Home
last modified time | relevance | path

Searched refs:vmull_s16 (Results 1 – 25 of 49) sorted by relevance

12

/external/libhevc/common/arm/
Dihevc_quant_iquant_ssd_neon_intr.c185 qtmp_0 = vmull_s16(abs_s0, sq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
186 qtmp_1 = vmull_s16(abs_s1, sq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
187 qtmp_2 = vmull_s16(abs_s2, sq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
188 qtmp_3 = vmull_s16(abs_s3, sq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
245 iqtmp_0 = vmull_s16(q0, siq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
246 iqtmp_1 = vmull_s16(q1, siq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
247 iqtmp_2 = vmull_s16(q2, siq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
248 iqtmp_3 = vmull_s16(q3, siq); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
294 r0 = vmull_s16(s0, s0); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
295 r1 = vmull_s16(s1, s1); in ihevc_quant_iquant_ssd_flat_scale_mat_neon()
[all …]
Dihevc_resi_trans_neon_32x32.c286 vmull_s16(vget_low_s16(g_ai2_ihevc_trans_32_01_8), eeee_00); in ihevc_resi_trans_32x32_neon()
290 vmull_s16(vget_low_s16(g_ai2_ihevc_trans_32_01_8), eeee_10); in ihevc_resi_trans_32x32_neon()
305 vmull_s16(g_ai2_ihevc_trans_32_4_04, vget_low_s16(eeo)); in ihevc_resi_trans_32x32_neon()
307 vmull_s16(g_ai2_ihevc_trans_32_12_04, vget_low_s16(eeo)); in ihevc_resi_trans_32x32_neon()
309 vmull_s16(g_ai2_ihevc_trans_32_20_04, vget_low_s16(eeo)); in ihevc_resi_trans_32x32_neon()
311 vmull_s16(g_ai2_ihevc_trans_32_28_04, vget_low_s16(eeo)); in ihevc_resi_trans_32x32_neon()
314 vmull_s16(g_ai2_ihevc_trans_32_4_04, vget_high_s16(eeo)); in ihevc_resi_trans_32x32_neon()
316 vmull_s16(g_ai2_ihevc_trans_32_12_04, vget_high_s16(eeo)); in ihevc_resi_trans_32x32_neon()
318 vmull_s16(g_ai2_ihevc_trans_32_20_04, vget_high_s16(eeo)); in ihevc_resi_trans_32x32_neon()
320 vmull_s16(g_ai2_ihevc_trans_32_28_04, vget_high_s16(eeo)); in ihevc_resi_trans_32x32_neon()
[all …]
Dihevc_inter_pred_chroma_vert_w16inp_w16out.s145 vmull.s16 q0,d0,d12 @vmull_s16(src_tmp1, coeff_0)
148 vmull.s16 q4,d2,d12 @vmull_s16(src_tmp2, coeff_0)
192 vmull.s16 q15,d0,d12 @vmull_s16(src_tmp1, coeff_0)
200 vmull.s16 q14,d1,d12 @vmull_s16(src_tmp2, coeff_0)
211 vmull.s16 q13,d2,d12 @vmull_s16(src_tmp2, coeff_0)
221 vmull.s16 q12,d3,d12 @vmull_s16(src_tmp2, coeff_0)
236 vmull.s16 q15,d0,d12 @vmull_s16(src_tmp1, coeff_0)
246 vmull.s16 q14,d1,d12 @vmull_s16(src_tmp2, coeff_0)
257 vmull.s16 q13,d2,d12 @vmull_s16(src_tmp2, coeff_0)
269 vmull.s16 q12,d3,d12 @vmull_s16(src_tmp2, coeff_0)
[all …]
