AFAIK, only the non-SSE2 code has been (noticably) changed. :mellow:

[QUOTE]Linux: [url]ftp://mersenne.org/gimps/mprime246.tar.gz[/url]
or [url]ftp://mersenne.org/gimps/sprime246.tar.gz[/url] (staticly linked)[/QUOTE]

BTW, almost a year ago GP2 and me began to build a first rpm version for mprime. I wrote the spec file for SuSE distribution, GP2 the RedHat part. AFAIK, the RedHat part has not been tested but the SuSE version ran in my systems since many months ago.

Mprime run as a daemon service at start time, following UNIX System V start scheme. The (user,group) is (mersenne, mersenne). You can get a SuSE 23.5 and 24.6 version here:

[url]ftp://ftp.oxixares.com/pub/mprime/[/url]

I don't know how it works for RedHat systems. You can rebuild rpm from source rpm in that directory.

Guillermo

FreeBSD 5.3 executables:

[url]ftp://mersenne.org/gimps/mprime246-freebsd.tar.gz[/url]
[url]ftp://mersenne.org/gimps/sprime246-freebsd.tar.gz[/url]

The K6-2 illegal instruction was reported earlier, I am not sure whether this is a new problem or the fix didn't make it into the Linux version:

CPU Information:
AMD-K6(tm) 3D processor
CPU speed: 549.68 MHz
CPU features: RDTSC, MMX
L1 cache size: 32 KB
L2 cache size: unknown

[CODE]
geoff@yel:~$ mkdir test
geoff@yel:~$ cd test
geoff@yel:~/test$ tar xzf ../sprime246.tar.gz mprime
geoff@yel:~/test$
geoff@yel:~/test$
geoff@yel:~/test$ ./mprime -v
Mersenne Prime Test Program, Version 24.6.2
geoff@yel:~/test$
geoff@yel:~/test$
geoff@yel:~/test$ ./mprime -t
Beginning a continuous self-test to check your computer.
Please read stress.txt. Hit ^C to end this test.
Test 1, 400 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
Illegal instruction (core dumped)
geoff@yel:~/test$
geoff@yel:~/test$
geoff@yel:~/test$ gdb --quiet ./mprime core
(no debugging symbols found)...Using host libthread_db library "/lib/tls/libthread_db.so.1".
Core was generated by `./mprime -t'.
Program terminated with signal 4, Illegal instruction.
#0 0x08126b48 in ?? ()
(no debugging symbols found)...
(gdb) #
(gdb) #
(gdb) info frame
Stack level 0, frame at 0xbffff750:
eip = 0x8126b48; saved eip 0x1
called by frame at 0xbffff754
Arglist at 0xbffff748, args:
Locals at 0xbffff748, Previous frame's sp is 0xbffff750
Saved registers:
eip at 0xbffff74c
(gdb) #
(gdb) #
(gdb) info registers
eax 0x0 0
ecx 0xb7932000 -1215094784
edx 0x44000 278528
ebx 0x0 0
esp 0xbffff74c 0xbffff74c
ebp 0x40 0x40
esi 0xb7d49040 -1210806208
edi 0x30 48
eip 0x8126b48 0x8126b48
eflags 0x210247 2163271
cs 0x73 115
ss 0x7b 123
ds 0x7b 123
es 0x7b 123
fs 0x0 0
gs 0x0 0
(gdb) #
(gdb) #
(gdb) x/10i 0x8126b20
0x8126b20: add %al,(%eax)
0x8126b22: fstpl 0x2108(%esi)
0x8126b28: fstpl (%esi)
0x8126b2a: fstpl 0x6308(%esi)
0x8126b30: fstpl 0x2100(%esi)
0x8126b36: fstpl 0x4200(%esi)
0x8126b3c: lea 0x10(%esi),%esi
0x8126b3f: add $0x40,%dl
0x8126b42: jae 0x8126a86
0x8126b48: prefetcht1 0x0(%ebp,%ecx,1)
(gdb) quit
[/CODE]

[QUOTE=geoff]The K6-2 illegal instruction was reported earlier, I am not sure whether this is a new problem or the fix didn't make it into the Linux version:
[/QUOTE]

Try downloading again. When I tar the object files over to linux they go "back in time" 5 hours. This can screw up make. I'm sure I've just done something boneheaded in setting up GMT / EST in Linux.

