srsieve/sr2sieve enhancements - Page 11

unconnected · 2013-10-28, 12:29

sr1sieve 1.4.5 and sr2sieve 1.9.3 are both compiled and work well.
I've "patched" srsieve sources, it is now able to compile but crashes with "Floating point exception" error.

rogue · 2013-10-28, 12:32

Quote:

Originally Posted by unconnected

Some problems with compilation v. 1.0.7. I've tried gcc 4.1.2 and 4.4.7

Code:

files.c: In function ‘add_seq_knc’:
files.c:373: error: called object ‘filter_k’ is not a function
files.c: In function ‘read_input_file’:
files.c:711: error: ‘null’ undeclared (first use in this function)
files.c:711: error: (Each undeclared identifier is reported only once
files.c:711: error: for each function it appears in.)
make: *** [files.o] Error 1

filter_k is declared on line 47. Try "NULL" instead of "null".

rogue · 2013-10-28, 17:20

Quote:

Originally Posted by unconnected

sr1sieve 1.4.5 and sr2sieve 1.9.3 are both compiled and work well.
I've "patched" srsieve sources, it is now able to compile but crashes with "Floating point exception" error.

I think I have some changes I made on my Mac to address some issues, but never ported them back to Windows. I'll take a look tonight.

pepi37 · 2013-10-28, 21:45

Can you in sr1sieve set -f option by default - so you dont have add it in command line ( like in sr2sieve)?

rogue · 2013-10-29, 00:47

Quote:

Originally Posted by pepi37

Can you in sr1sieve set -f option by default - so you dont have add it in command line ( like in sr2sieve)?

Not everyone wants and output file of factors from sr1sieve. I assume that most just use the -o option to write a file with the factors removed.

ValerieVonck · 2013-11-13, 08:59

Rogue,

when compiling on Mac Os X 10.9 I get the following error:

Code:

imac-van-cedric-vonck:srsieve_1.0.7 cedricvonck$ make
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o srsieve.o srsieve.c
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o arithmetic32.o arithmetic32.c
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o arithmetic64.o arithmetic64.c
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o bitmap.o bitmap.c
In file included from bitmap.c:18:
In file included from ./memset_fast32.h:21:
./asm-x86-64-gcc.h:394:41: error: fields must have a constant size: 'variable length array in structure' extension will never be supported
         "=m" (*(struct { uint_fast32_t dummy[count]; } *)dst)
                                        ^
1 error generated.
make: *** [bitmap.o] Error 1
imac-van-cedric-vonck:srsieve_1.0.7 cedricvonck$ gcc -v
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.9.sdk/usr/include/c++/4.2.1
Apple LLVM version 5.0 (clang-500.2.79) (based on LLVM 3.3svn)
Target: x86_64-apple-darwin13.0.0
Thread model: posix
imac-van-cedric-vonck:srsieve_1.0.7 cedricvonck$

Can you please help me?
Thank you

rogue · 2013-11-13, 13:28

Quote:

Originally Posted by CedricVonck

Rogue,

when compiling on Mac Os X 10.9 I get the following error:

Code:

imac-van-cedric-vonck:srsieve_1.0.7 cedricvonck$ make
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o srsieve.o srsieve.c
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o arithmetic32.o arithmetic32.c
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o arithmetic64.o arithmetic64.c
gcc -O2 -fomit-frame-pointer -ffast-math -march=k8 -m64  -Wall -DHAVE_CMOV  -DUSE_ASM -DNDEBUG  -c -o bitmap.o bitmap.c
In file included from bitmap.c:18:
In file included from ./memset_fast32.h:21:
./asm-x86-64-gcc.h:394:41: error: fields must have a constant size: 'variable length array in structure' extension will never be supported
         "=m" (*(struct { uint_fast32_t dummy[count]; } *)dst)
                                        ^
1 error generated.
make: *** [bitmap.o] Error 1
imac-van-cedric-vonck:srsieve_1.0.7 cedricvonck$ gcc -v
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.9.sdk/usr/include/c++/4.2.1
Apple LLVM version 5.0 (clang-500.2.79) (based on LLVM 3.3svn)
Target: x86_64-apple-darwin13.0.0
Thread model: posix
imac-van-cedric-vonck:srsieve_1.0.7 cedricvonck$

Can you please help me?
Thank you

I ran into the same problem. I just changed the routine to use memset, but ran into another problem. The application crashes. I suspect a compiler bug in the latest Xcode.

f1pokerspeed · 2014-01-12, 20:23

Has the stop on removal rate feature been added to sr1sieve? I have tried using -R and -r, and neither argument is handled by the program...

rogue · 2014-01-12, 22:42

Quote:

Originally Posted by f1pokerspeed

Has the stop on removal rate feature been added to sr1sieve? I have tried using -R and -r, and neither argument is handled by the program...

No. -r is used when sieving for RieselSieve and I don't even know if they use it for that project. -R was never added to sr1sieve.

Antonio · 2014-07-12, 17:52

While demonstrating srsieve to a friend we happened to try 2^n-1 for n=1 to 1M with the following results in the output file:

Code:

pmin=16481841017
2^n-1
1
3
7
13
17
19
31
61
89
.
.
.
etc

It would appear to have incorrectly factored out n=2 and n=5. while this, of it's self, is not a real worry I was wondering if it was a problem in the code which may be propagated into the 'normal' search areas.

(Note. Geoffrey Reynolds' srsieve ver. 0.6.17 doesn't have this problem)

rogue · 2014-07-12, 18:04

It seems to be something that was introduced in 1.0.7. 1.0.6 has the correct output. I'll look into it.

2014-07-12, 17:52	#120
Antonio "Antonio Key" Sep 2011 UK 213₁₆ Posts	srsieve 1.0.7 bug While demonstrating srsieve to a friend we happened to try 2^n-1 for n=1 to 1M with the following results in the output file: Code: pmin=16481841017 2^n-1 1 3 7 13 17 19 31 61 89 . . . etc It would appear to have incorrectly factored out n=2 and n=5. while this, of it's self, is not a real worry I was wondering if it was a problem in the code which may be propagated into the 'normal' search areas. (Note. Geoffrey Reynolds' srsieve ver. 0.6.17 doesn't have this problem)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Sieving twins with srsieve	henryzz	Twin Prime Search	0	2014-03-18 12:44
Intel announces multi-core enhancements for Haswell chips	ixfd64	Hardware	8	2012-02-10 20:32
LLRnet enhancements	kar_bon	No Prime Left Behind	10	2008-03-28 11:21
TODO list and suggestions/comments/enhancements	Greenbank	Octoproth Search	2	2006-12-03 17:28
Suggestions for future enhancements	Reboot It	Software	16	2003-10-17 01:31

2013-10-28, 12:29	#111
unconnected May 2009 Russia, Moscow 7·19² Posts	sr1sieve 1.4.5 and sr2sieve 1.9.3 are both compiled and work well. I've "patched" srsieve sources, it is now able to compile but crashes with "Floating point exception" error.

2013-10-28, 21:45	#114
pepi37 Dec 2011 After milion nines:) 3·463 Posts	Can you in sr1sieve set -f option by default - so you dont have add it in command line ( like in sr2sieve)?

2014-01-12, 20:23	#118
f1pokerspeed Jun 2012 1101010₂ Posts	Has the stop on removal rate feature been added to sr1sieve? I have tried using -R and -r, and neither argument is handled by the program...

2014-07-12, 18:04	#121
rogue "Mark" Apr 2003 Between here and the 61×101 Posts	It seems to be something that was introduced in 1.0.7. 1.0.6 has the correct output. I'll look into it.