/external/libvpx/libvpx/vpx_dsp/arm/ |
D | transpose_neon.h | 168 const uint32x4x2_t b0 = in transpose_u16_4x4q() 451 const uint32x4x2_t c0 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[0]), in transpose_u16_8x4() 453 const uint32x4x2_t c1 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[1]), in transpose_u16_8x4() 556 const uint32x4x2_t d0 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_8x8() 558 const uint32x4x2_t d1 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[1]), in transpose_u8_8x8() 680 const uint32x4x2_t c0 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[0]), in transpose_u16_8x8() 682 const uint32x4x2_t c1 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[1]), in transpose_u16_8x8() 684 const uint32x4x2_t c2 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[0]), in transpose_u16_8x8() 686 const uint32x4x2_t c3 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[1]), in transpose_u16_8x8() 847 const uint32x4x2_t d0 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_16x8() [all …]
|
/external/libgav1/libgav1/src/dsp/arm/ |
D | loop_restoration_neon.cc | 665 inline uint32x4x2_t Sum3_32(const uint32x4x2_t src[3]) { in Sum3_32() 666 uint32x4x2_t d; in Sum3_32() 722 inline uint32x4x2_t Sum5_32(const uint32x4x2_t src[5]) { in Sum5_32() 723 uint32x4x2_t d; in Sum5_32() 759 inline uint32x4x2_t Sum3WHorizontal(const uint16x8_t src[2]) { in Sum3WHorizontal() 761 uint32x4x2_t sum; in Sum3WHorizontal() 793 inline uint32x4x2_t Sum5WHorizontal(const uint16x8_t src[2]) { in Sum5WHorizontal() 796 uint32x4x2_t sum; in Sum5WHorizontal() 834 void SumHorizontal(const uint16x8_t sq[2], uint32x4x2_t* const row_sq3, in SumHorizontal() 835 uint32x4x2_t* const row_sq5) { in SumHorizontal() [all …]
|
D | common_neon.h | 623 const uint32x4x2_t d0 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[0]), in Transpose8x8() 625 const uint32x4x2_t d1 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[1]), in Transpose8x8() 649 const uint32x4x2_t c0 = vuzpq_u32(vreinterpretq_u32_u16(b0.val[0]), in Transpose8x8() 651 const uint32x4x2_t c1 = vuzpq_u32(vreinterpretq_u32_u16(b0.val[1]), in Transpose8x8() 716 const uint32x4x2_t c0 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[0]), in Transpose8x8() 718 const uint32x4x2_t c1 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[1]), in Transpose8x8() 720 const uint32x4x2_t c2 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[0]), in Transpose8x8() 722 const uint32x4x2_t c3 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[1]), in Transpose8x8() 798 const uint32x4x2_t d0 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[0]), in Transpose8x16() 800 const uint32x4x2_t d1 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[0]), in Transpose8x16() [all …]
|
D | intrapred_directional_neon.cc | 369 const uint32x4x2_t d0 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[0]), in DirectionalZone3_WxH() 371 const uint32x4x2_t d1 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[1]), in DirectionalZone3_WxH()
|
/external/libaom/libaom/av1/common/arm/ |
D | transpose_neon.h | 55 const uint32x4x2_t d0 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_8x8() 57 const uint32x4x2_t d1 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[1]), in transpose_u8_8x8() 356 const uint32x4x2_t c0 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[0]), in transpose_u16_8x8() 358 const uint32x4x2_t c1 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[1]), in transpose_u16_8x8() 360 const uint32x4x2_t c2 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[0]), in transpose_u16_8x8() 362 const uint32x4x2_t c3 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[1]), in transpose_u16_8x8()
