[llvm-commits] [llvm] r168641 - /llvm/trunk/include/llvm/Operator.h

[Owen Anderson]

On Nov 26, 2012, at 4:40 PM, Michael Ilseman <milseman at apple.com> wrote:

> 
> +/// Convenience struct for specifying and reasoning about fast-math flags.
> +struct FastMathFlags {
> +  bool UnsafeAlgebra   : 1;
> +  bool NoNaNs          : 1;
> +  bool NoInfs          : 1;
> +  bool NoSignedZeros   : 1;
> +  bool AllowReciprocal : 1;
> +
> +  FastMathFlags() : UnsafeAlgebra(false), NoNaNs(false), NoInfs(false),
> +                    NoSignedZeros(false), AllowReciprocal(false)
> +  { }
> +
> +  bool any() {
> +    return UnsafeAlgebra || NoNaNs || NoInfs || NoSignedZeros ||
> +      AllowReciprocal;
> +  }
> +};
> +
> +
> /// FPMathOperator - Utility class for floating point operations which can have
> /// information about relaxed accuracy requirements attached to them.
> class FPMathOperator : public Operator {
> public:
> +  enum {
> +    UnsafeAlgebra   = (1 << 0),
> +    NoNaNs          = (1 << 1),
> +    NoInfs          = (1 << 2),
> +    NoSignedZeros   = (1 << 3),
> +    AllowReciprocal = (1 << 4)
> +  }

Why do you need to use a bitfield (gross!)?  Can't FastMathFlags just be a wrapper around an unsigned?

--Owen