Dihevc_inter_pred_chroma_vert_w16inp.s145 vmull.s16 q0,d0,d12 @vmull_s16(src_tmp1, coeff_0)
148 vmull.s16 q4,d2,d12 @vmull_s16(src_tmp2, coeff_0)
193 vmull.s16 q15,d0,d12 @vmull_s16(src_tmp1, coeff_0)
201 vmull.s16 q14,d1,d12 @vmull_s16(src_tmp2, coeff_0)
212 vmull.s16 q13,d2,d12 @vmull_s16(src_tmp2, coeff_0)
223 vmull.s16 q12,d3,d12 @vmull_s16(src_tmp2, coeff_0)
239 vmull.s16 q15,d0,d12 @vmull_s16(src_tmp1, coeff_0)
250 vmull.s16 q14,d1,d12 @vmull_s16(src_tmp2, coeff_0)
262 vmull.s16 q13,d2,d12 @vmull_s16(src_tmp2, coeff_0)
275 vmull.s16 q12,d3,d12 @vmull_s16(src_tmp2, coeff_0)
[all …]
Dihevc_weighted_pred_neon_intr.c300 i4_tmp1_t = vmull_s16(pi2_src_val1, wgt0.val[0]); in ihevc_weighted_pred_chroma_uni_neonintr()
303 i4_tmp2_t = vmull_s16(pi2_src_val2, wgt0.val[0]); in ihevc_weighted_pred_chroma_uni_neonintr()
641 i4_tmp1_t1 = vmull_s16(pi2_src1_val1, wgt0.val[0]); in ihevc_weighted_pred_chroma_bi_neonintr()
644 i4_tmp1_t2 = vmull_s16(pi2_src2_val1, wgt1.val[0]); in ihevc_weighted_pred_chroma_bi_neonintr()
649 i4_tmp2_t1 = vmull_s16(pi2_src1_val2, wgt0.val[0]); in ihevc_weighted_pred_chroma_bi_neonintr()
652 i4_tmp2_t2 = vmull_s16(pi2_src2_val2, wgt1.val[0]); in ihevc_weighted_pred_chroma_bi_neonintr()
/external/libaom/libaom/av1/encoder/arm/neon/
Dquantize_neon.c60 vmull_s16(vget_low_s16(v_tmp), vget_low_s16(v_quant)); in av1_quantize_fp_neon()
62 vmull_s16(vget_high_s16(v_tmp), vget_high_s16(v_quant)); in av1_quantize_fp_neon()
86 vmull_s16(vget_low_s16(v_tmp), vget_low_s16(v_quant)); in av1_quantize_fp_neon()
88 vmull_s16(vget_high_s16(v_tmp), vget_high_s16(v_quant)); in av1_quantize_fp_neon()
125 vmull_s16(vget_low_s16(qcoeff), vget_low_s16(dequant)); in calculate_dqcoeff_lp_and_store()
127 vmull_s16(vget_high_s16(qcoeff), vget_high_s16(dequant)); in calculate_dqcoeff_lp_and_store()
158 vmull_s16(vget_low_s16(v_tmp), vget_low_s16(v_quant)); in av1_quantize_lp_neon()
160 vmull_s16(vget_high_s16(v_tmp), vget_high_s16(v_quant)); in av1_quantize_lp_neon()
183 vmull_s16(vget_low_s16(v_tmp), vget_low_s16(v_quant)); in av1_quantize_lp_neon()
185 vmull_s16(vget_high_s16(v_tmp), vget_high_s16(v_quant)); in av1_quantize_lp_neon()
Dav1_error_neon.c33 const int32x4_t err0 = vmull_s16(diff_lo, diff_lo); in av1_block_error_neon()
40 const int32x4_t sqcoeff0 = vmull_s16(coeff_lo, coeff_lo); in av1_block_error_neon()
75 const int32x4_t err0 = vmull_s16(diff_lo, diff_lo); in av1_block_error_lp_neon()
/external/webrtc/modules/audio_coding/codecs/isac/fix/source/
Dtransform_neon.c50 int32x4_t tmp0 = vmull_s16(vget_low_s16(tmpr), vget_low_s16(inre1)); in ComplexMulAndFindMaxNeon()
51 int32x4_t tmp1 = vmull_s16(vget_low_s16(tmpr), vget_low_s16(inre2)); in ComplexMulAndFindMaxNeon()
