[npugraph_ex]enable npugraph_ex by default (#6664)

### What this PR does / why we need it?

This pull request enables the `npugraph_ex` backend by default to
improve performance on Ascend NPUs, as proposed in the
[RFC](https://github.com/vllm-project/vllm-ascend/issues/6214).


### Does this PR introduce _any_ user-facing change?

Yes. `npugraph_ex` is now enabled by default. Users can disable it by
setting `enable: false` in the `npugraph_ex_config` section of the
`additional_config`.

### How was this patch tested?

CI passed. The changes are covered by existing and new E2E tests
(`test_aclgraph_accuracy.py`) and unit tests (`test_ascend_config.py`)
that have been updated to reflect the new default behavior. The tests
verify correctness and consistency with `npugraph_ex` enabled and
disabled, as well as with the new static kernel option.

Signed-off-by: huyuanquan1 <huyuanquan1@huawei.com>
Co-authored-by: huyuanquan1 <huyuanquan1@huawei.com>

vllm_ascend/patch/__init__.py

@@ -249,3 +249,17 @@
 #       make unquantized_gemm as a customop.
 #    Future Plan:
 #       Remove this patch when vLLM support the operator as customop.
+#
+# ** 13. File: worker/patch_npugraph_ex_triton.py**
+# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+#   1. `torchair.core._concrete_graph.ValuePack`,
+#      `torchair.npu_fx_compiler._unpack_meta`,
+#      `torchair.npu_fx_compiler._NpuGraphConverter._unpack_npu`
+#    Why:
+#       In the Triton scenario, npugraph_ex backend needs to process the value pack of the input parameters.
+#    How：
+#       Supplement the relevant processing logic through patches.
+#    Related PR (if no, explain why):
+#       https://gitcode.com/Ascend/torchair/pull/2575
+#    Future Plan:
+#       Remove this patch when the PTA version used by vllm-ascend has been upgraded.