[misc] move mxfp_compat into device to decouple from quantization init chain (#6918)

### What this PR does / why we need it?
`mxfp_compat` only provides dtype/symbol compatibility helpers for
different `torch_npu` versions, but it was placed under
`vllm_ascend.quantization`. Importing it from device/ops paths could
trigger `quantization/__init__.py` and pull in heavy quantization method
dependencies, increasing startup coupling and causing import-cycle risk
(especially on 310P paths).

### Does this PR introduce _any_ user-facing change?
No functional behavior change intended.

### How was this patch tested?
CI passed.

- vLLM version: v0.16.0
- vLLM main:
15d76f74e2

---------

Signed-off-by: linfeng-yuan <1102311262@qq.com>

vllm_ascend/device/device_op.py

@@ -18,7 +18,7 @@
 import torch
 import torch_npu
 
-from vllm_ascend.quantization.mxfp_compat import (
+from vllm_ascend.device.mxfp_compat import (
     FLOAT4_E2M1FN_X2_DTYPE,
     FLOAT8_E8M0FNU_DTYPE,
     HIFLOAT8_DTYPE,