[PATCH v3 06/17] ipa: libipa: vector: Generalize arithmetic operators

[Barnabás Pőcze]

Hi


2024. november 18., hétfő 23:16 keltezéssel, Laurent Pinchart <laurent.pinchart at ideasonboard.com> írta:

> Instead of hand-coding all arithmetic operators, implement them based on
> a generic apply() function that takes an operator-specific
> std::function. This will simplify adding missing arithmetic operators.
> 
> Signed-off-by: Laurent Pinchart <laurent.pinchart at ideasonboard.com>
> Reviewed-by: Milan Zamazal <mzamazal at redhat.com>
> ---
> Changes since v2:
> 
> - Include <algorithm>
> - Use std::plus, std::minus, std::multiplies and std::divides
> ---
>  src/ipa/libipa/vector.h | 50 ++++++++++++++++++++++++-----------------
>  1 file changed, 30 insertions(+), 20 deletions(-)
> 
> diff --git a/src/ipa/libipa/vector.h b/src/ipa/libipa/vector.h
> index 2abb5536361c..76fd2934b870 100644
> --- a/src/ipa/libipa/vector.h
> +++ b/src/ipa/libipa/vector.h
> @@ -6,8 +6,10 @@
>   */
>  #pragma once
> 
> +#include <algorithm>
>  #include <array>
>  #include <cmath>
> +#include <functional>
>  #include <optional>
>  #include <ostream>
> 
> @@ -78,36 +80,24 @@ public:
>  		return ret;
>  	}
> 
> -	constexpr Vector<T, Rows> operator-(const Vector<T, Rows> &other) const
> +	constexpr Vector operator-(const Vector &other) const
>  	{
> -		Vector<T, Rows> ret;
> -		for (unsigned int i = 0; i < Rows; i++)
> -			ret[i] = data_[i] - other[i];
> -		return ret;
> +		return apply(*this, other, std::minus<>{});
>  	}
> 
> -	constexpr Vector<T, Rows> operator+(const Vector<T, Rows> &other) const
> +	constexpr Vector operator+(const Vector &other) const
>  	{
> -		Vector<T, Rows> ret;
> -		for (unsigned int i = 0; i < Rows; i++)
> -			ret[i] = data_[i] + other[i];
> -		return ret;
> +		return apply(*this, other, std::plus<>{});
>  	}
> 
> -	constexpr Vector<T, Rows> operator*(T factor) const
> +	constexpr Vector operator*(T factor) const
>  	{
> -		Vector<T, Rows> ret;
> -		for (unsigned int i = 0; i < Rows; i++)
> -			ret[i] = data_[i] * factor;
> -		return ret;
> +		return apply(*this, factor, std::multiplies<>{});
>  	}
> 
> -	constexpr Vector<T, Rows> operator/(T factor) const
> +	constexpr Vector operator/(T factor) const
>  	{
> -		Vector<T, Rows> ret;
> -		for (unsigned int i = 0; i < Rows; i++)
> -			ret[i] = data_[i] / factor;
> -		return ret;
> +		return apply(*this, factor, std::divides<>{});
>  	}
> 
>  	constexpr T dot(const Vector<T, Rows> &other) const
> @@ -182,6 +172,26 @@ public:
>  	}
> 
>  private:
> +	static constexpr Vector apply(const Vector &lhs, const Vector &rhs, std::function<T(T, T)> func)
> +	{
> +		Vector result;
> +		std::transform(lhs.data_.begin(), lhs.data_.end(),
> +			       rhs.data_.begin(), result.data_.begin(),
> +			       func);
> +
> +		return result;
> +	}
> +
> +	static constexpr Vector apply(const Vector &lhs, T rhs, std::function<T(T, T)> func)
> +	{
> +		Vector result;
> +		std::transform(lhs.data_.begin(), lhs.data_.end(),
> +			       result.data_.begin(),
> +			       [&func, rhs](T v) { return func(v, rhs); });
> +
> +		return result;
> +	}

`std::function` is quite heavy. I think making `apply()` a template would be
beneficial wrt. performance.


Regards,
Barnabás Pőcze

> +
>  	std::array<T, Rows> data_;
>  };
> 
> --
> Regards,
> 
> Laurent Pinchart