Chiebot-Mirror
/
opencv
mirror of https://github.com/opencv/opencv.git
8
0
Fork 0
Code Issues Projects Releases Wiki Activity

Merge pull request #23929 from CNClareChen:4.x

Browse Source 
* Optimize some function with lasx.

Optimize some function with lasx. #23929

This patch optimizes some lasx functions and reduces the runtime of opencv_test_core from 662,238ms to 633603ms on the 3A5000 platform.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
pull/24438/head
CNClareChen 1 year ago committed by GitHub
parent 996b6c37c7
commit d142a796d8
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
9 changed files with 2800 additions and 414 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 11
      cmake/OpenCVCompilerOptimizations.cmake
  2. 15
      cmake/checks/cpu_lsx.cpp
  3. 9
      modules/core/include/opencv2/core/cv_cpu_dispatch.h
  4. 21
      modules/core/include/opencv2/core/cv_cpu_helper.h
  5. 6
      modules/core/include/opencv2/core/cvdef.h
  6. 6
      modules/core/include/opencv2/core/hal/intrin.hpp
  7. 605
      modules/core/include/opencv2/core/hal/intrin_lasx.hpp
  8. 2536
      modules/core/include/opencv2/core/hal/intrin_lsx.hpp
  9. 5
      modules/core/src/system.cpp

11
cmake/OpenCVCompilerOptimizations.cmake
Unescape View File

@ -53,6 +53,7 @@ list(APPEND CPU_ALL_OPTIMIZATIONS NEON VFPV3 FP16 NEON_DOTPROD)
list(APPEND CPU_ALL_OPTIMIZATIONS MSA)
list(APPEND CPU_ALL_OPTIMIZATIONS VSX VSX3)
list(APPEND CPU_ALL_OPTIMIZATIONS RVV)
list(APPEND CPU_ALL_OPTIMIZATIONS LSX)
list(APPEND CPU_ALL_OPTIMIZATIONS LASX)
list(REMOVE_DUPLICATES CPU_ALL_OPTIMIZATIONS)
@ -397,10 +398,16 @@ elseif(RISCV)
set(CPU_BASELINE "DETECT" CACHE STRING "${HELP_CPU_BASELINE}")
elseif(LOONGARCH64)
ocv_update(CPU_LSX_TEST_FILE "${OpenCV_SOURCE_DIR}/cmake/checks/cpu_lsx.cpp")
ocv_update(CPU_LASX_TEST_FILE "${OpenCV_SOURCE_DIR}/cmake/checks/cpu_lasx.cpp")
ocv_update(CPU_KNOWN_OPTIMIZATIONS "LASX")
ocv_update(CPU_KNOWN_OPTIMIZATIONS "LSX;LASX")
ocv_update(CPU_LSX_FLAGS_ON "-mlsx")
ocv_update(CPU_LASX_FLAGS_ON "-mlasx")
set(CPU_BASELINE "LASX" CACHE STRING "${HELP_CPU_BASELINE}")
if("${CPU_BASELINE_DISABLE}" STREQUAL "LASX")
set(CPU_BASELINE "LSX" CACHE STRING "${HELP_CPU_BASELINE}")
else()
set(CPU_BASELINE "LASX" CACHE STRING "${HELP_CPU_BASELINE}")
endif()
endif()

15
cmake/checks/cpu_lsx.cpp
Unescape View File

@ -0,0 +1,15 @@
#include <stdio.h>
#include <lsxintrin.h>
int test()
{
const float src[] = { 0.0f, 1.0f, 2.0f, 3.0f};
v4f32 val = (v4f32)__lsx_vld((const float*)(src), 0);
return __lsx_vpickve2gr_w(__lsx_vftint_w_s(val), 3);
}
int main()
{
printf("%d\n", test());
return 0;
}

9
modules/core/include/opencv2/core/cv_cpu_dispatch.h
Unescape View File

@ -172,6 +172,11 @@
# define CV_MSA 1
#endif
#ifdef CV_CPU_COMPILE_LSX
# include <lsxintrin.h>
# define CV_LSX 1
#endif
#ifdef CV_CPU_COMPILE_LASX
# include <lasxintrin.h>
# define CV_LASX 1
@ -376,6 +381,10 @@ struct VZeroUpperGuard {
# define CV_RVV 0
#endif
#ifndef CV_LSX
# define CV_LSX 0
#endif
#ifndef CV_LASX
# define CV_LASX 0
#endif

21
modules/core/include/opencv2/core/cv_cpu_helper.h
Unescape View File

@ -525,6 +525,27 @@
#endif
#define __CV_CPU_DISPATCH_CHAIN_RVV(fn, args, mode, ...) CV_CPU_CALL_RVV(fn, args); __CV_EXPAND(__CV_CPU_DISPATCH_CHAIN_ ## mode(fn, args, __VA_ARGS__))
#if !defined CV_DISABLE_OPTIMIZATION && defined CV_ENABLE_INTRINSICS && defined CV_CPU_COMPILE_LSX
# define CV_TRY_LSX 1
# define CV_CPU_FORCE_LSX 1
# define CV_CPU_HAS_SUPPORT_LSX 1
# define CV_CPU_CALL_LSX(fn, args) return (cpu_baseline::fn args)
# define CV_CPU_CALL_LSX_(fn, args) return (opt_LSX::fn args)
#elif !defined CV_DISABLE_OPTIMIZATION && defined CV_ENABLE_INTRINSICS && defined CV_CPU_DISPATCH_COMPILE_LSX
# define CV_TRY_LSX 1
# define CV_CPU_FORCE_LSX 0
# define CV_CPU_HAS_SUPPORT_LSX (cv::checkHardwareSupport(CV_CPU_LSX))
# define CV_CPU_CALL_LSX(fn, args) if (CV_CPU_HAS_SUPPORT_LSX) return (opt_LSX::fn args)
# define CV_CPU_CALL_LSX_(fn, args) if (CV_CPU_HAS_SUPPORT_LSX) return (opt_LSX::fn args)
#else
# define CV_TRY_LSX 0
# define CV_CPU_FORCE_LSX 0
# define CV_CPU_HAS_SUPPORT_LSX 0
# define CV_CPU_CALL_LSX(fn, args)
# define CV_CPU_CALL_LSX_(fn, args)
#endif
#define __CV_CPU_DISPATCH_CHAIN_LSX(fn, args, mode, ...) CV_CPU_CALL_LSX(fn, args); __CV_EXPAND(__CV_CPU_DISPATCH_CHAIN_ ## mode(fn, args, __VA_ARGS__))
#if !defined CV_DISABLE_OPTIMIZATION && defined CV_ENABLE_INTRINSICS && defined CV_CPU_COMPILE_LASX
# define CV_TRY_LASX 1
# define CV_CPU_FORCE_LASX 1

6
modules/core/include/opencv2/core/cvdef.h
Unescape View File

@ -281,7 +281,8 @@ namespace cv {
#define CV_CPU_RVV 210
#define CV_CPU_LASX 230
#define CV_CPU_LSX 230
#define CV_CPU_LASX 231
// CPU features groups
#define CV_CPU_AVX512_SKX 256
@ -342,7 +343,8 @@ enum CpuFeatures {
CPU_RVV = 210,
CPU_LASX = 230,
CPU_LSX = 230,
CPU_LASX = 231,
CPU_AVX512_SKX = 256, //!< Skylake-X with AVX-512F/CD/BW/DQ/VL
CPU_AVX512_COMMON = 257, //!< Common instructions AVX-512F/CD for all CPUs that support AVX-512

6
modules/core/include/opencv2/core/hal/intrin.hpp
Unescape View File

@ -206,7 +206,7 @@ using namespace CV_CPU_OPTIMIZATION_HAL_NAMESPACE;
# undef CV_RVV
#endif
#if (CV_SSE2 || CV_NEON || CV_VSX || CV_MSA || CV_WASM_SIMD || CV_RVV071) && !defined(CV_FORCE_SIMD128_CPP)
#if (CV_SSE2 || CV_NEON || CV_VSX || CV_MSA || CV_WASM_SIMD || CV_RVV071 || CV_LSX) && !defined(CV_FORCE_SIMD128_CPP)
#define CV__SIMD_FORWARD 128
#include "opencv2/core/hal/intrin_forward.hpp"
#endif
@ -242,6 +242,10 @@ using namespace CV_CPU_OPTIMIZATION_HAL_NAMESPACE;
#include "opencv2/core/hal/intrin_rvv.hpp"
#endif
#elif CV_LSX && !defined(CV_FORCE_SIMD128_CPP)
#include "opencv2/core/hal/intrin_lsx.hpp"
#elif CV_LASX
#if !defined(CV_FORCE_SIMD128_CPP)
#define CV_FORCE_SIMD128_CPP 1

