[Feat]Qwen3 Moe supports npu_add_rms_norm_quant op by default, update op with bias, resolve conflict with weight prefetch (#3465)

### What this PR does / why we need it?
1.qwen3 moe uses add_rms_norm_quant op instead of 'add_rms_norm op and
quant op' during quantization scene.
2.torch_npu.add_rms_norm_quant op fixed accuracy while model weights is
quantized by anti_method m4, m4 quantization is asymmetric outlier
suppression method, it will generate none-zero norm bias,
add_rms_norm_quant op updated to add this parameter to calculate.
3. add torch-npu check

### Does this PR introduce _any_ user-facing change?
new feature works if torch_npu version >= torch_npu-2.7.1.dev20250919

### How was this patch tested?
1.no special parameters to set, no new envs to set. new feature works if
torch_npu version >= torch_npu-2.7.1.dev20250919
2.use qwen3 moe quantization model to test ,such as
Qwen3-235B-A22B-W8A8, Qwen3-30B-A3B-W8A8,
Qwen3-235B-A22B-Instruct-2507-m4 (anti_method m4)

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: h30027576 <huangdong51@huawei.com>

vllm_ascend/ops/weight_prefetch.py

@@ -7,6 +7,7 @@ from vllm.forward_context import get_forward_context
 from vllm_ascend.ascend_config import WeightPrefetchConfig
 from vllm_ascend.ops.linear import (AscendQKVParallelLinear,
                                     AscendRowParallelLinear)
+from vllm_ascend.utils import version_check
 
 SUPPORTED_MODULES = ["attn", "mlp", "moe"]
 MOE_PREFETCH_TOKEN_THRESHOLD = 96
@@ -82,14 +83,15 @@ class WeightPrefetchMethod:
         if not self.moe.is_active_this_forward:
             return
         forward_context = get_forward_context()
+        if not version_check():
+            forward_context.layer_idx += 1
         weight = forward_context.model_instance.model.layers[
-            forward_context.layer_idx].mlp.experts.w13_weight
+            forward_context.layer_idx - 1].mlp.experts.w13_weight
         weight_size = weight.data.element_size() * weight.data.numel(
         ) * self.moe.prefetch_ratio.get(prefix, 0)
         torch.ops.vllm.prefetch_preprocess(weight=weight,
                                            start_flag=None,
                                            max_weight_size=int(weight_size))
-        forward_context.layer_idx += 1
 
     def maybe_prefetch_moe_weight_postprocess(self, stop_flag: torch.Tensor):
         if not self.moe.is_active_this_forward: