/external/libvpx/libvpx/vpx_dsp/arm/ |
D | idct32x32_135_add_neon.c | 93 int16x8_t in[12], s1[32], s2[32], s3[32], s4[32], s5[32], s6[32], s7[32]; in vpx_idct32_12_neon() local 205 s5[0] = vaddq_s16(s4[0], s4[3]); in vpx_idct32_12_neon() 206 s5[1] = vaddq_s16(s4[0], s4[2]); in vpx_idct32_12_neon() 207 s5[2] = vsubq_s16(s4[0], s4[2]); in vpx_idct32_12_neon() 208 s5[3] = vsubq_s16(s4[0], s4[3]); in vpx_idct32_12_neon() 210 s5[5] = sub_multiply_shift_and_narrow_s16(s3[7], s3[4], cospi_16_64); in vpx_idct32_12_neon() 211 s5[6] = add_multiply_shift_and_narrow_s16(s3[4], s3[7], cospi_16_64); in vpx_idct32_12_neon() 213 s5[8] = vaddq_s16(s2[8], s3[11]); in vpx_idct32_12_neon() 214 s5[9] = vaddq_s16(s4[9], s4[10]); in vpx_idct32_12_neon() 215 s5[10] = vsubq_s16(s4[9], s4[10]); in vpx_idct32_12_neon() [all …]
|
D | highbd_idct32x32_135_add_neon.c | 101 int32x4x2_t in[12], s1[32], s2[32], s3[32], s4[32], s5[32], s6[32], s7[32], in vpx_highbd_idct32_12_neon() local 216 s5[0] = highbd_idct_add_dual(s4[0], s4[3]); in vpx_highbd_idct32_12_neon() 217 s5[1] = highbd_idct_add_dual(s4[0], s4[2]); in vpx_highbd_idct32_12_neon() 218 s5[2] = highbd_idct_sub_dual(s4[0], s4[2]); in vpx_highbd_idct32_12_neon() 219 s5[3] = highbd_idct_sub_dual(s4[0], s4[3]); in vpx_highbd_idct32_12_neon() 221 s5[5] = sub_multiply_shift_and_narrow_s32_dual(s3[7], s3[4], cospi_16_64); in vpx_highbd_idct32_12_neon() 222 s5[6] = add_multiply_shift_and_narrow_s32_dual(s3[4], s3[7], cospi_16_64); in vpx_highbd_idct32_12_neon() 224 s5[8] = highbd_idct_add_dual(s2[8], s3[11]); in vpx_highbd_idct32_12_neon() 225 s5[9] = highbd_idct_add_dual(s4[9], s4[10]); in vpx_highbd_idct32_12_neon() 226 s5[10] = highbd_idct_sub_dual(s4[9], s4[10]); in vpx_highbd_idct32_12_neon() [all …]
|
D | highbd_vpx_convolve8_neon.c | 47 int16x8_t *const s4, int16x8_t *const s5, in load_8x8() argument 59 *s5 = vld1q_s16(s); in load_8x8() 69 const uint16x8_t s4, const uint16x8_t s5, in store_8x8() argument 81 vst1q_u16(s, s5); in store_8x8() 90 const int16x4_t s3, const int16x4_t s4, const int16x4_t s5, in highbd_convolve8_4() argument 101 sum = vmlal_lane_s16(sum, s5, filters_hi, 1); in highbd_convolve8_4() 109 const int16x8_t s3, const int16x8_t s4, const int16x8_t s5, in highbd_convolve8_8() argument 122 sum0 = vmlal_lane_s16(sum0, vget_low_s16(s5), filters_hi, 1); in highbd_convolve8_8() 130 sum1 = vmlal_lane_s16(sum1, vget_high_s16(s5), filters_hi, 1); in highbd_convolve8_8() 157 int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10; in vpx_highbd_convolve8_horiz_neon() local [all …]
|
D | vpx_convolve8_neon.c | 36 const uint8x8_t s4, const uint8x8_t s5, in store_u8_8x8() argument 48 vst1_u8(s, s5); in store_u8_8x8() 75 int16x4_t filter3, filter4, s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10, d0, in vpx_convolve8_horiz_neon() local 96 s5 = vget_high_s16(tt1); in vpx_convolve8_horiz_neon() 116 d0 = convolve8_4(s0, s1, s2, s3, s4, s5, s6, s7, filters, filter3, in vpx_convolve8_horiz_neon() 118 d1 = convolve8_4(s1, s2, s3, s4, s5, s6, s7, s8, filters, filter3, in vpx_convolve8_horiz_neon() 120 d2 = convolve8_4(s2, s3, s4, s5, s6, s7, s8, s9, filters, filter3, in vpx_convolve8_horiz_neon() 122 d3 = convolve8_4(s3, s4, s5, s6, s7, s8, s9, s10, filters, filter3, in vpx_convolve8_horiz_neon() 139 s1 = s5; in vpx_convolve8_horiz_neon() 143 s5 = s9; in vpx_convolve8_horiz_neon() [all …]
|
/external/llvm-project/llvm/test/MC/AMDGPU/ |
D | smem.s | 235 s_scratch_load_dword s5, s[2:3], s101 240 s_scratch_load_dword s5, s[2:3], s0 glc 308 s_atomic_add s5, s[2:3], s101 313 s_atomic_add s5, s[2:3], 0x0 318 s_atomic_add s5, s[2:3], s0 glc 328 s_atomic_and s5, s[2:3], s101 368 s_atomic_dec s5, s[2:3], s0 glc 378 s_atomic_inc s5, s[2:3], s0 glc 388 s_atomic_or s5, s[2:3], 0x0 398 s_atomic_smax s5, s[2:3], s101 [all …]
|
D | sop2.s | 241 s_lshl1_add_u32 s5, s1, s2 246 s_lshl1_add_u32 s5, -1, s2 251 s_lshl1_add_u32 s5, s1, 0 256 s_lshl1_add_u32 s5, s1, 0x3f717273 266 s_lshl2_add_u32 s5, 0xaf123456, s2 271 s_lshl3_add_u32 s5, 0x3f717273, s2 276 s_lshl3_add_u32 s5, s1, s101 281 s_lshl4_add_u32 s5, s1, 0xaf123456 286 s_lshl4_add_u32 s5, -1, s2 291 s_mul_hi_i32 s5, s101, s2 [all …]
|
/external/llvm-project/llvm/test/CodeGen/AMDGPU/ |
D | cc-update.ll | 30 ; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5 37 ; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0 47 ; GFX1010-NEXT: s_addc_u32 s5, s5, 0 49 ; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5 68 ; GFX803-NEXT: s_mov_b32 flat_scratch_lo, s5 71 ; GFX803-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12 79 ; GFX900-NEXT: s_addc_u32 flat_scratch_hi, s5, 0 84 ; GFX900-NEXT: s_addc_u32 s5, s5, ex@rel32@hi+12 93 ; GFX1010-NEXT: s_addc_u32 s5, s5, 0 95 ; GFX1010-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s5 [all …]
|
D | llvm.maxnum.f16.ll | 28 ; SI-NEXT: s_mov_b32 s1, s5 56 ; VI-NEXT: s_mov_b32 s1, s5 81 ; GFX9-NEXT: s_mov_b32 s1, s5 113 ; SI-NEXT: s_mov_b32 s1, s5 129 ; VI-NEXT: s_mov_b32 s1, s5 131 ; VI-NEXT: s_mov_b32 s5, s7 148 ; GFX9-NEXT: s_mov_b32 s1, s5 150 ; GFX9-NEXT: s_mov_b32 s5, s7 181 ; SI-NEXT: s_mov_b32 s1, s5 197 ; VI-NEXT: s_mov_b32 s1, s5 [all …]
|
