Reporting a Catastrophic Error

[sknobat]

Hi

I was trying to compile the following code (developedbyJ F Harper, Mathematics, Victoria University, Wellington). I got the following messsage:

------ Build started: Project: Console1, Configuration: Debug|Win32 ------

Compiling with Intel Visual Fortran Compiler XE 12.1.4.325 [IA-32]...

testpure.f90

0_12459

: catastrophic error: **Internal compiler error: internal abort** Please report this error along with the circumstances in which it occurred in a Software Problem Report. Note: File and line given may not be explicit cause of this error.

compilation aborted for C:\\Temp\\testpure.f90 (code 1)

Build log written to "file://C:\\Temp\\Console1\\Console1\\Debug\\BuildLog.htm"

Console1 - 1 error(s), 0 warning(s)

Code

! A Fortran 95 program that compiles and runs with some compilers, but

! others can't calculate lengths of both spf and dpf. The output should be

! "-1.000". If the line near the end that says

! real :: x = -1

! is changed to

! real(dp):: x = -1

! then the offending compilers can cope and give correct output.

!J F Harper, Mathematics, Victoria University, Wellington,NZ

!

module findbug

implicit none

integer,parameter:: dp=kind(1d0)

interface f

module procedure dpf,spf

end interface f

contains

pure function findf( x) result (out)! with length 50

real(dp),intent(in)::x

character(len = 50):: out

write(out,"(F0.3)") x

end function findf

pure function dpf( x) ! trimmed

real(dp),intent(in)::x

character:: dpf*(len_trim(findf(x)))

dpf = trim(findf(real(x,dp)))

end function dpf

pure function spf( x) ! trimmed

real,intent(in)::x

character:: spf*(len_trim(dpf(real(x,dp))))

spf = trim(dpf(real(x,dp)))

end function spf

end module findbug

program tryf

use findbug, only: dp,f

implicit none

real :: x = -1

print *,'"'//f(x)//'"'

end program tryf

Thank you

Sam

[TimP]

This is definitely a bug of Intel compilers; I reproduced this with a 6 week old ifort for intel64.
The workaround is as stated by the author.

[Steven_L_Intel1]

I can also reproduce this. The program looks familiar - it may have been reported before. I will check.

Yep - this is our issue DPD200175919 which has not yet been fixed. Pretty much identical program source.

[mecej4]

I note that the program performs recursive I/O (not to the same "unit", but once to an internal file and then to standard output).

[Steven_L_Intel1]

That's allowed.

[Steven_L_Intel1]

This problem was fixed in the 16.0 compiler. It was particularly difficult to solve.