/external/XNNPACK/src/f32-vbinary/gen/ |
D | vrsubc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vrsubc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vrsubc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vrsubc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vrsubc_ukernel__neon_x4()
|
D | vrdivc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vrdivc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vrdivc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vrdivc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vrdivc_ukernel__neon_x4()
|
D | vsubc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vsubc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vsubc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vsubc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vsubc_ukernel__neon_x4()
|
D | vaddc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vaddc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vaddc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vaddc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vaddc_ukernel__neon_x4()
|
D | vmaxc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmaxc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vmaxc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vmaxc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmaxc_ukernel__neon_x4()
|
D | vdivc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vdivc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vdivc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vdivc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vdivc_ukernel__neon_x4()
|
D | vmulc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmulc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vmulc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vmulc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmulc_ukernel__neon_x4()
|
D | vminc-neon-x4.c | 58 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vminc_ukernel__neon_x4() local 60 vst1_f32(y, vy01); y += 2; in xnn_f32_vminc_ukernel__neon_x4() 61 vy01 = vget_high_f32(vy0123); in xnn_f32_vminc_ukernel__neon_x4() 64 vst1_lane_f32(y, vy01, 0); in xnn_f32_vminc_ukernel__neon_x4()
|
D | vsub-neon-x4.c | 60 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vsub_ukernel__neon_x4() local 62 vst1_f32(y, vy01); y += 2; in xnn_f32_vsub_ukernel__neon_x4() 63 vy01 = vget_high_f32(vy0123); in xnn_f32_vsub_ukernel__neon_x4() 66 vst1_lane_f32(y, vy01, 0); in xnn_f32_vsub_ukernel__neon_x4()
|
D | vmax-neon-x4.c | 60 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmax_ukernel__neon_x4() local 62 vst1_f32(y, vy01); y += 2; in xnn_f32_vmax_ukernel__neon_x4() 63 vy01 = vget_high_f32(vy0123); in xnn_f32_vmax_ukernel__neon_x4() 66 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmax_ukernel__neon_x4()
|
D | vdiv-neon-x4.c | 60 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vdiv_ukernel__neon_x4() local 62 vst1_f32(y, vy01); y += 2; in xnn_f32_vdiv_ukernel__neon_x4() 63 vy01 = vget_high_f32(vy0123); in xnn_f32_vdiv_ukernel__neon_x4() 66 vst1_lane_f32(y, vy01, 0); in xnn_f32_vdiv_ukernel__neon_x4()
|
D | vmin-neon-x4.c | 60 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmin_ukernel__neon_x4() local 62 vst1_f32(y, vy01); y += 2; in xnn_f32_vmin_ukernel__neon_x4() 63 vy01 = vget_high_f32(vy0123); in xnn_f32_vmin_ukernel__neon_x4() 66 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmin_ukernel__neon_x4()
|
D | vmul-neon-x4.c | 60 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmul_ukernel__neon_x4() local 62 vst1_f32(y, vy01); y += 2; in xnn_f32_vmul_ukernel__neon_x4() 63 vy01 = vget_high_f32(vy0123); in xnn_f32_vmul_ukernel__neon_x4() 66 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmul_ukernel__neon_x4()
|
D | vadd-neon-x4.c | 60 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vadd_ukernel__neon_x4() local 62 vst1_f32(y, vy01); y += 2; in xnn_f32_vadd_ukernel__neon_x4() 63 vy01 = vget_high_f32(vy0123); in xnn_f32_vadd_ukernel__neon_x4() 66 vst1_lane_f32(y, vy01, 0); in xnn_f32_vadd_ukernel__neon_x4()
|
D | vrdivc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vrdivc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vrdivc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vrdivc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vrdivc_ukernel__neon_x8()
|
D | vsubc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vsubc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vsubc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vsubc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vsubc_ukernel__neon_x8()
|
D | vdivc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vdivc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vdivc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vdivc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vdivc_ukernel__neon_x8()
|
D | vmaxc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmaxc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vmaxc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vmaxc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmaxc_ukernel__neon_x8()
|
D | vmulc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmulc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vmulc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vmulc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmulc_ukernel__neon_x8()
|
D | vrsubc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vrsubc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vrsubc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vrsubc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vrsubc_ukernel__neon_x8()
|
D | vaddc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vaddc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vaddc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vaddc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vaddc_ukernel__neon_x8()
|
D | vminc-neon-x8.c | 63 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vminc_ukernel__neon_x8() local 65 vst1_f32(y, vy01); y += 2; in xnn_f32_vminc_ukernel__neon_x8() 66 vy01 = vget_high_f32(vy0123); in xnn_f32_vminc_ukernel__neon_x8() 69 vst1_lane_f32(y, vy01, 0); in xnn_f32_vminc_ukernel__neon_x8()
|
D | vmax-neon-x8.c | 66 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vmax_ukernel__neon_x8() local 68 vst1_f32(y, vy01); y += 2; in xnn_f32_vmax_ukernel__neon_x8() 69 vy01 = vget_high_f32(vy0123); in xnn_f32_vmax_ukernel__neon_x8() 72 vst1_lane_f32(y, vy01, 0); in xnn_f32_vmax_ukernel__neon_x8()
|
D | vsub-neon-x8.c | 66 float32x2_t vy01 = vget_low_f32(vy0123); in xnn_f32_vsub_ukernel__neon_x8() local 68 vst1_f32(y, vy01); y += 2; in xnn_f32_vsub_ukernel__neon_x8() 69 vy01 = vget_high_f32(vy0123); in xnn_f32_vsub_ukernel__neon_x8() 72 vst1_lane_f32(y, vy01, 0); in xnn_f32_vsub_ukernel__neon_x8()
|
/external/XNNPACK/src/q8-vadd/ |
D | neon.c | 92 uint8x16_t vy01 = vqmovun_high_s16(vqmovun_s16(vacc0), vacc1); in xnn_q8_vadd_ukernel__neon() local 95 vy01 = vmaxq_u8(vy01, vy_min); in xnn_q8_vadd_ukernel__neon() 97 vy01 = vminq_u8(vy01, vy_max); in xnn_q8_vadd_ukernel__neon() 100 vst1q_u8(y, vy01); y += 16; in xnn_q8_vadd_ukernel__neon() 143 uint8x16_t vy01 = vcombine_u8(vqmovun_s16(vacc0), vqmovun_s16(vacc1)); in xnn_q8_vadd_ukernel__neon() local 144 vy01 = vmaxq_u8(vy01, vy_min); in xnn_q8_vadd_ukernel__neon() 145 vy01 = vminq_u8(vy01, vy_max); in xnn_q8_vadd_ukernel__neon() 147 vst1q_u8(y, vy01); y += 16; in xnn_q8_vadd_ukernel__neon()
|