Add SSSE3 implementation for ipred_filter

Cycle times:

intra_pred_filter_w4_8bpc_c: 457.3
intra_pred_filter_w4_8bpc_ssse3: 66.7
intra_pred_filter_w8_8bpc_c: 1401.6
intra_pred_filter_w8_8bpc_ssse3: 208.6
intra_pred_filter_w16_8bpc_c: 2719.5
intra_pred_filter_w16_8bpc_ssse3: 431.3
intra_pred_filter_w32_8bpc_c: 6666.4
intra_pred_filter_w32_8bpc_ssse3: 936.7