[sidcraftscode]

Pull Request Details

Description

_set_direction() in host/lib/rfnoc/fft_block_control.cpp pokes REG_DIRECTION and then checks it with get_direction(), which returns the _direction property. That is the value the resolver just passed in, so the check compares the value with itself and "FFT Direction readback failed!" can never fire. Length, magnitude and scaling peek their register after the write; now direction does too.

Related Issue

None.

Which devices/areas does this affect?

Host side of the RFNoC FFT block. On compat 2.x/3.x blocks the peek reads 0x1C, which fft_core.sv returns as reg_fft_direction.

Testing Done

• Ran fft_core.sv in Verilator (FFT pipeline stubbed out) with the same poke-then-peek on 0x1C: it reads back 0 and 1 as written.
• Built the old and new _set_direction() into a small harness with a register that ignores writes: the old code returns, the new one throws "Expected: 0, Received: 1".
• clang-format is clean. fft_block_test and fft_block_test_v2 weren't run (no Boost here); both mocks return the poked value on peek.

Checklist

• I have read the CONTRIBUTING document.
• My code follows the code style of this project. See CODING.md.
• I have updated the documentation accordingly.
• I have added tests to cover my changes, and all previous tests pass.
• I have checked all compat numbers if they need updating (FPGA compat,
  MPM compat, noc_shell, specific RFNoC block, ...)

[github-actions]

CLA Assistant Lite bot All contributors have signed the CLA ✍️ ✅

[sidcraftscode]

I have read the CLA Document and I hereby sign the CLA

[mbr0wn]

Yeah this looks good, thanks.