- CLN-list - ginac.de

Results of CLN build
by Richard Haney 12 Jul '06

12 Jul '06

I have encountered problems with testing my build of CLN using mingw tools under Windows XP and cygwin-bash. In particular, the test "test_I_io" fails with a message "Illegal number syntax" evidently from function read_number_bad_syntax() in "cl_read_err_bad.cc", but when I comment out the "test_I_io" test in "test_I.cc", the rest of the tests are fine. I describe the considerable machinations and efforts I went through to see if I could determine specifically what the error was, but CLN proved horrendously complicated for the purposes of even preliminary debugging efforts. Those efforts and problems are discussed in more detail near the beginning of my post "Results of GiNaC (& CLN) builds" to the ginac-list at http://www.cebix.net/pipermail/ginac-list/2006-July/000857.html . Richard Haney __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

Question on algorithm for pi
by Barry Levin 09 Jun '06

09 Jun '06

I am not a programmer (yet), and I am doing a project for a Calculus class on the derivation of pi. Since this software managed to produce a large amount of digits for me, and although the code is open, I cannot see what it does due to my own lack of skills, I would like to know how it calculates pi.

3 2

Questions?
by Wilson Castro 02 Jun '06

02 Jun '06

Hello CLN list-users, I want to experiment with ring of the form: (Z mod r) + (Z mod s)*sqrt( z ). where r,s > 1 in Z, and z in Z with abs(z) is squarefree. I have done following little program in order to leave in the root the part of z that is quadrat free. // --------------------- Begin ----------------------------------- // void squarefree_descomposition( cl_I n ) { bool bNegative; cl_I raiz; if( n < 0 ) bNegative = true; else bNegative = false; if( bNegative ) n = -n; if( cln::isprobprime( n ) ) { cout << n << " is quadrat frei " << "-- if( cln::isprobprime( n ) ) -- " << endl; } else { if( cln::sqrtp( n, &raiz) ) { cout << n << " is a perfect square and the exact square root is " << raiz << endl; } else { cl_I m; cl_I factor; cln::isqrt( n , &m); while( m > 1 ) { factor = cln::floor1( n, m*m); if( factor*(m*m) == n ) { cout << "sqrt( "<< n <<" ) = " << m << "*sqrt( "<< factor << " )" << endl; return ; } m = m - 1; } cout << n << " is quadrat frei" << endl; } } } // --------------------- End ----------------------------------- // when I invoke the function as I show in following main. // --------------------- Begin main ----------------------------------- // int main(int argc, char *argv[]) { cl_I n; n = "9999999999"; // 10 { CL_TIMING; squarefree_descomposition(n); } n = "99999999999"; // 11 { CL_TIMING; squarefree_descomposition(n); } n = "999999999999"; // 12 { CL_TIMING; squarefree_descomposition(n); } n = "9999999999999"; // 13 { CL_TIMING; squarefree_descomposition(n); } n = "99999999999999"; // 14 { CL_TIMING; squarefree_descomposition(n); } n = "999999999999999"; // 15 { CL_TIMING; squarefree_descomposition(n); } n = "999999999999999"; // 15 { CL_TIMING; squarefree_descomposition(n); } return EXIT_SUCCESS; } // --------------------- End main ----------------------------------- // I obtain the following result. sqrt( 9999999999 ) = 3*sqrt( 1111111111 ) real time: 0.131 s, run time: 0.132 s sqrt( 99999999999 ) = 3*sqrt( 11111111111 ) real time: 0.493 s, run time: 0.456 s sqrt( 999999999999 ) = 3*sqrt( 111111111111 ) real time: 1.533 s, run time: 1.500 s sqrt( 9999999999999 ) = 3*sqrt( 1111111111111 ) real time: 4.798 s, run time: 4.760 s sqrt( 99999999999999 ) = 3*sqrt( 11111111111111 ) real time: 15.180 s, run time: 15.093 s sqrt( 999999999999999 ) = 3*sqrt( 111111111111111 ) real time: 47.838 s, run time: 47.623 s sqrt( 999999999999999 ) = 3*sqrt( 111111111111111 ) real time: 48.248 s, run time: 47.811 s QUESTIONS: 1. because if the length of the two last calls are equal (length (n) = 15), the times reported by “CL_TIMING” are different? it seems to me that the complexity of my approach depends on the primality test of n. 2. There is some limitation in the length of the algument n in the function "cl_boolean isprobprime (const cl_I& n)"? thank you very much, -- Ms. C. Wilson Castro Rojas Foundations of Informatics Group, Department of Informatics, University of Kaiserslautern, Germany, PO Box: 3049, Building: 34/423, Phone: ++49 631 205 2155, Fax: ++49 631 205 2156,

