[PYTHON] Added support for tuples (#116)

python/test/test_language.py

@@ -189,6 +189,52 @@ def test_index1d(expr, device='cuda'):
     triton.testing.assert_allclose(z_ref, z_tri)
 
 
+# ---------------
+# test tuples
+# ---------------
+
+
+@triton.jit
+def fn(a, b):
+    return a + b, \
+            a - b, \
+            a * b
+
+
+def test_tuples():
+    device = 'cuda'
+
+    @triton.jit
+    def with_fn(X, Y, A, B, C):
+        x = tl.load(X)
+        y = tl.load(Y)
+        a, b, c = fn(x, y)
+        tl.store(A, a)
+        tl.store(B, b)
+        tl.store(C, c)
+
+    @triton.jit
+    def without_fn(X, Y, A, B, C):
+        x = tl.load(X)
+        y = tl.load(Y)
+        a, b, c = x + y, x - y, x * y
+        tl.store(A, a)
+        tl.store(B, b)
+        tl.store(C, c)
+
+    x = torch.tensor([1.3], device=device, dtype=torch.float32)
+    y = torch.tensor([1.9], device=device, dtype=torch.float32)
+    a_tri = torch.tensor([0], device=device, dtype=torch.float32)
+    b_tri = torch.tensor([0], device=device, dtype=torch.float32)
+    c_tri = torch.tensor([0], device=device, dtype=torch.float32)
+    for kernel in [with_fn, without_fn]:
+        kernel[(1, )](x, y, a_tri, b_tri, c_tri, num_warps=1)
+        a_ref, b_ref, c_ref = x + y, x - y, x * y
+        assert a_tri == a_ref
+        assert b_tri == b_ref
+        assert c_tri == c_ref
+
+
 # ---------------
 # test atomics
 # ---------------

python/triton/code_gen.py

@@ -11,7 +11,7 @@ import triton._C.libtriton.triton as _triton
 import triton
 import sys
 import textwrap
-from abc import ABC, abstractmethod
+import collections
 
 
 class CodeGenerator(ast.NodeVisitor):
@@ -134,15 +134,20 @@ class CodeGenerator(ast.NodeVisitor):
         return node.arg
 
     def visit_Assign(self, node):
-        names = []
+        _names = []
         for target in node.targets:
-            names += [self.visit(target)]
-        assert len(names) == 1
-        name = names[0]
-        value = self.visit(node.value)
-        if not isinstance(value, triton.language.block):
-            value = triton.language._to_ir(value, self.builder)
-        self.set_value(names[0], value)
+            _names += [self.visit(target)]
+        assert len(_names) == 1
+        names = _names[0]
+        values = self.visit(node.value)
+        if not isinstance(names, tuple):
+            names = [names]
+        if not isinstance(values, tuple):
+            values = [values]
+        for name, value in zip(names, values):
+            if not isinstance(value, triton.language.block):
+                value = triton.language._to_ir(value, self.builder)
+            self.set_value(name, value)
 
     def visit_AugAssign(self, node):
         name = node.target.id