[gfx1250] Add 256x256 compute-bound GEMM (WIP)

[aoli26]

Motivation

Add 256x256 tile specific compute-bound A8W4 kernel for gfx1250. A4W4 and A8W8 WIP.

Technical Details

• gemm_a8w4_256x256_gfx1250.py (new): hand-scheduled for exactly one profile — 256x256x128 tile, 2x2 waves. Hand-placed s_wait_dscnt/READY fences, odd/even wave-parity loop clones, operand-major planar LDS with B in its own 64 KiB segment.
• tdm_ops.py: add update_tensor_descriptor_2d_lds_addr (re-point one persistent descriptor per ring slot); widen lds_memref to accept !fly.memref. Both additive.
• Also folds the a8w8 mx32/mx128/ptpc variants into one launcher via is_mxscale / block_size constexprs.

Test Plan

pytest tests/kernels/test_gemm_fp8fp4_gfx1250.py on gfx1250.

Test Result

All tests passed.

Submission Checklist

• Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.