[QUOTE=sdbardwick][...] Probably take a little less than a month for the first results.[...]
[/QUOTE]P95's estimated time of completion has dropped. Looks like 2 DC (one from each proc.) will be done 12/26 or 12/27.

The 24.6 version increases my CPU temps by about 5 degrees Fahrenheit!

ETA: Hey! I've acquired an avatar!

The ECM code with large exponents now works OK with the new version 24.6. Thanks a lot.

Results for my Pentium 4 3E & Prime95 v. 24.6 vs. 23.8.
 

V. 23.8:

Intel(R) Pentium(R) 4 CPU 3.00GHz
CPU speed: 2998.53 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: unknown
L2 cache size: 1024 KB
L1 cache line size: unknown
L2 cache line size: 128 bytes
TLBS: 64
Prime95 version 23.8, RdtscTiming=1
Best time for 384K FFT length: 13.474 ms.
Best time for 448K FFT length: 16.412 ms.
Best time for 512K FFT length: 18.467 ms.
Best time for 640K FFT length: 22.004 ms.
Best time for 768K FFT length: 26.435 ms.
Best time for 896K FFT length: 32.078 ms.
Best time for 1024K FFT length: 35.625 ms.
Best time for 1280K FFT length: 47.376 ms.
Best time for 1536K FFT length: 56.745 ms.
Best time for 1792K FFT length: 68.362 ms.
Best time for 2048K FFT length: 76.099 ms.

V. 24.6:

Intel(R) Pentium(R) 4 CPU 3.00GHz
CPU speed: 2998.53 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 16 KB
L2 cache size: 1024 KB
L1 cache line size: 64 bytes
L2 cache line size: 128 bytes
TLBS: 64
Prime95 version 24.6, RdtscTiming=1
Best time for 512K FFT length: 18.507 ms.
Best time for 640K FFT length: 21.895 ms.
Best time for 768K FFT length: 26.552 ms.
Best time for 896K FFT length: 32.006 ms.
Best time for 1024K FFT length: 35.698 ms.
Best time for 1280K FFT length: 47.067 ms.
Best time for 1536K FFT length: 56.520 ms.
Best time for 1792K FFT length: 68.287 ms.
Best time for 2048K FFT length: 76.137 ms.

Regards.

Marco Franceschini.

I came across this problem when I tried ECM on 10^311-1:

ECM found a factor in curve #1, stage #1
Sigma=7393313418913594, B1=110000000, B2=4290000000.
ERROR: Factor doesn't divide N!
10^311-1 completed 1 ECM curves, B1=110000000, B2=4290000000

I tried some smaller bounds and got mixed results, including one where the 'ERROR: Factor doesn't divide N!' message repeated over and over until I hit stop.

The prime.ini, local.ini and worktodo.ini that I used are attached. The machine is a Northwood P4 2.66GHz with Windows 2000, Prime95 is from the 14 December 2004 archive.

Finished and reported a DC done ~50% with v24.6 on an athalon.

[Thu Dec 16 16:15:45 2004]
UID: Team_Inspector/uncwilly, M13328993 is not prime. Res64: xxxxxxxxxxxxxxxx. Wa1: xxxxxxxx,xxxxxxxx,00000000

res blocked out by me.

[QUOTE=geoff]I came across this problem when I tried ECM on 10^311-1:[/QUOTE]

I'll look into the bug. BTW GMP-ECM should clean prime95's clock on this number.