mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Software

Reply
 
Thread Tools
Old 2010-05-21, 03:52   #1
NBtarheel_33
 
NBtarheel_33's Avatar
 
"Nathan"
Jul 2008
Maryland, USA

5·223 Posts
Question Bug in P-1 completion percentage report

I'm running v25.11 build 2. The box in question is a PIV, 3.06 GHz, with 1GB RAM. I have Hyperthreading enabled, so I'm running an LL test and a P-1 at the same time. OS is Windows 7, 64 bit, and I'm running Prime64.

Strange thing happened tonight - Prime64 is chugging along, reporting that the P-1 job is 100.00% completed to both the results.txt file and on the Prime64 console window...even though it has only processed something like 442 of 480 relative primes (which is more like 92%) and is obviously still working! What clued me in to this was that I checked results.txt and saw repeatedly the line "Mxxxxxxxx has completed P-1 to 100.00%". I thought something had perhaps "stuck", so I terminated the Windows service, and fired up the Prime64 console only to see that indeed, the job was still chugging along. So how does Prime64 think it's finished, when it's obviously not? Not a big issue, but one that George might be curious to check out.
NBtarheel_33 is offline   Reply With Quote
Old 2010-05-21, 03:54   #2
NBtarheel_33
 
NBtarheel_33's Avatar
 
"Nathan"
Jul 2008
Maryland, USA

5·223 Posts
Lightbulb Forgot to mention...

I did forget to mention that when the P-1 test began, I had 512 MB RAM allocated to Prime64, but as this was leaving me with essentially zero available memory, I had to drop back to 400 MB RAM for Prime64. Would this by chance have wrought havoc on the progress calculation?
NBtarheel_33 is offline   Reply With Quote
Old 2010-05-21, 06:21   #3
Batalov
 
Batalov's Avatar
 
"Serge"
Mar 2008
Phi(4,2^7658614+1)/2

2×7×683 Posts
Default

The threads probably stopped and started with new settings. Then this bug is something that was known (and reported) for a few years.
It is only a UI bug, not the library bug (screen reporting after a stop+[optional crash/reboot]+go).
Batalov is offline   Reply With Quote
Old 2010-05-21, 16:47   #4
NBtarheel_33
 
NBtarheel_33's Avatar
 
"Nathan"
Jul 2008
Maryland, USA

21338 Posts
Thumbs up

Quote:
Originally Posted by Batalov View Post
The threads probably stopped and started with new settings. Then this bug is something that was known (and reported) for a few years.
It is only a UI bug, not the library bug (screen reporting after a stop+[optional crash/reboot]+go).
Yep, that's it. I had to shut Prime95 down and restart it, in order to change the memory settings. But I just found out the other day that my folks have 2 x 1GB sticks of RAM laying around, so the next time I head home, I'm going to collect that and turn this machine into a P-1 beast. Should also help with general day-to-day tasks, too - I'm starting to see that 1GB RAM is just not enough for Windows 7 (or Vista, really, for that matter), though for a 2006 model, this machine is still plugging along pretty well (considering it had XP on it originally, and has gone through Vista, and now 7).
NBtarheel_33 is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
What percentage of CPUs/GPUs have done a double check? Mark Rose Data 4 2016-06-17 14:38
Mechanics of DC instead of LL percentage Fred PrimeNet 1 2016-02-29 02:19
Fines based on a percentage of net worth jasong jasong 30 2013-06-06 07:20
DC completion not noticed Mini-Geek GPU to 72 3 2011-12-20 16:12
Modifying Percentage-Done Display? Primeinator Software 4 2005-03-15 05:32

All times are UTC. The time now is 01:39.


Sun Oct 17 01:39:55 UTC 2021 up 85 days, 20:08, 0 users, load averages: 2.92, 2.94, 2.60

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.