605
modules/core/include/opencv2/core/hal/intrin_lasx.hpp
Unescape View File

@ -96,54 +96,22 @@ inline __m256d _v256_setall_pd(double f64)
inline __m256i _lasx_packus_h(const __m256i& a, const __m256i& b)
{
__m256i u8min = __lasx_xvreplgr2vr_h(0);
__m256i u8max = __lasx_xvreplgr2vr_h(255);
__m256i sat_a = __lasx_xvmax_h(a, u8min);
sat_a = __lasx_xvmin_h(sat_a, u8max);
__m256i sat_b = __lasx_xvmax_h(b, u8min);
sat_b = __lasx_xvmin_h(sat_b, u8max);
__m256i byteIndex = _v256_setr_b(0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30,
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
return __lasx_xvshuf_b(sat_b, sat_a, byteIndex);
return __lasx_xvssrarni_bu_h(b, a, 0);
}
inline __m256i _lasx_packs_h(const __m256i& a, const __m256i& b)
{
__m256i s8min = __lasx_xvreplgr2vr_h(-128);
__m256i s8max = __lasx_xvreplgr2vr_h(127);
__m256i sat_a = __lasx_xvmax_h(a, s8min);
sat_a = __lasx_xvmin_h(sat_a, s8max);
__m256i sat_b = __lasx_xvmax_h(b, s8min);
sat_b = __lasx_xvmin_h(sat_b, s8max);
__m256i byteIndex = _v256_setr_b(0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30,
0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30);
return __lasx_xvshuf_b(sat_b, sat_a, byteIndex);
return __lasx_xvssrarni_b_h(b, a, 0);
}
inline __m256i _lasx_packus_w(const __m256i& a, const __m256i& b)
{
__m256i u16min = __lasx_xvreplgr2vr_w(0);
__m256i u16max = __lasx_xvreplgr2vr_w(0xffff);
__m256i sat_a = __lasx_xvmax_w(a, u16min);
sat_a = __lasx_xvmin_w(sat_a, u16max);
__m256i sat_b = __lasx_xvmax_w(b, u16min);
sat_b = __lasx_xvmin_w(sat_b, u16max);
__m256i hwordIndex = _v256_setr_h(0, 2, 4, 6, 8, 10, 12, 14,
0, 2, 4, 6, 8, 10, 12, 14);
return __lasx_xvshuf_h(hwordIndex, sat_b, sat_a);
return __lasx_xvssrarni_hu_w(b, a, 0);
}
inline __m256i _lasx_packs_w(const __m256i& a, const __m256i& b)
{
__m256i s16min = __lasx_xvreplgr2vr_w(-0x8000);
__m256i s16max = __lasx_xvreplgr2vr_w(0x7fff);
__m256i sat_a = __lasx_xvmax_w(a, s16min);
sat_a = __lasx_xvmin_w(sat_a, s16max);
__m256i sat_b = __lasx_xvmax_w(b, s16min);
sat_b = __lasx_xvmin_w(sat_b, s16max);
__m256i hwordIndex = _v256_setr_h(0, 2, 4, 6, 8, 10, 12, 14,
0, 2, 4, 6, 8, 10, 12, 14);
return __lasx_xvshuf_h(hwordIndex, sat_b, sat_a);
return __lasx_xvssrarni_h_w(b, a, 0);
}
inline __m256i _v256_combine(const __m128i& lo, const __m128i& hi)
@ -191,7 +159,7 @@ inline _Tpvec v256_permute4x64(const _Tpvec& a)
{ return _Tpvec(_v256_permute4x64<imm>(a.val)); }
inline __m128i _v256_extract_high(const __m256i& v)
{ __m256i temp256i = __lasx_xvpermi_q(v, v, 0x31);
{ __m256i temp256i = __lasx_xvpermi_d(v, 0x4E);
return *((__m128i*)&temp256i); }
inline __m128 _v256_extract_high(const __m256& v)
@ -211,10 +179,7 @@ inline __m128d _v256_extract_low(const __m256d& v)
inline __m256i _v256_packs_epu32(const __m256i& a, const __m256i& b)
{
const __m256i maxv = __lasx_xvreplgr2vr_w(65535);
__m256i am = __lasx_xvmin_wu(a, maxv);
__m256i bm = __lasx_xvmin_wu(b, maxv);
return _lasx_packus_w(am, bm);
return __lasx_xvssrlrni_hu_w(b, a, 0);
}
template<int i>
@ -869,14 +834,11 @@ OPENCV_HAL_IMPL_LASX_BIN_FUNC(v_mul_wrap, v_int16x16, __lasx_xvmul_h)
inline v_uint8x32 v_mul_wrap(const v_uint8x32& a, const v_uint8x32& b)
{
__m256i ad = __lasx_xvsrai_h(a.val, 8);
__m256i bd = __lasx_xvsrai_h(b.val, 8);
__m256i p0 = __lasx_xvmul_h(a.val, b.val);
__m256i p1 = __lasx_xvslli_h(__lasx_xvmul_h(ad, bd), 8);
const __m256i b01 = __lasx_xvreplgr2vr_w(0xFF00FF00);
return v_uint8x32(__lasx_xvbitsel_v(p0, p1, b01));
__m256i p0 = __lasx_xvmulwev_h_bu(a.val, b.val);
__m256i p1 = __lasx_xvmulwod_h_bu(a.val, b.val);
return v_uint8x32(__lasx_xvpackev_b(p1, p0));
}
inline v_int8x32 v_mul_wrap(const v_int8x32& a, const v_int8x32& b)
{
return v_reinterpret_as_s8(v_mul_wrap(v_reinterpret_as_u8(a), v_reinterpret_as_u8(b)));
@ -963,14 +925,7 @@ inline v_uint16x16 v_mul_hi(const v_uint16x16& a, const v_uint16x16& b) { return
OPENCV_HAL_IMPL_LASX_SHIFT_OP(v_uint16x16, v_int16x16, h, __lasx_xvsra_h)
OPENCV_HAL_IMPL_LASX_SHIFT_OP(v_uint32x8, v_int32x8, w, __lasx_xvsra_w)
inline __m256i _v256_srai_dx(const __m256i a, const __m256i shift)
{
__m256i d = __lasx_xvreplgr2vr_d((int64)1 << 63);
__m256i r = __lasx_xvsrl_d(__lasx_xvadd_d(a, d), shift);
return __lasx_xvsub_d(r, __lasx_xvsrl_d(d, shift));
}
OPENCV_HAL_IMPL_LASX_SHIFT_OP(v_uint64x4, v_int64x4, d, _v256_srai_dx)
OPENCV_HAL_IMPL_LASX_SHIFT_OP(v_uint64x4, v_int64x4, d, __lasx_xvsra_d)
/** Bitwise logic **/
@ -979,7 +934,7 @@ OPENCV_HAL_IMPL_LASX_SHIFT_OP(v_uint64x4, v_int64x4, d, _v256_srai_dx)
OPENCV_HAL_IMPL_LASX_BIN_OP(|, _Tpvec, __lasx_xvor_##suffix) \
OPENCV_HAL_IMPL_LASX_BIN_OP(^, _Tpvec, __lasx_xvxor_##suffix) \
inline _Tpvec operator ~ (const _Tpvec& a) \
{ return _Tpvec(__lasx_xvxor_##suffix(a.val, not_const)); }
{ return _Tpvec(__lasx_xvnori_b(a.val, 0)); }
OPENCV_HAL_IMPL_LASX_LOGIC_OP(v_uint8x32, v, __lasx_xvreplgr2vr_w(-1))
OPENCV_HAL_IMPL_LASX_LOGIC_OP(v_int8x32, v, __lasx_xvreplgr2vr_w(-1))
@ -1224,11 +1179,9 @@ inline v_int8x32 v_reverse(const v_int8x32 &a)
inline v_uint16x16 v_reverse(const v_uint16x16 &a)
{
static const __m256i perm = _v256_setr_b(
14, 15, 12, 13, 10, 11, 8, 9, 6, 7, 4, 5, 2, 3, 0, 1,
14, 15, 12, 13, 10, 11, 8, 9, 6, 7, 4, 5, 2, 3, 0, 1);
__m256i vec = __lasx_xvshuf_b(a.val, a.val, perm);
return v_uint16x16(__lasx_xvpermi_q(vec, vec, 1));
__m256i vec = __lasx_xvshuf4i_h(a.val, 0x1B);
vec = __lasx_xvshuf4i_w(vec, 0x4E);
return v_uint16x16(__lasx_xvpermi_d(vec, 0x4E));
}
inline v_int16x16 v_reverse(const v_int16x16 &a)
@ -1236,8 +1189,8 @@ inline v_int16x16 v_reverse(const v_int16x16 &a)
inline v_uint32x8 v_reverse(const v_uint32x8 &a)
{
static const __m256i perm = _v256_setr_w(7, 6, 5, 4, 3, 2, 1, 0);
return v_uint32x8(__lasx_xvperm_w(a.val, perm));
__m256i vec = __lasx_xvshuf4i_w(a.val, 0x1B);
return v_uint32x8(__lasx_xvpermi_d(vec, 0x4E));
