[llvm] LoopVectorize/test: fix opt invocations with -march (NFC) (PR #74462)
via llvm-commits
llvm-commits at lists.llvm.org
Tue Dec 5 04:57:21 PST 2023
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-llvm-transforms
Author: Ramkumar Ramachandra (artagnon)
<details>
<summary>Changes</summary>
opt erroneously accepts the -march command-line argument, although this argument has no effect. Fix a couple of tests under LoopVectorize that invoke opt with this argument, to avoid confusing users.
---
Full diff: https://github.com/llvm/llvm-project/pull/74462.diff
2 Files Affected:
- (modified) llvm/test/Transforms/LoopVectorize/Hexagon/maximum-vf-crash.ll (+1-2)
- (modified) llvm/test/Transforms/LoopVectorize/Hexagon/minimum-vf.ll (+1-2)
``````````diff
diff --git a/llvm/test/Transforms/LoopVectorize/Hexagon/maximum-vf-crash.ll b/llvm/test/Transforms/LoopVectorize/Hexagon/maximum-vf-crash.ll
index 4c7a7d31c5e72..4966ddd299492 100644
--- a/llvm/test/Transforms/LoopVectorize/Hexagon/maximum-vf-crash.ll
+++ b/llvm/test/Transforms/LoopVectorize/Hexagon/maximum-vf-crash.ll
@@ -1,4 +1,4 @@
-; RUN: opt -march=hexagon -hexagon-autohvx -passes=loop-vectorize -S < %s 2>&1 | FileCheck %s
+; RUN: opt -mtriple=hexagon -hexagon-autohvx -passes=loop-vectorize -S < %s 2>&1 | FileCheck %s
; Check that we don't crash.
@@ -6,7 +6,6 @@
; CHECK: vector.body
target datalayout = "e-m:e-p:32:32:32-a:0-n16:32-i64:64:64-i32:32:32-i16:16:16-i1:8:8-f32:32:32-f64:64:64-v32:32:32-v64:64:64-v512:512:512-v1024:1024:1024-v2048:2048:2048"
-target triple = "hexagon"
; Function Attrs: optsize
define i32 @f() #0 {
diff --git a/llvm/test/Transforms/LoopVectorize/Hexagon/minimum-vf.ll b/llvm/test/Transforms/LoopVectorize/Hexagon/minimum-vf.ll
index 4ff7ad9ab4390..1ac556a2fda39 100644
--- a/llvm/test/Transforms/LoopVectorize/Hexagon/minimum-vf.ll
+++ b/llvm/test/Transforms/LoopVectorize/Hexagon/minimum-vf.ll
@@ -1,4 +1,4 @@
-; RUN: opt -march=hexagon -passes=loop-vectorize -hexagon-autohvx -debug-only=loop-vectorize -disable-output < %s 2>&1 | FileCheck %s
+; RUN: opt -mtriple=hexagon -passes=loop-vectorize -hexagon-autohvx -debug-only=loop-vectorize -disable-output < %s 2>&1 | FileCheck %s
; REQUIRES: asserts
; Check that TTI::getMinimumVF works. The calculated MaxVF was based on the
@@ -6,7 +6,6 @@
; CHECK: LV: Overriding calculated MaxVF({{[0-9]+}}) with target's minimum: 64
target datalayout = "e-m:e-p:32:32:32-a:0-n16:32-i64:64:64-i32:32:32-i16:16:16-i1:8:8-f32:32:32-f64:64:64-v32:32:32-v64:64:64-v512:512:512-v1024:1024:1024-v2048:2048:2048"
-target triple = "hexagon"
%s.0 = type { ptr, i32, i32, i32, i32 }
``````````
</details>
https://github.com/llvm/llvm-project/pull/74462
More information about the llvm-commits
mailing list