| /external/XNNPACK/src/f32-ibilinear-chw/gen/ |
| D | wasmsimd-p8.c | 83 const v128_t vtltr23 = wasm_f64x2_replace_lane(vtltr2, 1, vtltr3); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8() local
91 const v128_t vldrd23 = wasm_f32x4_sub(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8()
100 const v128_t vtl0123 = wasm_v32x4_shuffle(vtltr01, vtltr23, 0, 2, 4, 6); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8()
101 const v128_t vtr0123 = wasm_v32x4_shuffle(vtltr01, vtltr23, 1, 3, 5, 7); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8()
150 const v128_t vtltr23 = wasm_f64x2_replace_lane(vtltr2, 1, vtltr3); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8() local
154 const v128_t vldrd23 = wasm_f32x4_sub(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8()
159 const v128_t vtl = wasm_v32x4_shuffle(vtltr01, vtltr23, 0, 2, 4, 6); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8()
160 const v128_t vtr = wasm_v32x4_shuffle(vtltr01, vtltr23, 1, 3, 5, 7); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8()
|
| D | wasmsimd-p4.c | 64 const v128_t vtltr23 = wasm_f64x2_replace_lane(vtltr2, 1, vtltr3); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p4() local
68 const v128_t vldrd23 = wasm_f32x4_sub(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p4()
73 const v128_t vtl = wasm_v32x4_shuffle(vtltr01, vtltr23, 0, 2, 4, 6); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p4()
74 const v128_t vtr = wasm_v32x4_shuffle(vtltr01, vtltr23, 1, 3, 5, 7); in xnn_f32_ibilinear_chw_ukernel__wasmsimd_p4()
|
| D | neonfma-p8.c | 81 const float32x4_t vtltr23 = vcombine_f32(vtltr2, vtltr3); in xnn_f32_ibilinear_chw_ukernel__neonfma_p8() local
89 const float32x4_t vldrd23 = vsubq_f32(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neonfma_p8()
100 const float32x4x2_t vtl_t0123 = vuzpq_f32(vtltr01, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neonfma_p8()
151 const float32x4_t vtltr23 = vcombine_f32(vtltr2, vtltr3); in xnn_f32_ibilinear_chw_ukernel__neonfma_p8() local
155 const float32x4_t vldrd23 = vsubq_f32(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neonfma_p8()
161 const float32x4x2_t vtl_t = vuzpq_f32(vtltr01, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neonfma_p8()
|
| D | neon-p8.c | 81 const float32x4_t vtltr23 = vcombine_f32(vtltr2, vtltr3); in xnn_f32_ibilinear_chw_ukernel__neon_p8() local
89 const float32x4_t vldrd23 = vsubq_f32(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neon_p8()
100 const float32x4x2_t vtl_t0123 = vuzpq_f32(vtltr01, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neon_p8()
151 const float32x4_t vtltr23 = vcombine_f32(vtltr2, vtltr3); in xnn_f32_ibilinear_chw_ukernel__neon_p8() local
155 const float32x4_t vldrd23 = vsubq_f32(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neon_p8()
161 const float32x4x2_t vtl_t = vuzpq_f32(vtltr01, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neon_p8()
|
| D | neonfma-p4.c | 63 const float32x4_t vtltr23 = vcombine_f32(vtltr2, vtltr3); in xnn_f32_ibilinear_chw_ukernel__neonfma_p4() local
67 const float32x4_t vldrd23 = vsubq_f32(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neonfma_p4()
73 const float32x4x2_t vtl_t = vuzpq_f32(vtltr01, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neonfma_p4()
|
| D | neon-p4.c | 63 const float32x4_t vtltr23 = vcombine_f32(vtltr2, vtltr3); in xnn_f32_ibilinear_chw_ukernel__neon_p4() local
67 const float32x4_t vldrd23 = vsubq_f32(vblbr23, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neon_p4()
73 const float32x4x2_t vtl_t = vuzpq_f32(vtltr01, vtltr23); in xnn_f32_ibilinear_chw_ukernel__neon_p4()
|
| /external/XNNPACK/src/f32-ibilinear-chw/ |
| D | wasmsimd.c.in | 114 const v128_t vtl = wasm_v32x4_shuffle(vtltr01, vtltr23, 0, 2, 4, 6);
115 const v128_t vtr = wasm_v32x4_shuffle(vtltr01, vtltr23, 1, 3, 5, 7);
|
| D | neon.c.in | 113 const float32x4x2_t vtl_t = vuzpq_f32(vtltr01, vtltr23);
|