}
inline v_int32x8 v_reverse(const v_int32x8 &a)
@ -1266,17 +1219,19 @@ inline unsigned v_reduce_sum(const v_uint8x32& a)
__m256i t1 = __lasx_xvhaddw_hu_bu(a.val, a.val);
__m256i t2 = __lasx_xvhaddw_wu_hu(t1, t1);
__m256i t3 = __lasx_xvhaddw_du_wu(t2, t2);
return (unsigned)(((v4u64)t3)[0]+((v4u64)t3)[1]+((v4u64)t3)[2]+((v4u64)t3)[3]);
__m256i t4 = __lasx_xvhaddw_qu_du(t3, t3);
return (unsigned)(((v8u32)t4)[0]+((v8u32)t4)[4]);
}
inline int v_reduce_sum(const v_int8x32& a)
{
__m256i t1 = __lasx_xvhaddw_h_b(a.val, a.val);
__m256i t2 = __lasx_xvhaddw_w_h(t1, t1);
__m256i t3 = __lasx_xvhaddw_d_w(t2, t2);
return (int)(((v4i64)t3)[0]+((v4i64)t3)[1]+((v4i64)t3)[2]+((v4i64)t3)[3]);
__m256i t4 = __lasx_xvhaddw_q_d(t3, t3);
return (int)(((v8i32)t4)[0]+((v8i32)t4)[4]);
}
#define OPENCV_HAL_IMPL_LASX_REDUCE_32(_Tpvec, sctype, func, intrin) \
inline sctype v_reduce_##func(const _Tpvec& a) \
{ \
@ -1344,7 +1299,8 @@ OPENCV_HAL_IMPL_LASX_REDUCE_FLT(max, __lsx_vfmax_s)
inline int v_reduce_sum(const v_int32x8& a)
{
__m256i t1 = __lasx_xvhaddw_d_w(a.val, a.val);
return (int)(((v4i64)t1)[0]+((v4i64)t1)[1]+((v4i64)t1)[2]+((v4i64)t1)[3]);
__m256i t2 = __lasx_xvhaddw_q_d(t1, t1);
return (int)(((v8i32)t2)[0]+((v8i32)t2)[4]);
}
inline unsigned v_reduce_sum(const v_uint32x8& a)
@ -1367,13 +1323,13 @@ inline float v_reduce_sum(const v_float32x8& a)
inline uint64 v_reduce_sum(const v_uint64x4& a)
{
uint64 *pa = (uint64*)&a;
return pa[0] + pa[1] + pa[2] + pa[3];
__m256i t0 = __lasx_xvhaddw_qu_du(a.val, a.val);
return (uint64)(((v4u64)t0)[0] + ((v4u64)t0)[2]);
}
inline int64 v_reduce_sum(const v_int64x4& a)
{
int64 *pa = (int64*)&a;
return pa[0] + pa[1] + pa[2] + pa[3];
__m256i t0 = __lasx_xvhaddw_q_d(a.val, a.val);
return (int64)(((v4i64)t0)[0] + ((v4i64)t0)[2]);
}
inline double v_reduce_sum(const v_float64x4& a)
{
@ -1406,7 +1362,8 @@ inline unsigned v_reduce_sad(const v_uint8x32& a, const v_uint8x32& b)
__m256i t1 = __lasx_xvhaddw_hu_bu(t0, t0);
__m256i t2 = __lasx_xvhaddw_wu_hu(t1, t1);
__m256i t3 = __lasx_xvhaddw_du_wu(t2, t2);
return (unsigned)(((v4u64)t3)[0]+((v4u64)t3)[1]+((v4u64)t3)[2]+((v4u64)t3)[3]);
__m256i t4 = __lasx_xvhaddw_qu_du(t3, t3);
return (unsigned)(((v8u32)t4)[0]+((v8u32)t4)[4]);
}
inline unsigned v_reduce_sad(const v_int8x32& a, const v_int8x32& b)
{
@ -1414,7 +1371,8 @@ inline unsigned v_reduce_sad(const v_int8x32& a, const v_int8x32& b)
__m256i t1 = __lasx_xvhaddw_hu_bu(t0, t0);
__m256i t2 = __lasx_xvhaddw_wu_hu(t1, t1);
__m256i t3 = __lasx_xvhaddw_du_wu(t2, t2);
return (unsigned)(((v4u64)t3)[0]+((v4u64)t3)[1]+((v4u64)t3)[2]+((v4u64)t3)[3]);
__m256i t4 = __lasx_xvhaddw_qu_du(t3, t3);
return (unsigned)(((v8u32)t4)[0]+((v8u32)t4)[4]);
}
inline unsigned v_reduce_sad(const v_uint16x16& a, const v_uint16x16& b)
{
@ -1445,36 +1403,13 @@ inline float v_reduce_sad(const v_float32x8& a, const v_float32x8& b)
/** Popcount **/
inline v_uint8x32 v_popcount(const v_uint8x32& a)
{
__m256i _popcnt_table = _v256_setr_b(0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4,
0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4);
__m256i _popcnt_mask = __lasx_xvreplgr2vr_b(0x0F);
return v_uint8x32(__lasx_xvadd_b(__lasx_xvshuf_b(_popcnt_table, _popcnt_table, __lasx_xvand_v(a.val, _popcnt_mask)),
__lasx_xvshuf_b(_popcnt_table, _popcnt_table, __lasx_xvand_v(__lasx_xvsrli_h(a.val, 4), _popcnt_mask))));
}
{ return v_uint8x32(__lasx_xvpcnt_b(a.val)); }
inline v_uint16x16 v_popcount(const v_uint16x16& a)
{
v_uint8x32 p = v_popcount(v_reinterpret_as_u8(a));
p += v_rotate_right<1>(p);
return v_reinterpret_as_u16(p) & v_uint16x16(__lasx_xvreplgr2vr_h(0x00ff));
}
{ return v_uint16x16(__lasx_xvpcnt_h(a.val)); }
inline v_uint32x8 v_popcount(const v_uint32x8& a)
{
v_uint8x32 p = v_popcount(v_reinterpret_as_u8(a));
p += v_rotate_right<1>(p);
p += v_rotate_right<2>(p);
return v_reinterpret_as_u32(p) & v_uint32x8(__lasx_xvreplgr2vr_w(0x000000ff));
}
{ return v_uint32x8(__lasx_xvpcnt_w(a.val)); }
inline v_uint64x4 v_popcount(const v_uint64x4& a)
{
v_uint8x32 atemp = v_popcount(v_reinterpret_as_u8(a));
uint8_t *pa = (uint8_t*)&atemp;
uint64 v[4];
for (int i = 0; i < 4; ++i) {
v[i] = pa[i*8] + pa[i*8+1] + pa[i*8+2] + pa[i*8+3] + pa[i*8+4] + pa[i*8+5] + pa[i*8+6] + pa[i*8+7];
}
return v_uint64x4(v[0], v[1], v[2], v[3]);
}
{ return v_uint64x4(__lasx_xvpcnt_d(a.val)); }
inline v_uint8x32 v_popcount(const v_int8x32& a)
{ return v_popcount(v_reinterpret_as_u8(a)); }
inline v_uint16x16 v_popcount(const v_int16x16& a)
@ -1500,10 +1435,9 @@ OPENCV_HAL_IMPL_REINTERPRET_INT(double, int64)
inline int v_signmask(const v_int8x32& a)
{
int mask = 0;
int8_t *pa = (int8_t*)&a;
for( int i = 0; i < 32; i++ )
mask |= (reinterpret_int(pa[i]) < 0) << i;
__m256i result = __lasx_xvmskltz_b(a.val);
int mask = __lasx_xvpickve2gr_w(result, 0);
mask |= (__lasx_xvpickve2gr_w(result, 4) << 16);
return mask;
}
inline int v_signmask(const v_uint8x32& a)
@ -1516,10 +1450,9 @@ inline int v_signmask(const v_uint16x16& a)
inline int v_signmask(const v_int32x8& a)
{
int mask = 0;
int *pa = (int*)&a;
for( int i = 0; i < 8; i++ )
mask |= (pa[i] < 0) << i;
__m256i result = __lasx_xvmskltz_w(a.val);
int mask = __lasx_xvpickve2gr_w(result, 0);
mask |= (__lasx_xvpickve2gr_w(result, 4) << 4);
return mask;
}
inline int v_signmask(const v_uint32x8& a)
@ -1527,10 +1460,9 @@ inline int v_signmask(const v_uint32x8& a)
inline int v_signmask(const v_int64x4& a)
{
int mask = 0;
int64 *pa = (int64*)&a;
for( int i = 0; i < 4; i++ )
mask |= (pa[i] < 0) << i;
__m256i result = __lasx_xvmskltz_d(a.val);
int mask = __lasx_xvpickve2gr_d(result, 0);
mask |= (__lasx_xvpickve2gr_w(result, 4) << 2);
return mask;
}
inline int v_signmask(const v_uint64x4& a)
@ -1592,7 +1524,7 @@ OPENCV_HAL_IMPL_LASX_MULADD(v_float64x4, d)
