 |
 |
On May 3, 2004, at 7:21 PM, Ted Stern wrote: > On 3 May 2004 at 09:35 PDT, Drew McCormack wrote: >> >> Is the following legal fortran 90? >> >> >> module A >> contains >> subroutine sub( array ) >> real :: array(10) >> end subroutine >> end module >> >> >> program main >> use A >> real :: a, b(10), c(10,10) >> call sub(a) >> call sub(b) >> call sub(c) >> end program >> > > No. Sorry I can't cite chapter and verse for you. > > There are ways to get around this for specific compilers (e.g. Cray > pointers) > but they should be used with *extreme* caution. In your example, you > could > easily step out of bounds with the "call sub(a)" call. I can see that if I set a(10) in sub, that I will have an out of bounds problem, but if I tried to set array(1), would the compiler be entitled to cause problems (according to the standard)? Is the compiler obligued to make array(1) correspond to a in main, or is it just an error to pass a scalar for an array with an explicit interface? (If so, why isn't my compiler flagging this as an error? I know... I should buy the cray compiler ;-) ) Drew > > Ted > -- > Ted Stern Applications Group > Cray Inc. office: 206-701-2182 > 411 First Avenue South, Suite 600 cell: 206-383-1049 > Seattle, WA 98104-2860 FAX: 206-701-2500 > > Frango ut patefaciam -- I break that I may reveal > (The Paleontological Society motto, equally apropos for debugging) >