60 int32x4_t tmp2 = vmull_s16(vget_high_s16(tmpr), vget_high_s16(inre1)); in ComplexMulAndFindMaxNeon()
61 int32x4_t tmp3 = vmull_s16(vget_high_s16(tmpr), vget_high_s16(inre2)); in ComplexMulAndFindMaxNeon()
184 int32x4_t outr0 = vmull_s16(tmpr, xr); in PostShiftAndSeparateNeon()
185 int32x4_t outi0 = vmull_s16(tmpi, xr); in PostShiftAndSeparateNeon()
186 int32x4_t outr1 = vmull_s16(tmpi, yr); in PostShiftAndSeparateNeon()
187 int32x4_t outi1 = vmull_s16(tmpi, yi); in PostShiftAndSeparateNeon()
285 int32x4_t xr = vmull_s16(tmpr, inre_0); in TransformAndFindMaxNeon()
286 int32x4_t xi = vmull_s16(tmpr, inim_0); in TransformAndFindMaxNeon()
[all …]
Dfilters_neon.c42 tmpa0_v = vmull_s16(x0_v, x0_v); in WebRtcIsacfix_AutocorrNeon()
72 tmpa0_v = vmull_s16(vget_low_s16(x1_v), vget_low_s16(y_v)); in WebRtcIsacfix_AutocorrNeon()
76 tmpa1_v = vmull_s16(vget_high_s16(x1_v), vget_high_s16(y_v)); in WebRtcIsacfix_AutocorrNeon()
88 tmpa0_v = vmull_s16(x2_v, y2_v); in WebRtcIsacfix_AutocorrNeon()
Dpitch_estimator_c.c73 int32x4_t int_32x4 = vmull_s16(int_16x4_x, int_16x4_in); in WebRtcIsacfix_PCorr2Q32()
/external/tensorflow/tensorflow/core/kernels/
Dquantized_mul_op.cc89 vmull_s16(x_high_high_16x4, scalar_minus_offset_16x4); in ScalarMultiply()
91 vmull_s16(x_high_low_16x4, scalar_minus_offset_16x4); in ScalarMultiply()
93 vmull_s16(x_low_high_16x4, scalar_minus_offset_16x4); in ScalarMultiply()
95 vmull_s16(x_low_low_16x4, scalar_minus_offset_16x4); in ScalarMultiply()
167 vmull_s16(x_high_high_16x4, y_high_high_16x4); in VectorMultiply()
169 vmull_s16(x_high_low_16x4, y_high_low_16x4); in VectorMultiply()
171 vmull_s16(x_low_high_16x4, y_low_high_16x4); in VectorMultiply()
172 const int32x4_t z_low_low_32x4 = vmull_s16(x_low_low_16x4, y_low_low_16x4); in VectorMultiply()
254 vmull_s16(x_high_high_16x4, y_high_high_16x4); in VectorTensorMultiply()
256 vmull_s16(x_high_low_16x4, y_high_low_16x4); in VectorTensorMultiply()
[all …]
/external/libvpx/libvpx/vp9/encoder/arm/neon/
Dvp9_quantize_neon.c33 vmull_s16(vget_low_s16(qcoeff), vget_low_s16(dequant)); in calculate_dqcoeff_and_store()
35 vmull_s16(vget_high_s16(qcoeff), vget_high_s16(dequant)); in calculate_dqcoeff_and_store()
77 vmull_s16(vget_low_s16(v_tmp), vget_low_s16(v_quant)); in vp9_quantize_fp_neon()
79 vmull_s16(vget_high_s16(v_tmp), vget_high_s16(v_quant)); in vp9_quantize_fp_neon()
102 vmull_s16(vget_low_s16(v_tmp), vget_low_s16(v_quant)); in vp9_quantize_fp_neon()
104 vmull_s16(vget_high_s16(v_tmp), vget_high_s16(v_quant)); in vp9_quantize_fp_neon()
182 dqcoeff_0 = vmull_s16(vget_low_s16(qcoeff), dequant); in vp9_quantize_fp_32x32_neon()
