[Mlir-commits] [mlir] [mlir][arith] mul operation regressions (PR #96975)
Jacob Yu
llvmlistbot at llvm.org
Thu Jun 27 15:02:17 PDT 2024
https://github.com/pingshiyu created https://github.com/llvm/llvm-project/pull/96975
Regressions for the mul operation, a part of the original large PR https://github.com/llvm/llvm-project/pull/92272
>From 67998205165c1e733084431ad48ab283dc0e172b Mon Sep 17 00:00:00 2001
From: pingshiyu <pingshiyu at gmail.com>
Date: Thu, 27 Jun 2024 22:58:35 +0100
Subject: [PATCH] mul regressions
---
.../Dialect/Arith/CPU/multiplication.mlir | 44 +++++++++++++++++++
1 file changed, 44 insertions(+)
create mode 100644 mlir/test/Integration/Dialect/Arith/CPU/multiplication.mlir
diff --git a/mlir/test/Integration/Dialect/Arith/CPU/multiplication.mlir b/mlir/test/Integration/Dialect/Arith/CPU/multiplication.mlir
new file mode 100644
index 0000000000000..494bc2af1821e
--- /dev/null
+++ b/mlir/test/Integration/Dialect/Arith/CPU/multiplication.mlir
@@ -0,0 +1,44 @@
+// Tests mul operations and their variants (e.g. extended).
+// These tests are intended to be target agnostic: they should yield the same results
+// regardless of the target platform.
+
+// 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
+
+func.func @mulsi_extended_on_i1() {
+ // mulsi_extended on i1, tests for overflow bit
+ // mulsi_extended 1, 1 : i1 = (1, 0)
+ // CHECK: 1
+ // CHECK-NEXT: 0
+ %true = arith.constant true
+ %low, %high = arith.mulsi_extended %true, %true : i1
+ vector.print %low : i1
+ vector.print %high : i1
+ return
+}
+
+func.func @mulsi_mului_extended_overflows() {
+ // mulsi and mului extended versions, with overflow
+ // mulsi_extended -100, -100 : i8 = (16, 39); mului_extended -100, -100 : i8 = (16, 95)
+ // CHECK-NEXT: 16
+ // CHECK-NEXT: 39
+ // CHECK-NEXT: 16
+ // CHECK-NEXT: 95
+ %c-100_i8 = arith.constant -100 : i8
+ %low, %high = arith.mulsi_extended %c-100_i8, %c-100_i8 : i8
+ vector.print %low : i8
+ vector.print %high : i8
+ %low_0, %high_1 = arith.mului_extended %c-100_i8, %c-100_i8 : i8
+ vector.print %low_0 : i8
+ vector.print %high_1 : i8
+ return
+}
+
+func.func @entry() {
+ func.call @mulsi_extended_on_i1() : () -> ()
+ func.call @mulsi_mului_extended_overflows() : () -> ()
+ return
+}
More information about the Mlir-commits
mailing list