Date: Sun, 30 Mar 2003 02:39:26 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [Ux2bs_Archive] No. 120 ************************************************** Saturday 29 March 2003 Number 120 ************************************************** Subjects for today 1 Re: Autoconf error...? : Andreas Buening 2 Re: re: Building perl with gcc-3.2.1 : Stefan.Neis at t-online.de 3 Re: Autoconf error...? : John Poltorak 4 Re: M4 : John Poltorak 5 Re: M4 : Andreas Buening **= Email 1 ==========================** Date: Sun, 30 Mar 2003 01:26:44 +0100 From: Andreas Buening Subject: Re: Autoconf error...? John Poltorak wrote: > > On Sat, Mar 29, 2003 at 11:40:54PM +0100, Andreas Buening wrote: > > > The latest texinfo I have is 4.0. If you look into the configure.in > > v4.5 uses configure.ac... That's the same. > > file you find things like > > > > AC_CHECK_HEADERS(ncurses/termcap.h) > > > > The correct syntax would be > > > > AC_CHECK_HEADERS([ncurses/termcap.h]) > > > > If you add all those missing quotes then it should work. > > Do you mean I should change all the lines begining with AC_ ? > > What about AM_ ? All lines in upper case like "XX_XXX(...)" are macros. > How do I change this, if it needs changing:- ? > > AC_CHECK_FUNC(sigblock, , AC_CHECK_LIB(bsd, sigblock)) Every comma separated argument gets its own "[]": AC_CHECK_FUNC([sigblock],, [ AC_CHECK_LIB([bsd], [sigblock]) ]) (The line breaks are just for readability) Bye, Andreas -- One OS to rule them all, One OS to find them, One OS to bring them all and in the darkness bind them In the Land of Mordor where the Shadows lie. _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 2 ==========================** Date: Sun, 30 Mar 2003 12:52:52 +0100 From: Stefan.Neis at t-online.de Subject: Re: re: Building perl with gcc-3.2.1 Hi, > Use which C compiler? [gcc] > Uh-oh, the C compiler 'gcc' doesn't seem to be working. > You need to find a working C compiler. > [...] > > > Well! Looks like some digging around in the configure script... Presumably, it expects to find .obj files, however, with gcc-3.x all object files have the extension .o, independent on whether or not -Zomf is being used... Regards, Stefan _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 3 ==========================** Date: Sun, 30 Mar 2003 19:21:14 +0000 From: John Poltorak Subject: Re: Autoconf error...? --nbPbA20jtBMZpcdc Content-Type: text/plain; charset=us-ascii On Sun, Mar 30, 2003 at 01:26:44AM +0100, Andreas Buening wrote: > John Poltorak wrote: > > > > On Sat, Mar 29, 2003 at 11:40:54PM +0100, Andreas Buening wrote: > > > > > The latest texinfo I have is 4.0. If you look into the configure.in > > > > v4.5 uses configure.ac... > > That's the same. > > > > file you find things like > > > > > > AC_CHECK_HEADERS(ncurses/termcap.h) > > > > > > The correct syntax would be > > > > > > AC_CHECK_HEADERS([ncurses/termcap.h]) > > > > > > If you add all those missing quotes then it should work. > > > > Do you mean I should change all the lines begining with AC_ ? > > > > What about AM_ ? > > All lines in upper case like "XX_XXX(...)" are macros. > > > How do I change this, if it needs changing:- ? > > > > AC_CHECK_FUNC(sigblock, , AC_CHECK_LIB(bsd, sigblock)) > > Every comma separated argument gets its own "[]": > > AC_CHECK_FUNC([sigblock],, [ > AC_CHECK_LIB([bsd], [sigblock]) > ]) > > (The line breaks are just for readability) I've done my best with it but can't get it work any differently... Can you look at the attached configure.ac and say what needs changing? I'd be interested in seeing how your changes differ from mine... > > Bye, > Andreas > > -- > One OS to rule them all, One OS to find them, > One OS to bring them all and in the darkness bind them > In the Land of Mordor where the Shadows lie. -- John --nbPbA20jtBMZpcdc Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename="C:\\configure.ac" dnl Process this file with autoconf to produce a configure script. dnl $Id: configure.ac,v 1.22 2003/02/05 00:40:19 karl Exp $ dnl dnl This file is free software; as a special exception the author gives dnl unlimited permission to copy and/or distribute it, with or without dnl modifications, as long as this notice is preserved. dnl dnl This program is distributed in the hope that it will be useful, but dnl WITHOUT ANY WARRANTY, to the extent permitted by law; without even the dnl implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. dnl AC_PREREQ(2.57)dnl Minimum Autoconf version required. dnl AC_INIT([GNU Texinfo], [4.5]) AC_CONFIG_SRCDIR([makeinfo/makeinfo.c]) AM_CONFIG_HEADER(config.h:config.in)dnl Keep filename to 8.3 for MS-DOS. AM_INIT_AUTOMAKE([dist-bzip2]) dnl AM_CONDITIONAL(TEXINFO_MAINT, test -n "$TEXINFO_MAINT") dnl Checks for programs. AC_PROG_CC AC_PROG_GCC_TRADITIONAL AC_PROG_INSTALL AC_PROG_MAKE_SET AC_PROG_RANLIB AC_ISC_POSIX AC_MINIX # Needed on sysV68 for sigblock, sigsetmask. But check for it in libc first. AC_CHECK_FUNC(sigblock, , AC_CHECK_LIB(bsd, sigblock)) dnl Checks for header files. dnl Do not use unless we're linking with ncurses. if test "x$termlib" = xncurses; then dnl Use AC_CHECK_HEADERS so the HAVE_*_H symbol gets defined. AC_CHECK_HEADERS(ncurses/termcap.h) fi AC_HEADER_STAT AC_HEADER_STDC AC_CHECK_HEADERS(fcntl.h io.h limits.h pwd.h string.h strings.h \ termcap.h termio.h termios.h unistd.h \ sys/fcntl.h sys/file.h sys/stream.h sys/time.h sys/ttold.h sys/wait.h) dnl sys/ptem.h requires sys/stream.h for mblk_t on Solaris. AC_CHECK_HEADERS([sys/ptem.h],,, [#if HAVE_SYS_STREAM_H #include #endif ]) AC_SYS_POSIX_TERMIOS AC_HEADER_TIOCGWINSZ dnl Checks for typedefs, structures, and compiler characteristics. AC_TYPE_OFF_T AC_TYPE_SIGNAL AC_C_CONST AC_STRUCT_TM dnl Checks for function declarations. txi_CHECK_DECLS dnl Checks for library functions. AC_FUNC_ALLOCA AC_FUNC_STRCOLL AC_FUNC_VPRINTF dnl in theory only pre-sysvr3 systems needed this and it's not likely dnl that anyone compiling new texinfo still has such a thing? we'll see. dnl AC_FUNC_SETVBUF_REVERSED AC_CHECK_FUNCS(bzero getcwd memset setvbuf sigaction sigprocmask \ sigsetmask strchr) dnl strerror, xmalloc, xrealloc, probably others should be added. AC_REPLACE_FUNCS(memcpy memmove strdup strcasecmp strerror strncasecmp) dnl We want to recognize djgpp to avoid the useless warning about no dnl term library. AC_CANONICAL_BUILD # Some GNU/Linux systems (e.g., SuSE 4.3, 1996) don't have curses, but # rather ncurses. So we check for it. TERMLIBS= # Check for termlib before termcap because Solaris termcap needs libucb. TERMLIB_VARIANTS="ncurses curses termlib termcap terminfo" for termlib in ${TERMLIB_VARIANTS}; do AC_CHECK_LIB(${termlib}, tgetent, [TERMLIBS="${TERMLIBS} -l${termlib}"; break]) done # don't bother warning on djgpp, it doesn't have a term library, it # ports each termcap-needing program separately according to its needs. if test -z "$TERMLIBS" && echo "$build" | grep -v djgpp >/dev/null; then AC_MSG_WARN([probably need a terminal library, one of: ${TERMLIB_VARIANTS}]) fi dnl Checks for variables. # HP-UX 9 (at least) needs -lncurses which defines termcap variables PC etc. AC_MSG_CHECKING(for library with termcap variables) AC_CACHE_VAL(ac_cv_var_ospeed, oldLIBS=$LIBS for trylib in $termlib ${TERMLIB_VARIANTS}; do if test "x$trylib" != "x$termlib"; then LIBS="$oldLIBS -l$termlib -l$trylib" else LIBS="$oldLIBS -l$termlib" fi AC_TRY_LINK(, #ifdef HAVE_NCURSES_TERMCAP_H #include #else #ifdef HAVE_TERMCAP_H #include #else #undef PC char *BC; char **UP; char PC; short ospeed; #endif #endif /* Make sure all variables actually exist. AIX 4.3 has ospeed but no BC. --Andreas Ley , 24 Aug 2000. */ BC++; UP++ PC++; return ospeed != 0; , ac_cv_var_ospeed=$trylib; break) done LIBS=$oldLIBS ) AC_MSG_RESULT($ac_cv_var_ospeed) if test -n "$ac_cv_var_ospeed" \ && test "x$termlib" != "x$ac_cv_var_ospeed"; then TERMLIBS="${TERMLIBS} -l${ac_cv_var_ospeed}" fi AC_SUBST(TERMLIBS)dnl dnl gnulib. UTILS_FUNC_MKSTEMP # For teTeX and TeX Live. AC_CANONICAL_HOST AC_ARG_ENABLE(multiplatform, [ --enable-multiplatform put executables in bin/PLATFORM]) test "x$enable_multiplatform" = xyes \ && test "x$bindir" = 'x${exec_prefix}/bin' \ && bindir="$bindir/$host" dnl i18n support. To update to a new version of gettext, run: dnl gettextize -f -c --intl AM_GNU_GETTEXT_VERSION(0.11.5) AM_GNU_GETTEXT() AC_CONFIG_FILES([ Makefile \ doc/Makefile \ info/Makefile \ intl/Makefile \ lib/Makefile \ m4/Makefile \ makeinfo/Makefile \ makeinfo/tests/Makefile \ po/Makefile.in \ util/Makefile \ ]) AC_OUTPUT --nbPbA20jtBMZpcdc Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs --nbPbA20jtBMZpcdc-- **= Email 4 ==========================** Date: Sun, 30 Mar 2003 19:57:55 +0000 From: John Poltorak Subject: Re: M4 On Sat, Mar 29, 2003 at 11:41:14PM +0100, Andreas Buening wrote: > John Poltorak wrote: > > > > On Sat, Mar 29, 2003 at 09:15:02PM +0100, Andreas Buening wrote: > > [snip] > > > > Maybe m4.texinfo is defective. I don't know. > > > > There is no indication thatit is defective. > > > > Do you have this problem when building m4? > > Yes. Is it due to a bug? Presumably this error doesn't occur on Linux? > > > This means you define __GNU_LIBRARY__ somewhere which sounds > > > not really good. > > > > I'm only using a standard build process of running autoconf, configure and > > Make and not setting any specific variables. The build does include your > > diffs for m4 and I've tried to follow your build instructions as closely > > as possible. > > But somewhere you define __GNU_LIBRARY__, maybe in a header file. I can't find it anywhere... Would it be shown in config.log or somewhere like that? > > > > > In spite of this, m4.exe does get built and runs. Whether it works > > > > correctly I'm not sure. > > > > > > > > AIUI the default build should not use the included regex. I don't know why > > > > it ignores the regex library which exists in /usr/lib... > > > > > > There must have been a configure test for regex which you've > > > ignored. > > > > This is completely reproducible under UX2BS. No manual intervention is > > involved. > > I mean one line of configure output should tell you that > no external regex could be found. The only reference to regex whilst running configure is here:- checking if malloc debugging is wanted... no checking for library containing re_compile_pattern... -lregex checking for working re_compile_pattern... yes configure: WARNING: Included src/regex.c not used Does suggest it has been found? It seems as though some of the problems I've had may have been due to using Posix/2. I've just excluded it and now I get the following error:- make[1]: Leaving directory `U:/unixos2/workdir/m4-1.4/lib' making install in src make[1]: Entering directory `U:/unixos2/workdir/m4-1.4/src' gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ m4.