Re: Simple Program, Lots of Issues (F90)

On Tue, 29 Jul 2008 07:26:26 -0700 (PDT), AnotherSquid <mai@[EMAIL PROTECTED]
>
 wrote in
<4403cfeb-a2c1-4299-88b4-8dceab468535@[EMAIL PROTECTED]
>:
> On Jul 25, 7:57 am, e p chandler <e...@[EMAIL PROTECTED]
> wrote:
 
>> seeftp://ftp.swcp.com/pub/walt/F/Examples/sieve.f95
 
> I tried this simple program using six Fortran compilers but setting
> the integer parameter last_number to 10000000 (10 million). All
> compiled but four of the six had problems during exection.
 
>   gfortran:
>     Segmentation fault
>     Program received signal SIGSEGV, Segmentation fault.
>     0x08048875 in MAIN__ () at primes2.f90:29
>     29            nums(2*i:mxn:i)=0

Program received signal SIGSEGV, Segmentation fault.
0x0000000000400c34 in MAIN__ () at sieve.f90:27
27          numbers(2*i : last_number : i) = 0   ! eliminate all multiples

	Same as you.
GNU Fortran (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
Copyright (C) 2007 Free Software Foundation, Inc.

also:
Program received signal SIGSEGV, Segmentation fault.
0x0000000000400bd6 in MAIN__ () at sieve.f90:27
27          numbers(2*i : last_number : i) = 0   ! eliminate all multiples

GNU Fortran (GCC) 4.2.3 (Ubuntu 4.2.3-2ubuntu7)
Copyright (C) 2007 Free Software Foundation, Inc.
 
	Further printouts suggest this happens when i~last_number/2, for
last_mumber>32762.

>   Intel ifort:
>     Segmentation fault
>     Program received signal SIGSEGV, Segmentation fault.
>     0x0000000000402909 in sieve_of_eratosthenes () at primes2.f90:1
>     1       program Sieve_Of_Eratosthenes

Program received signal SIGSEGV, Segmentation fault.
0x0000000000402954 in sieve_of_eratosthenes () at sieve.f90:37
37      numbers(1:number_of_primes) = pack(numbers, numbers /= 0)

	Problem appears to be with PACK.
ifort (IFORT) 9.1 20061101
Copyright (C) 1985-2006 Intel Cor****ation.  All rights reserved.
 
> My conclusion: DO-loops still work better than array syntax.

	Given that my ifort is relatively old, I'd rather see someone repeat
it with the latest version.  The Ubuntu gfortran is relatively new, I
think.

-- 
Ivan Reid, School of Engineering & Design, _____________  CMS
Collaboration,
Brunel University.    Ivan.Reid@[EMAIL PROTECTED]
|cern.ch]    Room 40-1-B12,
CERN
        KotPT -- "for stupidity above and beyond the call of duty".