================ @@ -1,7 +1,7 @@ -// RUN: %clang_cc1 -triple x86_64-linux -std=c++98 %s -O3 -disable-llvm-passes -pedantic-errors -emit-llvm -o - | FileCheck %s ---------------- rjmccall wrote:
This is okay as long as you're sure the optimization will be performed and you make sure your test only checks the pattern you want out of the optimizer and doesn't cover other details that could potentially be changed by unrelated LLVM patches. Otherwise, we generally aim to make IR-generation tests "unit" tests that specifically test the output of Clang's IR-generation. For output that's only included in optimized modes, like TBAA metadata, that generally requires enabling optimization (with `-O<n>`) and then disabling all the LLVM optimizations (with `-disable-llvm-passes`). https://github.com/llvm/llvm-project/pull/75177 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
