[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] Intel compilers, revisited



I haven't heard a response to this question, but then I sent it late on
a Friday too.. so maybe early-ish Tuesday will catch peoples inboxes at
a better time :>

Anyone with insight to how condor_compile does its magic, and if its
magic would finally work with the Intel compilers?  My tests seemed to
say yes, at least somewhat, but I don't know if there's some hidden
gotcha just waiting to ruin the day (and output of someone's
long-running job).

On 3/6/09 5:32 PM, Steve Huston wrote:
> I have quite a few users here who need to use the Intel compilers for
> one reason or another (usually performance), however they would like to
> be able to use condor_compile to create jobs for the standard universe.
> 
> Since the only information I could find on this dates from 2005, I
> decided to take a drastic step - copying the condor_compile script,
> replace 'gcc' with 'icc' in the tests, then tell the makefile to compile
> each line with 'condor_compile icc'.  The compile worked, linked
> properly, and runs fine from everything I can tell.
> 
> While I have yet to try this with a Fortarn program, I wonder aloud -
> and in the direction of those who might know much better than I - if the
> problems with the Intel compilers have been fixed in one way or another,
> and one can actually use them with condor_compile now (and consequently,
> the condor_compile script could be updated to allow for use of icc).
> Anyone?
> 


-- 
Steve Huston - W2SRH - Unix Sysadmin, Dept. of Astrophysical Sciences
  Princeton University  |    ICBM Address: 40.346525   -74.651285
    206 Peyton Hall     |"On my ship, the Rocinante, wheeling through
  Princeton, NJ   08544 | the galaxies; headed for the heart of Cygnus,
    (609) 258-7375      | headlong into mystery."  -Rush, 'Cygnus X-1'