[LLVMdev] Bug 16257 - fmul of undef ConstantExpr not folded to undef

[Stephen Canon]

> I think you should try to get LLVM floating point experts involved, to find out their opinion about whether LLVM should really assume that snans always trap.

We definitely cannot assume that operations on sNaN always trap.  Actually, are there *any* llvm-supported platforms where the invalid exception is unmasked by default?

– Steve