Dvp9_error_neon.c31 const int32x4_t err0 = vmull_s16(diff_lo, diff_lo); in vp9_block_error_fp_neon()
/external/webrtc/common_audio/signal_processing/
Dcross_correlation_neon.c31 int32x4_t tmp0 = vmull_s16(vget_low_s16(seq1_16x8), in DotProductWithScaleNeon()
35 int32x4_t tmp0 = vmull_s16(vget_low_s16(seq1_16x8), in DotProductWithScaleNeon()
37 int32x4_t tmp1 = vmull_s16(vget_high_s16(seq1_16x8), in DotProductWithScaleNeon()
/external/libaom/libaom/av1/common/arm/
Dselfguided_neon.c271 d1 = vmull_s16(t1, t1); in boxsum2_square_sum_calc()
272 d2 = vmull_s16(t2, t2); in boxsum2_square_sum_calc()
273 d3 = vmull_s16(t3, t3); in boxsum2_square_sum_calc()
274 d4 = vmull_s16(t4, t4); in boxsum2_square_sum_calc()
275 d5 = vmull_s16(t5, t5); in boxsum2_square_sum_calc()
276 d6 = vmull_s16(t6, t6); in boxsum2_square_sum_calc()
277 d7 = vmull_s16(t7, t7); in boxsum2_square_sum_calc()
278 d8 = vmull_s16(t8, t8); in boxsum2_square_sum_calc()
279 d9 = vmull_s16(t9, t9); in boxsum2_square_sum_calc()
280 d10 = vmull_s16(t10, t10); in boxsum2_square_sum_calc()
[all …]
Dwarp_plane_neon.c361 res_0 = vmull_s16(src_0, fltr_0); in vertical_filter_neon()
372 res_1 = vmull_s16(src_1, fltr_1); in vertical_filter_neon()
386 res_0 = vmull_s16(src_0, fltr_0); in vertical_filter_neon()
397 res_1 = vmull_s16(src_1, fltr_1); in vertical_filter_neon()
414 res_0 = vmull_s16(src_0, fltr_0); in vertical_filter_neon()
425 res_1 = vmull_s16(src_1, fltr_1); in vertical_filter_neon()
439 res_0 = vmull_s16(src_0, fltr_0); in vertical_filter_neon()
450 res_1 = vmull_s16(src_1, fltr_1); in vertical_filter_neon()
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/integer_ops/
Dmul.h88 auto p1 = vmull_s16(input2_val_high_high, input1_val_high_high); in MulElementwise()
89 auto p2 = vmull_s16(input2_val_high_low, input1_val_high_low); in MulElementwise()
90 auto p3 = vmull_s16(input2_val_low_high, input1_val_low_high); in MulElementwise()
91 auto p4 = vmull_s16(input2_val_low_low, input1_val_low_low); in MulElementwise()
/external/libvpx/libvpx/vpx_dsp/arm/
Dquantize_neon.c22 vmull_s16(vget_low_s16(qcoeff), vget_low_s16(dequant)); in calculate_dqcoeff_and_store()
24 vmull_s16(vget_high_s16(qcoeff), vget_high_s16(dequant)); in calculate_dqcoeff_and_store()
170 int32x4_t dqcoeff_0 = vmull_s16(vget_low_s16(qcoeff), vget_low_s16(dequant)); in calculate_dqcoeff_and_store_32x32()
172 vmull_s16(vget_high_s16(qcoeff), vget_high_s16(dequant)); in calculate_dqcoeff_and_store_32x32()
Ddeblock_neon.c355 xy_low = vmull_s16(vget_low_s16(x), vget_low_s16(y)); in vpx_mbpost_proc_across_ip_neon()
356 xy_high = vmull_s16(vget_high_s16(x), vget_high_s16(y)); in vpx_mbpost_proc_across_ip_neon()
