fortran (once again)
Calvin H. Vu
calvin at dinkum.SGI.COM
Thu Mar 30 11:44:32 AEST 1989
In article <8903291158.aa16902 at SMOKE.BRL.MIL>, LAAKSONE at FINFUN.BITNET writes:
>
> Please tell me what I'm doing wrong. I have a short file with some
> fortran subroutines but I'm totally unable to compile the file.
> The fortran compiler grabs more and more of memory and after a while
> the system tells: DANGER: Out of swap space and I lose the control
> of the system totally. The system I'm using is a 4D/70GTB and the
> fortran compiler is version 1.31.
>
> I can get the same effect by just compiling the following subroutine:
>
> I have tried with and without optimization but always the same problem.
> ----------------------cut here--------------------------------------
> C
> C
> C PLOTTER OFF CALL ESC.)
> SUBROUTINE HPLOF
> C BEFORE OFF SEND A TERMINATING CHARACTER
> CALL HPOUT(char(125)//char(27)//'.)')
> RETURN
> END
>
> -----------------------cut here------------------------------------
> I have tried to compile the code on a VAX and a SUN and there was no
> problem.
>
> Thanks in advance,
>
> Leif Laaksonen
> CSC ESPOO FINLAND
I tried your test case with the 3.1 Rev C release which is the oldest
1.31 compiler I can lay my hands on but could not reproduce the
problem. It compiles successfully with all different versions
of the fortran compiler I currently have so I'm not sure what the
problem is/was.
I suggest the followings:
- check the version of the fortran compiler you have by typing
'versions ftn'.
[Rev C release should have 41 as the last two digits in the Version
column.]
- if your alpha number is less than 41 you can get Rev C from your
local SGI office. You might be able to get Rev D too.
- if your alpha number is 41 or greater you might have problems with
the distribution tape. It sounds far-fetched but it's the only
thing I can think of. You can try to get a replacement tape from
the SGI office if that's the case. I tend to believe this is what
happened though since this is the first time we have ever heard of
this problem. It would have caused an uproar if this actually
happened to everybody :-).
Calvin Vu
Silicon Graphics Computer Systems
More information about the Comp.sys.sgi
mailing list