用allocate重写数组声明后出现 Can't open module file gai'ru'he'jie'j...

xiejihong0306

已知在我把主程序和子程序的数组声明都改写成allocate形式之后（在此之前是直接指定的数组大小，可以正常编译），例如

[Fortran] 纯文本查看 复制代码

integer(kind=4) :: Iorder_, l, js(:)
real(kind=8), allocatable :: FmatrixCoeff_(:, :), FvectorRside_(:), x(:)
allocate(FmatrixCoeff_(Iorder_, Iorder_), FvectorRside_(Iorder_), x(Iorder_), js(Iorder_))

就出现以下报错:

Fatal Error: Can't open module file 'subroutines.mod' for reading at (1): No such file or directory
compilation terminated.
终端进程“C:\WINDOWS\System32\WindowsPowerShell\v1.0\powershell.exe -Command gfortran -g d:\_Personal__backup##\100__Project\VSCode_Fortran\Fortran_Alog_2nd_XuShiLiang\*.f90 -o 'D:\_Personal__backup##\100__Project\VSCode_Fortran\Fortran_Alog_2nd_XuShiLiang\1-1 AGAUS All-choice principal element Gaussian elimination method.exe'”已终止，退出代码: 1。


gFortran编译，算法没问题，可以忽略，肯定是语法上哪里细节错了，实在找不到问题了。

求助，感激不尽。源代码如下：


-----------------------------------------------------------------------------------

[Fortran] 纯文本查看 复制代码

!  All-choice principal element Gaussian elimination method
!  Solve the system of linear algebraic equations AX = B.

      module SUBROUTINES
        contains  ! Used to encapsulate functions or subroutines
        !implicit none  ! Used to encapsulate variables
        Subroutine SUB_agaus(FmatrixCoeff_, FvectorRside_, Iorder_, x, l, js)
          integer(kind=4) :: row = 0, col = 0
          integer(kind=4) :: Iorder_, l, js(:)          
          real(kind=8) :: b, t
          real(kind=8), allocatable :: FmatrixCoeff_(:, :), FvectorRside_(:), x(:)
          allocate(FmatrixCoeff_(Iorder_, Iorder_), FvectorRside_(Iorder_), x(Iorder_), js(Iorder_)) 

          l = 1
          Do k = 1, Iorder_ - 1
            d = 0.0
            ! Do i = k, Iorder_
            !   Do j = k, Iorder_
            !     If (abs(FmatrixCoeff_(i,j))>d) Then
            !       d = abs(FmatrixCoeff_(i,j))
            !       js(k) = j
            !       is = i
            !     End If
            !   End Do
            ! End Do
            forall ( row = k:Iorder_, col = k:Iorder_, abs(FmatrixCoeff_(row, col))>d ) 
              d = abs(FmatrixCoeff_(row, col))
              js(k) = col
              is = row
            end forall

            If (d+1.0==1.0) Then
              l = 0
            Else
              If (js(k)/=k) Then
                Do i = 1, Iorder_
                  t = FmatrixCoeff_(i, k)
                  FmatrixCoeff_(i, k) = FmatrixCoeff_(i, js(k))
                  FmatrixCoeff_(i, js(k)) = t
                End Do
              End If
              If (is/=k) Then
                Do j = k, Iorder_
                  t = FmatrixCoeff_(k, j)
                  FmatrixCoeff_(k, j) = FmatrixCoeff_(is, j)
                  FmatrixCoeff_(is, j) = t
                End Do
                t = FvectorRside_(k)
                FvectorRside_(k) = FvectorRside_(is)
                FvectorRside_(is) = t
              End If
            End If
            !If l=0, it means that the coefficient matrix of the system of equations is singular, and the solution fails; if L=0, it means that the normal return
            If (l==0) Then
              Write (*, 100)
              Return
            End If
            Do j = k + 1, Iorder_
              FmatrixCoeff_(k, j) = FmatrixCoeff_(k, j)/FmatrixCoeff_(k, k)
            End Do
            FvectorRside_(k) = FvectorRside_(k)/FmatrixCoeff_(k, k)
            Do i = k + 1, Iorder_
              Do j = k + 1, Iorder_
                FmatrixCoeff_(i, j) = FmatrixCoeff_(i, j) - FmatrixCoeff_(i, k)*FmatrixCoeff_(k, j)
              End Do
              FvectorRside_(i) = FvectorRside_(i) - FmatrixCoeff_(i, k)*FvectorRside_(k)
            End Do
          End Do
          If (abs(FmatrixCoeff_(Iorder_, Iorder_))+1.0==1.0) Then
            l = 0
            Write (*, 100)
            Return
          End If
          x(Iorder_) = FvectorRside_(Iorder_)/FmatrixCoeff_(Iorder_, Iorder_)
          Do i = Iorder_ - 1, 1, -1
            t = 0.0
            Do j = i + 1, Iorder_
              t = t + FmatrixCoeff_(i, j)*x(j)
            End Do
            x(i) = FvectorRside_(i) - t
          End Do
          js(Iorder_) = Iorder_
          Do k = Iorder_, 1, -1
            If (js(k)/=k) Then
              t = x(k)
              x(k) = x(js(k))
              x(js(k)) = t
            End If
          End Do
          Return
          100 Format (1X, ' FAIL ')
        End Subroutine SUB_agaus
      end module SUBROUTINES




!// ForQuill v1.01 Beta www.fcode.cn


! NOTE：
!
program PROGRAM_NAME
  use SUBROUTINES, only : SUB_agaus
  implicit none
  integer ExitKey
  integer(kind=4) :: i, n = 4, l
  integer(kind=4), allocatable :: js(:) !Declare a variable-size array  
  real(kind=8), allocatable :: a(:, :), b(:), x(:) !Declare a variable-size array
  allocate(js(n), a(n, n), b(n), x(n)) !Configure memory space

  Data a/0.2368, 0.1968, 0.1582, 1.1161, 0.2471, 0.2071, 1.1675, 0.1254 &
  , 0.2568, 1.2168, 0.1768, 0.1397, 1.2671, 0.2271, 0.1871, 0.1490/
  Data b/1.8471, 1.7471, 1.6471, 1.5471/
  
  Call SUB_agaus(a, b, n, x, l, js)
  If (l /= 0) write(unit=6, fmt="('X(', I2, ' )=', D15.6)") (i, x(i), i=1, 4)


! Avoid having the command window exit before the results are output
  write ( *, *) 'Press input any words to exit...'
  read ( *, *) ExitKey
  stop
end program PROGRAM_NAME

fcode

需要把 js 也定义为 allocatable 的数组，此外，请一定要写 implicit none，它非常重要