/external/libpng/arm/ |
D | filter_neon_intrinsics.c | 79 uint8x8x2_t *vrpt = png_ptr(uint8x8x2_t, &vtmp); in png_read_filter_row_sub3_neon() 80 uint8x8x2_t vrp = *vrpt; in png_read_filter_row_sub3_neon() 102 vrpt = png_ptr(uint8x8x2_t, &vtmp); in png_read_filter_row_sub3_neon() 159 uint8x8x2_t *vrpt; in png_read_filter_row_avg3_neon() 160 uint8x8x2_t vrp; in png_read_filter_row_avg3_neon() 165 vrpt = png_ptr(uint8x8x2_t,&vtmp); in png_read_filter_row_avg3_neon() 174 uint8x8x2_t *vppt; in png_read_filter_row_avg3_neon() 175 uint8x8x2_t vpp; in png_read_filter_row_avg3_neon() 180 vppt = png_ptr(uint8x8x2_t,&vtmp); in png_read_filter_row_avg3_neon() 196 vrpt = png_ptr(uint8x8x2_t,&vtmp); in png_read_filter_row_avg3_neon() [all …]
|
/external/pdfium/third_party/libpng16/arm/ |
D | filter_neon_intrinsics.c | 79 uint8x8x2_t *vrpt = png_ptr(uint8x8x2_t, &vtmp); in png_read_filter_row_sub3_neon() 80 uint8x8x2_t vrp = *vrpt; in png_read_filter_row_sub3_neon() 102 vrpt = png_ptr(uint8x8x2_t, &vtmp); in png_read_filter_row_sub3_neon() 159 uint8x8x2_t *vrpt; in png_read_filter_row_avg3_neon() 160 uint8x8x2_t vrp; in png_read_filter_row_avg3_neon() 165 vrpt = png_ptr(uint8x8x2_t,&vtmp); in png_read_filter_row_avg3_neon() 174 uint8x8x2_t *vppt; in png_read_filter_row_avg3_neon() 175 uint8x8x2_t vpp; in png_read_filter_row_avg3_neon() 180 vppt = png_ptr(uint8x8x2_t,&vtmp); in png_read_filter_row_avg3_neon() 196 vrpt = png_ptr(uint8x8x2_t,&vtmp); in png_read_filter_row_avg3_neon() [all …]
|
/external/libvpx/libvpx/vp8/common/arm/neon/ |
D | loopfiltersimpleverticaledge_neon.c | 19 const uint8x8x2_t result) { in write_2x4() 29 const uint8x8x2_t r01_u8 = vtrn_u8(result.val[0], result.val[1]); in write_2x4() 50 const uint8x8x2_t result, in write_2x8() 51 const uint8x8x2_t result2) { in write_2x8() 58 const uint8x8x2_t result, in write_2x8() 59 const uint8x8x2_t result2) { in write_2x8() 118 const uint8x8x2_t r01_u8 = vtrn_u8(vreinterpret_u8_u16(r02_u16.val[0]), in read_4x8() 120 const uint8x8x2_t r23_u8 = vtrn_u8(vreinterpret_u8_u16(r02_u16.val[1]), in read_4x8() 186 uint8x8x2_t d2u8x2; // d12, d13 in vp8_loop_filter_simple_vertical_edge_neon() 187 uint8x8x2_t d3u8x2; // d14, d15 in vp8_loop_filter_simple_vertical_edge_neon()
|
D | vp8_loopfilter_neon.c | 275 const uint8x8x2_t r01_u8 = vtrn_u8(vreinterpret_u8_u16(r02_u16.val[0]), in write_4x8() 277 const uint8x8x2_t r23_u8 = vtrn_u8(vreinterpret_u8_u16(r02_u16.val[1]), in write_4x8()
|
/external/libgav1/libgav1/src/dsp/arm/ |
