[v2,07/17] libcamera: matrix: Extend multiplication operator to heterogenous types

22985 diff mbox series

From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

It is useful to multiply matrices of heterogneous types, for instance
float and double. Extend the multiplication operator to support this,
avoiding the need to convert one of the matrices. The type of the
returned matrix is selected automatically to avoid loosing precision.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

---

Changes in v2:
- Added this patch
---
 include/libcamera/internal/matrix.h | 10 ++++++----
 src/libcamera/matrix.cpp            |  5 +++--
 2 files changed, 9 insertions(+), 6 deletions(-)

Quoting Stefan Klug (2025-03-19 16:11:12)
> From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> 
> It is useful to multiply matrices of heterogneous types, for instance
> float and double. Extend the multiplication operator to support this,
> avoiding the need to convert one of the matrices. The type of the
> returned matrix is selected automatically to avoid loosing precision.
> 
> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Posted by Stefan, so should probably have your SoB - (or at least RB tag
now :D)

But this sounds like something that will be helpful later in this
series...


Acked-by: Kieran Bingham <kieran.bingham@ideasonboard.com>

> 
> ---
> 
> Changes in v2:
> - Added this patch
> ---
>  include/libcamera/internal/matrix.h | 10 ++++++----
>  src/libcamera/matrix.cpp            |  5 +++--
>  2 files changed, 9 insertions(+), 6 deletions(-)
> 
> diff --git a/include/libcamera/internal/matrix.h b/include/libcamera/internal/matrix.h
> index 6e3c190286fe..688c6fd498f3 100644
> --- a/include/libcamera/internal/matrix.h
> +++ b/include/libcamera/internal/matrix.h
> @@ -8,6 +8,7 @@
>  
>  #include <algorithm>
>  #include <sstream>
> +#include <type_traits>
>  #include <vector>
>  
>  #include <libcamera/base/log.h>
> @@ -129,15 +130,16 @@ constexpr Matrix<U, Rows, Cols> operator*(const Matrix<U, Rows, Cols> &m, T d)
>         return d * m;
>  }
>  
> -template<typename T, unsigned int R1, unsigned int C1, unsigned int R2, unsigned int C2>
> -constexpr Matrix<T, R1, C2> operator*(const Matrix<T, R1, C1> &m1, const Matrix<T, R2, C2> &m2)
> +template<typename T1, unsigned int R1, unsigned int C1, typename T2, unsigned int R2, unsigned int C2>
> +constexpr Matrix<std::common_type_t<T1, T2>, R1, C2> operator*(const Matrix<T1, R1, C1> &m1,
> +                                                              const Matrix<T2, R2, C2> &m2)
>  {
>         static_assert(C1 == R2, "Matrix dimensions must match for multiplication");
> -       Matrix<T, R1, C2> result;
> +       Matrix<std::common_type_t<T1, T2>, R1, C2> result;
>  
>         for (unsigned int i = 0; i < R1; i++) {
>                 for (unsigned int j = 0; j < C2; j++) {
> -                       T sum = 0;
> +                       std::common_type_t<T1, T2> sum = 0;
>  
>                         for (unsigned int k = 0; k < C1; k++)
>                                 sum += m1[i][k] * m2[k][j];
> diff --git a/src/libcamera/matrix.cpp b/src/libcamera/matrix.cpp
> index 8590f8efeff3..f9117357dc7e 100644
> --- a/src/libcamera/matrix.cpp
> +++ b/src/libcamera/matrix.cpp
> @@ -138,11 +138,12 @@ LOG_DEFINE_CATEGORY(Matrix)
>   */
>  
>  /**
> - * \fn Matrix<T, R1, C2> operator*(const Matrix<T, R1, C1> &m1, const Matrix<T, R2, C2> &m2)
> + * \fn operator*(const Matrix<T1, R1, C1> &m1, const Matrix<T2, R2, C2> &m2)
>   * \brief Matrix multiplication
> - * \tparam T Type of numerical values in the matrices
> + * \tparam T1 Type of numerical values in the first matrix
>   * \tparam R1 Number of rows in the first matrix
>   * \tparam C1 Number of columns in the first matrix
> + * \tparam T2 Type of numerical values in the secont matrix
>   * \tparam R2 Number of rows in the second matrix
>   * \tparam C2 Number of columns in the second matrix
>   * \param m1 Multiplicand matrix
> -- 
> 2.43.0
>