inline v_int32x8 v_fma(const v_int32x8& a, const v_int32x8& b, const v_int32x8& c)
{
return a * b + c;
return v_int32x8(__lasx_xvmadd_w(c.val, a.val, b.val));
}
inline v_int32x8 v_muladd(const v_int32x8& a, const v_int32x8& b, const v_int32x8& c)
@ -1601,17 +1533,10 @@ inline v_int32x8 v_muladd(const v_int32x8& a, const v_int32x8& b, const v_int32x
}
inline v_float32x8 v_invsqrt(const v_float32x8& x)
{
v_float32x8 half = x * v_float32x8(0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5);
v_float32x8 t = v_float32x8(__lasx_xvfrsqrt_s(x.val));
t *= v_float32x8(1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5) - ((t * t) * half);
return t;
}
{ return v_float32x8(__lasx_xvfrsqrt_s(x.val)); }
inline v_float64x4 v_invsqrt(const v_float64x4& x)
{
return v_float64x4(1., 1., 1., 1.) / v_sqrt(x);
}
{ return v_float64x4(__lasx_xvfrsqrt_d(x.val)); }
/** Absolute values **/
#define OPENCV_HAL_IMPL_LASX_ABS(_Tpvec, suffix) \
@ -1629,28 +1554,18 @@ inline v_float64x4 v_abs(const v_float64x4& x)
/** Absolute difference **/
inline v_uint8x32 v_absdiff(const v_uint8x32& a, const v_uint8x32& b)
{ return v_add_wrap(a - b, b - a); }
{ return (v_uint8x32)__lasx_xvabsd_bu(a.val, b.val); }
inline v_uint16x16 v_absdiff(const v_uint16x16& a, const v_uint16x16& b)
{ return v_add_wrap(a - b, b - a); }
{ return (v_uint16x16)__lasx_xvabsd_hu(a.val, b.val); }
inline v_uint32x8 v_absdiff(const v_uint32x8& a, const v_uint32x8& b)
{ return v_max(a, b) - v_min(a, b); }
{ return (v_uint32x8)__lasx_xvabsd_wu(a.val, b.val); }
inline v_uint8x32 v_absdiff(const v_int8x32& a, const v_int8x32& b)
{
v_int8x32 d = v_sub_wrap(a, b);
v_int8x32 m = a < b;
return v_reinterpret_as_u8(v_sub_wrap(d ^ m, m));
}
{ return (v_uint8x32)__lasx_xvabsd_b(a.val, b.val); }
inline v_uint16x16 v_absdiff(const v_int16x16& a, const v_int16x16& b)
{ return v_reinterpret_as_u16(v_sub_wrap(v_max(a, b), v_min(a, b))); }
{ return (v_uint16x16)__lasx_xvabsd_h(a.val, b.val); }
inline v_uint32x8 v_absdiff(const v_int32x8& a, const v_int32x8& b)
{
v_int32x8 d = a - b;
v_int32x8 m = a < b;
return v_reinterpret_as_u32((d ^ m) - m);
}
{ return (v_uint32x8)__lasx_xvabsd_w(a.val, b.val); }
inline v_float32x8 v_absdiff(const v_float32x8& a, const v_float32x8& b)
{ return v_abs(a - b); }
@ -1740,28 +1655,8 @@ inline v_float64x4 v_cvt_f64_high(const v_float32x8& a)
return v_float64x4(__lasx_xvfcvtl_d_s((__m256)ahigh));
}
// from (Mysticial and wim) https://stackoverflow.com/q/41144668
inline v_float64x4 v_cvt_f64(const v_int64x4& v)
{
// constants encoded as floating-point
__m256i magic_i_lo = __lasx_xvreplgr2vr_d(0x4330000000000000);
__m256i magic_i_hi32 = __lasx_xvreplgr2vr_d(0x4530000080000000);
__m256i magic_i_all = __lasx_xvreplgr2vr_d(0x4530000080100000);
__m256d magic_d_all = _lasx_256_castsi256_pd(magic_i_all);
// Blend the 32 lowest significant bits of v with magic_int_lo
__m256i mask = _v256_set_w(0, -1, 0, -1, 0, -1, 0, -1);
__m256i v_lo = __lasx_xvbitsel_v(magic_i_lo, v.val, mask);
// Extract the 32 most significant bits of v
__m256i v_hi = __lasx_xvsrli_d(v.val, 32);
// Flip the msb of v_hi and blend with 0x45300000
v_hi = __lasx_xvxor_v(v_hi, magic_i_hi32);
// Compute in double precision
__m256d v_hi_dbl = __lasx_xvfsub_d(_lasx_256_castsi256_pd(v_hi), magic_d_all);
// (v_hi - magic_d_all) + v_lo Do not assume associativity of floating point addition
__m256d result = __lasx_xvfadd_d(v_hi_dbl, _lasx_256_castsi256_pd(v_lo));
return v_float64x4(result);
}
{ return v_float64x4(__lasx_xvffint_d_l(v.val)); }
////////////// Lookup table access ////////////////////
@ -1967,11 +1862,9 @@ inline v_float32x8 v_interleave_pairs(const v_float32x8& vec)
inline v_int8x32 v_pack_triplets(const v_int8x32& vec)
{
__m256i vzero = __lasx_xvreplgr2vr_w(0);
__m256i t1 = __lasx_xvshuf_b(vec.val, vec.val,
_v256_set_d(0xffffff0f0e0d0c0a, 0x0908060504020100, 0xffffff0f0e0d0c0a, 0x0908060504020100));
__m256i t2 = __lasx_xvshuf_b(vzero, t1,
_v256_set_d(0x1211100c0b0a0908, 0x0706050403020100, 0x1211100c0b0a0908, 0x0706050403020100));
return v_int8x32(__lasx_xvperm_w(t2,
__m256i t1 = __lasx_xvshuf_b(vzero, vec.val,
_v256_set_d(0x1211100f0e0d0c0a, 0x0908060504020100, 0x1211100f0e0d0c0a, 0x0908060504020100));
return v_int8x32(__lasx_xvperm_w(t1,
_v256_set_d(0x0000000700000007, 0x0000000600000005, 0x0000000400000002, 0x0000000100000000)));
}
inline v_uint8x32 v_pack_triplets(const v_uint8x32& vec)
@ -1980,11 +1873,9 @@ inline v_uint8x32 v_pack_triplets(const v_uint8x32& vec)
inline v_int16x16 v_pack_triplets(const v_int16x16& vec)
{
__m256i vzero = __lasx_xvreplgr2vr_w(0);
__m256i t1 = __lasx_xvshuf_b(vec.val, vec.val,
_v256_set_d(0xffff0f0e0d0c0b0a, 0x0908050403020100, 0xffff0f0e0d0c0b0a, 0x0908050403020100));
__m256i t2 = __lasx_xvshuf_b(vzero, t1,
_v256_set_d(0x11100d0c0b0a0908, 0x0706050403020100, 0x11100d0c0b0a0908, 0x0706050403020100));
return v_int16x16(__lasx_xvperm_w(t2,
__m256i t1 = __lasx_xvshuf_b(vzero, vec.val,
_v256_set_d(0x11100f0e0d0c0b0a, 0x0908050403020100, 0x11100f0e0d0c0b0a, 0x0908050403020100));
return v_int16x16(__lasx_xvperm_w(t1,
_v256_set_d(0x0000000700000007, 0x0000000600000005, 0x0000000400000002, 0x0000000100000000)));
}
inline v_uint16x16 v_pack_triplets(const v_uint16x16& vec)
@ -2018,24 +1909,21 @@ inline v_int32x8 v_dotprod(const v_int16x16& a, const v_int16x16& b, const v_int
inline v_int64x4 v_dotprod(const v_int32x8& a, const v_int32x8& b)
{
__m256i even = __lasx_xvmulwev_d_w(a.val, b.val);
__m256i odd = __lasx_xvmulwod_d_w(a.val, b.val);
return v_int64x4(__lasx_xvadd_d(even, odd));
return v_int64x4(__lasx_xvmaddwod_d_w(even, a.val, b.val));
}
inline v_int64x4 v_dotprod(const v_int32x8& a, const v_int32x8& b, const v_int64x4& c)
{ return v_dotprod(a, b) + c; }
{
__m256i even = __lasx_xvmaddwev_d_w(c.val, a.val, b.val);
return v_int64x4(__lasx_xvmaddwod_d_w(even, a.val, b.val));