D | loop_restoration_neon.cc | 38 inline uint8x8_t VshrU128(const uint8x8x2_t src) { in VshrU128() 262 inline uint8x8x2_t WienerVerticalTap7Kernel2(const int16_t* const wiener_buffer, in WienerVerticalTap7Kernel2() 267 uint8x8x2_t d; in WienerVerticalTap7Kernel2() 285 uint8x8x2_t d[2]; in WienerVerticalTap7() 329 inline uint8x8x2_t WienerVerticalTap5Kernel2(const int16_t* const wiener_buffer, in WienerVerticalTap5Kernel2() 334 uint8x8x2_t d; in WienerVerticalTap5Kernel2() 351 uint8x8x2_t d[2]; in WienerVerticalTap5() 392 inline uint8x8x2_t WienerVerticalTap3Kernel2(const int16_t* const wiener_buffer, in WienerVerticalTap3Kernel2() 397 uint8x8x2_t d; in WienerVerticalTap3Kernel2() 413 uint8x8x2_t d[2]; in WienerVerticalTap3() [all …]
|
D | loop_filter_neon.cc | 44 const uint8x8x2_t a = Interleave32(p0q0, p1q1); in OuterThreshold() 163 const uint8x8x2_t q0p1xp0q1 = Interleave32(Transpose32(p0q0), p1q1); in Horizontal4_NEON() 197 const uint8x8x2_t p1q1xq0p0 = Interleave32(p1q0, Transpose32(p0q1)); in Vertical4_NEON() 198 const uint8x8x2_t p1q1xp0q0 = {p1q1xq0p0.val[0], in Vertical4_NEON() 235 const uint8x8x2_t p1p0xq1q0 = Interleave32(p1q1_output, p0q0_output); in Vertical4_NEON() 360 const uint8x8x2_t q0p1xp0q1 = Interleave32(Transpose32(p0q0), p1q1); in Horizontal6_NEON() 408 const uint8x8x2_t p2q2xq1p1 = Interleave32(p2q1, Transpose32(p1q2)); in Vertical6_NEON() 434 const uint8x8x2_t q0p1xp0q1 = Interleave32(Transpose32(p0q0), p1q1); in Vertical6_NEON() 462 const uint8x8x2_t p1p0xq1q0 = Interleave32(p1q1_output, p0q0_output); in Vertical6_NEON() 608 const uint8x8x2_t q0p1xp0q1 = Interleave32(Transpose32(p0q0), p1q1); in Horizontal8_NEON() [all …]
|
D | common_neon.h | 304 const uint8x8x2_t b = {vget_low_u8(a), vget_high_u8(a)}; in VQTbl1U8() 409 inline uint8x8x2_t Interleave32(const uint8x8_t a, const uint8x8_t b) { in Interleave32() 413 const uint8x8x2_t d = {vreinterpret_u8_u32(c.val[0]), in Interleave32() 461 const uint8x8x2_t e = in Transpose4x4() 480 const uint8x8x2_t b0 = vtrn_u8(*a0, *a1); in Transpose8x4() 481 const uint8x8x2_t b1 = vtrn_u8(*a2, *a3); in Transpose8x4()
|
D | intra_edge_neon.cc | 250 const uint8x8x2_t output = {Upsample(src0, src1, src2, src3), src2}; in IntraEdgeUpsampler_NEON() 265 const uint8x8x2_t output_lo = {result_lo, vget_low_u8(src2)}; in IntraEdgeUpsampler_NEON() 274 const uint8x8x2_t output_hi = {result_hi, vget_high_u8(src2)}; in IntraEdgeUpsampler_NEON()
|
D | intrapred_directional_neon.cc | 853 const uint8x8x2_t b0 = vtrn_u8(value_v[0], value_v[1]); in DirectionalIntraPredictorZone3_NEON() 854 const uint8x8x2_t b1 = vtrn_u8(value_v[2], value_v[3]); in DirectionalIntraPredictorZone3_NEON()
|
/external/XNNPACK/src/x8-zip/ |
D | x2-neon.c | 25 uint8x8x2_t vxy; in xnn_x8_zip_x2_ukernel__neon() 33 uint8x8x2_t vxy; in xnn_x8_zip_x2_ukernel__neon()
|