|
/external/llvm-project/llvm/test/CodeGen/Thumb2/mve-intrinsics/ |
D | vld24.ll | 6 %struct.uint32x4x2_t = type { [2 x <4 x i32>] } 81 define arm_aapcs_vfpcc void @test_vst2q_u32(i32* %addr, %struct.uint32x4x2_t %value.coerce) { 90 %value.coerce.fca.0.0.extract = extractvalue %struct.uint32x4x2_t %value.coerce, 0, 0 91 %value.coerce.fca.0.1.extract = extractvalue %struct.uint32x4x2_t %value.coerce, 0, 1 97 define arm_aapcs_vfpcc i32* @test_vst2q_u32_post(i32* %addr, %struct.uint32x4x2_t %value.coerce) { 106 %value.coerce.fca.0.0.extract = extractvalue %struct.uint32x4x2_t %value.coerce, 0, 0 107 %value.coerce.fca.0.1.extract = extractvalue %struct.uint32x4x2_t %value.coerce, 0, 1
|
/external/XNNPACK/src/x32-zip/ |
D | xm-neon.c | 46 const uint32x4x2_t vxy = vzipq_u32(vx, vy); in xnn_x32_zip_xm_ukernel__neon() 47 const uint32x4x2_t vzw = vzipq_u32(vz, vw); in xnn_x32_zip_xm_ukernel__neon()
|
D | x2-neon.c | 26 uint32x4x2_t vxy; in xnn_x32_zip_x2_ukernel__neon()
|
/external/llvm-project/llvm/test/CodeGen/ARM/ |
D | arm-vlddup.ll | 24 %struct.uint32x4x2_t = type { <4 x i32>, <4 x i32> } 49 declare %struct.uint32x4x2_t @llvm.arm.neon.vld2dup.v4i32.p0i8(i8*, i32) 167 define %struct.uint32x4x2_t @test_vld2q_dup_u32(i8* %src) { 169 %tmp = tail call %struct.uint32x4x2_t @llvm.arm.neon.vld2dup.v4i32.p0i8(i8* %src, i32 4) 170 ret %struct.uint32x4x2_t %tmp
|
D | arm-vld1.ll | 24 %struct.uint32x4x2_t = type { <4 x i32>, <4 x i32> } 56 declare %struct.uint32x4x2_t @llvm.arm.neon.vld1x2.v4i32.p0i32(i32*) nounwind readonly 177 define %struct.uint32x4x2_t @test_vld1q_u32_x2(i32* %a) nounwind { 178 %tmp = tail call %struct.uint32x4x2_t @llvm.arm.neon.vld1x2.v4i32.p0i32(i32* %a) 179 ret %struct.uint32x4x2_t %tmp
|
D | arm-vst1.ll | 24 ; %struct.uint32x4x2_t = type { <4 x i32>, <4 x i32> } 51 %struct.uint32x4x2_t = type { [2 x <4 x i32>] } 262 define void @test_vst1q_u32_x2(i32* %a, %struct.uint32x4x2_t %b) nounwind { 264 %b0 = extractvalue %struct.uint32x4x2_t %b, 0, 0 265 %b1 = extractvalue %struct.uint32x4x2_t %b, 0, 1
|
/external/llvm-project/clang/test/CodeGen/arm-mve-intrinsics/ |
D | vld24.c | 55 void test_vst2q_u32(uint32_t *addr, uint32x4x2_t value) in test_vst2q_u32()
|
/external/llvm-project/clang/test/CodeGen/ |
D | arm-neon-vld.c | 1255 uint32x4x2_t test_vld1q_u32_x2(uint32_t const *a) { in test_vld1q_u32_x2() 2153 void test_vld2q_dup_u32(uint32x4x2_t *dest, const uint32_t *src) { in test_vld2q_dup_u32()
|
D | aarch64-neon-perm.c | 1205 uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) { in test_vuzpq_u32() 1589 uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) { in test_vzipq_u32() 1973 uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) { in test_vtrnq_u32()
|
D | arm-neon-vst.c | 2057 void test_vst1q_u32_x2(uint32_t *a, uint32x4x2_t b) { in test_vst1q_u32_x2()
|
D | aarch64-neon-ldst-one.c | 1160 uint32x4x2_t test_vld2q_lane_u32(uint32_t *a, uint32x4x2_t b) { in test_vld2q_lane_u32() 4473 void test_vst2q_lane_u32(uint32_t *a, uint32x4x2_t b) { in test_vst2q_lane_u32()
|
/external/clang/test/CodeGen/ |
