/external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
D | 5x5s2p2-minmax-neon-1x4.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4() 229 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4()
|
D | 5x5s2p2-minmax-neonfma-1x4-acc3.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc3() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc3() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc3() 231 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc3()
|
D | 5x5s2p2-minmax-neon-1x4-acc2.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc2() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc2() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc2() 230 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc2()
|
D | 5x5s2p2-minmax-neon-1x4-acc4.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc4() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc4() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc4() 232 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc4()
|
D | 5x5s2p2-minmax-neon-1x4-acc3.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc3() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc3() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc3() 231 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc3()
|
D | 5x5s2p2-minmax-neonfma-1x4.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4() 229 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4()
|
D | 5x5s2p2-minmax-neonfma-1x4-acc4.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc4() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc4() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc4() 232 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc4()
|
D | 5x5s2p2-minmax-neonfma-1x4-acc2.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2() 230 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2()
|
D | 5x5s2p2-minmax-neon-1x4-acc5.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc5() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc5() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc5() 233 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc5()
|
D | 5x5s2p2-minmax-neonfma-1x4-acc5.c | 73 float32x4_t vi1x0246 = vmovq_n_f32(0.0f); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc5() local 116 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc5() 117 vi1x0246 = vi1x8ACE9BDF.val[0]; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc5() 233 const float32x4_t vi1x68AC = vextq_f32(vi1x0246, vi1x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc5()
|
D | 5x5s2p2-minmax-wasmsimd-arm-loadsplat-1x4-acc2.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc2() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc2() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc2() 299 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc2()
|
D | 5x5s2p2-minmax-wasmsimd-arm-loadsplat-1x4-acc3.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc3() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc3() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc3() 300 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc3()
|
D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-1x4.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4() 298 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4()
|
D | 5x5s2p2-minmax-wasmsimd-arm-loadsplat-1x4.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4() 298 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4()
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc5.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc5() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc5() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc5() 282 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc5()
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc2.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2() 279 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2()
|
D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-1x4-acc4.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc4() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc4() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc4() 301 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc4()
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc3.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc3() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc3() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc3() 280 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc3()
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4() 278 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4()
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc2.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2() 279 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2()
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc3.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc3() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc3() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc3() 280 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc3()
|
D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-1x4-acc3.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc3() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc3() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc3() 300 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc3()
|
D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-1x4-acc2.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc2() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc2() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc2() 299 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc2()
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc5.c | 76 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc5() local 140 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc5() 141 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc5() 282 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc5()
|
D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-1x4-acc5.c | 96 v128_t vi1x0246 = vzero; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc5() local 160 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc5() 161 vi1x0246 = vi1x8ACE; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc5() 302 const v128_t vi1x68AC = wasm_v32x4_shuffle(vi1x0246, vi1x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_1x4_acc5()
|