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ builtin.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ debug.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ eval.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ format.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ freeze.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ input.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ macro.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ output.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ path.c gcc -c -DHAVE_CONFIG_H -I.. -I. -I./../lib -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ symtab.c gcc -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ -Zcrtdll -Zmt -s -Zlinker /exepack:2 -Zlinker /pmtype:vio -o m4.exe m4.obj builtin.obj debug.obj eval.obj format.obj freeze.obj input.obj macro.obj output.obj path.obj symtab.obj ../lib/libm4.a -lregex LINK386 : fatal error L1104: ..\lib\libm4.a : not valid library make[1]: *** [m4] Error 1 make[1]: Leaving directory `U:/unixos2/workdir/m4-1.4/src' make: *** [install] Error 1 What's causing this error? M4 itself appears to get built OK. > Bye, > Andreas > > -- > One OS to rule them all, One OS to find them, > One OS to bring them all and in the darkness bind them > In the Land of Mordor where the Shadows lie. -- John _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 5 ==========================** Date: Sun, 30 Mar 2003 23:30:15 +0100 From: Andreas Buening Subject: Re: M4 John Poltorak wrote: > > On Sat, Mar 29, 2003 at 11:41:14PM +0100, Andreas Buening wrote: > > John Poltorak wrote: > > > > > > On Sat, Mar 29, 2003 at 09:15:02PM +0100, Andreas Buening wrote: > > > > [snip] > > > > > > Maybe m4.texinfo is defective. I don't know. > > > > > > There is no indication thatit is defective. > > > > > > Do you have this problem when building m4? > > > > Yes. > > Is it due to a bug? Presumably this error doesn't occur on Linux? I don't think so. The file is rather old. The texinfo syntax may have changed or the parser may be more strict nowadays. [snip] > > But somewhere you define __GNU_LIBRARY__, maybe in a header file. > > I can't find it anywhere... Would it be shown in config.log or somewhere > like that? No idea. [snip] > checking if malloc debugging is wanted... no > checking for library containing re_compile_pattern... -lregex > checking for working re_compile_pattern... yes > configure: WARNING: Included src/regex.c not used > > Does suggest it has been found? In this case: yes. > It seems as though some of the problems I've had may have been due to > using Posix/2. I've just excluded it and now I get the following error:- Ah, this could be the reason why the included regex is used. The configure script checks whether you have a regex library. Posix/2 has one, but not GNU regex. So configure stops searching and uses the internal one. [snip] > gcc -Zomf -O2 -s -Zmt -D__ST_MT_ERRNO__ -Zcrtdll -Zmt -s -Zlinker /exepack:2 -Zlinker /pmtype:vio -o m4.exe m4.obj builtin.obj debug.obj eval.obj format.obj freeze.obj input.obj macro.obj output.obj path.obj symtab.obj ../lib/libm4.a -lregex This line looks okay. > LINK386 : fatal error L1104: ..\lib\libm4.a : not valid library > make[1]: *** [m4] Error 1 > make[1]: Leaving directory `U:/unixos2/workdir/m4-1.4/src' > make: *** [install] Error 1 > > What's causing this error? You may have forgotten "make AR=emxomfar". [snip] Bye, Andreas -- One OS to rule them all, One OS to find them, One OS to bring them all and in the darkness bind them In the Land of Mordor where the Shadows lie. _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs