compressible_ops_device.F90

Go to the documentation of this file.

! Copyright (c) 2025, The Neko Authors
! All rights reserved.
!
! Redistribution and use in source and binary forms, with or without
! modification, are permitted provided that the following conditions
! are met:
!
!   * Redistributions of source code must retain the above copyright
!     notice, this list of conditions and the following disclaimer.
!
!   * Redistributions in binary form must reproduce the above
!     copyright notice, this list of conditions and the following
!     disclaimer in the documentation and/or other materials provided
!     with the distribution.
!
!   * Neither the name of the authors nor the names of its
!     contributors may be used to endorse or promote products derived
!     from this software without specific prior written permission.
!
! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
! "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
! LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
! FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
! COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
! INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
! BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
! LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
! CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
! LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
! ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
! POSSIBILITY OF SUCH DAMAGE.
!
module compressible_ops_device
  use, intrinsic :: iso_c_binding, only: c_ptr, c_int
  use num_types, only: rp, c_rp
  use field, only: field_t
  use utils, only: neko_error
  implicit none
  private
 
#ifdef HAVE_HIP
  interface
     subroutine hip_compute_max_wave_speed(max_wave_speed_d, u_d, v_d, w_d, &
          gamma, p_d, rho_d, n) &
          bind(c, name = 'hip_compute_max_wave_speed')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: max_wave_speed_d, u_d, v_d, w_d, p_d, rho_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine hip_compute_max_wave_speed
  end interface
 
  interface
     subroutine hip_compute_entropy(S_d, p_d, rho_d, gamma, n) &
          bind(c, name = 'hip_compute_entropy')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: S_d, p_d, rho_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine hip_compute_entropy
  end interface
 
  interface
     subroutine hip_update_uvw(u_d, v_d, w_d, m_x_d, &
          m_y_d, m_z_d, rho_d, n) &
          bind(c, name = 'hip_update_uvw')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d
       integer(c_int) :: n
     end subroutine hip_update_uvw
  end interface
 
  interface
     subroutine hip_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
          p_d, ruvw_d, u_d, v_d, w_d, E_d, rho_d, gamma, n) &
          bind(c, name = 'hip_update_mxyz_p_ruvw')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d
       type(c_ptr), value :: p_d, ruvw_d, E_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine hip_update_mxyz_p_ruvw
  end interface
 
  interface
     subroutine hip_update_e(E_d, p_d, ruvw_d, gamma, n) &
       bind(c, name = 'hip_update_e')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: p_d, E_d, ruvw_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine hip_update_e
  end interface
 
#elif HAVE_CUDA
  interface
     subroutine cuda_compute_max_wave_speed(max_wave_speed_d, u_d, v_d, w_d, &
          gamma, p_d, rho_d, n) &
          bind(c, name = 'cuda_compute_max_wave_speed')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: max_wave_speed_d, u_d, v_d, w_d, p_d, rho_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine cuda_compute_max_wave_speed
  end interface
 
  interface
     subroutine cuda_compute_entropy(S_d, p_d, rho_d, gamma, n) &
          bind(c, name = 'cuda_compute_entropy')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: S_d, p_d, rho_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine cuda_compute_entropy
  end interface
 
  interface
     subroutine cuda_update_uvw(u_d, v_d, w_d, m_x_d, &
          m_y_d, m_z_d, rho_d, n) &
          bind(c, name = 'cuda_update_uvw')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d
       integer(c_int) :: n
     end subroutine cuda_update_uvw
  end interface
 
  interface
     subroutine cuda_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
          p_d, ruvw_d, u_d, v_d, w_d, E_d, rho_d, gamma, n) &
          bind(c, name = 'cuda_update_mxyz_p_ruvw')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d
       type(c_ptr), value :: p_d, ruvw_d, E_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine cuda_update_mxyz_p_ruvw
  end interface
 
  interface
     subroutine cuda_update_e(E_d, p_d, ruvw_d, gamma, n) &
       bind(c, name = 'cuda_update_e')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: p_d, E_d, ruvw_d
       real(c_rp) :: gamma
       integer(c_int) :: n
     end subroutine cuda_update_e
  end interface
 
#elif HAVE_OPENCL
  interface
     subroutine opencl_compute_max_wave_speed(max_wave_speed_d, u_d, v_d, w_d, &
          gamma, p_d, rho_d, n) &
          bind(c, name = 'opencl_compute_max_wave_speed')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: max_wave_speed_d, u_d, v_d, w_d, p_d, rho_d
       real(c_rp), value :: gamma
       integer(c_int), value :: n
     end subroutine opencl_compute_max_wave_speed
  end interface
 
  interface
     subroutine opencl_compute_entropy(S_d, p_d, rho_d, gamma, n) &
          bind(c, name = 'opencl_compute_entropy')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: S_d, p_d, rho_d
       real(c_rp), value :: gamma
       integer(c_int), value :: n
     end subroutine opencl_compute_entropy
  end interface
 
  interface
     subroutine opencl_update_uvw(u_d, v_d, w_d, m_x_d, &
          m_y_d, m_z_d, rho_d, n) &
          bind(c, name = 'opencl_update_uvw')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d
       integer(c_int), value :: n
     end subroutine opencl_update_uvw
  end interface
 
  interface
     subroutine opencl_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
          p_d, ruvw_d, u_d, v_d, w_d, E_d, rho_d, gamma, n) &
          bind(c, name = 'opencl_update_mxyz_p_ruvw')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d
       type(c_ptr), value :: p_d, ruvw_d, E_d
       real(c_rp), value :: gamma
       integer(c_int), value :: n
     end subroutine opencl_update_mxyz_p_ruvw
  end interface
 