2 3

fix m4 underquoting
by Ralf Wildenhues 30 May '06

30 May '06

Hello CLN authors, :-) Please apply this patch to fix some underquoted Autoconf macro. Cheers, Ralf * m4/gettimeofday.m4 (CL_GETTIMEOFDAY): Fix M4 quoting. Index: m4/gettimeofday.m4 =================================================================== RCS file: /home/cvs/cln/m4/gettimeofday.m4,v retrieving revision 1.1 diff -u -r1.1 gettimeofday.m4 --- m4/gettimeofday.m4 29 Aug 2005 13:18:40 -0000 1.1 +++ m4/gettimeofday.m4 30 May 2006 21:07:09 -0000 @@ -1,5 +1,5 @@ dnl -*- Autoconf -*- -dnl Copyright (C) 1993-2003 Free Software Foundation, Inc. +dnl Copyright (C) 1993-2003, 2006 Free Software Foundation, Inc. dnl This file is free software, distributed under the terms of the GNU dnl General Public License. As a special exception to the GNU General dnl Public License, this file may be distributed as part of a program @@ -22,7 +22,7 @@ ], [int gettimeofday (struct timeval * tp, struct timezone * tzp);], [int gettimeofday();], cl_cv_proto_gettimeofday_dots=no -cl_cv_proto_gettimeofday_arg2="struct timezone *", +cl_cv_proto_gettimeofday_arg2="struct timezone *", [ CL_PROTO_TRY([ #include <sys/types.h> #include <sys/time.h> @@ -31,7 +31,7 @@ cl_cv_proto_gettimeofday_dots=no cl_cv_proto_gettimeofday_arg2="void *", cl_cv_proto_gettimeofday_dots=yes -cl_cv_proto_gettimeofday_arg2="...")) +cl_cv_proto_gettimeofday_arg2="...")]) ], [extern int gettimeofday (struct timeval *, $cl_cv_proto_gettimeofday_arg2);]) if test $cl_cv_proto_gettimeofday_dots = yes; then AC_DEFINE(GETTIMEOFDAY_DOTS,,[declaration of gettimeofday() needs dots])

2 1

Python Interface to Class Library for Numbers
by Ray Buvel 06 May '06

06 May '06

The clnum package provides a Python interface to the Class Library for Numbers. There is now a binary installer and build procedure available for the Windows platform. For more information see http://calcrpnpy.sourceforge.net/clnumManual.html Ray Buvel

1 0

Re: Compiling CLN on Intel Macs
by Christian Fleck 28 Apr '06

28 Apr '06

Hi Richard, here is the backtrace: Program received signal EXC_BAD_ACCESS, Could not access memory. Reason: KERN_INVALID_ADDRESS at address: 0x06a0afe1 0xa0b02584 in std::wclog () (gdb) bt #0 0xa0b02584 in std::wclog () #1 0xa0b025f0 in std::wclog () #2 0x00009db7 in global constructors keyed to cl_module__cl_prin_globals__firstglobalfun () at ./base/output/ cl_prin_globals.cc:33 #3 0x00009c9a in __static_initialization_and_destruction_0 (__initialize_p=1, __priority=65535) at ../include/cln/io.h:91 #4 0x00009db7 in global constructors keyed to cl_module__cl_prin_globals__firstglobalfun () at ./base/output/ cl_prin_globals.cc:33 #5 0x000026f4 in __static_initialization_and_destruction_0 (__initialize_p=2, __priority=65535) at ../include/cln/io.h:91 #6 0x0000275d in global constructors keyed to _Z15test_elementaryv () at exam.cc:51 #7 0x8fe10168 in __dyld__ZN16ImageLoaderMachO16doInitializationERKN11ImageLoader11LinkCon textE () #8 0x8fe0a516 in __dyld__ZN11ImageLoader23recursiveInitializationERKNS_11LinkContextE () #9 0x8fe0ca08 in __dyld__ZN11ImageLoader15runInitializersERKNS_11LinkContextE () #10 0x8fe031b4 in __dyld__ZN4dyld24initializeMainExecutableEv () #11 0x0000264d in _call_mod_init_funcs () at /SourceCache/Csu/Csu-58/ crt.c:304 #12 0x00002579 in _start (argc=1, argv=0xbffff86c, envp=0xbffff874) at /SourceCache/Csu/Csu-58/crt.c:222 #13 0x00002529 in start () Christian

