[refactor] refactor weight trans nz and transpose (#4878)

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

Now `VLLM_ASCEND_ENABLE_NZ` will have three options:
0: disable nz;
1: only quant case enable nz;
2: enable nz as long as possible;

And `VLLM_ASCEND_ENABLE_NZ`=1 by default.

All cases are shown in the table below:

|  | W4A4 | W4A8 | W8A8 | fp16/bf16 | fp32 |
|---|---|---|---|---|---|
| trans nz | can't support nz | trans nz by default | trans nz by
default | trans nz when VLLM_ASCEND_ENABLE_NZ is 2 | can't support nz |
| transpose | only support not transpose case | only support transpose
case | only support transpose case | linear: only support not transpose
case<br>gmm: only support transpose case | same to fp16/bf16 |

Some exceptional cases:
1. MLAPO op need to do some additional processing on the weights,
including trans nz. If use MLAPO op, some weight will be transformed to
nz forcely;
2. MLA/SFA's weight `W_UV` will be used by op
`torch.ops._C_ascend.batch_matmul_transpose`, and this op can't support
nz currently;

### Does this PR introduce _any_ user-facing change?
Now fp16/bf16 weight will not trans nz by default.

### How was this patch tested?

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

Signed-off-by: zzzzwwjj <1183291235@qq.com>

tests/ut/ops/test_linear.py

@@ -1,3 +1,4 @@
+import os
 import unittest
 from unittest import mock
 from unittest.mock import MagicMock, patch
@@ -61,22 +62,24 @@ class TestAscendUnquantizedLinearMethod(TestBase):
         mock_dtype = mock.PropertyMock(return_value=torch.float16)
         type(self.layer.weight.data).dtype = mock_dtype
 
-    @mock.patch("vllm_ascend.ops.linear.is_enable_nz")
+    @patch.dict(os.environ, {"VLLM_ASCEND_ENABLE_NZ": "0"})
     @mock.patch("torch_npu.npu_format_cast")
-    def test_process_weights_after_loading_enable_nz(self, mock_format_cast,
-                                                     mock_is_nz):
-        mock_is_nz.return_value = 1
-        self.method.process_weights_after_loading(self.layer)
-        mock_format_cast.assert_called_once()
-
-    @mock.patch("vllm_ascend.ops.linear.is_enable_nz")
-    @mock.patch("torch_npu.npu_format_cast")
-    def test_process_weights_after_loading_disable_nz(self, mock_format_cast,
-                                                      mock_is_nz):
-        mock_is_nz.return_value = 0
+    def test_process_weights_after_loading_with_nz0(self, mock_format_cast):
         self.method.process_weights_after_loading(self.layer)
         mock_format_cast.assert_not_called()
 
+    @patch.dict(os.environ, {"VLLM_ASCEND_ENABLE_NZ": "1"})
+    @mock.patch("torch_npu.npu_format_cast")
+    def test_process_weights_after_loading_with_nz1(self, mock_format_cast):
+        self.method.process_weights_after_loading(self.layer)
+        mock_format_cast.assert_not_called()
+
+    @patch.dict(os.environ, {"VLLM_ASCEND_ENABLE_NZ": "2"})
+    @mock.patch("torch_npu.npu_format_cast")
+    def test_process_weights_after_loading_with_nz2(self, mock_format_cast):
+        self.method.process_weights_after_loading(self.layer)
+        mock_format_cast.assert_called_once()
+
 
 class TestAscendRowParallelLinear(BaseLinearTest):