__float128 compilation error

I’m trying to compile a copy of the Github clone of ROOT for PowerPC. I get these errors. Is there a way to turn on __float128 support?

Disabling gnu++: clang has no __float128 support on this target!

Dear @ChipKerchner ,

What version of ROOT are you trying to compile? What is the compiler version?

I invite @Axel , @pcanal that might be able to help, once you give us more information.


– Detected ROOT_VERSION 6.29.01

– The C compiler identification is GNU 14.0.0
– The CXX compiler identification is GNU 14.0.0

I see that a possible solution is to use “-mfloat128”.

Without validating our own error messages - this seems to indicate that clang itself does not support __float128 on power, which is why ROOT asks you to not build with -std=gnu++17 or whatever but -std=c++17 or whatever standard version you need.