}
// 8 >> 32
inline v_uint32x8 v_dotprod_expand(const v_uint8x32& a, const v_uint8x32& b)
{
__m256i even_m = __lasx_xvreplgr2vr_w(0xFF00FF00);
__m256i even_a = __lasx_xvbitsel_v(a.val, __lasx_xvreplgr2vr_d(0), even_m);
__m256i odd_a = __lasx_xvsrli_h(a.val, 8);
__m256i even_b = __lasx_xvbitsel_v(b.val, __lasx_xvreplgr2vr_d(0), even_m);
__m256i odd_b = __lasx_xvsrli_h(b.val, 8);
__m256i prod0 = __lasx_xvadd_w(__lasx_xvmulwev_w_h(even_a, even_b), __lasx_xvmulwod_w_h(even_a, even_b));
__m256i prod1 = __lasx_xvadd_w(__lasx_xvmulwev_w_h(odd_a, odd_b),__lasx_xvmulwod_w_h(odd_a, odd_b));
__m256i even = __lasx_xvmulwev_h_bu(a.val, b.val);
__m256i odd = __lasx_xvmulwod_h_bu(a.val, b.val);
__m256i prod0 = __lasx_xvhaddw_wu_hu(even, even);
__m256i prod1 = __lasx_xvhaddw_wu_hu(odd, odd);
return v_uint32x8(__lasx_xvadd_w(prod0, prod1));
}
inline v_uint32x8 v_dotprod_expand(const v_uint8x32& a, const v_uint8x32& b, const v_uint32x8& c)
@ -2043,14 +1931,10 @@ inline v_uint32x8 v_dotprod_expand(const v_uint8x32& a, const v_uint8x32& b, con
inline v_int32x8 v_dotprod_expand(const v_int8x32& a, const v_int8x32& b)
{
__m256i even_a = __lasx_xvsrai_h(__lasx_xvbsll_v(a.val, 1), 8);
__m256i odd_a = __lasx_xvsrai_h(a.val, 8);
__m256i even_b = __lasx_xvsrai_h(__lasx_xvbsll_v(b.val, 1), 8);
__m256i odd_b = __lasx_xvsrai_h(b.val, 8);
__m256i prod0 = __lasx_xvadd_w(__lasx_xvmulwev_w_h(even_a, even_b), __lasx_xvmulwod_w_h(even_a, even_b));
__m256i prod1 = __lasx_xvadd_w(__lasx_xvmulwev_w_h(odd_a, odd_b),__lasx_xvmulwod_w_h(odd_a, odd_b));
__m256i even = __lasx_xvmulwev_h_b(a.val, b.val);
__m256i odd = __lasx_xvmulwod_h_b(a.val, b.val);
__m256i prod0 = __lasx_xvhaddw_w_h(even, even);
__m256i prod1 = __lasx_xvhaddw_w_h(odd, odd);
return v_int32x8(__lasx_xvadd_w(prod0, prod1));
}
inline v_int32x8 v_dotprod_expand(const v_int8x32& a, const v_int8x32& b, const v_int32x8& c)
@ -2059,36 +1943,24 @@ inline v_int32x8 v_dotprod_expand(const v_int8x32& a, const v_int8x32& b, const
// 16 >> 64
inline v_uint64x4 v_dotprod_expand(const v_uint16x16& a, const v_uint16x16& b)
{
__m256i mullo = __lasx_xvmul_h(a.val, b.val);
__m256i mulhi = __lasx_xvmuh_hu(a.val, b.val);
__m256i mul0 = __lasx_xvilvl_h(mulhi, mullo);
__m256i mul1 = __lasx_xvilvh_h(mulhi, mullo);
__m256i p02 = __lasx_xvbitsel_v(mul0, __lasx_xvreplgr2vr_d(0), _v256_set_w(-1, 0, -1, 0, -1, 0, -1, 0));
__m256i p13 = __lasx_xvsrli_d(mul0, 32);
__m256i p46 = __lasx_xvbitsel_v(mul1, __lasx_xvreplgr2vr_d(0), _v256_set_w(-1, 0, -1, 0, -1, 0, -1, 0));
__m256i p57 = __lasx_xvsrli_d(mul1, 32);
__m256i p15_ = __lasx_xvadd_d(p02, p13);
__m256i p9d_ = __lasx_xvadd_d(p46, p57);
return v_uint64x4(__lasx_xvadd_d(
__lasx_xvilvl_d(p9d_, p15_),
__lasx_xvilvh_d(p9d_, p15_)));
__m256i even = __lasx_xvmulwev_w_hu(a.val, b.val);
__m256i odd = __lasx_xvmulwod_w_hu(a.val, b.val);
__m256i prod0 = __lasx_xvhaddw_du_wu(even, even);
__m256i prod1 = __lasx_xvhaddw_du_wu(odd, odd);
return v_uint64x4(__lasx_xvadd_d(prod0, prod1));
}
inline v_uint64x4 v_dotprod_expand(const v_uint16x16& a, const v_uint16x16& b, const v_uint64x4& c)
{ return v_dotprod_expand(a, b) + c; }
inline v_int64x4 v_dotprod_expand(const v_int16x16& a, const v_int16x16& b)
{
__m256i prod = __lasx_xvadd_w(__lasx_xvmulwev_w_h(a.val, b.val), __lasx_xvmulwod_w_h(a.val, b.val));
__m256i sign = __lasx_xvsrai_w(prod, 31);
__m256i lo = __lasx_xvilvl_w(sign, prod);
__m256i hi = __lasx_xvilvh_w(sign, prod);
return v_int64x4(__lasx_xvadd_d(__lasx_xvilvl_d(hi, lo), __lasx_xvilvh_d(hi, lo)));
__m256i even = __lasx_xvmulwev_w_h(a.val, b.val);
__m256i odd = __lasx_xvmulwod_w_h(a.val, b.val);
__m256i prod0 = __lasx_xvhaddw_d_w(even, even);
__m256i prod1 = __lasx_xvhaddw_d_w(odd, odd);
return v_int64x4(__lasx_xvadd_d(prod0, prod1));
}
inline v_int64x4 v_dotprod_expand(const v_int16x16& a, const v_int16x16& b, const v_int64x4& c)
{ return v_dotprod_expand(a, b) + c; }
@ -2126,20 +1998,11 @@ inline v_int32x8 v_dotprod_expand_fast(const v_int8x32& a, const v_int8x32& b, c
// 16 >> 64
inline v_uint64x4 v_dotprod_expand_fast(const v_uint16x16& a, const v_uint16x16& b)
{
__m256i mullo = __lasx_xvmul_h(a.val, b.val);
__m256i mulhi = __lasx_xvmuh_hu(a.val, b.val);
__m256i mul0 = __lasx_xvilvl_h(mulhi, mullo);
__m256i mul1 = __lasx_xvilvh_h(mulhi, mullo);
__m256i p02 = __lasx_xvbitsel_v(mul0, __lasx_xvreplgr2vr_d(0), _v256_set_w(-1, 0, -1, 0, -1, 0, -1, 0));
__m256i p13 = __lasx_xvsrli_d(mul0, 32);
__m256i p46 = __lasx_xvbitsel_v(mul1, __lasx_xvreplgr2vr_d(0), _v256_set_w(-1, 0, -1, 0, -1, 0, -1, 0));
__m256i p57 = __lasx_xvsrli_d(mul1, 32);
__m256i p15_ = __lasx_xvadd_d(p02, p13);
__m256i p9d_ = __lasx_xvadd_d(p46, p57);
return v_uint64x4(__lasx_xvadd_d(p15_, p9d_));
__m256i even = __lasx_xvmulwev_w_hu(a.val, b.val);
__m256i odd = __lasx_xvmulwod_w_hu(a.val, b.val);
__m256i prod0 = __lasx_xvhaddw_du_wu(even, even);
__m256i prod1 = __lasx_xvhaddw_du_wu(odd, odd);
return v_uint64x4(__lasx_xvadd_d(__lasx_xvilvl_d(prod1, prod0), __lasx_xvilvh_d(prod1, prod0)));
}
inline v_uint64x4 v_dotprod_expand_fast(const v_uint16x16& a, const v_uint16x16& b, const v_uint64x4& c)
{ return v_dotprod_expand_fast(a, b) + c; }
@ -2261,12 +2124,7 @@ inline v_int8x32 v_pack(const v_int16x16& a, const v_int16x16& b)
{ return v_int8x32(_v256_shuffle_odd_64(_lasx_packs_h(a.val, b.val))); }
inline v_uint8x32 v_pack(const v_uint16x16& a, const v_uint16x16& b)
{
__m256i t = __lasx_xvreplgr2vr_h(255);
__m256i a1 = __lasx_xvmin_hu(a.val, t);
__m256i b1 = __lasx_xvmin_hu(b.val, t);
return v_uint8x32(_v256_shuffle_odd_64(_lasx_packus_h(a1, b1)));
}
{ return v_uint8x32(_v256_shuffle_odd_64(__lasx_xvssrlrni_bu_h(b.val, a.val, 0))); }
inline v_uint8x32 v_pack_u(const v_int16x16& a, const v_int16x16& b)
{
