CUDA Fortran host device=device assignment

Accelerated Computing HPC Compilers Legacy PGI Compilers
1

Just to clarify…A host program cannot copy device array to device array?
I will write a “copy kernel”, but am I misunderstanding this part of the “CUDA Fortran Programming Guide & Reference”, v.12 section 3.4.1 or is this a bug?

An assignment statement with a device variable or device array or array section on both sides of the assignment statement will copy data between two device variables or arrays.

pgfortran 10.3-0 64-bit target on x86-64 Linux -tp nehalem-64

:~ pgfortran -c foo.CUF
PGF90-S-0155-more than one device-resident object in assignment  (foo.CUF: 5)
  0 inform,   0 warnings,   1 severes, 0 fatal for test
:~ cat foo.CUF

program test
  use cudafor	
  real, device, dimension(100) :: ha, wha
  call foo(ha,wha)
  ha = wha
  call foo(ha,wha)
  stop
end program
2

Hi Sarah,

This is one of the cases where the implementation is behind the spec. We’re adding this support now with it becoming available sometime later this year.

Sorry,
Mat