init v0.23.0

Signed-off-by: Sun Ruoxi <sunruoxi@4paradigm.com>

vllm_ascend/patch/worker/__init__.py

@@ -15,5 +15,75 @@
 # limitations under the License.
 #
 
-from vllm_ascend.patch.worker import patch_common  # noqa: F401
-from vllm_ascend.patch.worker import patch_main  # noqa: F401
+from vllm.triton_utils import HAS_TRITON
+
+from vllm_ascend.utils import is_310p, vllm_version_is
+
+# The v2 model runner is intentionally NOT made compatible with the v0.23.0
+# release. vLLM v0.23.0 and the verified main commit are diverged, and the v2
+# worker patches target main-only APIs; rather than maintain a separate v0.23.0
+# compatibility path we keep v2 main-only. With v0.23.0 installed this flag is
+# False, so none of the patch_v2.* / routed-experts-capture patches below are
+# imported and the v2 worker stays dormant (the release uses the v1 runner).
+if vllm_version_is("0.23.0"):
+    _V2_MODEL_RUNNER_SUPPORTED = False
+else:
+    _V2_MODEL_RUNNER_SUPPORTED = True
+
+if HAS_TRITON:
+    import vllm_ascend.patch.worker.patch_triton
+
+    if _V2_MODEL_RUNNER_SUPPORTED:
+        import vllm_ascend.patch.worker.patch_v2.patch_triton  # noqa
+
+
+import vllm_ascend.patch.worker.patch_process_weights_after_loading  # noqa
+import vllm_ascend.patch.worker.patch_weight_utils  # noqa
+import vllm_ascend.patch.worker.patch_distributed  # noqa
+import vllm_ascend.patch.worker.patch_minimax_m2  # noqa
+import vllm_ascend.patch.worker.patch_minimax_m2_linear_attn  # noqa
+import vllm_ascend.patch.worker.patch_mamba_utils  # noqa
+import vllm_ascend.patch.worker.patch_qwen3_next_mtp  # noqa
+
+if not is_310p():
+    import vllm_ascend.patch.worker.patch_qwen3_5  # noqa
+    import vllm_ascend.patch.worker.patch_qwen3_dflash  # noqa
+    import vllm_ascend.patch.worker.patch_qwen3vl  # noqa
+else:
+    import vllm_ascend.patch.worker.patch_idex_310  # noqa
+import vllm_ascend.patch.worker.patch_rejection_sampler  # noqa
+
+# torchair/npugraph_ex is only available on NPU; silently skip when missing
+# so that CPU-only environments (e.g. UT runners without torch_npu) can still
+# import this module without crashing.
+try:  # noqa: SIM105
+    import vllm_ascend.patch.worker.patch_npugraph_ex_triton  # noqa
+except ImportError:
+    pass
+import vllm_ascend.patch.worker.patch_kimi_k25  # noqa
+import vllm_ascend.patch.worker.patch_draft_quarot  # noqa
+import vllm_ascend.patch.worker.patch_eagle3_init  # noqa
+import vllm_ascend.patch.worker.patch_cudagraph  # noqa
+import vllm_ascend.patch.worker.patch_deepseek_mtp  # noqa
+import vllm_ascend.patch.worker.patch_deepseek_v2  # noqa
+import vllm_ascend.patch.worker.patch_gqa_c8  # noqa
+
+# vLLM's use_v2_model_runner may enable the v2 runner without the
+# VLLM_USE_V2_MODEL_RUNNER env var (e.g. based on model architecture).
+# We always patch it so that on Ascend the v2 runner is enabled only
+# when the env var is explicitly set.
+import vllm_ascend.patch.worker.patch_v2.patch_use_v2_model_runner  # noqa
+
+if not vllm_version_is("0.23.0"):
+    import vllm_ascend.patch.worker.patch_fused_moe  # noqa
+
+if _V2_MODEL_RUNNER_SUPPORTED:
+    import vllm_ascend.patch.worker.patch_v2.patch_uva  # noqa
+    import vllm_ascend.patch.worker.patch_v2.patch_input_batch  # noqa
+    import vllm_ascend.patch.worker.patch_v2.patch_model_state  # noqa
+    import vllm_ascend.patch.worker.patch_v2.patch_block_table  # noqa
+    import vllm_ascend.patch.worker.patch_v2.patch_attn_utils  # noqa
+
+# only patch routed experts capture in main2main.
+if _V2_MODEL_RUNNER_SUPPORTED:
+    import vllm_ascend.patch.worker.patch_routed_experts_capture  # noqa