Revert #75195

This is a short-term fix for a serious regression in functorch
(pytorch/functorch#989).

Additional things this PR does:
- the out= tests for nn.functional.linear fail after the revert. I added
some xfails. These xfails were present in the original PR (#75195).
- the profiler tests fail on the revert, so I updated the expecttests
for the profiler tests

Test Plan:
- test offline that the functorch regression was fixed

ghstack-source-id: f127f9be33ba35ceeabbc6a9a4d8c24654defad7
Pull Request resolved: #82504

.github/ci_commit_pins/xla.txt

@@ -1 +1 @@
-b3342319e96a0becd139019620d8665605b78475
+73c64a55fb096f1e132029d3decbb6f4e532cc7b

aten/src/ATen/native/LinearAlgebra.cpp

@@ -1760,7 +1760,8 @@ Tensor _matmul_impl(
   } else if (dim_tensor1 == 2 && dim_tensor2 == 1) {
     return has_out ? at::mv_out(out, tensor1, tensor2) : tensor1.mv(tensor2);
   } else if (dim_tensor1 == 1 && dim_tensor2 == 2) {
-    return has_out ? at::mv_out(out, tensor2.t(), tensor1) : tensor2.t().mv(tensor1);
+    return has_out ? at::mm_out(out, tensor1.unsqueeze(0), tensor2).squeeze_(0)
+                   : tensor1.unsqueeze(0).mm(tensor2).squeeze_(0);
   } else if (dim_tensor1 == 2 && dim_tensor2 == 2) {
     return has_out ? at::mm_out(out, tensor1, tensor2) : tensor1.mm(tensor2);
   } else if (should_fold(tensor1, dim_tensor2) || should_fold(tensor2, dim_tensor1)) {

test/test_profiler_tree.py

@@ -535,12 +535,14 @@ def forward(self, x: torch.Tensor) -> torch.Tensor:
                             aten::transpose
                               aten::as_strided
                           aten::matmul
-                            aten::t
-                              aten::transpose
-                                aten::as_strided
-                            aten::mv
-                              aten::empty
-                              aten::addmv_
+                            aten::unsqueeze
+                              aten::as_strided
+                            aten::mm
+                              aten::resolve_conj
+                              aten::resolve_conj
+                              aten::resolve_conj
+                            aten::squeeze_
+                              aten::as_strided_
                           aten::add_
                   nn.Module: ReLU_1
                     <built-in method _get_tracing_state of PyCapsule object at 0xXXXXXXXXXXXX>
@@ -576,12 +578,14 @@ def forward(self, x: torch.Tensor) -> torch.Tensor:
                             aten::transpose
                               aten::as_strided
                           aten::matmul
-                            aten::t
-                              aten::transpose
-                                aten::as_strided
-                            aten::mv
-                              aten::empty
-                              aten::addmv_
+                            aten::unsqueeze
+                              aten::as_strided
+                            aten::mm
+                              aten::resolve_conj
+                              aten::resolve_conj
+                              aten::resolve_conj
+                            aten::squeeze_
+                              aten::as_strided_
                           aten::add_
                   nn.Module: ReLU_1
                     <built-in method _get_tracing_state of PyCapsule object at 0xXXXXXXXXXXXX>

torch/testing/_internal/common_methods_invocations.py

@@ -13459,6 +13459,8 @@ def error_inputs_mean(op_info, device, **kwargs):
                             'TestCommon', 'test_noncontiguous_samples',
                             device_type='cpu'), ],
            skips=(
+               # Strides are not the same!
+               DecorateInfo(unittest.expectedFailure, 'TestCommon', 'test_out'),
                # https://github.com/pytorch/pytorch/issues/67470
                DecorateInfo(unittest.skip("67470!"),
                             'TestCommon', 'test_noncontiguous_samples',
@@ -14861,6 +14863,8 @@ def error_inputs_mean(op_info, device, **kwargs):
            check_batched_forward_grad=False,
            supports_expanded_weight=True,
            decorators=(
+               # Strides are not the same!
+               DecorateInfo(unittest.expectedFailure, 'TestCommon', 'test_out'),
                DecorateInfo(toleranceOverride({torch.float16: tol(atol=1e-02, rtol=1e-02)}),
                             'TestCudaFuserOpInfo', 'test_nvfuser_correctness'),
            )),