Lines Matching refs:vdup_lane_s16
83 …int16x8_t vprod0x0123c0 = vmull_s8(vb0123c0x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
85 …vprod0x0123c0 = vmlal_s8(vprod0x0123c0, vb0123c0x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
87 …int16x8_t vprod0x4567c0 = vmull_s8(vb4567c0x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
89 …vprod0x4567c0 = vmlal_s8(vprod0x4567c0, vb4567c0x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
91 …int16x8_t vprod0x89ABc0 = vmull_s8(vb89ABc0x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
93 …vprod0x89ABc0 = vmlal_s8(vprod0x89ABc0, vb89ABc0x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
95 …int16x8_t vprod0xCDEFc0 = vmull_s8(vbCDEFc0x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
97 …vprod0xCDEFc0 = vmlal_s8(vprod0xCDEFc0, vbCDEFc0x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
99 …int16x8_t vprod0x0123c1 = vmull_s8(vb0123c1x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
101 …vprod0x0123c1 = vmlal_s8(vprod0x0123c1, vb0123c1x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
103 …int16x8_t vprod0x4567c1 = vmull_s8(vb4567c1x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
105 …vprod0x4567c1 = vmlal_s8(vprod0x4567c1, vb4567c1x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
107 …int16x8_t vprod0x89ABc1 = vmull_s8(vb89ABc1x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
109 …vprod0x89ABc1 = vmlal_s8(vprod0x89ABc1, vb89ABc1x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
111 …int16x8_t vprod0xCDEFc1 = vmull_s8(vbCDEFc1x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
113 …vprod0xCDEFc1 = vmlal_s8(vprod0xCDEFc1, vbCDEFc1x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
115 …int16x8_t vprod0x0123c2 = vmull_s8(vb0123c2x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
117 …vprod0x0123c2 = vmlal_s8(vprod0x0123c2, vb0123c2x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
119 …int16x8_t vprod0x4567c2 = vmull_s8(vb4567c2x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
121 …vprod0x4567c2 = vmlal_s8(vprod0x4567c2, vb4567c2x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
123 …int16x8_t vprod0x89ABc2 = vmull_s8(vb89ABc2x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
125 …vprod0x89ABc2 = vmlal_s8(vprod0x89ABc2, vb89ABc2x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
127 …int16x8_t vprod0xCDEFc2 = vmull_s8(vbCDEFc2x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
129 …vprod0xCDEFc2 = vmlal_s8(vprod0xCDEFc2, vbCDEFc2x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
131 …int16x8_t vprod0x0123c3 = vmull_s8(vb0123c3x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
133 …vprod0x0123c3 = vmlal_s8(vprod0x0123c3, vb0123c3x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
135 …int16x8_t vprod0x4567c3 = vmull_s8(vb4567c3x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
137 …vprod0x4567c3 = vmlal_s8(vprod0x4567c3, vb4567c3x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
139 …int16x8_t vprod0x89ABc3 = vmull_s8(vb89ABc3x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
141 …vprod0x89ABc3 = vmlal_s8(vprod0x89ABc3, vb89ABc3x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
143 …int16x8_t vprod0xCDEFc3 = vmull_s8(vbCDEFc3x0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_s16_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
145 …vprod0xCDEFc3 = vmlal_s8(vprod0xCDEFc3, vbCDEFc3x1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
171 …const int16x8_t vprod0x0123c0 = vmull_s8(vb0123c0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
172 …const int16x8_t vprod0x0123c1 = vmull_s8(vb0123c1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
173 …const int16x8_t vprod0x0123c2 = vmull_s8(vb0123c2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
174 …const int16x8_t vprod0x0123c3 = vmull_s8(vb0123c3, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
179 …const int16x8_t vprod0x4567c0 = vmull_s8(vb4567c0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
180 …const int16x8_t vprod0x4567c1 = vmull_s8(vb4567c1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
181 …const int16x8_t vprod0x4567c2 = vmull_s8(vb4567c2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
182 …const int16x8_t vprod0x4567c3 = vmull_s8(vb4567c3, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
187 …const int16x8_t vprod0x89ABc0 = vmull_s8(vb89ABc0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
188 …const int16x8_t vprod0x89ABc1 = vmull_s8(vb89ABc1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
189 …const int16x8_t vprod0x89ABc2 = vmull_s8(vb89ABc2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
190 …const int16x8_t vprod0x89ABc3 = vmull_s8(vb89ABc3, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
195 …const int16x8_t vprod0xCDEFc0 = vmull_s8(vbCDEFc0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
196 …const int16x8_t vprod0xCDEFc1 = vmull_s8(vbCDEFc1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
197 …const int16x8_t vprod0xCDEFc2 = vmull_s8(vbCDEFc2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
198 …const int16x8_t vprod0xCDEFc3 = vmull_s8(vbCDEFc3, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
215 …const int16x8_t vprod0x0123c0 = vmull_s8(vb0123c0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
217 …const int16x8_t vprod0x4567c0 = vmull_s8(vb4567c0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
219 …const int16x8_t vprod0x89ABc0 = vmull_s8(vb89ABc0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
221 …const int16x8_t vprod0xCDEFc0 = vmull_s8(vbCDEFc0, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
230 …const int16x8_t vprod0x0123c1 = vmull_s8(vb0123c1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
232 …const int16x8_t vprod0x4567c1 = vmull_s8(vb4567c1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
234 …const int16x8_t vprod0x89ABc1 = vmull_s8(vb89ABc1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
236 …const int16x8_t vprod0xCDEFc1 = vmull_s8(vbCDEFc1, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
245 …const int16x8_t vprod0x0123c2 = vmull_s8(vb0123c2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
247 …const int16x8_t vprod0x4567c2 = vmull_s8(vb4567c2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
249 …const int16x8_t vprod0x89ABc2 = vmull_s8(vb89ABc2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()
251 …const int16x8_t vprod0xCDEFc2 = vmull_s8(vbCDEFc2, vreinterpret_s8_s16(vdup_lane_s16(vreinterpret_… in xnn_qs8_igemm_minmax_ukernel_1x16c2__neon_mlal_padal_dup()