Issue with Clang on Windows and compiler-rt builtins
>> When __int128 operations are used, Clang on Windows generate calls to
>> compiler-rt builtins (__udivti3 etc.) instead of generating inline
>> optimized assembly directly.
> LLVM does not inline division or modulo for 128bit operands. Way too
> much code.
Thanks for the pointer.
After checking with https://godbolt.org/g/HN7wf5 , I realized my mistake. GCC
and Clang on Linux will constant fold __int128 if possible, otherwise __udivti3
etc. function calls will still be emitted.