(bDetaQ1Q0, beta);
// Unsigned extend p0, p1, q0, q1 from 8 bits to 16 bits
MSA_ILVR_B4(v8i16, zero, p0, zero, p1, zero, q0, zero, q1,
p0_e, p1_e, q0_e, q1_e);
f = (v16i8)bDetaP0Q0 & (v16i8)bDetaP1P0 & (v16i8)bDetaQ1Q0;
// p0
p0_e = ((p1_e << 1) + p0_e + q1_e + const_2_h) >> 2;
// q0
q0_e = ((q1_e << 1) + q0_e + p1_e + const_2_h) >> 2;
MSA_PCKEV_B2(v8i16, p0_e, p0_e, q0_e, q0_e, t0, t1);
p0 = (v16u8)(((v16i8)t0 & f) + (p0 & (~f)));
q0 = (v16u8)(((v16i8)t1 & f) + (q0 & (~f)));
// Store data to pPixCr
MSA_ILVR_B(v16u8, q0, p0, p0);
MSA_ST_H8(p0, 0, 1, 2, 3, 4, 5, 6, 7, pPixCr - 1, iStride);
}
void WelsNonZeroCount_msa(int8_t* pNonZeroCount) {
v16u8 src0, src1;
v16u8 zero = { 0 };
v16u8 const_1 = (v16u8)__msa_fill_b(0x01);
MSA_LD_V2(v16u8, pNonZeroCount, 16, src0, src1);
src0 = (v16u8)__msa_ceq_b((v16i8)zero, (v16i8)src0);
src1 = (v16u8)__msa_ceq_b((v16i8)zero, (v16i8)src1);
src0 += const_1;
src1 += const_1;
MSA_ST_V(v16u8, src0, pNonZeroCount);
MSA_ST_D(src1, 0, pNonZeroCount + 16);
}