D | aarch64-neon-perm.c | 1272 uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) { in test_vuzpq_u32() 1736 uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) { in test_vzipq_u32() 2200 uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) { in test_vtrnq_u32()
|
D | aarch64-neon-ldst-one.c | 353 uint32x4x2_t test_vld2q_dup_u32(uint32_t *a) { in test_vld2q_dup_u32() 2338 uint32x4x2_t test_vld2q_lane_u32(uint32_t *a, uint32x4x2_t b) { in test_vld2q_lane_u32() 5651 void test_vst2q_lane_u32(uint32_t *a, uint32x4x2_t b) { in test_vst2q_lane_u32()
|
/external/libjpeg-turbo/simd/arm/ |
D | jidctfst-neon.c | 435 uint32x4x2_t cols_0415 = vzipq_u32(vreinterpretq_u32_u8(cols_01), in jsimd_idct_ifast_neon() 437 uint32x4x2_t cols_2637 = vzipq_u32(vreinterpretq_u32_u8(cols_23), in jsimd_idct_ifast_neon()
|
/external/libvpx/libvpx/vp8/common/arm/neon/ |
D | vp8_loopfilter_neon.c | 326 uint32x4x2_t q2tmp0, q2tmp1, q2tmp2, q2tmp3; in vp8_loop_filter_vertical_edge_y_neon() 438 uint32x4x2_t q2tmp0, q2tmp1, q2tmp2, q2tmp3; in vp8_loop_filter_vertical_edge_uv_neon()
|
D | mbloopfilter_neon.c | 298 uint32x4x2_t q2tmp0, q2tmp1, q2tmp2, q2tmp3; in vp8_mbloop_filter_vertical_edge_y_neon() 462 uint32x4x2_t q2tmp0, q2tmp1, q2tmp2, q2tmp3; in vp8_mbloop_filter_vertical_edge_uv_neon()
|
/external/webp/src/dsp/ |
D | dec_neon.c | 202 const uint32x4x2_t row04 = vtrnq_u32(vreinterpretq_u32_u16(row02.val[0]), in Load8x8x2T_NEON() 204 const uint32x4x2_t row26 = vtrnq_u32(vreinterpretq_u32_u16(row02.val[1]), in Load8x8x2T_NEON() 206 const uint32x4x2_t row15 = vtrnq_u32(vreinterpretq_u32_u16(row13.val[0]), in Load8x8x2T_NEON() 208 const uint32x4x2_t row37 = vtrnq_u32(vreinterpretq_u32_u16(row13.val[1]), in Load8x8x2T_NEON()
|
/external/libaom/libaom/aom_dsp/simd/ |
D | v128_intrinsics_arm.h | 546 uint32x4x2_t r = in v128_unziplo_32() 557 uint32x4x2_t r = in v128_unziphi_32()
|
/external/llvm/test/CodeGen/AArch64/ |
D | neon-perm.ll | 17 %struct.uint32x4x2_t = type { [2 x <4 x i32>] } 2646 define %struct.uint32x4x2_t @test_vuzpq_u32(<4 x i32> %a, <4 x i32> %b) { 2653 %.fca.0.0.insert = insertvalue %struct.uint32x4x2_t undef, <4 x i32> %vuzp.i, 0, 0 2654 %.fca.0.1.insert = insertvalue %struct.uint32x4x2_t %.fca.0.0.insert, <4 x i32> %vuzp1.i, 0, 1 2655 ret %struct.uint32x4x2_t %.fca.0.1.insert 2862 define %struct.uint32x4x2_t @test_vzipq_u32(<4 x i32> %a, <4 x i32> %b) { 2869 %.fca.0.0.insert = insertvalue %struct.uint32x4x2_t undef, <4 x i32> %vzip.i, 0, 0 2870 %.fca.0.1.insert = insertvalue %struct.uint32x4x2_t %.fca.0.0.insert, <4 x i32> %vzip1.i, 0, 1 2871 ret %struct.uint32x4x2_t %.fca.0.1.insert 3078 define %struct.uint32x4x2_t @test_vtrnq_u32(<4 x i32> %a, <4 x i32> %b) { [all …]
|
/external/llvm-project/llvm/test/CodeGen/AArch64/ |
D | neon-perm.ll | 17 %struct.uint32x4x2_t = type { [2 x <4 x i32>] } 2646 define %struct.uint32x4x2_t @test_vuzpq_u32(<4 x i32> %a, <4 x i32> %b) { 2653 %.fca.0.0.insert = insertvalue %struct.uint32x4x2_t undef, <4 x i32> %vuzp.i, 0, 0 2654 %.fca.0.1.insert = insertvalue %struct.uint32x4x2_t %.fca.0.0.insert, <4 x i32> %vuzp1.i, 0, 1 2655 ret %struct.uint32x4x2_t %.fca.0.1.insert 2862 define %struct.uint32x4x2_t @test_vzipq_u32(<4 x i32> %a, <4 x i32> %b) { 2869 %.fca.0.0.insert = insertvalue %struct.uint32x4x2_t undef, <4 x i32> %vzip.i, 0, 0 2870 %.fca.0.1.insert = insertvalue %struct.uint32x4x2_t %.fca.0.0.insert, <4 x i32> %vzip1.i, 0, 1 2871 ret %struct.uint32x4x2_t %.fca.0.1.insert 3078 define %struct.uint32x4x2_t @test_vtrnq_u32(<4 x i32> %a, <4 x i32> %b) { [all …]
|