Clock inversion is per slice (as BEL CLKINV)
Vivado GUI is misleading as it often shows it per FF, which is not actually true

CLB.SLICE_X0.AFF.FF_INV_CLK 00_51
CLB.SLICE_X1.AFF.FF_INV_CLK <0 candidates>
Clifford suggests X1 may be in adjacent tile
as he found with X1 INIT values
More research needed




CLB.SLICE_X0.AFF.FF_INV_CLK 00_51
CLB.SLICE_X0.FF_FDCE 00_21 00_24 00_25 00_26 00_29 00_35 29_01 29_12 30_01 30_03
CLB.SLICE_X0.FF_FDPE 00_21 00_24 00_25 00_26 00_29 00_35 29_01 30_01
CLB.SLICE_X0.FF_FDRE 00_21 00_24 00_25 00_26 00_29 29_01 29_12 30_01 30_03
CLB.SLICE_X0.FF_FDSE 00_21 00_24 00_25 00_26 00_29 00_35 29_01 30_01
CLB.SLICE_X0.FF_USED 00_21 00_24 00_25 00_26 00_29 29_01 30_01
CLB.SLICE_X1.AFF.FF_INV_CLK <0 candidates>
CLB.SLICE_X1.FF_FDCE 00_21 00_24 00_25 00_26 00_29 29_01 30_01 30_04 30_15
CLB.SLICE_X1.FF_FDPE 00_21 00_24 00_25 00_26 00_29 29_01 30_01
CLB.SLICE_X1.FF_FDRE 00_21 00_24 00_25 00_26 00_29 00_31 29_01 30_01 30_04 30_15
CLB.SLICE_X1.FF_FDSE 00_21 00_24 00_25 00_26 00_29 00_31 29_01 30_01
CLB.SLICE_X1.FF_USED 00_21 00_24 00_25 00_26 00_29 29_01 30_01


Following bits are always present with a FF
    00_21
    00_24
    00_25
    00_26
    00_29
    29_01
    30_01
Interesting bits are then
CLB.SLICE_X0.FF_FDCE 00_35 29_12 30_03
CLB.SLICE_X0.FF_FDPE 00_35
CLB.SLICE_X0.FF_FDSE 00_35
CLB.SLICE_X0.FF_FDRE 29_12 30_03