• Liwei Wang's avatar
    Add SSSE3 implementation for the 4x16 and 16x4 blocks in itx · bf659082
    Liwei Wang authored
    Cycle times:
    inv_txfm_add_4x16_adst_adst_0_8bpc_c: 2203.6
    inv_txfm_add_4x16_adst_adst_0_8bpc_ssse3: 198.7
    inv_txfm_add_4x16_adst_adst_1_8bpc_c: 2235.1
    inv_txfm_add_4x16_adst_adst_1_8bpc_ssse3: 199.7
    inv_txfm_add_4x16_adst_adst_2_8bpc_c: 2199.1
    inv_txfm_add_4x16_adst_adst_2_8bpc_ssse3: 199.9
    inv_txfm_add_4x16_adst_dct_0_8bpc_c: 2272.4
    inv_txfm_add_4x16_adst_dct_0_8bpc_ssse3: 50.0
    inv_txfm_add_4x16_adst_dct_1_8bpc_c: 2281.6
    inv_txfm_add_4x16_adst_dct_1_8bpc_ssse3: 163.7
    inv_txfm_add_4x16_adst_dct_2_8bpc_c: 2262.5
    inv_txfm_add_4x16_adst_dct_2_8bpc_ssse3: 164.7
    inv_txfm_add_4x16_adst_flipadst_0_8bpc_c: 2456.5
    inv_txfm_add_4x16_adst_flipadst_0_8bpc_ssse3: 204.3
    inv_txfm_add_4x16_adst_flipadst_1_8bpc_c: 2349.1
    inv_txfm_add_4x16_adst_flipadst_1_8bpc_ssse3: 198.5
    inv_txfm_add_4x16_adst_flipadst_2_8bpc_c: 2241.5
    inv_txfm_add_4x16_adst_flipadst_2_8bpc_ssse3: 198.7
    inv_txfm_add_4x16_adst_identity_0_8bpc_c: 1574.7
    inv_txfm_add_4x16_adst_identity_0_8bpc_ssse3: 117.0
    inv_txfm_add_4x16_adst_identity_1_8bpc_c: 1576.3
    inv_txfm_add_4x16_adst_identity_1_8bpc_ssse3: 116.6
    inv_txfm_add_4x16_adst_identity_2_8bpc_c: 1572.9
    inv_txfm_add_4x16_adst_identity_2_8bpc_ssse3: 116.7
    inv_txfm_add_4x16_dct_adst_0_8bpc_c: 2162.8
    inv_txfm_add_4x16_dct_adst_0_8bpc_ssse3: 187.6
    inv_txfm_add_4x16_dct_adst_1_8bpc_c: 2180.4
    inv_txfm_add_4x16_dct_adst_1_8bpc_ssse3: 185.6
    inv_txfm_add_4x16_dct_adst_2_8bpc_c: 2165.1
    inv_txfm_add_4x16_dct_adst_2_8bpc_ssse3: 184.9
    inv_txfm_add_4x16_dct_dct_0_8bpc_c: 2233.7
    inv_txfm_add_4x16_dct_dct_0_8bpc_ssse3: 49.5
    inv_txfm_add_4x16_dct_dct_1_8bpc_c: 2770.4
    inv_txfm_add_4x16_dct_dct_1_8bpc_ssse3: 148.4
    inv_txfm_add_4x16_dct_dct_2_8bpc_c: 2288.7
    inv_txfm_add_4x16_dct_dct_2_8bpc_ssse3: 149.0
    inv_txfm_add_4x16_dct_flipadst_0_8bpc_c: 2242.0
    inv_txfm_add_4x16_dct_flipadst_0_8bpc_ssse3: 185.8
    inv_txfm_add_4x16_dct_flipadst_1_8bpc_c: 2249.6
    inv_txfm_add_4x16_dct_flipadst_1_8bpc_ssse3: 188.4
    inv_txfm_add_4x16_dct_flipadst_2_8bpc_c: 2237.3
    inv_txfm_add_4x16_dct_flipadst_2_8bpc_ssse3: 185.1
    inv_txfm_add_4x16_dct_identity_0_8bpc_c: 1532.3
    inv_txfm_add_4x16_dct_identity_0_8bpc_ssse3: 63.7
    inv_txfm_add_4x16_dct_identity_1_8bpc_c: 1534.5
    inv_txfm_add_4x16_dct_identity_1_8bpc_ssse3: 63.6
    inv_txfm_add_4x16_dct_identity_2_8bpc_c: 1548.1
    inv_txfm_add_4x16_dct_identity_2_8bpc_ssse3: 101.6
