ZarkoCA marked 7 inline comments as done. ZarkoCA added inline comments.
================ Comment at: clang/test/CodeGen/altivec.c:7 + +// RUN: %clang -S -emit-llvm -maltivec -mabi=vec-extabi -target powerpc-unknown-aix %s -o - | FileCheck %s +// RUN: not %clang -S -emit-llvm -mabi=vec-default -target powerpc-unknown-aix %s 2>&1 | FileCheck %s --check-prefix=AIX-ATVER ---------------- Xiangling_L wrote: > When user specify `-maltivec / -target-feature +altivec` without using any > abi option, the compiler will assume default altivec abi. In this situation, > since default abi hasn’t been implemented, we should emit an error. So can we > also add testcases for : > > ``` > // RUN: not %clang -S -emit-llvm -maltivec -target powerpc-unknown-aix %s > 2>&1 | FileCheck %s --check-prefix=AIX-ERROR > and > // RUN: not %clang_cc1 -target-feature +altivec -triple powerpc-unknown-aix > -emit-llvm %s 2>&1 | FileCheck %s --check-prefix=AIX-ERROR > ``` That's a good catch, the error was able to be generated previously but reworking the logic in Clang.cpp with the previous diff caused it to not be emitted. I went back to the older logic which emits the error in cases like where `maltivec` is specified without `mabi=vec-extabi`. As for the cc1 error, currently the not vector types error catch that before we emit the Altivec ABI error since at this time there isn't a good way to check for `target-feature +altivec` in cc1. ================ Comment at: llvm/test/CodeGen/PowerPC/aix-vec-abi.ll:1 +; RUN: not --crash llc < %s -mtriple powerpc64-ibm-aix-xcoff -mcpu=pwr8 2>&1 | FileCheck %s --check-prefix=DFLTERROR +; RUN: not --crash llc < %s -mtriple powerpc-ibm-aix-xcoff -mcpu=pwr8 2>&1 | FileCheck %s --check-prefix=DFLTERROR ---------------- Xiangling_L wrote: > May I ask why we use `pwr8` for this test? Sorry for missing this earlier, I wanted to specify a CPU that has Altivec instructions enabled so that hasAltivec true without the user specifying it. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D89684/new/ https://reviews.llvm.org/D89684 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