D | xm-neon.c | 42 const uint8x8x2_t vxy = vzip_u8(vx, vy); in xnn_x8_zip_xm_ukernel__neon() 43 const uint8x8x2_t vzw = vzip_u8(vz, vw); in xnn_x8_zip_xm_ukernel__neon() 85 const uint8x8x2_t vxy = vzip_u8(vreinterpret_u8_u64(vx), vreinterpret_u8_u64(vy)); in xnn_x8_zip_xm_ukernel__neon() 86 const uint8x8x2_t vzw = vzip_u8(vreinterpret_u8_u64(vz), vreinterpret_u8_u64(vw)); in xnn_x8_zip_xm_ukernel__neon()
|
/external/libaom/libaom/av1/common/arm/ |
D | transpose_neon.h | 83 const uint8x8x2_t b0 = vtrn_u8(*a0, *a1); in transpose_u8_8x4() 84 const uint8x8x2_t b1 = vtrn_u8(*a2, *a3); in transpose_u8_8x4() 125 const uint8x8x2_t d0 = in transpose_u8_4x4() 177 const uint8x8x2_t d0 = in transpose_u8_4x8() 179 const uint8x8x2_t d1 = in transpose_u8_4x8()
|
/external/swiftshader/third_party/llvm-7.0/llvm/test/CodeGen/ARM/ |
D | arm-vlddup.ll | 16 %struct.uint8x8x2_t = type { <8 x i8>, <8 x i8> } 32 declare %struct.uint8x8x2_t @llvm.arm.neon.vld2dup.v8i8.p0i8(i8*, i32) 85 define %struct.uint8x8x2_t @test_vld2_dup_u8(i8* %src) { 87 %tmp = tail call %struct.uint8x8x2_t @llvm.arm.neon.vld2dup.v8i8.p0i8(i8* %src, i32 1) 88 ret %struct.uint8x8x2_t %tmp
|
D | arm-vld1.ll | 16 %struct.uint8x8x2_t = type { <8 x i8>, <8 x i8> } 48 declare %struct.uint8x8x2_t @llvm.arm.neon.vld1x2.v8i8.p0i8(i8*) nounwind readonly 133 define %struct.uint8x8x2_t @test_vld1_u8_x2(i8* %a) nounwind { 134 %tmp = tail call %struct.uint8x8x2_t @llvm.arm.neon.vld1x2.v8i8.p0i8(i8* %a) 135 ret %struct.uint8x8x2_t %tmp
|
D | arm-vst1.ll | 16 ; %struct.uint8x8x2_t = type { <8 x i8>, <8 x i8> } 45 %struct.uint8x8x2_t = type { [2 x <8 x i8>] } 194 define void @test_vst1_u8_x2(i8* %a, %struct.uint8x8x2_t %b) nounwind { 196 %b0 = extractvalue %struct.uint8x8x2_t %b, 0, 0 197 %b1 = extractvalue %struct.uint8x8x2_t %b, 0, 1
|
D | vuzp.ll | 531 %struct.uint8x8x2_t = type { [2 x <8 x i8>] } 532 define %struct.uint8x8x2_t @vuzp_extract_subvector(<16 x i8> %t) #0 { 545 %.fca.0.0.insert = insertvalue %struct.uint8x8x2_t undef, <8 x i8> %vuzp.i, 0, 0 546 %.fca.0.1.insert = insertvalue %struct.uint8x8x2_t %.fca.0.0.insert, <8 x i8> %vuzp1.i, 0, 1 547 ret %struct.uint8x8x2_t %.fca.0.1.insert
|
/external/webp/src/dsp/ |
D | upsampling_neon.c | 53 uint8x8x2_t A_B, C_D; \ 123 const uint8x8x2_t rgba4444 = ZIP_U8(rg, ba); \ 131 const uint8x8x2_t rgb565 = ZIP_U8(rg, gb); \
|
/external/libaom/libaom/aom_dsp/simd/ |
D | v64_intrinsics_arm.h | 374 uint8x8x2_t r = vzip_u8(vreinterpret_u8_s64(y), vreinterpret_u8_s64(x)); in v64_ziplo_8() 384 uint8x8x2_t r = vzip_u8(vreinterpret_u8_s64(y), vreinterpret_u8_s64(x)); in v64_ziphi_8() 470 uint8x8x2_t r = vuzp_u8(vreinterpret_u8_s64(y), vreinterpret_u8_s64(x)); in v64_unziplo_8() 480 uint8x8x2_t r = vuzp_u8(vreinterpret_u8_s64(y), vreinterpret_u8_s64(x)); in v64_unziphi_8()
|
/external/clang/test/CodeGen/ |