    inv_txfm_add_4x16_flipadst_adst_0_8bpc_c: 2205.2
    inv_txfm_add_4x16_flipadst_adst_0_8bpc_ssse3: 201.6
    inv_txfm_add_4x16_flipadst_adst_1_8bpc_c: 2222.0
    inv_txfm_add_4x16_flipadst_adst_1_8bpc_ssse3: 202.6
    inv_txfm_add_4x16_flipadst_adst_2_8bpc_c: 2205.2
    inv_txfm_add_4x16_flipadst_adst_2_8bpc_ssse3: 205.7
    inv_txfm_add_4x16_flipadst_dct_0_8bpc_c: 2294.9
    inv_txfm_add_4x16_flipadst_dct_0_8bpc_ssse3: 50.0
    inv_txfm_add_4x16_flipadst_dct_1_8bpc_c: 2304.2
    inv_txfm_add_4x16_flipadst_dct_1_8bpc_ssse3: 164.5
    inv_txfm_add_4x16_flipadst_dct_2_8bpc_c: 2292.7
    inv_txfm_add_4x16_flipadst_dct_2_8bpc_ssse3: 164.5
    inv_txfm_add_4x16_flipadst_flipadst_0_8bpc_c: 2281.3
    inv_txfm_add_4x16_flipadst_flipadst_0_8bpc_ssse3: 202.9
    inv_txfm_add_4x16_flipadst_flipadst_1_8bpc_c: 2258.7
    inv_txfm_add_4x16_flipadst_flipadst_1_8bpc_ssse3: 202.4
    inv_txfm_add_4x16_flipadst_flipadst_2_8bpc_c: 2261.0
    inv_txfm_add_4x16_flipadst_flipadst_2_8bpc_ssse3: 201.3
    inv_txfm_add_4x16_flipadst_identity_0_8bpc_c: 1580.5
    inv_txfm_add_4x16_flipadst_identity_0_8bpc_ssse3: 116.1
    inv_txfm_add_4x16_flipadst_identity_1_8bpc_c: 1578.7
    inv_txfm_add_4x16_flipadst_identity_1_8bpc_ssse3: 116.7
    inv_txfm_add_4x16_flipadst_identity_2_8bpc_c: 1590.8
    inv_txfm_add_4x16_flipadst_identity_2_8bpc_ssse3: 117.4
    inv_txfm_add_4x16_identity_adst_0_8bpc_c: 1949.0
    inv_txfm_add_4x16_identity_adst_0_8bpc_ssse3: 170.9
    inv_txfm_add_4x16_identity_adst_1_8bpc_c: 1947.4
    inv_txfm_add_4x16_identity_adst_1_8bpc_ssse3: 171.0
    inv_txfm_add_4x16_identity_adst_2_8bpc_c: 1948.7
    inv_txfm_add_4x16_identity_adst_2_8bpc_ssse3: 170.3
    inv_txfm_add_4x16_identity_dct_0_8bpc_c: 2022.3
    inv_txfm_add_4x16_identity_dct_0_8bpc_ssse3: 59.2
    inv_txfm_add_4x16_identity_dct_1_8bpc_c: 2020.8
    inv_txfm_add_4x16_identity_dct_1_8bpc_ssse3: 133.7
    inv_txfm_add_4x16_identity_dct_2_8bpc_c: 2020.2
    inv_txfm_add_4x16_identity_dct_2_8bpc_ssse3: 133.2
    inv_txfm_add_4x16_identity_flipadst_0_8bpc_c: 2024.7
    inv_txfm_add_4x16_identity_flipadst_0_8bpc_ssse3: 170.3
    inv_txfm_add_4x16_identity_flipadst_1_8bpc_c: 2021.8
    inv_txfm_add_4x16_identity_flipadst_1_8bpc_ssse3: 170.0
    inv_txfm_add_4x16_identity_flipadst_2_8bpc_c: 2022.5
    inv_txfm_add_4x16_identity_flipadst_2_8bpc_ssse3: 169.9
    inv_txfm_add_4x16_identity_identity_0_8bpc_c: 1328.4
    inv_txfm_add_4x16_identity_identity_0_8bpc_ssse3: 87.7
    inv_txfm_add_4x16_identity_identity_1_8bpc_c: 1330.9
    inv_txfm_add_4x16_identity_identity_1_8bpc_ssse3: 87.7
    inv_txfm_add_4x16_identity_identity_2_8bpc_c: 1327.3
