617be2756f
Like the similar legacy extension FLOAT(), DFLOAT() represents a conversion from default integer to DOUBLE PRECISION. Rewrite into a conversion operation. Differential Revision: https://reviews.llvm.org/D107489
16 lines
423 B
Fortran
16 lines
423 B
Fortran
! RUN: %flang_fc1 -fdebug-unparse %s 2>&1 | FileCheck %s
|
|
! Checks that a call to the legacy extension intrinsic function
|
|
! DFLOAT is transmogrified into a type conversion operation.
|
|
module m
|
|
!CHECK: d = 1._8
|
|
double precision :: d = dfloat(1)
|
|
contains
|
|
subroutine sub(n)
|
|
integer, intent(in) :: n
|
|
!CHECK: 2._8
|
|
print *, dfloat(2)
|
|
!CHECK: real(n,kind=8)
|
|
print *, dfloat(n)
|
|
end subroutine
|
|
end module
|