4238 - 4243 complete no new records
[CODE]4238-4239e15 103 kgaps, largest 1306 @ 4238963932934785987
4239-4240e15 101 kgaps, largest 1194 @ 4239006098440446269
4240-4241e15 105 kgaps, largest 1220 @ 4240724612258930333
4241-4242e15 106 kgaps, largest 1258 @ 4241565046295047843
4242-4243e15 110 kgaps, largest 1200 @ 4242350489749598393[/CODE]

[QUOTE=Antonio;460685]Sorry about the programming errors, the attached version should work ok for you now though.[/QUOTE]

That works. Thank you!

Still, at only 854e6 n/sec, I either have to get that C code running, or at least see what my newer PC can do.

Trying to run gap5_e.c, it gives me stackdump-files of the type
[CODE]Exception: STATUS_ILLEGAL_INSTRUCTION at rip=001004110CB
rax=0000000100000000 rbx=00000000FFFFCC30 rcx=0000000000000002
rdx=00000000FFFFCC30 rsi=0000000600028390 rdi=0000000000000000
r8 =00000006000281A0 r9 =0000000000000000 r10=0000000000000040
r11=0000000000000000 r12=00000000FFFFCC6C r13=0000000000000001
r14=00000000FFFFCC6C r15=0000000000000067
rbp=00000000FFFFCB30 rsp=00000000FFFFCAB0
program=C:\Eigene Programme\gap5_d_haswell.exe, pid 8456, thread main
cs=0033 ds=002B es=002B fs=0053 gs=002B ss=002B
Stack trace:
Frame Function Args
000FFFFCB30 001004110CB (00000000020, FF0700010302FF00, 001800477A1, 000FFFFCDF0)
000FFFFCCC0 00180047812 (00000000000, 00000000000, 00000000000, 00000000000)
00000000000 001800455E3 (00000000000, 00000000000, 00000000000, 00000000000)
000FFFFFFF0 00180045694 (00000000000, 00000000000, 00000000000, 00000000000)
End of stack trace
[/CODE]What I can tell from my system informations I have a Dell Studio XPS 8100 with an Intel i7 CPU 870 @ 2.93 GHz - can't figure out by what name that goes right now. My newer PC has an i7-4790 @ 3.6 GHz.

I hope I have a little more spare time in the next few weeks, I really need to catch up on how things work here...

[QUOTE=mart_r;460751]
Trying to run gap5_e.c, it gives me stackdump-files of the type
[CODE]Exception: STATUS_ILLEGAL_INSTRUCTION at rip=001004110CB
rax=0000000100000000 rbx=00000000FFFFCC30 rcx=0000000000000002
rdx=00000000FFFFCC30 rsi=0000000600028390 rdi=0000000000000000
r8 =00000006000281A0 r9 =0000000000000000 r10=0000000000000040
r11=0000000000000000 r12=00000000FFFFCC6C r13=0000000000000001
r14=00000000FFFFCC6C r15=0000000000000067
rbp=00000000FFFFCB30 rsp=00000000FFFFCAB0
program=C:\Eigene Programme\gap5_d_haswell.exe, pid 8456, thread main
cs=0033 ds=002B es=002B fs=0053 gs=002B ss=002B
Stack trace:
Frame Function Args
000FFFFCB30 001004110CB (00000000020, FF0700010302FF00, 001800477A1, 000FFFFCDF0)
000FFFFCCC0 00180047812 (00000000000, 00000000000, 00000000000, 00000000000)
00000000000 001800455E3 (00000000000, 00000000000, 00000000000, 00000000000)
000FFFFFFF0 00180045694 (00000000000, 00000000000, 00000000000, 00000000000)
End of stack trace
[/CODE]What I can tell from my system informations I have a Dell Studio XPS 8100 with an Intel [B]i7 CPU 870[/B] @ 2.93 GHz - can't figure out by what name that goes right now. My newer PC has an [B]i7-4790[/B] @ 3.6 GHz.[/QUOTE]

The exception is caused by the AVX/AVX2 instructions, which your first generation i7 870 doesn't understand. However, the [B]gap5_d_haswell.exe[/B] should run perfectly on your other (newer) PC with the i7-4790, which actually is a [I]Haswell[/I] cpu.

I could try to build an executable for your specific cpu once I get back to a Windows pc (maybe tomorrow). But, perhaps, Antonio could build one earlier. The compiler flags should be:
[CODE]-march=corei7 -mtune=corei7 -mno-avx[/CODE]

4090-4093 done.

[QUOTE=mart_r;460751]That works. Thank you!

Still, at only 854e6 n/sec, I either have to get that C code running, or at least see what my newer PC can do.

Trying to run gap5_e.c, it gives me stackdump-files of the type
[CODE]Exception: STATUS_ILLEGAL_INSTRUCTION at rip=001004110CB
rax=0000000100000000 rbx=00000000FFFFCC30 rcx=0000000000000002
rdx=00000000FFFFCC30 rsi=0000000600028390 rdi=0000000000000000
r8 =00000006000281A0 r9 =0000000000000000 r10=0000000000000040
r11=0000000000000000 r12=00000000FFFFCC6C r13=0000000000000001
r14=00000000FFFFCC6C r15=0000000000000067
rbp=00000000FFFFCB30 rsp=00000000FFFFCAB0
program=C:\Eigene Programme\gap5_d_haswell.exe, pid 8456, thread main
cs=0033 ds=002B es=002B fs=0053 gs=002B ss=002B
Stack trace:
Frame Function Args
000FFFFCB30 001004110CB (00000000020, FF0700010302FF00, 001800477A1, 000FFFFCDF0)
000FFFFCCC0 00180047812 (00000000000, 00000000000, 00000000000, 00000000000)
00000000000 001800455E3 (00000000000, 00000000000, 00000000000, 00000000000)
000FFFFFFF0 00180045694 (00000000000, 00000000000, 00000000000, 00000000000)
End of stack trace
[/CODE]What I can tell from my system informations I have a Dell Studio XPS 8100 with an Intel i7 CPU 870 @ 2.93 GHz - can't figure out by what name that goes right now. My newer PC has an i7-4790 @ 3.6 GHz.

I hope I have a little more spare time in the next few weeks, I really need to catch up on how things work here...[/QUOTE] ok I've compiled a version that should work on your i7-870 (Lynnfield) see attached.

Reserving 4510-4520

Sorry about the irregular reservation I meant to start 4500 but typed the wrong thing and realized a few hours later.

LATEST STATS
 

RANGES WHERE WORK HAS ALREADY STARTED
40-41 , 41-42, 43-44, 44-45, 45-46, 49-50


COMPLETED (ranges of e15)
210
RESERVED (ranges of e15)
158
NEITHER RESERVED NOR COMPLETED
632 (ranges of e15)

Back from my short holiday - wow what happened while I was away. We have acceleration! We have a new searcher!

I have updated post 3. It might be worthwhile for searchers to look at what I have done to make sure I have recorded everything correctly. Also where there were incorrect highest gaps in the range reported because of the software problem, can people let me know what the correct highest gap found should be.

Some statistics - we have now done over 25%!

[CODE]

searcher compl. res'd

Thomas11 72 38
antonio 54 7
robert44444uk 49 6
pinhodecarlos 18 7
danaj 56 45
ET_ 3 2
henryzz 0 10

Total 252 115


[/CODE]

A new version of the c code version 0.05.f and Windows executables is now available, see the attachments.
By hand tuning the next_prime & prec_prime routines and including the Hensel lifting code I have an additional 3% speed boost over 0.05.d
I have now included a Nehalem build for mart_r (or anyone else who wants to put a pre sandybridge i7/i5/i3 machine onto the project)
You will still need the Cygwin1.dll file from post 178 ([url]http://www.mersenneforum.org/showpost.php?p=460557&postcount=178[/url]) to run the executables.

P.S. Welcome back Rob, hope you had a good time. You may want to update the first post to point to the latest code.

Antonio

Could you let me know which posts contain the latest versions of code for each of the architectures. It has gotten quite complex now. Then I'll signpost these in post 1.

[QUOTE=robert44444uk;460787]Antonio

Could you let me know which posts contain the latest versions of code for each of the architectures. It has gotten quite complex now. Then I'll signpost these in post 1.[/QUOTE]
I build for all i7/i5/i3 that GCC 5.4.0 supports (I think), I did a quick special build for mart_r in post 203 but this is now replaced by the Natham build.

Post 207 includes the latest builds for Natham, Sandybridge, Ivybridge and Haswell processors.
Natham is compiled to use SSE4.2 instructions
Sandybridge and Ivybridge compiled to use AVX instructions
Haswell and later compiled to use AVX2 instructions
Posts before 207 were just missing the Natham build.

Pointing to post 207 should be sufficient for your purposes.
I am not going to include the Cygwin1.dll again so point to post 178 for that if you wish, but it is pointed to in post 207 anyway.