4 4

Use datarootdir
by Ralf Wildenhues 25 Apr '06

25 Apr '06

The other change for Autoconf-2.60: If you bootstrap CLN with Autoconf-2.59c, config.status will warn you: | config.status: creating Makefile | config.status: WARNING: ../cln/Makefile.in seems to ignore the --datarootdir setting | config.status: creating src/Makefile | config.status: WARNING: ../cln/src/Makefile.in seems to ignore the --datarootdir setting | config.status: creating tests/Makefile | config.status: creating benchmarks/Makefile | config.status: creating examples/Makefile | config.status: creating doc/Makefile | config.status: WARNING: ../cln/doc/Makefile.in seems to ignore the --datarootdir setting and turn on the backward compatibility setting for those Makefiles. The backward-compatible patch below fixes this. :-) Cheers, Ralf * Makefile.in: Set datarootdir. * doc/Makefile.in, src/Makefile.in: Likewise. Index: Makefile.in =================================================================== RCS file: /home/cvs/cln/Makefile.in,v retrieving revision 1.13 diff -u -r1.13 Makefile.in --- Makefile.in 10 Jun 2005 16:16:33 -0000 1.13 +++ Makefile.in 25 Apr 2006 16:05:08 -0000 @@ -8,6 +8,7 @@ exec_prefix = @exec_prefix@ srcdir = @srcdir@ bindir = @bindir@ +datarootdir = @datarootdir@ datadir = @datadir@ libdir = @libdir@ includedir = @includedir@ Index: doc/Makefile.in =================================================================== RCS file: /home/cvs/cln/doc/Makefile.in,v retrieving revision 1.7 diff -u -r1.7 Makefile.in --- doc/Makefile.in 26 Oct 2004 22:23:41 -0000 1.7 +++ doc/Makefile.in 25 Apr 2006 16:05:08 -0000 @@ -9,6 +9,7 @@ prefix = @prefix@ local_prefix = /usr/local exec_prefix = @exec_prefix@ +datarootdir = @datarootdir@ datadir = @datadir@ infodir = @infodir@ mandir = @mandir@ Index: src/Makefile.in =================================================================== RCS file: /home/cvs/cln/src/Makefile.in,v retrieving revision 1.18 diff -u -r1.18 Makefile.in --- src/Makefile.in 22 Oct 2005 09:09:02 -0000 1.18 +++ src/Makefile.in 25 Apr 2006 16:05:08 -0000 @@ -10,6 +10,7 @@ prefix = @prefix@ local_prefix = /usr/local exec_prefix = @exec_prefix@ +datarootdir = @datarootdir@ datadir = @datadir@ libdir = @libdir@ includedir = @includedir@

2 1

Compilation Error - `cl_2D_div.lo' is not a valid libtool object
by Sonia Singhal 22 Apr '06

22 Apr '06

Hi, These are my system specs >uname -a Linux frodo 2.6.11.4-20a-default #1 Wed Mar 23 21:52:37 UTC 2005 i686 unknown unknown GNU/Linux >g++ --version g++-3.4.3 (GCC) 3.4.3 I am using the latest version of CLN ie cln-1.1.11 while 'make'ing I get this error. Any help/inputs? I see that file is not compiled then why is it being linked ? /bin/sh ../libtool --mode=link g++ -o libcln.la -rpath /usr/local/lib -version-info 4:1:0 -objectlist libcln.objectlist libtool: link: `cl_2D_div.lo' is not a valid libtool object make[1]: *** [libcln.la] Error 1 make[1]: Leaving directory `/home/soniasin/software/cln-1.1.11/src' make: *** [all] Error 2 Thanks.

