[test] Avoid writing to a potentially write-protected dir

[karka228]

The test length-errors.hlsl don't check the output written to the current directory. The current directory may be write protected e.g. in a sandboxed environment.

This patch simply remove the -emit-llvm option as this testcase don't care about the outputed llvm IR.

[llvmbot]

@llvm/pr-subscribers-backend-x86
@llvm/pr-subscribers-hlsl

@llvm/pr-subscribers-clang

Author: Karl-Johan Karlsson (karka228)

Changes

The tests avx10_2_512minmax-error.c and length-errors.hlsl don't check the output written to the current directory. The current directory may be write protected e.g. in a sandboxed environment.

This patch replace the -emit-llvm option with -emit-llvm-only as these testcases don't care about the outputed llvm IR.

---

Full diff: https://github.com/llvm/llvm-project/pull/102073.diff

2 Files Affected:

• (modified) clang/test/CodeGen/X86/avx10_2_512minmax-error.c (+2-2)
• (modified) clang/test/SemaHLSL/BuiltIns/length-errors.hlsl (+1-1)

diff --git a/clang/test/CodeGen/X86/avx10_2_512minmax-error.c b/clang/test/CodeGen/X86/avx10_2_512minmax-error.c
index 15ed7a0b35d82..7a5f6b5b2345c 100644
--- a/clang/test/CodeGen/X86/avx10_2_512minmax-error.c
+++ b/clang/test/CodeGen/X86/avx10_2_512minmax-error.c
@@ -1,7 +1,7 @@
 // RUN: %clang_cc1 %s -flax-vector-conversions=none -ffreestanding -triple=x86_64 -target-feature +avx10.2-512 \
-// RUN: -Wno-invalid-feature-combination -emit-llvm -verify
+// RUN: -Wno-invalid-feature-combination -emit-llvm-only -verify
 // RUN: %clang_cc1 %s -flax-vector-conversions=none -ffreestanding -triple=i386 -target-feature +avx10.2-512 \
-// RUN: -Wno-invalid-feature-combination -emit-llvm -verify
+// RUN: -Wno-invalid-feature-combination -emit-llvm-only -verify
 
 #include <immintrin.h>
 
diff --git a/clang/test/SemaHLSL/BuiltIns/length-errors.hlsl b/clang/test/SemaHLSL/BuiltIns/length-errors.hlsl
index fe0046a2ab5a1..8153726367637 100644
--- a/clang/test/SemaHLSL/BuiltIns/length-errors.hlsl
+++ b/clang/test/SemaHLSL/BuiltIns/length-errors.hlsl
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm -disable-llvm-passes -verify -verify-ignore-unexpected
+// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only -disable-llvm-passes -verify -verify-ignore-unexpected
 
 void test_too_few_arg()
 {

[llvm-beanz]

This probably shouldn't have the -emit-llvm flag at all since it isn't actually checking that output.

[llvm-beanz]

Looks like neither of these tests should be passing -emit-llvm at all.

[FreddyLeaf]

I just realized `-emit-llvm is not required here. You can directly removed it. Sorry for mistake.

[karka228]

When rebasing the patch I removed the fix in avx10_2_512minmax-error.c as it was already fixed in:

commit 1b9d7dd9eb31974077b64d66404193bd7c4ad65e
Author: Augie Fackler <augie@google.com>
Date:   Mon Aug 5 17:08:06 2024 -0400

    tests: fix clang flags in avx10_2_512minmax-error test
    
    Introduced in 3d5cc7e1e632b74119af13824dabc346bd248c93, caught by a
    sandboxed test run.