Path: utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!cwjcc!hal!nic.MR.NET!umn-cs!hall!tdg
From: tdg@hall.cray.com (Terry Greyzck)
Newsgroups: comp.arch
Subject: Re: Lawerence Livermore Loops in C
Keywords: loops, livermore, kernel
Message-ID: <12223@hall.cray.com>
Date: 19 Dec 88 17:34:58 GMT
References: <584@gt-eedsp.UUCP> <335@taux02.UUCP> <12074@hall.cray.com> <321@helios.prosys.se>
Reply-To: tdg@hall.UUCP (Terry Greyzck)
Organization: Cray Research, Inc., Santa Fe, NM
Lines: 17

In article <321@helios.prosys.se> ath@helios.prosys.se (Anders Thulin) writes:
>In article <12074@hall.cray.com> tdg@hall.UUCP (Terry Greyzck) writes:
>>There seems to be some interest, so I'll post the 14-loop [Laurence
>>Livermore Loops] version here.
>
>Just a comment ... At the end of the main program there's a variable
>'s' which is computed, but does not appear to be used. As it isn't inside
>any of the loops, removing it doesn't change the outcome of the benchmark.
>But it would be interesting to know what it is.

Good observation, although I don't have a clue what it is.  Examining the
Fortran version of the code shows the same variable ('s') calculated in
the same manner, but it isn't used there, either.  Hmm.  I expect the C
version has it simply because the Fortran version has it... as it is not
part of any timed section of code, it can be deleted.

Terry Greyzck