2 1

[patch] do not use `exit' in tests
by Ralf Wildenhues 19 Apr '06

19 Apr '06

Upcoming Autoconf-2.60 will not provide a declaration for exit; it has turned out that tracking MSVC incompatibilites here is a lot of work. Instead, returning from main is suggested as a viable alternative (the systems where that was not portable are not likely to be used any more). The following patch does that for the macros in CLN. Cheers, Ralf * m4/longdouble.m4 (CL_LONGDOUBLE): `return', rather than `exit' from main. * m4/longlong.m4 (CL_LONGLONG): Likewise. * m4/times.m4 (CL_TIMES_CLOCK): Likewise. * m4/general.m4 (CL_CC_WORKS): Likewise, do this for a C test to work when users set `$CC' to a C++ compiler. Index: m4/general.m4 =================================================================== RCS file: /home/cvs/cln/m4/general.m4,v retrieving revision 1.1 diff -u -r1.1 general.m4 --- m4/general.m4 29 Aug 2005 13:18:40 -0000 1.1 +++ m4/general.m4 11 Apr 2006 10:04:22 -0000 @@ -93,7 +93,7 @@ [AC_CACHE_CHECK(whether CC works at all, cl_cv_prog_cc_works, [ AC_LANG_SAVE() AC_LANG_C() -AC_TRY_RUN([int main() { exit(0); }], +AC_TRY_RUN([int main() { return 0; }], cl_cv_prog_cc_works=yes, cl_cv_prog_cc_works=no, AC_TRY_LINK([], [], cl_cv_prog_cc_works=yes, cl_cv_prog_cc_works=no)) AC_LANG_RESTORE() Index: m4/longdouble.m4 =================================================================== RCS file: /home/cvs/cln/m4/longdouble.m4,v retrieving revision 1.1 diff -u -r1.1 longdouble.m4 --- m4/longdouble.m4 29 Aug 2005 13:18:40 -0000 1.1 +++ m4/longdouble.m4 11 Apr 2006 10:04:22 -0000 @@ -12,7 +12,7 @@ AC_DEFUN([CL_LONGDOUBLE], [AC_CACHE_CHECK(for long double type, cl_cv_c_longdouble, [ AC_TRY_RUN([int main() -{ long double x = 2.7182818284590452354L; x = x*x; exit (x==0.0L); }], +{ long double x = 2.7182818284590452354L; x = x*x; return x==0.0L; }], cl_cv_c_longdouble=yes, cl_cv_c_longdouble=no, [ dnl When cross-compiling, use the test from gnulib. AC_TRY_COMPILE([ Index: m4/longlong.m4 =================================================================== RCS file: /home/cvs/cln/m4/longlong.m4,v retrieving revision 1.1 diff -u -r1.1 longlong.m4 --- m4/longlong.m4 29 Aug 2005 13:18:40 -0000 1.1 +++ m4/longlong.m4 11 Apr 2006 10:04:22 -0000 @@ -19,11 +19,11 @@ miscompiled. */ #if defined(__m68k__) || (defined(_IBMR2) || defined(__powerpc)) #if defined(__GNUC__) && (__GNUC__ == 2) && (__GNUC_MINOR__ <= 7) - exit(1); + return 1; #endif #endif { long x = 944938507; long y = 737962842; long z = 162359677; - exit(!(((long long) x)*((long long) y)>>32 == z)); + return !(((long long) x)*((long long) y)>>32 == z); } }], cl_cv_c_longlong=yes, cl_cv_c_longlong=no, [ Index: m4/times.m4 =================================================================== RCS file: /home/cvs/cln/m4/times.m4,v retrieving revision 1.1 diff -u -r1.1 times.m4 --- m4/times.m4 29 Aug 2005 13:18:40 -0000 1.1 +++ m4/times.m4 11 Apr 2006 10:04:22 -0000 @@ -29,12 +29,12 @@ clock_t result2; int ticks; result1 = times(&buffer); - if ((result1 == (clock_t)0) || (result1 == (clock_t)(-1))) exit(1); + if ((result1 == (clock_t)0) || (result1 == (clock_t)(-1))) return 1; sleep(1); result2 = times(&buffer); - if ((result2 == (clock_t)0) || (result2 == (clock_t)(-1))) exit(1); + if ((result2 == (clock_t)0) || (result2 == (clock_t)(-1))) return 1; ticks = result2 - result1; - exit(!((ticks >= CLK_TCK/2) && (ticks <= 3*CLK_TCK/2))); + return !((ticks >= CLK_TCK/2) && (ticks <= 3*CLK_TCK/2)); }], cl_cv_func_times_return=yes, cl_cv_func_times_return=no, dnl When cross-compiling, don't assume anything. cl_cv_func_times_return="guessing no")

2 1

Help - more info
by Wilson Castro Rojas 21 Mar '06

21 Mar '06

for complete the information: these are the data-types: cl_modint_ring modintRing; // Z/mZ cl_I intArg1; cl_MI intMod;

2 1