D | llvm.minnum.f16.ll | 28 ; SI-NEXT: s_mov_b32 s1, s5 56 ; VI-NEXT: s_mov_b32 s1, s5 81 ; GFX9-NEXT: s_mov_b32 s1, s5 136 ; SI-NEXT: s_mov_b32 s1, s5 152 ; VI-NEXT: s_mov_b32 s1, s5 154 ; VI-NEXT: s_mov_b32 s5, s7 171 ; GFX9-NEXT: s_mov_b32 s1, s5 173 ; GFX9-NEXT: s_mov_b32 s5, s7 204 ; SI-NEXT: s_mov_b32 s1, s5 220 ; VI-NEXT: s_mov_b32 s1, s5 [all …]
|
/external/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/ |
D | llvm.amdgcn.image.store.2d.ll | 12 ; GFX6-NEXT: s_mov_b32 s3, s5 14 ; GFX6-NEXT: s_mov_b32 s5, s7 25 ; GFX8-NEXT: s_mov_b32 s3, s5 27 ; GFX8-NEXT: s_mov_b32 s5, s7 38 ; GFX10-NEXT: s_mov_b32 s3, s5 40 ; GFX10-NEXT: s_mov_b32 s5, s7 55 ; GFX6-NEXT: s_mov_b32 s3, s5 57 ; GFX6-NEXT: s_mov_b32 s5, s7 68 ; GFX8-NEXT: s_mov_b32 s3, s5 70 ; GFX8-NEXT: s_mov_b32 s5, s7 [all …]
|
D | llvm.amdgcn.image.load.1d.ll | 12 ; GFX6-NEXT: s_mov_b32 s3, s5 14 ; GFX6-NEXT: s_mov_b32 s5, s7 26 ; GFX8-NEXT: s_mov_b32 s3, s5 28 ; GFX8-NEXT: s_mov_b32 s5, s7 40 ; GFX10-NEXT: s_mov_b32 s3, s5 42 ; GFX10-NEXT: s_mov_b32 s5, s7 58 ; GFX6-NEXT: s_mov_b32 s3, s5 60 ; GFX6-NEXT: s_mov_b32 s5, s7 72 ; GFX8-NEXT: s_mov_b32 s3, s5 74 ; GFX8-NEXT: s_mov_b32 s5, s7 [all …]
|
D | llvm.amdgcn.image.atomic.dim.ll | 13 ; GFX6-NEXT: s_mov_b32 s3, s5 15 ; GFX6-NEXT: s_mov_b32 s5, s7 27 ; GFX8-NEXT: s_mov_b32 s3, s5 29 ; GFX8-NEXT: s_mov_b32 s5, s7 41 ; GFX10-NEXT: s_mov_b32 s3, s5 43 ; GFX10-NEXT: s_mov_b32 s5, s7 61 ; GFX6-NEXT: s_mov_b32 s3, s5 63 ; GFX6-NEXT: s_mov_b32 s5, s7 75 ; GFX8-NEXT: s_mov_b32 s3, s5 77 ; GFX8-NEXT: s_mov_b32 s5, s7 [all …]
|
D | llvm.amdgcn.image.getresinfo.ll | 12 ; GFX6-NEXT: s_mov_b32 s3, s5 14 ; GFX6-NEXT: s_mov_b32 s5, s7 26 ; GFX8-NEXT: s_mov_b32 s3, s5 28 ; GFX8-NEXT: s_mov_b32 s5, s7 40 ; GFX10-NEXT: s_mov_b32 s3, s5 42 ; GFX10-NEXT: s_mov_b32 s5, s7 59 ; GFX6-NEXT: s_mov_b32 s3, s5 61 ; GFX6-NEXT: s_mov_b32 s5, s7 73 ; GFX8-NEXT: s_mov_b32 s3, s5 75 ; GFX8-NEXT: s_mov_b32 s5, s7 [all …]
|
D | llvm.amdgcn.image.load.1d.d16.ll | 13 ; GFX8-UNPACKED-NEXT: s_mov_b32 s3, s5 15 ; GFX8-UNPACKED-NEXT: s_mov_b32 s5, s7 27 ; GFX8-PACKED-NEXT: s_mov_b32 s3, s5 29 ; GFX8-PACKED-NEXT: s_mov_b32 s5, s7 41 ; GFX9-NEXT: s_mov_b32 s3, s5 43 ; GFX9-NEXT: s_mov_b32 s5, s7 55 ; GFX10-NEXT: s_mov_b32 s3, s5 57 ; GFX10-NEXT: s_mov_b32 s5, s7 73 ; GFX8-UNPACKED-NEXT: s_mov_b32 s3, s5 75 ; GFX8-UNPACKED-NEXT: s_mov_b32 s5, s7 [all …]
|
D | llvm.amdgcn.image.atomic.dim.a16.ll | 11 ; GFX9-NEXT: s_mov_b32 s3, s5 13 ; GFX9-NEXT: s_mov_b32 s5, s7 25 ; GFX10-NEXT: s_mov_b32 s3, s5 27 ; GFX10-NEXT: s_mov_b32 s5, s7 45 ; GFX9-NEXT: s_mov_b32 s3, s5 47 ; GFX9-NEXT: s_mov_b32 s5, s7 59 ; GFX10-NEXT: s_mov_b32 s3, s5 61 ; GFX10-NEXT: s_mov_b32 s5, s7 79 ; GFX9-NEXT: s_mov_b32 s3, s5 81 ; GFX9-NEXT: s_mov_b32 s5, s7 [all …]
|
D | image_ls_mipmap_zero.ll | 11 ; GFX9-NEXT: s_mov_b32 s3, s5 13 ; GFX9-NEXT: s_mov_b32 s5, s7 25 ; GFX10-NEXT: s_mov_b32 s3, s5 27 ; GFX10-NEXT: s_mov_b32 s5, s7 44 ; GFX9-NEXT: s_mov_b32 s3, s5 46 ; GFX9-NEXT: s_mov_b32 s5, s7 58 ; GFX10-NEXT: s_mov_b32 s3, s5 60 ; GFX10-NEXT: s_mov_b32 s5, s7 77 ; GFX9-NEXT: s_mov_b32 s3, s5 79 ; GFX9-NEXT: s_mov_b32 s5, s7 [all …]
|
D | llvm.amdgcn.image.getresinfo.a16.ll | 11 ; GFX9-NEXT: s_mov_b32 s3, s5 13 ; GFX9-NEXT: s_mov_b32 s5, s7 25 ; GFX10-NEXT: s_mov_b32 s3, s5 27 ; GFX10-NEXT: s_mov_b32 s5, s7 44 ; GFX9-NEXT: s_mov_b32 s3, s5 46 ; GFX9-NEXT: s_mov_b32 s5, s7 58 ; GFX10-NEXT: s_mov_b32 s3, s5 60 ; GFX10-NEXT: s_mov_b32 s5, s7 77 ; GFX9-NEXT: s_mov_b32 s3, s5 79 ; GFX9-NEXT: s_mov_b32 s5, s7 [all …]
|
D | ssubsat.ll | 87 ; GFX8-NEXT: s_cselect_b32 s5, s3, s4 88 ; GFX8-NEXT: s_sub_i32 s5, s5, 0x7fff 92 ; GFX8-NEXT: s_sext_i32_i16 s4, s5 209 ; GFX8-NEXT: s_cselect_b32 s5, s3, s4 210 ; GFX8-NEXT: s_sub_i32 s5, s5, 0x7fff 214 ; GFX8-NEXT: s_sext_i32_i16 s4, s5 262 ; GFX6-NEXT: s_brev_b32 s5, 1 264 ; GFX6-NEXT: v_subrev_i32_e32 v5, vcc, s5, v5 273 ; GFX6-NEXT: v_subrev_i32_e32 v4, vcc, s5, v4 297 ; GFX8-NEXT: s_movk_i32 s5, 0x8000 [all …]
|
D | saddsat.ll | 87 ; GFX8-NEXT: s_cselect_b32 s5, s3, s4 88 ; GFX8-NEXT: s_sub_i32 s5, 0x7fff, s5 97 ; GFX8-NEXT: s_sext_i32_i16 s3, s5 209 ; GFX8-NEXT: s_cselect_b32 s5, s3, s4 210 ; GFX8-NEXT: s_sub_i32 s5, 0x7fff, s5 219 ; GFX8-NEXT: s_sext_i32_i16 s3, s5 257 ; GFX6-NEXT: s_brev_b32 s5, 1 261 ; GFX6-NEXT: v_sub_i32_e32 v5, vcc, s5, v5 272 ; GFX6-NEXT: v_sub_i32_e32 v4, vcc, s5, v4 292 ; GFX8-NEXT: s_movk_i32 s5, 0x8000 [all …]
|
/external/libaom/libaom/av1/common/arm/ |
D | convolve_neon.c | 28 const int16x4_t s4, const int16x4_t s5, in convolve8_4x4() argument 36 sum = vmla_n_s16(sum, s5, filter[5]); in convolve8_4x4() 50 const int16x8_t s3, const int16x8_t s4, const int16x8_t s5, in convolve8_horiz_8x8() argument 58 sum = vmlaq_n_s16(sum, s5, filter[5]); in convolve8_horiz_8x8() 76 const int16x4_t s3, const int16x4_t s4, const int16x4_t s5, in convolve8_horiz_4x1() argument 84 sum = vmla_n_s16(sum, s5, filter[5]); in convolve8_horiz_4x1() 102 const int16x8_t s3, const int16x8_t s4, const int16x8_t s5, in convolve8_vert_8x4() argument 109 sum = vmlaq_n_s16(sum, s5, filter[5]); in convolve8_vert_8x4() 123 const int16x4_t s3, const int16x4_t s4, const int16x4_t s5, in convolve8_vert_4x4_s32() argument 136 sum0 = vmlal_n_s16(sum0, s5, y_filter[5]); in convolve8_vert_4x4_s32() [all …]
|
