/external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
D | 5x5s2p2-minmax-neon-1x4-acc3.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc3() local 234 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc3() local
|
D | 5x5s2p2-minmax-neonfma-1x4-acc2.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2() local 233 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2() local
|
D | 5x5s2p2-minmax-neonfma-1x4-acc3.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc3() local 234 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc3() local
|
D | 5x5s2p2-minmax-neon-1x4.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4() local 232 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4() local
|
D | 5x5s2p2-minmax-neonfma-1x4.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4() local 232 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4() local
|
D | 5x5s2p2-minmax-neon-1x4-acc2.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc2() local 233 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc2() local
|
D | 5x5s2p2-minmax-neon-1x4-acc4.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc4() local 235 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc4() local
|
D | 5x5s2p2-minmax-neonfma-1x4-acc4.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc4() local 235 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc4() local
|
D | 5x5s2p2-minmax-neon-1x4-acc5.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc5() local 236 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4_acc5() local
|
D | 5x5s2p2-minmax-neonfma-1x4-acc5.c | 122 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc5() local 236 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc5() local
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc2.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2() local 282 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2() local
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc2.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2() local 282 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2() local
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4() local 281 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4() local
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4() local 281 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4() local
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc4.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc4() local 284 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc4() local
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc5.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc5() local 285 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc5() local
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc5.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc5() local 285 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc5() local
|
D | 5x5s2p2-minmax-wasmsimd-arm-splat-1x4-acc3.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc3() local 283 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc3() local
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc4.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc4() local 284 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc4() local
|
D | 5x5s2p2-minmax-wasmsimd-x86-splat-1x4-acc3.c | 146 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc3() local 283 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc3() local
|
D | 5x5s2p2-minmax-neon-2x4-acc2.c | 150 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_2x4_acc2() local 309 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_2x4_acc2() local
|
D | 5x5s2p2-minmax-neonfma-2x4-acc2.c | 150 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_2x4_acc2() local 309 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_2x4_acc2() local
|
D | 5x5s2p2-minmax-neon-2x4.c | 150 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_2x4() local 307 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_2x4() local
|
D | 5x5s2p2-minmax-neonfma-2x4.c | 150 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE9BDF.val[0], 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_2x4() local 307 const float32x4_t vi4x68AC = vextq_f32(vi4x0246, vi4x8ACE, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_2x4() local
|
D | 5x5s2p2-minmax-wasmsimd-arm-loadsplat-1x4-acc3.c | 166 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc3() local 303 const v128_t vi4x68AC = wasm_v32x4_shuffle(vi4x0246, vi4x8ACE, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_1x4_acc3() local
|