    inv_txfm_add_4x16_identity_identity_2_8bpc_ssse3: 87.6
    inv_txfm_add_16x4_adst_adst_0_8bpc_c: 2166.3
    inv_txfm_add_16x4_adst_adst_0_8bpc_ssse3: 186.3
    inv_txfm_add_16x4_adst_adst_1_8bpc_c: 2166.9
    inv_txfm_add_16x4_adst_adst_1_8bpc_ssse3: 184.9
    inv_txfm_add_16x4_adst_adst_2_8bpc_c: 2167.2
    inv_txfm_add_16x4_adst_adst_2_8bpc_ssse3: 185.2
    inv_txfm_add_16x4_adst_dct_0_8bpc_c: 2123.2
    inv_txfm_add_16x4_adst_dct_0_8bpc_ssse3: 172.1
    inv_txfm_add_16x4_adst_dct_1_8bpc_c: 2124.2
    inv_txfm_add_16x4_adst_dct_1_8bpc_ssse3: 171.2
    inv_txfm_add_16x4_adst_dct_2_8bpc_c: 2122.8
    inv_txfm_add_16x4_adst_dct_2_8bpc_ssse3: 171.8
    inv_txfm_add_16x4_adst_flipadst_0_8bpc_c: 2213.3
    inv_txfm_add_16x4_adst_flipadst_0_8bpc_ssse3: 189.6
    inv_txfm_add_16x4_adst_flipadst_1_8bpc_c: 2227.7
    inv_txfm_add_16x4_adst_flipadst_1_8bpc_ssse3: 188.4
    inv_txfm_add_16x4_adst_flipadst_2_8bpc_c: 2228.5
    inv_txfm_add_16x4_adst_flipadst_2_8bpc_ssse3: 188.4
    inv_txfm_add_16x4_adst_identity_0_8bpc_c: 1906.7
    inv_txfm_add_16x4_adst_identity_0_8bpc_ssse3: 154.3
    inv_txfm_add_16x4_adst_identity_1_8bpc_c: 1905.2
    inv_txfm_add_16x4_adst_identity_1_8bpc_ssse3: 155.6
    inv_txfm_add_16x4_adst_identity_2_8bpc_c: 1905.6
    inv_txfm_add_16x4_adst_identity_2_8bpc_ssse3: 156.3
    inv_txfm_add_16x4_dct_adst_0_8bpc_c: 2209.8
    inv_txfm_add_16x4_dct_adst_0_8bpc_ssse3: 37.4
    inv_txfm_add_16x4_dct_adst_1_8bpc_c: 2209.8
    inv_txfm_add_16x4_dct_adst_1_8bpc_ssse3: 157.9
    inv_txfm_add_16x4_dct_adst_2_8bpc_c: 2221.1
    inv_txfm_add_16x4_dct_adst_2_8bpc_ssse3: 158.5
    inv_txfm_add_16x4_dct_dct_0_8bpc_c: 2177.5
    inv_txfm_add_16x4_dct_dct_0_8bpc_ssse3: 29.6
    inv_txfm_add_16x4_dct_dct_1_8bpc_c: 2179.3
    inv_txfm_add_16x4_dct_dct_1_8bpc_ssse3: 144.9
    inv_txfm_add_16x4_dct_dct_2_8bpc_c: 2177.8
    inv_txfm_add_16x4_dct_dct_2_8bpc_ssse3: 143.7
    inv_txfm_add_16x4_dct_flipadst_0_8bpc_c: 2293.6
    inv_txfm_add_16x4_dct_flipadst_0_8bpc_ssse3: 38.3
    inv_txfm_add_16x4_dct_flipadst_1_8bpc_c: 2293.2
    inv_txfm_add_16x4_dct_flipadst_1_8bpc_ssse3: 163.9
    inv_txfm_add_16x4_dct_flipadst_2_8bpc_c: 2301.3
    inv_txfm_add_16x4_dct_flipadst_2_8bpc_ssse3: 163.7
    inv_txfm_add_16x4_dct_identity_0_8bpc_c: 1977.7
    inv_txfm_add_16x4_dct_identity_0_8bpc_ssse3: 39.9
    inv_txfm_add_16x4_dct_identity_1_8bpc_c: 1978.7
    inv_txfm_add_16x4_dct_identity_1_8bpc_ssse3: 126.8
    inv_txfm_add_16x4_dct_identity_2_8bpc_c: 1979.5
    inv_txfm_add_16x4_dct_identity_2_8bpc_ssse3: 128.1
    inv_txfm_add_16x4_flipadst_adst_0_8bpc_c: 2175.6
    inv_txfm_add_16x4_flipadst_adst_0_8bpc_ssse3: 185.1