@ -2276,13 +2134,8 @@ inline v_uint8x32 v_pack_u(const v_int16x16& a, const v_int16x16& b)
inline void v_pack_store(schar* ptr, const v_int16x16& a)
{ v_store_low(ptr, v_pack(a, a)); }
inline void v_pack_store(uchar* ptr, const v_uint16x16& a)
{
const __m256i m = __lasx_xvreplgr2vr_h(255);
__m256i am = __lasx_xvmin_hu(a.val, m);
am = _v256_shuffle_odd_64(_lasx_packus_h(am, am));
v_store_low(ptr, v_uint8x32(am));
}
inline void v_pack_store(uchar *ptr, const v_uint16x16& a)
{ v_store_low(ptr, v_pack(a, a)); }
inline void v_pack_u_store(uchar* ptr, const v_int16x16& a)
{ v_store_low(ptr, v_pack_u(a, a)); }
@ -2290,45 +2143,43 @@ inline void v_pack_u_store(uchar* ptr, const v_int16x16& a)
template<int n> inline
v_uint8x32 v_rshr_pack(const v_uint16x16& a, const v_uint16x16& b)
{
// we assume that n > 0, and so the shifted 16-bit values can be treated as signed numbers.
v_uint16x16 delta = v256_setall_u16((short)(1 << (n-1)));
return v_pack_u(v_reinterpret_as_s16((a + delta) >> n),
v_reinterpret_as_s16((b + delta) >> n));
__m256i res = __lasx_xvssrlrni_bu_h(b.val, a.val, n);
return v_uint8x32(_v256_shuffle_odd_64(res));
}
template<int n> inline
void v_rshr_pack_store(uchar* ptr, const v_uint16x16& a)
{
v_uint16x16 delta = v256_setall_u16((short)(1 << (n-1)));
v_pack_u_store(ptr, v_reinterpret_as_s16((a + delta) >> n));
__m256i res = __lasx_xvssrlrni_bu_h(a.val, a.val, n);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
template<int n> inline
v_uint8x32 v_rshr_pack_u(const v_int16x16& a, const v_int16x16& b)
{
v_int16x16 delta = v256_setall_s16((short)(1 << (n-1)));
return v_pack_u((a + delta) >> n, (b + delta) >> n);
__m256i res = __lasx_xvssrarni_bu_h(b.val, a.val, n);
return v_uint8x32(_v256_shuffle_odd_64(res));
}
template<int n> inline
void v_rshr_pack_u_store(uchar* ptr, const v_int16x16& a)
{
v_int16x16 delta = v256_setall_s16((short)(1 << (n-1)));
v_pack_u_store(ptr, (a + delta) >> n);
__m256i res = __lasx_xvssrarni_bu_h(a.val, a.val, n);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
template<int n> inline
v_int8x32 v_rshr_pack(const v_int16x16& a, const v_int16x16& b)
{
v_int16x16 delta = v256_setall_s16((short)(1 << (n-1)));
return v_pack((a + delta) >> n, (b + delta) >> n);
__m256i res = __lasx_xvssrarni_b_h(b.val, a.val, n);
return v_int8x32(_v256_shuffle_odd_64(res));
}
template<int n> inline
void v_rshr_pack_store(schar* ptr, const v_int16x16& a)
{
v_int16x16 delta = v256_setall_s16((short)(1 << (n-1)));
v_pack_store(ptr, (a + delta) >> n);
__m256i res = __lasx_xvssrarni_b_h(a.val, a.val, n);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
// 32
@ -2346,67 +2197,51 @@ inline void v_pack_store(short* ptr, const v_int32x8& a)
inline void v_pack_store(ushort* ptr, const v_uint32x8& a)
{
const __m256i m = __lasx_xvreplgr2vr_w(65535);
__m256i am = __lasx_xvmin_wu(a.val, m);
am = _v256_shuffle_odd_64(_lasx_packus_w(am, am));
v_store_low(ptr, v_uint16x16(am));
__m256i res = __lasx_xvssrlrni_hu_w(a.val, a.val, 0);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
inline void v_pack_u_store(ushort* ptr, const v_int32x8& a)
{ v_store_low(ptr, v_pack_u(a, a)); }
template<int n> inline
v_uint16x16 v_rshr_pack(const v_uint32x8& a, const v_uint32x8& b)
{
// we assume that n > 0, and so the shifted 32-bit values can be treated as signed numbers.
v_uint32x8 delta = v256_setall_u32(1 << (n-1));
return v_pack_u(v_reinterpret_as_s32((a + delta) >> n),
v_reinterpret_as_s32((b + delta) >> n));
}
{ return v_uint16x16(_v256_shuffle_odd_64(__lasx_xvssrlrni_hu_w(b.val, a.val, n))); }
template<int n> inline
void v_rshr_pack_store(ushort* ptr, const v_uint32x8& a)
{
v_uint32x8 delta = v256_setall_u32(1 << (n-1));
v_pack_u_store(ptr, v_reinterpret_as_s32((a + delta) >> n));
__m256i res = __lasx_xvssrlrni_hu_w(a.val, a.val, n);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
template<int n> inline
v_uint16x16 v_rshr_pack_u(const v_int32x8& a, const v_int32x8& b)
{
v_int32x8 delta = v256_setall_s32(1 << (n-1));
return v_pack_u((a + delta) >> n, (b + delta) >> n);
}
{ return v_uint16x16(_v256_shuffle_odd_64(__lasx_xvssrarni_hu_w(b.val, a.val, n))); }
template<int n> inline
void v_rshr_pack_u_store(ushort* ptr, const v_int32x8& a)
{
v_int32x8 delta = v256_setall_s32(1 << (n-1));
v_pack_u_store(ptr, (a + delta) >> n);
__m256i res = __lasx_xvssrarni_hu_w(a.val, a.val, n);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
template<int n> inline
v_int16x16 v_rshr_pack(const v_int32x8& a, const v_int32x8& b)
{
v_int32x8 delta = v256_setall_s32(1 << (n-1));
return v_pack((a + delta) >> n, (b + delta) >> n);
}
{ return v_int16x16(_v256_shuffle_odd_64(__lasx_xvssrarni_h_w(b.val, a.val, n))); }
template<int n> inline
void v_rshr_pack_store(short* ptr, const v_int32x8& a)
{
v_int32x8 delta = v256_setall_s32(1 << (n-1));
v_pack_store(ptr, (a + delta) >> n);
__m256i res = __lasx_xvssrarni_h_w(a.val, a.val, n);
__lsx_vst(_v256_extract_low(_v256_shuffle_odd_64(res)), ptr, 0);
}
// 64
// Non-saturating pack
inline v_uint32x8 v_pack(const v_uint64x4& a, const v_uint64x4& b)
{
__m256i a0 = __lasx_xvshuf4i_w(a.val, 0x08);
__m256i b0 = __lasx_xvshuf4i_w(b.val, 0x08);
__m256i ab = __lasx_xvilvl_d(b0, a0);
__m256i ab = __lasx_xvpickev_w(b.val, a.val);
return v_uint32x8(_v256_shuffle_odd_64(ab));
}
@ -2424,31 +2259,19 @@ inline void v_pack_store(int* ptr, const v_int64x4& b)
template<int n> inline
v_uint32x8 v_rshr_pack(const v_uint64x4& a, const v_uint64x4& b)
{
v_uint64x4 delta = v256_setall_u64((uint64)1 << (n-1));
return v_pack((a + delta) >> n, (b + delta) >> n);
}
{ return v_uint32x8(_v256_shuffle_odd_64(__lasx_xvsrlrni_w_d(b.val, a.val, n))); }
template<int n> inline
void v_rshr_pack_store(unsigned* ptr, const v_uint64x4& a)
{
v_uint64x4 delta = v256_setall_u64((uint64)1 << (n-1));
