[v0.11.0][refactor] refactor SequenceRowParallelOp forward (#3654)

### What this PR does / why we need it?
This PR refactors SequenceRowParallelOp forward. In order to further
expand the operator inclusion scope in dynamic judgment scenarios, this
PR customizes the entire matmul computation and communication as a
custom operator masking. With this refactor, it will support directly
writing code such as common operation fusion into the
SequenceRowParallelOp class's member function matmul_and_reduce, without
the need to register more redundant custom masking operators.

### How was this patch tested?
CI passed with new added/existing test.

Signed-off-by: rjg-lyh <1318825571@qq.com>

tests/ut/ops/test_linear.py

@@ -4,6 +4,7 @@ from unittest import mock
 from unittest.mock import MagicMock, patch
 
 import torch
+from vllm import config
 
 from tests.ut.base import TestBase
 from vllm_ascend import ascend_config
@@ -106,6 +107,9 @@ class TestAscendRowParallelLinear(BaseLinearTest):
         linear(input_tensor)
 
     def test_oproj_tp(self):
+
+        config._current_vllm_config = MagicMock()
+
         ascend_config._ASCEND_CONFIG = MagicMock()
         ascend_config._ASCEND_CONFIG.oproj_tensor_parallel_size = 2
 

vllm_ascend/ops/linear.py

@@ -26,6 +26,7 @@ import torch
 import torch.nn as nn
 import torch_npu
 from torch.nn.parameter import Parameter
+from vllm.config import get_current_vllm_config
 from vllm.distributed import divide
 from vllm.model_executor.layers.linear import (  # noqa
     WEIGHT_LOADER_V2_SUPPORTED, ColumnParallelLinear, LinearBase,
@@ -234,6 +235,13 @@ class AscendRowParallelLinear(RowParallelLinear):
         return_bias: bool = True,
         disable_tp: bool = False,
     ):
+        compilation_config = get_current_vllm_config().compilation_config
+        # TODO(shaopeng-666): Remove the visual check after the mm model reconstruction is complete.
+        if prefix in compilation_config.static_forward_context and \
+            "visual" not in prefix:
+            raise ValueError(f"Duplicate layer name: {prefix}")
+        compilation_config.static_forward_context[prefix] = self
+
         self.custom_op, self.tp_rank, self.tp_size = get_parallel_op(
             disable_tp, prefix, self, "row")
         # TODO(realliujiaxu): Replace the initialization code below with super().__init__ after linear of vllm supports custom comm group

vllm_ascend/ops/linear_op.py

@@ -366,14 +366,23 @@ class SequenceRowParallelOp(CustomRowParallelOp):
                                              input_parallel,
                                              bias=bias_)
         else:
-            output_parallel = self.quant_method.apply(self.layer,
-                                                      input_parallel,
-                                                      bias=bias_)
-            output = torch.ops.vllm.maybe_pad_and_reduce(output_parallel)
+            output = torch.ops.vllm.matmul_and_reduce(input_parallel,
+                                                      self.prefix)
 
         output_bias = self.bias if self.skip_bias_add else None
         return output, output_bias
 
+    def matmul_and_reduce(self, input_parallel: torch.Tensor,
+                          bias_: Optional[Parameter]) -> torch.Tensor:
+        assert self.quant_method is not None
+        output_parallel = self.quant_method.apply(self.layer,
+                                                  input_parallel,
+                                                  bias=bias_)
+        from vllm_ascend.ops.register_custom_ops import \
+            _maybe_pad_and_reduce_impl
+        output = _maybe_pad_and_reduce_impl(output_parallel)
+        return output
+
     def update_attrs(self):
         super().update_attrs()
         self.input_is_parallel = self.layer.input_is_parallel

vllm_ascend/ops/register_custom_ops.py

@@ -235,6 +235,31 @@ def _maybe_all_reduce_tensor_model_parallel_impl(
         return tensor_model_parallel_all_reduce(final_hidden_states)
 
 
+def _matmul_and_reduce_impl(input_parallel: torch.Tensor,
+                            layer_name: str) -> torch.Tensor:
+    forward_context = get_forward_context()
+    self = forward_context.no_compile_layers[layer_name]
+    assert self.custom_op is not None
+    bias_ = None if (self.tp_rank > 0 or self.skip_bias_add) else self.bias
+    output = self.custom_op.matmul_and_reduce(input_parallel, bias_)
+
+    return output
+
+
+def _matmul_and_reduce_impl_fake(input_parallel: torch.Tensor,
+                                 layer_name: str) -> torch.Tensor:
+    forward_context = get_forward_context()
+    self = forward_context.no_compile_layers[layer_name]
+    num_tokens = input_parallel.size(0)
+    if forward_context.sp_enabled:
+        num_tokens = num_tokens // self.tp_size
+    output = torch.empty(size=(num_tokens, self.output_size_per_partition),
+                         device=input_parallel.device,
+                         dtype=input_parallel.dtype)
+
+    return output
+
+
 direct_register_custom_op(op_name="maybe_all_gather_and_maybe_unpad",
                           op_func=_maybe_all_gather_and_maybe_unpad_impl,
                           fake_impl=_maybe_all_gather_and_maybe_unpad_fake,
@@ -282,3 +307,9 @@ direct_register_custom_op(op_name="maybe_all_reduce_tensor_model_parallel",
                           fake_impl=lambda x: x,
                           mutates_args=[],
                           dispatch_key="PrivateUse1")
+
+direct_register_custom_op(op_name="matmul_and_reduce",
+                          op_func=_matmul_and_reduce_impl,
+                          fake_impl=_matmul_and_reduce_impl_fake,
+                          mutates_args=[],
+                          dispatch_key="PrivateUse1")