    inv_txfm_add_16x4_flipadst_adst_1_8bpc_c: 2175.7
    inv_txfm_add_16x4_flipadst_adst_1_8bpc_ssse3: 185.7
    inv_txfm_add_16x4_flipadst_adst_2_8bpc_c: 2173.1
    inv_txfm_add_16x4_flipadst_adst_2_8bpc_ssse3: 185.0
    inv_txfm_add_16x4_flipadst_dct_0_8bpc_c: 2140.5
    inv_txfm_add_16x4_flipadst_dct_0_8bpc_ssse3: 172.0
    inv_txfm_add_16x4_flipadst_dct_1_8bpc_c: 2147.5
    inv_txfm_add_16x4_flipadst_dct_1_8bpc_ssse3: 171.9
    inv_txfm_add_16x4_flipadst_dct_2_8bpc_c: 2148.5
    inv_txfm_add_16x4_flipadst_dct_2_8bpc_ssse3: 172.0
    inv_txfm_add_16x4_flipadst_flipadst_0_8bpc_c: 2240.6
    inv_txfm_add_16x4_flipadst_flipadst_0_8bpc_ssse3: 191.3
    inv_txfm_add_16x4_flipadst_flipadst_1_8bpc_c: 2243.5
    inv_txfm_add_16x4_flipadst_flipadst_1_8bpc_ssse3: 193.2
    inv_txfm_add_16x4_flipadst_flipadst_2_8bpc_c: 2242.9
    inv_txfm_add_16x4_flipadst_flipadst_2_8bpc_ssse3: 192.0
    inv_txfm_add_16x4_flipadst_identity_0_8bpc_c: 1919.2
    inv_txfm_add_16x4_flipadst_identity_0_8bpc_ssse3: 155.1
    inv_txfm_add_16x4_flipadst_identity_1_8bpc_c: 1925.2
    inv_txfm_add_16x4_flipadst_identity_1_8bpc_ssse3: 155.2
    inv_txfm_add_16x4_flipadst_identity_2_8bpc_c: 2084.8
    inv_txfm_add_16x4_flipadst_identity_2_8bpc_ssse3: 155.0
    inv_txfm_add_16x4_identity_adst_0_8bpc_c: 1498.5
    inv_txfm_add_16x4_identity_adst_0_8bpc_ssse3: 107.6
    inv_txfm_add_16x4_identity_adst_1_8bpc_c: 1499.5
    inv_txfm_add_16x4_identity_adst_1_8bpc_ssse3: 107.0
    inv_txfm_add_16x4_identity_adst_2_8bpc_c: 1498.9
    inv_txfm_add_16x4_identity_adst_2_8bpc_ssse3: 107.9
    inv_txfm_add_16x4_identity_dct_0_8bpc_c: 1471.9
    inv_txfm_add_16x4_identity_dct_0_8bpc_ssse3: 45.4
    inv_txfm_add_16x4_identity_dct_1_8bpc_c: 1476.4
    inv_txfm_add_16x4_identity_dct_1_8bpc_ssse3: 45.5
    inv_txfm_add_16x4_identity_dct_2_8bpc_c: 1459.8
    inv_txfm_add_16x4_identity_dct_2_8bpc_ssse3: 92.3
    inv_txfm_add_16x4_identity_flipadst_0_8bpc_c: 1548.7
    inv_txfm_add_16x4_identity_flipadst_0_8bpc_ssse3: 112.1
    inv_txfm_add_16x4_identity_flipadst_1_8bpc_c: 1548.2
    inv_txfm_add_16x4_identity_flipadst_1_8bpc_ssse3: 111.7
    inv_txfm_add_16x4_identity_flipadst_2_8bpc_c: 1547.2
    inv_txfm_add_16x4_identity_flipadst_2_8bpc_ssse3: 114.1
    inv_txfm_add_16x4_identity_identity_0_8bpc_c: 1271.5
    inv_txfm_add_16x4_identity_identity_0_8bpc_ssse3: 74.5
    inv_txfm_add_16x4_identity_identity_1_8bpc_c: 1266.8
    inv_txfm_add_16x4_identity_identity_1_8bpc_ssse3: 74.5
    inv_txfm_add_16x4_identity_identity_2_8bpc_c: 1268.0
    inv_txfm_add_16x4_identity_identity_2_8bpc_ssse3: 74.6
    bf659082
itx_init_tmpl.c 6.57 KB