<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=ISO-8859-2"> <META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7650.28"> <TITLE>RE: AIX 4.x and 5.x compilation</TITLE> </HEAD> <BODY>  Thanks for quick reply, I have tried  gcc already, it compiles gmp but 3 tests crashed (core dump). Still I have to stay with xlC compiler, I will try to take a look at gmpxx.h and reorganize it so xlC could compile it. What I would like to do is to compile the same code (with respect to config.h and other files depending on the system) at Windows and Aix. I hope there is easy way through those templates :) Regards  Leszek -------------------------- <Leszek.Wlodarski@partners.brebank.pl> writes:   It seems to me that this is a problem with compiler. I am   little bit afraid of switching those unary ... tests off, Is   there any fast solution/patch for these problems ?   Use gcc, but beware of that most gcc releases will miscompile gmp, so you will very likely need to try several gcc versions. -- Torbjörn <UL>  -----Original Message----- From:   Włodarski, Leszek, (BRE/GLOBUS)  Sent:   Tuesday, August 22, 2006 11:16 AM To:     'gmp-bugs@swox.com' Subject:        AIX 4.x and 5.x compilation Hi, I'm trying to compile gmp under AIX and I have some problems. Compilation of gmp (simple make) causes no problems, running make check also goes smooth, but only with disabled C++. After enabling C++, --enable-cxx) make check failes. I'm using xlC compiler (IBM VisualAgeC). I got message: "../../../gmp-4.2.1/gmpxx.h", line 1918.3: 1540-1202 (I) No candidate is better than "__gmp_expr<__m pf_struct [1],__mpf_struct [1]>::__gmp_expr<__mpf_struct [1],__mpf_struct [1]>(const __gmp_expr<__mp f_struct [1],__mpf_struct [1]> &)". "../../../gmp-4.2.1/gmpxx.h", line 3645.1: 1540-1231 (I) The conversion from argument number 1 to "c onst __gmp_expr<__mpf_struct [1],__mpf_struct [1]> &" uses "the identity conversion". "../../../gmp-4.2.1/gmpxx.h", line 2196.7: 1540-0217 (S) "eval" is not a member of "class __gmp_expr <__mpq_struct [1],__mpq_struct [1]>". "../../../gmp-4.2.1/gmpxx.h", line 2194.13: 1540-0700 (I) The previous message was produced while pr ocessing "__gmp_set_expr<__mpq_struct [1]>(mpq_ptr, const __gmp_expr<__mpq_struct [1],__mpq_struct [ 1]> &)". "../../../gmp-4.2.1/gmpxx.h", line 1736.3: 1540-0700 (I) The previous message was produced while pro cessing "__gmp_expr<__mpq_struct [1],__mpq_struct [1]>::__gmp_expr<__mpq_struct [1],__mpq_struct [1] >(const __gmp_expr<__mpq_struct [1],__mpq_struct [1]> &)". "../../../gmp-4.2.1/gmpxx.h", line 3631.1: 1540-0700 (I) The previous message was produced while pro cessing "__gmp_expr<__mpq_struct [1],__mpq_struct [1]>::operator++(int)". "../../../gmp-4.2.1/gmpxx.h", line 2145.7: 1540-0217 (S) "eval" is not a member of "class __gmp_expr <__mpz_struct [1],__mpz_struct [1]>". "../../../gmp-4.2.1/gmpxx.h", line 2143.13: 1540-0700 (I) The previous message was produced while pr ocessing "__gmp_set_expr<__mpz_struct [1]>(mpz_ptr, const __gmp_expr<__mpz_struct [1],__mpz_struct [ 1]> &)". "../../../gmp-4.2.1/gmpxx.h", line 1568.3: 1540-0700 (I) The previous message was produced while pro cessing "__gmp_expr<__mpz_struct [1],__mpz_struct [1]>::__gmp_expr<__mpz_struct [1],__mpz_struct [1] >(const __gmp_expr<__mpz_struct [1],__mpz_struct [1]> &)". "../../../gmp-4.2.1/gmpxx.h", line 3673.65: 1540-0700 (I) The previous message was produced while pr ocessing "__gmp_expr<__mpz_struct [1],__gmp_urandomm_value>::__gmp_expr<__mpz_struct [1],__gmp_urand omm_value>(__gmp_randstate_struct *, const mpz_class &)". Removing constructor without template arguments makes it compile and even work (test programs work). The next problem I have is message that 'eval' is not a member of __gmp_expr<> It seems to me that this is a problem with compiler. I am little bit afraid of switching those unary … tests off, Is there any fast solution/patch for these problems ? Kind regards    Leszek Wlodarski </UL> </BODY> </HTML>