Hi all, pfgw -f100 -q"2^n+1" fails for n > 2755074: "The exception unknown software exception (0xc00000fd) occured in the application at location 0x0047c5b3" ...
This is working fine on my system. I have tried with and without the GMP dll. However, version 4/17 is old. It may have been a version where the GMP was...
Hi Jim, Have you considered using LoadLibrary so that the software loads up the correct version of the library automatically? Or do you not have control over ...
... LoadLibrary is used, and full control over which functions are loaded is being done. However, a check for correct version is NOT done, nor is there any...
Is there any sign yet of special code to speed up the search for Gaussian Mersenne primes, Chris? Yours hopefully :-) Mike Oakes In a message dated 03/05/02...
There is a new version of PFGW being uploaded right now which will use GMP dll's which do not use the alloca. These DLL have been uploaded into the primeform...
I believe i have finally tracked down a bug that has been giving me big head aches. I have confirmed the same thing happens when i run this on a P2, and a...
Complex number's may not be the issue. What you may have is a DLL getting hung. I have seen similar behavior on 2 different WinNT systems. I have to find out...
Ya i am sure it is the DLL. Because when i run your ECM application afterwards which uses the same DLL it gives me bad data as well. Both machines are...
PFGW should work fine without the DLL (make sure that the DLL is not in WinNT or WinNT\system32 or on the path (or in the current folder, or the folder of...
In http://groups.yahoo.com/group/openpfgw/message/737 Greg noted that he had kindly provided a linux build. This works fine, thanks. But it does not have the...
If someone is going to handle this, then contact me first. I have code newer than what is in the OpenPFGW file folder. What is there is one version prior to...
... I've fixed the bugs in erat now so the modular sieve appears to work to arbitrary ranges, though still with a 64k small-prime-window. If that level of...
I was curious if anyone can provide BeOS precombiled objects of the Woltman objects. I retrieved Win32/Linux ones from Woltman's original code but I fail to...
Louis Kenneth Helm
lhelm@...
Jun 14, 2002 4:14 pm
887
Hi there Louis ... The Linux ones should work just fine - they are in position-independent ELF object format, and I can link OpenPFGW with them using the...
Chris, Thanks for the prompt reply. I was hoping the ELF objects could be used but something told me it wouldn't be that simple. I'm excited to hear that it...
Louis Kenneth Helm
lhelm@...
Jun 14, 2002 5:37 pm
889
... One last thing, make sure mult.o is linked on a 32-byte boundary. You will get your best performance that way....
Things are finally returning back to normal on my end, so if the source is ready, I can compile a Linux binary. In the latest source in the files area, all of...
New Requirements Don't Miss This Exciting Opportunity! Fantastic Computer Based Training Offer There has never been a better time to get Trained or Certified. ...
I know this is an old bug, and has "magically" gone away, but I believe that I have found out why it was occuring, and now can look for other possible problems...
... Did I miss something? I still cannot find a linux dev_build with a big modular sieve. Sorry to be greedy, but it's frustrating to have linux cycles to hand...
Sorry, I sent the email asking for the latest source from Jim, then completely forgot about it. Jim: Is the latest source ready for a linux compile, or...
... What is in the files area (and the latest build which is on my PC's) is not ready for a Linux build at all. The configure scripts need editing (new files...
I should have said (I always forget the good news) that Greg's old linux build is A1OK for me in all respects save the modular sieve limitation to 31 bits. I...
If that is the case, and if Greg still has a buildable environment of files which built that version, then he may simply be able to insert the current...
Tonight I'll pull the stable source code off the website (20020129) and try it. Wait...isn't it in the 20020311 DEV source on the website? I'll try to...
The > 31 bit mod factoring should be in either of those "stable" releases I believe. I think that Phil worked on this last year (November or December rings a...
Greg, You may want to look long and hard and make sure that the "Phil - **** " text is not being output. I think it was wrapped in #ifdef (DEBUG) .... #endif...
Jim, That code is enclosed in #if !defined (NDEBUG) ... #endif so it has to be explicitly disabled. You may want to consider using an #ifdef instead so we...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
