vrend: Fix calculation of dual_src

The logic to calculate dual_src should be as close as possible to calculate dual_src_linked. Otherwise we could keep linking the same program in some corner case. For example, when the guest send blend_mode via alpha_src_factor, BLEND_SOFTLIGHT (0x9) would be treated as PIPE_BLENDFACTOR_SRC1_COLOR (0x9). The dual_src didn't match dual_src_linked in such case.

Signed-off-by: Lepton Wu lepton@chromium.org