[Mlir-commits] [mlir] [mlir][arith] wide integer emulation support for fpto*i ops (PR #132375)

[Jakub Kuderski]

================
@@ -0,0 +1,82 @@
+// Check that the wide integer `arith.fptosi` emulation produces the same result as wide
+// `arith.fptosi`. Emulate i64 ops with i32 ops.
+
+// RUN: mlir-opt %s --convert-scf-to-cf --convert-cf-to-llvm --convert-vector-to-llvm \
+// RUN:             --convert-func-to-llvm --convert-arith-to-llvm | \
+// RUN:   mlir-cpu-runner -e entry -entry-point-result=void \
+// RUN:                   --shared-libs=%mlir_c_runner_utils | \
+// RUN:   FileCheck %s --match-full-lines
+
+// RUN: mlir-opt %s --test-arith-emulate-wide-int="widest-int-supported=32" \
+// RUN:             --convert-scf-to-cf --convert-cf-to-llvm --convert-vector-to-llvm \
+// RUN:             --convert-func-to-llvm --convert-arith-to-llvm | \
+// RUN:   mlir-cpu-runner -e entry -entry-point-result=void \
+// RUN:                   --shared-libs=%mlir_c_runner_utils | \
+// RUN:   FileCheck %s --match-full-lines
+
+// Ops in this function *only* will be emulated using i32 types.
+func.func @emulate_fptosi(%arg: f64) -> i64 {
+  %res = arith.fptosi %arg : f64 to i64
+  return %res : i64
+}
+
+func.func @check_fptosi(%arg : f64) -> () {
+  %res = func.call @emulate_fptosi(%arg) : (f64) -> (i64)
+  vector.print %res : i64
+  return
+}
+
+func.func @entry() {
----------------
kuhar wrote:

Nice, thanks for testing this

https://github.com/llvm/llvm-project/pull/132375