[inductor] Don't specialize split on sizes parameter. (pytorch#14…

…1077)

Fix: pytorch#139936

This PR modifies the lowering of `split` operation, so that it won't generate guards,
specializing on the sizes parameter. Instead, it specializes on the number of output
tensors being generated (i.e. function of the size of the base tensor, and the sizes
parameter).

As a result, operations such as `chunk` (whose number of output tensors usually is
constant given a static chunk number) won't trigger recompiles when varying the size of
the base tensor.

Pull Request resolved: pytorch#141077
Approved by: https://github.com/ezyang

test/inductor/test_torchinductor.py

@@ -11798,6 +11798,37 @@ def f(m, x):
             f"Ref:\n{ref_grad_list}\nAct:\n{act_grad_list}",
         )
 
+    def test_chunk_recompiles(self):
+        def f(x):
+            return x.chunk(4)
+
+        def run(size):
+            input = torch.randn(size)
+            expected_out = f(input)
+            actual_out = optf(input)
+            self.assertEqual(expected_out, actual_out)
+
+        cnts = CompileCounterWithBackend("inductor")
+        optf = torch.compile(f, backend=cnts, fullgraph=True)
+
+        # The first run should compile once with static shapes.
+        run(4)
+        self.assertEqual(cnts.frame_count, 1)
+
+        # Varying the input size should trigger a recompilation.
+        # Since the input size is a multiple of 4 (i.e. all runs shall
+        # generate 4 output tensors), there should be no further
+        # recompilation.
+        for i in range(2, 12):
+            run(4 * i)
+        self.assertEqual(cnts.frame_count, 2)
+
+        # Input size: 9
+        # Yields one less output tensor, which should trigger a
+        # recompilation.
+        run(9)
+        self.assertEqual(cnts.frame_count, 3)
+
 
 @dataclasses.dataclass
 class TestFailure:

torch/_inductor/lowering.py

@@ -1670,13 +1670,11 @@ def select(x, dim, idx):
 @register_lowering(aten.split, type_promotion_kind=None)
 def split(x, sizes, dim=0, clamp=True):
     dim = _validate_dim(x, dim, 0)
-    if isinstance(sizes, sympy.Expr):
-        # TODO: We don't have to guard on sizes per se, but the number
-        # of splits must stay constant
-        sizes = V.graph.sizevars.evaluate_static_shape(sizes)
-    if isinstance(sizes, (int, sympy.Integer)):
-        x_size = V.graph.sizevars.evaluate_static_shape(x.get_size()[dim])
-        sizes = [sizes] * ((x_size + sizes - 1) // sizes)
+    if not isinstance(sizes, (list, tuple)):
+        chunks = V.graph.sizevars.evaluate_static_shape(
+            FloorDiv(x.get_size()[dim] + sizes - 1, sizes)
+        )
+        sizes = [sizes] * chunks
     result = []
     start = 0
     for size in sizes: