Revert "fix(fpga): IOB=TRUE on FT2232H pads to meet 5 ns FPGA launch budget"

This reverts commit 94bf6944a3.
This commit is contained in:
Jason
2026-04-20 20:20:02 +05:45
parent 94bf6944a3
commit bf89984f04
@@ -354,18 +354,6 @@ set_property DRIVE 8 [get_ports {ft_oe_n}]
set_property DRIVE 8 [get_ports {ft_siwu}] set_property DRIVE 8 [get_ports {ft_siwu}]
set_property DRIVE 8 [get_ports {ft_data[*]}] set_property DRIVE 8 [get_ports {ft_data[*]}]
# IOB packing: force output/tristate FFs to pack into the IOB to meet the
# FT2232H 5 ns FPGA launch budget (period 16.667 ns t_su 11.667 ns ≈ 5 ns).
# Without this, fabric→OBUFT routing eats 2-3 ns and Vivado starts
# replicating the tristate driver (ft_data_TRI[*]_repN) trying to rescue
# timing. The FSM flops in usb_data_interface_ft2232h.v already sit at
# the output boundary, so packing is safe and does not require RTL change.
set_property IOB TRUE [get_ports {ft_data[*]}]
set_property IOB TRUE [get_ports {ft_rd_n}]
set_property IOB TRUE [get_ports {ft_wr_n}]
set_property IOB TRUE [get_ports {ft_oe_n}]
set_property IOB TRUE [get_ports {ft_siwu}]
# ft_clkout constrained above in CLOCK CONSTRAINTS section (C4, 60 MHz) # ft_clkout constrained above in CLOCK CONSTRAINTS section (C4, 60 MHz)
# -------------------------------------------------------------------------- # --------------------------------------------------------------------------