[BugFix] Align default (C) tirx.round lowering to ties-to-even

src/target/intrin_rule.cc

@@ -33,6 +33,18 @@ namespace codegen {
 namespace intrin {
 using tirx::FLowerIntrinsic;
 
+// `tirx.round` is ties-to-even (see include/tvm/tirx/op.h), and constant
+// folding implements it with std::nearbyint. The C library's round()/roundf()
+// is ties-AWAY-from-zero, so lowering through FloatSuffix would disagree with
+// the folder and with every other backend. Rename to nearbyint before the
+// float suffix is applied, as the CUDA rule already does.
+struct FloatSuffixTiesToEven {
+  std::string operator()(const PrimType& ty, std::string name) const {
+    if (name == "round") name = "nearbyint";
+    return FloatSuffix()(ty, name);
+  }
+};
+
 TVM_REGISTER_OP("tirx.exp")
     .set_attr<FLowerIntrinsic>("default.FLowerIntrinsic", DispatchPureExtern<FloatSuffix>);
 
@@ -115,7 +127,8 @@ TVM_REGISTER_OP("tirx.ceil")
     .set_attr<FLowerIntrinsic>("default.FLowerIntrinsic", DispatchPureExtern<FloatSuffix>);
 
 TVM_REGISTER_OP("tirx.round")
-    .set_attr<FLowerIntrinsic>("default.FLowerIntrinsic", DispatchPureExtern<FloatSuffix>);
+    .set_attr<FLowerIntrinsic>("default.FLowerIntrinsic",
+                               DispatchPureExtern<FloatSuffixTiesToEven>);
 
 TVM_REGISTER_OP("tirx.nearbyint")
     .set_attr<FLowerIntrinsic>("default.FLowerIntrinsic", DispatchPureExtern<FloatSuffix>);

tests/python/codegen/test_target_codegen_c_host.py

@@ -193,7 +193,17 @@ def check_c():
         fround = m["test_round"]
         dev = tvm.cpu(0)
         n = nn
-        a = tvm.runtime.tensor(np.random.rand(n).astype("float32"), dev)
+        # Exact midpoints first: this is where ties-to-even (np.round, and the
+        # semantics every other backend and the constant folder use) differs
+        # from ties-away-from-zero. np.random.rand never produces them, so the
+        # random tail alone cannot exercise the tie rule.
+        midpoints = np.array(
+            [0.5, 1.5, 2.5, 3.5, -0.5, -1.5, -2.5, -3.5], dtype="float32"
+        )
+        a_np = np.concatenate(
+            [midpoints, np.random.rand(n - len(midpoints)).astype("float32")]
+        ).astype("float32")
+        a = tvm.runtime.tensor(a_np, dev)
         b = tvm.runtime.tensor(np.zeros(n, dtype="float32"), dev)
         fround(a, b)
         tvm.testing.assert_allclose(b.numpy(), (np.round(a.numpy()).view("float32")))