[pt2] add SymInt support for tensordot and inner (#100356)

Pull Request resolved: #100356
Approved by: https://github.com/ezyang

Author: nkaretnikov

aten/src/ATen/native/Linear.cpp

@@ -727,12 +727,12 @@ Tensor bilinear(const Tensor& input1, const Tensor& input2, const Tensor& weight
 Tensor tensordot(const Tensor& input1, const Tensor& input2, IntArrayRef dims1, IntArrayRef dims2) {
   TORCH_CHECK(dims1.size() == dims2.size(), "both dimension lists should have same length");
   TORCH_CHECK(input1.scalar_type() == input2.scalar_type(), "both inputs should have same dtype");
-  int64_t csize = 1;  // total size of the contracted dimensions
+  SymInt csize = 1;  // total size of the contracted dimensions
   Tensor t1 = input1;
   Tensor t2 = input2;
   for (const auto i : c10::irange(dims1.size())) {
-    int s1 = input1.size(dims1[i]);
-    int s2 = input2.size(dims2[i]);
+    SymInt s1 = input1.sym_size(dims1[i]);
+    SymInt s2 = input2.sym_size(dims2[i]);
     if (s2 == 1) { // broadcasted dimensions can be summed right away
       t1 = t1.sum(dims1[i], true);
     } else if (s1 == 1) {
@@ -746,19 +746,20 @@ Tensor tensordot(const Tensor& input1, const Tensor& input2, IntArrayRef dims1,
 
   auto cdims1 = at::dim_list_to_bitset(dims1, input1.dim());
   auto cdims2 = at::dim_list_to_bitset(dims2, input2.dim());
-  std::vector<int64_t> p1, p2, rsizes;  // p1, p2: input permutations, rsizes: sizes of the result
+  std::vector<int64_t> p1, p2;  // p1, p2: input permutations
+  std::vector<SymInt> rsizes;  // rsizes: sizes of the result
   p1.reserve(input1.dim());
   p2.reserve(input2.dim());
   rsizes.reserve(input1.dim() + input2.dim() - (int64_t) dims1.size());
-  int64_t size1 = 1; // number of non-contracted elements in input1
-  int64_t size2 = 1; // number of non-contracted elements in input2
+  SymInt size1 = 1; // number of non-contracted elements in input1
+  SymInt size2 = 1; // number of non-contracted elements in input2
 
   // fill the permutations and compute sizes
   for (const auto i : c10::irange(input1.dim())) {
     if (! cdims1[i]) {
       p1.emplace_back(i);
-      size1 *= t1.size(i);
-      rsizes.emplace_back(t1.size(i));
+      size1 *= t1.sym_size(i);
+      rsizes.emplace_back(t1.sym_size(i));
     }
   }
   for (const auto x : dims1) {
@@ -770,15 +771,15 @@ Tensor tensordot(const Tensor& input1, const Tensor& input2, IntArrayRef dims1,
   for (const auto i : c10::irange(input2.dim())) {
     if (! cdims2[i]) {
       p2.emplace_back(i);
-      size2 *= t2.size(i);
-      rsizes.emplace_back(t2.size(i));
+      size2 *= t2.sym_size(i);
+      rsizes.emplace_back(t2.sym_size(i));
     }
   }
   // permut and reshape for matrix multiplication
-  t1 = t1.permute(p1).reshape({size1, csize});
-  t2 = t2.permute(p2).reshape({csize, size2});
+  t1 = t1.permute(p1).reshape_symint({size1, csize});
+  t2 = t2.permute(p2).reshape_symint({csize, size2});
   // multiply and reshape to target size
-  return at::mm(t1, t2).reshape(rsizes);
+  return at::mm(t1, t2).reshape_symint(rsizes);
 }
 
 Tensor &tensordot_out(const Tensor& input1, const Tensor& input2, IntArrayRef dims1, IntArrayRef dims2, Tensor& result) {

aten/src/ATen/native/LinearAlgebra.cpp

@@ -1284,11 +1284,11 @@ Tensor inner(const Tensor& self, const Tensor& other) {
 
   // Last dimension should match (tensordot does not enforce this)
   TORCH_CHECK(
-      self.size(-1) == other.size(-1),
+      self.sym_size(-1) == other.sym_size(-1),
       "inner() the last dimension must match on both input tensors but got shapes ",
-      self.sizes(),
+      self.sym_sizes(),
       " and ",
-      other.sizes());
+      other.sym_sizes());
 
   return at::tensordot(self, other, -1, -1);
 }

test/functorch/test_aotdispatch.py

@@ -2502,7 +2502,6 @@ def forward(self, x):
     xfail('hsplit', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('i0', ''),  # aten.i0.default - couldn't find symbolic meta function/decomposition
     xfail('index_fill', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
-    xfail('inner', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('kron', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('kthvalue', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('linalg.cholesky_ex', ''),  # could not find kernel for aten.linalg_solve_triangular.default
@@ -2612,7 +2611,6 @@ def forward(self, x):
     xfail('svd', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('svd_lowrank', ''),  # could not find kernel
     xfail('take_along_dim', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
-    xfail('tensordot', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('trace', ''),  # Cannot call sizes() on tensor with symbolic sizes/strides
     xfail('triangular_solve', ''),  # aten.triangular_solve.default - couldn't find symbolic meta function/de...
     xfail('_upsample_bilinear2d_aa'),  # RuntimeError: isIntList() INTERNAL ASSERT FAILED  Expected IntList but got GenericList

test/test_proxy_tensor.py

@@ -1455,7 +1455,6 @@ def f(a, b, c, d, e):
     xfail('histogramdd', ''),  # aten._histogramdd_bin_edges.default - couldn't find symbolic meta function/decomposition
     xfail('hsplit', ''),  # aten.size.default - couldn't find symbolic meta function/decomposition
     xfail('index_reduce', ''),  # Float
-    xfail('inner', ''),  # aten.size.default - couldn't find symbolic meta function/decomposition
     xfail('isin', ''),  # aten.isin.Tensor_Tensor - couldn't find symbolic meta function/decomposition
     xfail('kron', ''),  # aten.size.default - couldn't find symbolic meta function/decomposition
     xfail('kthvalue', ''),  # aten.kthvalue.default - couldn't find symbolic meta function/decomposition
@@ -1559,7 +1558,6 @@ def f(a, b, c, d, e):
     xfail('stft', ''),  # argument 'size' must be tuple of ints, but found element of type torch._C.SymIntNode at...
     xfail('svd_lowrank', ''),  # aten.mm.default - couldn't find symbolic meta function/decomposition
     xfail('take_along_dim', ''),  # dtype of indices should be Long but got Float
-    xfail('tensordot', ''),  # aten.size.default - couldn't find symbolic meta function/decomposition
     xfail('triangular_solve', ''),  # aten.triangular_solve.default - couldn't find symbolic meta function/decomposition
     xfail('vsplit', ''),  # aten.size.default - couldn't find symbolic meta function/decomposition
     xfail('unique_consecutive', ''),  # aten.unique_consecutive.default - couldn't find symbolic meta function/decomposition