diff options
| author | Nikita Popov <npopov@redhat.com> | 2025-07-08 10:20:43 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-07-08 10:20:43 +0200 |
| commit | 18f7655178a69daa52f4d4ef56169b90efc7f601 (patch) | |
| tree | 838d27e41f3c0d81db1a131e8f0e735f1075bbd7 /clang/lib/CodeGen/CodeGenModule.cpp | |
| parent | ae496bfac0c34c511775229e94a1439fe8ee4867 (diff) | |
| download | llvm-18f7655178a69daa52f4d4ef56169b90efc7f601.zip llvm-18f7655178a69daa52f4d4ef56169b90efc7f601.tar.gz llvm-18f7655178a69daa52f4d4ef56169b90efc7f601.tar.bz2 | |
[Clang][Wasm] Set __float128 alignment to 64 for emscripten (#146494)
https://reviews.llvm.org/D104808 set the alignment of long double to 64
bits. This is also the alignment specified in the LLVM data layout.
However, the alignment of __float128 was left at 128 bits.
I assume that this was just an oversight, rather than an intentional
divergence. The C ABI document currently does not make any statement
about `__float128`:
https://github.com/WebAssembly/tool-conventions/blob/main/BasicCABI.md
Diffstat (limited to 'clang/lib/CodeGen/CodeGenModule.cpp')
| -rw-r--r-- | clang/lib/CodeGen/CodeGenModule.cpp | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/clang/lib/CodeGen/CodeGenModule.cpp b/clang/lib/CodeGen/CodeGenModule.cpp index e6d150f..c8866f1 100644 --- a/clang/lib/CodeGen/CodeGenModule.cpp +++ b/clang/lib/CodeGen/CodeGenModule.cpp @@ -388,8 +388,7 @@ static void checkDataLayoutConsistency(const TargetInfo &Target, llvm::Type::getFloatingPointTy(Context, *Target.LongDoubleFormat), Target.LongDoubleAlign); } - // FIXME: Wasm has a mismatch in f128 alignment between Clang and LLVM. - if (Target.hasFloat128Type() && !Triple.isWasm()) + if (Target.hasFloat128Type()) Check("__float128", llvm::Type::getFP128Ty(Context), Target.Float128Align); if (Target.hasIbm128Type()) Check("__ibm128", llvm::Type::getPPC_FP128Ty(Context), Target.Ibm128Align); |