Wierd bug popped up over the night. I was trialfactoring to lowlimits using 4 threads on my quad core, and it was running normally last night. Then today I noticed its not working just spamming this message over and over:

PrimeNet error 33: CPU identity mismatch
CPU identity mismatch: g=e8e86d15234d3b60fc7c2f907cf63be8 hg=0742e5edcca38c55615b8d12ca95f5fa wg=f2c8329a1206b1aacf4c344702ed9155
Updating computer information on the server

I didn't touch the computer over night, so no changes to local.txt or prime.txt or anything else.

[QUOTE=ATH;138560]PrimeNet error 33: CPU identity mismatch
CPU identity mismatch: g=e8e86d15234d3b60fc7c2f907cf63be8 hg=0742e5edcca38c55615b8d12ca95f5fa wg=f2c8329a1206b1aacf4c344702ed9155
Updating computer information on the server[/QUOTE]
I had something similar happen to a machine that I was setting up.

Can you use 2 or more threads on the same ECM curve?

I assume not since ThreadsPerTest=4 isn't working.

[QUOTE=ATH;138560]Wierd bug popped up over the night. I was trialfactoring to lowlimits using 4 threads on my quad core, and it was running normally last night. Then today I noticed its not working just spamming this message over and over:

PrimeNet error 33: CPU identity mismatch
CPU identity mismatch: g=e8e86d15234d3b60fc7c2f907cf63be8 hg=0742e5edcca38c55615b8d12ca95f5fa wg=f2c8329a1206b1aacf4c344702ed9155
Updating computer information on the server

I didn't touch the computer over night, so no changes to local.txt or prime.txt or anything else.[/QUOTE]

I've fixed the looping bug. I don't know what caused the mismatch that started your problems. I take it you haven't installed a new version of Windows or changed your hardware recently.

[QUOTE]I've fixed the looping bug. I don't know what caused the mismatch that started your problems. I take it you haven't installed a new version of Windows or changed your hardware recently.[/QUOTE]

No, didn't change anything, it stopped working between last night and today. I had to delete local.txt to get it to work, and now I have a new ComputerGUID, so it made a new computer on my account with same name but new ComputerGUID :( Oh well.

I just did a manual com with the V5 server. I checked the box to report expected completion dates. The worker threads stopped and restarted [B]on different [/B]numbers. I lost, maybe 69% on a double check.

[QUOTE=Uncwilly;138661]I just did a manual com with the V5 server. I checked the box to report expected completion dates. The worker threads stopped and restarted [B]on different [/B]numbers. I lost, maybe 69% on a double check.[/QUOTE]

The save file should still be in your directory. You can manually add the exponent back to your worktodo.txt file.

Does prime.log show any errors or weird messages that ight shed a clue as to why prime95 started on new exponents?

I just noticed very strange behaviour on a machine:
[code]
[Work thread Aug 1 14:04] Optimal P-1 factoring of M46010101 using up to 3600MB of memory.
[Work thread Aug 1 14:04] Assuming no factors below 2^68 and 4.6e+04 primality tests saved if a factor is found.
[/code]Assuming 46000 primality tests saved?

mprime is behaving strange. I've told it to use one thread on one CPU, saving one core for gmp-ecm, but it uses 100% on both cores (200%). It has been chewing on this exponent since the middle of june. When I restarted it now, it said:

[code]
[Fri Aug 1 14:02:45 2008]
M46010101 stage 1 is 0.02% complete.
[/code]A prime.spl file indicates that mprime can't communicate, but all communication seems to work fine:

[code]
[Fri Aug 1 14:04:02 2008 - ver 25.6]
Updating computer information on the server
URL: http://v5.mersenne.org/v5server/?v=0.95&px=GIMPS&t=uc&g=c6cc4d91020df9de24d692e404ad507c&hg=d5bc75ee6eae9567a6e20a04af3e8c44&wg=&a=Linux64,Prime95,v25.6,build+6&c=AMD+Athlon(tm)+64+X2+Dual+Core+Processor+4200%2B&f=RDTSC,CMOV,Prefetch,3DNow!,MMX,SSE,SSE2&L1=64&L2=512&np=2&hp=1&m=3822&s=997&h=24&r=1000&ss=42763&sh=87C56C9DC59B1C83611BEB305BE1B5A5
RESPONSE:
pnErrorResult=0
pnErrorDetail=SUCCESS
g=c6cc4d91020df9de24d692e404ad507c
u=S00113
un=Sturle Sunde
cn=express
od=4
==END==
[/code]I tried to kill mprime, remove prime.spl and restart. Same output and prime.spl was recreated. Any ideas?

[QUOTE=S00113;138676]Assuming 46000 primality tests saved?[/QUOTE]

I fixed a server bug getting P-1 factoring work. Thanks. In your worktodo.txt file change 46002 to 2.

The bad value is causing the optimal P-1 bounds code to take a really, really long time. I'll add some
sanity checks to the test_saved value to prevent this.

[QUOTE=S00113;138676]Assuming 46000 primality tests saved?[/QUOTE]I found something weird in worktodo.txt too:[quote]Pfactor=6B43FFBB23CB98FE024A2978888313FD,1,2,46018909,-1,68,[color=red]46002[/color][/quote]It got there all by itself, it first showed up as a problem when I tried to Test|Status and it just froze, so I terminated Prime95 and restarted it half a dozen times, even rebooted to try and fix the problem (Prime95 would load, but not start any work and lock 1 core to 100% doing no (apparent) useful work. Eventually I noticed the above and changed the "46002" to "2" and it seems happy enough now...

[QUOTE=James Heinrich;138705]I found something weird in worktodo.txt too.[/QUOTE]

This is the same bug S000113 just reported. It is now fixed.