v_pack_store(ptr, (a + delta) >> n);
}
{ __lsx_vst(_v256_shuffle_odd_64(__lasx_xvsrlrni_w_d(a.val, a.val, n)), ptr, 0); }
template<int n> inline
v_int32x8 v_rshr_pack(const v_int64x4& a, const v_int64x4& b)
{
v_int64x4 delta = v256_setall_s64((int64)1 << (n-1));
return v_pack((a + delta) >> n, (b + delta) >> n);
}
{ return v_int32x8(_v256_shuffle_odd_64(__lasx_xvsrarni_w_d(b.val, a.val, n))); }
template<int n> inline
void v_rshr_pack_store(int* ptr, const v_int64x4& a)
{
v_int64x4 delta = v256_setall_s64((int64)1 << (n-1));
v_pack_store(ptr, (a + delta) >> n);
}
{ __lsx_vst(_v256_shuffle_odd_64(__lasx_xvsrarni_w_d(a.val, a.val, n)), ptr, 0); }
// pack boolean
inline v_uint8x32 v_pack_b(const v_uint16x16& a, const v_uint16x16& b)
@ -2583,63 +2406,48 @@ template<int i>
inline v_float32x8 v_broadcast_element(const v_float32x8 &a)
{ return v_reinterpret_as_f32(v_broadcast_element<i>(v_reinterpret_as_u32(a))); }
///////////////////// load deinterleave /////////////////////////////
inline void v_load_deinterleave( const uchar* ptr, v_uint8x32& a, v_uint8x32& b )
inline void v_load_deinterleave(const uchar* ptr, v_uint8x32& a, v_uint8x32& b)
{
__m256i ab0 = __lasx_xvld(ptr, 0);
__m256i ab1 = __lasx_xvld(ptr + 32, 0);
const __m256i sh = _v256_setr_b(0, 2, 4, 6, 8, 10, 12, 14, 1, 3, 5, 7, 9, 11, 13, 15,
0, 2, 4, 6, 8, 10, 12, 14, 1, 3, 5, 7, 9, 11, 13, 15);
__m256i p0 = __lasx_xvshuf_b(ab0, ab0, sh);
__m256i p1 = __lasx_xvshuf_b(ab1, ab1, sh);
__m256i pl = __lasx_xvpermi_q(p0, p1, 0x02);
__m256i ph = __lasx_xvpermi_q(p0, p1, 0x13);
__m256i a0 = __lasx_xvilvl_d(ph, pl);
__m256i b0 = __lasx_xvilvh_d(ph, pl);
a = v_uint8x32(a0);
b = v_uint8x32(b0);
__m256i t0 = __lasx_xvld(ptr, 0);
__m256i t1 = __lasx_xvld(ptr, 32);
__m256i p0 = __lasx_xvpickev_b(t1, t0);
__m256i p1 = __lasx_xvpickod_b(t1, t0);
a.val = __lasx_xvpermi_d(p0, 0xd8);
b.val = __lasx_xvpermi_d(p1, 0xd8);
}
inline void v_load_deinterleave( const ushort* ptr, v_uint16x16& a, v_uint16x16& b )
{
__m256i ab0 = __lasx_xvld(ptr, 0);
__m256i ab1 = __lasx_xvld(ptr + 16, 0);
const __m256i sh = _v256_setr_b(0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15,
0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15);
__m256i p0 = __lasx_xvshuf_b(ab0, ab0, sh);
__m256i p1 = __lasx_xvshuf_b(ab1, ab1, sh);
__m256i pl = __lasx_xvpermi_q(p0, p1, 0x02);
__m256i ph = __lasx_xvpermi_q(p0, p1, 0x13);
__m256i a0 = __lasx_xvilvl_d(ph, pl);
__m256i b0 = __lasx_xvilvh_d(ph, pl);
a = v_uint16x16(a0);
b = v_uint16x16(b0);
__m256i t0 = __lasx_xvld(ptr, 0);
__m256i t1 = __lasx_xvld(ptr, 32);
__m256i p0 = __lasx_xvpickev_h(t1, t0);
__m256i p1 = __lasx_xvpickod_h(t1, t0);
a.val = __lasx_xvpermi_d(p0, 0xd8);
b.val = __lasx_xvpermi_d(p1, 0xd8);
}
inline void v_load_deinterleave( const unsigned* ptr, v_uint32x8& a, v_uint32x8& b )
{
__m256i ab0 = __lasx_xvld(ptr, 0);
__m256i ab1 = __lasx_xvld(ptr + 8, 0);
__m256i t0 = __lasx_xvld(ptr, 0);
__m256i t1 = __lasx_xvld(ptr, 32);
//const int sh = 0+2*4+1*16+3*64;
__m256i p0 = __lasx_xvshuf4i_w(ab0, 0xD8);
__m256i p1 = __lasx_xvshuf4i_w(ab1, 0xD8);
__m256i pl = __lasx_xvpermi_q(p0, p1, 0x02);
__m256i ph = __lasx_xvpermi_q(p0, p1, 0x13);
__m256i a0 = __lasx_xvilvl_d(ph, pl);
__m256i b0 = __lasx_xvilvh_d(ph, pl);
a = v_uint32x8(a0);
b = v_uint32x8(b0);
__m256i p0 = __lasx_xvpickev_w(t1, t0);
__m256i p1 = __lasx_xvpickod_w(t1, t0);
a.val = __lasx_xvpermi_d(p0, 0xd8);
b.val = __lasx_xvpermi_d(p1, 0xd8);
}
inline void v_load_deinterleave( const uint64* ptr, v_uint64x4& a, v_uint64x4& b )
{
__m256i ab0 = __lasx_xvld(ptr, 0);
__m256i ab1 = __lasx_xvld(ptr + 4, 0);
__m256i ab1 = __lasx_xvld(ptr, 32);
__m256i pl = __lasx_xvpermi_q(ab0, ab1, 0x02);
__m256i ph = __lasx_xvpermi_q(ab0, ab1, 0x13);
@ -2652,8 +2460,8 @@ inline void v_load_deinterleave( const uint64* ptr, v_uint64x4& a, v_uint64x4& b
inline void v_load_deinterleave( const uchar* ptr, v_uint8x32& a, v_uint8x32& b, v_uint8x32& c )
{
__m256i bgr0 = __lasx_xvld(ptr, 0);
__m256i bgr1 = __lasx_xvld(ptr + 32, 0);
__m256i bgr2 = __lasx_xvld(ptr + 64, 0);
__m256i bgr1 = __lasx_xvld(ptr, 32);
__m256i bgr2 = __lasx_xvld(ptr, 64);
__m256i s02_low = __lasx_xvpermi_q(bgr0, bgr2, 0x02);
__m256i s02_high = __lasx_xvpermi_q(bgr0, bgr2, 0x13);
@ -2686,8 +2494,8 @@ inline void v_load_deinterleave( const uchar* ptr, v_uint8x32& a, v_uint8x32& b,
inline void v_load_deinterleave( const ushort* ptr, v_uint16x16& a, v_uint16x16& b, v_uint16x16& c )
{
__m256i bgr0 = __lasx_xvld(ptr, 0);
__m256i bgr1 = __lasx_xvld(ptr + 16, 0);
__m256i bgr2 = __lasx_xvld(ptr + 32, 0);
__m256i bgr1 = __lasx_xvld(ptr, 32);
__m256i bgr2 = __lasx_xvld(ptr, 64);
__m256i s02_low = __lasx_xvpermi_q(bgr0, bgr2, 0x02);
__m256i s02_high = __lasx_xvpermi_q(bgr0, bgr2, 0x13);
@ -2717,8 +2525,8 @@ inline void v_load_deinterleave( const ushort* ptr, v_uint16x16& a, v_uint16x16&
inline void v_load_deinterleave( const unsigned* ptr, v_uint32x8& a, v_uint32x8& b, v_uint32x8& c )
{
__m256i bgr0 = __lasx_xvld(ptr, 0);
__m256i bgr1 = __lasx_xvld(ptr + 8, 0);
__m256i bgr2 = __lasx_xvld(ptr + 16, 0);
__m256i bgr1 = __lasx_xvld(ptr, 32);
__m256i bgr2 = __lasx_xvld(ptr, 64);
__m256i s02_low = __lasx_xvpermi_q(bgr0, bgr2, 0x02);
__m256i s02_high = __lasx_xvpermi_q(bgr0, bgr2, 0x13);
@ -2741,8 +2549,8 @@ inline void v_load_deinterleave( const unsigned* ptr, v_uint32x8& a, v_uint32x8&
inline void v_load_deinterleave( const uint64* ptr, v_uint64x4& a, v_uint64x4& b, v_uint64x4& c )
{
__m256i bgr0 = __lasx_xvld(ptr, 0);
__m256i bgr1 = __lasx_xvld(ptr + 4, 0);
__m256i bgr2 = __lasx_xvld(ptr + 8, 0);
__m256i bgr1 = __lasx_xvld(ptr, 32);
__m256i bgr2 = __lasx_xvld(ptr, 64);
__m256i s01 = __lasx_xvpermi_q(bgr0, bgr1, 0x12); // get bgr0 low 128 and bgr1 high 128
__m256i s12 = __lasx_xvpermi_q(bgr1, bgr2, 0x12);
@ -2756,81 +2564,60 @@ inline void v_load_deinterleave( const uint64* ptr, v_uint64x4& a, v_uint64x4& b