  interface
     subroutine opencl_update_e(E_d, p_d, ruvw_d, gamma, n) &
       bind(c, name = 'opencl_update_e')
       use, intrinsic :: iso_c_binding
       import c_rp
       type(c_ptr), value :: p_d, E_d, ruvw_d
       real(c_rp), value :: gamma
       integer(c_int), value :: n
     end subroutine opencl_update_e
  end interface
#endif
 
  public :: compressible_ops_device_compute_max_wave_speed, &
       compressible_ops_device_compute_entropy, &
       compressible_ops_device_update_uvw, &
       compressible_ops_device_update_mxyz_p_ruvw,&
       compressible_ops_device_update_e
 
contains
 
  subroutine compressible_ops_device_compute_max_wave_speed(max_wave_speed, u, v, w, gamma, p, rho, n)
    integer, intent(in) :: n
    real(kind=rp), intent(in) :: gamma
    type(field_t), intent(inout) :: max_wave_speed
    type(field_t), intent(in) :: u, v, w, p, rho
 
#ifdef HAVE_HIP
    call hip_compute_max_wave_speed(max_wave_speed%x_d, u%x_d, v%x_d, w%x_d, gamma, p%x_d, rho%x_d, n)
#elif HAVE_CUDA
    call cuda_compute_max_wave_speed(max_wave_speed%x_d, u%x_d, v%x_d, w%x_d, gamma, p%x_d, rho%x_d, n)
#elif HAVE_OPENCL
    call opencl_compute_max_wave_speed(max_wave_speed%x_d, u%x_d, v%x_d, w%x_d, gamma, p%x_d, rho%x_d, n)
#else
    call neko_error('No device backend configured')
#endif
  end subroutine compressible_ops_device_compute_max_wave_speed
 
  subroutine compressible_ops_device_compute_entropy(S, p, rho, gamma, n)
    integer, intent(in) :: n
    real(kind=rp), intent(in) :: gamma
    type(field_t), intent(inout) :: s
    type(field_t), intent(in) :: p, rho
 
#ifdef HAVE_HIP
    call hip_compute_entropy(s%x_d, p%x_d, rho%x_d, gamma, n)
#elif HAVE_CUDA
    call cuda_compute_entropy(s%x_d, p%x_d, rho%x_d, gamma, n)
#elif HAVE_OPENCL
    call opencl_compute_entropy(s%x_d, p%x_d, rho%x_d, gamma, n)
#else
    call neko_error('No device backend configured')
#endif
  end subroutine compressible_ops_device_compute_entropy
 
  subroutine compressible_ops_device_update_uvw(u_d, v_d, w_d, &
       m_x_d, m_y_d, m_z_d, rho_d, n)
    type(c_ptr), intent(inout) :: u_d, v_d, w_d
    type(c_ptr), intent(in) :: m_x_d, m_y_d, m_z_d, rho_d
    integer, intent(in) :: n
    integer :: i
 
#ifdef HAVE_HIP
    call hip_update_uvw(u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d, n)
#elif HAVE_CUDA
    call cuda_update_uvw(u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d, n)
#elif HAVE_OPENCL
    call opencl_update_uvw(u_d, v_d, w_d, m_x_d, m_y_d, m_z_d, rho_d, n)
#else
    call neko_error('No device backend configured')
#endif
 
  end subroutine compressible_ops_device_update_uvw
 
  subroutine compressible_ops_device_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
       p_d, ruvw_d, u_d, v_d, w_d, E_d, rho_d, gamma, n)
    integer, intent(in) :: n
    type(c_ptr), intent(inout) :: m_x_d, m_y_d, m_z_d, p_d, ruvw_d
    type(c_ptr), intent(in) :: u_d, v_d, w_d, e_d, rho_d
    real(kind=rp), intent(in) :: gamma
 
 
#ifdef HAVE_HIP
    call hip_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
         p_d, ruvw_d, u_d, v_d, w_d, e_d, rho_d, gamma, n)
#elif HAVE_CUDA
    call cuda_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
         p_d, ruvw_d, u_d, v_d, w_d, e_d, rho_d, gamma, n)
#elif HAVE_OPENCL
    call opencl_update_mxyz_p_ruvw(m_x_d, m_y_d, m_z_d, &
         p_d, ruvw_d, u_d, v_d, w_d, e_d, rho_d, gamma, n)
#else
    call neko_error('No device backend configured')
#endif
 
  end subroutine compressible_ops_device_update_mxyz_p_ruvw
 
  subroutine compressible_ops_device_update_e(E_d, p_d, ruvw_d, gamma, n)
    integer, intent(in) :: n
    type(c_ptr), intent(inout) :: e_d, p_d
    ! ruvw = 0.5 * rho * (u^2 + v^2 + w^2)
    type(c_ptr), intent(in) :: ruvw_d
    real(kind=rp), intent(in) :: gamma
 
#ifdef HAVE_HIP
    call hip_update_e(e_d, p_d, ruvw_d, gamma, n)
#elif HAVE_CUDA
    call cuda_update_e(e_d, p_d, ruvw_d, gamma, n)
#elif HAVE_OPENCL
    call opencl_update_e(e_d, p_d, ruvw_d, gamma, n)
#else
    call neko_error('No device backend configured')
#endif
 
  end subroutine compressible_ops_device_update_e
 
end module compressible_ops_device