tools(rdna): drop the TunableOp mirror and unused probes

docs/rdna2/V620-CANDIDATE.md

@@ -121,7 +121,6 @@ that selection.
 Reproduce the CPU checks without downloading weights:
 
 ```bash
-.venv/bin/python tools/rdna2/test_startup_plan_cpu.py
 .venv/bin/python -m pytest --noconftest -q \
   tests/distributed/test_rdna_p2p.py \
   tests/kernels/quantization/test_rdna2_w4a16_selection.py \

docs/rdna2/V620-QSA-LIVE-PREFILL.md

@@ -75,7 +75,7 @@ export PYTORCH_TUNABLEOP_ENABLED=0 PYTORCH_TUNABLEOP_TUNING=0 PYTORCH_TUNABLEOP_
 sdk=$runtime/.venv/lib/python3.12/site-packages/_rocm_sdk_core
 export PATH="$runtime/.venv/bin:/opt/rocm/core-10.0/bin:/opt/rocm/core-10.0/llvm/bin:$PATH"
 export LD_LIBRARY_PATH="$sdk/lib:$sdk/lib/host-math/lib:/opt/rocm/core-10.0/lib"
-source "$root/source/tools/rdna2/tunableop_env.sh"
+source "$root/source/tools/rdna2_028/tunableop_env.sh"
 configure_v620_tunableop "$runtime/.venv/lib/python3.12/site-packages/_rocm_sdk_libraries/lib/librocblas.so.5" "$runtime/tunableop"
 command=("$runtime/.venv/bin/python" -m vllm.entrypoints.openai.api_server
  --model ${V620_MODEL:?set V620_MODEL} --served-model-name active qwen3.8-flash-next

tests/tools/test_v620_tuning_coverage.py

@@ -77,7 +77,7 @@ def test_unknown_library_disables_tuning_without_stopping_service(self):
                     'printf "%s:%s" "$PYTORCH_TUNABLEOP_ENABLED" '
                     '"${PYTORCH_TUNABLEOP_FILENAME-unset}"',
                     "check-tuning",
-                    str(ROOT / "tools/rdna2/tunableop_env.sh"),
+                    str(ROOT / "tools/rdna2_028/tunableop_env.sh"),
                     str(library),
                     str(ROWS.parent),
                 ],

tools/rdna2/serve_v620_baseline.sh

@@ -36,7 +36,7 @@ export PYTORCH_TUNABLEOP_HIPBLASLT_ENABLED=0
 sdk=$runtime/.venv/lib/python3.12/site-packages/_rocm_sdk_core
 export PATH="$runtime/.venv/bin:/opt/rocm/core-10.0/bin:/opt/rocm/core-10.0/llvm/bin:$PATH"
 export LD_LIBRARY_PATH="$sdk/lib:$sdk/lib/host-math/lib:/opt/rocm/core-10.0/lib"
-source "$source_dir/tools/rdna2/tunableop_env.sh"
+source "$source_dir/tools/rdna2_028/tunableop_env.sh"
 configure_v620_tunableop "$runtime/.venv/lib/python3.12/site-packages/_rocm_sdk_libraries/lib/librocblas.so.5" "$runtime/tunableop"
 command=("$runtime/.venv/bin/python" -m vllm.entrypoints.openai.api_server
     --model "$model" --served-model-name active qwen3.8-flash-next

tools/rdna2/serve_v620_candidate.sh

@@ -94,8 +94,8 @@ export PATH="$venv/bin:/opt/rocm/bin:/opt/rocm/llvm/bin:$PATH"
 export LD_LIBRARY_PATH="$runtime_sdk/lib:$runtime_sdk/lib/host-math/lib:/opt/rocm/core-10.0/lib${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}"
 if [[ ${V620_TUNABLEOP:-0} == 1 ]]; then
     # V620_ROCBLAS_LIBRARY must identify the library loaded by this wheel SDK.
-    # shellcheck source=tools/rdna2/tunableop_env.sh
-    source "$source_dir/tools/rdna2/tunableop_env.sh"
+    # shellcheck source=tools/rdna2_028/tunableop_env.sh
+    source "$source_dir/tools/rdna2_028/tunableop_env.sh"
     rows_root=${V620_TUNABLEOP_ROOT:-$test_root/tunableop}
     if [[ ! -d $rows_root && -z ${V620_TUNABLEOP_ROOT:-} ]]; then
         rows_root=$source_dir/tunableop

tools/rdna2/serve_v620_piecewise.sh

@@ -69,8 +69,8 @@ fi
 # TunableOp: solution IDs belong to a rocBLAS build, not just a version number.
 # The rows root holds one rocblas-<sha256-first-12>/ directory per build.
 if [[ ${V620_TUNABLEOP:-1} == 1 ]]; then
-    # shellcheck source=tools/rdna2/tunableop_env.sh
-    source "$source_dir/tools/rdna2/tunableop_env.sh"
+    # shellcheck source=tools/rdna2_028/tunableop_env.sh
+    source "$source_dir/tools/rdna2_028/tunableop_env.sh"
     rows_root=${V620_TUNABLEOP_ROOT:-$source_dir/../pr17-tunableop}
     configure_v620_tunableop \
         "${V620_ROCBLAS_LIBRARY:-$rocm_lib/librocblas.so.5}" "$rows_root"