c = v_uint64x4(r0);
}
inline void v_load_deinterleave( const uchar* ptr, v_uint8x32& a, v_uint8x32& b, v_uint8x32& c, v_uint8x32& d )
inline void v_load_deinterleave(const uchar* ptr, v_uint8x32& a, v_uint8x32& b, v_uint8x32& c, v_uint8x32& d)
{
__m256i bgr0 = __lasx_xvld(ptr, 0);
__m256i bgr1 = __lasx_xvld(ptr + 32, 0);
__m256i bgr2 = __lasx_xvld(ptr + 64, 0);
__m256i bgr3 = __lasx_xvld(ptr + 96, 0);
const __m256i sh = _v256_setr_b(0, 4, 8, 12, 1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15,
0, 4, 8, 12, 1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15);
__m256i t0 = __lasx_xvld(ptr, 0);
__m256i t1 = __lasx_xvld(ptr, 32);
__m256i t2 = __lasx_xvld(ptr, 64);
__m256i t3 = __lasx_xvld(ptr, 96);
__m256i p0 = __lasx_xvshuf_b(bgr0, bgr0, sh);
__m256i p1 = __lasx_xvshuf_b(bgr1, bgr1, sh);
__m256i p2 = __lasx_xvshuf_b(bgr2, bgr2, sh);
__m256i p3 = __lasx_xvshuf_b(bgr3, bgr3, sh);
const __m256i sh = _v256_setr_w(0, 4, 1, 5, 2, 6, 3, 7);
__m256i ac_lo = __lasx_xvpickev_b(t1, t0);
__m256i bd_lo = __lasx_xvpickod_b(t1, t0);
__m256i ac_hi = __lasx_xvpickev_b(t3, t2);
__m256i bd_hi = __lasx_xvpickod_b(t3, t2);
__m256i p01l = __lasx_xvilvl_w(p1, p0);
__m256i p01h = __lasx_xvilvh_w(p1, p0);
__m256i p23l = __lasx_xvilvl_w(p3, p2);
__m256i p23h = __lasx_xvilvh_w(p3, p2);
__m256i pll = __lasx_xvpermi_q(p01l, p23l, 0x02);
__m256i plh = __lasx_xvpermi_q(p01l, p23l, 0x13);
__m256i phl = __lasx_xvpermi_q(p01h, p23h, 0x02);
__m256i phh = __lasx_xvpermi_q(p01h, p23h, 0x13);
__m256i b0 = __lasx_xvilvl_w(plh, pll);
__m256i g0 = __lasx_xvilvh_w(plh, pll);
__m256i r0 = __lasx_xvilvl_w(phh, phl);
__m256i a0 = __lasx_xvilvh_w(phh, phl);
__m256i a_pre = __lasx_xvpickev_b(ac_hi, ac_lo);
__m256i c_pre = __lasx_xvpickod_b(ac_hi, ac_lo);
__m256i b_pre = __lasx_xvpickev_b(bd_hi, bd_lo);
__m256i d_pre = __lasx_xvpickod_b(bd_hi, bd_lo);
a = v_uint8x32(b0);
b = v_uint8x32(g0);
c = v_uint8x32(r0);
d = v_uint8x32(a0);
a.val = __lasx_xvperm_w(a_pre, sh);
b.val = __lasx_xvperm_w(b_pre, sh);
c.val = __lasx_xvperm_w(c_pre, sh);
d.val = __lasx_xvperm_w(d_pre, sh);
}
inline void v_load_deinterleave( const ushort* ptr, v_uint16x16& a, v_uint16x16& b, v_uint16x16& c, v_uint16x16& d )
inline void v_load_deinterleave(const ushort* ptr, v_uint16x16& a, v_uint16x16& b, v_uint16x16& c, v_uint16x16& d)
{
__m256i bgr0 = __lasx_xvld(ptr, 0);
__m256i bgr1 = __lasx_xvld(ptr + 16, 0);
__m256i bgr2 = __lasx_xvld(ptr + 32, 0);
__m256i bgr3 = __lasx_xvld(ptr + 48, 0);
const __m256i sh = _v256_setr_b(0, 1, 8, 9, 2, 3, 10, 11, 4, 5, 12, 13, 6, 7, 14, 15,
0, 1, 8, 9, 2, 3, 10, 11, 4, 5, 12, 13, 6, 7, 14, 15);
__m256i p0 = __lasx_xvshuf_b(bgr0, bgr0, sh);
__m256i p1 = __lasx_xvshuf_b(bgr1, bgr1, sh);
__m256i p2 = __lasx_xvshuf_b(bgr2, bgr2, sh);
__m256i p3 = __lasx_xvshuf_b(bgr3, bgr3, sh);
__m256i t0 = __lasx_xvld(ptr, 0);
__m256i t1 = __lasx_xvld(ptr, 32);
__m256i t2 = __lasx_xvld(ptr, 64);
__m256i t3 = __lasx_xvld(ptr, 96);
__m256i p01l = __lasx_xvilvl_w(p1, p0);
__m256i p01h = __lasx_xvilvh_w(p1, p0);
__m256i p23l = __lasx_xvilvl_w(p3, p2);
__m256i p23h = __lasx_xvilvh_w(p3, p2);
const __m256i sh = _v256_setr_w(0, 4, 1, 5, 2, 6, 3, 7);
__m256i ac_lo = __lasx_xvpickev_h(t1, t0);
__m256i bd_lo = __lasx_xvpickod_h(t1, t0);
__m256i ac_hi = __lasx_xvpickev_h(t3, t2);
__m256i bd_hi = __lasx_xvpickod_h(t3, t2);
__m256i pll = __lasx_xvpermi_q(p01l, p23l, 0x02);
__m256i plh = __lasx_xvpermi_q(p01l, p23l, 0x13);
__m256i phl = __lasx_xvpermi_q(p01h, p23h, 0x02);
__m256i phh = __lasx_xvpermi_q(p01h, p23h, 0x13);
__m256i a_pre = __lasx_xvpickev_h(ac_hi, ac_lo);
__m256i c_pre = __lasx_xvpickod_h(ac_hi, ac_lo);
__m256i b_pre = __lasx_xvpickev_h(bd_hi, bd_lo);
__m256i d_pre = __lasx_xvpickod_h(bd_hi, bd_lo);
__m256i b0 = __lasx_xvilvl_w(plh, pll);
__m256i g0 = __lasx_xvilvh_w(plh, pll);
__m256i r0 = __lasx_xvilvl_w(phh, phl);
__m256i a0 = __lasx_xvilvh_w(phh, phl);
a = v_uint16x16(b0);
b = v_uint16x16(g0);
c = v_uint16x16(r0);
d = v_uint16x16(a0);
a.val = __lasx_xvperm_w(a_pre, sh);
b.val = __lasx_xvperm_w(b_pre, sh);
c.val = __lasx_xvperm_w(c_pre, sh);
d.val = __lasx_xvperm_w(d_pre, sh);
}
inline void v_load_deinterleave( const unsigned* ptr, v_uint32x8& a, v_uint32x8& b, v_uint32x8& c, v_uint32x8& d )
{
__m256i p0 = __lasx_xvld(ptr, 0);
__m256i p1 = __lasx_xvld(ptr + 8, 0);
__m256i p2 = __lasx_xvld(ptr + 16, 0);
__m256i p3 = __lasx_xvld(ptr + 24, 0);
__m256i p1 = __lasx_xvld(ptr, 32);
__m256i p2 = __lasx_xvld(ptr, 64);
__m256i p3 = __lasx_xvld(ptr, 96);
__m256i p01l = __lasx_xvilvl_w(p1, p0);
__m256i p01h = __lasx_xvilvh_w(p1, p0);
@ -2856,9 +2643,9 @@ inline void v_load_deinterleave( const unsigned* ptr, v_uint32x8& a, v_uint32x8&
inline void v_load_deinterleave( const uint64* ptr, v_uint64x4& a, v_uint64x4& b, v_uint64x4& c, v_uint64x4& d )
{
__m256i bgra0 = __lasx_xvld(ptr, 0);
__m256i bgra1 = __lasx_xvld(ptr + 4, 0);
__m256i bgra2 = __lasx_xvld(ptr + 8, 0);
__m256i bgra3 = __lasx_xvld(ptr + 12, 0);
__m256i bgra1 = __lasx_xvld(ptr, 32);
__m256i bgra2 = __lasx_xvld(ptr, 64);
__m256i bgra3 = __lasx_xvld(ptr, 96);
__m256i l02 = __lasx_xvpermi_q(bgra0, bgra2, 0x02);
__m256i h02 = __lasx_xvpermi_q(bgra0, bgra2, 0x13);

2536
modules/core/include/opencv2/core/hal/intrin_lsx.hpp
View File

File diff suppressed because it is too large Load Diff

5
modules/core/src/system.cpp
Unescape View File

@ -427,6 +427,7 @@ struct HWFeatures
g_hwFeatureNames[CPU_RVV] = "RVV";
g_hwFeatureNames[CPU_LSX] = "LSX";
g_hwFeatureNames[CPU_LASX] = "LASX";
}
@ -703,6 +704,10 @@ struct HWFeatures
have[CV_CPU_RVV] = true;
#endif
#if defined __loongarch_sx
have[CV_CPU_LSX] = true;
#endif
#if defined __loongarch_asx
have[CV_CPU_LASX] = true;
#endif

Write Preview
Loading…
Cancel
Save