D | jnt_convolve_neon.c | 333 int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, d0; in dist_wtd_convolve_2d_horiz_neon() local 364 s5 = vget_high_s16(tt1); in dist_wtd_convolve_2d_horiz_neon() 383 d0 = convolve8_4x4_s16(s0, s1, s2, s3, s4, s5, s6, s7, x_filter_tmp, in dist_wtd_convolve_2d_horiz_neon() 385 d1 = convolve8_4x4_s16(s1, s2, s3, s4, s5, s6, s7, s8, x_filter_tmp, in dist_wtd_convolve_2d_horiz_neon() 387 d2 = convolve8_4x4_s16(s2, s3, s4, s5, s6, s7, s8, s9, x_filter_tmp, in dist_wtd_convolve_2d_horiz_neon() 389 d3 = convolve8_4x4_s16(s3, s4, s5, s6, s7, s8, s9, s10, x_filter_tmp, in dist_wtd_convolve_2d_horiz_neon() 417 s5 = vext_s16(s4, s7, 1); // a5 a6 a7 a8 in dist_wtd_convolve_2d_horiz_neon() 421 d0 = convolve8_4x4_s16(s0, s1, s2, s3, s4, s5, s6, s7, x_filter_tmp, in dist_wtd_convolve_2d_horiz_neon() 433 int16x8_t s0, s1, s2, s3, s4, s5, s6, s7; in dist_wtd_convolve_2d_horiz_neon() local 459 s5 = vreinterpretq_s16_u16(vmovl_u8(t5)); in dist_wtd_convolve_2d_horiz_neon() [all …]
|
/external/llvm-project/llvm/test/MC/RISCV/ |
D | rv32a-valid.s | 56 # CHECK-ASM-AND-OBJ: amomax.w s7, s6, (s5) 58 amomax.w s7, s6, (s5) 59 # CHECK-ASM-AND-OBJ: amominu.w s6, s5, (s4) 61 amominu.w s6, s5, (s4) 62 # CHECK-ASM-AND-OBJ: amomaxu.w s5, s4, (s3) 64 amomaxu.w s5, s4, (s3) 84 # CHECK-ASM-AND-OBJ: amomax.w.aq s7, s6, (s5) 86 amomax.w.aq s7, s6, (s5) 87 # CHECK-ASM-AND-OBJ: amominu.w.aq s6, s5, (s4) 89 amominu.w.aq s6, s5, (s4) [all …]
|
D | rv64a-valid.s | 68 # CHECK-ASM-AND-OBJ: amomax.d s7, s6, (s5) 71 amomax.d s7, s6, (s5) 72 # CHECK-ASM-AND-OBJ: amominu.d s6, s5, (s4) 75 amominu.d s6, s5, (s4) 76 # CHECK-ASM-AND-OBJ: amomaxu.d s5, s4, (s3) 79 amomaxu.d s5, s4, (s3) 106 # CHECK-ASM-AND-OBJ: amomax.d.aq s7, s6, (s5) 109 amomax.d.aq s7, s6, (s5) 110 # CHECK-ASM-AND-OBJ: amominu.d.aq s6, s5, (s4) 113 amominu.d.aq s6, s5, (s4) [all …]
|
/external/llvm-project/llvm/test/MC/Disassembler/AMDGPU/ |
D | sop2_gfx9.txt | 3 # GFX9: s_lshl1_add_u32 s5, 0, s2 ; encoding: [0x80,0x02,0x05,0x97] 6 # GFX9: s_lshl2_add_u32 s5, 0xaf123456, s2 ; encoding: [0xff,0x02,0x85,0x97,0x56,0x34,0x12,0xaf] 9 # GFX9: s_lshl3_add_u32 s5, s1, -1 ; encoding: [0x01,0xc1,0x05,0x98] 12 # GFX9: s_lshl4_add_u32 s5, s1, 0xaf123456 ; encoding: [0x01,0xff,0x85,0x98,0x56,0x34,0x12,0xaf] 15 # GFX9: s_mul_hi_i32 s5, s101, s2 ; encoding: [0x65,0x02,0x85,0x96] 18 # GFX9: s_mul_hi_i32 s5, 0, s2 ; encoding: [0x80,0x02,0x85,0x96] 21 # GFX9: s_mul_hi_i32 s5, 0xaf123456, s2 ; encoding: [0xff,0x02,0x85,0x96,0x56,0x34,0x12,0xaf] 24 # GFX9: s_mul_hi_u32 s5, s1, -1 ; encoding: [0x01,0xc1,0x05,0x96] 27 # GFX9: s_mul_hi_u32 s5, s1, 0xaf123456 ; encoding: [0x01,0xff,0x05,0x96,0x56,0x34,0x12,0xaf]
|
/external/deqp-deps/glslang/Test/ |
D | spv.bufferhandle4.frag | 19 } s5; 22 x.k.h = s5.m.k.i.k.i.k.h; 25 s5.m = b ? s5.m : s5.m.k.i;
|