D | aarch64-neon-tbl.c | 539 uint8x8_t test_vtbl2_u8(uint8x8x2_t a, uint8x8_t b) { in test_vtbl2_u8() 781 uint8x8_t test_vtbx2_u8(uint8x8_t a, uint8x8x2_t b, uint8x8_t c) { in test_vtbx2_u8()
|
D | aarch64-neon-perm.c | 987 uint8x8x2_t test_vuzp_u8(uint8x8_t a, uint8x8_t b) { in test_vuzp_u8() 1451 uint8x8x2_t test_vzip_u8(uint8x8_t a, uint8x8_t b) { in test_vzip_u8() 1915 uint8x8x2_t test_vtrn_u8(uint8x8_t a, uint8x8_t b) { in test_vtrn_u8()
|
D | aarch64-neon-ldst-one.c | 563 uint8x8x2_t test_vld2_dup_u8(uint8_t *a) { in test_vld2_dup_u8() 2673 uint8x8x2_t test_vld2_lane_u8(uint8_t *a, uint8x8x2_t b) { in test_vld2_lane_u8() 5936 void test_vst2_lane_u8(uint8_t *a, uint8x8x2_t b) { in test_vst2_lane_u8()
|
/external/libvpx/libvpx/vpx_dsp/arm/ |
D | transpose_neon.h | 89 const uint8x8x2_t d0 = in transpose_u8_4x4() 236 const uint8x8x2_t d0 = in transpose_u8_4x8() 238 const uint8x8x2_t d1 = in transpose_u8_4x8() 409 const uint8x8x2_t b0 = vtrn_u8(*a0, *a1); in transpose_u8_8x4() 410 const uint8x8x2_t b1 = vtrn_u8(*a2, *a3); in transpose_u8_8x4()
|
/external/llvm/test/CodeGen/AArch64/ |
D | neon-perm.ll | 6 %struct.uint8x8x2_t = type { [2 x <8 x i8>] } 2514 define %struct.uint8x8x2_t @test_vuzp_u8(<8 x i8> %a, <8 x i8> %b) { 2521 %.fca.0.0.insert = insertvalue %struct.uint8x8x2_t undef, <8 x i8> %vuzp.i, 0, 0 2522 %.fca.0.1.insert = insertvalue %struct.uint8x8x2_t %.fca.0.0.insert, <8 x i8> %vuzp1.i, 0, 1 2523 ret %struct.uint8x8x2_t %.fca.0.1.insert 2730 define %struct.uint8x8x2_t @test_vzip_u8(<8 x i8> %a, <8 x i8> %b) { 2737 %.fca.0.0.insert = insertvalue %struct.uint8x8x2_t undef, <8 x i8> %vzip.i, 0, 0 2738 %.fca.0.1.insert = insertvalue %struct.uint8x8x2_t %.fca.0.0.insert, <8 x i8> %vzip1.i, 0, 1 2739 ret %struct.uint8x8x2_t %.fca.0.1.insert 2946 define %struct.uint8x8x2_t @test_vtrn_u8(<8 x i8> %a, <8 x i8> %b) { [all …]
|
/external/swiftshader/third_party/llvm-7.0/llvm/test/CodeGen/AArch64/ |
D | neon-perm.ll | 6 %struct.uint8x8x2_t = type { [2 x <8 x i8>] } 2514 define %struct.uint8x8x2_t @test_vuzp_u8(<8 x i8> %a, <8 x i8> %b) { 2521 %.fca.0.0.insert = insertvalue %struct.uint8x8x2_t undef, <8 x i8> %vuzp.i, 0, 0 2522 %.fca.0.1.insert = insertvalue %struct.uint8x8x2_t %.fca.0.0.insert, <8 x i8> %vuzp1.i, 0, 1 2523 ret %struct.uint8x8x2_t %.fca.0.1.insert 2730 define %struct.uint8x8x2_t @test_vzip_u8(<8 x i8> %a, <8 x i8> %b) { 2737 %.fca.0.0.insert = insertvalue %struct.uint8x8x2_t undef, <8 x i8> %vzip.i, 0, 0 2738 %.fca.0.1.insert = insertvalue %struct.uint8x8x2_t %.fca.0.0.insert, <8 x i8> %vzip1.i, 0, 1 2739 ret %struct.uint8x8x2_t %.fca.0.1.insert 2946 define %struct.uint8x8x2_t @test_vtrn_u8(<8 x i8> %a, <8 x i8> %b) { [all …]
|
/external/libaom/libaom/aom_dsp/arm/ |
D | sse_neon.c | 54 uint8x8x2_t tmp, tmp2; in aom_sse_neon()
|