From owner-freebsd-ports@FreeBSD.ORG Tue May 22 05:32:09 2012 Return-Path: Delivered-To: freebsd-ports@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id C48FF106564A for ; Tue, 22 May 2012 05:32:09 +0000 (UTC) (envelope-from rhurlin@gwdg.de) Received: from fmailer.gwdg.de (fmailer.gwdg.de [134.76.11.16]) by mx1.freebsd.org (Postfix) with ESMTP id 3F2D18FC12 for ; Tue, 22 May 2012 05:32:09 +0000 (UTC) Received: from p508c712f.dip.t-dialin.net ([80.140.113.47] helo=krabat.raven.hur) by mailer.gwdg.de with esmtpsa (TLSv1:AES256-SHA:256) (Exim 4.72) (envelope-from ) id 1SWhhX-00065r-1o; Tue, 22 May 2012 07:32:07 +0200 Message-ID: <4FBB24D3.4010209@gwdg.de> Date: Tue, 22 May 2012 07:32:03 +0200 From: Rainer Hurling User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:12.0) Gecko/20120503 Thunderbird/12.0.1 MIME-Version: 1.0 To: "coder.tuxfamily" References: <4FBA9620.3020207@gwdg.de> <4FBA997C.6010000@tuxfamily.org> <4FBAA203.90201@gwdg.de> <4FBAA388.1070906@tuxfamily.org> In-Reply-To: <4FBAA388.1070906@tuxfamily.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit X-Authenticated: Id:rhurlin X-Spam-Level: - X-Virus-Scanned: (clean) by exiscan+sophie Cc: freebsd-ports@freebsd.org Subject: Re: graphics/gdal 1.9.0 does not build on CURRENT X-BeenThere: freebsd-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting software to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 22 May 2012 05:32:09 -0000 On 21.05.2012 22:20 (UTC+1), coder.tuxfamily wrote: > Seems to be a problem with python. > > Here the diff that i made yesterday. > > Courtesly. This patch applies fine, but I get exact the same errors as before (see below). There seems to be a problem within extensions/gdal_wrap.cpp. Rainer > Le 21.05.2012 22:13, Rainer Hurling a écrit : >> On 21.05.2012 21:37 (UTC+1), coder.tuxfamily wrote: >>> Hello, >>> >>> It's a problem know by GDAL. >>> >>> In the 1.9.0 with PGSQL, gdal don't build because of include in >>> cpl_recode_iconv.cpp. >>> >>> You can corrige this with include "${CPPFLAGS}" in configure files. >>> >>> You can see the trac on OSGeo : http://trac.osgeo.org/gdal/ticket/4525 >>> That will be fix on 1.9.1 version. >> >> Wow, thanks for the quick response. So the following patch against the >> 1.9.0 port does it for me: >> >> --- files/patch-configure.orig 2012-05-19 12:04:43.000000000 +0200 >> +++ files/patch-configure 2012-05-21 21:59:47.000000000 +0200 >> @@ -1,6 +1,15 @@ >> ---- configure.orig 2011-12-23 09:51:45.000000000 +0800 >> -+++ configure 2011-12-23 15:30:42.540316668 +0800 >> -@@ -21140,10 +21140,10 @@ >> +--- configure.orig 2012-01-04 08:03:42.000000000 +0100 >> ++++ configure 2012-05-21 21:58:56.000000000 +0200 >> +@@ -18694,7 +18694,7 @@ >> + rm -f testiconv.* >> + echo '#include ' > testiconv.cpp >> + echo 'int main(int argc, char** argv) { iconv_t cd; return iconv (cd, >> (const char **) 0, 0, 0, 0); } ' >> testiconv.cpp >> +- if test -z "`${CXX} testiconv.cpp -c 2>&1`" ; then >> ++ if test -z "`${CXX} ${CPPFLAGS} testiconv.cpp -c 2>&1`" ; then >> + { $as_echo "$as_me:${as_lineno-$LINENO}: result: using >> ICONV_CPP_CONST=\"const\"" >&5 >> + $as_echo "using ICONV_CPP_CONST=\"const\"" >&6; } >> + ICONV_CPP_CONST="const" >> +@@ -21232,10 +21232,10 @@ >> if { test -x "$ncdump"; }; then >> { $as_echo "$as_me:${as_lineno-$LINENO}: checking libnetcdf version with >> $ncdump" >&5 >> $as_echo_n "checking libnetcdf version with $ncdump... " >&6; } >> >> >> But now I ran into the next problem: >> >> creating build/temp.freebsd-10.0-CURRENT-amd64-2.7/extensions >> /bin/sh /usr/local/bin/libtool --mode=compile --tag=CC cc -DNDEBUG -O2 >> -pipe -O2 -fno-strict-aliasing -pipe -msse3 -O2 -pipe -O2 >> -fno-strict-aliasing -pipe -msse3 -Wall -Wdeclaration-after-statement >> -DOGR_ENABLED -I/usr/local/include >> -I/usr/ports/graphics/gdal/work/gdal-1.9.0/port -I/usr/local/include >> -I/usr/local -I/usr/local/include -I/usr/local/include -I/usr/local >> -I/usr/local/include -I/usr/local -I/usr/local/include >> -I/usr/local/include -I/usr/local/include -I/usr/local >> -I/usr/local/include -I/usr -I/usr/include -fPIC -I../../port >> -I../../gcore -I../../alg -I../../ogr/ -I/usr/local/include/python2.7 >> -I/usr/local/lib/python2.7/site-packages/numpy/core/include >> -I/usr/ports/graphics/gdal/work/gdal-1.9.0/include -c >> extensions/gdal_wrap.cpp -o >> build/temp.freebsd-10.0-CURRENT-amd64-2.7/extensions/gdal_wrap.o >> libtool: compile: cc -DNDEBUG -O2 -pipe -O2 -fno-strict-aliasing -pipe >> -msse3 -O2 -pipe -O2 -fno-strict-aliasing -pipe -msse3 -Wall >> -Wdeclaration-after-statement -DOGR_ENABLED -I/usr/local/include >> -I/usr/ports/graphics/gdal/work/gdal-1.9.0/port -I/usr/local/include >> -I/usr/local -I/usr/local/include -I/usr/local/include -I/usr/local >> -I/usr/local/include -I/usr/local -I/usr/local/include >> -I/usr/local/include -I/usr/local/include -I/usr/local >> -I/usr/local/include -I/usr -I/usr/include -fPIC -I../../port >> -I../../gcore -I../../alg -I../../ogr/ -I/usr/local/include/python2.7 >> -I/usr/local/lib/python2.7/site-packages/numpy/core/include >> -I/usr/ports/graphics/gdal/work/gdal-1.9.0/include -c >> extensions/gdal_wrap.cpp -fPIC -DPIC -o >> build/temp.freebsd-10.0-CURRENT-amd64-2.7/extensions/.libs/gdal_wrap.o >> cc1plus: warning: command line option "-Wdeclaration-after-statement" is >> valid for C/ObjC but not for C++ >> extensions/gdal_wrap.cpp: In function 'PyObject* >> _wrap_VSIFTruncateL(PyObject*, PyObject*)': >> extensions/gdal_wrap.cpp:6845: error: 'VSIFTruncateL' was not declared >> in this scope >> extensions/gdal_wrap.cpp: In function 'PyObject* >> _wrap_MajorObject_SetMetadata__SWIG_0(PyObject*, PyObject*)': >> extensions/gdal_wrap.cpp:7220: warning: deprecated conversion from >> string constant to 'char*' >> error: command '/bin/sh' failed with exit status 1 >> gmake[2]: *** [build] Fehler 1 >> gmake[2]: Leaving directory >> `/usr/ports/graphics/gdal/work/gdal-1.9.0/swig/python' >> gmake[1]: *** [build] Fehler 2 >> gmake[1]: Leaving directory >> `/usr/ports/graphics/gdal/work/gdal-1.9.0/swig' >> gmake: *** [swig-modules] Fehler 2 >> *** [do-build] Error code 1 >> Stop in /usr/ports/graphics/gdal. >> >> >>> Loïc. >>> >>> >>> Le 21.05.2012 21:23, Rainer Hurling a écrit : >>>> Thanks for the update of graphics/gdal to version 1.9.0. >>>> >>>> It builds fine on 9.0 (amd64), but fails on 10.0-CURRENT (amd64): >>>> >>>> [..snip..] >>>> libtool: compile: c++ -O2 -pipe -O2 -fno-strict-aliasing -pipe -msse3 >>>> -Wall -DOGR_ENABLED -I/usr/local/include >>>> -I/usr/ports/graphics/gdal/work/gdal-1.9.0/port -I/usr/local/include >>>> -I/usr/local -I/usr/local/include -I/usr/local/include -I/usr/local >>>> -I/usr/local/include -I/usr/local -I/usr/local/include >>>> -I/usr/local/include -I/usr/local/include -I/usr/local >>>> -I/usr/local/include -I/usr -I/usr/include -DHAVE_LIBZ -c >>>> cpl_recode_iconv.cpp -fPIC -DPIC -o .libs/cpl_recode_iconv.o >>>> cpl_recode_iconv.cpp: In function 'char* CPLRecodeIconv(const char*, >>>> const char*, const char*)': >>>> cpl_recode_iconv.cpp:92: error: invalid conversion from 'char**' to >>>> 'const char**' >>>> cpl_recode_iconv.cpp:92: error: initializing argument 2 of 'size_t >>>> libiconv(void*, const char**, size_t*, char**, size_t*)' >>>> cpl_recode_iconv.cpp: In function 'char* CPLRecodeFromWCharIconv(const >>>> wchar_t*, const char*, const char*)': >>>> cpl_recode_iconv.cpp:243: error: invalid conversion from 'char**' to >>>> 'const char**' >>>> cpl_recode_iconv.cpp:243: error: initializing argument 2 of 'size_t >>>> libiconv(void*, const char**, size_t*, char**, size_t*)' >>>> gmake[1]: *** [cpl_recode_iconv.lo] Fehler 1 >>>> gmake[1]: Leaving directory >>>> `/usr/ports/graphics/gdal/work/gdal-1.9.0/port' >>>> gmake: *** [port-target] Fehler 2 >>>> *** [do-build] Error code 1 >>>> Stop in /usr/ports/graphics/gdal. >>>> >>>> >>>> It seems there is a portability issue with ICONV_CPP_CONST like it is >>>> described in port/cpl_recode_iconv.cpp:l76ff ? >>>> >>>> /* >>>> -------------------------------------------------------------------- */ >>>> /* XXX: There is a portability issue: iconv() function could be */ >>>> /* declared differently on different platforms. The second */ >>>> /* argument could be declared as char** (as POSIX defines) or */ >>>> /* as a const char**. Handle it with the ICONV_CPP_CONST macro here. */ >>>> /* >>>> -------------------------------------------------------------------- */ >>>> >>>> Please let me know, if you need more details. >>>> >>>> Thanks again, >>>> Rainer >>>> >>>> >>>> P.S.: ports/166605 should be obsolete now?