Return-path: [log in to unmask] Received: from SERV04.SLAC.STANFORD.EDU ([134.79.16.134]) by mailbox.slac.stanford.edu (PMDF V5.2-27 #34068) with SMTP id <[log in to unmask]> for [log in to unmask] (ORCPT rfc822;[log in to unmask]); Wed, 03 Mar 1999 09:02:40 -0800 (PST) Received: from DIRECTORY-DAEMON by SERV04.SLAC.STANFORD.EDU (PMDF V5.2-27 #34067) id <[log in to unmask]> for [log in to unmask] (ORCPT rfc822;[log in to unmask]) ; Wed, 03 Mar 1999 09:02:39 -0800 (PST) Received: from zeno.physics.lsa.umich.edu (gerdes@[141.211.101.161]) by SERV04.SLAC.STANFORD.EDU (PMDF V5.2-27 #34067) with ESMTP id <[log in to unmask]> for [log in to unmask]; Wed, 03 Mar 1999 09:02:37 -0800 (PST) Received: from localhost (gerdes@localhost) by zeno.physics.lsa.umich.edu (8.8.7/8.8.7) with ESMTP id MAA11342 for <[log in to unmask]>; Wed, 03 Mar 1999 12:01:54 -0500 Date: Wed, 03 Mar 1999 12:01:54 -0500 (EST) From: David Gerdes <[log in to unmask]> Subject: Re: continued ascii2lcd crashes (fwd) In-reply-to: <[log in to unmask]> X-Sender: [log in to unmask] To: Richard Dubois <[log in to unmask]> Message-id: <[log in to unmask]> MIME-version: 1.0 Content-type: TEXT/PLAIN; charset=US-ASCII Content-transfer-encoding: 7BIT X-Authentication-warning: zeno.physics.lsa.umich.edu: gerdes owned process doing -bs Hi, It looks like ~80% of the jobs crash somewhere. The crashes occur in ascii2lcd, when it choked on corrupt data, not in the simulation itself. >From this rate I'd conclude that something like 1-2% of events have this problem. Of course, if it happens to be the 2nd event in a file, ascii2lcd will not convert the other 98 events... I did not see this problem with the non-optimized code, at least I don't think so. In any case it was not nearly this bad. --Dave -------------------------------------- David Gerdes, University of Michigan (734) 647-3807 / (734) 936-1817 FAX [log in to unmask] http://umaxp1.physics.lsa.umich.edu/~gerdes/ On Wed, 3 Mar 1999, Richard Dubois wrote: > Hmmm. OK, so I guess optimization is still not quite working on Linux. You only > see this with the optimized code? And what fraction of jobs crash (roughly)? > > Richard > > David Gerdes wrote: > > > These crashes happened only on Linux, with the exe I compiled with > > optimization. That's the bad news. The good news is that, with > > optimization, the average time per event dropped on my 450 MHz machine > > from 4 minutes to about 90 seconds. I saw a few crashes in the Small > > detector, but most of them are with the Large. > > > > --Dave > > -- > Richard Dubois > SLD, Stanford Linear Accelerator Center > [log in to unmask] > http://www.slac.stanford.edu/~richard/ > 650-926-3824 > 650-926-2923 (FAX) > >