425 sumsq_low = vmull_s16(vget_low_s16(sum), vget_low_s16(sum_tmp)); in vpx_mbpost_proc_down_neon()
426 sumsq_high = vmull_s16(vget_high_s16(sum), vget_high_s16(sum_tmp)); in vpx_mbpost_proc_down_neon()
452 xy_low = vmull_s16(vget_low_s16(x), vget_low_s16(y)); in vpx_mbpost_proc_down_neon()
453 xy_high = vmull_s16(vget_high_s16(x), vget_high_s16(y)); in vpx_mbpost_proc_down_neon()
Dvariance_neon.c367 q7s32 = vmull_s16(d22s16, d22s16); in vpx_get4x4sse_cs_neon()
368 q8s32 = vmull_s16(d24s16, d24s16); in vpx_get4x4sse_cs_neon()
369 q9s32 = vmull_s16(d26s16, d26s16); in vpx_get4x4sse_cs_neon()
370 q10s32 = vmull_s16(d28s16, d28s16); in vpx_get4x4sse_cs_neon()
/external/libgav1/libgav1/src/dsp/arm/
Dmask_blend_neon.cc92 vmull_s16(vget_low_s16(pred_mask_0), vget_low_s16(pred_val_0)); in WriteMaskBlendLine4x2()
94 vmull_s16(vget_high_s16(pred_mask_0), vget_high_s16(pred_val_0)); in WriteMaskBlendLine4x2()
221 vmull_s16(vget_low_s16(pred_mask_0), vget_low_s16(pred_val_0)); in MaskBlend_NEON()
223 vmull_s16(vget_high_s16(pred_mask_0), vget_high_s16(pred_val_0)); in MaskBlend_NEON()
Dwarp_neon.cc273 vmull_s16(vget_low_s16(filter), vget_low_s16(intermediate)); in Warp_NEON()
275 vmull_s16(vget_high_s16(filter), vget_high_s16(intermediate)); in Warp_NEON()
/external/libaom/libaom/aom_dsp/arm/
Dvariance_neon.c388 q7s32 = vmull_s16(d22s16, d22s16); in aom_get4x4sse_cs_neon()
389 q8s32 = vmull_s16(d24s16, d24s16); in aom_get4x4sse_cs_neon()
390 q9s32 = vmull_s16(d26s16, d26s16); in aom_get4x4sse_cs_neon()
391 q10s32 = vmull_s16(d28s16, d28s16); in aom_get4x4sse_cs_neon()
/external/XNNPACK/src/qu8-dwconv/
Dup8x9-minmax-neon.c84 int32x4_t vaccX0_lo = vmull_s16(vget_low_s16(vxk0), vget_low_s16(vxi0)); in xnn_qu8_dwconv_minmax_ukernel_up8x9__neon()
85 int32x4_t vaccX0_hi = vmull_s16(vget_high_s16(vxk0), vget_high_s16(vxi0)); in xnn_qu8_dwconv_minmax_ukernel_up8x9__neon()
175 int32x4_t vaccX0_lo = vmull_s16(vget_low_s16(vxk0), vget_low_s16(vxi0)); in xnn_qu8_dwconv_minmax_ukernel_up8x9__neon()
176 int32x4_t vaccX0_hi = vmull_s16(vget_high_s16(vxk0), vget_high_s16(vxi0)); in xnn_qu8_dwconv_minmax_ukernel_up8x9__neon()
/external/libaom/libaom/aom_dsp/simd/
Dv64_intrinsics_arm.h143 vmull_s16(vreinterpret_s16_s64(x), vreinterpret_s16_s64(y))); in v64_dotp_s16()
146 vpaddlq_s32(vmull_s16(vreinterpret_s16_s64(x), vreinterpret_s16_s64(y))); in v64_dotp_s16()
292 vmull_s16(vreinterpret_s16_s64(x), vreinterpret_s16_s64(y))); in v64_mulhi_s16()
296 vmull_s16(vreinterpret_s16_s64(x), vreinterpret_s16_s64(y)), 16))); in v64_mulhi_s16()
306 int32x4_t t = vmull_s16(vreinterpret_s16_s64(x), vreinterpret_s16_s64(y)); in v64_madd_s16()

12