Fortran MPI_ALLGATHER错误并行化代码

我正在尝试并行化以下代码

subroutine log_likelihood(y, theta, lli, ll)
    doubleprecision, allocatable, intent(in)    :: y(:) 
    doubleprecision, intent(in)                 :: theta(2)
    doubleprecision, allocatable, intent(out)   :: lli(:)
    doubleprecision, intent(out)                :: ll
    integer                                     :: i
ALLOCATE (lli(size(y)))
lli = 0.0d0
ll = 0.0d0
do i = 1, size(y)
    lli(i) = -log(sqrt(theta(2))) - 0.5*log(2.0d0*pi) &
           - (1.0d0/(2.0d0*theta(2)))*((y(i)-theta(1))**2)
end do
ll = sum(lli)
end subroutine log_likelihood

为此，我尝试使用MPI_ALLGATHER。这是我写的代码

subroutine log_likelihood(y, theta, lli, ll)
    doubleprecision, allocatable, intent(in)    :: y(:) 
    doubleprecision, intent(in)                 :: theta(2)
    doubleprecision, allocatable, intent(out)   :: lli(:)
    doubleprecision, intent(out)                :: ll
    integer                                     :: i, size_y, diff
size_y=size(y)
ALLOCATE (lli(size_y))
!Broadcasting
call MPI_BCAST(theta, 1, MPI_DOUBLE_PRECISION, 0, MPI_COMM_WORLD, ierr)
call MPI_BCAST(y, 1, MPI_DOUBLE_PRECISION, 0, MPI_COMM_WORLD, ierr)

! Determine how many points to handle with each proc
points_per_proc = (size_y + numprocs - 1)/numprocs
! Determine start and end index for this proc's points
istart = proc_num * points_per_proc + 1
iend = min((proc_num + 1)*points_per_proc, size_y)
diff = iend-istart+1
allocate(proc_contrib(istart:iend))
do i = istart, iend
    proc_contrib(i) = -log(sqrt(theta(2))) - 0.5*log(2.0d0*pi) &
                    - (1.0d0/(2.0d0*theta(2)))*((y(i)-theta(1))**2)
end do
call MPI_ALLGATHER(proc_contrib, diff, MPI_DOUBLE_PRECISION, &
                   lli, diff, MPI_DOUBLE_PRECISION, &
                   MPI_COMM_WORLD, ierr)
ll = sum(lli)
end subroutine log_likelihood

当我试图运行我的程序时，我得到以下错误

$ mpiexec -n 2 ./mle.X 
Fatal error in PMPI_Allgather: Internal MPI error!, error stack:
PMPI_Allgather(961)......: MPI_Allgather(sbuf=0x7ff2f251b860, scount=1500000, MPI_DOUBLE_PRECISION, rbuf=0x7ff2f2ad5650, rcount=3000000, MPI_DOUBLE_PRECISION, MPI_COMM_WORLD) failed
MPIR_Allgather_impl(807).: 
MPIR_Allgather(766)......: 
MPIR_Allgather_intra(560): 
MPIR_Localcopy(357)......: memcpy arguments alias each other, dst=0x7ff2f2ad5650 src=0x7ff2f251b860 len=12000000

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   EXIT CODE: 1
=   CLEANING UP REMAINING PROCESSES
=   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================

谁能给我解释一下我做错了什么

---

谢谢

我终于解决了我的问题。我的代码的串行和并行版本在

---

您正在写入

proc\u contrib（istart:iend）

，但您已将其分配为从1:diff；等级1及以上将覆盖proc_contrib的结尾，这可能会导致问题。（还有一堆缺失的变量；numproc等）。此外，allgather尝试读取数组边界之外的内容。您可以

分配（proc_contrib（istart:iend））

并更改对allgatherv的调用，等等。此外，您实际上不需要执行allgather；安会帮你算数，这就是你在这里所需要的。事实上，我建议不要修改程序，而是在外面算数。理想情况下，您不会让每个处理器“知道”所有的

y（i）

s；他们会有自己的局部问题，调用

log\u likelion

例程，然后全部减少部分和（或者如果log likelion是最终结果，只需要打印出来，就减少部分和）。1。缺少的变量都在该子例程之外声明为全局变量。2.我理解了我的第一个错误，我改为分配（proc_contrib（istart:iend））3。你能给我解释一下怎么修Allgather吗？我需要一个Allgather，因为在我的实际程序中，我需要lli作为输出。谢谢正如乔纳森·杜西（Jonathan Dursi）在上文中指出的那样，你的“diff”论点可能会随着等级的不同而有所不同，即最后一个论点的值可能与所有其他论点的值不同。要处理这个问题，您需要使用allgatherv例程。

subroutine log_likelihood(y, n, theta, lli, ll)
    integer, intent(in)                         :: n
    doubleprecision, intent(in)                 :: y(n) 
    doubleprecision, intent(in)                 :: theta(2)
    doubleprecision, intent(out)                :: lli(n)
    doubleprecision, intent(out)                :: ll
    integer                                     :: i

do i = istart, iend
    proc_contrib(i-istart+1) = -log(sqrt(theta(2))) - 0.5*log(2.0d0*pi) &
                    - (1.0d0/(2.0d0*theta(2)))*((y(i)-theta(1))**2)
end do
 if ( mod(n,numprocs)==0 ) then
    call MPI_ALLGATHER(proc_contrib, points_per_proc, MPI_DOUBLE_PRECISION, &
                       lli, points_per_proc, MPI_DOUBLE_PRECISION, &
                       MPI_COMM_WORLD, ierr)
    else if (numprocs-1 == proc_num ) then 
    recvcounts(numprocs) = iend-istart+1
    call MPI_ALLGATHERV(proc_contrib, points_per_proc, MPI_DOUBLE_PRECISION, &
                        lli, recvcounts, displs, MPI_DOUBLE_PRECISION, &
                        MPI_COMM_WORLD, ierr)
 end if
 ll = sum(lli)
end subroutine log_likelihood